SP ees

=

. - = ‘ ja.
be one *e tid <5 iit Nig bay BOO 1 net Grd go oe

THE JOURNAL

Ob

BIOLOGICAL CHEMISTRY

EDITED BY

J. J. ABEL and C. A. HERTER

Baltimore New York

Assistant Editor, A. N. RICHARDS, New York

WITH THE COLLABORATION OF

R. H. CHITTENDEN, New Haven, Conn. JACQUES LOEB, Berkeley, Cal.
OTTO FOLIN, Waverley, Mass. GRAHAM LUSK, New York.
WILLIAM J. GIES, New York. A. B. MACALLUM, Toronto, Canada.
REID HUNT, Washington, D. C. J. J. R. MACLEOD, Cleveland, Ohio.
WALTER JONES, Baltimore, Md. A. P. MATHEWS, Chicago, IIL
J. H. KASTLE, Washington, D.C. L. B. MENDEL, New Haven, Conn.
WALDEMAR KOCH, Chicago, Ill. F. G. NOVY, Ann Arbor, Mich.
P. A. LEVENE, New York. W. R. ORNDORFF, Ithaca, N. Y

THOMAS B. OSBORNE, New Haven, Conn. . a es

FRANZ PFAFF, Boston, Mass. fo s 7

Oe (0

A. E. TAYLOR, Berkeley, Cal. 0
V. C. VAUGHAN, Ann Arbor, Mich. V9
ALFRED J. WAKEMAN, New York.

HENRY L. WHEELER, New Haven, conn. >

‘

VOLUME. tik
New Yor«K CITY

1907.

QP
S50 |
Br
\ee

Cop: aa

CopyRIGHT, 1907
BY

THE JOURNAL OF BIOLOGICAL CHEMISTRY.

WILLIAMS & WILKINS COMPANY PRESS,
BALTIMORE, MD.

CONTENTS OF VOLUME III.

Water Jones and C. R. Austrian: On thymus nucleic acid...
Harvey W. Witey: The excretion of boric acid from the human

S.S.Maxwe.u: Creatin asa brain stimulant.. a
R. A. Hatcuer and C.G. L. Woutr: The Pomiation of aeeeeen
ammEUTIOS RCN EN ee, botres 3/5 oc Lote ia. tas SU iade ey ee
H. Gipron Wetts and R. L. Benson: The relation of the thyroid
to autolysis, with a preliminary report on the study of
autolysis by determinations of the changes in freezing
point and electrical conductivity........0. 00024. 0 56.90%.
W. Kocu and Howarp 8. Rreep: The relation of extractive to
protein phosphorus in Aspergillus niger. . x
W. Kocu: The relation of electrolytes to ieciriwe sl ental
H. D. Dakin: Experiments bearing upon the mode of oxidation
of simple aliphatic substances in the animal organism. . .
Orro Foun: On the reduction of barium sulphate in ordinary
pravime me determimations..~ 42% 55h 6. keds eee doen
Orro Foun: On the occurrence and formation of alkyl ureas
pumice kepime emer A 59. 2s 2’. Ra enteekioe
ALONZO ENGLEBERT TayLor: On the synthesis of protein
pirouehithe action Of trypsin... .~ 2.2054 vo. es. b- oeeeek
T. BrartsrorD Ropertson: Note on the synthesis of a protein
PoTOM AM one ACLION OF PEPSINis.... 32.) 5.-... alee eo
STANLEY R. Benepicr: The detection and estimation of reduc-
ISDE SOC 8 ot al eee ee 4 Sr ee
Francis G. Benepicr and THomas B. OsporNnE: The heat of
combustion of vegetable proteins. . BAR ee ee
LAFAYETTE B. MENDEL and PRANK P. ieee rere ae the saliva

Puitrp H. Mircueu: A note on the behavior of uric acid toward
aIMalexteacie and alkalies. .. 02.2. 0. wns cade sata
Haroup C. BrapLEy: Manganese, a normal element in the tissues
of the fresh water clams, Unio and Anodonta............
W. Kocu: The quantitative estimation of extractive and pro-
SEE MOS NOTUS. cop .cs ir. 22 2. Ste Santos. «Pe eee ae

11
21

iv Contents of Volume III

E. Osterperc and C.G. L. Wotr: Day and night urines.......
H. C. Suerman, W. N. Bere, L. J. Conen, and W. G. WHITMAN:
Ammonia in milk and its development during proteoly-
sis under the influence of strong antiseptics............-
Orro Foun: On the separate determination of acetone and
diacetic acid in diabetic urines..........-.--+-++--++05-
Henry L. WHEELER and Treat B. Jounson: IV. Researches on
pyrimidins: On a color test for uracil andcytosin. Plate Le
FRANK W. Bancrorr: On the relative efficiency of the various
methods of administering saline purgatives.............
Tuomas B. Osporne and Isaac F. Harris: The proteins of the
Bae (PisWm SALUUM) 62:8 a)s 52.2 > vials ei tai ee
Tuomas B. Osporne and S. H. Ciapp: Hydrolysis of legumin
PPOWNEDE PER...» ox Sele Laie sak oe < ie een =
Watrer Jones and C. R. Ausrr1an: On the nuclein ferments of
BETO OR. co's c+ i Silene Ski 2 at ae
PROCEEDINGS OF THE AMERICAN Society OF BioLoGIcAL CHEM-
BRERA a a 00 Di so: onan SER eh Sly oe ee ee ee
Rosert Banks Gipson and Karuarine R. Couuins: On the
fractionation of agglutinins and antitoxin..............
Epwin J. Banzuar and Rosert Banks Gipson: The fractional
. precipitation of antitoxic serum: 2...0. 20). 6. s~yaeee
Suinkicui Suzuki: A study of the proteolytic changes occurring
in the lima bean during germination.................-.
HeRMANN ScHLESINGER and Witu1aAM W. Forp: On the chemi-
eal properties of Amanita-toxim ........-....-.+-...2%
Henry L. WHEELER: V. Researches on pyrimidins: On some
salts of cytosin, isocytosin, 6-aminopyrimidin and 6-oxy-
PYTIMIGINS Pao hae fas Seeee Sew cee oe ened ae
Treat B. Jounson: VI. Researches on pyrimidins: Synthesis
of thymin-4-carpORVlE Aid os. 6 as < in cs os ee
H. C. SHerman and J. Epwin Srncuair: The balance of acid-
forming and base-forming elements in foods............
S. Amperc and W. P. Morrinu: On the excretion of creatinin
in the new-born Iniaiits.. «eek. «Se ees. Se
Howarp D. Haskins: The effect of transfusion of blood on the
nitrogenous metabolism of dogs..............-.-.-5-
W. H. Warren and R. 8. Wetss: The picrolonates of certain
alkaloids.. Plates ILI-VIl 2 2... gas 3 ote

Contents of Volume III

WiuuiamM J.Gies: Further observations on protagon..
S.S. Maxwetu: Is the conduction of the nerve eaples a eee
MoU On ap OMSICAl PTOCESS! 65.22... + sun dah! > s+ ou dele.
WILFRED H. MANWARING: Quantitative methods with hemo-
By MCOCRIIM pret eNO Se Sie. . Bp Wee wk te ban ee
ATHERTON SEIDELL: A new standard for use in the colori-
metric determination of iodine. ............... 0.02002
PUMA SATRECATIGI-IUNABE. 2... ee oo ewe es
WiuuiaAM Savant: The influence of alcohol on the metabolism
UTS A GEV COMEW ike cc o's Ss ed a stare 2 ase don Soe eee
H. D. Dakin and Mary Dows Herter: On the production of
phenolic acids by the oxidation with hydrogen peroxide
of the ammonium salts of benzoic acid and its derivatives,
with some remarks on the mode of formation of phenolic
Pubsianeesun the orzanigm: 2). 2.02) 22 225 22o5 aes oe
H. D. Dakin: The action of arginase upon creatin and other
TUT POTEET (6 Fel 2957S aa ed a
Mary F. Leacuw: On the chemistry of Bacillus coli communis: 11.
he aon-POIsanONS POTION, . .. 20. ia... eons ce hee
Tomas A. RutHerForD and P. B. Hawk: A study of the com-
parative chemical composition of the hair of different

A.D. EmMmert and H.S.Grinpitey: Chemistry of flesh: Further
studies on the application of Folin’s creatin and creat-
inin method to meats and meat extracts............

403

459

PROCEEDINGS OF THE

AMERICAN SOCIETY OF BIOLOGICAL CHEMISTS

’

IN SESSION IN

Washington, D. C., May 8 and g, 1907.

EDITED BY THE SECRETARY.

Summary of Meetings.

1. Wednesday morning, May 8, at the George Washington
Medical College.

2> Wednesday afternoon, May 8, at the George Washington
Medical College.

3. Thursday morning, May 9, at the George Washington Medi-
cal College. Joint session with the American Physiological
Society.

4. Thursday evening, May 9, at the Cosmos Club. Joint ses-
sion with the Washington Section of the American Chemical

Society.

PROCEEDINGS OF THE AMERICAN SOCIETY OF
BIOLOGICAL CHEMISTS.

Washington, D. C., May 8 and 9, 1907.

First Meeting.

George Washington Medical College. _Wednesday morning,
May 8.
Presiding officer: The Vice-President, John J. Abel.

ON THE BEHAVIOR OF FROG’S MUSCLE TOWARD
ACIDS.

By JOHN J. ABEL.
(From the Pharmacological Laboratory of the Johns Hopkins Umiversity.)

This subject was taken up with the view of describing the
chemical steps involved in the toxic action of acids, or more pre-
cisely stated, to determine whether the altered state of the inor-
ganic constituents is more deleterious to muscle than the forma-
tion of a certain amount of acid-proteid compound, dissociated
or non-dissociated.

When frog’s muscle (gastrocnemius) is immersed in an isotonic
solution of sodium chloride (100 cc.) to which five or ten cc. of a
tenth-normal acid have been added, the problem becomes one
which deals with diffusion and progressive swelling, rather than
one which deals with osmotic pressure effects developed by a
diaphragm which prevents the diffusion of certain newly formed
substances from the interior of the muscle. This point of view is
forced upon one by the following experiments. A muscle that
has been immersed for two hours in an acidulated isotonic saline
solution is frozen and cut. The cross section shows an outer
layer more homogeneous, coagulated, and impregnated with acid
and water, while the core of the muscle is normal in appearance;
t.e., it differs in no respect from the core of a normal muscle that

Vill

Society of Biological Chemists 1X

has been frozen and cut. Muscles taken from frogs that have
received acid fuchsin are well adapted for the study of this diffu-
sion of acids and the progressive formation of the watery zone,
especially when the muscles are immersed in pure solutions of the
acids (4 to 7/5 normal).

Perfusion of frogs through the bulbus aorte with acidulated
saline solutions also gives results that lead to the above conclu-
sion, and furthermore shows that an important rdle must be
ascribed to their varying rate of diffusion when the acids are
applied as in Loeb’s experiments.

A muscle that has been perfused with an acidulated (HC1) iso-
tonic salt solution for ten or twelve minutes is quite normal in
appearance, butis little or not at all responsive to electrical stimu-
lationand willtake up little or no water when itis immersed inan
isotonic acid-free salt solution. The weight-time curve in this
case is practically a straight line. If, however, such a muscle is
immersed in 50 or 100 cc. of the acid solution that was used in per-
fusing, it will take up water exactly as do normal muscles, al-
though we have here no difference in the composition of the fluid
within and without the muscle. Muscles that have been per-
fused for an hour or more and which have swelled greatly, will
of course take up water less rapidly from an acidulated solution
in which they are immersed. The swelling proceeds from with-
out inwards in proportion as acid is absorbed from the outer fluid
and the weight-time curve is exponential in character, the rapid-
ity of water absorption at any given moment being proportional
to the amount of muscle substance that is still capable of binding
water.

A muscle that has been perfused with hydrochloric acid-saline
solution for ro minutes and which is then immersed in an isotonic
acid-free saline solution for from 50 to 70 hours, produces less acid
than a muscle which has been perfused with an acid-free solution
of sodium chloride and which has remained for the same length
of time in isotonic salt solution. The latter finally swells to a
considerable extent, probably in consequence of the appearance
of non-diffusible autolytic products, the development of which is
inhibited in the acid-perfused muscle.

The reversed reaction, that is, the loss of water by an acid-
perfused, swollen muscle when it is immersed in isotonic salt

x Proceedings

solution, and the rate of diffusion of acid from such muscles have
also been studied.

Two acids, hydrochloric and acetic, representing two extremes
of electrolytic dissociation, have been compared in respect to
their ability to cause swelling when distributed throughout the
substance of the muscles by the perfusion method. Loeb found
that these two acids among others differ greatly in their power to
induce swelling of muscles immersed in saline solutions that con-
tain these acids. Organic acids, he observed, act less powerfully
as a rule than the highly dissociated mineral acids. No good ex-
planation has been offered for this difference. Loeb found that
hydrochloric acid causes muscles to take up 9 per cent of water
in one hour, while acetic acid induces them to take up only 3.9
per cent in this time. A study of his data for an immersion of
eighteen hours, shows that this ratio alters with the time of immer-
sion, as the water intake under the influence of hydrochloric acid
now attains 47.5 per cent, while under the influence of acetic acid
itamountstoonly 14.2 percent. Thesenumbers stand inthesame
relation to each other as the diffusion constants of the two acids.
When muscles are perfused, the surfaces of contact between acid
and tissue are so enormously increased that the modifying influ-
ence of the varying diffusivity of the acids should disappear.
Pairs of frogs of equal weight were perfused with 100 cc. of iso-
tonic salt solution, made up in the one case with the hydrochloric
and in the other with acetic acid in equivalent amounts (V = 100,
Loeb), one gastrocnemius being removed in each case as a con-
trol. The duration of perfusion was one hour and the amount of
swelling varied from 18 to 40 per cent with both acids according
to the size of the frogs used. When the two acids are applied in
this way they show no difference in their power to cause swelling in
muscles of the same weight. The muscles perfused with hydro-
chloric acid were quite dead, while those perfused with acetic
acid still responded fairly well to electrical stimulation, thus
showing that swelling and toxic action do not take a parallel
course irrespective of the acid used.

Acids in general diffuse in the order of the migration velocity
of theiranions. When muscles are immersed in acidulated solu-
tion an amon effect may, therefore, be said to become manifest
through variations in the diffusivity of the acids used. Other

Society of Biological Chemists x1

pairs of acids will be compared in the same way and the partition
of the sodium ion between the acid which is to be tested and that
displaced from the neutral salt of the medium (NaCl or other salt)
as also the partition of the acids among the constituents of
muscle will be taken into consideration. The points raised in
the introduction, together with other questions pertaining to
diffusion and dissociation will be discussed more fully in a later

paper.

A NEW REAGENT FOR THE RECOGNITION AND ESTI-
MATION OF FREE HYDROCHLORIC ACID IN GASTRIC
CONTENTS.

By J. H. KASTLE anp H. L. AMOSS.

(From the Hygienic Laboratory, U. S. Public Health and Marine Hospital
Service, Washington, D. C.)

It was observed by Schénbein in 1854 that the coloring matters
present in flowers can be bleached by means of sulphur dioxide
and the color restored by strong acids. In 1905 Kastle con-
firmed these observations and pointed out that the affinities of
acids could be determined colorimetrically by means of vegetable
coloring matters which had been bleached by means of sulphur
dioxide. Evidently, therefore, such chromogens as result from
the action of sulphur dioxide on certain vegetable coloring matters
are reagents for the hydrogen ion. It therefore occurred to one
of us (Kastle) that such substances might be employed as re-
agents in the detection and estimation of free hydrochloric acid in
the gastric contents. Such has been found to be the case. Asa
matter of convenience the coloring matter of red cabbage has
been employed in the preparation of the reagent. This is pre-
pared by macerating the leaves of the cabbage with water, bleach-
ing the solution with sulphur dioxide, boiling to expel the excess
of sulphur dioxide, and filtering until clear. Sometimes small
amounts of white of egg are added to properly clarify the solution.
When a small amount of this solution is added to 4 hydrochloric
acid or to the clear filtrate from normal stomach contents a pur-
plish red color is developed in the solutions. In determining the
quantity of free hydrochloric acid in gastric contents 1 cc. of the
filtered gastric contents is brought together with 2 cc. of the

xii Proceedings

reagent and 1 cc. of water. This solution is then allowed to stand
10 minutes when it is compared in a tintometer with a solution
containing 1 cc. of 4; hydrochloric acid, 2 cc. of the reagent and 1
ee. of water. In our experiments a Dubosc-Pellin colorimeter
was used. In one experiment the scale of the instrument on the
side of the tube containing the gastric contents was arbitrarily
set at 10 divisions on the scale, the scale on the side of the instru-
ment containing the tube with the #4 hydrochloric acid was then
found to read 4.2 divisions when the color in the two half circles
of the instrument matched in depth and tint.

To find the quantity of free hydrochloric acid present in 1 cc. of
the filtered gastric contents we have therefore,

0.00305 X 0.42) = o.0015 32

Hence this particular specimen of gastric contents contained

1.533 parts of free hydrochloric acid per thousand.

PHENOLPHTHALIN AS A REAGENT FOR OXIDASES
AND OTHER OXIDIZING SUBSTANCES IN PLANT AND
ANIMAL TISSUES.

By J. H. KASTLE.

(From the Hygienic Laboratory, U. S. Public Health and Marine Hospital
Service, Washington, D. C.)

Phenolphthalin or dioxytriphenylmethane carbonic acid is
readily converted into phenolphthalein by various oxidizing
agents, such as lead peroxide, potassium permanganate, potassium
ferricyanide, etc. Several years ago Kastle and Shedd observed
that it is also readily oxidized by the plant oxidases and recom-
mended its use as a reagent for these substances. Since then it
has been employed by a number of observers in the study of plant
oxidases. Attention was called to the peculiar conduct of blood
and animal tissues toward phenolphthalin. Ordinarily blood and
animal tissues are incapable of oxidizing phenolphthalin. On the
other hand blood and animal tissues readily oxidize phenolphtha-
lininalkalinesolutions. In oxidizing power blood and the animal
tissues thus far studied stand in the following order:

oe — eh CO

Society of Biological Chemists X1ll

Suprarenal bce 2 LN Se A Pe dense ais

nee MN le Sy, JS a ets Se wr eandiava e101, Sars Se 25
IGG ng a re en Zak
INSEE ES nee a ae ei ee 2.5

The substances in the blood and animal tissues capable of oxi-
dizing phenolphthalin in alkaline solution are weakened in their
activity by boiling and completely destroyed by certain poisons,
such as chlorine, bromine, and hydrocyanic acid. and also by
incineration.

The oxidation of phenolphthalin by blood has been found to
vary with the concentration of all the substances involved in the
oxidation, viz: with that of the alkali, the phenolphthalin itself
and with the quantity of blood present. On drying in the air the
oxidizing power of blood towards phenolphthalin is gradually
weakened but not destroyed.

It is the intention of the writer to continue these studies with
the view of throwing light on the kinetics of this oxidation.

PROTEIN METABOLISM IN EXOPHTHALMIC GOITRE.

By PHILIP SHAFFER.

(From the Department of Experimental Pathology, Cornell University Medi-
cal College, New York City.)

The author has studied the protein metabolism in twelve typical
cases of exophthalmic goitre, by a fairly complete analysis of the
nitrogenous substances, and in some cases of the sulphur partition
of the urine. The diets were in all cases free from meat products,
and in most cases were fairly constant.

Two products of metabolism are of especial interest in the dis-
ease because of their very great deviation from the normal.
These are creatin and creatinin.

In spite of the pathologically increased tissue catabolism in
most of the cases, resulting ina fairlyrapid loss of body weight,
the amount of creatinin excreted was very low. From his own
results and those of Folin and others in the literature, the author
has concluded that a normal healthy individual excretes from 20
to 30 milligrams of creatinin per day for each kilo of body weight;
and this factor the author has termed the “‘creatinin coefficient”
(z. e., number of milligrams of creatinin per kilo body weight).

XiV | Proceedings

The creatinin coefficients in these cases varied from7 to 12 (one
mild case being 16.8),or were less than half the normal. This
fact is accepted as evidence that creatinin is not a product of total
tissue catabolism, but is a product of certain normal cell processes,
which in many diseased conditions may be extremely sluggish in
their intensity,even though, as in exophthalmic goitre, the total
tissue catabolism may be much increased. The low creatinin
coefficients in all marked cases of exophthalmic goitre—subjects
of which disease are especially prone to muscular weakness—are
also accepted in support of the author’s hypothesis that creatinin
is an index of muscular tonus or of muscular and perhaps of gen-
eral cellular efficiency.

Eight of the twelve cases excreted considerable amounts of
creatin; in some cases even more creatin than creatinin. The
amount of creatin appears to bear some relation to the severity
of the condition. Reasons were given for believing that creatin
in the urine is pathological (except when creatin is taken in food),
and has a significance different from creatinin; aside from the
chemical relationship there isas yet no experimental basis for the
common assumption that creatin and creatinin have a close
physiological connection. Their appearance in urine is the result
of different causes.

ON THE BACTERIAL PRODUCTION OF SKATOL AND
ITS OCCURRENCE IN THE HUMAN INTESTINAL
TRACT.

By C. A. HERTER.
(From the Laboratory of Dr. C. A. Herter, New York.)

Observations upon skatol produced in the course of putre-
factive decomposition are at present few and imperfect. This is
due largely to the difficulties incidental to the certain recognition
of this substance when presentin smallamounts. Bymeans of a
method described by Herter and Foster it is possible to detect the
presence of very small quantities of skatol in a putrefactive mix-
ture, to separate skatol from indol and to estimate the quantity of
skatol present. This method is based on the use of £-naphtho-
quinone sodium monosulphonate and paradimethylamidobenz-
aldehyde(Ehrlich’s aldehyde). By means of this method, studies

Bat

Society of Biological Chemists XV

have been made with a view to discovering what organisms are
chiefly concerned with the production of skatol and many obser-
vations have been made upon the presence of skatol in the human
intestinal tract. A large number of facultative and strict ane-
robic organisms have been studied with respect to their ability to
form skatol. The anerobes, B. putrificus (strain isolated by
Bienstock) and one strain of the bacillus of malignant edema
(obtained from Prof. Theobald Smith) were found to produce
skatol in peptone bouillon, although it was not possible to deter-
mine the conditions under which skatol could be regularly ob-
tained through the action of these bacteria. It was found that
skatol is rarely present in the intestinal tract except in conditions
of disease associated with intestinal putrefaction. Usually skatol
is associated with indol insuchconditions, but there are instances
in which the intestinal contents contain little or no indol, and,
relatively speaking, considerable skatol. This has been observed
heretofore only in putrefactive processes associated with pro-
nounced clinical manifestations.

THE CHEMICAL COMPOSITION OF THE LIVER IN ACUTE
YELLOW ATROPHY.

By H. GIDEON WELLS.

(From the Sheffield Laboratory of Physiological Chemistry, Yale University,
and the Pathological Laboratory of the University of Chicago.)

The report comprised a summary of certain of the results of a
complete analysis of a liver obtained shortly after death, from a
typical case of idiopathic acute yellow atrophy. The most im-
portant of the findings was the isolation, in sufficient quantity to
be identified, of a considerable number of amino-acids, some of
which had not previously been found free in human tissues.
These were leucin, tyrosin, glycocoll, alanin, pyrrolidin-carbonic
acid, glutaminic acid, aspartic acid, andlysin. Histidin was also
present, but not isolated. Arginin, phenylalanin and trypto-
phan, although sought for, were not found. Xanthin and hypo-
xanthin were also found free in the liver extracts,but no adenin or
guanin. Altogether over 8 grams of amino-acids were isolated
from the watery and alcoholic extracts of 700 grams of liver tissue,
corresponding to about 12 grams in the entire liver, which amount

XVi Proceedings

probably represents but a small part of the free amino-acids that
were actually present on account of the inefficiency of the avail-
able methods for separating amino-acids. On account of the
relatively large quantity of amino-acids that seems to have
occurred free in the liver in this case,the writer is inclined to agree
with Neuberg and Richter in doubting if all the amino-acids pres-
ent could have been derived from the autolysed liver cells. There
was a slight decrease in the diamino-nitrogen (according to Haus-
mann’s method), but not so great as observed by Wakeman in
phosphorus poisoning of dogs. Sulphur was normal,and phos-
phorus increased. The amount of fat, both free and combined,
was below normal.

THE APPEARANCE OF MILLON’S REACTION IN THE
URINE IN THE ABSENCE OF PROTEINS, AS A CRI-
TERION IN THE TUBERCULIN-REACTION.

By C. VOEGTLIN.
(From the Medical Clinic of the Johns Hopkins Hospital.)

Normal urine does not give under ordinary circumstances a
positive reaction with Millon’s reagent; unless the urine contains
protein of some sort which possesses the tyrosin or phenylalanin
molecule (oxyphenyl-reaction), the rose red tint to the precipi-
tate does not appear. It was found, that whereas injections
of tuberculin into normal individuals are not followed by the
elimination of substances in the urine which give a positive
reaction with Millon’s reagent, after tuberculin injections into
patients suffering from all sorts of tuberculous infections, the
urines yield positive reactions in the first 24 hours after injec-
tion. The test was applied inthe following manner: First, the
urine was tested for proteins, the absence of which was always
noted. Then the urine was treated with a solution of lead
acetate, to precipitate the coloring matter. After filtration a
practically colorless liquid was obtained. Millon’s reagent was
added drop by drop to the clear fluid until a pink color appeared.
Heating the solution is not necessary. From one of the
urines under examination that yielded a positive test, a small
amount of tyrosin was isolated. While the possibility of other
abnormal constituents in the urine of. such cases giving the

|
7
.
(

Society of Biological Chemists XVil

reaction is borne in mind, it is believed that tyrosin is prob-
ably the substance which determines the positive outcome of this
test. Still further evidence has to be collected to prove definitely
this supposition, and to rule out other possibilities. It may be
said that minimal quantities of a solution of tyrosin added to
normal urine will yield a positive reaction, with the reagent
under consideration. The test, therefore, seems to be extremely
sensitive. The intention is to continue the work in this line.

A METHOD FOR THE DIRECT DETERMINATION OF
HEATS OF REACTION.

By LAWRENCE J. HENDERSON anp CHARLES T. RYDER.
(From the Laboratory of Biological Chemistry of the Harvard Medical School.)

Most thermochemical problems involve the determination of
differences in heats of formation or heats of combustion, because
they aim to determine heats of reaction, actual or hypothetical.
Accordingly the need of a method for the direct determination of
heats of reaction, even of those which proceed slowly, or a method
in some way differential, had long been felt. The lackof such an
aid to research is mainly responsible for the incorrect ideas still
widely current concerning the regularities among the thermo-
chemical data of organic substances; for here the significant dif-
ferences are so small as often to fall within the limits of error of
the determination, and seldom greatly to exceed them. Physio-
logical reactions are not less subject to doubt on this score, as the
recent investigations from Tangl’s laboratory indicate.

The present investigation was aimed to test a new method for
the direct determination of the heats of reaction in which the
chemical change is a slow one. A thermostat filled with water
and maintained at about thirty-nine degrees Centigrade, with a
variation in temperature of not more than one or two hundredths
of a degree, was set up. This apparatus differed inno way from
instruments commonly in use in physico-chemical laboratories.
Thereaction mixture after being brought closely to the temperature
of the thermostat was placed in a Dewar flask which rested on a
tripod within the thermostat. A Beckmann thermometer was
passed down through a glass tube inserted in the rubber stopper
of the Dewar flask and projected into the reaction mixture. The

XVli Proceedings

course of the temperature change, as indicated by the Beckmann
thermometer, was recorded during several days. It was shown
by preliminary experiments that evaporation from the flask did
not occur in sufficient amount to influence appreciably the tem-
perature of the flask. It was also shown that Newton’s law of
cooling holds under these circumstances, so that the rate of cool-
ing or warming due to differences in temperature between the
contents of the flask and the thermostat can be calculated. For
the present apparatus this cooling amounts to less than one-
tenth of the temperature difference in twenty-four hours.

In the experiments so far undertaken the tryptic digestion of
casein has been studied, using concentrated solutions of casein
dissolved in dilute sodium carbonate. After the first hour of
the experiment the progress of the reaction as measured by the
rise in the temperature is regular. There are, however, marked
indications of a slowing of the process, very probably due to the
retarding influence of the end products of the reaction. This is
a new method for following the progress of a slow reaction, and
will be further investigated in this laboratory. According to the
data thus far obtained, the heat of this reaction is very small but
positive, unless indeed the observed temperature change be due
to secondary reactions. The magnitude of the heat of reaction
probably lies between two-tenths and four-tenths of a great calorie
per gram-molecule of water added.

With the aid of this method it is planned to study in detail
various biochemical reactions in the hope of characterizing them
more definitely than is now possible. The method will also be
extended to the study of reactions of organic chemistry.

It seems certain that in this way an accuracy from ten to one
hundred times greater than with the aid of differences in the
heats of combustion can be attained in the measurement of
reactions which involve little heat change. A peculiar advan-
tage of the method is that moderate slowness of reaction is a
favorable circumstance.

ON THE CONVERSION OF GLYCOGEN INTO GLUCOSE.

By, A: E. TAYLOR.

Society of Biological Chemists X1X

ON THE INFLUENCE OF THE CONCENTRATION OF
THE HYDROXYL IONS OF A SALT SOLUTION UPON
THE PHYSIOLOGICAL EFFECTS OF ITS CATIONS.

By JACQUES LOEB.

FATTY TRANSFORMATION IN THE LIVER.

By H. C. JACKSON anp L. K. BALDAUF.

THE EXISTENCE OF CHOLESTERYL ESTERS OF THE
FATTY ACIDS IN GALL STONES AND THEIR BEAR-
ING UPON THE FORMATION OF CHOLESTERIN GALL
STONES.

By J. G. ADAMI ann OSCAR KLOTZ.

Second Meeting.

George Washington Medical College. Wednesday afternoon,
May 8.
Presiding officer: The Vice-President, John J. Abel.

THE METABOLISM OF NITROGEN AND SULFUR IN
PNEUMONIA.

By ALEXANDER LAMBERT anp C. G. L. WOLF.

(From the Fourth Medical Division, Bellevue Hospital, and the Department of
Chemistry, Cornell University Medical College, New York.)

Tables showing the results of the complete analysis of the urine
in severe and fatal cases of pneumonia were exhibited.

On a non-nitrogenous diet of sufficient calorific value, 30 grams
of nitrogen were eliminated. The relative excretion of urea was
low. The creatinin excretion was high during the pyrexia, and
fell rapidly after the crisis. This was also the case with uric acid.
Very considerable quantities of creatin were eliminated during
the febrile period. This also diminished, or disappeared after
the crisis. Undetermined nitrogen, amounting in one case to 5.8

XX Proceedings

grams, was excreted in the twenty-four hours. The amount of
neutral sulfur was six to eight times greater than that observed
in normal individuals under similar conditions of diet.

A BRIEF NOTE ON A SOURCE OF ERROR IN THE USE OF
A CERTAIN PETROLEUM ETHER AS AN EXTRACTING
MEDIUM.

By JOHN MARSHALL.

(From the Robert Hare Chemical Laboratory, Department of Medicine,
University of Pennsylvania.)

The fraction from a commercial petroleum ether, derived from
the Pennsylvania oil field, which was obtained between 20°—50°C.,
was found, on evaporating at 200 cc. at room temperature imme-
diately after distillation, to leave no residue, but on standing 30
days at room temperature in a stoppered flask with an air space
of about one liter above the surface of the liquid and exposed to
diffused sunlight, it was found that on spontaneously evaporating
200 cc. of it at room temperature and then over sulphuric acid, a
cosmoline-like residue weighing 0.0072 gram remained. The re-
maining fluid was redistilled at 20°-50°C.and 200 cc. of the distil-
late on evaporation left no residue, but a portion of 200 cc. of the
remainder of this distillate on standing in a flask, as above de-
scribed, for 1o days left on evaporation spontaneously and over
sulphuric acid a residue of 0.0007 gram and 200 cc. of the remain-
ing liquid on standing 17 days left a residue of 0.0016 gram. On
making a blank test with 200 cc. of that which had stood 17 days
by using it in a Soxhlet extractor, containing a fat-free, paper
thimble filled with washed and dried sand, for 14 hours and filter-
ing the liquid through a fat-free filter and evaporating spontane-
ously and over sulphuric acid, a residue of 0.0138 gramremained.
Two hundred cc. of a distillate which had been twice distilled at
20°-s5o° C. and had stood 215 days in a flask, as above described,
left on evaporation at room temperature and at 100° C. for two
hours a residue of 0.0494 gram, which at the latter temperature
turned black. Nineteen hundred cc. of a distillate which had
been twice distilled at 20°-50°C. and which, after distillation, had
stood 217 hours, was redistilled at 20°-50°C. All but 135 cc. dis-
tilled over at the temperature stated and on evaporating this

Society of Biological Chemists XX1

spontaneously a yellowish, cosmoline-like residue remained which
on being heated for 3 hours at 100° C. turned black and weighed
0.5062 gram. The composition of the residues was not deter-

-mined. In every case of distillation care was exercised that the

liquid in the distilling flask should not be in ebullition but that
evaporation should occur in such a manner that the surface of the

_ liquid remained placid. Weighings were made until constant

weight was obtained. The cork stoppers employed had been
previously extracted with ethyl ether. In some cases Glinsky’s
fractionating bulbs were employed. In every case there was an
open space of about a liter above the surface of the liquid in the
corked storage flask and the liquid was exposed to diffused sun-
light. It is evident that a petroleum ether of this sort is not
adapted for use in making extractions.

A CLINICAL METHOD FOR DETERMINING THE ALKA-
LINITY OF THE BLOOD.

By HERMAN M. ADLER (by invitation).

(From the Laboratories of Clinical Pathology and Biological Chemistry
of the Harvard Medical School and the Boston City Hospital.)

The methods used for the determination of the alkalinity of
the blood have in the past been based on titration. These have
been inaccurate because they yield information regarding only
the absolute quantity of acid or alkali present and none regarding
the equilibrium between the two. Since the work of Salm on the
exact point of H ion concentration at which a large number of
indicators turn, it has been possible to select certain indicators
whose turning points correspond to the concentration of the
blood. Rosolic acid seemed to be the best of these. By pre-
paring filter paper with 0.1 percent solution of rosolic acid, the test
could be applied clinically to the serum from a large number of
cases. It was foundthat the blood wasmaintained at about 2—3
X10 ‘nH inhealth. Inthe terminal stage of acidosis a distinct
variation towards the acid color was observed. In coma from
meningitis no such change was observed. Theinterdependence of
H ionization and CO, content of the blood has been shown by
Henderson and Black; hence this method is alsoa measure of the
latter quantity.

xxii Proceedings

CALCIUM METABOLISM IN A CASE OF MYOSITIS
OSSIFICANS.

By A. E. AUSTIN.

(From the Medical Chemistry Laboratory of Tujts College and the Corey Hill
Hospital.)

From the literature the following principles are presented as a
standard of normal metabolism of calcium. Calcium equilibrium
has been maintained with anintake of 0.688—0.860 gram of calcium
daily, reckoned as the oxide, but it is betterto regard 1-1.5 gram
as the daily need of every healthy adult and less than this as
insufficient. No dependence can be placed upon the relation of
feces lime to urine lime on account of its great variations and any
results based upon urine lime alone are useless. Ordinarily 5-10
per cent of the ingested lime is found in the urine, while the actual
amount of lime found in the urine varies from o.15—0.5 gram,
depending upon character of the food; that is, whether vegetable
or animal food. The amount of lime eliminated in the urine may
be increased by increased ingestion of water (v. Noorden), by the
use of hydrochloric acid (Hammarsten), by lactic acid and sodium
lactate (Rumpf), by bodily rest (Hoppe-Seyler), and by calcium-
poor food (Rumpf). In disease impoverishment of the body in
lime has been found in osteomalacia (McCrudden), in pernicious
anemia (v. Moraczewski), in inanition (Ott), and in diabetes
mellitus (Forster). Physiologically, retention of calcium can be
attained by ingestion of large quantities without any harmful
influences—pathologically, retention undoubtedly occurs in arte-
rio-sclerosis. In order to determine whether a metabolism is
pathological or not, we must learn how the individual complies
with these conditions: Does he maintain a calcium equilibrium
with 1-1.5 gram of this material in his daily food? Does he
respond toa diminished intake with an increased procentual outgo
of calcium; in other words, with an impoverishment of the body in
lime. The individual under investigation, who had long been
affected with myositis ossificans, and in whom a retention of lime
was suspected, upon an ingestion of 1.246 gram calcium oxide
maintained practically an equilibrium.

The lime eliminated in the urine was 0.472 gram and in the
feces 0.787 gram, leaving a negative balance of 0.011 gram daily.

i —_ EEE

Society of Biological Chemists XXlil

When, however, in a subsequent period of seven days, the intake
was reduced to 0.572 gram daily, there was found urine lime
amounting to 0.425 gram and feces lime to 0.473 gram, leaving a
deficit of 0.327 gram daily. Under the latter condition the urine
lime formed a much larger percentage of the lime ingested than
under the former; an increase from 37.5 per cent to 47.2 per cent
was found. Ina similar investigation, Thayer and Hazen ob-
tained approximate results but attribute them to the carbo-
hydrate-free diet which was employed. Their diet, however, was
very poor in lime (0.625 gram daily) and their results seem also
to follow the same physiological law of loss of lime in the body
when insufficient lime is given. Inso far as the results of this
investigation go, no evidence of calcium retention in myositis
ossificans was found.

HYDROLYSIS OF SPLEEN NUCLEOPROTEIN.

By J. A. MANDEL ann P. A. LEVENE.

(From the Chemical Laboratory of the New York University and Bellevue
Hospital Medical College, and the Rockefeller Institute for
Medical Research.)

The comparative study of the chemical composition of chro-
matins has been directed thus far towards the study of the
nucleic acid radical. It is, however, made probable today that
more distinction will be found in the investigations on nucleo-
proteins than on the acids. In recent years the composition of
nucleoproteins was studied only once by Wohlgemuth’ on liver
nucleoprotein. The results obtained by this observer were very
unusual. In the present work the nucleoprotein of the spleen,
containing 14.155 per cent N and 1.605 per cent P (500 grams)
was subjected to hydrolytic cleavage by hydrochloric acid.

The products of hydrolysis were separated and purified by the
aid of different methods and the yield calculated for roo grams of
the nucleoprotein was as follows:

MGs stectiinl HACIA fey alts ge wa ae Sai amy 25.0 grams

Leucin |
Aminovalerianic acid {

1 Zeitschr. f. physiol. Chem., xxvii, xlii, xliv, and Ber. d. deutsch. chem.
Gesellsch., xxxvii, 1904.

XXIV Proceedings

Glycocoll |

Be eo OO a eee 2.0 grams
Alanin
ASHAChO ACI ;acar ie emieies aos ne mer Own
Pins sf crave ope let suxas eye eee renee friiate ts pe alee As not found
Phan yvislarim oss creas xs alee cy ea present
PVTORID Cacene eb tetats ee peta aha oon «ale Fin: ot r.o gram
LFYSinl CTA moe, Sent pels foie corns ea ales wag Whe
Arent TiGKOlOMBES cis ad bh cir nnn wie ee BAO
Pistidin PicrOlonanes cr arate kes Fee ea os Olah) he
PCOINIT cee ee a eee Te ee se sise wae OiAl = or
NT ELINIT hare ree Ne eee ee otto 3) «RN Srasschs eau Ps Ohad)
Suton este ee ae nein eee at date ot One sive wears On Gh) gat
ANT ies bu meter le cic PEL Ne eee Se ean On5 ene

A COLOR TEST FOR URACIL AND CYTOSIN.

By HENRY L. WHEELER anp TREAT B. JOHNSON.
(From the Sheffield Chemical Laboratory of Yale University.)

A characteristic purple or violet-blue color is produced when
uracil or cytosin is dissolved in bromine water,and the solutions
treated with an aqueous solution of barium hydroxide.

The formation of the color is explained as follows: Uracil and
cytosin react with bromine to form dibromoxyhydrouracil.
Barium hydroxide then converts this dibrompyrimidin into isodi-
aluric acid, which immediately undergoes a rearrangement in the
alkaline solution to give the purple barium salt of dialuric acid.

The presence of guanin, adenin, thymin, isocytosin and 6-
amino-pyrimidin does not interfere with the test.

Precise directions for the application of the test have been
published in this journal.

THE ROLE OF THE OXIDIZING POWER OF ROOTS IN
SOIL FERTILITY.

By OSWALD SCHREINER anp HOWARD S. REED (by invitation).

(From the Laboratory of the Bureau of Soils, Un'ted States Department of
Agriculture.)

Growing roots possess well defined powers of oxidation,due
principally to the activity of enzymes. The oxidizing powers
can be demonstrated by the use of reagents which produce a dye
upon oxidation. Alpha-naphthylamine, benzidine, and vanillin

Society of Biological Chemists XXV

form insoluble dyestuffs when oxidized by the roots. Phenol-
phthalin and aloin produce soluble dyes and the amount of color
produced is proportional to the amount of oxidation accom-
plished. The addition of certain substances used as fertilizers
promotes the oxidative activity of the roots. The powers of
oxidation are greater in fertile soils and their extracts than in
unproductive soils or soil extracts. The authors also pointed out
the value of oxidizing processes in aiding the decomposition of
vegetable matter in the soil.

THE PRODUCTS OF GERMINATION AFFECTING SOIL
FERTILITY.

By OSWALD SCHREINER anp M. X. SULLIVAN (by invitation).

(From the Laboratory of the Bureau of Soils, United States Department of
Agriculture.)

The waterin which seeds have germinated, although containing
nutrient salts, is by no means as good a culture medium for the
seedlings as is carbon-filtered distilled water. Even the water in
which seedlings have grown restrains the development of a
second crop planted therein. Apparently something soluble in
water and active in extremely dilute solutions accumulates in
the medium in which plants have germinated.

Considering that the juice expressed from germinating seed-
lings would represent a concentrated extract of the toxic material
present in the medium in which seeds have germinated, experi-
ments were made on the juice expressed from wheat which had
germinatedseven totendays. This juice was pale yellow, became
darker on exposure to air, was acid in reaction and was free from
microorganisms. If not too highly diluted it is toxic to growing
plants both in water and soil, and injures likewise the resting seed.
This juice acts in two ways, first, by inhibiting the life activity of
the germinating seed, leading to a diminished amount of leucin
and tyrosin, and to a lessening of the oxidation by the roots;
secondly, by diverting the metabolism somewhat.

On analysis the expressed juice was found to contain a trace
of cholin, xanthin bases and soluble organic phosphorus com-
pounds. Whether the toxic action of the juice is due to alkaloids,
such as cholin and neurin, to complex organic acids as phytic or

XXV1 Proceedings

nucleic, to simpler organic acids, or to anti-enzymes or to a com-
bination of all these is yet to be determined.

SOLUTION TENSION AND TOXICITY IN LIPOLYSIS.
By RAYMOND H. POND.
(From the Chemical Laboratory of the New York Botanical Garden.)

An effort has been made to ascertain whether lipolysis is
affected by toxic agents in a manner corresponding to that found
by Mathews! for fertilized eggs of Fundulus, by McGuigan? for
diastatic activity and by Caldwell* for proteolytic digestion.
Mathews announced as a general law that the toxic action of
cationsas such and of anionsas suchis aninverse function of and
isdetermined by their solution tension. The toxic action of any
salt is then aninverse function of the decomposition tension of that
salt which is the sum of the solution tension of the cation and of
the solution tension of the anion of that salt both values being
regarded as having the same sign.

An examination of Mathews’ evidence suggested the desir-
ability of additional dataand since his law had not been tested for
lipolytic reaction, a study of its applicability to the saponifica-
tion of ethyl butyrate by a commercial extract of pancreas has
been made. The selection of this product proved fortunate
because a solution can be prepared which still has lipolytic power
though is too dilute to be coagulated by boiling. The enzyme
activity can be determined accurately by the volume of potas-
sium hydroxide required to neutralize the acid arising from the
saponification. There is nothing to obscure the end point in
titration and the relative toxicity of a series of salts can be very
satisfactorily determined.

My results differ from those obtained by others, notably in the
relative toxicity of silverand mercury. Thelatterwasmuch more

1 Mathews, A. P.: ‘The Relation between Solution Tension, Atomic
Volume, and the Physiological Action of the Elements.”’ Amer. Jour. of
Phystol., x, pp. 290-323, 1904.

? McGuigan, Hugh: ‘The Relation Between the Decomposition Ten-
sion of Salts and their Antifermentative Properties.’’ Amer. Jour. of
Physiol., X, pp. 444-451, 1904.

5 Caldwell, J. S.: ‘‘ The Effects of Toxic Agents upon the Action of Bro-
melin.’’ Botan. Gazette, xxxix, pp. 409-419, 1905.

Society of Biological Chemists XXVII

toxic, while Mathews, McGuigan and Caldwell all found silver to
be more toxic. The relative toxicity of a series of metallic salts
is not constant with varying concentration of the enzyme as Cald-
wellfound. The salts seem to fall into two groups with regard to
the possible cause of toxicity. In one group there is evidence of
chemical action between the salt and the substance of the enzyme
solution and in the other group there is no evidence of such reaction.
My own results considered in connection with those obtained by
the men mentioned, justify the statement that the evidence upon
which Mathews’s law is based, is insufficient.

ON THE CAUSE OF A RED COLORATION IN THE IODO-
FORM TEST FOR ACETONE WHEN APPLIED TO DIS-
TILLATES OBTAINED FROM URINE PRESERVED
WITH THYMOL.

By WILLIAM H. WELKER.

(From the Laboratory of Biological Chemistry of Columbia University, at
the College of Physicians and Surgeons, New York.)

In applying to urinary distillates the iodoform test for acetone,
a pink to red coloration was observed in some instances. Dr.
Foster in this laboratory also independently noticed this colora-
tion. All these observations were made on specimens of urine
from pathological cases. The conditions under which the color-
ation was produced were such as seemed to eliminate the pre-
servative (thymol) as the cause. A study of biochemical litera-
ture on the acetone test failed to give any clue to the reason for
the red coloration.

On the assumption that the compound that gave the colored
substance in the test was a product secreted during disease, con-
siderable work was done to ascertain the nature of the substance.
On isolating it, the compound proved to be thymol. In spite of
its apparent removal (by filtration) from the preserved urines that
were subjected to distillation, thymol had nevertheless always
been present im solution in appreciable quantity and accumu-
lated in the distillates in proportions sufficient to permit of its
identification. It was found that thymol was alone responsible
for the coloration noted. On referring to the chemical literature

XXVIll Proceedings

on thymol, it was found that Messinger and Vortmann‘* described
a synthetic iodothymol compound that obviously was identical
with the product formed in the test alluded to and which evi-
dently accounted for the coloration observed. The nature of
the compound is indicated by the following formula ascribed to
it by Messinger and Vortmann:

ON THE GLYCOL FATS AND THE CHEMICAL AND PHYS-
ICAL RELATIONSHIP OF CROSS FATS.

By R. F. RUTTAN (by invitation).

Third Meeting.

George Washington Medical College. Thursday morning, May
9. Joint session with the American Physiological Society.

Presiding officers: The President of the American Physio-
logical Society, William H. Howell, and the Vice President of the
American Society of Biological Chemists, John J. Abel.

ON THE OCCURRENCE OF FERMENTS IN EMBRYOS.

By WALTER JONES anp C. R. AUSTRIAN.
(From the Laboratory of Physiological Chemistry, Johns Hopkins University.)

We have already called attention to the fact that the distribu-
tion of the nuclein ferments (guanase, adenase and xanthooxi-
dase) is different for different animal species. This was found
notably true in the case of the livers of the four species, dog, ox,
pig and rabbit, each gland being characterized and easily
distinguishable from the other three by the ferments which it
exhibits. This naturally suggests that the ferments in the adult

1Messinger and Vortmann: Ber. d. deutsch. chem. Gesellsch., xxii, p.
2316, 1889.

Society of Biological Chemists XX1X

glands of a given species will differ from those of the embryo.
We have found this tobe truein the case of pig’s liver. The livers
of small embryos (65~—75 millimeters) do not contain any of the
ferments. When the embryo has reached a length of 150-175
millimeters, adenase makes its appearance, but xanthodxidase
cannot be demonstrated. The adult liver as formerly stated con-
tains both adenase and xantho6xidase but not guanase.

PROTEIN METABOLISM IN CYSTINURIA.

By C. G. L. WOLF anp PHILIP A. SHAFFER.

(From the Chemical Laboratory and the Department of Experimental
Pathology, Cornell University Medical College, New York City).

Metabolism experiments upon a case of cystinuria. The results
wholly confirm the findings of Alsberg and Folin, that the sulfur
of hair- or protein-cystin fed to a cystinuric individual is nor-
mally oxidized to sulfuric acid; and are directly contradictory to
the conclusion of Loewy and Neuberg, that cystinuric individuals
are unable to oxidize ingested cystin. Both cystein and cystin
prepared from the patient’s urine were likewise oxidized when
given by mouth. The cystin excreted in the urine is evidently
not absorbed as such from the intestine, but must be absorbed in
the form of a larger molecule; because cystin absorbed as such
is oxidized. An increase of food protein leads to an increase of
eystin excreted but when the food protein is hydrolyzed outside
the body and the isolated cystin is given to a cystinuric patient,
the sulfur of this cystin is oxidized to sulfuric acid.

Cystin injected subcutaneously was excreted in the form of
neutral sulfur (probably as cystin). Cystein similarly injected
led to an increase of the total sulfur of the urine, the increase being
equally divided between inorganic sulfates and neutral sulfur.

The authors believe that the cystin excreted by subjects of
cystinuria has a double source. A part, and perhaps a greater
part on the usual diet, is derived directly from the food protein,
and is therefore strictly exogenous. But a second part of the
cystin appears to be independent of the food protein, and is
therefore not exogenous. At least one phase of the anomaly in
cystinuria appears to consist of the inability to oxidize that part
of the sulfur-containing protein which has not been split so far as

XXX Proceedings

the cystin stage in the intestine. That part which is absorbed as
eystin the cystinuric, as well as the normal individual, does

oxidize.
PROTEIN METABOLISM IN THE DOG.

By C..G.-L. WOLF,

(From the Department of Chemistry, Cornell University Medical College,
New York City.)

The nitrogen and sulfur metabolism in dogs in an early stage of
starvation wasexamined. Theanimals were fed on a non-nitrog-
enous diet of fat and carbohydrates containing 80 calories per
kilo. At the end of eight days a large quantity of protein in the
form of casein was administered, and the metabolism followed
during four days of subsequent starvation. In a second series,
160 calories per kilo were fed at the end of the 80 calorie period.
This was done in an attempt to change the distribution of nitro-
gen and sulfur as a consequence of the high caloric value of the
diet. The total nitrogen excretion was reduced to a level not.
heretofore observed in dogs of this weight. The ammonia nitro-
gen was relatively increased as a result of the starch and fat diet.
The oxidized sulfur was lower markedly, both absolutely and
relatively. The relative excretion rose at once on the adminis-
tration of the protein. The absolute creatinin excretion was
uninfluenced by any change in diet. The undetermined nitrogen
and neutral sulfur were increased with the administration of pro-
tein, but decreased relatively to the nitrogen and sulfur outputs,
respectively. There was no constant relation between the elimi-
nation of ethereal sulfur and indican.

ON THE GLOMERULAR EXCRETION UNDER CERTAIN
CONDITIONS.

By A. B. MACALLUM.
(From the Physiological Laboratory of the University of Toronto.)

When large quantities’ of distilled water (2-3 liters) were
swallowed in the interval of one hour and a half and the charac-
ters of the urine of each ten minute period were determined it was
found that 4 progressively diminished until it was only 0.09. At

Society of Biological Chemists XXK1

this point the sodium chloride was only 0.047 per cent. Potas-
sium salts were not present and only traces of phosphates were
found. The results seem to show in this case that the reflected
epithelium of Bowman’s capsule actively removes only pure water
from the blood stream through the glomeruli and that the salts
found in it ultimately are those washed out from the epithelial
cells of the tubules, ureters and bladder. The water, therefore, of
urine cannot be wholly the result of pressure filtration merely
but must be also a true secretion.

ON THE COMPOSITION OF THE HOURLY EXCRETION
OF URINE.

By CC) BENSON:
(From the Phystological Laboratory of the University of Toronto.)

The urine excreted hourly for twenty-four hour periods was
examined quantitatively for chlorides, phosphates and nitrogen
and also as to its conductivity, its specific gravity and the depres-
sion of its freezing point. As a result of these observations it was
found that while the constants of Bugarszky’s and Koranyi’s
formulae were exemplified in some of the hourly excretions they
were not uniformly or even generally obtained.

PEE INACBITION OF TETANY PARATHYREOPRIVA BY
PCreACis OF THE PARATHYROID GLAND:

By S. P. BEEBE.

(From the Department of Experimental Pathology, Cornell University Medi-
cal College, New York.)

The acute symptoms following the extirpation of the parathy-
roid gland in dogs have been completely inhibited by the hypo-
dermic administration of the nucleoproteid of beef parathyroid.
The symptoms may be alleviated for three days by a single injec-
tion, but it has not been possible to prevent ultimate death by
the continued injection of this proteid. If the alkaline solution
of the proteid is first boiled it fails to act. The active principle
is not destroyed by the peptic or tryptic digestion for forty-eight
hours.

XXX11 Proceedings

PROTEID SUSCEPTIBILITY AND IMMUNITY.

By VICTOR C. VAUGHAN.
(From the Hygienic Laboratory, University of Michigan.)

All the proteids with which we have worked, and this includes
bacterial, animal and vegetable proteids, can be split up by dilute
alkali in absolute alcohol into poisonous and non-poisonous
portions. The poisonous part, which is similar but not identical
in the different proteids, is freely soluble in absolute alcohol and
in water, more freely in the former than in the latter. It gives
the general proteid color reactions with the exception of that of
Molisch, and it does not contain a carbohydrate group. It is
also free from phosphorus. From its alcoholic solution, it is
precipitated by ether. Its aqueous solution is distinctly acid
to litmus and slowly decomposes sodium bicarbonate. The free
poison when administered intra-abdominally, subcutaneously or
intravenously kills guinea pigs in a few minutes. At first the
animal shows evidence of peripheral irritation, then 1t becomes
partially paralyzed, and finally it develops violent convulsions
that terminate in death.

Guinea pigs can be sensitized to any of the proteids with which
we have worked. In case of the vegetable and animal proteids
the first dose has no recognizable effect on the animal, but the
second dose, provided that there has been a proper time interval
which varies somewhat with the different proteids, profoundly
affects and may kill the animal. The symptoms induced by the
second dose of the unbroken proteid ina sensitized animal are
identical in character and sequence with those induced by the
first dose of the free poison. It is evident from this that the sen-
sitized animal splits up the proteid to which it has been sensitized
with the liberation of the poison, much as the proteid may be
broken up in the retort with the alcoholic solution of alkali.
Undoubtedly this cleavage is carried out much more efficiently
in the animal body than in the retort, and in the body it is most
likely due to the action of a specific proteolytic ferment, which is
called into existence by the first injection of the proteid, is stored
in the animal cell as a zymogen, and is activated by the second
injection.

Society of Biological Chemists XXXill

The non-poisonous portion of the split proteid contains all the
phosphorus present in the unbroken body, and when the original
proteid possesses a carbohydrate group this remains in the non-
poisonous split product. The non-poisonous portions of some
bacterial proteids immunize animals to living cultures of homol-
ogous bacteria, and that of some vegetable and animal proteids
sensitizes animals to the proteids from which they are derived.
Hypersusceptibility or anaphylaxia, as it has been called, and
immunity, although apparently antipodal, are in reality identical
in their essentials. In both there is developed in the animal body
the capability of splitting up the specific foreign proteid. If the
foreign proteid be a living one—a bacterial cell—and the speci-
ally prepared ferment splits it up before it has time to multiply,
the animal is immunized and its life is saved. If the foreign
proteid be a dead one, and if it be present in sufficient quantity to
furnish a fatal dose of the poison on being split up, the animal
dies.

ON THE CHEMICAL RELATION BETWEEN COLLAGEN
AND GELATIN.

By A. D. EMMETT anv WILLIAM J. GIES.

(From the Laboratory of Biological Chemistry of Columbia University, at the
College of Physicians and Surgeons, New York.)

We have been unable, by Hofmeister’s method of continuous
drying at 130°C., to convert either pure gelatin or commercial
gelatin to collagen. The resultant desiccated products were
somewhat less soluble than the original gelatin both in water and
in dilute sodium carbonate solutions at 40° C., but were digested
with apparently the same readiness in neutral trypsin solution at
40° C., in which even very minute fragments of collagen fibers
remained unaffected for hours.

It appears that Hofmeister attached too much significance to
the difference in solubility between the unmodified gelatin and
the desiccation product. The observed difference may have
been due to decomposition instead of to simple dehydration. Be-
sides, he did not apply the tryptic digestion test to his products to
ascertain whether they resembled collagen in resistance to tryp-
tolysis.

XXXIV Proceedings

When fresh tendon, ossein shavings, and pure collagen from
bone and tendon were boiled in water about two hours for the
production of gelatin, considerable ammonia was liberated from
each. When gelatin was subjected to the same treatment, ammoma
was not eliminated.

We believe that the so-called collagen obtained by Hofmeister
from gelatin by desiccation at 130° C. was not collagen, and that
his conclusion and the prevalent view that gelatin is a simple
hydrate of collagen are not well founded. That gelatin arises
from an intramolecular rearrangement of collagen on treating
the latter with boiling water and that the resultant gelatin is not a
simple hydrate of collagen, are shown by the fact that ammonia is
liberated from the collagen when the latter is converted into
gelatin.

EMBRYO-CHEMICAL STUDIES—THE PURIN METABOL-
. ISM OF THE EMBRYO.!

By LAFAYETTE B. MENDEL.
(From the Sheffield Laboratory of Physiological Chemistry, Yale University.)

In connection with a more extended series of chemical investi-
gations in embryonic growth conducted by the writer, Dr. P. H.
Mitchell has studied the occurrence and development of the
enzymes associated with the metabolism of the purins. The
more important conclusions already reached may be summarized
as follows:

1. The liver of the embryo pig contains adenase but no guan-
ase. In this respect it resembles the adult liver of the same species.

2. Anextract of embryo viscera exclusive of the liver readily
indicates the presence of guanase.

3. The view that guanase and adenase have an unlike distri-
bution and are therefore specific enzymes receives new corrobo-
ration from such results.

4. Extracts of the organsof embryo pigs have failed to demon-
strate any capacity for forming uric acid from free purins, even

' The expenses of this investigation were shared by the Carnegie Insti-
tution of Washington and the Sheffield Laboratory of Physiological Chem-
istry, Yale University. The data will later be published in detail.

_—

Society of Biological Chemists XXXV

after prolonged digestions of five days’ duration in the presence
of oxygen. The added purin bases, although deamidized, could
be quantitatively recovered. Xanthoodxidase was therefore not
recognized in such extracts.

5. On the other hand, extracts of the livers of adult pigs or of
very young pigs, are capable of forming uric acid.

6. The uricolytic enzyme has not been identified in any extract
of embryo organs under conditions in which its presence is easily
demonstrated in adult tissues.

7. Both the xanthodxidase and the uricolytic enzyme appar-
ently begin to functionate either in the last stages of embryonic
life or soon after birth.

The tardy appearance of some of these enzymes is of interest
in view of the peculiar character of the physiological processes
characteristic of young or growing organisms.

NOTES ON THE THYROID.
By REID HUNT.

THE DISTRIBUTION OF SULPHUR AND PHOSPHORUS
IN THE HUMAN BRAIN.

By WALDEMAR KOCH.

Fourth Meeting.

Cosmos Club. Thursday evening, May 9. Joint session with
the Washington Section of the American Chemical Society.

Presiding officers: The President of the Washington Section of
the American Chemical Society, Peter Fireman, and the Secretary
of the American Society of Biological Chemists, William J. Gies.

CHEMICAL AND BACTERIOLOGICAL STANDARDS NOW
IN USE IN WATER ANALYSIS.

By J. H. KASTLE.

(From the Hygienic Laboratory, U. S. Public Health and Marine Hospital
Service, Washington, D. C.)

The various chemical and bacteriological standards now in use
in wateranalysis were reviewed and discussed. Particular atten-

XXXVI Proceedings

tion was paid to the recent work of Leighton on this subject,’ in
which this author shows that the results of the chemical analysis
are ofttimes at variance with experience so far as the purity or
pollution of a natural water is concerned. In the light of all of
the facts, as they are known to us at the present, three things
must be known before we can form a correct opinion regarding
the purity or pollution of a natural water: (1) The source of the
water and the sanitary character of the drainage area; (2) the
number and kind of microédrganisms present in the water, and (3)
the nature and amount of the chemical impurities.

By way of illustration the several water supplies of the District
of Columbia were considered and also the purification of the
water of the Potomac river effected by sedimentation and sand
filtration, as shown by the chemical and bacteriological findings.
Special attention was called to the parallelism found to exist
between the turbidity of the Potomac water at various stages in
its purification by sedimentation and sand filtration, and the
number of bacteria per cubic centimeter and also to the parallel-
ism between the several amounts of nitrites and the number of
specimens of water showing the B. Colt.

AMMONIA IN MILK AND ITS DEVELOPMENT DURING
PROTEOLYSIS UNDER THE INFLUENCE OF STRONG
ANTISEPTICS.

By H.C. SHERMAN, W.N. BERG, L. J.COHEN anp W. G. WHITMAN.
(From the Department of Chemistry, Columbia University.)

Analyses of 28 samples of milk obtained from 7 different sources
showed an average of 0.0004 per cent of ammonia. Asarule the
better the milk and the fresher when examined the lower were
the percentages of ammonia found.

Addition of 3 per cent chloroform or 0.1 per cent formaldehyde
retarded but did not stop the proteolysis which resulted in the
productionof ammonia. Thegreaterthe freedom of the milk from
contamination the less apparent is this influence of the anti-
septic and in one sample of exceptional purity spontaneous sour-

"See Biological Studies by the Pupils of William Thompson Sedgwick,
Boston, pp. 36-53, 1906.

—

Society of Biological Chemists XXXVII

ing appeared to inhibit the production of ammonia to a greater
extent than did the chloroform or formaldehyde.

STUDIES ON APPLE JUICE,

By H.C, GORE.
(From the Bureau of Chemistry, United States Department of Agriculture.)

Successive analyses were given of the juices of the cull apples
employed for cider making. These juices were found to become
gradually richer in sugar and acid as the season progressed.
Analyses of the juices of standard varieties of apples grown in
Nebraska were given, and compared with analyses taken from
the literature, of apples grown in Eastern States. The compo-
sition of the Nebraska apple juice was found to be practically the
same as that of Eastern grown apples.

Summer apples were discussed in regard to their value as
vinegar stock. To secure the highest yield of acetic acid in the
vinegar it was found that the alcoholic fermentation should be
carried on in closed vessels in order to hinder the development
of acetic acid bacteria which interfere with the alcoholic fermen-
tation. Juices of decaying apples were examined with regard to
their use as vinegar stock. It was found that changes go on when
apples decay similar to those which occur in vinegar making,
except that the inversion of the sucrose, the alcoholic fermenta-
tion and the acetic acid fermentation go on more or less simul-
taneously. Acetic acid was found in the juices of decayed apples
and in view of the well known tendency of acetic acid to retard
alcoholic fermentation the use of decayed apples is to be avoided.
The question of the wholesomeness of vinegar prepared from
decayed apples was not touched upon.

SUGAR METABOLISM.

By HUGH McGUIGAN.
(From the Physiological Laboratory of the University of Chicago.)

Work on the oxidation of sugars im vitro and their metabolism
in the body corroborates the clinical assertion that levuloseis more
easily oxidized than glucose, and that it may be used in the body
when glucose cannot. The order of their ease of oxidation im

XXXVill Proceedings

vitro and in the economy is: levulose, galactose, glucose, mal-
tose, saccharose, levulose being the easiest oxidized.

In cases of perverted metabolism with glycosuria my results
favor the theory that the sugar in the blood normally exists in
combination with a colloid. An abnormal breaking down of this
compound causes glycosuria. The intravenous injection of cal-
cium salts inhibits the glycosuria by forming a more stable com-
pound, probably oa A or Cali
tissues of the body also differ in their power of oxidation. By
using the volume of O evolved from H,O, by the same weight of
the different tissues, as an index of their oxidizing power the
order found was: kidney, liver, lungs, spleen, pancreas, muscle,
brain.

Preliminary work on the oxidation of sugars by the tissues
gives substantially the same order, but as yet it is not conclusive.

The various

THE PRESENCE OF SECONDARY DECOMPOSITION
PRODUCTS OF PROTEIDS IN SOILS.

By OSWALD SCHREINER anp EDMUND C. SHOREY.

(From the Laboratory of the Bureau of Soils, United States Department of
Agriculture.)

Soil as distinguished from the disintegrated rock which forms
the basis of soils contains organic matter. A portion of this
organic matter is made up of nitrogenous compounds, which are
evidently decomposition products of proteids.

We have absolutely no definite knowledge of the composition
or constitution of these bodies and many difficulties are en-
countered in attempting to isolate them.

A method finally adopted is essentially solution in 2 per cent
sodium hydroxide at room temperature, removal of mineral bases
and so-called humus bodies by neutralizing the solution so ob-
tained, resulting in a neutral solution containing 50-60 per cent
of the organic nitrogen in the soil. This solution gives precipi-
tates with silver nitrate and ammoniacal lead acetate and by
decomposing the precipitate so obtained with hydrogen sulphide, a
crystalline body has been obtained which has been identified as
picoline carboxylic acid.

Society of Biological Chemists XXX1X

This compound has been obtained in small quantity from
several soils of widely different character.

The presence of this pyridin compound may be connected with
the following facts. The melanoidins or humus bodies yield
pyridin on reduction. Tryptophan, a constant decomposition
product of proteids, can be easily converted into a pyridin deriv-
ative. Pyrouvic acid on treatment with alcoholic ammonia
yields uvitonic acid which on heating splits up into carbon dioxide
and picoline carboxylic acid.

There have been obtained indications of the presence of pyro-
uvic acid in some soil solutions but the identification isnotas yet
conclusive.

The work here outlined is the initial stage of one phase of the
investigation of the relation of organic soil composition to soil
fertility now being carried on by the Bureau of Soils and a more
detailed account of the work will be published at a later date.

ON LYSYLGLYCIN.

By P. A. LEVENE anp W. A. BEATTY.
(From the Rockefeller Institute for Medical Research.)

In the process employed by the writers a year ago for pre-
paring the peptid prolinglycyl, a substance was produced by
tryptic digestion of egg albumen, which on further cleavage yielded
only lysin and glycocoll. The substance could not becrystallized.
It is a noteworthy fact that peptids of the hexon bases obtained
by Fischer and Suzuki synthetically also failed to crystallize.

SOME AZOLITMIN COMPOUNDS OF MUCOIDS, NUCLEO-
PROTEINS AND OTHER PROTEINS, WITH EXHIBI-
TION OF PRODUCTS.

By JACOB ROSENBLOOM anv WILLIAM J. GIES.

(From the Laboratory of Biological Chemistry of Columbia University, at
the College of Physicians and Surgeons, New York.)

It is well known that mucoid, completely freed from the acid

used to obtain it, is acid to litmus. All previous references to

this fact appear to depend upon reaction tests made simply with

x] Proceedings

litmus paper. We have recently made the following additional
observations.

When a sample of pure moist mucoid is mixed with a small
quantity of a commercial blue litmus or azolitmin product, the
entire mass assumes a raspberry red color. We have been unable
to remove this color by any washing process, although an excess
of litmus may be readily eliminated by treatment with water.

Very faintly alkaline azolitmin preparations yield mucoid pro-
ducts that are also raspberry red when moist and which are some-
what soluble in water. If, however, one removes the excess of
base by thorough dialysis in water, azolitmin forms a raspberry
red product with moist mucoid that appears to be entirely znsoluble
in water.

This moist raspberry red product turns blue when treated with
a trivial quantity of base, such as dry CaO, and the material
then dissolves in water yielding a fine blue solution. A blue
precipitate is obtained upon treating this solution with an excess
of alcohol. This precipitate dissolves readily in water. Alka-
lies intensify the blue color of the solution of this precipitate,
but acids, and even CO, gas, precipitate a flocculent pink
product, which is insoluble in water and from which none of the
color can be washed. This pink precipitate in turn dissolves
readily to a blue solution in aqueous basic liquids.

Similar experiments are under way with other coloring mat-
ters and additional proteins, and pure products will shortly be
made for intimate chemical study. We believe that mucoid, and
proteins such as nucleoproteins that behave in the same way,
form definite compounds with azolitmin. That the phenomena
observed are not adsorption effects is already indicated, although
not conclusively proved.

NEGATIVE EVIDENCE OF THE ADAPTATION OF DOG’S
SALIVARY SECRETION TO MEET THE DIGESTIVE
REQUIREMENT OF THE DIET.

By WALTER E. GARREY.
(From the Physiological Laboratory of the Cooper Medical College.)

1. In testing the salivas of forty-seven dogs fed upon a mixed
diet it was found that only four showed any diastatic action

Society of Biological Chemists xli

upon starch paste of from 1 to 5 percent strength. Only about
8 per cent of dogs, therefore, secrete an active diastatic enzyme
in their saliva, and these only in very small amounts.

The samples of saliva were obtained in several ways; perma-
nent fistulas being established in some cases, in others drippings
were obtained under the influence of various stimuli, e. g., the
‘“psychic”’ secretion, that obtained by pilocarpine injections, by
induction of nausea, by sensory stimulation (both mechanical
and chemical), and by faradic excitation of the glandular nerve
supply (parotid and submaxillary), the animal being under mor-
phine or chloretone anesthesia. Eleven hundred samples of
saliva thus obtained gave only the negative result mentioned
above.

2. A diet of bread alone, of bread and milk, or of bread and
bouillon failed to develop any diastatic power when it had pre-
viously been absent, or to increase the action in the few salivas
which had previously shown some slight diastatic action. The
carbohydrate diet was maintained in these cases from 18 to 4o
days.

3. The failure of dog’s saliva to digest starch paste cannot
be ascribed tothe fact that proferments only are present, for the
activity is not forthcoming upon changing the reaction of the
media or by adding extracts of the esophageal or gastric mucosa,
nor does starch paste mixed with saliva show any reducing action
upon Fehling’s solution after remaining in dogs’ stomachs from
20 to 45 minutes—no “‘activation”’ takes place.

4. Extracts of active and resting dog’s salivary glands showed
no quantitative difference in diastatic power, and this was no
greater than one would expect from tissues in general.

5. Hewlett’s observation that fat splitting enzymes are absent
from dog’s saliva was confirmed.

6. Proteolytic enzymes capable of attacking raw fibrin are
also absent.

7. Diets containing much fat, or meat exclusively, do not
induce the secretion of lipase or proteolytic enzymes, respectively.

8. Histologic examination shows the typical granular struc-
ture of the salivary glands of the dog. We conclude that these
granules are not ‘“‘zymogen”’ granules.

xlil Proceedings

ON THE QUANTITATIVE DETERMINATION OF MU-
COID IN URINE, BLOOD AND TISSUE EXTRACTS.

By CLARENCE E. MAY anp WILLIAM J. GIES.

(From the Laboratory of Biological Chemistry of Columbia University, at the
College of Physicians and Surgeons, New York.)

Connective tissue mucoids cannot be completely precipitated
from their neutral or alkaline solutions by any degree of acidifica-
tion, however cautiously acidity may be brought about and
increased. As much as ro to 1s per cent of a given amount of
mucoid, dissolved in lime water under ordinary conditions, may
remain in solution after apparent attainment of complete pre-
cipitation by acid, in the usual flocculent condition in a water
clear liquid. Treatment of the clear filtrate with alcohol in ex-
cess causes precipitation of practically all the remaining mucoid
material.

The acid precipitation method is very unsatisfactory for quan-
titative determination of mucoid in tissue extracts, and bio-
logical liquids in general, not only because the mucoid cannot
be completely precipitated by mere acidification, but also because
other proteins that may be present are apt to be carried down
with the mucoid in flocculent combinations that cannot be severed
by any washing process.

Observations in other connections indicate that practically all
the acid precipitable proteins are like mucoid in these Several
respects.

The results of numerous quantitative experiments have enforced
the above conclusions and make it apparent that all published
data for the proportions of mucoids in tissues or biological liquids
are inaccurate. We hope to devise a new method for the more
accurate quantitative determination of mucoids and other acid
precipitable proteins.

ON THE NATURE AND OBJECTS OF THE AMERICAN
SOCIETY OF BIOLOGICAL CHEMISTS.

By WILLIAM J. GIES (by invitation).

ON THYMUS NUCLEIC ACID.

By WALTER JONES AND C. R. AUSTRIAN.
(From the Laboratory of Physiological Chemistry in the Johns Hopkins

University.)
(Received for publication, December 16, 1906.)

A consideration of the earlier investigations of the nucleic
acids would lead us to the conclusion that these substances yield
among their hydrolytic products four purin derivatives, namely,
xanthin, hypoxanthin, guanin, and adenin. Of these four bases
hypoxanthin was the first discovered in this connection and was
for a long time looked upon as the most constant in its occur-
rence, and, therefore, the most important; so that chemists fre-
quently omitted identification of the bases and spoke of them
collectively as hypoxanthin. Kossel was so strongly convinced
of the presence of all four of these bases, among the decomposi-
tion products of nucleic acids that he employed the expression
‘“‘nuclein bases’’ to include them all, and formulated an hypothesis
to the effect that there are in reality four nucleic acids, each of
which produces one nuclein base, and that the material ordinarily
considered nucleic acid was in most cases a mixture of four nucleic
acids so that four bases were necessarily obtained as hydrolytic
products.

All of the earlier work, however, was carried on with quantities
of materials so small as to render exactness impossible; therefore,
as methods were discovered by which nucleic acids could be
prepared in any desired amount and as schemes became perfected
for the separation of the decomposition products, the existing
literature on the subject seemed to require some revision. Thus,
Kossel had regarded thymus nucleic acid as one of the four
nucleic acids required by his hypothesis and called the substance
adenylic acid to indicate that of the four nuclein bases, it yields
only adenin. But with better facilities, Kossel and Neumann'*
showed later that adenylic acid yields both guanin and adenin.

1 Zeitschr. f. physiol. Chem., Xxii, p. 74.
I

2 On Thymus Nucleic Acid

Osborne and Harris! obtained similar results with the nucleic
acid of the wheat embryo. Although they had at their disposal
an almost unlimited amount of material and employed for their
experiments specimens of nucleic acid prepared by various
methods, they were unable to find either xanthin or hypoxan-
thin among the hydrolytic products. Moreover, the quantities ot
guanin and adenin which they did obtain correspond to one
equivalent of the former to one of the latter: from this it would
naturally be concluded that the two bases are produced from
one nucleic acid.

Again, Jones and Whipple? found that the nucleic acids of the
suprarenal gland and the pancreas yield both guanin and adenin
but neither xanthin nor hypoxanthin. Here also it was observed
that the ratio of the quantities of guanin and adenin obtained
was a simple multiple of the ratio of their molecular weights.
Finally, Levene*® has recently introduced new methods for the
preparation of nucleic acids and has been able to make a very
accurate examination of the decomposition products of a large
number of substances of this class. His results are uniform:
guanin and adenin were always present but neither xanthin nor
hypoxanthin were ever found even in traces, although in several
instances special care was employed in the analysis to find these
two bases.

From the results stated we would be justified in concluding
that nucleic acids do not contain groups which give rise directly
either to xanthin or hypoxanthin and that the occurrence of
these two bases is referable to guanin and adenin groups in
nucleic acids. The presence of xanthin and hypoxanthin in
glands is in all probability due to the action of two ferments in
whose presence the two bases are formed by the removal of amido
groups from guanin and adenin, respectively. If xanthin and
hypoxanthin were once formed in a gland in this way, they
would probably contaminate the nucleic acid obtained from the
gland, and appear among the hydrolytic products. This seems
to be one reason for the finding of the oxypurins by the earlier

1 [bid., xxxvi, p. 85.

2? Amer. Journ. of Physiol., vii, p. 423.

5 Zeitschr. f. physiol. Chem., xlvii, p. 140; tbid., xlvi, p. 155; tbid., xlv,
P- 37°-

Walter Jones and C. R. Austrian 3

investigators. A second reason is to be found in the fact that in
splitting nucleic acids, hydrolytic agents may have been employed
which, while they have little if any power to convert the amido
purins into oxypurins, may nevertheless be able to remove amido
groups from the nucleic acid or form some intermediate hydro-
lytic product, thus changing a guanin group into a xanthin
group so that xanthin will be found among the final products.
However this may be, the fact remains that nucleic acid prepared
free from impurities and submitted to hydrolysis under proper
conditions, will yield both guanin and adenin but neither xan-
thin nor hypoxanthin. The matter has become of considerable
interest to us since we have found that as a general proposition,
both xanthin and hypoxanthin are found among the products of
the self digestion of glands while guanin and adenin are absent.
This apparent anomaly was explained by the discovery in the
tissues of two ferments: one of which (guanase) causes the con-
version of guanin into xanthin but is without action on adenin;
the other (adenase) causes the corresponding transformation of
adenin to hypoxanthin; but is without action on guanin.t Our
first reported experiments on this subject were made with the
thymus gland? which contains both ferments and which therefore
on self-digestion gives rise to xanthin and hypoxanthin. After
the lapse of several years since the publication of this work, when
the existence of the two ferments has many times been proven,
a contribution by Steudel would make it appear that thymus
nucleic acid yields all four of the nuclein bases. While we do not
consider that Steudel’s work has any bearing on the conclusions
which we formerly drew concerning the ferments of the thymus,
it is a matter of considerable interest to know whether this nucleic
acid which certainly differs from other nucleic acids in its conduct
toward hydrolytic agents also differs from them in reference to
its purin groups; that is, whether the xanthin found among the
decomposition products is really due to the presence of a xanthin
group in the nucleic acid or merely to the action of reagents on a
guanin group.

1Jones and Partridge: Zeitschr. f. physiol. Chem., xlvii, p. 343; Jones
and Winternitz: zbid., xliv, p. I.
2 Jones: zbid., xli, p. 102.

4 On Thymus Nucleic Acid

The Method of Hydrolysis

As already stated, Kossel obtained only adenin by hydrolysis
of thymus nucleic acid but afterward Kossel and Neumann ob-
tained also a small amount of guanin. In the more recent work
of Steudel,! three methods of hydrolysis were employed:

I. With iodide of phosphorus.
II. With hydrochloric acid and stannous chloride.
III. With dilute sulphuric acid.

The results of this work are given in the following table. The
numbers express amounts of nitrogen in percentages of the entire
nitrogen of the nucleic acid employed.

ig | Il. Ill.
Cent = eee ae et tenn 3.61 | oD 10.07
eatin ate oes ces sree 6.74 ? ?
ideriincPrs ns res ee oe ee 13.45 4.76 16.39
EP YPOMAN CIN. com 2 x'cev sn «is cys wie 5.20 ? ?

In describing Experiments II and III, Steudel makes no men-
tion of xanthin and hypoxanthin but by comparison of the
figures in the three columns it is clear that the bases must have
been found. Moreover, we satisfied ourselves that both xanthin
and hypoxanthin are formed in the hydrolysis of thymus nucleic
acid by Steudel’sthird method. It is, therefore, apparent that in
the case of this nucleic acid the products depend quantitatively
within wide limits upon the hydrolytic agent employed and it
therefore becomes necessary to study the products obtained under
conditions which admit of a transformation neither of products
formed nor of groups in the nucleic acid. Fortunately we have
of late come into possession of a method which meets these
requirements and which may be briefly described as follows:

There is almost universally distributed in animal organs a
ferment (nuclease) which in faintly acid media effects the hy-
drolysis of nucleic acids with the formation of the amido purins.?
Under the influence of two other ferments usually present the

1 Zeitschr. f. physiol. Chem. xlii, p. 165; ibid., xliii, p. 402; ibid., xlvi,

Pp. 331.
? Iwanoff: Zettschr. f. physiol. Chem., xxxix, p. 31.

Walter Jones and C. R. Austrian 5

amido purins are converted into the corresponding oxypurins.}
Of rarer occurrence is a ferment (xantho@xidase) which in the
presence of oxygen brings about an oxidation of hypoxanthin to
xanthin and this inturntouric acid.” It is, therefore, clear that
in the action of an aqueous extract of a gland on a nucleic acid
xanthin may be formed directly from guanin or indirectly from
adenin.

Nucleic Acid.

(action of nuclease)

Pyrimidin derivatives, Guanin Adenin
phosphoric acid, etc., etc. (action of guanase) (action of adenase)
Xanthin Hypoxanthin

(action of xanthodxidase)

In this connection the pig’s spleen is exceptional.? It contains
nuclease and adenase, but neither guanase nor xanthooxidase.
Thus the ferments of this gland, while capable of decomposing
nucleic acid, cannot cause a formation of xanthin either from
guanin or from adenin. Hence we have in an aqueous extract
of this gland a means of hydrolyzing a nucleic acid at the body
temperature in a fluid that is practically neutral and at the same
time free from conditions which could produce xanthin except in
so far as xanthin groups are present in the nucleic acid molecule.

The Action of an Aqueous Extract of Pig’s Spleen on Thymus
Nucleic Acid.

Thymus nucleic acid was used in the form of its sodium salt
which was prepared by the method of Neumann‘ and consisted of
a mixture of the gelatinous and the non-gelatinous compounds.

1 Jones and Austrian: zbid., xlvili, p. I11.

2Spitzer: Arch. f. Physiol,, \xxvi, p. 192; Wiener: Arch. f. exp. Path.
u. Pharm., xiii, p. 373.

8 Jones: Zeitschr. 7. physiol. Chem., xlv, p. 84. See also Schittenhelm:
abid., xlvi, p. 354; Jones and Austrian: loc. cit.

4 Arch. f. Physiol., 1899, suppl., p. 552.

6 On Thymus Nucleic Acid

The product was free from proteid and contained 14.18 per cent
of nitrogen. The solution of ferments was prepared as follows:
Pig’s spleen which had been carefully freed from external
membrane, was ground in.a machine to a smooth paste, treated
with five times its weight of water and allowed to stand for
36 hours at the room temperature in a well closed vessel with a
sufficient amount of chloroform to prevent putrefaction. The
fluid was then strained through linen and placcd with additional
chloroform in one large, well closed vessel from which portions
were removed as required. In order to determine the quantities
of purin bases produced by the self-digestion of this ferment solu-
tion, three portions of 1650 cc. each were treated with 3 cc.
of 20 per cent acetic acid and kept at the body temperature for
seven days. The products of the digestion were then heated to
boiling and the filtered fluid examined for purin bases by the
method described below in connection with thymus nucleic acid.
As in our former experiments no trace of either xanthin or adenin
could be found but guanin and hypoxanthin were both obtained.
The former was converted into its characteristic crystalline
hydrochlorate; the latter into its nitrate which appeared under
the miscroscope as a uniform mass of whetstone crystals, and
failed to resporid to the color test for xanthin with nitric acid and
caustic soda. The quantities of the two salts obtained are given
in the following table:

I. II. III. Mean.

Guanin hydrochlorate.......... 0.188
Hypoxanthin nitrate. fon. ar 0.260

0.176 | 0.191 | 0.185

0.269 | 0.273 0.274

The guanin hydrochlorate was analyzed with the following
results:

0.1761 gram lost 0.0285 gram at 105°.

0.1906 gram lost 0.0308 gram at 105°.

0.1587 gram of dehydrated salt required 7.7 cc. sulphuric acid. (1 cc. =
.0077 gram N).

0.1472 gram of dehydrated salt required 7.1 cc. of sulphuric acid.

Theoretical for Found.
C; H; N; O.H Cl.2 H2O. 1K 10 Ill. IV.
PAS ID ae Ne ke ic ojala L6E2S8" “L65L6

tS RA aes Le 37.33 = == 37.36 difgae

Walter Jones and C. R. Austrian 7

This salt is a most convenient form of guanin both for purifica-
tion and analysis. It crystallizes from very dilute solutions in
5 per cent hydrochloric acid in clusters of long silky needles:
from more concentrated solutions it is deposited in prismatic
needles. It it easily solublein ten parts of hot 5 per cent hydro-
chloric acid but requires 5000 parts of the acid to effect itssolution
in the cold and is not removed from its acid solution by animal
charcoal: so that it can be decolorized and recrystallized with
very small loss of material. The salt has little tendency to lose
its water of crystallization on exposure to the air, but suffers loss
when kept over sulphuric acid. Several specimens of the sub-
stance after exposure to the air during three months were found
to contain the required amount of water which was given up
sharply at 80°, no further loss occurring at 120°.

The hypoxanthin nitrate was converted into the free base and
analyzed.

0.2010 gram required 10.7 cc. of sulphuric acid. (1 cc. = 0.0077 gm. N.)
0.1802 gramrequired 9.6 cc. of sulphuric acid.

Theoretical for Found.
C5H4N40. 1 ae
INSEE aes ere on ks 41.18 40.99 41.02

Parallel with the three digestions described, three others were
carried on with the sole difference that to each of the latter was
added before the digestion an aqueous solution of 14 grams of the
sodium salt of thymus nucleic acid. When the digestion had
continued for seven days, the product was heated to boiling, and
the filtered fluid after addition of 1occ. of 25 percent sulphuric
acid, was evaporated on the water bath to about 250 cc. The
acid solution was boiled for ten minutes, diluted with water, and
made strongly alkaline with ammonia. The alkaline fluid was
treated with an excess of a solution of silver nitrate in ammonia
and the precipitated silver compounds of the purin bases were
decomposed with hydrochloric acid. The acid solution of the
bases was filtered from silver chlorid and evaporated to a syrup
for the expulsion of most of the free acid. The residue was
treated with hot water and without filtration. The fluid was first
neutralized with ammonia and then treated with such an excess

8 On Thymus Nucleic Acid

that the fluid contained 3 per cent of the reagent. The material
was digested for several hours at 50° and after standing over night
at the temperature of the room was filtered. In the filtrate we
should be prepared to find xanthin, hypoxanthin and adenin; in
the residue, magnesium ammonium phosphate and guanin but
no other purin base except in so far as it is included in the preci-
pitate. The residue was treated with 2 per cent ammonia and
after digestion at so° and standing over night the ammoniacal
fluid was filtered off and united with the first ammoniacal filtrate.

The mixture of guanin and phosphate was dissolved in hot
I per cent caustic soda and the guanin precipitated with acetic
acid. The base was crystallized from hot 5 per cent hydro-
chloric acid. On cooling, the solution deposited guanin hydro-
chlorate in the usual aggregates of feathery needles. Two of the
specimens were analyzed.

0.1262 gram lost 0.0203 gram at 105°.

0.1124 gram lost 0.0181 gram at 105°.

0.1086 gram required 4.4 cc. of sulphuric acid. (1 cc. = 0.0077 gm. N.)
0.1435 gram required 5.8 cc. of sulphuric acid.

Theoretical for Found.
C;H;N;0. HCl. 2H20. if me IONE IV.
QHsOh i nee eee fey tat 16.09 16.10 — —
IN ee Sea eee Slee a= —- 31.19 Sle Le

The ammoniacal filtrate from guanin was treated with silver
nitrate in ammonia and the silver compounds of the purin bases
decomposed with hydrochloric acid. The acid filtrate from
silver chlorid was evaporated to dryness, and the addition of
water with subsequent evaporation repeated until the free acid
had been expelled. The crystalline residue dissolved easily in
water at 40°, leaving only a trace of brown flocculent material, so
small that it could not be collected for a color test. Xanthin was
therefore not present even in traces... A small portion of the fluid
was tested for adenin with picric acid. As this test proved nega-
tive the hypoxanthin was isolated in the usual way and finally
converted into the nitrate. This appeared asa perfectly uniform
mass of whetstone crystals which were free from xanthin nitrate

1Kruger and Solomon: Zeitschr. f. physiol. Chem., xxvi, p. 350.

Walter Jones and C. R. Austrian 9

as shown by their failure to respond to the color test with nitric
acid and caustic soda. Two of the specimens were analyzed.

0.2131 gram required 11.3 cc. of sulphuric acid. (1cc. = 0.0077 gm. N.)
0.1808 gram required 9.6 cc. of sulphuric acid.

Theoretical for Found.
C;H4 N40. ie ite
JIN oh thd ke ei oe eee 41.18 40.83 40.88

The quantities of guanin hydrochlorate and hypoxanthin
nitrate obtained in the three experiments are given in the table
below. If we deduct the amount of these constituents obtained
from the self-digestion of the ferment solution the difference will
represent the bases produced from the added nucleic acid.

GUANIN HyDROCHLORATE. HyYpoxANTHIN NITRATE.
I. Il. III. | Mean.| I. Il. | It. | Mean.
| =e —— = ——
From 1650 ce. of ferment
solution + 14 gm. nucleic
ECU AMe tots (ols Gis (ore ece ore 1.258 | 1.286 | 1.278] 1.274] 1.601] 1.587 | 1.694] 1.627
From 1650 ce. of ferment
BOMIULON lei ctere.a)s! ois es ee 0.188 | 0.176} 0.191 | 0.185 | 0.260] 0.269 | 0.293) 0.274
From 14 gm. nucleic acid. . — = = 1.089 = -- — | 1.353

By an obvious calculation from these data it will be found that
184 per cent of the nitrogen of thymus nucleic acid formed hypo-
xanthin, 18 percent formed guanin. By comparing these figures
with those obtained by Steudel the following points appear:

By Action of} By Action | By Action | By Acti
lodide of of of Boiling | of Ferments.
Phosphorus.|}HC] + SnCl, H2SO4

3.15 10.07 18.0
? ?

(aaminwerenc cc sathee ogee es SOL

2K ERIN AW Ene Oe eee eae 6.74 ? ? none
ANGEIENTSY 5) pec AIOE CE ae 13.45 4.76 16.39 none
EA ORANG. 5c. Sais 5 «55 lense 2s 5.20 ? ? 18.5

1. By the action of nuclease, thymus nucleic acid produces
no xanthin. The xanthin formed by more violent methods of
hydrolysis must therefore originate in guanin groups or in guanin
itself.

10 On Thymus Nucleic Acid

2. Hydrolysis at high temperatures not only gives results that
are misleading but actually destroys a large amount of purin pro-
ducts. This is especially true of guanin. The amount of this
base formed by ferment action is nearly twice as great as the
sum of guanin and xanthin obtained by hydrolysis with chemi-
cal reagents.

3. The quantities of guanin and hypoxanthin (equivalent to
adenin) formed by the action of the ferments are as nearly pro-
portional to the molecular weights of the two bases as could
reasonably be expected with the use of the methods at our disposal.
This constitutes strong evidence that the two bases result from
the same nucleic acid.

THE EXCRETION OF BORIC ACID FROM THE HUMAN
BODY.

By HARVEY W. WILEY, M.D.

(From the Bureau of Chemistry, Washington, D. C.)

(Received for publication, December 15, 1906.)

In the studies which I have inaugurated for the determination
of the effect of various substances added to foods on health and
digestion, the first substances investigated were boric acid and
borax. The object of this investigation was to administer these
bodies in small quantities to young men over a long period of
time and determine the effect upon health, digestion and the
various metabolic processes. Incidental to this investigation, the
fate of the boric acid in the human body was studied. It was
found that, as had been pointed out by other investigators, by
far the larger quantity of the boric acid administered was excreted
through the kidneys. There was a marked difference in the
action upon the urine between the borax and boric acid. When
boric acid was administered the acidity of the urine was consider-
ably increased. When borax was administered, the acidity was
diminished, the reaction sometimes becoming amphoteric and
even alkaline.

The first experimental work was divided into five series,
twelve young men being observed during each of theseries. The
time of the series varied somewhat in length. The first series
began on the sixteenth of December, 1902, and ended on the
thirteenth of January, 1903. The second series began on the
nineteenth of January, 1903, and ended on the twenty-first of
February, 1903. The third series began on the nineteenth of
February, 1903, and ended on the nineteenth of March, 19032.
The fourth series began on the twentieth of March, 1903, and
ended on the twenty-second of April, 1903. The fifth series
began on the twenty-fourth of April, 1903, and ended on the
twenty-ninth of June, 1903.

iat

12 Excretion of Boric Acid

In the first series there were administered altogether 150 grams
of boric acid, half of it as borax. There were recovered in the
urine 124.58 grams, or an average percentage of recovery of 83.05.
In the second series the average percentage of recovery was
82.85, the total amount administered being 98 grams and the
amount recovered 81.19 grams. In the third series there were
administered 132.90 grams and there were recovered 84.90 grams,
an average of 63.88 per cent. In the fourth series there were
administered 99.50 grams and recovered 82.55 grams, equivalent
to 82.96 per cent. In the fifth series there were exhibited 127
grams and recovered in the urine 95.47 grams, an average of
75:17 per cent.

The total quantity of boric acid (and borax as boric acid)
administered during the investigation was 607.40 grams, of
which there were recovered in the urine 468.69 grams, the average
percentage recovered being 77.16.

It having been stated by some investigators subsequent to the
publication of these results that the amount of boric acid excreted
in the urine as given was too small, a supplementary experimental
determination was made with six young men over a period
extending from the twenty-first of January, 1905, to the four-
teenth of February, 1905. In each case after the cessation of the
administration of the substance the tests of the urine were con-
tinued, until for several days, by the method employed, no
determinable quantity of boric acid was noticed. The percent-
ages of boric acid eliminated in the six cases were as follows:

In No. 1, 83.51;in No. 2, 77.84; in No. 3, 91.55; in No. 4, 87.75;
in No. 5, 84.26; in No. 6, 77.38.

The Thompson method' employed here was the same as that
employed in our previous work.? Calcium hydroxid was used
throughout and the only departure from the regularly outlined
method was in redissolving and precipitating more times in
order to free the occluded boric acid from the precipitate. The
volume of solution which was obtained was kept constant
throughout.

The second method with a few modifications is that of Fendler*

1 Sutton: Volumetric Analysis, 8th ed., p. 98.
? Bureau of Chemistry, Bulletin 84, pt. 1, ‘‘Boric Acid and Borax.”
3 Zettschr. —. Untersuch. d. Nahrungs u. Genussmittel, xi, p. 137, 1906.

Harvey W. Wiley 13

and consists essentially of concentrating a portion of the sample in
the usual manner, acidifying and immersing the turmeric paper
strips init. After sufficient time has elapsed to allow the liquid
to travel up the paper it is compared with strips which have been
immersed in asolution whose exact content of boric acid is known.

It is seen that in three subjects of the series there was more
than 80 per cent recovered, while in three others of the series
there was less than $0 per cent recovered. During this investi-
gation experiments were also made to determine whether any of
the boric acid administered assumed a volatile state and escaped
through the expired air, or if any of it was found in the respira-
tion. The well known tendency of boric acid to assume the
volatile condition, especially when heated with methyl alcohol
and some other bodies, led to the supposition that it might be
reduced in the system to the form in which it would be volatilized
in the expired air. To determine this point the expired air was
forced continuously through a solution of lime water for three
hours. The lime water was afterward tested for boric acid with
a negative result. It appears, therefore, that no appreciable
quantity of boric acid escapes through the lungs. On the other
hand it was found that very considerable portions of the boric
acid are excreted through the perspiration.

From the above data it is seen that the average percentage of
administered boric acid excreted in the urine in this series is
83.71, which is considerably greater than that given for the first
series of determinations. When the data are looked over, how-
ever, it is seen that this discrepancy is easily accounted for. In
the first set of investigations there was one series where the pro-
portion of boric acid recovered in the urine was only 63.88 per
cent, while in the present series there was one instance where
the proportion excreted was 91.55 per cent. If these abnormal
results be omitted from each of the investigations, it is seen that
they are practically the same. Thus in the first series of obser-
vations, made in 1903, the average percentage of borax excreted
in the urine, excluding the one phenomenally low case, is 82.15,
while in the observations made in 1905, excluding the one phe-
nomenally high case, the average percentage excreted is 81.01.

These data show that the relative quantity of boric acid
excreted varies with the individual case and may extend from 63

14 Excretion of Boric Acid

to gt per cent of the whole amount ingested. These data seem
to establish very conclusively that the average percentage of
administered boric acid which is excreted by the kidneys 1s a little
more than 8o per cent.

In the spring of 1906 another series of determinations was
made, extending from the fourth to the twenty-second of April.
The object of this investigation was to determine the quantity
of boricacidsecreted in the perspiration, and the examinations of
the urine after the exhibition of a small quantity of borax were
continued until no further test for boric acid could be distin-
guished. On the first day of the observation the urine was
examined carefully for boric acid. None was found by the
usual method of determination and a trace only by Fendler’s
method. On the fifth of April one gram of boric acid was given
and on the sixth of April, two grams.

The examinations by the usual method showed that after the
eleventh of April, five days after the administration of the last
borax capsule, there was no longer a sufficient quantity of boric
acid in the urine to be determined. On the contrary, by
Fendler’s method there was still estimated to be on the twenty-
second of April 0.002 per cent. The total percentage of the 3
grams administered which was excreted in the urine in this case
was 65.9. This indicates that there was a retention of a very
large proportion of the boric acid administered in the system,
and that this retention was of somewhat indefinite duration,
small quantities of boric acid being given off in the urine 16 days
after its administration. A second subject, treated exactly in -
the same way, showed a percentage excretion of boric acid in
the urine, by the usual method, of 71.3. There was still an
appreciable quantity of boric acid in the urine by the Fendler
method on the sixteenth day after the administration of the
capsules. A third subject, treated in exactly the same way,
showed an excretion of 82.96 per cent of boric acid in the urine and
an estimated amount of 0.003 percent in the urine on the sixteenth
day after administration, by the Fendler method.

The average quantity of boric acid excreted in the urine in
these three instances is 73.3, showing that when only a small
quantity of boric acid is administered a very much smaller per-
centage of it is recovered ‘n the urine than when the administra-

Harvey W. Wiley 15

tion is continued over a long period, thus indicating very plainly
the tendency to accumulate the boric acid in the system. This
fact is accentuated by the further observation that even after
sixteen days an appreciable quantity of boric acid was still
excreted in the urine.

Excretion of Boric Acid in the Feces.

Considerable quantities of boric acid when administered as
described in the preceding pages, are excreted in the feces, but
the total quantity is so small, compared with that eliminated in
the urine, as to be of little importance from a merely chemical
point of view.

In the case of Subject No. 1 in the series of April, 1906, 1.12
per cent of the administered borax was recovered in the feces.
In the case of No. 2, 0.78 per cent, and in the case of No. 3, 0.68
percent. The average shows 0.86 per cent of the borax adminis-
tered excreted in the feces. This number must be regarded as a
minimum percentage of excretion, since only 3 grams of borax
were administered in these cases, altogether, and the observa-
tions continued until no appreciable quantities of boric acid were
found in the feces. Traces, however, still remained after the
above observations were concluded. It is evident, therefore, that
a weighable portion of borax, administered as above described,
is excreted in the feces, probably only about 1 per cent of the
total quantity, where only a small quantity is exhibited for a
short time.

Excretion of Boric Acid in the Pers piration.

It is evident from a general idea of the method of excreting
boric acid that weighable portions of it should be found in the
perspiration. This theoretical consideration was verified on
several occasions during the progress of the work in the earlier
periods.

In the spring of 1906 it was decided to determine, if possible,
quantitatively the amount of boric acid excreted in the perspira-
tion. This was the principal object of the series of the spring of
1906, with the three men above mentioned. The method of deter-
mination, while perhaps not capable of absolute exactness, is one

16 Excretion of Boric Acid

which at least determines approximately the proportional amount
of boric acid recovered. The three men above mentioned, after
having taken 1 gram of borax on the fifth of April, and 2 grams
on the sixth, one in the morning and one at noon, were conducted
to the hot room of a Turkish bath in the afternoon on the sixth of
April. Before entering, they were carefully washed in distilled
water and thoroughly dried with an extracted towel. They were
then placed in a suit of thick woolen under-clothing which had
been previously washed, steeped in distilled water and extracted
in this way until no further extract was removed, and then dried.
They remained in the Turkish bath at a temperature of from 130°
to 135° F. for an hour and thirty minutes. The suit of under-
clothing was then removed and the men, standing in a basin,
were thoroughly sponged with distilled water, the washings saved
and evaporated, and the residue added to the extract from the
woolen suit. After concentration, the quantity of boric acid
recovered was determined. The quantity of boric acid recovered
in the perspiration in the case of No. 1 was 0.0299 gram, equiva-
lent to 1 per cent of the whole amount exhibited. The quantity
recovered in the case of No. 2 was 0.0311 gram, equivalent to 1.03
per cent of the amount exhibited. The quantity recovered in the
case of No. 3 was 0.0777 gram, equivalent to 2.59 per cent of the
quantity exhibited. In the case of No. 3 it should be observed
that the washings in distilled water were accidentally lost, so
that the amount represented as recovered was simply that derived
from the woolen suit.

The average percentage of boric acid recovered, based upon
the amount exhibited, is 1.54. It is fair to assume that the
amount of perspiration during the hour and a half spent in the
hot room of the Turkish bath is practically equivalent to that of
24 hours of ordinary temperature, so that roughly we may assume
that the average percentage of the exhibited boric acid excreted
in the perspiration of 24 hours would be 1.54.

No determination was made of the quantity of boric acid
which could be secured on subsequent days, although it is evident
that as long as it remains in the system and is excreted in traces
in the feces and in appreciable quantities in the urine, traces
would also be found in the perspiration.

The general conclusion derived from these experimental data is

Harvey W. Wiley 17

that the total quantity of boric acid excreted in the feces and
perspiration is not much if any over 3 per cent of that adminis-
tered during the ordinary period of observation. It is evident,
therefore, that even including these quantities with those which
are excreted in the urine, not over 85 per cent of the total amount
of boric acid exhibited in these experiments is in the three excre-
tions mentioned.

Excretion of Boric Acid in Milk.

From theoretical considerations it is evident that a substance
of the character of boric acid would be found also in the milk as
well as in the other excretions of the body. Accordingly arrange-
ments were made with a hospital to examine the milk of young
mothers shortly after childbirth. The milk was secured by the
attendants of the hospital in the usual way. Preliminary exami-
nations were made in all cases to establish the presence of boric
acid inthe milk. In only one case was it found in the preliminary
examination and this was evidently due to the exhibition thereof
at some time prior to the entry of the patient into the hospital.

The samples were obtained by L. F. Keblerand J. K. Haywood,
of the Bureau of Chemistry, and the determinations were made
by Rudolph Hirsch and W. B. Smith.

In the case of the first subject the lacteal secretion was so
diminished before the time of the preliminary exhibition of the
boric acid as to render the experiment of little value. After five
days of preliminary examination, begun on the twenty-sixth of
June, 1906, 1 gram of boric acid was administered on the first of
July. The milk which up to this time had given no trace of boric
acid, on the first examination thereafter showed a decided trace
which, estimated colorimetrically, was equivalent to about one
part in 100,000 parts of the milk. After the second of July the
flow of milk became so diminished in quantity that the experi-
ment was discontinued.

In the case of the second subject the preliminary examination
of the milk was finished on the twenty-seventh of May,1906. On
the twenty-eighth of May 1 gram of boric acid was adminis-
tered, the same amount on the twenty-ninth, 2 grams on the thir-
tieth, and 3 gramson the thirty-first, the first of June, and second

18 Excretion of Boric Acid

of June. In this case no trace of boric acid was found in the
milk until the first of June, when a mere trace was discovered.
On the second of June, the last day of the administration of the
preservative, the quantity excreted in the milk was estimated at
one part in 165,000. On the following day, being the first suc-
ceeding the cessation of the administration of the boric acid, the
amount in the milk had fallen to a mere trace. On the fourth,
fifth and sixth of June there was not even a trace. Unexpect-
edly on the seventh of June, five days after the cessation of the
administration of the boric acid a very large quantity was found
in the milk, namely, one part in 25,000. On the eighth of June
there was one part in 60,000; on the ninth and tenth of June, one
part in 90,000, respectively. In this instance it isseen that with
the exception of the first day it was not until long after the adminis-
tration of the boric acid that it appeared in the milk. Immedi-
ately after ceasing the administration of the boric acid it disap-
peared from the milk and did not reappear until the fifth day,
when it was present in its maximum quantity, diminishing on
successive days until the end of the observation, at which time the
amount present in the milk was one part in go,000.

In the third subject the preliminary examination, which ended
on the twenty-seventh of June, having extended over four days,
exhibited a somewhat remarkable phenomenon, namely, that
on the first day of the preliminary examination, the largest
quantity of boric acid which was found at any time in the milk
was indicated, namely, one part to 3500. On the two succeeding
days the amount dropped to a trace while on the twenty-seventh
of June, the last day of the preliminary examination, not even a
trace was found. One gram of boric acid was exhibited on the
twenty-eighth and twenty-ninth of June, 1} gram on the thirtieth
of June, and 2 gramson July 1 and July 2. Onthe twenty-ninth
of June the amount of boric acid in the milk was one part in 100,-
ooo. On the thirtieth of June and first of July there was only a
trace of boric acid in the milk. On the second of July, which was
the last day of the administration of the boric acid, there was
one partin 100,000. On the third of July, which was the last day
of the observation, there was one part in 25,000.
|. These data indicate that appreciable quantities of boric acid
administered to the mother are found in the lacteal secretion.

Harvey W. Wiley 19

The quantity is quite variable and increases or decreases without
much relation to the exact date.of administration of the preserva-
tive. It is evident that the residue of boric acid which is stored
in the body may at any time be expected to appear in the milk.
Properly this investigation should have been completed by a
study of the animal body itself after the administration of boric
acid for a certain period to determine in what organs the part
which escapes excretion is principally stored. Theoretically,
from the results of the metabolic experiments, a large part of it
would be found in the bones, or other phosphatic tissues, since
it was seen that the administration of the boric acid largely
increased the excretion of phosphorus. It is our intention in
the near future to complete the experiment by feeding animals
borax or boric acid for a period of time and then examining
their bodies to determine the quantity of borax stored and its
distribution.

a

\

:

CREATIN AS A BRAIN STIMULANT.
By S. S. MAXWELL.

(From the Rudolph Spreckels Phystological Laboratory of the University of
Calijornia.)

(Received for publication, January 1, 1907.)

In a former paper’ I believe that I have proved that the sub-
stances which we may consider specific nerve stimulants, namely,
those salts whose anions tend to precipitate calcium, produce very
prompt, in fact almost instantaneous, effects when applied directly
to the white matter of the corona radiata of the motor areas; and,
further, that no such effects are produced when these salts are
applied to the gray matter.

A number of years ago Landois? had shown that muscular
tremblings and cramps are produced by the application of creatin
to the motor areas of the cortex. In hisexperiments the cramps
occurred only after a relatively long period, seven or eight min-
utes to three-fourths of an hour, and when once established they
often lasted for many hours.

The questions now arose, first, whether the action is on the
white or on the gray matter, and, second, if on the latter, why
the latent period is so long?

When powdered creatin was applied to the cortex of the motor
areas I obtained results agreeing in every essential detail with the
description given by Landois. When, however, a saturated
solution of creatin was applied to the brain surface the latent
period was in every instance much longer and the stimulating
effects were much less marked. The following is the record of a
typical experiment:

Rassirt: right motor area exposed and determined by electrical stimu-
lation. Surface kept continuously moistened with saturated solution of
creatin warmed to body temperature.

10:40 Began applying the solution.
10:50 Fright and slight hemorrhage.

1 This Journal, ii, p. 183, 1906.
* Deutsch. med. Wochenschr., xiii, p. 685.

21

22 Creatin as a Brain Stimulant

11:00 Forepart of body inclines slightly to left. A few chewing
movements.

11:07 Slight fits of trembling in muscles of neck and of hind limbs.
Soon discontinued.

11:15 Slight chewing movements, with momentary tremblings of
muscles of neck and left foreleg.

11:27. A few chewing movements. Slight fits of quivering, mainly
of muscles of left side, have occurred at intervals for the
past ten minutes.

11:28 Discontinued the solution and applied powdered creatin.

11:36 Severe tremblings.

11:40 Epileptiform twitchings of left foreleg.

11:43 Left hind leg also twitching.

11:45 Paroxysm has become general.

In the above experiment the solution had been applied for
twenty minutes before any effect could be perceived and the
effect of forty-eight minutes’ continuous application was compara-
tively slight, and did not apparently intensify the results of the
subsequent use of the powdered substance.

In my previous experiments! I had found that when the calcium
precipitants were injected to the level of the white matter stimu-
lation usually occurred very promptly. Saturated solutions of cre-
atin injected in the same way gave no indication of stimulation of
the white matter. Occasionally in my notes occur such expressions
as ‘‘twitching of nose,’”’ ‘‘ chewing movements a few seconds after,”’
but not oftener than the accidental coincidence of these move-
ments could be expected. Ordinarily the record is ‘‘no effect
observable.’’ No limb movements were seen at any time. It
must be remembered, however, that it is not possible to intro-
duce more than a fewdrops of solution in this way without doing
so much mechanical injury as to render the results untrust-
worthy. Moreover, the experiments described above on applica-
tion of creatin to the brain surface show that the degree of con-
centration has a very great effect upon the results and point toa
mass action. One could hardly expect, then, that the injection
of a few drops of solution, which would necessarily be greatly
diluted by the fluids contained in the tissues, could give any
definite results, and it would not be safe to conclude from evi-
dence obtained in this way that creatin can nct stimulate the
white matter directly.

1 Loc. cit.

LD ee eR ee

S. S. Maxwell 2A

This question can, however, be approached in another way.
The salts which I have found to stimulate the white matter of the
corona radiata, are those which were already known through the
work of Loeb to be powerful excitants of the peripheral nerves.
For this reason I investigated in the following way the effect of
creatin solution on the sciatic nerve of the frog:

Nerve-muscle preparations were suspended in a moist cham-
ber in such manner that any contractions which might occur
would be recorded on a slowly moving drum. The creatin solu-
tion was held in a bent glass tube so placed that the entire nerve
could be immersed in it, especial care being taken that the part of
the nerve nearest the muscle should have no chance to suffer
from drying. In every instance a control experiment was made
with the nerve of the companion nerve-muscle preparation
immersed in s sodium chlorid, it having been shown by Loeb!
that this solution is without stimulating effect on nerve. In part
of these experiments I used a saturated watery solution of creatin;
in others I added to the saturated creatin solution sodium chlorid
sufficient to secure an osmotic pressure practically equal to that
of the body fluids. The results of these experiments were abso-
lutely uniform. Neither contractions nor changes of tonus of the
muscle were produced by the creatin solution nor by the mixture
of the creatin and sodium chlorid solution. Powdered creatin
strewed along the nerve and upon its cut end was equally without
effect.

It follows from these experiments that creatin does not excite

1 Loeb: Festschrift fiir Fick, Braunschweig, 1899, p. 118.

2 Mathews reports that after a latent period of one hundred and sixty
to two hundred and forty minutes rhythmical contractions appear in a
muscle whose nerve is immersed in an ¥ NaCl solution (Amer. Journ.
of Physiol., ii, p. 455). In the course of a large number of these experi-
ments I saw contractions only in two or three exceptional cases. On
the other hand I have frequently, after a nerve had lain for from four
to six hours in a pure ¥ NaCl solution without causing a single twitch,
transferred it to an ™ citrate or oxalate solution and have then obtained
a perfectly characteristic series of contractions. It seems probable that
the exceptional twitches which may occur with the nerve in ™ NaCl
are due either to erroneous methods of experimentation or to secondary
changes of some kind within the nerve, and not directly to chemical stimu-

lation by sodium chlorid.

24 Creatin as a Brain Stimulant

the peripheral motor nerves and the probabilities are that it does
not stimulate the white matter of the brain.

If the action of the creatin is upon the gray matter, as appar-
ently it is, we may now ask why its effects appear so tardily as
compared with the effect of applying the nerve-stimulating salts
to the white matter?

Since powdered creatin is so much more effective than the
solution and since, as I have found,! epileptiform contractions
may be produced by applying crystals of cane sugar to the brain
surface, one might suppose that creatin acts by the extraction of
water from the tissues. That this is not so is indicated by the
following facts: Creatin is not markedly hygroscopic. The
quantity of sugar necessary to produce cramps is enormously
larger than that of creatin. A saturated solution of the creatin
employed in these experiments, could not according to freezing
point determinations, give rise to an osmotic pressure greater
than that of an + sodium chloride solution. And finally pow-
dered creatin applied directly to the motor nerve does not stim-
ulate.

That the effects of the creatin appear so slowly and that they
depend upon concentration to so marked a degree indicates that
they are the result of chemical changes brought about in the gray
matter.

There are apparently two classes of substances which act as
brain stimulants. The first class includes such substances as
sodium citrate, oxalate,etc. These excite the white matter only,
but they do this promptly, almost instantaneously, and their
action is probably a direct one. The second class of which
creatin is an example, produce effects only after a relatively long
latent period. They do this apparently by bringing about
chemical changes in the gray matter. Among the by-products
of these reactions may be small quantities of the specific nerve
stimulants and these when they have accumulated may give rise
to the stimulation. In other words substances like creatin prob-
ably do not stimulate directly but bring about secondary chem-
ical changes to which the excitation is due.

1 Loc. cit.

THE FORMATION OF GLYCOGEN IN MUSCLE.
Plate I.
By R. A; HATCHER anv C. G. L. WOLKE.
(From the Chemical and Pharmacological Laboratories, Cornell University
Medical College, New York City.)
(Received for publication, December 28, 1906.)

It would appear from theoretical considerations that muscle
tissue should be incapable of forming glycogen from a disaccharid.

As has been shown by a number of observers,' the amount of
inverting substance present in muscle or blood is small, or is
entirely absent. The injection experiments with saccharose,?
and perfusion experiments with the liver’ also tend to show that
saccharose is quite incapable of being used by the organism as a
direct glycogen former. Moreover, the direct transformation of a
disaccharid to a polysaccharid without the preliminary inversion
to a simple hexose has not, as far as we are aware, been shown to
take place.

Nevertheless E. Kulz,‘in a series of experiments undertaken to
decide this question some years ago, obtained results which in
effect showed that saccharose was indeed a direct former of glyco-
gen in muscle.

It seems that the work of Kulz, has never been repeated,
although the anomalous character of the findings as recognized
by him, and he asked that his observation be tested by other obser-
vers. On his authority the statement that saccharose is a direct
former of glycogen has passed into the text-books, and Hammar-
sten, among others, mentions saccharose as an immediate ante-
cedent in the formation of glycogen. Onreferenceto Kilz’s original
experiments, it will be seen that the results on which he bases his
statement are by no means unequivocal. Three experiments

1Tebb: Journ. of Physiol., xv, p. 421, 1893.

2 Bernard: Lecons sur le diabéte.

3 Pfliiger: Das Glykogen, p. 205; see also Jappelli and D’Errico: Att
Real. Accad. di Med. di Napoli, 1903, cited by Moscati, Zettschr. 7. physiol.
Chem., 1, p. 90, 1906. These authors believe that saccharose is trans-
formed into glycogen by injection into the portal vein.

4B. Kiuilz: Zettschr. jf. Biol., xxvii, p. 237, 1890.

45)

26 Formation of Glycogen in Muscle

only are reported which confirm his conclusions, and these are
not given with very great confidence. For this reason we have
thought it advisable to repeat Kiilz’s experiments with improved
technique, and try if possible to obtain his somewhat anomalous
results.

Through the exhaustive work of Pfluger’ we are now in pos-
session of a method which permits the more accurate determi-
nation of glycogen than the Briicke-Ktlz’ method employed by
Kilz,’ and using a method of perfusion which will be presently
described, we believe we have been able to keep a muscle in a
surviving condition for a period of time sufficient to decide the
question. At the same time we have been able to fulfill cer-
tain experimental conditions which are necessary to render the
results trustworthy.

The methods used by Kiilz were two. In the first, a limb was
perfused with blood to which o.1 per cent of saccharose had been
added, and glycogen estimations were carried out in the limbs
before and after perfusion, using the unperfused limb as a control.
This method is faulty, in that one is never certain that the glyco-
gen destruction which is continually taking place is going on
more rapidly or more slowly than the glycogen formation by the
sugar. It is quite possible to have a formation of glycogen
which would be completely masked by the simultaneous destruc-
tion which is taking place. It will also be seen that the figures
as presented by Kiilz do not differ markedly from the differences
which Cramer‘ obtained in his analysis of symmetrical muscles,
using the Briicke-Kulz method employed in the work under dis-
cussion. The objections to this method were realized by Kiulz
and he accordingly changed his experimental conditions to meet
the difficulty.

The second method which was used was the perfusion of two
symmetrical limbs of the same animal, one with blood containing
saccharose, the other with blood containing no foreign sugar.
Symmetrical muscles were analyzed. It is the results obtained

1 Pfitiger: Das Glykogen, p. 67.

2 Briicke: Sitzungsber. d. Akad. d. Wiss. Wien., Abth. 2, p. 63, ce.

8 Kiilz: Zeitschr. f. Btol., xxii, p. 191, 1886.

4A. Cramer: Zeitschr. f. Biol., xxiv, p. 70, 1888; Aldehoff: zbid., xxv,
p- 147, 1889; see also Pfliiger: Das Glykogen, p. 178.

beeeeriatener and C.:.G.-L:-Wwolf 27

from this series of experiments on which Kiulz bases his conclu-
sions.

This method, suitably changed, appeared to fulfill more closely
the experimental conditions, and was the one chosen by us.! :

There are two points which Kilz does not appear to have
taken into consideration, and which may have some bearing on
his results. The perfusion of a limb with blood which has not
been properly arterialized does not conform to the state of affairs
obtaining under normal conditions. The perfusing fluid was
alsoruninat constant pressure. This while not directly affecting
the results has unquestionably an influence on the metabolism of
other organs, as Sollman’ has shown in the case of the kidney.
For that reason, we have performed the experiment, using inter-
mittent pressure. In using the lung itself to arterialize the blood
instead of the less efficient mechanical oxygenation, one may be
guided to some extent by the recent experiments of Riehl,’ who
has shown that lung tissue is incapable of inverting lactose.

The method which Kilz used for analyzing the muscle is also
very seriously open to question. This is especially the case as
in the present instance where the amounts of glycogen are small.
In the course of a large number of controls which were undertaken
to test the accuracy of the Pfluger method, we convinced our-
selves of its entire suitability for the estimation of small quantities
of glycogen.

At the outset of this work, it was realized that a perfusion
apparatus, fulfilling the conditions of efficient arterialization
and certainty of action was absolutely necessary. A number of

1QOne of us, with Dr. B. J. Dryfuss, has attempted the perfusion, by
quickly extirpating all the organs of the abdominal cavity. The liver
was ligated off piece by piece. Practically no liver tissue was left.
There was very little hemorrhage. One iliac artery was then ligated.
Artificial respiration was employed, and the animal kept warm on a hot
plate, additional heat being supplied by electric lamps. The animal
survived over two hours. It was found that the amount of glycogen in
the ligated leg was almost unweighable, showing the complete disappear-
ance of the saccharid from ligation. Our principal object in this experi-
ment was to obviate the difficulty of using defibrinated blood. See Bar-
croft and Brodie, Journ. of Physiol., Xxxii, p. 19, 1904; Pfaff and Vejux-
Terode: Arch. 7. exp. Path. u. Pharm., xlix, p. 324, 1903.

2Sollman: Amer. Journ. of Physiol., xiii, p. 253, 1995.

3 Riehl: Zeitschr. f. Biol., xlviii, p. 309, 1906.

28 Formation of Glycogen in Muscle

preliminary experiments lead to the adoption of the apparatus
described below, which has accomplished admirably the purpose
for which it was designed. We also made experiments in mechan-
ically perfusing one limb, using the unperfused limb as a control.
It was soon found that the rapid destruction of glycogen in the
unperfused limb which sometimes occurred with rigor did not
permit of its use.

Kisch' has recently examined the destruction of glycogen
after death, and has shown that the process takes place with
decreasing velocity, owing possibly toa using up of the glycogen-
splitting ferment. In the course of this work, we have been able
in part to confirm his results. We are also able to add that the
destruction of glycogen seems to be dependent on factors of
which very little can be conjectured. It was found that in experi-
ments taking place under what appear to be identical conditions,
one obtained a complete disappearance of glycogen from the
ligated limb, while in the second experiment, the amount of
glycogen from symmetrical muscles of perfused and unperfused
limbs was identical. That this is not due to disappearance of
glycogen in the perfused limb is shown by the high amount of
glycogen which is often obtained under these conditions.

In designing this experiment, it was necessary to have an
apparatus which possessed two completed systems for perfusing
and arterializing each limb. The apparatus is a combination of
that used by Embley and Martin? with the chamber for contain-
ing the limbs and lungs and for keeping the blood and organs at
constant temperature as designed by Brodie.’

The accompanying plate (I) gives the general arrangement
of the apparatus, while the following schematic diagram explains
the individual parts of the apparatus.

Description of Diagram. The pumps representing the right
and left hearts are formed from ordinary syringe bulbs, L,M,N,O,
with hard rubber seat valves. Land N are the right ventricles,
pumping blood to the two pairs of lungs, C and D, which are
supplied with air through the tube &. After the blood has been

' Kisch: Beitr. z. chem. Physiol. u. Path., viii, p. 210, 1906.

? Embley and Martin: Journ. of Physiol., xxxii, p. 147, 1904.

3 We wish to acknowledge the great assistance which one of us received
in the technique of perfusion from Prof. T. Gregor Brodie.

Ill, PLATE I.

VOL.

4
N
=
O

ICAL

OG

JOURNAL OF BIOL

THE

—
I e
>
7 ee
rhe
=<
A Fea
>>
a
hs
y

——'
= BG

.]
pacer

a tae

ae ee
as
a |
"Es
F
~

Trae!

S
ey

>

Keo batcherand ©. G. L. Wolf 29

ZG
S|
aul 4 =e I=

S ; e MN « 6).
©: =m (

30 Formation of Glycogen in Muscle

arterialized, it returns through V and W to glass receivers G
and/. Thence it is pumped by means of the left hearts L and M
to the respective limbs A and B. Returning, deprived of oxygen,
it reaches the receivers H and J, and again completes a cycle.

The coils F are beneath the organs to be perfused, and are
completely submersed in water at body temperature. By means
of the connections between the reservoirs G and H, and J and J,
the levels in the venous and arterial reservoirs are kept equal.
At the same time the flow from one reservoir to another is insuf-
ficient to mix the two bloods appreciably. Between the heating
coils and the organs small air traps are interposed, which are not
shown in the figure.

The heating tank is 107 cm. long, 35 cm. wide and 41 cm. deep.
The upper part contains two separate trays pierced at the bottom
with four holes 2.5 cm. wide, flanged to admit of being closed
with two holed rubber stoppers, through which the artery and
vein for one organ pass. The water which fills the tank does
not enter the trays, but is in contact with the bottom, and supplies
the necessary heat to the organs. If the temperature is not
sufficiently high in the chambers, additional heat is furnished by
electric lamps, which are turned off and on at will.

The system of cams, which is shown in the photograph, and
also extremely well in the diagram in Embley and Martin’s
article, permit of a very accurate adjustment of the flow toany of
the organs. It will be observed that instead of the two cams
and compressors used by Embley and Martin, we have used four.

The Operation. An animal is killed by medulla puncture, and
bled at once from the carotids. The blood is defibrinated, and
filtered through glass wool. <A second dog is killed in a similar
manner, bled and the blood added to that of the first dog. The
lungs and hearts are removed from both. The hearts are opened,
and very large glass cannulas with heavy shoulders are intro-
duced into the pulmonary arteries and veins through the right
ventricles and left auricles. The cannulas are securely tied in,
and the excess of heart muscle cut away. Three-way cannulas
are placed in the trachee. The lungs are then placed in warm
saline solution till ready to be used.

The abdomen of the second dog is then opened, and cannulas
introduced into the right and left iliac arteries and veins. In

R. A. Hatcher and C. G. L. Wolf 31

order to prevent collateral circulation as much as possible, liga-
tures are tied around the profunda arteries and veins just beyond
the origin of the iliacs. The limbs are severed from the rest of
the body, the bladder tied off and bleeding points stopped with
ligatures or with hemostats. All the cut surface of muscle and
skin are cauterized thoroughly, by brushing lightly with a blow-
pipe flame. Thespinal canal is plugged with modeling wax. The
amount of capillary oozing which takes place from the limls is
small.

In some few instances we have been able to detect a certain
amount of collateral anastomosis, but there were few of these
anomalous cases.

After joining up the arterial system with the organ, being
careful to exclude air, and removing clotted blood with a small
feather, the pumps are separately set in motion by hand, and the
blood flowing from the iliac veins and pulmonary arteries collected
in basins. About 150 cc.of blood are washed through. This is
defibrinated, filtered through glass wool and returned to the
proper receiver. In this way the organs are washed out with
defibrinated blood, and one has no further trouble from intravas-
cular clotting.

The cams are adjusted, so that each heart makes about 7o
beats per minute. The respiration pump inflates the lungs
about 25 times per minute. By means of the adjusting screws,
the flow to the organ is regulated so that limb and lung receive
the same quantity of blood.

The sugar solutions added were isotonic with blood. In every
case where sugar solutions were added, an equal amount of isotonic
salt solution was added to the control side. In this way we be-
lieved we kept the conditions on the two sides more closely alike
than if the control side was allowed to receive undiluted blood.

Perfusion was continued for two hours, after which symmetri-
cal muscles were removed from the two limbs as quickly as possi-
ble for analysis.'

1 It may be noted that it was found to be extremely difficult to filter
the alkaline solution of muscle clear, and it was found preferable to let it
stand a day or more for sedimentation to take place: this very greatly
facilitates the subsequent portion of the analysis, and no destruction of
glycogen occurs even on prolonged standing, of which we satisfied our-
selves by experiments.

32 Formation of Glycogen in Muscle

Before undertaking this work it was necessary to repeat some
of the experiments which had been performed to determine the
difference which might exist in the glycogen content of symmet-
rical muscles. This question has been inquired into a number of
times since Cramer took the matter up at the suggestion of
Kulz. The following are some of the result which we obtained.

We give but one of several sets of results which all show
approximately the same difference for symmetrical limbs."

TABLE I.
: Glycogen. :
Fore limb Hind limb Pectoral
Reenter soe 0.523 0.570 0.533
LOST eee ee 0.516 0.561 0.526

The difference in the content of glycogen of symmetrical
muscles would appear from our results to be much less than that
observed by other investigators. Recent observations on the
glycogen content of the liver tend to show that the difference in
concentration is also less than that heretofore believed to be the
case.

The Perfusion of Muscle with Blood Containing Saccharose.

The following table (11) gives the results which we have
obtained by the method above described.

As will be seen, the perfusion of a muscle with a solution of
saccharose in no case led to an increase in the glycogen content of
the side through which the sugar solution was perfused. A
number of experiments with a single perfusion, after ligation
of one side are not detailed. The loss of glycogen on the ligated
side during the two hours of the experiment is sometimes quite
large. This is not always the case, however, and we have some-
times observed as great an amount of glycogen in the non-per-
fused side as in the perfused. The reason for the difference in
individual experiments we are at a loss to explain.

‘These analyses (for which we are indebted to Dr. B. J. Dryfuss) were
made by a method similar to that described by Ivar Bang (Festschrift fur
Olof Hammarsten, 1906), in which a centrifuge is used to separate out
the glycogen. This method will be found much more convenient than the

usual procedure of filtration and washing. Duplicate analyses are very
satisfactory.

R.A. ‘Hatcher and C. G. L. Wolf

[BOF UEpT AIOM SQUIIT OMY OY Ul UBTOOATS JO syunowe oy} ‘youy JO 1094vUI B SY
p sv ABA oles 04) Ul Ivodde AU sz/NSeL INO yey} Laps1O Ul AVA SIY} Ul UOTPBUTUTIOZOP OY} OAIT OAV

aUdIIYIP B UO Spusdep YOY UO ST es¥atOUT Jue. Jed ('QG OF JUNOT 0} poyE[NOTBO ST YOIYM ooUdIOYIp oY} J] A JUOUIIIOdXTT JO oSvo 94 UT %

‘SOA Oy

“WIBISI[[IUL B JO UOTJOVIJ GB JO Sulysiom ut

‘aul[es [BUIIOU PUB BSOON|H| 00'S jeuiou jeuiou SeiGGrt|eeclie Orta as aitO, 140 IX
‘oul[es [BUIOU PUB VSOON[y| 00'S jewa0Uu jeusou 8°6 +| SO'0 1¢‘0 9¢S°0 x
“OUT[eS
|BulioU pusB ssoreyooRg ‘“UlUYyDAIGG ‘SABP Bg paArIBys| 00'S IOBW IOBUL 0°0O 00°0 00°0 00°0 ral
“OuI[BS
[eullou puRw ssoon[yy “ulUYyIAIYG ‘“SABP QB poeAIB{S| 00'S IOBIL 10311 0°0 00°0 00°0 00°0 IIIA
“ouT[eS
[Buliou pue vsorByooRg ‘ulUyoOAIQ “SABP 8 P2AIeyS) 00'S Ios Ios 0'0S+!| zo'O+! +¥0°0 90°0 #ILA
‘OUT[BS [BUIOU PUB BsSOIBYOIBY) (90'S [eu0Uu jeuou 0'0 00°0 12°0 4z2°0 IA
‘OUT[VS [BULIOU PUB osOLBYyoOVGg! OTS osu [eyswedsosupenwed) Fy +) TO'O+! O20 | TL°0 A
‘OUT[BS [BULOU PUB dsOIBYIIVG) VG" yeuou yeuwou 00 00°0 SPO SFO AI
‘OUT[BS [BULIOU PUB BSOLBYOOIRG 00°S [eultou yeuou °*FI— GO'O OF’ O ce'o Ill
‘OUT[BS [BUIOU PUB VSOIBYIIGS| 00'S jeutou [euwi0Uu 0'6I—| 80°0 0¢°0 Cr O Il
‘SquIT] YIOQ UL pooTq [BUION] OPT jeuou [euiou 8°¢—| Tto'o—| 2ZI‘0 910 I
yued 19g
z J ‘ : ‘90u0 aoue ‘pu
our Boe led PHBYOOBS | -ragiq | -sregiq | PO SENG *quowllodx iT
‘Il ATAVL

34 Formation of Glycogen in Muscle

In order to test the conclusion that saccharose is not a direct
glycogen former in muscle, several animals were made glycogen-
free by starvation and the administration of sub-maximal doses
of strychnin. Where an animal was succumbing to the poison,
artificial respiration was resorted to. The animal was finally
killed by pithing, and the limbs used for perfusion. In the case
of saccharose no glycogen was formed (Experiments VII and
IX). What is perhaps more remarkable is that in the case of
glucose no glycogen was also formed (Experiment VIII).

This may be explained in one of two ways. Either the rigor
induced interferes with the normal physiological processes, or
the glycogen is consumed as fast as it is formed. In support of
the latter view two experiments may be cited.

TABLE III.
Muscles. Condition. Glycogen. | Duteiee.
See normal © isa5o en 424.4
ett eects rigor 0.45
LES Ted) a eRe nicks normal 0.28 + 55.6
ete ye arrest rigor 0.18

In an experiment in which both limbs were in rigor we found
a complete disappearance of the glycogen from both limbs.
The production of heat and the formation of carbon dioxid
during rigor also goes to show that glycogen is used up during
the process."

On the other hand when glucose was added to the blood, there
was a distinct formation of glycogen amounting in the one case to
9.8 per cent, in the second to 22.8 per cent.

SUMMARY.

Glycogen is not formed in the perfusion of muscle by blood
containing saccharose. Muscles rendered free from glycogen by
starvation and strychnin do not form glycogen either from glucose
or from saccharose. Glucose does form glycogen in muscle.
The content of glycogen in symmetrical muscles is practically
alike.

1G. N. Stewart: Manual of Physiology, 4th ed., pp. 247, 587, 1900.

THE RELATION OF THE THYROID TO AUTOLYSIS, WITH
A PRELIMINARY REPORT ON THE STUDY OF AUTO-
LYSIS BY DETERMINATIONS OF THE CHANGES IN
FREEZING POINT AND ELECTRICAL CONDUCTIVITY:.'!

By H. GIDEON WELLS anp R. L. BENSON.

(From the Pathological Laboratory of the University of Chicago.)
(Received for publication January 14, 1907.)

In a previous publication” were reported the results of a study
of the influence of extracts of the thyroid upon autolysis of liver
tissue. These experiments were made to ascertain, if possible,
the relation of autolysis to normal cellular metabolism; and they
were based upon the well-known fact that administration of
thyroid preparations is followed by an increased elimination of
nitrogenous katabolites. If it could be found that thyroid
extract exerts an augmenting effect upon autolytic decomposition
of tissues 7m vitro, such a result might be considered as indicating
that the autolytic enzymes are factors in proteid metabolism, and
also that the thyroid modifies proteid metabolism through an
effect upon the autolytic enzymes. The results of the experi-
ments were, however, entirely negative. Liver tissue was per-
mitted to undergo autolysis under toluol at 37° C. for periods of
from 7 to 22 days, and the amount of autolysis was determined by
ascertaining the proportion of the nitrogen compounds which
existed in the three forms: (1) Coagulable; (2) non-coagulable,
but precipitable by zinc sulphate in acid solution; and (3) that
‘which was neither coagulable nor precipitable. It was found that
the rate of autolysis of normal liver tissue of dogs was quite the
same whether the tissue underwent autolysis by itself, or with the
addition of various quantities of extracts of thyroid or kidney
tissue.

1 Presented before the Chicago Pathological Society, January 14, 1907.
2 Wells: Amer. Journ. of Physiol., Xi, p. 351, 1904.
35

36 Relation of Thyroid to Autolysis

Later Schryver' reported a series of experiments by a different
method, and with positive results. He fed cats with thyroid pre-
parations for several] days, and found that liver tissue obtained
from such animals underwent autolysis somewhat more rapidly
during the first twenty-four hours than did liver tissue obtained
from control animals, which had been kept under exactly the
same conditions except that no thyroid was given. However, it
was found that if the animals were fed thyroid for more than
seven days the results were just the opposite, autolysis being
slower in the thyroid-fed animals than in the controls.

In comparing the results obtained by Schryver with those pre-
viously described, an essential difference is noted, in that Schry-
ver studied the changes occurring during the autolysis of the first
twenty-four hours, whereas in the experiments im witro the
amount of autolysis was not determined until several days had
passed. It might well be that the difference in the results de-
pends upon this difference in time, for it would be possible that
the thyroid stimulates the rate of autolysis greatly during the first
twenty-four hours, and yet after several days, when the changes
are proceeding very slowly, the curve of the control autolysis
might meet that of the thyroid-stimulated autolysis. To com-
plete the studies previously reported, therefore, it was necessary to
ascertain what effect, if any, thyroid extract has upon the rate of
autolysis in the earliest stages.

Such a series of experiments, when properly controlled, requires
the making of a very large number of nitrogen determinations,
and, furthermore, requires so much material that it is often diffi-
cult to make as many experiments as desirable with material
obtained from a single animal, which alone gives comparable
results. Therefore determinations were made by cryoscopic
methods, which permit of the making of a large number of obser-
vations in a brief space of time, and with a minimum amount of
material. The details of the. methods employed, and their
advantages and disadvantages, will be discussed below. It may
be here stated that the results indicate: First, that valuable and
accurate results concerning the disintegration of proteid mole-
cules can be obtained by means of freezing-point determinations.

1 Journ. of Physiol., Xxzii, p. 159, 1905.

H. Gideon Wells and R. L. Benson a7

This can be seen by an examination of the diagrams given below;
thus, Fig. 1 shows how closely together run the curves of autolysis
of different specimens from the same liver, and how much more
slowly autolysis occurs in muscle tissue. Secondly, they show
that the addition of thyroid extract to liver or muscle emulsions
has no appreciable effect upon the autolytic disintegration of
cellular proteids, even during the first hours of autolysis.

THE MEASUREMENTS OF AUTOLYSIS BY CRYOSCOPY.

Determination of the rate of disintegration of proteids by
means of measurements of the change in the freezing point of
solutions containing them, has, so far as we can ascertain, been
seldom used. Cryoscopic measurements of the osmotic pressure
of living tissues seems to have been first suggested by Sabbatani!
in connection with a study of the relation of the osmotic pressure
of plant cells to their resistance to freezing, which study was
carried out by Cavara. Sabbatani himself studied the freezing
point of various animal tissues, his method being simply to embed
the bulb of the thermometer in a piece of tissue which had been
placed in a freezing mixture; the temperature was read at one
minute intervals, and the freezing point was indicated by a sudden
rise in the height of the mercury, after a period of supercooling, to
a point where it remained stationary for several minutes before
beginning to fall again. By this means fairly constant results
were obtained. Neither Cavara nor Sabbatini made any studies
of the effects of autolysis, although the latter noted that rapid
changes in the freezing point occur rapidly after death.

Leon Frédericq? introduced another method of determining the
freezing point of tissues. This consists of extracting the tissue
thoroughly with boiling water, evaporating the extract to dryness,
redissolving the residue in a quantity of water equal to the
amount of fluid originally contained in the fresh tissues, and
determining the freezing point of this solution. The coagulable
proteids, which are left out of consideration in this method, exert
no appreciable influence upon the freezing point, and therefore
may be safely disregarded. (The possibility of loss of volatile

1 Arch. ttal. de biol., Xxxvi, Pp. 440, IgoI.
* Bull. Acad. Med. Belg., Seance du 29 Nov., 1902.

38 Relation of Thyroid to Autolysis

products of autolysis, such as volatile acids or ammonia, does not
seem to have been considered.) Frédericq observed the effect of
autolysis upon the freezing point, and correctly attributed it to
the decomposition of large molecules of proteids and carbohy-
drates into smaller molecules with consequent greater effect upon
the freezing point.

Delrez' studied the effect of autolysis upon the freezing point of
muscle tissue, using generally the method of Frédericq; and
Liagre? made a similar study of liver autolysis. In the case of
muscle tissue, Delrez observed that the freezing point sometimes
sank slightly after the first hour, then rose quite rapidly for
about eight hours, after which the changes were slight until
putrefaction setin. In aseptically preserved muscle he found the
increase in the depression of the freezing point greatest during
the first twenty-four hours, and relatively insignificant there-
after. Liagre found that aseptically preserved liver tissue also
showed its most rapid autolysis during the first few hours after
removal of the tissues from the body. Both authors proved by
direct analyses that the change in freezing point corresponds
to the decrease in the amount of coagulable proteid present in
the tissues.

The principle involved in this method of studying the rate of
autolysis is extremely simple, being briefly as follows: The effect
of any substance upon the freezing point of a liquid in which it is
dissolved, depends directly upon the number of molecules and
ions that are in solution—the nature of the molecules and ions
does not modify their effect in the least. Consequently, proteids,
with their enormous molecules, have so little effect upon the
freezing point that it can be practically disregarded. However,
as the proteid molecules break down into their components the
number of molecules in solution is greatly increased, and a corre-
sponding effect upon the freezing point may be observed. Hence
by determining the changes in the freezing point of a solution
containing proteids undergoing digestion, we obtain an accurate
measure of the extent and rate of this disintegration. The sim-
plicity and rapidity of application of this method, and the small

1 Arch. internat. de phystol., i, p. 159, 1904.
2 Thad. | p.ib72.

H. Gideon Wells and R. L. Benson 39

amount of material needed for making the freezing-point deter-
minations, makes the method very advantageous for certain lines
of research.

It is also possible to ascertain something as to the nature of the
new-formed molecules which are increasing the depression of the
freezing point, by making simultaneous determinations of the
electrical conductivity of the same or similar material. The
electrical conductivity of a solution depends upon the number
of electrolytes present, and is not modified by the non-electro-
lytes. Hence,ifin a given specimen the curves of the freezing
point and the electrical conductivity run parallel to one another,
it is evident that the new-formed molecules which are being
derived from the autolyzing tissue are electrolytes, and not non-
electrolytes, such as sugars and fats. In the experiments so far
made on this point, we have found that the curves of freezing
point depression and of electrical conductivity parallel each other
very closely; indeed it is quite possible that subsequent experi-
ments will show that the conductivity method gives quite as
valuable information as to the rate and degree of autolysis as does
eryoscopy, and, on account of the influence of the conversion
of glycogen into sugar upon the freezing point, conductivity
figures may be more useful than cryoscopic results in studying
autolysis of tissue. If that is the case, the conductivity method
will be found to possess greater advantages of simplicity and
rapidity of determination, and to offer the possibility of making
an almost indefinite number of determinations in a single small
sample kept under constant conditions. This question will be
the subject of further investigations.

The results which we have obtained with our experiments
have been so uniform and have checked with one another so well,
that we feel great confidence in the value of the cryoscopic
method for detecting very slight changes in molecular concentra-
tion produced by autolytic disintegration, and we believe that
it is perhaps more accurate for such purposes than is the usual
method of determining the proportion of the nitrogen in coagul-
able and non-coagulable form. Our experiments are as yet not
extensive, this paper being intended merely as a preliminary
report, and it is hoped that further experience will permit of
improvements on the method, and extensions of its applicability.

40 Relation of Thyroid to Autolysis

DETAILS OF EXPERIMENTS AND METHODS.

Dogs were used as the source of material in all the experiments.
After being kept without food for twenty-four hours they were
narcotized with morphine, and killed by bleeding from both
carotids; the bodies were suspended by the hind feet, to remove
as much blood as possible. The organs to be studied were
removed as rapidly as possible, ground to a fine pulp ina hashing
machine, and the pulp squeezed through cloth. Since it was
desired to try the effect of thyroid extract upon autolysis, it was
not possible to adopt the method of Frédericq, which possesses
the advantage of permitting of studies of aseptic autolysis with
perhaps less difficulty than by any other method yet proposed.
All steps in our experiments were performed as rapidly as possi-
ble, so that the material was ready for the initial determination
of the freezing point in less than thirty minutes after the animals
had been killed.

Series I. Three portions of liver pulp, each weighing 50 grams, were
placed in flasks, and distilled water added to make up to 500 grams. To
flask No. 1 were added 25 cc. of a ro per cent emulsion of kidney tissue
from the same dog; to No. 2 were added 25 cc. of a 10 per cent emulsion
of fresh sheep thyroid, which had been removed from the animals about
three hours previously; to No. 3, for control, were added 25 cc. of a 10 per
cent emulsion of the original liver tissue. In a second set of flasks were
placed, in the same proportions as with the liver, the following mixtures:
No. 4, muscle+kidney; No. 5, muscle+thyroid; No. 6, muscle +muscle.
The freezing point of a 20 cc. sample of each mixture was determined at
once with the usual Beckmann apparatus; the remainder was placed in
stoppered flasks in the incubator at 35°C. At frequent intervals sam-
ples were taken from each flask, and the freezing point determined. The
results obtained were as follows:

No. 1, Liver plus Kidney. No. 2, Liver plus Thyroid.
No. 3, Liver plus liver.

Bs Il. 1B i
Time. d* Time. 4d Time. 4
12:35 p. m. 0.85 12:40 p. m. 0.86 12:50p.m. 0.80
1:35 0.95 1:40 0.96 1:50 1.10
4:40 1.39 4:45 1.30 4:50 1.40
9:05 = 9:15 1.68 9:20 1.66
11:50 Leet 12:00 1.81 12:10 a. m. 1.93
6:00 a. m. Psy | 6:05 a.m. 2.95 6:10 2.90
11:20 2.84 11:40 3.17 11:45 3.30

* 4 = the depression of the freezing point below o° C. multiplied by 10
on account of the dilution of the liver emulsion to ;)jth strength.

No. 4, Muscle plus Kidney.
No. 6, Muscle plus Muscle.
iN

Time.

1:10 p. m.

4:20
8:30
11:20

5:30 a.m.
10:50

From these figures have been constructed the curves shown in Fig. I.

50

2s

2.0

on

os LN no

AUTOLYSIS OF LIVER AND MUSCLE, WITHOUT ANTISEPTICS.

1, Liver extract plus kidney extract; 2, Liver extract plus thyroid
extract; 3, Liver extract plus liver extract (control); 4, Muscle extract
plus kidney extract; 5, Muscle extract plus thyroid extract; 6, Muscle

H. Gideon and R. L. Benson

4

Vv.
Time.

1:20 p.m.
4:25

8:40
11:30

5:40 a. m.
11:05

& ffours 72

extract plus muscle extract (control).

It can be seen from the results that the addition of either thy-
roid extract or kidney extract to an emulsion of liver tissue in

* Putrefaction,

No.

76

5, Muscle plus Thyroid.

WA

Time.

1:25 p.m.

4:30
8:50
11:40

SS

£0

9:50 a.m.

4

mOonMd0 w
Ore ST Or > bO

SSeS Se
*

I

42 Relation of Thyroid to Autolysis

distilled water! has no appreciable effect upon the rate of autoly-
sis, as determined by the changes in the freezing point; the
curves are Approximately the same in each of the three sets of
determinations. The relatively slight degree of autolysis of
muscle tissue is also well brought out. The somewhat higher
rate of autolysis seen in the specimens of muscles to which thyroid
and kidney extracts have been added probably depend upon the
autolysis of the thyroid and kidney extracts themselves, rather
than upon any effect upon the muscle tissue. If these curves are
compared with those obtained by analytic methods by Schryver
for the autolysis of liver during the first sixteen hours, they are
seen to agree quite closely. Certain authors have referred to a
latent period lasting for the first four hours or so, but this is not
shown by our results nor by the results of Schryver, at least for
the liver. Inthe case of the muscle there is first a slight increase
in the depression (4), followed by a distinct fall, which persists
from the eighth to the twelfth hour. Delrez obtained a similar
decreased depression for a brief period in one experiment, but
not so frequently as with our experiments. The significance of
this phenomenon is not clear, and it is perhaps not a constant one.
It may be related to the observations of Galeotti,? who found that
the electrical conductivity of cells decreases at the time of their
death, which he attributes to a decrease in the degree of ionization
of the cellular constituents, for the effect on depression of the
freezing point was much less than the decrease in conductivity.

Series I]. The Effects of Antiseptics upon Autolysis.

In the preceding experiments putrefaction made all results
obtained after the twentieth hour of autolysis, of doubtful value.
The muscles are affected by putrefaction earlier than the liver,
which has been found to be the rule in other experiments. This
difference between the time of putrefaction in the muscle and
liver may possibly depend upon the greater autolysis of the
latter, for, according to Conradi,’ the products of autolysis exert a

1 Salt solution cannot be used as the relatively great effect of the salt
upon the freezing point would mask to a large extent the effects of the
products of autolysis.

2 Zeitschr. f. Biol. xlv, p. 65, 1903.

3 Beitr. z. chem. Physiol. u. Path., i, p. 193, 1901.

H. Gideon Wells and R. L. Benson 43

distinct antiseptic action. In order to control this interference
by putrefactive organisms the use of antiseptics was tried. It
was found by experimentation that chloroform could be used as
an antiseptic with safety, as regards its modification of the freez-
ing point of the solution, provided the water used for preparing
the tissue emulsion was first saturated with chloroform at the
temperature of the experiment. If the chloroform is not added
until the time of the experiment, its slow solution in the water
modifies the curve greatly. As has been observed in studies of
autolysis by means of analytic methods, the rate of autolysis is
somewhat slower in the presence of antiseptics than without
them, undoubtedly because of an inhibitory effect of the antisep-
tics upon the enzymes. Studies of the influence of various
antiseptics upon autolysis, with reference to their applicability in
determinations made by physical methods, are now under way.
The effects of the presence of chloroform upon autolysis may be
seen in the following table and chart:

Nos. 1 and 2, Liver without chloroform.

Time. 4 Time. 4
1:05 p.m 0.84 3:10 p.m 0.80
2:00 0.87 4:05 0.89
2:30 0.87 5:05 0.91
3:45 0.99 7:05 1.00
4:30 1.02 10:45 1 says
5:30 1.04 1-30)p2me LA6o
Zao PAT 6:10 1.70
9:10 1b a7 8:45 2.42*

12:10 a.m pod
Paka \e5) 2.10
4:10 2.09
8:05 Dat
*Putrefaction.

Nos. 3 and 4, Liver with chloroform.

Time 4 Time A
3:00p.m. 0.83 AUN ioe, (0), “fl
4:10 ibe iy 6:05 0.71
5:15 TO 7 7:20 0.73
7:15 1.20 10:55 0.92
10:55 iy 3:40 a.m 1.15
1:45 a.m 1.40 8:00 1.42
6:25 1.60 1:45 p.m 1.73
9:00 2.00 5:10 ib ZOFi
3:15 p.m 2.02 2:25 a.m 1.98

Only Experiments 2 and 3 were made with material from the same liver.

44 Relation of Thyroid to Autolysis

aS

2.0

75

AN

Os

< Sfours 12

AUTOLYSIS OF LIVER, SHOWING EFFECT OF ANTISEPTICS.

1 and 2, curves of autolysis without chloroform; 3 and 4, liver with
chloroform. Only specimens 2 and 3 were from the same liver; 1 and 4
coming from different animals.

Series III. The Effect of Autolysis upon Electrical Conductivity.

As mentioned previously, it is possible to determine whether
the newly added molecules in an autolyzing mixture are electro-
lytes or non-electrolytes, by comparing the curves of freezing
point depression and of electrical conductivity. This was done
with material from the same liver as No. 4 of the preceding series,
the change in conductivity being determined in the usual way
with the Wheatstone’s bridge, and Arrhenius cell, the autolyzing
mixture being kept at the same temperature (35°) as the material
used for cryoscopic determinations. The freezing point and
conductivity experiments were made simultaneously, and with
material from the same animal. The effects of thyroid extract
were also studied in this series. The results obtained were as
follows:

a

~

H. Gideon Wells and R. L. Benson 45

A. Liver in chloroform water. Freezing point depression.
B. Liver in chloroform water. Conductivity increase.

A. B.

Time. 4 Time. = i
5:00p.m. 0.71 6:00p.m. 0.1439
6:05 0.71 6:20 OM1525
7:20 0.73 7:10 0.1635

10:55 0.92 10:45 0.1778
3:40 a.m. Tele 4:15a.m 0.2030
8:00 1.42 8:30 OE2350
1:45p.m. 1.73 2:20 p.m 0.3356
5:10 esther 5:00 0.3587

2:00 a.m 0.4263

* z =resistance of liver extract in Arrhenius cell; i= the conductivity of the extract.

C. Liver plus Thyroid, in chloroform water. Freezing point depres-

D. Liver plus Thyroid in chloroform water. Conductivity increase.

C. D.

Time. 4 Time. =
5:10 p.m 0.70 6:00 p.m 0.1398
6:10 0.69 6:20 0.1463
7:35 0.74 7:10 0.1577

11:00 0.82 10:45 0.1717
3:45 a.m 1.13 3:30a.m. 0.1785
8:10 1.32 8:30 0.1915
1:50 p.m 1.63 2:20 p.m 0.3065
5:15 1.64 5:00 0.3292

2:00 a.m 0.3348

The results of this series (Fig. III) indicate again the absence of
any accelerating effect upon autolysis, at least 7m vitro, by thyroid
extract. Indeed, in nearly all cases the curve runs slightly
lower in the specimens to which thyroid extract has been added.
This probably depends upon the very slow rate of autolysis of
the added thyroid extract itself, for thyroid cells undergo autoly-
sis very slowly, afact previously ascertained both chemically’ and
microscopically.2. The very close parallelism of the conductivity
and freezing point curves indicates that the depression of the
freezing point which results from autolysis depends upon the
liberation of electrolytes; these are presumably chiefly amino-
acids, and perhaps also to a slight extent free fatty acids, such as

1 Schryver: Biochem. Journ., i, p. 156, 1906.
2 Wells: Journ. of Med. Res., xv, p. 159, 1906.

46 Relation of Thyroid to Autolysis

lactic and acetic. Bayliss' suggests that the inorganic salts
which are. held in the proteids by absorption affinity, are also
liberated during proteolysis, and exert a considerable effect upon
the conductivity of solutions containing proteolyzing mixtures.

2.0
OF

15
as

Conductivity (Dotted Lines)

02

ao

A (Solid Lines) »—>
10

Hours »—>

AUTOLYSIS OF LIVER, AS SHOWN BY FREEZING POINT AND CONDUCTIVITY.

Freezing point, solid lines; conductivity, dotted lines. A and B, liver
extract alone; C and D, liver extract plus thyroid extract.

CONCLUSIONS.

It is possible to measure the rate and degree of autolytic dis-
integration of tissues by determining the change in the freezing
point and electrical conductivity of solutions containing the
autolyzing tissues, and the products of their autolysis. This
method possesses great advantages because of the ease of applica-
tion and the small quantity of material required, especially in
investigations in which numerous determinations are necessary
for the purpose of plotting curves, etc.

1 Quoted by Starling: Recent Advances in the Physiology of Dzgestion,
p- 25, 1906.

H. Gideon Wells and R. L. Benson 47

Autolysis is much more rapid with livér tissue than with muscle
tissue; but in either case it is most rapid between the twelfth and
twentieth hours, at 35°.

The addition of extracts of thyroid or kidney tissue, to emul-
sions of liver or muscle tissue, does not greatly modify the rate of
autolysis.

The effect of autolysis upon the depression of the freezing
point seems to depend upon molecules which are electrolytes,
and, therefore, presumably chiefly amino-acids.

‘ti at eS
ee

-, : AY i?
=~ nat ¢t ri ure
<1 Meet’ iol
» :

(a

THE RELATION OF EXTRACTIVE TO PROTEIN
PHOSPHORUS IN ASPERGILLUS NIGER.'*

By W. KOCH anp HOWARD S. REED.

(From the Laboratory of Physiological Chemistry of the University of
Missourt, Columbia, Mo.)

(Received for publication, January 23, 1907.)

The relation of the three main groups of phosphoric acid
derivatives, protein, lipoid, and extractive, to one another has
been the subject of considerable study. A comparison of the
chemical structure of these various combinations would suggest
that the more complex, like nucleins and lecithins which are col-
loidal in nature, are built up at the expense of simpler ones like
glycerophosphoric acid and diethoxydiphosphoric acid which are
soluble in water. The formation of lecithin at the expense of
phosphates in the germinating seed has been demonstrated by
Maxwell? and also by Stoklasa.? Their method of estimating
lecithin was not strictly accurate but the differences in the
amounts of lecithin found were sufficiently great to justify their
conclusions. More recently Wilfarth, Romer, and Wimmer* have
shown that a decrease in the phosphates of the straw, with a
corresponding increase of protein and lecithin phosphorus in the
seed, occurs during the growth of the barley tomaturity. Experi-
ments involving the transfer of material, however, from one
tissue to another obviously do not permit of such definite con-
clusions. The same difficulty applies to the interpretation of

1 These results are published in this preliminary form on account of
changes which make it impossible for us to continue the work together.

? Maxwell, W.: Amer. Chem. Journ., Xiii.

8 Stoklasa, J.: Sztzungsber. d. kais. Akad. d. Wissensch. in Wien., Math.
naturw. Klasse, civ, Ab. i, 1896.

4Wilfarth, Romer and Wimmer: Landwirtsch. Versuchstat., \xiii, p.
I, 1905.

49

50 Phosphorus in Aspergillus Niger

the results of Miescher' and Noel Paton? as to the transfer of phos-
phorus in the salmon during the growth of the sexual organs.
There is indeed an actual loss of phosphorus from the muscle
which corresponds to the gain of phosphorus in the testicle and
ovary, but the relative amounts of nuclein and extractive phos-
phorus in the muscle are changed very little. It seemed of
interest, therefore, to devise an experiment in such a way as to
demonstrate a change in the relative proportion of protein to
extractive phosphorus on one tissue. Aspergillus niger was
selected for the purpose, as it is known to yield relatively good
crops even with very small amounts of phosphates at its disposal.
Higher plants under the same conditions would be stunted in
their growth or arrested in some part of their development.
Grown on the usual culture media the phosphorus of Aspergillus
niger will be distributed about as follows, the figures expressing
per cent of total phosphorus:

IPEOneiDNOSONOCUS ois.<.cnic. 5 .2)¢ 2 ae ners ss ce ee 29.0
Pextrachive PUOSPBOLUS, 6; 6. sVo.0\2) sys .s « «90 = 3) a-ep al a pee eee 68.0
jee gil albayelalccy0) 1 oval fiat te a ePID oye coc 3.0

The experiments were carried on as follows:

The culture medium which represents the control solution for determin-
ing the amount of growth under favorable conditions, consisted of:

I.o gram acid potassium phosphate.
2.0 grams ammonium nitrate.
0.5 gram magnesium sulphate.

10.0 grams cane sugar.

200 cc. distilled water.

The other solutions were made up by reducing the amount of acid potas-
sium phosphate and keeping the other constitutents the same. The supply
of potassium was kept constant by the addition of a corresponding amount
of potassium nitrate to the solutions which contained less acid potassium
phosphate. The solutions were placed in wide mouthed jars (Mason jars)
a thick piece of cotton batting tied over the mouth and then sterilized
in an autoclave at 110° C. for twenty minutes. When cool the solutions
were inoculated with spores from a pure culture of Aspergillus niger and
kept at a temperature of from 21° to 25°C. The crops were all collected

1 Meischer, F.: Die histochemischen und physiologischen Arbeiten, Leipzig,
1897.

2 Noel Paton, D.: Report of Investigation on the Life History of the
Salmon in Fresh Water, Fishery Board for Scotland, p. 143, 1898.

W. Koch and Howard S. Reed 51

on the same day washed, placed on filter papers to drain, dried over
calcium chlorid, and weighed. For the chemical analysis the method of
Koch for protein and extractive phosphorus and of Koch and Woods! for
lecithin phosphorus was employed. As the amount of lecithin was, how-
ever, very small the figures are not given. It was found to be present in
appreciable quantity in every case.

The following table gives the results of the chemical analysis:

Wt. of Wt. of Composition in Per cent of Air dry Substance. |

Ratio of
a Sea SE : Protein P to
er mgm. | Fee ExtractiveP.| LecithinP, | Extractive P.
2.410 1000, | 0.22 0.66 0.020 | 100: 300
2.107 Z00R mee OFZ 0.60 |. present | 100°2 290
1.778 100 0.25 0.71 present 100 : 288
1.466 7 | 0.18 0.12 present 100 : 66
1.083 10% *O.18 0.03 present | 100: 17

In the first two experiments the extractive phosphorus may have been
influenced by the phosphates of any adhering culture medium. This does
not apply to the last three experiments however, as there the culture
medium was found to be free from phosphates at the end of the experi-
ment.

The amount of nuclein phosphorus remains fairly constant in
spite of the great decrease in the amount of phosphate supplied.
The relatively good yield of crop is also interesting considering
the extreme phosphorus starvation, namely, one-hundredth the
usualamount. It is surprising to note to what an extreme degree
the starvation (one-fiftieth) must be carried before there is any
decrease in extractive phosphorus.

In conclusion the following observations seem justified:

Protein, or, in this case, nuclein phosphorus is, as we already
know from histological evidence, the most important form of
phosphorus at the disposal of the cell. It is formed at the
expense of other forms (except lecithin) and is not decreased even
in extreme starvation.

Lecithin phosphorus is next in order of importance. In the
building-up of the nucleins, however, lecithin probably takes no
direct part. When lecithin is broken up in the course of its
metabolism some or all of its phosphorus may be built up second-

1 Koch, W. and Woods, H. S.: This Journal, i, p. 203, 1906.

52 Phosphorus in Aspergillus Niger

arily into nuclein, merely as a matter of economy on the part of
the organism.

The extractive, water-soluble forms of phosphorus, are undoubt-
edly the ones from which the others are built up. They would
represent the intermediary step between the phosphates and the
more complex combinations of phosphoric acid.

THE RELATION OF ELECTROLYTES TO LECITHIN
AND KEPHALIN.

By W. KOCH.
(From the Marine Biological Laboratory, Wood's Hole, Mass.)
(Received for publication, February 4, 1907.)

ln a previous study! of the precipitating action of calcium
salts on solutions of lecithin it was found impossible to explain
why sodium chloride should prevent the settling out of the pre-
cipitate. Since then the subject of the precipitation of colloids
by electrolytes has been greatly advanced by the work of
Mathews.” It seemed of interest, therefore, to again take up the
subject and apply his results on albumin to colloidal solutions
of lecithin and kephalin. According to Mathews the concen-
tration of an electrolyte which is just sufficient to precipitate
an electro-negative colloid bears some relation to the solution
tension of its ions. The greater the solution tension of the
kation the greater must be its concentration necessary to cause
precipitation. The anion in such a case exerts a dissolving
influence which also bears a relation to its solution tension.
Sodium chloride would, therefore, prevent the precipitation of
lecithin by calcium chloride on account of the dissolving action
of the chlorine ion.

The experiments were carried on with a sample of A. G. F. A.
lecithin made from eggs. The sample had been previously
analyzed by Woods? and found to consist of two-thirds kephalin.
The solution was made up by shaking 1.0 gram of lecithin with
500 cc. of distilled water in a shaking machine for several hours.
This insured a perfectly fresh solution which was then filtered
before using. The tests were all made in test-tubes of 20 cc.
capacity. A sufficient amount of the salt was measured out

1 Koch, W.: Zeitschr. f. physiol. Chem., xxxvii, p. 181, 1903; The Decenntal
Publications of the University of Chicago, x, p. 93, 1902.

2 Mathews, A. P.: Amer. Journ. of Phystol., xiv, Pp. 203, 1905.

3 Koch, W. and Woods, H. S.: This Journal, i, p. 211, 1906.

53

54 Relation of Electrolytes to Lecithin

from a standardized solution to bring the concentration up to
the one given in the table when diluted to 10 cc. Five ce. of
lecithin solution were used in every case and the total made up

to ro cc. The following is an illustration: 1.7 cc. i CaCl, +
50

3.3 ec. H,O + 5 cc. of lecithin solution gave ro cc. of solution
in which the concentration of the calcium would be 0.00345M.
In order to decide if precipitation had taken place or not the
tube was allowed to stand twenty hours. This forms a con-
venient time as a solution which will not cause the lecithin to
settle out in twenty hours will not bring this about if allowed
to stand several days longer, although the solution may be
more turbid than the control. The following table gives the
results.

TABLE I.
Concentration of
Kation Required to Concentration of Anion Required to Prevent Precipitation.
Precipitate.
KCl 0.3 M|NaOH 0.0066 m prevents ppt. by 0.33. m NaCl
NaCl 0.15 M|NaOH 0.0066 m does not prevent ppt. by 0.005 m CaCl,
SrCl, 0.006 mM|NaOH 0.01m prevents ppt. by 0.005 m CaCl,
MgSO, 0.004 mjNaI 0.1 Mm _ prevents ppt. by 0.005 m CaCl,
CaSO, 0.0033 m|NaBr 0.2 m_ does not prevent ppt. by 0.005 m CaCl,
CaCl, 0.0030 m|NaBr 0.05m _ prevents ppt. by 0.0033 m CaCl,

H.SO, 0.002 mjNaCl 0.1 M_ doesnotpreventppt.by 0.0033 m CaCl,
CuSO, 0.00125 mM\NaCl 0.33mé +NaOH 0.0066™ prevents
ppt. by 0.01 m CaCl,

The arrangement of salts in the above table is practically the
same as that found by Mathews for colloidal albumin solutions.
The sensitiveness of lecithin to this precipitation is so great that
it is possible to distinguish between two solutions which differ
from one another by only o.1 mg. of calcium in 10 cc.

If we measure the changes in viscosity of the solution the reac-
tion becomes even more delicate. The addition of an amount
of calcium chloride not sufficient to precipitate the lecithin in
twenty-four hours will cause a decrease in the viscosity of the
solution. The measurements were made with an Ostwald vis-
cosimeter and represent rate of flow through a capillary tube.

W. Koch 55

ieeRor distilled water. ce. es 3...) te eee seconde
Pee ROrretait| SionuOmlecruninins aa) se) sl 65 seconds.
3. OOOMZoM (CaCl a. .n eee 63 seconds.
4. + OMOZS mM CaCl! ocd meee 60 seconds.
RN, = OL00S5emiCa@ly 2h... sane 57 seconds.
6. SSO O0VS. Sch INEONRW Ss Sea tc cc 67 seconds.

From the above it will be seen that as soon as the concentra-
tion of the calcium becomes sufficient to cause a precipitation
of the lecithin (5) the viscosity becomes equal to that of distilled
water. Sodium hydrate has the opposite effect, increasing the
viscosity.

The previous observation! made on lecithin from sheep’s brains
to the effect that sodium and potassium chlorides do not precipi-
tate lecithin are apparently not confirmed by the results given
in the first column of Table I. A repetition of these experiments
with crude preparations of lecithin from human brains revealed
the curious result that a preparation from corpus callosum
behaved like egg lecithin while a preparation from cortical gray
matter behaved like the lecithin previously obtained from sheep’s
brains. This at first suggested that the lecithin from gray matter
might be different from that made from eggs or corpus callosum.
The fact that a concentration of 0.0066m sodium hydrate prevents
the precipitation of egg lecithin by sodium chloride and more-
over makes it possible for the sodium chloride to prevent pre-
cipitation by calcium (see second column of Table I) suggests
the possibility that the lecithin of the cortex may be in combi-
nation possibly as an ion-lecithin compound. Another expla-
nation was, however, found to be correct. Egg lecithin and
lecithin from corpus callosum are very rich in kephalin. When
the crude lecithin from corpus callosum was freed from kephalin
by dissolving in cold alcohol (which leaves most of the kephalin
behind insoluble in this case) a preparation was obtained which
more nearly resembles lecithin from the cortex. The following
table gives the results:

TABLE Il.
Mae tet (Galesimn Chlnpile | eodween Goleta
Eggs (2 kephalin) 0.00300 m 0.16 M
Corpus callosum (4 kephalin) 0.00345 M 0.20M
Cortical gray matter (4 kephalin) 0.00588 Mm 2.0 M (no ppt.)
Corpus callosum (purified) 0.0059 M 1.0 M

iKoch, W.: Loc. ci., p. F83-

56 Relation of Electrolytes to Lecithin

The greater sensitiveness of kephalin to precipitation by
kations can be explained by its more acid properties. It is
well known that the addition of a slight amount of acid to an
electro-negative colloid greatly increases the ease with which it is
precipitated. Kephalin among other things differs from lecithin
by the absence of two methyl groups, and the consequent change
of the tetra ammomium base of the cholin to a triad nitrogen is
sufficient to explain the decrease in basic properties.

CHs

H
RN Ons OUR N<o¢y,+2 CHa

L Oils
OH

The fact that preparations representing mixtures of lecithin
and kephalin from corpus callosum are precipitated by 0.2 M or
about one per cent of sodium chloride suggests an interesting
explanation of Macallum’s' observation, namely, that sodium
chloride (which is most probably the chloride present) is confined
to the axis cylinder of a nerve fiber and does not penetrate
the medullated sheath. The medullary sheath contains large
amounts of kephalin as well as cholesterin and cerebrin (in per
cent of dry matter of corpus callosum there are: Kephalin, 14
per cent; cholesterin, 15 per cent; cerebrin, 17 per cent; lecithin,
14 per cent). Such a mixture would be extremely sensitive to
precipitation by sodium chloride and would consequently form in
contact with sodium chloride solution a precipitation membrane
similar to that formed by a drop of copper sulphate, which has
been suspended in a potassium ferrocyanide solution. The
sodium chloride cannot diffuse through such a membrane any
more than the copper sulphate diffuses through the semiper-
meable membrane of copper ferrocyanide. The cholesterin no
doubt plays a role in rendering the kephalin more sensitive to
precipitation and in strengthening the membrane when formed.
The starch-like cerebrin acts as a filler and the neurokeratin net-
work gives further stability to the medullary sheath.

The application of these results to the red blood cells, which
are also quite rich in kephalin and cholesterin, seems very prom-
ising and will be continued.

1Macallum: Proc. of the Roy. Soc., xxvii, Series B, p. 165, 1906.

EEE eee

EXPERIMENTS BEARING UPON THE MODE OF OXIDATION
OF SIMPLE ALIPHATIC SUBSTANCES IN
THE ANIMAL ORGANISM

(ACETIC ACID, GLYCOLLIC ACID, GLYOXYLIC ACID, OXALIC ACID,
GLYCOCOLL AND GLYCOL.)

By H. D. DAKIN.
(From the Laboratory of Dr. C. A. Herter, New York.)

(Received for publication, February 4, 1907.)

The study of the mechanisms of the chemical oxidations carried
out in the animal body is a problem which has long been recog-
nized to be of great importance, but the prosecution of this line
of research has met with so many experimental difficulties that
at the present time our knowledge of the subject is extremely
meagre. It is not yet possible to form a complete picture of
the mechanism of the oxidation of simple aliphatic substances
containing two or three carbon atoms; in fact, more is known
of the fate of complex substances belonging to the aromatic
series than those of the aliphatic group.

Oxidation of saturated substances, both by direct chemical
means and through the agency of the living organism usually
consists in the replacement of hydrogen atoms by hydroxyl
groups with formation of products which may or may not undergo
further decomposition. In the case of oxidations carried out
by the organism, the primary products are so seldom capable
of resisting further change that the intermediate stages be-
tween the primary and end products of oxidation are usually
unknown. The possibility of actually isolating intermediate pro-
ducts of oxidation from animal tissues and excretions is in a
large measure dependent upon the stability of the intermediate
products and whether the rate of their formation is great com-
pared with the rate of their decomposition. Unfortunately these
conditions seldom prevail.

By far the most satisfactory experimental method for the in-
vestigation of the course of oxidations would be to submit

57

58 Oxidations in the Animal Organism

single, definite compounds, such as are normally dealt with by
the living organism, to the action of roughly purified oxidizing
enzymes. Unfortunately we have no grounds for believing that
the oxidizing enzymes (oxygenases, peroxidases, aldehydases,
etc.) that have been studied apart from living cells, have anything
to do with the main changes of catabolism, since they are without
action upon the chief groups of substances which furnish energy
to the organism.

In some cases, in which direct chemical methods have been
inapplicable, indirect pharmacological methods have proved of
service; an example of this kind is found in the demonstration
of the formation of glycuronic acid as an intermediate product
of the oxidation of glucose and its excretion in the urine in com-
bination with drugs, such as aromatic alcohols.

A more direct method for the detection of intermediate pro-
ducts of oxidation is that employed by Paul Mayer’ and Hil-
debrandt,? and others, who have administered to animals enor-
mous doses of substances normally present in but small quantity.
When sufficiently large doses are employed, part of the substance
escapes oxidation and appears in the urine. In this way it is
possible to so fully utilize the oxidative capacity of the organism
that in some cases intermediate products of oxidation are pro-
tected from further attack and so are excreted and may be
detected in the urine. In this way it has been shown that
oxalic acid appears in the urine when an alimentary glycosuria
is set up in rabbits by the administration of large quantities
(forty-five grams) of sugar.

Both the pharmacological method of investigation and that
involving the administration of very large quantities of sub-
stances must be regarded as very unphysiological, and in fact
some writers go so far as to ascribe no value to results obtained
by these methods. Although I think it can hardly be doubted
that very valuable results have been and may yet be obtained
by following these lines of investigation, there is another method
which, though beset with very real drawbacks, does in some
respects involve less departure from the normal course of events

1 Paul Mayer: Deutsch. med. Wochenschr., 16-17, 1901; Zettschr. f. klin.
Med., vixii, p. 68, 1902; Zeitschr. f. physiol. Chem., xxxviii, p. 135, 1903.
*Hildebrandt: Jbid., xxxv, p. 141, 1902.

Hi: DD. Dakin 59

occurring in the organism. This mode of investigation of the
intermediate products of oxidation of a substance consists first
of ali in a theoretical consideration of the different products
which may be formed, followed by examination of the separate
behavior of each of these products in the organism. In some
cases, at least, the results will be such as will give a clear indi-
cation of some of the steps in the course of the degradation of the
substance.

The simplest example of the application of this principle may
be seen in the determination of the fate of formaldehyde in the
organism. When only catabolic changes are considered it will be
seen that the substance may theoretically be oxidized in the
following ways:

O

Vi

EC +O — CO+H-;0 CO+0=CO,

Ze
H
O

or
O O
Ji VA Vi
H:c +04 HC H-C+0=H,0+C0,
NS ~ oo
H OH OH

That is to say, that although the final products in each case are
carbon dioxide and water, the intermediate products may be
either carbonic oxide or formic acid. Further experiment shows
that carbonic oxide is not only poisonous but is absolutely resist-
ant against further oxidation by the organism. On the other
hand it is found that formic acid is actually oxidized to carbon
dioxide in the organism. It is, therefore, concluded that formal-
dehyde is oxidized in accordance with the second equation and
this is confirmed by the fact that the normal formic acid excre-
tion in the urine is increased after consumption of formaldehyde
derivatives. The application of considerations based on the
assumption that certain substances are attacked with difficulty
in the organism and so cannot physiologically be considered as
precursors of carbon dioxide, has been made by Pohl,' to several
cases of oxidation.

1Pohl: Arch. f. exper. Pathol., xxxvii, p. 413, 1895-1896.

60 Oxidations in the Animal Organism

The main objection to the method under consideration lies
in the crudity of our methods of administration which are often
incapable of bringing a substance in suitable quantity to the
proper sphere of activity of the cells concerned in its catabolism.
It is probably this difficulty which accounts for the fact that
certain substances which are believed to undergo active metabol-
ism in the body are excreted unchanged when introduced from
without.

In many ways a more than superficial resemblance appears to
exist between combustion taking place in flame and in the living
organism, and it is not impossible that the changes by which
energy is liberated from a straight-chain paraffin and a satu-
rated fatty acid have something in common. Our knowledge
of the mechanism of the combustion of hybrocarbons is largely
due to the brilliant series of researches carried out by Bone and
his co-workers.!. It has been shown that the oxidation of a
hydrocarbon by both slow and explosive combustion may be
regarded as a process involving the initial formation of unstable
hydroxylated molecules which subsequently undergo decompo-
sition into simpler products. For example, formaldehyde and
steam may be identified at an early stage in the combustion of
methane, the formaldehyde decomposing at the high tempera-
ture into carbonic oxide and hydrogen which are then burnt to
carbon dioxide and water. The changes are believed to take
place as follows:

CH, — CH;(OH) — CH,(OH), > H,O +CH,O— CO + H,—CO,+H.0

In a similar way aldehydes are formed by the combustion of
ethane and propane and undergo further decomposition at high
temperatures.

The detection with certainty of these intermediate products
of a reaction which usually results simply in the production of
carbon dioxide and water is of great interest and may be con-
sidered as justification for the belief in the formation of sub-
stances as intermediary products of metabolism which have but
a transient existence and which cannot be isolated at least

1W. A. Bone: Trans. Chem. Soc., 1xxxi, p. 536, 1902; 1xxxiii, p. 1074,
1903; Ixxxv, pp. 693 and 1637, 1904; Ixxxvii, p. 1232, 1905; Ixxxix, pp.
652 and 660, 1906.

H-: D. Dakin 61

from an organism under physiological conditions. Although the
analogy between combustion in the living organism and in flame
must not be pushed too far, for some of the changes in the latter,
especially those due to thermal conditions, such as the breaking
down of formaldehyde into carbonic oxide and hydrogen would
not be likely to occur in the cell, yet it can be hardly doubted
that a careful comparison of the two kinds of phenomena would
be of interest.

But few experiments have been made upon the fate of the
simple acids when introduced into the body. Schotten! showed
that large quantities of the sodium salts of the saturated fatty
acids could be readily burnt in the organism. Ten to twenty
grams of the fatty acids from formic to caproicacid, with the ex-
ception of propionic acid, were given to dogs by mouth in the form
of sodium salts. It was found that the quantity of volatile acids
in the urine after the administration of caproic, valeric and buty-
ric acids was not materially increased, indicating practically
complete combustion. In the case of acetic and formic acids
the results were similar except that the combustion was less
complete. After administration of twenty-five grams of sodium
acetate, rather less than ten per cent was recovered from the
urine while after twenty grams of sodium formate about twenty-
Six per cent was recovered unchanged in the urine. In all cases
the urines contained much sodium carbonate. Later experi-
ments by Pohl,? Gréhaut and Quinquand, and by Pellacani, in
the main confirm the earlier results of Schotten and tend to prove
that the formates are practically completely burnt in small doses
but when larger quantities are administered a part is excreted
unchanged. It is perhaps not out of place to mention that
these experiments in themselves are not sufficient to warrant
the deduction which is commonly made to the effect that the
lower fatty acids are much more difficultly oxidized than their
homologues, unless the relative permeability of the kidney for
the different salts is considered at the same time.

The mechanism of the combustion of formic acid does not pre-
sent many possibilities and it is probably correct to assume that

1 Schotten: Zeitschr. f. physiol. Chem., vii, p. 375, 1883
2, Pohl: Loc. cit:

62 Oxidations in the Animal Organism

a direct conversion into carbon dioxide and water takes place, but
in the case of acetic acid several modes of combustion are theo-
retically possible. It is natural to suppose that the methyl group
is the seat of attack and if we assume that combustion in the
tissues like the combustion in flame is essentially a process of
hydroxylation we may anticipate the initial formation of glycol-
lic acid. The introduction of a second hydroxyl group would
yield dioxyacetic acid which would give glyoxylic acid and water.
The glyoxylic acid on further oxidation might be expected to
give oxalic acid which in turn would yield carbon dioxide and
water. The change may be represented as follows:

CH;.COOH — CH.OH.COOH — CH(OH)2COOH — CHO.COOH
(H:0) — COOH.COOH — 2 CO,+H:,0

The likelihood of these reactions is increased by the fact that
acetic acid and glycollic acid both yield glyoxylic acid in direct
oxidation with hydrogen peroxide and glyoxylic acid in turn
is readily oxidized to oxalic acid.

On the other hand the possibility must be considered of the
glycollic acid decomposing in other ways. A direct decomposi-
tion into formic acid and formaldehyde is conceivable although
improbable. Since, however, glycollic acid yields formaldehyde,
formic acid and carbon dioxide on oxidation with hydrogen
peroxide! a similar decomposition might take place in the organ-
ism. The aldehyde would then be oxidized to formic acid which
would ultimately give carbon dioxide and water:

CH;.COOH — CH,0H.COOH — CH.O.;(CO. + H.0) — H.COOH
—- COE -++ H.0.

Accordingly the possible intermediate products of the oxida-
tion of acetic acid which we have to consider are glycollic acid,
glyoxylic acid, oxalic acid, formaldehyde and formic acid.
Reference must here be made to the experiments conducted by
Pohl, with the object of elucidating the mechanism of oxidation
of acetic acid, especially as my results in most respects do not
tend to confirm this investigator’s conclusions. The possibility
of the reaction involving the formation of formic or oxalic acids
has been rejected by Pohl on the ground that formic acid and

1H. D. Dakin: This Journal, i, p. 271, 1906.

2 eich TT I ER ROE a EE SE +

a

bie Wali 63

oxalic acid do not appear in increased amounts in the urine after
administration of acetates, but this does not appear to me to be
altogether convincing proof. Pohl regarded oxalic acid as stable
against further oxidation by the organism and so any oxalic acid
found as a product of oxidation of acetic acid should appear in
the urine. This, however, can be no longer regarded as correct.

Pohl pictured the oxidation of acetic acid by the organism as
taking place in the stages represented by the following equations:

CH3.COOH + O = CH2.0OH.COOH
CH,0H.COOH + 0 =CH(OH)2.. COOH =COH.COOH + H,0
COH.COOH + 0 =2 CO.+ H.O

The basis for this view is the fact that administration of acetates
is not followed by oxaluria and Pohl believed from the results
of his experiments that glycollic and glyoxylic acids were burnt
in the organism without the intermediate formation of oxalic
acid. New experiments by the writer have shown, however,
that considerable amounts of oxalic acid are formed by the oxi-
dation of glycollic and glyoxylic acids in the body.

It appears that at present we have insufficient evidence to
picture completely the mechanism of the oxidation of acetic
acid, but it is not improbable that glycollic, glyoxylic and oxalic
acids are intermediate products since they are all capable of fur-
ther oxidation by the organism. It must be admitted, however,
that it has not been possible to induce oxaluria as the result of
acetate administration (Experimental Part 1). The intermediate
production of formic acid may also have to be considered, since
the fact that an increased formate excretion has not hitherto
been found to follow acetate administration, cannot be held to
exclude this possibility.

From the fact that glycollic acid on oxidation with hydrogen
peroxide yields glyoxylic acid, formaldehyde, and formic acid,
one might anticipate that these products would be formed in the
course of its oxidation by the organism. Since glyoxylic acid,
as will be shown later, is at least in part, oxidized by the organ-
ism to oxalic acid, the presence of the latter substance might
also be expected.

CH,.OH.COOH —» COH.COOH —» COOH.COOH — 2CO2+H:,0
CH,.OH.COOH — H.COH + (CO, +H.0) — H.COOH — CO.+H20

64 Oxidations in the Animal Organism

Experiments (Experimental Part 2) showed that no increase
in formic or other volatile acid in the urine followed the adminis-
tration of glycollic acid or its salts given either by mouth or sub-
cutaneously to rabbits and dogs; on the other hand, very definite
evidence of the formation of oxalic acid was obtained, especially
when the glycollic acid was given subcutaneously. In some cases
the oxalic acid excretion is relatively large but under certain con-
ditions it may be but slight and when small doses are given by
mouth to dogs no increase in oxalic acid may be apparent. These
results are easily explained on the grounds that the oxalic acid
formed undergoes further oxidation to carbon dioxide and water.
Special experiments (4) showed that moderate quantities of
oxalic acid given subcutaneously to rabbits are readily burned.
These results are in accord with those of Hildebrandt! and Bak-
hoven’ (see p. 67).

Wher moderately large doses of free glycollic acid are given
by mouth a part of the acid appears unchanged in the urine and
the proportion of the total nitrogen of the urine in the form of
ammonia is raised.

Since the oxidation of glycollic acid by the organism proceeds
at least in part through the stage of oxalic acid, the intermediate
formation of glyoxylic acid is to be inferred. As glyoxylic acid
has been stated by Eppinger to form allantoin in the body it might
be conjectured that an increased allantoin excretion would follow
the administration of glycollic acid or its salts. Inno case, where
the free acid or the sodium, ammonium, or calcium salts were em-
ployed could such a change be satisfactorily demonstrated.

The combined results may be taken to show that glycollic acid
is oxidized in the organism through the stages of glyoxylic and
oxalic acids:

CH2.0OH COH COOH
= Nap al — 2 CO.+H;0
COOH COOH COOH

The amounts of oxalic acid found in the urine under certain
conditions indicate that a very large part, if not all, of the gly-
collic acid is broken down in this way, for a large part of the oxalic

‘Hildebrandt: Zeitschr. j. physiol. Chem., xxxv, p. 141, 1902.
? Bakhoven: Jahresber, f. Thierchemie, xxxii, p. 362, 1902.

fe Oe alin 65

acid initially formed must be completely burnt to carbon dioxide
and water. No evidence was obtained tending to favor the
belief that formic acid is an intermediate product in the reaction
but the probability of the further oxidation of the formic acid
renders negative insufficient evidence to disprove the possibility
of such a change.

The fate of glyoxylic acid in the organism has been investi-
gated with different results by Pohl and by Eppinger. The
former investigator concluded from his experiments that gly-
oxylic acid in moderately large doses! was completely burnt in
its passage through the organism, after the administration of
sodium glyoxylate by mouth. No glyoxylic acid could be
detected in the urine and oxalic acid was not considered to be
a product of its oxidation, although an inspection of Pohl’s
analytical results reveals the presence of a distinct oxaluria. No
toxic effects were noted. Eppinger, on the other hand, gave cal-
cium glyoxylate by mouth to dogs and believed that an increased
oxalic acid and allantoin excretion resulted therefrom. The
formation of the latter substance would be of considerable inter-
est as furnishing an example of the synthesis from a simple acid
derivable from both proteid and carbohydrates of a substance
closely allied to the purin bases. The formation of allantoin
from glyoxylic acid necessitates the addition of two molecules
of urea:

NH, O NH, HN— CH— NH
| | | |
OFC SS CH se (C40) = OFC C@
NH, ONC:@ NH, HN—C:O NH;
Allantoin.

Eppinger’s dogs which had received the glyoxylates died after
some days but no definite lesions were detected.

My own experiments have been made with the free acid, with
the calcium, ammonium, ands odium salts and with the methyl
ester. These substances have been given to dogs and rabbits
both by the mouth and subcutaneously. The results have shown
that glyoxylic acid is, at any rate in part, oxidized to oxalic acid
in the body, and that the proportion of oxalic acid appearing in

1 Two grams of the free acid for small dogs (6 kilos).

66 Oxidations in the Animal Organism

the urine varies according to the mode of administration, being
least when the glyoxylic acid is given by mouth. Considerable
variations in the oxalic acid excretion are found in even the same
animal with identical doses and this is to be explained on the
assumption that the further oxidation of the oxalic acid to carbon
dioxide and water is more complete under some conditions of the
organism than under others. No indication was obtained of the
presence of glyoxylic acid in the urine after administration of the
free acid and this result is confirmed by the fact that the ratio
of the ammonia to the total nitrogen of the urine is not increased.
No evidence was obtained of an increased formic acid excretion
although for reasons already mentioned this cannot be taken to
preclude the possibility of a part of the glyoxylic acid break-
ing down with the production of formic acid as an intermediate
product.

With regard to the formation of allantoin I am unable to con-
firm the results of Eppinger. An investigation of the available
methods for the estimation of allantoin! shows that many sub-
stances other than allantoin are precipitated and estimated along
with that substance, and although the ‘‘allantoin nitrogen’”’ of
the urine is frequently increased by glyoxylic acid administration,
it has not been possible to isolate the allantoin in the pure state
in more than the traces which are normally obtainable from
urines. In the experimental portion of this paper a criticism
will be found of Loewi’s method for the estimation of allantoin,
which was that employed by Eppinger. Both glyoxylic acid
and its salts in moderate quantities? were found to be compara-
tively non-toxic, certainly less so than oxalic acid. This result
agrees with Pohl’s experience and is in opposition to that of
Eppinger. It is perhaps conceivable that this divergence of
results may be due to the possible presence of the extremely
poisonous glyoxal in Eppinger’s preparation.

In the view provisionally put forward that glycollic, glyoxylic
and oxalic acids are intermediate products of the oxidation of

1 Loewi: Arch. f. exper. Path. u. Pharm., xliv, p. 19,1900; Swain: Amer.
Journ. of Physiol., vi, p. 38, 1902; Poduschka: Arch. f. exper. Path. u.
Pharm., xliv, p. 59, 1900.

?Two grams of the acidin the form of calcium or sodium salts do not
produce toxic symptoms in rabbits.

He oD: Dakin 67

acetic acid is accepted, it is obviously necessary to have definite
information as to the powers of the organism to carry the oxida-
tion of the oxalic acid further, to carbon dioxide and water.
Very many investigations have been made to decide this point
but with widely differing results. According to Gaglio, Wiener
and others’ oxalic acid given subcutaneously to dogs is excreted
unchanged, and the same result was obtained by Pohl when the
acid was introduced into a dog with a Thiry-Vella fistula. Many
experiments have been made in which destruction of oxalic acid
was observed when the acid was taken by mouth,” but as there
appears to be evidence that oxalic acid is decomposed by putre-
factive agencies these experiments are not completely convincing.

More modern experiments by Hildebrandt! upon the fate of
oxalic acid injected subcutaneously into rabbits showed that
from 60 to go per cent of the acid was oxidized in the body
and experiments of my own (Experimental Part 4) have com-
pletely confirmed these results. Confirmation of these results
is afforded by the fact that oxaluric acid, parabanic acid, and
alloxan are completely oxidized in the body when given to dogs
subcutaneously. The wide differences between the older and
more recent experiments can, I think, be accounted for by the
use of the quite inadequate method for the estimation of oxalic
acid devised by Neubauer and which for long was the only avail-
able one. In the experimental portion of this paper an account
is given of the method of oxalic acid estimation employed in my
own experiments.

The result of these experiments upon the fate of glycollic,
glyoxylic and oxalic acids in the organism, showing that they are

1 Gaglio: Arch. 7. exper. Path. u. Pharm., xxii, p. 246, 1887; Pohl: zbzd.,
XXXVii, p. 415, 1895-18096; Faust: zbid., xliv, p. 236, 1900; Abeles: Wen.
klin. Wochenschr., Nos. 19 and 20, 1892.

2 Baldwin: Jour. of Exper. Med.,v, p. 27, 1900; Marfori: Jahresber. }.Thier-
chemie, xx, p. 70, 1890; xxvi, p. 74, 1896; Lommel: Deutsch. Archiv f.
klin. Med., 1xiii, p. 599, 1900; Barth and Autenreith: Zettschr. 7. physiol.
Chem., XXXV, Pp. 327, 1902.

$Stradomsky: Virchow’s Arch. jf. path. Anat. u. Physiol., clxiii, p. 404,
Igol.

* Hildebrandt: Zeitschr. f. physiol. Chem., Xxxv, p. 141, 1902.

5 Luzatto: ibid., xxxvii, p. 225, 1902; Koehne: Inaug. Diss., Rostock,
1894.

68 Oxidations in the Animal Organism

capable of oxidation and that the oxalic acid is a product of oxida-
tion of glycollic and glyoxylic acids in the organism, appear to
me to give some degree of probability to the view that the oxida-
tion of acetic acid, at least in part, takes place with formation
of these substances as intermediate products:

CH;.COOH — CH:0H.COOH — CH(OH):COOH — CHO.COOH
+H,0 — COOH.COOH — 2CO,+H:,0

The question as to whether formaldehyde or formic acid are also
intermediate products cannot be regarded as settled.

It might be urged that the formation of more than traces of
a moderately toxic substance,such as oxalic acid as a product
of normal metabolism would be unlikely; but this objection can,
I think, hardly have much weight, for it 1s not necessary to assume
that intermediate products have more than a transient existence.
The final proof of the formation of formaldehyde in the assim-
ilation of carbon dioxide by plants, furnished by the masterly
experiments of Priestley and Usher, is an example of the produc-
tion by the cell of an extremely powerful protoplasmic poison in
the course of its normal activities. It seems reasonable to assume
that in cases such as these the cell’s defense consists in the pres-
ence of enzymes capable of rapidly converting the toxic sub-
stance into other products so that the concentration of the former
is always infinitely small. It is also probable that the oxidation
of intermediary substances formed in this way by intracellular
enzymes may be much more readily accomplished than when the
same substance is introduced from without in comparatively
high concentration.

The steps in the progressive oxidation of substances may in
some cases be so rapid that one may prefer to assume that the
intermediate products have no separate existence but that the
molecule of the substance undergoing change is altered in such a
way that if the reaction could be stopped at that particular
point, an intermediate product would be formed as the result of
intramolecular rearrangement, but that normally the reaction is
carried beyond this stage. This, however, is at present at any
rate as regards reactions in the living tissues, of theoretical rather
than practical interest.

The demonstration of the formation of oxalic acid from the

Epes Dalen 69

oxidation of glycollic acid in the organism is of interest in clear-
ing up the mechanism of the oxidation of glycol. Pohl and later
Paul Mayer found that injections of glycol into rabbits produced
a marked oxaluria and this result has been confirmed by my own
experiments (Experimental Part 5). Mayer also found that
when large doses were given, glycollic acid appeared in the urine
and was identified in the form of its phenylhydrazid. This
result appeared somewhat anomalous in view of the statement
of Pohl’s that glycollic acid did not yield oxalic acid in the organ-
ism. Pohl explained the formation of oxalic acid as arising from
the simultaneous oxidation of both alcohol groups, but the new
experiments make it appear more probable that glycol breaks
down in accordance with the following scheme:

CHLOH CHO COOH COOH COOH COOH
CHO CHLOH CH.OH C=(0OH), C=O COOH
|

|
H H

Since glycocoll gives rise to glyoxylic acid on oxidation with
hydrogen peroxide, it might be thought that glycocoll would be a
source of glyoxylic acid and hence of oxalic acid in the organ-
ism. Since glycocoll is also attacked with the liberation of
ammonia by many tissue ferments, the possibility of the forma-
tion of glycollic acid and of its oxidation products must be con-
sidered. At present, however, I have not been able to satisfac-
torily demonstrate a marked oxaluria as the result of glycocoll
catabolism (Experimental Part 6). The probability that gly-
cocoll does break down to some extent in the way outlined is
strengthened when the fact is considered that gelatin, which
contains large quantities of glycocoll, when consumed in large
quantities does lead to a slight oxaluria in man and sometimes
in dogs, although the increase in oxalic acid excretion is usually
only five to fifteen milligrams. This observation was first made
by Lommel! and has been confirmed by Stradomsky,’ Mohr and
Salomon,’? and by the writer (Experimental Part 7).

1 Lommel: Loc. cit.
?Stradomsky: Loc. cit.
3 Mohr and Salomon: Deutsches Arch. f. klin. Med., Ixx, p. 486, rgor.

70 Oxidations in the Animal Organism

It may perhaps be not out of place to refer to a remarkable
paper by Klemperer and Tritschler' in which a single experi-
ment is described relative to the production of oxaluria in dogs
as the result of glycocoll injection. These authors conclude that
an increased oxalic acid excretion, equivalent to one hundred
and fifty-six milligrams of oxalic acid, resulted from the injection
of five-tenths of a gram of glycocoll. Such a result is astonish-
ing, but when it is seen that the supposed oxaluria was at its
maximum twenty-five days after injection, it may be inferred
that the experiment is of questionable value. The same authors
describe a blood ferment causing a destruction of oxalic acid,
which, from the authors’ figures, appears to be most active at
the boiling point of the oxalic acid solution. Many other criti-
cisms of this work may be made but they would serve no use-
ful purpose.

Another source of glyoxylic acid and hence of oxalic acid may
be sought in creatin and creatinin, which readily yield glyoxylic
acid on oxidation with hydrogen peroxide. The probability of
creatin being a precursor of oxalic acid was long ago realized by
Kithne and recently Klemperer and Tritschler have claimed to
have produced such an oxaluria experimentally. Stradomsky,
however, observed only negative results and it appears that even
if oxalic acid were an intermediate product of oxidation of creatin
it is improbable that any appreciable amount would escape fur-
ther oxidation.

Much has been written upon the source of the traces of oxalic
acid which normally appear in the urine but there is but little
unanimity of opinion upon the subject. The fact that oxaluria
may persist after long starvation shows that the oxalic acid is
not exclusively of alimentary origin. Wesley Mills’ found that
dogs excreted most oxalic acid upon a flesh diet while Haas®
found in some cases an increase of oxalic acid followed increased
carbohydrate consumption. Liithje* and Stradomsky on the
other hand found that the oxalic acid output was highest upon a

1 Klemperer and Tritschler: Zeitschr. f. klin. Med., xliv, p. 337, 1902.

2 Wesley Mills: Virchow’s Arch. f. path. Anat. u. Physiol., xcix, p. 305,
1885.

’ Haas: ‘‘Ueber Oxalurie,” Inaugural Dissertation, Bonn, 1894.

4 Lithje: Zettschr. f. klin. Med., xxxv, 1808.

H. D. Dakin 71

flesh diet, lowest on a carbohydrate-rich diet, and an interme-
diate result was obtained with diets containing an abundance of
fat. A careful consideration of what is known of the break-
down of fats and higher fatty acids in the animal body, especially
with regard to the production of f-oxybutyric acid as an inter-
mediate product of normal metabolism does not exclude the
possibility of the production of oxalic acid as a further inter-
mediate product of oxidation. Since, moreover, there is good
evidence for the possibility of its formation from carbohydrates
and from some of the constituents of flesh, such as glycocoll,
creatin, creatinin, and probably other substances, it appears
unsafe to attempt to refer normal oxalic acid excretion to one
particular class of food-stuffs. It is more probable that it may
be formed from all classes of food, but that under conditions of
normal metabolism only a minute fraction of it escapes further
oxidation.
EXPERIMENTAL PART.

Analytical Methods.

THE EsTIMATION OF Oxatic Acip. There can be little doubt
that much of the confusion which exists with regard to the origin
and fate of oxalic acid in the organism is due to the employment
of faulty methods of analysis. Of the various methods proposed
for the estimation of oxalic acid, those associated with the names
of Salkowski and Barth and Autenrieth are probably the most
trustworthy, and the process employed in the present investiga-
‘tion is a modification of their methods. As large a quantity of
urine as is available is heated for twenty minutes on the water-
bath after the addition of about 5 per cent hydrochloric acid. The
object of this procedure is to hydrolyze the oxaluric acid, for
in the present state of our knowledge there appears to be no
reason for differentiating between the free oxalic acid and that
combined with urea. Anexcess of calcium chloride is next added
and the liquid made fairly strongly alkaline with ammonia and
allowed to stand in a warm place over night. The bulky pre-
cipitate is filtered off through a folded paper and well washed
with boiling water. The filtrate should be preserved for a short
time longer in order to give the finely divided calcium oxalate
an opportunity to sediment. If necessary a second filtration is

72 Oxidations in the Animal Organism

made through Swedish filter paper and adhering precipitates are
placed in a beaker and warmed with a small quantity of dilute
hydrochloric acid and the liquid is then filtered through a small
folded filter. The insoluble residue must be thoroughly washed
by digesting with successive portions of hot water. The filtrate
is concentrated on the water-bath to about 5 to 7 cc. and is then
transferred to the extraction tube of an apparatus for continuous
extraction with ether. The extraction with ether is carried out
at a fairly rapid rate and usually was allowed to proceed for at
least five or six hours. It is probable that two hours’ extraction
in an efficient apparatus is more than sufficient. About 20 cc.
of water are added to the ethereal extract and the ether is then
distilled off. The aqueous solution is filtered from any trace of
flocculent insoluble residue and precipitated with calcium chloride
and ammonia. The liquid is finally made decidedly acid with
acetic acid and kept in a warm place for twenty-four hours. The
precipitate of calcium oxalate is filtered off on a small Swedish
filter paper, washed with hot water and either weighed as calcium
carbonate or titrated with decinormal permanganate in the
usual way. Ina properly conducted experiment the two methods
give identical results. If the calcium oxalate precipitated should
be contaminated with any obvious amount of pigment, it is
advisable to redissolve it in a few drops of hydrochloric acid and
to reprecipitate with ammonia after adding a little calcium
chloride. The addition of alcohol to the ether used for oxalic
acid extraction as has been recommended, is disadvantageous and
leads to the contamination of the calcium oxalate precipitate
with impurities.

The rapidity and completeness of extraction of the oxalic acid
by ether from aqueous solution was tested in the following experi-
ment. Pure recrystallized ammonium oxalate equivalent to
17.8 mgms. of anhydrous oxalic acid was dissolved in 10 cc. of
water to which a few drops of dilute sulphuric acid were added
and the solution was then extracted with ether in a continuous
extractor. The rate of extraction was such that the return flow
of ether to the reservoir was just insufficient to form a continuous
stream but gave a rapid succession of drops. After two hours
and twenty minutes the receiver was changed and extraction
allowed to proceed for a further period of three hours. The two

ea) Dakin 73

re)

ethereal extracts and the aqueous residue were then separately
analyzed for oxalicacid. The results were as follows:

UACIGICAKCN Ress alle ss kis k apuila eos @ econ 17.8 mg.
ee found in 1st ether extract (24 hrs.)17.8 mg. | Total 18.0
“ oe Od i (er hess) 0:2 f mg.
a = aqueous mesidues. 4 -aa- 00.0

This result shows that under the prescribed conditions, oxalic
acid in such quantities as would be found in urines, etc., may be
extracted quantitatively in a few hours.

THE EsTIMATION OF ALLANTOIN. Methods for the estimation
of allantoin have been described by Loewi! and by Poduschka?
depending upon the precipitation of the allantoin in the form of
silverand mercury compounds. The analytical results furnished
by Loewi would lead us to conclude that the method was one of
great accuracy. In four experiments in which allantoin was
added to urine, the estimated amount differed from the actual,
on an average, by less than two milligrams. In spite of the
greatest care I have been completely unable to confirm Loewi’s
results and have come to the conclusion that the method is very
misleading. It is unnecessary to go into the details of Loewi’s
method but it may be mentioned that it depends upon the
removal of certain interfering substances by previous precipi-
tation with mercurous nitrate, followed by the precipitation of
the allantoin in the form of the silver salt by means of silver
nitrate and magnesia. The magnesia is chosen in place of the
more customary ammonia on account of the fact that an excess
of the precipitant does not cause resolution of the precipitates.
Loewi states with regard to the precipitate so obtained ‘‘Es
halt von N-haltigen Korpern nur Allantoin and kann direct zur
N-Bestimmung benutzt werden.” Inno single case have I been
able to confirm this result. The silver precipitates have been
carefully decomposed with sulphuretted hydrogen and freed
from magnesia and it was found that in every case very consider-
able quantities of impurities were present. In every case sub-
stances precipitable by phosphotungstic acid in acid solution
were present in large amount, and smaller quantities of bodies

1 Loewi: Loc. cit., p. 1.
2? Poduschka: Loc. cit.

74 Oxidations in the Animal Organism

reacting in alkaline solution with diazobenzene sulphonic acid
with production of a red coloring matter. The latter reaction
is probably due to some substance containing the iminazol ring,
such as histidin or the purin bases. Tyrosin was not present.
Allantoin is not precipitated by phosphotungstic acid nor does
it react with diazo salts with formation of colored products. In
many cases the presence of urea was detected and it was found
that this substance is precipitated by silver nitrate and magnesia
to a considerable extent. The frequent freedom of the allantoin
precipitate from urea depends upon the presence of ammonia in
the urine. In many cases silver precipitates were obtained which
contained large amounts of nitrogen but from which no trace of
allantoin could be isolated in crystalline form either by Loewi’s
mercuric nitrate method or by other procedures. Many of the
criticisms applied to Loewi’s method hold equally for the other
methods for the estimation of allantoin and it is, therefore, con-
cluded that at present no satisfactory quantitative method is
known. It therefore follows that results which are not based
upon the actual isolation of the allantoin in crystalline form are
of questionable value.*

The procedure adopted in the present experiments consisted
in the precipitation of the allantoin along with urea and many
other substances by means of mercuric nitrate. The acidity of
the solution was kept low by addition of sodium carbonate.
The mercury precipitate was filtered off, washed and decom-
posed with sulphuretted hydrogen. The filtrate was concentrated
slightly and precipitated with silver nitrate and ammonia accord-
ing to Poduschka’s method. The silver precipitate was filtered
off, washed, decomposed with sulphuretted hydrogen, and the
filtrate concentrated to small bulk and examined microscopically
for allantoin crystals. An estimation of the total nitrogen was
also made by Kjeldahl’s method but much reliance cannot be
placed upon the figures so obtained as other substances are pre-
cipitated as well. If allantoin crystals can be detected in the

1A great many experiments were made to try and devise a satisfactory
method for the estimation of allantoin but hitherto without success. The
fact was established that allantoin could be quantitatively oxidized to
oxalic acid by boiling alkaline permanganate and this fact may possibly
be of analytical value.

|
|

le. Dalian 7,

syrupy solution, they may be freed from mother liquor by wash-
ing with a minimal amount of cold water and then recrystallized
from boiling water. The method involves considerable loss of
allantoin.

Ammonia determinations were made in the fresh urines accord-
ing to Folin’s method.

Experimental Results.

In almost every case, the results given are representative of a
much larger number of experiments than those actually recorded.

fr. THE EFFECT OF ACETATES UPON OXALIC ACID EXCRETION.

(a) Dog, 15 kilos, given 25 gm. of sodium acetate crystals dissolved in
water, subcutaneously. The urine passed subsequently was clear and
strongly alkaline. It effervesced vigorously on addition of acids.

Total oxalic acid in urine before injection (24 hrs.) . .0.0045 gram.
ae a after > (st24vhrss) 20006 5a
36 ac eS “(2d 24 hrs.).0.0009 *

(6) Dog, 7 kilos, given solution of 25 gm. of sodium acetate crystals,
subcutaneously.

Total oxalic acid in urine before injection (24 hrs.) . .0.0050 gram.
Ge ae after << S@ist 24shirss) = OLO072—aee
es ee oe se (2d 24 hrs.). 0.0021 ‘

The slight increase in the oxalic acid excretion on the first
day after the acetate injection is probably to be ascribed to the
diuretic effects of the salts and is followed by a decreased excre-
tion on the second day.

The experiments do not demonstrate the formation of oxalic
acid as the result of acetate combustion in the normal organism.

2. EXPERIMENTS UPON THE FATE OF GLYCOLLIC ACID IN THE
ORGANISM.

(a) Rabbit, 1550 grams, received subcutaneously 2.0 gm. of Kahl-
baum’s glycollic acid in the form of sodium salt. The rabbit had been
fed for some days on bread and water. It received 100 cc. of water each

day by mouth.
Urine 48 hours Urine 48hours Following 48
before injection. after injection. hours.

Volatile acids as formic acid.. 0.0165 0.0163 a=
Opcalietaci clear cyendeae ea avs scuarae 0.0009 0.0117 0.0049

76 Oxidations in the Animal Organism

(6) Same rabbit as Experiment (a). Two grams of glycollic acid as
sodium salt given subcutaneously,

Oxalic acid, 48 hrs. before injection
ih es after RE MEEMRD (08 Tay iaic nin ai ovo fn sie 0.0360

(c) Rabbit, 2 kilos, 2.0 grams of glycollic acid given subcutaneously as
the sodium salt.

Oxalic acid 48 hrs, before injection .......1...005see eevee 0.0056
as atten SADT rien atk kOe’ SS eaer ee owoue 0.0400

es PAMALS MLCT Me eet eT Sera sie. lac. soya. o ordip ele vere panedENe 0.0170

is “ ee ate oo, cog pebee 0.0067

(d) Dog, ro kilos, 2 gm. of free acid in roo cc. of water given by mouth
in two equal portions about two hours apart. There was no vomiting.

N of NH3
Total whentotal Oxalic
Isis NasNHz N=100. acid.
Urine 24 hrs. before acid given 5.40 0.191 3.54 0.0045
“ es after nf 4.07 0.192 4.72 9.0095
re rg a A 4.12 0.165 4.00 0.0017

(e) Dog, 11 kilos, 1.11 gram free acid in 100 cc. of water given by mouth
No vomiting.

N of NH;
Total when total Oxalic
N NasNH3. N=100. acid.
Urine 24 hrs. before acid given 3.17 0.121 3.32 0.0050
£ after Ry 4.93 0.223 4.52 0.0074
oi & TART te cca cd ve 2.80 0.150 5.36 —
U, as AE a aes eee ae 4.27 0.137 33074 | ~--

(7) Dog, 12 kilos, 3.34 grams of calcium glyoxylate given by mouth.

Oxalic acid in urine 24 hrs. before ................... 0.0083
as cs cs EE. Acie rs oiste. <8 Re 0.0078
a a as laberaaeie.. ss ae: eee 0.0020

The results show that a very decided increase in oxalic acid
excretion followed administration of glycollates to rabbits.
Small doses of the free acid or calcium salt given by mouth to
dogs does not produce marked oxaluria. No increase in formic
acid (volatile acid) was observed. The ratio of nitrogen of am-
monia to total nitrogen was increased by administration of free
glycollicacid. Itis probable that thisis due to part of the glycollic
acid escaping combustion and appearing unchanged in the urine.
The fact that glycollic acid has been detected in urine after glycol
administration is in accord with this view. Noincrease in allan-
toin excretion was observed.

bee Ds Dale sy

3. EXPERIMENTS UPON THE FATE OF GLYOXYLIC ACID IN THE
ORGANISM.

(a) One gram of the free acid was exactly neutralized with sodium car-

bonate and then diluted to 100 cc. and given by mouth to a rabbit weigh-

ing I500 grams.

The animal remained quiet for some time but no toxic
symptoms were noted.

Volatile
Oxalic acid as
acid. formic acid. Allantoin.
Urine 24 hrs. before.....
ae

0.0030 0.0110 no crystals obtained.
fame SALCET ss is - 0.0648 0.0648 <“

“cc “ce

(0) Same experiment except 1.5 grams of dry calcium glyoxylate given.

Oxalie

Volatile
acid.

acid. Allantoin.
Urine 48 hrs. before ....0.0003 0.0136

minute trace.
Gi aes aro ee 0.0275 0.0152 no increase.
(c) Rabbit given 1.842 gram of dry calcium glyoxylate subcutaneously

Oxalic acid in urine 48 hrs. before
“cc “i

Ret: SE ee 0.0050
72 hrs. after

(d) Rabbit given 1.25 grams of methylglyoxylate in roocc. of water by
mouth,

Oxalic
acid. Allantoin,
WirineSGihrs: before. «2.556 000% 00046 0.0058 doubtful trace.
el Om Cc omcatber:

fre ae IE see 0.0297 ts ae

(e) Dog, 17 kilos, given 2.8 grams of free glyoxylic acidin 130 cc. of water
by mouth. After a short time vomiting occurred and o.

5 gram of acid
was ejected. No further vomiting took place.

Nas NH3
Oxalie Total N per NasNH3_ when total
: acid. Allantoin. 100 ec. per 100cc. N=100.
Urine 24 hrs. before. .0.0121 trace 3.25 0.138 4.24
us a after

0.046 3.56
later

ae ae

..0.0170 nocrystals 1.29
.. 0.0045 nocrystals —

(f) Dog, 9 kilos, given 2.0 grams of free glyoxylic acid in 150 cc. water
in two portions five hours apart.

No vomiting occurred.

NasHN3
Total N per N asNH3

when total
100 cc. per 100 ce. N=100.
Urine 24 hrs. before (faint acid)

pera 5.98 0.274 4.58
after (neutral)........ 4.54 0.189 4.16
SS oA later (faint acid) ...... 5.40 0.191

3.54
The results show that a marked oxaluria follows the adminis-
tration of salts of glyoxylic acid in the case of rabbits. With

78 Oxidations in the Animal Organism

dogs the oxalic acid excretion may be scarcely increased. No
evidence was obtained of the formation either of allantoin or
of formic acid. The administration of the free acid was not
followed by an increase in the NH,:N ratio. It is improbable
that any material amount of glyoxylic acid is excreted un-
changed.

4. EXPERIMENTS UPON THE FATE OF OXALIC ACID IN THE
ORGANISM.
(a) Rabbit, 1600 grams; 3.3 cc. of ammonium oxalate solution (equiva-

lent to 0.0990 gram oxalic acid) injected subcutaneously. The rabbit
was fed on bread and water.

Oxahcacidain urine, 24hrs, before. 2-32 32h) 2 ye le ctl oi 0.0030
“ as ay 117 A ee ee CE re 0.0135
“se “ as POE CT tates sce. 2ek Ree a oe oh Sate 0.0051

(b) Rabbit, r550 grams, ammonium oxalate equal to 0.051 gram of
oxalic acid injected subcutaneously.

Oxanhe/acidin wrme24-hrs. before: ......:...--.<s.00s 0.0015
% es de EEGEIYS SP ters ors oe ee 0.0077
_ nS -t LET SN Re eee gee Ame re hte 0.0030

(c) Rabbit, 1000 grams, ammonium oxalate solution equivalent to
0.297 gram oxalic acid injected subcutaneously. Rabbit died after 48
hours.

Oxalheacidiunwumne 24 hrs before. ..2....-. 6+ «.s..sene- 0.0078
cS - aC ARC T Bee ite rs eee ee 0.0312
ee of ae | EH ne) act ae ee Re es) 0.0135

The results show that moderate amounts of oxalic acid are almost
completely burnt in the organism—larger amounts are not so
completely oxidized. Similar results were obtained with dogs.

5- OXALURIA PRODUCED BY GLYCOL ADMINISTRATION.
(a) Rabbit, 2250 grams, given 2.5 grams of glycol diluted with water,
injected subcutaneously.
Urine 45 Des) Dette jem se riers sisters es ws 0.0040 oxalic acid.
ge Malar 2 ee a) oy, on 0.0600 “ o
(6) Repetition of the preceding experiment; 2.5 grams of glycol by
subcutaneous injection.

Oxalictaciasnesirineseuhrssatien’. oc ae .vevoc core n 0.0750

CO es

Fie Dakin 79

The results show a marked oxaluria as the result of glycol
administration, confirming the previous results of Paul Mayer
and Pohl.

6. EFFECT OF GLYCOL ADMINISTRATION UPON OXALIC ACID
EXCRETION.
(a) Rabbit, 1550 grams; previous diet of bread and water; a solution

of 2.5 grams of glycocoll injected subcutaneously. Rabbit starved 48
hours after injection; 100 cc. of water per day given by mouth.

Urine 48 hrs. before..... een oes 0.0049 gram oxalic acid.
cE BRR eOLUCE Sfp? ws. ie exes Acer 0.0045 uf a
(6) Repetition of same experiment.

Wrme 45hrs-petore =. .... 2... eee 0.0015 gram oxalic acid.
ch SEMEL EGC Tyere yen oy ayo hc,» “oysn\ 5's Sent 0.0035 ie cs

The results show that in the case of rabbits no appreciable
oxaluria follows the administration of glycocoll by subcutaneous
injection.

7. EFFECT OF GELATIN-FEEDING UPON OXALIC ACID EXCRETION.

(a) Fifty grams of gelatin were dissolved in 1400 cc. of water with the
addition of a little sherry and sugar. It was consumed by a healthy man
within two hours. Before and after the experiment care was taken to
avoid food which would be liable to yield oxalic acid.

Wirine24:hrs: before . 2.2 .2.62 25.22 +s 0.0008 gram oxalic acid.
o SET a aiea 2,0, Seale kd let, 0.0190

(6) Repetition of same experiment; 40 grams of gelatin consumed.

Wirine 24 hrs*before... 5.2... 2 et we 0.0270 gram oxalic acid.

a RULE CLD RRR a Ree de 0.0332 o oS
(c) Repetition of same experiment; 40 grams of gelatin consumed.
Winintes24ars. before) 2a 6.4 wis sis ees 0.0093 gram oxalic acid.

ie 2) od ab 5 51 atch Ot, ne ae 0.0160 e eS

The results are similar to those obtained by other observers
and appear to show that a slight oxaluria does follow the con-
sumption of large. quantities of gelatin, at least in the case of
man.

ON THE REDUCTION OF BARIUM SULPHATE IN ORDI-
NARY GRAVIMETRIC DETERMINATIONS.

By OTTO FOLIN-

(Received for publication, February 4, 1907.)

By copying the title of my recent paper ‘‘On Sulphate and
Sulphur Determinations’’* (a device rarely adopted except in
purely controversial and polemical communications) S. F. Acree?
has rather vigorously called attention to his criticisms of my
brief statement that barium sulphate precipitates are not ordi-
narily reduced when ignited together with the filter paper upon
which they have been collected.

Acree maintains that barium sulphate is nearly always reduced
when filter paper is employed in sulphate determinations, and that
the reduction is often extraordinarily large. My statement was
made on the basis of what appeared to me an adequate number
of tests, but without particular emphasis, because I still found
it necessary for good results to use Gooch crucibles instead of
filter paper. Since the appearance of Acree’s paper the point
has again been tested, and as before only a variable mechanical
loss and no appreciable reduction of barium sulphate has been
found.

Forty cc. of approximately tenth normal sulphuric acid gave 470.6 mg.
barium sulphate as the average of three closely agreeing determinations
by the help of Gooch crucibles.

1. The precipitate from 40 cc. of the above solution collected on
“black ribbon,’”’ S. and S. filter (No. 589, diam. 9 cm.), dried, and ignited
in a platinum crucible, weighed 465.5 mg. After adding and carefully
boiling off 3 cc. of conc. sulphuric acid the residue weighed 465.7 mg.

2. Another precipitate from 40 cc. of the solution was collected on
three filters (folded together). After carefully drying, charring and
burning the weight of the precipitate was found to be 465.5 mg. After
treatment with six drops of 50 per cent sulphuric acid the weight remained
unchanged (465.6 mg.), nor did a second treatment with 1 cc. of conc.
sulphuric acid yield a different result (weight 465.3 mg.)

'This Journal, i, p. 131.
7This Journal, ii, p. 135.

82 Reduction of Barium Sulphate

3. A third precipitate collected on two filters weighed 469.8 mg.
After sprinkling 1.005 gm. of pure grape sugar over this precipitate and
carefully charring and burning the mixture a loss of 1.3 mg. due to reduc-
tion was obtained (weight 468.5 mg.) The residue was then moistened
with water and three drops sulphuric acid and re-heated. The loss due
to reduction was recovered (weight 470.2 mg.), and the escape of some
hydrogen sulphide was noted.

The experiments 1 and 2 recorded above were the only ones which gave
precipitates the original weights of which did not exclude any appreciable
reduction of sulphate by the burning of the filters. Several other deter-
minations were made by the use of filter paper, but in these the amount of
barium sulphate found corresponded so nearly to the theoretical figure
(470.6) as to be well within the limits of unavoidable experimental error.

The final three determinations made by the use of filter paper, for ex-
; ample, gave the following values: (1) 469.6; (2) 470.7, (3) 470.2.

From these experiments it would appear. that barium sulphate
is even more resistant to the action of reducing agents than I
had previously supposed and consequently I can not concede the
validity of Acree’s criticism of my earlier statement. The results
of Acree and his students furnish, of course, the proof, if any
proof is necessary, that it is possible to reduce barium sulphate
to sulphide at ignition temperatures; but, in the light of positive
experience to the contrary, they do not prove that it is difficult to
avoid such reduction.

I have also been unable to confirm Acree’s statement that
it is difficult to convert barium sulphide into the sulphate by
treatment with sulphuric acid. His attempt at explaining this
alleged difficulty on the basis of ‘‘ occlusion” is scarcely tenable in
view of the fact that barium sulphate is very soluble in hot
concentrated sulphuric acid. In his experiment No, 3, for exam-
ple (p. 138), he repeatedly added enough sulphuric acid (3 cc.)
to dissolve the entire barium sulphate precipitate (.45 g.)

ON THE OCCURRENCE AND FORMATION OF ALKYL
UREAS AND ALKYL AMINES.

By OTTO FOLIN.
(From the Chemical Laboratory of McLean Hospital, Waverley, Mass.)

(Received for publication, February 4, 1907.)

In connection with feeding experiments with creatin made
during the past year it seemed desirable to determine whether
any of the missing creatin could have appeared in the urine in the
form of methyl urea or methylamine. The results were negative,
but in the course of the investigation I found that all human
urines contain certain small amounts of a substance which might
very well be methyl urea, and that some urines, particularly
typhoid urines, also contain appreciable quantities of methyl
amine.

For the detection of these substituted ammonia derivatives
I have used Hoffmann’s iso-nitril reaction. The test is made as
follows: The ammonia-containing distillates obtained from urea
and from ammonia determinations are slightly acidified and
concentrated by boiling to 50 or 75 cc. The solutions are then
transferred to small Kjeldahl flasks, and further concentrated to
a volume of from 5 to 15 cc. To the remainder are added about
25 cc. of saturated alcoholic potash solution and a few drops
of chloroform. On gently boiling these mixtures the intense,
disagreeable odor of iso-nitrilis obtained. The odor becomes par-
ticularly characteristic after most of the ammonia has been driven
off and the remaining solution has cooled.

The possible occurrence of methyl urea in urine was the sub-
ject of considerable investigation and discussion some twenty-five
years ago, but the positive results recorded by some, notably by
Shiffer, were never accepted, and in recent years this problem
has, as far as I know, received no further attention.

Shiffer’s results may indeed appear conclusive (Zeitschr. f. physiol.
Chem., iv, p. 237, 1880) to the casual reader, for he claimed to have
shown that the methyl amine group of creatinin did not interfere with

83

84 Alkyl Ureas and Alkyl Amines

Hoffmann's reaction as he applied it. This contention of Shiffer’s was,
however, evidently never accepted; and rightly so, for creatin and
creatinin do split off methyl amine when boiled in caustic soda or potash
solutions. .

The question of the presence or absence of traces of methyl
urea in normal urine is perhaps not a very important one, except
in so far as it is a part of the broader question, as to the origin of
other methylamine groupsin urine, such as those of methyl purins
and of creatinin. If all methyl amine groups in urine are derived
from preformed methyl amine groups in the food, as Koch, for
example, has recently attempted to explain the formation of crea-
tinin from the trimethyl amine of lecithin, then one can attach but
little importance to the methyl urea and methyl amine of urine, for
then they may represent only a certain excess of such methyl
amine groups in the food just as ordinary urea to a large extent
represents the unnecessary nitrogen of the amido acids. On the
other hand, if the methyl amine groups in urine are not necessa-
rily derived from preformed methyl amine groups in the food but
can just as well be derived from the abundant amido acids of
ordinary proteins, then the occurrence of methyl urea and of
methyl amine salts is important because then we have no reason
to assume that they may not occasionally occur in very large
quantities.

The main point which I wish to bring out in the present com-
munication is that methyl or other alkyl amine groups are
obtainable from ordinary amido acids, and therefore from all
proteins.

The formation of ammonia and through ammonia of urea, in
protein catabolism is in all probability more a matter of hydroly-
sis than of oxidation.” Similarly in the decomposition of protein
derivatives by boiling sulphuric acid, as in the Kjeldahl method
for the determination of nitrogen, the formation of ammonia is
undoubtedly the result of hydrolysis rather than of oxidations.’

It occurred to me that there is no-binding reason why in hydro-
lytic cleavages of this kind the nitrogen atoms should be quan-
titatively separated from the carbon atoms to which they are

1W. Koch: Amer. Journ. of Phystol., xv, p. 15, 1905.
2 Folin: Amer. Journ. of Phystol., xiii, p. 124, 1905.
3 Folin: Zeitschr. f. physiol. Chem., xli, p. 239, 1904.

Otto Folin 85

attached. Schematically such cleavage can just as well be made
to yield methyl amine as ammonia.’

(1) =C—CH.! NH.+H (OH

@) =C—CH.NH,+ HOH

The experiments undertaken to test the correctness of this
point of view gave at once positive results. The distillates
obtained from ordinary Kjeldahl determinations of the total
nitrogen, when concentrated and further treated as described
above, give invariably intense iso-nitril reactions. Witte’s
peptone, urine, creatin, creatinin, glycocoll, aspartic acid, and
hippuric acid, all give it,? and it makes no appreciable difference
whether the decomposition is produced by sulphuric acid alone
or by sulphuric acid plus copper sulphate or mercury.

These results prove that methyl amine (or other alkyl amines)
can be formed by hydrolysis from ordinary amido acids, and in
the light of this fact there is, I think, no reason for assuming any
genetic relationship between the methyl amine groups of the food
and those of the urine where experimental evidence in favor of such
a relationship is wanting.*

Further, if methyl amine, and therefore also methyl urea,
can be formed by the catabolism of ordinary amido acids, then
it is neither impossible nor particularly improbable that a syste-
matic investigation of the urines from a large number of different
kinds of animals would show that some contain relatively large
quantities of such substituted ammonia derivatives; nor is their
exclusive formation and elimination as abnormal metabolic pro-
ducts in diseases excluded.

1 The fact that the Kjeldahl process ordinarily gives perfectly accurate
analytical results does not prove that the nitrogen has been quantitatively
split off as ammonia, because if any methyl amine were formed and not
subsequently oxidized it would come over together with the ammonia in
the distillation and would act as ammonia in the titration.

2 Pure urea, as was to be expected, gives only ordinary ammonia.

3 The recently discovered fact that the human organism does not even
convert food creatin directly into urinary creatinin (see Festschrift Jur
Olof Hammarsten, iii, 1906) illustrates how fallacious such assumptions
may be.

86 Alkyl Ureas and Alkyl Amines

The occurrence of tetra- and pentamethylene diamines in some cases
of cystinuria may be considered as an illustration of such an abnormal
metabolism. Lysin and arginin are in all probability, as C. Neuberg!
has pointed out, the direct precursors of these diamines, though no one
has yet been able to verify Neuberg'’s remarkable, not to say peculiar,
experimental results.

Unfortunately I have not yet been able to find a quantitative
method for the determination of methyl amine in the presence
of ammonium salts; and without such a method it is not pos-
sible to determine with any degree of certainty just what rdle
the formation of methyl amine and methyl urea may play in
protein metabolism. The data which I have so far obtained
in attempts at quantitative estimations indicate that certainly
not more than 2 or 3 per cent of the urea-nitrogen (corresponding
to 4 to 6 per cent of the urea) is normally eliminated on nitrogen-
rich diets as methyl urea. On low nitrogen diets the methyl
urea elimination diminishes but never wholly disappears. On
low nitrogen diets in typhoid fever the ammonia elimination is
very high, and this ammonia contains, I think, an admixture of
at least 5 or 6 per cent of methyl amine.

1C, Neuberg: Zeitschr. f. physiol. Chem., xliii, p. 352, 1904.

ON THE SYNTHESIS OF PROTEIN THROUGH THE ACTION
OF TRYPSIN.

By ALONZO ENGLEBERT TAYLOR.

(From the Hearst Laboratory of Pathology, University of California.)

(Received for publication, March 12, 1907.)

The attempts to verify in biological material the laws of mass
action, equilibrium and partition have not been uniformly suc-
cessful. In some instances the verifications were easily accom-
plished, in others success could not be attained. The difficulties
are resident in the complexities of the systems under experi-
mentation and in the lability of the reactions concerned. The
same comments apply to the verifications in the fermentations
of the lawsof catalysis. Inone particular, however, inthe exceed-
ingly important matter of synthesis by ferment action, the experi-
mental success has been unexpectedly uniform. Following the
first report of a successful reversion by Hill,! successful experi-
ments were reported by Kastle and Loevenhart,? Bernizone,* and
Emmerling,*‘ and these have been followed since by those of
Fischer and Armstrong,> Pottevin,® Wroblewski,” Acree and
Hinkins,® and Taylor.® For the fats the relations seem clear.
Esters of both glycerine and the monatomic alcohols and the

1 Hill: Journ. of the Chem. Soc., \xxiii, p. 634; Ber. d. deutsch. chem.
Gesellsch., xxxiv, p. 1380; Emmerling: [bid., xxxiv, pp. 600, 2206.

2 Kastle and Loevenhart: Amer. Chem. Journ., xxiv, p. 491.

3 Bernizone: Atti de Soc. lig. di Scien. nat. e Geograf. Genoa, xi, p. 327.

4Emmerling: Ber. d. deutsch. chem. Gesellsch., xxxiv, p. 3810.

° Fischer and Armstrong: Ibid., xxxv, p. 3144.

8 Pottevin: Compt. rend. de l’Acad. d. sci., CXxXXVi, Pp. 1152; CXXXViii,
P. 378.

7 Wroblewski: Journ. f. prakt. Chem., N. F., xiv, p. 41.

8 Acree and Hinkins: Amer. Chem. Journ., xxviii, p. 370.

® Taylor: Univ. of Cal. Publications, Pathology, i, p. 33.

87

88 Synthesis of Protein by Trypsin

acids of the series C,,H,,,0, (and of oleic acid also) may be formed
under appropriate conditions through the action of lipase. With
the carbohydrates the relations are less simple. All the success-
ful reversions definitely known comprise the formation of disac-
charides or glucosides. Cremer! has described the formation
of glycogen from d-glucose, and Roux, Maquenne, Fernbach
and Wolff? have adduced data tending to show that starch may
be formed from glucose through the action of amylase; in both
instances, however, the proper chemical confirmatory evidence
has not been supplied, though the general evidence tends to sup-
port the inferences of reversion. For the proteins no reversions
have been reported. Three years ago I* published the results
of a series of negative attempts to form protamin from amido-
acids under the action of trypsin. Later I attempted to form the
synthetic peptides of Fischer from the appropriate amido-acids
under the influence of trypsin, again to no result. Recently
Abderhalden and Rona‘ have published the details of a series
of similarly negative results. I am now able to report a success-
ful synthesis of protamin from amido-acids through the action
of trypsin.

I have again employed protamin, despite the previous negative
results, because it was clear that if the proper conditions could
be secured, the use of such a simple protein presented a better
prospect of a positive result than could be expected with the
use of a higher and more complexly formed protein. An inves-
tigation of the relations of equilibrium in the hydrolysis of pro-
tamin suggested that the maximum of concentration of the
amido-acids should be employed. I had in most of the earlier
experiments used the amido-acids in the form of the sulphates;
that is, the protamin sulphate was completely digested, and then
the attempt was made to regenerate protamin sulphate. It was
clear from the theory of the reaction that the strong combina-
tion of the amido-acids with the sulphuric acid was unfavorable

> 1Cremer: Ber. d. deutsch. chem. Gesellsch., xxxii, p. 2062.

? Roux, Maquenne, Fernbach and Wolff: Compt. rend. de l’ Acad. d. sci.,
CXXXVil, p. 718; CXxXVill, pp. 819, 849; Cxxxix, p. 1217; cxl, pp. 95, 1303,
1403, 1547, 1067.

3 Taylor: Univ. of Cal. Publications, Pathology, i, p. 65.

* Aberhalden und Rona: Zettschr. f. physiol. Chem., xlix, p. 31.

Alonzo Engiebert Taylor 89

to the reaction of synthesis. In the present experiment, there-
fore, the amido-acids were used in the free state or as carbonates.
When a solution of protamin sulphate is digested with trypsin
the cleavage is usually complete. If the mass of ferment be
small, a trace of the substrate will remain. It is possible how-
ever, to secure the apparent station of equilibrium that is to be
observed in fermentations in general, in the following manner.
A saturated solution of protamin sulphate is digested with tryp-
sin, and from time to time fresh portions of the substrate in
powder form are added. A time will come when an undigested
fraction of the substrate remains in the system, and cannot be
digested, even by the addition of fresh portions of ferment. If
now the system be diluted, the digestion will recommence. The
station of equilibrium is obviously determined by the concentra-
tion of the products of the reaction. The products of the hydro-
lysis of protamin, the amido-acids, are very soluble in water; the
substrate is not very soluble in water. Under these circum-
stances, it is clear that unless some relation intervenes to shift
the station of equilibrium or the protamin is removed from the
system, the synthesis of a large amount of the protamin cannot
be anticipated, even though the concentration of the amido-acids
be high.

Under our present conditions of experimentation at least, all
ferment reversions are reactions of slow velocity. Trypsins of
higher origin are, however, labile substances, and in solution
rapidly undergo hydrolysis which destroys the enzymic property.
Under these circumstances I cast about for a more stable trypsin.
From the liver of the star-fish I obtained a trypsin that was very
active, but not particularly resistant to hydrolysis. From the
liver of a large Pacific coast clam (Schizotherus Nuttallit) I have
obtained a trypsin that is very resistant to hydrolysis. This
ferment, in the form of a glycerine extract of the tissue, I have
used in the experiment to be described.

The protamin sulphate used in these experiments was derived
from the Roccus lineatus. It has, as I have previously deter-
mined, the composition C,,H,,N,,O,.2 H,SO,. Four hundred
grams of the sulphate were dissolved in about 15 liters of warm
water, the solution made alkaline, and digested with trypsin until
the solution was miscible with five volumes of acidulated alcohol,

90 Synthesis of Protein by Trypsin

and bore saturation with sodium chloride without the production
of any opacity or precipitation.'. It was clear from these tests
that the digestion was completed, and that the fluid then rep-
resented the solution of the amido-acids that are the products of
the hydrolysis of protamin. The solution was then heated to the .
boiling point, the sulphuric acid precipitated by the addition of
barium chloride, the slight excess of barium precipitated by
carbon dioxide, filtered hot, and the filtrations repeated until the
filtrate was clear. The filtrate remained clear after cooling, and
gave an alkaline reaction. The fluid was then concentrated on
the water-bath until a sample on cooling showed beginning pre-
cipitation. This cooled fluid then represented a concentrated
solution of free amido-acids and their carbonates, the products
of the cleavage of protamin. Some 700 cc. were preserved as
a control, and 4 liters were employed for the reversion experi-
ment. To this were added 300 cc. of the glycerine extract of
the clam livers; this mixture could be added to 4 parts of acidu-
lated alcohol without the production of any opacity. Twenty
grams of toluol were then added, the bottle sealed and set aside
at room temperature. As time passed the contents of the bottle
became opalescent, then cloudy and finally a small white pre-
cipitate settled upon the bottom of the bottle, the supernatant
fluid remaining cloudy. Five months after the experiment was
begun, the container was opened, and the contents tested bac-
teriologically, with negative results. To the contents of the
bottle sulphuric acid was then added until the reaction was acid,
whereby the precipitate and cloudiness disappeared. The solu-
tion was then heated to the boiling point, filtered, and the filtrate
poured into four volumes of acidulated absolute alcohol, with
the production of a white precipitate. This precipitate was
collected upon a filter paper, washed with alcohol, dissolved in
water, filtered, again precipitated with alcohol and this pro-
cedure repeated four times. The final precipitate was dried
with alcoholand ether, and dehydrated in the desiccator at go° C.

1 The test with alcohol was made as follows: 200 cc. of the solution were
acidulated with sulphuric acid and then mixed with one liter of absolute
alcohol. The solution remained clear. After several days, the solution
still remaining clear, the alcohol was removed by heating, and the remain-
ing solution returned to the original liquor.

Alonzo Englebert Tayler gI

It weighed 1.8 gram. Possibly one-fourth as much had been lost
in the operation of isolation and purification. This powder is
soluble in about thirty-five parts of water, is precipitated by
three parts of alcohol, and is salted out of solution by a sodium
chloride concentration of 10 percent. It is digestible by tryp-
sin, resistant to pepsin. The elementary analysis gave the fol-
lowing figures:

0.2752 gm. of substance yielded 0.3802 gm. CO, and 0.1708 gm. H,O.
rr i a 0.0556 “ nitrogen.

2578“ * e ‘t 0.0593 ‘ Ss

0.087 oe oe oe oe 0.0219 oe oe

O. 1592 ee oe ae se 0.0399 oe oe

0.4358 ‘“‘ “ ia Es 0.2146 ‘“* BaSQO,.

The theoretical values and the analytical results are as fol-
lows:
Calculated for

CzoH 60N1706-2H2SOu: Found:

C =387.85 per cent 37.68 per cent

= 6.72 5°“ G39" 9"

N=25.13_ +o 24.45
24.93 | Average
25.17 (| 24.9 per cent
25.06 J

lEls SO n= PX), (6{0) “2 20.69 per cent.

The values for the analyses are in good agreement with those
demanded by the formula for protamin sulphate and there can
be no doubt that the substance is protamin sulphate.

The portion of the solution of amido-acids that had been pre-
served for a control was found to be sterile and had undergone
no change.’ Theglycerine extract, a portion of which was ‘used
in the experiment, was found to be active after the completion
of the experiment.

The successful result of the experiment when contrasted with
the negative results in the previous experiments must be due to
one of two factors, or to both; to the use of the concentrated
solution of the free amido-acids and their carbonates; and to the
use of a stable trypsin. The velocity of the synthesis is very
slow. From about 400 grams of the amido-acids some 2 grams
of protamin were synthesised in five months. All the reported

1 The test was done as before described: 200 cc. of the solution were
acidulated with sulphuric acid and mixed with one liter of absolute alcohol;
no opacity or precipitation was produced.

92 Synthesis of Protein by Trypsin

syntheses have been slow. When the reactions of the reversions
are better understood, this will be an important question to be
investigated according to the mathematical principles of physical
chemistry; it has, however, no bearing upon the theory or fact
of the reversion.

The meaning of the experiment is clear. Like all proteins,
protamin is composed of amido-acids. The digestion of pro-
tamin is an hydrolysis; the synthesis of protamin from the
amido-acids is a condensation. The reaction runs:

Protein + Water <; Amido-acids.

Under appropriate conditions trypsin, in accordance with the
theory, will accelerate the reaction in either direction, in the
direction of synthesis as well as in the direction of hydrolysis.
From the point of view of general theory, the formation of the
blood protein from the products of the digestion of protein is to
be interpreted as a reversion. This experiment furnishes a con-
firmation of this theory.

The general medical and biological public is not as yet fully
conversant with the theory of the reversed reaction, or with the
theory of the laws of mass action and equilibrium, as has been
well illustrated by the reception that has been accorded in the
medical world to the quantitative measurements and calcula-
tions of Arrhenius in the question of the toxins and antitoxins,
as well as by the attitude to the theory of catalysis displayed by
all the text books on fermentation. It is the conviction of the
writer that this, for the matter of the reversed reactions here
under consideration, is due to the fact that the energy relations
concerned in such a system are not understood. The reaction
is progressing to a station of equilibrium, and there is always a
certain zone of reaction to either side of the station of equilib-
rium. For some systems, as in the reaction

Ester + Water S Alcoho! + Fatty acid,

the reaction on either side of the station of equilibrium is easily
demonstrable; in many others, however, the station of equilib-
rium is so near to a completed reaction that special conditions
of experimentation are required to demonstrate that the reaction
is limited and reversible. In every such reaction there is a driv-

Alonzo Englebert Taylor 93

ing force and an internal chemical resistance. When now a cat-
alysor or a ferment accelerates such a reaction, it does so solely
through a reduction in the internal chemical resistance; the driv-
ing force remains the same. ‘Obviously, therefore, a catalysor
or ferment can accelerate a reversion or synthesis to the station
of equilibrium even though that reaction be an endothermic one.
In whatever direction the reaction is progressing under the con-
ditions of the experiment, whether it be building up or breaking
down, the ferment must accelerate the reaction. The term syn-
thesis through ferment action is, in the direct sense, a misnomer.
The ferment of course simply accelerates the reaction of syn-
thesis. The synthesis is a reaction per se; the ferment simply so
accelerates the velocity as to make the result apparent within
experimental time.

(Postscript made at the time of proofreading, April 24, 1907.)

When in January I realized that a positive result had probably
been attained, another control experiment was arranged to ex-
clude the possible objection that some other substance than
trypsin in the glycerin extract might have been the cause of the
reversion. From the theoretical point of view such an objection
was most improbable; nevertheless, in view of the criticisms to
which reversion experiments are subjected, the direct experi-
ment wascarried out. Four hundred cubic centimeters of the orig-
inal solution of concentrated products of the protamin digestion
were mixed with some 40 cc. of the glycerin extract that had been
previously heated to the boiling point for ten minutes to destroy
the activity of the ferment. Toluol was then added and the
mixture set aside. After more than three months this test was
analyzed. The mixture was sterile. Two hundred cubic centi-
meters of the mixture were acidulated with sulphuric acid and
mixed with one liter of absolute alcohol; the liquid remained
clear, no opacity whatever resulted. This experiment was iden-
tical with the positive experiment except in that the ferment
had beeninactivated byheating. The experiment and the several
controls may be tabulated as follows:

Original digestion-solution + alcohol =no opacity or precipitation.

Concentrated digestion-solution-++ferment extract +alcohol=no opac-
ity or precipitation.

94 Synthesis of Protein by Trypsin

Concentrated digestion-solution after 5, respt. 8 months + alcohol=no
opacity or precipitation.

Concentrated digestion-solution+ferment extract, after five months,
+alcohol=precipitate, protamin.

Concentrated digestion-solution + boiled ferment extract, after three
months, + alcohol =no opacity or precipitation.

These controls indicate that the reversion must have been asso-
ciated with some thermolabile substance in the ferment extract.
There is every reason to attribute thisresult to the trypsin which
the extract was knownto contain. Thecontrol experiments indi-
cate also that the protamin recovered must have originated in the
particular experiment, and cannot have been an undigested resi-
due, since under such circumstances it would have been recover-
able from the controls, whereas they were all negative to the
method with which protamin was isolated from the test with the
active ferment.

NOTE ON THE SYNTHESIS OF A PROTEIN THROUGH THE
ACTION OF PEPSIN.

By T. BRAILSFORD ROBERTSON.

(From the Rudolph Spreckels Physiological Laboratory of the University of
California.)

(Received for publication, April 8, 1907.)

In arecent paper Taylor' has described the synthesis of a pro-
tein (protamin) through the action of a trypsin obtained from
the liver of the soft-shelled California clam. The concentrated
products of the tryptic digestion of 400 grams of protamin sul-
phate were converted into the carbonates and subjected to the
action of the trypsin; at the end of five months about 2 grams of
protamin (weighed as sulphate) were obtained from the solution.
By pursuing a method in some respects similar to the above I
have succeded in synthesising one of the first products of the pep-
tic digestion of casein, a substance, or mixture of substances, to
which the name “ Paranuclein”’ has been applied.’

Four hundred cc. of 4; potassium hydroxide were “‘saturated”’
with casein’ and the solution subjected to the action of pepsin for
several days at 40°C. It was then heated to 100°C. for about 10
or 15 minutes to destoy the ferment and filtered while hot. The
solution was then evaporated on a water-bath to a volume of
about 7occ. This concentrated solution of the products of the
peptic digestion of casein is a clear brown syrup which is strongly
acid and gives no precipitate or opalescence upon the addition of
acetic acid or upon the addition of acetic acid in excess after hav-
ing previously rendered the solution alkaline by the addition of
potassium hydroxide. Both casein and paranuclein are therefore
absent from the solution. To 7o cc. of this solution were added

* Univ. of Cal. Publications, pathology, i, p. 343, 1907. This Journal, iii,
P- 87, 1907.

For literature, consult Gustav Mann, Chemistry of the Proteids, 1906,
PP. 395-396.

’T. Brailsford Robertson: This Journal, ii, p. 337, 1907.

95

96 Synthesis of a Protein through Action of Pepsin

30 cc. of a concentrated (approximately 10 per cent) solution
of Grubler’s pepsin which had previously been filtered through
rapid filtering paper. The mixture of the two solutions was a
clear brown syrupy fluid which gave no trace of a precipitate with
acetic acid either before or after neutralization with alkali. The
mixture was set aside at 40°C. in the presence of excess of toluene
to prevent bacterial infection. Within two hours a thick white
precipitate had formed in the solution; after 48 hours the solution
was filtered and the precipitate thus collected on the filter was
dissolved in a minimal amount of sodium hydroxide and the
filter so arranged that the alkaline solution dropped directly into
water acidified with acetic acid. The precipitate thus obtained
was reprecipitated twice and was washed by decantation several
times. Finally it was collected on a filter, washed with alcohol and
with ether in a desiccator and dried at 60° C. In this way 1.02
grams of a friable, grayish-white, hygroscopic powder were ob-
tained; I do not think that more than 10 per cent was lost in the
process of purification. Theoriginal filtrate was tested for casein
by rendering alkaline with sodium hydroxide and then adding
excess of acetic acid—a very slight precipitate was formed, too
small in amount to collect and purify.

Paranuclein was prepared by partly digesting sodium caseinate
with pepsin, filtering off the precipitate, dissolving in sodium
hydroxide and precipitating with acetic acid. The paranuclein
was reprecipitated twice, washed with alcohol and ether in a
desiccator and dried at 60°. A powder exactly similar in appear-
ance to the synthesised substance was thus obtained. The
physical properties of the precipitates with acid, bulky, flocculent
and settling rapidly, were identical in both substances. °

The synthesised substance and the paranuclein were both
analyzed for phosphorus by Neumann’s method! with the follow-
ing results:

0.1014 gm. of substance yielded 0.001665 gm. P, O,
0.1010 “ re 0.001598 “ P, O.

H P,O, = (1) 1.64 |
ence a spe he a ) Ay erage—1.61 per cent.

0.1027 gm. of paranuclein es 0.004312 gm. P,O,
0. 0988 = 0. 004099 i PE. c

Mecuae Arch. f. Wgek und Ppajeiol., Pp. 159, 1900.

T. Brailsford Robertson 97

wines: 3 e 16 ap ane } Average—4.175 per cent.

Previous observers agree in ae that the percentage compo-
sition of paranuclein varies very greatly with the circumstances
under which it is prepared,' the percentages of phosphorus which
have been found in various preparations varying from 0.88 to
6.86.2 This fact leads us to suspect that the substance which has
been termed paranuclein is, in reality, a mixture of at least two
substances and the hypothesis which has suggested itself to me is
that during the hydrolysis of casein by pepsin an insoluble sub-
stance of high phosphorus content is first formed, that this sub-
stance splits off a soluble phosphorus-containing moiety leaving
another substance, insoluble in acid, of lower phosphorus content,
and that this second substance is in its turn attacked and splits
up into soluble substances. That this explanation is probably
the correct one, although, of course, several such steps may be
involved, is shown by the following experiment.

One gram of the paranuclein containing 4.175 per cent of POs
was dissolved in 400 cc. of 0.045 N Ca(OH)2and stood at 40° C. for
12 hours; acetic acid was then added in excess and the precipi-
tate washed and dried in the manner described above. This sub-
stance, which we may designate, Paranuclein A, on analysis for
P.O;, gave results as follows:

0.1071 gm. yielded 0.00174 gm. P,O,
0.0957 * rg O200130 Bs O.
Hence P,O,= S ae percent } Average—1.51 per cent.

Only a little over 0.2 gm. of this substance was obtained from
the gram of paranuclein originally dissolved in the lime-water.

This splitting off of phosphorus from paranuclein, particularly
in alkaline solution, has been commented on by other observers.?

The substance which is obtained synthetically, as described
above, resembles very closely Paranuclein A, both in physical
properties and in percentage of P.O;. The synthesised substance
is practically insoluble in acids, is readily soluble in dilute alkali,
precipitates protamin from a1 per cent solution of the sulphate

1See literature quoted in Gustav Mann, loc cit.

2W.v. Moraczewski: Zeitschr. 7. physiol. Chem., xx, p. 28, 1895.
3 Salkowski und Hahn: Arch. f. d. ges. Physiol., lix, p. 225, 1895.

98 Synthesis of a Protein through Action of Pepsin

at a reaction just alkaline to phenolphthalein (at which reaction
both the protamin and the paranuclein remain in stable solution
when not mixed), and a 2 percent solution in ;*,, sodium hydroxide
is precipitated by .“, ferric ammonium sulphate; these being all
well known properties of paranuclein.'

The synthesised substance also resembled my preparation of
paranuclein in the following properties: in approximately 2 per
cent solution in ;*, sodium hydroxide it gives the xanthoproteic,
Millon’s, Adamkiewicz and the biuret (violet) reactions; it is
precipitated by cupric chloride (1 vol. of * to 100) and by zinc
chloride but not by mercuric chloride (5 vols. of #5); it is pre-
cipitated by picric and by tannic acids, but the precipitate redis-
solves on rendering the solution alkaline; it is not precipitated by
the addition of five volumes of absolute alcohol; and the precipitate
at first produced by the addition of acetic acid is soluble in con-
siderable excess of glacial acetic acid.

li no pepsin be added to the concentrated solution of the peptic
digestion of casein, prepared as described above, the solution,
after keeping for three weeks at 40° C. remains perfectly clear
and homogeneous and gives no tests for paranuclein or for casein.
A to per cent, filtered solution of Griibler’s pepsin, on standing
for the same period at 40° also remains perfectly clear and
homogeneous. Yet these two solutions when mixed in the pro-
portion of 1 vol. of ferment to 5 vols. of the solution of the
casein products, gave a voluminous precipitate which remained
permanent throughout the period of three weeks.

The experiment has been repeated a number of times, always
resulting in the production in the solution of a precipitate resem-
bling in properties that which is described above. All the experi-
ments were carried out in the presence of an excess of toluene
(several drops to 50 cc.).

Further experiments are in progress.

CONCLUSIONS.

1. The substance, derived from casein by incomplete diges-
tion with pepsin, which has been termed paranuclein is probably

1 Salkowski and Hahn: loc. cit. Milroy: Zeitschr. f. physiol. Chem., xxii,
P- 307, 1896.

T. Brailsford Robertson 99

a mixture of at least two substances, the one containing a high
percentage of phosphorus and the other, in other respects similar
in properties to the former substance, containing a much smaller
percentage of phosphorus. Paranuclein containing 4.175 per
cent of P.O; digested with lime-water for twelve hours at 40° C.
yields a small quantity of a substance similar in properties to the
paranuclein but containing only 1.4 to 1.6 per cent of P2O;. This
I term, provisionally, Paranuclein A.

2. Byacting at 40° C.on an acid, concentrated solution of the
products of the peptic digestion of casein, containing no casein or
paranuclein, with a concentrated solution of pepsin, a substance is
precipitated which is identical in properties and in phosphorus
content with the above-mentioned Paranuclein A.

3. The concentrated solutions of the casein products and of
the pepsin, when kept separately at 40° C. give no precipitate,
remain clear and homogeneous for a period of over three weeks,
and at the end of that time yield no tests for paranuclein or for
casein.

THE DETECTION AND ESTIMATION OF REDUCING SUGARS.
By STANLEY R. BENEDICT.
(From the Sheffield Laboratory of Physiological Chemistry, Yale University.)
(Received for publication, March 23, 1907.)

THE DETECTION OF SUGARS.

Of the extremely large number of methods proposed for the
detection of reducing sugars there are very few which may be
regarded as specific for sugars alone. With one or two possible
exceptions, these tests indicate only the presence or absence of
reducing substances, and are inapplicable to the detection of
sugars when other reducing substances are present. The fact
that many of the sugars are powerful reducing agents in alkaline
solutions, while they exert, at most, slight action in neutral or
acid solutions, is very commonly recognized, and application
is made of this fact in almost every test which has been pro-
posed for the detection of these substances.

Gaud, Framm and others! have made studies of the effect of
alkali upon a number of the sugars. The substances formed
appear to be oxidation products, possibly preceded by a dehy-
dration and decomposition. The destructive action of alkalies
upon glucose is a common matter of reference throughout the
literature upon the estimation of sugars, and numerous sugges-
tions have been made to avoid this difficulty by means of the
substitution of ammonium hydroxide for potassium or sodium
hydroxide, and keeping the temperature somewhat below the
boiling point during the reaction, with other devices of a similar
nature.’

The following experiments were undertaken to throw light
upon the mechanism of the reducing action of the sugars in alka-

1See De Bruyn and Ekenstein: Rec. Trav. Chem., xvi, p. 274, 1897;
Framm: Arch.f.d. ges. Physiol., \xiv, p. 575, 1896, and Gaud: Compt. rend.
de l’ Acad. des sci., cxix, p. 650, 1894.

2 See Gaud: loc. cit.

Tol

102 Detection and Estimation of Sugars.

line solution as well as to test the relative destructive action of
certain of the alkalies upon various carbohydrates. A 1 per
cent solution of dextrose was boiled for about a minute with one-
half its volume of a ro per cent potassium hydroxide solution,
the resulting solution cooled to room temperature and added to
an equal volume of ordinary Fehling’s solution, also at room
temperature. No change occurred in the mixture in the cold,
and only a very faint reduction was obtained upon boiling.
Clearly the sugar had been practically completely oxidized, so
far at least as its power of affecting Fehling’s solution was con-
cerned. This result was rather what might have been expected,
inasmuch as we are here dealing with one of the most readily
oxidized of the carbohydrates.

In the next experiment a solution of lactose was substituted
for the dextrose. Upon adding an equal volume of Fehling’s
solution to the cooled solution a heavy reduction of the copper
occurred almost instantaneously 7” the cold.’

Five cubic centimeters of a 1 per cent solution of dextrose were
now heated to boiling with about 0.5 gram of sodium carbonate,
the resulting solution cooled, and mixed with an equal volume
of Fehling’s solution. An almost instantaneous reduction of the
copper compound occurred, as indicated by a heavy precipita-

1 The only references which I have been able to find relating to the for-
mation of a compound capable of reducing Fehling’s solution in the cold
as a result of warming sugars with alkali, are in Kiihne’s Lehrbuch der
phystologischen Chemie (p. 518, 1868), in the work of Worm Miiller and
Hagen (Arch. f.d. ges. Physiol., xxii, p. 391, 1880), and of Emmerling and
Loges (Arch. f. d. ges. Physiol., xxiv, p.184,1881). Kthnestates that dex-
trose solutions, heated to about 70° C. with sodium or potassium hydroxide
solution will yield a solution capable of reducing Fehling’s solution in the
cold. He suggests this fact as evidence that the sugars themselves do
not reduce directly and suggests this procedure asa test for dextrose in the
urine. Worm Miiller and Hagen discuss the bearing of this fact upon
Trommer’s test for dextrose in the urine. Emmerling and Loges come to
the conclusion that the reducing substance formed is acetone alcohol,
CH,CO-CH,OH. At the time my work was done the statement of
Kiihne on this subject was unknown to me, and inasmuch as neither he
nor Worm Miiller made comparisons of the action of the other alkalies and
other sugars, nor gave a full theoretical discussion or analytical appli-
cation of the fact which they stated, I have thought it desirable to give
a rather complete discussion of my results.

Stanley R. Benedict 103

tion of the yellow cuprous oxide. A solution of lactose substi-
tuted for the dextrose in the above experiment yielded a similar
result. In other words, a 1 per cent solution of glucose boiled
for a few moments with 1o per cent potassium hydroxide solu-
tion almost completely loses its reducing power. <A similar treat-
ment of a lactose solution yields a solution containing some sub-
stances capable of reducing Fehling’s solution in the cold. Boil-
ing either lactose or dextrose with sodium carbonate yields a
solution which reduces Fehling’s in the cold. It is particularly
to be noted here that the potassium hydroxide solution destroys
the reducing power of the dextrose, while the carbonate, under
similar conditions, does not. This fact will be referred to later.
The compound formed under these conditions is remarkably
strong in its reducing power. Ammoniacal silver solutions are
reduced instantaneously and completely in the cold. The prod-
uct formed either from dextrose or lactose is capable of reducing
Barfoed’s reagent upon boiling, even when the acidity is consider-
ably greater than that called for in Barfoed’s formula.

These results indicate that the sugars themselves are not
directly responsible for the strong reducing action of their alka-
line solutions. They also admit of some valuable analytical
applications. The first of such applications would be appar-
ently to afford a differentiation of lactose from dextrose, the
former substance in 1 per cent solution forming a compound upon
boiling with alkali which will reduce Fehling’s fluid in the cold,
while dextrose under similar conditions is destroyed. This appli-
cation can be made, and by means of such a method it is possible
to detect lactose in presence of dextrose, or to distinguish between
dextrose and lactose solutions. The procedure is as follows:
To about 5-10 cubic centimeters of the solution to be tested is
added one-half volume of 1o per cent potassium hydroxide solu-
tion and the resulting mixture boiled vigorously for about one
and one-half minutes. The solution is then cooled and added
to an equal volume of Fehling’s fluid. The immediate formation
of a heavy precipitate which is bright yellow and fills the entire
body of the solution indicates lactose. A slight precipitate
is to be disregarded. The solution used for this test should
contain between 0.5 and 1 percent of the sugar. It is not a par-
ticularly delicate reaction and will not give positive results where

104 Detection and Estimation of Sugars

the amount of lactose is under 25 per cent of the total quantity
of sugar present.

A much wider analytical application of these facts is possible.
Upon them we may base a reaction which is apparently specific
for sugars. Very few, if any other substances will, upon boiling
with alkalies, form substances capable of reducing Fehling’s solu-
tion in the cold. In just so far as this is true, this method of
detecting sugars is specific. The procedure is as follows: To
5-8 cubic centimeters of the fluid to be tested in a test tube, is
added about one-half gram of sodium carbonate. Heat the
solution to boiling and continue the boiling for about one-half
minute. Cool and add an equal volume of Fehling’s solution.
If sugar is present a reduction will occur more or less promptly,
one per cent or more giving a practically instantaneous reaction,
smaller amounts requiring from three to five minutes. As thus
carried out this method will detect sugar in solutions containing
0.1 per cent, or more, of sugar. Slightly warming the tube will
increase the delicacy to about 0.05 percent. These figures apply
to relatively pure solutions of the sugars.

While this method is not nearly so sensitive as the usual
Fehling’s method, it may become of material use in establishing
the presence or absence of sugars in mixtures containing other
reducing substances, and is delicate enough for many purposes.

Attention was called above to the fact that sodium hydroxide
solutions destroy the reducing properties of glucose much more
rapidly than does sodium carbonate solution.t Yet the alkalin-
ity given by sodium carbonate is amply sufficient for the glucose
to exhibit its full reducing action. Considering the bearing of
these facts upon sugar detection it will be evident that where
only very small amounts of dextrose are present, or where some

1 Tt is not to be inferred from this statement that a distinction is implied
between the action of the hydroxides and carbonates in any respect other
than may be accounted for by the difference in the degree of alkalinity
represented by the two substances. If we use very dilute alkali solutions,
stronger sugar solutions, or if the heating be more mild, it is perfectly pos-
sible, as Ktthne and Worm Miiller have shown, to obtain a solution which
will still exhibit reducing properties. If, however, we endeavor to use
hydroxide solutions sufficiently dilute to avoid the destruction of even
minute quantities of dextrose, a solution is obtained which is utterly
unfitted for ordinary qualitative or quantitative sugar work.

Stanley R. Benedict 105

substance is present which temporarily interferes with the reduc-
tion of the copper solution, a reagent in which the alkalinity is
secured by means of sodium carbonate, instead of hydroxide,
should give more satisfactory results. In order to test this hypoth-
esis the following work has been done.

Solutions of the following composition were prepared:

SOLUTION A.

CGuystallimedicopper sulfates. -<. shyt te a yobes wcvalerstenetet= ale 69.3 gm.

Wistille diya berrtOns! 5.1 sce ene ate keys epeveesiantce lens ser hm erin Opa 1000 ce.
So.uTion B.

EIRCRINOCM ENGEL tie fry tei teida ochre J cusuetess eco ta eceey ole 346 gm.

PATUVCLOUS SOCMIMICAEDONALE... 24 14 hie aise 42 3 Gt a> ete oe 200 “

WWistilled wAteEiOnee acy cai a ere ccs Me taies De nie ciel nts 1000 ce.

For use, these solutions are mixed in equal proportions, and
the resulting mixture diluted with an equal volume of distilled
water. When further diluted with an equal volume of distilled
water, heated to vigorous boiling and allowed to cool spontane-
ously, this solution should not show the slightest turbidity. The
reagent thus prepared was tested regarding its power of detecting
Sugars as compared with Fehling’s fluid, under the following
conditions. When distilled water solutions of dextrose were
used and the solution boiled as in the usual procedure, it was
found possible to obtain in most cases a perceptible reaction
with Fehling’s fluid' when the sugar present amounted to 0.001
per cent. The result here was in many instances uncertain and
marked the full limit of the test as I applied it. The method of
procedure was to add to 3 cubic centimeters of Fehling’s fluid
in a test tube an equal volume of the solution to be tested, the
resulting mixture being heated to vigorous boiling, which was
continued for about one-half minute. The solution was then
allowed to cool to room temperature, and unless precipitation
had already occurred, it was again heated to boiling and allowed to
cool, this process being sometimes repeated. Worm Miller and
Hagen’ state that Fehling’s solution will indicate the presence of
glucose in solutions containing 0.0008 per cent. I have been
absolutely unable to obtain any reaction with 0.0008 per cent
solutions of glucose with the procedure above described, while
solutions of o.oo1 per cent often gave negative results.

' The Fehling’s fluid used was made up according to Soxhlet’s formula.
? Arch. f. d. ges. Physiol., xxii, p. 383, 1880.

106 Detection and Estimation of Sugars

With the reagent in which the carbonate is substituted for
the hydroxide, according to the formula given above, and follow-
ing an exactly similar method of procedure to that described
for Fehling’s solution, I have obtained absolutely positive results
with solutions containing 0.00005 per cent of glucose. A dis-
tinct yellowish precipitate is obtained with glucose solutions of
this dilution, especially upon cooling. Further dilutions were
not tested. It thus appears that the substitution of carbonate
for hydroxide yields a solution many times as delicate for the
detection of sugars as we have in Fehling’s fluid. A compari-
son made with any dilute solution—for instance, about 0.01 per
cent solutions of dextrose—will speedily convince anyone of the
greater merits of the carbonate reagent.

The urine, containing as it does substances which either par-
tially reduce or else inhibit the reduction of Fehling’s fluid,
offers peculiar difficulties in the detection of small amounts of
sugar. It is, therefore, of special interest to determine whether
the carbonate solution possesses any advantages over Fehling’s
fluid for the detection of dextrose in the urine. Fehling’s
solution will usually indicate dextrose in the urine when present
up to, or in excess of, 0.1 per cent. Smaller amounts give abso-
lutely no indication of their presence, save possibly by confusing
changes in the color of the solution. When the amount is as
small as 0.1 per cent, the results are often very uncertain. With
the reagent in which the carbonate is substituted for the hydrox-
ide, it is perfectly possible to detect as small amounts of dextrose
in the urine as from 0.015 to 0.02 per cent. Even these small
quantities yield results which are almost invariably far more
positive than can be obtained with Fehling’s fluid in the presence
of ten times the amount of sugar.

The procedure for the detection of sugar in the urine is as
follows. Solutions A and B, made according to formule given
above, are mixed in equal proportions. The resulting solution
is then diluted by the addition of three times its volume of dis-
tilled water. To about 6 cc. of this reagent in a test tube are
added from 7 to 9 (not more) drops of the suspected urine. The
mixture is heated to vigorous boiling for about one-fourth to one-
half minute, and allowed to cool spontaneously to room tempera-
ture. This process may be repeated if desired, though it is usu-

an ee a ee ee

Stanley R. Benedict 107

ally unnecessary. In the presence of sugar a precipitate will
form which is often greenish or bluish green to begin with (in
case the amount of sugar present is small), and usually becomes
yellowish upon standing. This precipitate generally forms at
or below the boiling temperature if the sugar present exceeds
0.06 per cent; with smaller amounts it forms slowly, usually only
upon cooling. With larger amounts of sugar the reaction is
obtained very readily upon reaching the boiling temperature.
The precipitate is then generally reddish or yellowish in color.
The results obtained in this test, even with the smaller amounts
of sugar are extraordinarily definite, and according to my expe-
rience leave no room for uncertaininterpretation. Normal urine
will not, under these conditions, produce even the slightest tur-
bidity in the reagent. Professor Mendel, of this laboratory,
has suggested that the greenish precipitate obtained with urines
containing small amounts of sugar, may be a compound of copper
with the sugar, rather than a reduction product. While this
may be the case, it seems more probable to the writer that the
precipitate represents, if not a simple cuprous oxide or hydrox-
ide, a compound of some constituent of the urine with reduced
copper oxide. This appears the more likely, because, in the
absence of the urine, sugar solutions of equal dilution give a
definite reduction product of either the red oxide or the yellowish
hydroxide. ‘This latter compound, it may be remarked, is more
usually obtained as a product of the reduction of the carbonate-
copper solution, than is the case with Fehling’s fluid, particularly
when only small amounts of sugar are present, owing probably
to the less strongly dehydrating action of the carbonate solution.

Whatever may be the nature of the precipitate obtained as a
result of boiling the carbonate copper solution with sugar-con-
taining urines, it appears to be a most delicate and satisfactory
method for the detection of dextrose in this fluid.

It is interesting to note in this connection the modification of
Fehling’s test for sugar in the urine proposed by Worm Miller,!
in which he suggests the use of solutions of sodium hydroxide
more dilute than in ordinary Fehling’s fluid, small amounts of
copper solution, and the continued heating of the solution well

1 Worm Miiller: Arch. f. d. ges. Physiol., xxvii, p. 107, 18823

108 Detection and Estimation of Sugars

below the boiling point. This modification is capable of detect-
ing about 0.03 per cent of dextrose in the urine. Worm Muller
himself suggested that the superior delicacy of this test over the
ordinary Fehling’s method is due to the fact that the sugar
reduces at a lower temperature than the reducing substances
normally contained in the urine. In view of the facts brought
out in the earlier portions of this paper it would be inferred that
there are other factors here as well as the one he suggested. Thus
it seems probable that ordinarily the interfering substances in
the urine (chief among which is apparently creatinin, as pointed
out by MacLean') will inhibit the reducing action of the sugar
long enough for the strong alkali to completely destroy the small
amount of carbohydrate present. Lowertemperature and weaker
concentration of alkali would tend to prevent this destruction,
so that eventually the glucose would have its normal reducing
action. Similar results are apparently obtained where carbon-
ate is substituted for the hydroxide. The procedure is not
troublesome as in Worm Miiller’s test, and the results are more
delicate and conclusive.

Thus we see that the carbonate solution has the following
advantages over Fehling’s solution. It is many times more
delicate for the detection of sugarsin pure solutions. This is just
what we should expect from theoretical considerations. Feh-
ling’s fluid, containing as it does, a substance which is strongly
destructive to glucose, should not be used so long as we can sub-
stitute something which is more effective and has not this destruc-
tive action. The carbonate solution yields more definite results
than does Fehling’s fluid, since there are fewer substances which
interfere with its action than is the case with the other fluid.
This can be shown by comparative tests with the urine, and fur-
ther it can be most beautifully exemplified by comparisons of
the reducing action of chloroform upon the two solutions. A
solution of chloroform in water will reduce Fehling’s solution
most copiously, even slightly below the boiling point, whereas
its action is very slight upon the carbonate solution and only
occurs upon prolonged boiling. There are many other substances
which reduce Fehling’s solution more readily than the carbonate

1 MacLean: Bio-chem. Journ., i, p. 111, 1906.

t~ 7

Stanley R. Benedict 109

solution, and thus interfere with the use of the former solution
as a test for sugars to a greater degree than is the case with the
carbonate solution; but chloroform is the most striking example.
Regarding the stability of the carbonate solution the following
should be stated. For delicate work in sugar detection, either
in pure solutions or in the urine, the solutions making up the
reagent should be freshly mixed and diluted. If it is desired to
keep the mixed reagent on hand the two solutions should be
mixed in equal proportions and to every liter of the undiluted
mixture should be added from five to ten grams of sodium hydrox-
ide. This small amount of alkali serves to prevent decomposition
and does not affect the delicacy of the reagent to any great extent.
Such a mixture will remain for weeks more delicate as a reagent
than Fehling’s solution and is not nearly so caustic. It should
be remembered, however, that the best results are only above
obtained with freshly mixed solutions diluted according to the
directions.

A few copper solutions containing carbonates have been pro-
posed in the literature. The best known of these is Soldaini’s
solution,’ which contains 3.464 grams of crystallized copper sul-
fate and 297 grams of potassium bicarbonate dissolved in a
liter of water. Ost? later modified this solution by substituting
for a portion of the bicarbonate, a nearly equal weight of the
normal carbonate, as well as increasing the amount of copper sul-
fate present. These solutions are open to serious objections to
which the solution suggested in this paper is not. Soldaini’s
solution is more difficult to prepare,a portion at least of the cop-
per always being left undissolved as the carbonate, which must
be filtered off. Furthermore the boiling with the test solution
must be continued for some time, often continuously for ten or
fifteen minutes before smaller amounts of sugars indicate their
presence by even the slightest apparent reduction. Thereason
for this fact appears to be that bicarbonate is capable of holding
considerable amounts of cuprous oxide in solution, or else of pre-
venting the reduction of cupric compounds by dextrose. This is
readily proved by the following experiment. If to a portion of

1 Soldaini: Chem. Centralbl., p. 389, 1889.
? Ost: Ber. d. deutsch. chem. Gesellsch., xxiii, 1, p. 1035, t89go.

110 Detection and Estimation of Sugars

the carbonate copper solution described earlier in this paper a
considerable amount of bicarbonate of sodium is added, it will
be found that the resulting solution shows no reduction even
with large amounts of sugar except upon very long continued
boiling, while a similar solution to which the bicarbonate has
not been added gives a heavy precipitate even before the boiling
point isreached. Thus it will be seen that the bicarbonate solu-
tions are open to marked objection in the fact that they require
very prolonged boiling to obtain any result, and are, therefore,
scarcely applicable to qualitative work. Results are obtained
with the solution above recommended almost instantaneously,
unless the amount of sugar present is very small, in which case
it may require from four to five minutes.

A METHOD FOR THE VOLUMETRIC ESTIMATION OF SUGARS.

Owing to the depth of color of the precipitate obtained upon
the reduction of Fehling’s solution by means of sugars, the
application of Fehling’s process to the volumetric determina-
tion of sugar solutions is very difficult Even with long prac-
tice it is almost impossible to obtain satisfactory results by the
use of this method. Various modifications of the original Feh-
ling’s fluid have been proposed with a view to render it more
applicable to volumetric work. Most of these are based upon
an attempt to keep the cuprous oxide formed in solution. In
Pavy’s solution ammonia is added for this purpose. The objec-
tions to the latter method are the great ease with which the dis-
solved cuprous compound undergoes oxidation—hence the abso-
lute necessity of entirely excluding air if accurate results are to
be obtained—and furthermore the rapidity with which the ammo-
nia boils out of the solution. In Gerrard’s method potassium
cyanide is used as the solvent for the reduced copper compound.
As ordinarily performed this method requires two titrations and
must be carried out quite rapidly in order to avoid reoxidation
of the reduced copper. The potassium cyanide solution used is
unpleasant to work with and is quite unstable.

To the writer it seemed that a more satisfactory method of sol-
ving the difficulties offered by Fehling’s solution in sugar titra-
tion would be to obtain some modification of the solution, which
should yield upon reduction not the red cuprous oxide but some

Stanley R. Benedict III

colorless insoluble compound. In this case we should have the
reduced copper compound in such a form as would prevent its
reoxidation, while it would not obscure the end point of the
reaction. i

The well known insolubility of cuprous compounds of the hal-
ogens led to the attempt to obtain the precipitation of the reduced
copper as one of the haloid salts, through the addition of con-
siderable quantities of chlorides, bromides, or iodides to Feh-
ling’s solution. No satisfactory results were obtained with these
substances. Closely related in chemical behavior to the halo-
gens are the simple and complex cyanides, and the attempt was
next made to see whether any of these substances would yield the
desired results. Addition of potassium ferrocyanide to Fehling’s
solution causes, upon reduction, the precipitation of a white
compound, which, however, upon continued boiling becomes
dark in color. This substance was, therefore, unsatisfactory
except in a certain combination referred to later. If potassium
sulfocyanide be added, in small amounts, to Fehling’s fluid it
produces no appreciable change in the reduction product. If,
however, it be added in considerable excess, the cuprous oxide
formed will be held in solution. After the writer had found that
carbonate copper solutions are much less destructive in their
action upon dextrose than is the corresponding hydroxide solu-
tion, as has been shown earlier in this paper, it was only natural
to try the above suggested compounds in connection with this
solution. It was speedily found that in this case potassium sul-
focyanide yielded a very different result from that obtained with
Fehling’s solution. After addition of potassium sulfocyanide to
the carbonate solution it was found that upon reduction a chalk-
white compound of cuprous sulfocyanide was produced, instead
of the red oxide. This suggested at once the employment of
this solution for volumetric purposes.

This fact was discovered by the writer early in November,
1906. A somewhat hasty review of the literature at that time,
preliminary to working up the method, failed to reveal the fact
that potassium sulfocyanide had ever been previously applied
to sugar estimation, or even that the fact above stated had been
mentioned in the literature. About January 1, 1907, a paper

1p Detection and Estimation of Sugars

appeared by Bang,’ describing a volumetric method for sugar
estimation, depending upon the use of potassium sulfocyanide.
In this paper Bang referred to the fact that he had pointed out
in an earlier communication? that addition of potassium sulfocy-
anide to alkaline copper solution which contained no hydroxide,
caused the precipitation of white cuprous sulfocyanide instead
of the red suboxide. He had also suggested in this earlier paper
a method for the estimation of dextrose, based upon this fact.
Bang’s earlier paper had completely escaped my attention in the
review of the literature on this subject, owing very probably to
the fact that its title was in nowise concerned with sugar analy-
sis, being ‘‘Ueber die Verwendung der Zentrifuge in der Quan-
titativen Analyse.”’ The first method Bang proposed was based
upon anattempt to estimate the excess of potassium sulfocy-
anide remaining in the solution after the reduction of the copper
had taken place. This method appears to the writer obviously
unsatisfactory. The cuprous sulfocyanide formed is not com-
pletely insoluble in even a small excess of potassium sulfocyanide,
and unless some excess of this salt is present a portion of the cop-
per is precipitated as the suboxide. It would thus appear that
determination of the residual potassium sulfocyanide would
not yield very satisfactory results. Bang recognized the unsat-
isfactory nature of his earlier method and in his second paper
substitutes one which he considers much better. His method
in this case, briefly outlined, is as follows: The copper solution
employed is a modified Soldaini’s solution, to every liter of which
is added two hundred grams of potassium sulfocyanide. Toa
measured volume of this solution is added a measured volume
of the sugar solution to be determined,the mixture then being
boiled for three minutes. (No precipitation takes place owing
to the large excess of sulfocyanide present.) Then the solution
is cooled and the excess of cupric copper determined by titration
with standard hydroxylamine solution to a colorless solution.
The results obtained appear to be very satisfactory. There are
at least two objections to Bang’s method. The boiling copper
solution cannot be titrated with the sugar solution directly to

1 Biochem. Zeitschr., ii, p. 271, 1906 (December).

2 Festschrift fir Hammarsten, 1906 (Upsala Lakareférenings Forhand-
lingar. Ny Foljd, Elfte Bandet. Supplement).

Stanley R. Benedict LE

a colorless solution because of the employment of the bicarbon-
ate solution of Soldainiand Ost. This solution, as is mentioned
earlier in this paper, requires continued boiling before complete
reduction takes place, which renders impracticable any attempt
to titrate directly to an end point. The second objection to
Bang’s method is the employment of hydroxylamine for his final
titration, a substance which even in form of its salts, is unstable,
and not commonly available.

In the use of the writer’s method for the volumetric estimation
of sugar three solutions are required, which are made up accord-
ing to the following formule:

SOLUTION A.

Prystallezed copper sulfate... 2s... .i)d an. cee cieee 69.30 gms.

rice, Warletyne 2's «.20t52 2202) STS Sw sand RR ee 1000 ce.
SOLUTION B.

evsbatlimedwROchelle Salt | 2... 620% «14 node ae ee 346 gms.

Paureannydroussodium carbonate . 2. i. osh aoe eee 200) ss

Mita CW UBT COs 53.5 cpaye coon -s! x ove-02 2 5) a gehen le So 1000 ce.
SOLUTION C.

Retuacsiin SUTOCYAMIUS «22.25 = =< 2:1. = ane emo 200 gms.

esbiMed WaALEr tO, cians 2 ss. < 5 s/h la's x so pee Se eee 1000 cc.

For use these solutions are mixed in equal proportions in the
order indicated. To every 30 cubic centimeters of the solution
thus obtained are added from 2.5 to 5 grams of pure anhydrous
sodium carbonate.' The amount of this substance added should
roughly correspond to the dilution to which the solution will be

1 Regarding the addition of sodium carbonate in this connection the
following should be stated. The alkalinity secured by the addition of the
indicated portion of Solution B, is not great enough to give the most satis-
factory end point. This lack of alkalinity might be overcome either by
the addition of the solid carbonate as above suggested, or through the
substitution of a greater amount of potassium carbonate in the formula of
Solution B. (Solution B is practically saturated with sodium carbonate,
which is, it will be remembered, much less soluble than the correspond-
ing potassium compound.) I prefer the method above detailed for secur-
ing the increased alkalinity. Sodium carbonate is much commoner than
the potassium salt, keeps anhydrous better and is usually more available.
The addition of a small portion of the dry salt, which may be roughly
measured, is-certainly no great trouble. Furthermore, the formula of
Solution B,it will be noticed, coincides with the one found most satis-
factory for qualitative work, as detailed earlier in this paper, so that such
a solution is readily available for either qualitative or quantitative work.

114 Detection and Estimation of Sugars

subjected during the titration, 7. e., for titrating dilute sugar
solutions add greater quantity of carbonate and vice versa.
The solutions are mixed in a beaker of suitable capacity, the
requisite quantity of carbonate added, and the mixture heated
to boiling over a gauze until the carbonate completely dissolves.
Thirty cubic centimeters of this mixture (equivalent to 10 cubic
centimeters copper sulfate solution) are equal to approximately
0.073 gram of pure dextrose.

The titration is carried out as follows—the sugar solution is
run in from a burette rather rapidly—(not so rapidly as to inter-
fere markedly with continuous vigorous boiling) until a heavy,
chalk-white precipitate is formed and the color of the fluid begins
to lessen perceptibly. The last portions should be run in in
quantities of from two to ten drops (depending on depth of color
remaining and the relative strength of the sugar solution), with
a vigorous boiling of about one-fourth minute between each addi-
tion. The end point of the reaction is the complete disappear-
ance of the blue color. This point is sharp and satisfactory.
The precipitate obtained is chalk-white and is rather an aid than
a hindrance to the determination of the end point. While potas-
sium sulfocyanide could be added in large enough excess to
retain the precipitate in solution, I can see no advantage in such
a procedure, whereas it has the disadvantage of permitting some
reoxidation if the titration be carried on too slowly.

It may be of interest to mention a simple devise used by the
writer with much success to prevent the annoying bumping of
the solutions during the process of titration. This consists in the
introduction into the titration beaker of a medium sized piece
of pure, previously well washed, absorbent cotton. By stirring
this cotton about as the titration proceeds, it is possible to
entirely prevent the bumping which otherwise may become very
troublesome. Glass wool may be used in place of cotton but is
not much more satisfactory and is considerably less economical.

In order to test the accuracy and reliability of this method
a solution of dextrose in distilled water was prepared of approx-
imately one per cent. The sugar content of this solution was
determined by Allihn’s gravimetric method, duplicate analyses
being made in every case. Using this solution, the value of the
copper solution used in the above method was computed from

~—

ox,

EE

Stanley R. Benedict 115

the results of three titrations, in which were employed 10,20 and
30 cubic centimeters of the copper sulfate solution, respectively.
The results of these titrations were exactly concordant, giving
the value of one cubic centimeter of copper solution as equiv-
alent to .0o727 gram of dextrose. On the basis of this stand-
ardization the strength of about twenty-five sugar solutions
was determined, duplicate titrations being made in every case.
The actual amount of sugar present in each solution was also
determined by Allihn’s gravimetric method and the results thus
obtained compared with the results of the titrations. Where
the solution titrated was too strong for determination directly
by Allihn’s method, it was diluted with a measured volume of
water and then analyzed.

Duplicate analyses were invariably entirely concordant, the re-
sults differing either not at all or only by such slight error as is
experienced in reading the ordinary burettes. The titrated solu-
tions varied in strength from 2.2 to o.1 per cent of dextrose. To
some of the solutions were added small portions of the common
inorganic salts, such as might occur as impurities in ordinary
sugar solutions. The results were not affected by the presence
of these substances.

The values obtained for the various solutions were in every
case in very close agreement with the results found by Al-
lihn’s method. So far as could be determined, the value of the
copper solution in terms of dextrose remained absolutely con-
stant for strong or dilute solutions. In fact, in the case of many
of the dilute solutions, 7. e., under 0.3 per cent, the results were
more closely inagreement with the actual sugar content (known in
these cases because obtained by exact dilution of stronger solu-
tions) than could be obtained by Allihn’s method, which, while
giving very exact results and duplicates with solutions of about
five-tenths of one per cent, is not so satisfactory for more
dilute, or stronger solutions.

Attention is called to the fact that the value of the copper
solution remains the same for dilute as for stronger solutions.
This is not the case with many of the solutions employing the
alkali hydroxides. I believe this fact finds ready explanation in
the less strongly destructive action of the carbonates than the
hydroxides, as pointed out earlier in this paper.

116 Detection and Estimation of Sugars

A table is appended giving results of five typical titrations,
carried out in exact accordance with the directions given above.

GRAMS OF GLUCOSE IN 10 cc. OF SOLUTION.
Ne: at € Allihn’s | Average from | Writer's Titration Average of Results
Method. Allihn’s Method. Method. by Writer’s Method.
1 0.0280 0.0270 0.0280 0.0277
0.0260 0.0274
2 0.0530 0.0530 0.0528 0.0528
} 0.0531 0.0528
3° 0.1021 0.1016 0.1020 0.1021
0.1012 0.1022
4 0.1810 0.1814 | 0.1815 0.1817
0.1819 0.1820
5 0.0112 0.0107 0.0117 0.0115
0.0102 0

.0113

It should be stated here that where the sugar solution to be
titrated contains less than three-tenths of one per cent, the
writer has found the following method more expeditious and
satisfactory, as avoiding extreme dilution during the titration.
A measured amount (25 cc.) of the sugar solution to be deter-
mined is run into 30 or 60 cubic centimeters of the copper titra-
tion solution, the mixture boiled for from five to seven minutes
and the residual copper determined with a dextrose solution of
known strength, of approximately 1 per cent. While not an
essential procedure, this method is expeditious and yields ac-
curate results. In ordinary titrations the writer has used the
mixed fluid in quantities of 30 or 60 cubic centimeters, the
former volume being sufficient for sugar solutions of 1 per cent
or less.

In cases where it is not convenient to standardize the copper
solution against sugar solutions, of known strength, it is possible
to obtain quite satisfactory results through the employment
of an exact weight of pure crystallized copper sulfate. Where
this method is employed the copper sulfate should be freshly
recrystallized, dried upon blotting or filter paper and accurately
weighed out upon an analytical balance, 69.30 grams being dis-
solved in water and the volume made up to exactly one liter.
Ten cubic centimeters of this solution are equivalent to approx-

Stanley R. Benedict 117

imately .o73 gram of dextrose. This process was tried upon
three occasions by the writer and the results varied but slightly
from those obtained through standardization of the solution
with sugar solutions of known strength, though for absolute
accuracy the latter process is of course to be preferred.

Some titrations have been made to test the applicability of
this method to the determination of dextrose in the urine.!
The results obtained show a quite satisfactory approximation to
the actual sugar content, although, just as should be expected,
they are not exact, the variation usually being from five-hun-
dredths to two-tenths of one per cent of the actual sugar content,
depending upon the relative amount of dextrose present. Solu-
tions containing larger percentages of sugar yield more correct
results.

In conclusion it may be stated that the writer has met with
certain substances (notably traces of chloroform) which are cap-
able of causing a portion of the copper to precipitate as the red
oxide, even in presence of the sulfocyanide. While impurities
which will do this would be only rarely encountered, and although
the interference with the end point is usually not marked, it is
desired to offer an alternative formula for Solution C, which will
obviate this slight difficulty entirely. Theformula for this solu-
tion is as follows:

PeaeaSSipiiar feTLOCyAnide .¢......\0 42s see eee ee eee ee 30 gms.
Peaieissiiin syitocyanide: -\.) 2.4). is ap eee ee ie 125 *
maaToOus Sodium carbonate: « 53... ese ee eo Oe oe 100 “
PEPSEAUC VALET EO! 3 = 2:c 0) 2jcts,'e.5) 2) teers ee ee a ee 1000 cc.

The use of this solution does not change the value of the copper
in terms of dextrose, and may be used entirely in place of the
other formula if desired.

This method has not yet been tested regarding the other re-
ducing sugars except to find that they also yield the white
precipitate as a result of their action on the solution. There is no
apparent reason why there should be any difficulty in such an
application.

? Rudischand Celler (Journ. Am. Med. Assoc., Jan. 26, 1907, p. 324) de-
scribe a method for determination of sugar in the urine through the use of
ordinary Fehling’s fluid to which is added a large amount of potassium
sulfocyanide, thus preventing precipitation of cuprous oxide when reduc-
tion occurs.

Ree) *
MEPe cer (ed

be
can Lanett!
ee Lt: 4
- be é
é [
4 ¥
2y io 5
ties b ‘
Re
alt
e - >
1
v3" vo

THE HEAT OF COMBUSTION OF VEGETABLE PROTEINS.
By FRANCIS G. BENEDICT anp THOMAS B. OSBORNE.

(From the Chemical Laboratory of Wesleyan University and the Laboratory
of the Connecticut Agricultural Experiment Station.)

(Received for publication, March 16, 1907.)

Very few determinations of the heat of combustion of vege-
table proteins are to be found in the literature. The earliest
appear to be those made by Danilewsky,! who found for “ Pflan-
zen-fibrin’’ 6231, for legumin 5573 and gluten 6141 calories per
gram.

As these determinations were made according to Stohmann’s
method at a time when this method was not perfected and as
numerous precautions, which later were found to be important,
were not considered, Stohmann” made new determinations of the
heat of combustion of several proteins, among which were the
crystallized globulin of the squash seed, prepared by Grutbler,
for which he found 5598 calories, and conglutin, which gave 5362
calories per gram.

Berthelot and Andre,’ using the Berthelot bomb found for
“Vegetable fibrin” 5837, for crude gluten 5995 calories per gram.

Stohmann and Langbein,‘ next determined the heat of combus-
tion of various carefully prepared proteins using the calorimetric
bomb of Berthelot. Most of the preparations were made by
Grtibler and before combustion were protractedly extracted with
ether. The substances were burned in the air dry state and the
results calculated to a water and ash-free basis. They found for
‘“Pflanzen-fibrin”’ (glutenin of wheat) 5942, for legumin from
white beans (phaseolin) 5793, for the crystallized globulin of
squash seed 5672, and for conglutin from lupines 5479 calories per
gram. No other determinations of the heat of combustion of
vegetable proteins are known to us.

1 Danilewski: Centralbl. f. d. med. Wissensch., xix, pp. 465 and 486, 1881.
*Stohmann: Journ. f. prakt. Chem., xxxi, p. 273, 1885.

3 Berthelot and André: Aun. de chim. et de phys., XXii, p. 25, 1891.
*Stohmann and Langbein: Journ. 7. prakt. Chem., xliv, p. 336, 1891.

IIg

120 Heat of Combustion of Vegetable Proteins

In view of the very small number of vegetable proteins of which
the heat of combustion has been determined and of their impor-
tance in so much of the food consumed both by men and animals,
it seemed important to determine this factor for a large number
of these substances derived from many of the different kinds of
seeds in general use as food.

These proteins were all prepared with special reference not
only to their separation from all non-protein bodies but also from
other associated proteins. In other words, the preparations rep-
resent as nearly as possible definite chemical individuals. That
they are in fact definite chemical individuals we are unable to
assert for no method is yet available whereby this fact can be
established. These preparations, however, represent substances
that we have been unable to separate by fractionation into dif-
ferent parts, the composition or properties of which would indi-
cate a mixture. They must for the present be accepted as rep-
resenting the most definite protein products that can be prepared
from the seeds that yielded them. These substances were burned
in the Berthelot-Atwater calorimetric bomb.

The calorimeter room was kept at the most constant tempera-
ture possible and all the minor precautions, suggested by experi-
ence with over 10,000 of these combustions, were employed.

The bomb used in these determinations was so adjusted, as
regardsits hydrothermal equivalent, as to give the heat of com-
bustion of pure, anhydrous cane sugar as 3959 calories per gram
and of pure, fused benzoic acid as 6322 calories per gram.

From 0.5 to 0.8 gram of substance was used for each combus-
tion which was weighed after the thoroughly dried material had
assumed constant weight by prolonged exposure to the air.

As it is of the utmost importance that a definite state of desic-
cation should be established to which the heat of combustion
should be referred a series of preliminary experiments were tried
with the globulin edestin. The preparation of edestin, made
according to the method described by Osborne,! was neutral to
phenolphthalein and had been recrystallized several times.

Of this preparation 16 samples of the air dry material, each weighing
2 gms. were dried in a high vacuum for 9 days, 17 days, 24 days and 46

1 Osborne: Journ. of the Amer. Chem. Soc., xxiv, p. 39, 1902; also Zett-
schr. f. physiol. Chem., xxxiii, p. 240, 1901.

Francis G. Benedict and Thomas B. Osborne 121

days and weighed at the end of each period. The average percentage of
water, as determined by the total loss of weight at the end of each period,
was 5.64, 5.76, 5.95, and 5.99, respectively, the agreement between the
several samples being remarkably close. The average weight of material
in each dish was therefore 1.8802 gram. After the desiccated edestin
had been exposed to the air of the room the average weight in each dish
was 1.991; gram, the protein having regained very nearly all the moisture
lost in the desiccator.
The heat of combustion of this material was determined with the follow-

ing results:

5184 calories per gram.

5202“ e

S185) as e

5199 *

5196 4 1

ICE “i
ol94 ag

“ce ce

Average, 5193

Calculating this result to an ash- and water-free basis, as determined
by the above desiccation, we have 5507 calories per gram of edestin.

Of the same edestin 6 samples of 2 gms. each were weighed in shallow
aluminum dishes, provided with a tight fitting cover,! and dried at 110°
for successive periods of 5, 34 and 4 hours, weighed after cooling in a desic-
cator, and then placed in the high vacuum for 5 weeks. At the end of this
time the loss of weight was 8.37 per cent. After standing for some time
in the room, exposed to the air, 4 samples were burned with results shown
below:

Weight after Heat of Combustion, Heat of Combustion

Dry Weight. Standing. Calories perGram. Calculated to an Ash-

and Water-free Basis.
1.8353 gm. 1.9685 gm. 5228 5614
i 33583 1.9700 “ 5247 5639
So 200s: ILE PA oY 5226 5632
E8322) > OTA 5222 5633

Average, 5629

A comparison of these moisture determinations shows that
while the highest per cent of moisture found by desiccation at
the room temperature in a very high vacuum was 5.99 that found
by heating at r10° C. and subsequently drying in the vacuum
desiccator for a long time was 8.37 per cent or 2.38 per cent more.

While this large difference in the apparent water content was
found in this sample of edestin, in three other samples of edestin
the difference was found to be but 0.08, 0.42, and 0.39 per cent,

1 Cf. Benedict and Manning: Amer. Journ. of Physiol., xiii, p. 309, 1905;
XVili, p. 309, 1907.

122 Heat of Combustion of Vegetable Proteins

respectively. We are at present unable to explain the marked
difference in behavior of these samples of edestin. In a large
number of other vegetable proteins the difference was found to
be not far from 0.25 per cent.

It has previously been shown by one of us’ that protein pre
parations which have been dried by very long desiccation in a
high vacuum over sulphuric acid show a gain in weight when
subsequently dried at 110° C. in air. This oe in weight is
more than lost if the protein after drying at 110° is again dried
for a sufficient time in vacuo over sulphuric acid.

The fact that heating at 110° C. and subsequently drying ina
vacuum results in a slightly larger loss of weight indicates a slight
loss of water of constitution from the protein, which water, owing
to the very high hygroscopic power of the dry protein, is retained
as moisture,even at 110°, and is subsequently lost inthe vacuum
desiccator. That this explanation is correct is, however, ren-
dered doubtful by the results obtained in drying the prepara-
tions for the determinations of the heats of combustion that are
later described in this paper. A certain number of these prepara-
tions had, before using them for these experiments, been dried
at 110°, the others dried only over sulphuric acid at the room
temperature. The results, however, in most cases showed prac-
tically the same differences in moisture as determined by the
two methods of drying just described. It would seem probable
that if this difference was due to water of constitution lost at 110°
these samples which had previously been dried at 110° would
show no difference in the result of drying by the two methods, for
it has been our experience that the protein preparations assume
constant weight after drying for a few hours at 110°.

This excess of moisture, found after drying at 110° and then
in vacuo, over that found by drying in vacuo alone, is shown by
the following table:

Preparation Previously Dried over Preparation Previously Dried
Sulphuric Acid. at 110°C.
Conghstin i: .vsaseeess'- 0,18 ae cent. Amandin=, 727.4 0.28 per cent
Edestin No. Py ae 0.08 Pxcelsin, = 2.4.42. 0.29 a
si NO2io aad US: ARE Vionin® nate aise: O70)
< NO: o.oo = ate 0:39. * Giycinint 3525.6 0:297
Cotton- seed globulin ee. Legumin, lentil ...0.06 ‘

* CE Benedict and Manning: Amer. Journ. of Physiol., xviii, p. 213, 1907.

Francis G. Benedict and Thomas B. Osborne 123

Preparation Previously Dried over Preparation Previously Dried
Sulphuric Acid, at 110° C.

Phaseolin, adzuki bean ..0.21 percent. Legumin, horse

Prorder! 5. ssc ate we ae OF 20) ort. bean 7) ys e.ee 0.08 percent.
Bydestin AQMs san, <2 s0herers Pyapaie) | a Legumin, Vetch ..0.20 ‘
Canelutsny 8. 5 4.5 viewer 1.26 se Phaseolin, adzuki
Dea” ea eee Ostse
Phaseolin, kidney
Weal 52. eee 0.207 252
Congliutin as se OLOR A
Glutenim, ~ + 4.cee OF09i = =
Glycmin ile See On 25: oe
Edestin No.4 ....0.68 “

With the exception of Edestin No. 40 and Conglutin / the pre-
parations that had previously not been heated show about the
same differences as those that had already once been dried at 110°
and it would therefore appear that this difference is, in most cases,
due to some undetected condition of manipulation. No explana-
tion of the large difference shown by Edestin No. 40 is apparent.
This might be attributed to water of crystallization were it not
for the fact that the other samples of edestin were likewise crys-
talline.

In order to detect any possible effect of heating during the dry-
ing a large number of determinations of the heat of combustion
were made on material that had been dried only at room tem-
perature 7m vacuo and the heat of combustion calculated to an
ash- and moisture-free basis, the moisture being determined solely
in vacuo and for comparison with these the heat of combustion
was also determined in duplicate samples that had been dried at
r10° and then for a long time im vacuo, as just stated. Ina few
samples the moisture was determined simply by long drying in
the vacuum and the results in these cases are probably stated a
small fraction of 1 per cent too low.

Returning now to the sample, Edestin No. 40, it is to be noted
that the heat of combustion of the sample which had been dried
im vacuo was 5507 calories per gram, calculated to an ash- and
water-free basis as determined by drying in vacuo, whereas the
heat of combustion of the sample dried at 110° and afterwards
im vacuo was 5629 calories per gram, calculated to the ash- and
water-free basis as determined by drying at 110° and then im

124 Heat of Combustion of Vegetable Proteins

vacuo, If the first determination of the heat of combustion is re-
calculated on the assumption that the difference in water deter-
minations hy the two methods is due to moisture still present in
the sample dried only in vacuo we find the heat of combustion to
be 5630, which agrees exactly with the second determination,
5629 calories per gram. In confirmation of this result four other
samples of edestin were dried in the two ways and the following

results obtained, calculated to an ash- and water-free basis; a, —

moisture determined in vacuo; b, moisture determined by heating
at 110° for 124 hours and then drying in vacuum desiccator for
5 weeks. The result given under a is calculated under ¢ on the
assumption that the difference in moisture found between a and
b was due to moisture still retained in the samples dried by
method a.

a b c
Edestin No. 40 5500 5622 5623 calories per gram.
“ < 5570 5658 5575 * 2
ys ae 5575 5655 5600“ -
Cees 5629 5668 5653“ a
Cs 4 eis 5644 See: ake a

The difference in the moisture determinations in samples Nos.
1, 2 and 3, dried by the two methods, is much less than that found
for Edestin No. 40 and the difference in the heats of combustion
is correspondingly less.

In the following tables we give the results of the determina-
tions of the heats of combustions of a large number of different
vegetable proteins. The condition of these preparations as re-
spects drying before using in this work is indicated in Column
I by ‘‘dried at 110°” or ‘‘over H,SO,” respectively. Under VII
is given the heat of combustion as found for the preparation dried
only im vacuo calculated to a water-free basis, assuming that the
slightly greater loss of weight after drying at 110° and then im
vacuo is moisture.

AmanpiNn. This is a globulin forming the greater part of the protein
matter of the seeds of the almond (Prunus amygdalus var. dulcis).
Dried at 110° in air it has the following composition:!
C, 51.30; H, 6.90; N, 18.90; S, 0.43; O, 22.47 per cent.

1 The composition of all these proteins was determined after drying in
the customary way at 110° in air without subsequent drying in vacuo.

Francis G. Benedict and Thomas B. Osborne

I. ie

3

2

QA

ita)

n

5 ;

S 3

2 a>)

io (=)
at 110° in vacuo

-
_
Lal

Weight of Dry Sub-
stance.

1.8063
1.8054

at 110° and f 1.8001
thenin vacuo | 1.8002

|
s

stance when

Weight of Sub-
Burned.

1.9787 gm.
1.9784 “
1298225 *
1.9820 “

x

tion, Calories per

Heat of Combus-
Gram.

aT —

Ord Ordon
Soi i—)

and Water-free

Calculated to Ash- <
Basis.

<3 =
ference in Mois- = Ny
on

Corrected for Dif-

On
On
rs
(oe)

Corvin. . This globulin constitutes most of the protein substance of

the filbert or hazel-nut (Corylus avellana).

C, 50.72; H, 6.86; N, 19.03; S, 0.83; O, 22.56 per cent.

BLOG 1S
1 II. 10 Gt
3 F
o nD
A Pp
> sath
4 ; na
is z a8
2 ont oa
oy fa =
over. f 1.8237
ESO. vacuo 1.8238
at 110° and f 5
then in vacuo | 5202
EXCELSIN.

plates.

hexagonal plates.

gm. { 1.9980 gm.

INE

stance when

Weight of Sub-
Burned.

[2 Q000 1

ae

2.0046 “

=

tion, Calories per

Heat of Combus-
Gram.

| 5036 |
\ 5053 f
5044 |
5044 f

<
4

Calculated to Ash-
and Water-free '

Its composition, dried in air at

<
Lemal
=

ference in Mois-!

Corrected for Dif-
ture.

5983

Most of the protein of the Brazil or Para nut (Bertholletia
excelsa) consists of the globulin excelsin which crystallizes in hexagonal

C, 52.23; H, 6.95; N, 18:26; S, 1.09; ©} 21.47 per cent.

16 ic
as}
x
(a)
2
n
5 ;
— lao}
3 z
oy A
at 110° © in vacuo
at 110° and
then in vacuo
EDESTIN.

e
_
lanl
.

02)
S& Weight of Dry Sub-
Go stance

;
| 1.8552
1.8500
1.8485

HVE

stance when

Weight of Sub-
Burned.

=

ororo1 co Heat of Combus-
NW tion, Calories per
Gram.

—_™—

<
4

and Water-free +

Calculated to Ash-
Basis.

5719

The preparation used for these experiments consisted entirely of
Its composition is

Il.

Lama

ference in Mois-

Corrected for Dif- <
ture.

5737

This is the principal protein constituent of the hemp-seed

126 Heat of Combustion of Vegetable Proteins

(Cannabis Sativa). The preparations used for these combustions
all composed of octahedral crystals.

_
.

Previously Dried.

Prep.
No. 1

over
H,SO,

Prep.
No. 2

over
H,SO,

Prep.
No. 3

over
H,SO,

Prep.
No. 4

at 110°

GLOBULIN OF THE COTTON

composition of which is

_
.

Previously Dried.

over
H.SO,

C, 51.36; H, 7.01; N, 18.65; S, 0.88; O, 22.10 p

Il.

Dried.

in vacuo

+ at 110°
and then
in vacuo

in vacuo

} at 110°
and then
in vacuo

|

)

in vacuo

} at 110°
and then
in vacuo

in vacuo

—_ co aS

Lael
—

Weight of Dry Sub- -
stance.

. 8980
. 8980

. 8972
. 8972

. 9092
. 9087

oo co

oo

. 9072
. 9024

. 9098
.9102

oo oo

9045
9077

oo

0.9182
0.9114

0.

SEED

The composition of Edestin 1
er cent.
IV. v (on
is sg a8
2s 35 <4
ns 35 $3
jj Og a
® 2% as : SB a
Zgé -se sya
‘5 aia e305 24m
= ss) 6)
.9960 gm. 5006 |
9961 } Haat one
9867“ 5126
9867“ | $28 otee
LG
0083 “ \ f 5018
0st 7 { 2020 } 5575
9989“ 5113 5630
9925“ 5165 5681
0289 4973
(0296 } { 4971 } 5629
. 5146
“epee? 6 5673
5156 }
‘ 5138
toee. | 2135 | 5644
5135

(Gossypium herbacium).
nearly all of the protein matter of this seed consists of this globulin, the

C, 51.71; H, 6.86; N, 18.30; S, 0.62; O, 22.51 per cent.

Il.

Dried.

in vacuo

| at 110°
and then
| in vacuo

Ill.

stance,

Weight of Dry Sub-

—
7 ae
~S ee

1.8132
0.9054

co

IV. v.

& ©
2g aie
ne Be
So |
2oe o-g
Has oie
‘Sam B25
= 6}
~ 5179

pies eo: \ | ori3 |
-9741 | 5258 |

3020 |
.9939 “ \ 2059 |
9966 | | 9074 }

<
lanl

and Water-free ,

Calculated to Ash-
Basis.

5657

5596

ference in Mois- +

Corrected for Dif-
ture.

or
oO
I
or

5583

5653

Very

r=)
=
:

ference in Mois-

Corrected for Dif- S
ture.

5673

Francis G. Benedict and Thomas B. Osborne 127

ViGNIN. The greater part of the protein substance of the cow pea
(Vigna sinensis) consists of this protein which has the properties of a glob-
ulin and the following composition:

C, 52.64; H, 6.95; N, 17:25; 5, 0.42; O, 22.74 per cent.

io Wie III. IV. V. Viw Vil
: a ob ® qo am
8 Z ag ae ae Ae
E > Re ane gs §
> A ee og B= Deh sat
= Ba Sra BONY eS
5 “ = Bee Go eS aes
é 2 oe cet $26 28a ES
A Qa = = fd S) 6)
{ in vacuo 0.9206 gm. 1.0292 gm. | 5059 =a
Og1g2 “) Yeages ef > | 50g0nF een
at 110° at 110° 1 £5902
and then 0.9118 “ 0.9944 “ ( 5210 5703
| in vacuo \
{ in vacuo 0.9081 “‘ 1:02427 "~~ 5023,) 5688
OF90 gs “L0240Reran| | 5027
over
H.SO,4 at 110°* OR4 552 5.02499) { 5160 )
andthen 0.4551 “ 0.4999 “ f {5178 { 5691 5721
{ in vacuo

* This was the only preparation which showed a greater loss of weight after drying in
vacuo alone than after drying at 110° and thenin vacuo. No reason for this exception is
apparent. Were it not for the fact that the heats of combustion of the two samples agreed
so closely it might be taken as indicating oxidation.

GiyciniInN. This globulin forms the greater part of the protein matter
of the soy bean (Glycine soja). Its composition is

€, 52:01; H; 6.89; N, 17.47;'S, 0:71; 0,222.92 per cent:

If II. Ii. TV. V. VE VAT:
A £5 9 wh
a 7 ne oped ease
ae > Zz 5.2 o5 5
A fa) By OS Serie ae
2 Og ee ao = ae - B8
z ae S26 oo SBSH
= z oS mas 2#§8 Bea oes
o < ‘Ohh on B30 saa Eas
Ay A = = se S) o
( 5236 )
iets 1.8242 gm. 1.9620 gm.| | 2
yellow in vacuo { 1.8254“ 1.9655. “ } , oan + 5667 5687
SOY | 5255 J
bean
t 110° es J
Bett?) | 1.8181 1.9938 <“ 54 Taos
eee 1.8197 “ 1.9979 “ eter Du
Japan. [imvacuo 0.9184 “ 1.0248 « 5056 5665 5680
Ee aay O.one7 Hs T0247) 5016 5619 5636
pee, Pamidee, | Onto “8 al-0996 “S088 ) seas
so, | andthen 0.9151 “ 1.0211 “ f 5048 [° oo
ores in vacuo

128 Heat of Combustion of Vegetable Proteins

Lecumin. Most of the protein matter of seeds of the horse bean (Vicia
faba), lentil (Ervum lens), vetch (Vicia sativa), and pea (Pisum sativum)
consists of a globulin, the preparations of which have thus far appeared
to be identical,in respect to all their properties that have been examined.

This agreement is now found for the heats of combustion of preparations |

from the three seeds first named. No preparation from the pea was avail-
able at the time this work was done. The composition of legumin is

C, 51.72; H, 6.95; N, 18.04; S, 0.39; O, 22.90 per cent.

Legumin, lentil.

lL. Il. Ill. IV Vie VI VII.
: 4 2 a3 Sm
z Z ee pe 32 Be
® fo ok &
A R ae 35 “Ss “Se
2 -_. Oo og a 6: ie
S : ae 225 ae: ee 3
ant =~ a eae sU OR
c 2 ck San S20 EEE ESS
Py A = S iq +S) tS)
5224
0.9094 gm. 0.9771 gm. — [
at 110° S
at 110 e
O290se) O.98aD" .- Sled,
and then { . Lee { t 5630
TS ROT 0.9093 0.9850 5193
Legumin, horse bean.
a f0.9122 “* 0.9820 “ 5203
invacwo jo e112 “ 0.9896“ tae } 5625 5632
at 110
at 110°
Pali2 s (0.9872 = 5220
and then { sc ” { } 5656
ae A 0.9107 0.9869 sul7P-
Legumin, vetch.
if “ re , 5156
; 0.9088 0.9830
invacuo { ’9998 “ 0.9840“ 5143 5586 5600
at 110°
at 110° “cr sc if 5132
0.9063 0.9878
and then { o8 ik Bees, 5150} 5621
ae e077 0.9865 1 pts

PuHasEo.in. This is a globulin that forms the principal part of the
protein substance of the white or kidney bean (Phaseolus vulgaris), and
also that of the Japanese adzuki bean (Phaseolus radiatus). Its compo-
sition is

C, 52.57; H, 6.97; N, 15.84; S. 0.33; O, 24.29 per cent.

———Oo

Francis G. Benedict and Thomas B. Osborne 129

Phaseolin, kidney bean.

i Il. JOG k IV. V. WAL VIL.
é a ae a) ob
E i 45 Be ge AS
"a > as S = } & we
A 5 ze 8§ 88 &:
2 oo S oe Os Se 5 ES
: | Ee 235 See Ed See
5 € mS ms 3 aoe 3ea e255
2 E ‘OD oA eeO ga EES
0 fa) = = eo} és) 6)
a 1.8270 2.0409 5050
| invacuo { 18976 2.0403 5049 ~ 5875 5689
at 110° 4 at 110° 5083 }
Sedthen (18232 2.0273 5060 $ 5679
in vacuo 5086 J
Phaseolin, Adzuki bean.
| in vacuo { Anas sng a eee a { ae \ 5709 5726
au L102 }
| at 110° 0.9213 “ 1.0202 “ 5095
fades (Uz IS a F { } 5683
gud ey, 0.9225 9 106 5088
EEE Ronee i oa ae ea 5681 5693
over |
isl) | at LO? md “6 66
0.9045 1.0029 5097
Poe {0.9040 “ 1.0030. (B1iz } 5702
1n vacuo

CoNGLUTIN. The substance formerly called conglutin, which forms
almost all of the protein matter of the seeds of lupines (Lupinus) of various
species consists in those of the yellow lupine (Lupinus luteus) of at least
two distinct globulins. These are designated Conglutin a and Conglutin /.
Inthe seeds of blue lupine (Lupinus angustifolius) the conglutin is simi-
lar to Conglutin a. Thecomposition of conglutin from the blue lupine is

C, 513; FE 6:86: (N1S01- S,0:32- 0} 23.10) per cent:

Conglutin a hasa similar composition, namely,

C, 51.75; H, 6.96; N, 17.57; S, 0.62; O, 23.10 per cent.

Conglutin ? has the following composition !

C, 49.91; H, 6.81; N, 18.40; S, 1.67; O, 23.21 per cent.

130 Heat of Combustion of Vegetable Proteins

Conglutin, Blue lupine.

I. II. Ill.
z Q

a _ g
: ss
- z 44

2 se on
Re =) =

over 902

a: 0.8
in vacuo { 0.8905

Conglutin a, Yellow lupine.

in vacuo 1.8551
¢ a
at 110° ) at 110°
| and then 1.8537
| in vacuo

Conglutin 8, Yellow lupine.

in vacuo 1.8530
over
H.SO, at 110°
pee 1.8278
in vacuo
VICILIN.

bo

bo

IV. Vv. VI.
bs ad
4 Bo a5
ba: a ee *
SEE ni a8 255
Baa $30 2am
i 20) 6)
SH 4893
Oran} | 4882 | 5475
“9897 Toa |
40
0223 epee 5528
0203 1 2367 | 5548
4873
0109 4868 5302
4888
0275 { oun } 5341

<
-
Ls

ference in Mois-

Corrected for Dif-
ture.

5536

5376

This is a globulin associated with legumin in the seeds of lentil

(Ervum lens), horse bean (Vicia faba) and pea (Pisium Sativum). Its com-

position is

C, 52.29; H, 7.03; N, 17.11; S, 0.17; O, 23.40 per cent.

he Il. Ill.
E :
= As
> aes
= og
> 2a
om = 2
: 2 3a
Wy =) =
at 110° in vacuo 0.8900 gm.

LEGUMELIN.

(Ervum lens).

IV. V.
28
28
4s ER
Ra 55
Mm Po 3
288 SO g
ag ag
was *#OR
ch $55
es rm
0.9727 gm. 5183
: 8 { 5200

<
lal

and Water-free +

Calculated to Ash-
Basis.

a)
jor)
ee)
Ww

This is an albumin-like protein which is found in the
extracts of most leguminous seeds.

This preparation was from the lentil

The composition of legumelin is

C, 53.31; H, 6.71; N, 16.08; S, 0.97; O, 22.93 per cent.

Francis G. Benedict and Thomas B. Osborne 131

ie Il. Il. EV: Viz VI.

1 he od
se) 3a ae
- gE =§ Es o8
i) ‘3g 63 2 =e
D 3 oo wO = = a
23 Sas nee S33
: 3 a's es 5 268 ga
2 a ‘oD ‘3 AQ B50 ici
6 a) = = tr )

at 110° in vacuo 0.9154 0.9901 gm 5209 5676

Gu1aDIN. This is a protein soluble in 70-80 per cent alcohol, which
forms nearly one-half of the protein of wheat (Triticum vulgare) and rye
kernels (Secale cereale). The composition of gliadin is

252.723 EL. G86) Ni. l7A66s e038; ©} 2/3 percent:

Gliadin from wheat.

Ne 1 Til IV. Vv. VI.
3 a gs a:
A Le as ES es
ae ewe Og 3
2 o5 So aoe ue
=} Sie Og oO. 3S 2
xe) 5 sire} qeae s =| s =30 2
ay a S = ss} S)
5229
I. over : {1.7786 gm. 1.9596 gm. | e
EMO TES a. 7a87 ead Ona mee oie
II. over (1.8045 “ 179003 sa) yieatsir
H.SO, 2 CEO” \e029023)° Ovog2 sues { 5181 \ 5713
Gliadin from rye.
1.8095 “ 1.9925 ||
over . j . 992 2 2
FSO, eee { 1.8090 “ 1.9955 “ | 2177 Sau
5184

GLUTENIN. This protein (Gluten-casein according to Ritthausen)
forms about one-half of the protein matter of the gluten of most varieties
of wheat, and is insoluble in neutral solvents. Its ultimate composition
is nearly the same as that of gliadin but the structure of its molecule is
very different, as shown by a comparison of the proportion of decomposi-
tion products yielded by these two proteins. The composition of glutenin
is

C, 52.34; H, 6.88; N, 17.49; S,.1.08; O, 22.26 per cent.

132 Heat of Combustion of Vegetable Proteins

i TH: Ii. EY. Vie VI.
a3 43
' = eit
i b ag a3 38
= Aé Aa i) EB ae
@ ? ye OL UE
2 83 at: 38 3 bie.
3 23 BEE oda 20%
5 2 22 323 326 EEE:
a A - = 0 6)
ee 1.8072 gm. 1.9924 gm. { 5134 }
at TAO in vacuo 1.8054. 1.9977" {5137 ai
5111
5119 5703
5116

GLoBuLIN. Wheat (Triticum vulgare). This globulin forms only a
very small part of the total protein of the wheat kernel. It is contained
chiefly, if not wholly in the embryo of the seed. Its composition is

C, 51.03; H, 6.85; N, 18.30; S, 0.69; O, 23.13 per cent.

L Il. Ill. EV: WE VI-
1 he ‘oo

= Ba a
= 9 4g ERS o8
Qa As =A o>} P+
> Aye ns og ae
: 33 ae: meh gE
Ss . ~

5 : £2 soe ee S05
= a wl was =)

Fs 2 SA crt $20 24a
a Aa E S fH 6)

over : oF 4759

H.SO, invacuo 0.8925 gm. 0.9883 gm. { 4790 5358

Horpein. Barley (Hordeum vulgare). Hordein, which forms about
one-half of the protein substance of the barley kernel, is readily soluble
in 70-80 per cent alcohol. Its composition is

C, 54.29; H, 6.80; N, 17.20; S, 0.85; O, 20.86 per cent.

L. I. Il. IV. v. VI. VII.
é bo AB ks
2 g Ag Fre
- as ns 55 ies “5
os = eo . bes |
= os ° 29 «5 : oF of 38
Z 3 a4 a5 t+ ee
5 £ Sn cht. 850 28m ELS
ay A — = 5) +) 1)
over sok = 5316
H,SO, i vacuo 1.8082 gm. 2.0025 gm. { 5317 5902 5916
at 110° a ; “ 5349
and then 12-8043 1.9946 hee 5934
in vacuo

Bynin. Barley malt. Bynin is the alcohol soluble protein of barley
malt. Its composition is

>, 55.03; H, 6.67; N, 16.26; S, 0.84; O, 21.20 per cent.

ee

Francis G. Benedict and Thomas B. Osborne’ 133

if Il. III. IV. v. VI.
a 5 Lo
- oe oe
a ze 26 a3 <7
A Q o ae Ow on
cal oe a pa Pie Cs oS
3 Ones Ces Oe EN.
3 ; 2% BEE me 25%
SS % "eo ‘was 25 3
8 2 3a ‘Sem B20 23a
Ay a = = ss) 6)
at 110° invacuo 0.9006 gm. 0.9899 gm. { 2503) 5807

The results of these determinations are given in the follow-
ing table:

Calories

Cc H N Ss O per gram

AIRE GIGES Geo 5 nearer 51.30 6.90 18.90 0.48 22.47 5543
COR WII 5 Gat Slo ape nee 50.72 6.86 19.03 0.83 22.56 5590
DOO ha, 419.6 aeeces GORRIOI 52.23 6.95" 18026" Ie09" 21°47 “5137
OO ES Fear en 51.36 7.01 18.65 0.88 22.10 5635
Globulin (Cotton seed)...... 51.71 6.86 18.30 0.62 22.51 5596
LEST ie a 52.64 6.95 17.25 0.42) 22.74 5718
SEAM SESE joo eicrais. 3's, Soo, e Ses ss 52.01 6.89 7-47 O27) 222292." 5668
Bo tiandte tee set enckay ays) eas oper. 4) sens 51.72 6.95 18.04 0.389 22.90 5620
REMEESCIMUENG 3125 soc 's, sco <'s (o, Sls. 0.3 52.57 6.97 15:84 O233 24729 1/5726
Conglutin (Blue lupine) ..... 51.13 6.86 18.11 0.32 23.10 5475
Conglutin a (Yellow lupine) . 51.75 6.96 17.57 0.62 23.10 5542
Conglutin 8 (Yellow lupine) . 49.91 6.81 18.40 1.67 23.21 5359
WGI “Gece operat achen cs ocr Ear ae 52.29) 72038 Wiis Oni, 623-40 5685
WS SMNLC LIT o). 5, ers wie 2 a¥e eis) +12 53.31 6.71 16.08 0.97 22.93 5676
(Gea iivay” VR Si enete cree ech ener 52.72 6.86. 17266 63203) 21 73 okas
Ginnie Sg ais eee eke Orne 52.34 6.83. 17-49 | 1.08') 22.26 5704
Cjobulm (Wheat) .......:.- 61.03 6.85 18.30 0.69 23.13 5358
ROT CIIM Ela S<Psyra Arye) & Akensnsholo ws 54.29 6.80 17.20 0.85 20.86 5916
IESiy AMT ety 55) sts) oa oa) bo cy-sus oud ever s 55.03 6.67 16.26 0.84 21.20 5807

In general the higher heats of combustion are found for those
proteins which have a higher carbon content and similarly for
those with a lower oxygen content. Many irregularities, how-
ever, appear in the preceding table, which are doubtless due to
the different proportions of the various amino-acids which con-
stitute the molecules of the different proteins. As we have but
little knowledge of the relative proportions of these amino-acids
in the proteins burned, we are not able yet to draw any definite
conclusions in regard to these differences.

——
lees
re an

>, Sal
: -
2 » 2
Poa 4
ro - G "i ——P
i is ;
a ¢ .
oe
* 7 - F *
ae ad
- a ie. ood ~y
2: ewe
‘s ee 4 i

er

IS THE SALIVA OF THE DOG AMYLOLYTICALLY ACTIVE?

By LAFAYETTE B. MENDEL anp FRANK P. UNDERHILL.

(From the Sheffield Laboratory of Physiological Chemistry, Yale University.)
(Received for publication, March 19, 1907.)

The view that the saliva of certain animals, particularly the
carnivora, is devoid of amylolytic power, has been current since
the classic investigations of Bidder and Schmidt.!. With respect
to the dog their statement is quite specific.2, An equally positive
and definite conclusion was drawn by Claude Bernard from his
experimental observations.? Physiological literature since that

1 Bidder und Schmidt: Die Verdauungssdjte und der Stoffwechsel, 1852,
p. 14, fig.

2 Wir haben daher neuerdings nicht nur die reinen Sekrete der betref-
fenden Driisen aufgefangen, sondern auch das reine Secret der Mund-
schleimhaut von Hunden gewonnen. ... Die so erhaltenen Secrete
wurden, jedes fiir sich, mit Starkekleister vermischt und einer Temperatur
von 40° C. ausgesetzt. In keinem Falle war durch die Trommersche
Probe vor 8 Stunden und auch dann nur eine Spur von Zucker nachzuweisen,
und wir miissen hiernach auf’s Nachdrticklichste wiederholen, dass Keinem
der Secrete, durch deren Vermischung die Mundfltissigkeit gebildet wird,
allein ftir sich bei der Umsetzung des Starkemehls in Zucker irgend eine
Fermentwirkung zugeschrieben werden kénne. . . . (loc. cit., p. 19 fig.)

3 ““La salive de chien pure et a l'état frais n’agit pas sur l’eau d’amidon;
mais elle acquiert cette propriété lorsque, abandonnée 4a elle-méme, elle
vient a éprouver une certain degré d’altération. C’est ce que prouve
lexpérience suivante. Exp.—Des salives fraiches de chien sous-maxil-
laire et sublinguale, trés gluantes, ont été séparément mises en contact
avec de l’eau d’empois d’amidon, et n’ont exercé aucune action pour
changer cette substance en sucre. Au bout de deux jours, ces salives ayant
été abandonées a elles-mémes par un temps chaud et orageux, avaient
complétement perdu leur viscosité, et alors elles agissaient trés éner-
giquement sur l’eau d’amidon pour le transformer ensucre. De lasalive
parotidienne fraiche placée dans les mémes circonstances n’eut pas d’action
sur l’eau d’empois d’amidon, et acquit la propriété de la transformer
lorsqu’elle eut subi un commencement d’alteration: d’ou il faut con-
clure qu’a l'état frais, les salives pures ne transforment pas l’eau d’empois
d’amidon. (Claude Bernard: Legozsis sur les propriétés physiologiques et les
altération pathologiques des liquides de lV’ organisme, ii, p. 249, 1859.)

135

136 Amylolytic Properties of Dog’s Saliva

time has repeatedly reiterated these announcements,’ so that we
read in the recent lectures of Starling: ‘‘In dogs the mechanical
action of saliva is its only one.’”*

The experience in our laboratory has never afforded any occa-
sion to question the accuracy of this statement. It was with
some surprise, therefore, that we learned from the paper of Neil-
son and Terry,’ entitled ‘‘The Adaptation of the Salivary Secre-
tion to Diet,” that in all their experiments ‘‘the saliva of dogs
was found to be active, but varying considerably in its amylo-
lytic power, in different animals.’’ Dogs fed upon a diet consist-
ing principally of bread, with a small amount of meat broth and
some ground meat are reported as invariably having a saliva
with strong amylolytic powers. The authors claim that both
the saliva and gland extracts of dogs fed on a mixed diet contain-
ing bread show a much greater amylolytic power than those of
street dogs on an unknown diet. An experiment is reported in
which a dog was fed on meat for fourteen days, the salivary glands
on one side were taken out, and the animal subsequently put on a
bread diet for fourteen days before removal of the remaining
glands. No sugar was detected in any case within the first hour
in the digestion trials with the extracts and starch paste; but on
the basis of subsequent observations upon the digestive mixtures
the conclusion is drawn that the glands adapt themselves to
diet. It is further stated that ‘‘ probably in all dogs there is an
active ptyalin, but it is relatively inert as compared to human
saliva.”’

The importance and interest which attaches to a demonstra-
tion of adaptation in digestive secretions need scarcely be empha-
sized. Earlier alleged evidences of such reactions in the animal
organism have lately been subjected to severe criticism,’ so that
analogies in the case of the pancreas and intestine are, at the —

' They are found, for example, in the text-books of Gamgee, Neumeister,
Schafer, Hammarsten and Abderhalden.

* Starling: Recent Advances in the Physiology of Digestion, p. 41, 1906.
Fermi (Arch. f. Physiol., Supplementband, p. 65, 1901) also failed to find
ptyalin in dog’s saliva.

* Neilson and Terry: Amer. Journ. of Physiol., xv, p. 406, 1906.

“Cf. Plimmer: Journ. of Physiol., xxxiv, p. 93, 1906; xxxv, Pp. 20, 1906;
Bierry: Compt. rend. de la soc. de biol., lviii, pp. 700, 701, 1905.

Lafayette B. Mendel and Frank P. Underhill 137

present moment, not well substantiated. Aside from the work of
Neilson and Terry we recall only two other recent investigations
which attribute amylolytic properties to dog’s saliva. Henri
and Malloizel have incidentally noted variations in supposed
amylolytic activity in the submaxillary saliva of dogs when dif-
ferent stimuli were employed. They remark ‘‘La salive sous-
maxillaire est toujours trés peu active. Dans les cas de l’activité
maximum nous avons trouvé au bout de 5 heures, 6 a 7 milli-
grammes de sucre.”"! When we add that under comparable con-
ditions of experiment lymph formed five times as much sugar
and that there was a parallelism between the mucin content of
the digestive mixtures and their supposed activity, these experi-
ments have little positive significance. Aside from this incon-
clusive evidence we have further noted an experiment by Hem-
meter? in which dog saliva and trypsin solutions were subjected
in different and separate portions to various temperatures, and
thereafter their amylolytic and proteolytic power tested. Since
the investigator was studying a quite different problem and used
saliva as a type of amylolytic secretion one must assume that
he regards the canine saliva to be active upon starch.®

We have made a series of observations upon dogs and cats in
order to demonstrate, if possible, the native or acquired amylo-
lytic properties of the saliva. The experiments have failed to give
evidence of any marked or characteristic digestive action upon

1 Henri and Malloizel: Compt. rend. de la soc. de biol., liv, p. 331, 1902.
Our attention was directed to this paper by Prof. A. J. Carlson.

2Hemmeter: Journ. of the Amer. Med. Assoc., Dec. 9, 1905; Berl. klin.
Wochenschr., 1905, No. 44a (Ewald Festschrift). In the older literature
positive results are recorded by Astaschewsky. Cf. Jahresber. f. Thierchem.,
wit 250,-1877.

3 Professor Hemmeter has informed one of us (M.) that he has observed
the inactivity of dog’s saliva on boiled starch in some cases. He writes:
“T noticed in one dog that when the saliva is collected by a sponge from
his mouth and not by catheterization of the duct, that sometimes it is
active and sometimes it is not. Even in dogs in which the saliva is
obtained by catheterization of the duct and stimulation of the chorda
tympani, the saliva is of varying amylolytic power. This is true not only
of dogs, but also of human beings . . . some deep seated causes are under-
lying these discrepancies.’’ Professor Hemmeter adds that he has not
investigated the production of sugar in the digestions, but confined his
attention to the solution of the starch.

138 Amylolytic Properties of Dog’s Saliva

starch paste! Furthermore a careful study of the data submitted
by Neilson and Terry renders their positive deductions some-
what less convincing.?- The reacting solutions as a rule reduced
Haines’ solution only after comparatively long periods of diges-
tion, and then only slightly. For example, one extract in which
the reduction test showed ‘‘a trace of sugar on standing”’ after
four hours’ digestion with one gram of starch, was ascertained
to contain 0.029 gram of sugar (Experiment D); whereas a com-
parable digestion (Experiment B) with a presumably more active
extract is recorded as giving a ‘‘heavy reduction at once”’ after
one and one-half hours, while the amount of sugar was estimated
at only 0.054 gram from 1.3 grams of starch. In one experiment
(B) upon a bread-fed dog* the saliva itself was distinctly active.
With this exception, the amylase might be regarded as relatively
inert, at any rate if we use the observations on the pancreatic
amylase of the same animal or the saliva of other species as
standards of comparison.

In our experiments upon dogs the saliva was obtained from
cannulas in the ducts after stimulation of the chorda tympani,
with frequent intervals of rest, unless otherwise stated. The
animals were anesthetized with ether or A. C. E., after adminis-
tration of a small dose of morphine. The gland extracts were
prepared with toluene water and filtered through absorbent cot-
ton. A one per cent paste prepared from best grade arrowroot
starch was used in the digestions, which were carried on in the
presence of toluene, at 40° C. The progress of amylolysis was
noted by applying the iodine test and Fehling’s test. Where
quantitative estimations were attempted the Allihn gravimetric
method was applied, because we regard this procedure as rather
more satisfactory than the volumetric processes where such

+I am informed by Prof. E. H. Starling that he also has examined
both submaxillary and parotid saliva from the dog (after chorda stimu-
lation or pilocarpine injection) a number of times, but has never obtained
any amylolytic effect.—L. B. M.

? Plimmer also says: ‘‘ The results obtained by these authors, however,
are not very conclusive and further experiments on this point would be of
great interest.”’ (Journ. of Physiol., xxxv, p. 21, 1906.)

3In the original paper B, p. 409, is labeled ‘‘ Meat-fed Dogs’’—evidently
a typographical error. The weight of the second submaxillary gland on
p. 410 is also apparently printed incorrectly.

Lafayette B. Mendel and Frank P. Underhill 139

small quantities of reducing substances are present. The reduc-
ing power of the solutions is expressed in terms of copper. This
was necessary because in several of the trials the quantity of
maltose could not be calculated from Wein’s tables, as the amounts
of copper obtained fell below the limits of these tables.

PROTOCOLS.

I. A dog weighing 14 kilos was used. Diet unknown. The saliva
was obtained from both submaxillary glands, and portions
successively collected were mixed. The final records noted
below were made after three to four days digestion at about 40°C.

(a) 3.15 p.m. Mixed Saliva ....... 20 cc. | In two hours a quantita-
StAarChv easter. a, - 20/5 tive estimation indi-

Sterilized Water | cated 0.0137 gram Cu

and Toluene.... } 20 * in all. Three days

(b) 4.05

(c) 4.15

(d) 4.25

(e) 4.30

= | later this was prac-
60“ | ticaliy unchanged:
0.0141 gram Cu in
all. After four days
the iodine reaction
showed a reddish
tinge. Aslight reduc-
tion could be observ-
ed with Fehling’s so-

lution.
Sralivds 5. < s.2-0's shaq 2 40 cc. | The unchanged starch
SEAT aa hasnt eaten 25 interfered with the fil-
Water and Toluene.. .10 “ tration of the trace of
-+ cuprous oxide preci-
(ite pitated. After four

days the solution still
gave the blue color
with iodine solution
and only a slight re-

duction.
SLRUED, Ce Caen 6 ce. Four days later this gave
Starch and Toluene . 6“ a reddish blue iodine
— test anda slight reduc-
12 tion with Fehling’s so-
lution.

A similar trial gave less evidence of digestion after
four days.

5 cc. saliva+toluene. No starch added. On the fol-
lowing day this saliva alone yielded 0.0018 gram
Cu in an Allihn estimation.

140 Amylolytic Properties of Dog’s Saliva

It will be noted that the digestive action, even as indicated by the
reduction tests, was minimal at most. In (@), for example, despite the
large quantity of saliva present, it did not increase, thus suggesting exte-
rior causes for the reaction. The saliva itself, after standing, gave a visible
reduction sufficient to account for part of the other reactions.

II. From a dog weighing 15 kilos, saliva was collected alternately from
the right and left submaxillary gland, with varying strengths of
stimuli (coil distance) and intervals of rest, and it was tested in
portions of 5 cc. saliva+5 cc. starch paste+toluene. Next day
the nine tubes all still showed a blue reaction with iodine. Slight
reduction tests were obtained, becoming somewhat more distinct
in the portions of saliva last collected (after 2 hours).

Ill. A bitch weighing 9.5 kilos was anesthetized with ether and A.C. E.
and saliva was collected directly by blowing chloroform or ether
vapor into the mouth. Digestion trials were made: 44 cc.
saliva + 4} cc. starch paste (1) with and (2) without addition
of toluene. On the following day both tubes gave a blue reac-
tion with iodine solution and a slight reduction. Four days
later (1) was clear but still gave a blue reaction with iodine
(soluble starch) as wellasa marked reduction test; (2) was decom-
posed.

IV. For quantitative estimations submaxillary saliva was collected by
chorda stimulation from the same animal, and digestion trials
were arranged as follows:

(1) (2)
18 cc. Saliva. 18 cc. Saliva.
18 ‘* Starch Paste. 18 ‘‘ Starch Paste.
18 ‘* Water. 18 ‘* Water.
Toluene. No Toluene.

Two days later the mixtures still gave a blue reaction with iodine
despite the very large quantities of saliva used.

Copper found: (1) 0.0480 gm.; (2) 0.0696 gm.

The toluene was omitted from one of the trials in order to exclude
any possible retarding action of the antiseptic on the enzyme.
The slightly increased yield of copper in (2) may equally well
be ascribed to the action of microérganisms. At any rate no
serious inhibition is attributable to the toluene used in our
experiments.

V. The dog weighed 12 kilos. Cannulas were introduced into the
ducts of both submaxillary and sublingual glands. The periods
of stimulation and rest were continued over three hours; the
submaxillary and parotid glands were excised, minced and ~

Lafayette B. Mendel and Frank P. Underhill 141

extracted with toluene water and filtered through absorbent
cotton.

Successive portions of saliva, etc., were mixed, as soon as col-
lected, with sterilized water, starch paste and toluene. They
were examined 24 hours later.

Iodine Test. Reduction Test.
(1) 5cc. Submaxillary Saliva
20 *“‘ Starch Paste blue 18 mgm. Cu
5 *“* Water and Toluene
(2) 10 “ Submaxillary Saliva
10 ‘‘ Starch Paste blue 28 mgm. Cu
Toluene
(3) 5 “* Botled Submaxillary
Saliva |
20 ‘“‘ Starch Paste | blue none
5 “* Water and Toluene
(4) 1“ Sublingual Saliva
10 “ Starch Paste blue faint
5 ‘““ Water and Toluene J
(5) 10 “ Submaxillary Saliva
10 “ Starch Paste purple positive
Toluene
(6) 15 “ Parotid Extract
20 ‘‘ Starch Paste red heavy
Toluene
(7) Ditto, with Parotid Extract boiled, blue none
(8) 35cc. Submaxillary Extract ]
40 “ Starch Paste colorless heavy
Toluene
(9) Ditto with submaxillary extract boiled, blue none

(10) A further experiment was tried with 7 cc. of submaxillary saliva
which was allowed to stand two hours under cover before being
mixed with 5 cc. starch paste andtoluene. This mixture gave
no color with 1odine and a heavy reduction test on the following
day.

(11) A mixture of 4.6 cc. submaxillary saliva + 10 cc. starch paste +
toluene was frequently tested. After 1} hour it began to give
a faint reduction test which did not increase in the next 20
hours.

These trials indicate the nearest approach to active digestion
which we have recorded.
VI. Adaptation Experiment. A dog weighing 14.5 kilos was kept upon
a diet of bread, with a little milk and without any meat, during
4 to 6 weeks. Successive portions of submaxillary saliva were
examined alone, or with equal quantities of starch paste, with
and without toluene.

142 Amylolytic Properties of Dog’s Saliva

(1) (2) (3) (4)
Saliva 20 cc. Saliva 20 ce. Saliva 10 ce. Saliva 20 ce.

Starch Paste. Starch Paste- Starch Paste
Toluene. No Toluene. Toluene. Toluene.
Reduction test, next day. . .0117 gm. Cu faint 0.001 gm, Cu faint
lodine test, after 2 days blue blue
yp * sD) : decomposed blue
Reduction test, after 9 days none positive

There is no evidence here, in our opinion, that any adaptation has taken
place.

VII. Adaptation Experiment. <A bitch weighing 15 kilos had for months
’ been kept upon a diet of meat, cracker meal and lard, which was
regarded as very favorable for the development of an adapta-
tion of the secretion. Saliva was collected from both submax-
illary glands with varying rates of flow and intensity of stimu-
lation (coil distance). The different portions, usually about
5 cc., were mixed with equal volumes of starch paste + toluene.
After one day at 37°C. all still gave a blue iodine test together
with reduction tests on standing, which we can best describe
as varying from slight to distinct, without any apparent sequence.
Thus the most distinct reduction test was given by a specimen
obtained from the left gland with weak stimulation, while three
specimens secreted under comparable conditions from the right
gland showed almost no cuprous deposit.

VIII. The submaxillary saliva of a large dog was collected directly from
a duct by chorda stimulation. To 10 cc. portions of starch
paste an equal volume of (1) boiled and (2) unboiled saliva was
added in two flasks, with toluene. Next day both mixtures
still gave a blue color with iodine solution, 5 cc. gave a slight
reduction in each with Fehling’s solution. The deposited cu-
prous oxide was, if anything, larger in quantity in (2).

IX. Experiments with Cats. The saliva of the cat is generally regarded
as devoid of amylolytic properties.' We have collected mixed
saliva from this animal directly from the mouth by the use of
ether vapor stimulation. It fails to clear starch paste or give
evidence of amylolytic power. The possible influence of diet
has not been considered. —

We agree with Plimmer in assuming that adaptation is perhaps more
readily established in connection with functions present or dormant than
in the case of activities of which there are no direct indications in the
species.

*Cf. Cannon and Day: Amer. Journ. of Physiol., ix, p. 396, 1903.
Other references of interest in this connection are given in the intro-
ductory part of their paper.

Ne OD), eee EEE

Lafayette B. Mendel and Frank P. Underhill 143

The extracts from our protocols have been recorded here in
somewhat greater detail than they might otherwise deserve, in
order to enable the impartial observer to draw his own conclu-
sions. Our interpretation of the observations has already been
indicated. The almost uniform failure of dog’s saliva, collected
under variable conditions from the mouth cavity or directly
from the ducts, to convert starch paste completely to or beyond
the dextrin stage under the most favorable conditions of diges-
tion seems to speak against any distinctive amylolytic activity.
The more nearly positive results indicated in Experiment V
could not be obtained again. The experiments with the gland
extracts are not so significant when we recall the wide distribution
of amylases in small amounts throughout many tissues and in
the blood and lymph.* Furthermore the demonstration of a
slight reducing power must be interpreted with considerable
caution. The saliva itself was sometimes slightly effective in
this direction; and the presence of sugar in the digestions is an
uncertain inference when the quantitative data approach the
lower limits of accuracy. This is emphasized by the failure of
the slight reducing power at times observed to increase ade-
quately with the period of digestion. While these experiments
give no evidence of a specific adaptation of the salivary glands
effected by diet, the possibility of a more successful result under
still other conditions cannot be denied.

1Tf one accepts the idea of a specificity of amylases as advocated by
Ascoli and Bonfanti (Zeitschr. f. physiol. Chem., xliii, p. 156, 1904), it
might be objected that the test-starch selected was unsatisfactory. It
is, of course, also possible, though unlikely, that the canine saliva contains
enzymes which act upon soluble starch and dextrins, without converting
starch paste.

? Cf. for example, Bainbridge and Beddard: Bzto-chemical Journal, ii,

D291, 51007.

A NOTE ON THE BEHAVIOR OF URIC ACID TOWARD
ANIMAL EXTRACTS AND ALKALIES.

By PHILIP H. MITCHELL.

(From the Sheffield Laboratory of Physiological Chemistry, Yale University.)

(Received for publication, March 11, 1907.)

Various animal tissues possess the power of destroying uric
acid. Wiener,! Schittenhelm,? Burian*? and others! have shown
that this capacity varies in different species and in different
organs of the same species. Ina recent paper Austin’ gives the
results of an examination of ox spleen and kidney for a uricolytic
enzyme. He concludes: ‘“‘We have some evidence that there
is a uric acid splitting enzyme in both the spleen and kidney of
the ox; still the results obtained are so closely simulated by the
alkali used for solution of the uric acid that the proof is not posi-
five.”

Austin’s method of investigation consists in extracting the
organs with water, filtering, precipitating the filtrate with twice
its volume of saturated ammonium sulphate solution, removing
the precipitate so formed, then dissolving it in water and dialyz-
ing. The resulting enzyme solution was divided into 400-cc. por-
tions and digested with o.5 gram of uric acid dissolved in sodium
hydroxide. An equal volume of a solution of uric acid in the
same strength of alkali was used as a control. After three days
the uric acid was found to have diminished not only in the diges-
tions but to a nearly equal degree in the control uric acid solu-
tion. No controls with the boiled extracts were reported in any
of the protocols.

1 Wiener: Arch. jf. exp. Path. u. Pharm., xiii, p. 375, 1899.

* Schittenhelm: Zeitschr. f. physiol. Chem., xlv., pp. 121, 161, 1905.

3 Burian: [bid., xliii, p. 532, 1904.

* Ascoli: Arch. f. d. ges. Physiol., \xxii, p. 340, 1898. | Bendix and Schit-
tenhelm: Zettschr. f. physiol. Chem., xlii, p. 461, 1904.

> Austin: Journ. of Med. Research, xv, p. 309, 1906. Since the above
was printed Austin has published an additional paper on this topic: [bid.,
Xvi, p. 71, 1907.

145

146 - Behavior of Uric Acid

In view of these results Austin raises the question: How much
of the action of the various purin enzymes, adenase, guanase and
the uricolytic enzyme, is due to the alkali used as a solvent?
Such a criticism must certainly be considered if an excess of
alkali is used. It is true that uric acid dissolved in even the
smallest possible amount of sodium hydroxide suffers decom-
position under the conditions of these digestions. A different
condition exists, however, in the presence of proteid. Austin
suggests that in an albuminous solution, the formation of an
alkali-proteid combination may protect the uric acid, although
he recorded no experiments to illustrate this point.

The following simple experiment of the writer speaks for itself:

(a) 0.15 gram uric acid suspended in warm water was dissolved by
adding 4 cc. of 2 per cent sodium hydroxide solution and made up to
400 cc.

(6) 0.15 gram uric acid was similarly dissolved, the diluted, filtered
white of one egg was added, and the solution made up to 400 cc.

(c) 0.15 gram uric acid was dissolved in 40 cc. of 2 per cent sodium
hydroxide solution and made up to 400 cc.

To all the solutions considerable amounts of toluene were added and all
were digested at 40° C. for three days. Air was drawn through during
30 hours of that time and toluene was renewed in each digestion daily.
(a) and (c) were then treated with ro cc. of ro per cent hydrochloric
acid and concentrated to 15 cc. (b) was heated to boiling, coagulation
effected by adding dilute acetic acid, the fluid filtered, treated with ro cc.
of ro per cent hydrochloric acid and concentrated to 15 cc. The uric acid
was then allowed to separate by crystallization and filtered on a Gooch cru-
cible. It was washed with alcohol and ether and dried at 100° C.

Making correction for the solubility of uric acid the amounts recovered
were: (a) 0.1085 gram; (b) 0.1405 gram; (c) no crystals of uric acid. Con-
sidering the difficulty of removing uric acid from a proteid coagulum
(b) may be regarded as a practically quantitative recovery. The sepa-
rated crystals were submitted to Kjeldahl nitrogen estimation.

0.1200 gram of the material required 13.4 cc. of acid (1 cc. = 0.00296
gram N).

Found: N =33.03 per cent; calculated: N = 33.33 per cent.

The solutions were quite clear at the end of the digestion, show-
ing that the uric acid had remained in solution. Any acidity
which might be assumed to develop in an autolysis of animal
organs and protect the uric acid, could not be present here; for

1 Cf. Sundwik: Zettschr. f. physiol. Chem., xx, p. 335, 1894.

Philip H. Mitchell 147

the alkaline eggwhite does not readily autolyze. The experi-
ment illustrates the destructive action of excess of alkali, and
the protective influence exerted by proteids.

Experimental studies which the writer has been conducting
for some time with Professor Mendel! on the uricolytic power and
purin enzymes of embryonic tissues has made it necessary to
consider carefully the adverse criticism of Austin. In the light of
our experience, however, his conclusions seem unjustified. For
example, we have found: (1) that uric acid is not destroyed by
extracts of embryo pig’s livers; (2) that, under comparable con-
ditions, uric acid is destroyed by extracts of adult pig’s liver, and
(3) that uric acid is not destroyed by a boiled extract of adult
pig’s liver.

Metuop: Livers obtained fresh from the slaughter house
were macerated and extracted with five times their weight of
toluene water during 24 hours with frequent stirring. To 500
ec. of the filtered extract, 0.15 gram of uric acid was added
after being dissolved in the amount of sodium hydroxide necessary
to form disodium urate. The solutions were then digested at 38°
C., air being drawn through for periods varyingin different experi-
ments. The digestion mixtures were precipitated by heat with
addition of acetic acid and the coagulum was redissolved in dilute
alkali and reprecipitated with acetic acid. The filtrates were
united and precipitated by the Kriger-Schmid’ method with
copper sulphate and sodium bisulphite. The precipitate of
purin bodies so formed was decomposed with sodium sulphide,
acidified with acetic acid, and the filtrate from the copper sul-
phide was concentrated to 20 cc. after adding ro cc. of 10 per
cent hydrochloric acid. If the uric acid did not crystallize in a
pure condition, without color and admixture of foreign materials,
it was redissolved in concentrated sulphuric acid, according to
Horbaczewski,? and recrystallized by pouring into alcohol. The
uric acid was weighed in a Gooch crucible after washing and dry-
ing at 100° C.

Some idea of the relative age of the embryos was obtained
by measuring the length of the body. Only the largest embryos

1 With the aid of a grant from the Carnegie Institution of Washington,

2 Hoppe-Seyler-Thierfelder: Handbuch der chem. Analyse, Pp. 435.
3 Horbaczewski: Zeitschr. f. physiol. Chem., xviii, p. 341, 1893.

148 Behavior of Uric Acid

available in sufficient quantity for experiments were used. Their
length varied from six to eight inches, corresponding approxi-
mately to a foetal age of eleven to thirteen weeks.

Experiment 1. Eight-inch embryo livers extracted as above. 500 CC.
digested three days with o.15 gram uric acid dissolved in the smallest
possible amount of sodium hydroxide. 0.139 gram uric acid recovered.

Experiment 2. Seven-inch embryo livers similarly treated. 0.137 gram
uric acid recovered,

Experiment 3. Seven-inch embryo livers similarly treated. 0.122 gram
uric actd recovered.

Experiment 4. Seven-inch embryo livers similarly digested four days.
0.127 gram uric acid recovered. This material was not collected on a
Gooch crucible, but weighed in a drying bottle and subjected to Kjeldahl
nitrogen estimation. °

0.0820 gram of substance used 9.15 cc. acid (1 cc. = 0.00296 gram N).

Found: N =33.02 per cent; calculated: N =33.33 per cent.

This shows that the uric acid recovered was pure. In all cases
the murexide test was positive.

In contrast with this recovery of uric acid in embryo liver
digestions are the following experiments with adult pig’s liver.

Experiment 5. (a) 500cc. extract of adult pig’s liver prepared as above
plus 0.15 gram uric acid digested one day. 0.048 gram uric acid recovered.

(b) 500 cc. of same extract similarly digested 3 days. 0.021 gram
uric acid recovered.

(c) 500 cc. same extract, boiled, plus 0.15 gram uric acid digested
three days. 0.139 gram uric acid recovered.

Experiment 6. (a) and (6) prepared as (a) and (b) of preceding experi-
ment and digested four days. No trace of a substance giving the murex-
ide test was obtained.

Austin demonstrated as much uricolytic activity in ox spleen
as in ox kidney; while Schittenhelm' was unable to detect such
power in ox spleen. This discrepancy may be due to differences
in the alkalinity of the digestions. An examination of Austin’s
protocols shows that in every case he used more alkali than was
required to dissolve the uric acid. This excess of alkali would
find in his purer enzyme solution less proteid to combine with
than in a simple extract of tissues. Again, in a long process of
dialysis and digestion there is considerable opportunity for bac-
terial contamination. In this connection, it may be noted that

1 Schittenhelm: Zeitschr. f. physiol. Chem., xlv, p. 121, 1905.

Philip H. Mitchell 149

in one of Austin’s experiments, tyrosin, regarded as a product of
incipient putrefaction, could be detected.

It must be admitted that the methods of testing for an urico-
lytic enzyme, involving as they do somewhat variable factors of
alkalinity and possibility of bacterial action during a prolonged
digestion, are not ideal. But experiments such as those planned
by Austin in solutions poor in proteid and rich in alkali need not
necessarily call into question the results obtained by the current
methods. For although uric acid is destroyed by adult pig’s
liver, it is not changed by extract of embryo pig’s livers or by a
boiled extract of adult pig’s liver under precisely comparable
conditions of alkalinity, antisepsis, etc.

It seems more reasonable to attribute such specific differences
to the variations in the enzyme content of the organs examined
than to external factors.

MANGANESE, A NORMAL ELEMENT IN THE TISSUES OF
THE FRESH WATER CLAMS, UNIO AND ANODONTA,

By HAROLD C. BRADLEY.
(From the Laboratory of Physiological Chemistry, University of Wisconsin.)
(Received for publication, March 29, 1907.) =

Manganese in the tissues of normal animals may be regarded
asarareelement. In traces it is known to be present in a num-
ber of marine animals, as was shown by Pichard.1. Hoppe-Sey-
ler? is authority for the statement that it is not infrequently
present in human blood and in the liver: but it is neither constant
in its occurrence nor significant in amount. Among the lower
animals there is at least one instance on record of the normal and
physiological presence of manganese in the blood and tissues.
In the blood of the mollusk, Pinna squamosa, manganese is
believed to serve arespiratory function. According to the analy-
ses made by Griffiths,? manganese is present in the blood to the
extent of 0.35 per cent. So unique was this discovery, however,
that it has received little more than the passing interest given a
curious fact devoid of any real physiological interest. We have
to report in this preliminary notice what seems to be another nor-
mal example of its use in the metabolism of the fresh water clams,
Unio and Anodonta, so abundant in the lakes and rivers of the
Mississippi basin. While our specimens were obtained from the
Madison lakes of Wisconsin only, they are thoroughly represen-
tative of the two species in every other respect, so that we are
confident that specimens from other localities will exhibit this
same metabolic idiosyncrasy. This point, however, will be defi-
nitely settled as soon as more material can be obtained.

While the qualitative recognition of manganese in tissues con-
taining such considerable amounts as are present in this animal

1P. Pichard: Compt. rend. de l’ Acad. des. sci., cxxvi, pp. 550 and 1882,
1808.

? Hoppe-Seyler-Thierfelder: Handbuch der chem. Analyse, p. 39.

3 Griffiths: Compt. rend. de l’ Acad. des sct., cxiv, p. 840, 1892.

I51

152 Manganese in Unio and Anodonta

is not difficult or uncertain, we have found the standard quanti-
tative methods peculiarly inappropriate for the conditions of
this research. The large amounts of calcium, magnesium, and
phosphorus in the ash render the gravimetric separations tedious
and uncertain. Among the volumetric methods, that of Volhard
is easily the best, though as ordinarily carried out that too is un-—
necessarily slow. We have, therefore, modified the Volhard pro-
cess in such a way that while the accuracy of the determinations
is not seriously affected—if it is lessened at all—the time con-
sumed in carrying them out is materially shortened. The details
of the process are given in the hope that they may be tried out
or criticised by chemists more familiar with the practical esti-
mation of manganese, and also that biologists and biological
chemists may be led to take up this problem in other localities.

The tissue is first reduced to a carbon-free ash of stable weight and com-
position. It has been found that the ashing of considerable amounts of
the dry tissue in ordinary crucibles consumes a great deal of time; the car-
bon tends to assume the graphitic form and as such is but slowly oxidized,
especially in the center of the ignition mass. We have, therefore, sub-
stituted the common clay pipe for the crucible, with excellent results.
The bow] is filled with the ground tissue and heated over the blast lamp.
The stem is connected by a T tube to the air blast, and a stream of air
thus led through the ignition mass from the bottom. The amount of air
may be so regulated by a stopcock that a rapid and intense oxidation goes
on in the ash, accomplishing in a few minutes what requires a long period
of time in a crucible under the most favorable conditions. The ash now
consists chiefly of oxides, phosphates, carbonates, and sulphates of cal-
cium, magnesium, manganese, sodium and potassium. It is moistened
with nitric acid and ignited in a crucible—thus removing the carbon
dioxide and leaving an ash of constant weight and fairly uniform compo-
sition.

The ash, pulverized and sampled, is weighed out for analysis. From
1.0 too.5 gm. are taken asarulée. This is dissolved in 5 cc. of concen-
trated hydrochloric acid to which about 1 cc. of nitric acid is added. The
mixture is boiled till the bulk of free chlorine is removed and the solution
is a clear pale yellow. If the ashing process has been complete no insolu-
ble residues will remain. The solution is then transferred quantitatively
to a graduated flask and diluted to the mark. A convenient dilution is
found to be 250 cc., though a roo cc. flask or a 500 cc. flask may be used
equally well. Aliquots are then taken and transferred to Erlenmeyer
flasks or covered beakers and an excess of pulverized zinc carbonate is
added. The mixture is quickly brought to a boil, the spatters on the
cover washed back into the beaker, and the manganese precipitated as

Harold C. Bradley 153

manganese dioxide by titrating with 4, potassium permanganate till a

permanent pink indicates the completion of the reaction. In practice the
potassium permanganate is run into the first aliquot a few cubic centi-
meters at a time with stirring, until the reaction is nearly complete—and
this can be judged quite accurately by the color of the solution. The
contents of the beaker are then brought to a boil again. The hydrated
manganese dioxide separates out at once as a brown precipitate and
carries down with it the suspended zinc carbonate, leaving a clear super-
natant liquid. The final additions of potassium permanganate are made
drop by drop till the end point is reached. The second aliquot is added
to the first with more zinc carbonate, brought to a boil and titrated rapidly
as before. If the first two aliquots do not agree to o.2 cc. a third is taken.
Finally the total potassium permanganate used for the sum of the ali-
quots is made the basis for calculating the amount of manganese present
in the sample of ash.

If a 3, solution of potassium permanganate is used the calculation of
results is very readily made. From the reaction of potassium perman-
ganate in neutral solutions, it is evident that 1 cc. of the standard solu-
tion is equivalent to 0.00165 gm. of manganese.

EXAMINATION OF THE METHOD. The points of departure from the origi-
nal Volhard process are in omitting the concentration of the solution with
sulphuric acid, in substituting zinc carbonate for zinc oxide and in omit-
ting to filter off the suspended zinc carbonate or zinc oxide before per-
forming the titration. Further, the acid is neutralized only in the
aliquots taken instead of in the entire solution. Since the ash is readily
soluble in hydrochloric acid or aqua regia, there seems no advantage to be
gained from the evaporation with sulphuric acid and its omission saves
time. The substitution of zinc carbonate for zinc oxide has an advan-
tage in that the degree of acidity can be judged at once by the efferves-
cence, while this can be told much less readily when the oxide is used. If
ordinary precautions are taken during neutralization and the subsequent
heating, no loss by spattering need be experienced; the neutralization is
rapid, complete, and the excess of the carbonate does not in any way
affect the results, so far as we have been able to observe. The omission
of the filtration is obviously advantageous if the end point can be deter-
mined accurately in the presence of the suspended zinc carbonate. Our
experience leads us to believe that the end point is quite as sharp with the
zine salt there as when it is absent, while its presence insures the complete
neutrality of the solution—a most important point in this reaction.

Finally, however, we must judge of a process by its actual performance
under varying conditions, and while we have not subjected this to any
exhaustive series of tests, the following table will indicate its general
reliability under the conditions likely to be met in this work. No attempt
has been made to carry out the tests of the method more accurately or
carefully than the subsequent determinations on ash samples, since we
wish to know not so much the accuracy of the method when employed

154 Manganese in Unio and Anodonta

with extraordinary care and pains, but its limitations and the magnitude
of its errors when applied to every-day analyses, with every-day speed and
care. The figures, therefore, indicate its reliability when pushed to the full
speed of routine determinations.

The samples of manganese were taken by measuring out definite por-
tions of a standard potassium permanganate solution. This was decom-
posed by boiling with hydrochloric acid till pale yellow. The solutions
were then made up to known volumes and analyzed as described before.

The effects of various amounts of zinc carbonate were tried, and the
effect of varying the dilution. The amount of hydrochloric acid was kept
fairly constant. A blank determination was first carried out to deter-
mine whether the zinc carbonate and other reagents used were free from
reducing compounds. A single drop of the potassium permanganate
solution sufficed to give a permanent pink to the blank titration.

The results are all low and while the errors are considerable they are all
of the same order of magnitude and in the same direction. There is no
appreciable effect of varying the dilution or the amount of zinc carbonate.
The ultirnate error is on the average no greater than the normal variations
in samples. We may conclude, therefore, that the method as outlined is
reliable enough for the purposes of proximate analysis. The figures in
Table II will be uniformly low, but where from 1.0 to 0.5 gm. of ash are
taken, the error will approximate o.2 per cent of the total manganese—
an amount of no significance where the normal individual variations are
so large among the specimens.

Having established the efficiency of the method for our analy-
ses, a series of determinations was made on the samples obtained.
To establish the normality of the manganese in these species of
mollusks, the following questions must be definitely answered:
1. Is manganese always present inthe animal’s tissues? 2. Isit
present in significant amounts and subject to variation within
physiological limits? 3. Is it a characteristic that is trans-
mitted from generation to generation—that is, do the eggs or
young exhibit the same idiosyncrasy? These questions must be
solved as a preliminary to the further investigation of the physio-
logical functions of the element in the metabolism of these mol-
lusks.

The specimens were obtained from different localities of the
Madison lakes. The entire tissues of from 12 to 24 specimens
were dried and ground up for each of the samples I, II and III.
The ovaries of the females were found to be full of eggs and these
could readily be separated from the other tissues. In the other
samples, therefore, the eggs were removed and analyzed sepa-

155

Harold C. Bradley

03‘ &T i 00
(0) |! GG000°—| SPrrSO° 00 °§€ 00¢ 0g¢0°
02 ST 6 006
¢9°9 6 00%
9°0 Z1000-—|| S82700% 9¢ OT 00¢ GLZ0°
09°9 G 00G
¢9'°9 14 OOT
9°S 69000°—| T8920 cS ‘OT 093 ¢LZ0°
cg'9 1 OOT
0Z°8 v 0g
9°T ¥F000°' —| 90260" OF OT OOT GLE0°
OL'F G SG
00°8 G O¢
FG 99000°—| #8920" 9% OT OOT C160
’ 02 'F I cS
Oo. 8 i 0g
8°S 1L000°—| €2920° 02‘ 9T a a OOT GLZO°
= 2 G
0) <3 G 0g
0°S ¢¢g000°—| $6960" €e OT OOT CL60°
cl’? T GG
“sub “00 99 ‘sub 00 00 "sub
*u9yB} Uy, “SUN ;
~ OOT X 101g, oe punew poyentorey oun “x! *PPPPV | -syonbyy | ‘wound cae
ut LOLI esoUBSUBI, YOUNM “UOTPBIYLL, £O00Uuz : r a osoUBdUBIY

*1OLIG, OFBYUDDIOT

0g

GS

‘09

“uaBy

YOUN M

OL
N

‘ON

‘I GTaViL

156 Manganese in Unio and Anodonta

rately, the remaining tissues dried, mixed and pulverized as
before. Each of the samples IV and V represent the average
composition of the tissues of some two dozen clams, minus the
eggs of the females present. Samples III, IV and V were sub-
jected to a starvation period of two, four and six weeks, respec-
tively. This should insure the elimination of any food residues
containing manganese from the alimentary tract, which might
otherwise introduce a doubt as to the real significance of the
metalin I and II. Separate determinations of the ash, content
of each sample were made in platinum, igniting to constant
weight after moistening with nitric acid. The results are sum-
marized in the table below.

TABLE Il.
M
gample.| Poioatcitsh | PerCent of Mo | Fes Console) Remark
I 9 21.63 4.20 0.931 freshly collected
4.20
aa Il 26.00 4.55 1.19 freshly collected
4.60
Ill 21.94 4.35 .929 starved two weeks
4.24 |
at
| 4.03
IV 17.88 5.76 1.02 _ starved four weeks
| 5.64 |
5.82
V 13.88 4.24 .601 | starved six weeks
4.42
eggs
lV 39.55 1.49 .633 | starved four weeks
1.69 - |
a 35.20 2.32 .818 | starved six weeks

Decided variations in the ash content are found. In the first
three samples this is not significant since the eggs with their nor-
mally high percentage of inorganic salts were not separated.
Samples IV and V, however, seem to indicate the steady elimi-
nation of inorganic material during starvation. Up to the last
sample when the animals were reduced to a pathological condi-
tion by long starvation and the presence in the vessels contain-

Harold C. Bradley 157

ing them of their excretory products, no decided fluctuation of
the manganese was observed. |

We may conclude, therefore, that so far as the specimens from
the Madison lakes are concerned, manganese is a normal con-
stituent of the tissues; it has always been present in the specimens
examined, it is present in considerable quantities and varies
within comparatively narrow limits—well within the limits of
variation of the total inorganic constituents of the tissues. It is
also present in the eggs. Further extensions of this investi-
gation are contemplated when material is available, with a view
to determining variations produced by local environment, the
sources of the metal in the food of the clams, and the physiologi-
cal functions of the compounds containing the manganese in the
tissues.

rao
ah

Me Le

a

THE QUANTITATIVE ESTIMATION OF EXTRACTIVE AND
PROTEIN PHOSPHORUS.'!

By W. KOCH.
(From the Pathological Laboratory of the London County Asylums.)

(Received for publication, January 16, 1907.)

The different combinations of phosphorus to be found in a given
tissue may be divided into three main groups:

1. Protein phosphorus or phosphorus in combination with
protein, including nucleoprotein and phospho-proteins or nucleo-
albumins, insoluble in water especially after treatment with
alcohol;

2. Lecithin and kephalin phosphorus or phosphorus in com-
bination with fat and a nitrogen complex, soluble in alcohol and
ether, but insoluble in acid chloroform water;

3. Extractive phosphorus, including inorganic phosphates and
the simpler combinations of phosphoric acid, such as glycero-
phosphoric acid, phytin or diethoxy-diphosphoric acid, and a
number of related compounds as yet little investigated, all of
which are soluble in water and partly soluble in dilute alcohol.

In a previous paper”? a method for the estimation of lecithin
and kephalin phosphorus was described. In the following pages
are given methods for the determination of extractive and
nuclein phosphorus, which can be carried on at the same time
and with the same material as the lecithin estimation.

EXTRACTIVE PHOSPHORUS.

A considerable portion of this form is found in the filtrate from
the lipoids precipitated with acid chloroform as described in the
above mentioned publication. Whether any of this phosphorus

1 These methods were used in the investigation with H. S. Reed, published
in vol. iii, p. 49, of this Journal.
*Koch and Woods: This Journal, i, p. 203, 1905.

T59

160 Extractive and Protein Phosphorus

is inorganic cannot be determined. Schulze! in several of his
publications emphasized the fact that absolute alcohol and ether
do not dissolve inorganic phosphates. He is dealing, however,
with relatively dry plant tissues and not with moist animal tissues
which necessarily dilute the alcohol. The separate estimation of
inorganic phosphates has not been attempted in these methods, as
the danger of hydrolyzing simple organic combinations of phos-
phoric acid seemed too great to promise reliable results.

The following table gives an idea of the amount of phosphorus,
not lecithin or kephalin, dissolved out by alcohol and ether from
brain tissues.

TABLE I.

PHOSPHORUS IN FILTRATE FROM LIPOID PRECIPITATE. ALCOHOL-ETHER-
WATER-SOLUBLE PHOSPHORUS.

Number of In Per Cent of In Per Cent of In Per Cent of
Case. Dry Tissue. Total Extractive P. Total P.
349 0.68 37.8 4.8
359 0.63 San 4.5
449 0.75 36.0 5.4

The remaining portion of the extractive phosphorus is to be
found in the portion of the tissues insoluble in alcohol and ether
and must be removed by treatment with water to which a little
chloroform has been added to prevent bacterial action. Noél
Paton? recommends dilute acid for this extraction, but does not
make it clear whether he altogether avoids the possibility of
breaking up more complex substances. Control experiments
have shown that in the case of the brain about five or six extrac-
tions are sufficient to remove all the phosphates that can be
removed. The following table gives the results.

TABLE Izl.

WATER-SOLUBLE PHOSPHORUS IN RESIDUE INSOLUBLE IN ALCOHOL AND
ETHER. ALCOHOL-ETHER-INSOLUBLE, WATER-SOLUBLE PHOSPHORUS.

In Per cent of ~ In Per Cent of In Per Cent of
Case. Dry Tissue. Total Extractive P. Total P
349 1 a 62.2 7.8
352 1.30 67.9 9.5
449 1.308 64.0 9.6

‘Schulze, E.: Zeitschr. f. physiol. Chem., xx, p. 225, 1904.
? Noel Paton, D.: Report of Investigations on the Life History of the
Salmon in Fresh Water. Fishery Board for Scotland, Pp. 143, 1898.

W. Koch 161

PROTEIN PHOSPHORUS.

The phosphorus compounds present in the tissues after ex-
traction with alcohol, ether and water can only be nucleins,
phospho-proteins and tricalctum phosphates. The latter com-
pound is not usually supposed to be present in appreciable amount
in tissues except under pathological conditions and can therefore
be neglected in the case of brain tissues. If calcium is present
it would be more likely to exist as a calcium protein compound.
Extraction with dilute acid might be used where calcium phos-
phate is suspected but this procedure so swells the tissues that
complete removal of the adhering liquid becomes very difficult.
Besides there is the danger of rendering the alcohol-coagulated
protein again soluble.

The following table gives some of the results:

TABLE IIl.

PHOSPHORUS IN INSOLUBLE RESIDUE. ALCOHOL-ETHER-WATER-INSOLUBLE
OR PROTEIN PHOSPHORUS.

In Per cent of In Per Cent of
Case. Dry Tissue. Total P
349 0.81 5.6
359 0.86 6.1
449 0.85 6.1

A comparison of Tables I, II and III will show that about 80
per cent of the total phosphorus remains to be accounted for.
This is represented by lipoid phosphorus which, in the case of
corpus callosum here analyzed, is present in large amount.

DESCRIPTION OF METHOD.

About 10 grams of the moist tissue are extracted with alcohol
and ether as directed in the paper on the “Estimation of the
Lecithins.”’ The residue, insoluble in alcohol and ether, is dried at
102°C. to constant weight, transferred to a 300 cc. Jena flask and
extracted six times with about 100 cc. of water to each extraction.
Every extraction should extend over 24 hours; plenty of chloro-
form must be added and the mixture occasionally shaken to
prevent bacterial decomposition. The filtrates are evaporated
in a platinum dish and dried to constant weight. The residue
represents the salts and extractives, insoluble in alcohol and ether

162 Extractive and Protein Phosphorus

and soluble in water. The dried residue is ignited in the platinum
dish, surrounded by an outer larger platinum dish which is heated
to bright redness, until a nearly white ash is obtained. If the
inner dish does not come in direct contact with the outer dish
there is no danger of volatilizing chlorids. This residue is the
alcohol-ether-insoluble, water-soluble ash. The difference between
this ash and the residue on evaporation gives the alcohol-ether-
insoluble, water-soluble, organic extractives. The ash is moistened
with 1.5 cc. of nitric acid, dissolved in water, diluted to 100 to
200 cc. and phosphorus estimation made by the molybdate
method. This gives the alcohol-ether-insoluble, water-soluble
extractive phosphorus. (Table II.)

The residue left above, insoluble in water after six extractions,
is burned with nitric and sulphuric acids and the phosphorus
estimated. Incase calcium is present this must also be estimated
in a separate sample. The phosphorus method is described in
detail in a previous paper.!. This phosphorus is called the alcohol-
ether-water-insoluble, or protein phosphorus. (Table III.)

The alcohol and ether solutions obtained by the extraction
of the moist tissue are treated as directed in the paper above
referred to. If the emulsification and precipitation have been
properly carried on, the solution in the 100 cc. graduated flask
should be clear in two or three days. An excess of fat in the
tissue interferes seriously with this clearing and had best be over-
come by the presence of a large amount of chloroform (8 to ro cc.)
and the addition of 2 cc. instead of 1 cc. of hydrochloric acid.
The amount of chloroform added must be carefully measured and
recorded. After the solution has begun to clear and has been
made up to the 1oo cc. mark of the graduated flask, it is shaken
and allowed to stand until the precipitate has settled. After
settling the solution is filtered through a dry filter paper into a
dry roo cc. graduated cylinder. As much of the water as possible
is decanted from the chloroform, but it is better not to pour any of
the chloroform on the filter, as it may pass through and lipoids be
thus lost. Instead of washing the chloroform containing the
lipoids with acid water as previously directed, it is better to
allow the filter to drain and then read the volume of the filtrate,

1Koch and Woods: loc cit.

W. Koch 163

which should be perfectly clear and transparent. An aliquot part
of the filtrate, usually 80 cc., is-evaporated ina platinum dish and
dried to constant weight at 102°C. This gives the extractives and
salts soluble in alcohol, ether and water. The residue is ignited as
directed above and the ash is the alcohol-ether-water-soluble ash
while the difference between this and the residue obtained at
102° C. represents the alcohol-ether-water-soluble extractives. The
ash is again moistened with 4 cc. of nitric acid, diluted to 100 to
200 cc.and phosphorus estimated. Thisis the alcohol-ether-water-
soluble extractive phosphorus. (Table I.)

DISCUSSION AND CALCULATION.

In order to illustrate the method of calculation it is best to take
a sample analysis as follows:

RECORD OF ANALYTICAL RESULTS.
Case 342 Corpus Callosum.

Weight of sample............. 9.9038 grams
VIG eee A ee 70.37 per cent Weight of residue insoluble in
alcohol and ether........... 0.8886 “*
1. Insoluble residuet ..8.68 ‘“* ‘“ Weight of residue from six
extractions ees eee oe 0.0289 *“*
De CCLUNINS: sea srcs se 4.00 “ “ Weight of residue, insoluble in
: alcohol, ether and water..... 0.8597 *
oe Mephnalins) ..3..2..4.20 Phosphoric acid in this residue
: BBV Esc. ala: ofl sb baie exw 0.0085 *
4a Extractives....... OnI5. = & Mg2P207
5a Extractives ..... A030 ss te Residue from six water ex-
CLACHIONS 25 ca eceusie (ola Biewsnes 6 0.0289 grams
ADWPASH ios. e:s;h00 avec sae Ostay 5 4b Ash on ignition........... 0.0142 “*
DASH Pareie sé) -cte e <td 0.45 “° i
5 4a) extractives: so. oeee 0.0147
6 SulphurCompound*1.40 “
Phosphoric acid gave.......... OPFOR ee
MgoP207
Proteim! PP) .52sAs- 0.024 per cent.

fhotalvbPxtract, baer O.0bS) 5) os

To precipitate lipoids 3 ce. chloroform were used. The filtrate measured 76 cc.; 70 ce.
were evaporated.

MCSICIO MRE TOE ts GAs Mates clas Mase ote OS ayes oelota c cine bd ee DAS 0.1060 grams
SRE NS ETM OMIME EDL UL OL eee errr. = toys ere ety NOIRE Renee ng = deare eese 20325 ene
PERE ISULACULVES erates mot Cnr oiefeie rotsieus. <Patee sere ak tener oie ava Spee eGo OL073 ques
Phosphoriciacidin this residue gave: *......00.0.s.+2es.0s% 0.0051 “

ok Mg2P207

2, “Uifeeridatia Leesa cee ete ee oa oe Ee eee Bee eo 0.0551 grams
3) USGA RE eto are ele Bane CEI e eater 0.0697 “*

* Separate estimation. ‘
: + This essentially represents the total proteids and any glycogen that may be present in
the tissues.

The method of calculating the above results is as follows:
sa and sb alcohol-ether-water-soluble extractives and ash are calcu-
lated to 97 cc. (thus correcting for the volume of the chloroform)

0.0323 X 97 X I00

x I00
peel NODS £2) oy 03:

; 0.45
70 X 9.9038 70 X 9.9038

164 Extractive and Protein Phosphorus

Alcohol-ether-water-soluble, extractive phosphorus.

0.0051 X 97 X 100 X 62
— = 0.020

70 X9.9038 x 222.7

The same method applies to the calculation of the alcohol-ether-water-
soluble phosphorus. The filtrate obtained was only 74 cc., but as only
5 ec. of chloroform were used, the remaining watery solution must have
remained clinging to the rather spongy mass of fat and chloroform. Some
little solution goes to moisten the filter paper, but this partly counter-
balances whatever chloroform may have gone into solution. This method
of calculation is not absolutely accurate, but comes sufficiently near, con-
sidering the variations to which the material is liable in any case.

3. The kephalin receives a correction for the amount of phosphorus
in the liquid clinging tothe lipoid precipitate, in this case 95 - 74 =21 cc.,
equivalent to 0.06 milligram of phosphorus or 0.08 per cent of kephalin.

Theoretically the correction should be distributed between the lecithin
and kephalin, the values of both of which it affects. Asitisnotimprobable,
however, that the water-soluble phosphoric acid derivatives with which
we are here dealing form insoluble lead salts in ammoniacal alcohol solu-
tion, it was deemed best to apply the whole correction to the kephalin.
The kephalin in the case of brain tissues receives a further correction for
the phosphorus found in the sulphur compound.

In the light of later results it was found advisable to change the method
for the estimation of kephalin outlined in a previous paper as follows:

This Journal,i, p. 208, line 7. Substitute: allow to remain on water-bath
until there isno more smell of ammonia. The flask is then set aside to
cool, after replacing the alcohol which has evaporated.

p. 208, line1z. Read: and the precipitate washed once with hot alcohol.

p. 208, line 21. Experience has shown that it is preferable to burn the
filter paper with the precipitate.

In the muscle tissues of animals containing a solid fat the correction for
the kephalin on account of filtrate adhering tolipoid precipitate may
become rather large. It can, however, be easily reduced by making the
precipitate in a larger flask (200 cc. or 4oo cc.) and thus diluting the
adhering filtrate.

The method above outlined should be capable of more general
application to normal and pathological material. Sucha study
on the nervous system is soon to be published. This investiga-
tion was aided by grant from the Rockefeller Institute for
Medical Research.

DAY AND NIGHT URINES.
By E. OSTERBERG anp C. G. L. WOLF.

(From the Chemical Laboratory, Cornell University Medical College,
New York City.)

(Received for publication, March 1, 1907.)

The investigation of the excretory relationships during the
hours of sleep and of waking was one of the earliest problems
attempted in metabolism. Schweig,! in 1843, endeavored to
follow out the course of uric acid during eight-hour periods, and
came to the conclusion that this substance was eliminated in
least quantity during the night, and most during the midday
eight-hour period. Vogel? followed the hourly excretion, and
found that less urine was passed during the night than during
hours of activity. In commenting on Vogel’s results Speck?
attempted to show that the increase in the excretion during the
day was due to the intake of fluids during this time. On the
other hand, Bécker* observed a nocturnal polyuria, and asserted
that the elimination by the skin was more profuse during inac-
tivity.

Of the more recent work on this subject one has the observ-
ations of Quincke,® who determined that the urine was without
exception more abundant during activity. In following the
elimination, he found that excretion was most pronounced dur-
ing the first three hours of waking. He explains his results as
being due, either to decreased resorption of fluid from the intes-
tinal tract during sleep, or to retention in the organs, or that
the kidney secretion is itself diminished, either from lessened
blood supply, or contraction of the vessels. His results have

1 Schweig: Cited by Speck.

* Vogel: Cited by Speck.

3 Speck: Arch. f. exp. Path. u. Pharm., xv, p. 109, 1881.
* Cited by Speck: Ibid., p. 108.

5 Quincke: Ibid., vii, p. 115, 1877.

165

166 Day and Night Urines

since been confirmed by Breisacher,! Lahr’? and others. Air
appears, however, that in certain affections such as renal and
cardiac disease, one may have a nocturnal polyuria which is very
distinct.* ;

As far as we are aware no complete analyses of normal urine
in which the nitrogen elimination has been compared with that
of sulfur are on record. In the course of an investigation of
pathological urines in which it was often difficult to obtain com-
plete samples of the twenty-four hour excretion, we thought it
of interest to make complete analyses of twelve-hour periods
under ordinary conditions of diet. We hoped in this way to
obtain some data of comparative value which were not available.
For that reason the subject was not put on an accurate stand-
ard diet. In the first series he partook mainly of cereal food, and
in the second, a fairly liberal amount of lean meat was eaten.
The collection of the urines was made in alternating twelve-hour
periods, during one of which the subject rested in bed, and in the
other performed ordinary laboratory work. It was thought
advisable to collect the urines three hours after the period had
been finished, in order to compensate for lag in excretion.

The recent work of Leathes on the diurnal variation in the
excretion of uric acid was published some time after these experi-
ments were performed, and as our results corroborate those
obtained by him, it seemed worth while to publish the following
analyses of day and night urines, although it is realized that the
series is not as complete as it might be. We hope on a future
occasion to go into the matter in greater detail.

The subject of the experiments was a man, forty years of age,
weighing seventy kilograms. The amount of exercise consisted
in a walk of about four miles each day, and eight hours laboratory
work. The heaviest meal of the day was taken at 7 p. m., and
the rest period was from 8 p. m. to 8 a.m. The urines were
collected at 11 p. m. and 11 a. m.

The following tables give the results of the analyses of the
urine.

1 Arch. ¢. Physiol., p. 321, 1891.

2 Arch. }. Psychiatrie, \xvi, p. 286, 1899.

3 Deutsch. Arch. f. klin. Med., 1\xviii, P- 175, 1900; N. Y. Med. Journ.,
February, 1904; Munch. med. Wochenschr., No. 52, 1896.

M™
2a
No) ep ee PSR Cs ales A Sg aS ee elec Wey. We)
AY i NPM 4) forecast] PO DSHRADOASHHOHAS
mn UueZOIJIN pouTUIejopuL) Big FO 03.00.09 HO OBOE SL? SN) AQRSOHRASSORHRAYHHSS
‘ : SMONMWOHSROHOHOGON TOR FUEN || aes Gia ON ei Gy Sinois
2 — RS Se QS Se i Soe
toe eat oe SP Se Se :
eee poe ou SaASwSASXSSSAS YAN
os ; ‘ . | Ssesssosonossoen Sei OSea hig 08 1g SI SER AHID © Was SIG etiGo
o) = 4Og ‘ry | CM SCRHORHONHOHOHROHSS
= : Spy Epo ao tele ee Ree eS SRSORSCKROHSHSOOHOR
mg ANE ins SEN SoSmHSHSHSwSn Sane fe
e “YP UsFOIQIN, UlPBAT) ns = = -_—
rca ea esesesesesoesesss 1 B
DS a a
oO Te yp
. "OT he was q N ~~ ie) ~
GS Fx Pies 2 8 38 5 qe 08 y Pee eee eee e ee
| . ™~ i.)
mo ueFOA}IN ULUTPeAID Wig OAS AOARASAR A'S [2 *OS ‘IRV eee ee eee
: aS ms 1 OMNSMOMOCHOHOHORNOR] | OE | Sgoeceos pine, Se =f
(©) a ese
Oi -— 1 -- P =
ae Seale oy etal S 2 SS 8 BD R ao |
a ee seein FRIGABRNBDONGSAS OWS Seis ewes | corer. cei edie stie es oom tee
« >P 2 aa HSH ig di StGomnSoonnrs)| & ee SL + YOS | TIM SARONROHOOSODS
~
on oa CgmmOO I CG 1CO NGO N05) 00, u'og cals 4G [IGT les otras Soe veut camtacaneee eras
HS er ted CMORSOHOHSOHSHSSOHOS
. = ie : aor I = © 0 19 10 st = fleet Ses cea Colnoetc oh ie
oo cae, ee Beye RA AHAAARABARIOG BY go 3G
= nies ued0.jIN BluomMUIy | “. 27. oT. so. ae ~ —
Me i ae | SHONSRNSOHSHOHOGOH | cE = ar oe
Ee ae z A, ome aA SO D
© ‘Be Seoog wt th Oo a ‘SmMoL/ i + 6 WY EF 5S Oo H
; we win wmor | a &* 5 S$ 4S 8 KR Spheres ier ee)
ea) | mm @ 1 = ~ OO =
!
orp SRODNOROVONMOBOLO
OF D'S | FRSVANHONHDNONSAGR | eet Je ee tee See
heonny(oy;\ |] We) Wey We) Sb de) ey VS) | = me > So
: == Se ee ‘polleg | >
i}
Il |—,| I | eon! = _ | |
| Baie oe bec
ported = > B |

168 Day and Night Urines

Tue Votume. No regularity in the volume of the urine was
observed. On each diet, the excesses of excretion are equally
divided between the night and morning urines.

Tota. NirroGEN. The same effect is noted as with the vol-

ume. The elimination of nitrogen does not correspond to the
volume, for in,one period the greater amount of nitrogen was
excreted with the less volume of urine.

Urea. The urea practically follows the course of the total
nitrogen. In Periods V and VI the positions are reversed, but
the difference is almost inappreciable. In all cases the relation
of urea to total nitrogen is higher during sleep than in activity.

Ammonia. Inall instances the ammonia is higher during sleep
than during waking, except in one instance, which occurs during
the low protein diet.

CREATININ. It was thought that one might cbtain some
information regarding the influence of muscular activity on the
elimination of creatinin, which is still under discussion. The
results of von Hoogenhuyze and Verploegh' would lead one to
infer that this substance is not increased by muscular activity.
On the other hand, the recent estimations of Leathes? point to
an increase. Our own results point to an increase during the
day. The amount is always higher during hours of work. The
small number of observations here recorded are not sufficient
to do more than show that the question is not at all settled.
One difference between our experiments and those of the Dutch
observers may be noted. The latter made experiments which
showed the difference between severe muscular work and ordi-
nary activity. Our experiments are those of normal activity and
complete rest.

CREATIN. It will be observed that throughout this series a
small amount of creatin is present. This substance is not sup-
posed to be a constant constituent of normal urine, yet the sub-
ject of these experiments was as far as we are aware in perfect
health. The only anomaly in his metabolism was shown some
time previously in experiments in the respiration calorimeter in
a tendency to rapid elimination of nitrogen and what appeared

‘von Hoogenhuyze and Verploegh: Zeitschr. f. physiol. Chem., xlvi,
P- 415, 1905.
* Leathes: Journ. of Physiol., xxxv, p. 125, 1906.

E. Osterberg and C. G. L. Wolf 169

to be mental excitement.! In each set the elimination of creatin
was greater during the day than at night.

Uric Acip. The uric acid appears to follow to some extent
the course of the creatinin, although during one period, the
elimination is lower during a waking period. From what is
known regarding the excretion of this substance.after the intake
of purin-containing food, one can scarcely expect to find any
dependable regularity in experiments of such short duration as
these,” particularly when it is remembered that in the second set
a considerable quantity of meat was eaten.

UNDETERMINED NITROGEN. This fraction of the excretion
is eliminated in greater quantity during work, with the excep-
tion of one of the periods, where the excess was not very great.

SULFUR. The elimination of sulfur presents more regularity
than that of the nitrogen. The excretion of total sulfur, total
sulfates, and neutral sulfur is uniformly higher during waking
than during sleep. The ethereal sulfur is the exception. Dur-
ing low protein diet the elimination was higher. During high
protein diet the elimination was lower in the periods in which
work was done. The difference is, however, not considerable.

1 Atwater and Benedict: Transactions of the National Academy of
Sciences, viii, p. 426.
*Soetbeer: Zeitschr. f. physiol. Chem., xl, p. 25, 1903.

AMMONIA IN MILK AND ITS DEVELOPMENT DURING
PROTEOLYSIS UNDER THE INFLUENCE OF
STRONG ANTISEPTICS.

By H. C. SHERMAN, W. N. BERG, L. J. COHEN, AND
W. G. WHITMAN.

(Contributions from the Havemeyer Laboratories, Columbia University,
No. 136.)

(Received for publication, April 12, 1907.)

Both Raudnitz' and Stohmann’ quote Latschenberger® to the
effect that cow’s milk contains 0.02 per cent of ammonia. The
apparent acceptance of this estimate by such authorities, and the
fact that it does not seem to have been corrected in biochemi-
cal literature, leads us to record here some of the data obtained
in the course of an investigation, the ultimate object of which is
to throw light upon the nature of the changes occurring in milk
when subjected to different methods of preservation.

Ammonia in Normal Cow’s Milk.

For the determination of ammonia in milk we have used a
slight modification of the Boussingault-Schaffer method.* A
description of the method and of the experiments which led us to
adopt it for this work has already been published.® In outline
the process consists in adding to the milk an equal volume of
methyl alcohol and a small amount of sodium carbonate and
distilling the ammonia at 60° to 65° under reduced pressure into
receivers containing an excess of standard acid. No difficulty
is found in completely removing the preformed ammonia by

1 Bestandteile, Eigenschaften und Veranderungen der Milch, Ergebnisse
der Physiologie, 2, 1.

2 Milch- und Molkerei-produkte, Braunschweig, 1898.

3 Monatsh. jf. Chem.,v, p. 129; Jahresber. fj. Tierchem., p. 222, 1884.

4 Amer. Journ. of Physiol., viii, No. 4, 1903.

§ Journ. Amer. Chem. Soc., xxvii, p. 124, 1905.

171

172 Ammonia in Milk

this method; the only error which is likely to occur is that a
small amount of ammonia may be produced by cleavage of or-
ganic compounds during the distillation. To prevent this the
mixture of milk and methyl-alcohol may be saturated with salt
before heating. Fresh milk yields so little ammonia both when
distilled with and without salt that the observed differences may
be largely accidental. Stale milk, however, may show a con-
siderable difference, indicating the presence of notable amounts
of proteolytic products in which the amino-groups are less firmly
bound than in the original proteid.

The average percentages of ammonia yielded by commercially
fresh milk from different sources when distilled with and with-
out salt are shown in the following table:

Source. Number Ammonia.
of Samples. Without Salt. With Salt.
Per Cent. Per Cent,

High class dairy producing certi-

fied milk exclusively.......... 2 0.0003 0.00015
U5 oP 7 Re eae ee eee 1 0.0005 0.0003
JOE) ee ee 1 0.0004 0.0002
RI tree een ava woo of 0.0005 0.0002*
OSU A eee re 10 0.0008 0.0004
LSS Se il eee 6 0.0008 0.0005
Oe ee a 3 0.0008 0.00045

PATE POE siete ete alg alg o's «0s 28 0.0007 0.0004

* Only two of these five samples were distilled with salt.

The milk obtained from grocery was dipped from a large can,
all of the other samples were bottled milk.

Since each of the 28 samples examined probably represented the
mixed product of many cows it is reasonably safe to accept these
figures as showing the ammonia content of cow’s milk in general
as sold in New York. Such milk evidently averages less than
0.001 percent of ammonia and probably less than 0.0005 percent.
Perfectly fresh and clean milk doubtless contains considerably
less than commercial milk. We have rarely found over 0.001 per
cent and never over 0.002 per cent of ammonia in market milk.

Formation of Ammonia During Proteolysis.

When milk is kept at laboratory temperature so that the bac-
teria which it contains are allowed to develop unchecked there is
usually an increase both of (preformed) ammonia and of those ~

Sherman, Berg, Cohen, and Whitman 173

proteolytic products which yield ammonia by cleavage under the
conditions of distillation described above, so that the percentages
of ammonia yielded with and without salt are larger and the
difference between these percentages is also usually larger than
in the fresh sample. Thus a sample of milk analyzed as soon as
purchased and again five days later gave the following results:

Ammonia Found by Distillation.
With Salt, Difference.

(Preformed Am- Without Salt. (‘‘Cleavage”? Am-
monia.) monia.)
Per Cent. Per Cent. Per Cent.
Milk as purchased..... 0.0005 0.0008 0.0003
Same 5 days later..... 0.0027 0.0036 0.0009

The addition of an efficient antiseptic retards but does not stop
proteolysis. Examination of duplicate portions of the same milk
kept with and without chloroform yielded the following results:

Ammonia Found by Distillation.

With Salt, Difference.
(Preforme:t Am- Without Salt. (‘Cleavage’”’ Am-
monia.) monia,)

Per Cent. Per Cent. Per Cent.

Portion with no preservative
(examined when 5 months old) 0.0185 0.0267 0.0082
Portion with 3 per cent chlor-

oform (examined when 1

mame Old). 5.3.59... oc ss 0.0051 0.0043 0.0008

Here the suppression of bacterial action by means of chloro-
form diminished the formation of those proteolytic products
which yield the ‘cleavage’? ammonia, much more than it
diminished the formation of ammonia itself. This is confirmed
by the fact that other samples preserved with chloroform tend to
show a lower proportion of ‘‘cleavage”’ to preformed ammonia
than in the average of samples kept without preservative.

A similar effect was found by comparison of two samples from
the same source, one of which had received a sufficient, the other
an insufficient, amount of formaldehyde. When examined the
sampleswere three to four years old and in each case very exten-
sive proteolysis had occurred,! over 60 per cent of the proteids
having been digested to products not precipitable by tannin.

In one sample the action of bacteria appeared to have been

1 The nature of the proteolysis which takes place under such conditions
has been discussed in a previous paper. Journ. Amer. Chem. Soc., xxviii,
Pp. 189, 1906.

174 Ammonia in Milk

quite thoroughly inhibited, as the percentage of lactose had not
decreased and scarcely any odor had developed. The other sam-
ple hada very pronounced odor resembling that of strong cheese
and was found to contain less than half the original percentage
of lactose.

On distillation for ammonia with and without the addition
of salt the following results were obtained:

: Preformed Ammonia, “Cleavage” Ammonia. _ Total
Per Cent. Per Cent. Per Cent.
Sample containing sufficient
formaldehyde.........-..- 0.0281 0.0003 0.0284
Sample containing insuffi-
cient formaldehyde...... 0.0211 0.0175 0.0386

Relations of Initial Purity of Milk to Development of Ammonia
and Effect of Antiseptic.

A sample of ‘‘certified’? milk which had been kept for seven
days, much of the time at laboratory temperature, and which had
become sour, still yielded when distilled without salt only 0.0004
per cent, and with salt only 0.0002 per cent of ammonia. On
the other hand, a sample which had been contaminated with a
few drops of staleseparatorslime,' was found after six days to yield
without salt, 0.0084 per cent of ammonia. A duplicate sample
of the same contaminated milk to which formaldehyde had been
added in the proportion of 1:1000 yielded aftersix days (distilled
without salt) o.coro per cent ammonia or only one-eighth as
much as in the absence of the preservative.

The following indicates the effect of antiseptics as compared
with spontaneous souring in a sample of exceptional purity.

In March, 1905, triplicate samples were taken of the mixed
milk of a herd which was being managed with a view to the pro-
duction of milk of the highest possible purity and cleanliness.
One portion was treated with chloroform, three parts by weight
of the latter in 100 parts of sample; the second received formalin
to the extent of one part actual formaldehyde per 1000 parts of sam-
ple; the third portion received no preservative treatment what-
ever. The samples were kept in a room, the temperature of

1 This was obtained from a separator through which only milk of good
quality had been passed, but was allowed to become quite stale before
using.

Sherman, Berg, Cohen, and Whitman 175

which averaged about 15° for three months, after which they were
examined with the following results:

Ammonia Found by Distillation.
Acidity* | Without Salt. With Salt. Difference.
Per Cent. Per Cent. Per Cent,

Portion with 3 per cent

chloroform, 3 mos. old 10.0 0.0035 0.0026 0.0009
Portion with 0.1 percent

formaldehyde, 3 mos.

CLIN tee OEE 2.8 0.0014 0.0007 0.0007
Portion with no preserv-
auive, 3 mossold...... 22.0 0.0012 0.0001 0.0011

* Number of ec. of tenth-normal alkali required to neutralize 10 ce. of milk.

The portion which had been kept three months at room tem-
perature with no preservative had a perfectly clean and not un-
pleasant sharp taste, somewhat resembling that of very sour
kumyss. The portion to which formaldehyde had been added
was not sour but had a slightly musty and rather nauseous taste.
On testing it was found to still give a strong formaldehyde
reaction.

SUMMARY.

Analyses of a large number of samples of mixed milk as sold in
New York City showed an average of 0.0004 per cent of ammonia
preformed at the time of examination together with an additional
0.0003 per cent of what is here called “‘cleavage’’ ammonia.

When ordinary milk is allowed to become stale the amounts of
both preformed and ‘‘cleavage’”’ ammonia usually increase.

Addition of 3 per cent of chloroform or 0.1 per cent of formal-
dehyde retards but does not stop proteolysis which results in the
formation of ammonia. The production of those proteolytic pro-
ducts to which the ‘“‘cleavage’’ ammonia is due appears to be
retarded by these antiseptics to a greater extent than is the pro-
duction of ammonia itself.

The greater the freedom from contamination the less apparent
is the influence of the antiseptic upon the development of am-
monia, and in a sample of exceptional purity spontaneous souring
in the absence of preservative treatment appeared to inhibit the
production of ammonia to a greater extent than did the addition
of 3 per cent of chloroform or o.1 per cent of formaldehyde.

The effects of pasteurization, and of antiseptics in amounts
used as food preservatives, are being studied.

ON THE SEPARATE DETERMINATION OF ACETONE AND
DIACETIC ACID IN DIABETIC URINES.

By OTTO FOLIN.
(From the Chemical Laboratory of McLean Hospital, Waverley, Mass.)
(Received for publication, April 12, 1907.)

The Messinger-Huppert method is valuable for the determina-
tion of acetone and diacetic acid in urine, but the method gives
only the sum of these two products; and there is manifest need of
an additional quantitive method for the separate determination
either of acetone or of diacetic acid.

Although acetone is a liquid with a boiling point of 56°C. and
dissolves in water in all proportions, I have found that it can be
removed from its solutions (even more readily than ammonia) by
means of an air current and at ordinary room temperatures. The
acetone can be determined in about half an hour according to the
same principle and by the help of the same apparatus which I use
for the determination of ammonia.‘ The determination is made
as follows: .

Measure 20—25 cc.of acetone solution or urine into an aérom-
eter cylinder and add 0.2—0.3 gram of oxalic acid or a few drops
of ro per cent phosphoricacid, 8-1ograms of sodium chloride? and
a little petroleum. Connect with the absorbing bottle (as in the
ammonia determination) in which has been placed water and 4o
per cent potassium hydroxide solution (about ro cc. of the latter
to 150 cc. of the former), and an excess of a standardized solution
of iodine. Connect the whole with a Chapman pump and run the
air current through for 20-25 minutes. (The air current should
be fairly strong but not so strong as for the ammonia determina-
tion.) Every trace of acetone will now have been converted into
iodoform in the receiving bottle. Acidify the contents of the
latter by the addition of concentrated hydrochloric acid (10 ce.

1 Folin: Zeitschr. fj. physiol. Chem., xxxvii, p. 161, 1902.
2 Acetone is insoluble in saturated sodium chloride solutions.

ih ii

178 Determination of Acetone and Diacetic Acid

for each ro cc. of the strong alkali used), and titrate the excess of
the iodine, as in the Messinger method, with standardized thiosul-
phate solution and starch.

The determination of the acetone can be made simultaneously
with the determination of the ammonia by the use of the same air
current and even in the same sample of urine, but I do not recom-
mend the last named combination except for cases where the
amount of available urine is small.

Inasmuch as urines containing acetone and diacetic acid are
also the urines in which clinicians determine the ammonia, the
following combination of acetone and ammonia determinations
will probably prove practical and time-saving:

The ammonia determination is started first in the usual man-
ner, except that I make use of an ordinary Chapman pump in-
stead of the more rapid but not at all necessary air blast which I
ordinarily use for ammonia determinations. A second cylinder
and absorption apparatus are then arranged as described above
for the acetone determinations and connected with the cylinder
in which the ammonia is being determined. The air current is
then regulated for 20-25 minutes with special reference to the
acetone determination. The acetone apparatus is then discon-
nected and the acetone determination finished as usual. Thusthe
air current need never be stopped and the ammonia determination
is hardly if at all interfered with by the acetone determination.'

1 In this connection I may be permitted to state that in my opinion no
other method yet devised for the determination of ammonia is so accurate
for all kinds of urine as my air current method. An excellent alternative
method is the vacuum distillation method as described by Shaffer. Amer.
Journ. of Physiol., viii, p. 348, 1903.

The vacuum distillation method has an interesting history. It was origi-
nally described by Boussingault in 1850, but was criticised out of existence
by the leading German physiological chemists of histime. In recent years
many have made use of the method in one way or another, but no one
except Shaffer has shown any disposition to rectify the error committed
against Boussingault. The vacuum distillation method isas truly Bous-
singault’s as the desiccator method is Schlésing’s. In the current litera-
ture the method is accredited to Wurster; Nencki, and Zaleski; Séldner;
Steurer, Kriiger and Reich; Kriiger, Reich and Schittenhelm, etc., but
never to its real author. Recently Schittenhelm has even complained
because the “ Kriiger-Reich-Schittenhelm Methode’’ has not received
recognition!

OO

eS

a ee

Otto Folin 179

By combining this method with the usual Messinger-Huppert
method I have recently made a number of determinations in
diabetic urines. The urines were obtained from Dr. Elliott P.
Joslin, and were only a few hours old when taken for the acetone
determinations. The series given below may be cited as illus-
trations of the acetone and diacetic acid contents of such urines:

1 2 3 4 5 6 fi
2200 ec. 1980 ce. 3600 cc. 1990 cc. A750\ce: 1370/ce- 520 ec.
Acetone in 20 ce. 3.15 mg. 5-Ofme. “elieme: 4.8 me. 9 b26. me 4s3 ump 2 ame.
Total Esse 49 g. 1.28 g. -48 g. ie2ore- .29 .05 g.
Acetone + diacetic
acid in 20 ce. 13.5 mg. 13.8 mg. 38.8 mg. 15.8 mg. 29.6mg. 16.7 mg. 20.5 mg
Total 1.48 g. gator LORD es 1 tF/ fee 10.2 ¢g. 229 o2 53 g
Total NH; 3.93 g. 5.05 g. 2.47 g. eT ee ieoores Sig

For the sake of convenience I have in the above table considered 1 cc. of
tenth-normal iodine solution equal to 1 mgm. of acetone, whereas it actu-

58

ally corresponds to only 5 mgm. Urine No. 6 represents only a twelve
hour quantity of urineat the beginning of a lethal attack of diabetic coma.
Urine No. 7 represents the following complete twenty-four hour urine
from the same patient.

It will be seen that in my acetone titrations I have been dealing
with rather small quantities. On the other hand I am confident
that in no single case did the error in the determination exceed 0.1
ec. of tenth-normal iodine. Such an error in urine No. 7 is indeed
five per cent of the total, yet the absolute error for the whole 24-
hour quantity amounts to less than 3 mgm. The same error in
urine No. 5 where the total volume of urine is 4750 cc. would
amount to less than 2 per cent of the total, 7.e., about 24 mgm.

The values recorded in the above table are not uninteresting
because they represent the first definite information yet obtained
concerning the relative proportions of acetone and diacetic acid
in diabetic urines. In this paper I wish, however, to consider
those figures only from an analytical standpoint.

The acetone values vary from about one-third to less than one-
tenth of the total values given by the Messinger-Huppert method.
In the light of this fact it is of course important that the acetone
should be determined under conditions involving the least pos-
sible decomposition of the diacetic acid. Decomposition of 5
per cent of the diacetic acid may mean an error of 50 per cent
in the acetone determination. The relatively great preponder-
ance of the diacetic acid points also to the necessity or importance

180 Determination of Acetone and Diacetic Acid

of making the acetone determinations direct and independent of
any Messinger-Huppert distillations. In a differential method

based on distillation before and after the removal of the acetone —

by the air current an error of 5 per cent in each distillation
might mean an error of more than one hundred per cent in the
acetone determination. I emphasize these facts because the
method described above is capable of many variations and modi-
fications.

Future experience may show the necessity of using 35 or even
50 cc. of urine or other acetone solutions for the acetone determi-
nations; and I believe that the method will be found equal to
such anemergency. But I scarcely believe that more than 50 cc.
can be used in the case of solutions which contain not only acetone
but also diacetic acid. The following determinations of acetone
made in a pure acetone solution indicate the possibilities of the
method in this direction:

(1) 20 ee. acetone solution + 100 cc. H,O = by direct titration 23.95 ce. iG I
(2) “ ° . 100“ * ete . Mee
(3) “ = 55 LOD “ air current (20 min.) 23.8 s
(4 ) co oe ” 100 a “ “ “ 25 “ 93 : Ss “
(5) ™ 2 . 100 ‘ «, 305, +1230
(6) “ “ “ 100“ « “ “ 95“ 23/8 ¢
(7) “ ~ sf 30 “ & re sf 25 © 23.4 ss
( 8) . - ”“ 320 “ “ “ o 35 “ 23 ‘1 6 “

In the experiments 7 and 8 the solutions in the cylinders were rinsed
out and the acetone remaining determined by direct titration. No. 7
showed 0.5 cc. tenth-normal iodine, No. 8 showed o.2 cc. tenth-normal
iodine equivalent of acetone.

Schwarz who several years ago attempted to use air currents for the
removal of acetone from urine found that it required twelve hours to re-
move 50 mg. from roo cc. of liquid.!

Schwarz barely missed discovering the efficiency of air currents for
this purpose. He worked not only with large volumes of liquid (100 cc.),
but also with small volumes (20 cc.). Having, however, once found that
a twelve hour current was necessary in the case of 100 cc. he did evidently
not allow himself to hope that anything less than several hours could suf-
fice for 20 cc.?

The method described in this paper is not free from possible
sources of error and these should be clearly understood in order
to guard against mistakes that would certainly come in and ren-
der the results unreliable.

' Arch. f: exp. Path. u. Pharm., xl, p. 189, 1898.
* See Waldvogel’s monograph on Die Acetonkorper, p. 33, 34, 1903.

ee ee el

Otto Folin 181

The spontaneous decomposition of diacetic acid is not the only
circumstance which makes a rapid acetone determination neces-
sary. A fact, at least equally important, is the rapid spontane-
ous decomposition of the alkaline hypoiodite solution used for the
decompositon of the acetone. Schwicker’s' statement that alka-
line hypoiodite solutions are almost completely converted into
iodate solutions in the course of 30 minutes is not quite in accord-
ance with the facts. Such solutions can be kept for an hour or
longer and will still give considerable precipitates when acetone
is added. For practical analytical purposes Schwicker’s state-
ment may, however, be considered substantially correct. Since
the iodate is of no use for the formation of iodoform it is clear
that it is of the greatest importance that the acetone should be
driven from the urine into the alkaline iodine solution as rapidly
as possible, keeping in mind at the same time that there is also a
limit to the rapidity with which the alkaline iodine solution can
take up the acetone.

These difficulities are at their minimum—become perhaps
almost negligible—in actual urine work where the amount of
acetone involved is, as I have shown, very small. But the danger
is there, and anyone wanting to make acetone determinations
should be aware of it.

It goes without saying that no one should attempt to make acetone
determinations in urines or other unknown solutions until he has learned
to know his air current and his apparatus by working with known acetone
solutions. Such a solution can be made and standardized (by direct
titration) in the course of a few minutes. Ten cc. of acetone diluted to
one-fourth of a liter, and 20 cc. of this solution diluted to half a liter makes
a suitable test solution of acetone.

In working with such a solution the following hints may be of service:
(1) The excess of standardized iodine solution should not be too small (I
use an excess of ro-15 cc.). (2) No time should be wasted after the strong
caustic potash solution has been added to the diluted iodine solution in the
absorption bottle. (3) If the analytical figure obtained is too low the air
current has been either too fast for the absorption apparatus, or it has been
so slow asto give time for the loss of too much hypoiodite. A second deter-
mination by means of aslower air current will show whichis thecase. A
still lower result indicates that too slow a current was used in the first

experiment and the work must be repeated with a stronger air blast. (4)
The completeness with which the air current removes the acetone from its

1 Cited from Neubauer u. Vogel, Harnanalyse, roth ed., p. 76r.

182 Determination of Acetone and Diacetic Acid

solutions can best be determined by rinsing the solution into a beaker and
testing with the alkaline iodine solution. A twenty-minute air current
should remove the acetone from 20 cc. of its solution so completely that
the subsequent addition of iodine and alkali fails to give an appreciable
test for acetone. (5) The caustic potash added to the iodine solution
must be free from nitrites. The presence of nitrites and nitrates 1s Te-
vealed by the reappearance of the blue color after the titration is seemingly
finished. e-

A freshly prepared, strongly alkaline solution of hypoiodite is at least as
effective an absorbent of acetone as is a dilute solution of acid for the
absorption of ammonia. But as the acetone comes over (and must come
over) more rapidly than does the ammonia under similar conditions, it is

important that adequate provision be made for a thorough contact of the
air carrying the acetone, with the iodine solution. The double absorp-
tion tube which I use in connection with the ammonia determinations!
is satisfactory in this case also, except that the rubber stopper cannot be
used. Alkaline hypoiodite solutions act on rubber. The two parts of the
tube must therefore be sealed together with the blast lamp.? The cut
given above makes further description of this tube unnecessary. The only
special point to be observed in its construction is that the holes in the
outer tube should be decidedly larger than the holes in the bulb of the
inner tube.

* Zettschr. f. physiol. Chem., xxxvii, p. 169.
? I have obtained some excellent tubes thus sealed together from Eimer
and Amend, New York.

————

IV, RESEARCHES ON PYRIMIDINS: ON A COLOR TEST
FOR URACIL AND CYTOSIN,

PLATE II.
(Twenty-first Paper.)
By HENRY L. WHEELER anp TREAT B. JOHNSON.
(From the Sheffield Laboratory of Yale Untversity.)
(Received for publication, April 19, 1907.)

When uracil or cytosin is dissolved in bromine water and the
solution is treated with an aqueous solution of barium hydroxide
in excess, a purple or violet-blue precipitate or color is produced
even in dilute solutions.

The formation of the purple precipitate involves several inter-
mediate reactions that are explained in the following manner:
Uracil (I) and bromine water first react to form dibromoxyhy-
drouracil (II), and the same compound is also obtained when
cytosin (III) is treated with bromine water. Dibromoxyhy-
drouracil is very sensitive toward alkalies. When treated at
ordinary temperature with an excess of barium hydroxide the two
atoms of bromine are replaced by hydroxyl groups and isodialu-
ric acid (IV) is formed. Isodialuric acid then undergoes a rear-
rangement into dialuric acid (V).' Both isodialuric and dialuric
acids give a violet-blue precipitate with barium hydroxide as
observed by Behrend and Roosen.”

1 Behrend and Kéch: Ann. d. Chem. (Liebig), cccxv, p. 246, rgor.
2 Ann. d. Chem., (Liebig), ccli, p. 244, 1889.

18 3

184 Color Test for Uracil and Cytosin

N=C—NH,
ian
OC CH
| |
HN—CH
ae III
HN—CO HN—CO HN—CO
| | 0B
oC Ck. =——— oc cg 5 ~ OC CHa
ve | ‘OH
HN—CHOH HN—CHOH HN—CO
II nn IV V
a
HN—CO
OGY "CH.
| Il
HN—CH
I

That the present test involves the formation of dialuric acid
was shown as follows: The freshly prepared barium precipitate
was dissolved in hydrochloric acid and the barium was removed
by means of dilute sulphuric acid. On evaporating this solution
then in a desiccator we obtained crystals of alloxantin (VI).
Dialuric acid undergoes oxidation in the air to alloxantin,’ while
such behavior was never observed in the case of isodialuric acid.”

HN—CO HN—CO CO—NH

a
Gp enon ===>" .0C °CH—o—con £0

el |

HN—CO ' HN—CO CO—NH
VI

Behrend was the first to show that certain pyrimidins give
bromoxyhydro-derivatives. For example, he prepared dibrom-
oxyhydromethyluractil (VII)* from 4-methyluracil.

1 Baeyer: Ann. d. Chem. (Liebig), cxxvii, p. 12, 1863.
2 Behrend and Roosen: Loc. cit.
2 Ann. d. Chem. (Liebig), ccxxix, p. 18, 1885.

Henry L. Wheeler and Treat B. Johnson’ 185

This compound gives no color with barium hydroxide. Of
more interest in connection with the test, however, is the fact
that the similar compound from thymin, bromoxyhydrothymin
(VIII), described by Walter Jones,!also gives nocolor with barium
hydroxide. Obviously these compounds would not be expected
to yield dialuric acid on treating with baryta water.

HN—CO HN—CO
aN
OC CBr, OC C (CH,) Br
Po penal
HN—C (CH,) OH HN—CHOH
VII VIII

In Richard Burians’ interesting work’ on the question whether
cytosin is a primary product or whether it results by secondary
decomposition of some other substance when the nucleic acids
are submitted to hydrolysis, he boiled guanin and adenin mixed
with various carbohydrates in 30-40 per cent sulphuric acid. He
did not obtain cytosin by this treatment, but instead, from
guanin 2-amino-6-oxypyrimidin (isocytosin) was formed (IX).’
The synthesis of this pyrimidin has been described by us. On
the other hand, adenin (5 grams) gave 6-aminopyrimidin (X),°
(0.5 gram). These pyrimidins are therefore to be considered in
the test.

HN—CO N=C—NH2
H.N—C CH He) cH
| ll ll |
N—CH N—CH
Ie bs

When isocytosin is treated with bromine water it yields a
bromine derivative that is not identical with dibromoxyhydro-
uracil. This substance gives anintense blue color oncarefully add-
ing a solution of barium hydroxide. It is a more decided blue

1 Zeitschr. f. physiol. Chem., xxix, p. 20, 1900.

? Ergeb. d. Physiol. (Asher-Spiro), v, p- 794, 1905.

3 Amer. Chem. Journ., XxXix, p. 492, 1903.

* Bittner: Ber. d. deutsch. chem. Gesellsch., xxxvi, p. 2232, 1903.
5 Wheeler and Bristol: Amer. Chem. Journ., xxxiii, p. 458, 1905.

186 Color Test for Uracil and Cytosin

than that which results from dibromoxyhydrouracil and, what
is more important, it is readily distinguished from the latter by
immediately disappearing on adding an excess of the barium.
hydroxide solution. This behavior serves as a delicate test for
isocytosin.

Finally 6-aminopyrimidin was prepared by a new method;
starting with 2-thiouracil,! which can readily be obtained in
quantity, 2,6-dichlorpyrimidin was prepared by means of phos-
phorous pentachloride.? The dichlorpyrimidin then gave 2-chlor-
6-aminopyrimidin with alcoholic ammonia, and this was found to
reduce smoothly to 6-aminopyrimidin when warmed with con-
centrated hydriodic acid.

The material thus obtained gave no color whatever with bro-
mine water and barium hydroxide.

THE TEST.

Bromine water is added to about 5 cc. of the solution to be
examined until the color is permanent. Too much bromine is
to be avoided since a large excess interferes with the test.

It is advisable, especially when only small quantities of cytosin
or uracil are present to remove the excess of bromine by passing
a stream of air through the solution. Then on adding barium
hydroxide in excess the purple color is almost immediately pro-
duced.

Very dilute solutions do not give the test. In such cases on
evaporating to dryness and then taking up the materialin a little
bromine water, removing the excess of bromine, etc., a quantity
as small as 0.001 gram of uracil gives a decided bluish-pink or
lavender color.

In applying the test in the case of cytosin it is advisable to
warm or boil the solution with bromine water, cool, and then
apply the test as above, being sure to have a slight excess of bro-
mine present before adding barium hydroxide. Dibromoxyhy-
drouracil is decomposed by prolonged boiling with water into

‘Gabriel: Ber. d. deutsch.chem. Gesellsch., xxxviii, p. 1690, 1905; Johnson
and Menge: This Journal, ii, p. 115, 1906.
? Gabriel: Ber. d. deutsch. chem. Gesellsch., xxxviii, p. 1690, 1905.

Henry L. Wheeler and Treat B. Johnson 187

s-bromuracil,! which gives no color with barium hydroxide. If,
however, 5-bromuracil is treated with bromine water it 1s con-
verted back again into dibromoxyhydrouracil. Picricacid inter-
feres with the color and should be removed before applying the
tESE:

Dibromoxyhydrouracil, HN—CO

OG7 CBr

HN—CHOH

In preparing this compound for use in further experiments we
usually took 5 grams of uracil, suspended in 20 cc. of water, and
added a little over 15 grams of bromine. The uracil dissolved
completely on warming, and, on cooling, a crystalline mass sepa-
rated. The material thus obtained had a yellow color from
excess of bromine, and the yield that first separated was almost
go per cent of the calculated. On crystallizing once from water
colorless, large, flat prisms or blocks separated. The habit of
these crystals is shown in the microphotographs (magnified 60
times) Plate II.

The same substance was obtained wheno.6 gram of cytosin sul-
phate was suspended in water and bromine added until the salt
dissolved. The solution was then concentrated to a small volume
and cooled. The prisms obtained melted at 205-6° C. (Analy-
sis (111).

The analytical results were as follows:

Calculated for Found:

CyH4O3NoBro: Te iG Til. IV.
IN eee O72 9-59 9.63 9.46
Bie ss coke B5o 55 56.00

Dibromoxyhydrouracil melts with effervescence at 203-6°.
It shows signs of decomposition below this temperature. It is
more soluble in water than uracil. The solution of the pure
white crystals is neutral to litmus, but on boiling it has an acid
reaction and finally 5-bromuracil separates. In accordance with
this silver nitrate gives no precipitate in the cold but on warming
with this reagent silver bromide separates.

1 Wheeler and Merriam: Amer. Chem. Journ., xxix, p. 486, 1903.

188 Color Test for Uracil and Cytosin

Dibromoxyhydrouracil dissolves readily in alcohol. If boiled
with alcohol, 5-bromuracil separates. If the alcoholic solution is
treated with a solution of sodium in alcohol a purple percipitate
is at once produced similar to the barium hydroxide percipitate.
Alcoholic potassium hydroxide also produces a similar colored
precipitate. These colored alkali salts differ from the barium
hydroxide precipitate by being instantly decomposed and decol-
orized by treatment with water. The aqueous solution then
turns green and finally orange on standing.

Aqueous ammonia immediately dissolves the dibrom-deriva-
tive, and removes bromine; the solution slowly takes on yellow,
then a garnet color and if sufficient material is present a reddish-
brown precipitate separates.

Dibromoxyhydrouracil is almost insoluble in ether.

THE PURPLE PRECIPITATE.

The precipitate produced by adding barium hydroxide to an
aqueous solution of dibromoxyhydrouracil when exposed on
paper to dry in the air turned red. When treated with acetic
acid it changed to a bright red powder, while the precipitate when
freshly precipitated dissolved completely in acetic acid. The
analysis of the precipitate was therefore abandoned. It was
shown that the substance yields alloxantin on treatment with
acids as follows: Four anda half gramg of dibromoxyhydro-
uracil were dissolved in 40 cc. of water and added to 11 grams of
erystallized barium hydroxide in 100 cc. of water. The purple
precipitate was rapidly filtered but no attempt was made to wash
it. It was immediately dissolved in dilute hydrochloric acid and
the barium was removed by adding 20 per cent of sulphuric acid.
The colorless solution, which on testing a portion with barium
hydroxide again gave a purple precipitate, was allowed to evapo-
rate in a desiccator. On standing over night it gave small,
stout, colorless transparent prisms. These melted at 243° C. with
effervescence. Behrendand Friedrich! state that alloxantin melts
at 243-5°. This material was not dialuric acid since its aqueous
solution failed to decompose carbonate of sodium and it was not

' Ann. d. Chem. (Liebig), cccxliv, p. 11, 1906.

Henry L. Wheeler and Treat B. Johnson 189

isodialuric acid because it melted over 100° higher. A nitrogen
determination agreed with the calculated for alloxantin:

Calculated for F
CsHgOgsN4+ 2H.0: Found:
NB Pes ioeeasense' eee 17.40 per cent. 17.11 per cent.

6-Aminopyrimidin, N==C—NH2

HC CE

| ll
N—CH

One gram of 2-chlor-6-aminopyrimidin was dissolved in 20 cc.
of colorless, concentrated hydriodic acid and then evaporated to
dryness on the water bath. lIodine separated in abundance.
The residue was evaporated several times with a solution of sul-
phur dioxide. The colorless solution was then treated with an
excess of silver sulphate, filtered and the silver was then removed
with hydrogen sulphide. On concentrating the solution a syrup
was obtained, which, when taken up in boiling alcohol, gave well-
crystallized colorless prisms. The yield was overo.8 gram. The
crystals melted at 143° to clear oil, and nitrogen determinations
agreed with the calculated for an acid sulphate of 6-amino-
pyrimidin.

Calculated for Found:
Ca 5N2.H2SO,: it ib
IN eetercrchotots swiss, 3 20-710 PAGS Pit ert

The solution of this material was freed from sulphuric acid by
means of barium hydroxide and the excess of barium hydroxide
was removed with carbonic acid. The free base proved to be
extremely soluble in water. The aqueous solution was precipi-
tated by phosphotungstic and picric acid, and it gave a precipi-
tate of silver salt when treated with silver nitratein neutral solu-
tion. This precipitate was soluble in ammonia. With bromine
water and barium hydroxide it gave no color.

ON THE RELATIVE EFFICIENCY OF THE VARIOUS METHODS
OF ADMINISTERING SALINE PURGATIVES.

By FRANK W. BANCROFT.

(From the Rudolph Spreckels Physiological Laboratory of the University
of California.)

(Received for publication, April 10, 1907.)

In a recent paper Auer’ has come to the conclusion ‘‘that the
subcutaneous and intravenous injection of magnesium sulphate
and chloride, sodium sulphate, phosphate, and citrate does not
produce purgationinrabbits.’’ These results are just the reverse
of those arrived at by MacCallum in this laboratory for all these
salts except magnesium chloride. Since I had the opportunity
of seeing many of the experiments of Dr. MacCallum, and since
his death prevents him from stating his methods and results in
greater detail and from pointing out why Auer has arrived at
different results, | have attempted to supply these deficiencies.
Accordingly, I have repeated some of MacCallum’s experiments
to obtain the quantitative data which he aid not consider it
worth while to publish, and have also made some additional
experiments in which the effects of dosage by mouth and sub-
cutaneously were compared.

TERMS AND METHODS.

Obviously in an investigation of this kind the criterion of
purgation is of thefirst importance. Auer’ says “by purgation
is here understood the passage of soft and unformed feces in
amounts exceeding that which normal animals might conceivably
pass.”’ In accordance with this definition Auer does not con-
sider that mere increase in the amount of the feces without an
increase in fluidity constitutes purgation. MacCallum, on the
other hand, considers that both increas amount and in fluidity

,

1 Amer. Journ. of Physiol., xvii, p. 15, 1906.
Mpa: [). 17.

Ig!

192 Saline Purgatives

of the feces should be taken into consideration in estimating
whether purgation has taken place. This difference is partly
responsible for the difference in the conclusions of the two
authors. Now in this connection it should be pointed out that,
as in the rabbit, the feces are usually dry and well formed, and
since the action of small doses of weak purgatives is to increase
only the amount of feces eliminated in a given time, it is essential
in such investigations that the amount of the feces should be care-
fully considered.

Another and more important difference in the method of
experimentation of the two investigators concerns the nature
of the control experiments. This difference, I think, is mainly
responsible for the difference in conclusions. Auer seems in
general not to have kept control animals for each experiment.
For, in his paper, it is only on February 15 that he compares the
feces of experimental animals with the feces of control animals
observed at the same time. In general he seems to have de-
pended for his controls on animals which were not observed at
the same time as the experimental animals with which they were
compared; and which therefore may have been in a very different
condition from that of the experimental animals. Furthermore
he does not appear to have attached much weight to the controls
which he did keep. Thus, when the feces of the experimental
animals are slightly greater in amount or fluidity than the average
of his controls he does not conclude that the salts injected have
had a slight effect, but seems to think that, as indicated by his
definition of purgation, the purgative effect of the salts has been
proved only when the result produced is above the maximum of
the controls. MacCallum’s method of experimentation, when the
actual passage of feces was to be observed, was quite different.!
He always kept controls which were observed at the same time as
the animals experimented upon, and which were treated exactly
like the experimental animals in every respect except that no
salts were injected into them. The feces passed by the controls
and the experimental animals were compared. It was found that

* MacCallum: Amer. Journ. of Physiol., x, p. 103, 1904; On the Mech-
anism of the Physiological Action of the Cathartics. Univ. of Cal. Publi-
cations: Berkeley, The University Press, pp. 11-13, 1906.

a

Frank W. Bancroft | 193

the experimental animals passed more feces, and from this it was
concluded that purgation had taken place, and that the purgation
had been caused by the salts injected. When, in addition, it was
found that the feces eliminated by the experimental animals were
more fluid than those eliminated by the controls, this fact was
also taken to mean that purgation had taken place.

The necessity for simultaneous controls is seen when it is re-
membered that all that the purgatives can do, is to increase the
rate with which the feces already present are eliminated. The
milder purgatives will cause only an elimination of the feces
already present in the large intestine; while the more powerful
ones will in addition bring about peristalsis in the caecum and the
elimination of fluid feces. Now, as Auer used only small doses of
the salts which MacCallum had classed among the weaker purga-
tives the need for adequate control experiments was even greater
than usual. For, the nature of the food, the amount eaten, the
time after eating and drinking, exercise, etc.,may all affect the
amount and nature of the feces in the large intestine. To deter-
mine, therefore, whether a small dose of a weak purgative is hav-
ing any effect it is essential that the animals compared should
start with about the same amount of feces in their large intestine,
and the only way to insure this is to take similar animals and to
treat them alike.

EXPERIMENTAL RESULTS.

My experiments fall naturally into two groups:

1. Those which simply repeated MacCallum’s experiments,
small doses being used.

2. Those in which the largest possible doses were used, in
order to obtain fluid feces.

At the outset I may say that MacCallum’s results have been
confirmed in every respect. As an example of the experiments
where small doses were employed the following may be taken:

Experiment 1. Four rabbits of about the same size and weight were con-
fined in four similar cages. The animals were separated into two pairs.
In one pair the same rabbit was always the control, in the other pair one
animal was control one day, and the experimental animal on the next.
The rabbits had been fed on hay, grain, and vegetables, but during the

194 Saline Purgatives

experiment were fed only on carrots and water, one half of each carrot
going to each member of a pair. The purgative used was ¥ sodium citrate,
which was injected subcutaneously into the dorsal region in the amounts
specified in the table. The feces were collected in jars with false bottoms
of wire mesh, so that they were never soaked by the urine.

The most significant thing about these results is that the amount
of feces passed during the first 3 to 5 hours after the injection of the
citrate, in every case, was more than that passed by the control animal.
This happened even when the total 24-hour feces of the control
animal were more than those of the experimental one. The total
amount of feces passed by all the experimental animals during
these first 3 to 5 hours after injection was 109.33 grams, while
that passed by the controls during the same period was only 4.63
grams. Thus, the amount eliminated by the injected animals
during these first few hours was 23 times that eliminated by the
controls.

Another noticeable effect of the citrate is seen in comparing
Rabbits 1 and 2 which were injected on alternate days. In this
case the total daily amount was greater for the injected animal on
every day except one. The sum of the totals for the injected
Rabbits was 156.83, or about 4 times as much as the sum of the
totals for the controls (38.26). For quite a while the rabbits
passed almost all their feces on alternate days; and that this
alternation was due to the citrate and not to accident is indicated
by the fact that it immediately disappeared on the fifth and
tenth days when no citrate was given. It is evident that a period
of constipation follows the purgation due to the citrate.

But although these results appear so clear and convincing, it is
still probably worth while to spend some time in pointing out
that it is possible by means of an uncritical examination, to make
it appear that they mean little or nothing. Thus, when the total
amounts or averages of Nos. 3 and 4 are compared they are found
to be practically equal. From this it might be concluded that the
citrate had had no effect; and if the feces had been weighed only
once a day there would have been no escape from this conclusion.
What the citrate has done is not to make feces, but to accelerate
their elimination for a few hours after the injection. After that
the control animal may catch up again, as happened with both
controls on the first day. The total amount of the feces depends,

Frank W. Bancroft 195

TABLE I. EXPERIMENT I.

Rabbit 1. Rabbit 2. | Rabbit 3. Rabbit 4.
Hour. | Dose Se Dose of Dose of Dose of
> Sodium Feces. Sodium] Feces. Sodium Feces. |Sodium Feces.
A ‘Citrate. Citrate. | Citrate. | Citrate.
—EEE | te. ahs gm. | ee. a | ee, : gm. = | oath
1 | 12:00 m. 0 aks (eas) 30 |
Next 3$hrs. | 2.6 24.2 | 0.0 16.6
| Next 20thrs. | 22.65 | 12.5 | 30.83 16.93
| Total 24 hrs. | 25.25 | 36.7 | 30.83 33.53
2/2:05p.m. | 30 | 0 aor, | 30
Next 3hrs. | 7.74 | 0.0 | | 0.0 5.7
| Next 16hrs. | 18.7 | 0.0 | 30.65 | 22.76
Next 5hrs. | ORO) 0.0 | | 0.64} 0.0
Total 24 hrs. 26.44 0.0 | 31.29) 28.46
3|2:15p.m. | 0 | 30 Wena 30
| Next 3hrs. | OLN | 7.6 | 0.0 7.44
| Next 15 hrs. (2 | Pal 7 0.84 | O. 0
| Next 6hrs. 0.0 3.34 | 0.06 | 5.5
Total 24 hrs. 0.2 32.64. | 0.90. | 13.69
4 | 2:16 p.m. 30 0 fey | 30
Next 3hrs. 3.55 Te7eM | 0.0 | | 2.55
| Next 18 hrs. 4.9 0.10. | 0.0. Gn
Next 4hrs. | 0.05) OO | 0.0 | LSimalletes
Total 25 hrs. 8.50| 1.86 0.0 | 10.62
5 | Next19hrs. | 0.05 | 0.10. | 0.0 | | 0.0
6 | 10:00 a. m. 270) 30 | Cian l. -S0an
Next 3hrs. | P O20 |: “Ovary |. 0:17 7.58
Next 21 hrs. | 10.5 0.03 | 0.0) 6.6
Total 24 hrs. 110.5 Oe17,| 0.17] jel 18
7 | 11:00a. m. 20) | 0 EGY LH 36
Next 3hrs. 1.9 0.0 | 0.0 | 4.8
Next 18hrs. 19.35 0.0 32.29 0.0
Total 21 hrs 21.25 0.0 32.29 4.8
8 | 9:10 a. m. 0 30 0 30
Next dhrs. 0.0 13.43 0.0 ee
Next 19hrs. | 0.3 13.4 8.2 | 0.2
| Total 24 hrs. jes 26.83 Se | 5.4
9 | 10:20 a. m. 30 0 0 | 30
| Next 4hrs. | 0.3 (a). al 0.0 | 0.6
| Next 25hrs. | 4.4 0.05 Ceti 9.9
Total 29 hrs. 4.3 0.15 | Nae geoy 10.5
10 | Next 26 hrs. 0.6 0.1 ik 2) 3.2
Totals 97.39 98.55 122.68 124.38
Average 24 hrs 9.74 9.85 ae P42 12.44

196 Saline Purgatives

in the first place, upon the amount and the bulkiness of the food
eaten; anil only secondarily on the amount of water in the feces.
It is only by increasing the water that moderate doses of purga-
tives can increase the total amount of feces eliminated in an ex-
periment of several days’ duration. We should expect, however,
that the amount of water in the feces should be influenced by the
citrate, and the reason why the results do not show it is because
No. 3 for some unknown reason had the diarrhoea. After the first
day the feces of all the rabbits were always unusually moist
whenever they were eliminated in considerable quantities. This
may have been due to the citrate in Nos. 1, 2 and 4, and to some
other unknown cause in No. 3; or it may have been due to the
earrots in all four rabbits. At any rate our results do not allow
us to conclude that the citrate has increased the amount of water
inthe feces. In later experiments, however, it will be shown that
subcutaneous injection of saline purgatives does produce more
fluid feces.

A further consideration of the rate of the elimination of feces
will show how absolutely essential it is to have adequate control
animals. The total amount of feces eliminated by all four rabbits
during the 240 hours of observation was 443.0 grams, giving an
average rate of 0.46 grams per rabbit per hour. The 109.33
grams eliminated by the injected animals during the first 3 to 5
hours after the injections were eliminated in 27.75 hours; which
gives an hourly rate of 1.97 grams per rabbit. Were it not for the
controls we would be forced to conclude that the citrate had
accelerated the elimination only about 4 times, while a compari-
son with the controls shows us that the rate has been accelerated
23 times. The explanation appears to be that normally the
feces are not passed at a uniform rate, but much more rapidly at
certain times of day than at others; and the time of day at which
the injections were made was one during which ordinarily but
little feces would be passed. - But this would not have been
realized if adequate controls had not been kept.

Finally, the last thing in the results to which I wish to call
attention is the effect of the food on the amount of the feces.
Before the experiment was begun the rabbits had been fed largely

on hay, which contains much more cellulose than the carrots fed.

during the experiment. The cellulose is not digested and con-

CO EE

Frank W. Bancroft 197

stitutes the main bulk of the feces. So we should expect much
less feces on a diet of carrots than on a diet composed largely of
hay. The evident decrease in the daily elimination during the
course of the experiment is due to this change of diet.

The effect of the character of the food on the amount of feces
passed, and so indirectly upon the purgation is strikingly shown
when the results just discussed are compared with those of the
next experiment.

Experiment 2. Thesame rabbitsas in Experiment 1, but they have been
fed hay, grain, and water for several days. The same diet is continued
during the experiment. ¥ sodium citrate is given to No. 1 by mouth, and
to No. 2 subcutaneously.

TABLE II, EXPERIMENT 2.

| ee 1 1580 = | No.2,1501 gm. | No.3. No. 4.
: | 1697 | 1620 Ss
[on eter: ) aeubeut.- | Contos | Conteale
S Eour. | Dose of| ‘Dose of| |
3 Sodium| Feces. (Sodium Feces. | Feces. Feces.
A |\Citrate. | ‘Citrate. | A
| | Ges | gm. ce. | gm. | gm. | gm.
1 | 2:10 p.m. eee 30 | |
Banstek ln. 4.9 | 13.8 8.3 | 2.4
First 3 hrs. | S285 || 21.6 J) ela io28
| Next 154hrs. | ie: AG One OMe: 19.5
Total 183 hrs | 31.2 GSE a eiZes 25.4
2 | 9:20a. m. | S00" | 30 |
Hirst 1 hr. | 235i, | 0.0 S35 Ne wae
| First 4} hrs. 27.4 Sell 23.8 55
) Next 234 hrs. | 2.2 ayes) Ai Get |) BHee
| Total28 hrs. | 29.6 | | 32:6 |. S05S i ea0rS
3) || Ie4l() joyaaas 30 30
ieeerst 1 hr. | 13.6 8.8 7.8 8.8
| First 4hrs. | | 23.9 Jee ister’. 24.8 18.1
| Next 16hrs. Onl | 0).55 3052 eS ZO)
| Total 20 hrs. | 53.0 | = ket. 2 CLO oor
| | |
—|— aa - eae =
4 | 10:00 a.m. 30 30
| First 1hr. eae 10.0 2 lena mice
| First 4hrs. ye) gto 21.08 Gre 32.9
| Next 24hrs. | 69.8 30.4 622300) oO!
Total 28 hrs. | 84.4 OL74) | (8/4 oiegds20
| Totals
| 1ibesys I Jobe | 44.4 32.6 2085 28.9
| First 3-44 hrs. | ea ler? 69.5 79.6 62.4
| Daily Totals | 198.2 170.8 | 302.2 | 209.3

198 Saline Purgatives

Inthis experiment I used the same rabbits employed for Experi-
ment 1, but the diet had increased the average daily feces from
11.1 grams in Experiment 1 to 54.4 grams. This additional
amount of feces almost entirely obliterated any effect that might
be due to the citrate. It is only by comparing the totals that it
can be seen that in spite of the fact that the controls are larger
animals, and pass more feces, yet during the first hour after the
administration of the citrate the experimental animals passed
the greater amount of feces. It should be noted, however, that
there is no great difference, in this case, between the effects of the
citrate when given by mouth or subcutaneously. But this dimi-
nution in the effect of the purgative with the more bulky diet
should not surprise us. The rabbits are now constantly passing
pellets. On one day hourly observations were made on all four
rabbits from 10:00 a. m. to 4:00 p. m., and there was only one
hour for one of the rabbits during which no feces were passed.
On the average a pellet was passed every 6 minutes throughout
the 24 hours, and with such an output it is not surprising that the
effects of small doses of sodium citrate might be hard to detect.
But that decided purgation is possible, even on a diet of hay and
grain, will be shown in the following experiments in which much
larger doses were given.

EFFECTS OF MAXIMAL DOSES.

In this second series the attempt was made to obtain fluid
feces. The food given was usually hay and grain. In some cases
other food was given, but in every case the same food was given
to both control and experimental animals. Sodium citrate,
sodium sulphate, and barium chloride were given both by mouth,
subcutaneously, intravenously, and intra-abdominally. It was
found possible to obtain fluid feces by means of all four methods
of administration. They could be obtained in most cases when
the purgatives were given by mouth; but the doses required were
enormous. When, however, the same doses were given sub-
cutaneously or intravenously the rabbits were often killed before
fluid feces were obtained. In these fatal cases post mortems
showed that the salts had acted as purgatives, but that their
action had been cut short by death.

Frank W. Bancroft 199
EFFECTS OF SODIUM CITRATE.

The above statements were particularly true for this salt.
Divided doses of from 50 to 120 cc. of ¥ sodium citrate given
subcutaneously or intra-abdominally killed the rabbits in from 3
to g hours, according to the rapidity with which the solutions
were injected. When given by mouth 100 to 250 cc. given at the
rate of about 50 cc. per hour produced fluid feces in about one-
third of the experiments. When the citrate was given in other
ways than by mouth the rabbits survived a dose of 100 cc. or more
in only one case. In this case 180 cc. were given, and, as was to
be expected, it had the same effect as when given by mouth;
causing after 6 hours the elimination of perfectly fluid and un-
formed feces in considerable quantities. In this experiment the
animal that wastreated by mouthreceived 250cc. of citrate during
5 hours, but produced only pasty and sticky feces that were still
formed. The feces of the control were dry and hard.

Post mortems were made on the animals that died, and in
almost every case it was evident that if the intestines had con-
tinued to function for a little longer the fluid feces which had
replaced the formed pellets usually found in the large intestine
would have been extruded. It is thus seen that while it is much
easier to obtain fluid feces by giving sodium citrate by mouth
than by giving it subcutaneously or intra-abdominally this differ-
ence in results is due to the increase in toxicity when given in
these latter ways, and not because the salt has a different action
on the digestive tract.

EFFECTS OF SODIUM SULPHATE.

In these experiments, also, the main attempt was to obtain
fluid feces, and the experiments were not arranged to show the
greatest differences in the weight of the feces of control and ex-
perimental animals; but in spite of this disadvantage, striking
quantitative results were obtained. These results are shown in
the following summaries:

These summaries show that in spite of the fact that the injected
animals got such large doses that they were seriously injured and
often killed by them, thus of course diminishing the activity of

200 Saline Purgatives

the digestive tract, the injected animals passed on the whole
much larger quantities of feces than the controls. Thus, the
rabbits that were injected subcutaneously passed about 34 times

as much feces as the controls; while those that were injected
intravenously passed about 10 times as much. These results

TABLE Ill.
SUMMARY OF SUBCUTANEOUS INJECTIONS OF SODIUM SULPHATE.

Weight of Feces in Grams.

Date Sree rapt Dose of Nay SOx. | Dose by _ Dose
|‘ ontrol.| Mouth. subcut.

Jan. 1 10 hrs. 110 cc. m/2 18.6 | 8.3 62.4
ee | 80 ce. m/2 O52) |e utes 48.3
ys oe 80 cc. m/2 Ok 20.2 2.a*
Ss 23% ‘* 60 ce. m/2 0.9 14.3 0 .4*

Te | if 70 ce. m/2 0.2 87.5 24.0
a! i | AY le 80cc. m/2 | 0.0 37.0 25.5
12 >t a 250 cc. m/6 29:67) | 0050 11 .0t
Totals 50.1 29307 | Glee?

* Rabbit died during experiment.
+ Average of four controls.
t Rabbit very sick.

TABLE IV. SUMMARY OF INTRAVENOUS INJECTIONS OF SODIUM SULPHATE.

Weight of Feces in Grams.

Duration of

Date. E : Dose of Nay SO4.
eee * loontrol} “RSS aaa
Jan. 13 5 hrs. 520 cc. m/6 0.27) 9 _8*
“a 44 San 501 ce. m/6 0.0 $15 .8*
ates U3 Zr: 409 cc. m/6 6.2 72.8
“ 18 Mle 302 ce. m/6 129) |" Folal SLO
ma 49 Ye 267 cc. m/6 4.5 | 3.4*
Totals | 12784 132.8

* Rabbit died.
+ Average of four controls.
t Injection into abdomen.

show very clearly that the subcutaneous and intravenous admin-
istration of sodium sulphate increases markedly the amount of
feces eliminated for some time after. When given by mouth this
salt increases the amount of feces to an even greater extent.

Frank W. Bancroft 201

FLUIDITY OF THE FECES.

In addition to the increase in the amount of feces, the injection
of sodium sulphate also increases the proportion of water con-
tained in them. This proportion was not measured quantita-
tively, so the data are not so satisfactory. In three of the experi-
ments with subcutaneous injections the injected animals passed
pasty or semifluid feces, while none of the controls did so.

With intravenous injection, however, the effect of the salt on
the fluidity of the feces was most clearly shown. In these experi-
ments the rabbits remained tied down on the board for the whole
of the 4 to 8 hours during which the salt was given. They were
covered with cotton batting to keep their temperature up. A
warmed ¥ solution of sodium sulphate was allowed to flow from a
burette continuously into the marginal vein of the ear at the rate
of from 30 to 50 cc. per hour. In these experiments there was a
regular progressive increase in the fluidity of the feces. At first
there were well formed pellets, hard and dry. After from four
to six hours these pellets began to be more moist. At first they
were hard and well formed, but moister than usual on the surface,
as if the moisture had entered the intestine after the pellets had
been formed. Laterthe pellets became moist clear through, and
soft. Later still they were pasty and semifluid, the masses not
well formed at all, and so soft that after they had been collected
the masses could hardly bedistinguished. Finally, in some cases,
just as the first really fluid feces were being passed the animal
died. In these cases the post mortem showed that if the intes-
tines had continued to act as they had been doing for a little
longer fluid feces would have been obtained. For in these cases,
the lower part of the large intestine, which usually contains hard
well-formed pellets, was filled for distances of from 5 to 20 centi-
meters with perfectly unformed and fluid feces. In all the ex-
periments the controls never once showed such a regular increase
in the fluidity of the feces. These facts show very conclusively
that the intravenous injection of sodium sulphate increases markedly
the amount of water in the feces.

COMPARISON WITH ADMINISTRATION BY MOUTH.

When now we come to compare the effect of giving sodium sul-
phate by mouth with the results just described, a decided differ-

202 Saiine Purgatives

ence is observed. Fluid feces can easily and almost invariably
be obtained by giving sodium sulphate by mouth. From 60 to
110 cc. of ¥ sodium sulphate or 250 to 300 cc. of the “* solution
caused an elimination of fluid feces in every case except one, and
in no case had any injurious effect on the rabbit. Why this differ-
ence? In answering this question it is essential that the normal
anatomy and physiology of the rabbit's digestive tract should be kept
in mind. The food is ordinarily very bulky, containing much
cellulose, and the digestive tract is adapted to these conditions.
The stomach is large and is never empty; and between the small
and the large intestine is the enormous caecum, which is always
filled with rather fluid, partially digested material. The food
passes very slowly through the digestive tract. Thus, when 1 fed
carrots, the feces became reddish in color, but this color was not
decided until the carrots had been fed for several days and per-
sisted for several days after the diet had been changed. Whena
control animal is killed the greater part of the large intestine is

usually found to contain well-formed pellets, and it is only within.

the caecum or its immediate vicinity, that any considerable
quantity of fluid feces is seen. When sodium sulphate is given by
mouth the same regular increase in fluidity described for intra-
venous injections is obtained, and it takes about 6 hours before
fluid feces are obtained. In view of these facts, then, it is obvious
that the fluid feces have not been produced so much by the trans-
formation of solid feces in the large intestine, as by the rapid con-
veyance of fluid feces from the caecum to the exterior. Accord-
ingly unusual activity of the large intestine, and particularly of
the caecum, would appear to be the sine qua non for the elimina-
tion of fluid feces. As a matter of fact when fluid feces were
obtained, the peristalsis of the caecum was sometimes so violent
that it could easily be seen through the muscles, skin, and hair of
the abdomen.

We may now return to the explanation of why the sodium sul-

phate, given by mouth, is so much more effective in producing

fluid feces. The reasons for this are mainly three:

1. When given by mouth the salt solution is not rapidly
absorbed into the circulation, but is passed down into the caecum.
Thus the solution distends this organ, and dilutes its contents,
thus tending by its water and bulk alone to make the feces more
fluid, and to produce peristalsis of the caecum.

Frank W. Bancroft 203

2. When injected subcutaneously the solution is not rapidly
absorbed into the circulation, and so does not reach the intestine
in as great concentration as when given by mouth.

3. When injected intravenously it is diluted by the blood and
acts all over the body, so that only a small portion reaches the
intestine. But more important than this is the fact that it acts
as a diuretic, and is rapidly eliminated by the kidneys.

The evidence for these three reasons will be presented in order:

1. Post mortems made six or seven hours after the solution
has been given by mouth show the caecum very much distended
with unusually fluid feces. This fluid must have come directly
from that introduced into the stomach; for ifit had been absorbed
and secreted again it must have been in circulation for a while,
and in that case it would have caused diuresis, which did not
take place. That the mere volume of the solution has had con-
siderable effect is shown by the fact that 350 cc. of tap water
given by mouth caused the elimination of pasty andsemifluid
feces.}

2. That absorption is slow when sodium sulphate is injected
subcutaneously is shown by the persistence for several days of the
jelly-like oedematous mass that forms about the region of the
injection. An interesting behavior of this mass was noted when 5
sodium sulphate was injected into the dorsal region. When this
was done the mass would formas usual, and then, apparently acted
on by gravity, would slowly travel down the sides of the rabbit
until it came to rest in the midventral line where it would persist
forseveraldays. This migration usually took 4 or 5 hours to com-
plete. Post mortems showed a slight inflammation in the region
of the swelling.

3. The rapid excretion of intravenously injected sodium sul-
phate is easily shown. The rabbits thus treated were constantly
urinating, while the controls, and rabbits receiving the salt by
mouth urinated but rarely. In one experiment, in spite of con-

1The most striking result produced by the water was the phenomenon,
already noted by Claude Bernard, that pronounced trembling and shaking
and muscular twitching was produced. This shaking was much more pro-
nounced than similar phenomena obtained by the subcutaneous injection
of sodium citrate, and was not obtained at all with larger doses of sodium
citrate and sodium sulphate given by mouth.

204 Saline Purgatives

siderable loss of urine by evaporation and in other ways, 150 ce.
of the 267 cc; injected were recovered in the urine. That this
urine carried with it much of the sodium sulphate introduced is
self-evident, but direct evidence for it was also afforded by the
oedematous jelly-like masses in the vicinity of the kidneys, and
by the inflammation of kidneys and bladder.

This fact of the exceptional effectiveness of sodium sulphate
when given by mouth and the explanations of the same all go to
show what was already well known to us by MacCallum’s investi-
gations, that the mechanism for the production of purgation is
located in the walls of the digestive tract.

EFFECTS OF BARIUM CHLORIDE.

This salt, as has been pointed out by MacCallum in a number of
papers, is the most energetic of the saline cathartics. He says:!
‘For anyone to convince himself that a salt may act as a purga-
tive when injected subcutaneously or intravenously it is only
necessary to introduce a small amount of BaCl, into the blood
or under the skin of arabbit. The evacuation of large quanti-
ties of semifluid feces and the violent intestinal movements
leave no room for doubt as to the action of the salt. The fact
that the intravenous and subcutaneous administration of this
salt as a purgative by veterinarians is in general use should be
sufficient proof.’’ Auer did not test the purgative effect of
barium chloride. I have made a few experiments with this salt
in order to compare the effects of per os and subcutaneous adminis-
tration of a powerful saline purgative.

It has been shown that with the mild purgative sodium sul-
phate acts better by mouth, because the fluid is not rapidly
absorbed from the digestive tract, and so has time to accumulate
in the caecum, which it distends, and the contents of which it
dilutes. It acts less strongly when given intravenously because
it is rapidly excreted by the kidneys. Thus in both cases the
sodium sulphate reaches the purgative mechanism in the intes-
tines, but on account of the slow absorption from the intestines, it
remains in contact with this mechanism longer when given by

‘On the Mechanism oj the Physiological Action of the Cathartics. Univ.
of Cal. Publications: Berkeley, The University Press, pp. 20-21, 1906.

Frank W. Bancroft 205

mouth than when given intravenously; and since this is a mild
purgative a long contact isessential. With barium chloride, how-
ever, the conditions are quite different. Purgation can be ob-
tained with such small doses that the bulk of the fluid introduced
into the digestive tract cannot be of consequence in aiding purga-
tion. Furthermore, it is very probable that it is not rapidly
eliminated by the kidneys, for MacCallum has shown! that intra-
venous doses of 1 cc. of ¥ barium chloride stop the secretion of
the urine. For these reasons, then, it might be expected that
barium chloride would be more effective when given subcu-
taneously than when given by mouth, and such I have actually
found to be the case.

Subcutaneous doses of 2 cc. of * barium chloride never failed to
cause the elimination of fluid feces within two hours. The action
was very regular. Feces began to be passed within a few minutes
after the injection. At first the pellets were hard and dry, but
after about 30 to 40 minutes they were still hard but were covered
with a good deal of surface moisture. The amount of this mois-
ture increased, and the pellets became softer, until finally the
fecal masses were entirely unformed. They were rather lumpy,
some parts being pasty, while others were so fluid that they
flowed through wire netting with millimeter meshes, and wet the
hair about the anus. These feces were much more fluid than
those of any rabbits that I have seen that have not had purga-
tives, so that no matter what definition of purgation be adopted
it must be concluded that barium chloride does purge when given
subcutaneously.

When given by mouth the action of barium chloride takes
longer in manifesting itself, and is much less constant. In two of
the four experiments no fluid feces were obtained, the pellets
remaining distinct. In one experiment the feces were entirely
unformed and mushy, but were not fluid enough to run through
the millimeter meshes. In only one case were the feces really
fluid, but in this case they were more fluid than any which were
passed after the subcutaneous administration of barium chloride.
All the rabbits receiving barium chloride subcutaneously died
either during the experiment, or an hour or two after; while only

1 Journ. of Exp. Zodl., i, p. 186, 1904.

206 Saline Purgatives

one of the rabbits that were treated by mouth died, and that one
only after many hours. The quantitative data of the experi-
ments are contained in the following summary.

TABLE V. SUMMARY OF EXPERIMENTS WITH BARIUM CHLORIDE.

| ; Weight of Feces in Grams.

Duration of |

Date. Dose of “- BaCly | | Dose by | Dos
se Experiment. ee te Control.| Ree waned
Jan.4 2hrs. 4cc, 0.0 88.2 | 45.8

divided
Feb. 7} lhr. 25 min. 2c. | 9.2 13.5 | 30.3
“ 25 | Shrs.lSmin. | 2cc. 0.5 28.9 29.1
ea. 2 De.0 Min. 2 cc, 0.5 21.5 | 42.6
Totals | Ae 10.2 | 152.1 | ae

DISCUSSION OF RESULTS.

It has been shown, as may be seen in the following summary,
that the subcutaneous or intravenous injection of sodium citrate,
sodium sulphate, and barium chloride brings about a passage of
feces from 1.3 to 23 times the amount passed by the controls.

SUMMARY OF THE TOTALS.

Grams of ‘Feces Eliminated by

Experimental Animals for each
Usual |“ Gram Passed by Controls.
9 : ‘Duration of
Purgative. Usual Dose. Food. Experi-
ment leopose by | Dose sub-
Control., youth. cut. or
| Intraven.
Sodium citrate! 30 cc. m/6 |earrots. 4 hrs. 1 23.0
Sodium citrate} 30 cc.m/6 |hay,etc.| 4hrs.| 1 1.6 1.3
Na,SOx 80 cc. m/2 |mixed. | 20 hrs. 1 5.9 3.5
Na,SO, 400 cc.m/6| ‘ ci, 6 hirs, 1 3.9 10.4
BaCl, BCC TOG), 72" 2 hrs. 1 4.9 14.5

When given in large doses, the subcutaneous or intravenous
administration of all these salts also causes the passage of more
fluid feces than those passed by the controls. These results con-
firm in every respect the conclusions arrived at by MacCallum.
Our conception of purgation, then, is that the mechanism for this
process is located in the wall of the digestive tract. This
mechanism consists of both muscles and glands the stimulation

Frank W. Bancroft 207

of which produces peristalsis and secretion. This stimulus may
be supplied, among other things, by the purgative salts. One
method of demonstrating this stimulation which MacCallum has
described, is to open the abdomen and observe the peristalsis
after the injection of the purgative salts. Auer has confirmed
these observations of MacCallum, except for magnesium sulphate.
But as he has been unable to demonstrate any increase in the
feces eliminated after injection of these salts, Auer concludes
‘“‘Peristalsis and purgation are not synonymous terms; increased
peristalsis may occur during constipation.’’ According to the
conception of purgation here outlined the increased peristalsis
shows that at least part of the purgative mechanism has been
stimulated. But in order that this stimulus may result in the
increased elimination of feces it must be kept up for a consider-
able time. Consequently it is easier to demonstrate increased
peristalsis than increased elimination of feces. I have found that
fluid feces will be produced by barium chloride after it has been
in contact with the purgative mechanism for an hour and a half.
Sodium sulphate, which is a much weaker purgative, requires,
under the most favorable circumstances, not only 100 or 200
times as strong a dose, but s7x hours in which to produce fluid
feces. This shows us that time is an essential factor in the pro-
cess. Thus, in order to be effective in the elimination of feces
the salts must not only be brought in contact with the purgative
mechanism, but they must be kept there for a certain time. This
principle should always be kept in mind in considering the differ-
ent effects produced by the different methods of administering
saline purgatives.

Having discussed the reasons for the difference in the results
of MacCallum and Auer, I might close here were it not for the
emphasis that Meltzer and Auer have laid upon MacCallum’s
statement that magnesium sulphate acts as a purgative when
given subcutaneously.

ACTION OF MAGNESIUM SULPHATE.

In a series of papers on the action of magnesium salts Meltzer
and Auer have found that in all cases these salts have a narcotic
or inhibitory action, and never a stimulating effect. On the

208 Saline Purgatives

other hand they say that MacCallum’s results with saline purga-
tives, including magnesium sulphate, lead to the conclusion that
this salt does have a stimulating effect. Speaking of Mac-
Callum’s work Meltzer and Auer say: ‘‘Here we have, then, a
pronounced theory supported by experiments that purgatives,
magnesium sulphate included, produce peristalsis by increasing
the sensitiveness of nerve and muscle of intestines; in other words,
here we seem to have an instance in which magnesium salts do
not inhibit but stimulate activity.’ Now in this quotation we
have one example of a misconception that runs all through their
paper. They everywhere seem to consider magnesium sulphate
and magnesium salts as synonymous terms; and seem to think
that MacCallum considered the purgative action of magnesium
sulphate due to the magnesium which it contains. Nothing could
be farther from the truth. MacCallum was very familiar with
the way in which magnesium inhibits intestinal peristalsis. In
fact he was the discover of this inhibition, and has mentioned it in
a number of papers.* Concerning the purgative action of mag-
nesium sulphate he says in his last paper,* concluded but a short
time before his death: ‘‘This salt of course acts as a purgative
because it is a sulphate and not on account of the presence of
magnesium. As shown later on, magnesium chloride has an
effect quite opposite to this.”

Meltzer and Auer do not mention any of MacCallum’s results
on the inhibitory effect of magnesium in their paper on the rela-
tion of magnesium salts to peristalsis, in spite of the fact that
these results are exactly in accord with their main contention
that the effect of these salts is always an inhibitory or narcotic
one. On the contrary, starting with MacCallum’s statement
that magnesium sulphate acts as a purgative when given subcu-
taneously, they lead one to believe that MacCallum thought that
in this case magnesium stimulates peristalsis.

1 Amer. Journ. of Physiol., xvii, p. 314, 1906.

*MacCallum: Amer. Journ, of Physiol., x, p. 259, 1904; also, Arch. f.d.
ges. Phystol., civ, p. 425, 1904; Journ. of Exp. Zoidl., i, p. 179, 1904; Univ.
of Cal. Publications, Physiology, li, P- 47, 1905; this Journal, i, p. 339, 1906;
On the Mechanism of the Physiological Action of the Cathartics. Univ. of
Cal. Publications: Berkeley, The University Press, p. 38, 1906.

3 On the Mechanism of the Physiological Action of the Cathartics. Univ.
of Cal. Publications: Berkeley, The University Press, Pp. 21, 1906.

Frank W. Bancroft 209

As MacCallum has often stated, the theoretical ideas upon
which his conception of purgation was based are due to Loeb.'
In aseries of papers, the first of which was published in 1899, Loeb’
found that in producing muscular twitchings and some other
phenomena sodium and barium salts were particularly effective,
and were counteracted by calcium and magnesium salts. On
account of the counteracting effect of calcium the sodium salts
that were most effective were those that precipitate calcium.
For this reason sodium sulphate is more effective than sodium
chloride, becauseit not only adds sodium but takes away calcium
from the solution. Similarly, magnesium sulphate will remove
calcium while it adds magnesium. Now since calcium inhibits
more strongly than magnesium, the addition of magnesium sul-
phate should have something of a stimulating effect, not because
of the magnesium but because of the SO,-ion which precipitates
the calcium. In this way, then, the purgative effect of magne-
sium sulphate is explained; and it can be seen at once that, accord-
ing to this view, magnesium sulphate should be considered a
rather unsatisfactory purgative, as it contains both stimulating
and inhibitory components.

It is of course a fundamental precaution in interpreting the
effect of any salt, to consider the possibility of either 1on having
produced the effect, and Meltzer and Auer have on other occasions
been careful in taking this precaution. In the paper under dis-
cussion, however, they have on several occasions failed to do so.
Thus, when testing the effect of magnesium salts in stopping the
peristalsis caused by barium chloride they find in one experiment
that magnesium sulphate stops the peristalsis, and conclude from
this that magnesium salts do inhibit the peristalsis. Now while
this conclusion is right it is not justified by the fact, for it is well
known that magnesium sulphate reacting with barium chloride
will produce barium sulphate which is insoluble. The barium
would thus beremoved from solution and could no longer produce
its characteristic effect. The injection of magnesium chloride, on
the other hand, which was also tried by the authors, does show
that the magnesium inhibits the peristalsis.

1 Univ. of Chicago Decennial Publications, x, p. 10, 1902.
* Festschriit fur Fick, Braunschweig, p. 99, 1899.

210 Saline Purgatives

In conclusion, then, it may be stated that a re-examination of
the effect of intravenous and subcutaneous injection of sodium
citrate, sodium sulphate, and barium choride has shown that
these salts markedly increase the amount of feces eliminated,
and also, when given in larger doses, markedly increase the fluid-
ity of the feces. These results confirm in every respect the con-
clusions of MacCallum. The reasons why Auer, and Meltzer and
Auer have been unable to confirm this result seem to be: (1) they
used only the milder purgatives mentioned by MacCallum, and
(2) in using these they did not keep adequate controls and thus
failed to notice the increase in the amount of feces which they
were probably getting.

SUMMARY.

1. In any experiments dealing with the quantity and charac-
ter of the feces eliminated it is absolutely essential that the feces
of the experimental animals should be compared with those of
control animals kept at the same time and treated in the same
way as the experimental animals, except of course that no salts
are given to them. The neglect of this precaution is largely
responsible for the different results arrived at by Auer and Mac-
Callum.

2. When small doses of mild purgatives are given a bulky
diet will sometimes largely mask any effect that might be due to
the purgative. In such experiments, therefore, a more concen-
trated diet should be given.

3. The subcutaneous injection of small doses of sodium citrate
was found to increase the feces eliminated for the next 3 to 5
hours to 23 times the amount eliminated by the controls.

4. The intravenous injection of maximal doses of sodium
sulphate was found to increase the feces 10 times.

5. The subcutaneous injection of 2 cc. of barium chloride
was found to increase the feces eliminated for the next two
hours 144 times.

6. Large subcutaneous or intravenous doses of all three of
these salts make the feces eliminated unusually moist. After
barium chloride the feces even become partially fluid.

7. Fluid feces can be more easily obtained with sodium citrate
when it is given by mouth because it usually takes more than

Frank W. Bancroft 211

too cc. of an “ solution to produce fluid feces, and this amount
is almost always fatal when given subcutaneously.

8. With sodium sulphate fluid feces are more easily obtained
when the salt is given by mouth because:

a. It is but slowly absorbed from the digestive tract, and
so is passed on down into the caecum which it distends and the
contents of which it dilutes. It thus remains in contact with
the walls of the digestive tract for a long time and has a chance
to affect the mechanisms for peristalsis and secretion which are
contained in the walls.

b. It is but slowly absorbed from the subcutaneous tissue,
and so never reaches the walls of the digestive tract in sufficient
concentration to produce its maximum effect.

c. When introduced into the circulation it acts as a diuretic
and is thus rapidly eliminated and never reaches the intestines
in sufficient concentration and for a long enough time to pro-
duce its maximum effect before it has killed the rabbit.

g. With barium chloride fluid feces are more easily obtained
when the salt is given subcutaneously than when it is given by
mouth.

to. In considering the effect of any salt it should always be
borne in mind that the effect may be due to the activity of either
anion or cation. Magnesium sulphate acts as a purgative on ac-
count of its anion.

11. All of these results go to show that MacCallum’s con-
clusions are correct: The mechanism for causing purgation is
in the intestinal walls and is stimulated whenever any saline pur-
gative reaches it in sufficient concentration no matter in what
way that purgative may have been introduced into the body.

f + a
tas
: 4% fl “5 -
—_
Fg
5
?
y f 4
(
j . '
>. i
<= eel
i tae
- =? 5
7
7 .
=
« =
20
%
wr
: 5
. - a
=
ee
~~
‘
*.
‘
r)
A

Ss 7

fry

THE PROTEINS OF THE PEA (Pisum Sativam).'

By THOMAS B. OSBORNE anv ISAAC F. HARRIS.
(From the Laboratory oj the Connecticut Agricultural Experiment Station.)
(Received for publication, May g, 1907.)

According to former investigations made in this laboratory the
seeds of the garden pea (Pisum Sativum) contain three different
proteins: Legumin, vicilin and legumelin. The two former are
globulins of similar composition and properties which were sepa-
rated from one another by fractional precipitation from sodium
chloride solutions, the vicilin being more soluble than the legumin
in dilute saline solutions while the legumelin remained in solution
after dialysis and was separated by heating the solution to 80°.
The most marked difference between legumin and vicilin was shown
by the behavior of their solutions on heating. Sodium chloride
solutions of legumin remain perfectly clear when heated to 100°,
those of vicilin become turbid at 90°, at 95° a flocculent coagulum
separates and when heated at 100° for some time the vicilin is
almost completely coagulated. Legumin contains a little less
carbon and a little more nitrogen than vicilin and distinctly more
sulphur. Vicilin contains less sulphur than any protein thus far
isolated. By repeated precipitation the amount of sulphur was
found to diminish from 0.23 per cent to 0.08 per cent.?

The name legumin has long been used to designate various
protein preparations obtained from many of the leguminous seeds.
Most of these preparations were formerly obtained by extraction
with alkali and precipitation with dilute acid and therefore rep-
resented products of doubtful character. Later investigations
showed that most of these proteins were globulins which could be

1 The expenses of this investigation were shared by the Connecticut
Agricultural Experiment Station and the Carnegie Institution of Wash-
ington, D. C.

2 Cf. Osborne and Campbell, Journ. Amer. Chem. Soc., xx, p. 410, 1898.

213

214 The Proteins of the Pea

extracted by sodium chloride solution and that several of the
preparations; previously called legumin were certainly different
substances. The investigations of many different leguminous
seeds made in this laboratory have shown that from the pea,
Pisum Sativum, horse bean, Vicia Faba, lentil, Ervum lens, and
vetch, Vicia Sativa, preparations of the globulin can be obtained
which agree strictly in properties and composition with one
another but are distinctly different from those obtained from
seeds of the genus Phaseolus and other kinds of legumes. The
writer has therefore designated this protein as legumin, for it
undoubtedly represents the substance to which earlier investi-
gators most frequently intended to apply this name. In the
seeds of the pea, horse bean and lentil the legumin is associated
with another protein which has been called vicilin by the writer.
The history of legumin and the characteristics of the proteins of
the seeds above mentioned has been discussed in papers from
this laboratory.'

Legumelin is an albumin-like protein which is not precipitated
by dialysis and is coagulated by heating its solutions to about 80°.
The greater part of the legumelin separates between 60° and 65°
but its complete coagulation is not effected below 80°.

The composition of legumelin is distinctly different from that
of legumin or vicilin. In properties and composition it closely
resembles leucosin found in the embryo of wheat and it is proba-
bly a tissue protein rather than a reserve food protein of the endo-
sperm. Proteins of the same composition and properties as the
legumelin of the pea have been found in a large number of other
leguminous seeds, e.g., lentil, horse bean, vetch, adzuki bean, cow
pea and soy bean.

As a complete separation of legumin from vicilin by fractional
precipitation from sodium chloride solutions can be obtained only
by the sacrifice of a large part of the mixed globulins and the
expenditure of much time and labor we have studied the results
obtained by fractional precipitation with ammonium sulphate
and found that this method yields products of the same proper-
ties and ultimate composition as those previously obtained from

' Cf. Osborne and Campbell: Journ. Amer. Chem. Soc., xviii, Pp. 583,
1896. Ibid., xx, pp. 348, 362, 393, 406 and 410, 1808.

Thomas B. Osborne and Isaac F. Harris 215

sodium chloride solutions. By this method it is possible to pre-
pare large quantities of these proteins without much loss of mate-
rial and with comparative ease.

The results given by this method are shown by the following
experiment:

The pea-meal was extracted with 10 per cent sodium chloride
solution and the extract, filtered clear, was saturated with ammo-
nium sulphate. The precipitate thus produced was dissolved in
dilute ammonium sulphate solution and the resulting solution
dialyzed for five days. The precipitated globulin was dissolved
in sodium chloride solution and again precipitated by saturation
with ammonium sulphate, the precipitate dissolved in sodium
chloride solution and, after filtering perfectly clear, the solution
was dialyzed for ten days. The dialysis precipitate was then sus-
pended in 1000 cc. of water and dissolved by adding 76 grams of
ammonium sulphate. By adding 380 grams more ammonium
sulphate, thereby raising the concentration to six-tenths satura-
tion, a considerable precipitate was produced. This was filtered
out, washed with six-tenths saturated sulphate solution, dissolved
in dilute sulphate solution and the resulting clear solution dia-
lyzed for seven days. The globulin that precipitated was com-
pletely soluble in 10 per cent sodium chloride solution and gave no
coagulum when this solution was boiled.

The remainder of the globulin, when washed with water and
alcohol and air-dried, weighed 15.3 grams and, dried at 110°, had
the following composition: C, 51.74; H, 7.14; N, 17.77 per cent.

The filtrate from the precipitate produced by six-tenths satura-
tion was raised to seven-tenths saturation but, as only a trace of
precipitate formed, the saturation was raised to eight-tenths and
the resulting precipitate filtered out. This was dissolved in
dilute ammonium sulphate solution and the clear solution was
dialyzed for seven days. The precipitate that formed, when fil-
tered out, washed and air-dried, weighed 7.35 grams. It was
completely soluble in dilute sodium chloride solution and largely
coagulated when this solution was heated in a boiling water-bath.

Dried at 110° this preparation had the following composition:
C, 52.25; H, 7.28; N, 17.17 percent. The solution filtered from
the precipitate at eight-tenths saturation, when completely satu-
rated with ammonium sulphate, yielded a small precipitate which

216 The Proteins of the Pea

when redissolved and precipitated by dialysis, gave 0.69 gram of
substance which was wholly soluble in dilute sodium chloride
solution, coagulated by heating to 100° and when dried at 110°
had the following composition: C, 52.17; H, lost; N, 17.08 per
cent. A repetition of this experiment, making the separation at
seven-tenths saturation, gave essentially the sameresults, namely,
26.5 grams of legumin, containing C, 51.89; H, 6.83: N, 17.78
per cent, and 9.5 grams of vicilin, containing C, 52.35; H, 7.15;
N, 16.90 per cent.

A third extraction gave at five-tenths saturation 17.76 grams
of globulin containing 17.77 per cent of nitrogen; between five-
tenths and six-tenths 9.62 grams containing 17.99 per cent of
nitrogen; between seven-tenths and eight-tenths 10.03 grams
containing 17.18 per cent nitrogen, and between eight-tenths and
complete saturation 7.46 grams, containing 17.00 per cent of
nitrogen.

It is thus evident that separation by fractional precipitation
with ammonium sulphate yields products of the same composi-
tion and properties as those formerly obtained by fractional
precipitation from sodium chloride solutions and as this separa-
tion is effected without difficulty and with littlelossof substance,
it affords an excellent method for preparing large quantities of
these two proteins in as pure a state as it is possible to obtain
them in any way known to us.

We accordingly prepared large quantities of these proteins for
hydrolysis by extracting the pea-meal with ro per cent sodium
chloride solution, filtering the extract perfectly clear and dialyz-
ing until nearly all of the chlorides were removed. The dialysis
precipitate was then dissolved in one-tenth saturated ammonium
sulphate solution, the resulting solution filtered clear, and enough
ammonium sulphate crystals dissolved in it to bring it to six-
tenths saturation.

The precipitate thus produced was dissolved in sodium chloridé
solution and, after filtering perfectly clear, the solution was dia-
lyzed until free from chloride. The resulting precipitate of
legumin was then washed with water, dilute and absolute alcohol
and dried over sulphuric acid.

The filtrate from the precipitate produced by six-tenths satu-
ration with ammonium sulphate was completely saturated with

Thomas B. Osborne and Isaac uF. Elarris 217

this salt, the precipitated protein dissolved in sodium chloride
solution and, after filtering perfectly clear, the solution was dia-
lyzed until free from chlorides. After washing the dialysis pre-
cipitate with water and alcohol it was dried over sulphuric acid.
The product thus obtained formed our preparation of vicilin.
The filtrate from the precipitate produced by the first dialysis of
the original sodium chloride extract of the pea-meal was heated to
80° in a water-bath, the voluminous coagulum was washed with
water and dehydrated with absolute alcohol, giving a preparation
of legumelin. By this method 1840 grams of legumin, 865 grams
of vicilin and 790 grams of legumelin were obtained.

This legumin contained 17.75 per cent of nitrogen (Kjeldahl),
0.46 per cent of sulphur and 0.48 per cent of ash, and when dis-
solved in sodium chloride solution was not coagulated by heating
to 100°. The vicilin contained 17.15 per cent of nitrogen, 0.26
per cent of sulphur and 0.41 percent of ash. Dissolved in sodium
chloride solution it wasabundantly coagulated on heating to 100°.

These preparations of legumin and vicilin agreed in composition
and deportment on heating with those obtained by fractional pre-
cipitation from sodium chloride solution and undoubtedly repre-
sent the same fractions of the total globulin of this seed as were
formerly described under these names.

rs
- ay -
Leas
7
ie Ae
a er _
“i
_ >> ean,
We - :
_ a
= a
a. a *
: 4 —
as Be 7 <a.
a -_
cd 7

HYDROLYSIS OF LEGUMIN FROM THE PEA.'
By THOMAS B. OSBORNE anp S. H. CLAPP.

(From the Laboratory of the Connecticut Agricultural Experiment Station.)
(Received for publication, May 9, 1907.)

The legumin used for this hydrolysis was prepared in the way
described in the preceding paper.

Five hundred grams, equal to 449 grams of ash- and moisture-
free legumin, were dissolved by warmingin the water-bath with a
mixture of 500 cc. of water and 500 cc. of hydrochloric acid of
specific gravity 1.2. The solution was then boiled in a bath of oil
for 15 hours.

The hydrolysis solution was then concentrated under reduced
pressure to two-thirds of the original volume, saturated with
hydrochloric acid gas and allowed to stand at 0°.

The glutaminic acid hydrochloride, freed from ammonia by
boiling with baryta and recrystallized from strong hydrochloric
acid, weighed 77.34 grams, which is equivalent to 61.95 grams of
glutaminic acid. The free acid decomposed at 202°-203° with
effervescence to a clear oil and gave the following analysis:

Carbon and hydrogen: 0.2272 gm. substance gave 0.3380 gm. CO, and
0.1283 gm. H,O.
Calculated for C,H,O,N: C =40.81; H =6.12 per cent.
Boum dle: 75 ster eaten ienece Cl omsi7se kl — 0.27, < ee

The filtrate from the hydrochloride of glutaminic acid was con-
centrated under reduced pressure to a thick syrup and esterified
with alcohol and dry hydrochloric acid gas in the usual manner.
The aqueous layer remaining after extracting the esters with
ether, was then freed from inorganic salts and the esterification

The expenses of this investigation were shared by the Connecticut
Agricultural Experiment Station and the Carnegie Institution of Washing-
ton BD: .C.

219

220 Hydrolysis of Legumin

repeated. As considerable ester, soluble in ether, was obtained,
the whole operation was again repeated, but this third treatment
yielded only an insignificant quantity of ether-soluble ester.
The ether was then distilled off on the water-bath and the resi-
due fractioned under diminished pressure with the following

result:

Fraction. ec es Pressure. Weight.
I 55° 10 mm. 33.82 gm.
ll 85° to ate 29.88 “
A “ Ra es 27.06 ~
Ill {3 100 a es Fey
ia 140° a: | 65.08 :
\B 200° 0.65 “ 37-40

Total, 242.34 “

The undistilled residue weighed 107 grams.

FractionI. (Temperature of bath up to 55°; pressure, Io mm.;
weight, 33.82 grams.) This fraction, which consisted largely of
alcohol and ether, was evaporated to dryness with strong hydro-
chloric acid and the residue esterified with alcohol and dry hydro-
chloric acid gas. The glycocoll ester hydrochloride that sepa-
rated on standing at 0°, weighed 0.56 gram. The filtrate was
worked up conjointly with the glycocoll filtrate of Fraction II.

Fraction II. (Temperature of bath, up to 85°; pressure, 10
mm.; weight, 29.88 grams.) The ester of this fraction was
saponified by boiling with water until the alkaline reaction had
ceased. The dried amino-acids were freed from prolin with boil-
ing absolute alcohol and re-esterified with dry alcohol and hydro-
chloride acid gas. The weight of glycocoll ester hydrochloride
that separated at 0°, was 2.66 grams, equivalent to 1.43 gram of
glycocoll. It crystallized from alcohol in the characteristic
needles melting at 144°.

Chlorine: 0.3054 gm. substance gave 0.3174 gm. AgCl.
Nitrogen: 0.4008 gm. substance required 4.05 cc. am HCl.
Calculated for C,H,,O,NCI: Cl = 25.45; N = 10.04 per cent.
SISMINIGL ono tte sid ie hws Chea 2s5.70; NW = 10.50) oe

In the filtrate from the glycocoll the free amino-acids were -

regenerated and submitted to fractional crystallization. The less

|
|
|
|

Thomas B. Osborne and S. H. Clapp 221

soluble fractions yielded 4.23 grams of leucin, decomposing at
298°. :
Carbon and hydrogen: 0.1077 gm. substance gave 0.2166 gm. CO, and
0.0960 gm. H,O.
Calculated for C,H,,O.N: C =54.96; H =9.92 per cent.
Moun dertiaapekic ee ce sae ose, i 10.90 448 ue
The remainder of the fraction consisted mainly of alanin.
Valin, if present at all, was in such small quantity that its isola-
tion was not attempted.
The alanin weighed 9.34 grams.
Carbon and hydrogen: 0.1477 gm. substance gave 0.2184 gm. CO, and
0.1067 gm. H,O.

Calculated for C,H,0,N: C =40.45; H =7.86 per cent.
JEXCY Una seen Re cA A CH 10. 388 leleeeucp & He

The substance decomposed at about 290°.

Fraction III. (A: Temperature of bath up to 100°; pressure,
1omm.; weight, 27.06 grams. B: Temperature of bath up to
100°; pressure, 0.75 mm.; weight, 49.10 grams.)

This fraction consisted almost entirely of the esters of leucin
and prolin. After saponifying with boiling water, the solutions
were evaporated to dryness under reduced pressure and the prolin
extracted with boiling alcohol. The part remaining undissolved
yielded on fractional crystallization 31.65 grams of leucin.

Carbon and hydrogen: 0.1855 gm. substance gave 0.3742 gm. CO, and
0.1670 gm. H,O.
Calculated for C,H,,0,N: C=54.96; H = 9.92 per cent.
Leva SoVd | ape ce here sepa rane C= 55.023 Ikl==ne,o9 ~ ©

The substance decomposed at 298°.

The alcohol extracts containing the prolin of Fractions II and
III were worked up conjointly. After separating from a small
quantity of alcohol-insoluble substance by repeated evaporations
to dryness under reduced pressure, the substance was converted
into the copper salt and the laevo separated from the racemic
prolin copper salt with boiling absolute alcohol.

The insoluble racemic salt was freed from copper with hydrogen
sulphide and the filtrate from copper sulphide evaporated to dry-
ness under reduced pressure. The residue proved entirely soluble
in absolute alcohol. For identification the racemic prolin was

222 Hydrolysis of Legumin

again converted to the copper salt and the latter recrystallized
from water. The yield of the air-dried copper salt was 2.85 grams
equivalent to 2.00 grams of prolin.
Water; 0.1278 gm. substance, dried in air, lost 0.0138 gm. H,O at 110°.

Copper: 0.1134 gm. substance, dried at rro°, gave 0.0305 gm. CuO.
Calculated for C,,H,,O,N,Cu.2H,O: H,O = 10.99 per cent.

Ute er Co LA Sp techy ey Aa Ge 8 a HO = 10.80 25 4.2
Calculated for C,,H,,O,N,Cu: Cu=a21 .8r per cent.
ay ave Mitr Blair wl 8 LAR Cu=21.49 “ “

The copper salt of the laevo-prolin, soluble in alcohol, weighed,
when dried at 100°, 15.80 grams, equivalent to 12.47 grams of
prolin.

For identification the phenylhydantoin was employed. It
crystallized from water in the characteristic flat prisms, melting
at 143°.

Carbon and hydrogen: 0.2241 gm. substance gave 0.5456 gm. CO, and
0.1152 gm. H,O.

Calculated for C,,H,,O,N,: C =66.67; H =5.57 per cent.

(ipsjere(e ht Sg hea C=66.40; H=5.71 ~ .

Fraction IV. (A: Temperature of bath, up to 140°; pressure,
0.75 mm., weight, 65.08 grams. B: Temperature of bath, up
to 200°; pressure, 0.65 mm.; weight, 37.40 grams.)

From this fraction the ester of phenylalanin was removed in
the usual manner by shaking out with ether and the ether layer
washed repeatedly with an equal volume of water. After saponi-
fication with strong hydrochloric acid, there were obtained 20.54
grams of the hydrochloride of phenylalanin equivalent to 16.82
grams of free phenylalanin. For identification the copper salt
was employed.

Copper: 0.1422 gm. substance, dried at 110°, gave 0.0284 gm. CuO.

Calculated for C,,H,,0O,N,Cu: Cu = 16.24 per cent.

yepetal Pe, ee eee near Crit hOGA

The aqueous layer was saponified by warming with an excess
of baryta and the racemized aspartic acid isolated as the barium
salt. The yield was 13.43 grams.

Carbon and hydrogen: 0.2597 gm. substance gave 0.3451 gm. CO, and
0.1279 gm. H,O
Calculated for C\LH,O,N: C = 36.09; H =5.26 per cent.
CHIRRIED gis. his bo one wala C=36.24; H=5.47 “ 2

Thomas B. Osborne and S. H. Clapp 223

In the filtrate from the barium aspartate no glutaminic acid
could be isolated as the hydrochloride. The remainder of the
fraction consisted essentially of aspartic acid and serin and for
the isolation of the former substance the copper salt was em-
ployed.' The weight of air-dry copper aspartate was 21.46 grams
equivalent to 10.36 grams of aspartic acid.

Copper: 0.1317 gm. substance, dried in air, gave 0.0383 gm. CuO.
Nitrogen: 0.4428 gm. substance, dried in air required 2.33 cc. 2% HCl.
Calculated for C,H,O,NCu.44 H,O: Cu = 23.07; N =5.08 per cent.
MOET case tioys Speeders © ose 5 toms Cul 23-235) N 520 es

The filtrate from the copper aspartate was freed from copper
with hydrogen sulphide. By fractional crystallization there were
obtained 2.45 grams of pure serin, which decomposed at about

240°.

Carbon and hydrogen: 0.2300 gm. substance gave 0.2909 gm. CO, and
0.1389 gm. H,O.

Calculated for C,H,O,N: C = 34.29; H =6.67 per cent.
Roundiese tae aoe C4549) A — Ore «

Residue after Distillation. The residue that remained after dis-
tilling off the esters weighed 107 grams. It was dissolved in hot
alcohol. The solution separated on cooling a precipitate weigh-
ing 3.32 grams. The filtrate was evaporated to dryness, the resi-
due was taken up in water and saponified by warming with an
excess of baryta. After removing the baryta with sulphuric acid,
the solution was concentrated and saturated with hydrochloric
acid gas. After prolonged standing at 0°, no hydrochloride of
glutaminic acid could be made to separate, even at different con-
centrations of the solution.

TYROSIN.

Fifty grams of the legumin, equal to 44.89 grams ash-and mois-
ture-free, were hydrolyzed with three parts, by weight, of sul-
phuric acid and six parts of water for eleven hours and, after
removing the sulphuric acid with an equivalent quantity of
barium hydroxide, the solution was concentrated until a crystal-

‘Fischer and Dérpinghaus: Zettschr. f. physiol. Chem., xxxvi, p. 474,
1902.

224 Hydrolysis of Legumin

line product separated. This was filtered out and the filtrate
further concentrated and a second separation obtained. These
were then recrystallized from hot water and the crude tyrosin
which was thus obtained was dissolved in boiling water, decol-
orized with bone black and recrystallized. Dried at 110° this
tyrosin weighed 0.7 gram, equal to 1.56 per cent.
Nitrogen: 0.3162 gm. substance required 2.40 cc. . HCl.

Calculated for C,5H,,O,N: N =7.73 per cent.

ere Naw.5g * . *

The filtrate and washings from the tyrosin were concentrated
and histidin, arginin and lysin determined by the method of
Kossel and Patten.’

HISTIDIN.

The solution of the histidin contained nitrogen equal to 1.0870
gram of histidin, equal to 2.42 per cent.
Histidin solution =500 cc. 50 cc. required 5.90 cc. an HCl = 0.2950
gm. N in soo cc. = 1.087 gm. histidin or 2.41 per cent. ;
This histidin was identified as the dichloride. It decomposed
at about 232°—233° and gave, on warming, the biuret reaction.
Chlorine: 0.1481 gm. substance gave 0.1854 gm. AgCl.

Calculated for C,H,,O,N,Cl,: Cl = 31.14 per cent.
alot ie Pane Cl=go.95 “ “

ARGININ.

The solution of the arginin contained nitrogen equal to 4.542
grams of arginin or 10.12 per cent.
Arginin solution = 1000 cc. 5o0cc. required 7.2 cc. Sn HCl = 1.44 gm. N

in 1000 cc. = 4.47 gm. arginin. Adding 0.072 gm. for the solubility of the
arginin silver = 4.542 gm. or 10.12 per cent.

The arginin was identified as the copper nitrate double salt.

Water: 0.1937 gm. substance air dry lost'o.0189 gm. H,O at rro°.
Copper: 0.1019 gm. substance, dried at 110°, gave 0.0152 gm. CuO.
Calculated for C,,H,,O,N,Cu (NO,),.3H,0: H,O =9.16 per cent.

Roe MNR ae ails Eat saline Diy in ie. © ¥ vies ove H.O'=9,76,°% 36
Calculated for C,,H,,0,N,Cu (NO,),: Cu = 11.87 per cent.
MMEMR Rs eta Pade > fete sete o's 938 ws. 0 Cu=11.92 “ *

*.Kossel and Patten: Zettschr. f. physiol. Chem., xxxviii, p. 39, 1903.

9 er Se, ee ere See. a a

Thomas B. Osborne and S. H. Clapp 225
LYSIN.

The lysin was isolated and identified as picrate, of which 4.95
grams were obtained, equal to 4.29 per cent.

Nitrogen: 0.1232 gm. substance, dried at 100°, gave 20.6 cc. moist N, at
19.5° and 752.1 mm.
Calculated for C,H,,O,N,.C,H,0,N;: N =18.70 per cent.
IDOE is occa eA Cie Oa N25 00) oe
CYSEIUNI-

Owing to the very small amount of sulphur in legumin no
attempt was made to separate cystin.

The results of this hydrolysis are given in the following table:

HYDROLYSIS OF LEGUMIN FROM THE PEA.

RIED a et tet chlo o,f ufeis.siciis)'«\2 2 sldgele’s Sia ssatwayays o. 38 per cent.
AN erat MG Od Scio a: dacip ONO RCERONE CRORE REECE RENE RT Here Pyne;

VW DIETES a 2h 50 Ete sco a Sea A a not isolated
LRICLELT I oie Be ols 4 a ese Ser 8.00 per cent.
12 YREY Treyarch clue icy Cea eee oa Qt 22) tains
sre Singtel ATEN ECW Nis oe aif g 3/6) as “as, larg oa Seas ee SC a Mee
BRN PIEREIC ACI, pails. T5500) Shaysiin ciem aye fond Javahs SSS oe 52 Our ie
SecbibecAUEC CACHE: 5) dic. bdo Vien widen ot Sete eb a eee neater)
SERIA, 5 67, ceo HORS ELS UREA ORO CIO ee eine nrc O53 o | Gee
(COPRIBRD. 5 aig Acomhltei aero RCN Ne NE ea a not determined
Parra Sa taser Ns tare, 3 20s Sats fate a aid Bin atsid oS 00s 5 3 aed 1.55 per cent.
ANTERUTTIO eae aes a Oto ee ee et A ane ea LO.02 0s
ILAVGiia i. S ae Oren Ones ROIS SEO aOR RE ae a MRIs Ai. 20) Sana
GiGi niece eee rk As a A eee Seta Soo se RA 2.42 “ e
1S TSETTRE SATE tae OO Cts a eg ea ee EOOw") ase
“LUPE LES [0] AS te ee ee present

“ “

ON THE NUCLEIN FERMENTS OF EMBRYOS.

By WALTER JONES anp C. R. AUSTRIAN.
(From the Laboratory of Physiological Chemistry, Johns Hopkins University.)
(Received for publication, May 29, 1907.)

The existence in the tissues of amidases capable of causing the
conversion of the amidopurins (guanin and adenin) into the
oxypurins (xanthin and hypoxanthin) has been repeatedly dem-
onstrated.1 Since one of the amidopurins (adenin) was found
transformed by the spleen while the other under precisely the
same conditions was not, it became necessary to assume that two
ferments are concerned in those glands which can effect both
decompositions.” This exceptional behavior of the spleen, how-
ever, was afterwards found to characterize the animal species
(pig) as other spleens (ox) can cause the decomposition of both
amidopurins.? Having noted the different distribution of the
ferments in the spleens, we undertook an investigation to find
whether this difference is confined to the spleen or is extended
to other organs, and we included in our study the ferment xantho-
oxidase. We observed most remarkable differences of ferment
distribution in the livers of four animal species, ox, pig, dog and
rabbit. The matter is clearly set forth in the following diagrams,
in which solid lines represent the undoubted existence of the fer-
ment and dotted lines indicate that the ferment is totally absent,
exists in traces or can only occasionally be found.’ The first dia-
gram will explain the other four.

1 Jones: Zeitschr. f. physiol. Chem., xli, p. 101, 1904; xlii, p. 35, 1904.
Jones and Partridge: Ibid., xlii, p. 343, 1904.

? Jones and Winternitz: Ibid., xliv, p. 1, 1905.

3 Jones: Ibid., xlv, p. 84, 1905.

‘Spitzer: Arch. f. Physiol., \xvi, p. 192, 1899. Wiener: Arch. jf. exp.
Path. u. Pharm., xiii, p. 373-

5 Jones and Austrian: Zeitschr. f. physiol. Chem., xlviii, p. 110, 1906.

227

228 The Nuclein Ferments of Embryos
Guanin Adenin
o
3 3
= i)
a v
Fe uo]
c v=
Uric Acid Xanthin Hypoxanthin
< <

Xanthodxidase

fa) | | ee

+ || < < < ae
Ox Liver. Pig’s Liver. Rabbit’s Liver. | Dog’s Liver.

<

This striking difference of nuclein ferments in the livers of
different animal species at once suggests that the adult gland will
be found different from the embryo gland in regard toitsferments.
In order to test this hypothesis we selected the pig’s liver. As the
adult gland contains two of the ferments while the third is absent,
we would be in a position to observe differences whether the
embryo gland should take on enzymic functions as development
proceeds or lose those already acquired. Our results show con-
clusively that the former is the case. In the earlier stages of
embryonic development the pig’s liver contains none of the
nuclein ferments. As development proceeds, adenase makes its
appearance while xanthodxidase and guanase are still absent.
The adult liver contains both adenase and xanthodxidase, but
not guanase. Our attention was first directed to guanase.

Experiments with liver extract of pig embryos of various
lengths, from 80 to 200 millimeters, to which guanin had been
added, showed in every case the absence of a ferment which can
cause the transformation of guanin into xanthin, since in all
these experiments the added guanin could be recovered unchanged
after the digestion. The failure of guanase in these glands, how-
ever, is more forcibly shown in the experiments described below,
in which adenin was added to the gland extract. In all casesa
small amount of guanin was obtained, showing that there is no
trace of guanase present, since the gland cannot transform the

Walter Jones and C. R. Austrian 229

small amount of guanin which is formed from its own nucleic
acid. So far as guanase is concerned, the livers of pig embryos
of various ages do not differ from the adult liver.

For the work on adenase, pig embryos of various lengths were
sorted into groups which varied 5 millimeters in length. The
livers which had been carefully removed and trimmed were
ground with sand and a known weight of the paste was treated
with five parts of chloroform water and allowed to stand in well
closed vessels for 24 hours, after which the cloudy aqueous fluid
was strained through linen. To the fluid thus obtained was
added adenin sulphate in such an amount as would easily have
been transformed by a similarly prepared extract of the adult
liver. The results with embryos up to a lengthof 150 millimeters
were uniform. The introduced adenin could be recovered as
nearly quantitatively as one could reasonably expect, but no trace
of a purin body resembling hypoxanthin could be detected.

We will describe one experiment. Forty-one grams of liver
from embryos 145-150 millimeters in length were extracted with
water as described. To 205 cc. of this fluid were added 250 milli-
grams of adenin sulphate and 5 cc.ofchloroform. The material
was digested at 47° with frequent agitation for 7 days. The pro-
duct was then diluted with an equal volume of water and after
the addition of a drop of acetic acid, heated to boiling. The pale
yellow fluid was filtered from the coagulum, treated with 7 cc. of
25 per cent sulphuric acid, and evaporated to about 75 cc. The
material was then boiled for about ro minutes to decompose any
trace of proteid present, and the purin bodies precipitated in the
usual way with ammonia and silver nitrate. The silver precipi-
tate was decomposed with hydrochloric acid and the filtrate from
silver chloride evaporated to dryness. The residue dissolved
easily in water leaving a negligible quantity of pigmented material
which was submitted to a color test for xanthin, but with negative
results. The pale yellow acid solution was decolorized with a
little animal charcoal and evaporated again to dryness for the
expulsion of the greater part of hydrochloric acid. The crystal-
line product was finally taken up in water and treated with a
slight excess of picric acid, and the yellow precipitate of adenin
picrate was quickly filtered off. The picrate weighed when dry
402 milligrams, and was found to melt at 281°. This is practic-

230 The Nuclein Ferments of Embryos

ally a quantitative recovery of the added adenin (go per cent).
The adenin picrate obtained in three such experiments was con-
verted into the sulphate which was analysed with the following
results:

0.2016 gm. required 5.64 cc. sulphuric acid (1 cc. = 0.0124 gm. N)

o.1r46. 3 4. a7 Co; 4 “ (1 cc. = 0.0124 gm. N)
Calculated: Found:
Pavanes ShReaste 34.51 per cent. 34.68; 34.60 per cent.

The mother liquor from adenin picrate was treated with sul-
phuric acid and ether for the removal of picric acid, and the
xanthin bases precipitated with silver nitrate and ammonia. The
very small precipitate was treated in the usual way anda few milli-
grams of a base finally obtained which would not dissolve in 2.5
per cent ammonia and which gave characteristic microscopic
silky needles when treated on a slide with a drop of hot hydro-
chloric acid. Assuming that this gland contains the widely dis-
tributed ferment nuclease, the presence of a trace of guanin in
this connection speaks strong for the absence of guanase. Thus
the livers of pig embryos up to 150 millimeters do not contain
adenase.

A similar series of experiments was made with livers of pig
embryos of various lengths from 165 to 200 millimeters. These
gave practically uniform results; the presence of guanase is still
incapable of demonstration, but adenase has now made its appear-
ance. We will describe one experiment.

Three hundred and five cubic centimeters of extract of liver from
embryos 165-170 millimeters were treated with 250 milligrams of
adenin sulphate and the material digested at 38° for ro days.
The treatment of the products (except for the previous removal
of a trace of guanin) was the same as that described above to the
point where adenin was precipitated with picric acid. A speci-
men of this material, however, did not produce even a cloud when
treated with picric acid. The disappearance of the added adenin
being thus proven, the bases were again precipitated with ammo-
nia and silver nitrate and the silver precipitate decomposed
with sulphuretted hydrogen. The filtrate from silver sulphide
was evaporated to dryness, and the residue crystallized out of
6.4 per cent nitric acid. On cooling, the acid solution deposited

~

Walter Jones and C. R. Austrian 231

226 mg. of material which consisted uniformly of the character-
istic crystals of hypoxanthin nitrate, and which failed altogether
to respond to the xanthin color reaction with nitric acid and
caustic soda.

The finding of hypoxanthin in this experiment not only shows
the presence of adenase in the glands employed, but proves just as
conclusively that xanthoOxidase has not yet appeared in any
very appreciable amount, since we have satisfied ourselves that
extracts of adult pig’s liver under precisely the same conditions
would have changed this hypoxanthin to xanthin or to a mixture
of xanthin and uric acid.

We have described only two of a number of experiments made,
but our results have been so uniform that we can draw a sharp
conclusion. The liver of the pig embryo up to a length of rso
millimeters does not contain any of the nuclein ferments (guan-
ase, adenase, xanthodxidase). As the length of embryo in-
creases from 150 to 170 millimeters adenase makes its appearance
and this condition is maintained until a length of 200 millimeters
is reached. At some later period, perhaps after birth, xantho-
oxidase appears but guanase is not found even in the adult pig’s
liver.

The results described in this paper may throw some light on
certain differences between results which we have obtained and
those which have been described by others. One of us with Par-
tridge found among the products of the self-digestion of pig’s pan-
creas (the species was not named) xanthin and hypoxanthin, but
neither guanin nor adenin; and that ox pancreas (the species was
not named) is capable of slowly transforming added guanin into
xanthin. Schenck! later found among the products of self-diges-
tion of the pancreas of both species, guanin and hypoxanthin, but
neither xanthin nor adenin, and explained his results upon the
hypothesis that guanase and adenase are different ferments.
After this work on the pancreas had been repeated many times
with results essentially as originally described, we were somewhat
surprised upon one occasion to find a small amount of guanin
among the products of the self-digestion of pig’s pancreas. We

1 Zeitschr. f. physiol. Chem., xliii, p. 406.

232 The Nuclein Ferments of Embryos

have since obtained this result occasionally but always find in
the presence of this small amount of guanin a larger amount of
xanthin.

On one occasion we were able to show a trace of xanthodxidase
in pig’s spleen, but in spite of an enormous number of experiments
preceding and subsequent to this one we have not been able to
demonstrate the ferment in this tissue.

In his attempt to prove the identity of guanase and adenase,
Schittenhelm described experiments to show an appreciable
transformation of guanin into xanthin. In one of these experi-
ments, however, a perfectly enormous amount of guanin was
introduced and a mere trace of xanthin found afterthe digestion.

Finally: we stated in our last contribution to this subject that
rabbit’s liver contains very little, if any, adenase. After two
unsuccessful attempts (and we presume the results of these two
agree with our own) Schittenhelm finally succeeds in “‘activat-
ing” the ferment and claims to show that this tissue can cause
the conversion of adenin into hypoxanthin.!

Having in mind some of these differences, we concluded in our
last contribution that not only do animal species differ in the
@stribution of these ferments, but that individuals of the same
species may differ from one another. It is possible that some at
least of these individual differences may be accounted for by a
difference in age.

*Schittenhelm: Zeitschr. f. physiol. Chem., 1, Pp. 30.

ON THE FRACTIONATION OF AGGLUTININS AND
ANTITOXIN.

By ROBERT BANKS GIBSON anp KATHARINE R. COLLINS.

(From the Research Laboratory of the Department of Health, of the City of
New York, William H. Park, M.D., Director.)

(Received for publication, May 31, 1907.)

E. P. Pick' in 1901 associated a number of antisubstances
individually with the one or the other of the two serum globulin
fractions of the Hofmeister classification. In the pseudoglobulin
(3-4 to 4.6 saturated ammonium sulphate solution)? group of
antibodies he placed the diphtheria and tetanus antitoxins and
the typhoid agglutinin of horse serum; the lower or euglobulin
fraction (2.9 to 3.4 saturated) comprises diphtheria and tetanus
antitoxin and cholera lysin in the goat, rabbit and guinea pig, and
finally cholera agglutinin in the horse and goat. It becomes
possible, according to Pick, to separate the individual specifically
reacting antisubstances by fractioning appropriate mixtures of
sera. Such a possibility suggested the application of this method
to the further study of certain antibodies, especially of the rela-
tion of specific and group agglutinins developed by immunization
against a single strain of organism. Preliminary experiments
in the course of our investigation indicated the unreliability of
Pick’s differentiation, and attention was accordingly directed to
the actual possibility and practicability of distinguishing between
antibodies by fractionation of the globulin. The availability
of polyagglutinative sera for the work gave a chance for making
numerous and extended observations of the distribution of these
antibodies in the fractions.*

1 Beitr. z. chem. Physiol. u. Path., i, p. 351, 1901.

2 The degrees of saturation, as here expressed, indicate a concentration
equivalent to a content in 1o cc. of the precipitated solution of 3.4 and
4.6 cc. of saturated ammonium solution respectively.

3 A preliminary account of our results was published several months
ago in the Proceedings of the Society for Experimental Biology and Medicine,
iv, p. 15, 1906-1907.

233

234 Fractionation of Agglutinins and Antitoxin

The literature on the fractional precipitation of the antibodies
is not extensive. Ide and Lemaire! (in 1899) found the precipi-
tation limits of diphtheria antitoxin in horse serum to be from
2.8—4.4 saturation. Fuld and Spiro? (1900) associated the anti-
rennin of horse serum with the pseudoglobulin, while a milk-
coagulating action was possessed by the eu-fraction. The fail-
ures of Brodie and of Atkinson (of this laboratory) to separate
diphtheria antitoxin from the accompanying serum globulins are
referred to in the following paper. Porges and Spiro* without
giving any of their experimental protocols divide, according to
the distribution of the antibodies, the serum globulin into three
distinct fractions; the ammonium sulphate precipitation bound-
aries of these overlap unless the serum is greatly diluted. Land-
steiner* found that the antitryptic action of blood serum is pos-
sessed by the albumin precipitated by complete saturation with
ammonium sulphate after removal of the globulin. Cathcart?
also observed the antitrypsin to be associated with the albumin
but not with the globulin fraction. Glaessner® states that the
euglobulin fraction inhibits the action of trypsin, but the typical
protocol which he publishes and his statement of the Hofmeister
classification show a misconception and confusion of the identity
of his fractions.?. Glaessner apparently found that the globulin
remaining in solution on dialysis was antitryptic.

Very recently Simon, Lamar and Bispham’ found that the
opsonic substance in blood serum was precipitated with the serum

"Ide and Lemaire: Arch. internat. d. pharmacodyn., vi, p. 477, 1899.

? Fuld and Spiro: Zettschr. f. physiol. Chem., xxxi, p. 133, 1900.

* Porges and Spiro: Beitr. z. chem. Physiol. u. Path., iii, p. 277, 1903.

‘ Landsteiner: Centralbl. f. Bakt., xxvii, Abt. I, p. 357, 1900.

*Catheart: Journ. of Physiol., xxxi, p. 497, 1904.

* Glaessner: Beitr. z. chem. Physiol. u. Path., iv, p. 79, 1904.

? The error on his part has not heretofore been noted. Glaessner states
(p. 82): ‘‘ Das Globulin des Blutserums lasst sich nach den in Hofmeisters
Laboratorium in mindestens 3 Fractionen zerlegen: in das bei 25 Proz.
Sattigung mit Ammonsulfat ausfallbare Fibrino-globulin, in das Euglo-
bulin, das bei einer Sdttigung von 33 Proz. ausfallt und bei der Dialyse in
Lésung bleibt, und endlich in das bei 38 Proz. Sattigung ausfallbare bei
der Dialyse unldsliche Pseudoglobulin.’’ The confused nomenclature of
the degrees of salt saturation is discussed in the following paper (p. 254)..

‘Simon, Lamar and Bispham: Journ. of Exp. Med., viii, p. 651, 1906.

Robert B. Gibson and Katharine R. Collins 235

proteids which separated out on dialysis. Opie and Barker?
observed that proteolysis in.an alkaline medium by the enzyme
of the leucocytes is inhibited by the serum albumin.

In a paper just published on the relation of diphtheria anti-
toxin to the serum globulin, Ledingham? finds that the pseudo-
globulin of horse serum contains the greater part of the antitoxin.
The repeatedly precipitated euglobulin, however, in one horse
still contained fully 10 per cent of the antitoxin; in a second
horse, the euglobulin similarly treated contained practically none
of the antitoxin. Single precipitations (without further puri-
fication) showed that large amounts of antitoxin may be carried
down with the lower fraction; with the one horse, judging in ,
part from the tests on the pseudoglobulin fraction, over half the
units must have precipitated with the euglobulin.* Ledingham
also confirms our own observation here reported that the diph-
theria antitoxin of the goat is not invariably linked to the eu-
globulin fraction as maintained by Pick.

Recently observations have been made by Ruediger‘ on the
relation to the blood proteids of streptolysin, the hemolytic sub-
stance produced by the development of streptococci in heated
serum. The lysin was precipitated with the globulin by satura-
tion with magnesium sulphate; it was found with both the euglob-
ulin and pseudoglobulin of the fractioned undiluted serum and also
in both the insoluble proteid and the filtrate of the dialyzed half-

1 Opie and Barker: Journ. of Exp. Med., ix, p. 207, 1907.

?Ledingham: Journ. of Hygiene, vii, p. 65, 1907.

3 Brieger (Festschrift fur R. Koch, Jena, 1903) and also one of us (Gibson)
have already reported similar experiences. In some as yet unpublished
experiments carried on for another purpose by one of us (Gibson) and E.
J. Banzhaf of this laboratory, it has been found that if undiluted horse
serum be precipitated with half its volume of saturated ammonium sul-
phate solution and allowed to stand for 18 hours, the euglobulin precipi-
tate may contain over two-thirds of the total serum globulin. Precipita-
tion under the same conditions except that the precipitated mixture is
5 or ro times the volume of the original serum gives a euglobulin figure
only of from a fifth to a third the total globulin. The euglobulin at a
dilution of 1:10 is noticeably smaller than at 1:5. This fact explains
the diminished antitoxic content of reprecipitated or washed euglobulin
fraction and makes difficult any hard and fast division into eu- and pseudo-
globulins.

4Ruediger: Journ. of Infect. Diseases, iv, p. 377, 1997-

236 Fractionation of Agglutinins and Antitoxin

saturation ammonium sulphate precipitate. Moll,’ however,
has shown that such heating suffices to alter the chemical com-
position and to change the precipitation characters of the blood
proteids.

Owing to the difficulty of making a hard and fast division of
the serum globulins into the eu- and pseudo-fractions, our experi-
ments were not planned to be interpreted especially from the
quantitative occurrence of the agglutinins in the one or the
other fraction of specific agglutinating sera. We have aimed,
however, to determine if the relative proportion of the agglutinins
of polyagglutinative sera in the fractions remained constant as
regards the proportional distribution of all the agglutinins of the
serum in the eu- and pseudo-globulins. Ina word, by a differ-
ence of precipitation limits to ammonium sulphate, it should be
expected that the bulk of one or more of the agglutinins would
appear in the one fraction, as contrasted with the larger pro-
portion of each of the remaining agglutinins occurring in the
other fraction. Attention is particularly directed in studying the
results from our standpoint to the pseudoglobulin fraction (fil-
trate from the euglobulin fraction). Any loss from the euglob-
ulin fraction through solubility in the wash solution has been
considered only in the case of the antitoxins. Such loss may be
interpreted as due to the resolution of the mechanically precip-
itated proteids of the more soluble fraction; it may be considered
just as well in part as a not absolute insolubility of the eu-
fraction in 3.4 saturation ammonium sulphate. The content of
agglutinins in the washed euglobulin fraction is of interest, how-
ever, as it is more highly “purified” than the pseudoglobulin, so
that any relative differences in the distribution of the agglutinins
should be from this standpoint the more pronounced for the low
fraction.” The limitations in determining the agglutinative
potency of the serum and of the fractions, however, make diffi-
cult at times the interpretation of the readings obtained, and
do not permit conclusions being drawn from a single experiment.

* Moll: Bettr. z. chem. Physiol. u. Path., iv, p. 563, 1904.

* The euglobulin of 2 cc. of serum precipitated at a final dilution of 1: 5
was washed usually by being three times thoroughly suspended in ro cc.
of 3.4 saturation ammonium sulphate solution and recentrifuged.

EE —

Robert B. Gibson and Katharine R. Collins 237

It was found repeatedly in our experiments with rabbit and
goat sera that the agglutinins for the dysentery group of organ-
isms (Flexner Manila and Shiga), typhoid, colon and cholera,
were not confined to either the pseudoglobulin of the washed
(with 3.4 saturated ammonium sulphate solution) euglobulin
fractions; they were either split by the fractioning, the larger
portion occurring in the pseudoglobulin, or almost the entire
amount of the agglutinating substances recovered were in this
higher fraction in the original quantitative proportion to one
another. With antidysentery horse serum, the dysentery (Shiga
and Flexner) and B. colt agglutinins were fairly quantitatively
split between the pseudo- and euglobulin fractions, the latter
containing the lesser amount. With an anticholera and anti-
typhoid horse serum, the pseudoglobulin (two experiments) and
also the filtrates from two additional 3.6 and 3.8 saturation
precipitations contained the bulk of both the agglutinins. In
subsequent experiments with sera from other bleedings as well as
with the sera used above, the typhoid agglutinin was divided
between the two fractions with a somewhat larger proportion
occurring in the pseudoglobulin.

The results of exhaustion experiments on the two globulin
fractions were the same as those that would be obtained in the
use of the native serum, and failed to give any reason for believ-
ing that we were dealing with a separation of group and specific
agglutinins through fractioning.’

1Tmmunization with certain bacteria results in the development of
‘“‘sroup’’ or “common’”’ and of ‘‘specific’’ agglutinins in the serum of the
animal immunized. Group agglutinins are agglutinating substances
which are effective both on the homologous organism and on some allied
strains of bacteria; specific agglutinin is effective on the organism used
forimmunization. Animmune serum developed by immunization against
B. dysenterie (Shiga) might contain, for the sake of illustration, simply
(1) a group agglutinin effective for Shiga, Flexner Manila, Pfeiffer and
B. coli, and (2) an agglutinin specific for the Shiga strain. When immu-
nization has been developed simultaneously against two or three of the
above organisms instead of the Shiga strain alone, the number and agglu-
tinating scope of the agglutinins resulting becomes more complex, various
group agglutinins and the specific agglutinins for each organism being
present. The existence of the two types of agglutinins is demonstrated
by the agglutinating characters of the diluted serum after the agglutinins
for any desired strain of organism have been exhausted by adding sus-

238 Fractionation of Agglutinins and Antitoxin

Precipitation of antidiphtheria goat serum (three experiments)
showed that less than half the antitoxin remained in the pseudo-
globulin; practically none was found in the euglobulin while the |
3.4 saturated ammonium sulphate solution washings contained
the larger part. Our results with the antitoxic horse serum at a
dilution of 1:5 are essentially identical with Pick’s.

The results of the work accomplished have demonstrated the
untrustworthiness of any such differentiation of the antibodies
as those contained in the euglobulin and those of the pseudo-
globulin. No evidence has been adduced from our experiments
to show that the agglutinins developed in the rabbit, goat and
horse can be classed as belonging to either globulin, or that these
antibodies can be separated from one another by ammonium
sulphate fractioning of polyagglutinative sera.*

A description of the experimental procedure is given with the
protocols which follow:

FRACTIONATION OF POLYAGGLUTINATIVE RABBIT SERUM.

Combined immunization against Flexner Manila dysentery,
Shiga dysentery, Pfeiffer,> colon and cholera. Rabbit bled
II/27/06. Serum fractioned II/29/06.

Five cc. of the rabbit serum diluted with 11.5 cc. of distilled
water, were precipitated by 8.5 cc. of saturated ammonium sul-
phate solution. After standing 2 hours, 12.5 cc. of the uniform
mixture were removed and centrifuged in stoppered tubes. The
supernatant fluid contains the “pseudoglobulin”’ fraction. The

pensions of washed organisms; after contact for some hours the agglutin-
ated and excess of free organisms are removed by filtration. The agglu-
tinins for several related strains can thus be successively withdrawn. In
the example given above, exhaustion with either B. coli, Flexner Manila or
Pfeiffer would remove only the group agglutinin (1) so that the serum
would still agglutinate the Shiga, though at a diminished dilution; exhaus-
tion with the Shiga would take out both the group and the specific agglu-
tinins, and this serum would no longer have agglutinating properties for
any of the above organisms. Cf. Castellani: Zeitschr. —. Hyg. u. Inf., xl, p.
1, 1902; also Park and Collins: Journ. Med. Research, xii, p. 491, 1904.

‘In the paper by Banzhaf and Gibson following this article, it will be
shown that the globulins of the serum do differ markedly in their content
of antitoxin per gram proteid.

? The original Pfeiffer strain of B. typhosus.

Robert B. Gibson and Katharine R. Collins 239

precipitate (“‘euglobulin”) was washed three times by being
thoroughly suspended in 3.4 saturated solution and centrifuged;
it was once more suspended and the volume made up to 12.5 cc.
The agglutinating properties were then ascertained of (1) the
original serum; (2) the total globulin, a uniform sample of the
precipitated serum; (3) the pseudoglobulin fraction, and (4) the
washed 3.4 saturation precipitate or euglobulin fraction. Agglu-
tinations were determined microscopically and control slides
were examined. Dilutions are in terms corresponding to the
original serum. The charactersof theagglutinations at the vari-
ous dilutions are indicated as follows:

++++4+ agglutination with no free organisms.
+++ agglutination with relatively very few free organisms.
++ agglutination but with numerous free organisms.
+ incomplete agglutination, small loose groups and many free
bacteria.

+ tendency to agglutinate.

— noagglutination.

o observation lost.

The procedure was essentially unchanged in the other experi-
ments. The serum used in this and the following experiments
was roughly tested for orientation before the final agglutina-
tions were made. A + + + agglutination indicates usually
the highest dilution foran observed positive reaction. It should
be remembered that the character of the agglutination at any
dilution is often difficult to decisively determine; the actual
observations are by no means so exact as would be inferred from the
published experiments of Pick and of others.

The agglutinating properties of this rabbit serum (Table I)
were too low to be entirely satisfactory for fractioning, the weak-
est agglutinating action (1:50) being manifested on the Shiga
strain. The agglutinin for this organism drops out in the euglobu-
lin fraction. This lost agglutinin is not found in the pseudo frac-
tion. It is conceivable from this experiment that the agglutinin
of the Shiga dysentery is more soluble than that of the other five
strains; however, the Shiga shows no such differences in the two
following precipitation experiments on later bleedings of the
same rabbit. It is more likely that the content of serum in
agglutinin was originally so low that in the fractioned and washed

240 Fractionation of Agglutinins and Antitoxin

euglobulin the dilution of 1:20 failed to show the proportion of
agglutinin present. With the Flexner Manila dysentery, the
Pfeiffer typhoid strain, the colon and the cholera—all contained
in greater concentration than the Shiga agglutinin—the major
portions of the agglutinins occur in the pseudoglobulin, a smaller
amount being held by the low fraction.

TABLE I, FRACTIONING OF POLYAGGLUTINATIVE RABBIT SERUM.

Bleeding of II/27/06; fractioning I1/29/06.

Organism. | Fraction. | 20 | 50 | 100 200 | 500 | 1000

Flexner Original
Manila..| serum
Total gbl.
Pseudogbl.
Eugbl.*

Shiga ...| Serum
Total gbl.
Pseudogbl.
Eugbl.*

Pfeiffer ..| Serum
Total gbl.
Pseudogbl.
Eugbl.*

Colon....| Serum

| Total gbl.
Pseudogbl.
Eugbl.*

Cholera...) Serum
Total gbl.
| Pseudogbl.
| Eugbl.*

ee rae res te Ee

++ ++++
$it+ ++4++
i+ ++4++
++ +¢4+4++
ee tote
++ +4+4++
++ +444

|
tolls

“++++ Fttt+ +4+4+4+
++

+++ 4444+ +444
+++ +4+4+4+ +4474
++++ +4++4+ +4+++
ttt btt+ bet +
ttt ttt bett+
btett+ ttt bett+
ttt Fttt Pett

*Tested III/1/06 with a fresh culture.

Exhaustion of the two globulin fractions (Table II) with sus-
pensions of the Flexner Manila strain has not withdrawn the
agglutinins for the typhoid from either fraction. Exhaustion
with the Pfeiffer likewise is not complete for the dysentery in
either pseudo- or euglobulin.

This later bleeding of the same rabbit shows by far the greater
part of all the agglutinins in the pseudoglobulin (Table III), and
small though relatively proportional amounts of each in the low
fraction. The absolute dropping out of the Shiga does not occur
as in the preceding fractioning.. The results are also more uni-
form.

Robert B. Gibson and Katharine: R. Collins

TABLE II, EXHAUSTION EXPERIMENT. POLYAGGLUTINATIVE RABBIT.
Exhaustion of the Original Serum and the Fractions (cf. Table I).

Exhaustion with Flexner Manila.

Organism.

nlexneneMianilan wes. snc ees

JPARSTURI SOs ah ae ae

Fraction.

Serum
Pseudogbl.
Eugbl.

Serum
Pseudogbl.
Eugbl.

241

Exhaustion with Pfeiffer.

| Pseudogbl.

Eubgl.
Pseudogbl.

Eugbl.

TABLE III. FRACTIONING OF POLYAGGLUTINATIVE RABBIT SERUM.

Serum from bleeding IV/16/06; serum fractioned IV/18/06.

Organism. Fraction. 50 100 200 500
Flexner

Manila...| Totalgbl. |++4+4/++4+4+/4+4+44/+++

Pseudogbl. |+ +++) ++++)/++++4+)++4+

Eugbl. ++ - = AS

liga. 5)... Total gbl. |++++/+4+4+4/++4++4/4+4++4+

Pseudogbl. |+ +++/++++/++++4)\)+++

Eugbl. ++ as = a

Cholera.... | Totalgbl. |++++4++++4+/+4+4++4+4++4++4+

Pseudogbl. |+ +++|/++++|/++++/+++

Eugbl. a = | = =

Pfeiffer ....| Totalgbl. |++++ ++++ +4+4++ +++

| Psuedogbl. |++++)+++4+/)+++4+4+/) +++

Eugbl. eet = | = | =

ttt +

+

Ser ota oe Sse

| 2000

=

The third precipitation (Table IV) shows apparently a recovery
of all the agglutinins in the high globulin fraction.
globulin dilutions for the Flexner Manila, in fact, were read

slightly higher than were those of the total globulin.

The pseudo-

242 Fractionation of Agglutinins and Antitoxin

Pick’s single observation by the test-tube method on typhoid
rabbit serum, agglutinating at 1: 20,000, gave the limit of agglu-
tination of the pseudoglobulin at 1: 3000, of the euglobulin.
1: 20,000; after reprecipitating the fractions three and four times,
respectively, the limits of dilution for agglutination were at 1:20
and 1: 8000.

TABLE IV. FRACTIONING OF POLYAGGLUTINATIVE RABBIT SERUM.

Serum from bleeding on V/23/06.

Organism. Fraction. 50 100 200 500 a 1000
Flexner

Manila .| Total gbl. 4

Pseudogbl. |+

| Eugbl. w= — _ =

Pfeiffer...) Total gbl. tetti[tt+ttl(t+ttti[+t¢+4+) tt
| Pseudogbl. |++++/++++/++++/+++4+] 4+
| Eugbl. _ — = = =

Cholera...| Total gbl. +++4+)4+++4++4)/+4+4++4+) #++4 ++
Pseudogbl. |++++(/++++/+4+++| $44 | +4

POLYAGGLUTINATIVE GOAT SERUM.

Immunization against Flexner Manila, Shiga, Pfeiffer, cholera
and B. colt.

The proportion of the agglutinins for each organism is much
greater in the pseudoglobulin than in the euglobulin fractions
(Table V). This conclusion is confirmed for the Shiga, Pfeiffer
and cholera by the redetermination of the agglutinations in the
exhaustion experiment (Table VI). The Flexner Manila dysen-
tery strain has not exhausted the agglutinins for the other organ-
isms completely from the eu- or from the pseudoglobulin frac-
tions; nor have the agglutinins apparently been withdrawn to |
a relatively greater degree from the one than from the other frac-
tion. ;

Fractioning of the polyagglutinative serum (of lessened agglu-
tinating power) from the second bleeding of the goat immunized
against the mixed cultures showed that the agglutinins were
almost quantitatively contained in the pseudoglobulin fraction
(Table VII).

Robert B. Gibson and Katharine R. Collins 243

TABLE V. FRACTIONING OF POLYAGGLUTINATIVE GOAT SERUM.

Serum from bleeding II/27/06; fractioned III/9/06.

Organism. Fraction. 50 100 200 500 1000

Flexner

Manila | Serum +t+++}/++++/+++4] +++ + +
Totalgb. (|++++4+)+4++4++4/4++4+4+)] +++ ag
Pseudogbl. |++++4+/++4++4+] +++ ++ —
Eugbl. ++++) +++ ink

Shiga ....| Serum t++4+)+4+4+4)/4+4+4+4/)+#++#) ++
Totalgbl. |++++4+/++++4+|/+4++4+\)+++4+ 4
Pseudogbl. |++++4+/+++4+4+/+++4/)+++4+ =
Eugbl. ++++) $+ =

Pfeiffer...) Serum +t+4a4)t44 4 +- + fo Se
Totalgbl. {|++++4+/+++4) +4+4+ |*++4+ Zs
Pseudogbl. |++++4+/)%###) ++ + +
Eugbl. t+++) + =

Cholera...) Serum t++t++i+t+4+4)/+4+4+4)/4+4+4+4+)++++4+
Potala se ee ee aie ee ee eee
Pseudogbl. (+4+++/+++4+/)####) ++ +
Eugbl. ++ + +

Galon ....| Serum ++t+4]+4+4+4)/4+4+4+4+)/+4+4+4+|) ++
Totalgbl. |+++4+/++4+4+/+4+4++|/ttt4+) +
Pseudobgl. [|++++)/++4+4)/4744+/4ttF =
Eugbl t++t+ ++ —

TABLE VI. EXHAUSTION EXPERIMENT. POLYAGGLUTINATIVE GOAT SERUM.

Exhaustion with Flexner Manila of Fractions (Table V) on III/7/06.

Fraction. | Organism. | 20 | 50 100 =| 200 | 500 | 1000

Total gbl. | Flexner
| Manila |
| Shiga |
| Pleiffer.
Cholera |

Pseudogbl.| Flexner
Manila = = aig = a
Shiga 5 SP SP ae
Pfeiffer SPP AP SP
Cholera |For eiatal
Eugbl. ...| Flexner |
Manila -
Shiga aR ar
++

fea ae se ee
Sri mg mcd) owe oo 2
ciel siete ee

Pfeiffer
Cholera |

+

244 Fractionation of Agglutinins and Antitoxin

Pick’s corresponding experiment may be summed up as fol-
lows: An antityphoid goat serum agglutinating at 1:2600 was
precipitated in a series of progressively increasing concentra-
tions of ammonium sulphate. The initial precipitation was
at 2.6 saturation with a dilution of 1:5, agglutination being
evident at 1:500 of the unfiltered precipitated mixture. At
3.4 saturation over half and at 3.6 saturation all the agglutinin
was precipitated. Again, 20 cc. of the same serum were precip-
itated with 10 cc. of saturated ammonium sulphate solution ;
the euglobulin then agglutinated at 1:2400 (in terms of original
serum) the pseudoglobulin at 1:20. After two reprecipitations
the eu-fraction still reacted at 1:2400.

TABLE VII. FRACTIONING OF POLYAGGLUTINATIVE GOAT SERUM.
Bleeding of III/18/06; fractioned III/20/06.

Organism. Fraction. 100 200 | 500 1000

Flexner Manila......... Totalgbl. |++4++/+4+4++) ###/| +
Pseudogbl. |++++\/###tt ++ | -
IOS atic tin s «ee ot se Total gbl. Sater ar ar Ie oar 0 =
| Pseudogbl. |++4++/++4+4+)###+4) +
Pieter. (terse ose nas = Totalgbl. |++++|] *#* + sf =
| Pseudogbl. |++4+4+\*#### — =
Polar: sien ntaee es Totalgbl. (++++4+/4++++ > + 2s
| Pseudogbl. (++++/++++ ac =

Colgh: (aces eas serait Totalgbl. |++++) *++ ++
| Pseudogbl. -++++] #++ — au

POLYAGGLUTINATIVE HORSE SERUM.

(a) Antidysentery Horse Serum. Horse 284; immunized
against the Shiga, Flexner Manila and the Mount Desert strains.
Bleeding of X/3/06; fractioned X/4/06.

Here (Table VIII) the agglutinins are split between the frac-
tions, the larger part of each occurring in the pseudoglobulin.

(b) Anticholera and Antityphoid Horse Serum. Horse 254;
combined immunization against the original Pfeiffer and cholera.
Serum from bleeding on III/27/06; fractioned IV/1/06 and refrac-
tioned V/10/06. The results are given in Table IX.

With the cholera-typhoid serum, the agglutination values of
which for each organism were in the neighborhood of a dilution

Robert B. Gibson and Katharine R. Collins 245

of r:1000, the bulk of the agglutinins was found in the pseudo-
globulin (Table IX); the greater portion was also in the high
fraction when the serum was precipitated at 3.6 and again at
3.8 saturation. There is no evidence presented here that the
precipitation limits of the cholera agglutinin are in any way dif-
ferent from that of the typhoid. In the second fractionation
(V/10/06) it is seen that both the Pfeiffer and the cholera are
increased in the eu-fraction as contrasted with the result of the
first precipitation at 3.4 saturation.

TABLE VIII. FRACTIONING OF ANTIDYSENTERY HORSE SERUM.

Organism. | Fraction. | 50 100 200 500 | 1000 ; an
Flexner

Manila..... Serum tHttitteti+t¢tit+4t4t+) tet ++

Total gbl. |++++/++4++)/+4+4++| Bet] + +

Sees lefals) Baer anlar errr as eae ae | orate | Te =

Eugbl. Sigg te t| date teeta | en Se wy | —

s)he | Serum HHH tee tl+ets+l+tset| tet | -

Totalgbl. |+++4+,4+4+4+4+)4+4+4+4+)}F¢¢4+ + -

Pseudogbl. |++++4+4++4+ +++ _ _ —

Eugbl. lepers rrcarar | at: = 2 dees

Colas, 5:2). 2 Serum t++t+)/++4+4)4+4+4+4)/4+4+4+4/4+4+4++4| 0

LOH Oy Neer eo ot chy oh eos —

Pseudobel. |---|) Ft teeter) 6 = =

baer | | ee | ee = - —

The fractionation of the anticholera and antityphoid horse
serum is of especial interest because an experiment of this nature
is the most striking of E. P. Pick’s observations. Pick had found
that the typhoid agglutinin was precipitated with the pseudo-
globulin fraction in horse serum; the cholera agglutinin, on the
contrary, came down with the euglobulin. Pick, therefore,
mixed equal volumes of a typhoid and a cholera serum, and pro-
gressively precipitated 2 cc. amounts of the mixed sera at 2.8;
3.0, 3.2, etc., saturation (with a final volume of 10 cc. in each
case). A well marked separation of the cholera agglutinin into
the euglobulin and the typhoid agglutinin into the high fraction
resulted. As given in Pick’s tables, the observed agglutination
values are twice too much, since each serum was diluted a half
by the mixing. A direct separation at 3.4 saturation of the
agglutinin in a goat cholera serum (agglutinating at a dilution of

246 Fractionation of Agglutinins and Antitoxin

TABLE IX. FRACTIONING OF ANTICHOLERA-ANTITYPHOID HORSE SERUM.

1. Precipitation at 3.4 saturation.

Organism. Fraction. 50 100 200 500 1000
RE As Teach CP ee
Pfeiffer..... [Total gbl. j+++titttetittttit +++) bee
Pseudogbl. |++++|/++++/++++) +++ 2
Eugbl. ao + _ _ =
Cholera..... Total gbl. |+++4+/++4+4+\/++4+4+/+4+4+4+|F4+44
Pseudogbl. |++++/++++| +++ | ++ a
Eugbl. + + = = aes
|

2. Precipitation ; at 3. 6 saturation.

Organism. | Fraction. 50 100 00 00 1000

a | |
Pfeiffer... .. iTotalgbl. |++4+4+/)/+++4+/++++) bette ++
|\Pseudogbl. |+-+++)/++4++/++++) $44 =
Cholera. .... ‘Total gbl HHH Ft+etlt+tti+ttti teeter
|Pscudogbl. |++++/+++4+|/++44)+4+4 4)

3. ss at 3.8 saturation.

Organism. Fraction. 50 | 100 200 500 1000
Pfeiffer... .. Total gbl. $+ tt|+e44 ++t+; bee +
|Pseudogbl. |-++++/++++ Seta +
Cholera ..../Total gbl. .J}+++4+/++++/++4+4+/)++4+4|4444
cea dhe tet tr

4. The same serum was feaia fractioned on eae

Organism. Fraction. | 50

100 200 500 1000 | 2000
Pfeiffer...|Totalgbl. |++++/++++/++++/++++4
Pseudogbl. |++++ ++++/+++4+| $+
Eugbhl |+t+++)+++4+\bbtt —
Pfeiffer* . Totalgbl. |++++.4+++4+)4+4+4+4/++4+4)bbbitt+
Pseudogbl. |++++)+++4+4+/++++4+) be+
Eugbl. Jbettieeet — = —- |=
Cholera ..|Total gbl. |++++/++++/4++4+4/44+4+4/) +44
Pseudogbl. |++++ ++++/++4++) bet | +4
[Eugbhl (+++4++/)++++| +++ in 2
|

* A second fractionation.

Robert B. Gibson and Katharine R. Collins 247

1:1000) from the typhoid agglutinin in horse serum (1:20,000)
was made. The euglobulin was precipitated by adding a half
volume of saturated ammonium sulphate solution directly to the
mixed sera. The high agglutination values are again given as
direct observations, though probably calculated for the original
undiluted sera. The figures for the reprecipitated euglobulin
are for the typhoid, 1:3000, and for the cholera 1:1600, (an
increase over the value of the original goat serum); the pseudo-
globulin (3.3 saturation filtrate) agglutinated the typhoid at
1:16,000, the cholera at 1: 20.

TABLE X. FRACTIONING OF ANTICHOLERA-ANTITYPHOID HORSE SERUM.
Horse 254; bleedings of II/27/06 and V/29/06; fractioned X/8/06.
Fractions tested with the Mt. Sinai culture of typhoid.

Centrifuged. Fraction. | 100 | 200 | 500 | 1000 | 2000
11/27/06 | | |
2 hrs. after pre- | | | |
cipitation.../Totalgbl. |++4++/)/++4+4+/++4++/)-##*+ | +
Pseudogbl.|++++| +++ | #*#+ +
Eugbl. oe el ee ee
pbeeere tse) botal eb). | )- > ee ee ee tt ee
Pseudogbl.|++++]/ +++ |#t#t#F| £
Eugbl. et trika ee AARC cs oe dN sale
V/29/06. | | |
Serum PAH ttt+ ++ tt beet +
After 2hrs....|Totalgbl. |4++++)/++++/++++| ee | +
Pseudogbl. |++++/)+++4+/) tt ++ ) +
Eugbl. [geengret eerie eae ae aly ease | gic
After 12 hrs....|Totalgbl. |++++4+/+++4+)+++4+/ b+44 +
iPseudogbl.|/++++4+|) +++ ) #+#+ | ae
Eugbl. t+tt ++ Ee

In the following observations it is shown that a relatively large
proportion of the typhoid agglutinin may occur in the euglobulin
fraction.!

Six cc. of each serum were diluted with 13.8 cc. of water and
precipitated by the gradual addition of 10.2 cc. of saturated
ammonium sulphate solution. Uniform samples (15 cc.) of each

1 On resuming this problem in the fall of the year, it was found that our
cholera culture was spontaneously agglutinating; it could not therefore
be employed in testing the agglutination values of the fractions.

248 Fractionation of Agglutinins and Antitoxin

were centrifuged after two hours’ standing. The euglobulin pre-
cipitates were washed by thoroughly suspending the proteid in
1s cc. of 3.4 saturation ammonium sulphate solution and again
centrifuging; the precipitates were washed three times in this
fashion. Exactly similar precipitations were made on the two
sera but the precipitated mixtures were centrifuged after twelve
instead of two hours standing. The results are givenin Table X.

A high typhoid agglutinin content in the euglobulin was also
obtained on fractioning the last bleeding of the cholera-typhoid
horse. The results are shown in Table XI.

TABLE XI, FRACTIONING OF ANTICHOLERA-ANTITYPHOID HORSE SERUM.

Horse 254; bled X/3/06; fractioned X/4/06; tested only with typhoid.*

Organism. | Fraction. | 100 200 | 500 1000 2000
Mt. Sinai Ty- | |

BHO. cae Protaliebt ibe leteet che dah tett + 4+

Pseudogbl. |++++/++++4+/++++4+| ###+] —

Eugbl. bet tittt+| FS = --

St. Typhoid ...)/Total gbl. +++4+) +444) bttt ue fi

(Pseudogbl.|++++)/+++4+)httt ++ 2

Eugbl. ++4+4i+4+44+) $¢+4+ = —

* The Pfeiffer strainagglutinated spontaneously. The Mount Sinaistrain
has always shown the same agglutinations as the Pfeiffer in numerous
other observations.

FRACTIONATION OF DIPHTHERIA ANTITOXIC GOAT AND HORSE
SERUM.

Fresh goat serum and serum from horse 307, VII/5/06, were
fractioned as follows: 3 cc. of serum diluted with 6.9 cc. of
water were precipitated with 5.1 cc. of saturated ammonium
sulphate solution, and the precipitate from a1rocc.sampleof each
obtained by centrifuging. The precipitates were suspended in
3 cc. of 3.4 saturated ammonium sulphate solution and again
centrifuged; the washing was twice repeated and the wash solu-
tions united and made up to 1rocc. The precipitates were sus-
pended as usual in 3.4 saturated ammonium sulphate and made

up to rocc. The results calculated per cc. of the original undi-
luted serum follow:

Robert B. Gibson and Katharine R. Collins 249

Fraction. Goat. Horse.
SErumesyes scree seeks oto ares 90 units 250 units
Totalislotsalinwer ee oe COO 250 ~
Pseudoglobulina. +... .- 35) = +200 ‘“
Bavlobulin ee ae Oe OF
IWiashesolutionmeeereieel ate RQ) Se 2 bn Fs

* Tested for 5 units against 100 m.1.d.; the guinea pig died in 12 hours,
autopsy showing a typical diphtheria toxin picture.

The same results were obtained in two similar experiments.
The two sera certainly show a different behavior towards ammo-
nium sulphate precipitation. A relatively large proportion of the
goat antitoxin is precipitated with the euglobulin. The facility
with which the antitoxin can be washed out almost completely
(in a total united volume of wash solution less than the original
volume of the precipitated mixture) shows that the antitoxin is
not invariably linked to the euglobulin.

Pick’s experiment is given briefly for comparison with our
results:

Twenty cc. of antitoxic goat serum! (neutral reaction) were precipitated
with ro cc. of saturated ammonium sulphate solution. After two hours,
the precipitate was pressed out and dissolved in 30 cc. of water; it was
reprecipitated at 3.3 saturation and dissolved in 20 cc. of water (euglobulin
fraction). The filtrates were united and made up to half saturation, the
precipitate dissolved in the original volume (20 cc.) and refractioned
between 3.3 and 5.0. The euglobulin then obtained was united with the
above euglobulin solution and the mixture precipitated at 3.3 saturation.
The serum contained about 10 units per cc. The fractions tested as
follows:

1 Pick states (p. 361) ‘“Zu dem nun folgenden Trennungsversuche mit
Diphtherie immun Ziegenserum stand mir nur ein Ziegenserum zur Ver-
fiigung, von dem o,1 ccm eben im stande war, die 10 fache tédliche Gift-
menge eines Toxins zu paralysieren, das in der Dosis von 0,0098 ccm ein
Meerschweinchen von etwa 260 g in drei Tagen tétete.”’ This would
make the potency of the serum used only r unit perce. The control test
made actually gives the strength as 1o antitoxin units per cc.: 0.098 ec.
toxin (10 m.].d.) were neutralized by o.o1 cc. serum; therefore roo m.1.d.
toxin were neutralized by o.1 cc., or 1 cc. serum neutralized 10 X I00
m.l.d. toxin. Ledingham has passed this over: ‘‘The goat serum with
which Pick worked had a very low antitoxic value inasmuch as o.1 cc. was
required to neutralize 10 lethal doses of a toxin whose m.1.d. was only
about o.o1 cc.’’ Pick, himself, speaks of the antitoxic valueof this serum
incidentally ‘‘Man erkennt trotz der geringen Wertigkeit des Serums....”’

250 Fractionation of Agglutinins and Antitoxin

Pseudoglobulin:
0.05 cc, +10 m.].d. toxin (testing for 2 units). The guinea pig died on
the second day.

Euglobulin:
o.or cc. +10 m.1.d.toxin (testingforrounits). Diedon the third day.
0.017 cc. +10 m.l.d. toxin (testing for 6 units). Induration and loss
of weight, but survived.

Ledingham states in his conclusions that in the horse serum
the relationship of the diphtheria antitoxin to the pseudoglobu-
lin fraction ‘holds good only when the antitoxin content of the
serum is steadily rising.’ Horse 307 of this department had
been subjected to immunization for over five months; it attained
a maximum of over 300 units per cc. in three months and had
declined to 250 units two months later when the blood of the
serum used in our fractionation experiments was drawn.’ Ap-
parently Ledingham’s conclusion (from observations on a single
horse) is not of general application.

Tabulating Pick’s results by a somewhat different arrange-
ment than the one presented in his paper (p. 384), it is seen from
the following—

| . . |
. | Diphtheria Tetanus Cholera Lysin Typhoid Cholera
Animal. Antitoxin. | Antitoxin. ‘(Pfeiffer.) Aeaaain: Agglutinin.
GO8b. oon ka eugbl. eugbl. eugbl. | eugbl. eugbl.
Rabbit ..... eugbl
Guinea pig. .| | | eugbl.
Horse 23 x's 22 pseudogbl. pseudogbl. | pseudogbl. | eugbl.
|

—that there is noevidence of any differencesin the precipitation
limits of the antibodies in goat, rabbit and guinea pig sera. We
should hardly expect a priori, then, that a separation of any
other antibodies by fractionation of goat and rabbit serum would
be possible, and we have not found otherwise. Yet the prob-
ability of such a separation for horse serum was suggested by
Pick’s experiments. With this serum, even, the distribution of
the antibodies as determined by Pick, has been similarly homo-
geneous with the exception of the cholera agglutinin. Pick,
accordingly, gives one example, and one only, of an antibody

' The horse subsequently was killed as no longer of service for antitoxin
production.

Robert B. Gibson and Katharine R. Collins 251

differing from other antibodies in the serum of the same species
by its precipitation characters toward ammonium sulphate.
This observation of Pick’s we have been unable to verify when
polyagglutinative horse sera have been used. It is probable,
however, that the serum globulins of different animals or even
of various individuals of the same species may show a different
and inconstant behavior quantitatively toward fractional ammo-
nium sulphate precipitation. We have not found that any of the
antibodies in goat, rabbit and horse serum were invariably
associated with the euglobulin. Our results with goat diphtheria
antitoxin have been confirmed by Ledingham. At the same
time we have presented repeated observations with several
strains of typhoid showing that a large proportion (almost half
in some instances) of the typhoid agglutinin of horse serum may
be found in the thoroughly washed euglobulin fraction of both
old and fresh serum.

The results of our experiments have already been briefly sum-
marized in the preceding portion of the paper.

It is to be hoped that any future work on the fractionation
of the antibodies or of the proteids of the blood will not be under-
taken without a thorough comprehension of the nature and limi-
tations of the process. Salt fractionation is a valuable method
for the purification and preparation of proteid products; the
salt concentration precipitation limits, however, are not a reliable
means for differently classifying proteids the precipitation char-
acters of which are not widely separated.

“56 ee |

.

THE FRACTIONAL PRECIPITATION OF ANTITOXIC SERUM.
By EDWIN J. BANZHAF asp ROBERT BANKS GIBSON.

(From the Research Laboratory of the Department of Health of the City of
New York, Wiliam H. Park, M_D., Derector.)

(Received for publication, May 31, 1907-)

Comparatively little attention has been paid to the fractional
precipitation of antitoxin. Brodie’ in 1897 separated antitoxin
horse serum into four fractions by the progressive addition of
ammonium sulphate to half saturation; all four contained, how-
ever, relatively equal amounts of antitoxm. Atkinson” m this
laboratory saturated with sodium chloride a solution of the
moist serum globulin precipitate obtaimed with magnesium sul-
phate, and by then employing heat differentiated the globulm
into several fractions contaiming antitoxim. The protective
properties corresponded roughly to the quantities of serum globu-
lim im the precipitates. In some unpublished experiments he
found that alterations of the amounts of coagulated proteid m
the several fractions resulted if more magnesium sulphate was
added before heating; there were proportionate changes im the
distribution of the antitoxin. Owing to the destruction of 2
portion of the antitoxin at the higher temperature and possible
injury by exposing it to heat of less degree, this fractionation
must be considered as incomplete and does not exclude a purt-
fication of the antitoxin by salt fractionation. The work of E-
P. Pick on the ammonium sulphate fractioning of the anti-bodies
has been referred to in the preceding communication. Our own
experiments have resulted somewhat differently from either
those reported by Atkinson or by Pick, and have developed some
new and suggestive facts.

On the basis of the solubility of the antitoxic proteids m
saturated sodium chloride solution, one of us (Gibson) recently
devised a method for the partial purification and concentration

1 Brodie: Journ. of Path. and Bact., tv, p. 460, 1897-
2 Atkinson: Journ. of Exper. Med_, v, p- 67, 1g0r-

253

254 Fractional Precipitation of Antitoxic Serum

of antitoxin.! This consisted in precipitating the diluted plasma
with an equal volume of saturated ammonium sulphate and sep-
arating the antitoxic proteids by extracting the precipitate with |
saturated sodium chloride solution. We now have employed
the method of salt fractionation to study further the concen-
tration of antitoxin.

There exists at the present time considerable confusion in comprehend-
ing the methods and basic principles of ammonium sulphate fractional
precipitation of proteids. The nomenclature which we have employed
and which designates the number of cc. of saturated ammonium sulphate
solution in ro cc. of the precipitated mixture has been used by some author-
ities; it avoids the confusion developed by the use of such terms as “‘per
cent (NH,),SO, solution”’ and ‘‘ per cent saturation (NH,),SO,”’ “per cent
of saturated (NH,),SO, solution’’ and ‘per cent saturation (NH,),SO,
solution"’ and it seems the simplest and best practical expression of de-
grees of saturation yet suggested. We advise that this method be em-
ployed in future papers on fractional precipitation.

Mann, in his version of Cohnheim’s ‘‘ Chemie der Eiweisskorper,’’ states
(p. 292): ‘‘As the solubility of ammonium sulphate is 76.8° (per cent ?)
at room temperature, it is easy to calculate what percentage of ammonium
sulphate is required for bringing about incipient and complete precipita-
tion of any one albumin, as soon as we know what amounts of saturated
ammonium sulphate have to be added for any given quantity of fluid.”
The simplicity of the above method of calculating vanishes when atten-
tion is drawn to the fact that while 100 parts of water dissolve 76.8 gms.
of dry ammonium sulphate, the volume resulting is increased to 141 cc.
so that roo cc. of the saturated solution actually contain approximately
54 gms. of the salt, and the degree of saturation as indicated by the con-
tent of dry ammonium sulphate must be calculated with reference to the
latter figure. An example will make clearer the above statement: To
obtain a concentration of ‘‘half saturation’? ammonium sulphate, equal
volumes of the proteid solution and of saturated ammonium sulphate
solution are mixed; according to the apparent meaning of Mann's obscure
statement, 100 cc. of the resulting ‘“‘ half saturated’’ solution would con-
tain 38.4 gms. of the dry salt; it actually does contain 27 gms. of ammo-
nium sulphate.’

‘The literature on the purification and chemical characters of anti-
bodies has been briefly reviewed in a paper on ‘‘ The Practical Concentra-
tion of Diphtheria Antitoxin for Therapeutic Use,”’ this Journal, i, p. 161,
1906, and more recently by Ledingham: Journ. of Hyg., vii, p. 65, 1907.

? Because of the change in the volume of the solvent on adding the salt,
it is similarly not possible to add 38.4 gms. of ammonium sulphate to 100
cc. of water and have a solution at ‘‘half saturation.’’ In this case the
volume would be increased to 120.7 cc. and 100 cc. would contain 31.7
gms. of the salt.

Edwin J. Banzhaf and Robert Banks Gibson 255

E. P. Pick has fallen into the same error in his paper on the fractionation
of the anti-substances in the globulins of serum. He speaks (p. 356) of

the limits of the various serum fractions as follows: ‘dass das von Reye
aus ‘normalen Pferdeserum abgeschiedene Fibrinoglobulin entsprechend
einer Sattigung von 21.5 Proz. Ammonsulfat.... ...ein bestimmter Teil

(the euglobulin) des nun iibrig bleibenden Globulins keine antitoxische
Wirkung hatte und dass sich dieser aus dem Serum noch bequem abschie-
den liess, wenn die Flussigkeit einer Gehalt von 25.6 Proz. an Ammon-
sulfat enthielt. Es verblieb nunmehr ein Eiweisskérper in Lésung- (the
pseudoglobulin) der durch weiteres Eintragen der gesittigten Ammon-
sulfatl6sung bis zu einem Gehalte von 38 Proz. von dem Serumalbu-
min gut zu trennen ist und den Heilkérper in quantitativer Ausbeute
enthalt.”” The precipitation limits are distinctly designated here by
21.5, 26.6 and 38 per cents of ammonium sulphate in the precipitated mix-
ture. They actually mean a content of 2.9, 3.33, and 4.9 cc. of saturated
ammonium sulphate solution in ro cc. of the precipitated mixtures, which
would then contain, respectively, 15.67, 18.00 and 26.50 gms. of the dry
ammonium sulphate per roo cc.—figures which by no means or method
of interpretation can be logically expressed by 21.5, 25.6 and 38 percent-
ages of ammonium sulphate. Fortunately the fault lies in the nomencla-
ture only, the precipitations being accomplished by the use of saturated
ammonium sulphate solution.

Twenty liters of plasma (475 units per cc.) were diluted with
20 liters of water; by fractioning with saturated ammonium
sulphate solution, the three proteid precipitates were obtained
which separated at concentrations corresponding to 3.3 cc.,
3-3-3-8 cc. and 3.8-5.0 cc. of the saturated salt solution in 10 cc.
The saturated sodium chloride soluble (antitoxic) globulins of
these fractions and of the 5.0 saturation precipitate of a second
20 liters of the plasma were prepared as usual. Proteid deter-
aa (coagulations) and potency tests were duplicated.

Prep. 77

A. B. Cc: D.
Fractions. 0.0—5.0 0.0—3.3 3.3-3.8 3.8-5.0
Gielen CCl aS oa ere 5200 1440 1400 2050
PE BCECE icc in et - L450 1150 1350 1750
Times concentrated........... 3.05 2.42 2.84 3.68

RPercentirecovered..:........: 79.3 17/34! 19.9 37.8
Broteid, gms. per 100cc........ 11.66 ig ey | 9.87 9.70
mits; per em: proteid... 2.2... - 12436 10000 13666 18000

A second experiment with a 450 unit plasma gave the following results:

256 Fractional Precipitation of Antitoxic Serum

Prep. 82

; A. B. C. D:
Fractions. 0.0-5.0 0.0-3.3 3.3-3.8 3.8-5.0
WolumO C0. o.5. vandies aint oases = 6240 1350 1640 2550
Units per CC... icine wage wees 1050 900 1300 1600
Times concentrated........... 2.34 2.00 2.89 3.56

Per cent recovered. ...-...«»++ 72.8 13.9 22.6 45.3
Proteid, gms. per 100 cc........ 10.59 12.06 13.46 13.41
Units, per gm. proteid......... 9914 7464 9655 11930

These observations show that the antitoxic globulins of the
higher fraction are much more potent than those of the less sol-
uble proteids.

Both the preparations by the half-saturation ammonium sul-
phate method and by fractioning, when precipitated from the
saturated sodium chloride solution and dialyzed, contained’ a
probably partially denaturalized antitoxic globulin; this had a
diminished solubility and antitoxic potency (per gram proteid)
and was precipitated on slight acidification by diluting twenty
times. The filtrates from the water-acid precipitable globulin
coagulated at 73°, while saline solutions of the precipitates so
obtained showed varying and much lower coagulating tempera-
tures. The solutions of the high proteid fractions have a peculiar
green color. A redetermination of the precipitation limits of the
globulin in the three fractions after removal of the water-acid
precipitable proteid, showed that the different precipitation limits
were relatively characteristic for the fractions.

The following results were obtained on progressively fraction-
ing (in two experiments) by the addition of the dry salt! to a

' Calculations or reference tables for the amounts of salt to be added
to produce or raise a proteid solution to any desired concentration may
accurately be made by employing the following formula:

UP (s— 6)

10 — € PCy where x is the number of gms. of salt to

be added to give the required concentration, v the original volume
in cc., e the increase in the volume of the solvent by 1 gm. of salt,
p the gms. of salt per cc. of its saturated solution, and c, and c, are the
initial and desired degrees of saturation, expressed as cc. in 10 cc. For
(NH,),SO, e and p may be regarded as approximately 0.54; then

v(C.—C¢,) UC»
18.158— 0.54¢,’ es ae oe ce.

X= =
18.158— 0.54 ¢,

To raise the concentration by the addition of saturated salt solution the

Edwin J. Banzhaf and Robert Banks Gibson 257

liter of about 400 units antitoxic plasma. The initial dilution
was 1:5. The precipitates were pressed between filters and
extracted with saturated sodium chloride solution. The deter-
minations on the filtered extracts are given per cc. of the original
plasma. The results are roughly quantitative only, loss of the
filtrate in pressing out the precipitated globulins being disre-
garded. Proteid determinations and potency tests were dupli-
cated.
FRACTIONING OF PLASMA 300, 8/1/06.

Fractions. Proteid per cc. Units per cc. Units per Gram of

" Proteid.
0.0-3.4 0.00321 . 25 7788
3.4-3.6 0.00223 20 8968
3.6-3.8 0.00432 47 10879
3.8-4.0 0.00416 52 12500
4.0-4.2 0.00408 60 14705
4.2-4.4 0.00272 55 20220
4.4-4.6 0.00191 40 20942
4.6-4.8 0.00163 32 19632
4.8-5.0 0.00111 19 7,
5.0-5.6 0.00428 18 4205
0.0-3.4 0.00394 27 6853
3.4-3.6 0.00219 20 9132
3.6-3.8 0.00397 45 11335
3.8-4.0 0.00336 50 14880
4.0-4.2 0.00332 60 18072
4.2-4.4 0.00255 55 21568
4.4-4.6 0.00181 40 22094
4.6-4.8 0.00147 30 20408
4.8-5.0 0.00093 18 19355
DAO-ORGE etmek Miceeas 18 > »008 Paasee

In each instance there is a progressive increase in potency as
the antitoxic globulin becomes more soluble in the fractions
until a concentration of the salt of about 4.2 is reached. The
potency per gram remains then practically constant at about
three times that of the saturated sodium chloride extract of the
euglobulin fraction (0o.o—-3.4) until between 4.8 and 5.0 saturation;
above this limit the potency per gram rapidly diminishes to a
relatively very low figure. Between the 4.2 and 4.8 limits, over

amounts (cc.) of the original proteid solution and of the saturated salt
solution in the mixture are calculated; also the amouutt of the salt solution
necessary to bring the proteid solution to the desired concentration. Suf-
ficient excess of saturated salt solution over that already present is added
to make the required total.

258 Fractional Precipitation of Antitoxic Serum

half the units of the original plasma are precipitated, while the
antitoxin is contained in less than one-third of the total antitoxic
globulin. .

The fact that the major portion of the antitoxin remained
soluble at a concentration of 4.2 saturation, led us to investigate
whether the protective material was mechanically precipitated
with the proteid of the lower fractions. Such a result seemed
a priori improbable because the individual fractions were at
such frequent intervals and contained such a small amount of
the globulin precipitate as to make hardly conceivable a mechan-
ical inclusion of more soluble colloidal particles for more than a
few minutes’ duration. Our plan was to fraction the antitoxic
plasma at 4.2 saturation. The lower fraction (precipitate) was
to be dissolved in an added known volume of water, reprecip-
itated at 4.2 saturation and after standing 24-48 hours was to
be filtered. The procedure was twice repeated with the precip-
itates obtained at 4.2 saturation. The three filtrates and the
saturated sodium chloride soluble (antitoxic) globulin of the
final 4.2 saturation precipitate were examined for globulin and
antitoxic content. We were not able to separate by three times
repeated fractioning at 4.2, the antitoxin from the lower frac-
tion; over half the antitoxin brought down at first was pro-
nouncedly a constituent of the precipitate, the amount in the
filtrate from the final precipitation being very slight (though the
potency per gram of proteid was relatively high). The protocol
follows:

250 cc. of antitoxic plasma (305,8/10/06, 300+ units percc.) were diluted
with 475 cc. of water and precipitated with 525 cc. of saturated ammonium
sulphate solution. After standing 24 hours, the precipitated globulin
was filtered off. To the filtrate, tooo cc., was added 60 gms. of dry am-
monium sulphate after sufficient ammonium sulphate solution had been
employed to give 15oocc. at half saturation. The resulting precipitate
(4.2-5.5 saturation) was pressed out between filters, dissolved and made
up to 200 cc. ;

The precipitate at 4.2 saturation was pressed out between filter paper,
dissolved by the addition of 580 cc. of water and reprecipitated with 420
ce. of saturated ammonium sulphate solution. The total volume of the
precipitated mixture was slightly over rooo cc. After standing 24 hours,
the reprecipitated globulin was filtered off from ‘‘ Filtrate I’’ (goo cc.).

The precipitate from I was dissolved in 580 cc. of water and precipitated
with 420 cc. of saturated ammonium sulphate. Filtrate II was 930 cc.

Edwin J. Banzhaf and Robert Banks Gibson 259

Filtrate III similarly obtained amounted to 950 cc. The globulin pre-
cipitate was extracted with 1000 cc. of saturated NaCl solution.
Determinations of proteid and antitoxic content are given per ce. of

the original plasma.
Units per Gram of

Proteid per ce. Units per ce. Proteid.

Filtrate Wetec 0.00270 40 14810

e (Dg Ses eee 0.00332 50 15080

bi TEL eee eae 0.00075 12 16100

4.2 Ppt. gbl. (ext. NaCl) 0.01066 125 12100

Ae — OL OM is Fe nece Ke 0.00685 80 11680
0.02428 307

Further fractioning after complete removal of the water pre-
cipitable globulin was done on 50 cc. of the globulin solution, Prep.
77 A (cf. p. 255). The fractioning was made at a dilution of the
original preparation of 1:20. The results are expressed per
ce. of the original undiluted preparation.

REFRACTIONATION OF PREPARATION 774A.

Units per Gram of

Fraction. Proteid per ce. Units per ce. Proteids

0.0-4.0 0.0408 400 9791

4.0-4.4 0.0165 225 13667

4.4-5.0 0.0176 375 21306

5.0+ 0.0018 75 41722

4.8-5.5* 0.0046 150 34783
* Made on a second 50 cc. of the same preparation.

The refractioning of 77A from which the water-acid precipit-
able globulin had been removed, showed a marked progressive
increase in potency hand in hand with the greater solubility of
the proteid.

Fraction 3.8—5.0 of Prep. 82 was refractioned without remov-
ing the water-acid precipitable globulin. The dilution was 1:1o.

REFRACTIONATION OF PREPARATION 82D. (High Potency Fraction.)
Units per Gram of

Fraction. Proteid per cc. Units per ce. Proteids
0.0-4.0 0.07318 600 8136
4.0-4.2 0.01779 240 13490
4.2-4.4 0.02197 260 11840
4.-4-4.6 0.01232 160 12990
4.6-4.8 0.00708 90 PAA i
4.8-5.0 0.00511 80 15670
5.0-5.6 0.00197 90 45690

0.13941 1510

For 82D. 0.1341 1600

260 Fractional Precipitation of Antitoxic Serum

82 D contained a globulin of rather uniform potency per gram
from fractions 4.0-4.8; then a marked jump for the fraction
5.0-5.6 to about three times the original potency per gram was
observed. The portion of antitoxin in the highest fractions
of 77A and 82D was less than 6 per cent of the total units.
Prepared for administration as is the ordinary antitoxic globulin,
the resulting product would have had a potency of from 5000—
6000 units per cc.

The high antitoxic potency per gram proteid of the globulin
in the preparations precipitable between 4.8 and 5.6 saturation
led us to attempt to obtain such a product in bulk from the anti-
toxic globulin preparations (Gibson) and directly from plasma.
Our experiments have comprised (1) the influence of the reaction
of the plasma, (2) repeated extraction (with 4.8 saturation am-
monium sulphate) of the globulins to dissolve out the mechanic-
ally precipitated highly potent antitoxic substances, and finally
(3) progressive denaturalization of the globulin by repeated ex-
tractions with saturated sodium chloride solution and repre-
cipitation with thesulphate. Our results, however, have not been
encouraging. The protocols are given below:

(1) a. 250 cc. of antitoxic plasma (262, 10/22/06, 500 units percc.),
were diluted with 1050 cc. distilled water and precipitated at 4.8 with 1200
cc. of saturated ammonium sulphate solution. After standing three hours
it was filtered and the filtrate (2250 cc.) raised to 5.0 saturation by adding
28.6 gms. of dry (NH,),SO,. After 24 hours at room temperature, the
half saturation precipitate was filtered off, pressed between filters and
made up to 225cc. Of the 5.0 saturation filtrate, 2200 cc. were precipi-
tated at 5.6 saturation withg6 gms. of dry (NH,),SO, and filtered after 24
hours’ standing at room temperature. The precipitate was pressed out
and made up to 218.5 cc. in distilled water.

b. 250 cc. of the same plasma were made distinctly alkaline with
5 NaOH and then fractioned exactly as in a.
c. 250 cc. of the plasma were made distinctly acid with dilute
acetic acid, and then similarly fractioned.
Proteid coagulations and potency tests were made as usual.

Plasma Fractions Proteid per cc. Units per ce. Units pera
a. ‘Native: ...%; 4.8-5.0 0.0015 1 7333
5.0-5.6 0.0051 20 3921
b. Alkaline... . 4.8-5.0 0.0011 11 10000
5.0-5.6 0.0047 20 4255
Ge PCIe crema % 4.8-5.0 0.0012 iit 9175
5.0-5.6 0.0060 24 4000

Edwin J. Banzhaf and Robert Banks Gibson 261

(2) 5o0occ. of the plasma of the bleeding employed for reprecipitation
at 4.2 (305, 8/10/06, 300 units per cc.)! were diluted with 800 cc.
of water and precipitated at 4.8 saturation with 1200 cc. of saturated
ammonium sulphate solution. After standing for 24 hours, the precipi-
tatewasfilteredoff. Of thefiltrate, 2250cc. were precipitated, at about 5.6
saturation, with 116 gms. of dry (NH,),SO,; after standing, the precipitate
was separated, pressed out between filter paper and made upin solution to
450 cc. with water (fraction 4.8-5.6). The moist precipitate obtained at
4.8 saturation was thoroughly suspended in about 1500 cc. of 4.8 saturation
(NH,),SO, and filtered after standing for two days, during which time
the mixture was occasionally shaken up. The precipitate from the first fil-
trate (1) was reéxtracted as before, this procedure being carried on, in all,
four times. The precipitate then remaining was made up with saturated
NaCl solution to a volume of 1000 cc. Proteid determinations and the
antitoxin tests were made on the fraction 4.8-5.6, on the four filtrates
and on the NaCl extract of the residue, and are tabulated as before per
cc. of the original plasma.

Proteid per cc. Units per ec. Units per Gram of

roteid.
ened! BO Gisk awa) See 0.00191 12* 6316
Riltrate T. . 2. .05% 0.00240 10* 4170
lis) eee 0.00160 g* 15000
eS eee 0.00026 8* 30770
Ve eee 0.00030 4* 13333
IVESIGIIC. oo. 4s os eee 0.01784 250 14020
0.02431 293

* Because of the low antitoxic and high (toxic) (NH,),SO, content, the
tests were made with 25 or 50 m.1.d. instead of the 100 m.I.d. ordinarily
employed.

The amount of antitoxin and of proteid in the filtrates was so
small that slight errors in the determinations would influence
greatly the calculations of the antitoxin units per gram of pro-
teid. Yet the results obtained on Filtrate III, when the figures
on the preparations are recalled (p. 259), make it highly probable
that a very small portion of the antitoxin can be separated in a
much more highly potent form than is the case for the bulk of
this substance.

Progressive denaturalization of the proteid as a means of sep-
arating the globulin from the antitoxic substance, if other than
the serum globulin itself, has not proved successful. The method
used was to extract the ammonium sulphate globulin precipitates
of plasma or the antitoxic globulin preparation with saturated

1 Cf. pp. 258 and 259.

262 Fractional Precipitation of Antitoxic Serum

sodium chloride solution, to filter off the insoluble globulin residue
(after some days standing), and to reprecipitate the filtrate by
the addition of a little over half its volume of saturated ammo-
nium sulphate solution. The extraction of this last precipitate
with the sodium chloride and the precipitation of the filtrate
with ammonium sulphate followed. This procedure was carried
on 4-6 times on the 4.8 residue of (2), and on two antitoxic glob-
ulin solutions, which were obtained by the sodium chloride extrac-
tion of 3.8 saturation precipitates (less potent fraction), one of
which was already thoroughly denaturalized in preparation
because of the accidental partial desiccation of the acidified
saturated sodium chloride precipitated proteid before dialysis.
The final filtrates of one of the sodium chloride extracts of the
originally denaturalized antitoxic globulin preparations had a
potency of almost 1s5,ooounits per gramof proteid. Theglobulin
solution of the 4.8 saturation extraction precipitate and the
second antitoxic globulin preparation contained about 10,000
units per gram of proteid.

The injection of the antitoxic globulins of the various globulin
preparations sensitizes guinea pigs to subsequent, otherwise non-
fatal intraperitoneal administration of serum (Smith and Rose-
nau and Anderson). Injected intraperitoneally into sensitized
guinea pigs, the typical convulsions produced by serum are
incited, and the deaths of the animals may ensue. Rashes of
the urticarial character with little or no accompanying consti-
tutional symptoms may follow the therapeutic administration
of the several fractionally precipitated antitoxic globulins. Ther-
apeutically there is no difference in the results obtained with the
equivalent unit injections of either the high (3.8+saturation)
or low (3.3 saturation) fractions of preparations 77 and 82 (pp.
255 and 256).

CONCLUSION.

From the data presented, it appears that the saturated sodium
chloride soluble serum globulins of the higher fractions are uni-
formly much more potent per gram of proteid in antitoxin than
are those precipitated by lower concentrations of ammonium
sulphate. Between concentrations of the sulphate of 5.0 and
5-6, a small proportion of the total sodium chloride soluble glob-

Edwin J. Banzhaf and Robert Banks Gibson 263

ulin of the antitoxic globulin preparation (Gibson) or of a higher
fraction of the same is precipitated; the solution of this globulin
has a protective power of over 40,000 units per gram of proteid.
The direct fractioning of the plasma, however, does not yield
so potent a product; at a dilution of 1:5 of a 4oo unit plasma
the globulin remaining in solution at 4.2 and precipitated at 4.8
saturation has a potency of about 20,000 units per gram of pro-
teid. It is thus practicable to prepare an antitoxic solution of
over 2000 units per cc. from a relatively low plasma.

Whether or not this difference in the potency per gram of
proteid is associated with the presence of non-antitoxic globulins
having the same fractional precipitation limits as the protective
substance remains as yet undecided. It is possible that such
a variation in potency may be purely physical, associated with
the size or condition of aggregation of the colloidal globulin
particles—the less soluble larger masses having diminished anti-
toxic properties. Certainly, however, we find the antitoxin is
characterized by a wide range of the precipitation limits similar
to the soluble globulins, 7. e., in spite of repeated precipitations,
a part of the antitoxin is comparatively insoluble in concentra-
tions of ammonium sulphate in which the major portion of the
protective substance readily dissolves.

In concluding the present paper, we desire to express our
appreciation of Dr. Park’s suggestions and helpful criticism.

A STUDY OF THE PROTEOLYTIC CHANGES OCCURRING
IN THE LIMA BEAN DURING GERMINATION.

By SHINKICHI SUZUKI.

(1. Contribution jrom the Agricultural Chemical Laboratory oj the Univer-
sity of Wisconsin.)

(Received for publication, June 10, 1907.)

Assuming that during germination catabolic processes pre-
dominate in the cotyledon and anabolic processes are most active
in the growing stem, then there must be a great difference in the
nature of the various nitrogenous bodies of the seed, cotyledon
and stem at different stages of development.

In view of the interest and importance from a physiological
and chemical standpoint, of a better knowledge concerning these
proteolytic changes, the following investigation was under-
taken.

As a very satisfactory available material, the lima bean (Phase-
olus lunatus) was chosen for this study. Beans of about the
same size and in perfect ripeness were selected in order to obtain
as homogeneous a sample as possible. A part of the seeds was
ground and analyzed by the method detailed below and another
part placed in sand, kept moist and allowed to germinate at a
suitable temperature in darkness. After six days of growth the
etiolated plants were 8—12 centimeters long and had a single pair
of small leaves 1-1.5 centimeters in length. The temperature
of germination, go° F., seemed to have been favorable for rapid
growth, for thereafter the other samples did not grow so fast.
From these plants only perfectly normal specimens were selected
for analysis, while a part of the remainder was placed again in
darkness and the other part left to grow in sunlight. After
six days more of growth the plants were harvested as before.
They were now 25-30 centimeters long and had two pairs of
leaves. The green plants looked larger than the etiolated ones,
but the latter were not so perfectly etiolated as expected, the
covering having been imperfect, showing a light, pale color.

265

266 Proteolytic Changes during Germination

The difference in color and growth of the two samples, however,
was marked! The cotyledons were separated carefully from
their leaf-bearing stems.

The fresh samples of six-days-old plants were ground in a
mortar and extracted with cold water. The twelve-days-old
plants were dried at from 50-60° C. after the cotyledons had
been separated. During drying! at from 60-70° C. a slight loss
of nitrogen may have occurred. Schulze also found that after
drying the ammonia content showed a higher value, probably
due to a hydrolyzing of acid amide.

From our several materials we then had the following samples
for analysis:

L Seed.
Il. Cotyledon of 6-days-old etiolated plant.
I Il r Stem “ “ “ “ “
IV. Cotyledon of 12-days-old etiolated plant.
V. Stem “ “ “ “ “
VI. Cotyledon “ 3 Z : green plant.
VI I A St em “ “ “ “ “ “

METHOD OF ANALYSIS.

All nitrogen determinations were made by the Kjeldahl method.

A. Total nitrogen of substances.

B. Nitrogen of water extracts of substance. (100 parts of
cold water were added to one part of sample and after frequent
shaking for six or seven hours filtered.)

(1) The Total Nitrogen of the Water Extract. An aliquot part
of the water extract was taken for the estimation of nitrogen.

(2) The Nitrogen Insoluble in Water was calculated from the
result of A and (1).

(3). Nitrogen of Total Proteids. Of the various methods for
estimating proteids, Stutzer’s has been most commonly used.”
Since cupric hydroxide forms crystallizable, insoluble substances
with certain amino-acids, it is possible that under the conditions
of proteolysis some amino-acids may result and may be thrown
down along with proteids, as shown by Scherning’s work;? it is
doubtful, however, if this would be a disturbing factor in the
analysis of natural grains, fodders, etc. The chief objection to

‘E. Schulze: Zeitschr. f. physiol. Chem. xxiv, pp. 42, 43, 1898.
2 Journ. fj. Landw., p. 103, 1880.

3 Zeitschr. f. anal. Chem., xxxix, p. 545, 1900.

Shinkichi Suzuki 267

Stutzer’s method for this work is that the precipitation of albu-
moses and peptones by cupric hydroxide is not complete. Stut-
zer' states that cupric hydroxide completely precipitates albu-
moses but only partly precipitates peptones; yet in the author’s
experiments with the seeds of the lima bean it appeared to throw
down even albumoses incompletely, giving a slightly lower
result than the method in which a saturated solution of zinc sul-
phate was employed. N. Nedokutschajen’ also obtained a low
result with Stutzer’s method in an analysis of certain grains and
Frankfurth® has called attention to the incomplete precipitation
of albumoses by cupric hydroxide.

These results do not, however, lead to the conclusion that
cupric hydroxide does not precipitate completely all albumoses.
We are led therefore to consider whether Stutzer’s reagent does
or does not completely precipitate albumoses of some kinds and
only partially throws down or does not at all precipitate some
other kinds; as, for example, it has been shown that copper sul-
phate or copper acetate‘ precipitates primary albumoses (hetero-
and proto-) but does not precipitate deutero-albumoses.

In an experiment with a mixture of Witte’s peptone and the
acid decomposition products of casein, Stutzer’s method gave
a higher value than that with zinc sulphate. This mixture con-
tained:

In 100 ce. 031, * += peptones:

033 “ “ € amino-bodies.

015 gm. N by Stutzer’s method.

.012 “ “ “ zinc sulphate method.

.012 gm. N as albumoses.
In 100 ce. f :
\

To explain this result, it must be proved whether or not cupric
hydroxide precipitated part of the albumoses and part of the
peptones and their sum exceeds the albumoses precipitated by
zinc sulphate, or completely precipitated albumoses and partly
precipitated peptones, as anticipated from the above statement
of Stutzer. We must also consider what influence the presence
of amido-bodies may have had upon the yield of nitrogen. The

1 Zettschr. f. anal. Chem., XXXi, p. 505, 1892.

2 Landw. Versuchsstat., \viii, p. 273, 1903.

3 Tbid., xlvii, p. 451, 1895.

4 Hoppe-Seyler: Handbuch d. chem. Anal., p. 324. Folin: Zettschr. f.
physiol. Chem., xxv, p. 152, 1898.

268 Proteolytic Changes during Germination

above result suggests the question: Do the albumoses in seeds
have a close'relation to deutero-albumoses?

At least, we can say that cupric hydroxide does not perfectly
precipitate the albumoses of the lima bean and of some other
seeds, and consequently Stutzer’s method is not suitable for
the analysis of some plant seeds and young plants containing
albumoses.

The tannin salt method! proposed for the analysis of meat
extracts gave more satisfactory results in the hands of the author
than the cupric hydroxide methods.

(4) Nutrogen of Coagulable Proteids. The water extract was
boiled for a few minutes with the addition of one or two drops
of acetic acid, and treated in the usual way.

(5). Nitrogen of Ammonia. For the determination of ammo-
niacal nitrogen the filtrate from (4) was distilled with magnesium
hydroxide at 38-40° C. under reduced pressure. Ammonia in
the water extract of plants hascommonly been determined accord-
ing to the method of E. Bosshard? with the use of phosphotung-
stic acid. Bosshard obtained ammonia quantitatively from
ammonium chloride solution with the above reagent. In pre-
liminary tests by the author on ammonium sulphate solution,
phosphotungstic acid gave less accurate results, as shown by the
following table:

N per 100 ce. of (NH4)oSO,4 N precipitated from (NH4)2SO4
solution (in gram). solution by phosphotungstie acid.
(Per cent of total nitrogen.)
.0299 TPAe
0150 O21
0093 53.4
0047 73.0
0024 76.8
0012 60.6

These results led to the rejection of Bosshard’s method for
our purpose.

The question whether acid amide compounds, such as aspar-
agin and glutamin, lose their amide group by cleavage in heating
the solution with magnesium hydroxide has been conclusively
answered in an article by E. Schulze.? In digestion with mag-

' Journ. oj the Amer. Chem. Soc., p. 1497, 1906.
* Zettschr. f. anal. Chem., xxii, p. 3209.
* Landw. Versuchsstat., \xv, p. 237, 1906.

Shinkichi Suzuki 269

nesium hydroxide at boiling temperature under ordinary atmos-
pheric pressure, a part of the nitrogen splits off, but no cleavage
occurs in vacuum distillation at 40° C.

(6) Nitrogen of Albumoses; the solution from the determina-
tion of ammoniacal nitrogen was treated with zinc sulphate as
substituted by Bomer! for ammonium sulphate. He suggests
a correction of the results obtained because of the probable
formation of double salts of zinc sulphate with ammonium sul-
phate. In using this method on the mixtures of albumoses,
peptones and amid-bodies described under (3), containing am-
monium sulphate equal to 0.024 gram nitrogen per 100 cc., pre-
cipitation of ammonium salts did not occur, as proved by distilling
with magnesium hydroxide. The only difficulty with this method
was the slowness of filtration but this was modified so that one
worked with a definite volume of saturated zinc sulphate solu-
tion and estimated the nitrogen in an aliquot portion of the fil-
trate. According to the work of Pinkus, Shafer and Haslam?’
sodium sulphate has come into use as a substitute for zinc sul-
phate in the separation of albumoses. In experimenting with
a saturated solution of sodium sulphate, after the manner of the
zinc sulphate method, keeping the solution at 36—39° C.—above
the transition point of the two systems of sodium sulphate salts,
clear separation and rapid filtration was obtained. The fol-
lowing results indicate a close agreement with the zinc sulphate
method.

ZnSO, gave 0.284 per cent N of albumoses.
NaisO,. “0 27e ea ;

(7) The Nitrogen of Peptones was calculated by subtracting
the sum of the nitrogen of coagulable proteids and albumoses from
the nitrogen of total proteids found by the tannin salt method.

(8) Nitrogen of Diamino-Compounds; the solution from the
determination of ammoniacal nitrogen was submitted to Haus-
mann’s method for the separation of diamino-compounds from
monoamino-compounds.* In our tests the freshly prepared re-
agent was believed to have a higher precipitating value than

! Zeitschr. f. anal. Chem., Xxxiv, p. 562, 1895.
2G. Mann: Chemistry of Proteids, p. 185.
3 Discussion and criticism of this method will be found in Cohnheim’s

Chemie der Etwetsskorper.

270 ~—~Proteolytic Changes during Germination

older reagents and we thought it might be difficult to obtain
accurate results by the method; it was thought suitable, however,
for comparative work on diamino-compounds. The work was
carried out in accordance with the method proposed for the analy-
sis of cheese and milk by VanSlyke and Hart.’

(9) Nitrogen of Monoamino-Compounds; the total nitrogen
in the filtrate from the above phosphotungstic acid precipitate
was estimated as the nitrogen of monoamino-compounds.

(10) Nitrogen of Acid Amides; proteids were separated from
the water extract of the samples by lead acetate, the filtrate freed
from lead by hydrogen sulphide, the solution then digested with
dilute acid, neutralized with alkali and distilled with magnesium
hydroxide for ammonia. Subtracting from this result the value
obtained for ammonia under (5) we have the equivalent of nitro-
gen in acid-amide combinations.

TABLE I.
NITROGEN EXPRESSED AS PERCENTAGE OF DRY MATTER.

| CoTYLEDON. STrm.
pecae
Name of Nitrogen. | Seed | eaave | 12 sus aos | Fg days [eee Plant
| etiol. | etiol. | green | si | etiol. | green
N total.............| 3.023) 2.751) 2.908) 2.845) 5.116] 5.736) 5.081
N insoluble in water .| 0.698) 0.725) 0.979) 1.193) 0.177) 1.363) 1.917
N coagulable........ | 1.711) 0.701; 0.677} 0.543] trace | 0.314) 0.358
N albumoses ........ | 0.281) 0.205) 0.145) 0.148) 0.056) 0.082) 0.017
N peptones.......... | 0.005 0.021; 0.032) 0.005) 0.309 0.481) trace
N diamino-com- | |
POUNGS Ht e...« 5 | 0.023) 0.047) 0.024) trace | 0.017) trace | trace
N monoamino-com-
OHNE 0:2 = + sre si 0.307 0.920} 0.985) 0.899) 4.265) 3.330) 2.698
N ammonia......... | 0.007) 0.132) 0.066, 0.057) 0.292) 0.166) 0.093
N acid amides....... 0.016 0.112 | 0.652)
TABLE II.

THE NITROGEN EXPRESSED AS PERCENTAGE OF TOTAL NITROGEN.

1 Amer. Chem. Journ., xxix, 1903.

be Ee 1 ee ane ee 100.00 100. 00|100. 00100. 00/100. 00 100.00 100.00
N insoluble in water .| 22.79! 25.99] 33.66) 41.93] 15.91) 23.76) 37.72
N coagulable ....... | 56.59| 25.52| 23.28] 19.08] trace| 5.47) 7.04
N albumoses........ | 9.29] 7.45] 4.98) 5.20] 0.95) 1.42] 0.33
N peptones.......... | 0.16] 0.93] 1.10) 0.17] 5.14} 8.38) trace
N diamino-com- | |
POWNIS Hosi5 5 os.5:s | 0.76) 1.51] 0.82) trace} 0.28) trace | trace
N monamino-com-
pounds......... | 10.15| 33.79] 33.87) 31.59] 72.85] 58.051 53.09
N ‘ammonia ../.../.5-- | 0.23) 4.77) 2.26) 2.00) 4.85) 2.89) 1.83
| 0.52, 4.07 | | 11-14

N acid amides.......

Shinkichi Suzuki

TABLE III.

27

GRAMS OF NITROGEN IN PLANTS FROM 1000 SEEDs.

| CoTYLEDON. Str.
a 6-days l19-days Plant | 6-days |12-days| Plant
etiol. etiol. green etiol. etiol. | green
INEGOtAIRAE Es costae: * 28.50) 16.58) 7 56| 6.28] 5.93) 16.36 18.08
N insoluble in water .| 6.49) 4.37) 2.54) 2.63} 0.20) 3.95) 6.82
N coagulable ....... 16.13] 4.22| 1.76 1.20) trace| 0.91) 1.27
N albumoses........ 2.64, 1.23) 0.37) 0.32) 0.06) 0.23) 0.06
INE PeEpPtonesi sc gels, 7: 0.04). 0.12, 0.08) 0.01) 0.35) 1.39) trace
N diamino-com- |
pounds: = :.t0; 2: 0.21); 0.28} 0.06) trace} 0.02| trace | trace
N monoamino-com- |
POUDASSS . ele - 2.89} 5.54) 2.56) 1.98) 4.94) 9.65) 9.60
Neammonia. 2.22.2. - 0.06) 0.79) 0.17) 0.12) 0.33) 0.48) 0.33
INEacidramide= J... . 0.15) 0.67 |
EXPLANATION OF TABLES.

I. Cotyledons.

(1) Total nitrogen. The values of the nitrogen of dry mat-
ter are not greatly different but the figures for nitrogen of 1000
seeds and of 1000 cotyledons decrease with the growth of the
plants, simultaneously with a decrease of dry matter in the
cotyledon and an increase of that in the stem. The following
table shows the decrease of dry matter and nitrogen of the coty-
ledons.

| 12-days etiol-| 12-days green
6-days plant. fedianl : =
SEzL Cotyledon. Gotsedon, | Gaivinae
Per cent dry matter of seed . . 100.0) 75.4 32.5 | 27.6
N per cent of total nitrogen 100. 0) 58.1 26.5 22.0
Ratio of dry matter to nitro- | |
SIETAN eaten On tS, Cake UE le 20) oie pei 22 peel 26) 2

Growth continues with about the same ratio of dry matter
to nitrogen in the cotyledon of the etiolated plant, but in the
cotyledon of the green plant, this ratio increases showing that
a comparatively larger amount of non-nitrogenous substances
existsin the cotyledon while the nitrogen passes into the stem in
larger amounts than in the case of etiolated plants. This empha-
sizes an old truth that the synthesis of non-nitrogenous com-
pounds is more active in light than in darkness.

272 ~ Proteolytic Changes during Germination

(2) The nitrogen insoluble in water may be mostly that of
insoluble proteids. While the figures for the percentage of this
nitrogen to dry matter (Table I) and to total nitrogen (Table IT)
show an increase, this increase does not represent an increase of
actual weight of nitrogen in the cotyledon. In Table III the
quantities of nitrogen of 1000 seeds and of ro0o cotyledons de-
crease gradually from 6.49 grams in the seed to 4.37 grams in
6-days etiolated cotyledon; to 2.54 grams in 12-days etiolated
cotyledon; and to 2.63 grams in 12-days green cotyledon. The
following table shows the total weight of nitrogen of insoluble
proteids in 1000 plants (cotyledon and stem) :

SPR ee ea er re nee g FF an oid, Spe AE hanes e ee 6.49 grams
6-days old etiolated plant.............-----s eee e eens qe)

12-days old etiolated plant............------- eee eeee 6.49 “

i>-Gayeoil oreen plant... 2.3 0.5 ee ee we te ee ees 9.45 “

From the results of this table it appears that, after decompo-
sition, part of the insoluble proteids passes into the stem and
again rebuilds insoluble proteids. The fact that 9.45 grams of
nitrogen as insoluble proteids in the green plant is greater than
that of the seed proves that forms of nitrogen other than that of
insoluble proteids have entered into the formation of the insolu-
ble proteids in the stem. Briefly, the water-insoluble proteids
enter into physiological functions in germination, though it is
not known with certainty in what form they pass into the stem.

In view of the fact that there exist inthe plant proteids insolu-
ble in water, there is no justification for the opinion that the
water-insoluble portion cannot pass from cell to cell through
the cellulose wall or phloem. We must remember that plant
sap is a dilute salt solution, a medium in which at least globulins
and even certain albumins are to some extent soluble.

(3) Nitrogen of coagulable proteids. This group contains
the greater part of the proteids of the albumin-group and pos-
sibly some globulins, since the presence of soluble salts in the
plant can influence the composition of the water extract. The
decrease of these coagulable proteids, as shown by Tables I, II
and III, is very conspicuous, Apparently the proteids of this
class play a prominent part in the catabolic processes of the coty-
ledon. The following table gives the total nitrogen of coagulable
proteids in 1ooo seeds, and 1000 plants (cotyledon and stem):

Shinkichi Suzuki 273

COG Wyner is BE gg ayy owe Ik RY SAMS chen Santa et due sd bwelots guecaes 16.13 grams
G-daysroldremolated plantar yee ira. arcane) ete ae 4.22 5

12-days old etiolated plant ..... Sire aiaaaelciad ais Seeker chews 2.67 c

2zdaysioldioreenmlarita vests 0+ 20) pecusp es 2x3 caeusy bers es 2.47 i

From a study of this table it seems probable that the coagulable
proteids decompose in the cotyledon and pass into the stem to
again form protein substances; it would seem improbable, from
general physiological considerations, that the coagulable pro-
teids pass into the stem and there suffer proteolysis.

Osborne! states that an albumin contained in the wheat kernel
‘differs from animal albumin in being precipitated by saturating
its solution with sodium chloride or magnesium sulphate. In
this connection it was thought that it would be important to
make a similar observation on the lima bean. This test applied
to our work, gave the following results:

: Ee Dilute
| Water | Nacl
| Extract. pxtract.

Nitrogen of coagulable proteids 1 percent dry matter ....] 1.711 1.830
Nitrogen of proteids salted out by MgSO,............... 1.251

If o.119 (1.830—1.711) is the value of globulin, 1.132 (1.251—
0.119) should be the value for nitrogen of the proteids other than
globulin, such as albumins, which correspond to the proteids
found by Osborne. The above result is in harmony with that of
Osborne on the wheat in that the albumins of the lima bean are
precipitated by saturation with magnesium sulphate.

(4) Nitrogen of albumoses. The decrease of the easily solu-
ble albumoses in the cotyledon raises the question: Do the
albumoses pass into the stem without further cleavage and
form higher (more complex) proteids? This question cannot be
decided from.the results of the previous tables, but the figures
for albumoses of the cotyledons, as well as those for insoluble
and for coagulable proteids, decrease noticeably at the 6-day and
12-day stages.

(5) Nitrogen of peptones, diamino-compounds, mono-amino-
compounds and ammonia. Peptones have increased at the 6-
day stage and the 12-day stage shows a decrease. The small
amount of peptones in the cotyledons of the green plants at the

1 Amer. Chem. Journ., xv, No. 6, p. 77.

274. +Proteolytic Changes during Germination

latter stage is especially noticeable, while the green stems show
only a trace of these bodies. It seems reasonable to believe
that this dearth of peptones is the result of active chemical tran- _
sition.

From Table III it appears that the fluctuations of peptones
content follow those of diamino-compounds.

The decrease of monoamino-compounds and ammonia at the
12-day stage might result from translocation to the stem.

II. The Stem.

(1) Total Nitrogen. The nitrogen of the dry matter has
increased at the 12-days stage of etiolated stems while the 12-
days-old green stems have a little lower percentage of nitrogen,
a relation which illustrates a well known fact that the formation
of dry matter, is more rapid with than without sunlight. The
actual weights of total nitrogen and of dry matter of the stem
gradually increase, the following table showing their ratio:

Stem Stem
Stem | 12-days | 12-days
6-days | etiolated green

- plant. plant. plant.
Dry matter in 100 parts (stem of 6-days plant) 100 250 307
Nitrogen per cent of total nitrogen in stem of 6-

CS TANT ag ere ect ee eisinic encore 100 280 304

(2) Nitrogen of insoluble proteids. We have observed a
conspicuous decrease in the actual weights of coagulable proteids
of the cotyledon and now we find that the insoluble proteids
show a great increase in the stems of etiolated and especially
of green plants. This is a most interesting result and suggests
the possibility that the ‘‘fixed” or non-diffusible proteids play
an important part in the formation of the growing stem.

(3) The nitrogen of coagulable proteids increases in value
like that of the insoluble proteids though not so much. At the
6-days stage coagulable proteids were not found in the water-
extract on boiling, nor on adding acetic acid in the presence or
absence of sodium chloride. This fact had already been observed.
S. Frankfurth’s investigation of the wheat embryo! shows the
absence of coagulable proteids in the water extract but shows the
presence of albumoses. The author, thinking there might be a

* Landw. Versuchsstat., xlvii, p. 453, 1895.

Shinkichi Suzuki 275

stage of growth at which the coagulable proteids disappear
through cleavage or the formation of new substances, endeavored
to find such a stage in the development of lima and navy beans;
but plants of rapid growth, attaining a height of 8-12 centi-
meters in 6 days, were not obtained and the water extracts of
the plant of slow growth showed at each day of growth the pres-
ence of proteids coagulable on boiling with acetic acid.

(4) Nitrogen of albumoses. From Table III we see that the
albumoses have increased at the 12-days stage in darkness but
have decreased in the sunlight. It seems probable that the
albumoses took part in the formation of higher proteids in the
stem of the green plant. This supposition is supported by the
finding of traces of peptones and diamino-compounds in the
green stem and by the slightly further decrease of monoamino-
compounds and ammonia in the green stem as compared with
the etiolated stem at the above stage while at the same stage
the higher proteids show greater increase in the stems of green
plants.

(5) Nitrogen of peptones. Peptones have increased during
12-days growth in darkness and for the same time have decreased
in the sunlight. While the fluctuations of albumoses in the stem
resemble those of peptones, in the cotyledon, the value for pep-
tones resemble those for diamino-compounds.

(6) The nitrogen of diamino-compounds occurs only in
traces with both etiolated and green plants. It is supposed that
diamino-compounds, as we have assumed for albumoses, serve
in the formation of proteids.

(7) Nitrogen of monoamino-compounds. The increase of
monoamino-compounds in the two kinds of plants at the 12-
days stage does not show much difference but its amount is about
twice that of the 6-days stage, as shown in Table III. Appar-
ently monoamino-compounds are formed as by-products of
proteid synthesis; but the conception of continuous transloca-
tion of this group from the cotyledon to the stem offers some objec-
tions to any such assumption. The following table shows the
total nitrogen of monoamino-bodies in 1000 plants (cotyledon
and stem).

276 ~—-~ Proteolytic Changes during Germination

6-days 12-days ; 12-days

etiolated | etiolated green
plant. plant. plant.
Nitrogen of monoamino-bodies in 1000 coty-
| (Ca Fas ap ae reg ete ety cot ae ead ote Mea Re ee 5.54gm | 2.56gm| 1.98 gm
Nitrogen of monoamino-bodies in 1000 stem .| 4.94“ | 9.65“ | 9.60 “
Seve alts ta Wigs BIg acre orotate cee ieace e sivhe ts 10,48 112.01" ieee

In Table III, the fluctuation of amounts of insoluble and of
coagulable proteids in the 12-days etiolated and green plants
appears to have great influence on the values for albumoses,
peptones and diamino-compounds but not on the value for
monoamino-compounds. It suggests the idea that certain mono-
amino-bodies unprecipitated by phosphotungstic acid may not
be so active in the synthesis of proteids.

(8) Nitrogen of ammonia. This was estimated with the 6-
days plants in fresh condition but the 12-days plants were ana-
lyzed after drying at 50-60°C. Consequently the results on the
latter, as E. Schulze shows, should be high. Assuming that the
results obtained are higher than the actual weights, ammonia
still plays an active part in the germination, although we are
uncertain whether it is directly concerned in proteid synthe-
sis or aids in the formation of amino-compounds.

SUMMARY.

1. Cotyledon.

(1) In the cotyledon, all proteids except peptones show a
decrease at the 6-day and 12-day stages of growth which is most
conspicuous in the case of coagulable proteids.

(2) Peptones, diamino-bodies, monoamino-bodies and am-
monia show an increase at the 6 days stage and after that they
decrease, especially in the cotyledon of green plants.

(3) The increase of these substances must be due to the
decomposition of higher proteids. _

(4) The decrease of all nitrogenous substances at the 12 days
stage must be due to a translocation into the stem. Different
amounts of decrease for the two 12-day stages, in darkness and
in sunlight, satisfactorily explain the different degrees of growth
of etiolated and of green stems.

Shinkichi Suzuki 277

Il._ Stema:

(1) In the stem of the 12 days etiolated plant, all of the
nitrogen compounds show an increase, with the exception of
diamino-bodies, in comparison with those of the 6 days etiolated
plant; and the increase of insoluble proteids over the amounts
of this form of proteids at the 6-day stage is most remarkable.

(2) Table ILI shows that the amount of insoluble and coag-
ulable proteids in the stem of the 12-day green plant is notice-
ably higher than those for the stems of 12 days etiolated plants.
It is evident, therefore, that the formation of insoluble and
coagulable proteids is more active in sunlight than in darkness,
causing the decrease of albumoses, peptones and diamino-com-
pounds.

(3) The phenomenon of parallelism in the fluctuation of val-
ues for peptones and diamino-compounds in the cotyledon at
every stage of germination and growth is striking, while of equal
interest is the relation shown between the amounts of peptones
and albumoses in the stem; peptones from the standpoint of
molecular complexity, probably having a position between albu-
moses and diamino-bodies.

This work was done under the direction of Prof. E. B. Hart, to
whom the writer wishes to express his sincere thanks.

ON THE CHEMICAL PROPERTIES OF AMANITA-TOXIN.
By HERMANN SCHLESINGER anv WILLIAM W. FORD.
(From the Pharmacological Laboratory of the Johns Hopkins University.)
(Received for publication, July 1, 1907.)

It was pointed out by Kobert! in 1899 that Amanita phalloides
contains, in addition to the hemolytic ‘“‘toxalbumin,’’ Phallin,
described by him several years previously as the active principle
of this poisonous fungus, an alcohol-soluble substance killing
small animals acutely after subcutaneous administration of
minute doses. The second poison Kobert stated to be an alka-
loid, without, however, specifying any of its alkaloidal reactions,
but because of its failure to produce fatty degeneration of the
parenchymatous organs, the most striking lesion seen in fatal
cases in man, he could not accept it as the active principle.

At the same time he admitted that Phallin is not constantly
present in the plant and can therefore be no longer looked upon
as the essential poison. It was subsequently claimed by one
of us (Ford?) as the result of inoculating animals with extracts
of the fungus heated to 80° C. and subjected to artificial digestion,
whereby the hemolysins are completely destroyed, that a power-
ful toxin is present in this plant, in addition to the blood-laking
substance, ‘“‘Phallin.’’

This heat-resistant substance was called provisionally Aman-
tta-toxin.

Recently Abel and Ford? have shown that Phallin is not a
‘“‘toxalbumin”’ as supposed by Kobert, but a hemolytic gluco-
side, which is so sensitive to heat, and to a hydrochloric-acid-
pepsin mixture as to preclude its playing any réle in ordinary
human intoxications. If the aqueous extract of the fungi be
treated with ethyl alcohol, this hemolysin is precipitated while
the heat resistant Amanita-toxin is found in the alcohol filtrate.

1 Sitzungsber. d. naturforsch. Gesellsch. zu Rostock, p. 26, 1899.
2 Journ. of Exp. Med., viii, No. 3, May 26, 1906.
This Journal, ii, No. 4, January, 1907.

279

280 Amanita-toxin

Abel and Ford suggest that the Amanita-toxin is without doubt
the active principle of Amanita phalloides, probably identical
with the poison roughly studied by Kobert and assumed by him, .
on insufficient grounds, to be an alkaloid.

Finally it has been shown by Ford! that the heat resistant
alcohol-soluble Amanita-toxin when completely freed from the
hemolytic glucoside, can produce in animals the lesions found
in man, including the hemorrhage, necrosis, and especially the
jatty degeneration. It remains necessary, therefore, to take up
more fully the question of the purification of this Amantta-
toxin.

One hundred grams of fungi, previously dried over sulphuric
acid, were finely ground and the powdered material thoroughly
triturated with 300 cc. of 65 per cent ethyl alcohol. The residue
was twice treated in the same way and finally mixed with roo cc.
of alcohol of the same strength and allowed to stand over night.
The liquid of this fourth extraction was united with the first
three fractions, making a total of 1000 cc. After careful neu-
tralization with sodium carbonate the extract was evaporated
under diminished pressure until all the alcohol had been distilled
away and the volume of the remaining fluid had become reduced
to 150-200 cc. After filtering from deposited fatty acids, etc.,
the solution was made very slightly alkaline with sodium car-
bonate and treated with a 15 per cent solution of silver nitrate.
A voluminous precipitate was formed. This, being non-toxic,
was discarded. The filtrate was freed from the slight excess of
silver by means of sodium chloride and, now neutral, was treated
with a solution of basic lead acetate prepared in the usual way.
Another non-toxic precipitate was formed and was also discarded.
The filtrate? from basic lead acetate was treated with an excess
of a saturated sodium sulphate solution for the removal of lead
and to this filtrate phosphotungstic acid (10 per cent phospho-
tungstic acid in 5 per cent sulphuric acid) was added in slight

‘ Paper to be published shortly.

> Sometimes a considerable amount of toxic material was included in
the first lead precipitate. This should, therefore, be treated with a sat-
urated sodium sulphate solution and filtered. The filtrate is again pre-
cipitated with basic lead acetate. The precipitate may now be discarded
and the new filtrate treated as described above.

Hermann Schlesinger and William W. Ford 281

excess. The phosphotungstic precipitate was decomposed with
barium hydrate and the filtrate from the barium compounds
neutralized with sulphuric acid. The precipitate thus formed
was filtered off and the resulting fluid found to contain the
poisonous substance. On subcutaneous inoculation of both
rabbits and guinea pigs this fluid was highly toxic, 1 cc., con-
taining 0.0004 gram of organic and no inorganic material, killing
the animals acutely in from 24-48 hours and producing the
pathological changes characteristic of poisoning by Amanita-
toxin.

It will be observed that our toxic solution as thus obtained
is the product of a rather rigorous analytical separation. The
final fluid can only contain substances precipitable by phos-
photungstic acid, which at the same time do not form insoluble
compounds with either silver or lead and which are soluble in
65 per cent alcohol. The material must, therefore, be freed
from ordinary plant constituents. We nevertheless found it
useful to repeat the precipitation by phosphotungstic acid several
times in order to eliminate any error due to included matter.
With our purified poisonous product we were able to establish
the following points in regard to its chemical properties.

Amanita-toxin is very soluble in water, less so in 80 percent
alcohol and only very little soluble even in hot absolute alcohol;
it is insoluble in the ordinary organic solvents. Its aqueous
solution is optically inactive. It is a fairly stable compound
for it can be boiled in absolute alcohol and in aqueous solution
for some time without suffering serious loss in toxicity; it is only
very slowly affected by acids at room temperature, retaining its
toxicity for several days when thus treated. Boiling acids, how-
ever, rapidly destroy the poison. It does not reduce Fehling’s
solution either before or after prolonged boiling with 5 or 10
per cent hydrochloric acid. With the exception of phospho-
tungstic acid, this toxin reacts with none of the alkaloidal pre-
cipitants, nor does it respond to any of the alkaloidal color
reagents.!. It does not give the biuret test or Millon’s reaction.
We may, therefore, conclude that this poison is neither a gluco-
side, an alkaloid, nor a proteid in the generally accepted sense

1 For a list of these, see Kippenberger, Nachweis von Gijtstoffen.

282 Amanita-toxin

of these terms. The following reactions give us a clue to its
identity, and we are convinced that these reactions are due to
the Amanita-toxin itself because of our rigorous method of puri-
fication and because the tests become more pronounced as the
process of purification advances. Fusion with metallic potas-
sium and subsequent treatment in the usual fashion shows the
presence of nitrogen and sulphur. By boiling a concentrated
solution of the purified toxin with hydrochloric acid and sub-
sequently treating it with barium chloride the sulphur was
shown to be present as conjugate sulphuric acid.1. While mak-
ing the fusion with potassium a strong odor of fatty amines was
observed, and the gas evolved gave white fumes when a drop
of hydrochloric acid on a glass rod was brought near. To deter-
mine whether the toxin is a substance from which amines may
be split off by reagents ordinarily used for this purpose a small
portion of the dried material? was mixed in a test-tube with
powdered potassium hydrate. The amine odor was noticeable
at once, but after heating, the persistent and unmistakable
odor of indol completely masked that of the amines and a pine
splinter moistened with concentrated hydrochloric acid gave
the characteristic pyrrol red when held in the mouth of the test
tube. The application of the tryptophan test of Hopkins and
Cole gave negative results.

CONCLUSIONS.

From the reactions described above Amanita-toxin can not be
a proteid, a glucoside, or an alkaloid. Any attempts to classify
it definitely are made upon the assumption that the decomposi-
tion products we obtained are not due to very small amounts of
impurity but come from the toxin itself. Because we were
unable to obtain indol by boiling with concentrated solutions
of potassium hydrate, we cannot be sure that we have an indol
or pyrrol derivative. Nevertheless, since we are dealing with a
conjugate sulphate which, on fusion with dry potassium hydrate,
gives off pyrrol and indol we are led to conclude that Amanita-

' The solution gave no test for sulphates before boiling.
? Boiling with a solution of potassium hydrate gave no noticeable amine
odor or alkaline fumes.

Hermann Schlesinger and William W. Ford 283

toxin, although not necessarily an indol derivative, is at least
an aromatic phenol so combined with an amine group that it
readily forms an indol or pyrrol ring.

We wish to take this occasion to thank Dr. Abel for placing
material for this investigation at our disposal and for his many
valuable suggestions during the progress of the work.

V. RESEARCHES ON PYRIMIDINS: ON SOME SALTS OF
CYTOSIN, ISOCYTOSIN, 6-AMINOPYRIMIDIN
AND 6-OXYPYRIMIDIN.

(Twenty-second Paper.)

By HENRY L. WHEELER.
(From the Sheffield Laboratory of Yale University.)

(Received for publication, June ro, 1907.)

It has been shown by Richard Burian! when guanin (I), mixed
with carbohydrates, is heated with 30-40 per cent sulphuric acid
that hydrolysis and reduction take place at the same time, the
imidazole group is removed and isocytosin? (II) and uracil (IV)
result. It may be added that the probable formation and
decomposition of xanthin (III) would also give uracil. This
decomposition of guanin may be represented as follows:

HN—CO HN—CO
eee ae
H,N—C CNH. = ——+ 0C C—NH\
| cH i | oa
N= NF HN-C-a0he7
I Il
1
HN—co HN—_CO
|
HNC CH 2S ee eter
| | ||
N—CH HN—CH
u IV

When adenin (V) was treated in a similar manner, Burian
obtained 6-aminopyrimidin (VI), while 6-oxypyrimidin (VIII),
which would be expected to result in this case both by the

1 Asher-Spiro: Ergeb. d. Phystol., v, p. 795, 1905.
2 Wheeler and Johnson: Amer. Chem. Journ., Xx1x, p. 492, 1903.

285

286 Researches on Pyrimidins

decomposition of hypoxanthin (VII) and by the hydrolysis of
6-aminopyrimidin, escaped detection.’
N=C—NH, HN—CO

oa |
HC C—NH —__—_—> HC C— NE
i. \cH 1 | CH

ea see WC NH
V VI
N=C—NH, HN—CO
fan bias |
HC CH eek |= eR ere
|| | il
N—CH N—CH
VI VIII

The three pyrimidin bases, isocytosin, 6-aminopyrimidin and
6-oxypyrimidin, are probably formed in the energetic hydrolysis
of the nucleic acids by sulphuric acid and, since it has now been
found that the general reagents which precipitate cytosin also
precipitate these bases, an examination of the properties of the
compounds and some of their salts was undertaken, along with
the similar ones of cytosin. The statement of Burian that 6-
aminopyrimidin and isocytosin according to their entire behavior
must obstinately adhere to cytosin (‘‘dem sie ihrem ganzen Ver-
halten nach hartnackig anhaften missten’’) has been found to
be more especially true in the case of isocytosin.

Of the three bases 6-oxypyrimidin is new. Isocytosin was
first prepared synthetically in this laboratory”? and later it was
obtained in a different manner by Gabriel and Colman.3

In the case of 6-aminopyrimidin, Burian states that the base
was isolated in the form of the silver salt, by precipitating in
neutral solution with silver nitrate. This was decomposed by
means of hydrogen sulphide and the base precipitated by phos-
photungstic acid. A solution of the free base was then obtained
in the usual manner from which he prepared and analyzed the

‘In an article published while this paper was in press (Zeitschr. f.
physiol. Chem., li, p. 444, 1907), Burian describes the isolation of 6-oxy-
pyrimidin.

2 Loc. cst.

* Ber. d. deutsch. chem. Gesellsch., xxxvi, p. 3382, 1903.

ne

Henry L. Wheeler 287

picrate and chloroplatinate. He says nothing further in regard
to the base or its salts.

6-Aminopyrimidin, however, was first prepared by Ernst Butt-
ner, who obtained it from barbituric acid by means of a series of
operations.! This author described the free base and he pre-
pared the hydrochloride, easily soluble rhombic tables; the chloro-
platinate and picrate, needles, difficultly soluble. Our knowl-
edge of these pyrimidins stood at this point when the following
work was begun.

It has now been found that 6-oxypyrimidin is formed when
2-thiouracil (IX), which can be easily prepared? in any desired
quantity, is treated with hydrogen dioxide.

HN—CO HN—CO

| Pl

SC. CH+30+H,O= HC CH+H,SO0,
| |
HN—CH N—CH

1X

The writer finds, however, that 6-oxypyrimidin is more
smoothly obtained by warming 2, 6-dichlorpyrimidin (X), pre-
pared from uracil? or 2-thiouracil,t with hydriodic acid and red
phosphorus. The hydrogen iodide salt results from which the
pure base can be obtained in the usual manner.

N=CCl ' HN—CO
| swe
CIC - CH-+3H1-+-H,0O= HC CHIME 2He!

Pe | | |
N—CH N—CH

x

The 6-aminopyrimidin used in this work was prepared by a
shorter and more convenient method than the one employed by
Bittner mentioned above. 2, 6-Dichlorpyrimidin’ was heated

1 Ber. d. deutsch. chem. Gesellsch., XXxXvi, Pp. 2232, 1903-

2 Wheeler and Bristol: Amer. Chem. Journ., xxxiii, p. 458, 1905.

3 Gabriel: Ber. d. deutsch. chem. Gesellsch., xxxviii, p. 1690, 1905.

4 Johnson and Menge: This Journal, i, p. 114, 1906.

5 Care should be taken in working with 2, 6-dichlorpyrimidin not to
expose the material to the skin, since it has a very corrosive action. In
one case deep and painful blisters were formed on the hands, resembling
those produced by hydrofluoric acid.

288 Researches on Pyrimidins

with alcoholic ammonia. This gave a mixture of 2-amino-6-
chlorpyrimidin (XI) and 2-chlor-6-aminopyrimidin (XII).1_ It
was found, on boiling this mixture with water and zinc dust.
that 2-amino-6-chlorpyrimidin was reduced to the very soluble
2-aminopyrimidin? which could easily be removed, while 2-chlor-
6-aminopyrimidin remained unaltered. The pure 2-chlor-6-
aminopyrimidin then on warming on the steam-bath with hydri-
odie acid was smoothly reduced and the hydrogen iodide salt of
6-aminopyrimidin was obtained. Although Buttner worked
with small quantities of this base (0.25 gram) his results and
the writer’s agree in every respect.

N=CCI N=C—NH, N=C—NH,
Fara xt

H,N—C CH CIC CH Ss Ee ee
| | I ll Ih
N—CH N—CH N—CH
XI XII

EXPERIMENTAL PART.
6-Oxypyrimidin, HN—CO

HC CH

| |
N—CH

Thirteen grams of 2,6-dichlorpyrimidin were slowly added
to 50 cc. of concentrated hydriodic acid and 6 grams of red
phosphorus on the steam-bath. As soon as all was added the
mixture was boiled for a few minutes and then the hydrogen
iodide was removed, as far as possible, by evaporation in a
vacuum at 100°. The residue on taking up in hot water and
filtering from red phosphorus formed a syrup, which deposited
well crystallized needles, decomposing with effervescence when
kept at 300° (6-oxypyrimidin hydriodide). The whole was dis-
solved in water and an excess of silver sulphate was added and
filtered, the filtrate was precipitated with hydrogen sulphide,
the phosphoric and sulphuric acids removed by means of barium

1 Gabriel: Loc. cit.
2 Buattner: Loc. cit.

Henry L. Wheeler 289

hydroxide, and then, on removing the excess of barium with
carbon dioxide and evaporating to dryness, a very soluble
crystalline cake was obtained. This weighed 5.6 grams. On
crystallizing from ethyl acetate beautiful, long, thin, prismatic
needles separated melting to a clear oil at 164°-165°. The per
cent of nitrogen in this material agreed with the calculated
for 6-oxypyrimidin (Analysis I and II).

The same compound was also obtained as follows: Twenty
grams of 2-thiouracil were suspended in a liter of hot water and
about 530 cc. of commercial hydrogen dioxide solution were
added in portions. The thiouracil dissolved and after boiling a
few minutes, sulphur dioxide was added and the solution was
evaporated to a convenient volume. The sulphuric acid, which
had been formed in the reaction, was removed by means of an
excess of barium hydroxide and the excess of the latter was pre-
cipitated with carbon dioxide. The solution was then evapor-
ated to dryness and the residue was extracted with boiling alco-
hol. This dissolved the 6-oxypyrimidin, leaving a mixture weigh-
ing about 7.7 grams of uracil and a barium salt of an organic acid
that was not further examined. The barium salt was insoluble
in alcohol but readily soluble in water, it had the peculiar prop-
erty of swelling up to many times its original volume when
heated on platinum foil.

The alcoholic solution was evaporated to dryness and the
residue extracted with ethyl acetate. This gave about 6 grams
of crude 6-oxypyrimidin. When crystallized from a large
amount of benzene it formed colorless needles melting less sharply
than the above preparation (Analysis III). All the nitrogen
determinations in this paper were made by Kjeldahl’s method.

Calculated for Found:
4 40} 2: Ie ute III.
ING 29.16 29.02 28.78 28.85

6-Oxypyrimidin is insoluble in petroleum ether, very diffi-
cultly soluble in ether, somewhat more soluble in hot benzene
and still more soluble in ethyl acetate. This solvent is the best
to use to extract and crystallize the material. It is extremely
soluble in water and alcohol, and, in this respect, like 6-amino-
pyrimidin could not possibly be mistaken for the far more insolu-
ble cytosin and isocytosin.

290 Researches on Pyrimidins

This monoéxypyrimidin unlike the dioxypyrimidin (uracil)
gives no color with bromine and barium hydroxide.

It is precipitated by phosphotungstic acid, and, in neutral ©

solution, by silver nitrate or mercuric chloride. A strong solu-
tion is precipitated by picric acid and also by hydrochloro-
platinic acid. The platinum chloride double salt separates
slowly and forms prisms. It did not have a definite melting or
decomposing point.

6-Aminopyrimidin, N=C—NH,

HC CH
|
N—CH
2-Chlor-6-aminopyrimidin reduces smoothly to 6-aminopyri-
midin when warmed on the water-bath with concentrated
hydriodic acid and red phosphorous. The residue, after the
removal of hydrogen iodide by evaporation was treated with
silver sulphate, barium hydroxide, etc., as in the case of 6-oxy-
pyrimidin, for the preparation of the free base. It was found
that 2.7 grams of 2-chlor-6-aminopyrimidin, 27 cc. of concentrated
hydriodic acid and an excess of red phosphorus gave 1.6 gram
of crude base while the calculated is 1.9 gram. When crystal-
lized from ethyl acetate, snow white clusters of thin, leaf-like
crystals separated. The appearance of these clusters was similar
to those of 6-oxypyrimidin. The material melted at 151°—152°.
Bittner® gives the same melting point for this base. It is ex-
tremely soluble in water and alcohol and it is precipitated in
more dilute solutions by the same reagents which precipitate
6-oxypyrimidin.

The Acetyl Compounds.

A peculiar acetyl derivative of. 6-oxypyrimidin was formed
when the base was dissolved in acetic anhydride and evaporated
to dryness on the steam-bath. When the residue was crystal-
lized from alcohol, in which it is quite soluble, it formed colorless
needles or spikes. It is sharply distinguished from the other

Wheeler and Johnson: This Journal, iii, p. 186, 1907.
? Loc. cit.

Henry L. Wheeler 291

acetyl compounds by having two melting points. When heated
rapidly it melted to a clear oil at 180°, or a little below, then if
the temperature was kept at this point it solidified and on further
heating it remelted with effervescence at 215°-220°. (Analysis
Irand ff:)

Another sample of the acetyl compound was prepared and crys-
tallized from water. It then separated in the form of prismatic
scales which had the same behavior on heating as the above. It
was dried at 50°—55°.

(Analysis III.) The analytical results are low for a simple
acetyl derivative, but they agree with the calculated for the
expected acetyl compound witha molecule of water of crystalliza-
tion, or, if water of constitution, equally well for acetylformami-
dine acrylic acid, CH,CONH —CH =N —CH =CHCOOKH, or sim-
ply an acetic acid salt of 6-oxypyrimidin. The latter, however,
is excluded since the free base dissolves in glacial acetic acid
and on evaporation is recovered unaltered.

In order to determine whether the compound has water of
crystallization or water of constitution, a portion of the material,
which had been crystallized from water, was heated at 1og°—115°
forthree hours. It then lost 4 per centin weight This was due,
not to the fact that water was given off but that the substance
volatilized, since a nitrogen determination, after heating, gave
the same result as in the case of the previous determinations
(Analysis IV). This result makes it appear improbable that the
substance has water of crystallization. The view that the com-
pound is acetylformamidine acrylic acid, therefore, remains at
present as most probable. This, however, must be left for future
work to decide.

Caleulated for Found:
CgHsO3Ne: ie 108 Ill Vis

ING 21-32 17.94 ivi 17.64 17.88 17.86

~I

Acetyl-6-aminopyrimidin.—A quarter of a gram of the pure
base was dissolved in acetic anhydride and heated to boiling, then
evaporated to dryness on the steam-bath When crystallized
from about 5 cc. of water it formed an asbestos-like mass of fine
needles. About 0.2 gram separated. It melted at 202°, toa
clear oil, without effervescence. Analysis:

Calculated for
C.H;ON3: Found:

IN| CORB OEIC e OO me DD OOaciSd re 30.43 30.35

292 Researches on Pyrimidins

The Picrates.

The picrate of 6-oxypyrimidin is far more soluble than the.
picrates of cytosin and isocytosin. It also differs decidedly in
appearance from these salts. 6-Aminopyrimidin picrate, on the
other hand, closely resembles cytosin picrate both in regard to
solubility and crystalline form. The presence of this picrate
is possibly the cause of the picrate of cytosin from natural sources
invariably melting lower than that of synthetic cytosin."

6-Oxypyrimidin Picrate-—A saturated aqueous solution of
picric acid was mixed with a moderately strong solution of 6-
oxypyrimidin; as no precipitate was formed the solution was con-
centrated to almost the volume of picric acid solution employed.
On standing a long, flat, fern-like growth of crystals separated.
It melted to a clear oil at 190°. Analysis:

Calculated for
CioH7OsN;: Found:

IN ict sieves Bie Meteo a on eee eee 21.53 21.55

6-Aminopyrimidin Picrate-—Forty cc. of picric acid solution
were added to o.2 gram of 6-aminopyrimidin in a little water. A
bulky precipitate was formed at once which dissolved on adding
40 cc. of water and then boiling. On cooling, long bright, yel-
low, hair-like needles separated. On heating these showed evi-
dence of change a little below 200° and then suddenly melted
at 226° to a clear oil. This then turned brown and vigorously
effervesced at 270°-280°. Analysis:

Calculated for Found:

CipHg07Ne: I fe 1H
LBM | eae 25.92 25.76 25.92
The Hydrochlorides.

The hydrochlorides of 6-oxypyrimidin, 6-aminopyrimidin and
of isocytosin are even more soluble than the easily soluble cyto-
sin hydrochloride. The latter and 6-oxypyrimidin hydrochlor-
ide separate with water of crystallization. The hydrochlorides
are difficultly soluble in alcohol.

6-Oxypyrimidin H ydrochloride, C,H,ON,.HCl1.H,O.—Pure 6-oxy-
pyrmidin was dissolved in dilute hydrochloric acid and evapo-
rated to dryness on the steam-bath. The residue formed a syrup

' Amer. Chem. Journ., xxix, pp. 494, 500, 505, 1903.

Henry L. Wheeler 293

which solidified on cooling. It was dissolved in water and allowed
to crystallize by standing over sulphuric acid. Thick transparent
prisms or oblong blocks separated some ro-r5 millimeters in
length. These melted partially below 100°, finally melting to
an oil at about 205°-210°. Analysis:

Calculated for Found:
C4H;O2N,.Cl: iE Te
IN | ep aia eo eR eerie 18.60 18.54 18.64

A water determination was made by heating the material at
114° for one hour. This gave 13.1 per cent while the calculated
is 11.96. When reheated at 120° it was found that the material
slowly volatilized, which explains the high result.

6-Aminopyrimidin Hydrochloride, C,H;,N,.HCl.—The base was
evaporated to dryness with hydrochloric acid. The material
was then taken up in a little water and left to crystallize in a
desiccator. Transparent prisms or tables separated from the
syrupy solution. When dried over calcium chloride they melted
to an oil at 257° and then effervesced. A nitrogen determination
showed that the salt was anhydrous.

Calculated for
C4H 3N3.HCl: Found:

INI oe ns a) Antes oy osama clol- 31.93 31.49

Isocytosin Hydrochloride, C,H,ON,.HCl.—This salt has a de-
cided tendency to crawl up the sides of a dish when left to
crystallize. The crystals which separate in this manner are
prisms, when precipitated from an aqueous solution by the addi-
tion of alcohol it forms little square tables or blocks. When
heated it begins to change in appearance at about 250° and then
effervesces about 270°. Analysis:

Calculated for
C4H;ON3.H Cl: Found:

IN) es aR eRe ae ica 28.47 28.27

Cytosin Hydrochloride, C,H,ON,.HC1.H,O.—If cytosin is dis-
solved in strong hydrochloric acid and left to crystallize in a des-
iccator cytosin dihydrochloride is obtained, C,H;CN,2HCl1.' If
the acid solution is evaporated to dryness and the residue is
taken up in water and left to crystallize spontaneously, large,
transparent plates separate of the hydrous, 1:1 salt. This salt

1 Wheeler and Johnson: Amer. Chem. Journ., Xxxi, p. 598, 1904.

294 Researches on Pyrimidins

loses its water rapidly at 50° and in a few hours on exposure to
the air. It differs from the hydrous 6-oxypyrimidin hydro-

chloride since on standing over night the crystals become entirely ©

opaque. Itis more soluble than the dihydrochloride. The latter
and also the hydrous cytosin hydrochloride both melt at 275°-
279°. Analysis:

Calculated for
C4H;ON3.HCI1.H20: Found:

N SR Laat Seng s'p i,’ siehers gestation 25.37 25.52

Some of the material which had stood for about 3-4 hours was
dried to a constant weight at a little above 100°. It lost 9.5 per
cent of water while the calculated for one molecule of water is
10.8 per cent.

The Sulphates.

The sulphates of 6-oxypyrimidin and 6-aminopyrimidin are
very soluble in water. The neutral sulphate of isocytosin is less
soluble and it resembles the neutral sulphate of cytosin except
that it does not crystallize with water. The three sulphates
which cytosin forms have been prepared and some new facts are
given for their identification. The sulphates were prepared from
the hydrochlorides by treating the latter with an excess of silver
sulphate, removing the excess of silver by means of hydrogen
sulphide and then evaporating. Owing to their solubility, the
sulphates of 6-oxypyrimidin and 6-aminopyrimidin were crys-
tallized by means of alcohol. The sulphates of isocytosin and
cytosin were allowed to crystallize at ordinary temperatures in
order to determine their degree of hydration.

6-Oxypyrimidin Sulphate, (C,H,ON,),H,SO,—This salt was
purified by precipitating the strong aqueous solution with alcohol.
After the fourth precipitation it came down in the form of micro-
scopic prisms and melted with effervescence about 218°. Analy-
sis:

Calculated for F 1s
CgsHgO2.N4.HSO,: Ups ta Te
1 Beets 19.31 19.29 19.32

This neutral sulphate also separates from alcoholic solutions
containing a considerable excess of sulphuric acid.

Henry L. Wheeler 295

6-Aminopyrimidin Sulphate, C,H;N,.H,SO,, was described ina
previous paper.

Isocytosin Sulphate, (C,H;ON,),H,SO,.—This salt separated
from the aqueous solution containing a slight excess of sulphuric
acid, on standing, in the form of balls composed of radiating
clusters of small prisms. It melted at 276° with effervescence.
Analysis:

Calculated for Found:
CsgH i 9O02N5-H2SO.: I

2 Il.
Niort 26.25 26.30 25.98

Basic Cytosin Sulphate, (C,H;ON,),H,SO,.2H,O.—This hy-
drous salt has frequently been obtained by Levene.” It was
obtained in the anhydrous condition by Kossel and Steudel.’
It is the least soluble of the sulphates mentioned in this paper.
It has the highest decomposing point, the same as that of cyto-
sin itself, namely, 323°. It was obtained, in the present work,
when cytosin monohydrochloride was treated with a slight excess
of silver sulphate, and the silver then removed by means of hydro-
gen sulphide. On concentrating the solution the salt separated
in the form of sharply defined, long, needle-like prisms.

Neutral Cytosin Sulphate, (Cj,H,ON,),H,SO,.2H,O.—The anhy-
drous form of this salt was first obtained by Levene,* having
dried the material in a toluol bath. The hydrous form, here
described, separated on allowing the mother liquor from the
above basic salt to evaporate in the air. Stouter, more compact
masses of prisms formed which, on drying in the air, melted with
effervescence at 287°. Levene found 290°. The analyses now
show that this salt crystallizes with two molecules of water.

Calculated for Found:
CgH19O2N3.H2S04.2H20: Is we Ill.
INES as 23.59 23.66 23.45
ECO LO. 11 10.13

The water determination was made by heating the salt at 118°—
120° for an hour.

Acid Cytosin Sulphate, CSH,ON,.H,SO,—This salt has been
obtained by Kossel and Steudel; they simply mention that it is

1 Wheeler and Johnson: This Journal, 11, p. 189, 1907.

2 Zeitschr. f. physiol. Chem., xxxix, pp. 7, 135, 481, 1903.
3 [bid., XXXVili, P. 52, 1903.

4 Tbid., Xxxvili, p. 81, 1903.

296 Researches on Pyrimidins

easily soluble. It is easily obtained by dissolving the neutral
sulphate in 29 per cent sulphuric acid and allowing the material
to crystallize in a desiccator. The stout, transparent crystals
thus obtained appear to be rhombohedrous. These have the
characteristic property of becoming opaque when an attempt
is made to wash them with water. This is the most soluble
sulphate of cytosin. After pressing on paper and drying over
calcium chloride the material melted at 197° to a colorless oil.
Analysis:

Calculated for Found:
C,HsON3.H2SO,: is 10
|. eee 20.09 19.86 19.89

Acid Cytosin Phosphate, C,H,ON,.H,PO,—This salt was
obtained when cytosin monohydrochloride (7 grams) was boiled
with phosphorous oxychloride (50 cc.) for two hours. The phos-
phorous oxychloride was then evaporated and the residue treated
with ice. The material dissolved and, evaporating, a syrup was
obtained from which dilute alcohol precipitated well crystallized,
long, flat prisms. When crystallized from water, in which the
salt is very soluble, it melted at 236° with effervescence. Analy-

SIS <

Calculated for Found:
C,H;ON3.H3PO,: 1 Il.
IND Sect ats te 20.09 20.09 19.95

The melting or effervescing points of the substances com-
pared in this work are given in the following table. The effer-
vescing points even of the pure compounds may vary to a cer-
tain amount, perhaps several degrees, according to the rate of
heating, amount of substance used, etc. No especial accuracy
isclaimed. It is believed, however, that the melting points given
will be found to be sufficiently constant to serve as an aid for
the identification of the substances. There is in general a wide
difference in the melting points in each series. This should be
especially serviceable in the case of the sulphates of cytosin.

The picrolonates mentioned in the table are the most insoluble
salts prepared. Since some new facts have been observed in the
case of picrolonates these salts will be described in a later paper.

' Zettschr. j. physiol. Chem., xxxviii, p. 52, 1903.

IN

N

Henry L. Wheeler

‘S061 ‘1g ‘d ‘MAXxx “may ‘joisdyd *{ ayosaz :ousdaay ||

c0LG

of LG-o ELE

ol 9G

oso

9} BUO[OLILT

4.008
aAOqy

4086-5! LG

| #o9GG

el

| *oV61

aE

"So6r ‘egkl ‘d ‘1tAxxx “YISpasaH “mays “YISMap “p ‘dag :UeW]OD pur jalqey tt
‘oor ‘gr -d ‘it Qousnof siyy, xx
“bo6r ‘g6S -d ‘xxx “prqyt “fc6r ‘z6t -d ‘xxx “uanof ‘mayD ‘damp :ucsuyof pure ssjooy | ‘1oded styy y

of.

: [4 a
*0°Hs OS‘ H’CNO’H'O)

_ +4 206
*O°HS OS®

“L8G |
ELCNO Ho)

ee OL
* OSH “NO*H’O

QL
* OS°H’CNO‘H"O)

t oth ¢g LZ
xx OS HN HO

o8

Ic

* OS’ H’CNO'H’O)

SzBS

%OS*H

_ 06L8~oSL8 |
*O°H 10H” NO’H’O
OOLGTOSILG
LOG: NOW Oo 7

OL.
*IOH *NO'H"O

aleG
*IOH*N*H"O

_ 0016-0906
*O°H IOH“NO’H’O

1,008
VAOQYW

mele re

[Ajooy

+0888

tT1.922

*o LG

*o POL

ulsoqAo
ie
sar ifs
‘HNO N
uIsO} AOS
HO ime.
ea.
| | ‘eon
| OO= Ni
urpitutiAdoururr-y
“a 1

‘“HN—O=N
ulprwAdxo-9
HOaN
|
JH

VI.—RESEARCHES ON PYRIMIDINS: SYNTHESIS OF
THYMIN-4-CARBOXYLIC ACID,

By TREAT B. JOHNSON.
(From the Sheffield Laboratory of Yale University.)

(Received for publication, June 27, 1907.)

I shall describe, in this paper, the preparation and properties
of thymin-4-carboxylic acid,

|
NH—CCOOH

The study of the carboxylic acids of uracil, cytosin and thymin
is of interest on account of the possibility that these pyrimidins
may be linked in nucleic acids by acid amide groupings, —CO.NH.
It has already been shown! that uracil may exist in nucleic acids
asa 5-carboxyl compound since uracil-5-carboxylic acid can be
quantitatively converted into uracil if heated with 20 per cent
sulphuric acid at 160-170°. I now find that thymin-4-carbox-
ylic acid can be heated with 20 per cent sulphuric acid, under
the same conditions, without alteration. Thymin therefore can-
not exist in the nucleic acids as a 4g-carboxyl compound.

Wislicenus observed that diethyl oxalacetate and its homo-
logues show a wide difference in their behavior towards ammonia
and amines. Diethyl oxalacetate, for example, combines with
ammonia to form an unstable addition product, which is changed
to an ammonium salt of the lactone ester of oxalcitric acid (1)’
when warmed with alcohol.

Diethyl methyloxalacetate, on the other hand, does not form

1 Amer. Chem. Journ., Xxxvii, p. 392.
? Wislicenus and Beckh: Ber. d. deutsch. chem. Gesellsch., xxviii, p. 789;
Ann. d. Chem. (Liebig), cexcv, p. 339-
299

300 Researches on Pyrimidins

an addition product with ammonia but reacts at r1o°, giving
aminomethylmaleinimide (II).!

C6266 HN: C——co
So.n Ds
ae . N H, Ma 4
C,H,0,C: CH———C-CH,CO,C.H, CH,’ C60
!
CO,C,H,
I Il

Another striking example of this difference between diethyl
oxalacetate and diethyl methyloxalacetate was found when we
investigated the behavior of pseudothioureas towards these esters.
Pseudoethylthiourea combines with diethyl oxalacetate, giving
a stable addition product (Professor Wheeler).

On the other hand, diethyl methyloxalacetate reacts with
pseudomethylthiourea, in presence of potassium hydroxide, giv-
ing a good yield of the potassium salt of 2-methylmercapto-4-
carboxyl-5-methyl-6-oxypyrimidin (III). The condensation can
be represented as follows:

NH, COOC,H, NH—CO
|
| ae
CH SC choo CH, = CH,SC C-CH; + 2C,H,OH.
| | |
NH HO-C-COOC.H, N.— ©; COOr
Il

2-Methylmercapto-4-carbethoxy - 5 - methyl -6-oxypyrimidin (V)
was obtained, in one experiment, as a secondary product of the
condensation.
2-Methylmercapto-4-carboxyl-5-methyl-6-oxypyrimidin (III)
can be converted into thymin-4-carboxylic acid (VII), by digest-
ing with concentrated hydrochloric acid. When this acid was
boiled, in ethyl alcohol solution, with a small quantity of sul-
phuric acid it was converted into thymin-4-ethylcarboxylate
(VIII). This same ester was also obtained when 2-methylmer-
capto-4-carbethoxy-5-methyl-6-oxypyrimidin (V), was boiled in
alcoholic solution with sulphuric acid. When thymin-4-car-
boxylic acid (VII), is melted it undergoes complete decom-

' Wislicenus and Kiesewetter: Ber. d. deutsch. chem. Gesellsch., xxxi,
Pp. 194.

Treat B. Johnson 301

position. On the other hand, 2-methylmercapto-4-carboxyl-5-
methyl-6-oxypyrimidin (III), melts with evolution of carbon-
dioxide and is converted quantitatively into 2-methylmercapto-
5-methyl-6-oxypyrimidin (IV). This mercaptopyrimidin can
then be changed to thymin (VI) by boiling with hydrochloric
acid.?

The formation of thymin in this manner from 2-methylmer-
capto-4-carboxyl-5-methyl-6-oxypyrimidin (III), shows that the
condensation takes place as represented in the preceding equa-
tion and that the product is not a hydantoin derivative. The
above compounds and their various transformations are repre-
sented as follows:

NH—CO NH-—CO NH—CO

|
PP OemOCtrn = CHSC) CCH) > CHSC Cre
Neyer all lees i | ll

Ne (On N — CCOOH N= CCo. ens
| Iv ioe | v
i J
NH—CO NH—CO NH—CO
| |
CO. * CCH: COx CCH. >. “=, COm eatEs
| | | | | |
NH—CH NH—CCOOH NH—CCO,C,H,
VI VII VIII

Thymin-4-carboxylic acid is characterized by its insolubility
in cold water and by its property of crystallizing from aqueous
solution with and without water of crystallization. It gives very
insoluble barium and lead salts, and is not precipitated by phos-
photungstic acid.

Thymin-4-carboxylic acid reacts in the normal manner with
bromine water giving oxybromhydrothymin-4-carboxylic acid
(IX).

NH—COo NH—CO
: | cH
€or §C-CH, +. Br + HO = CO C< 2) '- MB
: | 4
NH—CCOOH NH—C:OH
COOH
IX

1 Amer. Chem. Journ., xxix, p. 487.

302 Researches on Pyrimidins

EXPERIMENTAL.

2-Methylmerca pto-4-carboxyl-5-methyl-6-oxypyrimidin,
NH—CO

CH,5*C GoGH-
| |
N — C:'COOH
Thirty grams of pseudomethylthiourea hydriodide and 50 grams

of the sodium salt of diethyl oxalpropionate were dissolved in
about 500 cc. of water and two molecular proportions of potassium
hydroxide (16 grams) added to the solution. The mixture was
allowed to stand on the steam-oven for 8—-1o hours and then
concentrated to a volume of 150 cc. After thorough cooling,
and acidifying with hydrochloric acid, the mercapto-pyrimidin
deposited in prismatic crystals. It was practically insoluble in
cold water and difficultly soluble in boiling water and alcohol.
It separated from water or alcohol in rectangular prisms that
melted at 243—244° with effervescence toa clearoil. It deposited
from glacial acetic acid in stout prismatic crystals. The yield
of acid corresponded to about 80 per cent of the theoretical.
Analysis (Kjeldahl):

Calculated for
C7H,O03N28: Found:

aia he Sateen meet ee 14.00 per cent. 13.88 per cent.

2-Methylmercapto-4-carbethoxy-5-methyl-6-oxypyrimidin,
NH—CO
|
CH,SC CCH,
| |
N — CCOOGH,

This ester was obtained, associated with 2-methylmercapto-
4-carboxyl-5-methyl-6-oxypyrimidin, when I used one instead
of two molecular proportions of potassium hydroxide in the
above condensation. It was difficultly soluble in cold water and
alcohol, but deposited from hot alcohol or water in slender needles
that melted at 201—202° toa clear oil without effervescence. An-
alysis (Kjeldahl):

Calculated for
CoH} ,05N8: Found:
IMA dWie hind Sabi g alent 12.30 per cent. 12.37 per cent.

Treat B. Johnson 303

Potassium Salt of 2-Methylmercapto-4-carboxyl-5-methyl-6-oxy-
pyrimidin,

NH—CO

CH,SC C-CH,-6H,0

Ee ll
N — CCOOK

A good yield of this salt was obtained under the following con-
ditions: Pseudomethylthiourea and diethyl oxalpropionate were
condensed as in the above experiment, in presence of two mole-
cular proportions of potassium hydroxide. After standing for
about twelve hours at 40—-45° the solution was then acidified with
acetic acid and concentrated to a volume of about 150 cc. On
cooling the potassium salt deposited in distorted needles that
decomposed with effervescence when heated above 230°. When
this acetic acid filtrate was treated with hydrochloric acid the
mercapto-acid separated melting at 243°. The potassium salt
was very soluble in hot water and deposited, on cooling, in
needles. They contained water of crystallization which was
determined by heating at 1oo-110° for two hours.

1.2424 gram substance lost 0.4032 gram water.

Caleulated for .
C;H;O;N.SK.6 H,O: Found:

| S(O GRR a eat MRR ROE neigh: 31.21 per cent. 32.4 per cent.

Nitrogen determination in the anhydrous salt (Kjeldahl):

Caleulated for FE :
C;H;03,N.SK: ound :

IN es erating Calne ct 11.76 per cent. 11.91 per cent.

Behavior of 2-Methylmercapto-4-carboxyl-5-methyl-6-oxypyrimi-
din on Heating—About two grams of the mercapto-acid were
heated in a sulphuric acid bath at 245° until all effervescence
ceased. I obtained a clear oil that crystallized, on cooling, in
large, prismatic crystals. When these prisms were heated they
melted sharply at 230—231° without effervescence to a clear oil.
The compound deposited from water in flat prisms that melted
at 233° and was identified as 2-methylmercapto-5-methyl-6-oxy-

304 Researches on Pyrimidins

pyrimidin.!| When mixed with this pyrimidin the melting point
was not lowered. Analysis (Kjeldahl):

Calculated for :
CyHsON2S: Found:
Nears sais a nas 17.94 per cent. 17.80 per cent.
Thymin-4-carboxylic Acid,
NH—CO

CO CCH,°H,O
|

NH—CCOOH

A quantitative yield of this acid was obtained when 2-methyl-
mercapto-4-carboxyl-5-methyl-6-oxypyrimidin was digested with
concentrated hydrochloric acid. The oxygen acid separated
from the acid solution as a granular powder that was difficultly
soluble in boiling water and practically insoluble in alcohol. A
most characteristic behavior of this acid is its property of crystal-
lizing from hot water in anhydrous condition and with one mole-
cule of water of crystallization. When a hot, saturated aqueous
solution of the acid was allowed to cool slowly the anhydrous
acid first deposited in balls of microscopic prisms resembling
very much the crystalline form of uracil. They decomposed
at 328-330° (Anschutz thermometer) and did not lose weight
when heated at 120°. Analysis (Kjeldahl):

Calculated for
36H pO4No:

PEE ee re a are 16.47 per cent. 16.37 per cent.

Found:

After filtering from the anhydrous acid and allowing the fil-
trate to stand, transparent, rectangular prisms of the hydrous
acid deposited. They decomposed at the same temperature
as the anhydrous acid (328-330°). The acid was not precipitated
by phosphotungstic acid.

Water determination: 0.6742 gram substance lost 0.0687 gram water
after heating one hour at 110-120°.

Calculated for
CsHoO4No. H20: Found:
HO a ie ieteasa eae ee 9.60 per cent. 10.1 per cent.

‘Amer. Chem. Journ., xix, p. 478.

Treat B. Johnson 305

Analysis of the anhydrous acid: 0.2592 gram substance gave 0.4002
gram CO, and 0.0854 gram H,O. Nitrogen determination (Kjeldahl):

Caleulated for

CeH.O4No: Found:
Gas Deter PRN ANS ip 8 Ol a 42.35 per cent. 42.11 per cent.
1G Lat OO OL ee ae pronve. SuGG. Sao
jt td a ae TORag en <= 16.33; “ee

Action of 20 per cent Sulphuric Acid.—One-half a gram of the
thymin acid was heated with 5 cc. of 20 per cent sulphuric acid
for two hours at 160-170°. When the tube was opened there
was no pressure and the unaltered acid was suspended in the
sulphuric acid. I recovered 0.45 gram of acid melting at 326—
329°. Analysis (Kjeldahl):

Calculated for
CeH,O4No: Found:

IN SPER: MERE Eos 16.47 per cent. 16.15 per cent.

Potassium Salt, C,H,0,N,K.2H,O0.—This salt was prepared
by dissolving molecular proportions of potassium hydroxide and
thymin-4-carboxylic acid in water. When the solution was
concentrated and allowed to stand for a few hours the salt
deposited in radiating prisms. It was dried for analysis in a
desiccator over sulphuric acid (Kjeldahl):

Calculated for Calculated for Found
Cg5H,O,NoK: CsH;O4NoK. 2H,0: ound.

IN eyecce 13.45 percent. 11.06 per cent. 10.97 per cent.

Lead Salt (C,H;0,N,), Pb.—This salt deposited in well devel-
oped prisms when a solution of lead acetate was added to a hot,
saturated, aqueous solution of thymin-4-carboxylicacid. It was
practically insoluble in cold water. Analysis (Kjeldahl) :

Caleulated for F 1:
(Cg6H;04N2)2Pb: ae

DS eis uae Bape 10.27 per cent. 10.00 per cent.

Barium Salt (C,H;0,N,),Ba.—Thymin-4-carboxylic acid gives
no precipitate when treated with a solution of barium chloride.
The barium salt was prepared by dissolving the acid in potas-
sium hydroxide and then adding the calculated amount of barium
chloride. It separated from water in corpuscular crystals.
Analysis (Kjeldahl):

Calculated for F “i:
(CgsH;04Ne2)2Ba: ound:

in Ce ee 11.77 per cent. 11.30 per cent.

/

306 Researches on Pyrimidins

Thymin-g-ethylcarboxylate,
NH—CO

CO CCH,
| |
NH—CCOOC,H,

was prepared by esterifying the acid with ethyl alcohol and
sulphuric acid. It was also obtained when 2-methylmercapto-
4-carbethoxy-5-methyl-6-oxypyrimidin was boiled in alcoholic
solution with a small amount of hydrochloric acid. The ester
deposited from hot water in distorted prisms that melted at

255° to a clear oil without effervescence. Analysis (Kjeldahl):
ca NS Found:
LS sony Bap CES RPE 14.14 per cent. 14.07 per cent.

Oxybromhydrothymin-4-carboxylic Acid,

NH—CO

| aie
~ CH;

CO "Cees

NH—C-OH
COOH

Three and five-tenths grams of finely pulverized thymin-4-
carboxylic acid were suspended in about 40 cc. of bromine water
and bromine slowly added until the acid had completely
dissolved. There was no evolution of carbon dioxide and when
the solution was allowed to evaporate spontaneously in the
atmosphere well-developed, prismatic crystals of the hydro-
derivative separated. It was purified for analysis by recrys-
tallization from bromine water. It crystallized in small, pris-
matic crystals that charred when heated above 270° and then
decomposed with violent effervescence from 295—-300° accord-
ing to the rate of heating. When thymin-4-carboxylic acid was
heated with bromine water at 146—152° it was completely decom-
posed with formation of bromoform. Analysis (Kjeldahl) :

Caleulated for Calculated for Found:
CyH,O;NoBr: C5H-O2,NoBr:

N..... 10.48 percent. 12.66 percent. 10.45 10.45 per cent.

THE BALANCE OF ACID-FORMING AND BASE-FORMING
ELEMENTS IN FOODS.

(Preliminary Paper.)
By H. C. SHERMAN anp J. EDWIN SINCLAIR.
(Contribution from the Havemeyer Laboratories, Columbia University, No. 138)
(Received for publication, June 19, 1907.)

Although mention is frequently made of the fact that some
foods contain an excess of acid-forming and others of base-form-
ing elements, no systematic quantitative study of the subject
appears to have been published. It is true that the determina-
tion of the alkalinity of the ash is a method much used by food
chemists in the examination of fruit products; but the alkalinity
of the ash may easily be greater than would correspond to the
excess of base-forming elements in the food itself, for even those
food materials which yield an alkaline ash are apt to lose sul-
phur and chlorine in burning.

As early as 1885, Bunge? in reporting an analysis of beef flesh,
made a partial computation which showed that the oxidation of
the sulphur in the flesh would alone yield sufficient acid to com-
bine with all of the bases present, and various writers have sug-
gested that the injurious effects of ‘‘ash-free’”’ food might be
partly due to the lack of a sufficient supply of bases to neutral-
ize the sulphuric acid produced in the protein metabolism, but
so far as we are aware there has been no attempt at a systematic
quantitative balancing of the acid-forming against the base-
forming elements in foods, although similar work has been
done with urine by several investigators. We have undertaken
a study of this subject which, on account of the large amount
of analytical work involved, will probably require a consider-
able time for completion. The present widespread interest in
all subjects connected with acidosis suggests, however, a brief
preliminary publication of some of the results already obtained.

1 Zeitschr. f. physiol. Chem., ix, p. 60.
3°7

308 <Acid- and Base-forming Elements in Foods

While postponing detailed descriptions either of the materials
examined or of the analytical methods employed, it may be
stated that calcium, magnesium, sodium, and potassium were
determined by the usual methods, and that chlorine, phosphorus
and sulphur were determined in separate portions of the dried
food material which were oxidized for the determination of
chlorine by burning at a low temperature in the presence of a
large excess of sodium carbonate; of phosphorus, by means of
nitric and sulphuric acids and ammonium nitrate; of sulphur, by
Liebig’s method, checked by several determinations in which the
oxidation was accomplished by burning in compressed oxygen.!

Of the different methods available for computing and express-
ing the results, the one which appears most satisfactory is to
calculate the amount of each element found to the corresponding
number of cubic centimeters of a normal solution of acid or base.
By then adding together the results obtained for all of the base-
forming and for all of the acid-forming elements, respectively, it
is easy to compare the totals and the result obtained shows the
excess of acid-forming or of base-forming elements in terms of a
familiar standard and in figures of convenient magnitude.

For the purpose of this calculation, it is assumed that phos-
phoric acid is “‘neutralized’’ when two hydrogen atoms are
replaced by bases. By this method of calculation the following
results were obtained:

EXCESS OF ACID-FORMING OR OF BASE-FORMING ELEMENTS PER 100 GRAMS.

Excess of acid-forming or of
base-forming elements in

Food material in the fresh or dried terms of a normal solution of
state as ordinarily purchased. Acid: Base:
ec. ce.

Oarmreal (airedig) males Mat) oti caine eee os ee stscet 12.93
Beef (fresh), free from visible fat ................ T2160
Wikest, entire. orain (air-dry) en.) oe tev onic 9.66
Maik; a composite sample. 22.4207)... 2... can 237
Gan ACnied ny merermre rs oad cmibeet. Lisa km ik Oe ma OF
Pranes (arr-cmeGiniine oc aes ss ee Se sey 24.4

The above figures may of course be as readily expressed on the
basis of pounds or ounces and would possibly be most conven-

* Berthelot: Compt. rend. de lV’ Acad. des sci., cxiv, p. 318; CXX1X, Pp. 1002;
Sherman: Journ. of the Amer. Chem Soc., xxiv, p. 1100, and Methods of
Organic Analysis, p. 19.

H. C. Sherman and J. Edwin Sinclair 309

ient in that form for dietary computations based on weights of
food material purchased or prescribed. It is probable, however,
that a better impression of: the relative acid-forming or base-
forming tendencies of foods will be given if the final figures are
expressed in terms of roo calorie portions rather than roo gram
portions. The above figures then yield the following:

EXCESS OF ACID-FORMING OR OF BASE-FORMING ELEMENTS PER IOOCALORIES,

Normal Normal
Solution. Solution.

BEI Tee) LOmMNVvVASI DLE haitegs:vs «cou cue tieidion) os) sels) eaten cate LOe

Warbmiteal acc s erntas, yo b aeepeuar oe cdaro © Hane ote gpa nO era Bes
Wihteatvermtire owraiils . 5 oc gems ts 5 os sas «hem Bers ervenebe tee 2.62

GAS Rt ieee RAI ee 50008 Aesth acai oleae ae ee 1.94
AVI cepey ce nae cuc ta ake Sts 5 ds oS ciatsoahe soz suse =y Joe one RNs eee Regi
BAAR OS a foo Sc, ira 5 Gite) eae ayn) 545 4a Sie, $a NSE 7.92

It is noticeable that peas, in spite of their high protein content,
contain an excess of base-forming or acid-forming elements.
Doubtless in most other vegetables this ‘“‘ potential alkalinity”
will be found to be much greater. When a larger number of
food materials have been examined it will be possible to compute
the extent to which the excess of acid-forming elements in meat
and cereal products is offset by that of bases in other food mate-
rials of ordinary mixed diets.

Remembering that the 100-calorie portion of any staple food
material may be multiplied many times in a day’s dietary it is
obvious that, by the free use of meats and breadstuffs on the one
hand or of fruits, vegetables and milk on the other, the net excess
of acid or base introduced into the body through the food may
be varied at will within wide limits. It is not our purpose, how-
ever, to discuss the possible physiological applications until the
data for a larger number of food materials have been obtained.

er ® :
a a
ree. 5
eee 4
7 Sa
mA :
7 7
2 2

e)

ON THE EXCRETION OF CREATININ IN THE NEW-BORN
INFANT.

(A Preliminary Report.)

By S. AMBERG anp W. P. MORRILL.
(From the Pharmacological Laboratory of the Johns Hopkins University.)

(Received for publication, June 21, 1907.)

Satisfactory data concerning the metabolism of the new-born
are relatively few. Inasmuch as it is very important to know
the peculiarities of the metabolism in the new-born infant and
in young animals, both from a clinical and from a physiological
point of view, we have undertaken investigations in this field
with the newer methods and in the light of the newer interpre-
tations of metabolism experiments. The results thus far ob-
tained are presented in their present incomplete state because
it was necessary to discontinue the work for some time. How-
ever, the results thus far obtained are of sufficient interest to
warrant publication at this time.

The investigations of Rietschel! concerning the creatinin meta-
bolism of infants led him to the conclusion that the urine of
normal infants does not contain any creatinin. In febrile diseases
a small amount of creatinin seemed to be excreted. In two
cases in which creatinin-zine chloride was given per os the urine
gave a strong reaction for creatinin. Reitschel used the crea-
tinin-zinc chloride method to demonstrate the presence or absence
of creatinin in the urine. The sediment was examined micro-
scopically and then dissolved in water in order to test for creatinin
by the Weyl reaction. No attempt was made to determine
quantitatively the creatinin in twenty-four hour specimens of
urine. Reitschel mentions that Grocco, an earlier investigator,
occasionally found creatinin in very small amounts in the urine

1 Zur Kentniss des Stoffwechsels beim Satiglinge, Jahrbuch fur Kinder-
heilkunde, 1xi, p. 615, 1905.

3It

312 Excretion of Creatinin in Infants

of infants. Von Hoogenhuyze and Verploegh' constantly found
creatinin in the infants’ urine and more distinctly with the
reaction of Jaffé than with that of Weyl. In only four cases
were exact colorimetric determinations by Folin’s method possi-
ble on account of the low concentration and small amount of
urine obtainable. In these four cases the creatinin content per
10.0 ce. of urine was (1) 1.11 mg., (2) 0.91 mg., (3) 0.41 mg., (4)
1.7 mg. The lowest value (3) was obtained from the urine of
a weak, artificially fed baby, while the other babies were vigorous
and breast fed.

Closson? reports the quantitative determination of creatinin
in the case of two boys. The first boy, 6 years and 9 months old
and weighing 22 kilos, excreted 0.42 gram and 0.32 gram of
creatinin in 24 hours on a meat-free diet.* The second boy, a
case of diabetes insipidus, 14 years of age and weighing 27.5
kilos eliminated 0.44 gram of creatinin in 24 hours. Further-
more this author demonstrated the presence of creatinin in the
urine of young puppies and of kittens.

Our first experiments consisted of qualitative tests for crea-
tinin in specimens of urine from infants; later the creatinin was
determined quantitatively according to Folin’s method, and in
five cases the total creatinin excretion in 24 hours was determined.

In several afebrile cases the Weyl reaction with the fresh urine
was negative and the Jaffé reaction doubtful. But while these
reactions may be doubtful or negative-with 5.0 cc. to 10.0 cc.
of fresh urine, the following instance demonstrates that such
results may be easily accounted for by the low concentration of
the urine. Thus, urine of a strong breast-fed infant did not give
a definite reaction when tested directly, but after evaporation
of adarger amount of urine and dilution up to 25 per cent of the
original volume, both reactions were strongly positive. The
dilution in this case was such as to bring the concentration up to
about that of normal adult urine. - By evaporating the urine or

* Von Hoogenhuyze and Verploegh: Beobachtungen iiber die Kreatinin
Ausscheidung beim Menschen, Zeitschr. 7. physiol. Chem., xlvi, p. 415,
1905.

*Closson: The Elimination of Creatinin, Amer. Journ. of Physiol.,
XVi, p. 252, 1906.

* The creatinin was determined by the Neubauer-Salkowski method.

S. Amberg and W. P. Morrill 214

by using larger amounts of fresh urine, we never failed to get
a positive reaction for creatinin.*

The quantitative estimations of creatinin were made according
to Folin’s method’ using the Duboscq colorimeter for the color-
imetric comparison. Since the total amount of creatinin in the
infants’ urine is small, a larger amount of urine must be used in
order to carry out the determination than is required in the case
of an adult. Preliminary experiments showed that it is not
feasible to work directly with large amounts of urine but that
the infants’ urine must be concentrated. For instance 5.0 cc.
of the urine of an adult gave 22.5 milligrams of creatinin per 10.0
ec. Then 5.0 cc. of this urine were diluted to 100.0 cc. with
water and now our first reading showed 20.4 milligrams per 10.0
ec. After standing for to-12 minutes the readings changed
suddenly and corresponded to 15.6 milligrams of creatinin per
1o.o cc. Ina second test the same urine diluted as before gave
immediately only 13.2 milligrams per 10.0cc. Ina second series
10.0 cc. urine gave 12.3 milligrams creatinin. Then 10.0 cc. of
the same urine were diluted to 100.0 cc. with water and after the
addition of the picric acid and sodium hydroxide the mixture
was allowed tostand 6 minutes before diluting to 500cc. Asecond
test was allowed to stand 12 minutes and a third 18 minutes.
The first test gave 6 milligrams of creatinin, the second 9.3 milli-
grams, and the third 10.2 milligrams. On account of these
results this mode of procedure was abandoned. Then 5.0 cc.
of the urine used in the last series were diluted with 95.0 cc.
of water, evaporated to dryness on the water bath, the residue
taken up in water, filtered and washed with water until the
filtrate had a volume of 10.0-12.0 cc. This filtrate was used for
a determination, giving a result of 12.4 milligrams, per 10.0 cc.
of urine as against 12.3 milligrams as determined directly. In
another experiment in which 1o.0 cc. of urine were diluted with
go.o cc. of water and treated in the same manner, the determina-
tion gave 12.1 milligrams of creatinin, showing that this method

1 Tt is interesting to note that 100.0 cc. of amniotic fluid after removal
of the proteid and evaporation to 5.0 cc. gave positive Weyl and Jaffé
reactions for creatinin.

2 Folin: Beitrag zur Chemie des Kreatinins und Kreatins im Harn,
Zeitschr. f. physiol. Chem., xli, p. 223, 1904.

314 Excretion of Creatinin in Infants

is accurate for urine containing small amounts of creatinin. In
all the quantitative determinations therefore the urine of the
infants was evaporated to dryness on the water bath, taken up
in water and filtered in the manner described, the resulting 10.0
ce. of filtrate being used for the colorimetric determination. A
further advantage of this method is that it excludes any possi-
bility of error that might arise from the presence of acetone or
hydrogen sulphide in the urine. None of the urine used for
these experiments contained albumin or reduced Fehling’s solu-
tion.

The results of the experiments, obtained in three healthy,
breast-fed infants 10-14 days old, were as follows:

Ce. of Urine Creatinin in Milligrams

Used. per 10 cc. of Urine.
Bee ee ic ap. ane pv hin oo ew pea ease e 100.0 Loves
Une eR a TWN gn pra plete le 2 aor. oie Tila) [e350
Se et ee eee PL NG Ne soc pi's, ds .aree) ot se 50.0 eos

Of some interest are the data obtained in several consecutive
experiments on an emaciated female child, 23 years old, suffering
from a chronic enteritis probably of tubercular origin. The
child died 10 days after the last determination. The urine was
obtained by catheterisation. In the first examination the speci-
fic gravity was 1005. Seventy-two cc. of urine gave a definite
Jaffié reaction but the color was not of sufficient intensity to
permit a quantitative determination of the creatinin as the urine
was used without previous concentration.

In the following determinations the urine was evaporated and
treated as above.

Temp. | __ Wetcnr. aa ee Specific | Creatinin per a

F. SSS aa Gravity. 2 * lec. Urine
Ib. oz. ] mg. mg.

99 16 10 30.0 3.9

98.2 15 fs 80.0 1092 0.65 0.11

99.8 14 8 17.95 - 1010 9.07

99.8 14 13 62.0 1004 | 1.87

102.4 14 9 59.0 | 1004 2.42

These results may to a certain extent bear out the observation
of Rietschel' that fever may influence the creatinin excretion

MOG. AGA

_———

S. Amberg and W. P. Morrill at5

and of Leathes,t who found an increased output of creatinin in
febrile diseases. In our case the creatinin excretion seemed to
vary more with the changes in the concentration of the urine
than with the changes in the temperature, but since we did not
have the 24 hour urine our observations do not permit us to draw
any definite conclusions thereto. Similarly the results of Von
Hoogenhuyze and Verploegh’ are of only limited value since they
also did not have the 24 hour urine and, while further investiga-
tions may confirm their observation that the weak, artificially
fed infant excreted less creatinin than the other vigorous breast-
fed babies, their results do not justify such a conclusion.

In order to determine the normal daily excretion of creatinin
the urine of five male infants was collected for 24 hours. All
these infants were inmates of the maternity ward of the Johns
Hopkins Hospital, and we desire here to acknowledge the kind-
ness of Professor Williams and of Dr. Goldsborough who placed
the necessary material at our disposal. The urine was collected
by a modification of the method described by Freund.*

Small round flasks of about 180 cc. capacity with a curved neck were
attached to the pubic region by means of a rubber collar. The flask was
enclosed in a muslin sack and these sacks were again tied around the
abdomen, thus affording greater security. This additional security was
very desirable since the babies had to be left entirely to the mothers’ care
during nursings and so were not always carefully handled. The flasks
were changed before each nursing in order to avoid any loss which might
be incurred in handling the baby. When in bed the legs of the child
should be crossed over the flask rather than under it as will happen if they
are not watched. This insured the best position of the flask and in this
position the babies remained perfectly quiet not requiring to be tied as
required by the method of Freund. By this method it was possible to
collect 24 hour specimens without disturbing the routine of the nursings,
etc., and in some cases in fact the mothers did not know anything unusual
was going on. During the time the collection was made, the total amount
of food was determined by weighing the child before and after each nurs-

ing.

In the following table the age and weight of the infant, the
food taken and the urine voided in 24 hours are given, the last

1 Leathes: On the Excretion of Nitrogen, Creatinin and Uric Acid in
Fever, Journ. of Physiol., Xxxv, p. 205, 1907.

Feo. C1.

3Chlor und Stickstoff im Satiglingsorganismus, etc., Jahrbuch fir
Kinderheilkunde, xviii, p. 137, 1898.

~)

316 Excretion of Creatinin in Infants

also being given in per cent of total ingestion. The weight of

the infant was taken at the end of the 24 hours.

give an idea of the general condition of the infants.

In a second

table the daily weight from date of birth is recorded in order to

rei Amt. of Food. | Amt. of Urine. |Amt,tohiieinen

No. | ks. i mas Amt, Of FOr ae eee. Pas Cae mea
1 10 4.033 793.0 320.0 40.4
2 10 3.463 | 620.0 350.0 56.5
3 14 2.723 | 252.0 168.0 66.7 |
{ 7 3.093 495.0 200.0 40.4
5 13 33 772.0 426.0 55.2

No Birth. ld 2d 3d , 4:08 5d. | 6 d. Urals

1 | 3.715 | 3.65 | 3.57| 3.625 | 3.69| 3.72 | 3.812] 3.92

2 3.19 | 3.095| 2.985) 2.95 | 3.085) 3.15 | 3.225] 3.275

3 2.45 | 2.35 | 2.32 | 2.29 | 2.265) 2.27 | 2.345] 2.389

4 3.12 | 2.97 | 2.89 | 2.95 | 2.985) 3.02 | 3.065) 3.093

5 o4 3.0 | 2.842] 2.86 | 2.885} 2.87 | 2.9 2.89

No. 8 d. 9d. 10 d. Gale ras Side 14d.

1 3.92 3.956 | 4.033

2 ase 3.423 3.463

3 2.422 | 2.495 | 2.535] 2.58 2°635)| 2 277398

4

5 2.947 | 2.96 2.995 | 2.98 3.04 3.1

In the tollowing table the total excretion of nitrogen and of
creatinin during 24 hours is given together with the amount of
creatinin per kilo of body weight and the per cent of the total

nitrogen appearing as creatinin. -

CREATININ. CREATIN.
N os a < 7 a |p ae Total N.
NO. Totai in Per K. - gms.
24 hr. Body Wt. Total. | ,PerK. |
mg. - mg. | Body Wt.
1 26.82 6.71 0.389
2 26.92 1342 | 03267
3 27.05 9.94 | O.2785
4 17.48 5.46 0.218
5 30.1 9.7 2.87 0.93 0.411

Creatinin N
Per Cent of

Total N.

S. Amberg and W. P. Morrill ee,

Besides the determination of the total nitrogen and the crea-
tinin it was our intention to follow the nitrogen distribution, tak-
ing into consideration the total nitrogen, ammonia, urea, uric
acid, creatinin, and creatin, but since the amounts of urine re-
quired for the quantitative determination of these substances
far exceeded our supply, we will have to defer the study of the
nitrogen distribution to a later publication. It may be men-
tioned that, with the exception of No. 5, we used 100.0cc. of urine
for the determination of creatinin, 20.0 cc. for ammonia deter-
mination (according to Folin‘), 100.0 cc. evaporated to half the
original volume for the urea determination (according to Schon-
dorf’), 20.0 cc. for the total nitrogen, and 150.0 cc. for the uric
acid (according to Folin). In the case of No. 5 in which the
creatin was determined, only 70.0 cc. were used for the creatinin
determination and the uric acid determination was omitted.
The results obtained in this case are as follows:

No. Total N. Urea N. Ammonia N. Creatinin N.
5 0.411 gm. 91.32 percent. 4.15 percent. 2.72 percent.

In order to secure enough urine to make all of these determin-
ations, we tried to collect the urine of two consecutive days, but
on the first day we were unable to watch the collection closely
enough and the amounts obtained certainly did not represent
the 24 hour urine. Since we had to mix the urine of both days
in order to carry out the urea and uric acid determinations we
will refrain from giving these results until we have further data.
The total nitrogen and the creatinin, however, were in every case
determined on the urine of the same day. In later experiments
weintend to take the determination of the nitrogen in the food
into consideration.

To return to our results obtained thus far, No. 4 showed a
very low excretion of creatinin and the excretion of total nitrogen
too is rather low. Therefore it is highly probable that in this
case some urine was lost and we can not consider it as giving
normal results. In the other four cases the amount of creatinin

1 Folin: Approximately Complete Analyses of Thirty ‘‘Normal’’ Urines,
Amer. Jour. of Physiol., xiii, p. 45, 1905.

2See Emerson: Clinical Diagnosis, p. 107, Lippincott, Phila., 1906,
and Schéndorff: Ezne Methode der Harnstofjbestimmung, etc.

318 Excretion of Creatinin in Infants

excreted in 24 hours showed remarkably little variation for the
different infants (26.82—30.1 milligrams) although there was
considerable variation in the total quantity of urine and there-
fore in the creatinin concentration. Furthermore, considering
the excretion of creatinin per kilo of body weight, we see that
No. 1 presents the lowest value, while No. 3 presents the highest.
No. 1 was a very fat baby while No. 3 was not very well nourished,
thus showing a close agreement to the statement of Folin’ that
fat or corpulent persons yield less creatinin per unit body weight
than lean ones.

There is a very considerable difference between the amount
of creatinin excreted by the infant and that excreted by the
adult. The values for the latter vary according to Folin between
20 and 24 milligrams per kilo of body weight depending on
the corpulency of the individual. Shaffer? more recently found
variations from 18.0 to 30.0 milligrams per kilo for the normal
adult. In our cases the elimination of creatinin per kilo varied
between 6.71 and 9.94 milligrams and since moderate muscular
exercise does not influence the creatinin elimination (see Folin,’
Von Hoogenhuzye and Verploeght and Otto of Klercker*) the
lack of it can not be responsible for these low values. Benedict
and Myers" studied the creatinin excretion in women in order
to throw additional light on the three factors which have been
claimed to influence the creatinin elimination, namely, the degree
of muscular development, the muscle tonus (Shaffer), and the
proportion of adipose tissue. Of interest are their observations
that in general the creatinin excretion of women is much lower
than that of men, and in two cases of advanced age the creatinin
coefficient, that is, the creatinin elimination in milligrams per
kilo body weight, fell to 7.4. In general though they found the
creatinin excretion proportional to the body weight and not to

' Folin: Laws Governing the Chemical Composition of the Urine, Amer.
Journ. of Physiol., xiii, p. 85, 1905.
*? Cited by Benedict and Myers.

‘Amer. Journ. of Physiol., xiii, p. 85, 1905.

RoC. ct.

* Otto of Klercker: Kreatin and Kreatinin im Stoffwechsel, Biochem.
Zeitschr., iii, p. 33, 1907.

* Benedict and Myers: The Elimination of Creatinin in Women, Amer.
Journ. of Phystol., xviii, p. 377, 1907.

S. Amberg and W. P. Morrill 319

the active mass of protoplasmic tissue, thus denying the influence
of inert adipose tissue. Furthermore, they were unable to show
a constant proportion between muscular development or muscle
tonus and the creatinin elimination. We cannot enter here into
a discussion of these conclusions, but it appears that they will
have to be supported by further investigations.

Our results appear to speak in favor of the theory advanced by
Folin and also for that of Shaffer, at least in so far as muscular
development is concerned. The body of the new-born child’
is relatively richer in water and fat, while poorer in nitrogenous
substance and ash than the body of the adult, and the muscles
of the new-born child constitute only 23 per cent of its total
body weight, while those of an adult constitute 43 per cent of
the total weight.

Another point must be mentioned. The diet of our babies
was one which, in the light of Folin’s studies, is rich in proteids.
This we may deduce from the relative amount of creatinin
nitrogen which represents an average of 2.99 per cent of the
total nitrogen. The proteid-rich diet yielded, as cited by Folin
and taken from one of his tables, 3.6 per cent creatinin nitrogen,
while the diet poor in proteid yielded 17.2 per cent. The rela-
tively high percentage of urea nitrogen in our cases would point
in the same direction. In accordance with Folin’s views then
the end products of the metabolism of the normal breast-fed
infant would demonstrate a pronounced exogenous nitrogen
metabolism; in other words, we would have an indication of a
‘Luxus’ consumption of proteid. Whether such a diet rich in
proteid is warranted by the demands of growth is a question
which can not be discussed on the basis of our experiments.
From the exposition of Camerer, however, it is doubtful that
such is the case.

In this discussion we have followed the theory of Folin’ that
in the creatinin excretion we have one of the most important
indicators of the endogenous nitrogen metabolism, or tissue meta-
bolism while only a small portion of the urea could be derived

1Camerer in Pfaundler und Schlossman: Handbuch der Kinderheil-
kunde. Leipzig,1906. Stoffwechsel und Ernahrung im Ersten Lebenjahr.

2 Folin: A Theory of Proteid Metabolism, Amer. Journ. of Phystol.,
Mispe-Li7, F905.

320 Excretion of Creatinin in Infants

from that source, though we are well aware that this conception
is not generally accepted. Thus Leathes' concludes that the
urea constitutes the principal nitrogenous product of tissue
metabolism. Therefore we will have to be extremely cautious in
concluding from our experiments that we are actually confronted
by a ‘Luxus’? consumption of proteid.

SUMMARY.

1. Creatinin is constantly present in the urine of normal

breast-fed infants.

2. The creatinin coefficient while only approximately one-
third as great as in adults is fully as constant.

3. Taking into consideration the relative percentages of
muscle substance in the adult and in the infant, we are inclined
to favor the theory that the creatinin excretion is an important

index of endogenous proteid metabolism.

1Loc. cit.

THE EFFECT OF TRANSFUSION OF BLOOD ON THE
NITROGENOUS METABOLISM OF DOGS.

By HOWARD D. HASKINS:.!
(From the Physiological Laboratory, Western Reserve University.)
(Received for publication, June 5, 1907.)

The subjects of the experiments reported in this paper were
healthy small dogs.

The operations for transfusion were performed by Dr. George
W. Crile and his assistants, Drs. Lenhart and Hitchings. The
surgical aspects of transfusion will be fully considered by Dr.
Crile in other journals.

MetHops: The dogs were put on to a uniform diet, consisting
of milk (230-265 cc.) and dog biscuit, the amount given being
proportionate to the body weight. A quantity of the dog bis-
cuit sufficient for all the experiments was ground up, and after
being thoroughly mixed was kept in a large bottle. Analysis of
samples of this showed 3.3 per cent of nitrogen.

It will be observed from the tables that the dogs were practi-
cally in nitrogenous equilibrium before the operations took
place. The dogs were kept in a large metabolism cage and were
catheterized every 24 hours.

As to the methods of analysis of the urine, Kjeldahl’s method
was used for total nitrogen, while Folin’s methods? were used for
estimating urea, ammonia, uric acid and creatinin. Inasmuch as
I did not have access to a suitable apparatus, Dr. Folin himself
kindly made the creatinin estimations for me.

CONSIDERATION OF ReEsuLts. Experiment 1: Dog A (see
Table I) was put on the milk and biscuit diet until the nitrogenous
excretion had become constant, when a control operation was
performed. In this, the dog was kept under ether for one hour,
during which time the femoral blood vessels were exposed but

1H. M. Hanna Research Fellow, Western Reserve University.
? Folin: Amer. Journ. of Phystol., xiii, p. 45, 1905.
321

322 Effect of Blood-transfusion on Metabolism

not injured. This operation resulted in a slight increase in the
excretion of tatal nitrogen and some increase of that of ammonia.
A few days later a transfusion was performed, 250 cc. of blood
being withdrawn and being replaced immediately afterwards by
230 cc. from Dog B. The latter dog had not been given the
preparatory diet, but was taken from the dog room to replace
the dog that had been specially dieted for this purpose but was
accidentally killed. Even after allowing for body weight and
for the effect of hemorrhage, Dog B was shown, by the urines
following operation (Table II) to be on a distinctly higher plane
of excretion than Dog A as regards total nitrogen, ammonia,
uric acid and creatinin. There was a marked change in the com-
position of the urine of Dog A after transfusion, namely, an
increase affecting total nitrogen and urea.

We might suppose that this change was due to the character
of the blood received. To decide this it was necessary to com-
pare the above effect with that of hemorrhage alone. When the
wounds had healed sufficiently and the composition of the urine
had come back almost to what it was before transfusion, another
operation was performed in which 130 cc. of blood were removed.

On comparing the effect of hemorrhage (+ anesthesia) with
the effect of transfusion alone in the same dog the following are
noted:

As regards nitrogen excretion, averaging two days before and
two after operation, hemorrhage caused 14.7 per cent increase,
transfusion 20 per cent; as regards the per cent of nitrogen
present as ammonia, averaging four days before and after, hem-
orrhage caused an increase amounting to 11 per cent, transfu-
sion 15 per cent; as regards the per cent of nitrogen excreted as
urea, a more marked decrease occurred from hemorrhage alone—
dropping from 85.7 per cent to 72 per cent—than from trans-
fusion, which caused a drop from 84.2 per cent to 76.8 per cent.
Uric acid was increased for one day in both cases, being 18 per
cent higher after transfusion and 14 per cent after hemorrhage,
as compared with the day before operation. Creatinin decreased
7.8 per cent after hemorrhage, and increased 3.7 per cent (aver-
aging several estimations before and after) after transfusion.
This increase of creatinin does not seem to be at all proportional
to the very high excretion of the donor (Dog B).

223

Howard D. Haskins

‘BH 9'¢ “YSIOM
“AyQpesy puno

‘UMPIP
pooyq ‘99 OST eseYysIIOWaY—uOT}eI90dQ

‘BY 2°¢ ‘74SI0 \\

pooj UT UasoIqIU ‘Ws G'Z ‘AYYeEaYy puNno\\
‘Suryeinddns puno
‘BH 19 ‘IYSIOM

‘gq WOdJj pasnjsuesy poolq

99 OFZ ‘UMBIP POOTq ‘do NEY ‘uOIsNjsuesy

‘1, [ PISOYyysouR ‘uo1yeIedo JoryUOD

‘BY C'9 ‘FYSIO MW

‘pooj Ul UasoI4IU “WIS 77
‘BY 99 ‘IYO \

“jolp yMosiq pure YIN

*SYIBUI9}]

|
|

“sues
‘ulUuly
-Ral)

£60"

860°
GEO"

860°
C60"
6GO-
ro Oe
0G0
610°

960°

GG0°

0¢0°
00°
LG0°
060°
OGO°

‘PLy
et)

‘vy 90d

8
(0) fe
GE
VG
(6)!
Gall
Ww) ks
6°
Lax
6 °F
8'§
Pi
GG
Ey
O) te
VG
LG
8°G
LG
IG
8 PF
Ly
GG
ONS
0G

‘suwia | N°HN | N B21)

Z80°
te IZ] (OXON
FOL ©
GFL | 9OT°
ORGZ | OL0)
Me ERS WHA)”
0°98 | PEL”
£80"
990°
Os leer cites
SF OKO) A)
ty tev, |) COIN
to) 42, ie L2(0)
| 6 PS | 860"
¢L0-
0°¢S8 | 0¢S0°
990°
SGxiGenlSSZ0r
€°F8 | 290°
& 18 | 690°
SOT”
ell
190°
GOL”
G8I-

“Sued

Ses || COSOIFUNy

“UsBOIQIN [B90],
JO JUuaD Jeg uy

al

WT V.L

B UOlWUIYy

Ven)

N

©
NANN

“suBIa
*‘(LaBOIQIN
BOL)

L6G
O8L°%
OGG &
O1F FP
16¢ ¢
Of €
VOL=S
LO8°G
O&L°G
GOL NG
[06 °6
[69°
O9T'€

CLES
9066
AOL
1G0°G
OEL SG
190 'G
LAGIG
VIS G
GE38 G
9GE'G
LE0°S
LEGS
“Ssueld
‘Ould
ul
UIZO.IYIN

O9T
CGI
COT
IST
OGI

006
C86
VST
GET
CVI
OFT
OOT
89T

O¢T
OST
CsI
O9T
OFE
066
LIG
OFT
CoOL
Cél
crI
SGl

"OO.

“euliy)
jo

IUINTO A |

LG ”
EG ss)
GG ”
1G ”
0G »”
(ont ”
61 ”
SI »”
9| ”
¢ l ”
be} ”
Me ”
9 ”
¢ ”
ii2 ”
€ ”
G ”
I ‘G9
I € ”
O¢ ”
64 ”
LG ”
inG ”
GG ”
real! ”
OT WL
FT ‘uel
“LO6BT
, ‘ayecT

9€
Lac
€€
GE
T€

OF
6G
8G
9G
6
GG
1G
0G

324 Effect of Blood-transfusion on Metabolism

The effect of hemorrhage in increasing total nitrogen excretion
had been previously observed by Hawk and Gies.’

Experiment 2; In this case both dogs (see Tables III and IV)
were under exactly the same dietary conditions and their excre-
tions were just about the same, when allowance is made for the
difference in body weight. The amount of blood drawn and
transfused was not determined.

Unfortunately the effect of hemorrhage on the excretion of
Dog D could not be studied very successfully because accidents
prevented the collecting of the urine for two days after the opera-
tion.

With Dog C, transfusion caused 35 per cent increase of nitrogen
(averaging two days before and after operation), 48 per cent
increase of ammonia nitrogen per cent (averaging three days
before and after), 22 per cent increase of uric acid, and 6.5 per
cent increase of creatinin (averaging two days before and after).

By comparing with Experiment 1, we find that the increases
are greater in the case of C than in that of A. Therefore, we
cannot argue that the blood of the higher (meat) diet dog, B,
produced any characteristic effect different from that produced
by the blood of the medium diet (milk and biscuit) dog, D. Diet
does not seem to be an essential factor. Judging from the results
in these two experiments, it would seem that transfusion (or
rather hemorrhage followed by transfusion) produces practically
the same effect on nitrogenous metabolism as does moderate
hemorrhage alone, but toa much greater degree, and further,
that the nature of the diet given the donor previous to transfusion
is of no importance in influencing the result.

A third transfusion was successfully performed, but unfor-
tunately the dog refused food and was quite sick following the
operation, so that further work on the urines was rendered value-
less.

In the case of all the other dogs, transfusion and hemorrhage
were quickly recovered from, in so far as could be judged by
their appearance and behavior.

The work reported here is hardly extensive enough to warrant

' Hawk and Gies: The Influence of External Hemorrhage. Amer. Journ.
of Physiol., xi, p. 171.

Howard D. Haskins

| . | . a

: POV rc0 "Gy Jeco0r. || sreuclmor Sie s,
‘SABP [BIOAOS IOF BULIN JOaT[OO YOU P[Noy | FCT" Le 020° | 87% 20° Oco'% | GIT | 8 » i
‘) 0JUI pasnjsuers} poolq Dee a IAO | | | ¢ - ra
‘By G9 ‘WY8IOM | OST” LOS ACIO Es NI Recall cli: AIG os. fe
‘pooj ut “uaSomiu | | | |
‘US J$'°Z ‘SAPP [B1IAaS JaIP WMosIq pu YIU UO uaeq Sv yy | SFL’ r 610° | OZ S70") O884G) 1-49 enor een
‘d 90@ ‘AI aTaVL
Ge" | es0" || Lae'c) Soviee ot eaaees
Colas) CO! eras: 820° 4066 | G8 |6 » Z
6°¢ Sél- 8GE°G | 8 » 9
TE | RAD BS Gam SoC G6) | O0G. aiuz » ¢
F Se eGGO! sleeeal 9¢0° 896°€ | OAT | 9 ” v
‘q Wodj uotsnjsuvs} Aq pamol[oj aseys0Way—uorjrsisdg | | ¢ 5
DAES AUS SS die ae) SIE GN ts
BAC OSM | Shi |t9co" | 12 | E20 | Oe-c | oar oy, j
‘pooj ut ueso1q1U |
"ms €2°¢ ‘sAep arenes ¥1P es ees tien uo aie SAP | Gy Oke L830" Gree s |G Youre 1
So Ae Senet GEG juafo.nIN *SUIBIS ‘SUB IG or) | a
“SY1IBU9 YY] ul Ulu} | ‘poy ee lmeaatan | eee tie | | jee Stal
EDO) PMY | deg UP BUOY) UaFOIN aUN[OA
2) D0G Stata
- ‘BY ET ‘IYysIOM eS Vee OM elec
"Y OUI posnjsursy poojq ‘99 | | |
08% 4SO[_ ‘uorvsedo 210j0q 12IP wou UO ¢'p | 0°18 | 80g’ | #ez°¢-| gT8‘9 | l¢@ ‘qe | 1
ane seen | gs | UNL NON | oon | aa | oan | ate | caper [log
"OT, BYR) [rota og a ce ROI). cenodin Sranan | ee omen
G Se) O Giga ane cai val

326 Effect of Blood-transfusion on Metabolism

the drawing of any conclusions further than that transfusion of
blood following hemorrhage has practically the same effect on
nitrogenous metabolism as hemorrhage alone has. In other
words, the transfusion of normal blood after hemorrhage does
not prevent the effect on nitrogenous metabolism of hemorrhage
alone.

THE PICROLONATES OF CERTAIN ALKALOIDS.
By W. H. WARREN anp R. S. WEISS.

(From the Chemical Laboratory of the Medical Department of Washington
University, St. Louis, Missourt.)

(Received for publication, June 25, 1907.)
(Plates III-VIII.)

Our attention was called to the use of picrolonic acid in pre-
cipitating alkaloids and we have studied its behavior with some
of the more important representatives of this group. Like picric
acid picrolonic acid precipitates many basic nitrogen compounds
and, since the latter has given better results in other instances,
we thought it advisable to determine the relative value of these
reagents as alkaloidal precipitants. They are widely different
as far as ease of obtaining them is concerned. Picric acid is
cheap and quickly prepared, whereas picrolonic acid is not listed
by manufacturers and its preparation is rather difficult.

Knorr! gave the name “ picrolonic acid’’ to 1-p-nitro-phenyl-
3-methyl-4-nitro-5-pyrazolon,

Ch — C=
| SN C,H,—NO,,
NO-_C—¢
=
OH

which he preferred to picric acid in studying 1,2-ethanol-amine
and certain morpholine derivatives, because of the greater insolu-
bility of their picrolonates and the ease with which these could
be recrystallized and identified. Recently Steudel’ has pointed
out the value of this reagent inisolating and identifying the two
hexone bases, arginin and histidin.

1 Ber. d. deutsch. chem. Gesellsch., xxx, p. gog, and xxxii, pp. 732 and 736.
2 Zeitschr. f. physiol. Chem., xxxvii, p. 219.

327

328 Picrolonates of Certain Alkaloids

PREPARATION OF PICROLONIC ACID.

Methyl-phenyl-pyrazolon, prepared from acetacetic ester and
phenylhydrazine, according to Knorr,’ was converted into picro-
lonic acid by Bertram’s method as described by Steudel.’ In
the hope of improving the yield we tried several modifications
of this method but had indifferent success. Finally we employed
the method of Braun* which gave a result so much better than
any we had obtained up to that time that we can recommend it
for the preparation of picrolonic acid. As this method is some-
what inaccessible, we have described our manner of carrying it

out.

Six hundred cc. of 90 per cent nitric acid (sp. gr., 1.49), prepared by
diluting 540 cc. of strongest fuming nitric acid (sp. gr., 1.52=99.7 per
cent) with 60 cc. of water, were kept cold with ice-water in a wide-mouth
bottle. Two hundred grams of methyl-phenyl-pyrazolon were added
gram at a time, the mixture being mechanically stirred all the while and
the temperature being kept below 15° C. Each addition of substance
caused a vigorous reaction attended with escape of red fumes. Halfway
in the process, a reddish-yellow, crystalline product began to appear. At
the end, the mixture was stirred for thirty minutes and then the crystal-
line mass was collected upon a small porcelain disc covered with asbes-
tos, freed from acid by suction and washed successively with dilute nitric
acid and with water. As this crude product gradually undergoes decom-
position, it is well to proceed at once with the next step in the process.

Fifteen hundred cc. of 33 per cent acetic acid, about six times the weight
of the crude product, were mechanically stirred in a beaker on the water-
bath. The crude product finely ground was added and the temperature
was not aliowed to exceed 60° C. This substance, which is heavy and
granular, tends to settle to the bottom of the beaker. At low tempera-
tures it remains unaltered but gradually changes as the temperature rises.
The change is rapid from 50° to 60°. A mass of light-yellow, flocculent
crystals replaces the reddish-yellow compound and fills the liquid. Sa-
ponification is complete in 15 to 4o minutes. The product may be exam-
ined occasionally under the microscope, if necessary, and the reaction is
complete when only needles of picrolonic-acid are to be seen.

The crude picrolonic acid, freed from acid on a Biichner filter, was
washed with water and then ground in a mortar with 150 grams of crystal-
lized sodium carbonate. Carbon dioxide was given off and the vellow
sodium salt was formed. This was pressed out and crystallized from a

* Ber. d. deutsch. chem. Gesellsch., xvi, p. 2597.
? Zeitschr. f. physiol. Chem., xliv, p. 157.
? Inaugural Dissertation, Jena, 1899.

W. H. Warren and R. S. Weiss 329

mixture of three volumes of strong alcohol and one volume of water.
Sodium picrolonate was then decomposed with concentrated hydro-
chloric acid. The acid was warmed in a beaker on the water-bath, stirred
as before and the sodium salt was added in small portions. Picrolonic
acid separated as a fine, mealy powder. It was thrown on a porcelain
disc covered with asbestos, freed from acid by suction and washed with
water to remove acid and sodium chloride. We obtained 166 grams of
picrolonic acid by this method from 200 grams of methyl-phenyl-pyra-
zolon. If further purification is desired, picrolonic acid may be crystal-
lized from strong alcohol from which it separates in fine yellow needles
which melt at 116.5° C.

USE OF PICROLONIC ACID.

A saturated alcoholic solution of picrolonic acid is the most
convenient form in which to use this reagent for ordinary pur-
poses. Such a solution may be employed to precipitate alka-
loids from water, alcohol or from other solvents miscible with
alcohol. Although picrolonic acid itself is only slightly soluble
in water, sufficient alcohol is present to prevent the reagent
from being precipitated when added in excess to aqueous solu-
tions.

In some cases 1t may be an advantage, after extracting an
alkaloid from an aqueous, alkaline solution with such solvents
as ether, chloroform, benzol, acetic ether and amyl alcohol, to
substitute for the alcoholic reagent a saturated solution of picro-
lonic acid in the particular solvent used to extract the alkaloid.
In isolating morphine, for example, the alkaline solution may
be extracted either with hot chloroform or with hot amyl alcohol
and this extract may be tested directly witha solution of picrolonic
acid in the corresponding solvent. The following table shows
the relative solubility of picrolonic acid in the solvents men-
tioned above:

SOLUBILITIES OF PICROLONIC ACID AT ORDINARY TEMPERATURE.

e Acid in 1 ee.

Solvent. in Grams.
\Wilgen ay capa st bar Meteo, Sie deg 8 tained Make Ses Di ale Gre ee 0.0017
Janlveyol stot eRe kt eae Mein Saree. Rie yoc aee ce os no eee 0.0209
| D8) dV oe ORT aS RCE Fe a eee ce fee 5 ee 0.005
BYES OVC) Lae eRe gue eA Ey Sl Re aE tn ade! Sak ine tine 0.0024
Ghiloroforias we ae ec ie cn ae ee ee 0.015
VCore CONST 7) O1 3) ey ES EE eters nea a EPO cnn Ce ORO ne 0.041

{irigg ite foo) 110) Ae ae Ae Ink AIR atm rr Bee Cope Re 0.0056

330 Picrolonates of Certain Alkaloids

To recover an alkaloid from its picrolonate, it is necessary
only to warm the latter with dilute sulphuric acid which dis-
charges the bright yellow color of the picrolonate, causing the
alkaloid to pass into solution and precipitating pale-yellow
picrolonic acid. By extracting with acetic ether, in which picro-
lonic acid is especially soluble, the aqueous solution of the alka-
loid is left colorless. This has suggested a method of purifying
alkaloids after their extraction in the usual way from animal
tissues. An alkaloid like strychnine, whose picrolonate is very
insoluble, may be precipitated from aqueous solution and thus
separated from other substances which interfere with its purifi-
cation. The precipitated picrolonate may be collected on a
filter, washed with water and then warmed with dilute sulphuric
acid. After extracting the free picrolonic acid as before, a very ©
pure solution of the alkaloid will remain.

COMPARISON OF PICROLONIC ACID AND PICRIC ACID.

Although picric acid may be used either in aqueous or in
alcoholic solution, we were limited in the case of picrolonic acid
to alcohol because of the slight solubility of this substance in
water. The two reagents were prepared by saturating go per
cent alcohol with these acids. Such a solution of picric acid con-
tains at ordinary temperature 0.0596 gram in 1 cc. and the cor-
responding solution of picrolonic acid contains 0.0209 gram in
rece. There is no precipitation of either reagent upon adding it
to 5 cc. of water volume for volume, or in smaller quantity. A
precipitate, therefore, in an actual test is due to alkaloid and
not to dilution of the reagent with water.

The solution of each alkaloid was prepared either by dissolv-
ing one of its salts in water, or by dissolving the alkaloid itself
in water with the aid of dilute acetic acid. We began in each
case with a 1 per cent solution, aliquot volumes of which were
diluted with water until a point was reached where the test
failed. Time is a factor to be considered in deciding whether
a test is positive or negative, especially when the solutions are
dilute. The precipitate may not be immediate but it will some-
times appear upon standing. We adopted fifteen minutes as an
arbitrary limit and called the test negative when there was no

W. H. Warren and R. S. Weiss 331
precipitate within that time. Frequently there was a turbidity
instead of a distinct precipitate but this sometimes became
crystalline upon standing. We decided to accept as positive
those tests where a turbidity gave such a result. The relative
delicacy of the two reagents toward the alkaloids studied appears
in the following table:

DELICACY OF REAGENTS AS ALKALOIDAL PRECIPITANTS.

Picric ACID. PICcROLONIc Acip.
ALKALOIDS. aa: Pos cae ee ne Seon
Fontive Remarks. | Postive: Remarks.
(Connne ss. fae 1-250 Turbidity with- | 1-800 | Very slight in
out precipitate 15 min.
Nicotine....... 1-20000 Slight crystalline | 1-20000 Slight crystalline
precipitate precipitate
Strychnine.... 1-40000 Very slight in || 1-75000 Very slight in
5 min. | 15 min.
Bggeine : 2% <5. 1-20000 Positive in 12 | 1-10000 Very slight in
min. 15 min.
Morphine......| 1—400 Very slight in | 1-500 Positive in 30
15 min. min. in 1—1000
Codeine....... 1-1000 Very slight | 1-500 Very slight
AGfOpine oi: .. 1-1000 Turbidity which | 1-1000 | Turbidity which
becomes crystal- | becomes crystal-
line | | line
GQamine 2.2.2 s.. _ 1-50000 Turbidity with- 1 1-50000, Turbidity with-
| out precipitate | | out precipitate
|
Hydrastine....| 1-10000 Turbidity with- | 1-10000| Turbidity which
out precipitate || | becomes crystal-

line

Our results show considerable variation in the limit of dilution
within which alkaloids can be detected with these reagents.
Strychnine, which can be detected by picrolonic acid in 1-75000,
affords the most delicate test, whereas coniine can barely be
detected by picric acid in 1-250. Yet the difference between
these reagents is not as marked as we had expected when we
began our experiments. Upon the whole picrolonic acid is the

332 Picrolonates of Certain Alkaloids

more sensitive, especially toward coniine, strychnine and mor-
phine. On the other hand, picric acid is the more sensitive
toward brucine and codeine. They appear about equally sensi-
tive toward nicotine, quinine, atropine and hydrastine. As a
matter of fact picric acid is a better reagent for nicotine, because
the precipitate is distinctly crystalline and quite characteristic
in r-20000. This reagent would seem to be well-adapted for
detecting this particular alkaloid.

The recognition of these alkaloidal picrolonates and picrates
by means of the microscope depends entirely upon the method
of preparing these compounds. It is difficult, if not impossible,
to distinguish most of these substances when they have been
precipitated from aqueous solution. We should hardly expect
it to be otherwise. We observed a difference in those cases
where there was a turbidity which became crystalline upon
standing. The crystalline form was more distinct. For this
reason we have always redissolved the substance, preferably
in alcohol, and allowed crystallization to take place slowly.
This applies more particularly to our later work where we had
occasion to examine crystals carefully either to describe them
or to take photo-micrographs. Since our results with alkaloidal
picrates are for the most part like those of Popoff,t who has
made an extensive study of picric acid as an alkaloidal precip-
itant, we have described and photographed only a few of the
more important of these compounds.

PROPERTIES OF ALKALOIDAL PICROLONATES AND ANALYTICAL
RESULTS.

In preparing alkaloidal picrolonates for analysis and for observ-
ing their properties, we have obtained the best results by adding
an excess of alcoholic reagent to a hot solution of the free alka-
loid in as small a quantity of alcohol as possible. Some picrolo-
nates appear immediately; others crystallize as the solution cools
and yet others require long standing or concentration of the
solution. Before being analyzed, these compounds were recrystal-
lized at least once and dried at 1oo—-110° C., or by standing in
desiccator over sulphuric acid.

* Le Laboratoire de Toxicologie: Brouardel et Ogier, pp. 203-220, 189.

W. H. Warren and R. S. Weiss 232

Contne Picrolonate—C,H,,N.C,,H,N,O,.

Coniine picrolonate was prepared by adding a slight excess
of reagent to Merck’s pure coniine dissolved in a small quantity
of alcohol. Since the precipitate, which appeared immediately,
did not show the characteristic crystalline form of this substance,
it was recrystallized once from alcohol somewhat diluted with
water. Coniine picrolonate forms large rhombohedrons when
it crystallizes slowly from alcohol, but when crystallization is
rapid it appears as short, thick, irregular crystals with pointed
ends and these are often in rosettes (Fig. 1). These crystals
are quite different from those of the corresponding nicotine com-
pound and serve to distinguish these two liquid alkaloids. Co-
niine picrolonate is light-yellow, it dissolves readily with heat
either in water or in alcohol and melts at 195.5° at which point
it becomes dark owing to decomposition. Upon analysis it gave
the following result:

0.2252 gram of substance gave 35.5 cc. of nitrogen gas at 21° and 748

mm.
Caleulated for
1. gHosN;O5: Found:

EN LsR oy -t.2 2st yack S206 5 17-9) per cent: 17.7 per cent.
Nicotine. Picrolonate—C,,H,,N,-C,,H,N,O;.-

As our available alkaloid showed signs of change due probably
to long standing, we prepared the dioxalate and recrystallized
this compound until it melted at 110° C. Nicotine picrolonate
was precipitated by adding an excess of reagent to an aqueous
solution of the dioxalate. The precipitate was crystallized from
alcohol somewhat diluted with water. This compound forms
long, thin, prismatic needles with square ends which are often
grouped in fan-shaped clusters (Fig. 2). It is light-yellow and
melts at 213° previously subliming somewhat and turning dark.
Upon analysis it gave the following result:

0.2015 gram of substance gave 35.6 cc. of nitrogen gas at 19° and

47 mm.
gen Calculated for

CocHo2N 605:

IN Gch s oss einer aera 19.72 per cent. 19.96 per cent.

Found:

Nicotine, though a diacid base, appears to combine with but
one molecule of picrolonic acid. On the other hand it combines
with two molecules of picric acid. This compound has been

334 Picrolonates of Certain Alkaloids

described by Pinner and Wolfenstein’ and, as stated elsewhere,
we have found it preferable to the picrolonate. Aqueous picric
acid precipitates nicotine picrate as small yellow needles which
melt at 218° and dissolve with difficulty. Crystallized from
hot alcohol it appears in short, hexagonal prisms often terminated
by hexagonal pyramids (Fig. 3). Some forms resemble quartz

crystals.
Strychnine Picrolonate—C,,H,.N,0,.C,,H,N,O;.

The picrolonates of the two nux vomica alkaloids, strychnine
and brucine, melt with difficulty, or not at all, and differ mainly
in crystalline form. Strychnine picrolonate was prepared by
adding the reagent in excess to a hot alcoholic solution of the
alkaloid. The precipitate consists of crystalline leaflets but,
when crystallized slowly from alcohol, it appears in rectangular
prisms, more or less irregular in form, which are sometimes
arranged in star-shaped clusters (Fig. 4). This compound is
light-yellow and very insoluble in water or in alcohol of any
strength. When heated it begins to darken at 256° and the color
increases as the temperature rises, making it difficult to deter-
mine the precise melting-point which seemed to be about 275°.
Upon analysis it gave the following result:

0.2016 gram of substance gave 24.9 cc. of nitrogen gas at 20° and
747 mm.

Calculated for
C3)H3yN 507:

3 felines Sh me ced Ae ae 14.04 per cent. 13.9 per cent.

Found:

In comparing strychnine picrolonate with the picrate, we
found the latter compound mentioned by Guareschi? who refers
to Gerock’s use of the picrates of strychnine and brucine in
separating these alkaloids but does not give their properties.
Strychnine picrate, prepared by adding an excess of alcoholic
reagent to a hot alcoholic solution of the alkaloid, separates from
solution in leaflets. Redissolved in alcohol and allowed to
crystallize slowly, it appears in short, slightly thickened trape-
zoidal plates, the non-parallel ends of which are usually irregular

* Ber. d. deutsch. chem. Gesellsch., xxiv, p. 66.
* Einjuhrung in das Studium der Alkaloide, p. 505, 1896

ome

W. H. Warren and R. S. Weiss 335

in outline (Fig. 5). It undergoes decomposition when heated
and up to 300° does not melt.

Brucine Picrolonate—C,,H,,N,O0,.C,)H,N,O,.

Brucine picrolonate was prepared by adding the reagent in
excess to an alcoholic solution of the alkaloid. The precipitate,
which is almost immediate, consists of cubical crystals many
of which look like rhombohedrons (Fig. 6). This compound is
very insoluble in water or in alcohol of any strength. When
heated it begins to darken at 234° and finally melts at 256°.
Upon analysis it gave the following result:

0.2011 gram of substance gave 22.8 cc. of nitrogen gas at 20° and
753 mm.

Caleulated for c *
C33H 34N 6Ov: Found:

BN eysimis: sits 5 =. cleicier areas 12'..7 per cent. 12.8 per cent.

In comparing brucine picrolonate with the picrate, we noticed
a peculiarity in the behavior of the latter. An alcoholic solution
of brucine was added to an alcoholic solution of picric acid.
There was an immediate yellow crystalline precipitate and, when
this was allowed to stand without being filtered, an orange-red
compound also appeared gradually. Upon heating the solution
to boiling, the yellow substance was dissolved and, as the solution
cooled, only the orange-red compound reappeared. This latter
substance crystallizes in cubical crystals and in elongated, rect-
angular prisms terminated by low pyramids (Fig. 7). When
heated up to 310°, neither the yellow nor the orange-red com-
pound melts though both undergo decomposition. At 213° the
yellow compound becomes orange-red. Judged by the follow-
ing analyses, there is not a wide difference in their composition:

Yellow Compound—C,,H,,N,O,.C,H,(NO,),0H.

o.2007 gram of substance gave 19.8 cc. of nitrogen gas at 18° and 739
mm.

Calculated for = oe
CogHogN 3011: ound:

DIN PR ay tieec, arate yet ale Breage 11.23 per cent. 11.08 per cent.
~ Orange-red Compound—C,,H,,N,0,.C,H,(NO,) OH.

©. 2016 gram of substance gave 20 cc. of nitrogen gas at 19° and 751 mm.

Calculated for Bae
CogHogN 5011: eter:

TN ORGS pe Hotere een 11.23 per cent. 11.32 per cent:

336 Picrolonates of Certain Alkaloids

Morphine Picrolonate—C,,H,,NO,.C,)H,N,O;.

This compeund was prepared by adding the reagent in excess
to an alcoholic solution of the alkaloid and evaporating the
solution until crystallization began. Morphine picrolonate forms
broad, flattened needles which resemble knife-blades and are
sometimes arranged in fan-shaped clusters (Fig. 8). It is quite
readily soluble in alcohol. When heated it melts at 186.5° and
turns dark-brown near its melting-point. Upon analysis it gave
the following result:

©. 2024 gram of substance gave 22 cc. of nitrogen gas at 16° and 749 mm.

Leet for * 5
CorHoyN;Oa: Found:

SEO. ee 12.7 per cent. 12.5 per cent.
Codeine Picrolonate—C,,H,,NO,.C,,H,N,O,.

This compound was prepared by adding the reagent in excess
to an alcoholic solution of the alkaloid. When the solution is
very concentrated and crystallization is rapid, codeine picro-
lonate appears in short, thick crystals which form rosettes; but
when it crystallizes slowly from hot alcohol it forms character-
istic crystals looking like truncated pyramids (Fig. 9). It is deep-
yellow and melts at 219°, turning orange-red just before melting
and becoming dark as it melts. Upon analysis it gave the
following result:

0.2054 gram of substance gave 22.4 cc. of nitrogen gas at 20° and’750
mm.

Calculated for :
CosH ooN 50g: Found:

RASA e nichts Sialege 3 12.4 per cent. 12.3 per cent.
Atropine Picrolonate—C,,H,,NO,.C,,H,N,O;.

This compound was prepared by adding the reagent in excess
to an alcoholic solution of the alkaloid. The yellow precipitate
was recrystallized from alcohol in which atropine picrolonate
is readily soluble. It appears as short, pointed crystals which
are grouped in thick, spherical masses (Fig. 10). It melts at
194° and turns orange-red. Upon analysis it gave the following
result:

9.2009 gram of substance gave 22.4 cc. of nitrogen gas at 20° and 752
mm.

Calculated for
Co;Hg)N Ox: Found:

tabetha Sa %s iid ccc ven 12.66 per cent. 12.63 per cent.

W.H. Warren and R. S. Weiss 3

ios)
“I

Quinine Picrolonate—C,,H,,N,0,.2 C,,H,N ,O;.

This compound was prepared by adding a hot alcoholic solu-
tion of quinine to a hot alcoholic solution of picrolonic acid con-
taining the reagent in the proportion of two molecules to one
molecule of the alkaloid. The picrolonate crystallized slowly
and seemed rather soluble in alcohol, but in recrystallizing it
from hot alcohol we found that a much larger volume of solvent
was required for complete solution. This compound appears in
hair-like needles which form spherical masses or sheaf-like clusters
(Fig. 11). It begins to darken at 223° and melts at 225°. Upon
analysis it gave the following result:

0.2007 gram of substance gave 28.8 cc. of nitrogen gas at 19° and 750
mm.

Calculated for ; -
C4pH4oN 100 12: Found:

BN eps ich Paksi ‘ote we & cl eeels 16.4 per cent. 16.28 per cent.
Hydrastine Picrolonate—C,,H,,NO,. C,,H,N O;.

This compound was prepared by adding the reagent in excess
to a hot alcoholic solution of the alkaloid. There was an imme-
diate crystalline precipitate which was recrystallized from hot,
alcohol somewhat diluted with water. Hydrastine picrolonate
is light-yellow and melts at 220°. It appears as long, flat, pris-
matic needles with square or pointed ends and these are some-
times in spherical clusters (Fig. 12). Upon analysis it gave the
following result:

o.2013 gram of substance gave 19.7 cc. of nitrogen gas at 21° and 747
mm.

Calculated for x i:
C31HoN5O11: Found:

SN Diep oa SO ner ea ee 10.81 per cent. 10.95 per cent.

Other Alkaloids and Picrolonic Acid.

In addition to the preceding nine alkaloids, which form definite
crystalline compounds with picrolonic acid, we have also studied
the action of this reagent upon cocaine, aconitine and caffeine.
We did not succeed in obtaining as satisfactory results with
these three alkaloids. Hager! found that a saturated aqueous
solution of picric acid would not precipitate caffeine and there
was some doubt about its giving a precipitate with aconitine.

1 Zeischr. fj. anal. Chem., ix, p. 110.

338 Picrolonates of Certain Alkaloids

Pure cocaine hydrochloride, dissolved in water and treated
with alcoholic picrolonic acid, gave only a thick resinous pre-
cipitate which did not show the least sign of becoming crystal-
line after long standing, or when dissolved in alcohol in which it
is readily soluble. The same result was obtained when the
alkaloid itself was dissolved in alcohol and treated with the
reagent. Consequently we did not make any further exami-
nation of this substance.

We had a hope that aconitine, for which the number of well-
defined tests is limited, might give a characteristic result with
picrolonic acid. When pure crystallized aconitine was dis-
solved in alcohol and treated with picrolonic acid, there was not
an immediate precipitate. As the solution evaporated, crystal-
lization began but it was evident that these crystals were not
homogeneous. Some of these were yellow and others were
white. A result quite similar to this was given also by caffeine.
We did not, therefore, continue the examination of the picrolo-
nates of these alkaloids.

In conclusion we desire to acknowledge our obligations to
Prof. H. Steudel who directed our attention to the subject con-
sidered in these pages and very kindly supplied us with a portion
of the picrolonic acid required for the work.

THE JOURNAL OF BIOLOGICAL CHEMISTRY. VOL. III. PLATE III.

Fig. rt. Coniine Picrolonate.

Fig. 2. Nicotine Picrolonate.

THE JOURNAL OF BIOLOGICAL CHEMISTRY. VOL. III. PLATE IV.

Fig. 3. Nicotine Picrate.

Fig. 4. Strychnine Picrolonate.

THE JOURNAL OF BIOLOGICAL CHEMISTRY. VOL. III. PLATE V.

Fig. 6. Brucine Picrolonate.

THE JOURNAL OF BIOLOGICAL CHEMISTRY. VOL. III. PLATE VI.

Fig. 7. —-Brucine Picrate:

Fig. 8. Morphine Picrolonate.

{AY

THE JOURNAL OF BIOLOGICAL CHEMISTRY. VOL, III. PLATE VII.

Fig. 9. Codeine Picrolonate.

Fig. ro. Atropine Picrolonate.

a)

THE JOURNAL OF BIOLOGICAL CHEMISTRY. VOL. III. PLATE VIII.

Fig. 11. Quinine Picrolonate.

Fig. 12. Hydrastine Picrolonate.

FURTHER OBSERVATIONS ON PROTAGON.

By WILLIAM J. GIES.

(From the Laboratory of Biological Chemistry of Columbia University, at
the College of Physicians and Surgeons, New York.)

(Received for publication, July 13, 1907.)

The last issue of the Bzo-Chemical Journal’ contains a very
interesting paper, by Lochhead and Cramer,” ‘‘onthe phosphorus
percentage of various samples of protagon.’’ Lochhead and
Cramer stated in an introductory paragraph that ‘‘In this com-
munication we [they] intend to bring forward some positive
evidence in favor of the view that protagon is an individual
substance and not a mixture.’”’ The “positive evidence’’ that
Lochhead and Cramer brought forward in fulfilment of that
intention consists simply of less than a dozen figures for the per-
centage amounts of phosphorus in a few dissimilar so-called
protagon’ products obtained from ox brains with the aid of sev-
eral new methods.* Lochhead and Cramer considered that

1 Issued June 20, 1907 (Nos. 7 and 8).

Lochhead and Cramer: Bio-Chém. Journ., ii, p. 350, 1907.

3] say ‘‘so-called protagon products’ because the paper fai's to show
that they were identical with the preparations universally regarded as
typical, e. g., Liebreich’s.

4 That these methods were valid substitutes for the processes by which
the classical protagon products were made does not appear to have been
determined by Lochhead and Cramer. That their methods were satis-
factory substitutes, the authors seem to infer merely from the fact that
most of their products contained proportions of phosphorus equal approxi-
mately to from 1.0 to 1.1 percent. They appear to assume that this is the
proportionate content of phosphorus in pure protagon. Do Lochhead
and Cramerq@vish us to understand that the name protagon may now be
applied to any white brain material having a proportionate content of
phosphorus equal to from 1.0 to 1.1 per cent and do they consider that
any method by which such material can be obtained is a satisfactory
process for the preparation of protagon?

339

340 Protagon

their figures were in ‘close agreement,’’ but the analytic data for
percentage phosphorus contents, of the preparations of their
‘‘most highly purified products” especially, z.e., A, B,C and D
(which had been recrystallized only three times, however) show
clearly that there were, for at least three of the four purest prod-
ucts, very significant disagreements on recrystallization, as may
be seen in the appended summary of their results (p. 354):

PERCENTAGE OF PHOSPHORUS IN PROTAGON PRODUCTS MADE BY LOCHHEAD
AND CRAMER, WITH THE AID OF NEW METHODS.

Chloro-

[NV she
Solvent Used for Extraction Ethylaleohol sada Ethylaleohol | Payee
Samples of protagon....... A | B } Cc D 1D} F
Crude crystalline product... -- — —= — = Z 1.18
(“i ¥ Sia ae Route Se
Recrystallization from ..... Ethyl- | Ethyl- Glacial Methyl- :
: aio slechal _aleohol| Acetic | alcohol Methy ilcohes
Acid
First a = =— meen | 1.22 , |
Second 1.34 | 1.25 | (1.25) | 0.94 |1.14 1-05)
AMY Tao by ieee, ea 1.07 | 1.05 0.96 | 0.97 |

Why was recrystallization of products A and B discontinued
after the third treatment, when the decrease of proportionate
phosphorus content was so noticeable, and that too in the prod-
ucts recrystallized from ethylalcohol, the alcohol employed for
purposes of preparation and purification in practically all pre-
vious protagon investigations? It is very probable that addi-
tional recrystallizations would have shown furtherlowering of the
percentage phosphorus content. It is significant, also, that the
first product (D) which was recrystallized from methylalcohol was
at once different, in percentage phosphorus content, from prod-
ucts A and B. The proportionate contents of phosphorus in
the latter were probably unusually high after the first recrystalli-
zation, for which, unfortunately, Lochhead and Cramer present
no analytic data.

Lochhead and Cramer laid great stress on the constancy of
the percentage content of phosphorus in Gamgee and Blanken-
horn’s protagons, even after repeated recrystallization (see foot-
note 2, p. 349). They regarded that constancy of proportion-
ate phosphorus content as evidence of the complete removal of
pseudocerebrin, by Gamgee and Blankenhorn, from their prota-

William J. Gies 341

gons, and as proof of the purity of the latter. Lochhead and
Cramer also mistakenly assumed that Posner and I obtained a
product of constant composition after repeated recrystallization.
If such constancy of proportionate phosphorus content is so sig-
nificant, however, and implies so much in favor of the purity of
the product, what opinion do Lochhead and Cramer entertain
regarding their own results in this connection which, in all cases
except one, showed a steady fall of phosphorus content after
recrystallization. It seems to me that Lochhead and Cramer
have indeed obtained, contrary to their own opinion, “different
mixtures by extracting brain tissue by different methods.”
The results of the treatment of products A and B are appar-
ently very significant, from several standpoints. The state-
ment made by Lochhead and Cramer regarding the origin of
““Protagon C’’ is not very clear, but, as I understand it, ‘‘ Prota-
gon B,” which contained 1.25 per cent of phosphorus after the
second recrystallization from ethyl alcohol was then divided,
and each of two portions of it was subjected to a third recrystal-
lization; one portion from “‘boiling absolute alcohol,” by which
treatment phosphorus content was reduced from 1.25 to 1.05
per cent (B); the second portion (C) from “boiling glacial acetic
acid,’’ which caused even a greater decrease of phosphorus con-
tent—from 1.25 to 0.96 per cent. Does not Cramer recall that
Koch! obtained phrenosin (‘‘cerebrin’’)? from the material which
is extracted from brain by boiling alcohol and separated when
the extract is cooled to o° C., 7. e., from _protagon; that by
recrystallization from glacial acetic acid two or three times,
Koch thus succeeded in isolating pure, phosphorus-free phren-
osin; in effect, prepared phosphorus-free protagon by the method
Lochhead and Cramer used forthe purification of protagon? The
drop in percentage phosphorus content from 1.25 to 0.96 per cent
that was caused by but one crystallization from glacial acetic acid,
suggests that if crystallization had been repeated in that reagent
several times, Lochhead and Cramer would have obtained phren-
osin in as complete a state of purity and as free from associated
matter containing phosphorus, as Koch did. Lochhead and

1 Koch: Zeitschr. f. physiol. Chem., xxxvi, p. 138, 1902.
2 Koch: Amer. Journ. of Physiol., xi, p. 310, 1904.

342 Protagon
Cramer said nothing about any of this, so that it is probable the
entire matter was overlooked by them.’

Special allusion is made to protagons E and F on pp. 351 and
352 of this paper.

The significance of the above noted disagreements? is obvious
when certain other data are taken into account. | Lochhead
and Cramer laid special stress, and properly so, upon Gamgee’s
results and methods. The following quotation from Gamgee’s®
remarks shows how little Lochhead and Cramer agree with Gam-
gee on the matter of phosphorus content of protagon as affected
by the process of recrystallization.

‘The amount of phosphorus in specimens of protagon which had
been crystallized from alcohol four or five times was not smaller than
that present in protagon which had only once been crystallized, though a
thorough treatment with ether preceded each recrystallization.”

Some of the duplicate results obtained by Gamgee in his phos-
phorus determinations were the following: (A) 1.094 and 1.107,
(B) 1.032 and 1.081. (See remarks on p. 344 of this paper.)

With regard to the results tabulated on p. 340 and which the
reader sees, in the light of the remarks just made, are notin par-
ticularly close accord, Lochhead and Cramer remarked as follows,
at the beginning of their summary of conclusions:

‘The close agreement(!) between the phosphorus percentage of various
samples of protagon prepared by the most diverse methods is strong evi-

‘ Besides indicating methods for the preparation of phrenosin (‘‘cere-
brin’’), lecithin and kephalin, Koch said at the conclusion of one of the papers
already referred to (Zeitschr. f. physiol. Chem., xxxvi, p. 140, 1902, in
which his observations on these three substances were summed up): ‘“‘ Das
Protagon braucht weiter nicht beriicksichtigt zu werden, da ich die
Theile (phrenosin, lecithin, kephalin), aus denen es aufgebaut, direkt
bestimme.’’ In the paper in the American Journal of Physiology (xi,
P. 324, 1904) Koch said a typical preparation of protagon (Noll’s, Zeitschr.
j. physiol. Chem., xxvii, p. 370, 1899) ‘‘contained about 75 per cent of
phrenosin (cerebrin); the remainder consisted of sulphur compound and
either lecithin or more probably kephalin.”’

* Designated by Lochhead and Cramer as examples of ‘‘close agree-
ment.”’

*Gamgee: A text book of the physiological chemistry oj the animal body,
1, Pp. 427, 1880.

*Gamgee and Blankenhorn: Zeitschr. f. physiol. Chem., iii, p. 279, 1879.

William J. Gies 343

dence in favor of the view that protagon is an individual substance of a
well-defined chemical composition.”

The second of the three sentences comprising their summary
of conclusions is the following:

‘“Even more conclusive evidence (than that referred to in the preceding
quotation) is afforded by the observation of Posner and Gies, that after

ten times repeated crystallization the protagon crystals separating out
have the same phosphorus percentage as the substance in the mother liquor.’”’!

At another place in their paper they refer to this particular
matter somewhat more accurately as follows:

“Posner and Gies have recrystallized protagon as often as ten times.
They found the phosphorus percentage of the crystals separating out
(0.93 per cent.) and of the substance in the mother liquor (1.02 per cent)
to be almost tdentical'—surely the most conclusive proof (!) of the homo-
geneous nature of the substance they were dealing with. The significance
of this result does not seem to have been recognized by the authors.' Accord-
ing to them even this sample contained considerable quantities of pseudo-
cerebrin as evidenced by the effect of their method of fractionation.”’

‘The significance of this result,” I am happy to state, was
duly ‘‘recognized”’ by us? when we came to the only unbiased con-
clusion the results could possibly justify, viz., that ‘‘the crystals
separating out (0.93 percent P)”’ and ‘‘thesubstance in the mother
liquor (1.02 per cent P)” were not 1dentical—that, even after a
tenth recrystallization under the conditions maintained by us,
protagon lost to the mother liquor, material which was chemically
unlike that crystallized from it. This conclusion was enforced
not only by a comparison of the above quoted figures for phos-
phorus contents, but also by the further observation that the
products were unlike in appearance and consistence.

The suggestion that ‘‘the significance of this result does not
seem to have been recognized” by Posner and me—a result which
Lochhead and Cramer think deserves more emphasis than any
of their own data, because ‘“‘even more conclusive’ than any
of theirs in showing that protagon is not a mixture—leads me to
believe that the following remarks, made by us? in this connec-
tion, were overlooked by Lochhead and Cramer:

1 Ttalics are mine.
2 Posner and Gies: This Journal, i, p. 108, 1905.
3 Posner and Gies: loc. cit., p. 107.

344 Protagon

“The analyzed protagon samples (of the last product) were purified
and dried as usual. Only necessary quantities were removed for this pur-
pose. So much depended on the exactness of the results for phosphorus
content of the protagon, especially of the final product, that our colleague
Mr. W. N. Berg, was invited to check the results. Mr. Berg was told
nothing beforehand about the characters of the products he was asked to
analyze. His results agreed perfectly with the writer’s. The results
recorded above for proportions of phosphorus in the protagons are those
obtained by Mr. Berg. We are indebted to him for his cordial codépera-

tion.”’

My own result for the proportion of phosphorus in the final
protagon product was 0.92 per cent; for that in the substance of
the filtrate it was 1.04 per cent. These results were as closely
in agreement with Dr. Berg’s as duplicate results obtained by
either of us, or any one else, doubtless would have been. Instead
of giving my own results only, or the averages of Dr. Berg’s re-
sults and mine, we presented only Dr. Berg’s, for the reason that
the difference between his two figures in this particular connec-
tion was somewhat less than between mine, and therefore slightly
less favorable to the view to which our previous results consist-
ently led us.

The agreement between Dr. Berg’s independent results and
ours emphasizes strongly the fact that the difference between the
quoted 0.93 per cent for the crystallized protagon, and the 1.02
per cent for the mass of solid substance in the mother liquor,
was a real difference and not merely the appearance of one created
by defects of technic that would make such a mathematical dif-
ference of no chemical significance.1 This was so obvious, we
thought, that we did not dwell upon the fact, nor did we imagine
it would be necessary todo so. The facts alluded to below show
further how much Lochhead and Cramer ignored in their jump
to the conclusion quoted above that is so wofully unappreciative
of the meaning of the results of ours to which they referred.

We found that, on subjecting. the specified protagon to an
eleventh recrystallization after ordinary drying? (by the fractionation
process from alcoholic solutions that were purposely made more

‘ Lochhead and Cramer failed to remark that by recrystallization the
phosphorus content of our product was reduced from 1.22 per cent to
0.93 per cent.

2 Jn vacuo over concentrated sulfuric acid at room temperature.

Wiiliam J. Gies 345

concentrated than those used for the first ten recrystallizations),
the percentage amount of phosphorus in the first fraction of the
protagon that had thus been recrystallized an eleventh time, but
which was filtered off at 28° C. (instead of at o° C. as theretofore),
was 0.95 per cent, whereas that of the second fraction from the
resultant filtrate (filtered off at o° C.) was 1.22 per cent, and
that of the solid substance in the final filtrate from these two
fractions was 1.68 percent (see p. 346 for asummary of these and
related facts). On the assumption that the solid matter in the
above mentioned ‘‘final filtrate’? was composed solely of unpre-
cipitated protagon' (0.93 per cent P) and material containing
about 1.68 per cent of phosphorus, it is obvious that, if the
mixture consisted of seven parts of the protagon (0.93 per cent
P) and one part (12.5 per cent) of substance having 1.68 per
cent of phosphorus, the proportion of phosphorus in the mix-
ture would have been 1.02 per cent, the figure obtained for it
by Dr. Berg.? If Lochhead and Cramer attach no importance to
differences between analytic figures which show such evident
disparity between two products as that just pointed out, I am
obliged to insist, nevertheless, that due attention be paid to the
real significance of our results that they would ignore and which,
seemingly, in their ardor to establish what at the outset they
stated they intended to bring forward, they would make appear
to be just the opposite of what it really was.

Lochhead and Cramer have shown singular partiality to the
only pair of results of the very many in our paper that could
possibly bear any semblance to evidence in favor of their view
that protagon is a chemical entity. They ignored dozens of more

1 Protagon cannot be completely precipitated from its solution in 85
per cent alcohol by refrigeration at 0° C. Posner and Gies: Loc. cit., p. 93.

? My own figures suggested such a mixture of five parts of protagon and
one of the same material with specially high phosphorus content.

3 After the ninth recrystallization of the same protagon, in a reduced
volume of solvent, the percentage of phosphorus in the filtrate was 0.96
per cent. No determination of phosphorus was made in the correspond-
ing product that had been obtained in crystalline form. The physical
qualities of the two products were different enough, however, to make it
evident that if the phosphorus contents were nearly the same the mixtures
were quite distinct. The phosphorus content in the filtrate material
after the sixth recrystallization was 1.42 per cent; that of the protagon,
even after the first recrystallization, was not above 1.22 per cent.

346 Protagon

significant results of our fractionation experiments, the methods
of which were, in effect, merely fractional recrystallizations of dry
protagon products by classical methods used for their preparation.
The protagon that had been recrystallized a tenth time and dried,
whose phosphorus content was 0.93 per cent, and which may
certainly be regarded as having been as pure as any protagon
ever prepared (if the Gamgee and Blankenhorn method is accept-
able for the purpose, and no one denies that), yielded the follow-
ing results on simple fractional recrystallization from 85 per cent
alcohol, in solutions somewhat more concentrated than those
employed for the first ten recrystallizations. (Posner and Gies:
Loc. cit., p. 109, table xiv.)

FRACTIONATION PRODUCTS OBTAINED AT VARIOUS TEMPERATURES BY
POSNER AND GIES FROM PROTAGON PREPARED BY THE GAMGEE AND
BLANKENHORN PROCESS AND RECRYSTALLIZED fen TIMES.

Weight in Grams. Percentage of

Fractional Product. —— a Phosphorus.
Separate. Combined.
I. Protagon fractions: 28° C0: c. 2s°c. oC
A—lI: Productiat 28° C....| 2.05 | 0.95
a ns ok We Ose 1.69 | 1.22
(peewee é oer Cee. 5, 1 UO OF. 0.89
4. = an an Cee 1.12) ihe
C—5. = ENS 32 | OS49 0.74
6. bs Boe, xsi 0.59 | (6291 0.98
II. Insoluble portion (residue)... | §.39 0.74
III. Solid matter in the corre-
spon ene filtrates from
the protagon products at
OFC:
- file Ta ache ens ean 0.60. 1.68
SS PA. 5 MO OE eat 0.31
a OC ean Sr 038 | 1.19 |} a
IV. Combined ether-washings of I.
AnG(UE se os coke oes | ae oe 0.47
Substance recovered.............. rear layer 2 a

SHpstance takene yc so nae | 16.20 0.93

*The total weights of protagon products at the two temperatures were: 28° C.—3.51
grams; 0° C.-—3.40 grams.
+ The residues B and C were combined for analysis because of the small quantity of each.

Asample of protagon prepared by Cramer’s coagulation method,
which, according to Cramer, yields typical protagon, was sub-
jected by us to the same simple fractional recrystallization

William J. Gies 347

process that was used for purification purposes, with the follow-
ing results (p. 96, table viii):

FRACTIONAL PRODUCTS OBTAINED FROM CRAMER’S PROTAGON (SHEEP
BE AENS) BY THE METHOD OF LESEM AND GIES.

Fractional Product. labcandeie saieke _Blementary Composition.
Separate pone | Bales, ee | Sulfur.
Il. Protagon fractions: | | 13.5 |
ieebirstproe@uct.: 22... +24) o-9 | 1.29 | 2.08 | 0.68
2./ Second product: jo20:.....2-. 3.2 || 1.02 _ 0.60
3. Third product...........| 2.3 | 1.00 | 1.90 |
Ase OUTtLOEpPLOGUCts ees -4)12:- 2-1 0.68 | 0.54
Il. Insoluble protagon (final resi-
“(CHEN a ee aera a7 | 3.2 || 0.39 | 1.96 | 0.22
IIL. Solids in the filtrates from, and |
the washings of, the protagon) |
products: 3.15)
He PET SE PRAGUCH eG oe) 2. - 1.31 | 1.92 |
PEESeCONGd Product. —- os 4ssa-- | 0.79 | lesa sea
3. Third product............ | 0.56 | 1-28 |
AP Pourthi product: -. 2. i...4.-. 0.49 | | 1.02
| |
SHAISEANCE TECGVETEG, «2 es wl sw v's 19.85 |
Original protagon: «5... ss) «21 sia = = 20.00 10% | 22207 Ose

Until our methods of fractional recrystallization are shown
to be inadequate, the foregoing results cannot be superseded by
data that are less significant. Why is it that those who have so
confidently supported the idea that protagon is a definite chem-
ical substance never deliberately subjected their products to me-
chanical fractionation and the resultant fractions to compar-
ative analysis? What more elementary or decisive test could be
applied to protagon products to prove the uniformity of their
characters? What is the matter with the fractionation process
that it should be so studiously shunned by all the “ protagon-
ist5r—

I desire to make this communication as brief as possible and
will single out for immediate comment only those remaining
points in Lochhead and Cramer’s paper which cannot satisfac-
torily be reserved for a future discussion in connection with the

348 Protagon

results of work related to protagon that has been in further prog-
ress in this laboratory.

The third of the three sentences comprising the summary of
conclusions at the end of the paper by Lochhead and Cramer
reads as follows:

“The view that protagon is a mixture of substances differing in their
solubility and in their phosphorus contents is not compatible with these
results.and cannot be accepted until the substances constituting the mix-
ture have been isolated.”

One such substance, 7. e., phrenosin (pseudocerebrin, cerebron,
cerebrin) was isolated by Gamgee, Worner and Thierfelder, Koch,
Lesem and Gies, and Posner and Gies. Koch has isolated others.
Thudichum claimed to have identified phrenosin and others in
protagon. See p. 354.

Lochhead and Cramer presented data showing that Gamgee’s
pseudocerebrin (1880) and Worner’s cerebron (1900) were practi-
cally identical. Why did they not go further and give facts
showing that both these products were practically the same as
Thudichum’s phrenosin (1874'), that was first described before
Gamgee discovered the difficulty of separating ‘‘pseudocerebrin”
from protagon and which separation does not appear to have
been made when his classical products were prepared.’

Lochhead and Cramer stated that ‘‘according to Gamgee,
pseudocerebrin exists together with protagon and is extracted
together with protagon from the brain. Being less soluble in
alcohol than protagon, Gamgee was able to separate it from prota-
gon by simple recrystallization. If his view is correct it is possible
to remove pseudocerebrin entirely from the phosphorized matter,
which can be obtained as a definite chemical compound of a
constant phosphorus percentage:’’ Did Gamgee ever prepare a
‘“‘pseudocerebrin”’ product that was freefrom phosphorus? Did
Gamgee ever prepare protagon that was free from ‘‘pseudocere-
brin?’’ If so, where may his data-on these points befound? In
this connection I invite from Lochhead and Cramer answers to
the following, which was published by us two years ago in a dis-
cussion of Gamgee’s method of preparing pseudocerebrin, of the

*Gies: This Journal, ii, p. 159, 1906.

? Which on repeated recrystallization retained the same proportion of
phosphorus.

William J. Gies 349

significance of his product, etc., but of which no correction has
yet been made that we know of (Posner and Gies: Loc. cit., p. 69):

‘““Gamgee has apparently made no further contribution to the subject.
It is difficult to understand exactly, from the very brief statement of
Gamgee’s that is quoted above, what is meant by the direction to subject
‘ampure’ protagon to ‘repeated’ crystallization, for it was shown by
Gamgee and Blankenhorn that the composition of ox-brain protagon,
which had been recrystallized only twice from 85 per cent. alcohol,! was
unchanged by a third recrystallization, and their results indicated in a
general way that protagon from dog-brains that was recrystallized only
once Was as ‘pure’ as protagon from horse-brains that was recrystallized
jour times.?, Gamgee and Blankenhorn did not take the subsequently-
stated precaution to recrystallize ‘repeatedly,’ unless only one to ‘four
or five times’ at most is considered sufficient to meet the requirements of
the direction given. Isit not highly probable, therefore, quite certain in
fact, that their protagons contained pseudocerebrin? Are not the remarks
of Gamgee in 1880 indicative of a logical departure from the position held
in 1879, and of a feeling that protagon, as previously prepared and de-
scribed, contained the new substance subsequently discovered by him,
1. e., pseudocerebrin?

“How can anyone know definitely when pseudocerebrin has been com-
pletely removed from protagon in the process of recrystallizing from 8o
per cent alcohol at 45° C.? Gamgee has said nothing that will answer
the question. He referred, it is true, to the nodular masses in which
pseudocerebrin crystallizes, but ‘pure’ protagon, that has been recrystal-
lized ten times, forms such masses if the solution from which it crystallizes
is not too dilute and the crystallization is not too slow. Is it not almost
certain, from the various facts recited above, that every protagon that
was described before Cramer drew attention to pseudocerebrin, contained
the latter substance? Who of the ‘protagonists,’ except Gamgee, knew
about pseudocerebrin, and who succeeded in removing it from his prota-
gons either by design or accident? If these questions be answered in the
negative, and we think they must be, what has protagon been if not always
a mechanical mixture of substances?

‘““Was Cramer’s protagon assuredly free from pseudocerebrin? If
pseudocerebrin was not contained in it, how did Cramer remove it? His
methods of preparation and purification warrant the conclusion, however,

1 The different concentrations of alcohol (80 per cent and 85 per cent)
could not account for the discrepancies noted.

2 ‘*The amount of phosphorus in specimens of protagon which had been
crystallized from alcohol jour or five times was not smaller than that
present in protagon which had only once been crystallized, though a
thorough treatment with ether preceded each recrystallization.’’ (Gamgee:
Text-Book, p. 427.)

350 Protagon

that pseudocerebrin was present in each of his protagon preparations.
We have separated a pseudocerebrin-like substance from every sample
of protagon made by us by the Cramer method. We believe that the
higher temperature prescribed in Cramer's method tends to favor the
presence of even a larger proportion of pseudocerebrin in the protagon
obtained by means of the new process than that in the protagon prepared
by the older and more cautious methods.

“What is the effect of the ‘repeated recrystallization’ (necessary,
according to Gamgee, to remove pseudocerebrin from protagon), on the
protagon itself, or, more correctly, on the phosphorized matter associated
with the pseudocerebrin? What is the composition of such associated
matter, or of ‘pure’ protagon, if we may call it such? Gamgee gave no
information on these points. No one else, we believe, is able to supply
a definite answer.

“That Gamgee's further studies, unfinished though they were (1880),
modified greatly his original views was shown also by his remarks on
the great stability of protagon in boiling alcohol' and ether, and on the
presence in the brain of a free, preformed ‘non-phosphorized cerebrin.’
His earlier, opposite views have been quoted repeatedly by successive
investigators, all of whom, except Thudichum, have remained unac-
quainted, until recently, with the remarks in the text-book. It is to be
regretted that Gamgee’s ultimate conclusions on the matters referred to
above were not published where they would have been sure to receive the
attention they merited. If they had been promptly published in the
journal, for example, in which the paper by Gamgee and Blankenhorn
appeared, we believe the protagon question would have been settled
before Parcus and succeeding authors could have made the obvious mis-
takes that have been recorded in abundance since.’’?

‘On p. 426 of his Text-Book Gamgee stated: ‘At higher tempera-
tures than 55° C. alcohol appears to decompose protagon’”’ On p. 429
the following may be found: ‘Pure protagon is remarkably rebellious
to the action of even boiling alcohol, though that action be continued
for hours."’ P. 440 records the following: ‘‘By boiling with alcohol for
many hours protagon appearsto be decomposed.”’ In his first protagon
paper he wrote: ‘‘Es verdient ferner hier noch erwaéhnt zu werden,
dass wir bei der Darstellung sowohl, als auch beim Umkrystallisiren von
Protagon die Lésungen nie tiber 45° C. erwdrmten, da wir tiberzeugt sind,
dass Liebreich mit Recht annimmt, dass sich dasselbe in alkoholischer
Lésung tiber 50° C. erhitzt, zu zersetzen beginnt’’ (Gamgee and Blanken-
horn: Zeitschrift fur phystologische Chemie, iii, p. 280, 1879).

* The paper by Gamgee and Blankenhorn was published in 1879 (Zezt-
schr. f. physiol. Chem., iii, p. 260). It is stated at the head of the paper
that it was received for publication May ro. Nothing was said in that
paper about pseudocerebrin, of which, a year later, however, Gamgee
stated (Text Book, i, p. 441, 1880) ‘‘a very considerable quantity was

William J. Gies 351

The following, which is quoted from Lochhead and Cramer’s
paper (p. 353), was evidently written to show that the method of
preparing the protagon referred to yielded typical protagon and
that fractional recrystallization had no effect on the composition
of the product:

“Protagon E.—A sample of protagon was prepared in the manner
described for samples A and B. After one recrystallization it contained
1.22 per cent P. It was recrystallized once more out of methylalcohol
in the following manner; a small quantity of the solvent, which was not
sufficient to dissolve all the protagon, was added first and decanted after
having kept the solution boiling for one minute. The same process was
repeated with another small quantity of methylalcohol. The crystals
separating out in these two fractions gave on analysis:

PSG LAC IO Ue 2 heen eee ke oo win ee E04. per centr:
SSO K0 a Se Sap es ie ee wee eee | TOs Pemeenin kw

In short, a substance having a composition of 1.22 per cent of P,
on recrystallization yielded two crystalline fractions, one of which
contained 1.14 percent of Pand the other, 1.05 percent of P. How

always extracted with protagon and other phosphorized mattersfrom brain
by alcohol at 45°C.” When did Gamgee first learn about pseudocerebrin ?
Until recently the only statement by Gamgee about it appeared in his
text book and few, as Thierfelder lately made evident, knew anything
about its presence there (Gies: This Journal, ii, p. 167, 1906). Three
years ago Thierfelder (Zezttschr. 7. physiol. Chem., xliii, p. 22, 1904) pub-
lished the substance of a personal communication to him by Gamgee to
the effect that his (Thierfelder’s) then recently described cerebron was
merely rediscovered pseudocerebrin; that the composition of the former
as given in Gamgee’s text book was the same as that of the latter pub-
lished by Worner and Thierfelder (Zeztschr. 7. physiol. Chem., xxx, p. 542,
1900); and, furthermore, that Gamgee by reference to his laboratory
journal, after the appearance of W6rner and Thierfelder’s paper, found,
under date of July 11, 1879, the note that pseudocerebrin melted at 210°
C. (The melting point of cerebron was found by Worner and Thierfelder to
be 209—-212° C.) Lochhead and Cramer have repeated the substance of
this in the form of a letter from Gamgee.

These facts explain, it seems to me, why Gamgee and Blankenhorn did
not mention pseudocerebrin in their paper. Pseudocerebrin was dis-
covered by Gamgee ajter the Gamgee and Blankenhorn products had been
isolated and described, and when it was too late to effect removal of
pseudocerebrin from them. And with pseudocerebrin present, Gamgee
and Blankenhorn failed to affect the percentage content of phosphorus
in their protagon products by recrystallization (!). In the preliminary
work begun as early as 1877, Gamgee, ‘‘assisted by Mr. Leopold Larmuth

352 Protagon

much material remained in the filtrates and what was the phos-
phorus content of each mass? Did any of the original protagon
remain undissolved in this treatment? If so, what was its con-
tentof phosphorus? The foregoing data appear to be the only ones
bearing on fractional reprecipitation that were obtained by Loch-
head and Cramer. What better evidence of the heterogeneity
of the protagon product under consideration could be required
to prove it a mixture? The results in this particular connection
are in complete accord with those for protagons (made by the
Gamgee and Blankenhorn process and by Cramer’s coagulation
method) obtained in this laboratory every time the fractional
recrystallization process was applied to them.

A further quotation from the Lochhead-Cramer paper is
appended (p. 354; also, p. 355):

“ Protagon F.—The dry powder was extracted with 500 cc. of boiling
chloroform. After cooling, 1500 cc. of ether were added. A white pre-
cipitate formed, which was analyzed without further purification, as only

found that the amount of phosphorus in specimens of prota-
gon which had been crystallized from alcohol four or five times, was not
smaller than that present in protagon which had only once been crystal-
lized, though a thorough treatment with ether preceded each recrystalli-
zation."’ (Text book, i, p. 427, 1880.) Of course pseudocerebrin was
present in these products also.

Why is it that Gamgee never candidly revised the statements in his
paper with Blankenhorn that his discovery of ‘“‘pseudocerebrin’’ made
necessary in the interest of the truth about protagon? Even his text
book treatment of the subject is vague and contradictory in places. How
can Lochhead and Cramer expect to use the data pertaining to pseudo-
cerebrin as a prop for their hypothesis that the protagons of Liebreich and
of Gamgee and Blankenhorn (the classical products) were something uni-
form chemically, without focusing attention upon the real facts in the
case; without emphasizing the apparent truth of the situation to which
Thudichum referred when he declared that Gamgee’s change of position
in 1880 from that assumed in 1879 was an outcome of Thudichum’s frac-
tionation studies of protagon in 1879 and that ‘‘Gamgee himself, by using
my [Thudichum’s] method for the preparation of phrenosin, extracted
this principle from protagon and to hide its identity called it pseudocere-
brin?”’ (Thudichum: The progress of medical chemistry, p. 191, 1896.) I
do not know that these statements by Thudichum have ever been refuted.
I have already endeavored to show the fact that phrenosin, pseudocere-
brin and cerebron were essentially the samesubstance. (Gies:This Journal.
i, p. 159, 1900.)

Mie

William J. Gies 353

small quantities could be obtained in this way. It contained 1.18 per
Cents eo who wanxht Sample F, prepared by precipitation with ether, is
of special interest, as, according to Gies, protagon contains a considerable
quantity of an ether soluble substance. poor in phosphorus. If this is
correct, the method employed for Sample F should yield a substance
richer in phosphorus than the other samples of protagon. This was not
found to be the case.”’

Sample F was found to contain 1.18 per cent of P, which zs rela-
tively high for protagon, if the tabulated data presented by
Lochhead and Cramer from the papers of the leading investiga-
tors is to be the basis for judgment. But how do Lochhead and
Cramer know that “‘Sample F”’ was protagon? Lochhead and
Cramer speak of this matter, however, so far as it pertains to
our observations, as if their own results were the only evidence
besides ours that could be brought forward, apparently being
quite oblivious of the fact that Kossel and Freytag obtained from
an ether extract, a protagon having a rather high content of phos-
phorus, 7. e., 1.35 per cent. Furthermore, the above quotation
disregards the fact that our remarks were especially directed to
“freshly prectpitated (hydrated)’’ protagon, a distinction which
was not noticed by Lochhead and Cramer. We stated (p. 111)
that ‘“‘long continued extraction of pure dry protagon with ether
in a Soxhlet apparatus failed to affect its composition—another
matter totally ignored by Lochhead and Cramer. Besides, we
said nothing at all about protagon solubilities in ether-chloroform
solutions. The solvent action of ether on one hand and of ether-
chloroform on the other may be very different, both in kind and
in degree.

Lochhead and Cramer present a summary of figures apparently
intended to show an impressive agreement between their lowest
data and those obtained by previous workers, for the proportion
of phosphorus in protagon. With the exception of the figure
for Liebreich’s protagon, all the others given by them range
between 1.13 per cent (Ruppel) and 0.93 per cent (Posner and
Gies). The quoted figure for Liebreich’s is 1.23 per cent, 7. é., an
average of his two results (1.1 and 1.1 per cent) for products from

1 Why was it called protagon? Because it was white, had been obtained
from brain and contained 1.18 per cent of phosphorus?

354 Protagon

domestic animals and the one (1.5 per cent)! for protagon from
human brain. The quoted figure for Kossel and Freytag’s prod-
ucts is 0.97 per cent, no account being taken of their protagon
from an ether extract having 1.35 per cent. Relatively high
results obtained by Lesem and Gies, e.g.,1.23 per cent, for prod-
ucts very carefully prepared by the Gamgee and Blankenhorn
method and particularly low figures for similar products, e. g.,
0.57 per cent, were not referred to by Lochhead and Cramer.
Zuelzer’s? low figure (0.72 per cent) as well as the low one obtained
by Gamgee and Blankenhorn (0.72 per cent), after treatment
with hot ether, is also not mentioned, and Ulpiani and Lelli’s®
high result (1.31 per cent) is ignored. The figure quoted by
Lochhead and Cramer for Posner and Gies’ products is 0.93
per cent, which fails to indicate the much higher and also the
decidedly lower results obtained by us for many typical protagon
products. Thudichum’s important data were entirely ignored.
It seems to me the reader is never particularly interested in only
one side of such questions, but invariably appreciates a presen-
tation of both. When all the figures for phosphorus content of
protagon products that have been obtained by the classical
methods are assembled, they at once suggest the question,
What is the percentage amount of phosphorus in protagon?
Who can answer this question?

Lochhead and Cramer say that ‘‘in order to definitely deprive
a substance like protagon, possessing charactertstic physical and
chemical properties,‘ of its existence as a chemical individual, it
is necessary to isolate and identify the various substances con-
stituting the mixture—a task which, in view of the differences
in the physical and chemical properties of these substances should
be a comparatively easy one.’ Would not the “isolation and

‘Did Liebreich have two different protagons or was the high figure
(1.5 per cent) expressive chiefly of analytic error? In eitherevent what
does the amount, 1.23 per cent, as given by Lochhead and Cramer, stand
for?

* Zuelzer: Zettschr. f. physiol. Chem., xxvii, p. 255, 1899.

? Ulpiani and Lelli: Gazetta chimica italiana, xxxii, Pp. 466, 1902.

‘Italics are mine.

* Who, since Thudichum first successfully ‘pointed out its heterogeneity,
has established the properties of protagon? Has protagon been deprived
of anything that ever justly belonged to it?

William J. Gies 355

identification”’ of only one of them settle the point? Phrenosin
(pseudocerebrin, cerebron, cerebrin) has already been isolated
from protagon and identified, has it not? As to the ‘‘compara-
tive ease’ of the task alluded to, we fear Lochhead and Cramer
have passed very lightly over facts too well known to require
discussion.?

Lochhead and Cramer suggested that “‘a final and decisive
proof of the homogeneous nature of a substance even of a less
complex nature than protagon, cannot be brought about until
the substance in question can be synthesized. In default of
this,’ they added “‘the constancy of its physical and chemical
properties? is generally considered to be sufficiently trustworthy
evidence.”

In view of the great uncertainty that has prevailed for many
years regarding the “‘ physical and chemical properties” of prota-
gon, it would surely help greatly to a speedy and correct solution
of the protagon problem if Lochhead and Cramer were to pub-
lish a detailed statement of the physical and chemical properties
which they appear to think are so constant and which, by im-
plication, must be so well known—the properties, for example,
that were so definite for and characteristic of their ‘‘Sample F,’’
let us say, to which I referred on p. 352. Has the “ homoprota-
gon’”’ accessory to the main hypothesis been abandoned? If not,
what are the properties of homoprotagons that distinguish them
from the protagon? Cramer did not return to this subject in his
last paper.

The work on protagon that Cramer has described thus far has
not, it seems to me, been directed at the heart of the question,
regarding protagon, that he appears to be desirous of solving.
That question is not: Can products closely resembling the pro-

ce

1“* There can be no doubt that protagon is accompanied in the brain by
large quantities of a body or bodies which may provisionally be conveni-
ently classed under the term of cerebrin, and likewise by smaller quanti-
ties of other phosphorized bodies, containing a percentage of phosphorus
very close to that found in lecithin, yielding the same products of decom-
position, and the separation of which presents extraordinary difficulties.”
(Gamgee: Text Book of Physiological Chemistry, i, p. 429, 1880). See
also Halliburton: The Oliver-Sharpey Lectures, 1907; British Medical
Journal, May 4 and 11, 1907.

2Italics are mine.

356 Protagon

tagon of Liebreich, and of Gamgee and Blankenhorn, be prepared
by new methods, but rather, are the protagon products prepared
by the Liebreich and the Gamgee and Blankenhorn methods or
any other, samples of a definite chemical individual? The prob-
lem is not, it seems to me, merely the production of something
new to look like the old and to which the familiar name may be
applied but the establishment, clearly and precisely, of the quali-
ties of the old classical protagon products of debatable characters;
ij they are in fact samples of a substance and not of mixtures.

The remarks in the appended quotation, published by us two
years ago, are quite as applicable now as we thought them
then:!

‘*We have already indicated the probability that all preparations of
protagon that were made previous to Cramer’s work contained phrenosin.
We obtained phrenosin-like products from every sample of protagon made
by us by Cramer’s method, that we examined. The fact expressed in the
last sentence of the quotation (below)? from Cramer’s remarks on the
results obtained by Lesem and Gies must also be explained away satis-
factorily before it will be possible for any one to insist that protagon pre-
pared by the classical methods in general use before the appearance of
Cramer's paper was not always a mixture. Cramer prepared protagon
by anew method but did not investigate the question whether prota-
gon was a mixture. He did not subject his new protagon to a fractiona-

' Posner and Gies: Loc. cit., p. 76.

2‘ By treating protagon, the analysis of which gave figures correspond-
ing to other authors’, with dilute alcohol for several times, Lesem and Gies
got substances which differed widely in their percentage of phosphorus
and in their solubility in alcohol. . . . . . The decreasing percentage of
phosphorus in the ‘precipitated protagon’ corresponds entirely with the
statement of Gamgee regarding pseudocerebrin. We may therefore
assume that the ‘protagon’ of Lesem and Gies is a mixture of protagon and
pseudocerebrin. Indeed, the precipitate of the first extract, where the
protagon should be purest, has a percentage of phosphorus equal to that
of protagon. It remains to explain how it is that this mixture of a phos-
phorus-free with a phosphorus-holding substance should on analysis give
a percentage of phosphorus corresponding with that of protagon, when
one would, of course, expect a lower figure. This deficit must have been
made up by a substance richer in phosphorus than protagon itself. This
substance is indeed present in Lesem and Gies’ protagon, for in the filtrate
from the freshly precipitated protagon of the first extract and in the

alcohol-ether washings we find a substance containing a higher percentage
of phosphorus.”

William J. Gies 357

tion process and consequently left untouched the fundamental point
alluded to in the last sentence in the remarks of his we have quoted.'

‘Suppose it be admitted that Lesem and Gies, and all other protagon
investigators to the time of Cramer’s work, analyzed products containing
phrenosin (pseudocerebrin, cerebron), and we think as indicated before
that this must be conceded, would it not also be admitted at the same
time that the idea contended for so long unavailingly by Thudichum, and
recently by Lesem and Gies, that protagon was a mixture, was correct?
But what was the nature of the phrenosin-free part of the mixture that
has always been called protagon? If now we decide to apply, as Cramer
does after Gamgee, the name protagon to the phrenosin-free part of the
old protagon, do we not merely shift our point of view and give to a part
the name that has heretofore been applied to the whole?

‘Did any one ever succeed in separating and identifying this hypo-
thetical protagon—the part of protagon that would remain if phrenosin
were removed from it? If we admit Cramer’s conclusion that Lesem and
Gies were dealing with mixtures of such hypothetical protagons and phren-
osin, then surely we are also obliged to admit that it was this newly con-
ceived protagon in their mixtures that yielded the products of relatively
high and distinctly variable phosphorus content that were alluded to in
the quotation above.? But if Cramer’s view were correct would not this
hypothetical protagon also have to be conceived as a mixture? It stands
to reason also that after removal of the phrenosin from the protagon
mixture, the phosphorus content of the remainder would be much higher
than any ever before recognized as typical of protagons.”’

I have lately completed, with Mr. Matthew Steel, an inquiry
into the general nature of Ulpiani and Lelli’s ‘‘ paranucleoprota-

7

gon.” Our results are about to be published. The non-protein
part of paranucleoprotagon behaves much like ordinary prota-
gon when subjected to mechanical fractionation, and it seems
to be a mixture of products similar to those in protagon.

I hope to publish in the near future, also, the results of a study
of the effects of desiccation on the fractional recrystallization of

1“Dr. Cramer informed us that the experimental part of his work was
completed before the publication of the paper by Lesem and Gies, and that
it was only lack of proper opportunity to take up the work again that
prevented him from going into this matter.”

[The only results of such fractional recrystallization that Cramer has
published since the above was first written, two years ago, that I know of,
are referred to on p. 351 of this paper.]

2 The products of high phosphorus content could not have been im-
purities, because the methods of Lesem and Gies, and their analytic
results, furnish conclusive evidence against that possibility.”

358 Protagon

protagon. That the solubility of protagon in 85 per cent alcohol
at 45° C., and in ether, was decreased by drying was made very
clear in the work described in the paper I published with Posner.
That preliminary conversion of protagon to the anhydrous con-
dition favored the fractionation results obtained by us was quite

obvious.

IS THE CONDUCTION OF THE NERVE IMPULSE A
CHEMICAL OR A PHYSICAL PROCESS?

By S. S. MAXWELL.

(From the Rudolph Spreckels Physiological Laboratory of the University of
California.)

(Received for publication, July 12, 1907.)
ip

The remarkable resistance of nerves to fatigue and the diffi-
culty of demonstrating satisfactorily the existence of chemical
changes as the result of stimulation of the nerve are too well
known to require restatement or discussion. The experiments
of Bowditch, Wedenski and others, in which nerves were stimu-
lated for many hours without indications of fatigue, are described
in the common text-books. The failure to detect the liberation
of heat from a stimulated nerve is equally well known. From
these experiments the conclusion has very frequently been drawn
that the transmission of the nerve impulse is a purely physical,
molecular, and not a chemical process. In connection with
these facts and the much investigated electro-motive phenomena
of nerve a mountain of physical hypotheses as to the nature of
the nerve impulse has been built up.

On the other hand it has often been argued that all other “‘life
processes” with which we are familiar are associated with some
sort of chemical change and that it would not be safe to decide
upon purely negative evidence that the propagation of the nerve
impulse is a unique exception. It is possible to conceive that
fatigue does occur, but that recuperation takes place so rapidly
that the recovery can be completed in the short intervals between
the successive shocks of the Faradic current. The attempts to
secure positive evidence of chemical change in stimulated nerve
have taken in the main four forms:

1. The attempt to prove fatigue of the nerve.

Passing by the earlier experiments, the results of which were
wholly negative or the methods admittedly faulty, I merely

359

360 Conduction of Nerve Impulse

mention some of the more recent attempts. Gotch and Burch’
found that when two stimuli separated by a very brief interval
of time were applied to a frog’s sciatic, no electro-motive response
to the second shock was obtained if the interval was less than
about 0.006 of a second. They also noted that the duration of
this ‘‘critical interval” was very greatly influenced by tempera-
ture.

According to Garten? definite fatigue could be produced in the
olfactory nerve of the Pike by the use of induction shocks occur-
ring as frequently as every 0.27 of a second. In this case the
evidence of fatigue was the failure of the electro-motive response.

2. The attempt to demonstrate the presence of the products of
chemical change in stimulated nerve.

Again I pass by the older observations. Waller? has shown
that tetanization of a nerve for a few minutes increases its sub-
sequent electro-motive response to induction shocks, and that a
similar increased response can also be obtained by exposing the
nerve for a short time to very dilute carbon dioxide. From these
experiments he argues the probability of the production of carbon
dioxide in the stimulated nerve. The observation is extremely
interesting, but the argument that the cause of the increased
electro-motive effect is the same in both cases is certainly not
convincing.

Bethe* has shown that certain substances, fibril acids, in the
neurofibrillz, have their distribution altered by the passage of a
constant current through the nerve, in such way that the readily
staining fibril acids migrate away from the anodal region and
collect at the cathodal. He assumes provisionally that the con-
ditions of anelectrotonus and catelectrotonus are connected with
this movement, and that changes in the fibril acids, orin their
relation to the neuro-fibrillea, have to do with the propagation
ofthe nerveimpulse. In order, however, to secure the character-
istic difference between anode and cathode, as brought out by
staining, the current must be allowed to act for an exceedingly

*Gotch and Burch: Journ. of Physiol., xxiv, p. 410, 1889.

? Garten, quoted by Biedermann in Ergeb. d. Physiol., ii, 2, p. 129.

* Waller: Lectures on Physiology, First Series, London, 1897.
‘Bethe: Allgemeine Anatomie und Physiologie des Nervensystems, Leip-
zig, 1903.

S. S. Maxwell 361

long time compared to that in which stimulation and conduction
can occur. It would not be safe to assume that these changes
are the essential factors in nerve excitation and conduction.

3. The attempt to demonstrate a change of temperature in stimu-
lated nerve.

Helmholtz’ working with a thermopile could find no evidence
of evolution of heat in the frog’s sciatic when stimulated.

Heidenhain’ obtained similar results.

Schiff? believed that he could show with the thermopile the
production of heat in stimulated nerves of the cat and rabbit.

Rolleston,* using an exceedingly sensitive electrical resistance
thermometer, could find no change of temperature in nerve as
the result of stimulation.

4. The attempt to prove the necessity of free oxygen to the nerve.

According to Baeyer® a nerve surrounded by an atmosphere of
pure nitrogen or hydrogen loses its irritability in about five hours,
and becomes again excitable within a few minutes after the read-
mission of oxygen. He compared the effect of lack of oxygen
upon stimulated and unstimulated nerves and could find no well
marked difference between them. So far as one may draw con-
clusions from his experiments they seem to indicate that stimu-
lation and conduction in the nerve are not oxidation processes.

Lr,

Neither the experiments mentioned above nor other work along
similar lines have forso far given conclusive evidence as to the
chemical or non-chemical nature of the nerve impulse. In the
effect of temperature on reaction velocity van’t Hoff® and Arrhe-
nius? have pointed out a simple and reliable criterion for the dis-
crimination between chemical processes and those which are usu-
ally designated purely physical processes. They have shown that
a rise of 10° of temperature increases the velocity of a chemical

1 Helmholtz: Arch. f. Anat. u. Physiol., p. 158, 1848.

2 Heidenhain: Studien d. physiol. Inst. zu Breslau, iv, p. 250, 1868.
3 Schiff: Arch. de physiol. norm, et pathol., p. 157, 1869.

* Rolleston: Journ. of Physiol., ii, p. 208, 1890.

§ Baeyer: Zeitschr. f. allgem. Phystol., ii, p. 169, 1903.

®Van’t Hoff: Etudes de dynamique chimique, Amsterdam, 1884.

7 Arrhenius: Zettschr. f. physikal. Chem., iv, p. 226, 1899.

362 Conduction of Nerve Impulse

reaction to two or three times its original rate, while no known
physical process is accelerated in anything like so great an amount
by a similar rise of temperature, probably not more than 5 to 15
per cent at the most. This principle has been used by Loeb’ to
show that the artificial maturation of the eggs of Lottia is really
a chemical process. Mr. C. D. Snyder was also asked by Pro-
fessor Loeb to investigate the effect of temperature on the re-
action velocity of the tortoise heart. These experiments gave
a high temperature coefficient and indicated definitely that
chemical processes underlie the rhythmic activity of the heart.”
His observations were fully confirmed by Robertson using the
heart of a crustacean, Ceriodaphnia.’ The same criterion has
been used by Loeb‘ to determine whether the production of
artificial parthenogenesis by the action of hypertonic sea water
depends upon changes of a chemical or of a physical nature.
The high coefficient obtained showed conclusively that the
process isachemical one. Professor Loeb also asked Dr. Burnett
and myself to investigate the effect of temperature on the
velocity of the nerve impulse, since it would be possible by this
means definitely to decide whether the propagation is due to
chemical or to non-chemical processes. Dr. Burnett found it
advantageous in the beginning of his experiments to study the
effect of temperature on the latent period of striated muscle.
This part was published early in 1906.° Afterward on account
of the pressure of other work Dr. Burnett was unable to continue
the investigation, and, at Professor Loeb’s request, I have
carried it through with the results herein reported.

The fact that the velocity of the nerve impulse is affected by
temperature was observed by Helmholtz*and has been frequently
noted since, but so far as I know, no investigations have hereto-
fore been made with the direct purpose of determining the tem-
perature coefficient of nerve conduction. Nicolai? made a

‘Loeb: University of California Publications, Physiology, iii, p. 1, 1905.
? Snyder: Ibid., ii, p. 125, 1905.

* Robertson: Biol. Bull., x, p. 242.

* Loeb: Unwwersity of California Publications, Physiology, iii, p. 40, 1906.
* Burnett: This Journal, ii, p. 195, 1906.

* Helmholtz: Arch. f. Anat. u. Physiol., p. 358, 1850.

" Nicolai: Arch. f. d. ges. Physiol., 1xxxv, p. 65, 1901; Arch. f. Physiol.,

1905, supplement, p. 341.

— a

S. S. Maxwell 363

series of experiments on the speed of the nerve impulse in the
olfactory of the Pike, in which, among other things, he con-
sidered the effect of temperature. He used the ‘‘negative varia-
tion” photographically recorded as the indicator of the passage
of the impulse. His results show indeed that the velocity is
affected by the temperature, but the experiments were not speci-
fically planned for the determination of the temperature coeffi-
cient and for several reasons are inapplicable to that purpose.
The length of nerve employed was relatively small, from 4 to
23 millimeters, allowing room for large experimental errors. The
temperature changes were, except in one experiment, all made
in one direction; that is, the temperature was either raised for each
successive observation, or lowered for each, so that it was not
possible to know whether the observed change in rate was all
due to the change in temperature, or whether it was also affected
by other variables. The importance of this consideration I
shall discuss more fully in connection with my own experiments.
Von Miram! studied the effect of high temperatures on the con-
ductivity of the frog’s sciatic. His results are given in the form
of curves and averages... On account of the great individual
variations a statistical method is absolutely out of the question
in dealing with the effect of temperature on nerve conduction
and hence von Miram’s paper is unavailable for the purpose.
His table shows, for example, at 15° a minimum rate of 25 meters
_and a maximum rate of 38.7 meters a second, and at 30° a mini-
mum of 38.7 and a maximum of 49.5 meters. Since, according
to this, one nerve at 15° may have a higher rate than another at
25° it is evident that one cannot safely draw conclusions from
averages except from an impossibly large number of observa-
tions. Besides, the high rate of the impulse in the frog’s nerves
makes it necessary to deal with differences of thousandths of a
second—differences in many cases smaller than the probable
experimental error.

Snyder? has based on Nicolai’s and v. Miram’s experiments
calculations of the temperature coefficient of the velocity of the
nerve impulse, and concludes that ‘“‘a comparison of tempera-

1Von Miram: Arch. f. Physiol., p. 533, 1906.
Snyder: Arch. —. Phystol., p. 113, 1907.

304 Conduction of Nerve Impulse

ture coefficients leaves us no longer in doubt as to the nature of
nerve conduction.”’ The methods employed in calculating these
coefficients are not free from objection. He says, ‘‘I have used
only the average values of the determinations of these authors.”
At first glance it is hard to see how these averages were made.
Of the eight lines in the table founded on Nicolai’s tables,’ I
quote as specimens the first two and the last.

10R

ty to R, Re Ga (h Sas
9.25 3.45 13.7 5.65 4.1
9.25 4.82 lisioyr 8.55 3.6
25.0 3.45 22.2 5.65 Ite}

In the above table at rate 13.7 temperature 9.25 is obtained
by averaging two consecutive observations in the experiment of
November 11 in which the stimulus was the closing of a constant
current. The single items are

O° rate 1476
Orbe) 1238

Compared with this is rate 5.65 at temperature 3.45 obtained
by combining and averaging as follows:

3.7° rate 6.4 from above mentioned experiment of November
See 0k Gy Alfero ait,

Pe: oa ry from an experiment of November 30, using
ee ee kD break induction shock as stimulus.

Average 3.45° “ 5.65

In the second line the same 13.7 rate is used as before. The

rate 8.55 at temperature 4.82 is secured by the following selec-
tion:

5.0° rate 7.6) f aS ?

4.7° * 7.1f rom above experiment of November 11.

4.8° “ 10.2 \from an experiment of November 28, break
4.8° “ 9.3 f{ induction shock.

Average 4.82° “ 8.55

1 Arch. f. d. ges. Physiol., 1xxxv, pp. 82 and 83.

S. S. Maxwell 365

Finally the rate of 22.2 at temperature 25° was the last response
of the dying nerve used in the experiment of November 11, and
the only observation recorded at that temperature. It will thus
be seen that the tables in Snyder’s paper are obtained by select-
ing and comparing a few observations made upon the nerves of
different animals and in different degrees of freshness. One of
these nerves, moreover, is expressly stated by Nicolai to have
been injured by lying onice. By a different selection of data a
different conclusion might be reached.

ETT.

Our own experiments herein described were all made upon
the pedal nerves of the giant slug, Ariolimax columbianus. The
importance of this selection is seen in the great length of nerve
available and the slowness of the impulse, Jenkins and Carlson!
having found a mean rate of 440 millimeters a second, while it
is possible to use a nerve roo millimeters or more in length. If,
on the other hand, the frog’s sciatic were employed one would
expect to deal with magnitudes of the following order: Suppose
a distance of 30 millimeters between the near and far points of
stimulation and at a certain temperature a rate of 30 meters a
second. This would mean a difference in length of latent periods
for the near and the far points of 0.001 of a second. Now sup-
pose at another temperature the rate is 45 meters a second; the
difference of the latent periods becomes one one-thousand-five-
hundredth of a second. The value of the desired quotient will
thus depend upon a time relation of the difference between one
one-thousandth and one one-thousand-five-hundredth, or one
three-thousandth of a second; or, if one uses to measure the time
a tuning fork of 200 double vibrations a second, it depends on
the measurement of one-fifteenth of a single wave. It is at once
apparent that the probable error in such a calculation is vastly
larger than in observations on the pedal nerves of Ariolimax in
which one may obtain differences between the latent periods
for the near and the far points of from four to thirty-hundredths
of a second.

t Jenkins and Carlson: Amer. Journ. of Physiol., viii, p. 251, 1903.

366 Conduction of Nerve Impulse

The methods at first employed were essentially those of Jen-
kins and Carlson! except that the moist chamber in which the

animal was placed was so arranged that its temperature could ©

be regulated by surrounding it with warm or cool water as
desired. It was soon found that this plan was impracticable
because when the muscle is cooled the latent period is enor-
mously lengthened, the muscle contracts very slowly, and the total
shortening is greatly reduced. It thus becomes difficult at low
temperatures to secure curves with sharply defined beginnings.
When the higher temperatures were used another difficulty was
found in the increasing tendency to spontaneous rhythmical con-
tractions. For these reasons it was necessary to leave the mus-
cle outside the moist chamber and to cool the nerve only.
After many trials the method finally adopted was as fol-
lows: a moist chamber was constructed of heavy copper foil,
80 X 25 X 25 millimetersinsize. This was soldered by one end to
the inner wall of a metal case 120 X 105 X 40 millimeters inside
measurement. The outer case was provided with an inlet tube,
through which water could be circulated, and a stirrer to
equalize the temperature of the water. In the attached end of
the moist chamber and the wall of the outer case was cut a V-
shaped slit about romillimetersdeep. The edges of this slit and
the end of the case in its neighborhood were coated with wax to
prevent any possible contact of the nerve or muscle with the metal.
The use of copper foil for the moist chamber made it necessary
very carefully to avoid touching it with the nerve, which rested
on the electrodes only. It was found that a moist chamber of
hard rubber did not respond with sufficient promptness to
changes of temperature for the purposes of this investigation.
The electrodes consisted of platinum wires sealed into slender
glass tubes to insure perfect insulation. These tubes were then
fastened in place by means of sealing wax. One pair of elec-
trodes was placed near the V-shaped slit about halfway from
bottom to cover of the moist chamber, the other pair was near
the opposite end of the moist chamber and at a similar depth.
The pairs of electrodes in part of the experiments were 32 milli-
meters, in the others 64 millimeters apart. They were connected

* Jenkins and Carlson: Amer. Journ. of Physiol., viii, p. 251, 1903.

S. 5. Maxwell 367

with the poles of an inductorium by means of a Pohl’s commu-
tator without cross wires, so that the stimulating current had the
same direction in both pairs of electrodes.

The stimulus employed consisted of the break induction shock
from a Zimmermann’s DuBois-Reymond inductorium of 10,000
windings, with one Edison-Lalande cell, Type Z, in the primary
circuit. The circuit was broken automatically by a peg pro-
jecting from the rim of the kymograph drum. The time was
recorded by a tuning fork making roo double vibrations a second.

Two large percolators connected by a Y-tube to the inlet tube
of the outer case served as reservoirs of warm and cold water for
the regulation of the temperature. A thermometer held by a
rubber support was so placed that its bulb was near the middle
of the moist chamber at the level of the nerve.

The muscle used consisted of about two or three centimeters
of the posterior end of the foot. This was pinned firmly at its
anterior end to a block which carried a small pulley. A hook
was placed in the free, posterior end of the preparation, and a
thread passed from the hook around the pulley, and up to a
light lever arranged to record the contractions on the drum of
the kymograph.

The muscle of the slug is so readily thrown into strong and
lasting tonic contraction that special care had to be taken to get
it into proper position before beginning the dissection. It was
also found much better to secure the slug in its elongated position
than to attempt after dissection to stretch the nerve to the length
which it would have in the relaxed animal. In doing the latter
one frequently finds that the adhering blood and mucus are much
stronger than the nerve itself and, stretching irregularly, they
tend to break or injure the nerve. For these reasons in making
the preparations a second block was placed in front of the one
bearing the pulley. The animal was placed on the blocks with
the posterior end directed toward the pulley and allowed to creep
until only about two centimeters of its length remained on the
block bearing the pulley. Pins were now driven firmly through
the body into this block very near its edge. The pinning usually
acted as a stimulus to cause the animal to creep straight forward.
When maximal elongation had been obtained in this way the
anterior end was suddenly pinned down firmly, the body was

368 Conduction of Nerve Impulse

slit open and the viscera removed. A thread was tied around
the pedal nerves near the pedal ganglion (the two nerves taken
together made the best preparation), and their side branches
were cut off as far back as the first pins. The nerve was then
held up by the thread, care being taken to have it fully extended,
and the body was cut through at the line of contact of the two
blocks. The block with the preparation was now placed in front
of the V-shaped slit in the moist chamber. A pledget of cotton
wet with the animal’s blood was laid in the bottom of the slit,
the nerve was extended across the two pairs of electrodes and
held in position by looping the thread around a pin fastened at
the far end of the moist chamber. The upper side of the slit
was closed by another loose pledget of cotton wet with blood.
The nerve touched only this moist cotton and the two pairs
of electrodes. The moist chamber, which had been lined with
thoroughly wet filter paper, was now closed with a glass plate
covered with wet filter paper.

The velocity of the nerve impulse was determined by the
method of Helmholtz. Inits application, however, to the nerve
of the slug certain difficulties must be mentioned. In order that
the curves obtained by stimulation at the near and the far points
respectively shall rise with equal steepness from the horizontal
and so avoid error in determining the exact beginning of each
curve, it is usual to employ a maximal stimulus. With the
slug muscle, subject as it is to enormous changes of tonus, it
is impossible to say what constitutes a maximal stimulus. The
only practical thing to do is to usea strength of shock which
experience proves to give fairly equal curves. If the curves
do not come out well the tracing must simply be thrown away.
Much time and many tracings were lost on account of this
difficulty. I found it best to use fairly strong induction shocks,
the secondary coil at from 275 to 300. If, however, the cur-
rent employed was too strong, while the curves obtained looked
very satisfactory, it often happened, that when the tuning fork
record was counted the latent periods for the far and the near
points were practically equal. What occurred was probably
this, that on account of the high potential of the induced
current and the poor electric conductivity of the nerve the effect
was essentially the same as in the so-called unipolar stimulation;

———EO

S. S. Maxwell 369
the whole length of the nerve was excited simultaneously whether
the near or the far electrodes: were used. How easily the limit
of allowable strength may be overpassed is shown by one experi-
ment in which with the coil at 260 the latent periods for the two
points of stimulation were equal while with the coil at 275 the
results were apparently rational.

The nerve-muscle preparation of the slug is astonishingly sen-
sitive to mechanical stimulation. I found, for example, that
when the tuning fork was set in vibration before starting the
kymograph the slight jar communicated to the table excited a
contraction. For this reason it became necessary to support
the tuning fork in a manner entirely independent of the table
on which the moist chamber rested. For the same reason it was
necessary to discard the ordinary brake used to start and stop the
Ludwig kymograph and to employ a device which would work
more smoothly.

The tendency to spontaneous rhythmical contraction frequently
interfered with the experiments. The rate of these contrac-
tions was always comparatively slow, usually not far from one a
minute, and it was mostly possible to secure curves in the quiet
intervals of relaxation which gave apparently rational results.
Still it was impossible to be sure that the state of excitability of
such a muscle would not so vary from moment to moment as to
affect the length of the latent period and thus render unreliable
the estimation of the rate of the impulse. On this account most
preparations showing a persistent tendency to spontaneous con-
tractions were thrown away as useless.

The contraction and relaxation of the slug muscle are rela-
tively slow, especially the latter, and at times the muscle when
once stimulated tends to remain for a considerable period in a
state of strong tonic contraction. When this occurs it, of course,
interferes seriously with accurate velocity determinations. I
found, for example, that if the second curve was taken not more
than half a minute after the first the results appeared usually
to be reasonably consistent, but if a period of one or two minutes
intervened absurd or poorly agreeing figures were obtained.
It would seem that changes taking place in the muscle, impossible
to control, affect the duration of the latent period. This makes
it necessary to avoid all unnecessary delay between the two

370 Conduction of Nerve Impulse

curves. After considerable experience with this sort of diff-
culty, I found it best to throw away those tracings which I had
marked “relaxation slow” without taking the trouble to count
the vibrations. In order to shorten the time between the
curves for the near and the far points I put two pegs for the
circuit breaking key on the rim of the kymograph drum 180°
apart and adopted a rate of speed which allowed the essential
part of the tracings to be taken within one half a revolution.
This did away with the necessity of a readjustment of the drum
between the two steps of the experiment. It also had the
advantage of allowing twice as many tracings to be taken on the
one drum.
In order to secure the quotient

Velocity at T +10 degrees
Velocity at T degrees

in which T and T + 10 are absolute temperatures, it was thought
preferable to make the observations at temperatures of exactly
10° difference rather than to make calculations from observa-
tions at other or random intervals. The plan usually followed
was to take two observations at some convenient temperature;
then by the method previously described to change the tem-
perature 10° and take two observations, after which the nerve
was brought back to the original temperature and two more
determinations were made. The object in taking two determin-
ations at each temperature was, of course, that they should
serve as controls on each other. The reason for not taking
more than two is that, while the nerve remains irritable for
some hours after preparation, it nevertheless undergoes changes,
degenerative or otherwise, which affect its conductivity. In
order then that the chief variable shall be the change of tem-
perature it is important that the observations shall follow each
other as rapidly as possible. It is not possible, however, to
secure a change of 10° without a very considerable interval
of time. It required usually five to eight minutes to bring
about the desired change and be sure that the temperature
would remain fairly constant during the observation. It was
of course necessary that the nerve should have time to take the
actual temperature indicated by the thermometer. As an

a

SS se S---rtéi‘ ‘Cé™;éts CO

S. S. Maxwell Bye

extreme illustration of the progressive changes which may occur,
apart from the change due to temperature, I quote from the
tables to follow the velocities found in experiment 8:

Temperature. Velocity in Centimeters.
212 46),
91° 38 | Average 42.
1S 25
11° a Average 23.5
21° 29 \
9)
21° 28 | Average 28.5

42 2or5
nor
oo 23-5

correctly represents the acceleration due to a rise of 10°, but
that for temperature 21° the mean of the first and last observa-
tions,

should be taken, giving a quotient of

che)
23-5

In a preceding section I stated that on account of the failure
to observe the order of experiment just mentioned—. e., a
return to the beginning temperature—the data given by Nicolai
and by v. Miram are not usable in calculating the temperature
coefficient. Jenkins and Carlson! working with the pedal nerves
of Ariolimax observed that ‘‘the rate of the nervous impulse
seems to vary with the freshness of the preparation.” Nicolai’
called attention to similar changes in the Pike’s olfactory and

1 Jenkins and Carlson: Loc. cit., p. 254.
2 Nicolai: Arch. f. Physiol., supplement, p. 376, 1905.

372 Conduction of Nerve Impulse

considered them to be the effect of the repeated stimulation.
I am strongly inclined from some observations of my own to
believe that these changes occur in the absence of repeated stim-
ulation, and that they are caused by degenerative processes in
the nerve fiber when separated from its cell body, but I have not
sufficiently investigated the matter. The fact that they do
occur renders the above order of procedure absolutely necessary.

In the course of my experiments I found that the preparations
remained irritable for many hours, but that it was not advisable
to make more than a few observations on each nerve, and then
only while it was in the freshest condition possible. It is obvious
that the nerve in the moist chamber is exposed to the influence
of many unknown or uncontrollable variables, and that the effect
of these must increase as the preparation becomes older. The
possible injurious effect of repeated stimulation, contact with
injurious substances, changes in moisture, etc., may occur.
Other, accidental, disturbances may easily arise. For example
I several times found that the nerve suddenly ceased to respond
to the strength of shock which had just been giving good curves,
and discovered that with the lowering of the temperature the
moisture condensing on the nerve and electrodes had short-
circuited the current. Changes also take place in the muscle as
well as the nerve. This was evidenced by the increasing ten-
dency to spontaneous contraction. These contractions were not
caused by the condition of the nerve in the moist chamber for
they were not affected by cutting the nerve off close to the
muscle.

In the following table of results of my experiments I have
given not only the calculated velocities and the temperature
coefficients obtained from them, but also the latent periods for
the near and far points in each experiment, in order that the
reader may have an opportunity to judge of the degree of con-
sistency and of variability in the individual experiments. Each
numbered experiment represents a fresh preparation, The
figures are given in the order in which the records were made.
The intervals between the time of stimulation at the near and
the far electrodes were, as stated above, not far from half a
minute. The intervals between any two observations at the
Same temperature were usually from one to three minutes; and

—

ee ee eee

S. S. Maxwell g78

the intervals between observations at different temperatures
were from five to eight minutes.

Column 1 is the number of the experiment. Column 2 is the tempera-
ture. Column 3 is the latent period for the far point, and Column 4 for
the near point, in hundredths of a second. Column 5 is the velocity in
centimeters per second. In Column 6 the average velocity for each tem-
perature is placed in line with the figure representing that temperature in
Column 2. Column 7 is the quotient

Velocity at T +10
Velocity at T

in which T and 7+10 represent the temperatures in the experiment.

Conduction of Nerve Impulse

374

Ge I

6E°T

L9°1

s¢c'l

ZL 98 Aq19019,4

Ol + £3 Aq10019.4,

L

GL'0G
CLS

COP
¢ 08

*A{DO9A OBnIPAy

9

*AZIOTVA

¢

e°9%
© 0% | 1¢
¢ 91 LZ
cl | ¢°2e
0z 6E
Cae Se
91 its
GI C°8z
—Le Sf
6Z ag
1G 8%
eS] C°6Z
£Z 9¢
Ze ore
ay 76
LI ¥Z
ial 0%
cl 9%
PI re
Or | ¢°9]
ial Ze
ot ig
0Z ke
LI te
“ABON | ay
‘poled Wu94wy
r £

=
G

AD AD LD AD

“GI
GG

1G
4!
IT
1G

‘Os “dw y,

3/5

S. S. Maxwell

61°T

Og 'T

Og 'T

oF T

66° T

KG)

GL

6G

4

c¢

“8%

‘0€

“6S

Té

wD a aD nD

OT

Conduction of Nerve Impulse

‘©
m™
Y

09'@

OPE!

80'1

Lv AQwop,

oLl+ Zw Ay9012 4

,
=

¢°0€ ce ¢'8T

96 ¢ OL

G°6L €¢ ¢’ SI

901 as

¥G eel!

0€ oT

cS LP €¢ €1

8& Cor

oF 8G OT

89 cI

OF Il

8¢ Git

GL EP | OF CT

¢' OF | 8 ¢'8I

89 GLI

8 cI

€¢ LI

oF | OF ra

ce PE ¢ 91

9 9T

| eh eI

If OF 61

1Z IZ o' PZ

9F ooRG

rare

' “AVWOPA eDwWAY “AQIOTAA — =
ee ee

‘poled quoquwy]

‘DO, “dway

z

ST

FI

tea!

II

ON “dx

DAS,

S. S. Maxwell

896 GL SZ O9T 0G GG 1G
€¢ ¢ SI GG 1G
G86 86 GG Cees ial
6G &G VE IT |
66 LT 8G 1G
&¢ 1G CLG 1G 0G
¢9 1 19 | 8g 1G ¢° 9G &G |
8g GCG 86 &G
LE 96 Go hG VE teal
86 GG ¢ 08 ral! |
$9 0G GG €6
-9 06 GG &G 61
oa SI a 6P eI ¢ 61 1G
OV ial! 1G 1G
G66 L1G GT LG ni!
G& OT 9G IT
OF ial! GG 1G
&P Gl ¢ 61 1G SI
41 G@ oS &¢ ¢ 6F 9G 1G
OF 8 9G IG
GG GG 1G vé Il
GS 8I 1é Il
Tk Cau 1% 1G
6g LY £6 1G LT
69 I a a OF CVay GHG GI¢
86 8z tI! 86 Cae
€P i GG Cae 91

—tt—~S —_— as 7 LE ee - 7 eee « 7

Conduction of Nerve Impulse

378

&P I ¢o°8E 8& SS) LG G16
9€ 61 8% ¢G'1%
LG os 0G €€ tif
6% ¢’6r ¢'0€ oe
OP 8T 9G Ba 4
OF tL 8 GSE GIG ine
611 ¢ OF £6 GG 9€ €G
%@ £06 PE &G
PE og 22 ee eI
8E 0G G86 €I
8¢ 8ST G°&% €%
€¢ oT GG £&@ €%
re T cL OF 6F GS G Ig 1%
OF 1G 6% 1G
¢ 9% GS CS ¢ 9€ Il
8% £@ GE II
OP Gl GL] 1G
PE GLI LG 1G GG
6c 1 ¢€é PE | Cl% I€ 1G
9€ 1% O€ 1a
1% 1a 1G 9€ Il
1a o° 0% gS Il
Of ot ¢° 9% 1G
FE ¢ ST c% 1% 1G
LL ¥8 AQRWopA “IBIaN Fae |
—__—— *Aj100]9,A OFuIVAY | *£PI00]9A, = a *‘). *duray, ‘ON *dxq
OT + ZL 98 Ap190[9,4 ‘pouag 3u9qV'T
¢ b z I

L 9

379

S. S. Maxwell

96'T

I9'@
80'T

€°Lé

wD

wD

86

LG

9G

Conduction of Nerve Impulse

380

6y'T

cel

oF I

Po T

LZ 98 Aq190]9,4

OL + J FB AWWOPIA
ZL

ce

Gc

“AV WO[9A OFWIOAY

9

6I
GG

rat
SI

Lz
9T

"AYWO]OA

g

¢ 61 8% ey
G’Lt ¢°9Z Cy
CF LP ¢
8% LY c
¢'8I C12 Cy
LI 9% CT its
0% LE ial
LI ee ial
8I jie 7
LT 0g 74
¢° 0% ¢'8€ ial
£Z OF ial
b SF p
rd Lb P
¢'6I 8 ial
81 rE v1 Of
ZI ee el
ctl e¢ €1
o°2S aa ; |
£2 © OF :
SI | CSE eI
SI | 6% €1 6%
"IVON | “IVT
— —_— — ‘DO. “duoy ‘ON “AXA

GL

1G

wD

1n19

S. S. Maxwell

Go"

“LO

LE

‘66

Og
GG

1 ud

ren)

19.19

wD

te)

cE

ve

€&

0%

L8°G cy

681 £%

68°T 9F

Conduction of Nerve Impulse

90°% ce

18 A4100]94
te Les AQ O0[9 4 OTWIDAY

L 9

382

g

“APWOLVA

GIG (5
CG Pp
LG cP
9% L¥
PG ¢'°8§
LT 83
LI GG
9% 9€
64 OF
66 ro
O€ Og
81 G'LG
9T ¢° 0G
0€ OF
I€ tr

¢'8T Fo
0% €¢
GS OP
£6 ¢ sé
SS €€
I€ IP
(oa OF
61 1¢
1G ¢ 1g

© 0% 6%

“IVAN IB
‘poled 3U0zB'T

LE

9F

‘ON “AXE

383

S. S. Maxwell

*sqUOUIII9dxe 19440 oY} Ul UB I[qQUI[AL SseT B2UENHasUOD Ul aI SUOTYBUTULIaJap OT} {,G°) 1B [[BUIS AJOA 219M SUOTJOVIZUOD OY} [fF JUIUIMOdX |] UT,

Il 'G
i |

981

VIS

bron)

LE 1 2 1 6 19

Il
OL
0
it
OL
Il

OL

lop)
af

eV

IP

384 Conduction of Nerve Impulse

In the tables above the highest quotient is 3.15 and the lowest
is 1.08. The average of the whole number of quotients obtained
is 1.78. Inspection of the table will show that in nearly all,
though not all, cases where the quotient is very much above
or very much below the general average the abnormal result
is due to some one, or in a few cases to some two, figures
which are notably out of accord with the other magnitudes
in the experiment. These numbers are printed in italics.
Thus in experiment No. 5 the high quotient, 2.85, is clearly due
to the effect of the velocity 128 at 22.5° which differs so far from
the other velocity at that temperature, namely, 43. Again in
Experiment No. 14 the low quotient 1.10 is the effect of the
apparently disproportionately high rate of 58 at 11°. The sig-
nificant fact is that out of 48 quotients, obtained from 43 dif-
ferent nerves, only two are low enough to come within the limit
of acceleration of a physical process by a temperature rise of 10°.
It seems logical to conclude from the outcome of these experi-
ments that the transmission of the nerve impulse is a chemical
process.

Admitting the chemical nature of the nerve impulse, the ques-
tion arises, can we go further and reach any conclusion as to the
particular kind of reaction involved? It would be dangerous to
carry speculation very far in this direction, but one thing appears
to be significant and that is the relatively low reaction velocity
as compared with that of those biological processes which are
known to involve oxidations or the giving off of carbon diox-
ide. Among such processes may be mentioned the germination
of seeds, the segmentation and development of the egg and
the contraction of muscle. According to the experiments of
Clausen‘ the quantity of carbon dioxide given off from germinat-
ing seeds for temperatures between o and 25° increases about 2.5
times for each rise of 10° temperature. Loeb? has found that
free oxygen is necessary to the production of artificial par-
thenogenesis, and that a rise of temperature of 10° hastens the
process three or more times. For the rate of contractions of the
terrapin heart Snyder* obtained at temperatures between o and

* Cohen: Lectures on Physical Chemistry, New York, p. 60, 1902.

? Loeb: University of California Publications, Physiology, iii, p. 39, 1906.

* Snyder: University of California Publications, Physiology, ii, p. 141,
1905.

S. 5. Maxwell 385

25° quotients ranging from 10.2 to 2.1 For the sinus venosus
of the frog’s heart he found quotients of similar magnitudes,
as well as for the rhythmic contraction of smooth muscle.” It
will be noticed further that in the experiments on contractions
of the heart and of smooth muscle the quotient for tempera-
tures below 15° become markedly larger. The quotients for
velocity of the nerve are not only lower throughout the range
of temperature which I have investigated, but they do not show
the excessive increase below 15°. The average of the whole 48
quotients is, as I have stated, 1.78; the average of the 19 cases
in which the upper temperature was not above 15° is 1.94—an
increase but not such a marked increase as is shown by con-
tractile tissue. The inference from these facts would seem to
be that the reaction involved in the nerve impulse is not of the
nature of an oxidation, an inference which is made more prob-
able by the fact that the nerve can remain irritable for several
hours in lack of oxygen,? whether the nerve is stimulated dur-
ing that time or allowed to remain at rest. One remembers
alsoin this connection the remarkable sparseness of blood supply
to the nerve trunks.

The results of this series of experiments may be summarized
in a single statement: The temperature coefficient of the veloc-
ity of the nerve impulse indicates definitely that the conduction
is a chemical process, but probably not of the nature of an oxi-
dation.

‘Snyder, Arch. f. Physiol., p. 118, 1907.
PiGcd- p. 426.
3 Baeyer: Zeitschr. f. allgem. Physiol., 11, p. 169, 1903.

QUANTITATIVE METHODS WITH HEMOLYTIC SERUM.'!
By WILFRED H. MANWARING.

(From the Pathological Laboratory of Indiana University.)
(Received for publication, June 13, 1907.)

In a previous paper,” it was pointed out that the application
of direct quantitative methods is at times impossible with hemo-
lytic serum; that, for example, the direct determination of the
amount of hemolytic amboceptor (specific thermostable sub-
stance) remaining in a heated* hemolytic goat serum, after the
serum has been exposed to washed sheep corpuscles, is out of
the question. Work has been undertaken, during the last two
years, to devise a possible tudirect method of analysis, applicable
to this problem.

Such amboceptor determinations are impossible, because heated
hemolytic goat serum is so changed by contact with sheep cor-
puscles, that duplicate titrations do not agree. To illustrate,
in one attempt, duplicate titrations gave results varying from
32 per cent to 77 per cent of the original amboceptor; in another
attempt, from 70 per cent to 113 per cent, depending in each case
on the volume analyzed.*

Heated hemolytic serum contains, in addition to amboceptor,
a large number of unknown substances. These for convenience
have been spoken of collectively as constituting the third serum
component.°

It was found that a change in the relative amount of ambo-
ceptor and third component in a heated hemolytic serum, is suf-
ficient to render the direct analysis of that serum impossible, by
causing a non-agreement of duplicate titrations. Duplicate
determinations with such altered sera gave results varying from
52.5 per cent to 64.0 per cent in one experiment, and from 5.0

1 Work aided by the Rockefeller Institute for Medical Research.
? This Journal, i, pp. 213-218, 1906.

3 55°-59° C., for 30-60 minutes.

4 Journ. of Infect. Dis., ii, p. 490, 1905.

5 [bid., iii, pp. 647-662, 1906.

387

388 Hemolytic Serum

per cent to 9.8 per cent in another, depending on the volume
analyzed; the actual amount of amboceptor present in the two
experiments being 50 per cent and 5.6 per cent.’

These disagreements, though sufficient to render quantitative
work valueless, are not as marked as the disagreements observed
in sera after exposure to corpuscles. Changes, other than those
of the relative amount of amboceptor and third component, were,
therefore, suspected.

It was found that pure third component (heated normal goat
serum) suffers marked qualitative changes when exposed to
sheep corpuscles. A third component originally hemolysis-
increasing (auxilytic) has its auxilytic power decreased by such
exposure, or even replaced by an antilytic power; while a third
component originally antilytic has its antilytic power increased.”

It was further found that washed sheep corpuscles give off a
powerful antilytic substance into physiological saline, and that
the addition of a proper amount of this exposed salt solution to
a third component, produces approximately (though not exactly)
the same changes as those resulting from exposure to corpuscles.

The amount of the exposed salt solution necessary to produce
this approximate change, however, is not the same with all sera.
With one third component, the desired change is most closely
simulated, by the addition of an equal volume of the salt solution;
with a second serum, a half-volume produced the closest approx-
imation; while, with a third sample, a double volume is required.

As a result of this work, the following changes are now con-
ceived to take place in a heated hemolytic goat serum during
exposure to washed sheep corpuscles: (z) A decreasein theamount
of amboceptor, due to the absorption of amboceptor by cor-
puscles. This, of course, is as yet a purely hypothetical change,
as an unchallengeable measurement of this decrease is at present
impossible. (iz) A resulting change in the relative amount of
amboceptor and third component. (iii) The giving off into
the serum of antilytic corpuscles products, the amount (and pos-
sibly the nature) of which is influenced by the nature of the serum

' Journ, of Infect. Dis., iti, p. 648, 1906.
* A detailed account of these experiments is now in press in the Journal
of Infectious Diseases.

Wilfred H. Manwaring 389

used. And (cv) certain minor changes, the nature of which is
not understood.

These conceived changes are so complex, and each has such
a marked influence on hemolytic power, that it does not seem pos-
sible, at present, to devise an indirect method of analysis by means
of which the residual amboceptor in an exposed heated hemolytic
serum can be measured.

A NEW STANDARD FOR USE IN THE COLORIMETRIC
DETERMINATION OF IODINE.

By ATHERTON SEIDELL.
(From the Division of Pharmacology, Hygie nic Laboratory, Washington, D.C.)

(Received for publication, July 19, 1907.)

Of the methods which have been proposed and used for the
determination of small amounts of iodine in such substances as
thyroid glands, blood, muscle, and various animal and plant
substances the one which has apparently been found the most
satisfactory is that adapted by Baumann and Roos! from the
old process of Rabourdin and at present known as. Baumann’s
method. According to this process the weighed sample of mate-
rial which should contain not more than a few milligrams of
iodine is moistened with water and mixed thoroughly with about
five times its weight of powdered sodium hydroxide. After
warming gently the heating is increased until the mass fuses to
a liquid or semi-solid paste. If much charred organic matter
is present a little sodium nitrate may be added for its more com-
plete oxidation, and the heating continued until a clear melt is
obtained. After cooling, water is added and the aqueous solu-
tion containing all of the iodine in the form of an inorganic salt
is filtered into a separatory funnel. The solution is acidified
with sulphuric acid while being kept cool and is then treated with
to-20 drops of 1 per cent aqueous sodium nitrite solution and
the liberated iodine shaken out with chloroform. The intensity
of the pink color in the chloroform layer is proportional to the
amount of iodine present, and it is only necessary to compare this
color with standards of known iodine content in order to ascertain
the amount of iodine in the unknown sample. The standards so
far used have been prepared from aliquot portions of standard
potassium iodide solution, shaking out the liberated iodine with
the same volume of chloroform used for the unknown sample.

1 Zeitschr. f. physiol. Chem., xxi, p. 489, 1895-96.

301

392 Colorimetric Iodine Determination

The experience of Baumann and of many other investigators,
has shown ,that the pink colors in the chloroform layers fade
rapidly and therefore it is absolutely necessary that the stand-
ards be freshly prepared for practically all determinations.
Many efforts have been made in this laboratory to prevent or
even retard the fading of the iodine color in the chloroform but
without success. In endeavoring to overcome this difficulty
in another way, an examination of the tints of the aqueous solu-
tions of a large number of pink dyes was made, in the hope of
finding one, which resembled near enough the color of the chloro-
form solution of iodine, to be used as a standard for the colori-
metric estimation of the latter. Although none of the several
dozen pink dyes which were examined gave solutions possessing
a tint matching exactly that of iodine dissolved in chloroform,
several gave colors which corresponded near enough for all prac-
tical purposes. Among these may be mentioned Neutral Red,
Rubin (patent acid) and Fuchsin S (acid fuchsin according to
Weigert), all of which dyes in the present case bore the labels
of Dr. G. Gribler & Co., Leipzig. The Fuchsin S was, on the
whole, found to be the most satisfactory and was selected for use
in the tests here mentioned.

Experiments in making the iodine determinations with stand-
ards of iodine in chloroform indicated that a very satisfactory
color was one having the intensity produced by 1.0 milligram
of iodine dissolved in to cc. of chloroform. To reproducesuch
a tint with Fuchsin S it was found necessary to dissolve 0.02
gram oi the dye in 200 cc. of water made acid with 3-5 per cent
of hydrochloric acid and dilute inturn ro cc. of this solution to
200 cc. with water similarly acidified with hydrochloric acid. A
standard prepared in this way and compared in a Duboscq color-
imeter against a solution of 1.52 milligrams of iodine dissolved
in 10 cc. of chloroform gave a series of readings the average of
which was 14.95 of the Fuchsin S solution equaled 10 of the chloro-

: See Tage

form solution, from which it is seen that = 0.102 per 1 eG:
14.92

or 1.02 milligram of iodine equivalent to ro cc. of the standard

Fuchsin S solution. Similar readings were made with other

Fuchsin S$ solutions prepared fresh and allowed to stand various —

lengths of time and it was found that the desired tint could be

,

Atherton Seidell 393

reproduced at pleasure and remained unaltered for at least three
months, provided the solutions were kept acid. In addition to
this standard corresponding to 1.0 milligram of iodine per ro cc.
of solution it was also found advantageous to prepare a series of
test tubes with gradually increasing intensities of the pink color
corresponding to 0.025, 0.05, 0.075, 0.10, 0.15, 0.20 and 0.25
milligram of iodine per 5 cc., and also in smaller test tubes a
series corresponding to 0.01, 0.02, 0.03 and o.04 milligram
iodine per r cc. Provided with such standards it is a simple
matter to shake out the liberated iodine from the solutions of
the melt of the samples examined, with 1, 5 or 10 cc. of chloro-
form as the amount of iodine might require and compare the
solution so obtained with the corresponding standards.

The several points of advantage, such as saving of chloroform,
permanence and uniformity of standards resulting through the
use of such pink dye solutions as above described will be readily
recognized by anyone making frequent colorimetric determina-
tions of iodine. The dye standards have been in use in this
laboratory with entire satisfaction for several months.

Attention is to be directed also to the applicability of the
above described method to the rapid estimation of iodine in
many pharmaceutical preparations—such as U. S. P. desiccated
thyroid glands, etc.

Finally, it is of interest to note that Dr. J. H. Kastle of the
Hygienic Laboratory has found that the pink colors obtained
with Griess’s reagent for nitrites closely corresponds with the
color of iodine dissolved in chloroform. He has therefore made
use of an acid solution of Fuchsin S as a permanent standard
for nitrites in potable waters with good results.

¥
K on. _ -
. e 7 le :
A . ' - L@
1 A :
7
‘ « —
) ; °
ae @

(5

ANTI-INULASE.

By TADASU SAIKI.

(From the Research Laboratory of the Department of Health of New York
City, Wm. H. Park, M.D., Director.)

(Received for publication, July 13, 1907.)

INTRODUCTORY.

The peculiar endoenzyme inulase was first described by Green!
in Helianthus tuberosus, although Dragendorff? had earlier noticed
the inulin-splitting process in several vegetable structures. Our
knowledge of this enzyme in certain fungi is largely due to the
investigations of Bourquelot.* Subsequently, in confirmation
of the result of Bourquelot, Dean* found inulase in the fungi
Aspergillus niger and Penicillium glaucum. Inulase has not yet
been found in the animal body.’

Since the discovery of anti-emulsin,® anti-enzymes have been
subjected to increased experimental study by several investi-
gators.

Anti-trypsin,’ anti-rennin,' anti-cynarase,* anti-coagulin,” anti-

1 Annals of Botany, i, p. 223, 1888.

7 St. Petersburg (1870), cited by Wortmann: Bzol. Centralbl., iii, p. 266.

3 Compt. rend. de l’Acad. des sct., cxvi, p. 1143, 1893; Compt. rend. de la
soc. biol., ix, p. 481 1893; zbid., ix, p. 653, 1893; Bull. soc. mycol., France,
1X, p. 230, 1893; zbid., x, p. 235. Alsocf. Chevastelon, Journ. Pharm.,
IZ, Os ty 2etsko)ise

* Botan. Gazette, KXXV, P. 24, 1903.

5 Chittenden: Amer. Journ. of Physiol., ii, p. xvii, 1898; Richaud:
Compt. rend. dela soc. biol., iii, p. 416, 1900; Bieri and Portier: zbzd., p. 423.

8 Hildebrand: Arch. 7. pathol. Anat. u. Physiol., cxxxi, p. 5, 1893.

7 Fermi and Pernossi: Zeitschr. f. Hyg., xviii, p. 83, 1894; Landsteiner:
Centralbl. f. Bakteriol., xxvii, p. 357, 1900; Achalme: Ann. de l’Inst.
Pasteur, xxv, p. 737, 1901; also von Dungern: Miinchner med. Wochenschr.,
1898, cited by Metchnikoff, 1901, and by Morgenroth, 1899.

8’ Morgenroth: Centralbl. f. Bakteriol., xxvi, p. 349, 1899; ibid., xxvii,
Pp. 721, 1900. Cf. Briot: Thése de la faculte des sci. de Paris, No. 4,
cited by Metchnikoff, 1901; also Rédén, Ref. in Maly’s Jahresber. f. Thier-
chem., xvii, p. 160, 1887.

® Morgenroth, cited by Nuttall: Blood Immunity and Blood Relationship,
Pp. 16, 1904.

10 Bordet and Gengou: Amn. de l'Inst. Pasteur, xv, p. 129, 1901.

395

396 Anti-Inulase

urease,! anti-pepsin,? and also anti-anti-rennin® were early con-
sidered.

The presence of many of the antibodies, as well as of the com-
mon enzymes also, normally in serum, complicates the study of
the development and specific relationship of the anti-enzymes
and the respective enzymes. The absence from the organism
of a substance like inulase and of its antibody also, then, presents
opportunities especially favorable for the study of the relation-
ship of enzyme to antibody in general and in particular of inulase
and anti-inulase. I have accordingly thought it of interest to
study the antiserum prepared by immunization with suspensions
of ground up aspergillus. By the comparative quantitative
inhibitory effects of the serum on the action of aspergillus
extracts in inulin and in sucrose digestions, it seemed possible
to individualize an inulase and a sucrase rather than to ascribe
the two effects of the aspergillus extracts to a single enzyme.
Further, by the specificity of the aspergillus serum for the asper-
gillus sucrase as contrasted with the sucrase of intestinal origin,
there was offered a means of differentiating the aspergillus and
intestinal sucrases.

EXPERIMENTAL.

In my experiments rabbits were employed. The serum was
diluted up to three or four times its original volume with water,
before using, to prevent the formation of a large coagulum by
heating. The digestive mixtures, containing sodium fluoride
and covered with liquid paraffine, stood for varying periods in
test tubes at a fixed temperature, usually in the incubation room
at 38°. Then they were filtered through small filters and 6 cc.
samples were pipetted into large test tubes after dilution with
water to a definite volume or without dilution when there was
only a small quantity of sugar present. The test tubes were
placed together in a net basket. - Fifteen cc. portions of Fehling’s

1 Moll: Beitr. z. chem. Physiol. u. Path., ii, p. 344, 1902 (reviewed by
Jacoby, Centralbl. fj. Bakteriol., xxxii, p. 439).

? Sachs: Fortschritte der Medizin, cited by Michaelis and Oppenheimer,
Pp. 399, 1902. Cf. Pugliese and Coggi, Maly’s Jahresber. f. Thierchem., xxvii,
p. 832, 1897.

* Korschun: Zeitschr. f. physiol. Chem., xxxvi, p. 141, 1902.

Tadasu Saiki 397

solution (previously diluted one-half with water) were heated
in a water-bath and rapidly transferred at the same moment into
the digestion mixtures. The combined portions were then
heated three minutes in a boiling water-bath; the precipitated
cuprous oxide was filtered off at once upon a Gooch crucible, and
weighed as cupric oxide.

The inulin employed was obtained from Merck. The white
preparation was further purified by repeated washing with water
and then with alcohol. The inulase was from Aspergillus niger
and was prepared by Dr. A. L. Dean, who kindly furnished me
with some of his product for the present work. The mould had
been grown on an inulin-containing medium until the spores
were well developed. The mycelia were filtered out of the cul-
ture fluid, disintegrated, washed in acetone, and after drying
below 50° were pulverized. The product acted strongly on
inulin and saccharose, but not on starch.

I. Js there an Inulin-splitting Enzyme or an Anti-inulase present
in Normal Rabbit Serum?

In all, nine observations on the serum (in part heat-inacti-
vated) of six normal rabbits were made. Some of the typical
protocols are given. Three additional experiments to determine
the influence of serum on the hydrolysis by weak hydrochloric
acid are of interest in reference to my results with the inulase.
An example of these, also, is included below.

The experiments show a partial inhibition of the action of the
inulase when the rabbit serum is present in the digestive mixture.
The same effect, however, was obtained with the heated serum
and in experiments with neutralized proteid-free filtrates of the
coagulated serum.

EXPERIMENT III, RABBIT 3.

CuO in 6 ee.
gm.
2 per cent inulin solution, 3.5 ec. + 5 per cent NaF solution, 1 ce.

+ Normal serum, 1 5— oo Olce tt water: OlD CG-.n tactics a acieie renters sis 0.0028
+ ileal —— 3 oY 2 orl) Sah gira (On Gcnaecucecocequacacas 0.0034
+ se “1.55.0 “ +1 percent inulase solution,0.5cc.......... 0.0084
+ “ —§1.5— 5.0 “ (boiled) + 1 per cent inulase solution,0.5cece... 0.0085
+ Water 5:0 cc. + 1 percent inulase solution,0.5cc..............0.2-02ee50+% 0.0139
ee Lialhe Oe Cee R EEOC CEO eC EO OSLO eGR CES ooo oy Prmiaem ce Goose none

The digestive mixtures were kept at 38° for 20 hours. Rabbit 3 was subsequently used
for immunization.

* In all my experiments, ‘1.5 —> 5.0cc.”’ indicates that 1.5 ec. of the original serum was
made up to 5.0 ce. by the addition of 3.5 ec. of water.

398 Anti-Inulase

EXPERIMENT VI, RABBIT 5. :
‘ CuO in 6 ee.

gm.
1 per cent inulin solution, 6.5 cc. + 5 per cent NaF solution, 1 ce.
+ Normal serum, 0.5 —> 1.5 cc. + 1 percent inulase solution, 1 cc...........-+++ 0.0087
+ “0.5 — > 1.5 “ (boiled) + 1 per cent inulase ‘solution, lcc...... 0.0082
+ i “ 0.5 —> 1.5 “ + 1 percent inulase solution (boiled). 1 cc...... 0.0017
+ : “0.5 —> 1.5 “ (boiled) +1 per cent inulase solution(boiled), lec. 0.0021
+ Water, 1.5cc. + 1 percent inulase solution, Lcc...... 2... eee eee ee eee eee 0.0151
The digestions were kept at 38° for 20 hours.
EXPERIMENT VIII, RABBIT 5.
2 per cent inulin solution, 5cc. + 5 per cent NaF solution, 1 ce.
+ Normal serum, 1 —> 3ce. (at 65° for 30 min.) + 1 per cent inulase solution, lec. 0.0097
+ 5 “" _1—53 “ (boiled) + 1 percent inulase solution, 1 cc.......... 0.0102
+ . “ -1—>3 “ + 1 percent inulase solution, 1 cc..............+-+. 0.0111
+ “ RL ay i eee tm ote roid cic case 3 0.0016

The digestions were kept at 38° for 20 hours.

EXPERIMENT IX, RABBIT 6.
1 per cent inulin solution, 6.5 ec. + 5 per cent NaF solution, 1.0 ce.

+ Normal serum, - 56— 1. = ce+ 1 per cent inulase solution, lce............... 0.0085

+ 0.5— 1.5 “ * (boiled) u : per cent inulase solution, 1 ce.. a oe
. I 1 ae 1 RRA CC c

+ Proteid-free * ; \ II Teor i 2 “ s es De eye 0.0081

+ Water 1.5 cc. + 1 per cent inulase solution, 1 Cc... ......+- ee eee eee eee neers 0.0129

The digestive mixture was kept at 38° for 16 hours.

* Three cc. of serum were diluted up to 9 cc. in a test tube, co: eyulsted by boiling and the
addition of a trace of dilute acetic acid; the filtrate was neutralized with dilute soda solution.

EXPERIMENT XI, RABBIT 7.

CuO in 2 ce.
gm.
0.1 per cent HC! solution,* 5.0 cc. + 2 per cent inulin solution, 3.5 ce.
a Normal ehceiie Min | ea 7 BSL Re CIC a eICe eI EROR EORTC SiGIOm Ic aS o GoasG on 0.0142
+ . 1 ep SF Ra (705 0%: D Re ES lt Gc © 0.0146
CU ER Lenco tate ie eos oleic leo laicecleldln die oo fe dueca (o.a,ale Ble dies one eR NOR 0.0148

The digestive mixtures were kept at room temperature for 20 hours.
* The digestions contained, therefore, 0.05 per cent HCl.

The results of my experiments with normal rabbit serum show
no evidence of the presence of an inulin-splitting enzyme. On
the contrary, the addition of the serum to the digestions retards
the hydrolysis considerably. This noteworthy inhibition of the
action of inulase by the normal serum is not due to the proteids
present or to the reaction of the serum; for when the proteids
are coagulated out, the neutralized filtrates still inhibit the split-
ting of the inulin. The retardation cannot be ascribed to the
presence of an anti-inulase of the common anti-enzyme type in
the serum.

Without considering for the present the nature of the products
formed, it is evident that the process of hydrolysis by acids and

that by the enzyme are different, since the acid hydrolysis is not ©

influenced by the serum.

Tadasu Saiki 399

II. The Effects of the Serum of Immunized Rabbits upon the
Digestion of Inulin by Inulase.

Suspensions of the aspergillus inulase preparation were ground
up with physiological salt solution and kept overnight in the
ice-box. Injections of the material thus freshly prepared were
made subcutaneously on alternate days into two rabbits (rab-
bits 2 and 3, weighing 1995 gms. each) which had been bled from
the ear veins for normal serum previous to the initial treatment.
These rabbits were given a total of o.42 gram each of the asper-
gillus powder in gradually increasing doses of from .oo5—o0.1
gram over a period of 26 days. Rabbit 2 had lost about 200
grams in weight and Rabbit 3 about 95 grams, when the two were
again bled five days after the final injection.

The digestion experiments were carried out in the same general
way as were the observations on the normal serum. In the four
experiments made with the serum of the immunized rabbits, it is
shown conclusively that the anti-aspergillus serum inhibits the
action of the inulase on inulin. An anti-inulase has been devel-
oped by repeated injections of the aspergillus. A typical pro-
tocol follows:

EXPERIMENT XIII, ANTI-ASPERGILLUS SERUM FROM RABBIT 2.
CuO in 6 ee.

1 per cent inulin solution, 6.5 ec. + 5 per cent NaF solution, 1 ce.

+ Serum 0.3— 1.5 ce. =) percentunulasesolutiony McC... .. >> 5-6 os oak 0.0045
0.3— 1. ae * (boiled ) +1 per cent inulase solution, ICGENE Neto Se 0.0132
a) i 0.3—>1. +1 i (polled): 0.0027

Digestions were 5 at 38° for 2 days.

III. The Effects of the Anti-aspergillus Serum upon the Inversion
of Sucrose.

The well known inverting action of yeast on cane sugar was
attributed to an enzyme by Berthelot,' who gave to it the name
“ferment glucosique.’”’ The sucrases of fungi have been de-
scribed in numerous papers” and the animal sucrases have been
studied by many investigators since the classic discovery of

1 Compt. rend. de l’ Acad. des. sct., ii, p. 980, 1860. Cf. Dubrunfant:
4D1d., XXill, p. 38, 1846.

2 Cf. Béchamps: ibid., xlvi, p. 44, 1858; Gayon: zbid., 1xxxvi, p. 52, 1878;
Kossmann: Bull. soc. chim., xxvii, p. 251, 1877; Duclaux; Chimie Bio-
liogique, Paris, 1883, cit. J. R. Green: The Soluble Ferments and Fermen-
tation, p. 115, 1899; Bourquelot: Compt. rend. de la soc. biol., p. 237, 1898.

400 Anti-Inulase

sucrase in the intestine. Sucrose injected intravenously reap-
pears in the urine unaltered, according to Cl. Bernard.? This
indirect evidence of the absence of sucrase in the blood is in
harmony with the experimental observations reported below.
Mendel and Mitchell,? however, have recently observed in one
experiment that cane sugar intraperitoneally injected may be
utilized in part.

It has already been suggested that the sucrases of yeasts and
fungi differ considerably in several properties from other plant
sucrases.‘ By immunizing animals with pancreatin, M. Ascoli
and Bonfanti® have succeeded in developing an antibody in the
inactive serum of the rabbit acting specifically against pancreas
diastase. They assume to have demonstrated that the serum
amylases are specific for various kinds of starch, thus supple-
menting the explanation of the different characters of the sali-
vary amylolysis described by Hamburger.® And they proceeded
even to show that the anti-amylase of one animal acts upon the
serum-amylase of another species in a different degree. Ina
paper’ prepared under Professor Mendel’s guidance, it was sug-
gested that enzymes of plant origin act upon plant carbohy-
drates more readily than those of animal origin.

The influence of the serum upon the aspergillus sucrase and
the sucrase of different origin was determined in the following
four series of experiments:

a. The influence of normal serum upon aspergillus sucrase.

b. The influence of the anti-aspergillus serum upon asper-
gillus sucrase.

c. The influence of normal serum upon intestinal sucrase.

d. The influence of the anti-aspergillus serum upon intestinal
sucrase.

From the results of several experiments in each series, it
appears that normal rabbit serum, even when boiled or heated

*Cl. Bernard: Legons sur le Diabéte, Paris, p. 259, 1887.
? Cl. Bernard: cit. Schiitzenberger, Intern. Wiss. Bibl., p. 259, 1876.
* Amer. Journ. of Physiol., xiv, p. 239, 1905.
‘Cf. Oppenheimer: Die Fermente u. ihre Wirkungen, p. 206, 1900.
* Zettschr. f. physiol. Chem., \xiii, p. 156, 1905.
" Jahresber. f. d. ges. Med., 1871.
Saiki: this Journal, ii, p. 263, 1906.

Tadasu Saiki AOI

at 65° for 30 minutes inhibits considerably the inversion of sucrose
by the aspergillus extract, a result to be expected from the
similar influence on the digestive action of inulase. The anti-
aspergillus serum has a yet more powerful sucrase-inhibiting
effect than has the normal serum, though thisis not so pronounced
as in the inulin experiments. When compared with the normal
serum controls, inversion with intestinal sucrase is not influenced
by the antiserum.

The greater inhibitory effects of the antiserum on the inulin
digestions as contrasted with the results obtained below with
sucrose, point to the existence of individual inulin-splitting and
sucrose-inverting enzymes in the aspergillus. The experiments
show, as indicated by the specificity of the anti-aspergillus serum,
a difference between the intestinal and the aspergillus sucrase.
For the differentiation of the sucrases, further confirmatory
observations with a more active anti-aspergillus-sucrase are desir-
able, as well as a determination of the influence of the anti-intesti-
nal sucrase on the sucrose-splitting enzyme contained in the mould.
The observations might, also, be extended to include the sucrases
of other species.

a. The Influence of Normal Serum and Heat Inactivated Normal Serum upon Asper-

gillus Sucrase.
EXPERIMENT XVII (SERUM FROM NORMAL RABBIT 8).

CuO in 2 ec.

gm.

2 per cent sucrose solution,* 6.5 cc. + 5 per cent NaF solution, 1 ce.
+ Serum 0.5 — 2ce. + 1 per cent inulase solution . 5 ce. Deven OLOLSE
ee ee 0.5 — Z ‘** (boiled) 2 Tees e ee Sse eee 0.0141
+ “ 0.5—> 2 “ +1“ if = (boiled) 6. 5 Gen ae) 10-0012
“Ry oats 0.5— 3 “ (boiled) +1 ‘f o - . OFS) Se 0023
+ ‘Water 2 cc. + l.per cent inulase solution, O.5cc..........-..+csccccseceres 0.0244
+ o ” SWiSG}G0s. DAE 6 ney bape on cS REO RBG SOSA a eee ee eee oe Sse trace
Digestions kept at 38° for 20 hours.
EXPERIMENT. XVIII (SERUM FROM NORMAL RABBIT 8). ;

CuO in 1 ee.

2 per cent sucrose solution, 5 cc. + 5 per cent NaF solution, 1 ec. i
+ Serum 1 —> 38 cc. (heated at 65° for 30 min.) + 1 percent inulasesolution,1 ce. 0.0111
— = 1— 3 “ (boiled) +1 per cent inulase solution, 1 ee 0.0102
+ M 1—3 “ (fresh) +1 ‘ 5 ea betes chet = ina 0.0098

Digestions kept at 38° for 20 hours.
* The sucrose was purified by recrystallization from water.

b. The Influence of the Anti-aspergillus Serum upon Aspergillus Sucrase.

EXPERIMENT XIX (SERUM FROM IMMUNIZED RABBIT 2).

CuO in 6 ce.
gm.
2 per cent sucrose solution, 6.5 ce. + 5 per cent NaF solution, 1 cc.
+ Serum 0.5 —1.5ce. + 1 per cent inulase solution 1 ee 0.0171
=aMay 0.5—> 1.5 “ (boiled) oe an une Mee peaStovys Sarat 0.0250
= a 0.51.5 “ ae Ll te $ « 1 “ (boiled)..... 0.0028

Digestions kept at 38° for 20 hours.

402 Anti-Inulase

c. The Influence of Normal Serum upon Intestinal Sucrase.

EXPERIMENT XXII (SERUM FROM NORMAL RABBIT 9).

CuO in 2 ee.
? gm.
2 per cent sucrose solution, 5.5 ec. + 5 per cent NaF solution, 1 ce.
+ Serum, 1 — 3ce + intestinal extract* 0.5cc......-.-.--+-2seeee 0.0133
+ ey 1—>3 “ (boiled) + * 5: OB eh Secs le choles oienetans 0.0137
+ t—>3° “ sg 2 . 0.5“ (boiled) 3. 650.-0eaee trace

Digestive Mixtures were kept at 38° for 20 hours.

* The fresh mucosa of a normal rabbit's small intestine was ground up with sand, and
extracted with water containing 0.05 per cent Na,CO3 and 0.5 per cent chloroform, then
filtered. and the clear filtrate employed for the intestinal sucrase experiments.

d. The Influence of the Anti-serum upon Intestinal Sucrase.

EXPERIMENT XXVIII (SERUM FROM IMMUNIZED RABBIT 2)

CuO in 2 ee.

2 per cent sucrose solution, 5.5 ce. + 5 per cent NaF solution, 1 ce. y
+ Serum, 1 — 3ce. - intestinal/extract 0.5'CG.0 6 10 2 w cle triers 0.0185
+ * 1— 3“ (boiled) + . ce 0) Biss jschecctess tsetse eee 0.0190
7 1—3 “ i: oF z 0:5" (boiled) icc. Son «etna trace

Digestive mixtures kept at 38° for 20 hours.
CONCLUSIONS.

1. In normal rabbit’s serum, neither an inulin-splitting
enzyme nor an antibody of inulase occurs. Sucrase and its anti-
body are likewise not found.

2. The addition of serum induces an inhibition of the diges-
tive action of inulase, independent of the presence of proteid or
alkali. The hydrochloric acid hydrolysis is not inhibited by
serum.

3. By subcutaneous injection of inulase an antibody for inu-
lase can be produced in rabbit’s serum.

4. The anti-serum exhibits different degrees of inhibitory
action upon the inulin-digesting and sucrose-inverting activities,
respectively, of inulase preparations. Presumably, therefore,
the inulin-splitting action and the sucrase activity exist inde-
pendently in the preparation of so-called ‘‘inulase’”’ from Asper-
gillus niger.

5. Upon the intestinal-sucrase, the anti-serum exerts no
noticeable influence, or at least very little under the conditions
of these experiments.

I desire to express my obligation to Doctors Park and Gibson
for the opportunity of conducting this investigation in their
laboratories.

THE INFLUENCE OF ALCOHOL ON THE METABOLISM
OF HEPATIC GLYCOGEN.'

By WILLIAM SALANT:2

(From the Laboratory of Biological Chemistry of Columbia University, at
the College of Physicians and Surgeons, New York.)

(Received for publication, August 20, 1907.)

INTRODUCTORY.

Inquiries concerning the accumulation and transformation of
glycogen in the liver have, within recent years, led investigators
to study the influences exerted on these hepatic processes by
various poisons when introduced in the body. Metallic com-
pounds, also substances of the aliphatic and aromatic series as
well as bacterial poisons, have been used. The results of such
researches indicate that inorganic as well as certain organic
poisons cause the removal of glycogen from the liver.

Thus Kaufholz* has shown that phosphorus poisoning in rab-
bits causes rapid transformation of glycogen in the liver. Koch?
obtained similar results with corrosive sublimate. Kissel,> work-
ing in the same laboratory, corroborated the findings of Koch,
and also made the interesting observation that the transforma-
tion of glycogen induced in the liver by the administration of
corrosive sublimate may be inhibited by means of alcohol. Gar-
nier and Lambert® stated that after the intravenous injection of
sodium chloride the liver was freed from glycogen. Kriukoff’

‘The results of some of the experiments have already been communi-
cated in preliminary reports: Proceedings of the Society for Experimental
Biology and Medicine, 1905-06, iii, p. 58; Journal of the Am. Med. Assn.,
xlvii, p. 1467, 1906.

? Research Fellow of the Rockefeller Institute.

3 Kaufholz: Dissertation, Wiirzburg, 1894. °

4 Koch: «bid.

5 Kissel: Centralbl. f. inn. Med., xvi, p. 613, 1895.

* Garnier and Lambert: Compt. rend. de la soc. de biol., xlix, p. 617,
1897.

7 Kriukoff: Dissertation, Moskau, 1902.

403

404 Influence of Alcohol on Glycogen-Metabolism

carried out alarge number of experiments with various sub-
stances to ‘test their action in this regard. He reported that
after subcutaneous injection of arsenic, phosphorus, corrosive
sublimate, anilin, phenyl hydrazine, pyrogallol, sulphuric acid,
sodium hydroxide, or carbolic acid, glycogen disappeared from
the liver in twenty-four hours. Alcohol had the same effect if
injected subcutaneously at intervals for a long period.

In. this connection the work of Drummond and Noel Paton!
may be mentioned. These investigators found that in acute
adrenalin poisoning in rabbits the liver glycogen was markedly
diminished. The same results were obtained by Doyen and
Kareff? with adrenalin chloride when they injected this sub-
stance into the portal vein. Likewise pilocarpine may, accord-
ing to these investigators, favor glycogen transformation in the
liver. Mohr’s’ studies with various gastrointestinal irritants,
such as aloin, arsenious acid and croton oil have led to the same
conclusion.

Claude Bernard‘ was the first to announce that during fever
the glycogen in the livers of animals decreased even if nourish-
ment was given. His observations were confirmed by Bouley.°
May’ carried out related experiments on dogs and rabbits. He
also found that in fever hepatic glycogen diminishes. May
stated that fifteen hours after feeding cane sugar to rabbits in
which fever was induced by injecting pathogenic bacteria, 1.69 to
5.05 per cent of glycogen was obtained from the livers of such
animals. In control rabbits, which received the same amount
of cane sugar, the quantity of glycogen in the livers varied between
g.18 and 11.93 per cent. His results were even more marked
when twenty-four hours were allowed to elapse after feeding 30
grams of glucose to rabbits in febrile condition. The liver
removed at the end of this time contained an average amount
of 0.42 per cent of glycogen. The controls contained 2.71 per

‘Drummond and Paton: Journ. of Physiol., xxi, p. 92, 1904.

? Doyen and Kareff: Compt. rend. de la soc. de biol., lvi, p. 716, 1897.

* Mohr: Dissertation, Wtirzburg, 1894.

‘ Bernard, Claude, quoted by Roger: Arch. de physiol. norm. et pathol.,
5th series, vi, p. 64, 1894.

5 Bouley: ibid.

* May: Zeitschr. f. Biol., xxx, p. 48, 1894.

William Salant 405

cent of glycogen. Ott’ obtained similar results. He likewise
induced infection in rabbits, by the method previously employed
by May. Cane sugar was then given such rabbits and only
those whose temperature was 40° C. were used for the experi-
ments. Fifteen hours later the livers of these rabbits were removed
and examined for glycogen. An average of 5.5 grams of glyco-
gen was found in these livers, while almost double this quantity
was found in the livers of the control animals.

That bacterial toxins hasten the disappearance of glycogen
from the liver is aiso made probable by the observations of
Luschi? which indicate that the glycogen of the livers of animals
brought to a maximum of glycogen accumulation diminishes
before the expiration of six hours after infection. Colla,? who
studied the effect of infectious diseases on glycogen, found that
glycogen disappears from the liver during tetanus, diphtheria,
anthrax or pneumonia. Healso examined the livers of a number
of children who died of diphtheria. Four, ten or twelve hours
after death the livers were free from glycogen. The results
obtained by Hirsch and Rolly,* however, do not agree with
those just mentioned. After inducing strychnine tetanus (which
was preceded by seven days’ fasting), 3 cc. of a twenty-four hour
bouillon culture of Bacillus coli communis were injected into
rabbits. Their livers as well as their muscles contained appreci-
able amounts of glycogen. Inone rabbit 0.332 gram of glycogen
was found in the liver, which weighed 45 grams. The livers of
the control rabbits were free from glycogen.

That some substances, although not glycogen formers, may,
nevertheless, favor the accumulation of glycogen in the liver
has been indicated by experiments with antipyretics and nar-
cotics. Lepine and Porteret® carried out a large number of
experiments on guinea pigs with antipyrin, acetanilid, quinine
sulphate and sodium salicylate. They found 20 per cent more
glycogen in the livers of these animals than in the controls.
The investigations of Nebelthau corroborated these results.

1 Ott: Deutsch. Arch. j. klan. Med., xxi, p. 267, 1901.

*Luschi: Jahresbericht fir Thierchemie, xxx, p. 449, 1900.

3 Colla: Archiv. Ital. de biologie, xxvi, p. 120, 1896.

* Hirsch and Rolly: Jahresbericht fur Thierchemie, xxxiii, p. 604, 1903.
5 Lepine and Porteret: Compt. rend. de l’acad. de sct., cvi, p. 1023, 1888.

406 Influence of Alcohol on Glycogen-Metabolism

After subcutaneous injection of kairin, antipyrin or quinine
into hens on the fifth or seventh day of fasting, he found 1.81
per cent to 3.66 per cent of glycogen in the livers. His work
with sulphonal, urethane, chloralamid and paraldehyde, injected
on the seventh day into fasting hens, likewise indicated the
accumulation of glycogen in the livers. That paraldehyde exerts
a similar effect in rabbits was shown by Nebelthau' and Cremer.?
Nebelthau' also carried out a series of observations on the effects
of ether, chloroform and alcohol on the metabolism of hepatic
glycogen in hens. The results he obtained with these substances
led him to the conclusion that they favor the accumulation of
glycogen in the liver, although in his experiments with alcohol
he obtained positive results in only four out of eleven experi-
ments.

The discordant results obtained with alcohol by Nebelthau‘
and Kriukoff® furnished the indication for the present study of
the action of that substance on the metabolism of hepatic glyco-
gen. The determination of the action of alcohol in this regard
is especially important also because clinicians frequently pre-
scribe its internal administration in infectious diseases. Colla®
maintained that resistance to bacterial invasion varies with the
amount of glycogen in the liver. Although his results were not
corroborated by the observations of Luschi,’ the truth of this
contention by Colla is made probable by the work of Roger,* who
claimed that the glycogen seems to be an important factor in the
liver in reducing the toxicity of alkaloids and other poisons of
organic nature when these enter the circulation. The con-
clusions of Roger were disputed by Vamossy.® The evidence,
however, which Vamossy brought forward against the view of
Roger is not convincing.

‘ Nebelthau: Zeitschr. f. Biol., xxviii, p. 138, 1891.

? Cremer: Ergebnisse der pe te i, p. 876, 1902.

® Nebelthau: loc. cit.

* Nebelthau: loc. cit.

® Kriukoff: loc. cit.

* Colla: loc. cit.

7 Luschi: loc. cit.

® Roger: Centralbl. f. klinische Medizin, ix, Pp. 5, 1888.

*Vamossy: Archives internationales de pharmaco- -dynamie et de therapie,
Xlil, p. 208, 1904.

William Salant 407

EXPERIMENTAL.

Methods.—The experiments were carried out on full grown
healthy rabbits, between the months of December and May.
Throughout the experimental period each animal was kept in a
metallic cage provided with a wire net bottom and drip pan to
allow drainage of the urine. The room was maintained at a
practically uniform temperature since, as Liithje,1 and later
Amalgia and Embden’ have shown, the metabolism of carbo-
hydrates is influenced by the temperature of the surrounding
atmosphere. Before an experiment the animals were kept under
observation for several days in the cages. If failure of adapta-
tion to the new condition was manifested by any of these rabbits
they were rejected. In this way some assurance was obtained
that the subject of each experiment was normal.

When liver was subjected to analysis, the following proce-
dure was always followed: The animal was quickly killed; its
liver was rapidly removed and weighed and placed at once in
hot 60 per cent potassium hydroxid solution. Glycogen in the
livers was isolated by the shorter method of Pfliger.2 The
amount of glucose obtained from the glycogen by hydrolysis was
determined by Allihn’s method. Later in the course of the
investigation, for reasons of economy of time, the amounts of
copper thrown down by reduction were determined volumetric-
ally by the iodin method. On account of the sharp end point
given by the starch iodin reaction, the latter process was pre-
ferred to the cyanide method that is recommended by some
investigators for the determination of copper.

The introduction of alcohol or glucose into the body was made
per os through a stomach tube.

Sertes I. On fasting rabbits, with small preliminary accumu-
lations of hepatic glycogen. The experiments of several workers,
which showed the accumulation of glycogen in the liver after
the administration of various narcotics and antipyretics, as well
as the observations of Nebelthau* on hens which indicated

1 Liithje: Beitrage zur chemischen Physiologie und Pathologie, vii, p. 309,
1906.

* Amalgia and Embden: zbid., p. 310.

8 Pfliger: Archiv fur die ges. Physiol., xciii, p. 163, 1902.

4 Nebelthau: loc. cit.

408 Influence of Alcohol on Glycogen-Metabolism

similar results in some cases with alcohol, suggested the advis-
ability of carrying out similar experiments with alcohol on other
animals. The possible synthesis of glycogen from alcohol was
also thought of, for the work of Goddard’ on dogs has shown
that when large doses of alcohol are given, an appreciable amount
of aldehyde may be produced in the blood. Again the high calor-
ific value of alcohol and its ready oxidation in the body led to
the expectation that even if alcohol failed to induce an accumu-
lation of glycogen in the liver, its hepatic depletion commonly
observed during fasting might, under the influence of alcohol, be
reduced or perhaps even entirely inhibited. Experiments were
therefore carried out on fasting rabbits, which were given ro cc.
of 30 per cent alcohol per kilo daily for four or five days.

EXPERIMENT IA, FEMALE RABBIT.

Noy. 20, 2 p.m. Weight 1180 grams. Received 10 cc. 30 per cent alcohol per kilo.
“ 21 ‘ 4 “ “ 1 180 “ “ 10 “ae 30 “ “ “
‘ 22 5 2 f 30 “ “ 1 120 “ “ 10 “ 30 “ “ “

Nov. 22, 3.30 p.m. The rabbit was killed. The weight of the liver
was 48 grams. Analysis failed to show the presence ofglycogen.

EXPERIMENT 2A, WHITE FEMALE RABBIT.

Nov. 20, 2. p.m. Weight 1280 grams. Received 10 ce. 30 per cent alcohol per kilo.
“ At 4 “ “ 1210 “ “ 10 “ 320 “ “ “
22 24.30: 5 < 1170 Gi i 10 “ 30 e s re
NG 10’ a.m. 6 1100 g os 10> 49°30 ue 4 -

Nov. 23, 10.30 a.m. The rabbit appeared to be exhausted and dying-
She was killed soon afterwards. The weight of the liver was 35 grams.
In this rabbit also the liver was free from glycogen.

As controls were used two female rabbits (1 and 2, Table I) to which
water instead of alcohol was administered by mouth, through a stomach
tube on four successive days. At the end of this period they were killed
and the content of glycogen in their livers determined.

This analysis, Table I, as well as that for both the alcoholized rabbits,
Table II, failed to show the presence of glyco en in the liver of any one of
these animals.

Our finding indicates, therefore, that alcohol, in the amounts
given, does not favor the accumulation of glycogen in the livers
of fasting rabbits; neither was there any manifestation of the

sparing effect of fats or carbohydrates commonly ascribed to
alcohol.

1 Goddard: Lancet, ii, p. 1132, 1904.

a |

William Salant 409

f:

Series II. On fasting rabbits, with normal preliminary accu-
mulations of hepatic glycogen. Since the foregoing general re-
sult might be due to the presence of only relatively low propor-
tions of glycogen in the livers of these animals, which were fed on
hay, oats and cabbage, previous to the alcohol period, a series of
experiments was carried out in which carrots were fed in large
quantities during the fore period of three daysin order to induce
accumulation of normal amounts of glycogen in the livers of the
animals selected. Alcohol (approximately ro cc. of 30 per cent
alcohol per kilo) was then administered daily for four, five or
six days.

EXPERIMENT 5A, WHITE RABBIT.

Dec. 13 Weight 1300 grams. Received 15 cc. 30 per cent alcohol.

“ 14 “ 1260 “ “ 13 “ec 30 “

“ 15 “ 1210 “ “ 13 “ce 30 “ “
“ 16 “ 1 180 “ “ 25 “ce 30 “ “
“ il 7 “ 1070 “ “ 1 2 “ec 30 “ “
“ 18 “ ] 100 “ “ 1 ae 30 “ “

Dec. 18. The rabbit was killed. The weight of the liver was 44 grams.
The amount of glucose obtained by hydrolysis of glycogen was 0.92 per
cent of the fresh tissue.

EXPERIMENT 6A, WHITE RABBIT.

Dec. 12 Weight 1580 grams. Received 16 cc. 30 per cent alcohol.
30 < “

“ 13 “ 1540 “ “ 15 “ ‘

“ 14 “ 1500 “ “ 15 “ 30 “ “
“ 15 “ 1440 “ “ 15 “ 30 “ “
“ 16 “ 1400 “ “ 25 “ 30 “a “
“ 1 74 a 1 300 “ “ 13 “ 30 “ “
“ 18 “ 1300 “ “ 13 “ 30 “ “

Dec. 18. The rabbit was killed. The weight of the liver was 48 grams.
The amount of glucose obtained by hydrolysis of glycogen was 0.31 per
cent of the fresh tissue.

EXPERIMENT QA, FEMALE RABBIT.

Dec. 24 Weight 1520 grams. Received 15 cc. 30 per cent alcohol.

ae 20 ; 1420 ¥ Sy el)

“ 26 “ 14 1 0 “ “ 1 5 “ 30 “ “
“ Di. “ 1350 “ “ i 3 “ 30 “ “
“ 28° “ 1 300 “ “ 13 “ 30 “ “
“ 99 “ 1280 “ “ 15 “ 30 “ “
“ 30 “ 1220 “ “ 19 “ 30 ne “

Dec. 30, 2.40 p. m. The rabbit was killed. The weight of the liver
was 47 grams. Only a trace of glucose was obtained.

410 Influence of Alcohol on Glycogen-Metabolism

EXPERIMENT 7A, MALE RABBIT.
Dec. 24 Weight 1670 grams. Received 17 cc. 30 per cent alcohol,
« 25 seer VL! aes " L7* SoU gets
Several hours after it received the last dose, the rabbit escaped from the
cage and was found eating hay andoats. The experiment was then con-
tinued, as follows:
Dec. a Ww eight 1620 grams. Received ihe( (fet 4 per cent alcohol.

1580“ bY ag
“ 28 “ 1480 “ “ ] 5 “ 30 “ “
“ 29 “ 1530 “ “ 20) “ 30 “ “
“ 30 “ 1 470 “ “ 19 “ 30 “ “
“ 31 “ 1500 “ “ 15 “ 30 “ “

Dec. 31, 5 p. m. The rabbit was killed. The weight of the liver was
53 grams. The amount of glucose obtained by hydrolysis of the glycogen
was 0.09 per cent of the fresh tissue.

EXPERIMENT 9B, GRAY FEMALE RABBIT.
Dec. 24 Weight 1410 grams. Received 15 cc. 30 per cent alcohol.
= 25 . 1450 . sy EY e810) iw
(At this point there was a fasting period of the same length and for
identical reasons as that for rabbit 7a.)
Dec. 26 Weight 1420 grams. Received 15 ce. 30 per cent alcohol.

eee 1350 “ 14 “ 30

“ 28 “ 1 3 70 “ “ 14 “ 30 “ “
“ 29 “ 1350 “ “ 15 “ 30 “ “
“ 30 “ 1300 “ “ 20 “ 30 “ “
“ 3 l “ l 270 “ “ 14 “ 30 “ “

Dec. 31, 5 p. m. The rabbit was killed. The weight of the liver was
43 grams. Only a trace of glucose was obtained.

EXPERIMENT IOA, GRAY RABBIT.

Jan. 1 Weight 1830 grams. Received 18 cc. 30 per cent alcohol.
“ 2 “ 1690 “ “ 17 “ 30 “ “
“ 3 “ 1540 “ “ 1 5 “ 30 “ce “
“ 4 “ 1 5 ] (3) “ “ 1 5 “ 33 “ “
“ 5 “ 1470 “ “ 15 “ 33 “ “

Jan. 5. The rabbit was killed. The weight of the liver was 53 grams.
The amount of glucose obtained by hydrolysis of the glycogen was 0.02
per cent of the fresh tissue.

EXPERIMENT IIA, GRAY FEMALE RABBIT.

Jan. 1 Weight 1700 grams. Received 18 ce. 33 per cent alcohol.
“6 0 or ss 1560 “ 16-2 333
“ 3 oe 1460 “ “ 15 “ 33 “ “
“ 4 “ 1450 “ “ 15 “ 33 “ “
“ 5 “ 1350 “ “ 14 “ 33 “ “

Jan. 5. Therabbit was killed. The weight of the liver was 54 grams
The amount of glucose obtained by hydrolysis of the glycogen was 0.16
per cent of the fresh tissue.

William Salant 41!

The results obtained in the experiments of this series, which
are also shown in Table II, likewise failed to give evidence of any
inhibitory action of alcohol on the depletion of hepatic glycogen.
Of the three rabbits which received alcohol daily for six days,
the liver in one contained less than o.1 per cent of glycogen; in
each of the livers of the other two rabbits only traces of glycogen
were present at the conclusion of the experiments. In one
rabbit, 8, Table I, which was used as a control and was given
water by mouth through a stomach tube for the same length of
time after preliminary feeding with carrots, the amount of
glycogen in the liver was a little more than 0.04 per cent. In
this connection it may be pointed out that the livers of two
normal rabbits, kept on a diet of carrots for three days, con-
tained 3 to 4 per cent of glycogen at the end of that time.

In two experiments alcohol was given during a period of five days.
Appreciably larger amounts of glycogen were found in the livers
of these rabbits than in the livers of the controls (see Tables 1
and 2, p. 416), which would seem to indicate that alcohol caused
retarded transformation of glycogen. This was improbable,
however, in the light of the results of Experiments 10a and
11a. Inthe latter experiments, very small amounts of glycogen
were found in the livers of these rabbits, which received alcohol
on four days after the usual preliminary feeding of carrots. It
was thought, however, that possibly larger quantities of alcohol
and consequently an increase in the number of calories might
spare the hepatic glycogen. Two experiments were carried out
to test this suggestion (Series III).

Series III. Same as Series II, with larger doses of alcohol.
EXPERIMENT I7A, WHITE RABBIT.

Fasted for six days, was then fed carrots on four days (Jan. 30 to Feb. 3.)
Feb. 3 Weight 1800 grams. Received 20 cc. 30 per cent alcohol.

fos pate 20 “ 60
“ 5 1 y F 30 a.m. “ 10 “ 60 “ “
“ 5 3 “ “ 10 “ 60 “ “
“ 6 10 a m. “ 10 “ 60 “ “
“ 6 - 3 p m. “ 12 “ 60 “ “
« 712 noon AS) 8? 60 ra A «

At 3.30 p. m. the rabbit was found dead, but was still warm. The
weight of the liver was 66 grams. A qualitative test failed to show the
presence of glycogen in the liver.

412 Influence of Alcohol on Glycogen-Metabolism

EXPERIMENT 158A, GRAY RABBIT.

Fasted six days. Carrots were given on four days. (Jan. 30 to Feb. 3).

Feb. 3 Weight 1130 grams. Received 15 cc. 30 per cent alcohol.
“ 4 3 ” “

p.m. Rj >= 160
: 5 11.30 a.m. . Bo ae) Me Le
“"b* 22-380) pm. 7 (gla « 2
o) (Gs1G a.m. . SOO Z fe
o* i Gide p.m. “ 3, “ 60 : %
es ly Sak be noon my LO) OC id i
“ 7 8.30 p.m. * 40,* 60 * a

Feb. 7, 10.30 p. m. The rabbit was killed. The weight of the liver
was 45 grams. As in the preceding experiment the liver was free from

glycogen.

The results of analysis in these experiments (Series III) like-
wise indicate that even somewhat larger quantities of alcohol
than those of the previous experiments failed to inhibit the dis-
appearance of glycogen from the livers of fasting rabbits, this
organ in each rabbit having been found free from glycogen.
The question whether alcohol may accelerate the transformation
of glycogen in the liver now presented itself with special force.

Series IV. Does alcohol accelerate the transformation of
hepatic glycogen. To answer this question various amounts
of glucose were given to rabbits by mouth through a stomach
tube. Alcohol was administered either immediately after the
glucose was given or, as in some experiments, a few hours were
allowed to elapse between the last feeding of glucose and the
succeeding dose of alcohol. In some experiments two doses of
alcohol were administered at intervals of eighteen to twenty-
four hours. In one experiment (28a), only one dose was given
while another rabbit (29a) received three doses of alcohol. As
controls I used rabbits which were fed varying amounts of glu-
cose. The rabbits were killed at the end of different periods
following the administration of glucose.

Experiment 24a, white rabbit. Fasted 6 days. Weight, 1800 grams.

March 19, 10 grams of glucose dissolved in water were given by mouth.
Immediately afterward the animal received 10 cc. of 60 per cent alcohol.

March 20, 10 grams of glucose were fed, then r5 cc. of 60 per cent alcohol
were given.

About sixteen hours later the rabbit was killed. The weight of the
liver was 52 grams. The amount of glucose obtained by hydrolysis of
glycogen was 1.9 per cent of the fresh tissue.

William Salant AlZ

Experiment 24b, black rabbit. Weight, 1850 grams. Fasted 6 days.

March 19, 10 grams of glucose given by mouth.

March 20, 10 grams glucose given by mouth.

Sixteen hours later the rabbit was killed. The weight of the liver was
51 grams. The amount of glucose obtained by hydrolysis of glycogen
was 4 per cent of the fresh tissue.

Experiment 25a, rabbit. Weight, 1550 grams. Fasted 6 days.

May 9, 3-30 p-m. Eight grams of glucose given by mouth.

May 9, tiop.m. Received 15 cc. of 60 per cent alcohol.

May 10, 11a.m. Weight, 1500 grams. Received 15 cc. of 60 per cent
alcohol.

The rabbit was killed May 11 at 3 p.m. About 15 minutes after
alcohol had been given, symptoms of intoxication appeared which lasted
several hours. At the time of death, the rabbit looked normal. No
glycogen was present in the liver.

Experiment 25. Rabbit fasted 6 days. May 10, io a.m. Weight
18s5ograms. Nine gramsof glucose were given by mouth. May 1o, 3 p.m.
Water was given by mouth. The rabbit was killed May 11 at 7 p.m.
The weight of the liver was 33 grams. The amount of glucose obtained
by hydrolysis of glycogen was 1.23 per cent of the fresh tissue.

Experiment 26a. Weight of rabbit, 1450 grams. Fasted 6 days.

May 9, 3.30 p.m. Seven grams of glucose, dissolved in water, were
given by mouth.

May 9,10p.m. Fifteencc. of 60 per cent alcohol were given by mouth.

May 10, 11 a.m. Weight, 1300 grams. Received 13 cc. of 60 per
cent alcohol.

Shortly afterward, the rabbit manifested symptoms of severe intoxica-
tion which continued all day. While still under the influence of alcohol,
May 10, 7 p.m., the rabbit was killed. The amount of glucose obtained
by hydrolysis of the glycogen was 3.33 per cent of the fresh tissue.

Experiment 27a. Rabbit fasted 6 days.

Mayo9,4 p.m. Weight, 1200 grams. Six grams of glucose were given
by mouth.

May 9, to p.m. Received 15 cc. of 60 per cent alcohol.

May to, 11a.m. Weight, 1100 grams. Received 11 cc. of 60 per cent
alcohol.

May 10,7 p.m. The rabbit was killed. Signs of intoxication developed
ten minutes after the administration of alcohol. At the time of death
the rabbit looked normal. The weight of the liver was 36 grams. The
liver of this rabbit was free from glycogen.

Experiment 28. The rabbit fasted 6 days.

May 27,6. p.m. Weight, 1450 grams. Fifteen grams of glucose dis-
solved in water were given by mouth.

May 28, 12 noon. Weight, 1450 grams. Fifteen grams of glucose were
given as before.

May 30. The rabbit struggled a good deal when he was fastened on
the holder.

414 Influence of Alcohol on Glycogen-Metabolism

May 30,6 p.m. Killed. The weight of the liver was 30 grams. The
amount of glucose obtained by hydrolysis of the glycogen was 0.11 per
cent of the fresh tissue.

Experiment 29a. Rabbit fasted 6 days.

May 27,6 p.m. Weight, 1450 grams. Fifteen grams of glucose were
given by mouth.

May 28. Weight, 1450grams. Fifteen grams of glucose were given as
before.

May 28, 6.15 p.m. Received 15 cc. of 60 per cent alcohol.

May 29, 5 p-m. Received ro cc. of 60 per cent alcohol.

May 30, 11 a.m. Weight, 1450 grams. Received 1o cc. of 60 per
cent alcohol.

May 30, 4.30 p.m. The rabbit was killed while apparently still under
the influence of alcohol. The weight of the liver was 44 grams. Analysis
did not show the presence of glycogen.

Experiment 30a. Rabbit fasted 6 days.

May 27,6 p.m. Weight, 1270 grams. Received 13 grams of glucose
dissolved in water.

May 28, 12.15 p.m. The same amount of glucose was given.

May 29, 5.30 p-m. Received 12 cc. of 60 per cent alcohol.

May 30, 11 a.m. Received 15 cc. of 60 per cent alcohol.

May 30, 4.30p.m. The rabbit was killed while still under the influence
of alcohol. This liver was likewise free from glycogen.

Experiment 31a. Rabbit fasted 6 days.

May 27, 6. p.m. Weight of rabbit, 1800 grams. Eighteen grams
of glucose were given. This was repeated next day. Fifty-two hours
later the rabbit was killed. The liver was removed and treated as in the
other experiments. The weight of the liver was 28 grams. The amount
of glucose obtained from the glycogen was 1.24 per cent of the fresh tissue.

Experiment 28a. Rabbit fasted 6 days. Weight, 1230 grams.

May 27 and 28. Thirteen grams of glucose were administered as in
previous experiments.

May 28, 6 p.m. Received 13 cc. of 60 per cent alcohol.

May 29,6 p.m. The rabbit was found in a dying condition. It was
then killed. The liver was removed immediately and subjected to
analysis by the usual method. Weight, 45 grams. The amount of
glucose obtained from the glycogen was 1.24 per cent of the fresh tissue.

The analytic results given in the table (III) show with one
exception (No. 26a) a marked diminution in the glycogen of the
livers of the alcoholized rabbits as compared with the controls.
Thus, sixteen hours after the last feeding of glucose the total
amount of glycogen obtained was 3.7 per cent in the control (24),
while in the rabbit given alcohol (24a), which received the same
amount of glucose (20 grams in twenty-four hours), the quantity

——_ a

9 ee ee ee se eee

William Salant A415

of glycogen found in the liver was 1.76 per cent, less than half
that found in the control.

In the next group each rabbit was givena single dose of 5 to
g grams of glucose per kilo. They were killed twenty-eight hours
later. Excepting rabbit 26a this series showed in a striking man-
ner the action of alcohol in hastening the transformation of
hepatic glycogen, for, whereas the amount of glycogen in the
controls was 1.14 per cent (25) and 2.3 per cent (26), the livers
of the two corresponding alcoholized rabbits were free from
glycogen. The administration of alcohol after feeding much
larger quantities of glucose (20 gm. per kilo in twenty-four hours)
was accompanied by similar results when the rabbits were killed
in from fifty-two to fifty-four hours after they received the final
dose of glucose. In the two controls of this group the amount
of glycogen obtained was 0.11 per cent in one (28), and 1.14 per
cent (31a) in the other, whereas the livers of the two alcohol
rabbits were free from glycogen. In another experiment of this
group (28a) the rabbit received the same amount of glucose per
kilo in twenty-four hours, but, as he was in a dying condition,
apparently from the effects of alcohol, he was killed thirty hours
after he had received the final dose of glucose. The liver of this
rabbit contained only 1.33 per cent of glycogen, which is only
0.2 per cent more than that found in the control rabbit (25) and
about 1 per cent less than in rabbit 26, which received one-
fourth the amount of glucose per kilo fed to rabbit 28a.

The discrepancy between the results obtained for rabbits 26a,
and in the other alcohol experiments of the same series, may
be explained by the following observations: It was noted that
in this rabbit symptoms of severe alcohol intoxication (which
set in soon after the second and final dose of alcohol was given)
persisted eight hours, while the other rabbits of this group simi-
larly treated behaved at the end of this time as if they had recov-
ered completely from the effects of alcoholintoxication. This dif-
erence in the reaction to alcohol is suggestive of the possibility that
during the stage of profound alcohol narcosis, glycogen metabolism
is not affected. Perhaps it is only when recovery from narcosis
is practically complete that metabolism of glycogen is resumed.

In the two alcohol rabbits 29a and 30a, each of which was
killed in a state of deep intoxication, this supposition apparently

416 Influence of Alcohol on Glycogen-Metabolism

TABLE I. AMOUNTS OF GLYCOGEN IN THE LIVERS OF CONTROL RABBITS.

ongad
| Beige
toe Cun
: ‘ c Pasti Treat t a
— Rabbit. Liver. ea ih | seetyag duis fasting! 3 ae 238
| | 26.822
o & os Sto
q
et SS Mas ot ee, eee eee
gms gms. days ;
1 820 22 Cori0-* 4 (Water given none
2 1320 | 41 = 5 by stomach «
5 1980 || 27 Carrots 5 tube. 0.139
| 3 days.
6 | 970 | 22 a 5 0.148
8 | 1370 42 MS 6 0.043
10 1265 | 38 4 0.127
11 1470 53 a 4 none

TABLE II. AMOUNTS OF GLYCOGEN IN THE LIVERS OF FASTING RABBITS
AFTER ADMINISTRATION OF ALCOHOL.

l l
| | Seese
| | bari BE
Oe ans
ae? | piapeie (| ave, ope Beton |. Banting |), Atechot pes 292058
coe Re Pee
so]
gms. gms. days | 30% incc.
la 1120 48 G80 4 10 none
2a 1100 35 a 5 10 S
5a 1100 44 Carrots
3 days 5 10 | 0.84
6a 1300 48 - 5 10 0.28
9a 1280 47 oi 6 10 trace
7a 1500 50 3 6 10 0.083
9b | 1270 43 4 6 10 trace
10a 1470 53 < 4 10 0.018
lla 1350 54 . 4 10 0.148
3a 800 35 Om) s HO 4 12 trace
60 % ince
17a 1800 66 | Carrots 34 12 none
3 days
18a | 1130 45 3 4 15 none

*C. H. O.—Cabbage, hay, oats.

was not borne out by the results of the analysis. As shown in
the table, their livers did not contain any glycogen. The proto-
cols show, however, that the intervals between the two successive
doses of alcohol in each case were eighteen or twenty-four hours,
thus giving sufficient time for recovery from its intoxicating effect.
In rabbit 26a, the interval was only twelve hours, which may ac-
count for the different results obtained.

William Salant 417

TABLE III. SHOWING THE EFFECTS OF COMPARATIVELY LARGE QUANTI-
TIES OF ALCOHOL ON THE METABOLISM OF HEPATIC GLYCOGEN.

| ongss
E Gt fed Nae Icohol No, hours) rae
e + ; | ucose fed | ; ies GH os
Ne: Rabbit. Liver. | per kilo in | oes : Inst fesdiiel 2 8.9 ° Z
| 24 hours. | daily. of glucose | $37 25 g
| and death. Bon 2 ae
gms. gms. gms. Ce:
24 | 1850 Bl 20 0 164 bo 70
24a Te 52 | 20 25 £6) hero
| Sete 9 0 28 1.14
Boa lool 48 | 8 30 28 |
26a | 1450 58 7 28 28 3.08
26} 1020 A 54 0 a8 2.30
27a | 1200 36 6 26 58° ¢|
28 1450 S04 29 0 BAT ah POS
29a | 1450 44 | 29 35 52 |
30a | 1270 Ziee| .26 27 52.
3la | 1800 28 | 36 0 52 | 1.14
28a 1230 45 | 25 13 30 | eo
2la 1050 41 32 20 12 8.65
23a | 815 35 25 | 12 i 9.17

The suggestion that alcohol does not hasten the transforma-
tion of glycogen of the liver during the stage of profound narcosis
gains further support from the results of experiments 21a and
23a, in which large quantities of glucose were followed by
alcohol (shortly after, in one experiment, or three hours later,
in another). The amounts of glycogen found six to twelve hours
after the final dose of glucose had been given were 9.17 per cent
and 8.65 per cent, respectively.

It is to be noted that in both these experiments the rabbits
were killed from six to eight hours after receiving large amounts
of alcohol in proportion to body weight, that is, before the stage
of intoxication passed off. The conclusion seems to be justified,
therefore, that large quantities of alcohol may hasten the proc-
ess by which glycogen is made to disappear from the liver and
that it apparently exerts this action only after the stage of
intoxication has been passed. This mode of action of alcohol
might explain the discordant results of Nebelthau’s experiments,
some of which showed disappearance of hepatic glycogen, while
others indicated the presence of considerable amounts, after
administration of alcohol. It may in the same way also account

418 Influence of Alcohol on Glycogen-Metabolism

for the different results obtained by Nebelthau and Kriukoff,
whose work was referred to on p. 406.

Objection to this conclusion may be raised on the ground that
there is a wide range of variation in the proportion of glycogen
found in the livers of the control rabbits. Thus in experiment
28a, only o.11 per cent was obtained, while in 31a, the pres-
ence of 1.14 per cent of glycogen was shown. This difference
may possibly have been due to the activity of rabbit 28a, when it
was tied on the holder before it was killed. The other rabbits
did not behave in this way. Again, in experiments 25 and 26
(two control rabbits) there was a difference of 50 per cent in the
liver glycogen found; as great a difference as there was between
the alcohol rabbit in experiment 24a, and the corresponding
control. The wide variation in the amount of hepatic glycogen
of these two normal rabbits does not invalidate the above con-
clusions, since in the alcoholized rabbits of the same group (25a
and 27a), which received as much glucose per kilo as the con-
trols, and were killed at the same time after feeding glucose, the
livers were free from glycogen. Moreover, the difference between
the amounts obtained from the two normal rabbits may possibly
be accounted for by a difference in muscular activity. It is of
importance to note, in this connection that, in the alcoholized
rabbits, intoxication and consequent loss of muscular power
set in about 15 minutes or sometimes even earlier after alcohol
was administered, which would tend to inhibit rather than
hasten the amylolysis.

That other toxic substances may exert an action on hepatic
glycogen comparable to that here attributed to alcohol was
shown by Roger.' He was the first to find that anthrax has
no effect on glycogen during the first stage of infection, but later
accelerates its disappearance from the liver.

I am very much indebted to Prof. William J. Gies for valuable
suggestions received in the course of this investigation.

* Roger: Arch. de physiol. norm. et pathol., 5th series, vi, p. 64, 1894.

ON THE PRODUCTION OF PHENOLIC ACIDS BY THE OXI-
DATION WITH HYDROGEN PEROXIDE OF THE AM-
MONIUM SALTS OF BENZOIC ACID AND ITS DERIV-
ATIVES, WITH SOME REMARKS ON THE MODE
OF FORMATION OF PHENOLIC SUB-
STANCES IN THE ORGANISM.

By H. D. DAKIN AND MARY DOWS HERTER.

(From the Laboratory of Dr. C. A. Herter, New York.)
(Received for publication, September 14, 1907.)

While continuing the investigation of the action of hydrogen
peroxide upon various amido-acids,' the attempt was made to
isolate the products of the oxidation of phenylalanin. Judging
from analogy with aliphatic amido-acids, such as alanin which
yields acetaldehyde, acetic acid, ammonia and carbon dioxide,
it was anticipated that the products would be phenylacetalde-
hyde, phenylacetic acid, ammonia and carbon dioxide.

COOH Oo H O OH
Ww \

CH (NH,) C C
ke ch tie
|

Vas VON Va \

| | — NH,+ CO, + | | — |

VA WA Ne

Although undoubted indications of the formation of small
quantities of phenylacetaldehyde were obtained, showing that
the reaction, at least in part, had proceeded as was anticipated,
it was found that acid products were formed at the same time,

1 Dakin: This Journal, i, p. 171, 1906.

419

420 Oxidation of Benzoic Acids

which were certainly not exclusively composed of phenylacetic
acid. :

Subsequent investigation showed that phenolic acids were
present, for the product gave a violet-blue coloration with ferric
chloride and also a strong reaction with Millon’s reagent, thus
proving that oxidation had occurred in the nucleus as well as in
the side-chain. The fact that it had been possible to effect the
introduction of one or more hydroxyl groups into an aromatic
nucleus by oxidation of an amido-acid at the ordinary tempera-
ture at once suggested that the reaction had some biological
significance, for it is known that similar changes occur in the
animal organism, as exemplified in the formation of homogentisic
acid from phenylalanin and tyrosin in alkaptonuria and in the
conversion by the organism of benzene into phenol, pyrocatechin
and hydroquinone.! The formation of adrenalin, which contains
the catechol nucleus from any of the known aromatic substances
ordinarily supplied to the animal organism (phenylalanin, tyro-
sin and tryptophan) would necessitate similar oxidations:

HO

¢ eH, cH NH,-COOH -> a cH, Coon

~ OH

A preliminary attempt to determine more exactly the nature
of the products formed by the oxidation of phenylalanin showed
that the reaction was a somewhat complicated one, in which the
first stage consisted in the formation of phenylacetaldehyde
(phenylacetic acid) ammonia and carbon dioxide. In order to
obtain information as to the further oxidation products, it was
decided to study the action of peroxide upon simpler nitrogen-
free acids. It was found that hydrogen peroxide could effect
the introduction of hydroxyl groups into the benzene nucleus of

* Schultzen and Naunyn: Arch. f. Physiol., p. 349, 1867; Baumann and
Preusse: Zeitschr. f. physiol. Chem., iii, p. 156; vi, p. 190.

H. D. Dakin and Mary Dows Herter 421

a series of aromatic acids, such as benzoic acid, phenylacetic,
phenylpropionic and their substitution derivatives. The ques-
tion of the oxidation products of phenylacetic and phenylpropi-
onic acids will be discussed ina future paper. The present com-
munication consists in an account of the oxidation of benzoic
acid and some of its substitution derivatives.

It was found that if benzoic acid be dissolved in water con-
taining hydrogen peroxide and a slight excess of ammonia and
the mixture allowed to stand at the ordinary temperature of the
air, it was easy to demonstrate the formation of salicylic acid.
The salicylic acid may be detected by simply acidifying, extract-
ing with ether and applying the ferric chloride or other test to
an aqueous solution of the ethereal extract. Further investiga-
tion showed that salicylic acid was not the only product of the
reaction but para- and meta-oxybenzoic acids were formed at the
same time and in not widely differing amounts.

The salicylic acid was separated from the other oxyacids by
means of its solubility in chloroform. The free acid was obtained
from the sparingly soluble basic calcium salt and was further
identified by conversion into tribromophenol bromide, melting
point 131°, and tribromophenol, melting point 93—95°. Themeta-
oxybenzoic acid on treatment with bromide water, gave 2,4,6-
tribromo-3-oxybenzoic, melting point 145-146°, while the para-
oxybenzoic acid was separated in the form of its very sparingly
soluble basic barium salt, from which the pure acid was easily
obtained, melting at 210-211°. It was also converted into
tribromophenol bromide and tribromophenol. The details of
the separation will be found in the experimental part of this
paper.

Further investigation showed that a second hydroxyl group
could be introduced into the mono-oxybenzoic acids by treat-
ment of their ammonium salts with hydrogen peroxide. In
each case the second hydroxyl took up a position ortho. to the
first (OH) group. The 2,3-dioxybenzoic acid and 3,4-dioxy-
benzoic acid were both obtained in the pure crystalline state,
melting at 202° and 199-200° respectively. The following scheme
represents the progressive oxidation of benzoic acid:

422 Oxidation of Benzoic Acids

COOH
3
avg
L N
|
COOH COOH COOH
\/0H Iw
OH
1 \ vA
COOH COOH

The action of the hydrogen peroxide in effecting the introduc-
tion of hydroxyl groups into the aromatic nucleus of benzoic
acid and its derivatives seems to be a general reaction, for experi-
ments carried out with a series of chlor-, brom-, nitro-, dinitro-,
and amido-benzoic acids, all resulted in the production of pheno-
lic acids, although the yields appear to be in all cases very small.

The simultaneous production of the three isomeric oxybenzoic
acids is of some interest from several points of view. In the
first place the simultaneous production in considerable amounts
of ortho, meta and para derivatives is not very common. It
might be expected that the direct substitution of a benzene
derivative, C.H:X, where X is anacidic group, such as COOH,
would result in the formation of a meta derivative, and although
this is true for most substituents, yet in the case of the direct
introduction of an (OH) group the tendency to form ortho and
para derivatives is at least as strong. In the case of the dihy-
droxy-acids the position of the second entering (OH) group
appears to be governed mainly by the position of that already
present in the ring, the second (OH) going into the ortho (and
para) position to the first. The reaction furnishes an additional

‘Hiibner: Ber. d. deutsch. chem. Gesellsch., viii, p. 873; Nolting: ibid., -
ix, DP. 1707.

H. D. Dakin and Mary Dows Herter 423

exception to Brown and Gibson’s generalization,! which states
that if a benzene substituent HX is convertible by direct oxida-
tion into a compound XOH, then substitution will result in the
formation of meta derivatives. If, on the other hand, direct
oxidation is not practicable, ortho and para substitution deriva-
tives will be produced. It is interesting to note that the re-
verse change to that under consideration, namely, the introduc-
tion of carboxyl groups into phenol, results in the formation of
acids in which the carboxyl takes up a position ortho and para to
the hydroxyl group.

OH OH OH OH
can ne (coor f Ye = Hooc/ \cooH
VV BO ei a

COOH
OH OH
es coon Hooc “ \cooH
COOH COOH

Excluding reactions taking place in the living organism, we
have thus far been able to find only two other cases of the direct
introduction, by oxidation, of hydroxyl groups into an aromatic
acid” not already containing hydroxyl groups. Both of these
are reactions occurring at high temperatures and therefore posses-
sing limited biological significance. Thus Etting showed that
copper benzoate on heating to 275° yields some copper salicyl-
ate,? while Barth and Schreder determined the pressure of oxy-
benzoic acidsamong the products of the fusion of benzoic acid
with potash.‘

A few instances are recorded of the direct introduction of the
(OH) group into the ring of aromatic hydrocarbons. Thus ben-
zene has been directly oxidized to phenol by means of ozone,’

1 Journ. Chem. Soc., \xi, p. 368.

2 Ann. d. Chem., liii, pp. 88, 91.

3 Monatsh. f. chem., 111, p. 799.

*Nencki and Giacosa: Zettschr. f. physiol. Chem., iv, p. 339.

5 See, however, the action of potassium persulphate upon hydroxy-
cids referred to later.

424 Oxidation of Benzoic Acids

hydrogen peroxide,! by air in the presence of water and palladium-
hydrogen,?, or copper or iron salts*—reactions which probably
depend on the production of hydrogen peroxide—also by the action
of sunlight in the presence of caustic soda‘ and finally by the
combined action of oxygen and aluminum chloride.°

The production of phenol by the oxidation of benzene with
peroxide of hydrogen has been questioned by Kingzett® but the
original statement has been substantiated by the work of Cross,
Bevan and Heiberg,? who showed that not only phenol but
catechol and hydroquinone were produced simultaneously when
benzene was digested at low temperatures with peroxide of hydro-
gen and a trace of an iron salt. Hydroquinone has also been
obtained by Gattermann and Friedrichs through the electrolytic
oxidation of benzene in the presence of alcoholic sulphuric acid.®
Although there can be no doubt of the power of peroxide of
hydrogen to oxidize benzene, the reaction seems to have received
no extended application to other compounds.

The only other reaction involving the introduction of hydroxyl
groups into the benzene nucleus that will be referred to is an
extremely interesting series of oxidations carried out with potas-
sium persulphate, which are comprised in a number of Patent
Specifications by Schering & Co. Thus the action of potassium
persulphate in alkaline solutions upon phenols® or upon oxyben-
zoic acids results in the formation of peculiar sulphur-containing
products which on treatment with acids yields dihydroxy deriva-
tives. Salicylic acid, for example, yields hydroquinone carbox-
ylic acid,’ (2,5-dioxybenzoic acid) while para-oxybenzoic acid
yields protocatechuic acid" (3,4-dioxybenzoic acid).

‘Leeds: Ber. d. deutsch. chem. Gesellsch., xiv, p. 975.

* Hoppe-Seyler’s: Zeitschr. f. physiol. Chem., p. 1552, 1879.
* Nencki and Sieber: Journ. f. prakt. Chem., xxvi, p. 25.

* Radzisgewski: Journ. f. prakt. Chem., xxiii, p. 96.

* Friedel and Crafts: Compt. rend. de biol., \xxxiv, Pp. 1460, 1879.
*Chem. News, xliv, p. 229.

7 Ber. d. deutsch. chem. Gesellsch., xxxiii, p. 2018.

* Ibid., xxvii, p. 1942.

* E. Schering: D. R. P. 81068 (1894).

” Tbid., 81297 (1894).

1 Tbid., 81298 (1894).

H. D. Dakin and Mary Dows Herter 425

By employing potassium persulphate in sulphuric acid solu-
tion, for the oxidation of oxybenzoic acids, A. G. Perkin and his
co-workers' have obtained substances resulting from the con-
densation of two molecules of the oxidation products of oxy-
benzoic acid. Thus para-oxybenzoic acid yields catellagic acid
(CuH-O.). Similar results are obtained by electrolytic oxida-
tion of the oxybenzoic acids in sulphuric acid solution.

A considerable amount of evidence is gradually accumulating
to emphasize the fact of the remarkably close similarity between
oxidations effected by peroxide of hydrogen and those occurring
in the organism, although it is not suggested that peroxide of
hydrogen is the active agent in tissue oxidations but merely that
the two types of reaction have much in common. It was there-
fore a matter of considerable interest to find that peroxide of
hydrogen, acting at low temperatures, could effect the intro-
duction of hydroxyl groups into the nucleus of aromatic acids,
for this is a reaction which, though doubtless occurring in the
animal organism, cannot be imitated by the more usual oxidizing
agents.?, Attempts to imitate in the laboratory reactions occur-
ring in the organism are of value not only in confirming results
already arrived at from a study of animal metabolism but also
by affording grounds for speculation as to the mode of origin of
other substances. For example, although no physiological evi-
dence is at present available as to the mother substances of
adrenalin,* yet the origin of this substance from tyrosin or phe-
nylalanin can be easily pictured as taking place by a series of
oxidations, entirely similar to those capable of being effected
with peroxide of hydrogen. If the amido group of tyrosin or

1 Proc. Chem. Soc., xxi, p. 185; Trans. Chem. Soc., 1xxxix, p. 251.

2 Attempts to oxidize ammonium benzoate, at low temperatures (50°)
with sodium peroxide, magnesium peroxide, barium peroxide, and potas-
sium persulphate and a mixture of potassium persulphate and silver
oxide in no case gave results in any way comparable with those obtained
with peroxide of hydrogen.

3 Experiments which have been hitherto published upon the formation
of adrenalin from tryptophan can hardly be considered convincing.

426 Oxidation of Benzoic Acids

oO OH O OH O OH
VA YY bee
GC. C Gc
|
HC:-NH-R HC:-NH-R HC-NH:-R
|
CH, CH, CHOH

—>
HO! HO!
OH OH
I II Ill

phenylalanin be assumed to be attached to some other carbon
chain, such as an amido-acid group, as in the polypeptides,
the tyrosin or phenylalanin grouping might well be protected
from oxidation at the a carbon atom, while still being capable
of undergoing nuclear and other oxidations. An example of
this protection is seen in the fact that, although glycocoll is
readily oxidized by hydrogen peroxide, yet if one of the hydrogen
atoms of the amido group be substituted by an acid radical, e. g.,
benzoyl, as in hippuric acid, the product is attacked with great
difficulty. Nuclear oxidation of the phenylalanin or tyrosin,
with the introduction of two hydroxyl groups in positions (3)
and (4) is easily intelligible from the similar formation of proto-
catechuic acid (3,4-dioxbenzoic acid) from benzoic acid, meta-
oxybenzoic acid and from paraoxybenzoic acid. Oxidation in
the side chain, with introduction of an (OH) group in the P
position (III) is a reaction with many parallels, for not only
is B-oxidation common in the tissue oxidations of fatty acids,
but actual examples of this- have been observed in the oxi-
dation of fatty acids, e. g., butyric acid, with peroxide of
hydrogen.? By removal of carbon dioxide the tyrosin deriva-
tives would differ from adrenalin only as regards the nitrogen
group. If the nitrogen were attached to the carboxyl group of
the amido-acid radical, hydrolysis would result in the formation

1 Dakin: This Journal, i, p. 271, 1906.
? An account of these experiments will appear in the next number of
this Journal.

H. D. Dakin and Mary Dows Herter 427

of a product with an (NH,) group instead of an NHCH, group,
as is present in adrenalin. It is known, moreover, that the
body can effect methylations, so that the substance might be
converted into adrenalin in this way. If, however, the amido-
group in the tyrosin molecule were originally combined with an
oxyamido-acid grouping such as exists in serin

© . OH
Va
C
els indy: ve
H—-C—NH HOCH,-CH-NH,.cC
oe Cpe
C
va
O OH
We

=e © NH CH,-CHNEH, COOH

O OH
4

— H=C—NH-CH, + NH, +260,

€
VAIN
oxidative changes might well result in the formation of a

Nc.NHCH,

ee

group. It is not impossible that the source of the methyl
groups formed in the body is frequently due to such oxyamido-
acid condensations. Many examples might be quoted of the
analogy between peroxide oxidations and those occurring in the
organism, but these will be reserved for future discussion when
more experimental results have been obtained.

It is interesting at this point to consider briefly some possible
modes of formation of phenolic substances in both the animal

428 Oxidation of Benzoic Acids

and vegetable organism. In the latter, especially, they figure
prominently as is seen from the extremely extensive list of
phenols and phenolic-alcohols, -aldehydes, -ketones, -acids, tan-
nins, etc., which are obtained from vegetable sources. There
are two obvious processes by which phenolic substances may be
produced, namely:

(1) Hydroxylation of preformed aromatic substances.

(2) Direct synthesis of phenols and their derivatives from
aliphatic substances.

There is abundant proof that the first type of reaction com-
monly occurs in the animal organism, and several examples have
already been quoted, e. g., the production of phenol (Schultzen
and Naunyn) and of catechol and hydroquinone from benzene,’
the conversion of phenylalanin, tyrosin and oxyphenyllactic

9

acid? into the homogentisic acid in alkaptonuria, the formation
of para-amidophenol from anilin,’ of acetylparamidophenol from
acetanilide,‘ and many other examples.’ But there is no evi-
dence pointing to a synthesis in the animal body of phenols from
aliphatic substances.®

1 Baumann and Preusse: Zeitschr. f. physiol. Chem., iii, p. 156; Vi, p. 190.

2 Neubauer and Falta: ibid., xlii, p. 81, 1904.

3 Fr. Miiller: Deutsch. med. Woch., 1887.

‘ Jaffe and Hilbert: Zeitschr. f. physiol. Chem., xii, p. 295; K. Morner,
tbid., Xiii, p. 12.

’ The following is a list of substances which in their passage through the
animal organism undergo nuclear substitution by hydroxyl groups and
are converted, at least in part, into mono- or poly-hydric phenols: Ben-
zene, chlorobenzene, bromobenzene, p-dichlorobenzene, m-dichloroben-
zene, m-cymene, isopropylbenzene, isobutylbenzene. methylethyltoluene,
mesitylene, diphenyl, p-dibromdiphenyl, diphenylmethane, naphthalene,
phenol, o-cresol, anisol, phenetol, guaiacol, thymol, phenylalanin, tyrosin,
oxyphenylactic acid, anilin, formanilide, acetanilide, o-toluidin, o-acet-
toluidide, phenylurethane, carbonyl-o-amido-phenol, carbazol, acridin.

* A number of such syntheses have been accomplished outside the body
but they are mostly of a kind such as could scarcely have any analogy
in the living cell. Thus, hydroquinone and hydroquinone-dicarboxylic
acid are formed by the action of potash upon succino-succinic ester, while
small quantities of quinhydronedicarboxylic acid are formed by boiling
free succino-succinic acid with water (Baeyer). Hydroquinone is also
produced in the distillation of salts of succinic acid (Richter, Journ.
fj. prakt. Chem., 2, xx, p. 207). Meta-oxyuvitic acid, C,H,(CH,)(OH)-
(COOH),, is produced by the action of chloroform upon the sodium

H. D. Dakin and Mary Dows Herter 429

In the plant cell, on the other hand, there is less evidence
available, although Gonnemann’s observation of the formation
of homogentisic acid as an intermediate product in the action
of tyrosinase upon tyrosin 1s an example of the first type of
reaction, involving the hydroxylation of preformed benzene
derivatives. Low! and later Emmerling and Abderhalden? have
observed the formation of protocatechuic acid by the action of
fungi upon quinic acid (tetraoxyhexahydrobenzoic acid). Since
the formation of quinic acid from carbohydrate is conceivable—
although not yet demonstrated—it may be that this represents
a direct synthesis of a phenol from aliphatic substances.

On the whole, it seems most probable that in both animal and
plant life, phenolic substances are mainly, although possibly not
exclusively produced by the hydroxylation of preformed benzene
derivatives and not by direct synthesis from aliphatic substances.
In other words, the introduction of hydroxyl groups into the
nucleus of aromatic substances is a typical ‘“biological’’ reaction.

EXPERIMENTAL PART.

Oxidation of Benzoic Acid. Benzoic acid (1 mol.) was sus-
pended in a small quantity of warm water, and dissolved by
addition of a slight excess of ammonia. Dilute neutral approxi-
mately 3 per cent hydrogen peroxide, the strength of which
had been determined by titration with permanganate, was added
in quantity equivalent to 14 molecules of peroxide. Although
the reaction proceeds at ordinary temperature, it is acceler-
ated by heat, and accordingly the mixture was gently boiled

derivative of aceto-acetic ester (Oppenheim and Pfaff, Ber. d. chem. Gesell-
sch, Vii, p. 929: Oppenheim and Precht, zbid., ix, p. 321) while phloroglucin-
tricarboxylic ester is obtained by heating malonic ester with sodium
(Baeyer, zbid., xviii, p. 3457) or with zinc alkyl derivatives (Lang, zbid,, xix,
p. 2038). Collie and Myers (Trans. Chem. Soc., \xiii, p. 124) haveobtained
orcin (= 3,5-dioxytoluene) by the action of alkali upon dehydracetic
acid or dimethylpyrone while Berthelot (Compt. rend. de l’ Acad. des sci.,
CXXVili, p. 336) has shown that phenol is produced in small quantity
by heating with potash the sulphonates resulting from the action of
fuming sulphuric acid upon aldehyde or paraldehyde.

1 Ber. d. deutsch. chem. Gesellsch., xiv, p. 450.

2 Centralbl. f. Bakt., II, x, p. 338, 1903.

430 Oxidation of Benzoic Acids

for several hours over a low flame with reflux condenser. A
not inconsiderable amount of carbon dioxide is evolved during
the boiling and since no free phenol appears to be formed, it is
probable that part of the benzoic acid is completely oxidized.
After heating the liquid was diluted, acidified with sulphuric
acid and filtered. The precipitate consisted largely of unchanged
benzoic acid but contained a considerable amount of salicylic
acid. It was recrystallized several times from water till it
reacted no more with ferric chloride. In this way about half
the benzoic acid originally employed is recovered unchanged.
The original filtrate and mother liquors are combined and repeat-
edly extracted with chloroform. The chloroform extract con-
tained benzoic and salicylic acids. Meta- and para-oxyben-
zoic acids are insoluble in chloroform but are readily recovered
by ether extraction.

The chloroform extract was evaporated and the residue gave
the reactions for salicylic acid with ferric chloride, Millon’s re-
agent, alkaline diazonium salts, etc. Part was dissolved in
water and precipitated with bromine water. The precipitate
was dried and recrystallized three times from chloroform.
Tribromophenol-bromide, melting point 130—-131°, was obtained
in the form of thick prisms identical with the product similarly
prepared from pure salicylic acid.

The ethereal extract containing the meta- and para-oxyben-
zoic acids was evaporated and extracted with a little hot chloro-
form to remove a trace of salicylic acid. The residue no longer
reacted with ferric chloride. It was next dissolved in a little hot
water and excess of barium hydroxide was added. The very
sparingly soluble basic barium p-oxybenzoate crystallized out
in the form of a sandy deposit.. It was filtered off, decomposed
with hydrochloric acid and the free acid recovered by ether
extraction. It was recrystallized from water and formed large
prisms, melting at 210-211°. The recorded melting points for
p-oxybenzoic acid vary from 210~—213°.1 On treating the
p-oxybenzoic acid in dilute aqueous solution with bromine water
and repeatedly crystallizing the dried precipitate from chloro-

‘Hartmann: Journ. jf. prakt. Chem., 2, xvi, p. 36; Negri: Gaz. chem.
atal,, xvi, Ip: 165.

H. D. Dakin and Mary Dows Herter 431

form, tribromphenolbromide, melting point 130—131°,' was readi-
ily obtained. The para-oxybenzoic acid showed the usual reac-
tions with Millon’s reagent, diazonium salts. It gave no colora-
tion with ferric chloride.

The filtrate from the basic barium p-oxybenzoate was acidified
and extracted withether. The extract was evaporated and crys-
tallized from water. Since the preparation of pure meta-oxyben-
zoic acid possessing the correct melting point presented some
difficulties, as some of the para-oxybenzoic acid was still present,
it was converted into its bromine derivative. The residue was
dissolved in water and excess of bromine water added; the pre-
cipitate of tribromphenol bromide, derived from the p-oxyben-
zoic acid was filtered off and the filtrate extracted with ether.
The ethereal extract was dissolved in hot water, treated with a
little charcoal and on cooling 2,4,6-tribrom-3-oxybenzoic acid,
melting point 145-146°, was readily crystallized out in the form
of tufts of needles. Herzig? gives 145-147° for the melting
point of tribrom-meta-oxybenzoic acid. The product obtained
was identical in appearance with that prepared from pure meta-
oxybenzoic acid.

The total yield of the three oxybenzoic acids was determined
by iodine titration and amounts to 15 to 20 per cent of theory.
The amount of each acid which can be separated in the pure
state is much smaller, however.

Oxidation of Para-oxybenzoic Acid. One molecular proportion
of p-oxybenzoic acid was dissolved in hot water and ammonia
added until about five-sixths of the acid had been neutralized.
Three per cent hydrogen peroxide (1.5 g. mol.) was then added
and the mixture digested on the water bath for some hours. A
considerable amount of carbon dioxide is given off so that prob-
ably a portion of the oxybenzoic acid is completely oxidized.
If the reaction of the fluid is allowed to be decidedly alkaline
considerable darkening results on oxidation. After oxidation
the liquid was acidified and the products of oxidation, along with
much unchanged p-oxybenzoic acid, were extracted with ether.

1 Auwers and Biittner: Aun. d. Chem., 302, p. 133, give 131° for the
melting-point of tribromophenolbromide. Previously the melting-point
had erroneously been recorded as 118°.

2 Monatsh. f. Chem., xix, p. 92.

432 Oxidation of Benzoic Acids

The residue was tested for 2,4-dioxybenzoic acid and 3,4-dioxy-
benzoic acid (protocatechuic acid), these being the only dioxy-
benzoic acids derivable from p-oxybenzoic acid. The former
acid was found to be absent for no reaction was obtained with
bleaching powder, and moreover the dioxy-acid formed was pre-
cipitable by lead acetate, whereas 2,4-dioxybenzoic acid is not
precipitated by this reagent. The residue was then dissolved in
hot water and on cooling a large amount of unchanged p-oxy-
benzoic acid crystallized out.! The filtrate was then precipi-
tated with basic lead acetate. The precipitate on decomposition
with sulphuretted hydrogen and concentration of the aqueous
solution yielded crystals of protocatechuic acid. After recrystal-
lizing from water the acid melted at 199-200°. The product
gave an intense blue-green coloration with ferric solution, which
turned dark red on addition of caustic soda solution. It reduced
ammoniacal silver and on fusion with potash it yielded catechol.
The yield of dioxybenzoic acid was not more than 5 per cent of
the theoretical amount.

Oxidation of Meta-oxybenzoic Acid. The reaction was carried
out, and the products of oxidation separated in exactly the same
way as that employed for the para-oxybenzoic acid. Of the
four dioxybenzoic acids which might theoretically be formed
only the 3,4-acids (protocatechuic acid) could be identified. It
was separated and purified as in the preceding case. The yield
as before was small. The acid gave all the reactions of proto-
catechuic acid. 2,5-Dioxybenzoic acid was tested for, with nega-
tive result, by means of the quinone reaction with ferric chloride.”
3,5-Dioxybenzoic acid, which gives no color reaction with ferric
chloride, could not be detected with the anthrachyrson reaction
for this acid. 2,3-Dioxybenzoic acid was excluded by reason of
the fact that the oxidation products gave no blue color with
ferric chloride, turning violet-red on addition of alkali.

Oxidation of Salicylic Acid. The oxidation was carried out as
with meta- and para-oxybenzoic acids. The liquid, which had
turned dark brown was acidified and much unchanged salicylic

' By again oxidizing this precipitate of unchanged para-oxybenzoic acid,
the yield of the dioxybenzoic acid may be much increased, especially if
the oxidation of the unchanged acid be repeated several times.

? Nef: Ber. d. deutsch. chem. Gesellsch, xviii, p. 3499.

H. D. Dakin and Mary Dows Herter 433

acid filtered off. The filtrate extracted repeatedly with ether.
The dry ethereal extract was then treated with warm chloroform
to remove unchanged salicylic acid. The insoluble residue was
small and somewhat discolored. It gave a strong blue-black
coloration with ferric chloride which turned violet-red on addi-
tion of sodium bicarbonate and was completely precipitable
by lead acetate. Since 2,4-, 2,5-, and 2,6-dioxybenzoic acids are
not precipitable by lead acetate, these acids could not have
been present in appreciable amounts. The reactions with ferric
chloride and with lead acetate agree with those of 2,3-dioxyben-
zoic acid. The acid was purified by means of itslead salt. By
decomposing the lead salt with sulphuretted hydrogen, the free
acid was obtained inthe form of colorless needles, m. p. 200—202°.
It strongly reduced ammoniacal silver solution, even in the cold
and also Fehling’s solution on boiling. It gave catechol on dry
distillation. The yield of acid was about 3 per cent.

Oxidation of Substitution Derivatives of Benzoic Acid. Several
substituted benzoic acids were oxidized by the same method that
.was employed for the oxidation of benzoic acid. After heating
the solutions were acidified, extracted with ether and tested for
phenolic (OH) groups by ferric chloride, Millon’s reagent and
with sodium diazobenzenesulphonate in the presence of excess
of sodium carbonate. No attempt was made to isolate the indi-
vidual products. The yields of hydroxy-acids in all cases
appears to be very small.

REACTION OF PropucTs OF OXIDATION WITH THE FOL-

Actps OXIDIZED. LOWING REAGENTS.
Ferric Chloride. Millon’s Reagent. Diazo Reaction.

Orthochlorobenzoic acid...... Violet color Strong red Orange-red.
Parabromobenzoic acid........ Violet-red color Very faint posi- Orange.

tive reaction.
Paranitrobenzoic acid.........Violet-red color. Positive but Orange-red.

faint.
Metadinotrobenzoic acid

(G13 5) ee eae Negative Negative Strong orange red.*

Ortho-amido-benzoic acid.f ...Violet-red Positive Orange-red much

deeper than the
amido benzoic
acid.

* The original acid gave no trace of color reaction with diazonium salts.
+ The usual addition of ammonia was not made in the oxidation of this acid.

It will be seen from the table that qualitative evidence of the
formation of phenolic acids was obtained in each case. It is

434 Oxidation of Benzoic Acids

possible that the oxidation of the amido-acid with peroxide
yielded a phenolic acid through replacement of the amido group
as is known to occur with some oxidizing agents, but no similar
replacement would be likely to occur with the other benzoic
acids.

SUMMARY.

I. Hydrogen peroxide acting upon the ammonium salts of
benzoic acid, or its chlorine, bromine, nitro- and amido-deriva-
tives, can introduce hydroxyl groups into the nucleus, although
the yield of phenolic acid is small. Hippuric acid undergoes
nuclear oxidation with difficulty.

II. In the case of benzoic acid itself, ortho-, meta- and para-
_ oxybenzoic acids are produced in not widely differing amounts.
On further oxidation both meta- and para-oxybenzoic acids give
protocatechuic acid (3,4-dioxybenzoic acid). Salicylic acid gives
2,3-dioxybenzoic acid. The second hydroxyl group in the dioxy-
benzoic acids takes up a position ortho to that already in the
ring. <A part of the benzoic acid is decomposed with formation
of carbon dioxide and possibly other products.

III. The reaction occurs in approximately neutral solution
and to some extent at least the reaction progresses at the ordi-
nary temperature. Although benzoic acid itself is not appreci-
ably oxidized in the body, possibly owing to its conversion into
hippuric acid, a close parallelism exists between oxidations
carried out under these conditions and those taking place in both
vegetable and animal tissue oxidations.

IV. A summary is given of the reactions which have hitherto
been successful in effecting the direct entrance of hydroxyl
groups into the nucleus of benzene derivatives outside the body,
together with a list of aromatic substances which undergo nuclear
hydroxylation in their passage through the animal body.

V. The possible origin of phenolic substances in vegetable and
animal tissues is considered and though there is ample proof of
the production of phenolic substances by the oxidation of pre-
formed aromatic substances, the conclusion is reached that as
far as the limited experimental results permit there is little evi-
dence that they originate directly from the condensation or
rearrangement of aliphatic substances.

THE ACTION OF ARGINASE UPON CREATIN AND OTHER
GUANIDIN DERIVATIVES.

Bye. DD. DAKIN.
(From the Laboratory of Dr. C. A. Herter, New York.)

Received for publication, September 1, 1907.
Pp p 9°7

Three years ago Kossel and Dakin’ described the action of an
enzyme which had the property of effecting the hydrolysis of
arginin, with production of urea and ornithin.

NH

|
H,N — C — NH.CH;.CH,.CH,.CH(NH,).COOH + H,O = H,N.CO.NH, +
H,N.CH,.CH,.CH,.CH(NH,)COOH

This reaction represented a new type of enzymic hydrolysis,
differing from that effected by other “‘related’’ enzymes, such as
pepsin, trypsin and erepsin, in the fact that the latter only effect
separation of amido-acid groupings at points where an imide
group is adjacent to a carbonyl group (I), while in the case of
arginase a C:NH group may occupy the place of the carbonyl

grouping (II):
Va

1 Scconucd% wash Ne EQoH: NH,.CH
ZA $: UNG ye ae NS

Action of trypsin, ete.

ue H,N.CNH NHLCZ —» H,N.CO.NH,, wH,.cZ
Action of arginase.

This enzyme named arginase was found to be most abundant
in the liver, but was present in the kidney, intestinal mucous
membrane, thymus and probably other organs. It was also
found by Shiga? among the enzymes obtained from yeast. In
addition to arginin, arginase acts upon the arginin complex

1 Zeitschr.f. physiol. Chem., xii, p. 321, 1904; xiii, p. 131, 1904; Munch.
med. Wochenschr., 1904, No. 13. gt
2 Zeitschr. f. physiol. Chem., xlii, p. 502, 1904.
435

436 Action of Arginase upon Guanidin Derivatives

present in certain protones, with the liberation of urea and pro-
duction of polypeptide-like substances which yield ornithin on
complete hydrolysis with acids. Arginase was found to be
without action on the protamines themselves as well as other
proteins.

The fact that arginase attacked the guanidin complex in
arginin at once suggested that the same enzyme might possibly
attack other guanidin derivatives. It was found, however,
that arginase had but little action on guanidin itself and negative
results were also obtained in a preliminary experiment with
creatinin, but the great interest that attaches to creatin and
creatinin metabolism made it desirable to determine more exactly
whether creatin and creatinin were attacked at all by arginase.
While these experiments were in progress a paper appeared by
Gottlieb and Stangassinger! in which enzymes are described
which act upon creatin and creatinin, respectively. These
enzymes are named creatinase and creatininase. The former
apparently converts creatin into creatinin, while the latter
decomposes creatinin. The products of decomposition of crea-
tinin are not described, but if the reaction is similar to the action
of arginase (and Gottlieb and Stangassinger state that a long
preserved arginase preparation exerted an appreciable action
upon creatinin) the products should be sarcosin (methyl gly-
cocoll) and urea. If this should be the case, it would appear
anomalous that the first step in the enzymic decomposition of
creatin should consist in its conversion into its anhydride—
creatinin—for the hydrolysis of the latter with formation of
sarcosin and urea would apparently necessitate the re-formation
of creatin as an intermediate product.

NH——co NH, NH,

| |

C:NH oa C:NH > CO+CH,NH.CH,.COOH
N.CH, —CH, NCH,.CH,.COOH NH,

The object of the following note is to place on record the
results of some experiments which indicate that arginase and
creatininase are different enzymes, for organ extracts prepared

1 Zeitschr. t. physiol. Chem., lii, p. 1, 1907.

Fie? D-.2Dakam 437

in a manner such as would furnish an extremely active arginase
preparation were found to be practically without action upon
both creatin and creatinin as well as some other guanidin deriva-
tives. The organs examined were the liver, kidney and duodenal
mucous membrane of the calf and dog. It is remarkable that
the decomposition of creatin and creatinin described by Gott-
lieb and Stangassinger was not observed in these extracts and
the cause of this failure is by no means clear. It is hoped, how-
ever, that further experiments will throw light on this discrep-
ancy. It is interesting to note that the changes induced by
Gottlieb’s enzyme are extremely slow in comparison with the
vigorous action of arginase, for while the former requires many
hours or days to effect considerable decomposition, a small
quantity of the latter is able to decompose several grams of
arginin in a few minutes.

In the following experiments the tissues were ground up and
shaken vigorously with five to ten times their weight of water.
with liberal addition of toluene. The fluids were then roughly
strained through gauze and definite volumes of the turbid extract
were employed. Similar results were obtained when more highly
purified enzyme preparations were employed, obtained by alco-
hol and ether precipitation in the manner adopted for the
preparation of arginase. Additions of creatin or creatinin were
made to these extracts, which were then digested at 37° for vary-
ing lengths of time. The creatinin was estimated by coagulat-
ing measured portions of the digested fluid with dilute acetic
acid at the boiling point, filtering, and then applying Folin’s
colorimetric method with picric acid and caustic soda. Creatin
was estimated in a similar way, after conversion into creatinin.
The latter change was effected by boiling the clear filtrate with
normal hydrochloric acid for two hours, evaporating on the
water-bath and determining the total creatin plus creatinin.
The difference in preformed creatinin and the creatinin after
treatment with acid was a measure of the creatin. The amount
of solution employed for colorimetric estimation was such that
its creatinin content varied from eight to twelve milligrams.
Appropriate blank experiments were carried out and the neces-
sary corrections applied.

438 Action of Arginase upon Guanidin Derivatives

Dog liver extract upon creatinin.
One hundred ce. of 10 per cent emulsion used.

Crestinin ACA cy tien sede srs Glen svemearelay eee oe = 0.0794 gm
* fOUNC ALLEN 72 DOUDS) ence ola stnaetet = 0.0797 “
e MA ACS a Vi Os tc pea = 0.0792 “
Dog liver extract upon creatin.
One hundred ce. of 10 per cent emulsion used.
Gravtinbalt Ne (Co ( [a Gea ey are Pepe ty RRR keg 28 = 0.0604 gm.
FOUN SLCC fe OULS wha tase cya aarti aoe = 0.0569 “
3 . Rel ae ee eo. ak oe et eee = 0.0582 “
Dog kidney extract upon creatinin.
One hundred ce. of 10 per cent emulsion used.
Grpatinin SOG oo cs stairs cous +>. os mine eee alee = 0.0796 gm.
s found after (2hours; .!5. 5 seae: ee = 0.0806 “
: 4 6 MEO, or aha s 5a Gotta mee = 0.0860 “
Dog kidney extract upon creatin.
One hundred cc. of 10 per cent emulsion used.
Greabimiadded 0-65, mitaevnence eck or sie coou tenner = 0.0604 gm.
Be TOUnCUATLED, V2) OUTS: cn. 46s a) een = 0.0660 “
- : gle 3: |) alli a ee eS, 0 = 0.0594 “
Calf liver extract upon creatinin
One hundred and twenty cc. of 20 per cent emulsion used.
Creatinim added es. oe ve ceneyesats es. el cnt ek ann ne = 0.115 gm:
fs found after 160 hours... 2:22) eee = 0. 12003
Duodenal mucous membrane from calf upon creatinin.
One hundred cc. of 12 per cent emulsion used.
Creatminvadded: ....s. te inde se ee eee = 0.123 gm.
s oundvafter —40shourse ys eens = 0.134 *
= 0.112955

be : pcieage |: i eRe PR SA 8 5.

All these extracts were prepared in a manner such as yields
extremely active arginase preparations, yet the results shown
above do not indicate that they had any perceptible action
upon creatin and creatinin. The small differences between the
amounts of creatin and creatinin added and those found after
digestion are mostly well within the limits of experimental error
and it is therefore concluded that in all probability arginase has
nothing in common with Gottlieb and Stangassinger’s creatinase
and creatininase.

Assuming therefore that arginase does not act upon creatin
or creatinin, it is interesting to inquire into the reason why such
similarly constituted guanidin derivatives are not attacked by
the same enzymes. The most obvious difference in the con-

H. D. Dakin 439

stitution of arginin and creatin, judged from the point of view
of their attack by urea-forming enzymes, is the fact that while
the former contains a guanidin complex, the molecule of the
latter contains a methylguanidin group. It is improbable,
however, that this constitutes the reason for the differing
behavior towards arginase, for in the first place guanidin itself
is not appreciably attacked by arginase, as was found by Shiga
for the enzymes from yeast and by the writer for liver arginase.*
The result is in agreement also with the fact that guanidin
injected into the animal organism is almost entirely excreted
unchanged in the urine. The recognition by Kutscher, of
methylguanidin as a urinary constituent, makes it probable that
arginase has also no action upon methylguanidin, although direct
experiments upon this point are still wanting. Incidentally it
may be noted that liver extracts are also without action upon
triphenylguanidin.

It might be urged that the explanation of these negative
results with guanidin and its alkyl derivatives is to be found in
the absence of an aliphatic chain of carbon atoms containing a
carboxyl group, such as is present in the arginin molecule; but
this again is improbable, for it was found that arginase was with-
out material action upon glycocyamin (guanido-acetic acid).
This substance was prepared by Nencki and Sieber’s method?
by heating guanidin carbonate with glycocoll. It was digested
with tissue extracts containing arginase as in the other experi-
ments and the digested fluids were examined for increased urea-
formation.®

NH

|
NH, —C —NH.CH,.COOH+ H,0 =NH,.CO.NH,+NH,.CH,.COOH

1Guanidin carbonate was digested with calf liver extracts. The
digested fluids were subsequently examined for increased ammonia and
urea production, with negative results.

2 Journ. f. prakt. Chem., 2, xvii p. 477.

3 Urea was estimated by removing the coagulable protein from the
liquids, concentrating at low temperatures and then treating with ether
and alcohol, etc., as in the Mérner-Sjéqvist method. The urea was sub-
sequently precipitated with mercuric nitrate and sodium carbonate and
finally estimated by Kjeldahl nitrogen determination.

440 Action of Arginase upon Guanidin Derivatives

The negative results with glycocyamin show that the absence
of action of arginase upon creatin is not due to the presence of a
methyl group in the latter and also that the reason of its failure
to decompose guanidin itself and its alkyl derivatives is not to
be sought in the absence of an aliphatic chain containing a car-
boxyl group.'

It is probable that arginase is essentially a specific enzyme
and that, as in the case of the sugars and glucosides, a very close
intér-relation of enzyme and substrate is necessary for hydroly-
sis to occur. It must be remembered that arginin is an optically
active substance and although the asymmetric carbon atom is
situated at a considerable distance from the guanidin complex
which is attacked by the arginase,

H,N.C.NH.NH.CH,.CH,.CH,.CHNH,.COOH
*

yet it undoubtedly exerts a determining influence upon the
course of the reaction. The proof of this lies in the fact that,
as Otto Riesser? has shown, if racemic arginin be digested with
arginase, the dextro-compound is hydrolyzed, whilst levo-
arginin remains unattacked. Regarded simply from the stand-
point of structure, excluding the stereo-chemical relationship,
there is no apparent reason why levo-arginin should not under-
go hydrolysis with arginase in the same manner as dextro-
arginin even though the rates of hydrolysis should be dissimilar,
as is the case with the hydrolysis of isomeric optically active
esters by lipase.*

‘It may not be out of place to draw attention to the close similarity
existing between arginase and guanase both in their distribution in the
body and also the fact that both attach substances containing the guani-
din grouping with formation of a urea-group and either ammonia or an
amido acid.

HN:—CO HN:—CO
| ; |
HN:C: C—NH > O:C: C—NH + NH;
| || cH i on
HN:—C— N HNi—c_ 7

? Zeitschr. f. physiol. Chem., xlix, p. 210, 1906.
* Dakin: Proc. Chem. Soc., xix, p. 161, 1903; Journ. of Physiol., xxx, Pp.
253, 1903; XXXli, Pp. 199, 1905.

Le tay en) ee —

He D. Dakin 441

The evidence at present available, therefore, supports the
belief that arginase is a specific enzyme adapted for the exclusive
hydrolysis, as far as is known, of dextro-arginin or of substances
containing the dextro-arginin grouping and that, as in the case
of the glucosides and sugars, the relation of the enzyme to the
substrate is of so intimate and finely adjusted a kind, that many
other substances structurally similar to arginin are incapable
of hydrolysis by arginase.

ON THE CHEMISTRY OF BACILLUS COLI COMMUNIS.
II. THE NON-POISONOUS PORTION.

By MARY F. LEACH.
(From the Hygienic Laboratory of the University of Michigan, Ann Arbor.)
(Received for publication, July 22, 1907.)

INTRODUCTION. In previous papers upon this subject! the
author has described the method in use in this laboratory for ob-
taining large amounts of bacterial cellular substance, and has
shown that the cell substance of Bacillus colt communis, after
thorough extraction with alcohol and ether, is wholly dissolved
by the successive action of dilute acid and alkali. The solutions
thus obtained show the presence of proteid, nucleo-compounds,
and carbohydrate, while the properties and reactions suggest
that the cell is largely composed of glyco-nucleo-proteid. Heat-
ing with 1 per cent sulfuric acid breaks off the carbohydrates,
and splits up the proteid, apparently causing cleavage along
definite lines. By digestion with stronger acid, xanthin and
hexon bases were obtained. Lysin was isolated as the picrate,
purified, and transformed into the chlorid, both of which were
proved to be identical with lysin picrate and chlorid from other
sources. Thus bacterial proteid, like all others thus far exam-
ined, contains the hexon group, and another point of resem-
blance is established to other proteids of both animal and vege-
table origin.

Additional work shows? that the cellular substance of the
colon bacillus may be separated into a poisonous and a non-
poisonous portion by the action of a dilute alcoholic solution of
sodium hydrate. By the same method other proteids have

1 Trans. of Assoc. of Amer. Phys., xvii, p. 274, 1902; Journ. of the Amer.
Med. Assoc., xlii, p. 1003, 1904; this Journal, 1, p. 463, 1906.

2 Wheeler: ‘‘The Extraction of the Intracellular Toxin of the Colon
Bacillus,” Journ. of the Amer. Med. Assoc,, xliv, p. 1271, 1905.

443

444 Chemistry of Bacillus Coli Communis

been split up into poisonous and non-poisonous portions.' By
means Of the non-poisonous portion of the colon bacillus, it
is possible to induce active immunity to virulent cultures of
the living germ.?, Moreover the non-poisonous portion of the
typhoid bacillus has been used with marked success in the treat-
ment of typhoid fever. Hence the chemical nature of the non-
poisonous part of the bacterial cell is of great interest and impor-
tance. Further a single injection of many proteids, both of
vegetable and of animal origin, so sensitizes the animal, that a
second treatment, after a certain time limit has passed, will prove
promptly fatal. The non-poisonous part of any proteids that
have been thus far examined, will sensitize toward the whole
proteid, but is entirely specific in its action.* It is the purpose
of this paper to give an account of the non-poisonous part of the
colon bacillus, and of some of its decomposition products, espec-
ially the relation of the nitrogen and phosphorus.

MaTeERIAL. Bacillus colt communis was raised upon agar in
large tanks, harvested and purified as described elsewhere. By
repeated boiling with sodium hydrate dissolved in absolute
alcohol,‘ about one-third of the cellular substance, including all
the poison, goes into solution. The insoluble residue is filtered
out, extracted in Soxhlets with alcohol for 20 to 30 hours, dried
and pulverized. This is the material used in the following inves-
tigation.

PROPERTIES. The non-poisonous portion of cell substance
remaining after repeated extraction with alcoholic soda as
described above, is a cream colored powder. On burning it
puffs up, gives off the odor characteristic of nitrogenous com-
pounds, and leaves a copious ash containing phosphate. The
substance is mainly soluble in water, giving an opalescent solu-
tion from which a light colored sediment settles out on standing,
leaving a clear golden brown solution. This sediment is not

‘Vaughan and Wheeler: ‘Experimental Immunity to Colon and
Typhoid Bacilli,” N. Y. Med. Journ., 1xxxv, p. 1170, 1907.

*V. C. Vaughan, Jr.: ‘The Production of Active Immunity with the
Split Products of the Colon Bacillus, ”’ Journ. of Med. Res., xiv, p. 67, 1905.

*Vaughan and Wheeler: “Effects of Egg-White on Animals;” Journ.
of Infect. Dis., iv, p. 476, 1907.

* Wheeler: loc. cit.

Mary F. Leach 445

dissolved by dilute hydrochloric acid or dilute sodium hydrate
in the cold, but is dissolved by boiling with alkali. The clear
aqueous solution is alkaline in reaction from sodium hydrate
either held mechanically, or combined chemically; it is precipi-
tated by mineral acids and by alcohol. It gives the biuret,
xanthoproteic, and Adamkiewicz tests; Millon’s test is not very
satisfactory, but is more pronounced if the alkali is first neutral-
ized. It does not reduce Fehling’s solution directly, but does
so after boiling with hydrochloric acid. Tests with a-naphthol,
phloroglucin and orcin give positive results. Ammonium molyb-
date gives an organic precipitate, but no evidence of free phos-
phate. Thus the preliminary tests show the presence of proteid,
nucleic and carbohydrate groups. Comparing these results with
work previously reported upon the portion of the cell which is
soluble in alkaline alcohol, the soluble part always gives a very
pronounced Millon test, contains no carbohydrate, and all the
toxophorous group; while the portion insoluble in alkaline
alcohol gives only a faint Millon test, contains all the carbohy-
drates, most of the phosphorus, and seemingly the haptophorous
group. (Haptophorous is used here in Ehrlich’s sense of readily
combining with the physiological constituents of living cells.)

CLEAVAGE Propucts. With bacterial cell substance we have
succeeded in establishing certain definite lines of cleavage;
various attempts were made to find lines of cleavage for the non-
poisonous portion. The most promising ones may be outlined
as follows:

I. The material was treated with acid alcohol giving a solu-
tion A and an insoluble residue B. A solution of B in aqueous
alkali, on addition of alcohol, gave Filtrate C and Precipitate D.

II. An aqueous solution of the material treated with dilute
acetic acid gave Precipitate K and Filtrate L. On treating L
with alcohol, Precipitate M and Filtrate N were obtained.

III. Again an aqueous solution of material gave with acid
alcohol, Precipitate G and Filtrate H.

As was to be expected, qualitative tests showed few marked
differences in these preparations, but quantitative differences
were found which will be discussed later.

Alcoholic Solutions. As the cell substance of the colon germ
is entirely dissolved by successive treatment with very dilute

446 Chemistry of Bacillus Coli Communis

aqueous acid and alkali, the portion insoluble in alcoholic soda
was treated with acid alcohol. Doubtless the prolonged boiling
in extracting the poison has pronounced effect upon the physio-
logical constituents of the cell, still the substance carrying
immunity is to some extent left intact. As so many such sub-
stances are sensitive to heat, the further extractions were carried
on at room temperature.

Fifty grams of haptophorous substance were mixed with one
liter of absolute alcohol to which 30 cc. of strong hydrochloric
acid had been added, and shaken in a mechanical shaker for 5
hours at room temperature. After standing over night it was
filtered. The residue, which will be designated as B, was washed
with alcohol until the washings were no longer acid to moist
litmus, dried and pulverized. The weight was approximately
45 grams. The filtrate was evaporated to dryness on a water
bath. Adark, sticky mass resulted, which will be designated
A. When dry it could be pulverized, but was very hygroscopic.
After pulverizing it weighed 7 grams. A is mainly soluble in
water. With the aqueous solution the xanthoproteic, a-naph-
thol, phloroglucin, and orcin tests are all positive, the Millon
test is faint, biuret negative. It burns with a suggestion of
burning feathers, leaving an ash with a greenish tinge. Tests
showed the presence of aluminum and of phosphate in the ash.
Thus qualitative tests show that acid alcohol dissolved out some
carbohydrate, inorganic matter, as well as nucleo- and proteid
decomposition products, but no undecomposed proteid, while
about nine-tenths of the substance was left intact. Ash, nitro-
gen and phosphorus were determined as described later.

On stirring with water, B forms an emulsion which is acid in
reaction. The addition of sodium bicarbonate to slight alkalin-
ity gives a clear,dark solution which responds to the biuret,
xanthoproteic and Millon tests. The biuret is reddish violet.
Attempts were made to precipitate the proteid from this
alkaline solution. Hydrochloric acid gave a suspension which
seemed to thicken on heating, did not settle, and passed through
filter paper. Ammonium sulfate solution gave no precipitate,
but onadding the dry salt to saturation, precipitation took place,
the precipitate settled well and filtered clear. Both aqueous and
alcoholic solutions of mercuric chlorid with a little acid gave

Mary F. Leach 447

suspensions which did not settle. Alcoholic solution of picric
acid, copper sulfate and alcohol, and alcohol alone, all gave
precipitates that settled well. As alcohol seemed most promis-
ing, a clear alkaline solution of B was acidified with hydro-
chloric acid, and poured slowly with constant stirring into 5
volumes, giving a voluminous, flocculent white precipitate, D.
This was filtered out and washed with alcohol. It was readily
soluble in water, gave Millon, xanthoproteic, and a-naphthol
tests. The ash contained phosphate.

The alcoholic filtrate from D was evaporated to dryness, and
the residue extracted repeatedly with alcohol. The residue con-
sisted mainly of sodium chloride. The extracts were evaporated
to dryness leaving a small amount of brown scale, C, upon the
dish. C burns with nitrogenous odor, leaving comparatively
little ash which is largely phosphate. The solution gave xantho-
proteic and Millon tests. Owing to lack of material no quanti-
tative determinations were made. As A,B,C and D showed no
very marked differences, other methods of cleavage were tried.

Aqueous Solutions. In preparing aqueous solutions it was
found best to add water to the material little at a time, with
careful stirring, first making a smooth stiff dough, then gradually
thinning it to the proper consistency. The mixture was then
shaken at room temperature for two or three hours.

Ten grams of air-dried material mixed with 300cc. of water was
filtered through a sterilized Pasteur filter, requiring 36 hours.
The insoluble portion was removed from the filter, using proper
precautions to avoid removing particles of the filter. It was
mixed with a large amount of water and filtered through both
hard and soft paper, with and without using suction. No device
was successful in getting the residue reasonably free from solu-
tion. By repeated treatment nearly all went into an opalescent
solution from which nothing separated out on standing. The
remainder when dry looked like agar and sand.

Aliquot portions of the above solution that had been filtered
through porcelain were used for determinations of nitrogen,
phosphorus, diamino- and monamino-nitrogen, according to
methods to be described later. Tencc. contained 0.0036 gram of
phosphorus and 0.008092 gram of nitrogen, corresponding to
0.090 gram of phosphorus and 0.2023 gram of nitrogen in the

448 Chemistry of Bacillus Coli Communis

whole. As the non-poisonous portion of the germ contains
5.556 per cent of nitrogen and 2.34 per cent of phosphorus, less
than half passed through the filter. The ratio of nitrogen to
phosphorus is 2.25, very close to the ratio in the original, and in
a number of the preparations included in a subsequent table.
Of the total nitrogen in the solution, 15.91 per cent was in the
form of diamino compounds, and 73.06 per cent as monamino
compounds.

As filtering through porcelain kept back most of the substance,
that was abandoned. A 5 per cent solution was filtered through
cotton with suction, leaving very little upon the filter. The
filtrate was not quite clear, so it was filtered twice through
a ‘‘Nutsche” with two soft filters, then twice with hard filter
without much effect. However as opalescence seemed to be
due to matter in solution rather than suspension, further samples
were filtered through a layer of cotton and then through soft
filters with suction.

The difficulty of filtration, the mucilaginous character of the
solutions, as well as other properties and reactions, suggested
mucin, while the presence of nucleic acid was also indicated. In
the hope of effecting a separation, dilute acetic acid was added
to an aqueous solution prepared as above, giving a brown precipi-
tate, K, and a filtrate, L. Precipitate K was washed with
alcohol and then with ether, and dried im vacuo. The yield
was 0.6 gram from 20 grams of substance. It gave xantho-
proteic tests but not the biuret, and contained phosphorus, hence
it was not mucin.

Again 50 grams of material was dissolved and acetic acid was
added to the solution, giving Precipitate K, and Filtrate L,.
The yield of K, was 8 grams. It does not give a clear solution
with water, but clears on the addition of a very little sodium
acid carbonate. K gives the xanthoproteic and a-naphthol
tests, a faint Millon, no biuret. Ammonium molybdate solu-
tion gives no evidence of phosphate, but there is phosphorus
present in organic combination. Many indications suggest
nucleic acid, while the precipitation with acetic acid points to-
ward guanylic acid.!

* Bang: ‘‘ Chemische und physiologische Studien uber die Guanylsaure;””
Zeitschr. f. physiol. Chem., xxxi, p. 411, 1900-1.

Mary F. Leach 449

Filtrate L was poured into two volumes of alcohol, giving
Precipitate M and Filtrate N. The precipitate was washed with
alcohol and ether. After drying in vacuo and powdering there
was 33 grams of a fine mealy powder almost white, readily soluble
in water. It turned yellow on heating with sodium hydrate,
the xanthoproteic test was positive, the biuret negative.

Filtrate L, was treated in the same way. The precipitate,
M,, was sticky at first, but the sticky substance was removed
by repeated treatment with alcohol and ether. The yield of M,
was 13 grams. It is not so soluble in water as M, but the solu-
tion clears on neutralization with sodium bicarbonate. Animal
experiments and determinations of nitrogen, phosphorus and
ash did not show sufficient difference between K and M to
warrant further use of acetic acid.

Again a 5 per cent aqueous solution of the immunizing portion
of the germ was acidified and tested with varying amounts of
50 per cent alcohol, absolute alcohol, also alcohol and ether.
As the result of these tests, the solution was poured into 4 volumes
of absolute alcohol containing 10 cc. of hydrochloric acid and
too cc. of ether per liter. After settling, the supernatant liquid
was siphoned off, the Precipitate G filtered with suction, washed
with alcohol containing ether, then with ether, dried and pul-
verized. The yield from 50 grams of material was 19 grams.
This precipitate was twice dissolved in water made faintly
alkaline with sodium acid carbonate, and reprecipitated by alco-
hol containing hydrochloric acid and ether. The final preci-
pitate G weighed 16 grams. With water G forms an emulsion
acid in reaction, cleared by the addition of alkali. The biuret
test is negative, Millon doubtful, xanthoproteic, Adamkiewicz,
a-naphthol, and orcin tests are all positive, the carbohydrate
tests being very marked. After boiling with acid there is copious
reduction of Fehling’s solution. The substance was tested for
glycogen with negative results.

The alcoholic filtrate from G was concentrated on a water
bath below the boiling point. A black, charred mass H was
left, easily pulverized, smelling strongly of hydrochloric acid.
It contained phosphate, and responded to xanthoproteic, Millon,
and a-naphthol tests, but did not reduce Fehling’s solution even
after boiling with acid.

450 Chemistry of Bacillus Coli Communis

Asu, NirrRoGEN AND PuospHorus. On heating, these sub-
stances puff up, forming liquid and volatile products, some of
which burn‘ with a flame. For determinations of ash a sample
was heated in a platinum crucible to low redness, using extreme
care to completely incinerate the organic matter, without volatil-
izing the inorganic portion. The results are tabulated under
the heading ‘“‘Ash.” All determinations were made in duplicate,
and figures given are the average of closely agreeing determina-
tions: In some cases the ash obtained as above was heated
with the full flame of a Detroit burner, and results are given
under the heading ‘‘Fixed Ash.”’ Solutions of ash gave immedi-
ate precipitates in the cold with ammonium molybdate, and
hence contain orthophosphate. If the phosphorus in the ash is
all combined as PO,, then

Wt: Pee WisbOn srs
or, PO? =32 5X 3-00

Now the phosphorus in the material is organic, and the PO,
should not be counted as mineral matter. The total ash less
P X 3.06 is given in the table under the heading ‘Inorganic
Ash,” and these figures are the ones used in calculating ash-free
N and P.

Nitrogen determinations were made by the Kjeldahl-Gunning
method. Phosphorus was determined by the Neumann! method
with some modifications. The sample was shaken with ammo-
nium nitrate in a 500 cc. Kjeldahl digestion flask,” and 10 cc. sul-
furic acid added. When frothing subsided it was heated care-
fully on a sand bath, more ammonium nitrate added, and heated
until colorless. The mixture was cooled, water added, more
ammonium nitrate, made alkaline with ammonia, then acid with
nitric acid adding 1 cc. in excess. It was then heated until bub-
bles rose. At the same time ammonium molybdate solution
was heated until bubbles rose, and poured slowly into the hot
phosphate solution with shaking or stirring. Prepared thus the

1‘*Einfache Veraschungsmethode,”’ Zeitschr .f. physiol. Chem, xxxvii,
Pp. 115, 1902-3; xlili, p. 32, 1904-5.

?Sherman: ‘Determination of Sulfur and Phosphorus in Organic
Materials,”’ Journ. of the Amer. Chem. Soc., xxiv, p. 1100, 1902.

Mary F. Leach 451

precipitate settled within 15 minutes,’ although it was always
left for an hour or more. The precipitate was washed with
0.1 per cent ammonium nitrate solution? until free from acid,
filter and precipitate returned to the flask, 150 cc. of water added,
half-normal sodium hydrate run in from a burette until the pre-
cipitate is dissolved, 5 or 6 cc. added in excess, and boiled until
all ammonia is driven off. After cooling, a few drops of phenol-
phthalein were added, half-normal sulfuric acid run in to slight
acidity, then titrated back with alkali. The total alkali, less
the amount of acid used, gives the amount of alkali required to
dissolve the precipitate. Each cc. of half-normal alkali corre-
sponds to 0.553 milligram of phosphorus.

TABLE I. PERCENTAGES OF ASH, NITROGEN AND PHOSPHORUS.

- Inor- v :

Ash. ee ganic N Pa ae ate ae
ash. | | free. | free. | ~~
Cell substance...... 8.61 10.65] 2.87] |

Immunizing portion| 33.25 26.08} 5.56) 2.34) 7.52) 3.99) 2.38
EGE Tea cas svenshe ree 26.76] 20.36) 15.66] 6.76) 3.61] 8.02] 4.28] 1.87
A S18) Stee ee Seer 35.34 30.74] 4.87) 1.50) 7.03] 2.16) 3.25

oD oe oe 15.38] 15.05) 8.48) 4.95|) 2.25) 5.41] 2.46) 2.2
Gi 9 ACs een a 6.99 1.66) 4.65] 1.74) 4.73] 1.77) 2.67
see purthed. ... 5 5 5.9 1236\Fo43) lado) oe48) Leotlecnos
18 De a er 35.91) 14-00) 27.67) 5.98) 2.68) 8.27) 3.71) 2.23
Go) Vikas ele eens Thats 2.08) 5.5 | 1.79] 5.62) 1.83) 3.07
REMMI eo otc e 55a 11.71) 11.71) 3.74) 3.16) 2.47) 3.28 2.70! 1.28
Si. 9-1 Se eaeeanaeatie gia 8.3 SoA lp on oll OS Or Oo! LeO4 mare

Explanation of Table I.

Ash, residue after heating at low redness.

Fixed ash, residue after heating to full heat of powerful burner.

Inorganic ash, ash less calculated amount of PQO,.

N, nitrogen by Kjeldahl-Gunning method.

P, phosphorus by the Neumann method.

N and P ash free, reckoned free from ‘‘inorganic ash.”

N: P, quotient of column 4 divided by column 5.

A, portion of immunizing substance dissolved by acid alcohol.

B, portion of immunizing substance not dissolved by acid alcohol.

D, substance precipitated by acid alcohol from solution of B in aqueous
alkali.

1 Treadwell: Lehrbuch der Analytischen Chemie, 2d ed., ii, p. 302.
2Koch and Woods: ‘Estimation of Lecithans,” this Journal, i, p.
206, 1906.

452 Chemistry of Bacillus Coli Communis

G, substance precipitated by acid alcohol from aqueous solution of
immunizing portion of colon cellular substance.

H, obtained by concentrating the alcoholic filtrate from G.

K, substance precipitated by dilute acetic acid from aqueous solution
of immunizing portion of germ.

K,, same as K, except that strong acid was used.

M, precipitated by alcohol from filtrate from K.

M,, precipitated by alcohol from filtrate from K,.

The large amount of ash in some of the above preparations
doubtless includes the sodium salts of various decomposition
products, in so far as they are insoluble in alcohol, as well as
practically all the inorganic matter of the whole germ substance,
both that which is an integral part of the cell, and that which is
mechanically carried down from the culture medium.

As these preparations are all mixtures, the absolute values
found are worth nothing taken singly, butt he comparative values,
especially the ratio of N : P as given in the last column, are of
interest. The determinations were made for the sake of tracing
the nucleo-compounds. There are many indications of nucleic
acid, but the amount of both nitrogen and phosphorus is much
too small. The ratio between them is, however, quite within
the range found for nucleic acids from other sources, as may be
seen by comparison with the following table. Moreover the
nucleic acid and the nucleates are the only nucleo-compounds

TABLE II. PERCENTAGES OF NITROGEN AND PHOSPHORUS IN NUCLEO-
COMPOUNDS.
Source. Observer. N P INP ee,
INTICIGIG BOIL. 5 ca,s eis.s)os.cuteine | Sperma salmon Miescher* 15.24 | 9.62 | 1.58
S ob PEC RO ACIS _ “ sea-urchin | Mathews 15.34 | 9.59 | 1.6
ss ithe CO Ren Seema Yeast Miescher 16.03 | 9.04 | 1.77
e : Brae Bere Pancreas Bang 18.2 7.67 | 2.37
2 Ste ene e eee eee Thymus Kostytschew 15.55 | 9.25 |) 1368S
. fb eee cence eee a Bang 15.26 | 9.3 1.65
BS eterna hones Wheat embryo Osborne and
bate Harris 15.88 | 8.7 | 1.83
Inosinic acid Sa kundaeveeldcreeict\| MISE Haiser 16. 8.6 1.86
Clupein nucleate............ Mathewst 21.06 | 6.07 | 3.48
Nucleo-histone............. | Thymus | Huiskamp 18.37 | 3.7 4.97
Nucleo-proteid............. a i 16.42 | 0.95 |17.3
J Std. Rew arity ieiwssts | Pancreas Umber 17.82 | 1.67 |10.65
Nuclein from above......... } GS 17.39 | 4.48 | 3.88
Ba a-nucleate..............| Thymus Kostytschewt 12.83 | 7.63 | 1.68
Baf-nucleate.............. | 5 = 10.16 | 8.48 | 1.38

* Mann, Chemistry of the Proteids, p. 442, 1906, where will be found references to the

orginial articles.
t Ibid., p. 454.

t Kostytschew: Zeitschr. f. physiol. Chem., xxxix, p. 556.

Mary F. Leach 453

in which the ratios are at all comparable with those given in
Table I. Nuclein contains a little less phosphorus than any of
these preparations from the germ, while other nucleo-com-
pounds are much richer in nitrogen and poorer in phosphorus.
It is perhaps worthy of mention that contact with mineral acid
apparently breaks up the nucleic acid, the phosphoric acid
going into solution; thus Preparation A gives evidence of phos-
phorus in inorganic combination, while G does not.

CARBOHYDRATES. In all these preparations qualitative tests
point to the presence of carbohydrates, especially pentose. As
the easiest and quickest way to obtain comparative data, samples
were hydrolyzed and titrated with Fehling’s solution. Gram-
samples of the immunizing portion of the germ were dissolved
in water containing a little alkali, neutralized, hydrochloric acid
added to definite strength, and heated on the water bath in a
flask with reflux condenser. The hydrolyzed solution was neu-
tralized and titrated with Fehling’s solution. Although there
is undoubtedly some pentose present, there is no proof that the
reducing substance is all carbohydrate. However for purposes
of comparison the reducing substance was calculated as xylose.
In order to find conditions giving the maximum yield, amount
and strength of acid, as well as time of boiling, were varied as
given in Table III.

TABLE III. REDUCING POWER OF IMMUNIZING PORTION.

No. of sample. | Amountof HCl | Strength of HCl) Hours boiled. éaloulntedes
ec. per cent. xylose.
Mepis caeve "sve ts 2 26 1 1 7.05
Meret Nore fans she 38.5 2.5 il 16.45
Bc o tage ere Rene 38.5 2.5 2 21.56
eR us ti 3 38.5 2.5 4 23.12
Bicuc 2 Bgoee epee 72 2.5 3 23.93
OPP R ciel ie ao: 72 2.5 9 23.53

As shown by the table, the maximum amount of reducing sub-
stances was obtained by using 2.5 per cent acid, and boiling for
three hours. Longer heating changes the results very little.

Samples of G were also hydrolyzed and titrated with results
as follows:

rt gram boiled 23 hrs. with 72 cc. of 2.5 per cent HCl gave 38.63 per cent
reducing substance calculated as xylose.

454 Chemistry of Bacillus Coli Communis

1 gram boiled 5 hrs. with 72 cc. of 2.5 per cent HCl gave 43.77 per cent
reducing substance calculated as xylose,

DistRIBUTION or NirroGen.! One gram samples of the
immunizing portion of the colon germ were suspended in 25 cc.
of water, and 25 cc. of strong hydrochloric acid added. After
standing 24 hours, they were boiled on a sand bath with a reflux
condenser for 9} hours. The biuret test was applied to one with
negative results. The others were concentrated on a water
bath, water and cream of magnesia added, and the ammonia
distilled off into standard sulfuric acid. This is reported as
amid nitrogen. The residues left in the flasks after distillation
were filtered with suction and thoroughly washed. The nitrogen
was determined in the precipitates, and is reported as humin
nitrogen. The filtrates were treated with sulfuric and phos-
photungstic acids. The nitrogen of the phosphotungstic pre-
cipitates appears in the table as diamino nitrogen, that of the
filtrates as monamino. Samples of G were treated in the same
way. The difficulty of determining nitrogen in the presence of
the large amount of phosphotungstic acid in the monamino
fraction is such that no great dependence is to be placed upon
that result. Still it is useful as a check upon the whole. In the
table, column 4 gives the sum of the nitrogen found in the four
fractions, while column 5 gives values found directly.

TABLE IV. DISTRIBUTION OF NITROGEN.

Amid Humin Monam- Diamino | TotalN| Total
N | N |inoN | N cal. | N det.

per cent per cent per cent percent |per cent|per cent

Imm. portion, sample 1. .} 0.860 | 0.707 | 3.35 | 1.33 6.24 | 5.56
- as o -2..| 0.846 | 02608 | 3.28 | 1.389 6.12
Ceres 8 Ses 2ce4 Sees 4 0.384 | 0.441 | 1.81 | 0.966 | 3.60
iG 7 aah Op ae ey re .| 0.355 | 0.540 | 1.78 | 0.917 | 3.99) ogee

‘Per cent of Total N.

10.6
12.2

Ofimm. portion.........
857 EE ae eee

53.7 | 22.
50.3 | 26.8

ANIMAL EXPERIMENTS. The immunizing power of some of
these cleavage products was tested as follows:

1 Hausmann: Zettschr. f. physiol. Chem., xxvii, p. 95, 1899.

———— Ee

Ne

Mary F. Leach 455

400 mg. of K was suspended in 30 cc. of water, and neutralized with solid
sodium bicarbonate, making a fairly clear solution. Guinea pigs were
inoculated intra-abdominally with 5 cc. of this solution. Thus each
received 66.6 mg. of substance. Three days later each received 100 mg.,
and six days later four received 125 mg. in 8 cc. of water. These four
died in one or two days, post mortem findings proving that death was due
to local irritation, not to toxic action. The other two received on the
seventh day 125 mg. of M,. Other guinea pigs were inoculated with
solutions of M and M,, prepared in the same way. April 12, 1906, one
animal from each set was inoculated with twice the fatal dose (2 cc.) of
a colon culture 16 hours old. A control was dead the next morning, all
the others recovered. Animals receiving twice the fatal dose on April
14 and 15 recovered. On April 16 animals succumbed to this amount,
the immunity having partially worn off. On April 17 and 18, animals
recovered after receiving 1 cc. of the germ culture. These results are
tabulated in Table V.

TABLE V. IMMUNITY EXPERIMENTS.

April 7. Mh i04. | oon culeure | Result.
100 mg K |125mg K | ‘Died April 11
“ “ “ “ 12
“ “ | “ “ a
“ “ | “ “ 12
“ 125 mg M,'Apr. 12, 2 |Recovered.
“ “ “ 14, 2 “
100mg M |116mgM! * 12,2 a
“ “ “ 14, 2 “
“ “ “ Lis) 2 “
e a “« 16,2 |Dead 7 a.m.,April 17
sf & “ 17,1 |Recovered.
“ “ “ lye, 1 “
100mg M, |125mgM,) * 12,2 a
“ “ “ 14, 2 “
“ i “ 15, D) “
90 mg M, - “« 16,2 |Dead April 17.
e « “ 17,1 |Recovered.
100 mg M, “ “ 18, 1 iT;

On January 17, 1907, seven guinea pigs were inoculated with 50 mg.
of Preparation G, carefully purified by repeated solution and precipitation.
On January 21 one of these received 3 cc. of colon culture, was quite sick
within half an hour, but soon recovered, was given 50 mg. of G on January
24, with no apparent effect. The other six pigs on January 21 received
50mg.eachofG. On January 24, one of these received 3 cc. colon culture,
and 15 minutes later 50 mg. G, with same result as the preceding. A
third animal received 4 cc. of colon culture, was very sick, but recovered.
The other four animals received a third dose of 50 mg. of G, and at vary-
ing intervals were given 4 and 5 cc. of germ culture, but were not able

456 Chemistry of Bacillus Coli Communis

to withstand that amount. All colon cultures used were virulent ones
of which the fatal dose was 1 cc.

Thus it will be seen that all these preparations confer a certain
amount of immunity, but not so much as does the whole of the
immunizing fraction of the cellular substance. All of them have
been subjected to the action of acid, and it may be that the
immune body has undergone chemical change, or it may be that
the larger part is to be sought in some other portion of the hapto-
phorous fraction.

SuMMARY ANDConcLusions. By theaction of sodium hydrate
and alcohol, part of the proteid of the bacterial cell goes into
solution in the alcohol, and part of the proteid remains undis-
solved. The alcoholic solution contains the poison of the cell,
while the insoluble portion includes carbohydrate, nucleo-com-
pounds, and immunizing substance.

Although the germ substance is entirely dissolved by succes-
sive treatment with dilute aqueous acid and alkali, only about
10 per cent of the portion insoluble in alkaline alcohol is dissolved
by acid alcohol. And this is made up largely of inorganic salts,
including considerable phosphorus separated from nucleo-com-
pounds.

An aqueous solution of the immunizing portion of the germ
was treated with acetic acid in the hope of separating mucin,
but it did not give a sharp separation. The precipitate was not
mucin, but resembles guanylic acid in its difficult solubility,
especially in acetic acid.

Preparations precipitated by acid and alcohol resemble nucleic
acids in appearance and properties. The amounts of nitrogen
and phosphorus are far too small, but the relative amounts
correspond very well with proportions found in nucleic acids
from other sources. The failure of the biuret test shows that
the proteid has been gradually broken up and it would seem
that the acid used gradually decomposes the nucleic acid, phos-
phorus appearing in the acid alcoholic solutions as phosphate,
while the alcoholic precipitates contain phosphorus in organic
combination only. Decomposition is slow, for 19 grams of
Preparation G yielded 16 grams after being twice dissolved and
re-precipitated. There was no evidence of inorganic phosphorus
in the final precipitate.

Mary F. Leach ! 457

Most of the reducing substance from the non-poisonous por-
tion of the germ is to be found in the acid precipitate G. Rec-
koned as xylose, since the pentose tests are very marked, we have
23.93 per cent in the material, and 43.77 per cent in Preparation
G. Bang reports' 27 per cent of carbohydrate in guanylic
acid, while Osborne and Harris’? found evidence that tritico-
nucleic acid contains three molecules of pentose, corresponding
to 32.2 per cent xylose.

Preparation G then probably contains nucleic acid, and some
of the decomposition products of the same, notably pentose.
Comparing figures obtained for distribution of nitrogen in this
preparation and in the immunizing portion of the germ it would
appear that diamino compounds are in excess in the former.

These preparations, containing nucleic acid, possessimmunizing
power. Work has not progressed far enough to show that one
depends upon the other, but the fact is of interest in view of
the use of nuclein as a curativeagent. Casein, containing only
paranuclein, gives no sensitizing body.?

In conclusion I wish to express my thanks to Dr. V. C. Vaughan
for his valuable suggestions during the progress of the work.

EDIEOG? GtE.

2Osborne and Harris: ‘“‘Die Nucleinsiure des Weizenembryos,”’
Zeitschr. f. physiol. Chem., xxxvi, p. 85, 1902.

3 Vaughan and Wheeler: ‘‘Effects of Egg-White upon Animals,”’ Journ.
of Infect. Dis., iv, p. 476, 1907.

tay

oan
ako
roe oe Pal

pet

Winds ;

i

A STUDY OF THE COMPARATIVE CHEMICAL COMPOSITION
OF THE HAIR OF DIFFERENT RACES.

By THOMAS A. RUTHERFORD anp P. B. HAWK.

(From the Laboratory of Physiological Chemistry of the Department of
Medicine of the University of Pennsylvania.)

(Received for publication, September 27, 1907.)

tog PURGE OUR yi Ga fee 0 AS ae a 459
mie sources ot the keratm analyzed....... 5... ses cde cet const ees 460
BMeP ceparation of the hairfor analysis. .... 0.0. .6s.sscnce cerca ee 461
1. BUGER GS GHETaN GS > 3 at ire ieee Riri. 462
Wee DI SCUSSIOINO 1s CSUALSI Terie cilia s wlc’s <icie sGusie was claves stv euiaioe ae 462
r Chemical composition of indian hair... ....-..5.2j.-..05 462
2. Chemical composition of indian hair as influenced by the

sex of the individual and the purity of breeding.......... 463
ge Chemical composition of negro hair... .0. 2.24. 6506s 00255 467

4. Chemical composition of negro hair as influenced by the
sex of the individual and the purity of breeding......... 468

5. Comparison of the chemical composition of the indian
AelGTI RGEC) AMI te ope cs) 2 2a PME wis «oie ws nies) ee asso % slelereia 474
6. Chemical composition of japanese hair.................. 476
7. Chemical composition of caucasian hair...............-- 477
a. Chemical composition of the hair of adult caucasians.. 477

b. Chemical composition of the hair of adult caucasians
as influenced by the color of the hair and the sex
Gi? TovSanayethyare heel bee Acie oo emo cmeiomn cn Oe ooo 480
c. Chemical composition of the hair of caucasian children, 483
d. Chemical composition of the hair of caucasian children
as influenced by the color of the hair and the sex of
HOS Clivlicl 6 neue cosien apo DE pO Come e euoe Epi d.6 ec oc 487
WL, Concise: c <s0con on me Oo Gen eine nemo Oc ce 489

I. INTRODUCTION.

Apparently the first authentic observations upon the chemical
composition of hair were made by Vauquelin* and reported in
1806. He found that upon dissolving the hair in superheated

1 Vauquelin: Aun. de Chem. et de Phys., lviti, p. 41, 1806.

459

460 Chemical Composition of Hair

steam or in mineral acids a transparent solution resulted which
could be precipitated by an infusion of nut galls and which pos-
sessed the further property of blackening silver salts. He con-
cluded that hair contained silica, carbonates, iron, manganese,
calcium phosphate, and finally a considerable quantity of sul-
phur. Vauquelin found a lower percentage of sulphur in black
hair than was present in white, blond, or red hair.

The investigations of v. Laer! upon the chemical composition
of human hair are widely quoted. He could detect no carbonate
or manganese in the hair but, like Vauquelin, he found a high
sulphur content. v. Laer found the percentage of sulphur in the
hair of adult males to vary from 4.63 to 5.44.

Kthne and Chittenden? claim that neurokeratin has a higher
content of carbon and hydrogen and a lower content of nitrogen
and sulphur than the keratin prepared from the hair of the white
rabbit.

v. Bibra’ found only 3.83 per cent of sulphur in the hair of a
boy ten years old whereas the red hair of a man of thirty con-
tained 8.23 per cent. The same investigator found the average
sulphur content of fifteen samples of human hair to be 4.62 per
cent. He further found 3.7 per cent of sulphur in the hair of a
Peruvian interred four hundred years and 4.4 per cent in the
hair of a Bolivian interred one thousand years. Asa result of a
series of analyses of the hair of different animals v. Bibra deter-
mined that 4.25 per cent was the average sulphur content.

Mohr* found the hair of adult females to contain 4.95 per cent
of sulphur whereas that of a girl nine years of age contained 5.34
per cent. In the hair of a boy four years old Mohr found 4.98
per cent of sulphur, while the red hair of a boy six years of age
contained 5.32 per cent.

II. SOURCES OF THE KERATIN ANALYZED.

Our investigation included the analysis of the hair of the
following races: indian, caucasian, negro and japanese. The
specimens of indian hair, ten in number, were furnished through

1 v. Laer: Ann. d. Chem. u. Pharm., xlv, P- iT49; Taae

* Kithne and Chittenden: Zeitschr. f. Biol., xxvi, p. 291, 1890.
*v. Bibra: Ann. d. Chem. u. Pharm., xcvi, pp. 289-302.
‘Mohr: Zeitschr. f. physiol. Chem., xx, pp. 403-406, 1895.

Thomas A. Rutherford and P. B. Hawk 461

the courtesy of W. A. Mercer, Captain 7th Cavalry, Superintend-
ent of the Indian Industrial School, Carlisle, Pa., and were obtained
from students at the school. The negro hair, with the exception
of samples 1-5, was kindly furnished us by Dr. M. M. Waldron,
of the Hampton Institute, Hampton, Va. The hair of the living
adult caucasians was obtained from various students of the
University of Pennsylvania, and the children’s hair was obtained
through the courtesy of Richard Binder, Jr., of Philadelphia.
The japanese hair was courteously furnished us by Mr. Okumura
of the University of Pennsylvania.

The purpose of our investigation was to determine whether
there were any differences in the chemical composition of the
hair of the various races mentioned above. The samples of
hair analyzed varied in color from an extremely light yellow to
jet black and the ages of the individuals from whom the samples
were obtained ranged from two and one-half years to sixty years.
The average data for the hair of the different races are shown in
Table X, p. 488.

III. PREPARATION OF THE HAIR FOR ANALYSIS.

The hair, after being thoroughly washed in distilled water,
was placed in a large volume of artificial gastric juice and kept
at 40°C. for forty-eight hours. The digestion mixture was then
thrown upon cheese cloth and after draining thoroughly the
hair was washed free from gastric juice by means of distilled
water. The gastric fluid was now replaced by artificial pancre-
atic juice and the mixture again placed at 40° C. for forty-eight
hours. At the end of this period the hair was removed from the
pancreatic juice and washed with distilled water as before. It
was now introduced into a round-bottomed flask, sufficient
alcohol added to cover the sample, the flask provided with a
reflux condenser and the mixture boiled on a safety water-bath
for thirty-six hours. The alcohol was now removed by filtration
and the hair boiled in ether for twenty-four hours. The sample
was then removed from the ether, cut into exceedingly short
lengths by means of scissors, and placed at 100° C. for four hours.
It was then transferred to glass stoppered bottles ready for
analysis.

2 The authors wish to express their thanks to Mr. John Thomson, who
kindly assisted in the preparation of the hair for analysis.

462 Chemical Composition of Hair

IV. METHODS OF ANALYSIS.

Sulphur was determined after fusing with sodium hydroxide

and potassium nitrate. Nitrogen was determined by a modifi- -

cation of the Kjeldahl method in which the preliminary digestion
is accomplished by means of sulphuric acid and cupric sulphate.
Carbon and hydrogen were determined by Benedict’s modifica-
tion' of the Liebig method. In this modification the carbon
dioxide is absorbed by soda lime and the water by sulphuric
acid.

The figures for the percentage composition of the various speci-
mens analyzed are the averages of from two to six determina-
tions.

V. DISCUSSION OF RESULTS.

1. Chemical composition of indian hair. Ten samples of
indian hair were analyzed, the specimens being obtained from
the representatives of ten different tribes. Six of the samples
were from full-blood indians and the remaining four were from
half-bloods. Seven of the specimens analyzed were from female
subjects and three from males. The indians whose hair was
analyzed were from twelve to nineteen years of age.

The data obtained from the analysis of the ten samples of
indian hair are summarized in Table I, p. 464. The lowest sul-
phur content (4.55 per cent) is found in the hair of the half-blood
female Klamath, whereas the highest sulphur content (5.03 per
cent) is found in the hair of the half-blood male Shawnee. The
lowest nitrogen content (14.80 per cent) is possessed by the full-
blood female Pima and the highest nitrogen content (15.72 per
cent) by the half-blood female Alaskan. Carbon is present in
smallest percentage (41.68 per cent) in the hair of the half-blood
male Shawnee, while it is present in largest percentage (45.69
per cent) in the hair of the fuli-blood female Shoshone. Hydro-
gen follows the carbon in showing the lowest percentage (6.04
per cent) in the hair of the half-blood male Shawnee but differs
from the data for carbon in showing the highest content (7.10
per cent) in the hair of the full-blood female Chippewa. The
average percentage composition of the ten samples of indian

‘ Benedict: Elementary Organic Analysis, p. 49.

Thomas A. Rutherford and P. B. Hawk 463

hair is given at the bottom of Table I, p. 464. By referring to
these data it will be observed that the average percentage of
sulphur in the ten samples of indian hair is 4.82, the average per-
centage of nitrogen is 15.40, the average carbon value is 44.06
per cent and the average percentage of hydrogen is 6.53.

Data for the ratio, S: N, are given in the last column of Table I,
p. 464. An examination of these ratios reveals the fact that the
lowest ratio, 1:3.0,is found in connection with the full-blood
female Pima indian, whereas the highest ratio, 1:3.4, occurs in
connection with the half-blood female Klamath indian.

2. Chemical composition of indian hair as influenced by the
sex of the tndividual and the purity of breeding. A comparison
from the standpoint of the sex of the individual and the purity
of breeding is shown in Table II, p. 465. By referring to this
table it will be noted that the two extremes of sulphur data occur
in connection with the hair of the half-blood indian, the lowest per-
centage (4.69) being present in the hatr of the halj-blood female and
the highest percentage (5.03) occurring in the hair of the half-blood male
indian. On the other hand the hair of the full-blood male and
female indians contains practically the same percentage of sul-
phur (4.85 and 4.86). Compared as to sex without reference
as to whether the specimen is from half- or full-blood indians we
find that the hair of the female indians contains somewhat less
sulphur than that of the male indians, the percentage of sulphur
in the hair of the females being 4.79 and in that of the males being
4.91. It will be further observed that, when compared as to
the purity of breeding without reference to the sex, the hair of
the full-blood indians contains a higher percentage of sulphur
(4.86) than the hair of the half-blood indians (4.77). It is
interesting to note that whereas the hair of the full-blood indians
of both sexes contains practically the same percentage of sulphur
(4.85 and 4.86) the hair of the half-blood female indian shows
a lower percentage of sulphur (4.69) and the hair of the half-
blood male indian shows a higher percentage (5.03).

By an examination of the data contained in Table II, it will be
observed that the lowest percentage of nitrogen (15.20) occurs
in the hair of the full-blood female indian where as the highest
nitrogen content (15.60 per cent) is found in the hair of the half-
blood male indian. In this connection it may be remembered

464

Nn
ea)
HSH SH SH Sh Xt tt SH XH 1

ao
Dt

62°
€6°

ov)
i? 6)

1D Py oD OD
ae york)

|

S co ea 616% | €9°9 | 90° FF | OF ST
le ape sue? (eater sae ie. oe
= feet cg'Ig | #0°9 | 89°IF | 09'ST
5 e272 T 88°08 | 21°9 | Sh sr | F9°ST
ms Sec 61°0E | 19°9 | 66°F | bF'ST
5 pe: T 28S | L2°9 | 99°SF | 8z'ST
o ee. T 8I°8z | 12°9 | 2Z9°FR | ZL°ST
“m ee: T $9°6Z | #9°9 | GO'eh | LZP'ST
fe) Ze: } g€°LZ | 69°9 | 69'SF | 8F'ST
i ek ea | 79°2c | OL Z| S8° FF | FO'ST
& Le: 1 | 21°82 | 09°9 | 99°9% | 88°71
5 Oss T 16°6Z | 92°9 | $9°SF | O08 FT
4 es O | H +o) N

& ae

Ee Onee "HIVH AO NOLILISOdWOD GDOVINGOUGd
Vv

xs ss green SPs

O

‘*"Irey UvIpur jo soydurvs (QT

Jo uorpIsoduioo esvzusoied aseioay

JIPH
Tha
Thal
eH
eH
Je
Wha
Ma
Tha
aan F

(saeoA)

“Pood

‘nV

STC

ae

ale
a VUIO J
a[ BULA yy
a[BUld YT
a[ BULA YT
ayewoay
a BULA T
a[eulay

“Loardas Ao

NOILdTHOSad

aouMPYS
ojqond
soyedeiy
YEW] ST
ueysel[y
XNOoIg
auoysoys
emoddiy)
210
Bulg

OqIEyL

OL

PAN OD HID SS I~ CO OD

‘oj dureg
10. "ON

“MIVH NVIGNI AO NOILISOdNOD IVOINAHO

‘I aTav4

Thomas A. Rutherford and P. B. Hawk = 465

that the lowest percentage of sulphur occurs in the hair of the
half-blood female indian instead of the full-blood female indian
which contains the lowest percentage of nitrogen. Comparing
the data obtained from the analyses of the hair of the full-blood
indians we observe that the hair of the male contains consider-
ably more nitrogen (15.54 per cent) than the hair of the female
(15.20 per cent), and the same observation holds true for the
hair of the half-blood indians, 7. e., the percentagé of nitrogen in
the hair of the half-blood male (15.60) is somewhat higher than
the percentage of nitrogen in the hair of the half-blood female

TABLE II. CHEMICAL COMPOSITION OF INDIAN HAIR AS INFLUENCED BY
THE SEX OF THE INDIVIDUAL AND THE PURITY OF BREEDING.

AVERAGE PERCENTAGE COMPOSITION OF
HAIR. Ratio
Sex. Blood. S:N
s N Cc | H O
Female Full ANSGn | 1oeZO44.901) | sOnvomboseo si leo
Female Half 4.69 | 15.49] 44.63) 6.50 | 28.69/1 :3.3
Female Full and
half AED) | L522) (442 79) 6565) | 28e4 5 eee
Male Full 4.85 | 15.54] 42.71] 6.37 | 30.53 ]1 :3.2
Male Half 5.03 | 15.60] 41.68] 6.04 | 31.65/11 :3.1
Male Fulland
half 4.91 | 15.56 | 42.37] 6.26 | 30.90|1 :3.2
Male and
female Full 4786) | oot n44 li 6.62) 29504 ieee
Male and
female Half AVG ize | ayes PA|| 2B yates) || ese) || Hey Zyl a 33.8:

eee Pee bai} 4.82 |15.40| 44.06] 6.53 | 29.191 -3.2

of all specimens of indian hair

(15.49). Asthe result of a comparison of sexes without reference
to the purity of breeding it is seen that the hair of the male
indians is much richer in nitrogen than the hair of the females,
the percentage of nitrogen in the hair of the male being 15.56
and that in the hair of the female being 15.32. On the contrary,
a comparison as to the purity of breeding without considering
the sex of the individual, shows a balance of nitrogen in favor
of the half-blood indians, the percentage of nitrogen in the hair
of the full-blood indians being 15.31 and that in the hair of the
half-bloods being 15.52.

466 Chemical Composition of Hair

Carbon is seen to be present in smallest percentage (41.68) in
the hair of the half-blood male indian and present in the largest
percentage (44.91) in the hair of the full-blood female indian,
and it is very interesting to note that the order is exactly the
reverse of that in which the nitrogen occurs, 7. e., the largest per-
centage of nitrogen is found in the hair of the half-blood male and
the smallest percentage in the hair of the full-blood female
indian. Considering the female indians apart from the males
it may be observed that the carbon follows the sulphur in showing
a higher percentage in the hair of the full-blood female (44.91
per cent) than in the hair of the half-blood female (44.63 per
cent). In the case of the male indians however, the carbon fol-
lows a course the reverse of that taken by the sulphur and the
nitrogen since it exhibits a higher percentage (42.71) in the hair
of the full-blood male than is present in the hair of the half-blood
male (41.68). A comparison of sexes shows the percentage of
carbon in the hair of the female indians (44.79) to be considerably
higher than that in the hair of the male indians (42.37) and in
this point the occurrence of the carbon content again is the reverse
of that of sulphur and nitrogen. Compared from the standpoint
of purity of breeding, without reference to the sex of the indi-
vidual, we observe that the carbon is present in larger percentage
in the hair of the full-blood indians than in that of the half-bloods
and in this it exhibits a similarity to the sulphur content but is
the reverse of that of nitrogen. Hydrogen runs parallel with the
carbon in every instance.

The relation of the sulphur content to the nitrogen content is
stated in the last column of Table II, p. 465. The most striking
thing observed in this connection is the fact that the ratio, S: N,
for the specimens of hair from the male indians, without respect
to the purity of breeding, is exactly the same as the ratio in the
case of the specimens of hair from the female indians, 7. é., 1:3.2.
Considering the female indians. alone we observe that the ratio
varies with the purity of breeding, being 1:3.1 for the full-blood
females and 1:3.3 for the half-blood females. In the case of the
male indians the ratios run differently since they show a ratio of
1:3.2 for the full-blood indians and 1:3.1 for the half-bloods.
It is interesting to note that the ratio 1:3.2 is common to four
of the eight classes included in the table.

Thomas A. Rutherford and P.B. Hawk 467

3. Chemical composition of negro hair. Fourteen specimens
of negro hair were analyzed, five of them being obtained from
dead individuals and the remaining nine from living individuals.
Five of the latter were full-blood negroes ranging from seventeen
totwenty-three years of age, two were half-blood negroes eighteen
years of age, and two were three-quarter blood negroes eighteen
and twenty years old, respectively. Of the fourteen specimens
of negro hair analyzed eleven were obtained from females and
three from males.

In Table III, p. 469 are summarized the data from the chemical
analyses of the fourteen samples of negro hair. The first impres-
sion gathered from an examination of the table is the general
similarity of the analytical results with those from the analyses of
indian hair, as set forth in Table I, p. 464. There ts also observed
to be a well marked tendency for both the sulphur and the nitrogen to
occur in higher percentage in the hair of full-blood negroes than in
the hair of the negroes less purely bred, this higher percentage occur-
ring trrespective of the sex of the individual. It will also be noted
that the analyses of the hair of dead negroes show a rather lower
average percentage of sulphur and a somewhat higher average per-
centage of nitrogen than 1s shown in the data from the analyses of
the hatr of the living negroes. These three points, mentioned in
this general way, are more fully discussed on pp. 470-476.

Considering the sulphur content of the various specimens of
negro hair we observe that the lowest percentage (4.53) occurs
in the hair of a dead female negro (sample 2), whereas the highest
percentage (5.15) occurs in the hair of a living full-blood female
negro (sample 8). Nitrogen follows the sulphur in being found
in the largest percentage (15.55) in the sample containing the

‘greatest abundance of sulphur but differs from it in showing the
lowest percentage (14.51) in the hair of the living, three-quarter
blood female negro. The lowest percentage of carbon (42.43)
is found in the hair of a living male half-blood negro (sample 14),
and the highest percentage (45.70) in the hair of a living, full-
blood female negro (sample 6). In this connection it is exceed-
ingly interesting to note that an examination of the data for
carbon reveals the singular fact that the hatr of the impurely bred
negroes, 1.e., the half- and three-quarter bloods, contains, uniformly
considerably less carbon than the hair of any of the other negroes,

468 Chemical Composition of Hair

irrespective of the sex of the individual (see data for samples 9, to, 11,
and 14).

Hydrogen follows the carbon in showing the lowest per-
centage (5.60) in the hair of a living, half-blood male negro
(sample 14) and the next lowest (5.97) in the hair of a living,
half-blood female negro (sample 11), but differs from the carbon
in the fact that it is found in greatest amount (6.90 per cent) in
the hair of a living full-blood female negro (sample 7).

The average percentage composition of the fourteen samples of
negro hair is given at the bottom of Table III, p. 469. From this
it will be observed that the average percentage of sulphur is
4.84 and the average percentage of nitrogen is 14.90 whereas
the carbon and hydrogen averages are 43.85 per cent and 6.37
per cent respectively.

The data for the ratio, S: N, in the various samples of negro
hair, will be found in the last column of TableITI, p. 469. Among
the interesting observations in relation to this ratio is the general
uniformity of the ratios for the samples of hair of the living,
full-blood negroes and the decided variation of this ratio from
that obtained from the data for the samples of the hair of dead
negroes. The ratio for the hair of the living, full-blood individuals
is 1:2.9 fortwosamplesand 1 :3.0 for three samples whereas the ratio
for the hair of the dead individuals is 1:3.1 for two samples, 1:3.2
for twosamples, and 1:3.4 for one sample, making a variation from
I :3.1 to 1:3.4 as contrasted with the lesser variation from 1:2.9 to
1 :3.0 noted in the case of theliving, full-blood negroes. The two
instances in which the ratio for the living individuals is above
1:3.0 are both impurely bred negroes, one being a three-quarter-
blood female and the other a half-blood male. The hair of each
of these individuals shows a ratio of 1:3.2. An examination of
the ratios for the various samples shows conclusively that the
average ratio for the impurely bred individuals is considerably
above the average ratio for the full-blood individuals. These
average ratios are further discussed on p.474. The average ratio
for the entire fourteen samples of negro hair is 1:3.1.

4. Chenucal composition of negro hair as influenced by the sex
of the individual and the purity of breeding. In Table IV, p. 471,
is portrayed the chemical composition of the hair of living and
dead negroes as influenced by the sex of the individual and the

Se eS eee

= —~:

469

Thomas A. Rutherford and P. B. Hawk

|
ee res aa es | Sie ie 5 Wife Gdice hoc QiorcrOrG ue O10" Oebd ITBU oigou Io said
ee Papoose caer e Mere re | ene rel aan FI jo uorptsoduros sseveaoreel og areare
GAS SW tent ve 09'S SP GP OS FI 69 SUIAY] ST yey oP al
‘sudU | |
-peds % ‘od (SS teh Ay Oks SIE 8) le ey v9 VI 66° 7 BUIATT |Ezpueoz| Tha aye WW eI
-uresopisoduio) |
(je [fi tks Oks | OG © SV &P €6° FT Sie BUIATT | 02 mM ae ZI
GaGa by Los L6°G | PP GP 89 FI 00°¢ BULATT ST eH qewoy | II
G&-1T | 89 Té S79 | 06°GF TG FI 99° PT SUIAV] 0Z Joyrenb
-so1y, | aJewog | OL
0O°€: I | 46°08 99°9 Il &7 oo Wi SUIALT ST Joyrenb
| -s01Iy, | oyewog | 6
“SUSU
-ods ¢ ‘afd | Oese La leeeG S629) 02a Cosi GIG BUTAYT 61 IMq_ | eeuoyq | g
-uresoyisodwo0g
_ ‘“suour
-peds 7 ‘afd | Overy Talks Os 06°9 | G4°Sh | 92°FT G6°h | Sutary SI ing | eeweg | 2
-uresoyIsoduro)
*‘SUOUL |
oeds ¢ ‘fd | 6°: T | 09°82 GI'9 | OL SP | 69°FT 96°F | Sutary LT ng |epeueg | g
-tuvsoyisodtuo)|
Ge 1) Ga 6c 969 19 PP VL ST Lilia peoq | 0S-0F qeulay | g
Ee L\sS) 66 c& 9 Sl UP ST ST 8 7 peoq | 0S-OF grew | FF
eS Wt |] Ges Ge LL’9 08 FP 92° FT 18 peoq | 0S-OF gewmoy | ¢§
vy €:T | 20°6¢ 699 9L° FP Lena a peoq | 0S-OF gewoyq | Z
Goo) a 00606 OF'9 08 €P tei] TCSIL (Qa Ay prod 0S-OF qemu | T
i: "pve siBok
ae 0 H 9 N 8 10 Seanrch ‘od V POSTE os Lp 2
eat ¥ BS
ie a “HIVH AO NOILISOdWOO TOVINGOUAd “Logdradoas JO NOIWdalNoOsaa Zo
‘MIVH OUDAN AO NOILISOMWOO IVOINAHO “Ill ATAVL

470 Chemical Composition of Hair

purity of breeding. Considering first the sulphur content, one
of the most striking points we observe is the similarity between
the data for the living, full-blood female negroes and the living
full-blood male negroes, the percentage of sulphur being 5.02 for
the former and 5.03 for the latter. The value 5.03 for the full-
blood male negroes is the highest percentage of sulphur found in any
of the samples of negro hair analyzed. The lowest percentage of
sulphur (4.64) is found in the hair of the living, three-quarter
blood female negro and the next lowest (4.69) occurs in the hair
of the living, half-blood male negro. Grouping the female and
the male negroes together for purposes of comparison we observe
that the average percentage of sulphur is 5.02 for the full-blood
negroes and the lower value of 4.84 for the half-blood negroes.
The most striking variation, however, in the sulphur content of
the hair of the full-blood and half-blood negroes is noted in the
case of the male subjects in which connection it will be observed
that the hair of the half-blood individuals contains only 4.69 per
cent of sulphur whereas the hair of the full-blood individuals contains
5.03 per cent, which is, as has just been mentioned, the maximum
percentage of sulphur for the negro hair examined.

Comparing the data from the analyses of the hair of the dead
female negroes with the data from the analyses of the hair of the
living, female negroes we observe that the percentage of sulphur
in the hair of the former (4.73) is somewhat lower than that in
the latter (4.89). One of the most interesting points emphasized
by this table is the very close agreement in percentage of sul-
phur shown in the analyses of the hair of living, male and female
negroes without reference to the purity of breeding. In this
connection it will be noted that the hair of the female negroes
contains 4.89 per cent of sulphur and the hair of the male negroes
contains 4.92 per cent of sulphur, the variation being but 0.03
per cent which is within the limit of experimental error for such
determinations. The average percentage of sulphur in all the
specimens of negro hair without reference to the sex of the indi-
vidual or the purity of breeding is 4.84 and is identical with that
determined for the living, half-blood male and female negroes
and within the limit of experimental error of that determined
for the living and dead females (4.82), for the living, full-, half-
and three-quarter blood female negroes (4.89) and the living,
full- and half-blood male negroes (4.92).

47I

ehp) eleylaveie: sieveveceletererajajeta.eie) ear Shove Pibae. ev. sere mnn Les se 88) Irey o1s0U
LsGr ok FO 0€ LE'9 G8" SP 06 FI saa, es suouttoeds [[e Jo uorptsoduroo aseyuooied asvi9Ay
O;€: T 61 GE 64°S PV GV PL FT v8 '¥ BULATT FIP o[VUlof PUP O[PI
(OS oF VL LV’ 6G tr '9 SL PFT 68 FI c0'S SUTATT TM o[eUlof PUL OPIN
OFS Tt Gals 86°¢ 90° SP 62°71 c6°P SUTATT yey pure [hy STP IN
6% 0. [I SP SE 09'S Ev GP 08 FI 69°F SUTATT FTP ; STP IN
GinGrel: €9'°0€ SI'9 LE SP 62° FT €0'°¢ SUIAYT TH eTeI
[etsel! CLG LV'9 90° FF Somme GS peop pu surAry sTeulo yy
Gaara 82 62 9F'9 GV PV IT'ST ELV peod AACS NS f
Oy ts ol $0°0§ 6F 9 MP ASIP LL° 01 68° Pf SUIAVT ‘yey pure 104
-renb-901y} [ny a[eUld yy
Gicemal 16° LE L6°¢ PP GP 89°F 00°¢S SULAYT FIPH o[BUlo yy
Lean Sai Te Gg'9 00° &F e¢ FI 69'F SUIAYT Joqyrenb-oo1y L, oT PUI9
(Se | 69°83 199 (GQ VAY 96 FT c0'¢ SUIAVT TM oTeUlo
O H 10) N $8
N:S s =e > 3 : | “peep io Surry “poold *xog
one a “"WIVH JO NOILISOANOOD GANVINAOUAd ADVUAAV

Thomas A. Rutherford and P. B. Hawk

‘ONIGGAAUA AO ALINNd AHL AGNV TVOCIAIGNI
HHL JO XAS AHL AM GAONANTANI SV SHOUIDAN AVA ANV ONIAIT AO MIVH AHL JO NOILISOUWOO TVOINAHO ‘AI ATAVL

472 Chemical Composition of Hair

The most noteworthy observation in connection with the nitro-
gen content of the samples of negro hair is the variation of the
percentage of nitrogen in the hair of the dead female negroes
from the percentage of nitrogen in the hair of all the other classes
of living subjects whether male, female, full-, half- or three-quarter
blood. The hair of the dead females contains 15.11 per cent of
nitrogen whereas that of the other classes of subjects varies from
14.53 per cent to 14.96 per cent. One of the most interesting
points concerning the nitrogen content of the samples of negro
hair is the uniformity in the percentage of this element which is
present in the hair of the various classes of living subjects as
grouped in Table IV, p. 471. By referring to the table it will be
seen that, with the exception of the hair of the three-quarter
blood female negroes, all the percentage data for the nitrogen
of the hair of all the classes in the table may be grouped together
as duplicates which are almost within the limit of experimental
error. Of such variations as occur in these data the most striking
is the tendency of the nitrogen content to follow that of sulphur
in being higher in the hair of the full-blood negroes than in that
of the half-bloods, e. g., 14.89 per cent of nitrogen in the hair of
full-bood male and female negroes and 14.74 per cent in the hair
of the half-blood negroes. The average content of nitrogen for
the entire fourteen specimens of negro hair analyzed is 14.90
per cent, a value which is practically duplicated by all classes in
Table IV except the dead female negroes and the living, half- and
three-quarter blood female negroes.

Carbon shows a well developed tendency to be present in the
greatest amount in the hair of the purest bred negroes. This is
most strikingly indicated in the data for the living female sub-
jects where the hair of the full-blood individuals is shown to
possess 44.72 per cent of the element, that of the three-quarter
blood female 43.00 per cent and finally the half-blood’s hair
showing 42.44 per cent of carbon. The series shows a decreasing
carbon content as the purity of breeding is lowered, the maximum
percentage occurring in the hair of the purely bred individual and
the minimum in the hair of the half-blood. This series also prac-
tically includes the two extremes of the eleven classes as grouped
in Table IV, since 44.72 per cent is the highest of those tabulated
and 42.44 per cent is practically the lowest since it is almost

= “Se

Thomas A. Rutherford and P. B. Hawk 473

identically the same as the value (42.43 per cent) for the carbon
content of the hair of the living half-blood male negroes which
is the actual minimum percentage of carbon. This tendency to
exhibit a variation in carbon content according to the purity of
breeding is shown likewise in the case of the male negroes. In
this instance the hair of the full-blood individual contains 43.37
per cent of carbon while that of the half-blood contains but 42.43
percent. Grouping the males and females together we obtain as
the averages 44.18 per cent of carbon in the hair of the full-blood
negroes and 42.44 per cent in the hair of the half-bloods. In this
last particular the carbon follows the tendency already noted
in connection with sulphur and nitrogen. Dividing the female
negroes into living and dead subjects we observe that the hair
of the dead subjects contains considerably more carbon than that
of the living subjects, e. g., 44.42 per cent being present in the
hair of the former and 43.77 per cent in that of the latter. In
this regard the carbon again follows the course of the nitrogen.
Comparing the living females with the living males we note that
the hair of the former contains a higher percentage of carbon than
the hair of the latter, the averages being 43.77 per cent and 43.06
per cent respectively. The average percentage of carbon in the
whole series of samples of negro hair was 43.85, a value most
nearly duplicated by the hair of the living, female negroes (43.77),
the living, full-blood male and female negroes (44.18) and the
living and dead female negroes (44.06).

There is a very well defined tendency for the hydrogen to
follow very closely the same scheme of distribution as that noted
for the carbon content. First of all we observe the same general
decrease in the percentage of hydrogen coincident with a lessen-
ing of the purity of breeding. This feature is particularly well
illustrated in the case of the living, female negroes, where, it will be
noted, the hair of the full-blood individuals contains 6.61 per cent of
hydrogen, that of the three-quarter blood 6.55 per cent and that
of the half-blood 5.97 per cent. This variation according to the
purity of breeding is also well illustrated by the data for the analy-
ses of the hair of the male negroes. Here the hair of the full-
blood individual possesses a hydrogen content of 6.18 per cent as
contrasted with a content of 5.60 per cent for the hair of the half-
blood negro. The hydrogen again follows the carbon in showing
its maximum percentage in the hair of the living, full-blood

474 Chemical Composition of Hair

female and its minimum percentage in the hair of the living,
half-blood male. However, like the sulphur it is present in
slightly higher percentage in the hair of the living, female negroes
than in that of the dead female negroes. In a comparison of the
living females with the living males it is noted that the hydrogen
follows the carbon in being present in largest percentage in the
hair of the former, the percentage being 6.49 for the hair of the
females and 5.98 for the hair of the males.

The noteworthy uniformity with which the hair of the full-
blood negroes, irrespective of sex, contains a higher percentage
of each of the constituents determined by analysis, than the
hair of the half-blood negroes is very strikingly portrayed in the
lower portion of Table IV. Sulphur, nitrogen, carbon and hydro-
gen are found with absolute regularity in higher percentage in the
hair of the full-blood negroes than in the hair of the half-blood
negroes.

The ratio, S: N, is the lowest (1:2.9) for the living, half-blood
female negroes and living, full-blood male negroes, and highest
(1:3.2) for the dead female and the living, half-blood males. The
average ratio for all specimens of negro hair is 1:3.1.

5. Comparison of the chemical composition of indian and negro
hair. By careful examination of Tables II and IV, pp. 465 and
471, some interesting observations may be made upon the com-
parative chemical composition of the hair of the indian and
negro. Referring first to the sulphur content it will be observed
that the percentage limits are virtually identical in the two in-
stances, the highest percentage of sulphur in the indian hair being
5.03 and the lowest being 4.69, whereas the similar data for negro
hair are 5.03 and 4.64. It is of interest in this connection to note
that in each instance the hair of the male subject contained the
higher percentage of sulphur, one of the most striking comparative
observations portrays the influence of purity of breeding upon
the chemical composition of the hair of the individual. An
examination of the tables will show the great uniformity with
which the hair of the full-blood male or female indian or negro is
shown to contain a higher percentage of sulphur, carbon and hydro-
gen than the hatr of the half-blood, and, in the case of negro hair,
nitrogen also. The only exception to this rule is the nitrogen
content of indian hair where the hair of the impurely bred male

Thomas A. Rutherford and P. B. Hawk = 475

or female subject is, in every instance, observed to contain a
higher percentage of nitrogen than that of the full-blood subject.
The variation in the chemical composition according to sex is
clearly shown in the case of the sulphur content also. In both
tables it will be seen that the hair of the male subjects contains
a higher percentage of sulphur than the hair of the female sub-
jects, the data for indian hair being 4.91 and 4.79 and the data
for negro hair being 4.92 and 4.89. The average percentage of
sulphur is practically the same for each variety of hair, e. g.,
4.82 for indian hair and 4.84 for negro hair.

The most striking lack of uniformity is shown in the compari-
son of the nitrogen content. The hair of the indian contains
uniformly a higher percentage of nitrogen than that of the negro.
This fact is emphasized by an examination of the tables which
reveals the fact that the lowest percentage of nitrogen deter-
mined for indian hair (15.20) is nevertheless somewhat higher
than the maximum percentage of nitrogen (15.11) in negro hair.
The percentage of nitrogen in indian hair varies from 15.20 to
15.60 whereas the percentage of nitrogen in negro hair varies
from 14.53 to 15.11. Indian hair contains an average of 15.40
per cent of nitrogen while the average nitrogen value of negro
hair is 14.90 per cent.

An examination of the data for the carbon determinations
reveals another interesting similarity. It is there shown that
the maximum percentage of carbon in the hair of both the indian
and the negro is found in the hair of the full-blood female and
the minimum percentage of carbon is found in the hair of the
half-blood male. The tendency ts well developed jor the hair of
the full-blood individual of either race to contain a higher percentage
of carbon than the hair of any individual of that race less purely
bred. The average percentage of carbon for the indian hair
(44.06) is somewhat higher than the average for negro hair
(43.85) but the difference is almost within the limit of experi-
mental error for carbon determinations. Contrasted according
to sex it will be observed that in each table the hair of the females
contains a higher percentage of carbon than that of the males.

Hydrogen follows carbon in the main characteristics. For
instance we note the same marked tendency for the percentage of
hydrogen to decrease as the purity of breeding is lowered, as well as

476 Chemical Composition of Hair

the strict uniformity with which the sex appears to a degree to regu-
late the chemical composition of the hair. Thus we find that the
hair of the full-blood individuals contains a higher percentage
of hydrogen than that of the half-blood individuals, and it is also
noted in every instance, that the hair of the females contains a
higher percentage of hydrogen than that of the males. In com-
mon with carbon, hydrogen is also present in maximum percent-
age (6.75) in the hair of the full-blood female indian and in
minimum percentage (5.60) in the hair of the half-blood male
negro. The average percentage of hydrogen for the whole series
of samples of indian hair (6.53) is somewhat higher than the
similar average for negro hair (6.37).

The ratio, S : N, for the indian hair is higher throughout than
the same ratio for the negro hair. This is due to the uniformly
higher nitrogen content of the former.

TABLE V. COMPARATIVE CHEMICAL COMPOSITION OF THE HAIR OF FULL-
BLOOD MALE INDIANS, NEGROES AND JAPANESE,

PERCENTAGE COMPOSITION OF HAIR. f

Subject. ; Ratio
s we H |. 6 52m
Negro.............. | 5.03 | 14.79 | 43.37'| 6.18 | 30°63: )gigoaem
Japanese........... | 4.96 | 14.64 | 42.99 | 5.91 | 31.50 | 1:3.0
BTICIAT ee hone wets Ae 4.85 | 15.54 | 42.71 | 6.37 | 30.53 | 1:3.2

6. Chemical composition of japanese hair. Unfortunately
but a single sample of japanese hair was available for analysis,
this being the hair of a full-blood male japanese. The data for
the analysis of this specimen are given in Table V, p.476, in which
table are also given, for the purpose of comparison, the average
data for the chemical composition of the hair of full-blood male
indians and negroes. It is rather interesting to note the similar-
ity in the chemical composition of these samples of hair from
three distinct races, in fact, the agreement of the data, with the
exception of the nitrogen content, is as close as that previously noted
as existing in some instances between the data for different sexes of
the same race, or between those noted for full-blood and half-blood
individuals of the same sex. Considering, for example, the sul-
phur content, we observe that the percentage of this element in
the japanese hair is an approximate mean of the percentage of

Thomas A. Rutherford and P. B. Hawk 477

sulphur in the hair of the indian and the percentage of sulphur
in the hair of the negro, being a slightly lower percentage than
that obtained from the analysis of negro hair and slightly higher
than the determined percentage of sulphur in indian hair. We
do not note the same uniformity in the nitrogen content of the
three varieties of hair. There is rather close agreement between
the data for negro and japanese hair, the percentage in the former
case being 14.79 and in the latter case 14.64, but the percentage
of nitrogen in the hair of the indian is considerably higher (15.54),
in fact, as may be seen from a comparison of Tables II and IV,
the percentage of nitrogen 1s uniformly high for all the specimens of
indian hair as compared with the similar data for the negro hair.
In connection with the carbon content the same conditions
obtain as those noted in the discussion of the sulphur content,
4. e., the percentage of carbon in the hair of the japanese is an
approximate mean of the similar data for the negro hair and
indian hair, being somewhat lower than the percentage of carbon
in the hair of the negro and somewhat higher than the percentage
of carbon in the indian hair. The percentage of hydrogen is
similar to that of nitrogen in showing the highest percentage
(6.37) in the hair of the indian and the lowest (5.91) in the hair of
the japanese, the percentage of hydrogen in the hair of the negro
being an approximate mean of those percentages. The ratio,
S :N, is lowest (1:2.9) for the hair of the negro, slightly higher
(1:3.0) for the hair of the japanese, while the maximum ratio
(1:3.2) occurs in connection with the indian hair.

7. Chemical composition of caucasian hair. There were, in
all, twenty specimens of caucasian hair subjected to analysis,
thirteen of which were obtained from adults. The remaining
seven specimens were the hair of children ranging from two and
one-half to twelve years of age. In order to be able to interpret
the results more satisfactorily we will consider the analyses of
the caucasian hair in two divisions, 7. e., adults and children.

(a) Chemical composition of the hair of adult caucastans.
Thirteen specimens of hair obtained from adult caucasians were
analyzed, five of them being from dead females and eight from
living males.

The data for the analyses of the hair of adult caucasians are
given in Table VI, p. 478. Referring to this table and consider-

1g

i He

| Composition o

Chemica

478

lor]
N

NONFMOnMMNSCHWO
CO OD OCI ANNMOAAN

N:S
oney

Se i oo Be ce i eB oe oe Oe Oe ee

‘82 | bh'9 | 6b OF
eh'4z | 99°9 | L0°SF
06°93 | 99°9 | 8I°SF
86°22 | 99°9 | L9°bF

| 20°26 | L8°9 | 6L°S>
DORE 9 \eLe" ey
z9'6Z | 929 | SIF
| ¥°8z | 92°9 | IP FF
9€°8Z | 29°9 | FP FP
Tebe | .29°S | e877
926% | S¢°9 | Z0'FP
16°6Z | 98°9 | 96°&F
8°82 | 0&9 | 99°FF
82°82 | #29 | SL°FP
O H 9)

61ST
DE<SE
61 ST
SI ST
9S5Sir
66 FI
GéST
Le St
Te "ST
OL FT
G8 FI
46° FI
GG ST
P8 ST
N

€Z°
L0°
Si
9G"
CL"
£9"
69°
Lo"
6F°
68"
PL
00°
F6°

ae {

SSH OALQ SH SH SH LD 1D 1D HD 19 6 10

rey
uvisvoneo jo safdures ¢] Jo uorztsoduioo aseyusoiod asvisoay

SUTAY'T
SUIAVT
SULATT
SUIAYT
SUTATT
SUIATT
SUIAYT
SUIATT

pve
IO DSULATT

*HIVH dO NOILISOdWOO GADVINGONAd

“SNVISVOOVO LTINGV AO UIVH AHL AO NOILISOGWOOD IVOINAHO

‘Oo V

Se
oP
a(R]
ae
ae
ae]
ae
ae
a[VULd J
a[BULI
a[BUla yy
a[VULI J
a VU J

*xag

“~Loardoas dO NOIsdINoOsaa

‘IA ATAVL

pew
pew
UMOIQ YIeqd
UMOIG
YsVT
VYySrT
WYysrT
rysVT
UMOL q
UMOIG
Avr)
APIn)
ABIX)

‘ojdureg
jo 10]OH

ra OD sH 19 6 P00

‘oydureg
JO “ON

Thomas A. Rutherford and P. B. Hawk = 479

ing first the sulphur determinations we are struck at once by the
general tendency of the percentage of sulphur in these samples
to be much higher than the percentage of sulphur in the hair of
any of the races already considered. The maximum percentage
for the series is 6.07 which was found in the hair (red) of a living
male twenty-six years of age, whereas the minimum percentage
for the series is 4.49 which was found in the hair (brown) of a
dead female of unknown age. This tendency for the higher per-
centages of sulphur to be found in the hair of the living males is
quite well marked as will be seen from an examination of the
table. This point is more fully discussed on p. 482. The fact
that each of the samples of red hair analyzed contained a higher
percentage of sulphur than any of the other samples of hair, all races
and sexes included 1s particularly worthy of note. The nitrogen
determinations, as a rule, are seen to be higher than they were in
the case of the negro hair and lower than were found in the indian
hair. The maximum percentage of nitrogen (15.84) is found in
the hair (gray) of a dead female of unknown age, whereas the
minimum percentage of nitrogen (14.70) occurs in the hair (brown)
of a dead female of unknown age. After examining the data for
the carbon we note that they follow the sulphur in showing the
same well marked tendency to be higher than the similar data
for the hair of any of the other races examined. The maximum
percentage of carbon is 45.79 which occurs in the hair (brown) of
a male twenty-seven years of age, whereas the minimum per-
centage (43.15) occurs in the hair (light) of a male twenty-one
years of age. There is no striking difference in the general
trend of the percentage of hydrogen in the specimens of cauca-
sian hair from the percentages as noted for the hair of the three
other races examined.

The average percentage composition of the thirteen samples
of the hair of adult caucasians is given at the bottom of Table
VI, p. 478. It will be noted that the average percentage of
sulphur for this variety of hair is higher than that of any of the
other varieties of hair examined. This value for caucasian hair
is 5.22 per cent and is followed in descending sequence by japan-
ese hair (4.96 per cent), negro hair (4.84 per cent) and indian
hair (4.82 per cent). The average percentage of nitrogen is 15.19,
a percentage which is exceeded only by the percentage of nitrogen

480 Chemical Composition of Hair

in the hair of the indian (15.40). The similar data for negro
hair is 14,90 per cent and for japanese hair 14.64 per cent. In
common with the data for the percentage of sulphur the average
percentage of carbon in the hair of adult caucasians is higher
than that for the hair of the three other races considered, the cau-
casian hair containing an average of 44.49 per cent of carbon and
being followed in descending sequence by indian hair (44.06),
negro hair (43.85), and japanese hair (42.99). The average per-
centage of hydrogen in the caucasian hair is 6.44 and is higher
than the percentage obtained for japanese hair (5.91) and negro
hair (6.37) but is lower than the average of hydrogen in indian
hair (6.53).

The ratio, S : N, undergoes more pronounced variations in the
analyses of caucasian hair than were observed in the data for
the analyses of the hair of other races. This ratio varies from
1:3.3 to 1:2.s, the higher ratio being that obtained from the
analytical data for the hair (brown) of a female of unknown age
and the lower ratio being possessed by the hair (red) of a male,
twenty-six years of age. The next lower ratio, 1:2.6, also occurs
in connection with the red hair of a male twenty-six years of age.
The ratios 1:2.5, 1:2.6 and 1:2.7 arelower ratios than were obtained
for any of the specimens of hair of the other races. The average
ratio for the thirteen samples of hair of caucasian adults is 1:2.9
and is the same ratio as was determined for the hair of the full-
blood male negroes and is lower than the average ratios for the
hair of the japanese (1:3.0) and the indian (1:3.2).

From what has gone before it will be observed that the data
for the average percentage composition of the thirteen samples
of caucasian hair show a well marked tendency for the average
values to run higher for sulphur, carbon and hydrogen than the
similar data for the hair of the other racesmentioned. The per-
centage of nitrogen is also high, being exceeded alone by the
percentage of nitrogen in indian hair.

(b) Chemical composition of the hair of adult caucasians as
influenced by the color of the hair and the sex of theindividual. Data
showing the influence of the color of the hair and the sex of the
individual upon the chemical composition of the hair may be
found in Table VII, p. 481. The most striking point brought
out in this table is the pronounced variation in the sulphur con-

———————— OO Oe es

—_

und

481

Thomas A. Rutherford and P. B. Hawk

yeas a 99'°8Z% PPO 6h PP 61ST 22g ** *‘suourtoads []ev Jo uorpyIsoduroo o8eqyusor10d aseIOAY
Sn ER tsK6 br'9 GG PP ¥S ST 6P'S SUTAVT suountoeds [Ty DN
Og seas 9122 99°9 SI oF GIST 06'S SUIAYT poy ae
Hever I 86°22 99°9 L9 FP SI ST I¢°¢ SULAYT UMOIG IVC o[e IN
Over 20° LZ GE) 6L°SF 9¢°ST 92°¢ SULAT'T UMOIG ae W
Sage 9% 62 Te°9 18° &F &2 ST ee'¢ SULAVT qYysVyT a[eW
Pee cap GG 66 GPO PP PP GIST OSPF prod suouutoeds [Ty o[PUlo yy
EAS iah PS 64 199 GP PP Jaf iA 99°P prod UMOLE so) MLOCSENS |
(Get Ss sGm 1h 10°62 0g 9 OF PP VE GI 68° Ph prod Avr) oTewuoy
O H 0 N 8
N:S “gootqns “IIBEY JO LOTOD *xog
One ey ‘MIVH LO NOILISOAWOO MOVINGOUGd ADVUAAV Pas C Lope o NC
“IVOGIAIGNI AHL JO xasS AHL

ANV YIVH AHL AO YOTOD AM CAONANTANI SV SNVISVONVO LINGV JO UIVH AHL AO NOILISOMNWOO IVOINAHO ‘IIA AITAVL

482 Chemical Composition of Hair

tent of the samples analyzed. As will be seen by an examination
of Tables I and III, pp. 464 and 469, the percentage variation in
the amount of sulphur in the hair of the indian and negro is only
about o.5 per cent (0.48 per cent for indian hair and 0.62 per
cent for negro hair), whereas here in the case of the hair of the
caucasian adults the percentage of sulphur varies from 4.66 per
cent to 5.90 per cent, a variation of 1.24 per cent. The brown
hair of dead females contains the smallest percentage (4.66) of
sulphur, whereas the red hair of living males contains the largest
percentage (5.90). Another very interesting point shown by the
analyses is the high sulphur content of the hatr of the males as com-
pared with the hair of the females the average percentage of sulphurin
the hair of the males being 5.49 and the similar datum for the hair
of the females being 4.80. It must be remembered, however,
that the hair of the females was all obtained from dead subjects
whereas that of the males was taken from the living subjects.
Unfortunately we have no data obtained from the analyses of
the hair of living females or dead males for comparison. Work
along these lines is contemplated. However, we are inclined to
believe that the color of the hair, the age and sex of the individual
as well as the condition of the individual when the sample of hair
is obtained, 7. e., living or dead, contribute to influence the chem-
ical composition. From an examination of Table III, p. 469, it
will be seen that the data there recorded seem to indicate that
the samples of negro hair obtained from the dead subjects con-
tain rather less sulphur than those samples which were obtained
from the living individuals of the same sex.

Considering the sexes separately we find that the highest per-
centage (4.89) of sulphur in the hair of the females 1s present in the
samples of gray hatr, and the lowest percentage (4.66) 1m the brown
hair. In the case of the samples of hair from the males the
minimum percentage (5.26) of sulphur is found in the brown hair
and the maximum percentage (5.90), as already mentioned,
occurs in the red hair.

An examination of the data for the percentages of nitrogen in
the hair samples fails to reveal the pronounced and uniform
variation between the specimens of hair of the two sexes which
was observed in the case of the sulphur. However, the brown
hair of the males contains the maximum percentage of nitrogen

Thomas A. Rutherford and P. B. Hawk 483

(15.56) whereas the brown hair of the females contains the min-
imum percentage of nitrogen (14.77). Taking the average nitro-
gen values for each sex separately, however, we find that the
males show an average nitrogen value of 15.24 per cent and the
females show an average of 15.12 per cent. There are no note-
worthy variations to be observed among the nitrogen data for
the samples of caucasian hair.

The data for carbon and hydrogen follow the nitrogen in being
devoid of any pronounced variations. The average percentage
of carbon in the hair of the females is 44.44 and the similar value
for the hair of the males is 44.52. In the case of hydrogen the
agreement is much closer being 6.42 per cent and 6.44 per cent
for the hair of the females and males, respectively.

The ratio, S: N, appears torun ona lower plane for the caucasian
hair than for the specimens of hair obtained from the represen-
tatives of the other races mentioned. The ratio for the hair of
the females is uniformly 1:3.1, whereas the ratio for the hair of
the males varies from 1:2.6 to 1:3.0 the minimum ratio occurring
in connection with the red haired individuals and the maximum
ratio occurring in connection with the brown haired individuals.
Comparing the individuals by sex it will be seen that the average
ratio for the females is 1:3.1 and the average ratio for the males
is the rather lower ratio 1:28. It has already been pointed out
that the percentage of nitrogen in the hair of the two sexes does
not vary to any considerable degree, therefore this rather wide
variation in the ratio, S : N, shown by the data for the analyses
of the specimens of hair of the two sexes must be due, almost
entirely, to the pronounced and uniform tendency of the hair
of the males to contain a quantity of sulphur notably in excess
of that contained in the hair of the females.

(c) Chemical composition of the hatr of caucasian children.
Seven samples of this variety of hair were analyzed, six of them
being from females. Three of the samples from females were
composite samples made up of from four to twelve different
specimens of hair, the single sample of male hair was also a
composite sample, being composed of hair from nine differ-
ent individuals. Those from whom the various specimens of
children’s hair were obtained ranged from two and one-half to
twelve years of age. Analytical data for these samples of hair

484 Chemical Composition of Hair

may be seen in Table VIII, p. 485. Taking the table as a whole
the most notable feature is the somewhat lower values obtained
for the various constituents of the hair as contrasted with the
similar data (Table VI) for the hair of the caucasian adults.
This point is especially to be noted in the data for sulphur, nitro-
gen and carbon.

An examination of the data for sulphur fails to show any such
variation according to the sex as is shown in the examination
of the hair of adults (Table VI). In the case of the children we
observe that the sulphur values range from 4.77 per cent to 5.10
per cent, a difference of only 0.33 per cent as contrasted with a
variation of 1.24 per cent which is noted in the discussion of the
data from the analyses of the hair of adults. Here for the first
time we find the sulphur content of the hair of the male to be lower
than that of the female. In all the other tables, as will be seen by
an examination, the data show the females to possess hair of a
lower sulphurcontent than thatofthemales. Withthecaucasian
children, however, this order is reversed and the lower sulphur
content is shown to be possessed by the hatr of the males. It should
be mentioned, however, that this conclusion is based upon the
analysis of a single sample of boys’ hair and that further analyses
of other specimens of hair may serve to show the average sulphur
content of such samples to be higher than that of the samples
of hair from the females and thus conform with the data of the
analyses of the samples of hair of individualsof otherraces. How-
ever, the dependability of the data from the analysis of this single
sample is enhanced from the fact that it was a composite sample
containing hair from nine different individuals ranging from
three to seven years of age. Another point to be borne in mind,
in this connection, is the fact. that the samples of hair under con-
sideration were obtained from individuals much younger than
those from whom the other samples of hair were obtained. It
may therefore be possible that the relative chemical composition
of the hair changes with the development of the individual and
that in childhood the hair of the male may contain less sulphur
than that of the female, but that the condition is reversed with
the maturing of the individual and that later in life the hair of
all races would show uniformity in that the hair of the male
would contain the larger percentage of sulphur. It will be

485

Thomas A. Rutherford and P. B. Hawk

-roods 6 ‘atdures oytsodurog

-ioods 6 ‘atdures oytsodurog
-roods zy] ‘a[dures oytsodurog

-toods fF ‘atdures oytsodurog

“SYIBUIO YY

{

‘Iley $,Uuerp[Iyo Jo soyduies » jo
uortsoduroo oseyuso10d asevisAy

02:1 | 08 08 | 9F°9 | Ec: er | 8E°FT | S6'F
Bee cores fh Py 1a) O79 ol cL: er. 19 en | 86 P| 6 Onee seo
sue 1S: 0 | 69708 | eeuo) | GGker OL Tr | LLr 2 ong.) shog
poe 6°3: 1 | 29°08 | 249°9 | 92: eF | Zo FL | 0's | 2109 F | SID
vem OSL S920 06-9 leech SserL | e8e FG OO) | spr)
6.20 \ereee | Slo | South | ro VL Ong @ raza)
8°2: T | £g°6z | 99°9 | PE°FP | 89°F | Gorge D [ID
O2eh ob |etSe Os "Oc On| Yee | OG PL 08" F Z [IID
x ,
ap O H Pe ee Wess

0178 YY ‘MIVH JO NOILISOdWOO ADVENTOUTE “SLOara as

“NGUGTIHO NVISVONVO dO MIVH AHL AO NOILISOTWNOO IVOINGHOD

HO NOILdIyOsSaa

‘IIIA HIaVL

VysVy
UMOIC
yep OF

uMoIq YYSYT

por 04
uMOIq Ye

uMOoIq xed

uMOIq yIeC
uMOIq ye
uinqny

sa] dureg
Jo 10[OD

‘ejdures
TOON!

486 Chemical Composition of Hair

remembered that Mohr! found a lower percentage of sulphur in
the hair of a boy of four than in the hair of a girl of nine.

Two striking points are brought out in connection with the
data for nitrogen, as shown in Table VIII, p. 485. It will be
noticed that the nitrogen percentage in these samples of hair is
uniformly lower than that of any of the other forms of hair
examined, those forms of hair most nearly approaching it in
point of nitrogen content being the hair of the japanese and the
negro. The second point of interest is the close agreement in the
data for the nitrogen content of the varioussamples. The mini-
mum percentage of nitrogen (14.38 per cent) is found in the com-
posite sample of the dark brown hair of girls who were from six
to nine years of age, and further, this is not only the minimum
percentage of nitrogen for the hair of caucasian children but for
each of the varieties of hair investigated. The maximum per-
centage of nitrogen (14.76) is found in the composite sample of
the brown hair of boys who were from three to seven years of
age. The close agreement of the nitrogen values obtained from
the various samples is shown by the fact that the maximum
variation is only 0.38 per cent, 2. e., from 14.38 per cent to 14.76
per cent. This actual difference is very little larger (0.05 per
cent) than that observed in the case of sulphur, and when calcu-
lated upon the basis of the total amount of sulphur and nitrogen
present in the hair the difference between the variation in the
nitrogen values is only a little over one-third as great as the
variation in the sulphur values. The data for carbon do not
show the uniformity observed in the case of nitrogen and sulphur.
The values vary from a maximum of 44.34 per cent for the dark
brown hair of a girl, seven years of age, to a minimum of 41.65
per cent for the dark brown hair of a girl three years of age. The
data for hydrogen also fails to show any particular uniformity.
The minimum percentage of hydrogen (6.18) is found in the
sample of hair containing the least amount of carbon, 7. e., the
dark brown hair of a three year old girl; whereas the maximum
percentage of hydrogen (6.90) is found in the composite sample
of the hair of girls who varied in age from six to nine years.

The ratio, S: N, is rather uniform varying only from 1:2.8 to

1 Mohr: Loc. cit.

ee e—eEeEeEEet

a oe Oe ACF ass at

Thomas A. Rutherford and P. B. Hawk 487

1:3.1. The minimum ratio is possessed by the dark brown hair
of a seven year old girl, and the maximum ratio is possessed
by the hair of the boys. ;

The average percentage composition for the whole number of
samples of children’s hair analyzed is given at the bottom of
Table VIII, p. 485. It will be noticed that the average percent-
age sulphur content (4.93) 1s the same as that determined for
the light hair of girls who were from two and one-half to nine
years of age. The average percentage of nitrogen (14.58) is
practically duplicated (14.56) by the data from the analysis of
the auburn hair of the seven year old girl. The average percent-
age of carbon (43.23) isin very close agreement with the carbon
value obtained from the analysis of the samples of hair obtained
from girls from four to twelve years old, whereas the average
hydrogen value (6.46 per cent) agrees quite closely with the
similar data for the analysis of the light hair of girls two and one-
half to nine years of age. The average ratio, S:N (1:3.0), is
duplicated by three different samples of hair as will be seen by
referring to the tabulated data.

(d) Chemical composition of the hair of caucasian children as
influenced by the color of the hair and the sex of thechild. The data
for consideration here will be found in tabular form in Table
IX, p. 488. An examination of this table will reveal the fact
that the variations among the data for the several specimens of
hair analyzed are not so great in the case of caucasian children
as were noted in the consideration of the hair of the caucasian
adults (see Tables VI and VII, pp. 478 and 481), particularly as
far as sulphur and nitrogen are concerned. In Table IX we see
the maximum percentage of sulphur (5.00) is found in the brown
hair of the girls while the minimum percentage (4.77) is found
in the hair of the same color obtained from the boys. Of the
hair of the females the auburn variety contains the lowest per-
centage of sulphur (4.80) while the brown hair contains the maxi-
mum amount (5.00 per cent) of the element. An examination
of the data for the content of nitrogen reveals the fact that the
samples of female hair contain amounts of nitrogen which are
closely comparable, whereas the specimen of male hair contains
rather more of the element. In this connection it is interesting
to note that the brown hair of the boys contains the largest

488 Chemical Composition of Hair

amount of nitrogen (14.76 per cent) as wellasthesmallest amount
(4.77 per cent) of sulphur. The data for the carbon content of
the various samples analyzed do not run uniformly with either
sulphur or nitrogen but show the maximum percentage (43.87) to
be present in the auburn hair of a girl seven years of age and the
minimum percentage (42.72) to be present in the light colored
hair of girls ranging in age from two and one-half to nine years.

TABLE IX. CHEMICAL COMPOSITION OF THE HAIR OF CAUCASIAN CHILDREN
AS INFUENCED BY THE COLOR OF THE HAIR AND THE SEX OF THE CHILD.

AVERAGE PERCENTAGE COMPOSI-
TION OF HAIR. Ratio

S:N

Subjects. Color of Hair. s Remarks.

Saleen |) Co | Fae en

Girl Auburn 4.80/14. 56/43.87 6.26/30.51

233(0)

1
Girls Brown 5.00/14. 56/43 .12/6.57/80.75)1 :2.9
Composite
Girls Light 4.93/14.51)/42.72/6.40/381.44|1 : 3.0) 4 sample of 9
specimens,
Girls (All specimens) |4.96)14. 55/43. 18/6 .49/30.82|1 : 3.0
Composite
Boys Brown 4.77|14. 76/43. 56/6 .22/30.69)1 : 3.1) 4 sample of 9
specimens.
Average percentage compo- a es Gee ee aie ie. |.
sition of all samples of ¢ |4_93/14.58/43.23/6.46|30.80|1 : 3.0
children’s hair..> 2... .+.
TABLE X. AVERAGE PERCENTAGE COMPOSITION.
| PERCENTAGE COMPOSITION OF HAIR. pane
Subject. | ae i j S:N
Ss | N Kr adG: H O \ente
tien. Mehta ok eee 4.82 |15.40| 44.06| 6.53 | 29.19] 1:3.2
Fpanesd.. ie as ea | 4.96 | 14.64 42.99| 5.91 | 31.50| 1:3.0
Negro. . AS Ay AA ae Ree | 4.84 | 14.90 43-85 | 6237 |30204)| Mise
Caucasian (adulis)\Roee eee 5.22 | 15.79) 44.49| 6.44 | 28.66) 1:2.9
Caucasian (children)...... 4.93 | 14.58 | 43.23] 6.46 | 30.80| 1:3.0

Hydrogen, on the other hand, resembles sulphur in being present
in largest amount (6.57 per cent ) in the brown hair of girls and
in smallest amount (6.22 per cent) in the brown hair of boys.
The ratio, S :N, runs from 1:2.9 to 1:3.0 for the hair of the females
while the male subjects show the slightly higher ratio, 1:3.1.

It is interesting to note the close agreement between the results
of the analyses of the hair of the caucasian children and the hair

Thomas A. Rutherford and P. B. Hawk 489

of the japanese. This is shown clearly by a comparison of the
data for the japanese hair, as given in Table V, with the average
data for the children’s hair as given in Tables VIII or IX.

VI. CONCLUSIONS.

The data obtained from the analyses of specimens of hair from
the representatives of various races indicate that the chemical
composition of the hair is influenced by six factors, as follows:

(t) Race of the individual.

(2) Sex of the individual.

(3) Age of the individual.

(4) Color of the hair.

(5) Purity of breeding of the individual.

(6) Whether the hair sample was obtained from a dead or liv-
ing person.

> ae a we a) rw) =v 3 ¥ \
6 4) ree 7 as h ATT On a
oe ee
Pe ie
Ne
. rian ‘er A ve.
Ad) am yy an» (he
} i be
‘ Paks! oy
* eer’ ype A re
, ” j
Ny wens ap
i oe it ert ee
oe ae i
= ve ; jal - , ’
ey ; :
| alalg
a . 4 '

é sty .

— a i

) sans

i isp ipa

% oe.) ie i

7 . Lair ¥s

*
“a7

CHEMISTRY OF FLESH.
(SIXTH PAPER)!

FURTHER STUDIES ON THE APPLICATION OF FOLIN’S CREATIN AND
CREATININ METHOD TO MEATS AND MEAT EXTRACTS.

By A. D. EMMETT ann H. S. GRINDLEY.

From the Laboratory of Physiological Chemistry, Department of Animal
Husbandry, University of Illinots.

(Received for publication, October 15, 1907.)

Since the publication of the first paper® from this laboratory
upon the determination of creatin and creatinin in meats and
meat extracts further studies have been made to test the accuracy
of the method formerly used. This additional work has been in
part prompted by the paper of Otto Hehner,? who takes the atti-
tude that Folin’s creatin method if directly applied to commercial
meat extracts does not give accurate results. Hehner’s chief
criticism is that 15 cc. of the 1.2 per cent picric acid solution used
in Folin’s method as applied to the determination of creatin in
urine is not sufficient for the determination of this constituent
in commercial meat extracts. In his article Hehner states that
in such cases an increased quantity of the picric acid solution
should be used and he recommends the use of a total of 25 cc. of a
1.01 per cent solution of picric acid together with ‘‘a quite small
amount of alkali,’’ and he says further that ‘‘an excessive quantity
of the alkali diminishes thecolor.’’ This differencein the quantity
and strength of the picric acid which is represented by 6.1 cc. of
a 1.2 per cent solution, gave results according to Hehner which
showed commercial meat extracts (Lemco, Lazenty’s, Armour’s

1H. S. Grindley: Journ. of the Amer. Chem. Soc., xxvi, p. 1086, 1904;
H. S. Grindley and A. D. Emmett: Jbid., xxvii, p. 658, 1905; A. D. Em-
mett and H. S. Grindley: Ibid., xxviii, p. 25, 1905; P. F. Towbridge and
H. S. Grindley: [bid., xxviii, p. 469, 1906; H. S. Grindley and H. S. Woods:
This Journal ii, p. 309, 1907.

2 Grindley and Woods: This Journal, ii, p. 309, 1907.

3 Pharm. Journ., \xxviii, p. 683, 1907.

491

492 Chemistry of Flesh

Baron Liebig’s and Army and Navy) to contain ro to 12 percent
of combined creatin and creatinin, whereas Bauer and Barschall,’
and Grindley and Woods? found by using Folin’s proportion of the
reagents 4 per cent and 1 to 6.5 per cents, respectively, of these
combined extractives. Such differences in the quantities of
creatin and creatinin thus obtained in meat extracts by different
analysts certainly demand thorough investigation, if this con-
venient method of determining creatin and creatinin in such
preparations is to be retained. In this connection it should be
said, that Hehner gives none of his detailed analytical data and
the description of his analytical procedure is so meager that it is
impossible to accurately decide the details of his method.
Further, as a result of the work already published by Folin’ and
by Benedict and Meyers‘ in testing the method with pure creatin,
Hehner’s criticism would at first sight seem to be almost without
any foundation. On the other hand, the extent to which this
method of determining creatin and creatinin is already being
used both in commercial and scientific work demands that the
conditions under which it gives accurate results should be fully
determined experimentally beyond any reasonable doubt. With
this object in view, the work here reported was undertaken.
Folin,® in first presenting this method for the quantitative
determination of creatinin and creatin in urine, recommended the
use of 15 cc. of a 1.2 per cent solution of picric acid and 5 cc. of a
ro per cent solution of sodium hydroxide. Later,® in modifying
his method he increased the amount of alkali from 5 cc. to 9 cc.
on account of the increased quantity of normal hydrochloric acid
he found desirable to use in changing the creatin to creatinin.
Bauer and Barschall,’ in their work upon beef extracts, used 15 cc.
of the picric acid solution and 5 cc. of the alkalisolution. Bene-
dict and Meyers,* in using the method with urine and with beef

1 Arbeiten aus dem katserlichen Gesundheitsamte, XXiv, p. 552.

2 Loc. cit.

3 Zeitschr. f. physiol. Chem., xli, p. 223, 1904; Festschrift f. Olaf Ham-
marsten, iii, 1906.

* Amer. Journ. of Physiol., xviii, p. 4, 1907.

5 Zeitschr. f. physiol. Chem., xli, p. 223, 1904.

® Festschrift f. Olaf Hammarsten, tii, 1906.

7 Arbeiten aus dem kaiserlichen Gesundheitsamte, xxiv, p. 552.

PLL OGACTH.

————

A.D. Emmett and H. S. Grindley | 493

extracts, took 15 cc. of the saturated picric acid solution and rocc.
of the alkali solution. In the previous publication’ upon this
subject from this laboratory, 15 cc. of the saturated (1.2 per cent)
solution of picric acid and 5 cc. of the ro per cent sodium hydrox-
ide solution were taken. In the work of this laboratory the
details of the method as perfected and thoroughly tested by
Folin for the determination of creatinin and creatin in urine were
followed as closely as possible. Further, in the previous com-
munication, it was clearly stated that the color of the unknown
creatinin picrate solution was compared with 8.0 mm. of the %
solution of potassium bichromate (24.54 grams per liter) and
not with the color obtained by allowing the picric acid and alkali
to act upon a standard creatinin solution as Hehner states as
the method employed by us. In all our work we have accepted
the formulas and equivalents, which were so carefully worked out
by Folin.

EXPERIMENTAL.

In testing the method, the influence of varying the quantity
of both the picric acid solution and the alkali solution was con-
‘sidered. The differences in the intensity of the color, if there be
any within narrow limits, due to the length of time the solutions
were allowed to stand were not taken up sufficiently in connection
with the present study to warrant consideration in this paper.
In all the determinations herein reported, the time allowed for
the development of the color was as nearly as possible five min-
utes. The purity and strength of the picric acid solution were
thoroughly tested. The picric acid was found to be quite pure
and the strength of the picric acid solution used in this work was
1.2 per cent. A ro per cent solution of sodium hydroxide was
used as the alkali. The volumes of the 1.2 per cent solution of
the picric acid used were 15, 30 and 45 cc., and 5, 10 and 15 cc.
of the ro per cent solution of the sodium hydroxide were taken.
All the readings of the colorimeter were made by at least two
persons each working independently of the other. In making
the observations each person recorded three or four of those
agreeing most closely. One of the analysts in making the read-

1This Journal, ii, p. 309, 1906.

494 Chemistry of Flesh

ings was not informed as to the nature of the solutions under
examination. The average of the readings of the two analysts
was taken as the true value for the data reported in the tables
here given. In cases where the colorimeter was used continu-
ously for a considerable length of time, the standard bichromate
solution was renewed once or twice to avoid any possible error due
to evaporation.

Detailed description of the method used. About 10 grams of the
commercial meat extract were dissolved in water and the result-
ing solution was diluted to exactly 500 cc. and thoroughly
mixed. In some cases the solutions thus prepared were filtered
through dry filters while in other cases the original solutions
were not filtered before being used for the determination of
the creatinin and the creatin. For the determination of the
creatinin, aliquot portions of the sample solution which would
give a reading of 7 to g mm. in the Duboscq colorimeter were
transferred to 500 cc. measuring flasks. To these portions, the
measured amounts of the 1.2 per cent picric acid solution and
the ro per cent solution of sodium hydroxide were each added.
The mixture was then shaken and allowed to stand exactly five
minutes when it was immediately and quickly diluted to the 500
ec. mark with distilled water and thoroughly mixed. Readings
upon this solution were made at once in the colorimeter, compar-
ing the depth of color of the solution to be tested with that of
a standard half-normal bichromate (24.54 grams per liter) solu-
tion set at 8 mm.

For the determination of the creatin measured portions of the
sample solution were transferred to small Jena beakers. In case
the volumes of the solution taken amounted to more than Io cc.,
they were evaporated upon the steam bath to this volume and
then treated with 1o cc. of normal hydrochloric acid. The crea-
tin of the solutions was changed to creatinin by the autoclave
method of Benedict and Meyers! which has proved to be con-
venient and accurate. To do this, the beakers containing the
solutions were placed in an autoclave and heated at a temperature
of 117° C.-120° C. for 30 minutes. After these solutions were
taken from the autoclave they were diluted to a definite volume

' Amer. Journ, of Physiol., xviii, p. 398, 1907.

A. D. Emmett and H. S. Grindley 495

and measured amounts of the diluted solutions which would
give a reading of 7 to 9 mm. in the colorimeter were transferred
to 500 cc. measuring flasks. The details of procedure were then
continued exactly as directed above for the estimation of the
creatinin. The data of the experiments are given below in full.

Experiment 1. For this experiment 9.2899 grams of a well-known
brand of commercial beef extract were dissolved in water and the result-
ing solution was diluted to 500 cc. and thoroughly mixed. The solution
was then filtered through dry filters and portions of 25 cc. each were taken
for the determination of creatinin by the method described in detail above.
The results of the determination, in which varying amounts of the picric
acid and the sodium hydroxide solutions were used for the same volume,
of the sample solutions are given in the following table:

TABLE 1. CREATININ RESULTS OBTAINED IN EXPERIMENT 1.

A ne | a

METHOD. 2 | ® ° 3

° 8 | 5 45 h

Z S| wah =

: ao) Sle So aye | a

is SAMPLE. sz e Sia

5 atai| ee achat 2 Se FS aie

a |2 |e = 4
| ce. cc mm. mgr. mgr. p. ct.
2278a| Beef extract...........| 15 | 9.0 | 9.000 | 180.00 | 1.94
2278a; “ ps tors oe 15) O° }28.9°) 9-101 | 182-02) 1296
2278a; “ CC emp eet eo 6 30 5 | 8.5 | 9.529 | 190.58 | 2.05
2278a| “ ae Perea 30 3 (8820-1 9. 108 162702) 0-96
2278a| “ Span Pash ce sr ter 30 | 10 | 8.4 | 9.643 | 192.86 | 2.08
2278a| “ oA ee eee 30) JOS. 2) | 9: S23) | 19%-56 | 2. fo
2278a| “ ee as eae nee <b | 45 5 | 9.0 | 9.000 | 180.00 | 1.94
2278a| =“ ios kas Se | 45 | 10 | 8.9 | 9.101 | 182.02 | 1.96
BATOPERE 7S, 225% Bs) accel 2.00

496 Chemistry of Flesh

Experiment 2. For this experiment the same filtered solution of beef
extract was taken for the determination of creatin as was used above in
Experiment 1 for the determination of creatinin. For the estimation of
the creatin in the solution, five portions of 25 cc. each of the original
filtered solution were measured out and the creatin which they contained
was converted into creatinin in the autoclave as described above. The
five solutions representing 125 cc. of the original solution of the extract
after removal from the autoclave were united and diluted to 500 cc.
Fifty cubic centimeter portions of this diluted solution were taken for
the determination of creatinin by the method described in full above.
The analytical results obtained, where varying amounts of the picric
acid solution and the sodium hydroxide solution were used for the same
volume of the sample solution, are given in the following table:

TABLE 2. CREATIN RESULTS OBTAINED IN EXPERIMENT 2.
it ORIGINAL CREATININ| 4 4b © fe eyey
METHOD.) © | PLUS CREATININ DUE/ "3 |S5 |Zon
S S TO CREATIN. Eg |59 pies
Z | Ge . 5 oa et Ax
‘3 SAMPLE. ae) Sie Fo ‘6 PB SB OS log
se2|28| $6) BE | 2 | 5 | sa ipeesce
Bia }ma|o = a |e ie le
| ce. | cc. | mm. | mgr. mgr. D: Cb.|\ DLCL. | Ds Cha \\gaatte
2278a |Beef extract.......] 15] 5/16.1/5.032'201.28)2.17/2.00\0.17/0.20
2278a)| “ el et Rakes 15 | 10 10.5|7. 714/308. 56/3 .32)2 .00|1.32)1.53
2278a| “ Sah aid sicher 15! 15| 9 .6)8. 438/337 . 52/3 .63|2 .00|1.63/1.89
2278a| “ SY Bape 30/}10) 9.0/9.000/360.00/3.88|2.00)1.88/2.18
2278a| “ Sod Vix SB: 30}10) 9.1/8.901/356.04/3. 83/2 .00/1.83|2.12
2278a| “ eee 30/15} 9.0/9.000|/360.00)/3.88/2.00|1.88/2.18
2278a| “ Pee LHe 30| 15] 9.2/8.804/352.16/3.79/2.00|1.79/2.08
2278a| “ peasy Sheik 45|10| 8.9/9.101 364. 04/3.92/2.00 1.92)2.23
2278a| “ gee rare 4510) 9.0/9.000/360.00 3.88/2.00)1.88)/2.18

A. D. Emmett and H. S. Grindley

497

Experiment 3. For this experiment two solutions were prepared from
the same beef extract as that used in Experiments Nos. 1 and 2.
first, namely, No. 2278b, 11.3131 grams of the extract were dissolved in
water and the resulting solution was diluted to 500 cc. and thoroughly
In the second solution, No. 2278c, 10.3424 grams of the extract
were dissolved in water and made up to a definite volume of 500 cc.
These solutions were not filtered but they were used directly for the work
Portions of 25 cc. each
of these unfiltered solutions were taken for the determination of creatinin.

The results of the tests are given in the following table:

mixed.

reported in this and the following experiment.

In the

TABLE 3. CREATININ RESULTS OBTAINED IN EXPERIMENT 3.

METHOD. 8 3 6 3

9 S 5 2 5

‘sl ae) ro) 3 ee a)
~ SAMPLE. is 0 oe | a 2
A 20 =5 38 as 35 $.8
b3/28| sé 3.8 $5 58

ey P| = B oy
ce. ce. mm. | mgr. } mor. p. ct.
Desi Beet extract: .......<. 15 5 Negi |) TSAO I) BASE |) 20
PTSD: || “ Ling ne ee AS ORE VER 30 Hl) He) || WSR) |) Slo 74s || Boas
DPS | Bow © garetts 30 | 10 | 6.9 | 11.739 | 234.78 | 2.08
Deasie|) << a be, ents Bs 45 Hl (59) |) abs rekey |) Bayless | 2s!
eS) || A Ne he Gor ts 45 | 10 | 6.9 | 11.739 | 234.78 | 2.08
PU UCNO OE an oh a ae o's 2.07
2euse |Beet extract.........:.. 15 Sle se LOnSSor 20% 70) 220i!
2278c a Le Oe AS ne 2 15 Hl 7 I) WO | AO. 7) | 22.01
Z2Se..| “ ee ac NS tee 15 | aD 9/3) WOR |) 2Oe AD || 2 Ol
Pee) << Se es eae oe mel OMe On LOs253| 20 Se OGn unos
DSc | Gh Not See na oe Oil | WO i 72 i) UO. ces |) 207s ow) | 2 Ol
DOSee | Oe te SN Fer en, Pi | MON Hats I WO. || LW. 70) | 270i
D2Se.| A eae Dik We Ths} FPS ORS ss) POR Ay Pee
2278c a SAN reese APS hcg cists Dit |) aes | WS | Ose UAW C70) |) 2 Ou
PSE |.“ UE Piper eran SOON ese Ons Son 2 Ove ON ee Ol
2278e.|. “ LOR Mh TAS fs Sa aye Sx) 1 IO | 7531) MOSS |) BOS 7Al) |) OL
AUOFALC. vax = 25 2.01

498 Chemistry of Flesh

Experiment 4. For this experiment the same unfiltered solution, No.
2278c, was taken for the determination of creatin as was used above for
the determination of creatinin in Experiment 3. For the estimation of
the creatin four portions of 25 cc. each of the original unfiltered solution
were measured out and the creatin which they contained was changed
into creatinin as usual. The four solutions representing 100 cc. of the
original solution of the extract after removal from the autoclave were
united and diluted to 200 cc. Twenty-five cubic centimeter portions of
this diluted solution were taken for the determination of creatinin by the
usual method. The results of the tests are given in the following table:

TABLE 4. CREATIN RESULTS OBTAINED IN EXPERIMENT 4.

& |orreman creatinin| § | 6.8 (226
METHOD:|| "9 | PLUS CHEAT EN IA = | Ss (gen
& | DUE TO CREATIN. 64/22 |go7
6 | ea l 38 |oo |g.8X%
Z SAMPLE. z eed a £ < 23 eS Peele
el eg|ae|ss| SF] 8 | & | 85) 8s [883
era g5| 25) 8) ge) SF | 5 | s4| 54 lees

| & |< |e | Oo a Pi | Ay [Ax (Ay
Co: Cos mm. mgr. mgr. 'p 2 Cle | > Cl. ial. MupeiGrs
ae ‘Beet extract... 15) | 20 (8.7 Ne 310372. 403. 60/2 .01)1.59)1.84
278b ..-| 15 | 10 | 8.8 |9. 205/368. 20/3 .56/2.01/1.55)1.80
SOrBb * . ...| 21 | 10 |8.7 |9.310)372.40/3 .60|2.01)1.59)1.84
2278b| “ S ...| 21 | 10 |8.8 |9.205/368.20/3. 56/2 .01/1.55)1.80
2278b| “ x ...| 30 | 10 | 8.2 |9.878/395. 12/3. 82/2 .01/1.81/2.10
2278b| “ e ...| 80 | 10 | 8.2 |9.878/395.12/3.82/2.01 1.812210

| 1 !

OO

A. D. Emmett and H. S. Grindley 499

Experiment 5. For this experiment a cold water extract of 150.5287
grams of lean beef round was prepared by the methods described in a
former paper.! The resulting extract was made up to a volume of 7500
cc. Five portions of rooocc. each of this cold water extract were meas-
ured out and evaporated to a volume of about 50 cc. The coagu-
lated proteid resulting was removed by filtration and thoroughly washed
with hot water. The filtrates were now evaporated to a volume of 10 cc.
and the creatin? which they contained was converted into creatinin by
use of the autoclave. The five solutions representing five liters of the
original cold water extract after removal from the autoclave were united
and diluted to 500cc. Fifteen cubic centimeter portions of this last solu-
tion were taken for the determination of creatinin by the usual method.
The results of the tests are given in the following table:

TABLE 5. CREATIN RESULTS OBTAINED IN EXPERIMENT D9.

x a3 ee ‘ox ®
METHOD. EE $5 2 3 £
z 2 f= a We ecg
s SAMPLE. z orem cous | ‘O-5 ee <
4 Aeneas lec Meat
a 28| aS (Ss | Mag] ae | e874] os
oP a°|\s" | Sas 2° oe a
Cé2)\. ce | mm. | mgr. mgr mgr. | Dp. ct.
2279 | Lean beefround........ 15} 5 | 7.3/11.096)554.80 643 .57\0.428
2249 | “ eS eee See 15} 5! 7.7!10.519|525.95.610.1010.405
2279 “4 rece 15| 5 | 7.4)10.950/547.50|635.10,0.422
2279 x OAD OU le rocotacea ts 15| 5 | 7.510.800 540.00626.40 0.416
2279 - meme tee ae | 15) 5t| 7.7\10.520526.00\610.16.0.405
2279 | a ed ol gS tee eas 15] 10 | 7.5/10.800 540.00/626.40 0.416
2279) “ SP ie ee 15| 10*| 7.810.380 519.00 602 .04'0.400
2219') . “ SF rg ee 30) 10 7.011.570 578.50 671.06 0.446
2219) ~* Sra peas See 30, 10 7.0/11.570|578.50/671.06)0. 446
2279 fe SS neg PORE a aren 30) 10+ 7.2)11.250 562 .50652.500.434
2279 £ Sak t Meter 30] 15 7.011.570 578.50 671.060. 446
2279 z eda Wee. See rocR 30) 10 7.0)11.570 578 .50/671.06,0. 446
2279 y 7 eae ee 45| 15 | 7.0|11.570\578.50/671.06|/0.446

* Also added 10 ce. > HCl.
¢ Also added 10 ec. NaCl, 6 per cent.

1H. S. Grindley and A. D. Emmett: Journ. of the Amer. Chem. Soc.,
XXKVli, p. 658, 1905.

2 Former experiments made in this laboratory demonstrated the fact
that fresh meats contain only the slightest trace of creatinin if any at all.

500 Chemistry of Flesh

Experiment 6. For this experiment five different brands of beef ex-
tracts were taken. They were given Laboratory Nos. 2306, 2307, 2308,
2309 and 2310. The following weights of these extracts were respectively
taken for the work, 11.7990, 10.1120, 9.5615, 11.2308, and 9.6105 grams.
Each of these weighed portions was dissolved in water and the solutions
thus obtained were diluted to 500 cc. and thoroughly mixed. The result-
ing solutions were not filtered but they were used directly for the work
reported in this and the following experiment. The portions of these
diluted solutions which were taken for the several determinations are
given below in the table giving the results of the tests of this experiment.

TABLE 6. CREATININ RESULTS OBTAINED IN EXPERIMENT 6.

FE METHOD b 3 las 3

2 lacs ie ~ x

c ei ——| 2 2 ee

a SAMPLE. Ej Wo hea so. © Se 6
= a¢|35|3| #8] 2, | °s | 2,
2 BS EL \$3/38| oe | $8 | a

oe Fe ie ear = = a
ce. | cc. | cc. | mm.| mgr. mgr. p. ct.
2306, |\Beekextract... 5 5-..- Se Delo 5 | 725) 10-80) 635 225ieoeos
2306 | : Sen, > Cees 8.5) 15 | 10 | 7.4] 10.95) 644.08] 5.47
2306 s CE a ee 8.5) 30 | 10 | 7.2) 11.25) 661.73) 5.61
2306 | “ o. aeaereee: 8.5] 30 | 10 | 7.3] 11.09] 652.31) 5.53
| Awerage: ee 5 | 5.45
230% |Beefextract....- 42 23.0] 15 5 | 8.0] 10.13] 220.26] 2.18
FeV hh ae Or a Pee a | 23.0| 15 | 10 | 7.8) 10.38) 225.66) 2.23
2307 - HO) | ae aeons | 23.0} 30 | 10 | 7.6) 10.66) 231.75) 2.29
Aperage 3 25 cmee | | 2.28
2308 |Beef extract.......... | 50.0) 15 | 10 | 7.8} 10.39) 103.85) 1.09
2308 z nig ee 50.0) 21 | 10 | 7.8) 10.39) 103.85} 1.09
2308 | “ Ae wee | 50.0) 21 | 10 | 7.8) 10.39) 103.85) 1.09
2308 Zz 3 ~++-+.-.| 50.0) 30 | 10 | 7.8) 10.39) 105. S5ieiae
2308 is Ln | 50.0) 30 | 10 | 7.8) 10.39} 103.85] 1.09
AUET ALG eee eee | 1.09
2309 |Beefextract.......... | 13.0/ 15 | 5 | 7.3/11.096) 426.75] 3.80
2309 | “ ee Use ees oe | 13.0) 15 | 5 | 7.3)11.096| 426.75] 3.80
2309 | “ RS ag ae | 13.0) 15 | 10 | 7.2)11.225) 432.68) 3.85
2309 : a epee | 13.0) 15 | 10 | 7.0|11.571)| 445.02)"3296
2309 si ie bie eee | 13.0] 30 | 10 | 7.0)11.571) 445.02] 3.96
2309 | “ - _..--+..| 12.5) 30 | 10 | 7.3/11.096) 443-84) ooo
2309 - oa uae Ree 13.0, 30 | 10 | 7.1/11.408 438.75) 3.91
2309 : en, ey epee | 13.0) 30 | 10 | 7.1)11.408) 438.75) 3.91
2309 “ a ell a taten eye ) 13.0) 380 | 15 | 7.1)11.408) 488.75) 3.91
UCT ALE: oa | 3.89
2310 |Beefextract.......... | 17.0} 15 | 5 | 7.0] 11.57| 340.27] 3.54
210: | 4 i oeee+ee-| 17.0) 15 | 10 | 7-0) 11.57) S40 eee
2310 - SiR oid ie | 17.0; 30 | 10 | 6.9) 11.74) 344.09) 3.58
PEUET DO Me ae | | 3.56

A. D. Emmett and H. S. Grindley 501

Experiment 7. For this experiment the same unfiltered solutions of
the five beef extracts used above in Experiment 6 were taken for the
determination of creatin. For the estimation of creatin Nos. 2306, 2307,
2309 and 2310, five portions of 25 cc. each were measured from each of
the four solutions of the extracts and the creatin which they contained
was converted into creatinin as usual. The five solutions representing
125 cc. of the original solutions of each of the extracts after removal from
the autoclave were united and diluted to 500 cc. For the estimation of
creatin in No. 2308 four portions of 25 cc. each were measured from the
original solution of this extract and the creatin contained in the same
was converted into creatinin as usual. The four solutions representing
too cc. of the original solution after removal from the autoclave were
united and diluted to 200 cc. The portions of these diluted solutions
which were taken for the determination of the creatinin by the usual
method are indicated by the following table which gives the detailed
analytical results for the three samples of beef extract.

TABLE 7. CREATIN RESULTS OBTAINED IN EXPERIMENT 7.

¢ | |, |orreman creatinin|3 [44 [89%
g| METHOD. | 5 PLUS CREATININ DUE | ‘5, g2 soe
} £3) we) TO CREATIN. oc ap ie
Zz 8 Nees = Our ae mex
& SAMPLE. e| 3 38 _s + + ce ad baer
ale a" a fs BF | 5 5° haba a
a | a en ee ee ee p. ct. | p. ct
2306 |Beef extract../50 15 | 7.0)11.571/462.84/3.92/5.45
2306 s my SO) Lon co | 7.1\11.408)456.32/3.87|5.45
2306 it 7 30| 30 | 10 | 7.0)11.5711771. 7916. 54|5.45/1.09|1.26
aes
2307 z o 150) 15 5 |11.0| 7.364/294. 56/2 .91|2.23,0.68)0.79
2307 2 (0) 15 9} 105 | 6.9]11 . 594/331 . 23/3 . 28/2 .23)1.05)1.22
2307 x oa 170; 30 | 10 | 6.4/12.656 361.58 3.58 2.23 1.385.1.57
| | |
2308 | “ “ (60 30 | 10 | 9.0) 9.000 150.001.57|1.09 0.48 0.57
2308 is ad 67, 30 | 10 | 8.3] 9.759|145.70)1.52/1.09 0.43,0.50
2309 | “ “ (37/15 | 5 | 9.1] 8.901/481.10)4.28/3.890.390.45
2309 oS I37| 15 5 | 9.0) 9.000 486.45 4.33/3.890.440.51
2309 e 37| 15 5 | 8.9} 9.101/491.91/4.38/3.89 0.49 0.57
2309 ie < 37| 15 | 10 | 8.0/10.125|547.26/4.87/3.89,0.98)1.14
2309 s < 37| 15 | 10 | 7.9]10.253/554.18!4.93/3.89|1.04)1.21
2309 S be joa) Lo. je 10. | 7.8)10.385/561.31/5.00/3.89)1.11)1.29
2309 a Zs Seo Ome) 7.710.519 568.55 5.06/3.89 1.17)1.36
2309 e 3 37 30 | 15 | 7.9|10.253 554.18)4.93/3.89)1.04)1.21
, | | } | |
2310 & S 50 15 | fay tej AI 10.000 400.004.16/3.54 0.62)\0.72
2310 fe vf SO elon LOM Pons 13.966/558.64|5.81/3.54/2.27|2.63
2310 - - 50| 30 | 10 | 5.6/14.464'578.56\6.02/3. 54/2 .48/2.88
2310 a £ 35| 30 | 10 | 7.9/10.253'585.4516. 0913 .5412.55|2.96

502 Chemistry of Flesh

Experiment 8. For this experiment three samples of urine from three
different persons were taken. These samples were given Laboratory
Nos. 2319, 2320 and 2321. The specific gravities of the urines were
respectively, 1.033, 1.029, 1.018. In each sample the creatinin was
determined by the usual method, using however different quantities of
picric acid and varying quantities of alkali. The detailed results of the
experiment are given in the table that follows:

TABLE 8. CREATININ RESULTS OBTAINED IN EXPERIMENT 8.

S) EB 4 x) ma
\& METHOD. | ‘5 | g “ g
as ee 2. 1s
> SAMPLE. ior Z | fe S ae s
a oe soil ase | se) |) ates a | 8g
3 |e | £3 | 2: 38 | 38 26 |
(oO |e | = ie = = rw
=f ce. | ce. ce. mm. mgr. mgr. Dp. ct.
Pep Mate et ssc rs xs 14.0] 15 | 5 | 6.9-| 11.740 | 205.45 | 0.28
2319 Ces | 4.0] 15 | 5 | 7.0] 11.570 | 202.48 | anes
2319 ee 7 2 | 4.0 | 15 | 10 | 6.9 | 11.740 | 205.45 | 0.28
2319 Roe. te ce i | 4.0 | 30 | 10 | 6.4 | 12.660 | 221.55 | 0.31
2319 Be ieee Ae | 4.0 | 30 | 10 | 6.5 | 12.460 | 218.05 | 0.30
Average.....| 0.29
BeOM Oats 5 te ws /4.3|15] 5] 7.8 | 10.380 | 142747 )Mieee
2320 ie sean eh 4.3 | 30 | 10 | 7.4 | 10.950 | 150.23 | 0.25
Average..... 0.24
» Sayheel |
Gorn (Orme lei tt. |6.5| 15 | 5 | 7.4 | 10.950 | 17e.7onumene
BU ore 2 Sd | 6.5 | 15 | 10 | 7.3 | 11.100 | 179.20 | 0.17
7 Mu Wee a 6.5 | 15 | 10 | 7.5 | 10.800 | 174.42 | 0.17
eet Age tec: 6.5 | 30 | 10 | 7.3 | 11.100 | 179-20 30.9%
2321 CADE ae ee 6.5 | 30 | 15 | 7.3 | 11.100 | 179,20 |, Oeas
Average..... 0.97

a mm a ee ——ee

A. D. Emmett and H. S. Grindley 503

Experiment g. The same samples of urine as were used above in
Experiment 8 were taken for the determination of creatin. For the esti-
mation of the creatin in Laboratory Nos. 2319 and 2321 four portions of
Io cc. each were measured from each sample, and the creatin which they
contained was changed into creatinin as usual. The four solutions of
each sample representing 40 cc. of the original urine after removal from
the autoclave were united and diluted to 250 cc. The portions of the
diluted solution which were taken for the determination of the creatinin
are indicated in the table. For the estimation of creatin in Laboratory
No. 2320 three portions of 10 cc. each were taken for changing the creatin
into creatinin. The three solutions representing 30 cc. of the original
urine after removal from the autoclave were united and diluted to 250
ec. The portions taken for the tests are indicated in the table. The
complete analytical data for the creatin determination in the three sam-
ples of urine are given in detail in the table below:

TABLE 9. CREATIN RESULTS OBTAINED IN EXPERIMENT 9.

=

ORIGINAL CREATININ |

] 3

8 | METHOD, | & PLUS CREATININ DUE) -E

Fo} = } lS TO CREATIN. j=
Z (= ipo ious
3 SAMPLE. | & zs kore 2 : : $8
} a = |} we =a i =) ~~.
4 \ee jot [as lse| 2 | = | 8 | 8a
ce. cc. | cc. |mm.| mgr. | mgr. |p. ct. | p. ct.
Bema cine...20........ 20.0} 15 | 5 | 8.4] 9.643/210.94/0.29/0.28
(ili 2 re 20.0).15 | 10 | 8.2| 9.878/216.08/0.30)0.28
210 (Se 20.0) 15 | 10 | 8.2| 9.878/216.08/0.30/0.28
2319 | “ Pe... ..12020| 30 | 10:| 8. 1110000/218 7 75\02 3010030
3819 | “ _...........-|25.0| 30 | 10 | 6.4]12.656/221.48/0.31/0.31
Lit 20.0] 30 | 15 | 8.3 9.759 213.48)0.30)....
el || 9. ai ae en gies) tos! (58.8 aoe cele
PaI0.|. “ v,++-+--+--./81.5] 15 | 10 | 8.7) 9.310/145.33/0.24/0.24
es i 31.5| 15 | 10 | 8.8] 9.205|143.6910.24/0.24
C2) a crs 25.0} 30 | 10 |10.0| 8.100)159.33/0.260.24
20) | er 31.5| 30 | 10 | 8.5] 9.529/148.75|0.250.24
no na wn 31.5] 30 | 10 | 8.5] 9.529'148.75,0.25 0.24
2201 | ri 34.5| 15 | 10 | 8.8] 9.205/175.08|0:16/0.17
Loo) a 34.5| 15 | 10 | 9.0] 9.000/171.18/0.16|0.17
Lili 34.5] 30 | 10 | 8.8] 9.205175.08/0.16\0.17
225 |) eee 34.5) 30 | 10 | 9.0} 9.000 171.18|0.160.17
eM ones 25.0: 30 | 10 |10.9| 7.431.195.06/0.18,0.17

504 Chemistry of Flesh

Experiment 10. Asa further check on this method, we were fortunate
enough to obtain through the kindness of Dr. Folin some pure creatin
(Kahlbaum). Crystallized creatin contains one molecule of water of
crystallization and in our sample the moisture content was ascertained
and duly allowed for in the subsequent data in Table 10. Several por-
tions of the sample, given Laboratory No. 2289, were weighed off and
each dissolved in an appropriate quantity of distilled water. The follow-
ing weights of the dry substance were taken:

(a) .1290 grams (b) .2217 grams (c) .1165 grams
(d) .2705 “% (e) .2217 “ (ff) 460080

All of the samples were dissolved in 100 cc. of distilled water, except in
7, which was dissolved in 400 cc. From these solutions, five 10 cc. portions
of samples, b, d, e and fj, and nine ro cc. portions of samples a and ¢ were
transferred to small beakers. To each beaker 10 cc. of normal hydro-
cbloric acid was added, after which the dehydration was carried on at 117—
119° C. in the autoclave for one-half hour. Each of the resulting creatinin
solutions of the respective samples was transferred to a 500 cc. measuring
flask and diluted to the mark. After thoroughly mixing, portions of 50
ec. were taken for the usual colorimetric determinations. Special men-
tion should perhaps be made at this point for subsequent discussion that
solutions a and ¢ contained the equivalent of go cc. of normal hydro-
chloric acid and that the others, b, d, e and fj, contained 50 cc.
The following table gives the results of the several tests:

TABLE 10. SUMMARY OF RESULTS OBTAINED IN EXPERIMENT 10.

| E . | ao Soe s

5 I mle be SX =

| > : 5 |MerHop s ae ae a 3

¢ |8 | £ fe \_—|® [2s |g ee

. ie) P Ph lps Aa igo SG | Os 2

eh 5 2 fee? | (82 | eos oes eee

LJ fs) Ala |sa|e |F B = a
E | = aaa | 7
| gms. ce. cc. | cc. | cc. | mm. mgr. mgr. mgr. p. ct.
2289A 0.1290 100 50) 15] 5 (22.6) 3.580 39.82} 46.19/35.81
2289A |0.1290 100 50} 15) 5 |22.6) 3.580) 39.82) 46.19/85.81
| Average (2) [sae 22.6| 3.580) 89.82) 46.19|85.81
2239 0.1290 100 50| 15| 10] 9.1 8.901| 98.89|114.72/88.93
2289A |0.1290 100 50] 15} 10) 9.1) 8.901] 98.89)114. 72/88. 93
| | Average (2) 9.1 8.901| 98.89|114.72\88.93
2289A |0.1290 100 50] 30! 10] 8.4) 9.643/107.14/124.29/96.34
2289A 0. 1290 100 50} 30} 10] 8.5) 9.529/105.88]122 .82/95.21
Average (2) | 8.5 9.586\106.§1|123.56|95.78
2289A (0.1290 100 50} 30) 15} 8.5, 9.529)105.88/122.82/95.21
2289B |0.2217 100 50} 15} 10} 9.5 8.526]170.52/197.80)89.21
2289B |0.2217 100 50] 15] 10) 9.4) 8.617|172.34)199:91|90.17
Average (2) |_| | 9.5 8.572\171.43\198.86\89.69

a i, i

i

A.» Emmett and T.S. Grindley 505

2 g METHOD 5 32 ss = &

Ee E el eee ean ie linwe

4 or gi Pea Wesigr Woot line), (cere a ir cee 1 Wie

3 =8 rc 38] 8 | Cacia eseres ent ese al Stage. pues

3 & = Oalerelies. | Seen gai) ae | oes

E Seen see erie es ilsciee| et ale
2289B |0.2217| 100 50} 30 10 9.0 9.000 180.00 208.80 94.18
2289B (0.2217; 100 50| 30 15 9.0, 9.000 180.00 208.80 94.18
2289C |0.1165| 100 63] 15 10) 9.0) 9.000) 89.291103.58 88.91
2289C |0.1165| 100 63] 15) 10| 8.8] 9.110) 90.37|104.8389.99
Average (2) | 8.9| 9.055) 89.83)104.21\89.45
ee ee ee a feos poe SS
2289C 0.1165] 100 63] 30| 10| 8.5] 9.529] 94.54|109.66194.13
2289C 0.1165} 100 63] 30) 10) 8.4| 9.643) 95.67/110.97/95.25
Average (2) | | 8.5| 9.586) 95.11|110.32\94.69
2289C 0.1165] 100 63| 30, 15| 8.5) 9.529) 94.54/109.66|94. 13
2289D 0.2705] 100 50| 15 5) 9.0) 9.000/180.00 208. 80|76.89
2289D 0.2705} 100 50| 15) 10| 7.7|10.519|210.38|244.04/90.21
2289D 0.2705] 100 50| 15| 10| 7.6)10.658|/213. 16|247.27|/91.04
Average (2) 7.7|\10.589\211.77\245.65|90.63
2289D |0.2705| 100 50| 30| 10] 7.2|11.250/225.10|261.11|96.49
2289D 0.2705} 100 50| 30) 10) 7.3/11.096|221 .92/257.43/95.17
2289D |0.2705| 100 50] 30 10] 7.3/11.096|221.92/257.43/95.17
2289D |0.2705| 100 50} 30, 10| 7.410.946 218. 92)253.95|93 .88
Average (4) 7.8\11.097\221.97\257. 48|95.18
2289D 0.2705} 100 50| 30) 15| 7.4/10.946|218.92|253.95/93.88
2289E |0.2217; 100 50; 15| 511.1) 7.300)145.94)169.29/76.36
2289E |0.2217| 100 50} 15 10, 9.0} 9.000/180.00 208. 80)94.18
2289E |0.2217| 100 50| 15| 10) 8.9| 9.101|182.02|211.14/95.28
Average (2) | 9.0| 9.051|181.01\209.97|94.71
2289E |0.2217| 100 50| 30 10, 8.9| 9.101/182.02211.14|95.24
2289E |0.2217| 100 50| 30| 10} 8.9] 9.101|182.02|211.14|95.24
2289E |0.2217| 100 50| 30) 10| 8.9] 9.101|182.02/211.14|95.24
Average (3) | 8.9| 9.101 182.02\211. 14195. 24
2289F 0.4699 200 50| 15| 5110.2) 7.940/317.64/368.46/78. 47
2289F |0.4699 200 50| 15) 510.4] 7.788|311.52|361.36/76.88
Average (2) 10.3| 7.864)314.58\3864.91\77.68
2289F |0.4699 200 50] 15 10, 8.4) 9.643/385.72 447. 44/95 .22
2289F 0.4699 200 50| 15, 10) 8.3) 9.759/390.36|452.82/96.38
| Average (2) | 8.4) 9.701\388 .04| 450. 13\95.80
2289F 0.4699 200 ~50| 30, 10) 8.0 10. 125/405 .00/469. 80/99. 98
2289F (0.4699 200 50| 30 10) 8.0 10.125/405.00 469.80/99.98
; | Average (2) | 8.0 10. 125|405 .00| 469. 80|99.98
2289F 0.4699 200 50 30, 15 3.0 10. 125|405.00 469.80 99.98

| |

506 Chemistry of Flesh

Experiment 11. For this experiment several tests were made upon
solutions of creatinin (Merck’s). This product was found to be only about
85 per cent pure, there being also present approximately 1o per cent of
creatin and 2.07 percent of moisture. Asa result, the material in question
could not be used to serve the double purpose intended, namely, the check-
ing of the method and the standardizing of the bichromate solution.

However, to ascertain the variations, if any, in the percentage of
creatinin in the sample when different proportions of the picric acid and
alkali were used three portions were weighed out. Sample a weighed
0.1286 grams (dry); b, 0.1127 grams (dry), and c, 0.2055 grams (dry).
The first two samples were dissolved in 500 cc. of distilled water and after
thoroughly mixing, 50 cc. were taken for the usual procedure. Sample
c was dissolved in 250 cc. of water and 100 cc. of this solution were taken
and diluted to 500 cc. after which 75 cc. were used for the regular deter-
mination.

The detailed data resulting are given in the following table:

TABLE 11. SUMMARY OF RESULTS OBTAINED IN EXPERIMENT II.

LY) iS 1 _
3 g 3 a METHOD. & 3 $
eee ae Raper:
° Ge > ° < One ‘Ss
a * = a4 3 } 80 2 : 28
= a | ae 2 = Oo aa a8
E 2 ‘> | wo | B| 3 38 BA a§
4 e 5 en & = o-- cae oF
sa eh) a < ee = a
gms. ce. ce. ce. ce. mm. mgr. mgr.
2353a | .1286 | 500 | 50 | 15 5 Sz 310 93.10
2353a .1286 | 500 | 50 | 15 5 8.8 9.205 92.05
2353a .1286 | 500) 50 | 15} 10 7.6 | 10.658 | d06aue
2353a | .1286! 500| 50 | 15] 10 7.6 | 10.658 | 106.58
2353a 1286 | 500 | 50 | 30| 10 7.4 | 10.946 | 109.46
2353a 1286 | 500 | 50 | 30! 10 7.5 | 10.800 | 108200
2353a 1286 | 500 | 50 | 30| 15 7.4 | 10.946! 109.46
2353a 1286 | 500 | 50 | 30| 15 7.5 | 10.800 | 108.00
2353b 1127 | 500 | 50 | 15 Br tOe2 7.941 79.41
2353b 1127 | 500-1 50° |) Se 0 8.8 9.205 92.05
2368b | 1127 |.500 | 50 |) ciSele 10 es 9.205 92.05
2353b 1127 |.500 | 50. | 30°. 10 8.6 9.415 94.15
2353b 1127 | 500 | 50 | 30{ 10 8.7 9.310 93.10
2353b 1127 | 500 | 50 | 30| 15 8.8 9.205 92.05
2353c 2055 | 250| 75 | 15] 10 9.0 9.000 | 150.03
2353c | .2055 | 250| 75 | 15 | 10 9.0 9.000 150.03
2353c 2055 | 250 | 75, | 30 | 10 8.0 | 10.125). 168-78
2353c 2055 | 250 | 75 | 30] 10 8.0 | 10.125 | 168.78
2353¢ 2055 | 250 | 75 | 30] 15 8.0 | 10.125 | 6s s 7s

A. D. Emmett and H. S. Grindley 507

DISCUSSION OF RESULTS.

It is quite apparent from the data reported in the preceding
pages that the Folin method of estimating creatinin and creatin,
when properly modified, is as applicable to meat and meat extract
as itis to urine. Further, it is evident that if the details of the
method are properly followed, reliable and concordant results
can be obtained.

ORIGINAL CREATININ. (a) Influence of alkali. By compar-
ing the data in Tables 1, 3, 6 and 8 which relate to the per-
centage of creatinin in meat extracts and urine, it is seen that in
_ so far as the initial quantity of creatininis concerned, the influence
of increasing the amount of alkali is practically nil. If 15, 21 or
30 cc. of picric acid of 1.2 per cent are used with either 5, 10 or
15 cc. of sodium hydroxide of ro per cent strength the readings
of the colorimeter for the same volume of the solutions under
examination, are in general slightly higher for the 5 cc. quantities
than for the ro cc. portions. Stating these facts in terms of
percentages of creatinin, the 10 and 15 cc. quantities of sodium
hydroxide, which are approximately 8 and 13 cc., respectively,
in excess of the necessary amount to cause neutralization and
solution of the precipitate, gave but very slightly higher percent-
ages than did the 5 cc. quantities. These actual differences in
the case of meat extracts were 0.02 to 0.17 per cent, being on the
average 0.05 per cent. In the case of the creatinin (Merck’s),
which as previously stated contained 10 per cent of creatin, the
data in Table 11, show the variations in amount of alkali to
produce an appreciable difference. The weights of original
creatinin resulting from the use of ro or 15 cc. of alkali are the
same, but for those resulting from the use of 5 cc., they are dis-
tinctly lower, being in the former cases on an average of 106.58
mgms., and in the latter 92.58 mgms., a difference of 14.0
mgms. No positiveexplanation can be given for this large varia-
tion. It is easily seen that the differences in some cases in the
readings due to the amounts of alkali might be from 2 to 4 mm.
as was the case with the beef extracts and urine, and this would
mean with such a strong solution, a difference of 2 to 4 per cent.

From the above discussion, the data show that, contrary to
Hehner’s statement, an excess of alkali, using 5, 10 and 15 cc. of

508 Chemistry of Flesh

a ro per cent solution, does not diminish the depth of the color
produced with creatinin picrate but rather increases it. It would
seem, therefore, that it would be best to use 10 cc. of a 10 per
cent solution of alkali in all instances, it having been shown that
this amount of alkali has no detrimental effect whatever.

(b) Influence of picric acid. In studying the data in Tables
1, 3 and 6 on beef extract, 8 on urine, and 11 on pure creatinin,
it will be seen that the variations in the percentages of
original creatinin are extremely slight when the quantity of
picric acid is increased from 15 to 21, 30 or 45 cc. in the cases
where 10 cc. of the alkali are used. In fact, it can be stated, as
far as the creatinin sample and also the meat extract samples
are concerned, that the 15 cc. test of picric acid (1.2 per cent)
gave as high a percentage of creatinin as did the 21 cc. test
(which is the equivalent of Hehner’s 25 cc. portion, 1.01 per
cent), or even the 30 cc. test. This fact is most clearly brought
out in Tables 3 and 6 where the analysis of six different
samples of meat extract are reported. Sample 2306 has a varia-
tion, taking the average of the readings of the 30 cc. tests, of
0.10 per cent on a total of 5.47 per cent of original creatinin.
Sample 2307 shows an increase of 0.06 per cent for the 30 cc. of
the picric acid against that of 15 cc. on a total of 2.23 per cent.
Similarly, the data for sample 2310 gave a gain of 0.04 per
cent for the 30 cc. of picric acid when compared with the 15 cc.
on a percentage of creatinin of 3.55. These slight differences are
no greater than the errors which might be due to the matching
of the color, to the carrying out of the technique of the method
or to the sampling of the products.

In the case of the urine, Table 8, the variations in the readings
for sample 2319 are greater than in the cases just considered
but those for sample 2321 show the same variations as the meat
extracts. When calculated to the percentage of creatinin these
differences in the urine are insignificant. Further, when the
analysis of the three samples of the pure creatinin are con-
sidered, it is seen that the readings of the two tests, 15 and 30 cc.
of picric acid are almost identical. These small differences are
no greater than would be expected. However, if these values
are calculated to percentages, the variations are about 1.0 per
cent. The cause of this difference is not to be found as Hehner

A. D. Emmett and H. S. Grindley 509

states in the greater amount of creatinin in the solution which is
being treated with picric acid, because in this, and the previous
work of this laboratory, whether the sample under examination
was meat extract, urine, or creatinin itself, such an aliquot por-
tion of the original solution was taken that when it was treated
in the usual manner with picric acid and alkali, and diluted,
the resulting color gave a reading of 7 to 9 mm. on the scale of
the colorimeter when compared with the standard bichromate
set at 8mm. In other words, all the solutions tested at the time
of the reading contained approximately the same quantity of
original or converted creatinin. The real cause of the influence
of the slight differences in the readings upon the variations cal-
culated in percentage, naturally lies in the differences in the
weights of the substances taken for the samples, since the pro-
portion of the dilutions were approximately the same. The
weights of the samples were always about 150 grams for meats,
1o grams for meat extracts and 0.12 grams for the creatinin, and,
consequently, it is plainly seen that any variation in the readings
would be very much less apparent in the percentage composition
in the case of the meats than in the meat extract or the impure
creatinin.

From this discussion upon the influence of varying the quantity
of picric acid in cases where ro cc. of a 10 per cent solution of
alkali were used, it seems safe to state that it makes no difference,
when determining the original creatinin, whether 15, 21 or 30 cc.
of the 1.2 per cent solution of picric acid are used. The final
percentages are practically identical.

CREATININ DUE TOCREATIN. (a) Influence of alkali. Inthe
former paper, it was stated that the creatin in the samples was
changed to creatinin by using 25 cc. of ~4 hydrochloric acid
and that the resulting solutions after proper dilution were treated
with 15 cc. of picric acid (1.2 per cent) and 5 cc. of alkali (10 per
cent). The data herein reported were ascertained by using 10 cc.
of normal hydrochloric acid for the dehydration or four times the
amount previously taken. Following this change in the amount
of acid, the quantity of alkali was also increased. These modifi-
cations were adopted on account of the recoimmendations of
Folin in his second paper,' that in the determining of creatin in

1 Festschrift f. Olaf Hammarsten, iii, 1906.

510 Chemistry of Flesh

abnormal urine 1o cc. of normal hydrochloric acid, 15 cc. of
picric acid (1.2 per cent) and 9 cc. of sodium hydroxide (10 per
cent) should be taken.

In studying the influence of different proportions of alkali
when varying amounts of dehydrated creatin and original cre-
atinin were present, two series of tests were carried out. Inthe
first, 5 and 1occ. of the alkali were used against 15 cc. of the picric
acid, and in the second, ro and 15 cc. of the alkali were employed
with 30 and sometimes 45 cc. of the picric acid. It was ascer-
tained in the case of meat extracts that approximately 1 to 2 cc.
of the 10 per cent alkali were sufficient to dissolve the precipitate
and to produce a red coloration.

The data in Tables 2, 4 and 7, relating to meat extracts,
show the effect of using 5 and 10 cc. of the alkali with 15 cc. of
the picric acid. It is very apparent that the 5 cc. portion is
entirely too small. This fact is perhaps shown most markedly
in sample 2306, where the total creatinin is 3.90 per cent while
that for the original creatinin is 5.45 percent. In other words,
the 5 cc. of alkali produced a color which, when compared with
the standard, represented a value of 29.4 per cent less than the
amount obtained before dehydration. A comparison of the 5
and ro cc. tests is best brought out in the case of samples 2278a,
Table 2; and 2309 and 2310, Table 7. The differences in the
readings with the two quantities of alkali vary from 1.1 mm. in
2309 to 5.6 mm. in 2278a. These data when calculated to the
percentage of creatinin as creatin, show a range of 0.7 to 1.33,
respectively, in favor of the 10 cc. portion of alkali.

Very little can be said regarding the influence of the quantity
of alkali in the case of the meats and urines since the data are too
few for consideration. The slight differences in those instances
that are reported are inappreciable when calculated to their
final percentages. However, in connection with the meat extract
the facts show very plainly that the ro cc. of alkali when used
with 15 cc. of picric acid gives higher results than the 5 cc. portion.

In the second trial where 10 and 15 cc. of alkali were used with
30 or 45 cc. of picric acid, the resulting data for the beef extracts
and meat are fairly constant. The few variations amount in
the maximum to only 0.2 mm. This would indicate that the
slight differences were not due to the excess of alkali, but rather

A. D. Emmett and H. S. Grindley 511

to errors in the technique. However, it should perhaps be
stated that these differences are in the main in favor of the lower
readings for the ro cc. test, but when the data in Table ro, relat-
ing to pure creatin, are also taken into account the evidence seems
to be such that it can be stated that there are no differences
resulting in using either ro or 15 cc., of alkali with 30 cc. of the
picric acid, beyond the experimental errors.

From the above consideration, the data show that the quantity
of alkali does influence the depth of color, that a small quantity
does not yield as high a percentage as a large one; that a large
excess does not give low results, and that the accepted ro cc.
portion, yields better results than the 5 cc. and similar results to
the 15 cc. portion.

(b) Influence of picric acid. Inthe preceding pages, the effect
of using varying amounts of alkali and picric acid has been
considered where preformed creatinin was to be determined, and
in the above paragraphs, the influence of different quantities of
alkali has been discussed in the case where both the preformed
and dehydrated creatinin were present. It is the purpose of this
section to digest the data herein reported which refer directly to
the effect of using larger quantities of picric acid than has been
customary in determining creatin as creatinin.

Hehner noticed in his work that an increase of the picric acid
from 15 cc. to 25 cc. (1.01 per cent) caused a marked difference in
the amount of creatinin found to be present in meat extracts.
Instead of obtaining with 25 cc. of picric acid (1.01 per cent) 6 to 7
per cent of creatinin, which he got with the equivalent of 15 cc.
of 1.2 per cent acid, he reported 10 to 12 per cent. He found,
further, that a larger amount of picric acid had no increased
effect.

In our work, the amounts of picric acid used were 15, 21, 30
and 45 cc., and since it was found that 10 cc. of the sodium
hydroxide worked satisfactorily, this quantity of alkali was
taken throughout for the comparison. The data in Tables 2, 4,
and 7 on meat extracts, Table 5 on meat, and Table 10 on pure
creatin, show that in the case of 30 cc. of the picric acid the gen-
eral tendency is to produce a lower reading and hence a higher
percentage of creatinin as creatin. These differences in the read-
ings for the meat extracts are practically nothing in sample 2309

512 Chemistry of Flesh

and 1.4, 0.6, 0.5 and o.2 mm. in samples 2278a, 2278b, 2307 and
2310, respectively; for the meat they are o.5 mm., and for the
pure creatin 0.1 to 0.6 mm. Concerning the data for the urine,
Table 9, the differences due to the increased amount of picric
acid are very slight, being in samples 2319 and 2321 almost
nothing and in sample 2320, 0.3 mm.

In general then, it may be stated for the meat extracts, meat
and pure creatin that the additional quantity of picric acid may
have no effect in some cases and in others it may cause a decrease
in the readings of 0.2 to1.4mm. From the data in the following
Table 12, which gives a summary of these facts, the differences

TABLE 12. SUMMARY OF RESULTS ON INFLUENCE OF PICRIC ACID.

(Creatin.)
READINGS OF COLOR- PERCENTAGE OF CREATIN.
IMETER.

Z SAMPLE. = qe, ¢ = sf ¢

4 a3./.88 | 93 { $8 | Se.cie

F fee }ee| 8 | es | eo | &

7 Ay fa 7 v A
r; mm. mm. mm. p. ct. De iCt. pacts
22 Sa Beer extract. oii iet || Olrom|PsO eel | 1 (3333 2). Lom ORG2
2278c ogee a eee ere SSeS ORO 1.82 2.10 | 0.28
2307 4 Sh Sea Sa 6a9r W642 7085 22, Say AN (W)o3)5:
2309 s De Ti iar cle Hee) | Pes | Weal 1.29 1.369) On07
2310 OU ee Neh sears SE Suo0) || Oez 2.88 2.96 | 0.08
2209) ~ || Meat. a2 asc hoteeeeee Haar (ree! |) Wed 0.42 0.45 | 0.03
2519 UGE. . 2 1.0. 2s aot ea CORE OL [Ov 0.30 0.30 | 0.00
OA Beil nial. fe cs.h 0 NE SS eleseon| Oss 0.24 OnZ 5a OE Om
2321 TRONS Son oe nie Sea 8.9 | 8.9 | 0.0 0.16 0.16 | 0.00
2289a \Pure creatin..<.. 02.4.4) 921 | 825) |NORG |FS8eos al Con iSemoras
2289b ie ae et oy ee 9.5 | 9.0 | 0.5 | 89.69 | 94.18 | 4.49
2289c ‘- ee ate Me Sy Bx 8.9-) 7825 | 0.4 |) 89.455) | O4569Rltaeze:
2289d se HO Pains eee 7.0 7,3 | O.4 190263 | Soe uSaltaaoo
2289e be Rani aay ee 920} $39 | O11 | 945715 |) 952245 BOs
2289f 4 Oa Fa, Eee ee 8.4 | 8.0 | 0.4 | 95.80 | 99.98 | 4.18

in the percentage of creatin in the meat extracts are seen to
vary from 0.07 to 0.62, the total percentage of creatin being 1.4
to 3.0. The meat shows a corresponding difference, a gain of
0.03 per cent on a total of 0.446 percent. In the case of the urine
the variations are practically nothing, and in that of the pure

A. D. Emmett and H. S. Grindley 513

creatin, they are on the average 4.4 per cent greater for the 30
cc. portion of acid.

A second point should be considered, as to the influence of
using still more of the picric acid. The data for the sample of
beef extract 2278a, Table 2, and that for the meat 2279, Table 5,
show that an additional 15 cc. of acid, or 45 cc. in all, does not
cause any further change than that brought about by the 30 cc.
test.

The above two facts agree in general with Hehner’s conclusions
that 25 cc. (1.01 per cent) of picric acid should be used and that
an additional amount produces no different effect. However,
mention should be made that in our case where 21 cc. of the 1.2
per cent picric acid, which is the equivalent of 25 cc. of a 1.01
per cent solution was used, no decided change was produced.
This fact is shown in the data in Table 4. Further, while the
additional amount of picric acid seems in general to cause a deeper
color, the authors wish to emphasize the fact that this difference
is by no means as great as Hehner states it to be in his paper.
After calculating the data, obtained by using 30 cc. of picric acid
and ro cc. of alkali, it will be seen that the combined percentages
of creatinin and creatin in the six different samples of beef extract
reported in Tables 2, 4, and 7 amount to 4.15, 4.11, 6.71, 3.80,
1.63, 5.24 and 6.46. These samples correspond, respectively, to
those for Laboratory Nos. 2278a, 2278c, 2306, 2307, 2308, 2309
and 2310. When these percentages are compared with those
reported in the previous paper which ranged from 1.38 to 6.56
per cent, it can be stated that the data do not differ materially
in the two cases, and in as much as the samples used for this
work are both representative of those reported formerly and also
of Hehner’s the evidence seems to indicate that Hehner’s results
which varied from ro to 12 per cent for the combined creatinin
and creatin were entirely too high.

From the above consideration of the data, it is evident that
in the majority of cases the method gives higher results for the
converted creatin in meat extracts, meat and pure creatin when
the quantity of picric acid (1.2 per cent) is increased from 15 to
30 cc., although several instances are reported where the smaller
amount of acid served equally well. No definite explanation can
be given at the present time for this apparent anomaly. The

514 Chemistry of Flesh

tendency seems to indicate that it is more difficult for the 15 cc. of
picric acid to overcome the resistance of the resulting converted
creatinin than that of the preformed creatinin which shows that
the former must exist in a different condition than the latter.
The amount of hydrochloric acid used in dehydrating the creatin
does not seem to influence the results as is shown in Experiment
ro which relates to the pure creatin. Here, some solutions had
go cc. of hydrochloric acid and others 50 cc., yet the general effect
of the picric acid was the same in each case. Jaffe,’ states that
by using the zinc chloride method his maximum yield was 94.83
per cent and adds that this seems to show that the creatinin
resulting from converted creatin is broken down to some extent
by the strong acid. In general the results reported in this paper
confirm Jaffe’s conclusion. Benedict and Meyers? in using Folin
method with creatin obtained from 96 to 98.9 per cent, while the
data here reported show a variation in six determinations of
from 94.2 per cent to 95.8 per cent, and in one test the result was
99.98 per cent. These facts, however, do not necessarily reflect
upon the Folin colorimetric method as applied to urine, meat and
meat extract. Since normal urine contains no creatin, meats only
0.44 per cent, and meat extracts from 1 to 6 per cent, it will be
seen that a yield of 95 per cent is sufficiently accurate for all
practical purposes. Further, in as much as the modified method
seems to give uniform results throughout, the data should be
comparable in all cases and be of extreme value in giving impor-
tant information as to the quantity of these extractives in meat
products.

OUTLINE OF METHOD AS NOW USED IN THIS LABORATORY.

The following brief outline of the method as now used is given:
For the preformed creatinin, transfer aliquot portions of the
sample solution to 500 cc. measuring flasks, add 15 cc. of a 1.2 per
cent picric acid solution, mix, add 10 cc. of a 10 per cent sodium
hydroxide solution, shake thoroughly, and allow the mixture
to stand 5 minutes and then dilute to the mark at once and after
mixing, compare the depth of the color of the solutions with that

1 Zeitschr. f. physiol. Chem., xiviii, p. 436, 1906.
2 Amer. Journ. of Physiol., xviii, p. 4 1907.

A. D. Emmett and H. S. Grindley 515

of a half-normal bichromate solution set at 8mm. According to
Folin, the correct reading in millimeters of the colorimeter
divided into 81 gives the number of milligrams of creatinin
contained in the portion of the solution taken for the treatment
with picric acid and sodium hydroxide. In other words, 10
milligrams of pure creatinin after the addition of the picric
acid and the alkali and dilution to 500 cc., gives a reading of
8.1 mm. when compared with 8 mm. of } potassium bichromate
solution (24.54 grams per liter). For the combined creatinin,
transfer aliquot portions of the sample solution to beakers,
if the quantity is more than Io cc., or to roo cc. measuring
flasks, if the quantity is 10 cc. or less. Inthe former instance,
evaporate the solution on the water-bath to1occ. In either case
make the volume of the liquid up ro cc., if necessary, and add ro
ec. of normal hydrochloric acid. Rotate the vessels to mix the
liquids. Transfer the acid solutions to an autoclave and heat
them at a temperature of 117 to 119° C. for 30 minutes. After
removal, cool and dilute to the mark. If beakers were used,
transfer the contents to roo cc. measuring flasks and dilute. To
aliquot portions of the converted creatinin solution add, in 500 cc.
flasks, 30 cc. of 1.2 per cent picric acid, shake and then add 10 cc.
of the 10 per cent sodium hydroxide. Mix thoroughly and after
standing exactly 5 minutes dilute, and read the depth of color
of the solution. In order to convert milligrams of creatinin
into creatin multiply by the factor 1.16.

It was found that by using a black cloth, to shut out the sur-
rounding light from the eyepiece of the instrument, the colors
appeared more distinctly, and that the comparison could be
made more accurately and rapidly and with less strain on the
eye:

CONCLUSIONS.

From this study upon meat extracts and meat, the following
conclusions can be made in regard to the applicability of the
Folin method for determining creatinin and creatin:

(a) That an increase in the quantity of picric acid, according
to Hehner’s suggestion causes no difference in the so-called
original creatinin determinations; but it generally does produce
an appreciable difference when the converted creatin is also

516 Chemistry of Flesh

present, and, further, that the quantity of picric acid (1.2 per
cent) recommended for use in meat extract, meat and urine
should be left at 15 cc. for the original creatinin determinations
and be increased to 30 cc. for the dehydrated creatinin.

(b) That in the determination of the original creatinin, the
use of a small or large amount of 10 per cent alkali makes almost
no difference, the 5 cc. quantity giving slightly lower results than
the 10 and 15 cc. quantities; that, for the converted creatin, the
previously accepted quantity of alkali, ro cc., gives better results
than 5 cc. and equally as good results as the large excess, 15 cc.;
and further, that these facts are contrary to those found by Heh-
ner who states that a quite small amount of alkali gives better
results than a large quantity which he maintains diminishes
the depth of color.

(c) That the data reported are representative of the percent-
ages of creatinin and creatin in meats and meat extracts, being
practically the same for the combined extractives as those pre-
viously published, 0.45 per cent for the former and 1.4 to 6.5
per cent for the latter, whereas Hehner found the total percent-
age of creatinin and creatin in meat extracts to be 10 to 12 per
cent.

(d) That the Folin method when properly modified is as
applicable to meat extracts and meats as it is to urine, and that
it gives reliable and concordant results in the hands of different
analysts of this laboratory.

The authors wish to acknowledge their appreciation of the
assistance rendered by Messrs, H. H. Mitchell and D. L. Weather-
head.

ERRATA, VOLUME III.

Page 85, line 26, for exclusive read extensive.

Page go, line 7, for chloride read hydroxide.

Page 180, Nos. (3), (4), (5) and (6) of table of acetone determinations
should read, ‘‘20 cc. acetone solution + o cc. H,O, etc.”

INDEX TO VOLUME III.

Acetone and diacetic acids, the
separate determination of, in dia-
betic urines, 177

Acid-forming and base-forming ele-
ments in food, balance of, 307

Agglutinins and antitoxin, fraction-
ation of, 233

Alcohol, influence of, on the meta-
bolism of hepatic glycogen, 403

Aliphatic substances, oxidation of,
in the animal organism, 57

Alkaloids, picrolonates of, 327

Amanita-toxin, chemical properties
of, 279

AMBERG, S. and W. P. Morritv:
On the excretion of creatinin in
the new-born infant, 311

Amines, alkyl, occurrence and for-
mation of, 83

6-Aminopyrimidin, salts of, 285

Ammonia in milk and its develop-
ment during proteolysis under
the influence of strong antisep-
GCS 17

Amylolytic activity of saliva of the
dog, 135

Anodonta and Unio, manganese, a
normal element in the tissues of,

I51

Anti-inulase, 395

Antitoxic serum, fractional precipi-
tation of, 253

Antitoxin, fractionation of agglu-
tinins and, 233

Arginase, action of upon creatin
and other guanidin derivatives,

435

Aspergillus niger, relation of ex-
tractive to protein phosphorus
in, 49

AusTRIAN, C. R., see JoNES and
AUSTRIAN, I, 227

Autolysis, relation of the thyroid
60> 35

Autolysis, study of by determina-
tions of changes in freezing-
point and electrical conductivity,

35

Bacillus coli communis, chemistry
of, 443

Balance of acid-forming and base-
forming elements in foods, 307

BANCROFT, FRANK W.: On the
relative efficiency of the various
methods of adminstering saline
purgatives, IgI

BanzuHaF, Epwin J. and RogBerr
Banks Gipson: The fractional
precipitation of antitoxic serum,
253

Barium sulphate, reduction of, in
ordinary gravimetric determina-
tions, 81

Base-forming and acid-forming ele-
ments in foods, balance of, 307

Bean, lima, proteolytic changes
occurring in, during germination,
265

Behavior of uric acid toward animal
extracts and alkalies, 145

BENEpiIcT, FRaNcis G. and THomas
B. OsBorNE: The heat of com-
bustion of vegetable protein, 119

BENEDICT, STANLEY R.: The detec-
tion and estimation of reducing
sugars, 101

Benson, R. L., see WELLS and
BENSON, 35

Benzoic acid derivatives, produc-
tion of phenolic acids by the
oxidation of, 419

Berc, W. N,. see SHERMAN, BERG,
CoHEN and WHITMAN, 171

Boric acid, execretion of, from the
human body, 11

BraDLey, Harotp C.: Manganese,
a normal element in the tissues
of the fresh water clams, Unio
and Anodonta, 151

Chemical properties of Amanita-
toxin, 279
Crapp,, (S: JEL,
CLAPP, 219
CouEN, L. J., see SHERMAN, BERG,

CoHEN and WHITMAN, I71

see OSBORNE and

517

518

CoLuins, KATHARINE R., see GIB-
son and COLLINS, 233

Colorimetric determination of io-
dine, 391°

Composition, chemical, of the hair
of different races, 459

Conduction of nerve impulse, nature
of, 359

Conductivity, electrical, study of
autolysis by determinations of
the changes in, 35

Creatin, action of arginase upon,
435

Creatin as a brain stimulant, 21

Creatinin, excretion of, in the new-
born infant, 311

Cytosin, a color test for, 183

Cytosin, salts of, 285

Daxin, H. D.: Experiments bear-
ing upon the mode of oxidation
of simple aliphatic substances
in the aninal organism, 57

Dakin, H. D., The action of argi-
mase upon creatin and other
guanidin derivatives, 435

Dakin, H. D., and Mary Dows
HerTER: On the production of
phenolic acids by the oxidation
with hydrogen peroxide of the
ammonium salts of benzoic acid
and its derivatives, with some
remarks on the mode of forma-
tion of phenolic substances in the
organism, 519

Diacetic acid and acetone, the
separate determination of, in dia-
betic urines, 177

Electrolytes, relation of, to lecithin
and kephalin, 53

Embryos, nuclein ferments of, 227

Emmett, A. D., and H. S. Grinp-
LEY: Chemistry of flesh, 491

Excretion of boric acid from the hu-
man body, 11

Excretion of creatinin in the new-
born infant, 311

Ferments, nuclein, of embryos, 227

Flesh, chemistry of 491

Foun, Orro: On the occurrence
and formation of alkyl ureas and
alkyl amines, 83

Fo.in, Orro: On the reduction of
barium sulphate in ordinary
gravimetric determinations, 81

=H AWE, ~P). Bs

Index

Fotin, Ortro: On the separate
determination of acetone and
diacetic acid in diabetic urines,

I “

ieee) balance of acid-forming and
base-forming elements in, 307

Forp, WiILLiam W., see SCHLES-
INGER and ForD, 279

Fractional precipitation of anti-
toxic serum, 253

Fractionation of agglutinins and
antitoxin, 233

Freezing-point, study of autolysis
by determinations of changes in,

35

Germination, proteolytic changes
occurring in the lima bean dur-
ing, 265

Gipson, RoBert Banks and Katu-
ARINE R. CoLtiins: On the frac-
tionation of agglutinins and anti-
toxin, 233

GiBson, Rosert Banks, see BANz-
HAF and GIBSON, 253

Gies, WivuraM J.: Further obser-
vations on protagon, 339

Glycogen, hepatic, influence of alco-
hol on the metabolism of, 403

Glycogen, the formation of, in
muscle, 25

GRINDLEY, H. S., see EMMETT and
GRINDLEY, 491

Guanidin derivatives,
arginase upon, 435

action of

Hair of different races, the com-
parative chemical composition of,
459

Harris, Isaac F.,
and Harris, 213

Haskins, Howarp D.: The effect
of transfusion of blood on the
nitrogenous metabolism of dogs,

321

HatcHer, R. A. andC. G. L. Wor:
The formation of glycogen in
muscle, 25

see OSBORNE

see RUTHERFORD
and Hawk, 459

Heat of combustion of vegetable
proteins, 119

Hemolytic serum, quantitative
methods with, 387

Herter, Mary Dows, see DAKIN
and HeERTER, 419

Hydrogen peroxide, oxidation of
benzoic acid derivatives by, 419

Index

Hydrolysis of legumin from the pea,
219

Infant, new-born,
creatinin in, 311

Iodine, colorimetric determination
of, 391

Isocytosin, salts of, 285

Jounson, Treat B.: VI. Researches
on pyrimidins: Synthesis of thy-
min-4-carboxylic acid, 299

Jounson, TREAT B., see WHEELER
and JOHNSON, 183

Jones, WALTER, and C. R. Aus-
TRIAN: On the nuclein ferments
of embryos, 227

Jones, WaLTER, and C. R. Aus-
TRIAN: On thymus nucleic acid, 1

Kephalin, and lecithin, relation of
electrolytes to, 53

Kocu, W.: The quantitative esti-
mation of extractive and protein
phosphorus, 159

Kocu, W.: The relation of elec-
trolytes to lecithin and kephalin,

53

Kocu, W. and Howarp S. REEp:
The relation of extractive to pro-
tein phosphorus in Aspergillus
niger, 49

Leacu, Mary F.: On the chemistry
of Bacillus cola communis, 443
Lecithin and kephalin, relation of
electrolytes to, 53
Legumin from the pea, hydrolysis
of,219 ;

Manganese, a normal element in
the tissues of the fresh water
clams, Unio and Anodonta, 151

MANWARING, WILFRED H.: Quanti-
tative methods with hemolytic
serum, 387

MaxweELlt, S. S.: Creatin as a brain
stimulant, 21

MaAxwELI, S. S.: Is the conduction
of nerve impulse a chemical or a
physical process, 359

MENDEL, LAFAYETTE B., and FRANK
P. UNDERHILL: Is the saliva of
the dog amylolytically active, 135

Metabolism, nitrogenous, effect of

transfusion of blood on, 321

Metabolism of hepatic glycogen,
influence of alcohol on, 403

Methods, quantitative, with hemo-
lytic serum, 387

excretion of

519
Milk, development of ammonia in,
during proteolysis under the

influence of strong antiseptics,
MitcHELL, Puitip H.: A note on
the behavior of uric acid toward
animal extracts and alkalies, 145
MorriLi, W. P., see AMBERG and
MorRILL, 311
Muscle, formation of glycogen in,

25

Nerve impulse, nature of the con-
duction of, 359

Nucleic acid, thymus, 1

Nuclein ferments of embryos, 227

OsBORNE, THomas B. and S. H.
CLapp: Hydrolysis of legumin
from the pea, 219

OsBORNE, THOMAS B. and Isaac F.
Harris: The proteins of the pea
(Pisum sativum), 213

OsBORNE, THomas B., see BENE-
Dict and OSBORNE, 119

OSTERBERG, E. and C. G. L. Wotr:
Day and night urines, 165

Oxidation of benzoic acid deriva-
tives, the production of phenolic
acids by, 419

Oxidation of simple aliphatic sub-
stances in the animal organism,

27)
6-Oxypyrimidin, salts of, 285

Pea, hydrolysis of legumin from,
219
Pea (Pisum sativum), proteins of,

BER

Pepsin, synthesis of protein through
the action of, 95

Phenolic acids, production of, by
oxidation of benzoic acid deriva-
tives, 419

Phenolic substances, remarks on
the mode of formation of in the
organism, 419

Phosphorus, extractive and protein,
the quantitative estimation of,

59

Phosphorus, relation of extractive

to protein in Aspergillus niger,

49 , E
Picrolonates of certain alkaloids,
ae aoe ;
Proceedings of the American So-
ciety of Biological Chemists, vil
Protagon, further observations on,
339

520

Protein, synthesis of, through the
action of pepsin, 95

Protein, synthesis of, through the
action of trypsin, 87

Proteins of the pea (Pisum satévum),
213

Proteins, vegetable, heat of com-
bustion of, 119

Proteolytic changes occurring in
the lima bean during germina-
tion, 265

Purgatives, saline, on the relative
efficiency of the various methods
of administering, 191

Pyrimidins, researches on: On a
color test for uracil and cytosin,
183

Pyrimidins, researches on: On some
salts of cytosin, isocytosin, 6-
aminopyrimidin and 6-oxypyri-
midin, 285

Pyrimidins, researches on: Synthe-
sis of thymin-4-carboxylic acid,
299

Reduction of barium sulphate in
ordinary gravimetric determina-
tions, 81

REED, Howarp S., see Kocu and
REED, 49

Rosertson, T. Brattsrorp: Note
on the synthesis of protein
through the action of pepsin, 95

RUTHERFORD, THomas A. and
B. Hawk: A study of the com-
parative chemical composition
of the hair of different races, 459

SalkI, Tapasu: Anti-inulase, 395

SALANT, WILLIAM: The influence of
alcohol on the metabolism of
hepatic glycogen, 403

Saline purgatives, on the relative
efficiency of the various methods
of administering, 191

Saliva of the dog, amylolytic activi-
ty of, 135

SCHLESINGER, HERMANN and WIL1-.

LIAM W. Forp: On the chemical
properties of Amanita-toxin, 279

SEIDELL, ATHERTON: A new stand-
ard for use in the colorimetric
determination of iodine, 391

Serum, antitoxic, fractional precipi-
tation of, 253

Serum, hemolytic, quantitative
methods with, 387

Index

SHERMAN, H. C., W. N. Bere. L.
J. CoHen and W. G. WHITMAN:
Ammonia in milk and its develop-
ment during proteolysis under
the influence of strong antisep-
fics; 570

SHERMAN, H. C. and J. Epwin S1n-
CLAIR: The balance of acid-form-
ing and base-forming elements in
foods, 307

SINCLAIR, J. EDwin, see SHERMAN
and SINCLAIR, 307

Sugars, detection and estimation
of, Ior

SuzukI, SHINKIcHI: A study of the
proteolytic changes occurring in
the lima bean during germina-
tion, 265

Synthesis of protein through the
action of pepsin, 95

Synthesis of protein through the
action of trypsin, 87

Synthesis of thymin-4-carboxylic
acid, 299

TayLor, ALONzO ENGLEBERT: On
the synthesis of protein through
the action of trypsin, 87

Thymin-4-carboxylic acid, synthe-
sis of, 299

Thymus nucleic acid, 1

Thyroid, relation of, to autolysis,

5

Tecetasen of blood, effect of, on
the nitrogenous metabolism of
dogs, 321

Trypsin, synthesis of protein
through the action of, 87

UNDERHILL, FRANK P., see MENDEL
and UNDERHILL, 135

Unio and Anodonta, manganese, a
normal element in the tissues of,
Eyr

Uracil, a color test for, 183

Ureas, alkyl, occurrence and for-
mation of, 83

Uric acid, behavior of, toward ani-
mal extracts and alkalies, 145

Urines, day and night,,165

WarREN, W. H. and R. S. WeErtss:
The picrolonates of certain alka-
loids, 327

Weiss, R. S.,
WEISS, 327

see WARREN and

Index

We tts, H. GIDEON,
BENSON:

anayel IS. JL,
The relation of the

thyroid to autolysis, with a pre-_

liminary report on the study of
autolysis by determinations of
the changes in freezing-point and
electrical conductivity, 35

WHEELER, HENRY L.: V. Re-
searches on pyrimidins: On some
salts of cytosin, isocytosin, 6-
aminopyrimidin and 6-oxypyri-
midin, 285

521

WHEELER, Henry L. and TREAT
B. Jounson: IV. Researches on
pyrimidins: On a color test for
uracil and cytosin, 183

WuitmMan, W. G. see SHERMAN,
Berc, CoHEN and WHITMAN, 171

WILEY, Harvey W.: The excretion
of boric acid from the human

body, 11
Wotr, C. G. L., see HATCHER and
WOLF, 25

Wotr, C. G.:‘L., see- OSTERBERG
and Wo tr, 165

—

Pre 5 ~~, ne? es A o

«

QP The Journal of biological
501 chemistry

J77

Ved

COper

Biological ,
& Medical
Serials

PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET

UNIVERSITY OF TORONTO LIBRARY te