‘ Fa¥y.
ti adit y Wik
4 Hh! wy Risin in
aM

ny Nts? uy

eee

DEE go Sas
eee Ss
ee =

TS

pasate
Oe
ae

<>

s

Se
Sia eae

Pe
=
s

i) ny ath! eI Wik ASA Rae i Vk A ‘4 ae ‘ ed) i " Dy A ies
NY ; A I lye te 4
Me { ye pimeeT isd Vg nh} AL thei p. JX) FAB I naa
Wa : piataalels’ LOM 4 ph iy itn iy Hy Hy a pies
SMM ae Daten A
ny '; x tan ¥ Kort _ is
ih iy fut raisin! os ; reeeot tt iy Yn Se et
uit ay * Wn ti it ; Nite Basal ate bitte a
tt 1 uy

cry
—~
——

-) uphe

SS gee
ot

: es :
ete
SASS

M "

o as

ves Sa fh DUDS eat ara Ne Heh Sf iting nt rite Lire ' Wet (es tl Ke et ‘
Hai a i Mii 1 iF Sani Ay , ne senna Mpa At La a,
Ri aN NAO EH CRT a a i ao ea
ft Cnty Lae AE A Dyeity Hye NTH ia vith peat
Sf aR whe Ka tb ha UA oh Hi ay aut
By Wad Hat Weare CTL Al { } CAR PON At VEEN YON Wy Yi \ \ ;
; ; maagety ; 44 init ' Sh 1
AS .
Aue,

Hi Wi , ei
Ha!

aa he RPL

=

io

oe

eee

ee
<<

cial ~ —~ Sar
tre a x
See
Se ee
SEE
ste
=
=
~“

eee
Seas

Xf x

i
¥
t
’,
\
4
a3

SS

Wn
tari
‘ uf)

HH

Ny; ‘ i

ui i Vania

ye ena Me i
i Wy i

SaaS.

= as
—— == =
Geek

oh:
ne
RIB ;
Le
YANN Fat
wae ha y dita M Hi
ch 8 wine ay
4 roe UUUHy eH
ARO en nine a tary
misainiarnte te Rete
ren Aa
wot

sso

2a
=a
=—s
ee
a

= =
aS

abe

Aman oat ne}

=

i
i

A ve) pais {
ny ts 4} wy Hy HN ! 4
iW) — Ay i i} RR La
* yay

=
ex

te
“j
o fe
Doe f a

ae aN ne
i 4 ai is
oe 5 AY | ea

= ve.
one
SS Se
eee ety é #5 3

iy

red} senees
pay May a

ae

toca
=
ie

ak

~ Eee = o

Pe

ait } bs i
ai ‘ nh ae aay)
tt Le ; :
G

aren nut
ey,

Sete

hae
*
9
{ rol
- ‘ ,
\
j ’ a i
bd
he
; t
Pi a
+ '
‘
. =
© ive
4 her
bY.
= i i
@ ai
yd
2
- o>
1, Ore
t~e
wt
>
t I
n
a:
-
-
fal
LZ i
&
e
'
*
5 ‘

THE JOURNAL

OF

BIOLOGICAL CHEMISTRY

EDITED BY

J. J. ABEL and C, A. HERTER

Baltimore New York

Assistant Editor, A. N. RICHARDS, Chicago

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, IL

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.
FRANZ PFAFF, Boston, Mass.

A. E. TAYLOR, Berkeley, Cal.

Vv. C. VAUGHAN, Ann Arbor, Mich.

ALFRED J. WAKEMAN, New York. =) /

HENRY L. WHEELER, New Haven, Conn. A
sa ad \®

vv
\
:! \9
VOLUME V_, v\
BALTIMORE

1908—1909

CopyRIGHT 1908 AND 1909
BY

THE JOURNAL OF BIOLOGICAL CHEMISTRY.

WILLIAMS & WILKINS COMPANY PRESS
BALTIMORE, MD.

CONTENTS OF VOLUME V.

Water Jones: On the identity of the nucleic acids of the
thymus, spleen and pancreas

P. G. Heinemann: Note on the Jotceaeation of farelieer
toxin prey Ns

T. BRAILSFORD eee a c i A. ee On the Bart
played by the alkali in the ee eles of proteins by
trypsin 5s Ss Ry SS

Treat B. JOHNSON and SAMUEL H. CLAPP: IX. Researches on
pyrimidins: Synthesis of some nitrogen- eae deriva-
tives of cytosin, thymin and uracil

MatTtHEew STEEL and Wi iam J. Giles: Some oie on tie

efficiency of the Folin method for the quantitative
determination of urinary ammonia

Martruew STEEL: A study of the influence of inaeniestet on
fate on metabolism

Joun F. Lyman: A note on the chemtenee of 4he anise sad
liver of reptiles ;

H. Gipzon Wetts: The EA aiciny Be ihe eee in chienonne
necrosis (delayed chloroform poisoning)

T. BRAILSFORD RoBeRTSON: On the influence of feapemue
upon the solubility of casein in alkaline solutions

T. BRAILSFORD Ropertson: Note on the applicability of the
laws of amphoteric electrolytes to serum globulin

Treat B. Jounson and SamueL H. Crapp: X. Researches on
pytimidins: The action of diazobenzene sulphwsie acid
on thymin, uracil and cytosin —

H. D. Daxin: Further studies of the caods on ee idaticn ot
phenyl derivatives of fatty acids in the animal organism

Tuomas B. OsBporNE and FrEDERICK W. Hey: Hydrolysis of
vicilin from the pea (Pisum sativum) See at Ie

Tuomas B. OsporNne and FrepertcK W. Hey: Hydrolysis
of legumelin from the pea (Pisum sativum)

ili

27

31

49

iu

iv Contents of Volume V

O. F. Brack: The detection and quantitative determination
of B-oxybutyric acid in the urine

Puitip A. SHAFFER: A method for the quanrtaeae detenage
ation of f-oxybutyric acid in the urine « oe

FRANK P. UNDERHILL and Tapasu Saiki: The influence of
complete thyroidectomy and of thyroid feeding upon
certain phases of intermediary metabolism :

C. L. Atsperc and E. D. Crark: On a globulin from the eee
yolk of the spiny dog-fish, Squalus Acanthias

Racuev H. Cotwe ty and H. C. SHERMAN: Chemical emitenee
of peptonization in raw and pasteurized milk

Mary F. Leacu: A preliminary study of the sensitizing por-
tion of egg-white

Grorce A. Oxson: Milk proteins a Sai

C. A. Herter and A.I. Kenpari: The use of the fermentation
tube in intestinal bacteriology :

C. A. Herter and A. I. Kenpati: An Shion ae on thie fae
of B. Bulgaricus (in Bacillac) in the digestive tract of
a monkey (Plates I-III)

H. D. Daxin: Further studies of the ae e: oxidation of
phenyl derivatives of the fatty acids in the animal
organism. III. Synthesis of some derivatives of
phenyl propionmiciacid \y5 |. se! eae

ALONZO ENGLEBERT TAyLoR: Chemical studies in cytolysis

ALONzO ENGLEBERT TAYLOR: On the conversion of glycogen
into sugar in the liver ci ty

Atonzo ENGLEBERT Taytor: On the antagonism of alcohol
to carbolic acid vo tee. | Arr

C. L. AtsBEeRG and E..D. One “The blood clot of Limulus
Polyphemus ae

CHARLES Hucu NEILson and M. H. efeeete: “The effect of
diet on the maltose-splitting power of the saliva

Juiia T. Emerson and Witiiam H. WELKER: Some notes on
the chemical composition and toxicity of Ibervillea
Sonore (Plates IV and V) :

TuEo. C. Burnett: The inhibiting effect of potsuer chiomas
in sodium chloride glycosuria 3

Pau J. Hanzurk and P. B. Hawk: The uric aad excretion of
normal men

207 —

211

225

243 |

247

253
261

393
311

315

323

331

Contents of Volume V

Amos W. Peters: Studies on Enzymes; I. The adsorption
of diastate and catalase by colloidal protein and by
normal lead phosphate ee ct Aen ae ae

Atonzo ENGLEBERT Taylor: On the synthesis of protamin
through ferment action anata enair a lets

Atonzo ENGLEBERT TAYLOR: Onthe composition and deriva-
tion of protamin ia tay PO ee ee

Atonzo ENGLEBERT TAYLOR: On the pee: of the peas
of pepsin and chymosin : :

Atonzo ENGLEBERT TAYLOR: On the inversion cane oes
and maltose by ferments

H. D. Dakin: Note onthe oxidation of ieee ae Heeamus
acids by means of hydrogen peroxide ;

H. D. Dakin: The action of glycocoll as a detoxicating secs

Artuur I. Kenpatt: Bacillus Infantilis (n. s.) and its rela-
tion toinfantilism . Sa SAS i openers ees

C. A. Herter and A. I. ence Note on the products of
B. injantilis grown in artificial media (Plates VI and
VII) MR ee 2, TN) See Ee es

J. J. R. Macteop: A comparison of the methods of Reid and
Schenck for the quantitative estimation of the reduc-
ing substances in the blood :

R. H. Nicuott: The relationship between fea ionic pabetiiels
of salts and their power of inhibiting lipolysis
Wituam F. Boos, M.D., Ph.D.: On the reducing compomeat

of yeast nucleic acid EO): Thee eg,

Pau. E. Howe and P.B. Hawk: Comparative fests af Salva’
and Folin’s methods for the determination of ammonia
and urea RAS 1 Oe ng ek Ya i

SraNLEyY R. Benepicr: Areagent for the detection of reduc-
ing sugars Sa is tice. Lg a i

Curistian A. Herter: Note on the occurrence of skatol and
indol in the wood of Celtis Reticulosa (Miquel)

T BrartsForD ROBERTSON: On the synthesis of paranuclein
through the agency of pepsin and the chemical mechan-
ics of the hydrolysis and synthesis of proteins through
the agency of enzymes Soe oe Se ers

INDEX

439

493
525

4e
i)

Haare Yt ae

f
Po Rau a et oe
yest’ ¢ ' ‘sve 4 k wa ag Lie Regt tsa a
, i .
> ‘ ? ‘ ‘3 Marat lee? os
J Th \exhrade: a 4 ; it! Paap ait <i

ON THE IDENTITY OF THE NUCLEIC ACIDS OF THE
THYMUS, SPLEEN AND PANCREAS.

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

(Received for publication, May 1, 1908.)

A study of the nucleic acids is conceded to be one of the most
puzzling tasks that confront the student of physiological chemis-
try and for the reason that the literature on the subject reveals
a mass of contradictions, corrections and inconsistencies which
it would seem almost impossible to reduce to any satisfactory
scientific order. As is well known these substances yield among
their hydrolytic products various derivatives of purin and the
older literature records four substances of this type (guanin,
adenin, xanthin and hypoxanthin) any combination of which
might be expected among the decomposition products of a
nucleic acid. From the fact that certain nucleic acids produced
one of these purin derivatives considerably in excess of the
other three and that no simple stoichiometric relation was found
to exist between the various purin derivatives found in this
connection, Kossel advanced his hypothesis of the four nucleic
acids each giving one purin base. On the other hand Schmiede-
berg contended that salmon nucleic acid contains in the one
molecule groups of both guanin and adenin. Investigations
which followed Kossel’s earlier work have shown that in many
instances xanthin and hypoxanthin are secondary products and
need not be considered in this connection. Thus it was found
that a number of nucleic acids yield guanin and adenin but
neither xanthin nor hypoxanthin when prepared from fresh
glands and hydrolyzed under conditions which do not admit of
the alteration of the purin derivatives first formed. This was
shown by Kossel and Neumann!’ to be true of thymus nucleic
acid, by Osborne and Harris? of triticonucleic acid, by Jones

1 Kossel and Neumann: Zeitschr. f. physiol. Chem., xxii, p. 74.
2 Osborne and Harris: [bid., xxxvi, p. 85.

I

2 Identity of Nucleic Acids

and Whipple! of the nucleic acids of the pancreas and suprarenal
gland, and by Levene? of a large number of nucleic acids of
various origin. The confusions in the older literature are in
great measure due to the secondary action of the ‘hydrolytic
agent but in part also to the presence in the glands of ferments
which liberate the amidopurins from the nucleic acid and convert
them into the oxypurins, the latter probably being present as
such in the nucleic acid preparations which were employed for a
study of the hydrolytic products. At a time when difficulties
of this nature seemed to be removed, Steudel? made a contribu-
tion which unnecessarily threw the question of purin bases again
into confusion. He submitted the copper salt of thymus nucleic
acid to a severe method of hydrolysis and naturally obtained all
four purin bases.

It is not necessary to go beyond Steudel’s papers themselves to
see wherein the difficulty consisted, but as might have been
expected Steudel is now quoted as authority for the statement
that thymus nucleic acid produces all four purin bases. How-
ever, Jones and Austrian® found that no xanthin is produced
when thymus nucleic acid is decomposed by the action of the
ferment nuclease at the body temperature and ina practically
neutral fluid. At the same time Steudel® ignored his own find-
ings and stated that both salmon nucleic acid and thymus nucleic
acid yield exclusively guanin and adenin and that the two nucleic
acids produce these two bases in the same relative quantities.
It would seem, therefore, that at the present there is no well
established work which is at variance with the general statement
that all ordinary nucleic acids whether of plant or animal origin
yield two and only two purin derivatives, viz: guanin and adenin.’

* Jones and Whipple: Amer. Journ. of Physiol., vii, p. 423.

* Levene: Zeitschr. f. physiol. Chem., xxxvii, p. 402; xxxix, P. 479;
xlv, p. 370; 1, p. i; Levene and Mandel; /bid., xlvi, p. 155; xlvii, P. 140;
xlix, p. 262.

3 Jones and Austrian: Ibid., xlviii, p. IIo.

*Steudel: Jbid., xlii, p. 165.

* Jones and Austrian: This Journal, iii, Dp: o

* Steudel: Zeitschr. f. physiol. Chem., xlix, p. 406; liii, p. 14.

The expression “ordinary nucleic acid’ is used in this paper in its
narrow sense and does not include the ‘‘guanylic acids” which form a
distinct class of themselves.

Walter Jones 2

Our ideas in regard to pyrimidin derivatives have undergone
similar changes. Thymin was first obtained from thymus
nucleic acid by Kossel and Neumann! and their selection of this
term leads to the suspicion that they regarded the compound
as characteristic of the gland. But thymin has been frequently
found among the hydrolytic products of nucleic acids’ and had
come to be regarded as a constant decomposition product when
Ascoli? observed that in yeast nucleic acid the thymin group
is replaced by a uracil group. Osborne and Harris* showed the
same to be true of the nucleic acid of the wheat embryo while
Mandel and Levene’ obtained uracil but no thymin from the
nucleic acid of haddock eggs. Following Ascoli’s discovery
Levene obtained uracil (in addition to thymin) from various
animal nucleic acids but finally ascribed the substance to second-
ary action. This position is now taken by Steudel with reference
to uracil in general.

In 1894 Kossel and Neumann’ obtained from thymus nucleic
acid an organic base which they called cytosin and which was
subsequently proven to be a pyrimidin derivative.’ In the
meantime Levene® showed that spleen nucleic acid gives rise
to a base having the empyrical composition of amido-oxy-pyri-
midin. He afterwards obtained the substance from a number
of nucleic acids and was able to prove its identity with cytosin.

Again, Kossel® showed that levulinic acid results from the
hydrolysis of thymus nucleic acid and as the substance is not
formed from pentose Kossel’s discovery indicates a hexose group
in this nucleic acid molecule. Subsequent work by Levene’
failed to establish the presence of this group in other nucleic
acids and it seemed for a time that thymus nucleic acid was

1 Kossel and Neumann: Ber. d. deutsch. chem. Gesellsch., Xxvi, p. 2753.

2 Levene: Loc. cit., and Zeitschr. f. physiol. Chem., xxxix, pp. 4 and 133.

3 Ascoli: Zeitschr. f. physiol. Chem., xxviii, p. 426.

Osborne and Harris: Loc. cit.

5 Mandel and Levene: This Journal, i, p. 425.

8 Kossel and Neumann: Ber. d. deutsch. chem. Gesellsch., xxvii, p. 2215.

7 Kossel and Steudel: Zeitschr. f. physiol. Chem., xxxviii, p. 49.

§ Levene: [bid., xxxvii, p. 402.

*Kossel: Arch. f. Physiol., p. 157, 1893; Noll: Zettschr. f. physiol. Chem.,
RAV, P. 430.

10 Levene: Zeitschr. f. physiol. Chem., xxxvii, p. 402.

4 Identity of Nucleic Acids

thus shown to have a point of distinction from other members
of this group, but by the adoption of new methods Levene’
convinced himself of the presence of the levulinic acid group in
several nucleic acids which he examined.

When we consider the quantitative differences presented by
the decomposition products of nucleic acids we notice the same
two opposite tendences, i. e., a number of researches indicate
that all nucleic acids yield the same relative quantities of their
decomposition products, and on the other hand an equal num-
ber seem to prove that this relation is different for the various
nucleic acids. Moreover, this phase of the subject has become
complicated to a most unusual degree by the presence of guanylic
acid in the tissues. Ivar Bang? discovered this substance in the
pancreas. He stated that it yields glycerin phosphoric acid,
furfurol and guanin but neither adenin nor any pyrimidin
derivative. The matter of glycerin phosphoric acid seems to be
discredited? .but the other points in connection with the sub-
stance have been sufficiently confirmed. When it is remembered
that substances of the guanylic acid type are not confined to the
pancreas‘ it will appear that the presence of this substance in
varying quantities might account for some at least of the dif-
ferences presented by preparations supposed to be exclusively
“ordinary nucleic acids.”’

Thus while the weight of recent evidence has inclined to the
belief that certain supposedly different nucleic acids are the
same substance and while the support.for these supposed dif-
ferences has been much weakened or entirely removed, a con-
siderable amount of positive evidence has accumulated to show
the chemical identity of certain nucleic acids from sources zodlog-
ically distant from one another. Herlant® has prepared speci-
mens of nucleic acid from the thymus and from fish spermatozoa
and has shown that the two preparations have the same chemical
composition.

‘ Levene: Zeitschr. j. physiol. Chem., xliii, p. 109.

* Bang: Ibid., xxvi, p. 133} EXKi, Pp. 411.

* Steudel: Jbid., liii, p. 539.

‘ Jones and Rowntree: This Journal, iv, p. 280.
*Herlant: Arch. f. exp. Pathol. u. Pharmakol., xliv, p. 148.

Walter Jones 5

Again Steudel' has lately proposed a very accurate method of
estimating the purin derivatives obtainable from a nucleic acid.
The substance is treated with concentrated nitric acid when a
violent reaction ensues with evolution of the brown oxides of
nitrogen and on cooling the products, guanin and adenin are
quantitatively deposited as crystalline nitrates. The removal
of the two nitrates from the other decomposition products and
their separation from one another can be accomplished with
great accuracy, and in this manner Steudel has obtained identical
results from the nucleic acids of the thymus gland and salmon
testicle.’

Finally, Schmiedeberg® in a critical review concludes that
a large number of nucleic acids of various origin which have been
examined in his laboratory are identical chemical substances.
It is, therefore, probable that the following summary of the nucleic
acid question is not far from correct.

(x) All “ordinary nucleic acids” of plant or animal origin
yield the same two purin derivatives (guanin and adenin) and in
the same quantitative ratio.

(2) All ‘“‘ordinary nucleic acids’ yield cytosin.

(3) All “ordinary nucleic acids’ yield levulinic acid.

1Steudel: Zeitschr. f. physiol. Chem., x\viii, p. 425.

2 This work of Steudel appears to me a little curious. I have had fre-
quent occasion to use the method both before and after Steudel’s announce-
ment and can agree with him in regard to the quantitative ratio of the
guanin and adenin groups in thymus nucleic acid. But one will obtain no
trace of either guanin or adenin in the crystalline deposit when the method
is executed as described. The product is a mixture of xanthin and hypo-
xanthin nitrate and I base my conclusion regarding the guanin and
adenin groups of the nuleic acid on the relative quantities of xanthin
and hypoxanthin obtained. One would scarcely expect to find either
guanin or adenin after a violent evolution of the oxides of nitrogen
and as a matter of fact no trace of either is to be found. Nucleic
acid may be hydrolyzed with diluted nitric acid from which all traces of
the oxides of nitrogen have been removed and nitrates of guanin
and adenin beobtained. But under these conditions there is no evolution
of the oxides of nitrogen and the deposition of the nitrates is very slow,
requiring several days for completion. Steudel specifically states concen-
trated nitric acid describes the violence of the reaction, and repeats concen-
trated nitric acid in subsequent references, (Zeztschr. f. physiol. Chem.,
liii, p. 14.)

3 Schmiedelberg: Arch. f. exp. Pathol. u. Pharmakol., \vii, p. 3009.

6 Identity of Nucleic Acids

(4) Nucleic acids from plants and fish eggs yield uracil while
those from animal organs and fish spermatozoa yield thymin.

(5) There is considerable evidence for regarding the nucleic
acids of different mammalian organs as identical substances, and
there is no insurmountable evidence against the assumption.
The statements in the literature which are at variance with
these conclusions are satisfactorily explained on the following
grounds.

(6) In some cases where it was difficult to obtain material
the experimenter has endeavored to study the purin and pyri-
midin derivatives with the same specimen, and in order to insure
the complete hydrolysis of the nucleic acid, hydrolytic agents
have been employed which remove amido groups from both
purin and pyrimidin rings.’ In this manner, xanthin hypo-
xanthin and uracil have resulted by secondary action from guanin,
adenin and cytosin.

(7) The presence in the tissues of ‘‘amidases’’ was not known
to the earlier investigators, who in many cases must have worked
with glands that had stood for a time and in which the removal
of amido groups had proceeded to a considerable extent.

(8) There exists in various glands a curious nucleic acid (if
it be properly so called) which yields an excessive amount of
both pentose and guanin. This substance may well be responsi-
ble for the furfurol reaction given by ordinary nucleic acids and
has undoubtedly produced a confusion in regard to the excessive
amount of guanin supposed to result from the hydrolysis of some
members of the class of ordinary nucleic acids.”

The object of this communciation is to describe some experi-
ments with preparations of ordinary nucleic acid of various
origin (spleen and pancreas of the pig and thymus of the calf)
which have been obtained in such a manner as to avoid all errors
from any source that has been mentioned above, and to furnish
what should be considered conclusive evidence of the identity
of these particular nucleic acids. Upon finding that guanylic
acid is not confined to the pancreas but is present in the spleen,
the mammary gland and probably in other organs, various

‘Inouye: Zettschr. f. physiol. Chem., xlviii, p. 181; Inouye Katsuji
and Katoke: Jbid., xlvi, p. 201.
2 Jones and Whipple: Loc. cit.

Walter Jones 7

glands were first treated in such a manner as to remove the
guanylic acid and ordinary nucleic acids were then prepared
from the residual material. While the primary object of this
procedure was the removal of guanylic acid (i. e., its nucleo-
proteid) it has the great advantage of removing other objection-
able constituents of the glands and especially coloring matter
so that the ordinary nucleic acids obtained from the residues
yield sodium salts which are peculiarly adapted to polarimetric
measurement. Thus the nucleic acid of the spleen which when
prepared directly from the organ is always somewhat colored,
when prepared by this new procedure is so nearly colorless that
the polarimetric measurement of a 4 per cent solution in a 2
decimeter tube offers no greater difficulty than an empty tube.

Various nucleic acids prepared in this manner have the follow-
ing points in common:

(1) Their neutral sodium salts are all dextrorotatory and all
have the same specific rotation.

(2) The variation of the rotation with concentration is in each
case according to the same law.

(3) The variation of the rotation by the addition of ammonia
or alkalies is according to the same law in the case of each nucleic
acid but different from the variation by dilution with water.

(4) The variation of the rotation of the neutral sodium salt
by the addition of acids obeys a third law which, however, is the
same for each nucleic acid.

(5) All three of the nucleic acids examined, form gelatinous
neutral sodium salts.

(6) In the case of all three preparations the gelatinous charac-
ter of the neutral sodium salt disappears upon the addition of
either acid or alkali and the specific rotation changes accordingly.
This easy passage of the gelatinous into the non-gelatinous
form and the reverse offers a simple explanation of the physio-
logical localization and migration of nucleic acid.

(7) It is intended to confine this paper to the polarization and’
gelatinization of the sodium salts of the three nucleic acids
mentioned, yet it may be stated here that all three preparations
yield both guanin and adenin and in relative quantities which were
found to agree as closely as could be expected from the methods
at our disposal.

8 Identity of Nucleic Acids

PREPARATION OF ORDINARY NUCLEIC ACIDS FROM SPLEEN, THYMUS
AND PANCREAS.

The best specimens of the nucleic acids of spleen and pancreas
have been prepared and analyzed by Levene. In his method a
solution of the nucleic acid in dilute acetic acid was treated with
picric acid for the removal of the proteids and the nucleic acid
was precipitated from the filtered solution by the addition of
copper chloride. The method is an excellent one for the elimina-
tion of carbohydrates and other constituents of glands which
were known at the time the method was proposed but whether
by this process guanylic acid is removed is still a question.
Levene found spleen nucleic acid characterized by the relatively
large amount of cytosin which it yields. Thymus nucleic acid
has been the subject of a large number of researches but no
one has prepared a nucleic acid from this gland by the method
to be described. I make no claim that this gland contains gua-
nylic acid but in case it does the substance will surely be removed
by the method described. Steudel has recently confirmed the
existence of guanylic acid in the pancreas and states that after
the removal of this substance the residue answers the require-
ments of Neumann’s method and yields an ordinary nucleic acid
which produces both guanin and adenin but which differs from
thymus nucleic acid in that its sodium salt does not gelatinize.
I have not observed such a difference. The neutral sodium salt
of pancreas nucleic acid forms a 6 per cent solution which on
cooling to the room temperature gelatinizes so solidly that the
vessel may be inverted without losing its contents. Nor do I
understand that Neumann was accustomed to boil out his glands
with water before extracting the nucleic acid. He boils with a
very dilute acetic acid solution which is done for the coagulation
of the proteids so that the glands can afterwards be finely ground.
Such a procedure with the pancreas or spleen would of course shut
in the guanylic acid which latter would surely contaminate the
final nucleic acid preparation.

The method of preparation which was employed for the
material described below was essentially the same for all three
glands and may be briefly described as follows: Portions of
ground gland of one kilo each were well mixed with distilled

Walter Jones 9

water at the room temperature and the mixture brought slowly
to the boiling point. The fluid was roughly decanted from the
coagulum for the removal of all suspended as well as soluble
material and the latter was freed as far as possible from the fluid
by pressing with linen. The residue was again finely ground
in a machine for the purpose, boiled out with 24 parts of distilled
water, and the aqueous fluid roughly decanted and pressed out
withlinen. The grinding and extraction with water was repeated
a third time. The final residue was added in small successive
portions to a solution of 10oo grams of sodium acetate and 3o
grams of caustic soda in 2 liters of water which had been pre-
viously heated to the boiling point and which was kept hot during
the addition of the gland.

After the gland residue had all been added to the alkaline
solution the product was kept just below the boiling point for
20 minutes and while still hot was treated with 25 per cent
acetic acid until the fluid failed to give any indication with
red litmus. This is considerably past the point where the fluid
begins to turn blue litmus. A fluid brought to this condition of
acidity will usually filter rapidly, giving a perfectly clear filtrate
and it is necessary that this condition be reached either by test-
ing with litmus or by frequent attempts at filtration with small
quantities of fluid and subsequent addition of small amounts of
either acid or alkali as the case requires. The filtration is best
accomplished with a hot water funnel, since the phosphates are
precipitated as the solution cools and soon stop an ordinary fil-
ter. With the thymus gland the hot water funnel is necessary to
prevent gelatinization, since we are here dealing with a more
concentrated solution of nucleic acid. In any case the filtrate
is exactly neutralized and evaporated to about 700 cc. (in case
of the pancreas to a smaller volume) and while warm poured
into an equal volume of alcohol. A snow-white, voluminous
precipitate of the neutral sodium salt of nucleic acid is thrown
out which on standing adheres to the bottom and sides of the ves-
sel so that the colored alcoholic fluid may be completely decanted.
The product was washed in turn with alcohol of increasing con-
centration until 95 per cent alcohol was employed, when the
material was pressed as dry as possible with linen and dissolved
in 200 cc. of hot 1 per cent caustic soda. At this point differ-

10 Identity of Nucleic Acids

ent procedures were adopted with different preparations, the
heating of the alkaline solution being discontinued in some
cases as soon as the phosphates had separated, leaving a clear
interstitial fluid while in other cases the heating of the alkaline
solution on the water bath was continued for six hours. This
variation was adopted after some knowledge of the optical
properties of the products had been acquired and its object was
to see what influence the heating with dilute alkalies would
have on these optical properties and also on the gelatinization
of the neutral sodium salts. The results will show that the
alkaline treatment has no influence in either respect.

The alkaline fluid was filtered from the phosphates (there is
no necessity to use a hot water funnel at this point) and exactly
neutralized with acetic acid. When the neutral point is ap-
proached the fluid becomes viscous and just before it is reached
the material stiffens to a solid jelly even in the warm, and
this is the case no matter what the length of time during which
the substance was heated with alkali. The neutral salt gel-
atinizes, the other salts do not. It is, therefore, necessary to
make the final neutralization with a hot solution. The neutral
fluid is poured into alcohol and the sodium salt dried with
alcohol and ether as before described. This product dissolves
in water to a practically colorless solution but an insignificant
amount of insoluble phosphate gives the faintest indication of
a cloud. The material was, therefore, made into a 6 per cent
solution in water and allowed to stand over night. The cloud
collected to a well defined flocculent deposit which was easily
filtered off (after warming the solution) leaving a perfectly trans-
parent, almost colorless fluid. This was poured into alcohol and
the product ground in a mortar with successive portions of
alcohol, which were used for drying. In this way the material
is brought to a fine state of division so that the final product
after the use of ether falls immediately to a perfectly dry dust-
like powder which has no odor of alcohol. In order to be sure
of equally dry preparations, this material was not allowed to
stand before use, but portions were immediately weighed out
and made up into solutions intended for the polarimetric work
that follows.

A description of any one of the three preparations will answer

ee

Walter Jones II

equally well for the other two. All are snow-white, dustlike,
non-hygroscopic powders, which form aqueous solutions that
are perfectly transparent, practically colorless, and neutral to
litmus. Even concentrated aqueous solution gave only an
opalescence with alcohol in equal volume, but the addition of
sodium acetate in traces causes the sudden appearance of a
flocculent precipitate leaving a clear interstitial liquid. None
of the preparations give any response to the biuret reaction
even when made by the accurate method which Schmiedeberg
has lately proposed.t A 2 per cent solution has little viscosity
at the room temperature; a 24 per cent solution is about as mobile
as concentrated sulphuric acid; a 3 per cent solution drops very
slowly from a burette with stopcock open and gelatinizes on
standing over night in a cool room, and a 5 per cent solution
gelatinizes completely as soon as the room temperature is reached.

THE OPTICAL PROPERTIES OF THE ORDINARY NUCLEIC ACIDS OF
THE THYMUS, PANCREAS AND SPLEEN.

Preliminary experiments had shown that on account of the
gelatinous property of the sodium salts, the maximum concen-
tration which could be satisfactorily employed was a 4 per cent
solution. Solutions of this concentration were prepared and
the more dilute solutions were obtained from this by the addition
of water (or reagent) in the quantities indicated in the tables.
This dilution cost some trouble owing to gelatinization. The
original 4 per cent solution may be brought to the room tempera-
ture and allowed to stand for an hour or so without solidifying
but on standing several hours gelatinization occurs and the
material must be warmed for further use. In making up the
4 per cent solution, the material was usually dissolved in hot
water and transferred to a measuring flask, which was filled
nearly to the neck with hot water and the materials mixed as
well as possible while still hot. The solution was allowed to
stand over night and filled to the mark with cold water. The
gelatinous material in the bottom of the flask was then melted
by slight warming and the fluids thoroughly mixed. The dilu-
tion was afterwards done with both water and nucleic acid solu-

1Schmiedeberg: Loc. cit.

12 Identity of Nucleic Acids

tion at a temperature slightly above that of the room. The
polarizations were made in 2 dm. tubes and the angles recorded
are in each case the mean of five readings. In the cases of the
more dilute solutions the recorded angle is the mean of a large
number of readings selected near the mean,

The observations recorded in the tables were made with the
three neutral sodium salts of the nucleic acids of the thymus,
spleen and pancreas. For studying the effects of heating thymus
nucleic acid with dilute alkalies a series of experiments were
made with a specimen of sodium salt of this nucleic acid which
had been heated on the water bath for 6 hours with 1 per cent
caustic soda. This specimen is designated in the tabies as
‘Rhynnis TT

As will be seen the rotation of each sodium salt varies with the
concentration and also by the addition of acids or alkalies. The
laws of these variations are brought out more clearly in the
diagrams which have been constructed from the tables. In
these diagrams ordinates represent the observed angles of rota-
tion divided by half the percentages of the solutions. This
value which has been used for its convenient magnitude is of
course a function of the specific rotation [¢?] and if the rotation
of the substances did not vary under the conditions employed,
all curves would be straight horizontal lines. Abcissas represent
the percentages of the solutions multiplied by five-thirds.

Diagram I shows curves of dilution with water. All three
curves are slightly ‘‘convex upward,” although they are nearly
straight lines. The irregularities near the origin are to be
expected where one is dealing with such dilute solutions that an
error of 2’ in reading the polarimeter would make a considerable
difference in the location of the point on the diagram. The
spleen and thymus curves are practically coincident while the
slight but uniform departure of the pancreas curve might reason-
ably be attributable to an error in making up the original 4 per
cent solution. This assumption is not borne out, however, by
the pancreas curves in the diagrams which follow. The pancreas
solutions were all allowed to stand over night before polariza-
tion while the spleen and thymus solutions were not. A slight
multirotation or change in temperature might therefore account
for the small difference. The points of interest are that the

Walter Jones 13

curves are as close as one might reasonably expect and all have
practically the same curvature.

Diagram II shows curves similarly constructed for the three
nucleates but instead of water the diluting fluid used wasa 5 per
cent solution of acetic acid. The curve for the pancreas nucleate
crosses the other two and then falls slightly below. Asin the
curve above the points marked 7 and 8 are of little value. All
three curves show a sufficient approximation to coincidence and
all contrast with the curves in Diagram I, in that they are ‘‘con-
cave upward.”

Diagram III shows similar curves where the diluting fluid used
was 24 per cent ammonia. Here the curves show a marked fall
in the beginning but soon approach the horizontal position and
become evidently straight lines. This indicates that the fall
in rotation is connected with salt formation, an explanation
which applies equally well to the phenomena brought out in
Diagram II.

Diagram IV shows similarly constructed curves of nucleates
from spleen and thymus, where the diluting fluid used was 5 per
cent ammonia. If the hypothesis in regard to salt formation
be correct the complete replacement of all the acid hydrogen
atoms occurs more rapidly in this case than when more dilute
ammonia is used. This is just what is indicated. The curves
which are practically coincident to point 4 descend abruptly in
the beginning and reach the low level much sooner than was the
case with the use of the morediluteammonia. Experiments with
solutions of ammonia between 24 and 5 per cent produce curves
whose position lies between those of Diagrams III and IV.

Diagram V shows similar curves constructed from data
obtained in experiments with two specimens of thymus nucleic
acid. The one specimen was heated for half an hour with 1 per
cent caustic soda and the other for 6 hours. In the longer heat-
ing no sodium acetate was used to protect the nucleic acid from
the action of the alkali. The fair coincidence of the curves
would indicate that the decomposing effect of alkalies, at least
on specimens of nucleic acid that have been previously treated
with alkalies, is much less than is commonly supposed.

It is difficult to see how the relations between these curves
could be those described unless we are here dealing with chemical

14 Identity of Nucleic Acids

substances that are identical even to their space relations. It
is remotely possible that the small differences observed are to
be attributed to slight differences in chemical constitution
between the nucleic acids but it is more reasonable to assume
that they represent deficiencies in method.

ON GELATINOUS AND NON-GELATINOUS SALTS OF NUCLEIC ACIDS.

In 1898 Neumann! found it possible to prepare from the thymus
gland two different nucleic acids. The one of these (a-nucleic acid)
forms a sodium salt whose solution in warm water gelatinizes
on cooling while the sodium salt of the other (@-nucleic acid)
exhibits no such property. Neumann did not believe that
these two acids existed independently of one another in the
gland but that the @-acid is formed at the expense of the a-acid
by the action of the hot caustic soda which he used in his method
of preparation. The existence of these two thymus nucleic
acids has been accepted by later investigators and it has usually
been granted that thymus nucleic acid is characterized as the
sole nucleic acid which yields a gelatinous sodium salt. Thus
Steudel’ describes a nucleic acid of the pancreas as very similar
to thymus nucleic acid but different from the latter in that its
sodium salt does not gelatinize. Kostytschew® by a study of
the barium salts confirmed Neumann’s discovery of the two thymus
nucleic acids, and Araki‘ professed to demonstrate the existence
in the thymus of a ferment which causes the conversion of the
a-nucleic acid into the f-nucleic acid. ©

Schmiedeberg remarks in his-recent communication® that he
has frequently noticed the tendency to gelatinize on the part of
copper compounds of various nucleic acids, a property which
was lost when the copper compounds were further treated with
chemical reagents. While he does not claim to have discovered
the relation between the a and # acids he denies that there is
any essential difference between the two and expresses the
opinion that the one acid is simply the anhydride of the other.

‘Neumann: Arch. f. Physiol, suppl., p. 552, 1899.
* Steudel: Zeitschr. }. physiol. Chem., liii, p. 530.
* Kostytschew: Ibid., xxxix, p. 544.

* Araki: Ibid., xxxviii, p. 98.

5 Loc. cit.

Walter Jones 15

Schmiedeberg’s denial of the independent existence of the two
thymus nucleic acids is perfectly correct, but the explanation of
the relation between the gelatinous and non-gelatinous sodium
salts is even simpler than the one which he proposes. We are
in fact here dealing merely with two different sodium salts of the
same acid.

It has been shown above that the optical properties of the
sodium salt of thymus nucleic acid prepared by moderate treat-
ment with alkali do not differ from those of a sodium salt prepared
by excessive treatment with alkali, and yet this one difference
in the method of preparation has been supposed to be the deter-
mining factor in the production of the two substances. It may
now be added that there is no easily observed difference in the
viscosity of solutions of the two preparations with which the
polarimetric work was done. This was determined by dropping
the two solutions in succession from the same burette with open
stopcock and counting the drops that fall in a given time. The
differences noted in a number of experiments were no greater
than were noted in the polarimetric work.

Thymus nucleic acid is evidently a polybasic acid which forms
several series of salts. One of those salts has been called in this
paper the neutral salt on account of the behavior of its aqueous
solution to litmus. This salt (which properly should be called
an acid salt) is the one which gelatinizes easily in a 5 per cent
solution and shows the maximum rotation to polarized light. If
a trace of either acid or alkali be added to a solution of this salt
evidently the relation between base and nucleic acid is disturbed,
a more acid, a less acid or a neutral salt is formed, as the case may
be, the specific rotation falls and the gelatinous character 1s lost.
Abundant opportunity to notice the loss of gelatinization in the
dilution of solutions with acid and ammonia was offered by the
materials used in the polarimetric work.

In precipitating the sodium salt from its aqueous solution with
alcohol, the non-gelatinous form will be obtained unless the
aqueous solution is neutral to litmus. This non-gelatinous salt
will not react neutral to litmus and will show a low specific
rotation. On dissolving this salt in water neutralizing the solu-
tion and again precipitating with alcohol the gelatinous salt is
obtained. It happened on several occasions that a preparation

16 Identity of Nucleic Acids

which was intended for polarimetric measurement did not
possess the proper gelatinization and its rotation was found too
low. On treatment with ammonia the rotation rose in the
beginning instead of falling. A little alkali was added to the
solution when the viscosity was increased so that the solution
gelatinized solidly on standing and the rotation rose to the proper
point. This phenomenon has been occasionally observed with
preparations that were neutral to litmus so that it must be con-
cluded that in dealing with these salts the rotation and gelatiniza-
tion are more sensitive means of determination than is litmus.
These points relative to gelatinization apply equally well to all
three nucleic acids here under discussion.

Finally, a number of preparations of thymus nucleic acid
which were made years ago were examined. These materials
were labeled “gelatinous sodium salt’ and ‘‘non-gelatinous salt,”
in accordance with the method that had been employed in their
preparation. But many of those called non-gelatinous were found
gelatinous and vice versa. Those which actually gelatinized
showed a higher rotation than the others and the one salt could
be converted into the other by the method already suggested.

The various chemical differences between the animal nucleic
acids of different origin may of course be due to differences in the
chemical constitution of the nucleic acids themselves, but the
history of the development of the subject suggests that some at
least and perhaps all of those differences are due to the great
difficulty of making chemical examinations especially in regard
to the relative amounts of the-decomposition products. It is
difficult to understand how the three nucleic acids described in
this paper could possess the same specific rotation which varies in
each case according to the same law in whatever way the variation
is produced, or how the sodium salts could show the same degree
of viscosity which disappears and reappears under precisely the
same conditions unless the three nucleic acids in question are
identical.

17

00°%
ol '%
es GES
o 19'S
& 08°2
wa 66%
o 60'¢
=)
ic
= “guao 10d
0G

09°
90°
88°
06°
66"
96°

Ot nrt a st ©

‘0 o[suRB
paaresqo

*(X) SVHUONVa

Os I 9&0
08 T g¢'0
00'S | OOT
666 e8°T
LV'G GL G
L9°% vL €
08S 94°F
06°% o08
*queo red 0 o]3uB
DZ peAresqoO
‘(o) Nadas

a
i=)
ANA A

0S”
TZ"
¢c8"
86°

NNN

“quad 10d

VG

€0 | $9'T 0
1¢°0 061 L4¢°0
LOM 40'S 60 'T
81 0% 8 T
GL % Oa CLS
61°§ OLS 81°
c8 7 88° 06°P

of6 G [6G 006 'S
‘oazue | Ue0t0d | -y ague
pearosqg | Zz ‘| paasesqo

‘(Y) IT SOWAHL

*(.) I SQWAHL

OF”
09°
00°
09°
0G
08”
OF
00°

NN res oc

be)

*93B} U9019 qT

OST 0G
0 LT O'€
0 ZI OF
06 09
GL 8°8
GP ¢ Or
ans 9° €T
UOT}NOS | [BUIZIIO
09 "99
“19PBA “uolyn[og
“-NorLaTia

“UOYDAQJUBIUOD YUN UOYDIOL UL WOYDIULDL Buanoysy

‘I ATAVL

co.)

ot

HN

rN

*(.) I SQWAHL

610 OF 0
Te 0 09° 0
190 00'T
SIT 0o'T
68° 1 0G
L8°% 08 °%
CoP OnE
006° ¢ 00°F
“20 o[suB
PeATssqO | -asujueo10g

18

‘pip ayaon quad wad g yun uoynpp fig woynjos ur WoYMLiDA BuLnoys

7 06°0 81 0 $80 AM 0) 080 910 ¢6 0
5S} €6°0 8c 0 16°0 6a 0 Olt e¢'0 60° T
< OL'T g¢'0 FIT L¢°0 0c T 09°0 ce I
Wiz Sie cy T OT T cyt Hv yh SFT
= 69 T 6L I ih | Gia OZ LSial CLT
ha 10% 18% L0°G 06°% ites [6% c0'°G
S| 0S '% GOP 9F'G ro 4 SoG SEV OSG
ZA FOE 80°9 06°G o08 ¢ 86% of6 G [6G
i
o) r As be |e ahlagt eon | deta Se]
iA “queo 130d “20 o[3uB ‘quad 19d “0 a[sue *queo 139d “2 a[3uB *quo0 19d
ay DG paAresqo 0B pearesqg DZ pearesqo 0Z
iG *(X) SVGUONVd *(@) NEaTaS CY) Il SAWAHL ee
—_

‘Il AIAVL

OST 0%
0 LI Os
0'GI OP
0'6 09
Gee 88
a ¢ OL
VG | 9 €T
uorqznjos | [Burgio
oe) | 29
*ploe o00V “uOIyN[oS

“NOILOATIG

Hip Oo fr ©

19

NOwooanranrd
=a 00D Or ty ©
MANNA OOOO So

0s”
FO"

Walter Jones

‘quod 190d

0G

*(X) SVauONVa

i i

"2 o[3uB
pearasqo

‘puowmwup quao sad $e ym uoynpp fiq uoynjos ur uoynrior Burnoyg

0G

*que0 190d

*” o[3u8
peAresqo

*queo 19d

Cid

¢9'0 €l'0
010 120
ol 0 9¢°0
82°0 c9'0
96° 0 90 T
LE 1. 66 T
o8 1 60°€
61% r6 €
GPG 09°F
06°G 008 '¢
*queo 10d “p a[su8
0G pearesqO
*(o) N@GITdS

“(Y) I SOWAHL

0
0
0
0
OTT
SG
€
€
LLY
006 '¢

‘vp a[suB
pearesqg

*(.) I SAWAHL

|
|
|

i=)
ie)
HMNHMMANRA AOS

*97B] Ud018 7

|
|
|

WDAAHNNOCOOO SO

mwoMnyos
"99

“BIUOUIUIY

*NOILATIC

tO wMmodoorce

GST
[surs110
le)

“uoTyN[os
|

ie)

N OD HID © P~

“TH GTA

Identity of Nucleic Acids

20

‘queored |.» azue
DZ pearesqo

*(X) SYSUONVd

080 eae (0) OL 0 FLO 09°0
jel AY, €@'0 19°0 02 0 19°0
PL 0 Lé'0 PL 0 L€°0 49 0
8L4°0 69 0 8Z°0 69 0 OL 0
880 96 °0 ¢6 0 gO T 88 °0
tlk 8¢o'T Sila 8c T OTT
SV TI EPG os T Go'G LV I
69 T 16% 69 I 66'S G9 T
06 T toes €0°G c3'§ 161
06° 08 ¢ 86°G of6 G C6'G
“guegied | -yajgue | 3ue0ted | -y ajgue | “yue0 Jed
DZ pearesqg DZ pearesqg | Dz
‘(@) Naaids *(Y) 10 SOWAHL |

69°
o06—

i
Ot ee
WOMAN ooo Oo

70 a[suB
peAresqo

‘(.) I SQWAHL

OF
09°
00°
09°
OG
O8”
OF
09°
OS”
00°

HWTMNOMMA ANA AO ©

‘ode] U9010g

0°8T 0%
O'LT (ite

| “OF81 0'F
pO 0°9
GL 88
cP ¢ OL
bG 9 &I
IL v FI
8°0 G ST
uOIynos | [BUIZLIO

29 0)

“BIUOWULY “uoIynjos
*NOILOATIIG

“‘pUuoMUD quad tad g yyun UoYNpp fiq UoYD}O. Ur WOYDLLDA Buinoyy

“AI GIAVL

Identity of Nucleic Acids

Identity of Nucleic Acids

DIAGRAM IV.

Identity of Nucleic Acids

a

NOTE ON THE CONCENTRATION OF DIPHTHERIA TOXIN.
BY P.G. HEINEMANN.
(From the Bacteriological Laboratory of the University of Chicago.)

(Received for publication, May 7, 1908.)

It is well known that a high potency diphtheria toxin is
advantageous in the production of diphtheria antitoxin. The
antitoxin potency depends to some extent on the injection of
increasing doses of toxin. The amount of toxin which can be
given to a horse with safety is limited, however, by the volume;
as a tule the use of a toxin of less potency than 0.005 is not
advantageous. A toxin, the m.l.d. of which lies in the fourth
decimal, has been obtained occasionally, but not with regularity.
Such powerful toxins would enable us to give higher doses, so
that sera of high potency could perhaps be produced. By
precipitation we can also remove some of the constituents of
the toxin broth, which are useless, if not injurious.

In the literature several methods have been described for
preparing a concentrated and purified toxin, mostly, however,
in small amounts and under experimental conditions. Thus
Brieger and Fraenkel’ precipitated toxin with ethyl alcohol
acidulated with acetic acid. They stated that ammonium
sulphate, sodium chloride, and sodium sulphate also precipi-
tate toxins. Wassermann and Proskauer? precipitated toxin
by addition of eight to ten times the volume of acidulated abso-
lute alcohol. Another method used by these authors consists
in evaporation to one-tenth the volume, dialysis to remove the
peptones and then precipitation by use of acidulated 60 to 70
per cent alcohol. Brieger and Cohn? precipitated tetanus toxin
by supersaturation with ammonium sulphate and stated that
the precipitability of tetanus toxin with ammonium sulphate

1 Berl. klin. Woch. p. 241, xxvii, 1890.
2 Deutsche med. Woch., xvii, p. 585, 1891.
8 Zettschr. f. Hyg., xv, p. 1, 1893.

a7

28 Concentration of Diphtheria Toxin

decreases as the toxin becomes more free from foreign substances.
Brieger! then used neutral lead acetate and precipitated the toxin
almost quantitatively. The lead acetate is rather difficult to
get rid of. This method is better adapted for diphtheria than
tetanus toxin. Finally Breiger and Boer? prepared a double salt
by precipitating the toxin with a 1 per cent zinc chloride solu-
tion and removed the zinc by means of ammonium phosphate.
The toxin was then purified by redissolving and reprecipitating
with ammonium sulphate. Banzhaf? has used alcohol as pre-
cipitating agent, but later states‘ that the advantages of con-
centrated toxins do not justify the use, when the expense of
preparation is considered. This would also be true as regards
the use of some other precipitating agents, as acetone. The
experience of other observers gives one the impression that toxins
suffer in potency by use of alcohol and acids. Salts, however,
do not seem to affect them. Banzhaf also stated that his
experiments with horses do not prove that a higher potency is
obtained with concentrated toxins than is obtained with ordinary
toxins. The chances of abscess formation, however, are probably
lessened with a reduction of volume injected and the possibility
of saving a weak toxin is an additional advantage.

In preparing a concentrated toxin to immunize horses it is
not as essential to get a toxin as nearly chemically pure as possible
as to prepare it with small loss and as inexpensively as possible.
I have had some success in concentrating diphtheria toxin in
a series of experiments by the following method: To a given
volume of toxin is added twice the volume of a saturated
solution of ammonium sulphate with constant stirring. Less
ammonium sulphate solution does not precipitate the toxin
completely and more than twice the volume was found to be
superfluous. The mixture is allowed to stand for twenty-four
hours, after which a brownish-yellow scum forms and sometimes
a sediment of the same color. Both the scum and sediment are
then collected on a large folded filter. Twenty liters and more

1 Zeitschr. f. Hyg., xix, p. 101, 1895.

* Deutsche med. Woch., xxii, p. 783, 1896.

3 Collected Studies from the Research Laboratory of the Department of
Health of New York City, p. 35, 1905.

4 Tbid., p. 101, 1906.

P. G. Heinemann 29

can be filtered on one filter in a few hours. The precipitate
is scraped from the paper before it is entirely dry, since it comes
off more readily while moist. The precipitate is then dialyzed
in a collodion or parchment sac. Collodion was used exclu-
sively at first because Oppenheimer' has stated that part of the
toxin is lost by dialyzing through parchment. If heavy parch-
ment is used, however, this does not seem to be true, at least
one experiment made seemed to show this. In this experiment
the precipitate was divided into two equal parts and one-half
dialyzed in a collodion sac, hardened with chloroform and alcohol,
and the other half in parchment. The two lots of toxin were of
substantially the same potency. Since it is difficult to prepare a
collodion sac and seal it to a glass tube without sometimes having
a small leak, it may be more advisable to use heavy parchment
for dialysis.

Similarly to the precipitate a small portion of the filtrate in
each experiment was dialyzed under identical conditions and two
to four cubic centimeters injected into guinea pigs. The guinea
pigs showed no effect resembling diphtheria poisoning whatever,
but steadily gained weight, except in one case, where the animal
died in nine days and cultures from the heart’s blood showed
the presence of staphylococci, but no lesions of diphtheria
poisoning.

The toxin solution is removed from the sac after five or six
days. It is filtered through paper and then is a clear dark brown
fluid. As a preservative I add 0.3 per cent carbolic acid. I
found that by adding a larger amount of carbolic acid to the
toxin a heavy precipitate forms which seems to contain much
of the toxin. Whenever this happens it is advisable to add
enough water to dissolve the precipitate although the potency
is diminished proportionately.

The following table shows the results of the six experiments
made so far:

1 Die Toxine und Antitoxine.

30 Concentration of Diphtheria Toxin

Original m.1. d. Mldsct Pia Theoretical m.].d.* | Dialyzed through—
0.003 0.0007 0.0004 Collodion.
0.006 0.0005 0.0004 .

0.003 0.0003 0.0002

0.009 0.0007 0.0006 e

0.009 0.0007 0.0006 Parchment.
less than 0.005 0.000385 less than 0.00025) Collodion.

* Theoretical m.l.d. is the m.l.d. which would result if there had been no loss what-
ever by manipulation.

The table shows that the loss is not great and undoubtedly
can be reduced by perfecting the technic. By this method it is
possible to make use of a toxin, which is not potent enough to
be used without concentration.

ON THE PART PLAYED BY THE ALKALI IN THE
HYDROLYSIS OF PROTEINS BY TRYPSIN.

By T. BRAILSFORD ROBERTSON and C. L. A. SCHMIDT.

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

(Received for publication, May rs, 1908.)
I. INTRODUCTION.

The part played by the alkali and acid in protein hydrolysis
by enzymes is by no means clear. It is known that the activity
of pepsin is greatly enhanced by acid while that of trypsin is
greatly enhanced by alkali, and, since acids and alkalies are well-
known hydrolyzing agents, the most natural assumption regard-
ing their influence upon protein hydrolysis by enzymes is to
suppose that they play the part of accessory catalysors. As
regards the influence of acids upon the hydrolysis of proteins by
pepsin, however, there are many well-known facts which speak
against this view. The influence of acids is by no means propor-
tional to their degree of dissociation as would be expected were
their influence due to a catalytic action of the hydrogen ion."
On the contrary hydrochloric acid has an almost specific action
upon the hydrolysis of proteins by pepsin, so that many observers
have inclined to the belief that the real ferment is in this instance
a compound of pepsin with hydrochloric acid. As regards the
influence of alkalies upon the activity of trypsin the facts, so
far as has been hitherto ascertained, are much more in harmony
with the view that the alkali acts as an accessory catalysor, since
it is stated that the hydroxides of the alkalies and alkaline earths
act in proportion to their degree of dissociation,* although

1Cf. Klug: Arch. f. d. ges. Physiol., Ixv, p. 330, 1896; A. E. Taylor:
“On Fermentation,” University of California Publications, Pathology,
1907, p. 249; Berg and Gies: This Journal, ii, p. 489, 1907.

3C. Schmidt: Ann. d. Chem., 1xi, p. 311, 1847.

®A. E. Taylor: Loc. cit., p. 251.

31

B2 Hydrolysis of Proteins by Trypsin

divergencies from the proportionality of the action to the degree
of dissociation have been found.!

It has been frequently pointed out that the progress of the
hydrolysis of proteins is accompanied by marked changes in the
acidity or alkalinity of the solutions in which the hydrolysis
occurs’ and it has been suggested that this phenomenon is to be
attributed to the fact that proteins split up, during hydrolysis,
into amino-acids, so that where one molecule of amino-acid
(the protein itself) initially existed in the solution and partially
neutralized the acids or bases present, after the digestion has
proceeded for some time, severa! molecules now exist, each of
which neutralizes as much or nearly as much acid or base as
the original molecule—hence any excess of acid or base tends to
disappear during the hydrolysis and the system tends to approach
neutrality.’

Since the concentration of hydrogen and hydroxyl ions changes
markedly during the progress of hydrolysis, we would expect this
fact to exert a marked influence upon the progress of the reaction.
It happens, however, that the majority of the hitherto pub-
lished investigations upon the progress of protein hydrolysis by
trypsin have been conducted under such conditions that the
alkalinity of the system remained practically constant through-
out the experiment. Thus Bayliss’ experiments were conducted
at a high alkalinity (i. e., 2 cc. } added to 6 ce. of the system)
and any changes due to the alkali-binding power of the products
must have been negligible in comparison with the total concen-
tration of alkali.‘

Robertson’s experiments were conducted in neutral solutions
in which no perceptiblechange in Ht or OH~ concentration occurs
during the progress of hydrolysis.* Henri and his pupils also

1 Berg and Gies: Loc. cit.

? Gillespie: Journal of Anatomy and Physiology, xxvii, p. 195, 1893;
A. E. Taylor: Univ. of California Publications, Pathology, i, p. 7, 1904;
Greenwood and Saunders: Journal of Physiology, xvi, p. 441, 1894; T.
Brailsford Robertson: This Journal, ii, p. 368, 1907.

3T. Brailsford Robertson: Loc. cit.

‘Bayliss: Arch. des science biol. (St. Petersburg), xi, p. 261, 1904,
reprinted in the collected papers of the University College Physiological
Laboratory, London, xiii.

°T. Brailsford Robertson: Loc. cit.

T. Brailsford RobertsonandC. L.A. Schmidt 33

worked with neutral solutions. (Gelatin and neutral casein-
ates.)!

All these observers obtained results in excellent agreement
with the formula

iF

k= ; log. aaah
indicating that the hydrolysis follows closely the course of a
monomolecular reaction. In Taylor’s experiment? the alkalinity
of the system changed markedly with the progress of hydrolysis,
but, as our own experiments show, the influence of the alkali is
of such a nature that its decrease in concentration would not,
under the conditions of Taylor’s experiments, have much effect
upon the progress of the hydrolysis, and in fact Taylor’s results
also confirm the monomolecular formula.

The above quoted investigation by Taylor shows clearly that
the alkali, in the hydrolysis of protein by trypsin, does not
play the part of an accessory catalysor, for if this were the case.
since a considerable portion of the alkali disappeared during
hydrolysis, the monomolecular formula could not have been
obtained. Nevertheless our own investigations were initiated
in the belief that the alkali acted as an accessory catalysor and
that, therefore, in cases where the alkali diminishes appreciably
during the reaction the curve of the hydrolysis would be found
to obey the formula representing the progress of an auto catalyzed
reaction in which the catalysor is consumed as the reaction pro-
ceeds; that is, the formula,

b(@ —*%)

me ka zy

where a and b are, respectively, the initial concentration of the
substrate and of the catalysor. Somewhat to our surprise,
however, we found, at an early stage in the investigations, that
the course of the hydrolysis is by no means represented at all
stages by this-formula nor by any probable modification of the
formula which suggested itself, and our results are such as to

1 Henri and Des Bancels: Compt. rend. de l’Acad. des sct,, CxXxxvVi, p.
1581, 1903; Compt. rend. soc. biol., lv, pp. 563, 787 and 789, 1903.
7A. E. Taylor: Univ. of Calif. Publ. Pathol., i, p. 7, 1904.

34 Hydrolysis of Proteins by Trypsin

indicate that the part played by the hydroxyl ions is somewhat
more complex than that of an accessory catalysor.

II. EXPERIMENTAL.

The method employed was to follow the changes in OH con-
centration during the progress of the hydrolysis with the aid
of the gas chain. The apparatus employed was similar to that
used by one of us in the determination of the potential of a hydro-
gen electrode in acid and alkaline borate solutions! and is fully
described in the paper referred to. The hydrogen was generated
by the electrolysis of water, and subsequently passed over
platinized asbestos heated sufficientiy to rid it of any hydrogen
peroxide or oxygen. The hydrogen then passed through the
very fine meshes of a cylinder of platinized platinum gauze which
formed the electrode. In this way the three phases, platinum,
hydrogen and solution were brought into most intimate contact
and equilibrium was quickly established. The other extremity
of the chain was formed by a standard calomel electrode using
jo KCl Connection between the calomel electrode and the
cell containing the protein solution was made through two small
beakers filled with 58, KCl, the second beaker being connected
with the protein solution by means of a U-tube filled with agar-
agar which had been made up in a saturated solution of KCl.
The cell containing the solution under investigation had aside cup
attached into which the U-tube filled with agar-agar was dipped,
thus preventing contamination of the main body of the solution by
KCl. Any contact differences of potential in the system due to
the fact that KCl was not introduced into the solution under
investigation were eliminated by the interpolation of the satur-
ated KCl.2 The remainder of the apparatus was fitted up in
the usual manner, a sensitive galvanometer being, however, used
as the zero instrument instead of a capillary electrometer.

The experiments were conducted upon the hydrolysis of casein
and of protamine. The casein was Eimer and Amend’s C. P.

*C. L. A. Schmidt and C. P. Finger: Journal of Physical Chem. (about
to appear).

* Richards: Zeitschr. f. physik. Chem., xxiv, p. 37, 1897.

’ Bjerrum: Jbid., liii, p. 428, IQOS.

ee

T. Brailsford Robertsonand C.L.A.Schmidt 35

preparation “‘nach Hammarsten;” for the purpose of this inves-
tigation it was further purified by the method described by
Robertson.!' It was then shaken up with the various concentra-
tions of NaOH until no more would dissolve at 25° C. The solu-
tion was then filtered through glass wool and filter paper. Most
of the filter papers in the laboratories were found to be slightly
acid, but this source of contamination was eliminated by reject-
ing the first portion of the filtrate. In this way solutions of
neutral sodium caseinate were obtained.’

The protamine used had been prepared from salmon sperm by
Dr. A. E. Taylor and was very kindly supplied to us by the
Department of Pathology. It was in the form of the sulphate.
This was directly dissolved in distilled water in approximately
the concentration desired. In both sets of experiments the
procedure was as follows: The requisite amount of Grubler’s
pancreatin (nach Spateholz)* was dissolved in dilute alkali and
the solution filtered through glass wool and immediately added
to the casein or protamine solution under investigation. The
mixture thus prepared was then placed in the cell and hydrogen
allowed to bubble through at a slow rate. To all protein solu-
tions toluol was added to prevent bacterial infection, and a
layer of toluol was, in addition, placed upon the surface of the
mixture in the cell to further secure freedom from bacterial infec-
tion and to prevent foaming. A previous determination with a
borate mixture showed that the toluol has no influence upon the
equilibrium between the electrode and the solution. Hydrogen
was passed through at a slow rate and after about half an hour
the first reading was taken.

In alkaline solutions, when hydrogen is first passed through
the electrode and solution, the solution pressure of the hydrogen
in the electrode is considerably greater than the osmotic pres-
sure of the hydrogen ions in the solution—hence as the system
approaches equilibrium, the apparent alkalinity, as indicated by
the instrument, progressively increases; on the other hand, the
real alkalinity of the solution under investigation is progressively
decreasing on account of hydrolysis. The curves representing

1T. Brailsford Robertson: This Journal, ii, p. 317, 1907.
?7T. Brailsford Robertson: Loc. cit.
Cf. A. E. Taylor: Univ. of Calif. Publ. Pathol., i, p. 7, 1904.

36 Hydrolysis of Proteins by Trypsin

these two processes will evidently meet at some point and equilib-
rium between the electrode and the solution will be established
more rapidly than if hydrolysis were not taking place. From
the point of intersection of these two curves, the potential of
the electrode simply follows the decrease of alkalinity of the
solution, since any displacement in equilibrium is rectified by
the changes occurring in the solution and the velocity of hydrol-
ysis was, in all our experiments, low compared with the velocity
with which the electrode came to equilibrium except, possibly, in
the initial stages where, however, the changes in the solution are
as explained above, aiding the electrode in attaining equilibrium.

The experiments were carried out at the temperature of the
laboratory (average 16° C. ) which, while fairly constant for
periods of one or two hours, probably varied over an extreme
range of 3° C. to 4° C. throughout the day. The storage battery
was checked against the standard Weston cell after nearly every
reading.

All the sodium hydrate solutions employed were only of ap-
proximate strength and the concentration of the protein and of
the trypsin employed were likewise approximate.

The following is a tabular representation of our experimental
results: In the first column is given the time in minutes after
mixing the trypsin and alkali with the solution of protein. In
the second column is given the time (¢—?#,) in minutes after the
electrode has come to equilibrium with the solution; this is
indicated as explained above, by the occurrence of a maximum
in the potential. Since the exact position of this maximum
might readily escape detection in the readings, the time of the
first reading subsequent to the maximum was taken to repre-
sent the time at which the electrode had come to equilibrium
with the solution. In the third column is given the observed
E.M.F. in volts, in the fourth is given the OH~ concentration
corresponding to these voltages,’ in the fifth is given the change
(x) in hydroxyl concentration since the first reading at which
the electrode had come to equilibrium with the solution.

1 The H° concentration at the neutral point was taken as.75 X 10 ', and
the OH™ concentrations were calculated from the Nernst formula:
0002, ae c
£99 Meee .log. +
n e

a

T. Brailsford Robertsonand C. L. A. Schmidt 37

TABLE I.

EXPERIMENT 1. CASEIN. a NaOH “saturated” with casein to which was

added an equal volume of =e NaOH. 10 milligrams of trypsin in 50 cc.
Temperature, 19°.
Time in | k k
oer t—h. he aes Con. pe , (Monomolec- | (Bimolec-
mixing. ular). ular).
34 0.798 | | a95x10-* |
38 OU 02796+)' 5.50107 |.0
51 13 | 0.794 | S206><10m) | OF44 510m" 269510? Pils >alOm
65 27 | 0.791 | 4.50X107 | 1.00107 | 316X10°° | 146x10°
88 DOO Son 4200 10m! | 12 50>10n? | 275X<10m° 136 X10 -*
126 88 | 0.784 | 3.40x10-7 | 2.1010" | 236X10°° | 127x105
158 120 | 0.782 | 3.15X1077 | 2.35xX10™" | 201X10~ | 1131075
182 144 | 0.779 | 2.79x1077 2. 7110-7 | 204410, | 123321052
209 TN OST TT 2.56X10~7 | 2.941077 | 194X10~> | 123x105
215 iaWOLdso 2.38X10~7 | 3.121077 | 205x1075 | 1384x107
247 POOMEON WAT \PaeoO 10m? | oe 2010m" | ISI 10ne | I2tsaloms
347 309 | 0.770 | 1.96X10~" | 3.54X1077 | 145x10°* | 107x10™°
410 372 | 0.767 | 1.72X1077 | 3.78X10" | 135 10m 107 a10m2
530 492 | 0.766 1366.10" | 3.84X10-" | 106X10"* |. 86K10=
614 | 0.763 | 1.47X1077 | 4.031077 9310-5 | 811x107

We may reasonably assume that the change in OH concen-
tration is proportional to the amount of substrate which has
undergone hydrolysis. Since at the end of the experiment,
nearly all the hydroxyl is neutralized, we may assume that at
equilibrium the solution would be neutral. If, therefore, the
hydrolysis followed the course of a monomolecular reaction, it
should obey the formula:

kt—t%) = mtog gf
where f is the concentration of the OH ions at the time (¢,) when
the electrode first comes into equilibrium with the solution. On
the other hand, if it followed the course of a bimolecular reaction,
it should obey the formula:

x

ne 3(@-*

ae

38 Hydrolysis of Proteins by Trypsin

where f and t, have the same significance as before. In the sixth
column of the tables are given the constants calculated upon
the assumption that the reaction obeys the monomolecular
formula, while in the seventh are given the constants calculated
upon the assumption that the reaction obeys the bimolecular
formula.

In the tenth and eleventh experiments (with protamine: cf.
‘Tables X and XI) the reaction was allowed to proceed to a certain
point and then additional alkali was added to the solution in
the cell and the mixture shaken. As at the beginning of the
experiment, the indicated potential at first rises and then falls
after this procedure. The first reading after the maximum

TABLE II.
EMPHRIMENT 2. CasEIN. —N_ NaOH “saturated” with casein to which was
100 . . . . ~-
added an equai volume oj .X, NaOH. 10 milligrams of trypsin in 50 ce.

Temperature 15° to 18°.

Time in SoG k | k
Berean Oe alta: Con 2 (Monomolec- (Bimolec-
mixing. ular), | ular).
51 ORS3valezSe4e5ol Ome |
59 0 | 0.836 | 2723 X10-7 | 0 |
69 0) | OSs || AO << lore L.1-xX10- | L80>Ome iba aGee
73 140: 833) | 24.2 10m? 321 KIO 375 x10 mee poco elon
99 AQ) || (OseYAS) || BOG << lOe4 6.7 <<10m7 | 305 LO e296)
126 67 || 02823) 1622) X10m7 |e S10) 0° 338) Oe avian erome
153 94.) O- 817411228) <10m4 |) lab x 10m | 352 ClOm a esos atoms
197 | 138 | 0.809 9.3 x1077 | 18.0 x1077 | 340X107 | 516x10-*
225 | 166 | 0.804 7.6 1057) 19072 < 1067 ||) 33610 maa pore
246 | 187 | 0.801 6.7 10m | 2056 ° 10>” "326 <1Ome eG00 sero
276 | 217 | 0.798 5.951077 | 21.35x1077 | 305x10 | 604xX10~
303 | 244 | 0.795 5451077 | 21285 10m! (92938 X1On a G2 oom
338 | 279 | 0.792 4.7 1077 |-22.6 X10.7 | 275X109 feanseties
368 | 309 | 0.788 40° 1057 | 2303) X1087 |) 270) X10 s soo atm
408 | 349 | 0.786 3.7 X10-7 | 23.6 %1077 | 250X104) 7GiG ee
443 | 384 | 0.785 3.5° 107? | 238.8 107" 1.232 1055 Gee ee
470 | 411 | 0.783 3.3'X107—7 | 24.0 1077 | 2241075 | 65a eite
547 | 488 | 0.779 2.8 <105" |°24.55' 10-7 |) 203K 10 G60 0m
COM 5425 On iad 2.6 X1057 | 2407 SCl0n7 | 190K 10s y asG palms
647 | 588 | 0.776 2.5 K1077 | 248. 1077 | L780 | e275
(22) \663) 02776 2.5 <1057 | 24-8) 1047 lb lO oob> cl Ome

-
9,

T. Brailsford Robertson and C. L. A. Schmidt

EXPERIMENT 3. CASEIN.
added an equal volume of Oe

Time in

minutes | ¢—f. | in voits.

mixing.
41 0.790
56 0 | 0.790
96 40 | 0.785
112 56 | 0.783
133 77 | 0.779
188 132) | 0.777
205 149 | 0.772
237 181 | 0.770
271 215 | 0.769
305 249 | 0.766
332 276 | 0.765
388 332 ; 0.764
486 430 | 0.762
626 0.759

EXPERIMENT 4. CASEIN.

added an equal volume of .%

Be Ree Re RE ENON N Wwe

TABLE III.

ag

09 NaOH “saturated” with casein to which was
10 milligrams of trypsin in 50 cc.

c k _k
OH. z. (Monomolec- | (Bimolee-
ular). ular).
351077 |
.35X1077 | 0
5410-7 | 0.81X10—" | 23010 | 1810
.28X1077 | 1.071077 | 218x10-* | 1341075
AXIO? | 1.56 X10) 251x102 i ie7 seta
ESO [Om to 1 7910. |) L750 me OOS ctOm
HOS SLOme 221 < 10m 215 10m lGy com
96> 1Omie | 2239X 105" |) 1ST xX 1Oss ts 5sctOme
soa CLOm 2-52 10N | 7x 1Oe a7 SCO
POS 10m! 92. 71 On” | L7OSORe a et5SScolOme
EDO al 22 1S>CLOn? I) 16OS1Ome e471 Ome
A9S<10m | 2286 X10n7)) WOxX1On? Ni ossselOne
VASO 2 95 Xow? | 115 >CLOme | SScrOme
241077 | snaked ui 96107 | 102«1075
TABLE IV.

_N
A ae

Temperature 12° to 14°.

Time in

NaOH “ saturated’ with casein to which was

minutes | ;_ ay E.M.F
after in volts
mixing,
33 0.838
44 0 | 0.836
61 7a) Ons8ol
97 53 | 0.822
121 77 ; 0.816
181 | 187 | 0.804
228 | 184 | 0.797
260 | 216 | 0.793
285 | 241 | 0.791
315 | 271 | 0.789
345 | 301 | 0.786
387 | 343 | 0.784
431 | 387 | 0.782

aN, NaOH. 20 milligrams of trypsin in 50 ce.
c k k
OH z. (Monomolec- ;} (Bimolec-
ular). wlan)
a
an EOF 0
mee inet 4.8 X10™7 | 51410 | 483 10-*
16) 10m? | LEG x1LOm | AGISC1Ome Si 8>c10ne
13> X<1057 | 14.9 X10 | 4525105" 585 x 10
6 xX107—"*| 19.6 X10 | 407X10~° | 695x10-*
751057 | 21.4510" S0G 6 10m) |la747ocl Ome
9 x107-7 | 22.3 1077 | 8461084, | 775x 107°
Rel Om tater LOme 32D xX 10n> |) 769<10m*
2 1077 | 23:0 X1077 | BOL 105 1 750% 10F*
Tec lOm i 2oco oe LOme | 288X105 | 77610 °
ASSO? | 23.8 S10? | 26410-® | 75010-*
151077 | 24.05X10~ | 2483x1075 | 7251078

40

Hydrolysis of Proteins by Trypsin

EXPERIMENT 5. CASEIN.
added an equal volume of =X, NaOH.
Temperature 12° to 15°.

TABLE YV.

N_ NaOH “saturated” with casein to which was
10 milligrams of trypsin in 50 cc.

100

Time in
eae t—h. Fea Con. <r (Moaariaiees @inaer
mixing .| | | ular). ular).
38 0.849 | 46.0 x10? |
57 02853 '-|/ 54.0" x1057
132 0 | 0.847 | 42.4 x10°7 |] O
160 28 | 0.842 | 34.6 x10" |. 7:8 K10" | 315x103 ieee
197 65 | 0:834 | 25.2 .x1077 4 17:2 X10" | 347X104) soe
216 84 | 0.831 | 22.4 x107"'| 20:0 X10—" | 330X105" 7 25s
243 111 | 0.827 | 19.0-K10-" | 23.4 X10="| 316X103 aha
264 132 | 0.823_| 16:2. X1077 |°26.2 X10™" | 317 xX 103- 9) aaa
284 152° | 02821 | 15.0. xX 10-*. | 27:4. K107"-| 297 x 10 eae
299 167 | 0.818 | 12.85xX107—7 | 29.55xX10°7 |.310X10° | 326X10 *
322 190 | 0.815.| 11.8 X10" | 30.6 X10™ | 293X10= | 32270
359 227 | 0.809 | 9.3 X107'* 33.1 X10~' | 290X10 >))e ie
397 265 | 0.805 | 7:9 107-7 | 34.5 x10? | 275x104 |) 38a) iG
509 377 | 0.797 | 5.75xX1077 | 36.6510 | 230K10= 9290)
581 449 | 0.791 4.5° X10 |.37.9. X10 | 217X10= ase
1301 |1169 | 0.772 | 2.1 x10" | 40.3 X10" | 112X103 oe
} {
TABLE VI.
EXPERIMENT 6. PROTAMIN. 2 grams of protainin sulphate to 1 litre of solu-

tion. 10 milligrams of trypsin and 1.4 cc. of NX. NaOH added to 50 cc.

an

Time in
pag th. oles | Con us | Qisaccoiee (BiteAee:
mixing. | ular). ular).
58 0.819 | 13.8 x107
76 0: | 0281500 TRS a 5G iret ay
98 22 | 0.810 OV DOE 1 eal SCS |) BRS Oe 83 xX 1075
128 52 | 0.803 7.3 X<10°7 | 4.5 <X10m? | 403X105 S102
169 93 | 0.796 5.6 10-7 | 6.2 <10m? 349 10s OR 1 Oe
193 |:117 | 0.793 4:9 X<10-" | 6.9 X1OR 327 10m 02st
217 | 141 | 0.790 4-3) K10=7 | 7-5 X10n" | SIO 1OnS Seat Oe
242 | 166 | 0.787 3.85X10074) 7.95 10m! | 293510 Sa aes
271 | 195 | 0.785 3.55X1007 1-8. 25X1087 4) 2675108 | OE etme
307 | 231 | 0.7838 3.3 X10!) B25 Sc lOm aie sae Oe 95x10
410 | 334 | 0.780 2.9 XK1ORT S29 SclOm a els zeolOme 18X10
590 | 514 | 0.777 2.6 X10" | O82 Kk 10m BelOm oos Ome

T. Brailsford Robertson and C. L. A. Schmidt

TABLE VII.

4l

EXPERIMENT 7. PROTAMIN. 2 grams of protamin sulphate in 1 litre of solu-

tion.
Time in |
minutes a
after
mixing.
23 |
35 0 |
54 19 |
76 41
118 83
167 | 132
206 | 171
240 | 205 |
2/0 | 235
309 | 274
aon. | 320
415 | 380 |
460 | 425
|

|

|

10 milligrams oj trypsin and 1.5 cc. of X. NaOH to 50 ce.

EMF. 4 | k ead:
in volts. OH : (Monomolec- (Bimolec-
ular). | ular).
|

OFS47 4225) <105! | |
0.843 S002) SoLOm! 0 | |
D640 | o2-8 X10"? | 4.1 K10"" |. 279 K10-> | tes 10-
0.836 27.3. *10°"'| 8.9 x10°7 | 208x10= | 220210
OSs laie22 eo lOnm >| 1358 K10n? | 253 10m 20sec lOm
OnoceelGso >< lOm. 19.3. X105* 1 250<10a | 23910
ORs nielaasosclOmil 22551057 | 245 5c] Ome 260 salle
OES ploesapel Oman 23 4) X10e% | 2208 10m oso ins
0.815 | sec lOm? | 2454 =< 10" ||) 207 LO ie 4splOms
0.813 | 1029 10-7 |-25.3 X10 | 190<X 1052 | 23410
OnGizs | WlOFS <l0n 125.7 IO | 16810 Se 2s
0.809 OFS <10m* | 126-9 X10—7 | 156sl0ns ozs e10ms
0.809 ORO a 2629 3 CLOm | 139s 0m | 190 10~*

TABLE VIII.

10 grams of protamin sulphate im 1 litre of

EXPERIMENT 8. PROTAMIN.

Time in
minutes |
after
mixing.

|

solution. 10 milligrams of trypsin and

(

2 cc. of N, NaOH to 25 ce.

5 et a)

E.M.F. | : k k
| #—4- | in volts. Con a. (Monomolec- | (Bimolec-
ular). ular).
Oe O9) eG. 2. SOLO
OnwOr7196 | 5.6) 10-7 | 0
20 | 0.790 Aes SANs! | L.3 X10n? | Soot 2 70 sc lO
SLIMOnTSL | o-80x%10°¢ | 1.75xX107* | 479K 10 | 238X105
RSelOsise, | o.15 x10 "| 2.45107 | 4382 K10m | 239X105
SOR I oeoexX 10,. 1.3.0. <10m” | S821 0m 287 Oe
143°| 0.771 | 2.0 <10-7 | 3:6 x10—7 | 318X10— | 227x107
187 | 0.768 | 1.8 10-7 | 323° 10m 263m S202 5105
Zon 02765, | 1.6 X10 7" |-4:0) X08" | 231E<10R 1895105
Oo Onves: | Lb. <10-* (| 4.10 X0m? - L9G KOR | SIGs clon
soo | 02761 | 1.4 X1077 | 4.2 LOS Re IS iolOss

42 Hydrolysis of Proteins by Trypsin

TABLE IX.

10 grams of protamin sulphate in 1 litre of
N NaOH to 25 ce.

10

EXPERIMENT 9. PROTAMIN.

solution. 10 milligrams of trypsin and 2.5 cc. of

: | |
Time in |
ance t—h. ae Con” 20 | (fenonoe | (Bimolec-
mixing. ular). ular).
60 | 0.837 | 28.5107?
85 0.| 01838 | 24-35c10=%8| 0
108 | 23.) 0.829 | 20.7%10°7.| 3.6X10—7 | 30510 gare
136 51 | 0.821 | 15.0X10-7 | 9.31077 | 4111075 | 500x107*
181 96 | 0.812 | 10.5X1077 | 13.8x1077 | 8380x1075 | 56410-*
241 | 156 | 0.802) 7.0X10-7*! 17.3xX1077 | 346x10-* | 647x107°
307 | 222 | 0.795 | 5.3X1077 | 19.0x107~7 | 298x10-* | 66410
356 | 271 | 0.791 | 4.51077 | 19.8x107-7 | 270x10-* | 668x10-*
400 | 315 | 0.788 | 4.0X10~? | 20.3x10~7 | 249x107 | 664X10
453 | 368 | 0.785 | 3.5x107-? | 20.8x10-7 | 227x107 | 663X10-*
502 | 417 | 0.782 | 8.1x10—7 | 21.2x10-* | 213X105 | eGye ate
TABLE X.

EXPERIMENT 10. PROTAMIN.

10 grams of protamin sulphate in 1 litre oj

solution. 10 milligrams of trypsin and 2 cc. of N. NaOH to 25 ce.

Time in

wate t—t. ales | Con. As (Monomolee- caimoine

mixing. ular). ular).
33 0.825 |
43 0 | 0.822)| 15:6 x10=7 |e
60 17 | 0.816 | WP Sei 2 sco LOR | GLE 10m a Oe atte
76 SEU Oeil |) NO Oss lO ¢ 545X100" | 577 xX LOme a eb pel Ome
90 47 | 0.806 8.2 KOT 7.4 X<105 7 | 596 10men eS aes
102 59 | 0.803 Uae Ske? 8.3 X10~" | 560X 10m |Sieaalone
113 70 | 0.800 | GG SU 4 ).0) Sears | 533 10="" | 225s
19 OFSS4E eZ onsmeel Ome
29 On| MOPRS33 7 bao oLOm 0
42 B34) Mstsent | 24 Scugjy—e 1.9 X10-7 | 2731075 PALES UIST
59 60:)0-827 E1708 Salone? Ged) SSlOms! 4505 One 50xX105¢
80 D1 OLS24a LG soem Th See SL SOO 35 >c Om
104 75 | 0,821 | 15.0 home 9.3 <107 7 | 27O310% of x<10m
121 92°) OL 807 | 1228 <0 me eSoL Ome a Oar alOme 40X10
138 | 109 | 0.815 | 11.8 10 Roe Oine 288 ol Ome 405c 10m
162 | 133.) 0.812 | 10:5 K10-7 | 1378 410 1 eo 41510
202 | 173 | 0.808 8:9 <1077 | Toe 4e Ss lOm 252 0m 4NSx<c10ss
277 | 248 | 0.801 65761077 | Ci-6 SalOm e226 ele 441075

0 ee ee re

EXPERIMENT 11.

PROTAMIN.

TABLE XI.

T. Brailsford Robertson and C. L. A. Schmidt

0

43

10 grams of protamin sulphate in 1 litre of

solution. 10 milligrams of trypsin and 2.1 cc. of N. NaOH to 25 cc.

|

Time in

eae t—t. aa | Con. “2c (uduemoee: (Binvolec-

mixing. | ular). ular).
29 OF832' 23.38 X10) 7
38 a0 Sol. | 22.4 10-7 0
63 PoaiO.s2e.| 16.25<10~* 6.15X10~ | 559X10"° | 675105
78 40 | 0.818 | 138.3 x10~" 9.1 K107" | 5671055) 165105
93 Someta 1OkS <1 Om LES SclOn |) GOIS<10R8) | FO205clOm:
109 71 | 0.808 829 S1lOn | 13-5) X 105" | 565K 105 |) Got ae
124 86 | 0.804 eG S<AlOm 14. SeSclOs® | 545>C1Ome rots elie
132 Ose. f.0 X1077? 15.4 107-7 | 537x107> Ge
13 0.1836 | 27.3 X1057
23 (iO .835. | 26.3* x10-7 0
33 10 | 0.833 | 24.3 x1077 2:0: X10°7 | 384X<10n Slee
51 26 | 0.829 | 20.7 x10™ 6.6 X1077 «| S710) shaectoe
72 49 | 0.826 | 18.3 X10~’ 8.0: X10~ | 32310 | 339 105*
90 67 | 0.824 | 16.9 x10~" 9.4 X10 | 286X10™ | 316x10-§
111 Sees }ia.8 KI0-° |) 12.5 <10-7 | 317xK10 | sorx 10
ae 0-817 | 12.8 X10? | 13.5 <1077 | 31410 | 40110
amet 20) 0814 | 11.3 10° 15.0 X1077 | 285x10™ [-391 X10"
170 | 147 | 0.812 | 10.5 1077 | 15.8 x1077 | 2721075 | 389x107*
199 | 176 | 0.810 957 10-7 |.16.6 X10 | 246X105" | S69 10
222") 209 | 0.805 | 7.9 <1077 | 18.4 x10~7 | 25010 | 423x10*
251 | 228 | 0.804 TH 10? 1 18.7) X10-*)| 238651008! ras

was again taken as indicating the initial concentration and both
the time and the changes in alkalinity were reckoned afresh from
this point. The constants were calculated for these latter
observations in the same way as for the observation made before
the further addition of alkali, only the values of # and of 1,
used were those determined after the addition of the alkali.

III. THEORETICAL CONSIDERATION OF THE RESULTS.

It is at once evident that under certain conditions the trans-
formation follows very closely the monomolecular formula, k
calculated for this formula only varying within the extent of

44 Hydrolysis of Proteins by Trypsin

the experimental error, while k calculated for the bimolecular
formula progressively increases. This is true for Experiment 5
(cf. Table V) and at once shows that the alkali does not play the
part of an accessory catalysor, for in this experiment the con-
centration of the alkali changes several hundred per cent while
the value of k for the monomolecular formula remains nearly
constant throughout.

On the other hand, under certain conditions, the transforma-
tion follows very closely the bimolecular formula, k calculated
for this formula being practically constant, while k calculated
for the monomolecular formula progressively decreases. This
is true for Experiments 1, 3, 6, 8 (cf. Tables I, III, VI and VIII).
Under certain intermediate conditions the reaction is mono-
molecular in its initial stages and bimolecular in its latter stages.
This is true for Experiments 2, 4, 7 and g (cf. Tables II, IV, VII
and” 1X).

It is at once evident that the condition which determines
whether the transformation is monomolecular or bimolecular
in character is the hydroxyl concentration. In the tables
embodying the results for those experiments in which the order
of the reaction changes during its progress, the approximate
OH concentration at which the change in the order of the reaction
occurs, is indicated by an asterisk. It will be seen that it is
always in the neighborhood of ro ®OH~. At hydroxyl concen-
trations in excess of this, the reaction is monomolecular, at
hydroxyl concentrations which are less than this the reaction
is bimolecular. For a small range of intermediate alkalinities,
the order of the reaction is indeterminate. This interpretation
is further confirmed by Experiments 10 and 11. In these the
reaction evidently approaches in its initial stages the monomole-
cular type, but as the OH~ concentration approaches the limit
to ° OH the order of the reaction becomes indeterminate.
Were the system left to itself, judging from the results of the
other experiments, the reaction would thereafter approach more
and more closely the bimolecular type. Upon adding alkali,
however, the reaction again becomes of the monomolecular type,
and only as the alkali again approaches the limit ro * OH does
the order of the reaction become indeterminate or bimolecular.

The bimolecular formula indicates a greater decrease in the vel.

_—

T. Brailsford Robertson and C. L.A. Schmidt 45

ocity of transformation than is indicated by the monomolecular
formula. That this greater decrease in velocity is not due to
hydrolysis and consequent destruction of the trypsin, isshown by
the fact that it occurs at the same OH™ concentration independ-
ently of the time consumed in reaching that point, and that it
occurs from the very beginning in precisely those solutions which
contain the least hydroxyl, andin which, therefore, the destruction
of the trypsin is slowest. That it is not due to the depressing
effect of the products, is shown by the fact that it occurs from the
very beginning of the transformation in certain solutions and
that, moreover, if, instead of taking the total amount of alkali
present at the beginning of the experiment as proportional to
the amount of substrate hydrolyzed, we take the amount initially
present less the amount still left in the solution after a consider-
able lapse of time, we obtain no better agreement with the mono-
molecular formula. That it is not due to a “lagging behind”’
of the equilibrium between the electrode and solution is shown
by the fact that it occurs in just those solutions in which the
change in OH concentration due to hydrolysis is least rapid.
That it cannot be due to purely experimental errors in the
determination is shown by the fact that the displacement in the
estimated value of the constant for the monomolecular formula
is always in one direction while the experimental error would be
indifferently positive and negative.

The precise degree of alkalinity at which the reaction changes
character is evidently independent of the nature of the substrate,
since it is unaltered by substituting for the predominantly acid
protein casein, the predominantly basic protein, protamine.
The phenomenon is one, therefore, which primarily depends upon
some relation which subsists between the alkali and the trypsin
and not upon properties peculiar to the substrate or its products.
_ While we do not intend to offer this as a final explanation of these
phenomena we may point out that they are similar to those which
would be obtained were the true catalysor a hydrolyzable com-
pound of trypsin with NaOH (or other base). Were this the
case it is evident that so long as there were enough NaOH
present to combine with all the trypsin, the reaction would be
monomolecular (provided the hydrolysis of the substrate were
itself a monomolecular reaction, a fact which the observers

46 Hydrolysis of Proteins by Trypsin

quoted in the introduction have established), while if the NaOH
present were insufficient to combine with all the trypsin then
the amount of catalysor present in the system would be propor-
tional to the concentration of sodium hydrate, and the constant
k, in the monomolecular formula would vary directly as the OH—
concentration; that is, the transformation would be represented
by a bimolecular formula.

An examination of the experimental data shows that in every
case an apparent equilibrium is attained, the rate of change of
hydroxyl concentration becoming inappreciable; as this is
approached the constants calculated from both formule rather
suddenly fall off. This equilibrium is, however, probably not a
true equilibrium between the substrate and its products for
the following reason: The position of equilibrium, as judged
by the amount of alkali neutralized, varies very much with the
initial concentration of the alkaliin the system—the same amount
of casein or protamine will evidently neutralize very varying
amounts of alkali before reaching apparent equilibrium, accord-
ing to whether the alkali is initially added in great or small
amount. This varying alkali binding power of the system is very
much greater than could be accounted for by the action of OH
ions in increasing the alkali-binding power of an amphoteric
electrolyte; and moreover occurs when casein is used as a sub-
strate, a protein which attains its maximum alkali-binding
power in practically neutral solutions,’ while one, at least, of its
products is an even stronger acid than casein itself (paranuclein).?

Furthermore, not only is the position of apparent equilibrium
determined by the quantity of alkali initially added to the system
but it 1s likewise shifted if the alkali be added during the progress
of the reaction.

We are thus left with two alternatives: either the alkali shifts
the point of true equilibrium between the protein and its prod-
ucts and hence also shifts the observed equilibrium, or else the
observed equilibrium is a ‘‘false’”’ equilibrium, depending upon
the sum of the relations between the protein, its products, the
trypsin, and the alkali. The former alternative cannot be

+ Since its neutral salts obey Ostwald’s dilution law. Cf. T. Brailsford

Robertson: Journal of Physical Chem., xi, p. 542, 1907.
?Cf. Gustav Mann: ‘‘Chemistry of the Proteids,’’ p. 395, 1906.

—————

e

T. Brailsford Robertsonand C. L.A. Schmidt 47

accepted because in that case work would be performed in
changing the OH concentration and the decomposition of the
substrate could not, since the OH~ concentration varies during
hydrolysis, be monomolecular in character at any stage of the
reaction or at any (variable) alkalinity. Taylor’s investigations,'
which were carried out at an alkalinity considerably greater
(¢350) than 107°, as well as our own rule out this alternative
and we are forced to the conclusion that the station of rest
in the system trypsin + protein + alkali corresponds to a com-
plex equilibrium in which the protein, trypsin and alkali are
all involved.

The investigations described above show clearly in what the
“optimum” alkalinity of a tryptic digest consists. Evidently
at concentrations of alkali below 10-® (at the point at which
rosolic acid changes from rose to red)? the velocity of the reaction
falls off more rapidly than it would were the alkalinity higher;
hence alkalinities higher than this secure a greater amount of
hydrolysis in a given period of time. A lower limit of the
“optimum” alkalinity is thus established. At alkalinities

‘greater than this, further increase in OH™ concentration should

be without effect until its action in accelerating the destruction
of trypsin becomes appreciable. Thus an upper limit of the
“optimum” alkalinity is also established; this has been estimated
by Taylor to be about ;35, OH ?

Some experiments upon the progressive diminution in acidity
of a peptic digestion of casein were also carried out, but while
they succeeded in demonstrating that such changes do occur,
the total change was so small that accurate determinations could
not be successfully made, since any experimental error would
form a considerable proportion of the total change indicated.
Moreover, in this case the curves indicating the approach to
equilibrium of the electrode and the changes in the H* concen-
tration of the solution do not cut one another, but tend con-
tinually to diverge and it is, therefore, uncertain to what extent
the readings are influenced by the incomplete equilibrium of
the electrode; for, in any minimal displacement of equilibrium,

‘A. E. Taylor: Univ. of Calif. Publ. Pathol,. i, p. 7, 1904.

*Cf. E. Salm: Zeitschr. f. physikal. Chem., lvii, Heft. 4, 1906.
7A. E. Taylor: Loc. cit.

48 Hydrolysis of Proteins by Trypsin

the changes taking place in the solution instead of rectifying it,
as in alkaline digests, would only magnify the displacement.

IV. SUMMARY.

(1) The progressive changes in alkalinity of tryptic digests of
casein and of protamine have been followed by means of the
gas-chain.

(2) It is found that the progressive changes in OH concen-
tration of the digests can be expressed by a monomolecular
formula when the total OH concentration is greater than 10 ®
and by a bimolecular formula when the OH™ concentration lies
between this and the concentration at neutrality. For a short
range of intermediate alkalinities the order of the reaction is
indeterminate.

(3) It is pointed out that the above facts are inconsistent with
the view that the OH~ ions in a tryptic digest play the part of
an accessory catalysor, and, while we do not advance this as an
explanation of the phenomena, we have nevertheless pointed
out that they are such as would be observed were the real catalysor
in these systems a hydrolyzable compound of trypsin with the
NaOH or other base present in the system.

(4) Reasons have been advanced for believing that the apparent
condition of equilibrium, as regards changes in OH concentra-
tion, which is sooner or later reached in these systems, does not
consist in a true equilibrium between the protein and its products
but in a “‘false’’ equilibrium depending upon the sum of the
relations between the protein, the trypsin, and the alkali.

IX. RESEARCHES ON PYRIMIDINS: SYNTHESES OF SOME
NITROGEN-ALKYL DERIVATIVES OF CYTOSIN,
THYMIN AND URACIL.

(Thirty-third Paper.)
By TREAT B. JOHNSON anp SAMUEL H. CLAPP.
(From the Sheffield Laboratory of Yale University.)

(Received for publication, May 28, 1908.)

The nitrogen-methy] derivatives of cytosin, thymin and uracil
are of especial interest because of the occurrence of methyl-
purins in nature. Purins may be considered as compound rings
in which imidazol has been grafted into pyrimidins. A similar
relationship exists between 3-methyluracil, IV, 1,3-dimethyl-
uracil, VI, and the purins, theobromin, III, theophyllin, V,
and caffein, VII, as exists between uracil, II, and xanthin, I.

NH—CO
. NH—co
O CNH [vt 2]
7 4 Nu = CO CH
| pee Hegel
| =
NH—cN 7 Se ea:
I I
NH—CO NH—CO
CO CNCH, CO CH
7 =
\cH | |
| » CH,N——CH
CH,N——CN7
II IV
CH,N—CO CH,N—CO CH,N—CO
| Pea i
CO CNH CO CH CO CNCH,
m —> cs \)
| CH ya | \cx
| 4 CH,N—CH beck
CH,N—CN CH,N—CNZ
Vv VI VII

49

50 Kesearches on Pyrimidins

These alkyl derivatives are also of interest because of the
possibility that future investigations may show the presence
of methylpyrimidins in animal or vegetable organisms. It is
interesting to note here that Suzuki, Aso and Mitarai,’ in a
paper titled ‘‘Ueber die chemische Zusammensetzung der
japanischen Sojasauce oder Schoyu,” have described a decom-
position product of Schoyu to which they have assigned the
empirical formula C,H,N,. They state that the compound is
probably an isomer of aminodimethyl pyrimidin.?

We shall describe in this paper the syntheses and properties
of 3-methylcytosin, VIII, 1-methylthymin, IX, 3-methylthymin,
X, 1,3-dimethylthymin, XI, and 1,3-dimethyluracil, VI.

N=C.NH, CH,N Co NH=—CO@
ee | | |
CO CH CO CCH, CO. CCH,
eat | | | | |
CH,.N— CH NH—CH CH,.N——CH
Vill exe xX
CH,N—CO CH,N—CO
|
[=| eae
CO. CCH, CO CH
| eee |
CH,N—-CH CH,N—-CH
xa ‘VI

Two nitrogen-methyl derivatives of thymin and uracil have
been described in the literature, viz: 1,3-dimethylthymin, XI,
and 1-methyluracil, XII. Dimethylthymin*? was prepared by
heating the mono-potassium salt of thymin with methyliodide
at 150°. We prepared this compound and also 1,3-dimethyl-
uracil, VI, by warming thymin and uracil respectively in alco-
holic solution with the required proportions of potassium
hydroxide and methyliodide.

1-Methyluracil, XII, was described in a paper from this
laboratory and was prepared in the following manner:

'C. Blatt: ii, 1649, 1907. Bull. College Agr., Tokio, vii, 477.

* Schwarze: Journ. f. prakt. Chem., xlii, p. 1; Schlenker: Ber. d. deutsch,
chem. Gesellsch., Xxxiv, p. 2819; Schmidt: ibid., xxxv, p. 1577.

3 Steudel: Zeitschr. physiol. Chem., xxx, p. 539-

* Johnson and Heyl: Amer. Chem. Journ., xxxvii, p. 628.

Treat B. Johnson and Samuel H. Clapp 51

1-methyl-2-pseudoethylthiourea was condensed with the sodium
salt of ethyl formylacetate giving 1-methyl-2-ethylmercapto-6-
oxypyrimidin, XIV. This same mercaptopyrimidin was also
obtained by treatment of 2-ethylmercapto-6-oxypyrimidin, XIII,
with methyliodide in presence of alkali. Hydrolysis of this mer-
captopyrimidin with hydrochloric acid gave 1-methyluracil, XII.

NH—CoO CH,N—CO CH.N=co
| | |
erce CH => “HSC CH =< CO (CH
| | | | |
N= CH N—CH NH—CH
XIII XIV Paw

We now find that 2-ethylmercapto-5-methyl-6-oxypyrimidin
XV, reacts with methyliodide in presence of potassium hydroxide
giving about equal proportions of the two isomeric pyrimidins—
1,5-dimethyl-2-ethylmercapto-6-oxypyrimidin, XVI (m. 65°) and
3,5-dimethyl-2-ethylmercapto-6-oxypyrimidin, XVII (m. 156°).
Hydrolysis of these mercaptopyrimidins with concentrated
hydrochloric acid gave 1-methylthymin, XVIII (m. 202° to
205°) and 3-methylthymin, XIX (m. 280° to 282°), respectively.

NH—CO
|
GiEse CCH:
| |
NCH
XV

N—CH CH,N—CH
XVI XVII
7 Ls
CiEN-——=CO NH—CO
|
Eee CCH. CO “CGH:
| | | |
NH— CH CH,N——_CH
XVII p<

52 Researches on Pyrimidins

Thestructures of the isomeric methylthymins and incidentally
the corresponding mercaptopyrimidins were established in the
following manner: Wheeler and Johnson! have shown that cyto-
sin reacts smoothly with bromine water giving oxydibromhydro-
uracil, XX. When this hydropyrimidin was digested with
alcohol it was converted quantitatively into 5-bromuracil,
XXI. Since the 6-amino radical is removed by this treatment,

N=—CNH, NH—CO NH—CO

| besa Pie? |

CO = CHO Gos" .CO-= CBr > _ CO eee
|

| | | | | |
NiCr NG —— © EOvEr INGE Ot
OX XX

it seemed probable to the writers that substituted cytosin
derivatives would behave in a similar manner giving substituted
uracils. Our experimental data confirms this assumption. We
find that cytosin reacts with methyliodide in presence of po-
tassium hydroxide giving 3-methylcytosin, VIII. This pyri-
midin reacted with bromine water giving a quantitative yield of
3-methyloxydibromhydrouracil, XXII. When this hydropyri-
midin was warmed with alcohol it was converted into 3-methyl-
5-bromuracil, XXIII, melting at 255° to 260°. The isomeric
1-methyl-5-bromuracil? XXIV, melts at 228° to 229°. 5-Methyl-
cytosin apparently reacts with methyliodide in a similar man-
ner as cytosin giving 3, 5-dimethylcytosin, XXV. When this

N——CNH, NH—CO NH—CO CH,N——CO
| eon AG url | eee
co. cH 4 co Ce, 4”) co Car aan
(aed | eel | | | em
CH,N——cCH CH,N———CHOH CH,N=—¢H NH—CH
VIII XXII SXLIT XXIV

pyrimidin was treated successively with bromine water and
alcohol, it was converted quantitatively into 3-methylthymin,
XIX, melting at 280° to 282°. This result also proves that the

1 This Journal, iii, p. 183.
? Johnson and Heyl: Loe. cit.

Treat B. Johnson and Samuel H. Clapp 53

N=CNH, NH—CO NH—CO
| | Vs
PanccH.. —> CO oe =¥ COACCOER
yal | | SCH, | |
CH,N—-CH CH,N CHOH CH,N———CH
DP XIX

mercaptopyrimidin, XVII, which melts at 156°, is a 3-methyl
derivative since it gives 3-methylthymin on hydrolysis.

The introduction of methyl groups into uracil, thymin, and
cytosin, has a similar influence on their physical properties as
in the case of purins. They increase the solubility and lower
the melting points. For example: while uracil and thymin are
difficultly soluble in alcohol, the methyl derivatives of these
pyrimidins dissolve easily in this solvent and are moderately
soluble in cold water. The 3-methyl derivatives of 2-ethyl-
mercapto-5-methyl-6-oxypyrimidin, thymin and 5-bromuracil
are more soluble in water than the isomeric 1-methylpyrimidins.

It is also interesting to note that the 3-methylpyrimidins
melted higher, in every series examined, than the isomeric
1-methylpyrimidins:

2-Ethylmercapto-1,5-dimethyl- 2-Ethylmercapto-3,5-dimethy]l-
6-oxypyrimidin. 6-oxypyrimidin.
(65°) (156°)

1-Methylthymin. 3-Methylthymin.

(202° to 205°) (280° to 282°)
1-Methyloxynitrohydrothymin. 3-Methyloxynitrohydrothymin.

(135° to 136°) (178° to 181°)
1-Methyl-5-bromuracil. 3-Methyl-5-bromuracil.

(228° to 229°) (255° to 260°)

1-Methylthymin and 3-methylthymin reacted in a similar
manner with fuming nitric acid as thymin giving characteristic
oxynitrohydrothymins,! XXVIII and XXIX.

CH,N———-CO NH—CO
| | Hs | Hs
CO C “ and CONS &
| \no, | \vo,
NH— CHOH CH;N CHOH
XXVIII XXIX

* Johnson: Amer. Chem. Journ., xl; This Journal, iv, p. 407.

54 Researches on Pyrimidins

Conductivity measurements on thymin and its methyl deriva-
tives disclosed the interesting facts that thymin and 1,3-dimeth-
ylthymin gave practically constant conductivities at 25°. On the
other hand, 1-methylthymin and 3-methylthymin gave abnormal
conductivities which increased with the length of time these
pyrimidins were kept in solution (see Appendix). This inter-
esting behavior is possibly due to a slow hydrolysis of the
pyrimidin ring giving, in solution, (-uraminoacrylic acids,
XXVI and XXVII.

I 2> Sj dees 6
N(CH,).CO.NH.CH:C(CH,).CO —> CH,NH.CO.NH.CH:C(CH,).COOH.
|

XXVI

in eo 40 455 2 40
NH.CO.N(CH;)CH:CH.CO -—» -H,N.CO.N(CH,).CH:C(CH,).COOH
| |

XXVII

The methyl derivatives of thymin, described in this paper,
should be of interest to the pharmacologist. It is a well known
fact that the methylated dioxypurins possess a pronounced
diuretic action. Sweet and Levene! have recently shown that
the administration of thymin to a dog also caused a most pro-
nounced diuresis. Whether the methylated thymin will possess
a higher diuretic action than thymin must be decided by further
study.

EXPERIMENTAL PART.

2-Ethylmercapto-1, 5-dimethyl-6-oxypyrimidin:
CH_N==Co

Five grams of 2-ethylmercapto-5-methyl-6-oxypyrimidin and
a molecular proportion of potassium hydroxide (1.6 grams)
were dissolved in boiling 95 per cent alcohol. An excess of

1 Journ. of Exper. Med., ix, p. 229.

Treat B. Johnson and Samuel H. Clapp 55

methyliodide was then added and the solution boiled for about
one hour when it no longer reacted alkaline to turmeric. The
undissolved potassium iodide was filtered off and the filtrate
heated on the steam-bath to remove the excess of alcohol. We
obtained an oily residue, which deposited a mixture of 2-ethyl-
mercapto-1,5-dimethyl-6-oxypyrimidin and unaltered 2-ethyl-
mercapto-5-methyl-6-oxypyrimidin, when triturated with cold
water. The filtrate contained the isomeric 2-ethylmercapto-
3,5-dimethyl-6-oxypyrimidin (see below). 2-Ethylmercapto-
1,5-dimethyl-6-oxypyrimidin was freed from the unaltered
material by treatment with a small volume of a cold, dilute
solution of sodium hydroxide. The weight of the crude pyri-
midin was 2.1 grams or 30 per cent of the theoretical. It was
purified for analysis by crystallization from hot water, and
separated, on slow cooling, in long, slender prisms which melted
at 65° to a clear oil without effervescence. It did not contain
water of crystallization. It was dried for analysis over sulphuric
acid (Kjeldahl):
Calculated for
CgH,2ON,S: Found:
LY oy ws 6 SOnt a CRE OVE CLARE net ae ane 15.22 Loy,

2-Ethylmercapto-3,5-dimethyl-6-oxypyrimidin:
N—CO

||
tists CCH.

In order to isolate this pyrimidin from the above filtrate, the
solution was evaporated to dryness and the residue extracted
several times with cold chloroform. When the chloroform
was evaporated, at ordinary temperature, 1.8 gram of the
crude pyrimidin were obtained, or 33.4 per cent of the theo-
retical. The compound crystallized from benzene in prisms
which melted at 156° to a clear oil without effervescence. It
was dried for analysis over sulphuric acid (Kjeldahl):

Calculated for
CgH;2ONo8: Found:

IN] occ. ate sl aoe 0 oC Coo 15.22 15.20

56 Researches on Pyrimidins

1,5-Dimethyl-2,6-dioxypyrimidin (1-methylthymin):

CH N= =Cco
| |
CO CCH,
| |
NH—CH

A quantitative yield of this pyrimidin was obtained when
2-ethylmercapto-1,5-dimethyl-6-oxypyrimidin was digested with
hydrobromic acid until the evolution of ethylmercaptan ceased
(8 hours). The acid solution was then evaporated to dryness
and the pyrimidin crystallized from water. It separated in
aggregates of stout prisms which melted at 202° to 205°, with
effervescence, to a clear oil. The compound was readily soluble
in boiling alcohol and acetone. It did not contain water of
crystallization. Analysis:

0.2680 gram of substance gave 0.5081 gram of CO, and0.1379 gram H,O.

Nitrogen determination (Kjeldahl):

Calculated for Found:
6HsQ02No: is DES 18 be
rcs ae Glass vacouenee eed avers es 51.43 Ol
aye snd @ eek ee eee 5.71 5.71
Nite co et ee 20.00 20.07 S20007,

3, 5-Dimethyl-2,6-dioxypyrimidin (3-methylthymin) :

NH—CcoO

COT ACGH:

Badr
CH,N——-CH

This pyrimidin was not the only product formed when
2-ethylmercapto-3,5-dimethyl-6-oxypyrimidin was digested with
hydrobromic acid for 5 hours. When the acid solution was
evaporated to dryness, we obtained a mixture of 3-methyl-
thymin and a compound which was difficultly soluble in cold
water. The latter was soluble in warm alcohol and crystallized
from hot water in needles melting at 229° to 230°, without
effervescence to a clear oil, It was soluble in alkalies and gave
a strong test for sulphur. The analytical determinations indi-

Treat B. Johnson and Samuel H. Clapp 57

cated that the compound was 2-thio-3,5-dimethyl-6-oxypyrimidin.
Analysis (Kjeldahl) :

NEC)
Gor nCCE:
aca
CH,N CH
Calculated for Found:
CsHsON.S: ile ie
Dl, 6 0 De CtIS eee eee 17.95 18.4 18.3

3-Methylthymin is more soluble in cold water than this
2-thiopyrimidin and separated, nearly pure, when the aqueous
filtrates (above) were concentrated and cooled. It was purified
for analysis by recrystallization from water and melted, when
heated slowly, at 280° to 282° to a clear oil. This pyrimidin
showed a very characteristic behavior when crystallized from
hot water. When the hot, saturated, aqueous solution was
quickly cooled, the pyrimidin separated immediately as a bulky,
homogeneous mass of long, prismatic needles. On standing,
these prisms soon disintegrated, apparently redissolved, and
were replaced by characteristic octahedral prisms. The trans-
formation was complete in a few minutes. The compound
did not contain water of crystallization. Analysis (Kjeldahl):

Calculated for

Cy.Hs,O2No: Found:
IO oo 6.6.6 6.5 CENChG CRORE NCIC EnEnC REN aeeaeeae 20.00 19.91
1-Methyl-5-brom-4-oxyhydrothymin:
CH “Co
Be aahezer,
CO Cc
| | Nps
NH— CHOH

This compound was obtained when 1-methylthymin was dis-
solved in an excess of bromine water and the solution allowed
to evaporate in a vacuum over sulphuric acid. It crystallized
from bromine water in stout prisms and melted, on slow heating,

58 Researches on Pyrimidins

at about 123° to 125° to a clear oil. It was dried for analysis
over sulphuric acid (Kjeldahl):

Caleulated for

CgH gO3Ne2Br: Found:
lh peer eer ee eS A Got ee Le Si 92
1-Methyl-5-nitro-4-oxyhydrothymin:
CH.N— CO
(aaeren:
CO ce
=. =| Sno
NH— CHOH

This compound was prepared by dissolving 1-methylthymin
in a small volume of fuming nitric acid (sp. gr. 1.5) and allowing
the solution to evaporate spontaneously in the air. It deposited
in well-developed prisms melting at about 135° to 136° with
effervescence. Analysis:

I. 0.0844 gram substance gave 15.5 cc. moist N, at 21° and 728 mm.
II. Nitrogen determination (Kjeldahl):

Calculated for Found:
CgH,O;N3: Lh, a.
INGE eek: eee oe 20.69 20584 21a"

3-Methyl-5-nitro-4-oxyhydrothymin:

NH— CO
[oven
CO ch .H,O

| \NO,

CH,N CHOH

This pyrimidin was prepared in the same manner as its isomer.
It separated in large prisms which decomposed at about 178°
to 181° with effervescence. This decomposition point varies
according to the rate of heating. A nitrogen determination
indicated that the compound contained one molecule of water of
crystallization. It slowly underwent decomposition when heated
at 100°. Analysis (Kjeldahl):

Caleulated for Calculated for
CsH,O;N3.H20: CgHoO;N3: Found:
INE coved Pe Sat cote 19.00 20.69 18.74

Treat B. Johnson and Samuel H. Clapp 59

Mono-potassium salt of thymin:

NCO
|
CO CCH,
| |
NH—CH

Finely pulverized thymin and a molecular proportion of
potassium hydroxide were dissolved in boiling absolute alcohol
and the solution boiled for four hours. The potassium salt
separated, on cooling, in long needles. It was purified for analy-
sis by recrystallization from 95 per cent alcohol. Nitrogen
determinations in the salt, dried to a constant weight at 110°,
agreed with the calculated value in a mono-potassium salt of
the closed ring (Kjeldahl):

Calculated for Found:
C;H;02N2K: i I.
IN| 2 .c.9 S¢e oe REENEI Cee eee ae 17.07 17.22 17.32

This pyrimidin was first described by Steudel.t’ We obtained
the same derivative, in a smooth manner, under the following
conditions: Five grams of thymin and 4.6 grams of potassium
hydroxide were dissolved in go cc. of 95 per cent alcohol and
an excess of methyl iodide added to the warm solution. The
solution was boiled for twenty minutes, then evaporated to
dryness, and the crystalline residue extracted several times
with cold chloroform. When the chloroform was evaporated
we obtained 2.7 grams of the dimethyl pyrimidin. It was
very soluble in water and chloroform and difficultly soluble in
ether and petroleum ether. It crystallized from alcohol in

eaGs Ct.

60 Researches on Pyrimidins

long needles melting sharply at 153° to a clear oil. Analysis
(Kjeldahl) :

Calculated for
C7HiyO2N2: Found:

Nisais g eiade ee oc. se) vd one ee ee eee 18.18 18.32

1,3-Dimethyl-5-brom-4-oxyhydrothymin:

CH,N—CHOH

Was prepared by dissolving 1, 3-dimethylthymin in bromine
water. When the aqueous solution was concentrated in a
vacuum, the pyrimidin separated in prisms melting at 132° to
133° to a clear oil. Analysis (Kjeldahl):

Calculated for

C;H,,03NeoBr: Found:
INI ee Beet heat gs Seg ce cle ee aie ee ein EG 11.25
Mono-potassium salt of uracil:
KN—-—CO
CO “CH.H,O
| !
NH— CH

One molecular proportion of potassium hydroxide was dissolved
in 450 cc. of absolute alcohol and 13 grams of finely pulverized
uracil suspended in the solution. After digesting for eight hours
the uracil was completely changed into the potassium salt.
It was difficultly soluble in absolute alcohol and very soluble
in cold water. It separated from dilute alcohol in balls of long
needles. The yield was quantitative. The salt contained one
molecule of water of crystallization which was determined by
heating at 120°.

I. 0.3365 gram substance lost 0.0351 gram H,O. 7
Tbe 0) OFSAl “ “ “ 9.0302 “ “ ‘
Calculated for Found:
CyHzOoNoK. H.O : .
1 let © ieee ict, CR ec an : 10.70 10.438 10.52

ee a ee

Treat B. Johnson and Samuel H. Clapp 61

Nitrogen determination in salt dried at 160° (Kjeldahl):

Calculated for Found:
C4H202NoK: . if.
le oot CORRE De EO eee 18.67 18251) 18255

Potassium determination in the hydrous salt:

0.1727 gram substance gave 0.0775 gram KCI.

Calculated for Caleulated for
C4,H302.N2K.H20: C4H202.NoK : Found:
on 2 6b Ae 23.21 25.98 23.55

Nitrogen determination in hydrous salt (Kjeldahl):

Calculated for Calculated for

C4HO2N2K.H20: C4H3O.N2K: Found:
Ih) as oes een 16.67 18.65 16.81
1,3-Dimethyluractl:
CH,N—CO
CO Me
CH,N— ue

This compound was prepared by warming, in alcoholic solution,
molecular proportions of potassium hydroxide and the potas-
sium salt of uracil with an excess of methyliodide. After
boiling for three hours the solution was evaporated to dryness
and the pyrimidin extracted with chloroform. It crystallized
from a mixture of alcohol and ether in long, slender prisms which
melted at 121° to 122°. It was extremely soluble in cold water,
alcohol and chloroform, but insoluble in ether and petroleum
ether. Nitrogen (Kjeldahl):

Calculated for Found:
CgHsO2No: i: LE INA
Wo. dict aide ee 20.00 20.05 20.06 20.24

1,3-Dimethyl-dibromoxyhydrouractl:
CH, N—CO

CO CEr,

CH,.N—CHOH

62 Researches on Pyrimidins

This pyrimidin crystallized from bromine water in micro-
scopic prisms with curved outline. It melted at 135° to 136°
to a clear oil. Analysis (Kjeldahl):

Calculated for
CgHsO3NoBro: Found:

IN 5 aw: Shaina ctave ches oslo, Spek ROR ene eeeeaon 8.86 9.0

1, 3-Dimethyl-5-bromuracil:

CH,N—CO
CO CBr
eae

CH,N— CH

This compound was prepared by digesting the above hydro-
uracil derivative with absolute alcohol. It was purified for
analysis by recrystallization from water and melted at 181° to
182° toaclear oil. Analysis (Kjeldahl):

Calculated for
CgH7O2NoBr: Found:

IN She Spe ts ee eae eee eee ane ee ee 12.79 13.0

2-Oxy-3-methyl-6-aminopyrimidin (3-methylcytosin) :

N=C.NH,
CO CH
|

CH,.N—CH

Six and eight-tenths grams of anhydrous cytosin and 3.4
grams of potassium hydroxide were dissolved in 60 ce. of boil-
ing, absolute alcohol, and 16 grams of methyl iodide added to
the solution. After boiling for two hours the solution was
evaporated to dryness and the residue of pyrimidin and potas-
sium iodide dissolved in cold water. The iodine was removed
with silver sulphate and the excess of silver by precipitation with
hydrogen sulphide. The sulphuric acid was then quantitatively
removed with barium hydroxide and the filtrate from barium
sulphate concentrated to a small volume. The base was pre-
cipitated from this solution with a hot, saturated solution of
mercuric chloride and the mercury precipitate decomposed in

ee

Treat B. Johnson and Samuel H. Clapp = 63

the usual way with hydrogen sulphide; the chlorine removed
with silver sulphate, the excess of silver with hydrogen sulphide,
and the sulphuric acid with barium hydroxide. When this
solution was evaporated to dryness we obtained about 1.1 gram
of the pyrimidin associated with a small amount of unaltered
cytosin. The small yield is partly explained by the fact that the
base volatilizes with aqueous vapors. It was purified by recrys-
tallization from methylalcohol. It separated in beautiful,
distinct prisms which decomposed at about 278° to 279° to a
dark oil. This decomposition point varies according to the
rate of heating. The base was extremely soluble in water and
did not contain water of crystallization. Analysis (Kjeldahl):

Calculated for Found:
C;H;ON3: Il.
gk 5 OO Oe 33.60 SBI) Sine)

The chloroplatinate of 3-methylcytosin:

Was prepared by adding a solution of platinum chloride to a
hydrochloric acid solution of the base. It crystallized from
water in long, slender prisms. The salt contained two molecules
of water of crystallization which was determined by heating at
p20".

I. 0.0513 gram salt lost 0.0028 gram H,O.

WO. 2103 “ oe OF OME as
it 0.0841 . “ agen O.OO45r x 4
Calculated for Found:
(C;H7ON3.HCl)o.Pt Cl4.2H2O: i id. Ill.
UN oa 5.18 5.45 5.28 5.35

Platinum determination in anhydrous salt:

I. 0.0478 gram salt gave 0.0142 gram Pt.

HMeO-1991- « eels 020591; °S is
LE. 0 : 0793 “ “ “ 0 s 232 “ “
Calculated for Found:
(C5H;ON3.HCl)2.PtCls: i Lk: TLE.
i, eh Ae ee 29.54 29.71 29.68 29.26

Picrate of 3-methylcytosin:
This salt was very insoluble in cold water, and crystallized
from hot water in long prisms. The decomposition point varies

64 Researches on Pyrimidins

greatly according to the rate of heating. When heated slowly
it decomposed at about 280° with effervescence. Analysis:

0.1351 gram substance gave 28.7 cc. N, at 25° and 770 mm.

Calculated for
C;H;ON3.CgH30;N3: Found:

ING. sig Soke hs Se hohe eco ene 23.73 24.0

3-Methyl-5-bromuracil:

NH— CO
CO CBr
| |
i
CH,N===cH

When 3-methylcytosin was dissolved in a little cold water and
liquid bromine added to the solution, a heavy precipitate was
obtained. This dissolved immediately, on warming, and after
heating about ten minutes on the steambath, the solution was
evaporated to dryness under diminished pressure. The residue
obtained was then dissolved in a small amount of absolute
alcohol and the solution boiled for five hours. On cooling, the
brompyrimidin separated in needles. It crystallized from hot
water in long, slender, distorted needles. They decomposed
at about 255° to 260° to a clear oil with practically no effer-
vescence. The compound was soluble in cold, dilute sodium
hydroxide solution and was precipitated again by acetic acid.
It gave a strong test for bromine. A mixture of the pyrimidin
and the isomeric 1-methyl-5-bromuracil’ melted at 175° to
195°. A mixture of the pyrimidin and 5-bromuracil melted
from 230° to 270°. It did not contain water of crystallization.
Analysis (Kjeldahl):

Calculated for Found:
C;H;O2NoBr: i
ING ndas fas cee ee eee 13.66 13.53 J13.08

2-Oxy-6-methylphenylaminopyrimidin:
N——C.N(CH,) (C,H)

1 Johnson and Heyl: Loc. cit.

Treat B. Johnson and Samuel H. Clapp 65

Ten grams of 2-ethylmercapto-6-chlorpyrimidin and two
molecular proportions of monomethylaniline (twelve grams)
were dissolved in dry benzene and the solution boiled for eight
hours. The benzene was then evaporated, the residue dissolved
in ammonia and the excess of monomethylaniline removed by
distillation with steam. When this solution was concentrated
we obtained the mercaptopyrimidin as an oil which did not
solidify on standing. The crude mercapto derivative was con-
verted into the oxygen derivative by digesting with hydrobromic
acid. The yield was 80 per cent of the theoretical. The com-
pound is very insoluble in water and chloroform and crystallizes
from alcohol in beautiful hexagonal tables which do not decom-
pose below 285°. Analysis (Kjeldahl):

Calculated for
1 Hy,ON3:: Found:

2-Oxy-3-methyl-6-methylphenylaminopyrimidin:
N—CN(CH,) (C,H,)

CO) CH
see

Seven and five-tenths grams of 2-oxy-6-methylphenylamino-
pyrimidin, 2.1 grams of potassium hydroxide, and 12 grams
of methyliodide were dissolved in 150 cc. of absolute alcohol
and the solution boiled for two hours. The solution was
then evaporated to dryness and the residue extracted with
chloroform. When the chloroform was evaporated we obtained
the hydriodide of the pyrimidin base. The base was obtained
by decomposing the salt with sodium hydroxide and weighed
5-5 grams or 69 per cent of theoretical. It crystallized from
water in long, striated prisms which melted sharply at 186° to

187° to a clear oil without effervescence. Analysis (Kjeldahl):
Calculated for

Cy2Hy30N3: Found:
ie SS 19.53 19.78
2-Oxy-3,5-dimethyl-6-aminopyrimidin (3,5-Dimethyl-cytosin) :
N=C.NH,
|
CO. CCH,

66 Researches on Pyrimidins

5-Methylcytosin! was converted into this base by alkylation
with potassium hydroxide and methyliodide in the usual manner,
and the new pyrimidin isolated in the same way as 3-methyl-
cytosin. The compound was extremely soluble in water and
volatilized with aqueous vapors. It separated from methyl-
alcohol in prisms. It had no definite melting point, but decom-
posed from 300° to 310°, according to the rate of heating, with
effervescence. Analysis (Kjeldahl) :

Calculated for Found;
CgH ON3: Ve Il.
INES aaron ete eee 30.22 30.09 29.76

Conversion of 3,5-dimethylcytosin into 3-methylthymin:

Some 3,5-dimethylcytosin was dissolved in strong bromine
water, the solution warmed on the steambath for ten minutes,
and then evaporated to dryness in a vacuum. We obtained
a crystalline deposit which was digested with absolute alcohol
for five hours. The alcohol was then removed by evaporation
and the residue redissolved in hot water. Needle-like prisms
separated, on cooling, which melted at 160° with effervescence
to a colorless oil. This oil immediately solidified on cooling
and then melted at 280° to 282° to an oil without effervescence.
They did not contain bromine and were soluble in alkalies. We
did not obtain enough of this material for analysis but it was
probably $-methyluramino-a-methylacrylic acid:

NH, COOH

When the filtrate (above) was concentrated and cooled the
characteristic crystals of 3-methylthymin separated. They melted
at 280° to 282° to an oil. A mixture of pure 3-methylthymin
and this compound melted at the same temperature. Analysis
(Kjeldahl):

Calculated for
CgHsO2No: Found;

Nig ssotbueuss he'd ayaa els Soe ee 20.00 20.3

1 Wheeler and Johnson: Amer. Chem. Journ., Xxxi, p. 591.

Treat B. Johnson and Samuel H. Clapp 67

CONDUCTIVITY MEASUREMENTS ON THYMIN, 1-METHYL-
THYMIN, 3-METHYLTHYMIN, 1,3-DIMETHYLTHYMIN AND
4-METHYLURACIL.

BY N. A. MARTIN.

In a series of conductivity measurements on thymin and its
nitrogen-methyl derivatives it was found that the conductivities
of thymin and 1,3-dimethylthymin remained nearly constant.
The very slight increase in conductivity observed was probably
due to small amounts of impurities absorbed from the glass and
air. The conductivity of water was found to increase at about
the same rate.

On the other hand, duplicate determinations on 1-methyl-
thymin and 3-methylthymin did not give agreeing results nor
could dissociation constants be calculated. The investigation
showed that the conductivity of these two pyrimidins rose with
the length of time they were kept in solution. The greatest
rise in conductivity was observed in the determinations on 3-
methylthymin.

The measurements were carried out by the customary Kohl-
rausch method in a thermostat at 25°. The water used was
redistilled over barium hydroxide, rejecting all that gave a test
for ammonia with Nessler’s reagent, until it had a specific con-
ductivity of approximately 2 xX 10 °, and the conductivity
vessel, pipettes, etc., were thoroughly steamed before using.

The rise in conductivity was determined by preparing 20 cc.
of exactly ~; solution of the substance directly in the vessel
and measuring its conductivity, at intervals, until it remained
practically constant. The rise in specific conductivity of pure
water, due to absorption of air, solution of soda from the walls
of the vessel, etc., was also determined and found to be prac-
tically zero (0.0,2 per hour).

In calculating results the amount of hydrolysis could not be
determined as the amount of dissociation of the resulting ura-
minoacrylic acids (?) was unknown. The specific and molec-
ular conductivities were calculated and the latter plotted as
ordinates with the time in hours as abscissas. As a reference
line the rise in conductivity of water was also plotted, using 64,000
times the specific conductivity to agree with the molecular
conductivity of the solutions (A 4, = 64,000 x).

68

Molecular conductivity of the pyrimidins i

Researches on Pyrimidins

n %_ solution at 25°

64

due to hydrolysis.

, showing increase

Time.

Thymin.

(+2) 0.

OF
To)

1,3-Dimethly-
thymin.

0.23

0.37

.68
.69
4
74
75

ooo o& ©

(+4)

Or Or Gr

i-Methylthymin. |

.49
04
Ue
36
4

.98

.02

3-Methyl- | 4-Methyl-

thymin. |

0.38
0.90

|
|
|
|
}
|

8.90
10.19

13.53

uracil.

0.35
0.75

MPs

Treat B. Johnson and Samuel H. Clapp

69

Specific conductivity of X. pyrimidin solutions at 25°, showing increase due
to hydrolysis.

Time.

6

0.0,21
0.0,22 |

0.0,24 |
0.0,26 |

0.0.55

Thymin.

(+4) 0.0,65
0.0,66
0,67
0,69
10.7
.0;72 +
0,74

SS ere &

(+4) 0.0,89

0.0,91
0.0,92
0.0,93

1,3-Di-
methyl-
thymin.

0.0,36

0.057

0.0,107
0.0,109

0.0,113
0.04115
0.0,116

1-Methyl-
thymin.

0377
0,162
0,182
0,212
0,240

0,308

3-Methyl-

thymin.

0.0,60
0.0,14

4-Methyl-
uracil.

0.0,55
0.0,12

7O

Researches on Pyrimidins

No
IS
/7 (wh Water

(ll Thy lola
/3 (212 Met byt Thy mia
eee Y- [Let hyl-lraes/

(7) L-lferty| Thyrae
yy AD Sieh Thyiae

Ne SRS Se ST Gh Nh ey SOE CS

Sp

S /0 /[§ 20 2F 30 35 Yo ¥F FO FF fours

SOME NOTES ON THE EFFICIENCY OF THE FOLIN METHOD
FOR THE QUANTITATIVE DETERMINATION OF URI-
NARY AMMONIA.

By MATTHEW STEEL anv WILLIAM J. GIES.

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

(Received for publication, June 1, 1908).

I. INTRODUCTION.

Last fall, during the progress of the metabolism research that
is described in the paper immediately following this, certain
anomalous results were obtained in our quantitative determina-
tions of the urinary ammonia. In the earlier periods of that
research ammonia had been determined by the Folin method,
in the urines in duplicate for 38 days, with thoroughly con-
cordant results. Shortly after the beginning of a metabolism
period, however, during which magnesium sulfate was injected
subcutaneously’ every twenty-four hours, the titrations, in dupli-
cate (at the conclusion of the Folin process as applied to the
daily urines), were strikingly discordant, the disagreements
amounting to from 1 to 2 cc. of # potassium hydroxid per 25 cc. of
urine.

Our inability to obtain satisfactory duplicate results for urinary
ammonia content after the magnesium sulfate treatment, or to
explain the analytic discrepancies by any probable fault of tech-
nique, led us to make two general suppositions regarding the
cause of the analytic disagreements observed :

(1) That magnesium was eliminated into the urines in ques-
tion in relatively large quantities as ammonio-magnesium phos-
phate, which separated, in part at least, in typically crystalline
masses.

1 Period iv, p. 95 (this volume).
71

72 The Quantitative Determination of Ammonia

(2) That the crystalline ammonio-magnesium phosphate thus
deposited was not thoroughly decomposed by sodium carbonate,
as used in the Folin process, whereby ammonia, in variable
amounts, remained in its solid form as triple phosphate in the
urines under investigation.

General examination of the urines that gave the anomalous
quantitative results for ammonia content showed at a glance
that our first supposition was correct—triple phosphate had
crystallized in abundance. In separating portions of the urines
for analysis, care had always been taken to isolate fractions of
the thoroughly shaken and evenly mixed daily samples. Conse-
quently, we had no reason to believe that any of the above men-
tioned anomalous results of the ammonia determinations were
due to transferal of unequal amounts of the deposited ammonio-
magnesium phosphate in the duplicate fractions of the urine
taken. We therefore proceeded to test very carefully, and in
many trials, the validity of the second supposition stated above.

II. ANALYTIC OPERATIONS.

Furst Series. Is the amount of sodium carbonate (1 to 2 grams)
that is usually taken with 25 cc. of urine in the Folin process
sufficient to completely liberate the ammonia from small quan-
tities of crystalline ammonio-magnesium phosphate?

Method A. We endeavored to answer this question directly by the fol-
lowing special adaptation of the Folin process: Portions of pure, crystal-
line ammonio-magnesium phosphate, in different amounts between 50
and 500 milligrams inclusive, were quickly and very accurately weighed
on a watch glass and transferred quantitatively to aérometer cylinders of
the usual size, through a small dry funnel from which the tube had been
removed. All fragments adherent to the watch glass and funnel were
brushed into the cylinders. No losses of substance could have occurred
in the process. In all the tests the crystalline matter was a comparatively
coarse powder. About 25 to socc. of water were poured into the cylinders
on the powder, which quickly formed a loose sediment in the undisturbed
water. A layer of kerosene was then poured over the liquid in each cylin-
der merely to duplicate the conditions of the Folin process, although no
special frothing could have occurred to require its use. Solid sodium
carbonate in definite quantities, ranging between 1 and 4 grams inclusive.
was then added to the phosphate-water-kerosene mixtures in the cylinders.
The apparatus recommended by Folin was employed for aération. More

Matthew Steel and William J. Gies 73

powerful pumps than those recommended by Folin were kept in operation
for from five to fifteen hours, so that aération was unusually effective. In
all cases aération was continued at least five hours. In the groups desig-
nated B and C (Table I), aération was conducted during a second five
hour period, or ten hoursinall. The aérometer cylinders were not opened
between the two periods, but the acid of the first period of absorption was
titrated and a new portion of acid substituted for ammonia absorption
during the second aération period. In the sixth determination (Group
B) aération was carried in the same manner through a third period of five
hours, or fifteen hours in all. Approximately fifth-normal sulfuric acid
was used for the absorption of the ammonia. Congo red was used as the
indicator. Our results in the first series of tests are given in Table I.

TABLE I. FIRST SERIES, GROUPS A TO C.

Pure, crystalline ammonio-magnesium phosphate, 0.05 to 0.5 gram. Sodium
carbonate, 1 to 4 grams. Total periods of aération, 5 to 15 hours. Loss
of ammonia: maximum, 48.1 per cent; minimum, 31.1 per cent.

= %, VOLUME OF STANDARD ACID SOLUTION +
= So | Ss ia REQUIRED TO NEUTRALIZE— -
Group. 4 eC, 2° £ : 5
5 ot 2 a AFTER AERATION. If all NH; Ss
2 Peer lr mil | ae ee Es
my 5hrs. |5hrs.(2) 5hrs.(3) Total. “|
no. gram 9 rams ce. ce. ce. Cla ce. per ct.
Bees... Hen! 0.05: 2" 0:67 | | 0.6 1.05 | 42.9
Sie Pe eit2., || ae 2.05 | 41.5
Sra Ore 1 Dele a 2.4 4.10 | 41.5
peers |) til 4.2] | | 4.2 6.15 | 31.7
ee ,, 5 | 0.4 ee 4-1 10.55 4.65 8.25 | 43.6
ee 015 a) 4-6" | 1.20 | 1.20) 7.10), 10;s0 nae
C.. meres) oi | 4.3 | 1.05 | 5.35.1 10:30) | 43.1
ae 2 esol. | 0.80 | 5.90 | 10.30 | 42.7
10.5 | 4 5.6 | 0.65 | 6.25 | 10.30 | 39.3

* NH4MgPO,6H.O. A pure crystalline powder obtained from Eimer and Amend. The
theoretical content of nitrogen is 5.707 per cent. The average result of seven closely con-
cordant determinations of nitrogen contents by the Kjeldahl method was 5.727 per cent
(5.697, 5.766, 5.780, 5.669, 5.725, 5.725, 5.725).

tSeven determinations of the volume of our standard acid solution that was required to
neutralize the ammonia liberated in the Kjeldahl process from 0.5 gram samples of the
triple phosphate used in this series gave the following results (ce.): 10.25, 10.40, 10.40,
10.20, 10.30, 10.30, 10.30, or an average of 10.30 cc. (See footnote in the summary on
page 76).

The first result in Table I shows that one gram of sodium car-
bonate was unable, after five hours of very strong aération,
completely to eject the ammonia from 50 milligrams of the triple

74 The Quantitative Determination of Ammonia

phosphate. The results for Groups B and C show that, after ten
hours’ aération of 0.4 or o.§ gram samples of triple phosphate
with 1 to 4 grams of sodium carbonate, large proportions of
ammonium were undisturbed in the crystalline material. Even
after fifteen hours’ aération of a 0.5 gram sample of the phos-
phate with four grams of sodium carbonate, practically one-third
of the ammonium remained undisplaced.

These observations gave strong support to our second conclu-
sion regarding the cause of the anomalous ammonia results that
prompted this study.

Second Series. In the second series of tests, we endeavored
to ascertain the effect of previous solution of the triple phosphate
on the analytic outcome.

Method B. In this process Method A was employed, except that the
weighed portion of triple phosphate (0.5 gram) was dissolved in a beaker in
a small amount (2 to 4 cc.) of acetic acid (25 per cent) or hydrochloric acid
(10 per cent). The acid solution was diluted with about 25 cc. of water
and transferred quantitatively to the aérometer cylinder, where it was
treated with sufficient sodium carbonate to neutralize it and to afford an
excess of 2 to 4 grams of the carbonate. From this point forward Method
A was followed to completion. The first two periods of aération were five
hours each or a total of ten hours. In the last four tests of the series
(Group E) a third aération period of fourteen hours was maintained—a
total of twenty-four hours of aération for the individuals of Group E.
Results are summarized in Table II.

The data in Table II show clearly that preliminary solution
of the triple phosphate did not particularly favor ultimate dis-
placement of the ammonium but that, even in the presence of 4
grams of sodium carbonate and after very strong aération for
twenty-four hours (17), a large proportion of the ammonium
remained fixed in the original phosphate.

These observations suggested that the Folin method may be as
inadequate for the complete removal of ammonia from small
quantities of dissolved ammonio-magnesium phosphate as the
data of Table I indicated it is for the isolation of ammonia from
the crystalline form of that substance.

Third Series. The outcome of the first two series of tests led
us to check the previous results with additional preparations of
triple phosphate and also to ascertain the effects of larger pro-
portions of sodium carbonate.

| uroup.

Matthew Steel and William J. Gies 75

TABLE II. SECOND SERIES, GROUPS D AND E.

Pure, crystalline ammonio-magnesium phosphate (from the supply used in
the first series), 0.5 gram. Acid: acetic (25 per cent), 3 to 4 cc. or hydro-
chloric (10 per cent), 2.cc. Sodium carbonate, 2 to 4 grams. Total periods
of aération, 10 to 24 hours. Loss of ammonia: maximum 52.4 per cent;
minimum, 16.5 per cent.

AMMONIU-MAGNESI™M VOLUME OF STANDARD ACID SOLUTION

: PHOSPHATE. S REQUIRED TO NEUTRALIZE—

g a rey
3 soe zZ ____APTER ABRATION. i. 3
& : DISSOLVE IT. iS a a | mee ees

g < Sed ae ee = . | 2a8 é
2 a! me om Z Zz 2 3 | 333
8 s Kind. |Volume. S = a < 3 | S35 a
no. | gram per cent. ce. grams ce. cc. cc. cc. | ce. | perct.
10 | 0.5 |25 acetic 3 3 5.55 | 1.00 6-55 |) 10334) S6e4
11 OS 25 a 4 2 6.2 0.70 6.905 lOxoeieooet
12 OFouizo. 4 4 3.6 1.30 4.90 | 10.3 | 52.4
13 Ova 20 HCl Pe 2 6.9 105 6295 TAOESR P2278
mo) O.5 10° « 2 O57 | S00 7701023" 25 2
15f| 0.5 |25 acetic 4 4 6 Jo 0:90} 05407 |) 7 260") Oss 262
Giese \25 “ 4 2 4.4 0.85 | 0.90 | 6.15 | 10.3 | 40.3
FO; 5 10 HCl 2 4 6.55 | 1.60 | 0.45) | 82604}. 1Os3eelGeS
18t| 0.5 Orn 2 2 500%) 20) | Os60) 7-304) Oss e2SeL

\

*See the second footnote, Table I.
t+ These determinations were started by the junior author and completed independently
by the senior author, who knew nothing of the particular conditions of the tests.

Four preparations were employed for the purpose. The product used
in the first two series, purchased from Eimer and Amend, was again
selected. A second product was obtained from the Mallinckrodt Chemical
Company, a third was a Merck preparation and a fourth was our own
make. The Mallinckrodt product was a somewhat finer powder than the
others.

We prepared our own product by allowing very dilute ammonium
hydroxid to drop slowly from a burette into a crystallization dish con-
taining a fairly concentrated solution of magnesium sulfate and disodium
hydrogen phosphate. Deposition occurred gradually and the triple phos-
phate crystals were well formed and relatively large—much larger than
those in the powders obtained from the sources indicated above. The
crystalline precipitate was washed with water until it was free from am-
monium hydroxid. It was then dried in the open air at room tempera-
ture.

The following results were obtained in determinations, by the Kjeldahl
method, of the nitrogen contents of our several ammonio-magnesium
products:

76 The Quantitative Determination of Ammonia

VOLUME OF STANDARD ACID SOLUTION REQUIRED TO P ;
NEUTRALIZE THE LIBERATED AMMONIA FROM 0.5 ercentage 0}

SRODUGE. GRAM OF NH4MgPO4 nitrogen.
1 | 2 3 | AS oN onl) 16 if Avg. Found Theory
ec: ce. ce. | ce. | CComnluntCc: ce. | cee Avg. |
Eimer and Amend* | 10. 25110. 40/10.40'10.20 10.3010.3010.3010.30 5.72715.70%
Mallinckrodt ......| 13. 50/13 .50) | 13.50|7.503|5.707
IMerckt vee ee 110.05 /10.10|5.614)/5.707
Ourtown $2.3 10. 00) 9.95, 10.00|5.522|5.707

* The actual titration results were: 0.1 gram, 2.05 ec.; 0.2 gram, 4.15 ec.; 0.5 gram (four
samples), 10.4, 10.2, 10.3, and 10.3 cc.; 1 gram, 20.6 ce.

With the exception of the Mallinckrodt product, the preparations were
quite pure. The nature of the conspicuous impurity in the Mallinckrodt
product was not determined. It had no important influence on the tests,
however.

In this series of tests, on the four products referred to in the above
summary, Method A was employed. Two successive five-hour periods
of aération were conducted in all the tests except one, in which the sample
was subjected to uninterrupted aération for eighteen hours. Results are
summarized in Table III.

Table III makes it evident that the results of the first and sec-
ond series of tests with the Eimer and Amend product were not
due to any peculiarities of that particular sample of triple phos-
phate. In each group of tests the percentage loss of ammonia
decreased somewhat with each increase in the amount of sodium
carbonate in the aérometer cylinder, but it was always above
20 per cent, even when 1o grams or more of sodium carbonate
were used for the decompositions, and aération was continued ~
ten hours.

Fourth Series. At this point it occurred to us that possibly
we were overlooking some defect in technique that could be
detected by carrying out some of the tests with caustic alkali
as the decomposing agent. Furthermore, such a check on our
former results promised also to give us additional evidence per-
taining to the total ammonia contents of our products. Method
A was employed for the tests, but sodium hydroxid was sub-
stituted for sodium carbonate. Aération was continued only
three hours. Results are given in Table IV.

TABLE III. THIRD SERIES, GROUPS F TO I.

Crystalline ammonio-magnesium phosphate (four different preparations), 0.5
gram. Sodium carbonate, 1 to 20 grams. Total period of aération, 10
hours or 18 hours. Loss of ammonia: maximum, 53.3 per cent; mini-
mum, 22.8 per cent.

| | VOLUME OF STANDARD ACID

|| NHatepo. é| 22 ae
| & OS | APTER ABRATION. | 38 %
ts E e le | 2] 812 | seer
o Q | a, ES] Ss] is | el eee
no. | gram. grms.| ce. | ce. cc. ce cc. |pr.ct
F......| 19 | Eimer and Amend) 0.5 2 |4.50)1.05/5.55
20 : 0.5 | 2 /4.20|2.10/6.30)
21 | : | 0.5 | 2 |4.80/1.10/5.90)
22 | a 0.5] 2 4.90/0.95/5.85 5.90)10.3|42.7
| 23 | “ (0.5 | 5 6.10/0.856.95
| 24 | | 0.5 | 5 |6.90/0.15/7.05)7.00/10.3)32.0
| 25 | « (0.5 | 10 6.500.507.00
| 26 | a 0.5 | 10 |7.05)0.70\7.75)
27 | ‘ 0.5 | 10 |7.750.75.8.507.75/10.3)24.8
28 | ; | 0.5 | 15 |7.10|0.80/7.90|
| 29 | eg 0.5 | 15 |7.20|0.65/7.85/7.87|10.3]23.8
—-30* : Ord.) 15. | \7.85|
—31* “ 0.5 | 15 | 7.95\7.90|10.3/23.3
Peeeenies2 | Mallinckrodt... .. 0.5 | 5 '5.05)1.20/6.25)
| 33 | « 0.5 | 5 |5.05/1.306.35/6.30|13.5/53.3
34 | “ | 0.5 | 10 |7.30|0.60/7.90)
35 : 0.5} 10 |6.60/1.508.108.0013.5)40.7
36 | “ | 0.5 | 15 |7.20|1.20/8. 40] |
a7 | Gs 0.5 | 15 /8.00/1.209.20'8.80/13.5.34.8
38 | G | 0.5 | 20 |9.15|0.60/9.75)
39 z | 0.5 | 20 |8.05]1.109.15)9.45|13.5)30.0
H Mi Merck. 2... 5... 0.5 | 1 |5.70/0.70)6.40|
ipl 4 0.5 | 1 |4.90/1.30)6.20/6.30|10. 1/37.6
42 | i 0.5 | 2 |6.30/0.45/6.75|
43 | S O35}. 2 5.90/0.95'6 .85/6.80/10.1 32.7
44 | % 0.5 5 |6.050.55 6.60!
45 | ¢ | 0.5 | 5 |5.95/1.25|7.20/6.90|10. 1/31.7
46 4 | 0.5 | 10 |6.30/0.95/7.25|
47 | e 0.5 | 10 |8.10|0.25)8.35/7.80|10.1/22.8
a Asa OMIT OWN, .-0.2- > - | 0.5 | 5 [4.20/1.05/5.25)
49 f 0.5] 5 4.35)1.40|5.75
50 | @ 10.5 | 5 5.60(0.60/6.20
igy,} 0.5 | 5 |7.05|0.15/7.206.1010.0|39.0
52 | “ 0.5 | 10 |5.05/1.15|6.20
53 | ¥ 0.5 | 10 |4.90|1.20/6.10;
54 | ‘ | 0.5 | 10 |5.60)0.55|6. 15)
55 és 0.5 | 10 6.05'0.906.95 6.35 10.0)36.5
| |

* 30 and 31. Two samples of the Eimer and Amend product of 0.5 gram each were
aérated for eighteen hours. They required on titration respectively 7.85 cc. and 7.95 cce.;
average, 7.99 cc. The ammonia loss was 23.3 per cent.

78 The Quantitative Determination of Ammonia

TABLE IV. FOURTH SERIES, GROUP J.

Crystalline ammonio-magnesium phosphate (four different preparations as in
the third series),0.5 gram. Sodium hydroxid (1 gram) instead of sodium
carbonate. Period of aération, 3 hours. Loss of ammonia: none.

§ VOLUME OF STANDARD ACID. 2
3 NH4gMgPOg,. SOLUTION REQUIRED FOR | ~
5 NEUTRALIZATION. ee
afoiies eee
2 8 5 Af lf all NH, bad}. 5
Of 2 |) ee. : 2 After aération If a ad.
o | ea EEA. Weight. iz | 3 hours. |been liberated. &
| | | =
| | |
no. gram gram ce. cc.
J...| 56 | Eimerand Amend| 0.5 1ORZ5 | 10.30 | none
57 | Mallinckrodt ....| 0.5 1 Bes) | 13.50 | 4
58-1) Merete ae 9 ae eas 0.5 1 10.10 | 10.10 ;
59 | Ourown =. ....-. 0.5 1 9.95 | 10.00 | :
| |

The data in Table IV prove conclusively that the basis for all
our calculations in the previous tests was correct, that our tech-
nique was not necessarily responsible for the variations in, nor the
extent of, the ammonia losses, and that the total ammonia con-
tents of our products were easily discharged completely in a short
time by the caustic alkali employed.

These results reinforced the conclusions drawn irom the data
obtained in the first three series of tests and seemed to show con-
clusively that sodium carbonate, in the amounts ordinarily used
in the Folin process as applied to urine, may fail to eject com-
pletely the ammonia from the contained ammonio-magnesium
phosphate.

Practically all the previously recorded data were obtained by
the junior author under the senior author’s guidance. In order
to exclude any previously undiscovered error in technique, nearly
all the remaining series of tests were conducted either by the
senior author unaided, or, as will be indicated here and there, in
part by each author; in which special cases the one worker had
no knowledge of the conditions of the tests as imposed by the —
other. In this way it was hoped to exclude entirely every influ-
ential error of manipulation or deduction. The results seem to
show that this hope was realized.

The checking process was conducted with a new standard acid

Matthew Steel and William J. Gies 79

solution, which was not quite equal in strength to the acid solu-
tion used in the first four series. The second series of results
for nitrogen content, as determined by the Kjeldahl method, are
appended:

VOLUME OF STANDARD ACID

SOLUTION REQUIRED TO
Provocts eps a
ifs 2. {Average

cc. cc. cc
2 SEG rr 10.80 | 10.70 | 10.75
LETT en a rr 13285) || 13-85) | loesD
en se Se eI 10.30 | 10.40 | 10.35
Ree a 10220) |) 102207 | 10220

* Same as those used in the third and fourth series of tests.

Fifth Series. Method A was employed in this series. The same
amounts of phosphate and carbonate were used in all the tests.
The amount of carbonate was relatively larger than in most of
the previous groups. Each set of determinations was subjected
to a single definite period of aération (5 to 15 hours). Results
are recorded in Table V.

The results summarized in Table V make it very evident that,
in harmony with the previous results, large proportions of sodium
carbonate (30 of the carbonate to 1 of the phosphate), even after
fifteen hours of very vigorous and thorough aération, failed to
effect removal of all the ammonia. It is noticeable that the loss
of ammonia after five hours of aération of 0.5 gram of triple phos-
phate with 15 grams of sodium carbonate was appreciably greater
than that in the tests which included aération for ten hours.

Sixth Series. The influence of different proportions of sodium
carbonate on o.5 gram samples of the same triple phosphate (our
own product) is shown unmistakably in the results of our sixth
series of tests (Table VI).

Method A was employed for the tests of this series, but, instead of add-
ing solid sodium carbonate to the usual amount of phosphate-water mix-
ture, the weighed quantities of the triple phosphate were transferred to
the aérometer cylinders and, after all connections had been closed, too
ec. of a solution containing the desired amount of sodium carbonate were
poured down the inlet tube upon the phosphate, and aération was imme-

80 The Quantitative Determination of Ammonia

diately begun in the absence of the layer of oil (p. 82). This volume of
decomposition liquid which was larger than usual, was necessary for the
complete retention of sodium carbonate in solution. During aération of
the mixtures containing 20 or 25 grams of sodium carbonate, a special tend-
ency to incrustation of the tip of the inlet tube was manifested after sev-
eral hours. In a similar series that could not be carried to completion,
aération was entirely stopped in such mixtures by the filling of the tip of
the inlet tube with solid sodium carbonate, which formed from the dry-
ing of layers of the solution that were deposited in the lower end of

TABLE V. FIFTH SERIES, GROUPS K TO M.

Crystalline ammonio-magnesium phosphate (from different products, as in the
third and fourth series), 0.5 gram. Sodium carbonate, 15 grams. Periods
of aération, 5, 10 or 15 hours. Loss of ammonia: maximum, 32.1 per cent;
minimum, 7.8 per cent.

4 VOLUME OF STANDARD ACID SOLU-
' NH,MePO,. S | TION REQUIRED TO abies

; Ss | | 2 | AFTER AERATION. | 38 3

Group. 3 | Se |e nel =

& ° | | Wo ae

E | Seales Shs , | 4s] 6s

s | Product. | | 6) 8 | See

B | ele i2)e |S) 2eneeuee
no. | gram \grms.| cc cc ce. ce. cc. |pr.ct.
lies eee 60 Eimer and Amend) 0.5) 15 9.5) 9.5 |10.75)11-6
61 Mallinckrodt are en Oss mals 12.3112.3 |13.85}/11.2
1562 wal Merc kere aereae ONS alo | 9.2; 9.2 |10.40)11.5
| 63)4|\Ouriowni...-.2. Ozbipalo | 9.4) 9.4 110.20] 7.8
L......| 64 |Eimer and Amend) 0.5] 15 9.65) 9 .65/10.75|10.2
65 |Mallinekrodt..... | 0.5) 15 (Lea ear ss 11.15]13.85| 9.5
66:4) Mercksa eee OFS a5 9.45 9.45/10.40) 9.1
| 67 |Ourown....... | 0.5) 15 9.00 9.00/10.20/11.8
M.....| 68 |Eimer and Amend! 0.5} 15 |8.95 8.95}10.75)16.7
69 |Mallinckrodt..... 0.5} 15 |9.40 9.40)13 . 85/32 .1
70s | (Merckens ee | 0.5) 15 |8.35 8.35)10.40)19.7
TAL AO Wee ON o 4 5 ao ee 0.5} 15 |8.65 8 .65)10.20/15.2

the inlet tube as the solution rose and fell with fluctuations of pres-
sure. In the sixth series such incrustation was prevented from inter-
fering with aération by an occasional cessation of the process for a few sec-
onds. The alkaline solution then rose in the inlet tube and the carbonate
crust was dissolved. Some time elapsed, after aération had been resumed,
before discontinuance for the same reason again became advisable.
These facts are mentioned merely to show clearly that in this series we
used excessive proportions of sodium carbonate.

Matthew Steel and William J. Gies 81

TABLE VI. SIXTH SERIES, GROUPS N AND O.

Pure, crystalline ammonio-magnesium phosphate (our own product, p. 75), 0.5
gram. Sodium carbonate, 5 to 25 grams. Periods oj aération, 4 hours or
10 hours (in two periods of 5 hours). Loss of ammonia: maximum, 34.3
per cent; minimum, 3.4 per cent.

. VOLUME OF STANDARD ACID SOLUTION
g e REQUIRED FOR NEUTRALIZATION. 3
Sn) | ee | 8 ae , 8
Group. | F ae Q | AFTER ee & | Bs | E
¢ | BE a SI a > i, Be ies
a ra a ra a <a) =
4 a a IG
ar Ye 1D apo)
no. gram. | grams. cc ce. cc cc. ce. pr. ct.
GF e c -. a AOs5: || 20° | 8.75 9.75-| 10,29|) 454
fol O:o 5 6.70 6.70 | 10.2 | 34:3
be 74 O25 10 6285) | O70") fzo0n lt 1OR2Ze E2620
75 0.5 15 8°40 | 0-55 | 8.95 | 10227 |) 1233
Toma Oxo 20 | 9.50 | 0.10 | 9.60 | 10.2 5.9
ie O59 25 9.85 | 0.00 | 9.85 | 10.2 374

* The carbonate-phosphate mixtures in this series showed increasing clearness with
the rising concentration of carbonate. In test 77 the turbidity was almost negligible, show-
ing that practically all the triple phosphate had gone into solution. In dilute NaOH
(Table III) the product dissolved completely.

The results summarized in Table VI make it apparent that
even in the presence of 50 times its weight of sodium carbonate,
and after to hours of thorough aération, crystalline ammonio-
magnesium phosphate cannot be completely decomposed by the
Folin method of ammonia determination.

Seventh Series. At this point we made additional checks of
our manipulations and experimental conditions as is indicated
in Table VII. Method A was used—-without triple phosphate
in Group P; with ammonium chlorid, instead, in Group Q.

The data in Table VII finally removed our suspicion that our
results were due to impure chemicals, faulty technique or unsus-
pected errors.

Eighth Series. This investigation was concluded with a deter-
mination of the effects of relatively very great excesses of sodium
carbonate in the aération process. The results of this final test,
which are given in Table VIII, merely confirmed the conclusion

82 The Quantitative Determination of Ammonia

TABLE VII. SEVENTH SERIES, GROUPS P AND Q.

Tests of the purity of the sodium carbonate, and of the possible disturbing influence

of the kerosene, used in the first six series.

and on the handling of the Folin method in general.

Checks on the aération process

STANDARD SOLUTION OF
ACID.
g | <3
S Aération 4 hrs. in the absence of | : | 3 | so
2 NH,MgPO, but in the pres- | 3 ee 8
: a ence of— 2 ea &
2 =I - §
2 | 3 | 8 | 2 eae
oO A =) jon Z <
no. cc cc. cc.
P 78 lWateralone psa. .-05 eke Sel Lon) elo, Om arene
79 \Water + 15gramsNa,CO, ....| 15.0 | 15.00 “
80 |Water + 15 grams Na,CO, +! |
KEROSENE ee aol te tee 15.0 | 15.05 bg
ay Shae LOce: SIN EC eaay i arene re re
81 a. Without kerosene ....... 15.0 | 10.00 ! 5.00 | none
82 b. Without SS ge Be ae 15.0 | 10.05 | 4.95 oS
83 c. With oi Tag ey eee 15-0) L000) | Fas00 a
84 d. With amy Se a ere 15.0 ; 10.00 | 5.00 «

TABLE VIII. GROUP R.

Pure, crystalline ammonio-magnesium phosphate (Eimer and Amend product),

0.05 gram. Sodium carbonate, 2 to 16 grams. Periods of aération (2), 5 hours.

Loss of ammonia: maximum 35.19 per cent; minimum, 12.04 per cent.

a
2 tn os
= sO One
Group. zi eo 28
5 ot ee
S Bo =
a)
no gram. gram
R 85 0.05 2
86 0.05 4
87 | 0.05 8
88 0.05 16

|
VOLUME OF STANDARD ACID SOLUTION

REQUIRED TO NEUTRALIZE—

AFTER AERATION. | |

If all NH,

en WHR GAGs eee

5hrs. |S hrs. (2)| Total. {liberated |

Conte ience: ce. ce.

0.65 0.05 0.70 1.08
0.60 0.15 0.75 1.08
0.70 0.10 0.80 1.08
0.80 0.15 0.95 1.08

Ammonia lost.

—a

Matthew Steel and William J. Gies 83

already drawn that Folin’s splendid method fails, in the case of
triple phosphate, to give perfectly accurate results for ammonia
content.

We hope in the near future to report a simple modification
of the Folin method that will yield all the ammonia from am-
monio-magnesium phosphate in urine without producing ammo-
nia from non-ammoniacal radicals.

A STUDY OF THE INFLUENCE OF MAGNESIUM SULFATE
ON METABOLISM

By MATTHEW STEEL.

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

(Received for publication, May 7, 1908.)

SE NEMNCEANGS sh aiccie st cls 2 9 oye afeie +o oe + chee Dae ee ee 85
fee General description of the experiments. ............2sse8s50 86
Ill. Firstexperiment. First metabolism experiment............. 88

IV. Second experiment. Does bone ash in the diet offset the
catharticinfluenceof magnesium sulfategiven withthe food? 98
V. Third experiment. A determination of the conditions most
favorable for intravenous injection of large doses of magne-
RAAIRE DER IOLO MOPS 8s oo a5 ain = S00 go ees 3 2s 2 erent Oa IOI
VI. Fourth experiment. Second metabolism experiment......... 102
VII. Fifth experiment. Does the intramuscular injection of large
doses of magnesium sulfate into dogs cause the develop-

RPE SSO OCRSEOE (7 Suita Tooo0s 3S caid 2 oe! oi aid 2 wi aie» gal gene pede 118
VIII. General comparison and discussion of the results............. 118
ie smimimary of general conclusions... ......-2.-20++2s0+reense 123

I. INTRODUCTION.

Magnesium is normally present in the tissues and fluids of all
plants and animals. Comparatively little is known, however,
regarding the rdle that magnesium plays in the manifestations of
life phenomena, particularly in animals. Some of its salts are
widely used therapeutically, to produce purgative effects espe-
cially.

Many investigators have studied the influence of magnesium
salts upon animals, particularly after the injection of such salts
by one method or another. The greater number of the injection
experiments with magnesium salts have been carried out during
the past thirty-five years by investigators who have, in effect,

85

86 Influence of Magnesium Sulfate on Metabolism

inquired into the truth of the statement by Luton’ that subcu-
taneous injections of small doses of magnesium sulfate produce
purgation. Very discordant conclusions have been drawn mean-
while by the authors of the subsequent papers on this particular
phase of the subject. Recke? and Hay*® reported that subcu-
taneous injections of magnesium sulfate in relatively large doses
may be fatal to animals. Haystated that death occurs under
such circumstances from respiratory paralysis.

Lately Meltzer* and collaborators have shown that magnesium
sulfate, if subcutaneously injected in relatively large though non-
lethal doses, produces profound anesthesia without excitation,
but with complete relaxation of all the voluntary muscles, and
with abolition of many reflex acts. The animal ordinarily
recovers completely from the anesthesia produced by such non-
lethal doses.

These and the many additional observations by Meltzer and
others led Dr. Gies to suggest an investigation of the meiabolic
effects of magnesium sulfate. A careful review of the litera-
ture indicated that no extensive metabolic study of this kind
had ever been attempted before this work was inaugurated.°

II. GENERAL DESCRIPTION OF THE EXPERIMENTS.

This investigation consisted of two long metabolism experi-
ments, in which food and excreta were analyzed: and several
collateral experiments, in which no analyses were made. The
metabolism experiments were carried out by the methods usu-
ally employed in this laboratory.®

ANIMALS AND ENVIRONMENT. All the experiments were performed on
full grown, healthy dogs carefully selected to suit the special conditions

1 Luton: Gazette hebdomedaire de médecine et de chirurgie, Xxviii, p. 455,
1874.

? Recke: Dissertation, Gdttingen, 1881.

’ Hay: Journ. of Anat. and Physiol. vxii, p. 512, 1883.

‘Meltzer and Auer: Proc. of the Soc. for Exper. Biol. and Med., ii, p. 81,
1905; Amer. Journ. of Physiol., xiv, p. 366, 1905.

5Yvon: Arch. de Physiol., xxx, p. 304, 1898; Winterberg: Jahresber.
der Tierchemie, xxxiii, p. 799, 1903; Malcolm: Journ. of Physiol., p. 183, 1905.

® Mead and Gies: Amer. Journ. of Physiol., v, p. 106, 1901; also, Gies
and collaborators; Biochemical Researches, i, p. 419 (Reprint No. 21), 1903.

i

Matthew Steel 87

of the experiments. In each case we were particularly fortunate in our
selections.

The animals under observation were confined in cages devised in this
laboratory and recently described.!

Foop. The daily food consisted of a mixture of hashed lean beef, lard,
cracker meal, bone ash,and water. The raw beef was taken from large
supplies preserved in a frozen condition.’

The daily portion of cracker meal was obtained from large quantities of
the commercial product kept in big bottles with ground glass stoppers.
Lard was purchased in small amounts and kept in a refrigerator. It was
always devoid of rancidity. The bone ash consisted of the thoroughly
incinerated, carbon free commercial product. Ordinary tap water was
used. In the first experiment the dog was given his daily portion of food
at 12:15 p.m.; in the second, at 11:30 a.m.

PERIODS, WEIGHTS. In our metabolism experiments each day of the
first experiment ended at 12:15 p.m.; of the second experiment, at 11:30
a.m. The dogs were weighed just before they were fed. A new period
was always begun with the day on which the dog was subjected to new
conditions. The periods were made relatively long to afford more reliable
‘‘daily average” results of analysis.

COLLECTION OF EXCRETA. We collected the urine and feces daily just
before feeding. No artificial means such as catheterization were employed
in this connection. The cage was washed at the end of each period and
the total content of nitrogen in the washings determined. This total was
added to the total nitrogen of the urine for the corresponding period.

ANALYsIs. The total content of nitrogen in the ingredients of the food
was determined. The urine was carefully preserved with thymol. Uri-
nary nitrogen, in the leading forms, was determined as follows: ammonia,
each day; total, urea and creatinin, every other day; and allantoin, uric
acid and purin bases, at the end of each period.

ANALYTIC METHODS. Total nitrogen was determined by the Kjeldahl
process, oxidation being effected by means of concentrated sulfuric acid,
aided by a little cupric sulfate. Ammonia, urea, creatinin and uric acid
were determined by the methods devised by Folin.‘ Allantoin was deter-
mined by the Loewi® method. Nitrogen of purin bases was determined
by a combination of the Arnstein’ and Salkowski’ methods.

1Gies: Amer. Journ. of Physiol., xv, p. 403, 1905.

2 Gies; Amer. Journ. of Physiol.,v, p. 235, 1901; also Gies and collabora-
tors: Biochemical Researches, i, p. 69 (Reprint No. 1), 1903.

3 Steel and Gies: Amer. Journ. of Physiol., xx, Pp. 343, 1907.

4 Folin: Amer. Journ. of Physiol., xiii, p. 45, 1905.

5 Loewi: Arch. f. exp. Path. u. Pharm., xliv, p. i, 1900.

® Arnstein: Zeit. f. physiol. Chem., xxili, p. 417, 1897.

7 Salkowski: Salkowski’s Manual of Physiol. Chem. and Path., Transl.
by Orndorff, p. 24, 1904; Arch. f. d. ges. Phystol., 1xix, p. 268, 1897-98.

88 Influence of Magnesium Sulfate on Metabolism

Magnesium in bone ash and feces was determined as follows: 2.5
grams of the fine powder that had been dried to constant weight, were
dissolved in 50 cc. of HCl (sp. gr. 1.15), evaporated to dryness and desic-
cated at 110° C. The dry residue was dissolved in a few cc. of HCl
and about socc. of water. The solution was filtered, the filtrate made
up to 250 cc. and 50 cc. portions of this solution taken for analysis. A
quantity of ferric chlorid sufficient to precipitate the phosphoric acid
was added to the solution, after which the mixture was made slightly
alkaline with ammonium hydroxid. About two grams of sodium acetate
were next added, the solution then boiled a few moments and filtered.
The filtrate was treated with a moderate excess of ammonium oxalate
to precipitate the calcium, heated, filtered hot, the filtrate evaporated
to about 50 cc. and neutralized with HCl. A moderate excess of di-
sodium hydrogen phosphate was added; ten minutes later 10 cc. of
strong ammonium hydroxid were added, and the mixture allowed to stand
overnight. The product was incinerated and weighed as Mg,P,O,.

MAGNESIUM SULFATE. The magnesium sulfate employed in all the
experiments was a sample of a Kahlbaum (‘‘K’’) product.

Ill. FIRST EXPERIMENT. FIRST METABOLISM EXPERIMENT.

The dog selected for this experiment was a healthy male, weighing
about to kilos.

Diet. The character of the diet was practically the same throughout
the whole experiment. Three samples of meat were used, but all had
essentially the same nitrogen content. The amount and composition
of the daily diet is indicated in Table I.

TABLE I.
Composition of the daily diet.

INGREDIENTS. HASHED LEAN BEEF. |CRACKER | BONE |

MEAL. LARD.* ASH. | eas
Taya seh yee an eee | 1-7 | 8-40 | 41-62 | 1-62
e | 2
| grams. |grams.| grams. | grams. grams. | grams. | cc.
Daily amounts.......; 150 150 150°} 40,. | 30 | -205s ae
Nitrosen. sn eeee | 5.31 |5.91]| 5.34 | 0.650 | 0.008 | 0.006 |

| ee.

*Only one sample of lard was analyzed. The figure obtained for content of nitrogen
was practically the same as that heretofore noted for the same brand of lard that has long
been in use in this laboratory.

PREPARATORY PERIOD. Eleven days onthe above diet sufficed to accus-
tom the dog to the food and environment, and afforded ample opportunity
to observe the general characteristics of the animal. During this time

Matthew Steel 89

the weight of the dog rose from 9.80 kilos to 10.16 kilos. The preparatory
period ended on November to, 1907, and on that day, at 12:15 p.m., the
accumulation of analytic data was begun. The main part of the experi-
ment was continued uninterruptedly thereafter until noon of December
27, 1907 (47th day). A general supplementary period was maintained
until January 11, 1908 (62d day).

The experiment was divided into six periods of different lengths and
conditions as follows:

First PERIOD. Normal conditions. Days 1-10; November 11-20,
1907. During this period the animal was entirely accustomed to his envir-
onment and normal conditions were maintained uniformly.

SECOND PERIOD. Influence of magnesium sulfate administered per os.
Days 11-21; November 21-December 1. On the 11th day the first
dose of magnesium sulfate was given per os just before feeding. Dosage
in this instance, as in all subsequent administrations by mouth, was
effected by enclosing the magnesium sulfate in a little ball of the weighed
hashed meat, which the animal swallowed quickly without masticating;
and consequently, also without tasting the magnesium compound. The
remainder of the food was offered immediately afterward and eaten at
once.

We desired to ascertain the effects of magnesium sulfate on the partition
of nitrogen in urine eliminated during a fairly long period. Consequently
only a very small dose (0.513 gram) was given on the 11th day (ist day
of the period), but the dose was gradually increased until the 21st day, when
7.001 grams were given. By that time the purgative effect was so pro-
nounced that we were forced to discontinue dosage. Otherwise the dog
seemed perfectly normal. He was cheerful and showed no signs of dis-
tress.

THIRD PERIOD. Intermediate period. Days 22-29; December 2-9. The
recovery of the dog from the diarrhea was almost immediate. On the
22d day diarrhea was most pronounced as an effect of dosage on the last
day of the previous period, but on the 23d day the feces were only moder-
ately diarrheal. On the 24th day they were almost normal. Thereafter
in this period there were no diarrheal tendencies.

FourtTH PERIOD. Influence of magnesium sulfate when injected subcu-
taneously. Days 30-41; December to-21. By the end of the preceding
period the dog had practically returned to normal conditions. We next
attempted to ascertain the influence of magnesium sulfate injected sub-
cutaneously. We used very small amounts at first but gradually in-
creased the daily dose until the general effects were so decided that it
was obviously inadvisable to continue the treatment.

Schedule of injections of magnesium sulfate. 30th day. At 12:45 p.m.,
0.01 gram dissolved in 1 cc. of water was injected on the right side. There
were no appreciable effects. 3jrst day. At 12:35 p.m., 0.025 gram dis-
solved in 13 cc. of water were injected on the left side. Again there were
no particular effects. The feces for the day were a trifle softer than
normal. 32d day. At 12:50 p.m., 0.05 gram dissolved in 1 cc. of water were

go Influence of Magnesium Sulfate on Metabolism

injected on the right side. The animal was uneasy after the injection,
squirming about and rolling on his back. He defecated at 1 p.m. The
feces were soft, clay-like. 33dday. At12:50p.m., 0.1 gram dissolved in 2
cc. of water was injected near the shoulder on the left side. The feces had
the consistency of putty and the urine was turbid. 34th day. At 1:10
p-m., o.2 gram dissolved in 1 cc. of water were injected on the right side.
The feces were a trifle softer than on the day before. The turbidity of
the urine was about the same. 35th day. At 12:45 p.m., 0.4 gram dis-
solved in 2 cc. of water were injected on the right side. The feces were
softer than on the previous day. The turbidity of the urine was also
greater. 36th day. At 1:05 p. m., 0.6 gram dissolved in 1.5 cc. of water
were injected on the left side. The feces were partly soft and partly hard,
and the turbidity of the urine was about the same as on the previous day.
37th day. Ati p.m., 0.6 gram dissolved in 1.5 cc. of water were injected
on the right side. The dog was more decidedly affected than before.
After the injection he was very restless at first and then became unusually
sleepy. The feces were hard. The urine was very turbid and had con-
siderable sediment. 38th day. At 12:50 p.m., 1 gram dissolved in 2.5
cc. of water was injected on the left side. After the injection the dog was
very uneasy and apprehensive at first, but then gradually became dull
and drowsy and remained so until evening. The feces were hard and dry
and the urine had a heavy sediment. 39th day. At 12:40 p.m., I gram
dissolved in 2.5 cc. of water was injected on the right side. At 4 p.m.
another gram was injected on the left side. After the first injection the
dog showed the symptoms exhibited on the day before, namely, at first
he was very restless and apprehensive and then soon became drowsy.
After the second injection he immediately assumed a recumbent position,
then went to sleep and slept all afternoon and evening. On being awakened
at 9:15 p.m., he urinated. The next morning there were signs of edema
on the left side in the lumbar region and he was sore and stiff in his hind
legs. The feces were dry and brittle. The urine had a heavy sediment.
goth day. At12:40p.m., 5 grams dissolved in 8 cc. of water were injected
on the left side. The dog became very drowsy within a couple of minutes
and could not be persuaded to arise. He cuddled up in the corner of the
cage and shivered as if he were cold. Occasionally he got up to urinate.
The total amount of urine for the day was almost double the amount of
water given with his food. The feces were very hard and brittle. Next
morning the dog was bright and cheerful. rst day. At 12:50 p.m., 10
grams dissolved in 12 cc. of water were injected on the right side. The
animal exhibited practically the same symptoms as those shown on the
40th day, and was even more sleepy. The sleepiness had not worn off by
the next morning. The dog then acted as if in great pain when he was
touched in the region of the last injection. His appetite for the first time
was poor.

FIFTH PERIOD. After period. Days 42-47; December 22-27. During
this period the dog was permitted to recover from the effects of the dosage.
On the right side in the lumbar region, where the last injection was made,

Matthew Steel gI

a swelling gradually developed which opened on the 44th day, revealing
a considerable area of necrosis. Sloughing was prompt and complete.
No treatment of the wound was necessary, as the surface tended to become
dry and healed rapidly. By the 47th day the dog seemed to be normal
again in all respects with the exception of the sore on the side.

SIXTH PERIOD. Supplementary period. Days 48-62; December 28,
1907—January 11,1908. The dog was kept under general observation for
two weeks in order that possible ‘‘after effects’’ might be detected.

ANALYTIC RESULTS. The more important data in our records
of the first metabolism experiment are given in Tables IJ and III.

DISCUSSION OF THE RESULTS. Changes in the weight of the
dog. Examination of the data recorded in Table II shows that
there was a gradual and continuous increase in weight during
the fore period (1st), throughout the period during which magne-
sium sulfate was given per os (2d), during the intermediate period
(3d), and during the first half of the period in which magnesium
sulfate was administered subcutaneously (4th). During thesecond
half of the latter period there was a rapid decrease of weight, due,
undoubtedly, to the diuretic effect of the larger subcutaneous
doses of magnesium sulfate. The decline in weight reached its
lowest ebb on the last day of the period (4th) of subcutaneous
injection. During the fifth or ‘‘after’”’ period, the weight quickly
rose, but did not quite return to the maximal point (10.70 in the
4th period) before the end of the supplementary period.

Fluctuations in the daily volume of urine. The amount of water
in the food remained the same each day, yet, as an examination
of Table II will show, the volume of urine fluctuated through
quite a wide range. The minimal amount was 155 cc. on the
44th day, during the “‘after’’ period, when the animal was rapidly
recovering from the effects of the diuresis of the subcutaneous
injection period. The maximal amount was 716 cc. on the goth
day (next to the last day of the injection period). The relatively
large doses of injected magnesium sulfate in the latter part of the
fourth period appeared to exert a marked diuretic influence.
In the fifth period there was a corresponding retention of water,
and a consequent decrease of urine volume, as the weight of the
animal rose.

The comparatively uniform daily volume of urine in the second
period indicates that the doses of magnesium sulfate that were

92 Influence of Magnesium Sulfate on Metabolism

administered per os during that period had very little influence
on the volume of urine. The slightly reduced average daily vol-
ume of urine for that period (Table III) is explained by the moder-
ate diarrhea that resulted from the corresponding doses per os.

Alterations of the specific gravity of the urine. With a few
exceptions the specific gravity of the urine showed little variation
from the average except in the second and fifth periods, when the
excreted proportions of magnesium sulfate were large, while at
the same time the volumes of urine were at or below the average,
or when the volumes of urine were particularly small while the
nitrogen catabolism was comparatively extensive.

Effects on the elimination of feces. Aside from the fecal effects
already noted, little can be said in this connection. Fairly large
doses per os made the feces softer. Intheearly part of theinjection
period doses subcutaneously administered appeared to cause slight
softening. This may have been a mere coincidence, however.
The subsequent larger doses seemed to make the fecal masses
more brittle than they were during the normal period. The aver-
age daily quantity of dry fecal matter increased slightly per period
until the fourth, inclusive, then fell off rather decidedly during
the last two periods.

Effects on the elimination of total nitrogen of the urine. The
object of this experiment was primarily to ascertain the effect
of magnesium sulfate on the partition of urinary nitrogen.
Therefore we did not analyze the feces nor cast off hair in this
experiment and consequently made no attempt to ascertain
whether the animal was in nitrogenous equilibrium. Examina-
tion of Table III shows, however, that in the post-subcutaneous
injection period (5th) the total amount of urinary nitrogen was
considerably higher than in any of the other periods, and greater
than the total quantity of ingested nitrogen. In the injection
period (4th) the average daily elimination of urinary nitrogen
was practically the same as during the previous period (3d) and
less than the amount of ingested nitrogen. That the increased
elimination of total nitrogen in the urine of the fifth period cannot
be ascribed to the mere diuretic influence of the injected magne-
sim sulfate is evident from the fact that during that period (5th)
in which the greatest elimination of nitrogen occurred the
volumes of urine were the smallest of the experiment. During

iii

Matthew Steel 93

the period in which magnesium sulfate was given per os, there
Was no increase in the total urinary nitrogen.

Effects on the partition of the total urinary nitrogen. The most
noticeable effects on the distribution of urinary nitrogen among
the products named at the head of Table III were (1) a fairly
uniform absolute average daily excretion of urea but a striking
decrease in its relative elimination, beginning with the period
after administration of magnesium sulfate per os and continuing
to the end of the experiment; and (2) cumulative increases in
both the absolute and relative excretions of nitrogen in the
forms of allantoin, uric acid, purin bases and undetermined
products. ‘There was a slight increase in the relative urea excre-
tion in the second period, but it hardly deserves special attention.
Creatinin elimination was fairly uniform throughout the experi-
ment, although there were indications of slightly increased output,
both absolute and relative, in the dosage periods. The ammonia
nitrogen was exceptionally high during the period in which
magnesium sulfate was given per os and again in the last (5th)
period. It was low in the subcutaneous injection period. The
anomalous results especially in the subcutaneous injection period
were doubtless due to the sources of error referred to in the
preceding paper.’ Ammonia elimination appears to have been
very perceptibly increased in the second period as the daily
doses of magnesium sulfate per os were increased

See the discussion on p. 118 for further reference to these
results.

1 Steel and Gies: This journal, this volume, p. 71.

94 Influence of Magnesium Sulfate on Metabolism

TABLE II.

Some of the daily records of the first experiment.
The first metabolism experiment.

First periop. Normal conditions. Days 1-10; November 11-20, 1907.

URINE. FECES.
. | Body | Dose of & | q | EI | |
S| weight. | MeSO,.| g | SS gh 2%. 6 | Bh | pe
2 Ss | $e | SS | 6s | 82) Cea
Q > TD bh Be a <A O's
kilos. grams. cc. 1022. grams. | grams. grams. grams. | grams,
1} 10.19 280 15 0.123 | 2D
PAN UO) SIZ 478 13 10.224 8.482 ; 0.152 | 0.228 | 30°74
3 10.14 380 14 0.182 | 54.7
4 | 10.13 390 13 9.964 8.686 | 0.1380 | 0.247 | 28.4
5 | 10.09 418 13 0.142 Silt.
Gm pel On tes 290 15 9.662 8.398 | 0.124 | 0.227 | 41.4
T0226 355 15 0.1382 0
8 | 10.20 352 16 10.376 9.134 | 0.133 | 0.226 | 4472
9 | 10.24 360 13 0.094* 25.4
10 | 10.25 380 16 9.800 8°657 | 0.175 | 02228) taro

* This amount is small, but was correctly determined.

SECOND PERIOD. Effect of magnesium sulfate administered per os.

Days 11-21; November 21—December 1.

11 | 10,25 | 0.518 | 342 15 0.158 46.4
12 | 10.29 | 0.504 | 350 | 15 9.662 | 8.467 | 0.122 | 0.229; O

13 | 10.34 | 0.999 | 360 15 0.127 LG oar
14 | 10.44 | 1.506 | 310 18 9.789 | 8.662 | 0.171 | 0-219 | 2i2
15 | 10.37 | 2.001 | 400 15 0.219 55.7
16 | 10.33 | 2.499 | 382 16 10.298 | 9.089 | 0.221 | 0.256 | 39.0
17 | 10.40 | 3.000 | 366 17 0.259 0

18 | 10.42 | 3.999 | 348 | 18 10.278 | 8.975 | 0.269 | 0.239 | 47.2
19 | 10.42 | 5.001 | 340 18 0.279 40.5
20 | 10.44 | 6.006 | 360 | 20 0.307 23.2
PANEL. 46.197 O01 a Ss2bu) ee 14.895 | 12.866 | 0.284 | 0.379 | 29.5

|

Matthew Steel 95

TABLE II—CONTINUED.

THIRD PERIOD. Intermediate period. Days 22-29; December 2-9.

URINE. _ FECES.
} | Body Dose . P : 3 g
7 | weight. | MgSO, | o os, a ae a8 ‘aie ;
2 q | ae a 2 g2 | 3 re
5 = |) gece | 52 | Se ie glues
aa > | oo a Rue wae 24. a lee
kilos. grams. cc. | 10xx grams. grams | gram gram. | grams.
22 | 10.40 388 15 | 0.253 26.1
Bs | 10.39 350 15 12.169 | 10.481 | 0.200 | 0.236 | 37.9
24 | 10.48 300 15 0.132 fe 22ril
25 | 10.45 330 16 10.812 9.268 | 0.138 | 0.186 | lil
26 | 10.42 412 13 0.101 | 34.6
27 | 10.47 | 314 | 16 10.270 9.201 | 0.124 | 0.223 | 28-0
28 | 10.55 290 | 18 0.135 29.2
29 | 10.54 385 14 10.733 8.368 | 0.134 | 0.200 | 29.1

FourtuH Periop. Effects of magnesium sulfate injected subcutaneously.
Days 30-41; December 10-21.

30 | 10.66 | 0.010 | 268 | 16 0.152 | | 13.9
31 | 10.61 | 0.025 | 348 | 16 9.747 | 7.582 0.165 0.227 | 20.4
32 | 10.70 | 0.050 | 306 _ 16 0.136 \ 37.2
33 | 10.66 | 0.100 | 310, 16 9.806 | 7.999 | 0.142 | 0.232 | 27.0
34 | 10.67 | 0.200 | 340 | 17 0.124 | | 44.2
85 | 10:70 | 0.400 | 270 | 19 | 10.025 | 7.741 | 0.155 | 0.243 | 35.1
36 | 10.60 | 0.600 | 420 | 14 | 0.125 36.9
37 | 10.64 | 0.600 | 230 | 24 Din 8.350 | va 0.212 ree
38 | 10.61 | 1.000 | 498 | 15 0.222 332
39 | 10.50 | 2.000 | 430 | 18 | 11.829] 9.378 | 0.165 | 0.263 | 33.3
40 | 10.15 | 5.000 | 716] 15 0.112 2108
41 | 10.03 {10.000 | 479 | 24 | 14.329] 7.698 | 0.149 | 0.419*| 26.5
|
Firtu Periop. Ajter period. Days 42-47; December 22-27.
42 10.20 225 33 | 0.175 id
43 | 10.30 224! 38 | 16.082 | 10.097 | 0.374 | 0.264 | 33.2
44 | 10.42 155 | 37 0.161 41.2
45 | 10.37 300 | 25 | 15.054 | 10.009 | 0.315 | 0.223 | 42.9
46 | 10.41 260 | 23 | 0.286 20.4
47 | 10.52 290 | 21 | 12.224 6.827 | 0.219 |) 0,208 | 2100

* This amount is large, but was correctly determined.

96 Influence of Magnesium Sulfate on Metabolism
TABLE II—CONTINUED.
SrxtH PERIOD—Supplementary period.
Days 48-62; December 28, 1907 to January 11, 1908
: |

Number of the day......... 48 | 49 | 50 | 51 | 52 | 53* | 54 | 55
Body weight (kilos)........ (10.50 10.54 10.50)10.52 10.49,10.42)10.44,10.50
Urine: Volume (cc.) .......| 340] 358 | 376 | 312 | 420 | 360 | 345 | 314

Specific gravity (10xx)... | 16 17 15 16 13 17 16 itz
Feces: weight (grams)...... 136.4 118.5 |83.0°|25.1 (31.8 138.6 |15.2 |f573
Numberof theday:. 2.22. 2.2. 56»-|.-57 =| 58° | 59.) (GOss Gee
Body weight (kilos)............. 10.41}10.47/10.50/10.53)10.54/10.59)10.61
Urine: Volume (cc.) ......... 363 | 420 | 310 | 334 | 342 | 330 | 350

Specific gravity (10xx)......... iG 16 18 17 ly 18 15
Feces: weight (grams)........... 37.0 |23.4 |21.3 154.2 |14.0 |21.9 |26.8

* The total nitrogen was 29.90 grams for days 48 to 53, inclusive, or an average of 4.983

grams per day.

‘(ES-8b) SABpP XIs JsIY OY} 10j yuNOUIe OFBIOAW |

*(€9-gp) sAup XIs 481g 04} 10} AyIWUEND [BOT

< 226% £S'0 CLT es'T 6S°T zoe TT Op ee | ev ect Rosictars een ctapaya she *<-yeqsyl A
oy) LE'8t 8¢°0 €¢°0 F0'T PPG IL’°S 6h FL sree esssssssssnoeusjnogns OSs] AT
oP 6 F1'0 ch 0 49°0 G61 9L°S OSS sy ie aj ames RS ass yeh: ree eyetpeunezuy; TIT
Ch'F 22°0 820 29°0 OFZ SEF 0S" Lg Cre. e: ie! 4! C0 ayW wile) biey.sl ee) ele, Skye ea 80 dad FOQ3 Il
10°2 FIO seo ZL°0 Ze Bez 99°98 Debate tele Jel cov oracle Mant cteker cnere ‘sss -reuION| I
quaa wad |*7Uad ed|"yuaa wad|-quao sad | yuan sad \"quao wad\ yuao sad
“UuaBoujvu 70j0} 24) 07 UWoUDpaaT :
0S' 4% | s¢e FESOER I POOR Re ae cress bhai Arejuoureyddng) TA
GE°Ss | SE GILG 6€0°0 | Z2T°O | 060°0 | STT'O | Ssz‘o0 68h FP LOG L ALON Vel fears Rc en mee SIE 9 72.00 “"IOIFV) A
08°62 | F8E Z00'T TZ0°0 | 620°0 | 490°0 | $sTt°O | 6FT'O £90°F $SP'¢ 6LP'9 | “tess tsss + -snosueznogns "Og3W] AT
€T'6S | OFE 81c°0 800°0 | €20°0 | £80°0 | 901°0O | ZSt'o ¢99'P 86F'S 12 IS Cil lemmea tegen <p sabe PAR wa eons oo oyBIpsuieyUy) TIT
ras €1°6% | ege G26 0 T10°0 | 610°0 | Tg0°O | OZT'O | 61z'0 698 °F £66'°P PPA OA ye a en ES oan Nena so ad YOST! IT
Oo 60°62 | 898 Tg¢°oO 400°0 | 610°0 | 980°0 | 9TT‘°O | 6ET'O 9ee'F €00°¢ 1 TI oe? Lal We ae eee ag Sai oe oot oy eS [eunoN) T
_—)
Yn *sabpiaan fijiwg
= = = See a = —— = — = = a = — —s — = —— —
= bZIP PLE «06°62 090° 06 Wide. ee a a 6 pe evmKeuay ee Vie. ae Arejuawelddng IA
ate; L OLE SPE PL9.eR | S8s3/0) | 09270 ||| OFS 0) || 0890) | OS9°T | Ge6"Oa. | OOS" Sr || PaOsOe es es aera ee wakes eieroie TOV) A
(a) 9° LSE | PI9P | GIO'ZT | 9%Z°0 | OSE'O | 989°0 | LEST | 962'°T | FEL'Sb | QPR'SO | SPL Lz “****-snosueynoqns FOS) AT
= T° €€% | 6922 | FFI F SOON PAS TON R6G On SPS 0 P20S DSL ere) | PRB Cr il cGarca | ee oe eeicratae “*"o7,eIpeuiezUy! TIT
POZE | S88 | LHP'S OG OR RG LalOn FORE TO neces Orso lmGa Os SpanlecaG pone clei: eave ket nee cement so wad YOSSI) IT
6°06Z | #898 | 80S’ 990°O | *61°0 | 8SE°0 | OST'T | zZ8E"T | zge°Eeh | 9200S | OFG'T9 |-°° °°: VE oes, Coa ae ek [BUION)| T
*sumDpib ee) “smDpib “UD Una ‘wmpib | -swo.8 | *swps8 “suipiB *swmpib *smpib
< Q
S = | 8 5 e s 2 g 5 g
S 5 jee 5 5 es st. 8 ° ‘ueF0141U)| ‘ues01310 ‘ON
2 Y aS ef 8 oF B. 5 ; : “uoyIpuoD
&. TS al & By 5 5 5. [BI0y, | [Roy
og o 5 » * :
a a
co
; —JoO WHOJ oy4 Ul UsTOI4IN
“saoga ‘ANIUD *aooa Mecopericts

NS re Sl

"nor1ad wana sol emmom afhanyn In Rimaame

*TATOWT Wo renw roauerd

98 Influence of Magnesium Sulfate on Metabolism

IV. SECOND EXPERIMENT. DOES BONE ASH IN THE DIET OFFSET
THE CATHARTIC INFLUENCE OF MAGNESIUM SULFATE
GIVEN WITH THE FOOD?

It was thought that possibly the calcium of the bone ash in
the food exerted a neutralizing effect on the influence of magne-
sium. In view of the fact, however, that calcium salts do not
appear to be appreciably absorbed from bone ash under the
conditions of our experiments,' the possibility of a suspension of
magnesium influences in the manner indicated was rather remote.
Nevertheless, before proceeding farther, we made a direct test
of the matter, and attacked the problem at the point where
results, if any were attainable, would most likely be revealed.
It is a fair assumption, we think, that if bone ash in the food has
no effect on the catharsis induced by accompanying doses of
magnesium sulfate per os, it is likewise without influence on the
behavior of magnesium sulfate administered subcutaneously or
intravenously.

The animal selected for this experiment was a plump, vigorous, short
haired male dog, weighing about 8 kilos. The daily diet was similar to
that of the first experiment and contained the following proportions:
Hashed lean meat, 120 grams; cracker meal, 32 grams; lard, 24 grams;
water, 260 cc. The food was given daily at 11:30 a.m.

First Periop. Without magnesium sulfate and without boneash. Days
I-9; December 1-9, 1907. The dog was kept on the above-mentioned
diet for nine days, without magnesium sulfate and without bone ash, dur-
ing which period the soft, black feces exhibited the usual qualities of the
excrementitious matter from the bowels of dogs on a diet consisting
largely of meat. Defecations occurred only once about every third day.

SECOND PERIOD. Influence of bone ash on the cathartic effect of magne-
sium suljate. Days Io-21; December ro-21. In this period the animal
was given daily with the food a gradually increasing dose of magnesium
sulfate. without bone ash, until decided diarrhea occurred. The dog was
then kept on this dose without bone ash for a few days, during which time
the diarrhea was fully maintained. When we were certain that the daily
dose of magnesium sulfate would invariably produce diarrhea, 15 grams
of bone ash were added to the daily food. The appended summary pre-
sents the plan of treatment and the results in some detail (Table IV).

Plan of treatment with magnesium sulfate. roth to I7th days, inclusive;
magnesium sulfate without bone ash. roth day (first day of dosage). Dose:

1 Steel and Gies: Amer. Journ. of Physiol., xx, p. 343, 1907.

Matthew Steel 99

0.501 gram. The dog did not defecate. sithday. Dose: 1.002 grams.
No feces. r2th day. Dose: 2.001 grams. No feces. 13th day. Dose:
3.003 grams. The dog defecated. The black fecal mass was formed
but quite soft. srgth day. Dose: 4.005 grams. The black fecal mass
was unformed and like thick mud in consistency. r5th to 17th day.
Daily dose: 4 grams. Marked diarrhea occurred daily.

Eighteenth to 25th days, inclusive; magnesium sulfate with bone ash.
18th day. Dose: 4 grams with 15 gramsof boneash. The dog defecated
at about 3 p.m. The feces were thin diarrheal discharges like those of
the past few days. At 1o a.m. shortly before the end of the day of record,
the dog defecated again. This time the fecal discharge was a gray mass,
which although unformed, was somewhat thicker than the feces eliminated
during the 15th-17th days. It appeared to bein effect the usual diarrheal
mass made grayish and thicker by the mechanical admixture with the
bone ash. 19th to 21st days. Daily dose: 4 grams with 15 grams of bone
ash. The dog defecated daily. Each discharge was of approximately
the same consistency as that of the second elimination on the 18th day.
The degree of diarrhea was practically the same as in the first experiment
under similar conditions of dosage.

These results make it evident that bone ash in the diet will not
appreciably offset the cathartic effect of magnesium sulfate
except in a purely mechancial way.

THIRD PERIOD. Wzuth bone ash, but without magnesium suljate. Days
22-23; December 22-23. No magnesium sulfate was administered, but the
daily quantity of bone ash (15 grams) was continued. The recovery
from the cathartic effect was almost immediate. On the 23d day the
feces were entirely devoid of diarrheal consistence.

(This animal was used again in the special experiment de-
scribed in the footnote on p. 104.)

Influence of Magnesium Sulfate on Metabolism

100

0G P6L 00° ST 8° 2 &%
rat ZOE 00° ST OL'L (£6
vxOT 29 *supiB "80714
“AQIABIN, ae aah
Sau 7
oyroedg ee “ysy auog ee “ABp oy} Jo Jaquinyy
“ANIHO t
"86-8e saquaveg “Ee-ee slog
‘powwad sajf{p “aqoruad auiny, “TIT
- ae aes ) As
&6 €1 &%@ CT GG CT GS 61 0& Ce ee OT ** (XxQT) “ad dg
0¢% | O8E | SIZ c6E 81Z ZOF 002 ¢8Z O9T G9 | GZE OLE [°° ***((90) euNfo, :oul—)
00° ST) 00° ET; 00ST’ O00 ST "++" (sured) yse ouog
£00 F |666'€ |100'F | G00'F | T00'F | 00 F | FOO'F | S00 F | €00°€ | 100°% | Z00'T | TOGO | °° °°" *(sureI) "OS3V
€8°2Z | 9272.) Z8°2 | (082 Owes 86°2 €0°8 ee 8 | pe 8 | 862 a g | a, = "+ ***(sopLy) yysIem Apog
1Z Ocee| or ST LI | 9T cI lee ZI r | or Bee Aep ay} Jo Joquiny
{

‘T@-O] saquaseg “Té-o] shivg

‘ayof{ns wnisaubou fo joalfa xyupyyno ay} uo ysp auog fo anuanyuy

‘adoludd GNOOUG ‘JT

CT CT €1 CT
GCE 096 9EE C8
62 €0°8 00°S ¢0'°8
6 8 L 9

LI ST 9T al ST

T¥G O&% O8< VEE 193
OT'$ G08 00's | +¥0'8 60°8

a v € G U

(xxQT)AqAvI3 oyIoadg

EMO (00) eUINJOA :eulI,
afele (ahalalelantiey oitstte faites We (sopry) qySIoM Apog
Deceewe Ch cietiet ech? LieO wet). Cle ia Aep oy} jo Joaquin

‘6-[ saquaseq “6-] shvg
yquaursadxa puoras ay) fo spiovas fijvop ay fo awog

“SUOT]UPUOI JDULO NY

“AL GTEaVL

‘doludd Lsulq ‘|

Matthew Steel IOI

V. THIRD EXPERIMENT. A DETERMINATION OF THE CONDITIONS
MOST FAVORABLE FOR INTRAVENOUS INJECTION OF LARGE
DOSES OF MAGNESIUM SULFATE INTO DOGS.

Before subjecting the dog of the fourth experiment (p. 102)
to the influence of an intravenous injection of magnesium
sulfate, we endeavored, in a preliminary trial experiment, to
ascertain the procedure that could be followed to the greatest
advantage in an effort to administer large doses directly into the
circulation of adog. The following abstract from the protocol of
that experiment confirms the observation by Meltzer and Auer’
that large doses of magnesium sulfate may be introduced intra-
venously provided injection is conducted very slowly.

The dog used in this experiment weighed about 10 kilos and had been
kept on a diet very similar to that given the animal of the first metabolism
experiment. Ten grams of magnesium sulfate, dissolved in 100 cc. of
water, were injected into a saphenous vein, after anesthetizing locally
with cocaine (1 cc. of a 2 per cent solution) at 4:21 p.m. A summary
of operations and observations is appended.

4:22 p.m. Injection of the magnesium sulfate solution was started
from a burette at the rate of about 1 cc. in 35 seconds. This rate was not
increased but occasionally was decreased, as the effects on respiration
became more decided. Among the early effects was a marked gastric
disturbance, though vomiting did not occur. 5:00. Much salivation, aug-
mented nasal secretion, and watering of the eyes. 5:10. Anesthetic effect
quite marked. 5:25. Respirationweak. Reflex responses slight. 5:50.
Profoundly anesthetized. 5:53. The last of the solution (100 cc.)
injected. The animal was placed in a cage after rapid suturing of the
skin at the site of operation. 5:58. The dog tried unsuccessfully to
stand up. 6:02. Got up but was too weak to remain on his feet. 6:15.
Arose and wagged his tail. Able to walk without difficulty. 6:20.
Recumbent and going to sleep. 10:25. The dog urinated, and after that
seemed almost normal again. During the first 24 hours after the injection
there was diuresis. The urine was also exceedingly turbid and soon yielded
a heavy sediment. The first samples of urine after the injection were
very much paler in color than the portions eliminated previous to the
injection. On the next day the dog seemed to have entirely recovered
from the effects of the treatment. The anesthesia seemingly had no
unfavorable sequelle.

1 Meltzer and Auer: Amer. Journ. of Physiol. xv, p. 389, 1905-06.

102. Influence of Magnesium Sulfate on Metabolism

VI. FOURTH EXPERIMENT. SECOND METABOLISM EXPERIMENT.

The general conditions of the first metabolism experiment were dupli-
cated in this experiment. There was also a study of intravenous effects.

The dog used in this experiment was a healthy, short haired female,
weighing about 10.6 kilos. The diet was practically the same as that of
the dog in the first experiment. This dog received 25 cc. more water per
day than the first.

TABLE V.
Composition of the daily diet.

]
INGREDIENTS. HASHED LEAN BEEF. Doeroe neater | paps SSA
|
Day siscticocth cs wetem miss: | 1-14 | 15-45) 46-73 | 1-73
| grams. cane grams. grams. grams. | grams. bere
Daily amounts........ | 150°4 150°} 150 40 30 20 375
Nitrogen.............| 5.34 | 5.55| 5.43 | 0.651 | 0.008 | 0.006

PREPARATORY PERIOD. The dog was kept on the above-mentioned
diet for nine days, before any analyses were made, in order to accustom
her to the cage and diet. In four days her weight increased from 10.56
kilos to 10.72 kilos and then remained at about 10.70 kilos during the next
five days. The preparatory period ended on February 17, 1908. On
that day, at 11:30 a.m., the accumulation of analytic data was begun.
The experiment was continued uninterruptedly until April 30 (73d day)
and was divided into eleven periods of different lengths and conditions
as follows:

First PERIOD. Normal conditions. Days 1-7; February 18-24, 1908.
During this period the weight of the dog remained remarkably constant.
She seemed to be perfectly satisfied with her surroundings.

SECOND PERIOD. Influence of magnesium sulfate administered per os.
Days 8-14; February 25-March 2. With our previous experience in
mind we began the treatment with a fairly large dose.

Schedule of administrations of magnesium suljate. With the food of
the 8th day (1st day of the 2d period), 3.001 gramsweregiven. No
effects could be observed. There was no defecation. gth day. 3.998
grams were given. The feces were normal. sothday. 5.002 grams were
given. The feces were very soft. The diarrhea was so pronounced that
the dose was not increased, and the animal was kept on this daily amount
on the succeeding (17-14) days. On the rgth day (last of the period) the
feces were so soft that they ran into the urine receiver, which, however,
had no urine init at the time. The diarrheal effect of 5 grams was greater
in this dog than that of 7 grams in the first dog. The average volume of
urine was also considerably lower.

Matthew Steel 103

THIRD PERIOD. First intermediate period. Days 15-22; March 3-10.
The recovery from the diarrhea was quite rapid. On the rsth day (1st
of this period) the animal did not defecate. On the 16th day, however,
defecation occurred twice; the first mass was like stiff clay, the second
discharge was hard and brittle. The urine did not show a marked tend-
ency to increase in volume until the rgth day (5th of the period). After
that the urinary volume was relatively large.

FourtH PERIOD. Influence of magnesium sulfate when injected subcu-
taneously. Days 23-26; March 11-14. A 40 per cent solution of magne-
sium sulfate was used. Each day’s portion was diluted to 20 cc. before
injection.

Schedule of injections of magnesium sulfate. On the 23d day (1st of the
period), at 12:50 p.m., 1 gram was injected on the right side. The initial
effect was restlessness, followed by sleepiness until about 11 p.m., when
300 cc. of turbid urine were eliminated, after which the animal appeared
to be more lively. On the following morning, she was bright and cheerful.
The feces were normal. 24th day. At 12:55 p.m., we injected 2 grams
on the left side. At 1.07 she was very restless and apprehensive; at 1:15
drowsy. She slept all the afternoon and at 8:15 p.m. urinated 300 cc. of very
turbid urine; then she cuddled down in the corner of the cage as if try-
ing to keep warm, although the room temperature was normal. Atg a.m.
she was lively and cheerful again. The feces were normal. 25th day.
At 12:57 p.m., 4 grams were injected in the right side. As soon as the
injection had been made, the dog rolled on her back and bit viciously for
a moment or two at the place injected, then she cuddled down in the
corner of the cage and could not easily be aroused. She slept all the
afternoon and evening. At 11:15 p.m. she urinated 315 cc.; afterwards
she was lively and more cheerful. The urine had a very heavy sediment
and was alkaline to litmus. On the following morning an abundance of
fine triple phosphate crystals had deposited on the sides of the vessel con-
taining the urine. No feces were passed. 26th day. At 12:55 p.m.,
8 grams were injected on the left side. At 1:10 p.m. she was lying on
the bottom of the cage, panting and breathing rapidly with her tongue
protruding. At 1:20 p.m. she was perfectly motionless as if in heavy
slumber; her respiration was deep and slow. She slept all the afternoon,
refusing to arise when spoken to. The next morning marked depression
Was a conspicuous symptom. The feces were very hard and dry and
were passed with considerable difficulty. There was decided swelling on
the left side at the place where the last injection was made, and the dog
appeared to suffer considerable pain as a consequence of it.

FIFTH PERIOD. Second intermediate (post-subcutaneous injection) period.
Days 27-31; March 15-19. On the first day of this period (27th day)
the dog was very much depressed, and had to be coaxed to eat, whereas
theretofore, she had always been eager for her food. She looked very
apprehensive, shivered a great deal and was sleepy. Towards evening
she was much brighter and more responsive when spoken to. The urine
had a very heavy white sediment. The next day (28th) the animal ate

104 Influence of Magnesium Sulfate on Metabolism

her food with relish. The necrotic mass on the left side had opened dur-
ing the night. The opening was about three inches long and one and a
half inches wide. The wound was enlarged to that size by the dog in her
endeavor to remove dead material.! All of the latter was evidently
swallowed by the animal, for none of it could be found inthe cage. Re-
covery was rapid. Return to normal conditions was complete, before
the beginning of the next. period.

SrxtH PERIOD. Influence of magnesium sulfate injected intravenously.
Days 32-36; March 20-24. With the results of the fourth experiment to
guide us, we proceeded with an intravenous injection as follows: Ten grams
of magnesium sulfate were dissolved in 100 cc. of water, and injected into
a saphenous vein, after anesthetizing locally with cocaine (1 cc. of a 2
percent solution). Asummary of operations and observations is appended:

Schedule of operations. 32d day. At 3:17 p.m., the injection was
started at the rate of about 1 cc. in 50 seconds, and maintained at about
that rate throughout the entire procedure. 3:24. The dog began to
show effects by dropping her head and gagging. Breathing was irregular.
3:28. The dog vomited. (The vomit was fully recovered and fed to
heratrop.m.) 3:35. Breathing was very rapid. Anesthesia increased.

1 The dogs of both metabolism experiments were affected in the same
way locally by the subcutaneous injections of magnesium sulfate, but
necrosis and sloughing occurred more promptly in the dog of the second
metabolism experiment. The following special experiment, on the dog
used in the second experiment (p.98), may be cited as giving additional
evidence of the marked local effects of magnesium sulfate when introduced
under the skin in relatively large doses. ;

First PERIOD. Normal conditions. Days 1-2; December 22-23, 1907.
(See Table VI.)

SeconpD PERIOD. Local influence of subcutaneous doses of magnesium
sulfate. Days 1-2; December 24-25, 1907. Ist day. At12:40p.m., 8
grams, dissolved in 1o cc. of water, were injected subcutaneously on the
right side. The dog exhibited the same symptoms that the other dogs
showed under similar circumstances, i. e., at first he was very restless and
then gradually became drowsy and slept all the afternoon. Late in the
evening, slight edema was noticeable a few inches below the point of injec-
tion. 2d day. At 1 p.m., 10 grams, dissolved in 12 cc. of water, were
injected on the left side. The symptoms exhibited the day before were
again elicited, the only difference being that the anesthetic effect was a
little more pronounced. The next morning, at 10 o’clock, the animal
had apparently entirely recovered from the anesthetic effect. The edema
on the right side had disappeared, but on the lejt side, a few inches below
the place of injection, there was considerable swelling and general soreness.
The urine was pale in color and deposited a heavy sediment of fine triple
phosphate crystals. The feces were dry and brittle.

THIRD PERIOD. After period. Days 3-14; December 26, 1907—Jan-
uary 6, 1908. The dog rapidly recovered from the general soreness.

Matthew Steel 105

4:17. Breathing weakly. 4:21. Deep anesthesia. No conjunctival
reflex. 4:26. Anesthesia profound. Respiration almost entirely sus-
pended. 4:27. Injectionstopped. By this time exactly 80 cc. had run in,
which contained 8 grams of the magnesium compound. 4:28. Breath-
ing had nearly stopped. Respiration was artificially aided for several
minutes by appropriate foreleg and tongue movements and chest com-
pressions. 4:30. Conjunctival reflex sluggish. 4:35. Breathing more
regular. 4:40. Decided conjunctival reflex. 4:53. After stitching the
wound the animal was placed in a cage. She immediately arose but was
too weak to stand. Wassleepy. 5:45. Awake and bright, very respon-
sive when spoken to. 7:30. Lies cuddled up in a corner of the cage
shivering as if cold. Room temperature normal. When spoken to, she
got up. She soon urinated. The urine was very light colored. At to
o’clock on the following morning, the animal was active and apparently
had recovered fully from the anesthetic effects.

The swelling on the left side opened on the 4th day. Sloughing was rela-
tively extensive and an unsightly wound resulted. Healing was in rapid
progress on the 14th day, when observation was discontinued.

TABLE VI.
SPECIAL EXPERIMENT.

IJ. Seconp PEriop. Local influence

| of magnesium suljate injected

I. First periop. Normal Conditions. | subcutaneously.
Days 1-2; December 22-23, 1907. | Days 3-4; December 24-25, 1907.

Number of day (period)... 1 2 Number of day (period).| 3 | 4

Body weight (kilos)...... \7.70|7.83 | Body weight (kilos).....|7.82) 7.68
Dose of MgSO, (grams) . | Dose of MgSO, (grams) .|8.00)10.00
Urine: Volume (cc.)......|302|194 | Urine: Volume (cc.)..... 317 | 405

Specific gravity (10xx).. 17'| 20 |+ Specific gravity (10xx)| 27 | 24

Ill. Turrp reriop. A/jter period.
Days 5-16; December 26, 1907—January 6, 1908.

|

No. of day | | | | |

(period)...| 5 6 Uf lke OF) 10) | EE | 12) | Ss ea os eto
Body weight | |

(kilos).....|7.65]7.61|7.72|7 .82.7.63/7.80 7.72|7.80)/7.78|7.93|8.00|8.04
Urine: Volume | |

(eB) is eee | 280 | 268 | 185 | 185 | 356 | 168 | 304 | 210 318 | 188 | 225 | 267
Specific grav- |

ity (10xx)..| 21 Dee a5 LAL) 25s) 14 We V7 aia I Wes se Mas Lee 4) e210

106 Influence of Magnesium Sulfate on Metabolism

The different degrees in the effects of the intravenous injections in the
two dogs was quite noticeable. In the animal of the third experiment the
anesthesia developed gradually and was not nearly fatal, even when 10
grams had been injected, whereas in the dog of this experiment the effect
was not nearly so strong at first, but came on quickly and was nearly fatal
when only 8 grams had been injected. The rate of injection was some-
what slower in this experiment than in the third experiment. In the latter
experiment the dog did not vomit, although gastric disturbance was
noticeable. In this experiment, as recorded above, vomiting occurred
soon after injection was started. There was also a marked difference in
the nature of the urine from the two animals on the night and day follow-
ing the injections. The first dog excreted urine containing a heavy sedi-
ment. The second eliminated a larger volume, containing only a very
trifling amount of sediment. As in the case of the animalof the third
experiment, the dog in this one recovered rapidly and showed no lasting
effects of the intravenous injection. The feces were normal at each elimi-
nation, certainly not softer than usual.

SEVENTH PERIOD. Second intermediate period. Days 37-45; March 25—-
April 2. During this part of the experiment the general condition of the
animal was noted. Normal conditions appeared to prevail, as recovery
from the previous treatment had been very prompt. The ammonia in the
urine was determined daily and the total amount of nitrogen in the urine
of the whole period of nine days was ascertained. The dog increased in
weight from 10.53 to 10.68 kilos. The average volume of urine was normal.

EIGHTH PERIOD. New Normal period. Days 46-51; April 3 to 8. In
this period the general condition of the dog was noted and analytic data
collected as in an ordinary fore period. In anticipation of further observa-
tions on the effects of intravenous injections, the usual analytic routine
was followed in detail. The weight of the animal continued to rise a little
and was highest on the sist day (the last day of the period). The average
volume of urine was a trifle higher than it had been in the previous
period. By the end of this period, the wound caused by the subcutaneous
injection on the 26th day (the last day of the fourth period) had almost
entirely healed.

NINTH PERIOD. Second intravenous injection of magnesium sulfate.
Days 52-57; April g-14. We attempted to duplicate as closely as possible
all conditions attending the first intravenous injection in this experiment
(6th period, 32d day). Accordingly, 10 grams of magnesium sulfate
were dissolved in roo cc. of water, and injected into a saphenous vein, after
anesthetizing locally with cocaine (1 cc. of a 2 per cent solution). A
summary of the operations and observations is appended.

Schedule of operations. 52d day. At 3:25 p.m., the injection was
started at the rate of about 1 cc. in 47 seconds, and maintained at about
that rate throughout the entire procedure. 3:30. The dog begins to
show effects by panting and breathing rapidly. 3:34. The heart is
beating very rapidly. The dog showed considerable signs of gastric dis-
turbance but did not vomit. 4:00. Shows signs of increasing anesthesia.

a. 6

Matthew Steel 107

Heart beats slowly. Respiration weak. 4:10. Anesthesia increasing
rapidly. Conjunctival reflex sluggish. 4:18. No conjunctival reflex.
Breathing in gasps. 4:23. Respiration almost entirely inhibited. 4:24.
Profound anesthesia. Injection stopped. By this time exactly 70 cc.
had run in, which volume contained 7 grams of magnesium sulfate. 4:30.
Respiration more regular. 4:34. Slight conjunctival reflex. a2ss:
After closure of the wound, the animal was placed in the cage, where she
remained practically motionless for about 20 minutes and then stood up,
but seemed very weak. In a minute she lay down again and went to
sleep. While sleeping one of her paws twitched spasmodically, 9:15.
She urinated 600 cc. of very light, clear urine; afterwards was bright and
cheerful. Later in the night she urinated 65 cc. of urine. This was like-
wise very clear. There was no sediment. The specific gravity of the
first portion eliminated (the 600 cc.) was 1016 and the reaction was ampho-
teric to litmus. The specific gravity of the second portion (65 cc.) was
1023, and its reaction decidedly acid. When the two portions were united
the specific gravity was 1017 and the reaction faintly acid to litmus. The
dog’s weight fell considerably as was to be expected, from the fact that a
particularly large volume of urine had been eliminated. The weight of
the dog steadily rose again, however, and by the 57th day (last of the
period) was at the normal point. No lasting effects of the injection were
noticeable. The feces at each elimination were normal, perhaps drier
and more brittle than usual.

TENTH PERIOD. Third intermediate period. Days 58-63; April 15-20.
Only the general condition of the animal was noted in this period. Her
weight remained very constant, between 10.76 and 10.79 kilos. The vol-
ume of urine was comparatively large and the specific gravity uniform.

ELEVENTH PERIOD. Influence of an intravenous injection of magnesium
sulfate on the magnesium content of the feces. Days 64-73; April 21-30.
Before discontinuing the experiment we concluded to conduct a final
intravenous injection with a view primarily of determining some facts
tegarding the excretion of magnesium in the feces after introduction of a
soluble salt into the circulation. Special care was taken to prevent con-
tact between urine and feces. Ten grams of magnesium sulfate were
dissolved in 100 cc. of water and the solution injected into a saphenous
vein after locally anesthetizing with cocaine (1 cc. of a 2 per cent solution).
| Schedule of operations. 64th day. At 4:05 p.m., the injection was

started at the rate of about 1 cc. in 36 seconds, and maintained at approxi-
mately that rate throughout the entire procedure. 4:18. The dog showed
marked signs of gastric disturbance. 4:20. Vomited. 4:24. Heart
beating very rapidly. 4:26. Increased signs of anesthesia. 4:29. Heart
fluttering. Conjunctival reflex sluggish. 4:39. Marked abdominal
respiration and increased fluttering of the heart. 4:41. No conjunctival
Teflex. 4:50. Profound anesthesia. Respiration almost entirely sus-
pended. 4:55. Injection stopped. By this time exactly 85 cc. had run
in, which volume contained 8.5 grams of magnesium sulfate. Recovery
Wasprompt. At 5:10 after stitching the lips of the wound the dog was put

108 Influence of Magnesium Sulfate on Metabolism

in a cage. At 5:18 the dog stood up in the cage and began licking the
wound; after a moment she lay down again and went to sleep. As in
the former experience one of the paws twitched spasmodically for some
time at this stage. 9:30. The dog was bright and cheerful, and urinated
365 cc. of very clear urine. The vomit was then offered. Sheateit ail
very readily. Late at night she urinated 190 cc. of very turbid urine.
The reaction of the first portion eliminated (the 365 cc.) was alkaline to
litmus. The reaction of the second portion (the 190 cc.) was amphoteric
to litmus. When the two portions were united the reaction was faintly
alkaline. During most of the period the feces were analyzed daily for
magnesium.

ANALYTIC RESULTS. The essential data of this experiment
are presented in Tables VII-IX.

DiscUSSION OF THE RESULTS. Changes in the weight of the
dog. The weight of the dog was comparatively uniform until
the fifth period. On the 28th day (the second day of the post-
subcutaneous injection period there was a noticeable fall in
weight. On the next day the decrease was unusual. The decline
continued for a few days and reached the lowest point in the ex-
periment on the 32d day (first day of the first intravenous injec-
tion period). The weight then very slowly returned by the end
of the eighth period (51st day) to that at the beginning of the
experiment. It fell sharply immediately after each of the sub-
sequent intravenous injections but rapidly returned to the normal
each time (gth and 11th periods).

Administration per os was without effect on the weight of the
dog, but there was a marked decline in weight in the period
immediately after the subcutaneous injection. After the intra-
venous injections there was always a sharp decrease in weight
which was most pronounced within twenty-four hours after the
treatment, and which was not cumulative, as appeared to be the
case after subcutaneous administration.

Fluctuations in the daily volume of urine. The daily volume
of urine fluctuated between 270 cc. on the secoud day (33d) of
the first intravenous injection period (6th) and 665 cc. on the
first day (52d) of the second intravenous injection period (9th).
The average daily quantity was relatively high during the period
(sth) succeeding the subcutaneous injection and during the last
period. Administration per os appeared to cause perceptibly
decreased volume of urine, evidently in compensation for the —

Matthew Steel 109

loss of water in the diarrheal stools. The largest daily volumes
were eliminated on the days of intravenous injection.

Alterations of the specific gravity of the urine. The extreme
registrations of specific gravity of the urine were ro12 in the
opening period and 1038 in the fourth period (subcutaneous
injection period). The former value pertains to a relatively
high volume and the latter to a comparatively small one, so that
the figures quoted are not particularly significant. Ina general
way, specific gravity was higher after administration of magne-
sium sulfate per os then during the normal period. The reduced
volume of urine during the second period fully explains that
fact. The higher specific gravity of the volumes excreted shortly
after subcutaneous injection (4th period) were seemingly due
chiefly to the eliminated magnesium sulfate. The fluctuations of
specific gravity during the last seven periods were slight. Few of
the figures went above the normal ones. Although the specific
gravity of the urine eliminated on the first day of each intra-
venous injection period was somewhat above normal, the diure-
sis that immediately resulted after such injections prevented the
figures for specific gravity from registering at the higher points
they would otherwise have attained, because of the prompt
elimination of the injected magnesium sulfate.

Effects on the elimination of feces. Administration of magne-
sium sulfate per os made the feces softer, as has already been
noted. The injections, on the other hand had no effect at all
on their consistency or else actually rendered the fecal discharges
drier and more brittle. In one case after intravenous injection,
as has already been noted, the feces were passed with great diff-
culty, whether from inhibition of the normal muscular movements
or because of particular dryness alone, can hardly be said.

The average daily quantity of dry fecal matter was somewhat
greater than normal in the second and all subsequent periods.
In a general way the figures for dry fecal matter indicate specially
increased daily output promptly after administration of magne-
sium sulfate per os, subcutaneously or intravenously. In only
one case was the increased amount of dry fecal matter accom-
panied by an increased content of nitrogen in the feces. The
average daily elimination of nitrogen in the feces per period was
quite uniform throughout the entire experiment, although slightly
subnormal in all but the post-subcutaneous and first intravenous
injection periods.

110 Influence of Magnesium Sulfate on Metabolism

Effect on the total nitrogen of the excreta. The datain Table IX
show that our animal was not in nitrogenous equilibrium at the
beginning of the experiment. The retention balance of nitrogen
diminished per period until the fourth (subcutaneous injection
period) when the dog was in practical nitrogenous equilibrium.
In the fifth or post-subcutaneous injection period, perhaps as an
“after effect,’’ there was a large minus balance. The develop-
ment and opening of the abscess may have influenced the out-
come. Retention was the rule again in the subsequent periods,
with progressive period increase toward the highest plus balance
in the seventh period (period after the first intravenous injection)
and regular recession to a small plus balance in the last period.

These results might be assumed to indicate that the doses of
magnesium sulfate increased the total elimination of nitrogen,
but the last four plus balances, those in the sixth and seventh
periods particularly, make such an inference difficuit to maintain.
The explanation of the outcome in this regard is not at all clear.

Effects on the elimination of total nitrogen in the urine. In
almost perfect harmony with the trend of the period balances,
as given in Table VIII, there was a progressive increase in the
daily average amount of urinary nitrogen to the maximum excre-
tion in the fifth period (post-subcutaneous injection period),
succeeding which there was cumulatively decreased elimination
until the last or second intravenous injection period, in which
there was again a comparative increase. The significance of
these results is not at all clear, although the development and
opening of the abscess probably affected the data appreciably.
In this experiment, and in the first metabolism experiment, the
maximal period outputs of urinary nitrogen occurred during the
fifth or post-subcutaneous injection period. In the preceding
metabolism experiment the maximal excretion of urinary total
nitrogen was associated with the smallest average daily volume
of urine for the period. In this experiment, on the other hand,
the maximal elimination of urinary total nitrogen occurred dur-
ing the period in which the largest volumes of urine were excreted.

Effects on the partition of the total urinary nitrogen. The abso-
lute excretion of urea nitrogen ran parallel with the urinary total
nitrogen except in the eighth period. There was also the same
general harmony in the relative elimination. The trend in urea
elimination during the first five periods was practically the

Matthew Steel LI

reverse of that shown in the previous metabolism experiment.
Ammonia nitrogen was eliminated in large absolute amounts in
the first two dosage periods and in its maximal quantity during
the post-subcutaneous injection period (fifth). During the last
four periods the elimination was uniform and above the normal.
The percentage figures in Table VIII agree with those for the
absolute eliminations of ammonia. Creatinin, absolutely, was
much the same quantitatively for each period, with the greatest
elimination in the last period. Relatively the smallest excretion
of creatinin occurred in the post-subcutaneous injection period
(sth). There was relative period increase of creatinin to the
third period, decrease to the minimum percentage value in the
fifth period, and rise to the maximum percentage excretion at
the end. Allantoin, both absolutely and relatively, was in-
creased to the maximum values in the subcutaneous injection
period (4th) and decreased slightly toward the end of the ex-
periment. In all the periods except the second (per os admin-
istration) the quantity of allantion was above normal. Uric acid
was much the same in amount in each period. It was increased
somewhat during the last two periods. Relatively it was least in
amount in the post-subcutaneous injection period (5th) and highest
in the last two periods. Purin bases, both absolutely and rela-
tively, decreased in quantity sharply to the third period (per os),
rose somewhat though still remaining at a subnormal point in
the fifth (post-subcutaneous) and fell to very low amounts during
the last two periods. The nitrogen in undetermined products
fluctuated in absolute amounts to subnormal values in all but
the third (ante-subcutaneous) and fifth (post-subcutaneous)
periods. The absolute amounts of undetermined nitrogen in the
last two periods were especially small. Relatively there was a
fall to a low figure for undetermined nitrogen in the fourth (sub-
cutaneous injection) period, a rise nearly to the normal percentage
in the sixth (first intravenous injection) period, and a sharp fall
to very low figures at the end.

Elimination of magnesium per rectum. The average daily
eliminations of magnesium in the tenth and eleventh periods
(p. 107), were the following: Tenth period, 6 days, 0.096
gram; eleventh period (after intravenous injection of 8.5 grams
of MgSO,); 5 days, 0.103 gram; 7 days, 0.096 gram; ro days,
0.096 gram.

I12

NOP WD

| Day No.

TABLE VII.

Some of the daily records of the fourth experiment.
The second metabolism experiment.

FIRST PERIOD.

Body
weight.

kilos.

10.73
10.69
10.68
10.69
10.76
10.68
10.65

Dose of
MgSO,.

grams.

Normal conditions.

URINE.

‘ : di d & a | Be
rea ase ars ee B'S. ee
> DM bo a | Om |) —is
ce. 10xx. grams. grams. gram gram.
435 12 0.190
334 14 8 887 7.068 ; 0.158 | 0.144
346 14 0.148
340 16 9.383 6.986 | 0.183 | 0.1385
322 | 16 0.144
442 14 0.148
390*, 13 14.664 8.737 | 0.340 | 0.122

Influence of Magnesium Sulfate on Metabolism

Days 1-7; February 18-24, 1908.

FECES.

Dry
weight.

grams.
|

48 .2
30.5
29.0
tS 0
43.9
29.0

|

* By mistake 32 cc. of urine, which should have been added to the 8th day portion, were

added to that of the 7th day, thereby changing the volume from 358 to 390 cc. and the
weight of the dog from 10.68 to 10.65 kilos.

SECOND PERIOD.

Influence of magnesium sulfate administered per os.
Days 8-14; February 25—March 2.

10.73
10.78
10.76
10.68
10.75
10.73
10.73

3.001
3.998
5.002
5.004
5.000
5.002
5.001

310
312
330
390
300
346
345

We
19
17
16
18
20
18

9.371

| 10.007

14.737

THIRD PERIOD.

15
16
17
18
us)
20
21
22

10.77
10.78
Oi
10.78
10.79
10.75
10.68
10.69

330
285
392
315
385
408
455

380 |

17
18
17
16
15
14
14
14

10.763

10.776

10.534

10.878

|

First intermediate period.

10.382

0.183 |

0.208

0.328

0.202
0.252
0.279
0.278
0.257
0.293
0.288

Days 15-22. March 3-10.

qooocooq}cocqc &

.236
.189
.168

|
|
| 53.

Matthew Steel 113

TABLE VII—CONTINUED.

FourtH Periop. Influence of magnesium sulfate injected subcutaneously.
Days 23-26; March 11-14.

URINE. FECES.

Body | Dose of | : : s abe
6 | weight. | MgSOx,. rs) os, ae A 38 ag
Z g es Sb 3 i 5 o ea)
» RS ok A t= £2 32 fe es
A pias | a Se <5 Es
kilos. | grams. | cc. | 10zz. | grams. | grams. | gram. gram. | grams.
De, WO AL MOOR Sa ET | 0.254 45.9
24 | 10.68 2.00 | 415 15 11.079 8.113 0.237 | 0.266 16.2
2a LO wt 4.00 | 395 | 23 0.268 | 0
26 | 10.78 8.00 | 280 | 38 Miley 9.252 | 0.197 | 0.203 |44.9

27 | 10.71 | 450 | 23 | 0.395 | 33.5
28 | 10.64 400 | 15 | 15.817 | 11.914 | 0.263 | 0.444 | 15.0
29 | 10.53 433 | 21 | | 0.636 | 32.9
30 | 10.54 | 389 | 17 | 0.348 | 42.5
31 | 10.52 419 | 16 | 21.375 | 18.006 | 4 0.380 | 27.1
J |
SrxtH Periop. Influence of magnesium sulfate injected intravenously.
Days 32-36; March 20-24.
|
32 | 10.47 | 8.00 | 460| 23 0.184 | 0
33 | 10.48 270 | 21 | 11.698 | 8.852 | 0.239 || 0.258) 63:5
34 | 10.48 325 | 19 0.236 | 26.0
ab.) 10.53. | 362 | 15 0. 219s) io
36 | 10.49 | 340 | 15 | 15.444 | 10.701 | 0.345 | 0.204 | 54.4

114 Influence of Magnesium Sulfate on Metabolism
TABLE VII—-CONTINUED.

SEVENTH PERIOD. Third intermediate period. Days 37-45; March 25-April 2.

URINE. FECES,

; > 3. 4

g|weisnt. [Meso | § | gs | g8 | of | £8 | BE | #

2 3°) 83 | gf | Sa | 38-4

A > | ae 3 5 OF a5 A
kilos. grams. ce. 10xz. grams. grams. gram. gram. | grams.

37 | 10.53 364 15 0.204 0
38 | 10.58 303 bz 0.214 | 56.9

39 | 10.64 350 16 0.243 0
40 | 10.65 | 375 14 0.218 | 25.7
41 | 10.64 435 | 14 0.237 | 29.8
42 | 10.64 390 13 0.211 | 52.0

43 | 10.68 370 13 0.198 0
44 | 10.67 370 13 0.206 | 47.5

45 | 10.68 ao 14 | 43.70 0.206 | 26.0

EIGHTH PERIOD. New normal period. Days 46-51; April 3-8.

| |
46 | 10.70 395 | 13 0..185 0
47 | 10.74 360 14 9.233 7.636 | 0.253 | 0.200 | 37.5
48 | 10.70 465 | 13 0.225 | 53.0
28) We 10), 7A 405 14 10.049 | 8.743 | 0.249 | 0.208 | 12.8
50 | 10.75 310 13 0.202 | 24.9
51 |} 10.78 330 15 9.128 7.738 | 0.241 | 0.206 | 40.7

NINTH PERIOD. Second intravenous injection period. Days 52-57; April 9-14.

52 | 10.57 7.00 | 665 L7G 0.141 | 27.0
53 | 10.59 355 14 10.968 | 9.589 | 0.290 | 0.272 | 27.9
54 | 10.65 365 14 0.223 | 24.5
55 | 10.71 368 15 10.788 | 9.039 | 0.273 | 0.237 | 25.5
56 | 10.76 345 14 ; 0.202 | 52.5
57 | 10.78 375 14 10.477 | 9.035 | 0.245 | 0.233 | 21.5

:

Matthew Steel 115

TABLE VII—CONTINUED.

TENTH PERIOD. Fourth intermediate period. Days 58-63; April 15-20.

Number of the day(period)| 58 59 60 2 | 68 63
Body weight (kilos) ......| 10.77 | 10.79 | 10.76 a 7% a ee 110.77
Urine: Volume (cc.)....... 403 | 345 | 416 | 410 | 400 | 390
Specific gravity (10xx).; 14 14 13 13 13 14
Feces: weight (grams)....| 17.7 | 28.9 | 27.0 | 18.4 | 18.6 [64.0
Magnesium (gram)....... 0.576*

*Total for days 58-63,

ELEVENTH PERIOD. Influence of an intravenous injection of magnesium sul-
jate on the content of magnesium in the feces.
Days 64-73; April 21-30.

|

Number of day

Ppetionyes.-|G4-| 6b | 66 | 67 | 68 | 69 | 70 | 7L | % | @
Body weight

Galasys.. .... 10.59/10.64/10.69)10.72|10.75 10.77|10.78)10.77|10.78)10.78
Dose of MgSO,

(grams) ....! 8.50 | |
Urine: volume

(Ue) ar 555 | 352 ; 300 | 338 | 377 | 372 | 310 | 400 ; 374 | 384

Sp. gr. (10xx).| 19 14 15 15 14 14 14 | 13 14 14
Feces: weight

(grams)..... 0 (26.3 |49.0 |30.0 |37.5 |31.5 |27.0 |49.2 |23.0 |21.0
Magnesium
(gram)....; O* |0.137/0.146/0.105|0.127 156+ 0.289

*No defecation.
{Days 69-70.
{Days 71-73.

0°?

P9ULIG IU)

cress sss snousABizUl rOSIIN

Ge°0z ZL°0 62°0 s¢e°0 cls GO°SL ces IA
80°9T 28°0 Sz°0 ce°o 6S°T T6°¢ SP 08 soe es SYBTIPOULIOZUT PUODEG | A
89° FI $9°0 3e°O 9F°O $6'°T CPP OG LL sr eeeesssssnoauBynogns FOSIW | AT
& O€ 06 69°0 82°0 $20 TE°s 6P°e 6E° LL verses “SYBIPIULIOFAL ISIN |TIT
ee SOLOGA" BPE) Ger Or!) GE ONh cba: baver GU SESGE ILO oie eae emi enna Re ice 80 42d YOSSW II
o Te" ez SLT 9£°0 cl‘0 90° ze ©2769 Bua Apmttacersn ses eeessreauON IT
a) quao wad |*2ua0 dad)quao wad |*yuaa sad|"7uao sad|"yuaa sad |"yuaa sad "7
< *UaboujvUu 70}02 9Y}2 07 UOLPM)A—
= POF 0 | 866 GIP T¢¢'0 610°0 | #20°0 | Z10°0 | SET‘O | 81Z°0 | II9'F GlE°s SOOO es |i? ae eee © “""SsnousABIyULPOSSW | XT
c chr 0 | T 82 SLE see 0 OT0°O | 120°0 | 810°O | FZT°O | F0Z°0 | 6IO'F GEL’ P S60E 9) Seance ein ear ees [euriou MON |TITA
) TPP 'O | 96 O06 GIz°0 998° Ph HGR? la a Dee Fen ePEIpoULIozUI PATGT, jITA
6cS ‘0 | 8°86 TSE FOL T cv0'0 | 910°0 | 610°0 | ZITO | 022'0 | II6's 8cr'S GIZ'9 |i CCC Tt snousasiqUuLrOgsy ITA
At €0¢°0 | 2°0€ OTP 682 °0 190°0 | 610°0 | 920°0 | 6110 | ThF'O | F86°S SEP L CTLC2OM fica een “*o}BIPoUL1e}UT pUOd9g | A
& 98h 0 | 8°92 99¢ 6280 9£0°0 | 810°0 | 920°0 | GOTO | 8Fz'0 | IFE'F 009°S GIZ9 | ct tt snoaueynogns ’FOSsW | AT
3 GSP O | 92 69¢ O61 T 6£0°0 | SI0°0 | &10°'0 | FZT'O | Z6T°O | 862° 69¢°¢ cIé°9 eepe od ooo" OPVBIPOULIO}Ul ISIN (TTT
Y OPP 0 | 2°82 LEE 416°0 éL40°0 | #10°0 | 900°0 | SOTO | F92°0 | LEF'e e238) F. SOO LOR baw tars Sa ies so ad ¥OS3W {IT
L6r°0 | 8° SG 69€ 460° T ¥80°0 | 410°0 | 200°0 | 460°0 | ZF 'O | 992° GOL °F G00 20) 4S ae ie “o***[eULIoN |T
2 ‘sabpsaan fijvoqg
“DH G8L°S | 6 SLT | E2h% | 6OL'S? 9I1°O | Sh1°O | FOL'O | 808°0 | 80E'T | E99°z2z | TSs°sE | OLG'9E |** °° °° °° SnoTeABIZUTFORSZW |XT
S 099°% | 6° 89T | G9de | 8za°S? SQOKO OGIO HeLZOTO NSE ON O2GuD. |) Abba PeemOLPESGa|MOLGCOE ls siesta eens “ss *[eur10u MON |TITA
iy 2496°€ | 6° LES | SEEE 286° T QO EP ESE CES SIs eae tan “*o7VeIpaulezUr ply, TTA
(0) G19'°% | 6°ShT | LZGLT | 1eg°¢ OLS107 | 162050") 7605 OF | FSSs0) | LOLs | ReSSeG GrileLont SZONLEN i ee bee eas “*snouaABI}ULPOSSA, TA
= GI¢°% | O' IST | G80c | Fr6'S L0g°0 | ¥60°0 | TET'O | §6S°0 | 06'S | 026°6S | Z6T'LZE | GLO TE | °° °° 89BIpeurte}ur puosas) | A
ste cr6' T | O'LOT | SOFT | B8a'E cyIt°O | 220°0 | 9010 | FEh°O | 166°0 | SOE ZT | TOP ce | O98 FS | snosueynogns FOSS | AT
re) 819°€ | T° IT@ | O¢62 | 029°6 PSGuOn Secale ON cORsOl lh C6GROneeeSied MES SeOGa a eSOm Gra OG Grn een cette ncne 6 9}PBIPSOULIO,UL SINT |TIT
o Z80°€ | T° 10Z | S9E% | 6E8'°9 S0SS0F LOIS Os | @6E0n0n SG RZOs SGESels GOO mrGn may et Gellas Omciras ecm wee es et so dad OSS [IT
O L6h°€ | 9'O8T | LL49% | 629°2 LSQsOn SLieOs| 200.05 eSOuOn Lenin me LGZnGG. \shSOnGor||scGUNGi et tt cuen ss uhcmeen sees" [BULlON |T
. *swoi8 | ‘supsB 90 | sswpi6 | wps6 | -wos6 | ‘wo18 | -woib | -swpsb | -swns6 | -swo1b | -swo.6 |
= me os | | ee
Pee es (oa “(rs [aby lanes este el ceca
= 3 4 Bi lnseo@ ellen e 5 e B »
=a o. ® a8 o 5. 9. = g *ued01}1U|*uadOI4TU| aman oe
No) a 33 % ps FE eee a: 130], | [eIoy, | FPUCD N
re ee aes ss ee |
a P jO WO} 049 Ul UaTOI4IN |
“saoga *ANIUQ “dood “adoludd

‘po.ad yopa 10{ s]0}0 J,

“LNGWIYEdXo HLYNOW

‘TWA FIAVL

Matthew Steel 117

TABLE IX.

FourTH EXPERIMENT. Analytic totals and averages for nitrogen in each period |

Period No. .. I Il Ill IV v VI Wait VIII IX
7 8 5 9 6

| grams. | grams. | grams.| grams. | grams. | grams. | grams. | grams. | grams.
INaetol Meee 42.04! 42.04) 49.72/24.86 | 31.08] 31.08} 55.94) 36.57) 36.57
Excreta...| 36.42) 37.20) 46.57) 24.35) 39.71) 29.75) 47.67, 31.07| 35.04
Balance... .|+ 5.62|/+ 4.84/+3.15/+0.51 —8.63|/+ 1.33/+8.27/4+-5.50/+1.53

TABLE X.
Some of the daily records of the fijth experiment.
First periop. Normal conditions. Days 1-9; March 30-April 8, 1908.

Number of the day |

co ray 2 3 4 5 6 if 8 9
Body weight (kilos). .|15.83,15.77/15.88|15.74|15.73/15.88/15.9415.95 15.96
Urine: Volume (cc.) .| 275 | 470 | 320 | 540 | 370 | 340 | 370 | 325 | 270
Specific gravity |

Sreconp preriop. Intramuscular injection of magnesium sulfate.
Days 10-18; April 9-17.

Number of the day, |
(eros. ais. ee. 1074) 14 12) AB PA PG eee ees
Body weight (kilos). ./15.30|/15.37/15.90/15.81,15.85 16 .20)15.90|15.92|15.95
Dose of MgSO,

OMANI) o: oreke ve rags». < 12.00 |
Urine volume: (cc.)..| 860 | 220 | O | 455 | 350} O | 680 | 495 | 308
Specific gravity |

CL er 27 | 26 36 | 35 | 24 |) 26. |. 28

118 Influence of Magnesium Sulfate on Metabolism

VII. FIFTH EXPERIMENT. DOES THE INTRAMUSCULAR INJECTION
OF LARGE DOSES OF MAGNESIUM SULFATE INTO DOGS
CAUSE THE DEVELOPMENT OF ABSCESSES?

Meltzer and Auer’ found that in rabbits intramuscular injec-
tions of concentrated solutions of magnesium salts caused the
development of abscesses. Subcutaneous injections of such
solutions into rabbits did not cause abscesses. In dogs, however,
Meltzer and Auer observed the reverse effects, i.e., abscesses after
subcutaneous, but not after intramuscular, injections. In our
experiments relatively large doses introduced subcutaneously
always caused necrosis and sloughing in dogs. We took no
aseptic precautions, however, in connection with the hypodermic
treatment.

The outcome of the following experiment confirms the related
observations by Meltzer and Auer:

A healthy, long haired, male dog weighing about 15.50 kilos was given
a daily diet consisting of hashed lean beef, 225 grams; cracker meal, 60
grams; lard, 45 grams, bone ash, 30 grams; and 525 cc. of water.

First PERIOD. Normal conditions. Days 1-9; March 30-April 8,
1908. The dog was kept on the above-mentioned diet for twenty-two
days before the injection was made. Table X gives data pertaining to
the last nine of the twenty-two days referred to.

SECOND PERIOD. Intramuscular injection of magnesium sulfate. Days
10-18; April g-17._ On the 23d day (first of the injection period), 12
grams of magnesium sulfate, dissolved.in 30 cc. of water, were injected
into the right hind leg. Immediately after the injection the dog was very
restless. He squirmed about on the bottom of his cage, panted excessively
for about half an hour, and then became drowsy and slept practically all
evening, rising occasionally to urinate. On the next day he was still
sleepy and was so lame that he would not rise of his own accord. During
the night he urinated 860 cc. of very turbid urine; reaction to litmus,
amphoteric. The dog’s lameness continued for several days. There was
some temporary swelling, but no abscess formation nor any sloughing.
There was no diarrhea at any time. Recovery was prompt.’

VIII. GENERAL COMPARISON AND DISCUSSION OF THE RESULTS.

A comparative examination of the data of all our experiments
warrants the following condensed summary of the results of the
treatment with magnesium sulfate.

'Meltzer and Auer: Amer, Journ. of Physiol., xiv, p. 374, 1905.

* A second experiment under essentially the same conditions gave the
same negative result.

Matthew Steel 119

Bopy weiIcuHT. On a rising scale of weight in the normal
period of the first experiment, administration per os and some of
those subcutaneously were accompanied by further increase in
the body weight. The larger doses subcutaneously caused
diuresis and decreased weight. There was increased weight in
the post-subcutaneous period. Ona rising scale of weight in the
normal period of the special experiment referred to in the footnote
on p. 104, administration subcutaneously caused diuresis and
decreased weight. In the post-subcutaneous injection period
body weight returned to and rose above normal. In the fourth
experiment on a falling normal scale, administration per os and
subcutaneously were associated with stationary or perhaps slightly
increased weight, but there was decreased weight in the post-
subcutaneous period with prompt recovery. Intravenous injec-
tion always caused immediate diuresis and decrease in body
weight, with prompt recovery in a few days. In the second
experiment the largest doses per os caused sufficient diarrhea to
reduce body weight slightly. In the fifth experiment, intra-
muscular injection caused immediate diuresis and reduction of
body weight. Recovery of weight was prompt.

VOLUME OF THE URINE. Disregarding normal variations in
quantity, the volume of urine was generally decreased by adminis-
tration per os, evidently because of the resultant diarrhea. In
the second metabolism experiment there was apparently no
effect on volume. Urinary volume was usually increased, at
once or in a few days, by the larger subcutaneous doses. There
was immediate diuresis after intravenous or intramuscular
injections followed by very prompt compensatory reduction of
volume. Asarule diuretic effects were soon completely balanced
by corresponding retentions.

SPECIFIC GRAVITY OF THE URINE. Aside from its normal
fluctuations specific gravity was relatively high after subcutaneous
injections of large doses per os, obviously because of the rapid
elimination (of the magnesium compound) that occurred usually
without special diuretic effect. The marked diuresis that followed
intramuscular or intravenous injections prevented special rise
in specific gravity after such introductions. The specific gravity
was usually high for a few days after the cessation of diuresis,
evidently because of compensatory retention of water, as shown
by the simultaneous elimination of relatively small urine volumes.

120 Influence of Magnesium Sulfate on Metabolism

CONSISTENCY, AMOUNT, AND NITROGEN-CONTENT OF THE FECES.
Administration of relatively large doses per os caused diarrhea.
Subcutaneous intramuscular as well as intravenous injections
failed to elicit diarrhea; on the contrary, such introductions
seemed to make the feces drier and harder.!

In the first metabolism experiment the dry fecal matter in-
creased per period to the maximum average daily elimination
in the subcutaneous injection period (4th) and decreased to sub-
normal amounts in the post-subcutaneous injection and supple-
mentary periods. In the second metabolism experiment the
average daily quantity of dry matter in the fecal discharges was
above the normal elimination in the second (administration per
os) and all subsequent periods. The greatest eliminations in the
second metabolism experiment appeared to occur as a rule after
treatment with magnesium sulfate, whatever its channel of intro-
duction.

The average daily content of nitrogen in the feces was ascer-
tained in only one experiment—the second metabolism experi-
ment. See the remarks in this connection on p. tog.

TOTAL NITROGEN BALANCE. The nitrogen balance was deter-
mined in the second metabolism experiment, but in no other.
See Table IX; also p. r1o.

TOTAL URINARY NITROGEN. In all periods except one? the
total amount of recorded urinary nitrogen was above that of the
normal period. In both metabolism experiments the maximal
period excretion of urinary nitrogen occurred in the post-hypo-
dermic injection period after cumulative increase to that period.
Possibly the local disintegrating effects of the subcutaneous
treatment was the prime factor in this result. Diuresis had
evidently nothing to do with the outcome in this respect in the
first metabolism experiment, but may have greatly influenced
the result in the second metabolism experiment. Administra-
tion per os had no immediate effect if any on the elimination of
total nitrogen. The high excretion values for total nitrogen,
after subcutaneous injection, were not fully maintained in the
subsequent intravenous injection periods, although as stated

‘There were possible exceptions after subcutaneous injection in the

fourth period of the first metabolism experiment (p. 89).
? First metabolism experiment, administration per os (second period).

Matthew Steel 121

above excretion of total nitrogen was still supernormal in each
of them, and there was evidence of increased output of nitrogen
as a consequence of the treatment intravenously.

PARTITION OF THE TOTAL URINARY NITROGEN. The compara-
tive effects on the distribution of nitrogen among the leading
urinary nitrogenous constituents is shown at a glance in Table
XI, where the + sign meansa value above that for the first normal
period; a — sign shows a subnormal quantity; an = sign stands
for no appreciable increase or decrease, as compared with the
normal value and an asterisk indicates that no determination
was made. A blank signifies that the experiment was not dupli-
cated and that no comparative result was attained.

Ammonia. Ina general way the tables show that ammonia
was increased both absolutely and relatively in every period
after the normal one.

Urea. The absolute amounts of urea were appreciably above
normal in nearly every period although the relative excretion
was subnormal after the period of per os administration in the
third experiment.

Creatinin was essentially the same as urea in appearing in
larger absolute amounts than the normal in practically all post-
normal periods. Its relative excretion showed somewhat less
uniformity than urea’s.

Allantoin was like urea in passing into the urine in supernormal
amounts in nearly all the post-normal periods. Relatively its
excretion tendency was also similar to urea’s.

Uric acid was increased both absolutely and relatively in a
bare majority of the post-normal periods.

Purin bases. The nitrogen of purin bases exhibited diametric-
ally opposite tendencies in the two experiments, both absolutely
and relatively. Purin base nitrogen was in the main increased
absolutely and relatively in Experiment III, but decreased in both
respects in Experiment VI.

Undetermined. The remarks made about the nitrogen of purin
bases apply to that of the undetermined products.

Influence of Magnesium Sulfate on Metabolism

PIP

—

ee Weer hee a et 6 ORAS AER a

= is ae BE He ie + "+ A[snousaesjzul OSs XI
= 7 a oe i ds e Vs sseseeessremrou MON] IITA
* * *k * * a k ‘** operpoulseyul pag y| TTA
= = ae ae + ae +- "+ AJsnousarsyul "OSST] TA
-~+;),-4+]/-+4+]+ +4 )--)] +4 4 )4- *-ayerpeuttayut puoveg) 7
ea ea es pel Blea mete me ete +e + — ‘A[snoouvynoqns *OSSI| AT
= lls ras rots NY See = te Re ete grate fetes wel a eee Mie aYBIPIULIAzyUT SILA) TI]
eee aul, = Se ap Ae seit sie SSE |S © oes "80 vad ’QS3| IT
“sanyou aayppy g

= at ah ae ae = + + * Ajsnousaesjyut "OSS|) XI
ae =F We th SLi ohh oF an ees s )* RUIOU MON) TTA
* * * * * + * ag “""* eyeIpeulteyal part}! TTA
i. = = a alts ae ate + * AJsnousaerjut OSS) TA
—~+]),—- 4+] 4 +] + t+) + = | + + | + + | + + | coe peuttozuy paooag) A
eat Thai hai Siri acte rigiie at Sees Dees a Po + + | 4jsnooweynoqns’OS3W| AT
Stor ate ag ok He RUE ete Gir ae Fg se eal QPEIPSULIOFUT ISAT) TIT
ce Ee ee | ec a ee eo ad Sela
Satie TAY Senin TAS Ut |) en VNU DE ee oe WAG aut Te eT Aer ID IA III Orman ae
-lejyepuy ‘aseq ULINdg “plo® o1() “UlozpURI LV “ULUT} Ball) BIuOUULYy “BOl/) [BIOL he :

‘NHDOULIN AUVNIUN : SS

“‘sjuUNOWD ajnjosqyY YP

‘(TA pun JI] stuauiwadxg) sjuaursadxa

UsYognjau puodvas pup psy ayy wr Uabospu hupur_sn yoj07 fo uorpjsvd aayojas pup ajnjosqo ay} Uo DIDp aaynipduos fo fiumumungy

‘IX GITAVL

Matthew Steel ee)

GENERAL COMPARISONS. Considering the data pertaining to
the partition of nitrogen from the standpoint of direct agreement
in the two experiments it is obvious that the increased elimina-
tion of ammonia in all the post-normal periods is the most strik-
ing and perfect concordance. That this increase was largely due
to the combination of ammonia with magnesium in the form of
ammonio-magnesium phosphate appears to be certain. The
increase indicated by our figures in the third experiment was
doubtless below the full increase that would have been exhibited
by a perfect method of determining ammonia nitrogen when in the
form of “‘triple phosphate.” In the sixth experiment, however,
possibly all the ammonia was recovered by the modified method
employed.

The nitrogen of urea, creatinin and allantoin, taken collectively
appeared in increased amounts, both absolutely and relatively
in a majority of the post-normal periods.

The results for nitrogen of uric acid, purin bases and undeter-
mined products were too discordant to permit of any such deduc-
tion one way or the other.

IX. SUMMARY OF GENERAL CONCLUSIONS.

Disregarding matters of detail that have been referred to in
connection with the discussion of the data of each experiment,
the following conclusions sum up the main results of the research:

In these experiments, in which relatively large doses of magne-
sium sulfate were given to dogs, abscesses and sloughing followed
subcutaneous injections, but were not caused by intramuscular
or intravenous injections nor by administration per os. Very
large doses of magnesium sulfate could repeatedly be injected
intravenously without causing death, when care was taken to
conduct the process slowly.

Fluctuations in the weight of the animals, as well as in the
volume and specific gravity of their urines, resulted chiefly from
diuretic or diarrheal influences and the consequent compensatory
tendencies.

Administration per os caused diarrhea. Bone ash in the food
appeared to exert only a mechanical diminution of such diarrheal
tendencies. Injections under the skin or into a muscle or into

124 Influence of Magnesium Sulfate on Metabolism

the circulation failed to elicit any evidence of diarrhea, except
in one doubtful case after subcutaneous application. On the
contrary, such injections appeared to make the feces drier and
harder than ordinarily, and the urine volumes greater.

If there was any effect on the quantitative elimination of solid
matter in the feces, it was not more than a slight increase. The
same may be said of the contents of magnesium and nitrogen in
the feces. :

In a general way elimination of nitrogen in the urine was
increased after the normal periods, but the increase was not
sufficient to warrant the conclusion that it was a direct effect
of the dosage. The absolute increase of urinary nitrogen
observed was registered chiefly in the form of urea, although the
relative excretion of the latter was below normal in Experiment
III.

The most striking and consistent effect on the partition of the
urinary nitrogen was the continued absolute as well as relative
increase of ammonia elimination throughout the whole of the
dosage part of each metabolism experiment, in spite of the fact
that the Folin method does not permit of complete recovery
of ammonia from crystallized ammonio-magnesium phosphate.
The increased elimination of ammonia nitrogen may be attributed,
in large part at least, to special formation and elimination of
ammonio-magnesium phosphate under the prevalent conditions.
The observations pertaining to the remaining nitrogenous con-
stituents of the urine have been summarized on p. rat.

It is especially noteworthy that recovery from dosage with
magnesium sulfate, however profound the immediate effect of
such treatment may have been, was always prompt and apparently
complete so far as general observation and our data indicated.
That magnesium sulfate exercises surprisingly little effect on
nitrogen metabolism under the conditions of these experiments
has also been shown by the results.

In conclusion it gives me great pleasure to acknowledge my
sincere gratitude to Prof. William J. Gies, at whose suggestion
these experiments were begun and to whose constant advice
I owe their completion.

A NOTE ON THE CHEMISTRY OF THE MUSCLE AND LIVER
OF REPTILES.

By JOHN F. LYMAN.

(From the Sheffield Laboratory of Physiological Chemistry, Yale University.)
(Received for publication, May 2, 1908.)

The paucity of data regarding the chemistry of the lower
vertebrates affords the occasion for recording a few observations
made at the suggestion of Prof. Lafayette B. Mendel, on the
muscles and liver of pythons. Aside from the studies of Kruken-
berg! we know of no investigations on these tissues in reptiles.
This author states that he found creatin and hypoxanthin in the
muscle extract of Python sebe, Python molurus, and an alligator.
Through the kindness of Dr. L. J. Cole, we obtained 4 kilograms
of fresh muscle and the livers from two specimens of Python
reticularis (?) which had just died. The water extracts of the mus-
cles were precipitated with alcohol, the filtrates concentrated under
diminished pressure, treated with lead acetate, and the lead-free
solution allowed to concentrate until a crop of crystals was
obtained. At subsequent stages in the examination of the
extract more crystals were separated, making a total of 8.2
grams which were identified as creatin.’

Found: Calculated:
ER@OrmncC HoN.O, HO... 12.15 percent 12.08 percent
Pine EON AO," 2. es S220 < S2 eae te

The filtrate from the bulk of the creatin was concentrated to a
syrup, which was repeatedly extracted with hot alcohol. After
removal of the alcohol the solution was acidified with sulphuric
acid and extracted with ether in the Kutscher-Steudel apparatus.

1 Krukenberg: Vergleichend-physiologische Studien, ii Reihe, ii Ab-
theilung, p. 81, 1882.

2 Professor Mendel has separated creatin from the muscles of the turtle
Chelone midas.

£25

126 Chemistry of Muscle and Liver of Reptiles

From the extract 2.23 grams of zinc paralactate were prepared
in the usual manner and identified by analysis:

Calculated for

Found: (C3H;03)oZn.2H,0:
HO... Soe eee 13.0 percent 12.9 percent
/ Sy OP ke 2A Sim Sete 23.8 a

The zinc salt was levorotatory. It is interesting to note in this
connection that fermentation lactic acid has been found in the
muscles of invertebrates by Henze! in Octopus, and by Mendel
and Bradley? in Sycotypus.

After removal of sulphuric acid with barium carbonate the
purins were separated from the residual extract by precipita-
tion with an ammoniacal solution of silver chloride. As in the
case of the skeletal muscles of the higher vertebrates, nonstriated
vertebrate muscle,? and embryonic vertebrate muscle,‘ hypo-
xanthin was found to be the conspicuous purin derivative. Guanin,
adenin, and xanthin could not be isolated from the material
available. The hypoxanthin (0.33 gram) was identified as the
nitrate:

Found: C NAO : NOC H.O:
IND atic ee tas eee watese toe 32.4 percent” 32.3 percent

The fresh livers, weighing 440 grams, were comminuted and
boiled 4 hours with 5 per cent sulphuric acid. From the filtrates
the purin derivatives were separated by the copper sulphate-
sodium bisulphite method according to the direction of Schitten-
helm.®° The products isolated were:

Uric acid, 0.092 gram, recrystallized according to Horbaczewsk1’s
method:

‘Henze: Zettschrijt fur physiologische Chemie, xliii, p. 477, 1904.

? Mendel and Bradley: American Journal of Physiology, xvii, p. 169,
1906.

3 Saiki: This Journal, iv, p. 483, 1908.

* Mendel and Leavenworth: American Journal of Physiology, xxi, p.
102, 1908.

® Schittenhelm: Deutsches Archiv fiir klinische Medizin, 1xxxi, p. 450,
1904.

John F. Lyman 127

Calculated for
Found: C;H4N4O3:

eee Bee ene iat 33.99 percent 33.33 per cent

Guanin, 0.39 gram, which was analyzed as the hydrochloride:

Found: Calculated:
H,O in C,H,N,O.HC1.2H,O... 16.8 percent 16.1 percent
MemePOTC!.......... 37.2 “ 37.4 ee

Adenin, separated as the picrate, 0.62 gram being obtained.
It was still impure after recrystallization, melting at about 265°
C. instead of 281°. Hypoxanthin and xanthin could not be
separated in amount sufficient for identification. From the
preceding observations it is apparent that the same purins are
present in the nucleoprotein complexes of these species as in the
higher vertebrates. The presence of uric acid is associated
with the well known synthetic capacity of reptiles in relation
to the end products of nitrogenous metabolism.

THE CHEMISTRY OF THE LIVER IN CHLOROFORM
NECROSIS (DELAYED CHLOROFORM POISONING).

By H. GIDEON WELLS.
(From the Pathological Laboratory of the University oj Chicago.)
(Received for publication, June 23, 1908.)

Recently I reported the results of a complete analysis of a
liver exhibiting the typical conditions of acute yellow atrophy,’
and while engaged in this analysis had the good fortune, through
the kindness of Dr. F. S. Tufts, to perform an autopsy upon the
body of a young man who had died a few hours before from the
condition of acute necrosis of the liver that occasionally follows
the administration of chloroform. This liver has now been ana-
lyzed in the same manner as was the acute yellow atrophy liver,
and certain steps have been duplicated with two normal livers
for comparison, all four livers having been obtained from robust
young men in the third decade of life. Some of the final. figures
of all four analyses were included in the paper upon acute yellow
atrophy, but in the following report will be considered particu-
larly the chemistry of the liver in chloroform necrosis.

Delayed chloroform poisoning is an extremely serious, often
fatal sequence of the use of chloroform as an anesthetic, which has
only recently become generally recognized. Attention was called
to it in this country particularly by the articles of Ballin and of
Bevan and Favill,? and in England by Guthrie,’ while the chief
recent article in German literature ‘s that of Guleke.* It is not
my purpose here to go into the literature, or to discuss the clin-
icaland anatomical features of the case, which has been done more
extensively in another place.* Briefly, the features of the effects

¥ Journ. of Exper. Med., ix, p. 627, 1907.
2 Journ. Amer. Med. Assoc., xix, September 2, 1905.
3 Lancet, July 4, 1903.
4 Arch. f. klin. Chir., \xxxiii, p. 602, 1907.
> Archives of Internal Medicine, i, p. 589, 1908.
129

130 Chloroform Necrosis

of delayed chloroform poisoning are as follows: After an opera-
tion, which may involve any part of the body but which is most
frequently abdominal, and in which chloroform has been used
as the anesthetic,! the patient begins as if upon the road to an
uneventful recovery; but after a period of from 12 to 72 hours
(generally about 24 to 36 hours), symptoms of restlessness, fear
and delirium appear, passing into coma which terminates after
two to four days, as a rule, in death. The persons are almost
always young persons, never past middle life, and the cases
reported in the literature seem to fall into two groups. In one,

which is seen chiefly in children, the manifestations are similar _

to those of acid intoxication, with the so-called acetone odor of
the breath and the presence of acetone and diacetic acid in the
urine, more or less cyanosis, and sometimes distinct air hunger;
anatomically in these cases there is usually found a well marked
fatty degeneration of the kidneys, myocardium and liver, the
changes in the latter being described as chiefly in the periphery
of the lobules. The other type of poisoning has been observed
in young adults, and differs in that with the onset of delirium
jaundice appears, and usually becomes profound, cutaneous hem-
orrhages often develop, there is more or less tenderness in the
region of the liver, the area of liver dulness decreases, leucin
and tyrosin may be found in the urine, and the clinical picture is
similar to that of acute yellow atrophy; at autopsy the liver is
found decreased in size, friable, yellow, extremely degenerated,
with necrosis of by far the majority of the liver cells, and there
is more or less necrosis and fatty degeneration of the kidneys.

The case that I have studied is of the second type, in which the
hepatic changes are for the most part similar to those of acute
yellow atrophy, and may be described briefly as follows: A vigor-
ous young man, 28 years of age, was operated upon for gallstones,
chloroform being given in a rather liberal manner. The opera-
tion, which amounted merely to an exploratory laparotomy, was
of short duration, and for the first 48 hours after the patient was
in good condition; then restlessness began, rapidly passing into
delirium and coma; jaundice appeared and became pronounced,

‘There are, perhaps, a few instances in which a similar effect has
been produced by ether.

4
-

H. Gideon Wells E31

and death occurred too hours after the operation. At the
autopsy, performed twelve hours after death, the chief changes
were found in the liver, which had been reduced in size so that it
weighed but 1050 grams, this being rather under than over two-
thirds the normal weight for a man of this size and age. It hada
wrinkled capsule, and was strikingly shrunken, flabby, soft and
friable; it appeared as if badly decomposed, except that it had a
sweetish odor, and it was of a yellow color on the cut surface.
Microscopically the cells in the center of each lobule were found
necrosed, greatly decreased in size and number, granular, contain-
ing many small droplets, and causing an extremely disorganized
appearance of at least two-thirds to three-fourths of each lobule;
only the cells at the very periphery of each lobule still retain the
power of taking the nuclear stains, and even they are very
granular and contain numerous large droplets of fat. These
changes are quite similar to those seen in the early stages of
acute yellow atrophy, differing chiefly in the considerable amount
of fatty change that is present.

CHEMICAL ANALYSIS.

Immediately after the autopsy the liver was cut into thin
slices, and 850 grams (which constituted 81 per cent of the entire
liver) was placed in a large volume of 95 per cent alcohol. After
hardening was completed the tissue was ground up in a meat
chopper and extracted repeatedly with fresh quantities of boiling
alcohol under a reflux condenser, the alcoholic extracts being
filtered off and united. Twice the residue was dried and ground

to a powder in a mill during the alcohol extraction, and after this
_ it was extracted in a Soxhlet apparatus with absolute ether until
nothing more could be removed. The ether and alcohol extracts
Were united and evaporated to dryness under vacuum at a
temperature not exceeding 40°, the residue obtained weighing
127.5 grams, or 15 per cent of the entire fresh weight of liver
taken. A sample of the fresh liver tissue was also weighed and
dried, and the water content was found to be 72.4, leaving 27.6
‘per cent of solids. The liver residue left after this extraction
was then extracted in a shaking machine repeatedly with fresh
quantities of water, then with water at 50° to 60°, these extracts

132 Chloroform Necrosis
being united and evaporated to dryness at 40° to 45° im vacuo;
the weight of the dried extract was 5.03 grams. The residue of
liver tissue was then extracted with boiling water,-changed fre-
quently, until the extracts were practically colorless. These
three distinct extracts (alcohol and ether, cold and warm water,
hot water) and the insoluble residue of coagulated liver tissue
were now available for separate analysis.

The hot water extract was concentrated to 100 ec., and four
volumes of absolute alcohol added; the precipitate was filtered
off, washed in 95 per cent alcohol, redissolved in hot water and
reprecipitated with alcohol. On dissolving this precipitate again
it was found to give a good Millon’s reaction, a bluish biuret
reaction, but not the Hopkins-Cole reaction for tryptophan. On
cooling it gelatinized, and when dialyzed for three days under
toluol in a celloidin sac the diffusate contained no appreciable
amount of solids and did not give the biuret reaction, indicating
the absence of proteoses and peptones. Therefore the precipitate
was pure, or nearly pure, gelatin. Determination of nitrogen in
an aliquot part of the solution indicated a total of 0.302 gram
nitrogen, which corresponds to 1.68 gram of gelatin (assuming
that gelatin contains 18 per cent of nitrogen); this amounts to
but 0.72 per cent of the total solids of the liver.

Evaporation of the alcoholic filtrate from the gelatin precipita-
tion yielded 3.81 grams. of solids, which was added to the 5.03
grams of residue from the cold and warm water extracts; deter-
mination of the nitrogen in an aliquot part of this mixture indi-
cated the presence of 0.894 gram of nitrogen, or ro per cent of the
total solids in these watery extracts. This solution gave good
biuret and Millon reactions, and a faint reaction for tryptophan.
On addition of several volumes of alcohol to the watery solution a
granular precipitate was obtained, which was found to consist
chiefly of arsenic from the embalming fluid! The filtrate gave
a strong biuret reaction, and upon concentration to 50 cc. and
addition of five volumes of absolute alcohol there was obtained a —
typical hygroscopic proteose-peptone mixture, which weighed

‘

|

‘An attempt at embalming had been made, but fortunately for my
purposes this was very unsuccessful; the liver was not at all hardened —
except a small part of the surface where embalming fluid injected into
the peritoneal cavity had come in contact with it.

H. Gideon Wells 133

three grams whendry. This proteose-peptone mixture was then
dissolved. in 50 cc. of water and precipitated fractionally with
ammonium sulphate according to the method of Pick, with the
following result:

First fraction (half saturation). A brown gummy mass, weigh-
ing about 0.5 gram. This gave a slightly purplish biuret reac-
tion, and was probably a mixture of gelatin and proteoses.

Second fraction (two-thirds saturation). About 0.2 gram more
of brownish gummy precipitate, which gave a good pink biuret
reaction and a good tryptophan reaction.

Third fraction (saturation). Only a very slight precipitate,
giving a good biuret reaction and fair tryptophan reaction.

Fourth fraction (filtrate). Gave only a faint biuret reaction,
therefore probably only a small amount of peptone present.

COMPOSITION OF THE INSOLUBLE COAGULATED RESIDUE. The
residue of insoluble liver substance left after all the extractions
had been completed, was dried to a constant weight and analyzed.
The total nitrogen was determined by the Kjeldahl method in a
sample, and found to amount to 14.48 per cent of the entire
weight, indicating that the substance is nearly pure protein. The
distribution of this nitrogen in the form of monamino, diamino
and amid nitrogen was determined in two samples according to
the Hausmann method as used by Osborne and Harris! with the
following results:

Per cent
if TaN Average. total

nitrogen.
PUN MIDFOMEN,.... 0.2... 025.4. 0.52 0.58 |. OF56 3.9
MENS OUETOP EN. 2.25.22. oe ee ee se 0.82 0.80 0.81 5.7
yamine mitropen™..........:..... ori | ec ed Wai Ge Sy 30.0
Monamino nitrogeny............-. 8.45 | 8.45 | 8.45 60.3

(OURS Seas ea ee 14.04 14.00 14.02

* This fraction also contains part of the purins.

+ The monoamino nitrogen was determined directly in an aliquot part of the filtrate from
the phosphotungstic acid precipitate, and not by difference as is usually done. On this
account the total nitrogen obtained, 14.02 per cent is a trifle less than all the nitrogen
present, 14.48, but this statement of results is probably more nearly accurate than when
the determination is made by difference.

1 Journ. Amer. Chem. Soc., XXv, Pp. 323, 1903.

134 Chloroform Necrosis

Other portions of this liver residue were also analyzed for sul-
phur, iron and phosphorus, with the following percentage results:

i. 1 Average.
Sulphires)s, . (once ees La ee ee eee | 0.80 0.78 0.79
Phosphorus*: : Ge. > serena oe ee eee 0.85 0.83 | 0.90
Phosphorus...) ae see ee | 0.98 0.93 f
Tirta esi te Sears eee eee oe ae ee ee | 0.42 6.57 0.50
INST ai Shi, a cane aad hate cesar ale eae ee ae 0.93 0.98 0.95

* On account of the unexpectedly high figure for phosphorus in this analysis, in which
the combustion was made by fusing with sodium carbonate and sodium nitrate, another
pair of determinations was made by Neumann’s method, which gave the slightly higher
figures seen in the second row.

FATS AND LIPOIDS. ‘The extracts with alcohol and ether, which
had been evaporated to dryness 7m vacuo, weighed 127.5 grams,
and this material was in turn thoroughly extracted with ether,
which dissolved out 75 grams, presumably fats and lipoids, this
constituting 8.8 per cent of the total fresh weight or 32 per cent
of the solids of the liver. Therefore 18.8 per cent of the entire
weight of the fresh liver was proteins and extractives, 8.8 per
cent fats and lipoids, and 72.4 per cent water. The ether
extract was evaporated to dryness, dissolved in absolute alcohol
and made up to 250 cc., of which two 20 cc. samples were taken
for cholesterin determinations and two ro cc. samples for lecithin.

Cholesterin was determined according to the method of Ritter,’
in which the fats are saponified with sodium alcoholate, dried out
in a large volume of salt, and extracted thoroughly with ether
which dissolves out the cholesterin but not the soaps. Traces of
sodium alcoholate, soaps, and glycerin escaping into the ether
extract are removed by shaking out with water, and the chol-
esterin weighed directly after evaporating off the ether. The
amount found in the samples taken (0.353 gram in 3% of the
extract) corresponds toa total of 4.4 grams in the 850 grams of
fresh liver substance analyzed, or 5.4 grams in the entire liver;
this indicates that cholesterin constituted o.52 percent of the fresh
weight; 1.9 per cent of the total dry weight, 2.9 per cent of the fat-
free dry substance, or 5.9 per cent of the ether-soluble material.

1 Zeitschr. f. physiol. Chem., xxxiv, p. 430, 1902.

ee

H. Gideon Wells 135

Lecithin was determined by the method recommended by Koch
and Woods! in which the lecithin is precipitated from an aque-
‘ous emulsion with acid chloroform, combusted by Neumann’s
- method, and the phosphorus determined as magnesium pyrophos-
phate. The amount of phosphorus present, 0.0212 gram, in the
samples analyzed, corresponds to 0.558 gram of lecithin in the
sample, or 12.95 grams in the extract from 850 grams of liver,
representing a total of 16 grams of lecithin in the entire liver.
This indicates that lecithin constituted 1.5 per cent of the fresh
weight, 5.5 per cent of the total solids, 8.1 per cent of the fat-
free solids, and 17.3 per cent of the ether-soluble material.

AMINO-ACIDS AND PURINS. These were sought in the part of
the original alcohol extract that was left after ether extraction,
and in the non-protein portions of the watery extracts. These
were united, dissolved in water, and made up to rooo cc., and the
nitrogen in a 10 cc. sample determined by Kjeldahl’s method,
46.41 mg. of nitrogen was found in the sample, corresponding to
4.641 grams of nitrogen in the entire extract, the total weight of
which was 45 grams. This material gave a strong Millon reac-
tion, but only a very faint biuret reaction. It was examined
for amino acids and purins, following with slight modifications
the method used by Schumm?’ in his study of autolysisin leukemic
spleens. As Schumm has given his procedure in detail it is un-
necessary to give here the details of the analysis. Briefly, the
diamino acids and purins were separated from the monamino
acids by precipitating in 5 per cent sulphuric acid solution with
phosphotungstic acid. From this precipitate, which contained
1.386 gram of nitrogen, were isolated the following substances:

I. Free purins. The total quantity obtained contained 0.0974
gram nitrogen,and from it was isolated 0.16 gram of hypoxanthin
silver nitrate, and a smaller quantity of xanthin silver. Adenin
and guanin could not be found.

II. Diamino acids. After removal of the purins from the
phosphotungstic acid precipitate, the diamino acids were sought
according to the method of Kossel and Kutscher. In the his-
tidin fraction was found 91 mg. of nitrogen, which corresponds to

1 This Journal, i, p. 203, 1906.
2 Beitr. z. chem. Physiol. u. Pathol., vii, p. 175, 1906.

136 Chloroform Necrosis

0.34 gram of histidin. The arginin fraction contained but 32.2
mg. of nitrogen, and yielded a small amount of long rhombic
crystals, but the amount was so small that it is impossible to
tell whether this was arginin or not.

The fraction that should have contained the lysin contained
0.404 gram of nitrogen, but from it no lysin picrate could be
obtained; instead there was a gummy mass, which gave a strong
biuret and good Millon and tryptophan reactions. This was
soluble in strong alcohol, and only a small part was precipitated
by saturating the solution with ammonium sulphate; the filtrate
from this precipitate gave strong biuret and Millon reactions, and
a faint tryptophan reaction, and therefore was either peptone or
polypeptid, most probably chiefly the latter.

III. Monamino acids. After removal of the phosphotungstic
and sulphuric acids from the filtrate from the diamino acid precip-
itate, the filtrate was concentrated and by fractional crystalliza-
tion was obtained 2.70 grams of what seemed to be chiefly a
mixture of leucin and tyrosin. By extraction with glacial acetic
acid 0.26 gram of pure tyrosin was obtained from the insoluble
residue, which was identified by its typical cottony appearance
when crystallizing out after purification, 1ts insolubility in glacial
acetic acid, and its intense Millon reaction.

By careful recrystallization of the filtrate from the tyrosin
after removal of the acetic acid and decolorization with animal
charcoal, leucin was obtained to the amount of 1.5 gram. This
was converted into a copper salt by boiling with fresh copper
oxide, and the typical bluish-white copper salt of leucin was readily
obtained in pure condition. After drying this at 112° a nitrogen
determination was made, and 8.77 per cent of nitrogen was
obtained, the theory for the copper salt of leucin being 8.67 per
cent of nitrogen. Copper was determined, and found to be
19.9 per cent, the theory for the copper salt of leucin being 19.6
per cent.

Upon saturating with hydrochloric acid gas the concentrated
filtrate from the leucin-tyrosin crystallization, a considerable
amount of crystalline material came out, but on further examina-
tion this was found to be entirely inorganic, and not glutamic
acid hydrochloride. The filtrate from this was then esterified
three times according to the method of Emil Fischer, and about

= ee

er

H. Gideon Wells £37

10.5 grams of raw esters was obtained. This was fractionated
in the usual manner except that only the water pump was used,
the pressure throughout the distillation being 9 to12mm.,and the
following fractions were obtained.

Grams.
Fraction I. Temp.upto60°; weight of esters............... 4.0
S ie “~~ 60°-100°; oi Bid oie: ea 1.9
= 100 “ 100°-188°; 4 ee LS
z IV. Residue E RRS 5 95.0) 2.5

The first three fractions were hydrolyzed by boiling in water
eight hours with inverted condenser, and then evaporated to
- dryness 1m vacuo over sulphuric acid at 45°. Fraction I yielded
0.58 gram of a crystalline white substance having a sweet taste
and melting at 240°; probably chiefly glycocoll. Fraction II
yielded 0.99 gram of a substance similar to fraction I, except that
it had a mixed sweet and bitter taste and contained a few waxy
crystals resembling leucin; it melted at 250° and was presumably
a mixture of glycocoll with small amounts of higher amino acids.
Fraction III yielded 0.81 gram of a brownish, semi-solid sub-
stance, smelling like prolin and slowly forming a small amount of
crystalline substance.

Fraction 1V was shaken out with ether and water until the
ether-soluble portion was removed. This ether extract was
hydrolyzed by evaporating with concentrated hydrochloric acid,
but only a very small amount of crystalline material, mixed with
a brownish amorphous substance, was obtained, insufficient
for purification and identification as phenylalanin which would
be found in this fraction if present. The watery extract was
hydrolyzed by boiling with barium hydroxide, and the barium
Salts allowed to crystallize out, but after removing the barium
from the crystalline material no aspartic acid or other crystalline
organic substance could be found. The filtrate from the crystal-
line barium salts was freed from barium, concentrated to a few
cubic centimeters, and saturated with hydrochloric acid gas;
after standing on ice over night without crystallization the solu-
tion was inoculated with a crystal of glutamic acid hydrochloride
and a considerable amount of typical crystals of this salt came
out. This salt was separated by filtration through asbestos in
a Gooch crucible, and weighed 0.58 gram. It was recrystallized,

138 Chloroform Necrosis

and a nitrogen determination made, 7.74 per cent of nitrogen
being found, agreeing well with the theory for the hydrochloride
of glutamic acid, which calls for 7.63 per cent of nitrogen.

The chlorine was removed from the filtrate from the glutamic
acid with lead oxide, and after removal of the lead with hydrogen
sulphide a very small amount of slightly crystalline greenish
material was obtained. This was redissolved and a copper salt
was made, which was of a greenish yellow color and manifestly
impure; therefore it was impossible to determine whether aspartic
or any other amino acid was present in this fraction.

Fractions I and II were united, dissolved in ro cc. of hot water,
and on addition of three volumes of hot absolute alcohol there
was obtained no precipitate, and none appeared on cooling,
indicating the absence of any appreciable amount of leucin or
alanin. Therefore the solution was evaporated to dryness,
dissolved in a minimum amount of hot absolute alcohol, saturated
with dry hydrochloric acid gas and cooled. After inoculating
with crystals of the hydrochloride of glycocoll ethyl ester a
small amount of crystals of this type appeared. The yield
being unsatisfactory the chlorine was removed with lead oxide,
and a copper salt was made of the amino acid in the filtrate. On
analysis of the first fraction of this salt obtained on crystallization
13.58 percent of nitrogen and 30.2 percent of copper were obtained,
the theory for the copper salt of glycocoll being nitrogen, 13.40
per cent, copper 30.8 per cent. Therefore, most if not all of the
first two samples, amounting to 1.57 gram, is glycocoll.

The third fraction, which resembled prolin, was made into a
copper salt, and the salt treated with absolute alcohol which
dissolved out a greenish, noncrystalline material, leaving the
crystals of the copper salt. This was analyzed for nitrogen, and
8.0 per cent found. The theory for the copper salt of prolin
is 9.62 per cent, for aspartic acid 7.22 per cent, and for leucin
8.67 percent. Possibly the salt obtained was an impure mixture,
but the amount available for analysis was too small to permit
of duplicate analyses or for copper determinations; therefore
the presence or absence of prolin cannot be determined, but from
the appearance and odor of the material it is probable that prolin
was present.

Ee

,

H. Gideon Wells 139

DISCUSSION OF RESULTS.

This analysis has most interest when compared with the results
obtained by the analysis of two normal human livers and a
liver from a case of typical ‘‘idiopathic”’ acute yellow atrophy,
which has been previously published,’ all four livers being from
young men of about the same age. In the acute yellow atrophy
liver the most interesting result was the isolation of a compara-
tively large number of amino acids in sufficient purity for their
identification; in the chloroform necrosis a somewhat smaller
number and smaller amounts were isolated, as shown by the
following table:

Acute Chloroform
atrophy necrosis
TEUISIS CGI) os oes ete aie ce OR Ea ee 0.64 0.34
ANP TUTE, « 2.3e°SUn oe er eIRe ae Decks Oy er eee ?
OL. Ot ea a a 1.04 ?
0 DAD cos See Sa 0.70 0.26
ILGUTVIG, 6 2 sei tS pte Dae Se nee 4.16 1.50
RTE Pee i eke are GaN dgi div vg wien ae « 0.20 1.57
JTS WIS. 0. o:cies eo rek ey coerce eee 0.30
2 2.) 2S Ee eee 0.35 present ?
MEEEDEETISATEN UCI Clee Wnt es fa cases, Ade lah oreis Soya steuccatin 1.00 0.58
2. DL, aE eS ees eee 0.28
EDT) DEG Sa 2
TDETLZAL 7 och eeke eee Nn ae ee 8.67 4.25

With these results may be incorporated the results obtained
by A. E. Taylor, who has also examined for amino acids the
extracts from a case of acute yellow atrophy” and one of chloro-
form necrosis.2 From the acute yellow atrophy liver he iso-
lated 0.35 gramof leucinand 0.612 gramof asparticacid; from the
chloroform necrosis 2.3 gram of arginin nitrate, 2.2 grams of tyro-
sin, and 4.0 grams of leucin.

These figures of themselves indicate nothing as to the actual
quantity of free amino acids present, on account of the inade-
quacy of the analytic methods that are available for their isola-
tion; in each case the amino acids recovered account for only a

1 Journ. of Exper. Med., ix, p. 627, 1907.
2 Journ. of. Med. Research, viii, p. 424, 1902.
3 Umiv. of Calif. Publ. (Pathol.), 1, 43, 1904.

140 Chloroform Necrosis

small fraction of the nitrogen present in the solutions containing
them, and less than one-fourth of the total weight of the esters
obtained from the chloroform necrosis liver could be recovered
as amino acids. Nevertheless it is of interest to find that so
many of these constituents of the protein molecule can be found
free in degenerated livers, even if only in small amounts. These
two analyses furnish the only instances that I can find in the
literature of the isolation of free glutamic acid and free prolin
from either human or animal tissues or excretions; the identity
of the glutamic acid was completely established in both cases,
but the prolin, although almost certainly present in the extracts
from the chloroform necrosis liver, could not be isolated in suffi-
cient amount for analysis and identification.

It might be expected that more free amino acids would be
found in the chloroform necrosis liver, in which the reduction in
size of the autolyzing liver took place in a few days, than in
the acute yellow atrophy liver in which the process was of some
six weeks’ duration. Thefact that a smaller quantity of amino
acids was isolated from the chloroform necrosis liver may only
depend upon less success with the analysis, but it may mean that
there actually was a larger amount of free amino acids in the
acute yellow atrophy liver. If the latter explanation is correct
then we should be obliged to consider it as added evidence that
in acute yellow atrophy the free amino acids found in the liver and
secretions are not derived solely, or even chiefly, from the auto-
lyzing liver cells. Neuberg and Richter’ found larger amounts
of free amino acids in the blood in acute yellow atrophy than
could be accounted for by the destruction of liver tissue going on
at the time, and concluded that there must be some other source
for them, possibly the intestine. The rather large amount of
amino acids isolated from the liver in the case of acute yellow
atrophy mentioned above is in favor of the same idea, and the
smaller amount present in the more rapidly digesting liver with
chloroform necrosis might be looked upon as of similar signifi-
cance, were not the value of quantitative results obtained with
such materials and methods so very questionable.

Proteoses and peptones were also present, and apparently
much more abundantly in the choloform necrosis liver, in the

1 Deutsch. med. Woch., Xxx, p. 499, 1904.

H. Gideon Wells 141

extract from which were probably also considerable amounts of
substances related to the polypeptids. The relative abundance
of these substances intermediate between proteins and amino
acids in the chloroform necrosis liver is presumably dependent
upon the fact that here the autolysis was less advanced than in the
acute yellow atrophy liver.

In both livers the presence of free xanthin and free hypoxan-
thin was established, and in about the same amount in each.
Free guanin and adenin were absent from both livers, presumably
because they are so readily converted into xanthin and hypo-
xanthin by the hepatic enzymes.

The composition of the coagulated and insoluble proteins of the
liver left after thorough extraction with alcohol, ether, cold and
hot water, is found to be quite the same in chloroform necrosis
as in normal livers, as shown by the following table giving the
results of analysis by Hausmann’s method:

l ay
|

Acute Normal Normal Chloroform
atropny. (anemic). | (congested). | necrosis.
IGM LEO ON: =... .. 0.20. ae ee DD Shall 4.8 | 3.9
EUUIMAUISMTCTOP ED, «<0... coe ce eos 3.6 Soa 49-7} 6 7
Minmmime Wtrogen.........-...... 26.2 32.8 30.0 | 30.0
Monamino nitrogen..............| 64.8 60.3 60.2 60.3

In considering these figures it must be borne in mind that
we.are dealing with liver tissue from which not only the extract-
ives, fats and lipoids have been removed, but also the greater
partifnotall the gelatigenous substances. Analysis of liver tissue
by other observers in which these extractions have not been made
are, it seems to me, of uncertain value, especially in pathological
livers in which the great variation in fat as well as extractives and
connective tissue can by themselves produce great alterations
in the percentage figures, which are then incorrectly ascribed to
the essential constituents of the liver cells themselves. Further-
more, it must be considered that the degree of regenerative pro-
liferation and leucocytic invasion that is taking place in the
liver will modify greatly the amount of purins and nucleoproteins.
Taking these figures at their face value, however, they may be
interpreted as meaning that the decrease in diamino nitrogen

142 Chloroform Necrosis

which Wakeman’ found to be so striking in the livers of dogs
poisoned with phosphorus, and which was found to a less extent
in my case of acute yellow atrophy, was not exhibited by the
liver showing extensive necrosis from chloroform. More recently
Wakeman’ has analyzed a liver said to show acute yellow atro-
phy, in which there was found no decrease in the nitrogen of the
bases. To draw any conclusions from these isolated observa-
tions, however, would not be warranted.

Determination of insoluble sulphur, phosphorus and iron in the
extracted residues of these livers, diseased and normal, gave
results that are difficult of interpretation. The following table
gives the percentage amounts of these inorganic elements in the
residues:

Acute | Normal }) Normal Chloroform

atrophy. (anemic). | (congested). necrosis.
Sulphur. aewte ciaee oeeae eel O Ge 0.75 0.77 0.79
Phosphoruss (25.8000 eo o. oee ae ORO 0.27 0.21 0.90
LIC) CRU Set Ane ie MIE nie Sa ea eee Nhe ee 0.2 0.4 0.5

* Average of four analyses of each specimen.

While the sulphur is practically constant in amount in all four
Specimens, in spite of the great structural changes in the two
diseased livers, the insoluble phosphorus in each of the latter
is increased to about four times the amount present in the nor-
mal livers. The increase in the phosphorus in the acute yellow
atrophy may be readily explained as the result of the great pro-
liferative activity exhibited by the cells of the stroma and bile
ducts in areas where regeneration is taking place, which causes the
presence of large numbers of new cells rich in nucleic acid. No
such explanation is available for the increase of phosphorus in
the chlofoform necrosis liver, however, for in this specimen there
is not only no proliferation, but also by far the majority of hepatic
nuclei have disappeared, making it doubly hard to account for
this decided increase of insoluble phosphorus. It is barely
possible that some of the lecithin phosphorus has been so fixed

1 Journ. of Exper. Med., vii, p. 292, 1905.
2 This Journal, iv, p. 119, 1908.

ing

H. Gideon Wells 143

that it cannot be extracted from the cells, in view of the fact that
there has been some loss of lecithin in both livers, but it is not
probable that such a change could account for more than a small
fraction of the increase in phosphorus that wasobserved. Inany
case these figures show that necrosis of the liver cells, with dis-
appearance of the majority of stainable nuclei, is not necessarily
associated with a decrease in the amount of insoluble phos-
phorus as would be expected.

In agreement with the histological picture, the strikingly large
amount of gelatin obtained from the acute yellow atrophy liver
as a result of the connective tissue proliferation present in this
condition was not found in the liver of chloroform necrosis.
While the acute yellow atrophy liver yielded 13.8 grams of gela-
tin or 10.1 per cent of the dry, fat-free tissue, a normal liver
yielded but 3.2 per cent of the dry fat-free tissue as gelatin, and
in the chloroform necrosis but 1.5 per cent was gelatin.

There was no such increase of the proportion of water in the
liver as is found constantly in acute yellow atrophy, as shown in
the following table:

Fat-free dried

| Water. | Fat. substances,

Moumal liver (Quincke)...............+5+. | 76.1 | 3.0 20.9
Meemmmabliver(Wells)...................5- ame | S.0 Li
memneamopoy (Perls)..............<4..... 81.6 8.7 Sar

= 8 MBCrIS) ieee Serle Salts ee ore | 76.9 7.6 WS

Ms ‘ VAD EECKS) ps cits as ors ois aX ¢ 80.5 4.2 aye,

ss & (UNS ACTS) oe Sec. Sagi cn ee Sa.8 | 2.0 ZZ

cs MVVIAIKETIVAN™)) seis ces ohana » 79.3 |

: a GUViclls\ metrics fort a | 83.8 | 2.5 sya 0

x Me Mcgee) lesa ects 78.0 | «626 15.4
Phosphorus poisoning (v. Starck) .......... 60.0. | 2058 10.0
Fatty degeneration (v. Starck)............. | 64.0 25.0 11.0
Chloroform necrosis (Wells)............... | 72.4 8.8 18.8

* This Journal, iv, p. 119, 1908.
+ Johns Hopkins Hospital Bull., xix, p. 50, 1908.

The chloroform necrosis liver stands between the typical acute
yellow atrophy liver and the ordinary fatty liver, according to its
analytic figures as well as according to its histology. That

144 Chloroform Necrosis

is, there has been some replacement of water by fats, but not so
much replacement of protein by water as in acute atrophy.
These figures emphasize the fact that in acute yellow atrophy there
is no increase of fat in the liver, and that in chloroform necrosis
the amount of fat is distinctly increased, although not so much
is found chemically as might be expected from the microscopic
findings.

The amount of lecithin and cholesterin in this liver is by no
means so greatly altered from normal as was found to be the case
in acute yellow atrophy, as shown by the following table:

LECITHIN. | CHOLESTERIN.
| ates | | :
| eee et eh abo sol | Saati 3
et Er@— || eu SN iesst | ehs | Se eB
FS| G2 | 52) £3 | £2) £2 | Se | $8
Be | 69 | tae | 88 | 68 | Gal eee
5 | whe | |e ee
Per cent | | | |
of fresh weight...) 1.6) 1-4) 1.5) <0u49) 0.26, 0:37) 0202) Oss
of total dry weight.| 6.3) 6.25, 6.2; 2.9|1.0|1.7] 1.9) ais
of dry fat-free ma- | | | | | |
teriali..........| -7.7|) 8.0°/" 8.1). 3:2 | 1-25) 2 ee ee
of ether-soluble sub- | et a at |
StanGesee eee oes 1335.3) 2Zeu0 | 1723) Lieb | 5.7 | 74 |a589s\ Sie
entireliver:......- DSN se AN GAO eee 4: | 3.8 | 5.95) 5.4 3.38
| | | | |

While the total amount of lecithin has decreased, this is only
in proportion to the decrease in the total size and weight of the
liver; this proportional decrease has gone on in spite of a relative
increase in the amount of simple fat, showing the same lack of
correlation between the lecithin and the neutral fat which has
been observed by others who have determined the lecithin con-
tent of organs showing fatty degeneration. Evidently, there-
fore, the increase in the fat content of the liver in chloroform
necrosis is due entirely to simple fats. The cholesterin, on the
other hand, has apparently remained in about the normal amount,
and has not decreased with the lecithin and proteins; this is quite
what might be expected from what we know of the tendency of
cholesterin that is liberated by degenerating cells to remain at
the place where it is formed.

ES ee ee ee

E

H. Gideon Wells 145

The only other analysis of a liver showing chloroform necrosis
that has been recorded in the literature is the one published by
A. E. Taylor.! This case was very similar to the one described
above, both clinically and anatomically. The liver weighed
1200 grams, was very soft, friable, and ‘‘putty-like,” showing
microscopically widespread degeneration of the liver cells. It
contained over 200 grams of fat, and from the extracts were
obtained 4 grams of leucin, 2.2 grams of tyrosin, and 2.3 grams
of arginin nitrate. The other constituents of the liver were not
determined.

These two analyses corroborate one another in showing the
presence of free amino acids in amounts large enough for identi-
fication in the liver of chloroform necrosis. The amino acids
are presumably derived from autolysis of the liver cells, although
it is by no means certain that part of the free amino acids found
in the liver may not have come from some other source.

SUMMARY.

In the necrosis of the liver which occasionally follows chloro-
form anesthesia there is a rapid autolysis of the liver cells,
resulting in a loss of as much as one-third or more of the solidsin
three or four days, and indicated chemically by the presence of
free amino-acids, purins, proteoses, peptones and polypeptids in
the liver. Several of the amino-acids were present in quanti-
ties large enough to permit of their isolation and identifica-
tion. Despite the loss of nearly all the nuclear structures of
the liver the amount of insoluble phosphorus was found in the
specimen examined to be increased, without alteration in the
amount of insoluble sulphur. The distribution of the nitrogen
as mono- and diamino acids in the insoluble coagulated liver
proteins is not different from that of the proteins of the normal
liver. There is a moderate degree of fatty metamorphosis, the
microscopic and chemical findings corresponding in this respect;
this increase in ether-extractive material being due to infiltration
of simple fats, while there is a slight decrease in the lecithin and
no alteration in the amount of cholesterin. There is less replace-
ment of proteins by water and more fatty infiltration than in
acute yellow atrophy.

1 Umi. of Calif. Publ. (Pathol.), i, 43, 1904.

‘ \
f b 4 aol an ‘ E ne

¥ on Harice
aa vi 4

Lira, ie

ON THE INFLUENCE OF TEMPERATURE UPON THE
SOLUBILITY OF CASEIN IN ALKALINE SOLUTIONS.

By T. BRAILSFORD ROBERTSON.

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

(Received for publication, June 10, 1908.)

In previous papers! I have suggested that asolution of a protein
may be regarded as a system of polymeric modifications of the
amphoteric electrolyte HXOH, the point of equilibrium being
shifted by any variation in the conditions, such as the addition
of acid, alkali, salts or the application of heat, which influences
the concentrations of the protein ions; just as the simplest pos-
sible amphoteric electrolyte, namely, water, consists of a mix-
ture of polymeric modifications of the molecule HOH, the point
of equilibrium being shifted by alterations in temperature.’
From this point of view the process of heat-coagulation would be
regarded somewhat as follows; by repeated condensations of the
type

HXOH + HXOH = HXXOH + H,O

larger and larger molecule-complexes are formed until the molecu-
lar aggregates assume the properties of matter in mass and the
solution assumes the character of a suspension which is usually
unstable, the protein particles being thrown out of solution in the
form of coagula or flocculi. If this point of view be correct then
it follows that one of the effects of applying heat to a protein
solution must be the shifting of an equilibrium of the type:

HXOH + HXOH @ HXXOH + H,O

towards the right; were this so it would follow from van’t Hoft’s
“principle of mobile equilibrium” that the hydrolysis of pro-

17. Brailsford Robertson: Journ. of Physical Chem., x, p. 524, 1906;
Xi, p. 453, 1907; This Journal, iv, p. 23, 1908.
2Svante Arrhenius: Text-book of Electro-chemistry, Trans. by John
McCrae, p. 116, 1902; Wm. Sutherland: Phil. Mag.,1, p. 460, 1900.
147

148 Influence of Temperature upon Solubility of Casein

teins is accompanied by the evolution of heat, which conclusion
isin complete accord with experimental observation in so far as
positive results have been obtained.! The hypothesis is further-
more in accord with the view that heat-coagulation is accom-
panied by the withdrawal of water from the protein.” Views
in many respects similar to these have been expressed by a
number of authors. Mann, indeed, states that in his opinion
heat-coagulation is ‘“‘brought about by one portion of the albu-
min molecule precipitating the remainder,’’ a view which is
essentially similar to that expressed above.’ In this connection
it is also of interest to note that many authors have considered
that the initial stages of protein hydrolysis consist in the “‘de-
polymerization”’ of the protein molecule.* Sutherland has also
pointed out that the weight of different proteins which com-
bines with a gram equivalent of a heavy metal is a simple
multiple of the lowest observed weight and he deduces therefrom
that there is a large amount of ‘internal salt’’ formation in pro-
teins; he has also expressed the view, practically identical with
that put forward above, that coagulation of a protein is the
result of polymerization through the neutralization of “‘valencies
which are usually latent.’”®

If the heat-coagulation of a protein consists in the polymeriza-
tion of the amphoteric protein molecule with the elimination of
water, according to the equation given above, then the influence
of heat upon a compound of a protein with a base should be of
the following type:

NaXOH + NaXOH — NaXXOH + NaOH

1 The heat of reaction of protein hydrolysis is extremely small, observers
have either failed to detect any change in the heat-content of the system
or else have observed a very slight disengagement of heat. Tangl: Arch.
f. d. ges. Physiol., cxv, p. 1, 1906; v. Lengyel: Ib¢d., p. 7; Hari: Ibsd.,
p: 11, fo¢d., ‘CKXx1, pii4so, 1008.

? Michailow: Chem. Centralb., p. 1088, 1887; Starke: Zeitschr. f. Biol.,
Jubelband z. Ehren v. C. Voit, p. 206, rgor.

3 Gustav Mann: Chemistry of the Protetds, London, p. 318, 1906.

4Maly: Arch. f. d. ges. Physiol., ix, p. 585, 1874; XX, Pp. 315, 1879; Herth:
Zettschr. f. physiol. Chem.,i, p. 277, 1878; Poehl: Ber. d. deutsch. chem.
Gesellsch., p. 1355, 1881; p. 1152, 1883; Loew: Arch. f.d. ges. Physiol.,
SKI, Dp. GOs, Lass.

5 Wm. Sutherland: Proc. Roy. Soc., London, p. 130, 1906.

a

T. Brailsford Robertson 149

and the solution of the compound should become more alkaline
on heating. The marked increase in the alkalinity of solutions of
the caseinates of bases, which occurs on heating, has been observed
by Osborne,’ but his interpretation of the phenomenon is quite
different to that which is suggested above. He considers that
the action of heat consists in increasing the hydrolytic dissocia-
tion of the caseinate. Since the free casein is very insoluble and,
presumably, very slightly dissociated, an increase in the degree
of hydrolytic dissociation of the salt would lead to an increase
in the alkalinity of the solution, and furthermore, might be reason-
ably expected to lead toa marked opalescence of the solution or
even to the precipitation of casein, since the free casein is insoluble
in water. A marked increase in opalescence, on heating to 35°
to 45° C. was observed by Osborne in solutions of calcium, barium,
magnesium and lithium caseinates but it was not observed in
solutions of sodium, potassium or ammonium caseinates; if the
opalescence were due to undissociated casein being set free by
hydrolytic dissociation it is difficult to see why it does not occur
in solutions of sodium and potassium caseinate and especially
in solutions of ammonium caseinate in which, as Osborne him-
self points out, since the ammonium hydroxide is a very weak
base, hydrolysis might be expected to be especially intense. It is,
however, possible to decide between the two hypotheses in a very
simple way. I have shown in a previous paper” that a given
amount of alkali dissolves just sufficient casein to form the
“neutral caseinate”’ of the base (such that 8 cc. of ;¥, alkali = 1
gram of casein) and that the resulting solution is neutral to lit-
mus. If the influence of the application of heat upon this solu-
tion consisted in increasing the hydrolytic dissociation of the
caseinate it would follow that the power of the given amount
of alkali to bind casein is diminished by heat and therefore the
solubility of casein in a given concentration of alkali would be
diminished by an increase in temperature; on the contrary if the
influence of the application of heat upon a solution of a neutral
caseinate consists in the shifting of the equilibrium of the sys-

tem in the direction:
NaXOH + NaXOH — NaXXOH + NaOH

*W.A. Osborne: Journ. of Physiol., xxvii, p. 398, 1901.
*T. Brailsford Robertson: This Journal, ii, p. 317, 1907.

150 Influence of Temperature upon Solubility of Casein

then the alkali set free should be capable of dissolving more
casein, or in other words, the solubility of casein in a given con-
centration of alkali would be zucreased by an increase in temper-
ature. From either hypothesis the increase in the electrical con-
ductivity of the solution upon heating, which was observed by
Osborne, would necessarily follow. The following experiments
were undertaken with a view to ascertaining which of the alter-
native hypotheses represents the facts more accurately.

EXPERIMENTAL.

It has been pointed out by Osborne, in the paper referred to
above, that the influence of temperature upon solutions of casein-
ates is reversible, 1. e., thatthe opalescence and increase in alka-
linity which appear on heating disappear on cooling and reappear
on heating again. This is probably true for all heat coagulations
but where the protein is thrown down in coagula the hysteresis
of the system (owing to the excessive internal molecular friction
of large aggregates) prevents its reattaining equilibrium witha
measurable velocity on cooling. That the equilibrium charac-
teristic of low temperatures is rapidly regained if heat-coagula-
tion is not pushed too far, so that the molecular aggregates which
are formed are not too large, has been shown by Corin and An-
siaux' who find that the first traces of coagulation disappear on
quickly cooling and shaking the solution. We may, therefore,
assume that a solution of a caseinate which has been heated
regains, on cooling, its original condition and power of neutraliz-
ing bases. The method of procedure was as follows. Five, ten
etc., cc. of = potassium or lithium hydrate or of a saturated
solution of calcium hydrate (approximately 4), were diluted
to 100 cc. with distilled water, placed in tightly stoppered Erlen-
meyer flasks and warmed in a thermostat to the desired tempera-
ture. Three times the amount of casein which would be dis-
solved by the given amount of alkali at room temperature (1. e.,
3 grams to every 5 cc. of =) alkali) was then introduced and the
mixture left in the thermostat for from thirty to forty minutes,
being vigorously shaken at frequent intervals. The resulting
solution was then filtered in the thermostat (the filter and the

receiving vessel having been previously warmed to the desired —

‘Corin and Ansiaux: Bull. de l’acad. roy. de Belg., No. 21.

tale wt! ',

T. Brailsford Robertson I51

temperature) and the filtrate was allowed to cool or was cooled
by immersing the containing vesselin tap-water. The tempera-
tures were defined to within 4°. An aliquot part (25 cc.) of
the solution was then titrated against the alkali which had been
used to dissolve the casein, phenolphthalein being used as indi-
cator. Since 8 cc. of 4, alkali neutralize one gram of casein to
phenolphthalein' if the amount of alkali originally contained
in the volume of solution titrated is known and the amount
which it is necessary to add in order to secure neutrality to phe-
nolphthalein is ascertained the amount of casein contained in the
solution can immediately be deduced from the relation 1 cc. 79
alkali = .125 gram of casein. Ina previous paper I have shown
that this method of determination yields reliable results.’
The following were the results obtained:

Grams of casein dissolved per 100 cc. solution at—

Concentration of the

alkaline solution. 21° | 36° 46° 54° 60° 66° 81° ge
2310-4 N KOH | 0.46 | 0.52
46<10™ a 0.92 | 0.92 | 1.04 L384 | 227 | lesa ees
eaxlo+ “ 1.38 | mi a
O27 107-* ¢ 1.85 PAE T | S05
441074 LiOH 0.89 | 0.86 1,28. | 114 | De6Os i eteZO
Sel Op*. “ a LAF FTE 2.62 | 2.95
451074 Ca(OH),| 0.90 | 0.72 0.65 | 0.63 0.64 | 0.63
9010-4 4 1.80 1.35 | |

It is evident that the power of the bases, potassium hydroxide
and lithium hydroxide, to dissolve casein is greatly increased by
increasing temperature. In all cases, the solutions, at the tem-
peratures indicated, were acid to phenolphthalein and approxi-
mately neutral to litmus so that the effect of heating a solution
of a caseinate cannot be to increase its hydrolytic dissociation.
The facts are much more readily explained on the supposition
that the effect of temperature consists in shifting the equilib-
rium:

HXOH + HXOH @ HXXOH + H,O

1Van Slyke and Hart: Amer. Chem. Journ., xxxiii, p. 461, 1905; T.
Brailsford Robertson: This Journal, ii, p. 317, 1907.
2T. Brailsford Robertson: This Journal, ii, p. 317, 1907.

152 Influence of Temperature upon Solubility of Casein

towards the right so that a given amount of alkali, since it is
associated with a molecule of nearly double the weight, neutral-
izes nearly twice as much casein at 66° as it does at room tem-
perature (21°). The marked diminution in the solubility of
casein in calcium hydrate solutions which occurs on raising the
temperature can be explained by supposing that the salt
Ca(X XOH), is insoluble while the salt Ca(XOH), is soluble, and
this explains also why Osborne obtained an increase in opal-
escence upon heating solutions of calcium, magnesium and ba-
rium caseinates but not upon heating solutions of potassium,
sodium and ammonium caseinates. Since calcium hydrate
solutions do not fail, at any of the temperatures investigated, to
dissolve some casein it is evident that the conversion of the casein
molecule HXOH into double, triple or higher polymers cannot
be complete but that an equilibrium between the two forms exists
at every temperature. Since calcium hydrate solutions of dif-
ferent concentrations dissolve different amounts of casein at the
same temperature (as can be deduced by interpolation from the
above table) the amount which is dissolved by any giver con-
centration of calcium hydrate cannot represent, merely, satur-
ation of the solution with the salt Ca(X XOH), or Ca(XXXOH),
but must represent also the solution of the calcium salt of unpoly-
merized molecules in equilibrium with a saturated solution of the
calcium salt of the polymerized molecules. It appears probable
in the light of these results, that the view which I have pre-
viously expressed is the correct one, namely, that the influence
of heat upon proteins consists, among other effects, in shifting
the equilibrium among the polymeric modifications of the am-
photeric protein molecule in the direction of higher complexes.
An alternative hypothesis, which would cover the above facts,
is that casein acts as a dibasic acid and that at room temperatures
salts of the type Na,X(OH), are formed while at higher tempera-
tures acid salts of the type NaH X(OH), are formed. The fact that
solutions of both the neutral and ‘‘basic’’ caseinates obey Ostwald’s
dilution-law for a salt of a monobasic acid, however, excludes
this possibility. A possible source of error in the above experi-

1T. Brailsford Robertson: Journ. of Physical Chem., xi, p. 542, 1907;
Ibid. (shortly to appear).

T. Brailsford Robertson 153

mental determinations may be mentioned. Solutions of the
neutral caseinates undergo fairly rapid auto-hydrolysis, about
one-third being hydrolyzed in twelve hours at 37°. This effect
would of course be negligible in the short period during which
the solutions are being prepared, but at higher temperatures the
velocity of hydrolysis would probably be increased and this
might conceivably vitiate the accuracy of the titrations. Special
determinations made with a view to estimating the magnitude
of the error thus introduced showed that it was nearly inappre-
ciable. Thus too cc. of 46 X 10° *N KOH at 88° dissolves 1.13
gram of casein, the solution having been allowed to stand in
the thermostat for one-half hour. After three hours in the thermo-
stat the titration indicated that 1.25 gram had been dissolved.
The error at 88° in half an hour would therefore be, in this solu-
tion, about .o4 gram and at lower temperatures and in more
dilute solutions it must, of course, be considerably less. It may
here be noted that in none of the solutions in which more casein
was dissolved at higher temperatures than would be dissolved
at room temperature was any appreciable tendency towards
precipitation of casein on cooling observed, although in many
cases there was a marked increase in the opalescence of the solu-
tion. This is, however, not surprising since an appreciable
amount of acid may be added to a solution of alkali “‘saturated”’
with casein at room-temperature before precipitation occurs.
Such solutions are possibly to be regarded as being “‘super-
saturated’”’ with casein and in a condition of unstable equilibrium.

CONCLUSIONS.

(1) The solubility of the casein in alkaline solutions is con-
siderably augmented by carrying out the process of solution
at temperatures above 40° C.

(2) It is pointed out that this fact is not in harmony with the
view that a rise in temperature increases the degree of hydrolytic
dissociation of solutions of the caseinates.

(3) In explanation of this fact and of the increase in alkalinity
and electrical conductivity of caseinate solutions upon heating,

1T. Brailsford Robertson: This Journal, ii, p. 317, 1907.

154 Influence of Temperature upon Solubility of Casein

which were observed by Osborne, it is suggested that the influence
of heat upon proteins consists, among other effects, in shifting
equilibria of the type:

HXOH + HXOH @ HXXOH + H,O

in the direction of higher complexes, and that heat-coagulationis
a result of repeated condensations of this type.

(4) The solubility of casein in solutions of various concentra-
tions of potassium hydroxide, lithium hydroxide and calcium
hydroxide at various temperatures has been determined.

NOTE ON THE APPLICABILITY OF THE LAWS OF AMPHO-
TERIC ELECTROLYTES TO SERUM GLOBULIN.

By T. BRAILSFORD ROBERTSON.
(From the Rudolph Spreckels Physiological Laboratory of the University
of California.)
(Received for publication, June 3, 1908.)

In a recent number of this Journal, H. Lundén!' has raised a
number of objections to statements and computations made by
me in my paper on the dissociation of serum globulin at vary-
ing hydrogen ion concentrations.” While in dealing with his
objections, I should have preferred to bring forward additional
experimental matter, yet as pressure of other work forbids
this just at present and as Dr. Lundén’s statements are, in
many cases, very misleading, I have thought it advisable not to
delay publication of the following short review of some of the
more important points raised by him. The first of my state-
ments to which Lundén takes exception is that to the effect
that the method which has frequently been used to determine the
dissociation-constants of amphoteric electrolytes, namely, by
estimating the hydrolysis of a salt and proceeding to calculate
the dissociation-constant in the usual manner for non-amphoteric
electrolytes ‘“‘can only give even approximately accurate values
for the larger function if it be sufficiently large compared with
the other, otherwise the dissociation-constants obtained in this
way are subject to considerable error,’’ and he proceeds to demon-
strate that the method can be used even for determining the
smaller constant without appreciable error. That an error
is introduced by this procedure, however, he admits, and he
proceeds to estimate the magnitude of the error under varying
conditions. The computations thus tabulated by Dr. Lundén are
of considerable interest and value and I am gratified that my
imperfectly accurate statement should have elicited them. Itis

1H. Lundén: This Journal, iv, p. 267, 1908.

2T. Brailsford Robertson: Journal of Physical Chem., xi, p. 437, 1907.

I)

156 ‘‘Amphoteric Electrolytes ”’

a question, however, to what extent my objection to the method
has been invalidated by Dr. Lundén’s computations in so far
as the particular problem under consideration is concerned,
namely, the equilibrium of proteins in acid or alkaline solution.
Lundén estimates that if the dissociation-constant which is
being determined is less than 10 * while the other dissociation-
constant is less than 10° and the dilution is less than 100 then
the correction which must be applied to the constant as deter-
~ mined by the usual method is less than 1 per cent. As he points
out, however, at dilutions greater than this or if the dissociation-
constant which is not being determined is greater than 107 * then
the correction may assume considerable importance and he illus-
trates this fact by referring to thecaseofasparticacid. Nowequi-
libria in protein systems are, of necessity, usually investigated
at high dilutions and consequently the correction may, in these
cases, be expected to be of appreciable magnitude. Moreover,
Dr. Lundén’s statement (p. 280 of his paper referred to above)
that cases in which one of the constants is greater than 107° are
not likely to occur among the proteins betraysa lack of familiarity
with these bodies. Thus casein is practically insoluble in water!
yet a suspension of casein in water will displace carbonic acid
from carbonates.? Since the molecular concentration of dissolved
casein, under these conditions, must be exceedingly minute
casein must be a considerably stronger acid than carbonic acid.
Serum globulin, judging by its general behavior, cannot be far
inferior in strength to casein while some of the mucins and nucleic
acids are not improbably stronger.

In my paper upon serum globulin the various symbols employed
have the following significance:

The protein molecule ( = HXOH) is in equilibrium with the
acid HCl, in the solution we have the following ions and mole-
cules in the concentrations mentioned below:

Ht OE ORES HXt HXOH XX Cl> (ia
a b c d e ij r Le

1T. Brailsford Robertson: This Journal, ii, p. 317, 1907; van Slyke and
van Slyke: Amer. Chem. Journ., Xxxviii, p. 383, 1907; this Journal, iv, p.
259, 1908.

2'W. A. Osborne: Journ. of Physiol., xxvii, p. 398, 1901.

T. Brailsford Robertson We7

The ionic velocities of the H*+ and Cl~ ions in cm-sec. are writ-
ten, respectively, U and V while that of the protein ions (assumed
to be equal for the positive and negative ions) is written v. The
total amount of hydrochloric acid present, combined and uncom-
bined, is written a, while the quantity a, — a which is the amount
of acid neutralized (determined by the gas-chain) is written m.
The difference between the measured conductivity of the solution
and the calculated conductivity of the uncombined HCl is writ-
ten A while the constant quantity

is written H.

On p. 281 of the paper referred to Lundén states that ‘‘from the
equations [16], [13] and [10] in the paper of Robertson it follows
that:

A eee
Pee ie aa
k#=m—d+c
jeience -
| eb
and

PHO 1 d—c—a+ta—-a=a,

Robertson consequently makes the supposition that the salt is
completely ionized (4 = 0) and he also makes the supposition
that the concentration of the Cl~ ion is equal to the total con-
centration of Cl (y = a,). These are the same suppositions as
I have made in the former part of this paper.”

Dr. Lundén is evidently endeavoring to convey the impression
that I have overlooked or omitted to mention certain assumptions
hidden in my equations. Were Lundén’s quotation accurate
this would certainly be the case but it is not and although criti-
cism is welcome I must beg leave to protest against misquotation.

1TIn Lundén’s paper this equation is erroneously written # = m—d—c.

158 ‘‘Amphoteric Electrolytes”’

The statements actually made in my paper (p. 440) are as fol-
lows:

; 2u Ww
2 ie ae 9 \
2U of or ee V2 oy
i A V?-—v a’ — H
TE Sa Gee i a? arn

In practice it is found that the term:

a’ — H

is negligible, when the ampholyte is chiefly acid, except at com-
paratively high concentrations of the acid HA (HCl), so that
for low concentrations of the acid HA (HCl) we have:

This latter equation being the one quoted by Lundén. It will
be seen that the term which experiment showed could be rejected
is equal to » (by substituting from equations [10] and [11] in
my paper) but at no step in my analysis was the supposition
made that it was equal to zero: on the contrary, for the values
of the constants given in my paper it was an experimentally
ascertained fact that the value of “ was very small and it is
Lundén who makes the supposition that it is negligible and not
I. In fact in Table III of my paper I evaluate yu for the various
dilutions employed; this would hardly have been a consistent
procedure on my part I had previously assumed that “ was zero.

In what follows I cannot subscribe to Dr. Lundén’s logic.
Adopting the value for v which I found to fit the equations
evaluated above and assuming that it is correct he proceeds to
demonstrate, by its help, that these equations do not apply to
the system under consideration; later on he uses my constant

Yield
: a for the same purpose; but if, as Dr. Lundén supposes,
“b

w

T. Brailsford Robertson 159

the above equations are not applicable to this system then the
constants deduced by employing them certainly cannot be
accurate and Dr. Lundén is not justified is assuming that they
are correct in order to demonstrate that the equations are
erroneous. It isnot necessary, however, to impugn Dr. Lundén’s
logic in order to account for certain of the contradictions which
he has discovered between my experimental data and his equa-
tions. If Dr. Lundén will peruse the ninth line of print on p.
446 of my paper he will discover that in applying equation [21]
of his paper he has equated quantities determined at 25° to
quantities determined at 18°; this being the case it is not improb-
able that Dr. Lundén’s statement that ‘‘the values of d and c
calculated from equations [21] and [23] are all absurd”’ will meet
with general acceptance. Asa matter of fact Lundén’s equa-
tion [21] cannot, with accuracy, be directly applied to these
observations because although we know with considerable accu-
racy the temperature-coefficient of the conductivity of solutions
of hydrochloric acid we do not, with such accuracy, know the tem-
perature-coefficient of the ionic velocity of Cl~ nor the manner in
which this depends upon the concentration of the solution. In
the equations which I employed, on the contrary, since the only
temperature-coefficient which had to be introduced was that for
hydrochloric acid of concentration = a the introduction of the
correction for temperature involved no appreciable inaccuracy.

Dr. Lundén further endeavors, using the value given in my

wk
paper for the constant —“—" to apply the following equations

Ry

to the system:
d a’ ky
ea aay

and obtains values for d and ¢ which, although closely commen-
surate with those given by me in Table VI of my paper are
nevertheless negative for the lower concentrations of H* ions;
from this fact, which owing to the method of calculation adopted
by me, had escaped my notice, and from the fact that his equa-
tions [21] and [23] do not yield values commensurate with those

160 ‘“Amphoteric Electrolytes”’

obtained from these equations Dr. Lundén concludes that “‘the
formule used by Robertson do not represent the chemical equi-
libria in globulin solutions.’’ Two alternative possibilities, how-
ever, which do not appear to have suggested themselves to Dr.

Ww

Ry
that the observation in question isinaccurate. A simple criterion
enables us to decide which of these two alternatives is the correct
one. From equation [13] of my paper it follows that in all cases
A must be greater than m(V + v) and since V + v cannot be less
than V it follows that 4 must always be greater than mV ; other-
wise the conductivity of the solution would be less than the con-
ductivity due to the H* and Cl~ ions present in the solution,
which is impossible. Now V at 18° is 65.8 X 10 °cm-sec. and
at 25° it cannot be less than this, hence in all cases A must be
greater than 65.8 X 10 ° m, thatis, for the observation m = 14.9 X
to * (for which Lundén finds negative values of d and c) can-
not be less than 9.8 X 107’, whereas the observation actually
recorded is A = 7.8 X 107‘; hence this observation must be
rejected ; the remaining observations do not violate this criterion.
Since the observation immediately following that rejected above
is also untrustworthy (cf. p. 448 of my paper) there remain four
observations which, as Lundén’s and my computations show,
obey the equations evaluatedin my paper fairly accurately and
lead to consistent and intelligible values of c andd. I do not
consider, however, that four observations are sufficient to enable
us to state that serum globulin obeys the laws for dissociation of
an amphoteric electrolyte in other than a qualitative manner;
still less do I consider them adequate to sustain the far-reaching
conclusion of Dr. Lundén that the equations evaluated in my
paper do not represent the chemical equilibria in globulin solu-
tions.

Dr. Lundén objects to the endeavor to apply the laws for the
dissociation of a monovalent amphoteric electrolyte to proteins
because they contain more than one COOH or NH, group.
While this fact is undeniable it nevertheless.does not follow that
the laws which apply to the dissociation of a monovalent elec-
trolyte cannot apply to the dissociation of these bodies. It is
well known that many di- or tri-valent acids and bases behave,

Lundén are that my value for the constant = is too large or

7 a

T. Brailsford Robertson 161

in solution, essentially as if they were monovalent, the reason
being that the second and third hydrogen (or hydroxyl) ions are
split off with increasing difficulty and in indefinitely small quan-
tities! Thus succinic acid obeys Ostwald’s dilution-law for a
monovalent acid and in titrating against bases with methyl-
orange indicator can also be regarded as monovalent. As a
matter of fact I have shown that solutions of the neutral casein-
ates of ammonium and sodium obey Ostwald’s dilution-law
for a salt of a monovalent acid;? since, therefore, it has been
definitely shown in the case of one protein that only one COOH
group is appreciably concerned in determining the equilibrium
of the system it is not illogical to assume, until experiment proves
otherwise, that similar conditions prevail in the equilibria of
other proteins.

In the latter part of his paper Lundén devotes some space to
demonstrating that my estimates of the molecular weight of
globulin from conductivity and gas-chain data can only be
approximate in character. I think a perusal of my paper,
referred to above, will make it clear that I was sufficiently aware
of this fact and that the essentials of Dr. Lundén’s criticism of
these computations were adequately dealt with therein. Since
unionized and uncombined serum globulin is only inappreciably
soluble* the ‘“‘internal salt’? and the unhydrated form of the
serum globulin could only be present in inappreciable quantities
and Lundén’s contention that their presence invalidates the deter-
mination of the molecular weight by these methods fails to
apply to the case under consideration while it again reveals
the author’s lack of familiarity with the proteins. The error
due to the ‘“‘association”? which the protein undergoes in solu-
tion is fully discussed in my paper.’

1Cf. E. E. Chandler: Journ. of the Amer. Chem. Soc., xxx, p. 694, 1908,
in which the literature is also given.

*T. Brailsford Robertson: Journ. of Physical Chem., xi, p. 542, 1907.

$ Clarence Quinan: Univ. of Calif. Publ. Pathol., i, p. 1, 1903; Cf. also
Gustav Mann: Chemistry of the Proteids, London, pp. 361, etc., 1906,

4It may here be noted that the potential difference between a hydro-
gen electrode in + N hydrogen ion solution and the calomel electrode (with
*, KC) is, by a clerical error, given in my previous paper as .326 volt. It
should of course, be .336 volt.

X. RESEARCHES ON PYRIMIDINS: THE ACTION OF
DIAZOBENZENE SULFONIC ACID ON THYMIN,
URACIL AND CYTOSIN.

(Thirty-fourth Paper.)
By TREAT B. JOHNSON AND SAMUEL H. CLAPP.
(From the Sheffield Laboratory of Yale University.)

(Received for publication, June g, 1908.)

We have, at the present time, practically no knowledge of the
way in which the pyrimidins—thymin, uracil and cytosin—are
linked in the nucleic acid molecule. The question whether they
are actually as such contained in the nucleic acids, from which
they are obtined, has not been settled. The recent work of
Osborne and Heyl! on triticonucleic acid, and of Levene and
Mandel? would seem to indicate that they do not result from the
purin bases,’ but that the idin nucleus is present in nucleic
acids in the simple form.

In order to obtain new data, which might prove of service in
settling the question of the nature of the linking of pyrimidins
in nucleic acids, we undertook this investigation. We shall
describe the behavior of diazobenzene sulfonic acid on thymin,
uracil, cytosin and some of their alkyl derivatives. A sum-
mary of the results of our experiments, and their significance, 1s
given at the end of this paper.

Burian® has investigated the action of diazobenzene sulfonic
acid on several nucleic acids and purins. He examined the
nucleic acids from sperma, thymus, yeast and spermatozoa of the

1 Amer. Journ. of Physiol., xxi, p. 157.

? Biochem. Zeitschr., ix, p. 233.

8 Asher-Spiro: Ergeb. d. Physiol., v, p. 795, 1905; Burian: Zeztschr. f.
physiol. Chem., li, p. 438, 1907.

‘Johnson and Clapp: This Journal, v, p. 40.

5 Ber. d. deutsch. chem. Gesellsch., xxxvii, p. 708; Zettschr. f. physiol.
Chem., li, p. 435.

163

164 Researches on Pyrimidins

herring; and states that they do not react with this reagent. He
showed, on the other hand, that the purins—xanthin, hypoxan-
thin, guanin, adenin and theophyllin—in which the hydrogen in
position 7 is unsubstituted, react with diazobenzene sulfonic
acid, in presence of alkali, giving intensely colored compounds.
He regards the compounds formed as diazoamino derivatives of
the general formula I. Substitution in the pyrimidin (alloxan)

LN (60

2€ 5c —N—N: NC,H,SO,H
| pee 2 o.
3N———4C —N
9

u

ring apparently had no influence on the reaction. On the other
hand, he observed that purins substituted in position 7—theo-
bromin, caffein—and also uric acid, do not react with the diazo
acid. Burian concludes. from these results that@the xanthin
bases are linked in the nucleic acids at the nitrogen atom in posi-
tion 7.

So far as the writers are awa vans' was the first investi-
gator to observe that a aw derivative reacts with a
pyrimidin giving a colored compound. He found, for example,
that 2-oxy-4,6-dimethylpyrimidin, II,

N=—=C.CH,
|
CO CH
| |
NE ==G6Ee
i
combines with diazobenzene chloride, in presence of alkali, giv-
ing a red compound. He says: ‘‘Es ist dies ein sehr kraftiger
Farbstoff.’? The compound was unstable and no formula was

assigned it.
Steudel*® afterwards observed that natural thymin reacts with
diazobenzene sulfonic acid, in alkaline solutions, giving an intense

1 Journ. f. prakt. Chem., x\viii, p. 489.
? Zettschr. f. physiol. Chem., xlii, p. 170.

Treat B. Johnson and Samuel H. Clapp 165

red color. Pauly also mentions this diazo reaction in a later
publication! and states that it cannot be used to distinguish
between the pyrimidin and imidazol rings. He observed, for
example, that 4-methyluracil gives as intense a color with diazo-
benzene sulfonic acid as histidin.

We now find that not only thymin and 4-methyluracil, but
also uracil, 4,5-dimethyluracil, cytosin, 5-methylcytosin and 5-
bromuracil react under proper conditions with diazobenzene sul-
fonic acid giving red colored solutions. Thecolors obtained with
thymin, 4-methyluracil, 4,5-dimethyluracil and 5-methylcytosin
are much more intense thant those obtained with uracil, cytosin
and 5-bromuracil. Apparently, the character of the groups
occupying the 4 and 5 positions has a decided influence on the
intensity of the color.

We have also made the interesting observation that the forma-
tion of red colors is entirely inhibited by substitution in the 3
positions of the uracil, cytosin and thymin molecules. The
assumption that the diazo compound might react at the 4 posi-
tion of the pyrimidin ring”? giving azo derivatives isexcluded by
Pauly’s observation? and also by the fact that 4, 5-dimethyluracil
reacts with the diazo acid giving a red color. Substitution in
position 1 of thymin does not prevent the formation of a red color.
t-Methylthymin,‘ for example, gave as intense a color as thymin
itself. On the other hand a red color was not obtained with the
diazo acid when the hydrogens in positions 1 and 3 or 3 alone of
thymin were substituted by methyl groups.

Similar observations were also made in the cases of uracil and
5-bromuracil. While these pyrimidins gave red colors with the
diazo acid, no colors were obtained when 1,3-dimethyluracil’
and 3-methyl-s-bromuracil were tested with the reagent. The
corresponding 3-methyluracil is unknown. The formation of a
color in the cases of cytosin, III, and 5-methylcytosin, VI, is not
dependent upon the presence of the free amino radical. 2-Oxy-

1 Zettschr. f. physiol. Chem., xiii, p. 512.

? Burian: Zeitschr. f. physiol. Chem., xiii, p. 297; Mann’s Chemistry of
the Proteids, p. 431, 1906.

EQitoc. cit.

4Johnson and Clapp: Loc. cit.

5 Ibid.

166 Researches on Pyrimidins

6-methylphenylaminopyrimidin V,' gave as intense a color as
cytosin, III, while, on the other hand, no color was obtained with
3-methylcytosin, IV, 3,5-dimethylcytosin, VII, and 2-oxy-3-
methy1l-6-methylphenylamino-pyrimidin, VIII.

N=—=CNE. N — CNH, N— CN (CH,) (C,H,)
| | | |
CO! "CH CO” CH CO CH
| | | | | |
NH—CH CH,N——CH NH — CH
UI IV V
N ——CNH, N =CNH; N==—CN (CHG
| | | |
COR. CCH. Aa Oe 9 CORTE CO CH
| | | | |
NH — CH CH,N CH CH N= = =0r
VI VII VIII

The remarkable tendency of certain pyrimidins to react with
diazobenzene sulfonic acid, and the apparent inertness of the
same ring in purins are of especial interest. According to Bu-
rian’? substitution in the pyrimidin nucleus of purins did not influ-
ence the reaction with the diazo acid. While purins are closely
related in structure to pyrimidins it is important to point out
in this connection that positions 1 and ‘3 of the pyrimidin ring
do not necessarily correspond to positions 1 and 3 in purins.
While the pyrimidin nucleus is symmetrical, the purin, on the
other hand, is unsymmetrical with respect to the 2 position.
Two purins can theoretically be formed from a pyrimidin ring
according as the glyoxalin ring is joined at the 4,5 or 5,6 posi-
tions. An inspection of the formulas below will show that this
involves a change in the numbering of the atoms in the pyri-
midin nucleus.

6 4
iN——C N——C SN C—N?®
| eset tat | L a
2c 5C—N?7 da. C C as aC 5C—N
| ec Leal | ka
3N—-—_C —N N——C 1N oA
4 9
Purin. Pyrimidin. Purin.

1 Johnson and Clapp: Loc. cit.
2 Loc. cut.

Treat B. Johnson and Samuel H. Clapp 167

It is also of interest to note here that 6-oxypyrimidin' and
6-aminopyrimidin’ do not give a red color with diazobenzene
sulfonic acid. Whether this reagent reacts to form colored com-
pounds only with pyrimidins having a —CO-NH— grouping in
the 1, 2 or 2, 3 positions must be decided by further experi-
ments.

The compounds formed by the action of diazobenzene sul-
fonic acid of pyrimidins appear to be more unstable than those
formed in the case of purins. Two attempts to isolate the reac-
tion-product in the case of thymin were unsuccessful.

The diazo acid used in our work was prepared according to the
directions of Pauly.* That the test shall not be found to be
capricious it is absolutely necessary that the acid be pure and
freshly prepared.

METHODS OF APPLYING THE TEST.

1. With ro per cent sodium hydroxide solution.

Five to ten milligrams of the pyrimidin are dissolved in 0.5 cc.
of 10 per cent sodium hydroxide solution and about 5.0 milli-
grams of diazobenzene sulfonic acid then added to the solution.
If no reaction takes place the solution usually assumes a yellow
or orange color. When a red color is formed it usually develops
quite rapidly. The red color is usually quite permanent, lasting
in some cases (thymin) for several hours.

2. With 3, sodium hydroxide solution.

Dissolve 5.0 to 10.0 milligrams of the pyrimidin in 2 ce. of
py Sodium hydroxide solution and add 5.0 to 10.0 milligrams of
the sulfonic acid. This method of testing is not as reliable as
Method 1.

3. Testing on a watch glass.

Five to ten milligrams of the pyrimidin and an equal weight
of the sulfonic acid are mixed together, with a glass rod, on a dry

1 Wheeler: This Journal, iii, p. 285, 1907.

? Bittner: Ber. d. deutsch. chem. Gesellsch., xxxvi, p. 2232, 1903: Wheeler:
Loc. cit.

3 Zeitschr. f. physiol. Chem., xiii, p. 516.

168

watch glass.

is then allowed to flow into the mixture.
immediately under these conditions.

Researches on Pyrimidins

A drop of to per cent sodium hydroxide solution
The red color develops
This method of applying

the diazo test is recommended on account of its delicacy and

The results of our experiments are given, for comparison, in the

TABLE

reliability.
following tables 3
Pyrimidins.
NH——CO
he ae
NEE —CH
(Thymin).
CH,N———CO
co CoH
ae

NH——CO
| |
CO CCH
| |
CH,N CH
(3-Methylthymin.)
CH,N——CO
|
CORRCCH:
| |
CH,N CH
(1, 3-Dimethylthy-
min.)

0.5 ee. 10 per cent
NaOH solution.

|
Intense red color,
which disappears
on diluting with
water.

|

Intense red color,
which disappears
on diluting with
water.

No color.

No color.

2.0cc. X NaOH

10
solution.

Light red which
fades rapidly.

No color.*

No color.

No color.

Test on watch-
glass.

Intense red color.

| Intense red color.

No color.

No color.

|

*N ote—The statement—no color—indicates that no red color was formed.

Treat B. Johnson and Samuel H. Clapp

169
TABLE II.
= eee 05 cen10 2.0ce. N. NaOH T, 2
eine: Ravidatieion solution, |. 0s eaten
NH——CO
a Le Red color but not Red color heel
| I as intense as with) developed slowly. Red color.
NH CH thymin. | Color permanent
(Uracil.) |
CH.N co | |
| |
CO CH
l Red color. Red color. Red color.
NH CH
(1-Methyluracil).
NH——CO
| |
CO, = CH
| |
CH,N CH
(3-Methyluracil)
CH, CO
|
CO CH
| I No color. No color. No color.
CH,N. CH
(1, 3-Dimethyluracil) |

170

TABLE IIt.

Pyrimidins.

0.5 ec. 10 per cent
NaOH solution.

aco
| |
CO CBr
| |
NH CH

(5-Bromuracil.)

CH,N CO
|
CO CBr
| |
NH——CH
(1-Methyl-5-brom-
uracil. )
NH——CO
| |
CO CBr
| |
CH,N CH
(3-Methyl-5-brom-
uracil).
CH,N——CO
| |
CO CBr
| |
CH,N———CH

(1, 83-Dimethyl-
5-bromuracil.)

Red color.

Red color.

No color.

No color.

2.0 ce. =F NaOH
solution,

Faint red color
which develops:

slowly.

Researches on Pyrimidins

Test on watch-
glass.

Faint red color.

No color.

No color.

No color.

No color.

ee

Treat B. Johnson and Samuel H. Clapp 171

TABLE Iv.
| 2.0cc.™ NaOH as i
Pyrimidins. | uOiadetioe. solution. | ae io
nN — C.NH, |
CO CH.H,O /|Red color which : Red color like
oT developed slowly. Faint red. uracil.
NH es CH |
(Cytosin.) |
N=CNH, |
CO CH 2 = P
I No color. No color. No color.
CH,N — CH |
(3-Methyl-cytosin).
CH, | |
N= one
| | \C.Hs . |
CO CH Red color. No color. | Red color.
| | |
NH — CH |
CH, |
oc ong” |
| ei No color No color Maaer as
CO CH No color. No color. No color.

| |
CH.N — CH

172 Researches on Pyrimidins

TABLE V.
rites 0.5cc.10percent | 2-0 ce. at NaOH Test on watch-
Pyrimidins. NaOH solution. solution. glass.
IN CME
CO CCH, Intense red color.; Faint red which Intense red.
| | faded. |
el —= Osl
(5-Methyl-cytosin
NI CNH
CO CCH, No color. No color. No color.
|
CHEN — CH
(3, 5-Dimethyl-
cytosin.) |
SUMMARY.

(1) Thymin, uracil and cytosin react with diazobenzene sul-
fonic acid, in presence of alkali, giving red colored solutions.

(2) The color is given by thymin with greater intensity than
by uracil and cytosin.

(3) Substitution in position 3 of the pyrimidin ring prevents
the formation of a red color.

(4) Accepting the statement of Burian,’ that nucleic acids
do not react with diazobenzene sulfonic acid, the foregoing ob-
servations seem to indicate that thymin and probably uracil
and cytosin as well are linked in nucleic acids at position 3.

N= 66
nea
COawe
(3) N C

(5) Whether the pyrimidins are linked to phosphorus, a car-
bohydrate complex or otherwise must be decided by further
study.

Ber. d. deutsch. chem. Gesellsch., Xxxvii, p. 708; Zeitschr. f. physiol.
Chem., li, p. 435.

FURTHER STUDIES OF THE MODE OF OXIDATION OF
PHENYL DERIVATIVES OF FATTY ACIDS IN THE
ANIMAL ORGANISM.

(PHENYLBUTYRIC ACID, PHENYL-3-OXYBUTYRIC ACID, PHENYLACE-
TONE, PHENYLISOCROTONIC ACID, PHENYL-/,;-DIOXY-
BUTYRIC ACID.)

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

(Received for publication, August 12, 1908.)

In previous papers! it was shown that phenylpropionic acid,
at least in part, underwent oxidation in the animal organism in
accordance with the following scheme:

C,H, CH,.CH,-COOH
(phenylpropionic acid)

C,H, Ci (OH):CH,. COOH
(phenyl--oxypropionic acid)

[C,H,.CO.CH,.COOH]?
(benzoylacetic acid)

C,H,.CO.CH,
(acetophenone)

C.H,.COOH
(benzoic acid)

Benzoylacetic acid was not detected but its formation was in-
ferred from the production of acetophenone, into which it readily
passes through loss of carbon dioxide. At the same time it was
thought probable that some benzoic acid was formed without pass-
ing through the stage of acetophenone.

The close analogy between the apparent mode of catabolism of
phenylpropionic acid and that of butyric acid made it desirable

1 This Journal, iv, p. 419, 1908; Beitr. z. chem. Physiol. u. Pathol., xi, p.
404.
ii

174 Oxidation of Phenyl-fatty Acids

to extend the investigation to other aromatic derivatives of the
fatty acids in the hope of obtaining further insight into the
processes of tissue oxidation of the fatty acids of physiological
importance. The present communication deals with the fate in
the body of a number of derivatives of phenylbutyric acid.

The fate of phenylbutyric acid in the organism has already
been investigated by F. Knoop! who administered it by mouth
to a dog and observed the subsequent excretion of phenaceturic
acid in the urine. This result was interpreted in the light of
Knoop’s well-known hypothesis of $-oxidation and indeed was
one of the most important facts upon which his theory was based.
The conversion of phenylbutyric acid into phenaceturic acid in-
volves the intermediate formation of phenylacetic acid which is
then paired with glycocoll:

C,H,:CH,.CH,.CH,.COOH — 3 C,H;.CH,.COOH — C,H,.CHe COumIEn
CH,.COOH

It is clearly desirable to ascertain the intermediate steps in
the conversion of phenylbutyric acid into phenylacetic acid. So
far as I am aware no picture of the mechanism of the reaction
has hitherto been put forward. In order to elucidate this
mechanism it was natural to inquire first of all whether the
catabolism of phenylbutyric acid did not follow upon the same
line as that of phenylpropionic acid. Judging by analogy one
might expect the change to be as follows:

C7H; CH, CH? CHe- COOH
(phenylbutyric acid)

C,H;-CH,.CH(OH)-CH;: COOH
(phenyl-§-oxybutyric acid)

C,H, -CH,:€O:CH; COGe
(phenylaceto-acetic acid)

C,H; CHe-CG:GH,
(phenylacetone)

C,H;.CH,.COOH
(phenylacetic acid)

* Beitr. z. chem. Physiol. u. Pathol., vi, p. 155, 1904.

H. D. Dakin pA

The occurrence of the first step in this series of changes, namely,
the formation of phenyl-f-oxybutyric acid, is highly probable,
as is shown by the following facts:

(1) Asmall quantity of a levorotatory substance giving the
reactions of phenyl-$-oxybutyric acid was isolated from the urine
of dogs that had received subcutaneous injections of sodium
phenylbutyrate in fairly large doses. The f-oxy-acid could not,
however, be isolated in a state of purity.

(2) Phenyl-G-oxybutyric acid injected in the form of its
sodium salt was excreted in the form of phenaceturic acid, i. e.,
the same end-product as phenylbutyric acid itself yields.

There was little hope of isolating the substance corresponding
to the second hypothetical step in the reaction, namely, phenyl-
aceto-acetic acid, for this body has not yet been synthesized and
would doubtless be unstable. It is improbable, however, that it
was present in the urines of the dogs which had received injections
of sodium phenylbutyrate, because in this case phenylacetone
would have been found in the distillates, for a @-ketonic acid of
the type of phenylaceto-acetic acid doubtless would lose carbon
dioxide on boiling, with formation of the corresponding ketone.

The third possible stage in the catabolism of phenylbutyric
acid involving the formation of phenylacetone, corresponding
to the intermediate production of acetophenone from phenyl-
propionic acid, was definitely excluded on the basis of the follow-
ing experimental results:

(1) No trace of phenylacetone could be detected in the urines
of animals which had received injections of considerable quan-
tities of phenylbutyric acid. ;

(2) Phenylacetone cannot be an intermediate stage in the
catabolism of phenylbutyric acid for when administered to dogs
it results in the excretion of hippuric acid. Phenylbutyric acid
under similar conditions gives phenaceturic acid.

To sum up: Evidence has been obtained that phenylbutyric
acid when oxidized in the body passes through the stage of phenyl-
B-oxybutyric acid. No evidence could be obtained of the for-
mation of phenylaceto-acetic acid and it could not be detected
inthe urine. The possibility of its formation as an intermediate
product is not, however, excluded. Phenylacetone is certainly
not a product of the catabolism of phenylbutyric acid.

176 Oxidation of Phenyl-fatty Acids

It is clear, therefore, that the modes of catabolism of phenyl-
propionic acid and of phenylbutyric acid, though similar as re-
gards the primary formation of a §-oxy-acid, differ in that ketone
formation takes place in the case of phenylpropionic acid but
not in that of phenylbutyric acid. The catabolism of phenyl-
butyric acid is therefore to be represented as follows:

CyHy.CH,.CH,.CH;.COOH
CupcrienoncriconH
[C,H,.CH,.CO.CH,.COOH] ?
etn aa

The mechanism of the catabolism of phenylbutyric acid ap-
pears to be of interest from several points of view. In the first
place it furnishes an additional example of the primary oxidation
of the hydrogen attached to the §-carbon atom of a phenyl-fatty
acid with formation of a leavorotatory $-oxy-acid. In the second
place it indicates the possibility, which has been insisted on in
previous papers, of a f-oxy or ~-ketonic acid undergoing oxida-
tion without intermediate ketone formation.1 Indeed it would
appear as tf ketone formation were restricted to the simplest members
of the fatty acids and phenyl-fatiy acids, theoretically capable of
ketone formation, namely, butyric acid and phenylpropiome acid:

R.CH,.CH,.COOH — R.CH(OH).CH,.COOH >
R.CO.CH,.COOH — R.CO.CH,

This hypothesis furnishes an explanation of why under con-
ditions such as diabetes in which large quantities of ketones are
excreted only acetone has keen detected. If the above scheme
represented a perfectly general type of reaction, it would be
hard to explain why in diabetes the catabolism of acids such as
caproic and valeric do not give rise to the excretion of propyl-
methyl ketone or ethylmethyl ketone respectively :

CH,.CH,.CH,.CH,.COOH. — CH;.CH,.CO0.CH,cOOH —>
CH, CH CO:CH:

1 Unpublished experiments upon the mode of catabolism of phenylvaleric
and phenyl-$-oxyvaleric acid indicate that the substances resemble phe-
nylbutyric acid in this respect.

——e eT

-

EL Dy Dakin 177

It is at present impossible to decide what rdle, if any, the
f-ketonic acids, other than aceto-acetic acid play in intermediary
metabolism. Apart from aceto-acetic acid and benzoylacetic
acid few if any of these acids of a type which might be formed in
metabolism, have been prepared. It may be that they are
extremely unstable or even incapable of more than momentary
existence. It is well, however, to bear in mind that not only
are §-ketonic acids capable of undergoing hydrolysis according
to the well-known scheme

hm CO.CH, COOH + H,O = R.COOH + CH,COOH

a reaction which fits in well with the progressive degradation of
long-chain fatty acids by the loss of two carbon atoms at a time,
but also that they are extremely easily oxidizable substances.*

If the normal course of metabolism of a straight-chain fatty
acid other than butyric acid and phenylpropionic acid, proceeds
through the £-oxy-acids and possibly the {-ketonic acids, but does
not involve ketone formation, is it not probable that the cata-
bolism of butyric and phenylpropionic acid in part does not in-
volve the intermediate formation of acetone and acetophenone
respectively? There is a certain amount of physiological evi-
dence in support of this view, and so far as I know, there is none
opposed to it. The results of oxidation experiments in wttro
furnish complete chemical analogy for these reactions.’

The excretion of hippuric acid following the administration of
phenylacetone is of interest especially when it is considered that
its lower homologue, acetophenone, also yields hippuric acid.
These changes are similar to the action of ordinary oxidizing
agents (including hydrogen peroxide) which oxidize both ketones
to benzoic acid. It will probably be found that most aromatic
methyl ketones primarily undergo oxidation in the body, so as
to yield acids with two less carbon atoms, except in the case of
acetophenone, in which the carbonyl! group is directly attached
to the nucleus:

C,H,.CO:CH, — C,H,.COOH

C,H,.CH,CO.CH, > C,H,.COOH

1 A communication upon this subject will be made shortly.
2 This Journal, iv, p. 77.

178 Oxidation of Phenyl-fatty Acids

The fate of phenylisocrotonic acid was determined by injecting
solutions of the sodium salt (=2.0 grams of acid) subcutaneously
into cats. The fate of this substance had already been deter-
mined by Knoop! who fed it by mouth to a dog and found that
it was excreted in the urine in the form of phenaceturic acid.
The explanation of this rather remarkable result might con-
ceivably depend upon reduction in the intestinal canal. That
this was not the case was proved by obtaining the same end-
product when the acid was subcutaneously injected.

The fate in the organism of another derivative of phenlybutyric
acid, phenyl-§,7-dioxybutyric acid was also determined. This
acid is certainly not an intermediary product of the catabolism
of phenylbutyric acid since on oxidation in the animal body hip-
puric acid together with a little phenyl-f-oxybutyrolactone pro-
duced by the removal of the elements of water from the original
substance, were isolated from the urine.

a CsH;.COOH > C,H,CO. NH.CH,.COOH
C,H,.CHOH.CHOH.CH,.COOH
Ty OAS Pa OI SINC STONSMOIIALO)
ss

Whether the lactone was excreted as such or formed from un-
changed acid during the analysis is not clear. Lactone formation
is very easily brought about by merely boiling with water or
dilute acid and the lactones of aromaticacids as aclass appear to
be rather resistant to animal oxidation.

The formation of hippuric acid from phenyldioxybutyric acid
is of special interest since it is an example of tissue oxidation tak-
ing place at the y-carbon atom. So far as I am aware it is the
only example of its kind so far recorded. Judging by analogy
with some other tissue oxidations,? one might anticipate the for-
mation of mandelic acid according to the following scheme:

C,Hs.CHOH.CHOH CH,.COOH —+ C,H,.CH(OH).COOH

No definite indication of the presence of this acid could be ob-
tained, however.

i Loc. tt, Pause:
* E.g., phenyl-@-oxypropionic acid and phenyl-8-oxybutyric acid.

la. DD) Dakin 179

EXPERIMENTAL PART.

Preparation of phenylbutyric acid. Phenylbutyric acid has
been synthesized by several methods, the most direct of these
being the reduction of phenylisocrotonic acid with sodium amal-
gamas described by Fittig and Jayne.’ The reduction, however,
is not very easily carried out as was found by Jayne and by
Knoop and also in experiments of my own. Kipping and Hill,?
on the other hand, apparently had no difficulty in effecting the
reduction at the ordinary temperature. On the whole, it was
found more advantageous to employ the following method,
mainly, based on the investigations of Fittig and his pupils.
Phenylparaconic acid, prepared by means of Perkin’s reaction
from benzaldehyde, sodium succinate and acetic anhydride,’ is
distilled 7m vacuo. The distillate, consisting mainly of phenyl-
isocrotonic acid with a little phenylbutyrolactone, is boiled with
twenty-five parts of hydrochloric acid (1 part concentrated acid,
3 parts water by volume) under a reflux condenser, for six hours.
The result of this procedure is to convert about 65 per cent
of the phenylisocrotonic acid into phenylbutyrolactone.t The
acid is separated from the lactone by adding sodium carbonate
till faintly alkaline to the ethereal extract containing the two
substances.® The acid remaining in the aqueous layer is again
treated with hydrochloric acid. In this way an 80 per cent
yield of phenylbutyrolactone is readily obtained. The pheny]l-
butyrolactone is reduced to phenylbutyric acid according to
Shield’s method® by boiling it with ten parts of hydriodic acid
(b. p. 127°) and 1.5 parts of red phosphorus for ten hours. The
phenylbutyric acid is obtained by ether extraction after previous
dilution with water and after removing the ether by evapora-
tion readily crystallizes. The method gives an excellent yield.
Twenty-five grams of phenylparaconic acid gave on the average

¥ Ann. d. Chem., ccxvi, p. 108.

* Trans. Chem. Soc., \xxv, p. 147.

3 Ann. d. Chem., ccxvi, p. 100.

*Fittig and Hadorff: [bid., cccxxxiv, p. 117

5 This method of converting phenylisocrotonic acid into phenylbutyro-
lactone was found far superior to that of Erdmann who employed 33 per
cent sulphuric acid. Jbid., ccxxviii, p. 178.

§ [bid., cclxxxviii, p. 207

180 Oxidation of Phenyl-fatty Acids

12.5 gramsof crude phenylbutyric acid, melting at about 46° to 49°.
After a single recrystallization it melts at 49° to 50°. The
synthesis may be represented as follows:

COOH

|
C,H,.CH —CH =CH, —CO —>' C,H,.CH : CH.CH, COOBs—
| i (phenylisocro tonic acid)

(phenylparaconic acid)

€.H, CH.CH;CH,.CO-5, CH. CH, CH, CH, Cem
|

ik (phenylbutyric acid)

(phenylbutyrolactone)

Fate of phenylbutyric acid in the organism. Phenylbutyric
acid in amounts varying from 5 to 6 grams was converted into
the sodium salt and injected in aqueous solution subcutaneously
into a small dog weighing about 6 kilos. The site of injection
was usually the loose tissue at the back of the animal’s neck and
no ill effects followed the injection. The urine passed during
the next three days was collected and analyzed. It was first
distilled in order to test for the presence of phenylacetone (and
indirectly phenylaceto-acetic acid). In no case could any indi-
cations of this substance be obtained... The distillates were ex-
amined with the aid of paranitrophenylhydrazine,' with the
sodium nitroprusside reaction and with the iodoform test. On
one occasion a minimal iodoform reaction was obtained but it
was not due to an aromatic ketone nor was the amount more
than a negligible trace. A portion of the urine was tested with
ferric chloride but no color reaction was obtained, such as a
8-ketonic acid would be expected to yield. Phenylacetone and
phenylaceto-acetic acid were therefore absent.

The urine after distillation was concentrated, acidified with
phosphoric acid and extracted with ether in a continuous ex-
tractor for thirty-six hours. The ether residue was distilled in
steam and the aqueous solution decolorized with charcoal, con-—

1Phenylacetone forms a beautifully crystalline paranitrophenylhydra-
zone melting after crystallization from alcohol or pyridin at 145° to 145.5°.
It usually crystallizes in rosettes of platelets which are only moderately
soluble, even in hot alcohol. The substance serves well for the identi-
fication of phenylacetone. s

i a ee

A. DP. Dakin 181

centrated and allowed to crystallize. A little over 2 grams of
well crystallized phenaceturic acid were obtained in this way.
The substance crystallized in well formed platelets and melted
after a single recrystallization at 142° to 143°. The mother
liquors from the phenaceturic acid were examined in the polar-
imeter and found to be decidedly levorotatory (0.35° to 0.51°).
The optically active substance was found to be insoluble in ben-
zene but readily soluble in chloroform, in this respect agreeing
with the known properties of phenyl-$-oxybutyric acid. This
substance, however, could not be isolated in a state of purity.
Its presence was made almost certain, however, by the following
reactions: (1) Part of the solution was neutralized with am-
monia and distilled with hydrogen peroxide in the same way as
was employed for the detection of phenyloxypropionic acid.'
On successively redistilling the distillate with ammoniacal silver
solution to remove aldehydes and then with phosphoric acid, a
liquid was obtained with a strong aromatic smell, similar to
phenylacetone and the solution gave a strong iodoform reaction
and gave with sodium nitroprusside deep red coloration in both
alkaline and acid solution, identical with those obtained by using
pure phenylacetone. With paranitrophenylhydrazine in acetic
acid solution a yellow precipitate was obtained. The amount
was too small for complete purification. (2) On treating
another portion of the solution with a little sodium carbonate
and then adding a little strong, cold potassium permanganate
solution, a strong odor of an aromatic aldehyde similar to phenyl-
acetaldehyde was at once obtained.

These reactions, combined with the observed levorotation of
the solution make it extremely probable that phenyl-{-oxy-
butyric acid was present. However, until the substance can be
isolated in the pure state, complete proof must be considered
lacking.

Preparation of phenyl-§8-oxybutyric acid. Great difficulty was
experienced in obtaining this acid. A small quantity was even-
tually obtained by a modification of Fittig and Luib’s methods.?
Phenylisocrotonic acid (40 grams) was boiled with ro mol. parts
of 10 percent caustic soda solution. The solution was then

?This Journal, iv, p. 430.

2 Ann, d. Chem., cclxxxiii, p. 30

to

182 Oxidation of Phenyl-fatty Acids

acidified and extracted with ether. The ethereal residue dis-
solved almost completely in carbon bisulphide so that the quan-
tity of oxy-acid formed must have been very small. The residual
acids consisting of a mixture of phenylisocrotonic acid and phe-
nyl-a,3-crotonic acid were then crystallized from water which
removed the bulk of the former acid. The more soluble acid
remaining in solution was then extracted with ether. On evapo-
ration of the ether the residue, without further purification, was
allowed to stand for three days with 4 parts of glacial acetic acid
saturated with hydrobromic acid gas. On pouring the solution
into water an oil separated which was boiled for three hours with
15 parts of water to which a little sodium acetate had been added.
After cooling, the aqueous portion was filtered off and extracted
with ether. The ethereal residue readily crystallized and was
purified by washing with carbon bisulphide. The yield of pure
acid was only about 3 per cent.' The changes may be repre-
sented as follows:

C,H,;-CH:CH.CH,.COOH — C,H,.CH,.CH:CH.COOH —

(phenylisocrotonic acid) (phenyl-a-/-crotonic acid)
C,H, CH,,CHBr-CH, COOH — C,H;.CH, CHO teem
(phenyl-$-bromobutyric acid) (phenyl-f-oxybutyric acid)

Fate of phenyl-$-oxybutyric acid in the organism. One gram of
the acid was converted into the sodium salt and injected sub-
cutaneously into a cat weighing about 2.5 kilos. The urine
during the next three days was carefully collected and analyzed
exactly as in the case of phenylbutyric acid. No phenylacetone
or phenylaceto-acetic acid could be detected. Four-tenths of a
gram of phenaceturic acid was obtained in the form of crystals
which melted at 142° after a single recrystallization. A very
small amount of unchanged substance appeared to be present,
for the mother liquor from the phenaceturic acid was feebly
levorotatory (—o.10°) and faint indications were also obtained
of its presence by oxidation with hydrogen peroxide and with
potassium permanganate.

1 The acid obtained crystallized in platelets and melted at 98° to 100°. I
have been unable to find any record of the melting point of this acid which
was obtained by Fittig and Luib.

A De Dakin 183

Fate of phenylacetone in the organism. Phenylacetone (Kahl-
baum) was injected subcutaneously into dogs in dilute alcoholic
solution in doses varying from 3 to 4 grams. The absorption of
the ketone appeared to be slow and in one case comparatively
little hippuric acid was obtained (0.3 gram). In all the other
cases an abundant yield of hippuric acid was obtained—in one
case as much as 3.0 grams being recovered. A small amount of
unchanged ketone was excreted in the urine. The hippuric acid
after recrystallization, melted sharply at 187°.

Fate of phenylisocrotonic acid tn the organism. Phenylisocro-
tonic acid was prepared by distilling phenylparaconic acid im vacuo
and crystallizing the distillate from carbon bisulphide. Two
grams of the acid were dissolved in alcohol and almost neutralized
with caustic soda. The solution was injected subcutaneously
into a cat (two kilos). The urine was collected for three days.
It contained a trace of acetone but no aromatic ketone. The
urine was concentrated to about too cc., acidified with phos-
phoric acid and extracted in a continuous extractor with ethyl
acetate. After purifying the ethyl acetate extract by steam
distillation and by boiling the aqueous solution with charcoal,
0.65 gram of pure phenaceturic acid, melting at 142°, was ob-
tained. No hippuric acid could be detected.

Fate of phenyl-8,7-dioxybutyric acid. This acid was prepared
by Fittig and Obermiiller’s method! by oxidizing phenylisocro-
tonic acid in dilute alkaline solution at 0° with dilute potassium
permanganate. The product obtained was a mixture of the free
acid and lactone and was converted into the sodium salt by boil-
ing with excess of caustic soda, neutralizing with acetic acid and
injecting the dilute solution subcutaneously into a cat of about
2.5 kilos weight. One and a quarter grams of the acid in the
form of sodium salt gave about 0.45 gram of pure crystalline
hippuric acid (m.p. 185° to 187°) and about o.2 gram of phenyl-
B-oxybutyrolactone. The lactone was separated from the
hippuric acid as follows: The urine was concentrated, acidified
and extracted with ether in the usual way. The ethereal resi-
due was distilled in steam for a very short time only, then
decolorized with charcoal and concentrated to about 5 cc. The

1 Liebig’s Ann. d. Chem., cclxviii, p. 44.

184 Oxidation of Phenyl-fatty Acids

solution was then made just alkaline with sodium carbonate
solution and extracted with ether to remove the lactone. The
alkaline residue was acidified and again repeatedly extracted
with ether. On evaporation, hippuric acid crystals were obtained
in abundance. The aqueous solution of the ethereal extract was
optically inactive and no positive indication of the presence of
mandelic acid could be obtained.

SUMMARY.

The subcutaneous injection of phenylbutyric acid in the form
of its sodium salt in aqueous solution results in the excretion of
phenaceturic acid as found by Knoop. In addition a small quan-
tity of a levorotatory acid possessing the properties of phenyl-£-
oxybutyric acid was excreted. No phenylacetone was excreted
and phenylaceto-acetic acid could not be detected.

Phenyl-f-oxybutyric acid administered under similar con-
ditions results in the excretion of phenaceturic acid. No phenyl-
acetone could be detected. A part of the oxy-acid is apparently
excreted unchanged and is levorotatory.

Administration of phenylacetone results in the excretion of
hippuric acid, no phenaceturic acid being formed. Phenylace-
tone cannot therefore be an intermediate product of the cat-
abolism of phenylbutyric acid. The probable mode of oxidation
of phenylbutyric acid in the body may be represented as follows:

C,H,.CH,.CH,.CH,.COOH — C,H;.CH,.CHOH.CH,.COOH —>
[C,H;.CH,.CO.CH,.COOH]? — C,H,.CH,.COOH

The phenylacetic acid is excreted in the form of phenaceturie
acid.

A comparison is made between the mode of oxidation of phenyl-
butyric acid and that of phenylpropionic acid. The first step in
the catabolism of both acids apparently consists in the formation
of a §-oxy-acid but in the case of phenylbutyric acid there is no
formation of the corresponding ketone as a product of further
oxidation. The intermediate formation of ketones observed in
the catabolism of butyric and phenylpropionic acids is probably
confined to these two acids and is not a general reaction.

Gl ie La

He} D. Dakin 185

These results are in harmony with the view that in normal me-
tabolism probably only part of the butyric acid and phenylpro-
pionic acid undergoing oxidation passes through the stages of
acetone and acetophenone respectively.

Phenylisocrotonic acid administered subcutaneously to cats in
the form of its sodium salt is excreted in the form of phenaceturic
acid.

Pheny1-,7-dioxybutyric acid administered to cats in the form
of its sodium salt resulted in the excretion of hippuric acid to-
gether with a little phenyl-f-oxybutyrolactone. No indications
could be obtained of the formation of mandelic acid. Phenyl-
dioxybutyric acid therefore does not undergo f-oxidation but
oxidation takes place at the 7-carbon atom. Phenyldioxybutyric
acid is not a product of the catabolism of phenylbutyric acid.

Phenylacetone is readily identified by conversion into its para-
nitrophenylhydrazone which crystallizes from alcohol or pyridin
in sparingly soluble rosettes of platelets melting at 145° to 145.5°.

Note added during proof correction. The investigation of the
fate of phenylvaleric and phenyl-8-oxyvaleric acid has shown
that while the end product of catabolism in both cases is hip-
puric acid, cinnamylglycocoll, C,H;.CH:CH.CO.NH.CH,.COOH
m. p., 193°, is an intermediate product of their catabolism.
This substance is also produced in the oxidation of phenylpro-
pionic acid in the animal body. These observations throw
considerable light upon the mechanism of fatty acid metabol-
ism and will form the subject of a separate communication.

HYDROLYSIS OF VICILIN FROM THE PEA (Pisum Satibum).'
By THOMAS B. OSBORNE anp FREDERICK W. HEYL.

From the Laboratory of the Connecticut Agricultural Experiment Station.)
y g P
(Received for publication, July 23, 1908.)

Vicilin is a globulin which occurs in the seeds of the pea, horse-
bean and lentil. Its isolation, properties and composition have
been given in a series of papers from this laboratory.” It differs
from legumin in containing a little more carbon, a little less nitro-
gen and less than half as much sulphur, and also in being soluble
in much more dilute saline solutions.

Vicilin presents a special interest because it contains the
smallest proportion of sulphur yet found in any protein. A
large number of preparations made and analyzed in this labor-
atory, contained from o.10 to 0.20 per cent of total sulphur.
This difference is not analytical, for the extreme figures were con-
firmed in several cases by very closely agreeing duplicate deter-
minations. The sulphur content of vicilinas heretofore prepared,
is not constant and deserves further careful study, for it would
appear that either the sulphur-containing complex of this pro-
tein is very easily split off, or that the preparations were mixtures
of protein free from sulphur and protein containing sulphur.
Owing to the great solubility of vicilin, it has not yet been sub-
jected to sufficiently extensive fractionation to determine this
latter point. That it is a mixture of sulphur-free protein and
legumin is not probable, for the differences in sulphur content are
not accompanied by the differences in carbon and nitrogen which
would be expected if this were the case.

1The expenses of this investigation were shared by the Connecticut
Agricultural Experiment Station and the Carnegie Institution of Washing-
fon DD. C.
? Osborne and Campbell: Journ. Amer. Chem. Soc., xx, pp. 348, 362,
393, 410, 1898.
187

188 Hydrolysis of Vicilin

It is probable that vicilin is a distinct protein or mixture of
proteins and not an altered product of legumin, for the seeds of
the vetch which contain legumin yield no vicilin. The resem-
blance in properties and composition between vicilin and legumin
makes it desirable to know whether a more positive difference can
be found in the nature and proportion of the amino-acids which
each yields when hydrolyzed.

The results of the hydrolysis are given in the following table
together with those previously obtained for pea legumin.

Vicilin. Legumin.
per cent. per cent.
Giiyicocolll enmncit aie a ees eee hae eee 0.00 0.38
MlaminGax ayia, sania Vat etna aera 0.50 2.08
Wallines ts £ivs pa tlncd pee dency ere eae 0.15 ?
CUCINA nescence 9.38 8.00
IPTOMM Oye. ah inate teeny Ree ee 4.06 3.22
Phenlyalanine.<¢ snus ne cee ree 3.82 3.75
AS parties ACld, ons .0. cleat. pene eeae 5.30 5.30
Glutamainic aeid<ev ty wae a 21.34 16297
STINE ay, Gcryheesien dcr Meaveuanee cee taate : ? 0.53 |
(OWS EIINOM <2 teats eats cei bene ieee not det. not det.
ORY PLONE irene TN teat vere 8 i
FRVAOSIING' “seie hae Meine ok testa ncaaseeneeete cnet 2.38 1.55
SATII I 2 GAN at pee eat gatet Ss facet : 8.91 abet
FUistiGine? ote) senate ae elec 7 ee 1.69
STG VSUING sss, cccalan er Saves ee mega sieie one ue UA oe 5.40 4.98
I NindVoaLeROVes Maier erate OR She atonnOnS Gece Ce 2,,03* 2.05*
Avo top Mame tyres sere sc seven ones present present
ALO Gellk os 8 Bier Recaete ts atoa es wee Rre 65.44 62.22

* Osborne and Harris: Journ. Amer. Chem. Soc., xxv, p. 323, 1903.

The results of these hydrolyses are very similar, the most
marked difference being that vicilin yielded no glycocoll. Less
alanine and arginine and more glutaminic acid were also found
than in legumin. Although special effort was made to isolate
glycocoll from vicilin no trace of it was found, and it is probable
that vicilin in fact yields none. Itis to be noted however that
glycocoll can be missed when present in small amount. We have
recently made two hydrolyses on the same preparation of legu-
min from the vetch, in one of which a small quantity of glycocoll
was found while from the other none whatever was obtained.

— a

Thomas B. Osborne and Frederick W. Heyl 189

This difference we attribute to the fact that the esters of the
latter hydrolysis, owing to an accident to the liquid air apparatus,
had to be kept in ether for some weeks before distilling, during
which time the ester of glycocoll presumably underwent change.
That this occurred in the hydrolysis of vicilin is improbable, for
the esters were distilled soon after they were made. The differ-
ence in the proportion of alanine in legumin and vicilin is rela-
tively great, and it is probable that an actual difference exists
between these proteins in this respect. The vicilin used for this
hydrolysis was prepared in the way described in a former paper*
from this laboratory.

HYDROLYSIS OF VICILIN.

A quantity of vicilin was taken for hydrolysis which weighed
552.8 grams, ash and moisture-free. This was treated with 1200
ec. of hydrochloric acid of 1.10 specific gravity. The vicilin was
brought into solution by heating on the water-bath for about
three hours, and hydrolysis was completed by boiling in an
oil-bath for twenty hours.

The solution was then concentrated under reduced pressure,
saturated at o° with hydrochloric acid gas, and allowed to stand
on ice for four days. The glutaminic acid hydrochloride which
separated weighed 127.77 grams, which after deducting the
amount of ammonium chloride which it contained, is equiva-
lent to 102.38 grams of glutaminic acid, or 18.52 per cent of the
protein.

The glutaminic acid hydrochloride, after one recrystallization,
was analyzed, with the following result.

Chlorine, 0.2830 gm. subst. gave 0.2232 gm. AgCl.
Nitrogen, 0.3076 gm. subst. required 16.55cc. ~, N.HCl.

Calculated for

C;H,O,N.HC!:; Found:
FL ee hc ee 19.32 19.49
Mo ea BoM ea oo 28 Ee Sid CDI SCE Io ere tee Feo: (sors

The free glutaminic acid decomposed at 202° to 203°.

1 Osborne and Harris: This Journal, iii, p. 213, 1907.

190 Hydrolysis of Vicilin

Carbon and hydrogen, 0.2821 gm. subst. gave 0.4234 gm. CO, and
0.1604 gm. H,O.

Calculated:for-G, HO INE Ea ee C = 40.81; H =6.12 per cent.
Pound a23 ae eee eee © — 4079045 He — oR a

The filtrate from the glutaminic acid hydrochloride was con-
centrated under greatly reduced pressure and esterified with
alcohol and dry hydrochloric acid gas, as directed by Emil
Fischer. The esters were liberated, shaken out and dried in the
usual manner, and the aqueous layer freed from inorganic salts
and again esterified.

A third esterification was also carried out, but only a little more
ether soluble ester was obtained.

The ether was distilled off on the boiling water-bath, and the
residual esters distilled under diminished pressure into the follow-
ing fractions.

Temp. of Pressure, Weight,

Fraction. bath up to mm, grams.
1 Acero aaron : 65° 12.00 14.83
| Ee arene tm lee 100° 10.00 49.33
aMiae seh € eee 100° 1.00 76.27
JUNO 110° 1.00 13r5u
Vite parce ees ae 147° 0.85 56.92
VS Ro cna Ohi emer 158° 0.75 33.85
Vl 260° 0.70 - 28.75
ATO Gell. oe Bites ee retiree pene cee coe 273.52

The undistilled residue weighed 76 grams.

Between 65° and 87° at 10 mm. nothing was distilled.

Fraction I. From this fraction no glycocoll could be separated
as the hydrochloride of the ethyl ester. The ether which had
been distilled from the esters also yielded a negative result. The
esters were saponified and after removing chlorine, added to the
corresponding part of Fraction II.

Fraction II, This fraction was saponified by boiling with six
volumes of water for seven hours. The solution of the amino-
acids was evaporated to dryness under reduced pressure and the
proline extracted by boiling with absolute alcohol.

The amino-acids, insoluble in alcohol, were taken up in water
and the greater part of the leucine was separated by fractional
crystallization. The leucine weighed 13.82 grams.

Thomas B. Osborne and Frederick W. Heyl 191

Owing to the persistence of the characteristic leucine-valine
mixture, the filtrate from the leucine was racemized by heating
with an excess of baryta in an autoclave for twenty-four hours
ay ae

After removing the barium quantitatively with sulphuric acid,
the amino-acids were fractionally crystallized, and 0.84 gram of
substance obtained, which had the composition of amino-vale-
Tianic acid.

Carbon and hydrogen, I, 0.1255 gm. subst. gave 0. 2349 gm. CO, and

0.1107 gm. H,O. MII, 0.1612 gm. subst. gave 0.3012 gm. CO, and
0.1332 gm. H,O.

Calculated for C,H,,O,N....C = 51.28; H = 9.40 per cent.
TERONSERIG |. © Bp eee Ree Ce — ra O4 ee — Oe 80 os
iC. — (50505: Hi = 9.18 di

The filtrate from the valine yielded 2.80 grams of alanine. The
alanine crystallized from dilute alcohol in bunches of needles
and when recrystallized from water, was obtained in characteris-
tic prisms, which decomposed above 290°.

Carbon and hydrogen, 0.1445 gm. subst. gave 0.2165 gm. CO, and
0.1040 gm. H,O.

Calenlated for C,H,O;N..... C
attendant icicle ps ev ove o0 eo C

AORAD 3 Et
40.86; H

= 7.87 per cent.

ht: 9S)

The more soluble part of this fraction was examined for glyco-
coll before it was racemized, but none was found.

Fraction IIIa. This fraction was boiled, with five volumes of
water for seven hours with return condenser. The leucine which
had separated was filtered off, the filtrate evaporated to dryness
under diminished pressure, and the proline extracted with boil-
ing absolute alcohol.

The amino-acids, insoluble in alcohol, yielded 35.59 grams of
leucine which decomposed at 298° C.

Carbon and hydrogen, 0.1256 gm. subst. gave 0.2533 gm. CO, and
0.1130 gm. H,O.

Calculated for C,H,,0,.N....C = 54.96; H = 9.92 percent
7 So SR eee EC =" 55-01. re — 100i

The alcoholic proline solutions obtained from Fractions II and
ITI were united and evaporated to dryness under reduced pressure.

192 Hydrolysis of Vicilin

The residue was taken up in boiling absolute alcohol, and this
process repeated several times until the residue was completely
soluble in cold absolute alcohol. This solution yielded 27.96
grams of levoproline copper, dried at 110°, and 0.60 gram race-
mic proline copper, equivalent to 22.48 grams proline, or 4.06 per
cent.

The racemic salt was analyzed as follows.

Water, 0.1429 gm. subst. air dry lost 0.0151 gm. H,O at 110°.
Copper, 0.1065 gm. subst. air dry gave 0.0259 gm. CuO.

Calculated for C,,H,,O,N,Cu. 2H,O..H,O = 10.99 p.c.; Ca. = 19.40 p. ©
Round ee Ate ye Le eee ieee H,O = 10.57 “ Cu. = 19.43

Il

The amorphous levoproline copper was converted into the
phenylhydantoine, which after one recrystallization from a large
volume of water, melted at 143°, and had the following composi-
tion.

Carbon and hydrogen, 0.1294 gm. subst. gave 0.3170 gm. CO, and
0.0667 gm. H,O.

Calculated for C,,H,,O;N,:-......C = 66.67; = HoSo5 pemecue
Oude bab eve ated otros eer C.=:66.81; = Bee 7

Fraction IIIb. This fraction was shaken out with an equal
volume of ether, and the ether layer was washed with water in the
usual manner. From the ether layer there was obtained 3.09
grams of leucine hydrochloride which when converted into leucine
gave the following analysis.

Carbon and hydrogen, 0.1326 gm. subst. gave 0.2655 gm. CO, and
0.1158 gm. H,O.

Caleulatedtion Grr OsNem anne C = 54.96; H = 9.92 per cent.
| royiina\(c WN Spence Spee OSE ROP DUE A J C545, 60 Ee — 19a ‘

The aqueous layer was added to the corresponding solution
from Fraction IV.

Fraction IV. By a similar treatment, this fraction yielded
7.96 grams of phenylalanine hydrochloride; 20.05 grams of aspar-
tic acid as the barium salt; and 1.46 grams of copper aspartate.

The aspartic acid reddened but did not decompose at 300°.

After one crystallization from water, the following analysis was
obtained.

|

|

Thomas B. Osborne and Frederick W. Heyl 193
Carbon and hydrogen, 0.2185 gm. subst. gave 0.2914 gm. CO, and
0.1100 gm. H,O.

Calculated for C,H,0,N.........C = 36.09; H = 5.26 per cent.
= sli.dfs lal = 5,40 fa

hy
fe)
rat
S)
Qu

OQ
|

Fraction V. This fraction yielded 9.55 grams of very pure
phenylalanine hydrochloride. The free acid decomposed at 270°.

Carbon and hydrogen, 0.1515 gm. subst. gave 0.3628 gm. CO, and
0.0936 gm. H,O.

ealeulated for C,H,,0,N...-...- Cy= 65.45; Hi
PASTING MER rage! a0 spec sie sod eerste s Ce G5. Ik =

6.67 per cent.
6.8 “

lor)

There were further obtained from the aqueous layer, in the
usual manner, 4.85 grams of aspartic acid, as the barium salt.
The aspartic acid reddened at 300°, but did not decompose.
The filtrate from the barium aspartate was freed from barium,
concentrated to a small volume under diminished pressure, and
saturated with hydrochloric acid gasato°. There separated 3.87
grams of glutaminic acid hydrochloride.

After recrystallization, it decomposed at 198°, and had the
following composition.

Chlorine, 0.1382 gm. subst. gave 0.1069 gm. AgCl.

Caltematedifor C,H O,N-HGl.....:....2-.2-- Cl = 19:32 per cent:
LFS OURO. 6 cles econ, < erate Eenoaee ci are Cl 192 .

The filtrate from the glutaminic acid hydrochloride was freed
from chlorine and 6.79 grams of air dry copper aspartate were
obtained from it.

Copper, 0.1552 gm. subst. air dry gave 0.0451 gm. CuO.

Nutrogen, 0.3484 gm. subst. air dry required 13.1 cc. S = HCl.
Calculated for C,H,O,NCu.44H,O ..Cu = 23.07; N = 5.08 per cent.
YP OIETHGls spa ae eg Res Be rn eee Cu = 23.21: N = 5.26 £

Fraction VI. From this fraction there were obtained 8.46
grams of phenylalanine hydrochloride, 2.52 grams of glutaminic
acid as the barium salt, 5.95 grams of glutaminic acid hydrochlor-
ide, and 1.09 grams of air dry copper aspartate. The glutaminic
acid from the barium salt gave Analysis I and that from the
hydrochloride gave Analysis II.

194 Hydrolysis of Vicilin

Carbon and hydrogen, I, 0.3027 gm. subst. gave 0.4514 gm. CO, and
0.1622 gm. H,O. II, 0.2445 gm. subst. gave 0.3685 gm. CO, and
0.1419 gm. H,O.

Caleulatediior CAl.O Nee 2 eee C = 40.81; H = 6.12 percent.
Fownds,. » a 925 Scere I, C = 40.67; H = 5.95 e
II, C = 41.10; H = 6.44 _

Residue after Distillation.

The residue which remained after the distillation of the esters
yielded 4.4 grams of diketopiperazines, and 6.47 grams of gluta-
minic acid hydrochloride. The free acid decomposed at 202° to
203° C.

Carbon and hydrogen, 0.1500 gm. subst. gave 0.2255 gm. CO, and
0.0834 gm. H,O.

Calculated for Cp ON ss 2 oc .35 23% C = 40.81; H = 6.12 per cent.
| ERO TE Da KG PIN AIA ena, 5 Sanam Nain ehenn Ataee © = 41-007 Ei — 624

Cystine.

Owing to the very small proportion of sulphur in vicilin, no
attempt was made to determine cystine.

Tyrosine.

A quantity of vicilin equal to 36.57 grams ash and moisture-
free was hydrolyzed by boiling with a mixture of 120 grams of
sulphuric acid and 240 cc. of water for twenty-four hours in an
oil-bath. After freeing the solution from sulphuric acid, it was
concentrated to crystallization and cooled. The substance which
separated was recrystallized from water and yielded 0.8708 gram
of tyrosine, equal to 2.38 per cent.

Nitrogen, 0.1090 gm. subst. required 5.8 cc. X HCL.
Caleutateditor COs Nees oe eee ee N = 7.73 per cent.
Pound oh. d:0d8 ieee ¢ ane ka es Oe eee N = 7.45 k;

The filtrate from the tyrosine was used for determination of
the bases according to the method of Kossel and Patten.

Histidine.
The solution of the histidine = 500 cc.

Nitrogen, 50 cc. sol. required 2.15 ec. 2 HCl = 0.2150 gm. N in 500
cc. = 0.7934 gm. histidine = 2.17 per cent.

Thomas B. Osborne and Frederick W. Heyl 195

The histidine was identified as the dichloride.

Chlorine, 0.1377 gm. subst. gave 0.1718 gm. AgCl.
Bomemiatearor © EOIN, . ce ss oe ee ote eo wwe Cl = 31.14 percent.
SS 5 ee RR en ee Cl = 30.86 $

Arginine.
The solution of the arginine = 1000 cc.

Nitrogen, 50 cc. sol. required 5.07 cc. * HCl = 1.0140 gm. N in
1000 cc. = 3.1505 gm. arginine + 0.1102 gm. = 3.2607 gm. or 8.91 per
cent.

The arginine was converted into the copper-nitrate double salt
for identification.

Copper, 0.1713 gm. subst., air dry, gave 0.0235 gm. CuO.
Calculated for C,,H,,0,N ,Cu(NO,),.83H,O. ..Cu = 10.79 percent.
Fertile Men er ee SNS aridoee ioe hones! 8 codes Sy'c Cu = 10.96 .

Lysine.

The lysine picrate weighed 5.0707 grams equal to 1.9740 grams
lysine or 5.40 percent. The lysine was identified as the picrate.
Nitrogen, 0.1355 gm. subst. required 18.25 cc. HCl.

Caienulated for C,H,,0,N,.C,H,0,N;..... ...N = 18.70 per cent.
0) RES RED ea eer eee a Cte

HYDROLYSIS OF LEGUMELIN FROM THE PEA
(Pisum Satibum)

By THOMAS B. OSBORNE anp FREDERICK W. HEYL.

(From the Laboratory of the Connecticut Agricultural Experiment Station,
New Haven, Conn.)

(Received for publication, July 23, 1908.)

Legumelin is a name that has been applied to the albumin
which occurs in the seeds of a number of different leguminous
plants. As the properties, composition and occurrence of legu-
melin, the method employed in preparing the material used
for the hydrolysis here described, and also the literature of the
proteins of the pea, are given in a recent paper from this labora-
tory,” they will not be repeated here.

The results of this hydrolysis show legumelin to be distinctly
different in constitution from the other proteins with which it is
associated in the pea. The properties and ultimate composition
of legumelin closely resemble those of leucosin from the wheat
embryo, and this resemblance extends not only to the general
proportion of the several amino-acids, but also to the total quan-
tity obtained from each of these albumins as may be seen from
the accompanying table.

It is not improbable that legumelin is a constituent of the
physiologically active tissues of the seed, rather than a reserve
food substance for the developing seedling.

This supposition in regard to leucosin is supported by the fact
that this albumin is located almost entirely in the embryo of the
seed and that it resembles more closely, both in properties and
composition, the proteins of physiologically active tissues of

1The expenses of this investigation were shared by the Connecticut
Agricultural Experiment Station and the Carnegie Institution of Washing-
ton, D. C.

? Osborne and Harris: This Journal, iii, p. 213, 1907.

197

198 Hydrolysis of Legumelin

animals, than those which unquestionably form the reserve food
protein of the endosperm of other seeds.

As the leguminous seeds do not contain a sharply differentiated
embryo and endosperm, but are composed of tissues which com-
bine the functions of both, it is not possible to locate any one of
their proteins in any particular part of these seeds. The just
mentioned similarity, however, of legumelin to leucosin is sugges-
tive of similarity also in their physiological functions.

The results of the hydrolysis of legumelin are given in the fol-
lowing table, and in order to facilitate the above comparison
the results obtained for legumin, vicilin and leucosin are also
given.

Legumin Vicilin Legumelin Leucosin
pea pea pea wheat
per cent. per cent. per cent. per cent.
Givicocoll Bere cmierseee eae 0.38 0.00 0.50 0.94
UU Wath oC tare ae ane ea Men ni coke cot 2.08 0.50 0.92 4.45
Aen thay Sait ans Sper ai teg sieby sy he i 0.15 0.69 0.18
WCUCINe Ss craters eis ni IA 8.00 9.38 9.63 11.34
Proline yee kway eaten sae cm ey eter eeu 3,22 4.06 3.96 3.18
Phenlyalamimese erase eee 3.75 3.82 4.79 3.83
INS DAIbICACLin aa ser ae eer 5.30 5.30 4.11 3.39
Glutamimicacidee nese ae en Ono 21.34 12.96 6.73
Sertinee 5 f dco cide eka betty ae ten’ 0.53 ? ? 2
Gry atime ny Seki i NONE tie ele oy not det. not det. not det. not det.
Oxyproline ss 42 eibeeeeele a % ag *
SEV TOSI? HH ct tie on cialeateh a cien: 1.55 2.38 1.56 3.34
TAT OIMIME Meu oly aw ae enone eins al 8.91 5.45 5.94
DOTS} Coll aYses eA Mes ae ed ale Ot ala Abies rae 1.69 2.17 2.27 2.83
TG ViSITLO UN A slaedtd ata dntus uae thea ae 4.98 5.40 3.03 2.75
ATMO MMA |: Seales eeenen ey ae ctase 2.05* 2.03* 1.26* 1.41*
DEY PLOPWAME shasta ke esetsl one present present present present

62.22 65.44 51.13 50.53

* Osborne and Harris: Journ. Amer. Cherv. Soc., xxv, p. 323, 1903.

Hydrowsis of Pea Legumelin.

There were taken for this hydrolysis 550 grams, equal to 479.5
grams of ash and moisture-free legumelin. After warming for
four hours with a mixture of 550 cc. of water and 550 cc. of
hydrocholoric acid (sp. gr. 1.19) on the boiling water-bath, the
legumelin was finally dissolved. The hydrolysis was then com-

Thomas B. Osborne and Frederick W. Heyl 199

pleted by boiling for nineteen hours in an oil-bath. The solution
was concentrated to a volume of about 700 cc. and saturated at
0° with dry hydrochloric acid gas. After standing on ice for one
week, the separated glutaminic acid hydrochloride was filtered
off. The yield, after deducting the ammonium chloride present,
was equivalent to 50.45 grams glutaminic acid, or 10.52 per cent
of the protein. This glutaminic acid hydrochloride, after dis-
solving in water and reprecipitating with strong hydrochloric
acid, gave the following analysis:

Chlorine, 0.2127 gm. subst. gave 0.1695 gm. AgCl.

Saicuated torC HO,NHCl........2.. .... Cl = 19.35 per cent:
lPVexGea7a! = 5 Sey kee CORNER IPC Ee (CAR ae Cl = 19.70 *

The filtrate from the glutaminic acid hydrochloride was con-
centrated, under reduced pressure, to a very heavy syrup which
was esterified in the manner often described. After the esters
had been shaken out with ether, the aqueous layer was freed
from inorganic salts and the esterification was once more repeated.
The combined ethereal extracts were thoroughly dehydrated
with sodium sulphate and, after removing the ether at 760 mm.
the residual amino-acid esters were fractioned under diminished
pressure with the following results:

Temp. of Pressure, Weight,
Fraction. bath up to mm. grams,
Ld a 85° 23. 5.65
lU! s aioe Bee 80° 4. 37.81
HIM: 35605 Seance 110° 0.85 97.27
IG DE 136° 0.73 54.75
Ws, oe ee 195° 0.65 58.93
TiGhiPilsn6 3 ie Sot os Be ee ee ei ee ec on 254.41

The undistilled residue weighed 65 grams.

Fraction I. This fraction (including the ether removed at
ordinary pressures) yielded 4.49 grams of glycocoll ethyl-ester
hydrochloride, which at 145° melted sharply to a clear oil.

Nitrogen, 0.1974 gm. subst. required 14.3 cc. X HCl.

To
eaenisted gor CW OLNCI.......- sees es N = 10.04 per cent.
MITRE Ger Nat i ofthis bo. bain. ee no 55" waiinige N=10.14 “

The free amino-acids in the filtrate from the glycocoll ethyl-ester
hydrochloride were regenerated, and 0.50 gram of leucine and

200 Hydrolysis of Legumelin

o.go gram of alanine were obtained. The alanine crystallized in
needles having the following composition:

Carbon and hydrogen, 0.1739 gm. subst. gave 0.2583 gm. CO, and
0.1273 gm. HO,

Calculated for: ©, ACOlNbe ea Cc
Pound's ities ac: oe Ric ees ee Ieee CS 40), 09 TEL

40.45; H = 7.86 per cent.
8.13 a

Il

Fraction II. This fraction was saponified by boiling with
water for seven hours. The solution was then evaporated to
dryness under diminished pressure, and the dried amino-acids
were boiled with absolute alcohol, thus extracting the proline.
After removing the proline, the remaining amino-acids, which
weighed 23.75 grams, were subjected to a careful fractional cry-
stallization. Glycocoll was not found. There were isolated,
9.95 grams of leucine, 3.31 grams of valine and 3.6 grams of
alanine. The leucine weighed was analytically pure. The de-
composition point was 298°.

Carbon and hydrogen, 0.1533 gm. subst. gave 0.3090 gm. CO, and
0.1356 gm. H,O.

Calculated tor Cane OsNe. =r. re C = 54.96; H = 9.92 per cent.
POUNGs: a. ic CeOR ene ene oe C = 54.98; H = 9.82 &

The valine when dissolved in 20 per cent hydrochloric acid had
°

: : 20 ;
a specific rotation of (a) fie + 26.1°.

Carbon and hydrogen, 0.1391 gm. subst. gave 0.2626 gm. CO, and
0.1162 gm. H,O.

Calculatedition C7EE Os Near C = 51-28; El —) 940 pegeenn
51.48; H=9.28 “

hy
fo)
ie
Ss}
a
o)
I

The alanine crystallized in needles.
Carbon and hydrogen, 0.1471 gm. subst. gave 0.2191 gm. CO, and
0.1060 gm. H,O.

Calculated for GO LN ene aan C = 40.45; H = 7.86 per cent.
Found: cc eek eee C= 40562-5o — 78:00

Recrystallized once again from dilute alcohol, this alanine
gave beautiful needles which decomposed at 290° on rapid heat-
ing.

a

Thomas B. Osborne and Frederick W. Hey] 201

Carbon and hydrogen, I, 0.2619 gm. subst. gave 0.3908 gm. CO, and
0.1889 gm. H,O. II, 0.1490 gm. subst. gave 0.2227 gm. CO, and 0.0996
gm. H,O.

Calculated for C.H-O,N: . 2... - C = 40.45; H = 7-86 percent.
IEPETESTIG | ine Oe eee Re Sa a are Re LC = 40-69-70 — enon 5
i: € =40.76; Wo =7.42

Fraction III. This fraction was boiled with water for seven
hours and the solution concentrated to dryness under reduced
pressure. Proline was extracted from the dried residue with
absolute alcohol. The amino-acids insoluble in alcohol weighed
45.71 grams, from which there were isolated 35.73 grams of leu-
cine. The alcoholic extracts containing the proline from Frac-
tions II and III were joined, and the alcohol removed under
diminished pressure. The residue was extracted with boiling
alcohol and the solution allowed to stand over night. After
filtering out 1.15 grams of substance which had separated on
standing, this process was repeated until nothing more was depos-
ited after standing for several hours. The proline yielded 2.03
grams of the racemic copper salt, and 22.24 grams of the levo-
salt dried at 110°. The levo-proline was coupled with phenyliso-
cyanate, and the resulting phenylhydantoine melted sharply at
143° to a clear oil.

Carbon and hydrogen, 0.1162 gm. subst. gave 0.2840 gm. CO, and
0.0635 gm. H,O.

ealemiated for C,.H,,0,N,......+- C = 66.67; H = 5.57 percent.
=6.07 “

=
°
rat
3
a
fo
I
a
jor)
(op)
a
an
|

The racemic salt, when once recrystallized from water, separ-
ated in the characteristic plates.

Water, I, 0.2429 gm. subst., air dried, lost 0.0268 gm. H,O at 110°

i0t 0.6486 “ “ “ “ OU. 0712 “ “ “a
Calculated for C,,H,,O,N,Cu.2H,O.....H,O = 10.99 per cent.
ATUL EMME Te Mewar cot REET Soros os Siw 210 00:8 - HOO saa

TE, 10299) 4

Copper, 0.3538 gm. subst., dried at 110°, gave 0.0961 gm. CuO.

ealcwlated tor, ,ly,O,N,Cu......2.5..-... Ca = 218) per cent.
aE AIR SER WO pM Ts Mag ay oh a od n'a Gites Cu=21.70 “

202 Hydrolysis of Legumelin

Fraction IV. This fraction was suspended in five volumes of
water and shaken with an equal volume of ether. The ether
yielded, by the usual treatment, 9.90 grams of phenylalanine
hydrochloride. The aqueous layer was saponified by warming
on the water-bath for five hours with 100 grams of baryta. There
were obtained 13.54 grams of aspartic acid as the barium salt.
After once recrystallizing the free acid from water, the crystals
reddened but did not decompose at 300°.

Carbon and hydrogen, 0.2400 gm. subst. gave 0.3211 gm. CO, and
0.1171 gm. H,O.

CalenlateditonG,H- ON aoq 15 C=

= 5.26 percent.
TER eter neo Tay i es ieee pa ral st C= 5

26
42 “

The filtrate from the barium aspartate was quantitatively
freed from barium and the solution was concentrated to a syrup
under diminished pressure. No glutaminic acid could be separ-
ated from this solution as the hydrochloride. The free amino-
acids were regenerated and converted into their copper salts,
from which 2.93 grams of air dry copper aspartate were separated.
The copper was removed from the filtrate from the copper
aspartate, but no serine could be isolated from the solution.

Fraction V. This fraction yielded 18.07 grams of phenyl-
alanine hydrochloride, 3.55 grams of aspartic acid, 11.69 grams
of glutaminic acid hydrochloride and 2.54 grams of copper aspar-
tate. The free phenylalanine was analyzed as follows:

Carbon and hydrogen, 0.1637 gm. subst., gave 0.3940 gm. CO,.

Calculated for €,H,,NO,........C =65.45; H = 6766 peqeear
Jelvonbhnao vere meu eey en anaeae sieve icone C= OS, (CH18 Tel = lest

The aspartic acid obtained from the barium salt reddened
but did not decompose at 300°.

Carbon and hydrogen, 0.2553 gm. subst. gave 0.3377 gm. CO, and
0.1248 gm. H,O.

Calenlateditori Cie O Ne earner C=

36 —
RO Wt) silsahAeseeeae ere ae wee ee Coe Sth. =o

.26 per cent.
ge lis

The glutaminic acid hydrochloride after one recrystallization
from strong hydrochloric acid, decomposed at 199°. With an
equivalent quantity of potassium hydroxide, it yielded free glu-
taminic acid which decomposed sharply at 202° to 203°.

Es =

Thomas B. Osborne and Frederick W. Heyl 203
Carbon and hydrogen, 0.1252 gm. subst. gave 0.1866 gm. CO, and

0.0700 gm. H,O.

Caleulated for C,H,O,N ......... C = 40.81; H = 6.12 percent:
LENGE CES © Bee Seeley ee ea een ene C= 40°64: Ho 6221

The copper aspartate crystallized from a large volume of water
in the characteristic sheaves.

Nitrogen, 0.2558 gm. subst. required 9.7 cc. X. HCl.

10
Calculated for C,H,O,N Cu. 44H,O ...Cu = 23.07; N= 5.08 per cent.
JRiSNaVS), 25 big ie een ee Cu — 23-26-59 5e30) 4
% GLUTAMINIC ACID.

This protein resembles leucosin in regard to the difficulty of
determining glutaminic acid directly, as is shown by the follow-
ing.

Eighty-four grams of pea legumelin, equivalent to 73.23 grams
ash and moisture-free, were hydrolyzed by boiling with a mixture
of 84 cc. of water and 84 cc. of hydrochloric acid (sp. gr. 1.19)
for twenty-six hours. The solution was sharply concentrated to
a heavy syrup and saturated at o° with hydrochloric acid gas.
After prolonged standing on ice, 3.13 grams of glutaminic acid
hydrochloride separated. The filtrate was again saturated and
2.15 grams further was obtained. This is 5.76 per cent of the
protein, while we obtained in the hydrolysis of the larger quantity
first described, 12.96 per cent.

THE RESIDUE AFTER DISTILLATION.

The undistilled residue which weighed 65 grams was dissolved
in boiling absolute alcohol. On standing over night, 2 grams of
substance separated, which was filtered out. After removing
the alcohol and heating the residual solution with an excess of
baryta for five hours, the baryta was removed, and glutaminic
acid separated as the hydrochloride, in the usual manner. This
weighed 2.91 grams and decomposed at 199°. The free acid de-
composed at 203°.

TYROSINE.
A quantity of legumelin equal to 43.63 grams of ash and mois-

ture-free substance was hydrolyzed with 150 grams of concentrated
sulphuric acid and 300 cc. of water, by digesting on a boiling

204 Hydrolysis of Legumelin

water-bath until the legumelin was dissolved and then boiling the
solution for 12 hours in an oil-bath. After removing the sulphu-
ric acid and extracting the barium sulphate by boiling out with
water, the solution was concentrated to crystallization. The
substance that separated was then redissolved in boiling water,
decolorized with animal charcoal and the tyrosine recrystallized
by concentrating and cooling the solution. The tyrosine thus
obtained weighed 0.68 gram, equal to 1.56 per cent.

Carbon and hydrogen, 0.1573 gm. subst. gave 0.3460 gm. CO, and
0.0894 gm. H,O.

Calculated fon.GcH ON... ....-- G
Rounds acs) yeni een eee G

Il

59.67; H = 6.08 percent.
59.98; H=6.31 “

ll

The filtrates from the tyrosine were used for determinations
of the bases according to the method of Kossel and Patten.

CYSTINE.

No attempt was made to determine cystine, on account of lack
of sufficient material.
HISTIDINE.

The solution of the histidine = soo cc. .

Nitrogen, 100 cc. solution required 5.36 cc. “ HCl = 0.2680 gm. N

in 500 cc. = 0.9889 gm. histidine = 2.27 per cent.
The histidine was converted into the dichloride for identifica-
tion.
Chlorine, 0.1000 gm. subst., air dry, gave 0.1252 gm. AgCl.
Calenlatedtion Grrl. © eis Cleese ie Cl = 31.14 per cent.
Eh oar lies ace itetnitctisy Spetau a (es cack ice te ete eee pea aston Ci—30296 s

The histidine dichloride decomposed at 232°.
ARGININE.
The solution of the arginine = 1000 cc.

Nitrogen, 50 cc. solution required 3.71 cc. “ HCl = 0.7420 gm. N in
1000 cc. = 2.3054 gm. arginine + 0.0720 gm. = 2.3774 gm. = 5.45
per, cent.

;
7
q
i
\
.

Thomas B. Osborne and Frederick W. Heyl 205

The arginine was converted into the copper nitrate double
salt for identification. The greater part of the solution was lost
by an accident, therefore only a little substance was available.

Copper, 0.0366 gm. subst. air dry gave 0.0050 gm. CuO.
Calculated for C,,H,,0O,N, Cu (NO,),.3H,O..Cu = 10.79 per cent.

IS@RiNG | 55:55 Bog bare ole aad Oe Cane
Calculated for C,,H,,0,N, Cu (NO,),.3H,O..Cu = 10.79 per cent.
TE oaninG |... Soro Sicnois iAP RBIERO IC SRC Co — 10892 >

LYSINE.

The lysine picrate weighed 3.400 grams = 1.3224 grams
lysine = 3.03 percent. The lysine was identified as the picrate.

Nitrogen, 0.3000 gm. subst., dried at 100°, required 5.58 cc. &
lal(Gie

Catcmated for C.H.,O,N,.C,H,0,N,........N
Jere igGAVGl, cues ff Aleeso eRe ee N

18.70 per cent.
18.60

I

i
i : ee
Ul ‘L ,
=|
ee .
3 5
5 f
- a
:
or »
b
°
Z
' - 7 n
‘ he fe hi
x ~
-~ ait ~ -
=) . F 3
oo
Q é
4 4 A
- ~ -
A , «
liye
o>
,
rd e - M
bili vebs
ho
‘ ‘a 4
ue | i?
’
j ¢ + is
i
y, ’
of
bis’
i . { rh i
be a ‘ : TAN
, Th J Wid /
iy a ; Mag | tes AA Ry ;
| ’ ; : 2 : Vie ph eye PEP yt
z ; ra fa), fis :
‘ ; j
Adie se oN ‘
é T ig OS
; Hs eo aS ts cap ; i
i j bis
‘ a
i
; =

THE DETECTION AND QUANTITATIVE DETERMINATION OF
B-OXYBUTYRIC ACID IN THE URINE.'

By O. F. BLACK

(Contribution from the Laboratory of Biological Chemistry of the Harvard
Medical School.)

(Received for publication, July 29, 1908).

It is generally recognized that @-oxybutyric acid, on oxidation
in the body, yields first acetacetic acid, which then decomposes
into acetone. This reaction can be carried out in the laboratory
by the use of hydrogen peroxide as the oxidizing agent, and was
observed by me some months before a description of it was
recently published by Dakin.” The intermediate product, acet-
acetic acid, can be tested for in a solution by the addition of ferric
chloride which produces a bright red color, the well-known Ger-
hardt’s test. If @-oxybutyric acid be tested with hydrogen per-
oxide in the presence of ferric chloride one obtains this same
color reaction, which thus becomes a test for 6-oxybutyric acid.

Procedure. Two or three drops of the ordinary commercial
hydrogen peroxide are added to the dilute solution of S-oxybuty-
ric acid in a test tube and mixed by shaking, then a few drops of
5 per cent ferric chloride containing a trace of ferrous chloride.
On standing a few seconds a beautiful rose color develops which
slowly intensifies until it reaches a maximum and then gradually
fades owing to the further oxidation of the acetacetic acid. The
chief precautions to be observed in carrying out the test are to be
sure that the solution is cold and nearly neutral, and to avoid a
large excess of hydrogen peroxide and iron. If too much of the
oxidizing agents is added and but little §-oxybutyric acid be
present the color developed is transitory or fails to appear. By
starting with a small quantity and then adding more ferric
chloride at intervals of a few minutes until no further color is
produced one is able to observe the full intensity of color, and
thereby get a rough idea as to the amount of @-oxybutyric acid

1 Done with aid from the Proctor Fund.
2 This Journal, iv, p. 1.
207

208 Determination of §-Oxybutyric Acid

present. The test has a considerable delicacy; a solution contain-
ing o.1 of a milligram per cubic centimeter, or one part in 10,000,
gave an easily recognized color.

In the urine. The color of urine and the presence of acetacetic
acid and sugar all interfere with the test when applied to the
urine directly. The following procedure, however, has been
found to give the required result, and can easily and quickly be
carried out. Five to 1o cc. of the urine under examination are
concentrated in an evaporating dish at a gentle heat to one-third
or one-fourth of the original volume which eliminates the acet-
acetic acid. The residue is then acidified with a few drops of con-
centrated hydrochloric acid and made to a thick paste with plaster
of paris and allowed to stand until it begins to set. It is then
stirred and broken up in the dish with a blunt ended stirring rod.
The porous meal thus obtained is extracted twice with ether by
stirring and decantation. The ether extract which contains
$b-oxybutyric acid is evaporated spontaneously or on the water-
bath. The residue is finally dissolved in water, neutralized with
barium carbonate and the test applied as already described.

The reaction occurring in this test, namely, the oxidation of ?-oxybutyric
acid to acetacetic acid and thence to acetone has been investigated in
this laboratory in its quantitative aspect with the hope that it might offer
a means for the estimation of f-oxybutyric acid. The results were not
encouraging for, although a certain degree of constancy was observed in
the amounts of acetone produced by known quantities of $-oxybutyric
acid, they fell far short of what was required by calculation. Moreover,
sugar was found to yield products which interfered. So it was concluded
that oxidation by hydrogen peroxide is not suited for a quantitative
method.

QUANTITATIVE.

As described above in the qualitative method §-oxybutyric
acid is readily extracted from urine after evaporating and drying
with plaster of paris. It seemed probable that this might offer
a means for the quantitative estimation of the acid more rapid
than the methods that arenowinuse. Of the numerous methods
proposed for determining this acid the most reliable are undoubt-
edly Geelmuyden’s and Magnus-Levy’s. These authors extract
the urine, after strongly acidifying with a mineral acid, with
ether in a continuous extractor for periods ranging from 48 to 72

a
f
‘
+
a
‘
>

O. F. Black 209

hours. I have found that by previously drying the urine with
plaster of paris a more complete extraction may be accomplished
in two hours.

A known solution of active f-oxybutyric acid was prepared by
concentrating some diabetic urine, acidifying the residue, drying
with plaster of paris, and extracting with ether. This is an effec-
tive method for the preparation of the acid as fairly large volumes
of urine can be concentrated and the product is but slightly dis-
colored. The yellow syrup obtained was taken up with water
decolorized with bone black, and its strength determined by the
polariscope. This was used as a known solution to add to a nor-
mal urine as follows: 50 cc. of a normal urine was treated witha
definite amount of the above solution of 8-oxybutyric acid, made
faintly alkaline with sodium carbonate and boiled down to one-
third of its volume and then concentrated to about 10 cc. on the
water-bath. The residue was cooled, acidified with a few drops
of concentrated hydrochloric acid until it reddened litmus dis-
tinctly, and mixed with plaster of paris to a thick paste. The
mixture on standing a few minutes begins to harden. It is then
stirred and broken up with a stout glass rod and more plaster is
added, if necessary, until it has the consistency of a fairly dry,
coarse meal. This is transferred to a continuous extraction
apparatus where it is extracted with ether for two hours. The
ether extract is then evaporated, either spontaneously, or in an
air current, the residue taken up with water, treated with a little
bone black if necessary, filtered until perfectly clear and made to
known volume (25 cc. or less). The 3-oxybutyric acid is deter-
mined with the polariscope. The following figures were obtained
with normal urines treated with known quantities of the acid.

Polariscope. Calculated grams Grams aid

of acid. added.

Leitth ee AG? ae 0.8150 0.8053

iAG woe Ase 0.7824 0.8053 :

‘Deke ae 1.46° 0.7990 0.8053 .

coe eee 0.70° 0.3830 0.3840

Rs shsy es euinsto ¢ 0.70° 0.3830 0.3840

3 oo See 0.68° 0.3720 0.3840

LL St A022 0.5590 0.5760

2), Se ROE Ree 1.00° 0.5470 C.5760

210 Determination of 6-Oxybutyric Acid

A series of determinations on a diabetic urine containing much
sugar gave these figures in grams.

ils 2. 24-hour urine.
January p20 seems 27.3 29:7 3720
dehoeiny Alea roascse 28.0 29.3 3940
AEWA Psion o bisa o ac 24.6 23.9 3210
January 9:2 One se eee 26.7 25.5 3210
January “3000.02. o 34.6 32.0 3190
January, Si. eee 50.5 52.0 4600
Kebruary. \b0. ee: 38.8 39.5 4050
February 2240-5... 36.5 37.0 3510

For purposes of comparison the method of Magnus-Levy was
tried on one of the above urines, that of January 31 and two deter-
minations which were extracted, one for 60 hours, and the other
for 80 hours, gave respectively 48.8 and 45.5 grams, as against the
50.5 and 52.0 grams obtained by the plaster of paris method ina
two-hour extraction.

A Soxhlet apparatus was used in the extraction, and also a
modification of it which I have devised for this particular process

and which proved very effective. This appa-
ratus consists of an outer jacket (a) fitted toa
small flask and to a condenser. Inside the
jacket is a tube of slightly smaller bore (6)
with a one-hole cork stopper in the lower
end, through which passes the siphon (c) to
which is attached the funnel (d) by a twist
of platinum wire. A pad of glass wool is
packed about the open ends of the siphon and
funnel in the lower part of (b) then the
material to be extracted is poured upon it, not
filling over the top of thesiphon. The advan- |
tages of this apparatus are, first, that the inner
tube is constantly exposed to the warm
vapors of ether, then, that the ether returns
to the boiling flask hot so there is no interrup-
tion in boiling asin a Soxhlet, finally that
by using siphons of various lengths large or
small amounts of substance may be extracted
with little ether.

A METHOD FOR THE QUANTITATIVE DETERMINATION OF
6-OXYBUTYRIC ACID IN URINE.

By PHILIP A. SHAFFER.

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

(Received for publication, August 8, 1908.)

We have at present for the determination of $-oxybutyric
acid, no method which is not open to serious criticism, either on
the point of accuracy or on account of time, labor and chemicals
involved. The method of separately determining the inorganic
acids and bases of the urine, and so calculating the amount of
organic acid present, was first used for this purpose by Stadel-
mann’ who adopted it from Gahtgens,? and from its use Stadel-
mann was led to the discovery of the organic acid, thus verifying
the statement made three years earlier by Hallervorden that the
high ammonia content of diabetic urines is due to the presence of
an organic acid.*

This procedure has since been used in the study of (-oxybu-
tyric acid in relation to acidosis by Magnus-Levy* and others.
The method is very laborious, necessarily inaccurate, on account
of the many operations required, and still more important, it
does not determine the amount of §-oxybutyric acid, but merely
the amount of total organic acidity. And for this latter purpose
the simpler and more accurate method proposed by Folin’ is to

1Stadelmann: Arch. }. exp. Path. u. Pharm., xvii, p. 419, 1883.

*Gahtgens: Zeitschr. f. physiol. Chem., iv, p. 36, 1880.

$ Hallervorden (Arch. fj. exp. Path. u. Pharm., xii, p. 237) in 1880 con-
firmed the finding of Boussingault of high ammonia excretion in diabetes,
and suggested that it was caused by the excretion of an organic acid (lactic
or glycuronic acids). Stadelmann in 1883 confirmed Hallervorden and
isolated crotonic acid. In 1884 Kiilz, Zeitschr. f. Biol., xx, p. 165; Arch. f.
exp. Path. u. Pharm., xviii, p. 290, and Minkowski, Arch. f. exp. Path. u.
Pharm., xviii, pp. 35 and 147, independently showed the acid to be f-oxy-
butyric.

*Magnus-Levy: Arch. f. exp. Path. u. Pharm., xiii, p. 149, 1899.

* Folin: Amer. Journ. of Physiol., ix, p. 265, 1903.

2Ir

212 Determination of §-Oxybutyric Acid

be preferred. With one exception, the other methods for the
determination of 6-oxybutyric acid are based upon the optical
activity of the levorotary acid or its salts.

The rotation of the fermented urine is read in the polariscope
(Kulz) with or without preliminary treatment with basic lead
acetate; or the acid is extracted by ether from the evaporated
urine, and the rotation of an aqueous solution of the residue
from the ether is determined, and the amount of oxybutyric acid
calculated.

Direct reading of the fermented urine is easy enough but the
results are worthless because of, first, the great percentage error
in reading dilute solutions of the acid or its salts; second, the
probable presence of other optically active substances in urine
even after fermentation; and third, if basic lead acetate be used,
the action of this substance in increasing the levorotation of
salts of the acid (Magnus-Levy).? Extraction of the acid by
ether, and subsequent polarization of the aqueous solution of the
residue from the ether wasapparently first used by Wolpe? in 1886.
Without altering the principle of the method, it has been very
materially improved by Magnus-Levy, Bergell,‘ and most recently
by Black. With the use of the latter’s modifications which con-
sistin dehydrating the evaporated urine with plaster of paris, and
the use of an improved continuous ether-extraction apparatus,®
the method is fairly quick and the results may be fairly satisfac-
tory. The principle of the method is still however open to the
objection of difficulty of complete extraction, and that other
optically active substances may be extracted from the urine,

1 The specific rotation of the free acid is —24.12° and of the sodium salt
—14.35° (Magnus-Levy: Arch. f. exp. Path. u. Pharm., xlv, p. 393 and 397,
1901). The specific rotation is different for the salts with different bases.
Magnus-Levy points out that an error of o.10° in reading the polariscope
would amount to about 15 grams in 5 liters of urine (zbid., xlii, p. 170,
1899). An error of 0.10° in reading a 2 per cent solution of $-oxybutyric
acid (which is frequently obtained from urines containing only a little of
this acid) would amount to more than ro per cent of the total.

?Magnus-Levy: Arch. f. exp. Path. u. Pharm., xlv, p. 393, 1901.

3'Wolpe: Arch. f. exp. Path. u. Pharm., xxi, p. 138.

4 Bergell: Zeitschr. f. phystol. Chem., xxxiii, p. 310, 1901.

’ Bergell used anhydrous copper sulphate and extracted in a Soxhlet
apparatus.

Philip A. Shaffer aes

and either increase or decrease the levorotation of the final
solution.

The method proposed by Darmstadter' is, from its author’s
description, very easy and exceedingly accurate; though in my
hands it has not been successful. After evaporating the urine
made alkaline with sodium carbonate Darmstadter distills it with
a constant concentration of 50 per cent sulphuric acid thus con-
verting the f-oxybutyric acid into crotonic acid which distills over.
The distillate of 300 to 400 cc. is extracted three times with ether
which, he claims, removes all of the crotonic acid. The ether is
distilled and the residue, containing the crotonic acid, after being
heated on a sand-bath toremove the volatile acids, is dissolved in
water and titrated with alkali. Darmstadter in this way obtained
results from 99.36 per cent to 99.70 per cent of the amountof
synthetic $-oxybutyric acid added to urine. Unfortunately he
does not state just how he determined so accurately the amount
of oxybutyric acid used in his experiments. Of the various objec-
tions to this method as its author describes it, the most obvious
is perhaps the difficulty in completely extracting the crotonic
acid from 300 cc. of liquid, by two or three portions of ether.
The method has not apparently been received with favor.

It occurred to me that it might be possible to utilize as the
basis for a new method a property of $-oxybutyric acid long ago
mentioned by Minkowski—its oxidation with the formation of
acetone and carbon dioxide. I hoped that the reaction might,
under certain conditions, proceed after the following well known
scheme.

CH, CH, CH,

H — { — OH — : = 0 = { = (0
ne +O te 12 EO ie
COOH COOH GO;

B-oxybutyric acid — diacetic acid —> acetone and carbon dioxide.

Distilling with sulphuric acid and potassium bichromate, under
the conditions to be described, it is easily possible to obtain from
$-oxybutyric acid the theoretical amount of acetone, the quantity

1 Darmstadter: Zeitschr. f. physiol. Chem., xxxvii, p. 355, 1903.

214 Determination of §-Oxybutyric Acid

of which may be accurately determined by means of standard
iodine and thiosulphate solutions.

The conditions necessary for a maximum yield of acetone
concern the concentrations of sulphuric acid, and of bichromate.
Too little suphuric acid, even with an excess of bichromate liber-
ates the acetone very slowly and perhaps incompletely; while a
very great excess of sulphuric acid decomposes f-oxybutyric acid
with the formation of crotonic acid. The optimum concentration
appears to be between 3 per cent and 7 per cent sulphuric acid.
The following experiments show the effect of varying amounts of
sulphuric acid upon the speed with which the acetone is formed.

The determinations were made in a fairly pure solution of inactive
f-oxybutyric acid, made by the reduction of acetacetic ester by sodium
amalgam (Wislicenus).1 About go per cent of the titrated acidity of
the solution was due to oxybutyric acid (from the determination of acetone
under conditions giving the correct results).

The volume of each distillate was about 300 cc.; more water was added
to the distilling flask before each subsequent distillation or the volume was
kept constant at about 500 cc. by means of water from a dropping funnel.

The acetone in the distillates was determined as usual with standard
(103.5 per cent %) thiousulphate and iodine solutions, of which 1 ce. =

10
I mg. acetone. The results are given in milligrams of acetone.

With 1.0 gram potassium bichromate:

ec thee: SHES Or se stb tee on ean 1st dist. 16.0 mg. acetone.

PAG aed LS <3 .

adic 7.0 >

4th “ 2.4 .

44.8 =

Tl. 2) Sa COWES Oya to once eee eae 1st dist. 21.5 7
2d) S* Oe x

3d “ Gee “

4G ee a7 -

48.4 :

FIT.» § ce! WSO eked oe eee eee Ist dist. 39.1 ‘
2d," OG $

49.7 =

1 Wislicenus: Ann. d. Chem., cxlix, p. 205, 1869.

Philip A. Shaffer 215

OE VG s 5) 6 ee ee Ist dist. 47.0 mg. acetone.
Peis 0.0 me
47.0 7
“i. LGrOl) SES) © (saan ee tea Ist dist. 40.6 3
zd: <8 GEO %
40.6 :
Wil HO @e: 1SIBS Oa ec Gros eae Ist dist. 43.0
Aa O20) -
43.0 :

For bichromate the danger lies in use of too large a quantity,
which gives very low results: probably because of a further oxida-
tion of the acetacetic acid, first formed. This danger may be
averted by adding the potassium bichromate in a dilute solution
from a dropping funnel during the distillation. At the same time
the concentration of sulphuric acid is thereby kept practically
constant.

The following results show the effect of varying quantities of
potassium bichromate; in these experiments both sulphuric acid
and bichromate were added before starting the distillation.
The original volume in each distilling flask was about 500 cc. of
which about 300 to 350 cc. was distilled.

With 10 ce. H,SO,

I. With 0.8¢m.K,Cr,0, = 40.0mg. acetone.
ie Ui e 40.0 =
Il. Oe Ser : 41.0 #
1\ fe 20) 16 a 36.0 .
Ws 10-4 Z 35.3 ie
WAL 7p | ale ‘5 36.8 rs
Vil. Eye 1) eee fe 23.7 .

With 15 cc. H,SO,

eee Withee toon KoCrO> 36.2 ss
Ju Sealine a 23.5 S
Te 105 es 20.5 ¢
IV. 20)" 3 @20'ce.B55O) 1329 5

216 Determination of §-Oxybutyric Acid

The following are the results of similar determinations except
that the bichromate (in dilute solution, usually 0.2 to 0.5 per cent)
was added from a dropping funnel during the distillation; the
solution dropping in about as fast as the distillate collected.

I. 10cc.H,50,0.5 gm. K,Cr,0, 40.0mg. acetone:
JUL. us ORS i e 40.0 5
Es 7% A 0.5 y @ 41.0 as
TEV 87. Zi 0.5 = 3 40.0 te
We a0) i 3.0 a 38.0 us
Vii ao “ 3.0 S ‘ 38.4 aS
VII. 10 e 3.0 ‘ . 36.7 “
WAUHTS 10) i 3.0 us t 38.7 f
Pe TOUS 3) (0) B e 36.8 s
en 20" LS 3.0 - ae 38.1 <
In a different solution:
Mg. acetone
per 100 cc. oxy-
butyric aci
solution.
I... 20:cc. oxybutyric acidsol.. 15ce,H50, 1% KC eee
tie sOr os as - iy © oe 1 a 84.5
TEE 201 es ie Heyes 1 . 81.5
De. aOr se s Pos zs 1 > 79.5
VWeaeZ0 IS & a G 20 oy 2 4 78.5
WA fe 40 “ “ “ “ 20 “ “ ' Oy, “ 79 ; 3
AWA yA Doe # s s PAN), 3 2 < 82.5
WADA) g PAYEE “ 2 “ 83.0
ID See fe f <i PAN): 2 os Oe 82.5
Ke DOES a = 20 nes a 2. S 79.5

These results show that even when a dropping funnel is used
for the addition of the bichromate, the solution should not be too
concentrated, or, what is the same thing, the bichromate should
be added slowly. When this is done the results are accurate. In
order to prove this point sodium salt of the inactive acid (already
about 90 per cent pure) was prepared and recrystallized three
times from absolutealcohol. Thisrecrystallized salt was dissolved
in water, excess of sulphuricacid added, cooling with ice, and the
solution dehydrated with plaster of paris. The plaster was pow-
dered and extracted with dry ether in a Soxhlet apparatus. The
aqueous solution of the ether residue was boiled with pure bone
black, filtered, cooled and a portion titrated with {4 alkali (phenol-
phthalein).

Philip A. Shaffer 219

25 cc. = 6.85 cc. X XK 10.4 = 71.2 mg. f-oxybutyric acid = 39.6 mg.
acetone.

Five determinations under identical conditions in 25 cc. por-
tions of this solution gave

40.2 mg. acetone = 101.3 percent.

= a e ek. i
410 .% x 103.3) in
40:7 ™ iy 2S ee
39.2 * * 1 De |

The distillations were carried out with 15 cc. of concentrated
sulphuric acid and keeping the volume nearly constant at about
400 cc. by dropping in water or 0.2 per cent potassium bichro-
mate from a dropping funnel. A total of about 1 gram of potas-
sium bichromate was added, and about 800 cc. distilled in each
case.

When this method was applied to urine, difficulties were at once
encountered, but were satisfactorily overcome.

Normal urines distilled with chromic acid give acetone although
in small quantities. This was shown in 1885 by Flickiger,}
who found the chief source of the acetone in the conjugated
glucuronic acids. The glucuronic acids may however be removed
by basic lead acetate and ammonia, which reagents do not pre-
cipitate f-oxybutyric acid. The small quantities of formic and
butyric acids present in normal and pathological urines would
interfere with the determination of $-oxybutyric acid by this
method, but these volatile acids may be removed by distilling
the urine with sulphuric acid before the addition of bichromate?

1 Fitickiger: Zeitschr. f. physiol. Chem., ix, p. 343-

2 8-Oxybutyric acid in the concentration present during the distilla-
tion (never over 0.05 per cent) is not decomposed by 1 per cent toro per
cent sulphuric acid. Tencc. f-oxybutyric acid solution + 500 cc. water

+ 15 cc. sulphuric acid was distilled until 200 cc. distillate had collected.
This distillate contained 0.35 cc. acidity, the equivalent of 3.6 mg. of
f-oxybutyric acid. The liquid in the distilling flask was again distilled,
_ this time dropping in 0.5 per cent potassium bichromate. This distillate
contained 108.4 mg. acetone. Duplicate determinations in the same oxy-
butyric acid solution, but without previous boiling with sulphuric acid,
gave 107.8 mg. and 108.6 mg. acetone. Araki: Zettschr. f. physiol. Chem.,
XVili, p 1, 1893) found on distilling 200 cc. of ar per cent solution of
§-oxybutyric acid that about 1 per cent (0.019 gm.) passed over as cro-
tonic acid in the first 100 cc. distillate.

218 Determination of §-Oxybutyric Acid

Any formic acid passing into the distillate containing the acetone from
$-oxybutyric acid would use up iodine in the subsequent titration; while
butyric acid, on treatment with bichromate is partially oxidized, pre-
sumably to f-oxybutyric acid, thus increasing the yield of acetone.

Glucose also presents obstacles, since on treatment with chromic
acid small quantities of what I believe to be formic aldehyde
and formic acid are formed. This trouble is very easily over-
come by redistilling the distillate supposed to contain acetone
from f-oxybutyric acid, with hydrogen peroxide and alkali.
The acetone is not attacked! but the aldehyde (?) is oxidized
to the corresponding acid, which is held back by the alkali.

Lactic acid, when present, also interferes, presumably by the
formation of acetic aldehyde which passes over with the acetone,
and yields iodoform on treatment with hypoiodite; but this diffi-
culty is likewise wholly removed by subsequent distillation with
hydrogen peroxide and alkali.

The following experiments illustrate the above statement
concerning glucose and lactic acid.

2 grams glucose in 500 cc. water + 1o gm. sulphuric acid distilled,
dropping in 3 per cent potassium bichromate (12 gm.). 600 cc. of the
distillate titrated with iodine and thiosulphate gave the equivalent
of 19.7 mg. acetone; but no iodoform was found, the solution remaining
clear.

I gram glucose in 500 cc. water + 5 cc. @-oxybutyric acid solution +
ro cc. sulphuric acid. Distilled, dropping in 3 per cent potassium bichro-

mate (9 gm.). Distillate = 46.2 mg. acetone. Duplicate = 47.3 mg.
acetone.

No glucose, 5 cc. f-oxybutyric acid solution, etc. Distillate = 36.7
mg. acetone. Duplicate = 38.7 mg. acetone.

r gm. glucose + 50 cc. of a different @-oxybutyric solution + roc.
sulphuric acid + water. Distilled, dropping in 2 per cent potassium
bichromate. First distillate redistilled with 25 cc. 3 per cent hydrogen
peroxide + 5 cc. ro per cent sodium hydroxide. This distillate = 41.6

1 Dakin: This Journal, iv, p. 81,1908. The following experiment shows
the stability of acetone in the presence of hydrogen peroxide. 25 cc. of
a dilute acetone solution titrated direct with standard iodine and thio-
sulphate = 36.3 mg. acetone. 25 cc. acetone solution + 400 cc. water

Se

+ 25 cc. 3 per cent hydrogen peroxide + 5 cc. ro per cent sodium hy- a

droxide. Distillate contained 36.3 mg. acetone. Duplicate distillate
contained 36.9 mg. acetone.

a ae

Philip A. Shaffer 219

mg. acetone. Control determination in pure solution without glucose
= 42.8 mg. acetone.

2.5 gms. calcium lactate in 500 cc. water + 10 cc. sulphuric acid.
Distilled, dropping in 2 per cent potassium bichromate. 400 cc. of the
distillate gave the equivalent of 80 mg. acetone, with the formation of
much iodoform.

1 gm. calcium lactate in 500 cc. water + 25 cc. acetone solution
(= 36.3 mg. acetone) + 1occ. sulphuric acid. Distilled 500cc., dropping
in 2 per cent potassium bichromate (4 gms.)

(a) 125 cc. distillate titrated direct = 18.3 X 4 = 73.2 mg. acetone.

(b) 250 cc. distillate redistilled with hydrogen peroxide + sodium
hydroxide = 17.9 X 2 = 35.8.

(c) Gave strong positive reaction with ammoniacal silver nitrate solu-
tion containing sodium hydroxide.

By the use of basic lead acetate and ammonia, and of the sec-
ond distillation with hydrogen peroxide and alkali it is possible
to overcome all of the more serious drawbacks to this method
which I have so far encountered.

Phenol and skatol from the conjugated acids may still cause
some inaccuracy but the greater part of these substances pass
over in the preliminary distillation with sulphuric acid alone; and
in any event this error is very small—from 20 to perhaps 100 mg.
in a 24-hour urine.

Other substances, such as leucin,! sometimes present in urine
in relatively small quantities, may be capable of forming acetone
under the circumstances, but it is not likely that this possible
error is of sufficient size to materially detract from the usefulness
of the method.

A number of different normal and pathological urines, not con-
taining f-oxybutyric acid, which I have examined have given
results equivalent to less than 0.100 gram f-oxybutyric acid for a
24-hour urine (0.010 gram to 0.080 gram).

In the routine use of this method we determine at the same
time the preformed acetone plus acetone from diacetic acid, with
very satisfactory results. The three distillations of the Mes-
singer-Huppert method are for practical purposes unnecessary,
though two distillations are desirable.

The method which I propose for the determination of acetone
+ diacetic acid, and of B-oxybutyric acid is carried out as follows:

1 Dakin: This Journal, iv, p. 63, 1908.

220 Determination of 8-Oxybutyric Acid

From 25 to 250 cc. of urine, depending upon whether much or
little G-oxybutyric acid is expected,! is measured into a 500 cc.
volumetric flask and an excess of basic lead acetate and 10 cc.
of concentrated ammonium hydroxide are added. This is diluted
to the mark with water, shaken and filtered. An aliquot part
of the filtrate (usually 200 cc.) is diluted with water to 500 to
600 cc., 15 cc. of concentrated sulphuric acid and talcum added,
and the mixture distilled until 200 to 250 cc. of distillate has col-
lected. (Distillate A.)

The distilling flask (800 cc. Kjeldahl’s are convenient) must be
fitted with a dropping tube and water run in to prevent the vol-
ume in the flask from becoming less than 4oo cc.

Distillate A contains the acetone, preformed and from diacetic
acid, and also the volatile fatty acids present; to remove the
latter, including the disturbing formic acid, distillate A is redis-
tilled after adding a little fixed alkali (5 cc. of 10 per cent sodium
hydroxide). This distillate (A,) is titrated with standard iodine
and thiosulphate.

The residue of urine + sulphuric acid from which Distillate A
was obtained is again distilled, dropping in 400 to 600 cc. of o.1
per cent to 0.5 per cent potassium bichromate solution. (Ordi-
narily o.5 gm. potassium bichromate -will be sufficient; with

1uch sugar or when much urine is used even 2 or 3 gms. may be
necessary.) The potassium bichromate solution should not be
added faster than the distillate collects unless the boiling liquid
turns pure green, thus indicating that the bichromate is being
used up more rapidly. When about 500 cc. of distillate (B) has
collected, 20 cc. of 3 per cent hydrogen peroxide is added to the
distillate (or the distillate may be collected in a flask containing
the hydrogen peroxide) together with a few cc. of sodium hydrox-
ide solution and this distillate (B) is again distilled. This dis-

1 Of diabetic or other urines giving a strong ferric chloride reaction for
diacetic acid, or when 5 to 10 grams or more f-oxybutyric acid is expected,
25 to 5occ. or even less urine will be found a suitable quantity; when little
or no £-oxybutyric acid is expected, 125 cc. or 250 cc. may be used. In
either case the amount taken is sufficient for duplicate determinations.
The aim should be to use such a quantity of urine as will give between
25 and 50 mg. of acetone from f-oxybutyric acid.

a

Philip A. Shaffer 221

tillate now obtained (B,; 300 cc. should be distilled) is titrated
as usual with iodine and thiosulphate.

A good condenser must be used for the distillations, but it is
not necessary to cool the distillates with ice, as is sometimes
recommended.

We find it convenient to make our thiosulphate and iodine
solution 103.4 percent §;. Eachcc. of the iodine solution is then
equal to 1 mg. of acetone or to 1.794 mg. of §-oxybutyric acid.
The thiosulphate is accepted as the standard, and is restandard-
ized from time to time by ;4, potassium bi-iodate.

A few determinations in urines containing no f-oxybutyric
acid and also of normal urines to which were added varying amounts
of synthetic f-oxybutyric acid, are given below.

Lab. No. 622 Mixed urine of 5 days from a case of epilepsy. Average
for 24 hours.

Acetone + diacetic = 0.018 gm. acetone.
B-Oxybutyricacid = 0.018 x
0.025 fe

Lab. No. 626. 24-hour urine from a case of Grave’s disease. 1220 cc.
Sp. gr. 1.017.

Totals: Acetone + diacetic = 0.0

0.0

5 gm. acetone.
8-Oxybutyric = 5

1
2 “
Lab. No. 637. 24-hour urine from woman 5 months pregnant.

Totals: Acetone + diacetic = 0.032 gm. acetone.
§-Oxybutyric = (Des y 4

Lab. No. 638. 24-hour urine from woman 7 months pregnant, who
had a moderate acidosis some weeks earlier.

Acetone + diacetic = 0.006 gm. acetone.
8-Oxybutyric = 0.007 z

Lab. No. 641. Specimen of urine representing probably about 12 hours

_ from a case of puerperal eclampsia. Urine passed about the time of the

convulsions 400 cc., sp. gr. 1.019; 0.25 per cent albumen.

Acetone + diacetic = 0.027 gm. acetone.
B-Oxybutyric = 0.042 bs

Mixed normal urine.

For 1000 cc.: f-oxybutyric = 0.025 gm. acetone.

To the same urine was added 1.50 gm. §-oxybutyric acid per liter of
urine.

222 Determination of §-Oxybutyric Acid

For 1000 ce. of urine, 8-oxybutyric acid = 0.860 gm. acetone.
“ “ “ “ “ O 2 865 “
“ “ “ “ “ O i 890 “
“ “ “ “ “ (0) i 860 “
Average = 0.869 gm. acetone.
1.56 gm. f-oxybutyric acid.

Determinations in same #-oxybutyric acid solution: (a) 0.825 gm.;
(6) 0.8380 gm.; (¢) 0.840 gm.; () 0.840.
Average = 0.834 gm. acetone,
1.50 gm. fZ-oxybutyric acid.

I

Normal urine: sp. gr. 1.021.
Results are given in grams of acetone per 1000 cc. of urine.
I. Without treatment with basic lead acetate and ammonia.
Acetone + diacetic: Distillate A titrated direct = 0.060 gm.
Distillate A redistilled with

sodium hydroxide = (0,0ilGm

f-Oxybutyric: Distillate B titrated direct = 0.28 “
Distillate B redistilled with
hydrogen peroxide and

sodium hydroxide = Os1S rae

Il. After treatment with basic lead acetate and
ammonia.

Acetone + diacetic: Distillate A titrated direct = 0.042 “
Distillate A redistilled with

sodium hydroxide
B-Oxybutyric: Distillate B titrated direct
Distillate B redistilled with
hydrogen peroxide and
sodium hydroxide = 0.018 “

III. Same urine + #-oxybutyric acid solution (2000
cc. f-oxybutyric acid solution contained ac-
cording to determinations in the solution 0.686
gm.acetone). 2000 cc. of @-oxybutyric solu-
tion to 1000 cc. urine. Treated with basic lead
acetate and ammonia.
Acetone + diacetic: Distillate A titrated direct = 0.036 “
0

“

|
oo
SS
=z S
> 00

Distillate A redistilled with

sodium hydroxide. = 0

f-Oxybutyric: Distillate B titrated direct = 0.706 “
0

Distillate B redistilled with
hydrogen peroxide and
sodium hydroxide = 02670

Philip A. Shaffer 233

IV. Same proportions of urine and §-oxybutyric
acid asin III; 8 gms. glucose added per 100 cc.
urine. Treated with basic lead acetate and

ammonia.
Acetone + diacetic: Distillate A redistilled with
sodium hydroxide = 0.016 gm.
B-Oxybutyric acid: Distillate B redistilled with
H,O, + N,OH = 0.664 “

V. Same as III, but 4 gms. calcium lactate added
per 100 cc. urine. Treated with basic lead
acetate and ammonia.
Acetone + diacetic: Distillate A titrated direct = 0.042 “
Distillate A redistilled with
sodium hydroxide == {UU v
f-Oxybutyric: Distillate B titrated direct = 2.338 “
Distillate B redistilled with
hydrogen peroxide and
sodium hydroxide =) (I) (gef
VI. Sameurine + one-half the amount of $-oxybu-
tyric acid (= 0.343 gm. acetone per 1000 cc.
urine, according to previous determinations
in pure solution). Treated with basic lead
acetate and ammonia.
Acetone +diacetic: Distillate A redistilled with

sodium hydroxide. =. 070m =
8-Oxybutyric: Distillate B redistilled with
hydrogen peroxide and

sodium hydroxide. = 0.330 “

0.334 “

0.340 “

Lab. No. 627. Diabetic urine, 1710 cc., sp. gr. 1.025. Determinations
made as described on p. 220.

Total acetone + diacetic = 0.81 gm. acetone.

Total Z-oxybutyric = 2.00 gms. acetone or 3.60 gms. $-oxybutyric acid.

If 250 cc. of this urine (= 0.525 gm. $-oxybutyric acid) were used for
a determination by any of the ether extraction methods, the ether residue
dissolved in 50 cc. water and the resulting 1.05 per cent solution read in a
200 mm. tube in the polariscope, an error of 0.05° in reading would be 9.5
per cent of the total.

I am indebted to Mr. E. A. Reinoso for carrying out many of
_ these experiments.

THE INFLUENCE OF COMPLETE THYROIDECTOMY AND
OF THYROID FEEDING UPON CERTAIN PHASES
OF INTERMEDIARY METABOLISM.

By FRANK P. UNDERHILL anv TADASU SAIKI

(From the Sheffield Laboratory of Physiological Chemistry, Yale University.)
(Received for publication, July 10, 1908.)

Despite the overwhelming mass of literature recorded! rela-
tive to the influence of the thyroid upon the nutritional changes
in the animal organism the function of this gland still remains in
practical obscurity. Attempts toward a solution of the problem
have been made from various viewpoints, such as investigations
into the metabolic processes attendant upon pathological condi-
tions of the gland and of the action induced in such cases by the
administration of fresh and dried glands and of various com-
mercial preparations. A large number of observations are on
record of feeding or injection of the thyroid or its constituents to
normal man and animals with subsequent determination of the
effects exerted upon the general metabolic activities. Studies
have also been carried out for the purpose of ascertaining differ-
ences in nutritional processes after removal of the organ or a por-
tion of itin an attempt to correlate gland function with specific
symptom of metabolic activity, and finally the endeavor to sepa-
rate thyroid function from that exercised by the parathyroids is
at present a theme of much interest and scientific activity.”

The present investigation was undertaken with a view of ascer-
taining possible changes in the intermediary metabolic processes
after complete thyroidectomy and after thyroid feeding. The
occurrence of unusual variations in the excretion of one or more
urinary nitrogenous constituents might indicate a line of attack

1 For a review of the literature on thyroid, see von Noorden, Handbuch
der Pathologie des Stoffwechsels, ii, 1907.

? Cf. MacCallum and Voegtlin: Bulletin of the Johns Hopkins Hospital,
Kix, p.g1, 1908.

225

226 Thyroid and Metabolism

which followed would possibly render thyroid function less
obscure. Accordingly the composition of the urine, that is, the
relationship of certain of the nitrogenous constituents, has been
determined under the experimental conditions outlined.

EXPERIMENTAL.

Methods. The methods employed for the analysis of the urine
were those commonly in use in this laboratory.1 The feces were
not examined. Sugar determinations were carried out according
to the Allihn method and the procedure of Pfliiger? was followed
for the estimation of glycogen. The thyroidectomy experiments
were performed upon male dogs and catherization was not prac-
ticed. Inasmuch as only the relationships of the nitrogenous
substances were desired exactly twenty-four hour specimens
were not essential. Bitches were employed in the feeding experi-
ments and the urine was obtained in twenty-four hour periods
by catheterization, care being taken against the possibility of
bladder infection. The thryoid preparation fed, desiccated thy-
roid, was obtained from Armour & Co., and contained 12.75 per
cent of nitrogen and o.12 per cent of iodine. Whenever thyroi-
dectomy was performed the operation was carried out under con-
ditions as aseptic as possible and both the thyroid and para-
thyroids were presumably completely removed. None of the
animals operated upon gave evidence of any disturbance due to
the operative technique itself and all wounds healed rapidly.

The Composition of the Urine after Complete Thyroidectomy.

Aside from the observation® that extirpation of the thyroid
causes a decrease in nitrogen elimination little investigation has
been made concerning protein metabolism under these conditions.
A study of the partition of the urinary constituents indicative
of the more important lines of nitrogenous metabolism has there-
fore been carried out, the details of which are to be found in
Tables, 1,273, 4 andss.

‘Cf. Underhill and Kleiner: This Journal, iv, p. 165, 1908.

* Pfliger: Archiv fur die gesammte Physiologie, cxi, p. 307, 1906.

3 Dutto and Lo Monaco: Archives italiennes de biologie, xxiv, p. 196,
1895; Roos: Zettschrijt fir physiologische Chemie, xxi, p. 19, 1895-96.

227

Frank P. Underhill and Tadasu Saiki

‘TeuLouqe Suryjou payeoa
-o1 Asdoyny ‘a8vo ul pvep punoy 30q
‘us $0'0 ‘oun UI plov olueau
-AyY “‘puryq pue yeop Apuoredde yeunuy
‘aAtjisod prove
o1j008Ip puv UvOIpUI ‘uIOJOI1d OF sysaT,
‘soyein jo yisodep Aavoy pourezuo0s oul,
“I0}JUM “09 YE YURI
‘posuvyoun ATTeoyovId UoIpuood s,;euNUy
“TOVOM
‘00 00G yurIq ‘yoo} puv yoou ‘yynow
jO sojosnut jo ssurojimy ArezunToAuyT
‘aAOGB pejoU SUOIyOVEI DAVE oULI)
‘s]80} 9AOGR OATS
pue ulejoid peurejuo0o outm {paqru0A
Ajoyelpowuat ynq ATrpoesS yunap svar YW
‘S]S0} plore
d1Jo08IP PUB ULOIPUT osudjzUT VALS OUI

“SUUV WAU

te)
ei

Hef

Git

*quao wad

"N uluyeer9

——_————— ss Et

‘sony OF ‘ayhian {2 saqo}9Q Bop woul parowas spunjy :~ bog
‘ANOLOMGIOUAHL ALATAWOO UALAV ANINA JO NOILISOAWOO

9°6

wD 1d
wD 0

9°8

8°6

a9

"quad wad

"N—HN

°

ae)
al
i) o

iS
S
q
7)

*N ulayeero
peuLiojalg
[840.L

*N ulunyeve10

ye

“snuq}I]
0} UOTOBaYy

“AYIABIZ OYTOedg

cE

FE

| | ©
om
™

‘BUIUDAD OY} Ul pvep
punog ‘posoqey uoresud
-sar ‘UOIyIpuoD 100d ul SOG

“Hla 99 007

yuriq, ‘sapsnur jo ]o.1}
-u0d dAvy jou pIp [eUlUy
‘IvSNs OU PoUTe}ZU0D 9UIIL] |
“IVSNS POUTe}UOD OULL/)|

“SMUVNGY

Thyroid and Metabolism
|

228

AO We Sot 6'T 174 SOnO GL 0) -Sh Os S255 0s coe ” Lv | COL | SZ
. . | .
OT 0% O'€ 88 80:0 |, $i'0) |-e6, 0) Z9 0) 19 2 ” Gy | Z9OL | ve
i 0% ts 88 GO'0 | OT'0 | SI'0 | S¥'0 | 80° | PHB | 9E | OZ | EZ
¢°0 Gs Ons LOD | CORO) OL OF Gh ON OF OF kG) S55 es Roce eZOcaican ©
*quaa wad | *yuwao wad | ‘quad wad | *yuao wad Unid wo..b. wpb woib | *wou6 —O1 "29 ‘LOGI
2) at 8 2 i) aig | oR < eS sn lignes y
gine? | (hey | eee ee | be | ee le eee eee
i=} i) ry -
gee) eo dale Wea ee eeehere cea
aa a av | 2 Be
S
‘TE 4290199 ‘sony 6 fo bop wo parowas spunjy :g bog
“AWOLOUGIOUAHL ALATAWOO UALAVY ANINUN JO NOILISOdNOO °% WIAVL

229

Frank P. Underhill and Tadasu Saiki

” ” ”
480} plow oeORIp Suo14s

OAS OUI) ‘posnjot poo
“YIU “99 YOE pus pveiq
Aap A19A swuievis oOg—
Jol, ‘posueyoun uorytpuoy
“W[fur "09 ONE pus
pvoiq Arp A19A sueis OS
—jIq ‘saudsXp ‘Auvyoy,
“y[fur *09 OLE YURI
‘9]80} plow 91}008
-Ip puv urojoid oAes oun)

“SMUVANEU

0°0
£°0
0°0
OT
wo

OT

quad wad

"N UlBeID

19

“NU UlutyBaI0

'G

“I!

“quad wad

peullojaig

|

8°%

BG

au

quao sad

*N uruljeer0
12701

9°8

yr
oO 0

ee

"quad sad

‘NHN

oo

°

i
i
Ss
q
i)

‘N uljverp

peuLlojaig

"N Ulu BaI0
1330.1,

“N Ulut}e010

Gc

CF

9G"

So

4

60° €

06°€

9c 'P

tT SG

“wpb

"N [870.1

*snuryt]
0} UOTVORAY

I
*£4IAB1Z OYIOedg | iS) oF

89

"9g 4290709 ‘sony g fo Bop D wouf panowas aam SpuvD]D § boq

‘XWOLOAGIONAHL ALATAWOO YALTVY ANINN FO NOILISOdWOO

‘© AIAVL

Thyroid and Metabolism

230

= ee eee

“P29
‘uoltpuoo dood AtoA ul yeutuy| T['0 Ta (0) T1100 | OOT'O | 260'0 | €8'8 ” €¢ | S8T |Lz
‘poavoddesip sey vaoudsfq] g'0 | €°0 | gt || $e0°0 | e80°0 | get'o | go'or { go, | Ox | AOL 18e
ple) OF | SFI |S%
‘waoudskq “1oyBM “00 QOS YURI) ZO 10 VG T10°0 | IT0°0 | 8ST°0 | 2¢°9 HIe | €S | SOL |he “Ie
"$6 youn py ‘sony gg {o Bop wouf paaowas spunj,y :,7 bog
i a : “Poy
‘uontpuos 100d ATOA Ul [BUY Z'0 9081} ” §10'0 | 62°F ” €¢ | 06 |T&
ia) ” ¢10 0 | 09'€ ” 8F | 08 j\0€
10 v0 200° 0 ” Gc0 0 | IF ” LY | SEL 16%
60 eoel} | 9081} | 1200 | 62'S ” O€ | $8 |8z%
‘[BLIO}VUI Po}TULOA PoUTe}UOD 41
90UIS PopIVOSTp SVM OZ YOIB JOoulI)| §'O Z 0 £°0 ¥c0 0 | 1200 | €20°0 | 12°9 ” 6G | 8ZL |2e
‘UlOJUBT[B JO speysA10 pourezuod oul,
“TOYEM “99 NOS + Y[lUL “99 OST = oI 9'T 8900 1g § prow’ | co | €9 |SZ
£°0 LG €c0 0 | SPIO | 88°9 HI? | Cf | C8 |hS JPW
quad wad | *zUa0 wad | “quad wad woniBb “wnib “mp8 “mp6 —OL 29 *SO6I
= > y eS > 9 3 i |i ia b
e 5 af a 5 g. Bee ed ee See
‘SMUVWA . g. Z é, S. z Z eh 4 Fs
a p : A B ; 4 25 09 :
=

Se youwyy ‘sony FT {o Bop v worl paaowas aan spunjy :Q7 bog

“AWOLOGMGIOWUAHL ALATINOD YUALAV ANINO AO NOILISOdNOO

‘p @IAVL

231

Frank P. Underhill and Tadasu Saiki

'€ AVBL Ul woars ‘g Soq jo
SULN WOlZ paurezqo o1aM soinSy osory,
‘eAOgR sv [BUIUR sues WOI
‘oulin sep
omy JO aztsodui0d ev uo pouTezqo
S}[NSod oie sosoyjuered ur somsnyy
% 9yq8T, ur poekojdwa yeunue jo
oULIN WOLF pourezqo a1aM sorn3y asouy,

——————

“SMUVANAY

8° OF

quad Jad

>)
[Br}NON

6°9
“quad wad

‘8

[Bo19U4 A

0°2¢
“yuan wad

)
ormBed.10UT

“AWOLOMUGIOUAHL GALATINOO UAITV WAHATAS AUVNINA AO NOAaIULstIa

180°0 L10'0 180°0 S8t'0 | T “AON
860°0 $20°0 Z6T 0 PIE'O | SZ

okt ‘oO 2200 9910 6180 | #2» €2°290
(09Z°0) | (SFO'0) | (gee°0) | (8E9"0)

“mnie *Upab “und “up "L061

i 2 i ie
[e1yneN [eo1eqyg «=| olmes10UuT [B10 *ay8q
‘G WIAVL

222 Thyroid and Metabolism

The most significant change apparent in the distribution of
the urinary substances determined is the increased ratio of nitro-
gen in the form of ammonia. ‘This cannot be entirely accounted
for by the condition of partial or complete inanition since the
increase is greater than occurs in the latter conditions.t More-
over, feeding does not affect this ratio as may be seen from Table
3 where the percentage of ammonia is as high after the intake
of food as it is in Table 1 where the animal was fasting. It is
therefore probable that a corresponding decrease in the output of
urea occurred. Whether any relationship exists between the
increased output of ammonia and the practically constant posi-
tive test for diacetic acid was not determined. An interesting
feature with respect to the extirpation of the glands in the pres-
ent experiments is the almost invariable alkalinity of the urine
for a few days following the operation. While this work was in
progress a paper appeared by Coronedi and Luzzatto? in which it
was shown that the urine is alkaline after thyroidectomy owing
to an increased output of ammonia. These conditions were
entirely independent of diet and were not due to infection from
catheterization. Our own experiments confirm those of Coronedi
and Luzzatto with regard to an increased output of ammonia but
differ in that the urine remains alkaline, only for a day or two
although it must be admitted that the acid reaction is never as
strong after thyroidectomy as that usually found associated with
dog’s urine.

The total creatinin excretion is practically the same in the
three animals examined although Dog 1 weighed twice as much
as Dog 3. Whereas the percentage of nitrogen in the form of
creatinin is fairly constant, that existing as creatin varies widely;
for example, the urine of Dog 1 contains little or no creatin nitro-
gen while in that of Dog 2 creatin nitrogen is present at times to
the extent of 1 percent. The elimination of nitrogen as purin
bodies and allantoin isabout the same as the excretion of these sub-
stances noted in the urine of a fasting dog. The few phosphorus
determinations made show a small output, possibly indicating,

1 Oesterberg and Wolf: Biochemische Zeitschrijt, v, p. 304, 1907; Under-
hill and Kleiner: loc. cit., and also cf. Richards and Wallace: This Journal,
iv, p. 179, 1908.

?Coronedi and Luzzatto: Archives italiennes de biologie, xlvii, p. 286,
1907.

Frank P. Underhill and Tadasu Saiki 33.3

when considered with the low purin and allantoin excretion, a
low rate of nuclear disintegration. Sulphur partition in the urine
gives evidences of little change from the normal although the
number of estimations made are too few to draw definite conclu-
sions. According to Ducceschi' the total and neutral sulphur of
the urine is increased after thyroidectomy which has been inter-
preted as significant of a lowered rate of oxidation.

In cachexia thyreopriva’ pathological changes have been
observed in certain of the nervous structures and Vassale and
Donaggio® noted degeneration of the pyramidal tract following
thyroidectomy in dogs. Attendant upon certain acute degenera-
tive diseases of the nervous system Rosenheim* has shown the
presence of significant quantities of cholin in the blood, cerebro-
spinal system and tissues. The possibility therefore suggested
itself that if thyroidectomy were responsible for degeneration of
nervous structures the presence of cholin in the blood could be
taken as evidence of disturbed nervous metabolism. To test
this possibility two animals were allowed to live after thyroid
extirpation until death seemed imminent when they were chloro-
formed and blood withdrawn fromthecarotid. Cholin was tested
for by the method outlined by Rosenheim’ but the results were
negative.

Certain Aspects of Carbohydrate Metabolism ajter Complete
Thyroidectomy.

A former communication® from this laboratory corroborated
Scott’s’ demonstration of the ability of the animal body to
utilize large quantities of dextrose subcutaneously introduced.
After doses varying from 5 to 7 grams of dextrose per kilo of body
weight injected hypodermically into dogs only traces reappear
in the urine and then only for the first day or two subsequent to

! Ducceschi: Jbid., xxvi, p. 209, 1896.

*Schafer: Text Book of Phystology, i, p. 941.

3Vassale and Donaggio: Archives italiennes de biologie, xxvii, p. 129,
1896.

* Rosenheim: Journal of Physiology, xxxv, p. 465, 1906-07.

5 Rosenheim: Jbid., xxxili, p. 220, 1905-06.

® Underhill and Closson: This Journal, i, p. 117, 1906.

™Scott: Journal of Physiology, xxxiii, p. 107, 1902.

234 Thyroid and Metabolism

the injection. Comparable experiments (see Table 6) have been
carried out upon two dogs after complete thyroidectomy.

The results obtained under these experimental conditions,
however, present an entirely different degree of utilization of
the sugar introduced. Under normal circumstances little more
than a trace of the dextrose injected is eliminated in the urine,
whereas with Dog 3 almost one-half the quantity introduced
reappeared before the animal died. That this did not represent
the entire quantity which would have been excreted had the animal
lived longer is shown by the observation that the urine found in
the bladder after death possessed strong reducing powers. These
results would appear therefore to indicate a diminished ability
on the part of the body to utilize carbohydrate material, at least
when subcutaneously introduced. The lack of utilization may be
regarded as a lessened power of oxidation or glycolysis, or as a
decreased ability to transform dextrose into glycogen.

An experiment was performed to test the latter hypothesis, the
details of which follow: A dog of 10.2 kilos was allowed to fast
for one week and then complete thyroidectomy was performed.
On the two following days 40 grams of dextrose (80 grams in all)
were injected subcutaneously. The animal was killed next day
with chloroform and glycogen determinations were made on the
liver. The urine contained 1.6 gram of dextrose. The liver
weighed 261 grams and contained 0.43 gram glycogen. A control
experiment with a normal dog, after a like period of inanition and
a similar injection of dextrose yielded 0.64 gram glycogen in a
liver weighing 270 grams. The quantities of glycogen obtained
in the two cases force one to the conclusion that complete thyroi-
dectomy does not appreciably reduce the glycogenic function of
the animal body. A lessened oxidative or glycolytic activity sub-
sequent to complete extirpation of the thyroids would therefore
appear to be indicated. This assumption is in harmony with the
observations of Magnus-Levy! who has shown that the respira-
tory quotient is low in myxcedema and cretinism, and is high in
Basedow’s disease and after feeding thyroid preparations to nor-
mal man.

1Magnus-Levy: Zeitschrijt fur klinische Medizin, xxxiii, p. 269, 1897,
and lx, p. 177, 1906.

Poo

Frank P. Underhill and Tadasu Saiki

‘uorjonper su0.14s

9ARS JOppLlg UL eUM/) “odvo UI peop punoj Soq) §6T'0 oS 9T 0 66% ” €€ GIL | 96
SG AON JOF oUlIN ON “JoPVM YUBIp Teuluy; Tg 'T Gg 020 69° € ” cP OLE” 1-86
‘oAtyowut AT[eoydo sem ouLin uoTy
“BPUSUIIO} JOIFY “AO}PVM “00 OG Yuvip [ewuMluy|) GF T €'s 1a'0 ve 9 ” IP 6ZT | &
‘uOIyNOs yusd sed QE Ul OTL
Jod 9sso1jxop sures ¢ JO uoTjoolul snoouRynoqng eZ ro 0) 1S°¢ poe O9T | 3 ‘AON
‘sony Y'g = rwybhram ‘Tg saquaao yy ‘bop woul paaowas spun)y :% bog
‘uolonpet
Buoys 9AVS Joppr[q UL ou, ‘pap punojsoq) ¢Z'l | 26 Cie peor ‘9 Rel
” ” ” or & vs 0€ 0 rg'¢ ” SP Oot | Ee *0r
‘ulojord poureyuoo ouNy)) pT L'6 0&0 60°€ ” 6P o8 6
‘wut OOT UF o4°0 —
uoleyuouLey Joyye oAoe Ajpeoydo sung
‘(qua puew proiq) pooy urejzor OF a_qevun jeulluy|) Z6'T +6 T2'0 Sze “4 eC 6L 8
‘oatjovut ATeoydo svA puw uoonpodl
OU OAVS JULIN UOTYVJUDWIOJ 109JB ‘}S0} plow
d1j90¥IP SUOI}S VAVS puv UldJOId pouleyUOD OULIY)| SPs 6° OL Te 0 $3'Z e eG Z8 L
‘uoIyNpOS yued Jed OZ Ul OTDT
tod osoryxep survis g Jo uoooful snooueynoqng Z'6 820 FO’ 5; TF t6 9
"§ SGBL al peal]
-qno yey} JO UOIVENUTVUOD & st yUoUTLIedxe SIT, € Or 6&0 06°€ pre OF OZI Gg “AON
Unsb ‘quad sad “UDiB “Unb —Oor ms) “L061
‘SuUV WAY -asoryxeq | "N—°HN |‘N—HN | ‘N [8104 | ronoton oraeda, ‘aumnjod | “938

‘g a1qnT, ur sv bop aupg :¢ bog

236 Thyroid and Metabolism

The Composition of the Urine of Normal Dogs after Thyroid Feeding.

As a consequence of the large number of investigations! that
have been carried out with thyroid feeding both with normal man
and animals the idea has gradually been evolved that the thyroid
gland acts as a stimulus to protein catabolism. In experiments
with animals Roos’ and Voit? in particular have been largely
responsible for the theory as to this mode of action of the thyroid
on metabolism. In their investigations it was impossible to
maintain animals in nitrogenous-equilibrum, more nitrogen being
eliminated than was introduced. Schdndorff* obtained similar
results but offered the explanation that the increased nitrogen
output was due to an increased excretion of urea and other extrac-
tives present in the body fiuids. According to this author pro-
tein in the body is not attacked if sufficient fat is present. The
more recent work of Schryver’? demonstrating the influence of
thyroid feeding upon autolytic processes in the liver would tend to
corroborate the view that thyroid feeding has a stimulating action
upon catabolic processes although Wells® failed to get evidence
of such an action when thyroid extract was directly added to liver
pulp.

The experiments here recorded were made with two bitches fed
upon a standard diet composed of meat, cracker meal and lard
with a definite volume of water. The animals received this diet
several days previous to the beginning of the experiment and were
practically in nitrogenous equilibrium. Different quantities of
thyroid were added to the standard diet at intervals and the influ-
ence of this administration upon the composition of the urine
is shown in Tables 7 and 8.

From these tables it is readily seen that urinary nitrogen excre-
tion is but little influenced by thyroid feeding. When thyroid

1 A review of the literature is given by Andersson and Bergmann: Skan-
dinavisches Archiv Jur Physiologie viii, p. 326, 1897-98. The literature
is brought up to date by von Noorden, loc. cit.

? Roos: Zeitschrijt jir physiologische Chemie, xxi, p. 19, 1895-96; Xxv,
Pp. 1, and p. 242, 1898; xxvii, p. 46, 2809:

3 Voit: Zeitschrift Jur Biologie, xxxv, p. 116, 1897.

4 Schéndorff: Archiv fur die gesammte Physiologie, Ixvii, p. 395, 1897.

§Schryver: Journal of Phystology, XxXxli, p. 159, 1905.

® Wells: Americal Journal of Physiology, xi, p. 351. 1904.

Se

237

Frank P. Underhill and Tadasu Saiki

” ” €°2/6 O16 SiS SP OS ESE OlGT 01800122 '0/Z6T O|0FOO/ZE‘Olge'8 » | VP 61 9%
” ” 6° ZIT T0 Fe aS Of FP 92'°0 91' 060 OTE O89T' 0 0FO O'ZE O'E9 LZ » | SS |09G| SG
‘sus ¢°g = poo} UI N
"(Biyxe) OF H 99 OL +
‘prorAy} ‘suas QZ + JoIp prepuryg ESP iL e600 ah F \0€ 0 ZI‘ 080°002'08F0'0 0ZT'0 FG'0 ZFS » | VE |S8T| FS
” ” 6 TIS LP eis 10 Zr F 8Z' 0 0L' 0/80 OST 0/290 O|}FOL OES 00G'S » | OF |SOT| €%
” ” Z ZS LP £8 0S O16 E|\2S OST 0\20 O/6T OSFO O\ZTO Oe OTS G » | 09 |2Z8 | &2
” ” VciS 16 € I'ZE'00'°F| FZ O|ZT 0/80 0.0Z OOTT 0/0c0'0 TZ OIST' S| ,, | 6¢ \S8 | Te
” ” 9° Zr Til FE O10 OL PF 1Z' 0\LT' 080° 0/§2' 0.080 0,9T0' 0.€Z OES" s » | FE ISOT| 0G
‘suis (9 = pooj UI N |
‘goIp prepueqg |¢°Zil 1/9 €/6 TP 1/9 € 1Z' O|8ST' O|\L0' O/€Z' OTST O/F60 O/ES'O\ZE'9 oy Mh keel OTT, 61
” ” Ge GlOeLiGesOesgilOZes 9€°0,0Z'0|20°0.22 0|LFZ'0,90T 0.96098 9 » | SP \OFT| ST
” ” QSL TLS OF O12 €|/9 OST 080092 O9EF 0, ZE0 0/96 0.689 Sl VSeiGe Lez
” ” 9°Zj0 T/L |e Sis OL EEE 0\8T' 0\20' 0\Sc 0 ZGE‘O|9TO' OE 0€2L'9 op || We) GIT OL
” ” 9°Z/0'T|L°€/S' T|T' 0/0 Ee O|LT 020° 0/7 0080 OOTO 002° 08F'9 |) acho Tet) GT
” ” IP ZIT TG Ele OL OL E||9E OST 0|80 0/9% 0\FZ0' 0\S¢0 Oeo OLE L » | GS \OFL| FI
” ” 6 LP Te Si6 ae OZ € 9¢ OSL OT O/E% 06ST OSTO 092'068'9 Fos STT| €T
‘sums 9°9Q = pooj UI N
‘prorsy} swS ¢ + JIG prepurrg |9'Ze 16 EF S909 'E| SF OST 080 OSs OFFI OZEO OTS OS'S 5 | CY GEalck
” ” €°Z/9 116 EIT Si9 OF Si|9F O/T 0/60 O}LS OlZZZ Ol\FEO OIST OS'S} >, | SS |FIT! TT
‘SUID (Q'°9 = POOj UI N
“yetp piepueyg
‘SOIL, O'8 = qysiem Apog |¢'Z¢'1\0'F/6 Ti8 119'€ 9£'O'£T' 0/80 0/TZ'0,00T 0 960'061'0,94' Ss; PHB oG |SOL| OT “FH
*9 ‘d\'a‘d|'a ‘d\'a‘d\'a *d\'a ‘d|| ‘wb mb wes | ‘we “wb “ub ‘wb | ‘we —O[ | °99 “S061
eeelz sii elezizie (gl el glal ele) ¢
Eigsig2| 8) 5) 4 Be ayia oie see Waban ea: |e leis Fe
; Pisels. |.) 2] 2 Bete es dh ee ies ioe See 8
SAUVANTH z 5.0 5 5 ° s zm 5.o g 5 . a np 93
Saliba chan Bree | heel Bless
a Az Byles EY
A |% : 5 2
"6 bog

— ee a on | ee. Re ee. Oe ee 8 meovarwr

Thyroid and Metabolism

238

‘SOI, €'L= FYsI0M Apog

"SUIS ("9 = pooj ULN
‘qoIp paepuryg
” ”

”) ”)

‘SUIS Q'9 = pooy ul N
‘jaIp paepuryzg

‘sus €E° ZT = pooy ut N

*(B1}XO) 10VOM 100 OT +
prorky} suis OG + JoIp pavpuryg

”) ”)

) ”
+B)

‘SUIS “2° = poo}; UL N
‘joIp paepuryg

“SMUV WOU

O'S TZ Sio'SiF'OlZ ellFe'O}
GUS TL eZ SL OTF] Fe 0}
9°29 TZ Heo FS OF Eire 0
¢'ZE LS Sie LP OF FILS 0
LTO 14° 3s S202 €|/68'0
0°29 Tig e8'Zi¢ Ole FlFz'O}
6'11¢ LP ge'08' 019 Sl1z'0
6 IF Te sio't| ls ‘eliezo
L GP IL" H/0' Th 016 g||Fz 0}
i@-zs'oles| e-oleelizz'0l
‘a‘ditadi:a‘d:a'd‘a‘da-d\| -w6 |
alexis |e |e) 2] »
| 5 IBS |B |
Hee WEVa WE BS |e ||
| 2 |2~i4 | 2

)

OG

“N unverg | S

oo 416

oo 656

*N uljBer0
pemiojelg

| (UW

ooo

6G"

*N UlUI}BeI0
[210.L

aa)
OFT
sian
880°

CIS

ror

910°
SPO-

“N UI0JURILY

SSeS

| wb

069°
"0.19"
“0.92
009°

‘0 0T-

“Ub

“N [8301

© 0 19 9

ple

0g
&¢

Gg

|
S
™

|
|
|

“snwqy]
0} UOTI Ova

“AQIABIZ DYTOIdS

wig o ky

‘CHANILNOO—) WTAVL

259

Frank P. Underhill and Tadasu Saiki

‘ued}voUN JoIp Jo UOIWOg “ws

OF JNoqe = poojy ul N “orp pavpuryig
‘uayoun Jorp Jo Uol.0g

‘Ud 8p = pooyurnN “yorp pavpuryg

‘us 9°G = pooJUulN “qolp paepuryg
‘WS FET = poojurN “qoIp

0} popp® OFF 199 0G + prlordyy ‘sws OF

7) ” ) ”

) ” ”) ”)
” ” ” )

” ” ”) ”)
US 9'6§ = pooj uN “qolp parpuryg
‘SUS $' ST = pooj
UL N ‘UozvO JOU Poo] oATQUAT = —“9yVB AM
09 0G + JoIp 07 poppe prordy} “sus OF
) ) 9 7

‘WS 9°6 = pooyur nN “4orp prepurig

“SuMUV NAY

i=)
-

a ono

€ Zi €| STIS SE E||LE 0/90 O/FL 0/0Z 0/920 O|89T 0102 O;S6'¢ ” FF |O8T) 8Z
; OF |SLZ) Ld
L’T\6'@| €° 1/9 €/6'S\9S OST O/ZT 0\6Z O|\OST 0,098 0/62 '0\16'6 ” 67 |S6c} 96
9° TZ Z| $ 8i8 E8 EFF OTL O/9T 0/27 0/9S8 0/828 O}8E'0/96'6 ”
TT8'T| T'T2°8/0'F|/09 0/80 O|FT O\ZS O/9ET O|FSF O|6F O|6T ZT, GF S63! 9S
Lrg 9 OL 10'S 190/80 '0/ZZ' 0.08 0\8Z8'0'0Z2' 0.89’ O/SF' ZI ” CE |69E) FS
T'dic c| O TIS TIO SSE O}LO OSL OFT O;T80 0/960 O\TE O;\ZT'9 ” OF |SST| €
TG6'&| G 0/6 1/9 F LT O|9T O\OE O|OFO O}FFL O;SE O|TS 2 ” €F |88T| Ze
G GIL SI oe pF L¢°0)€0'0/9T O\6L 0)060' 0/960 '0/TE°0)96'°9 | PlEv| OF |<TT| Tz
| | | (0Z
$ T/0' €/90 O}T O}L €\/82Z° 1/07 OO OOS O/OTO 0\0ZO O\ZE O|TS 9T|_,, OG \OSF lace
$8119 € LO|\F 0/69) GL TILT O|LZT O/FE' 0/020 ' 0/0F0'0|99 O/SF'6 | dure | 9g |0GZ| ST
P'1/6' S| 2°0)/2°0|6'S/|Z0' TIST' O6T 0\28 0/980 ' 0/680 °0\92°082°ZT| _,, TS |O€F) LT
136% €°0/0'F||\S6 0/20 O|ZT OlFZ'0| 5, (920° Olge'0/60'8 ” CF \00ZG} OT
6 116% G 0/0 F 60° O/}6T 0/80} 5, |920 0168 O|€o 6 ” 68 |S9G} ST
6 1h SG 6 OL € C0'0/22' 0/22 '0)99819 |ZZ0' OTF OFT IT pee) ch ZZ FI “Ure
ALL)
a‘d\'a°d| ‘ad |'a‘d\"a°d|| wb | -wb |} *wb | -wéb “wp “mB | wb “ub —O[ | 99 *S06I
ich yel|ife b> tg | ao) ie) Cog ing > a>] 2 4 bg io) < i]
B88) Fg. | g|esiga| e | 8 | 2/2 | eB] ea |e] &
s.9 58 ae e. ae ge R a ‘| = Be |e ié y
i=} A 4 ° °
BAER) 2 | A) 2 w B'S |\B Bel el Sieceet BBS |e ole
zea | 2 seg eek aha S| 8
é
8 bog

240 Thyroid and Metabolism

is first fed the nitrogen excretion in the urine may exceed the
quantity introduced into the body but this influence is apparently
soon lost on cessation of thyroid administration. A small dose
(5 grams) appears to exert as great an influence upon nitrogen
excretion as a much larger quantity.

Unlike the results obtained with thyroidectomized animals
no change occurs in the elimination of ammonia nitrogen. With
small doses of thyroid the output of purin and allantoin nitrogen
is variable but after continued large doses the quantity of purin
nitrogen is largely increased. Allantoin nitrogen is extremely
variable and appears to bear little relation to thyroid intake or
purin excretion. ‘The phosphorus excretion is low. The elimi-
nation of creatinin is scarcely affected by thyroid feeding. Crea-
tin was a constant constituent of the urine of both dogs. Indeed
in Dog g, Table 7, the nitrogen existing in this form is in excess of
that in the form of creatinin. Reducing substances were never
present in the urine. Thyroid feeding, therefore, whether in
small or large doses, single or repeated has little influence in chang-
ing the ratios normally existing between the urinary nitrogenous
constituents. Our experiences with thyroid feeding impresses.
us that the action of this gland on protein catabolism should not
be over-emphasized, but that more attention should be directed
toward ascertaining the influence of the thyroid on the metabo-
lism of the carbohydrates and fats, and especially its relation to
various inorganic! constituents of the body.

SUMMARY.

After complete thyroidectomy (and parathyroidectomy?) in
dogs the ammonia output in the urine is increased, even beyond
what is observed in starving animals.

The urinary nitrogen eliminated as creatinin, purin bodies
and allantoin is about equal in amount to what has been noted in
fasting dogs.

As in normal fasting dogs a variable and frequently not incon-
siderable output of creatin was observed.

Thyroidectomized dogs are incapable of utilizing subcutane-
ously introduced dextrose in anywhere near the same degree as

‘MacCallum and Voegtlin: loc. cit.

Frank P. Underhill and Tadasu Saiki 241

normal animals. It is thus suggested that the loss of the glands
may occasion a change in the gaseous metabolism similar to what
has been observed in myxoedema, cretinism, etc.

Thyroid tissue fed to normal dogs causes a slight increase in
urinary nitrogen excretion. This influence soon disappears on
cessation of thyroid administration. Small doses of thyroid
appear to have as pronounced an influence upon the nitrogen
elimination as larger ones.

A larger output of purin-nitrogen and a low output of phos-
phorus were observed after the continued administration of large
doses of thyroid tissue.

Thyroid feeding produces little change in the interrelation of the
mitrogenous urinary constituents. This is in harmony with the
accumulating evidence that these relationships are upset only
by exceptionally profound disturbances in the nutrition of the
cells.?

1 Jackson and Pearce: Journal of Experimental Medicine, ix, p. 552,
1907; Underhill and Kleiner: loc. cit.; Richards and Wallace: loc. cit.

ON A GLOBULIN FROM THE EGG YOLK OF THE SPINY
DOG-FISH, SQUALUS ACANTHIAS L.

By C. L. ALSBERG anp E. D. CLARK.

(From the United States Fisheries Laboratory at Woods Hole, and the Depart-
ment of Biological Chemistry of the Harvard Medical School.)

(Received for publication, July 6, 1908.)

The yolk of the hen’s egg contains a great deal of an iron-con-
taining phosphoprotein with the physical properties of a globu-
‘ lin, known as ovovitellin. It has been repeatedly made the
subject of investigation. The only other vitellins that have
been studied were obtained from the eggs of teleost fishes. A
study of the vitellins from other classes of animals seems desir-
able because it might conceivably throw light upon some of the
processes which go onin the yolk during development. It seems
that vitellin has the important functions of furnishing iron? and
pyrrol® for the hemoglobin formation, as well as some of the
phosphorus needed for growth. Even among the vitellins of a
single class, the teleosts, there seem to be considerable differences.
The vitellin of the carp’s egg is a glycoprotein yielding a reducing
substance on hydrolysis. It also contains phophorus and iron.‘
The vitellins obtained by Levene*® from cod-fish eggs and by
Hammarsten® from perch eggs did not yield a reducing substance.

It is therefore not without interest to investigate the egg
of a selachian as to its vitellin, particularly one that is viviparous
as is Squalus acanthias. It is conceivable that the carrying of
the eggs by the female may affect the composition of the yolk
even when there is no true placentation.

1Cf. Hammarsten: Text-Book, pp. 503-504, 1908.
? Bunge: Zeitschr. f. physiol. Chem., ix, p. 49.

5 Levene and Alsberg: This Journal, ii, p. 133.
‘Walter: Zeitschr. f. physiol. Chem., xv.

5 Levene: [bid., xxxii.

®§ Hammarsten: Skand. Arch. f. Phystol., xvii.

243

244 Globulin from Egg Yolk of Squalus Acanthias

The eggs of this species are quite round and about 2.5 cm. in
diameter. The membrane is elastic and when it is ruptured the
yolk flows out as a slightly yellowish cream much thinner than
the yolk of the hen’s egg. The yolk was mixed with an equal
volume of to per cent sodium chloride solution and extracted
with ether in a separatory funnel until but little fat passed into
the ether.’ It was then precipitated by dilution with water and
the precipitate allowed to settle. In a few hours the water was
siphoned off and the white precipitate centrifugated from the
remainder of the solution. Centrifugation is much to be pre-
ferred to filtration in dealing with vitellins, for the time consumed
is much shortened and the loss as the result of the substance
becoming insoluble upon the filter much diminished. The pre-
cipitate is then redissolved in 10 per cent sodium chloride and
filtered. The filtrate is reprecipitated by dilution and washed by
decantation with ether-water until the latter is free from chlorides
and biuret giving substances. It is then collected on a filter,
washed with alcohol, transferred to a boiling flask with reflux
condenser and boiled out with alcohol many times. Only slowly
and after a long time is all the alcohol-soluble material removed.
It is then boiled out with ether, and dried iu vacuo over sulphuric
acid at 70° C. The yield from a relatively small amount of yolk
was very considerable.

The material thus obtained is a more or less yellowish or
brownish white powder giving the biuret reaction and a distinct,
though weak, Hopkins-Cole reaction for tryptophan. Millon’s
reaction is exceedingly weak or possibly negative. It differs
therefore markedly as regards the protein color reactions from
vitellin. The orcin and Molisch reactions are negative. On
hydrolysis with mineral acid it does not yield a reducing sub-
stance, resembling in this respect the majority of the vitellins.
On boiling with alkaline lead solution it gives a very faint
blackening of the solution, though the material itself as it passes
into solution turns black. The fresh material is soluble in
dilute salt solutions, insoluble in water. Its precipitation limits
as well as its coagulation temperature were not determined. We
have therefore no guarantee that we are not determined. We
mixture of two or more globulins.

1 Levene and Alsberg: Zeitschr. f. physiol. Chem., xxi.

~

Gok Alsberg and E. DaClack 245

On analysis it yields the following results:

PREPARATION I.

0.2117 gm. gave 3.40 cc. dry N at 768 mm. and 21°: N = 16.85 per cent.
0.2280 gm. gave 0.4240 gm. CO, and 0.1450 gm. H,O: C = 50.72 per cent.
Ee — 7112 per cent.

PREPARATION II.

0.1756 gm. gave 25.05 cc. dry N at 758 mm. and 20°: N = 16.58 per cent.
0.1884 gm. gave 0.3471 gm. CO, and 0.1219 gm. H,O: C = 50.25 per cent.
H = 7.24 per cent.

0.5515 gm. gave 0.0388 gm. BaSO,:S = 0.914 per cent.

(Determination by fusion with sodium peroxide according to Folin.)

PREPARATION III.

0.1965 gm. gave 27.90 cc. dry N at 772mm.and19°: N = 16.86 percent.
0.1886 gm. gave 0.1200 gm. H,O: H = 7.12 per cent (CO, determination
was lost).

0.2143 gm. gave 0.0032 gm. ash: Ash = 1.88 per cent.

As only one ash determination was done, none of the above
analyses are calculated for the ash-free substance. The ashing
was done in platinum and the ash was a fusible one of a slightly
brownish color. ‘Treated with hot hydrochloric acid it did not
apparently dissolve and in the hydrochloric acid extract no
phosphorus and only the merest trace of iron (potassium sulpho-
cyanide) could be discovered. To confirm the absence of phos-
phorus the substance was repeatedly decomposed by boiling
with concentrated nitric acid and this element searched for among
the decomposition products. Finally 0.2 gram were fused in
silver with sodium hydroxide and potassium nitrate, the fusion
mass dissolved, acidified with nitric acid, enough ammonium
nitrate added to make 15 per cent and ammonium molybdate
added. Not a trace of ammonium phosphomolybdate precipi-
tated though the solution was finally concentrated and stood for
many days. Plainly phosphorus is absent in our preparations.
The iron content, judging by the qualitative tests; is so exceed-
ingly slight that it is doubtful whether it is really a component or
an accidental contamination.

To confirm the absence of phosphorus and perhaps clear up the
question as to the occurrence of iron, an attempt was made to
prepare hematogen, in which the iron and phosphorus if present

246 Globulin from Egg Yolk of Squalus Acanthias

ought to be concentrated. About 4 grams of the material were
suspended in 200 cc. of artificial gastric juice and digested at
40° C. for 40 hours. The material went almost completely into
solution and at the end of the digestion only an exceedingly
small amount of sediment was present. Further digestion did
not increaseit. It was collected on an ash-free filter washed with
water and alcohol and ashed. There was less than 1o mg. of
ash which was completely soluble in hydrochloric acid but con-
tained no phosphorus. It did, however, contain a little iron, in
fact considerably more than the originai material. Still even
here the amount was very small. Plainly our globulin differs
from vitellin in not yielding a phosphorus-rich hematogen, and it
is doubtful whether it contains iron in its molecule.

As all our preparations were made from a single batch of eggs,
we can not be sure that some disturbing factors did not affect our
preparations; however, this seems very improbable. Almost cer-
tainly the eggs of Squalus acanthias contain no vitellin in the
strict sense of the word; but in its place a globulin or a mixture of
them, containing no phosphorus and perhaps no iron. It is,
so far as we are aware, the only vertebrate egg in which this has
been found to be the case. It would be interesting to investi-
gate whether there is any relation to the fact that Acanthias
is viviparous. This might easily be done by comparison with
other viviparous and oviparous species of selachians. It would
also be interesting to see whether the acanthias egg contains any
considerable amount of iron, and if so, in what form. A similar
investigation of the phophorus would also be worth while.

CHEMICAL EVIDENCE OF PEPTONIZATION IN RAW AND
PASTEURIZED MILK.

By RACHEL H. COLWELL anp H. C. SHERMAN.

(Contributions from the Havemeyer Laboratories of Columbia University,
No. 157.) ©

(Received for publication, July 13, 1908.)

It is well-known that milk which has been heated to a sufficient
temperature to destroy most of the acid-forming bacteriamay
still contain putrefactive or ‘‘peptonizing”’ organisms and it has
been suggested that this may constitute an objection to the pas-
teurization of milk as a preservative measure since the latter
forms may grow more readily after destruction of the former
and the absence of a sour odor or taste may result in the milk
being used as food after objectionable products have been formed
by the action of these putrefactive or “‘peptonizing’”’ bacteria
upon the milk proteids. On the other hand, it is held by some
bacteriologists that the danger from this source is comparatively
unimportant, the common proteolytic bacteria being largely ren-
dered inert by the heating.

In the belief that the bacterial decomposition of proteins in
milk might perhaps be measured by means of the ammonia pro-
duced, a method for the determination of ammonia was adapted
to milk and a somewhat extended study of the development of
ammonia in raw, chemically preserved, and pasteurized milk
was carried out in this laboratory. In the last and most sys-
tematic of these series of experiments, ten samples representing
the milk sold by six different dealers were taken at intervals dur-
ing a period of two months, February to April, 1907. Each of

1 Berg and Sherman: The Determination of Ammonia in Milk, Journ.
Amer. Chem. Soc., xxvii, p. 124; Sherman, Berg, Cohen and Whitman:
Ammonia in Milk and its Development during Proteolysis under the Influ-
ence of Strong Antiseptics, this Journal, iii, p. 171; Whitman and Sher-
man: The Effect of Pasteurization upon the Development of Ammonia
in Milk, Journ. Amer. Chem. Soc., Xxx, p. 1288.

247

248 Peptonization in Milk

these samples was mixed and divided into three portions, one of
which was kept raw, one pasteurized at 65° C. and one at 85° C.,
after which all were brought to a temperature of 15° to 20° C. and
examined regularly after standing for two, four and seven days.
From the results of these experiments which are given in full
elsewhere!’ it appeared that the pasteurization was less efficient in
checking the development of ammonia than in checking the pro-
duction of acid and this was especially true of the milk pasteur-
ized at the higher temperature (85° C.) which before becoming
sour often showed an amount of ammonia considerably in excess
of that contained in unpasteurized milk of the same age and ori-
gin. It was found, however, that the determination of ammonia
in milk cannot serve as an exact measure of the extent of protein
decomposition in all cases, since in raw milk there was often, and
in milk pasteurized at 65° C. there was sometimes, a decrease in
the ammonia content between the fourth and the seventh days.
In addition to the ammonia determinations each sample was reg-
ularly examined as to odor and taste and it was found that the
samples kept raw always developed within a few days a sour
taste and an odor which was always sour, sometimes “‘musty”’
or ‘“‘cheesy,’’ but never putrid or offensive, while the samples
which had been pasteurized, especially at the higher tempera-
ture, frequentlydeveloped an offensive, putrid odor and a bitter
taste. The occurrence of bitter tastes in the milk pasteurized
at the higher temperatures, together with the observation that
the ammonia content could not serve as a strictly quantitative
means of comparing the decomposition of protein in raw and pas-
teurized milk, suggested the desirability of supplementing the
study of the ammonia content by comparative tests for peptone
in samples kept raw and after pasteurization at different tem-
peratures.
EXPERIMENTAL.

During April and May, 1908, nine samples representing the
milk of six retail dealers were studied. Each sample was brought
to the laboratory as purchased, mixed and divided (in sterile
receptacles with precautions against secondary contamination)
into four portions, one of which was kept raw, the others pas-

1 Whitman and Sherman: Loc. cit.

Rachel H. Colwell and H. C. Sherman 249

teurized by heating for 20 minutes at 60°, 75° and go° C., respec-
tively. The pasteurized samples were then cooled and all were
allowed to stand at room temperature. After two days, and
usually again after four days standing, the odor was observed,
the acidity determined, and a test was made for peptone. In
testing for peptone, the coagulable protein was first removed,
the proteoses precipitated by zinc sulphate and an aliquot part
of the filtrate was treated with sodium hydroxide until the zinc
was precipitated and redissolved and was then submitted to the
biuret test. The volume of milk tested and the quantities of re-
agents used having been kept uniform, it is believed that the com-
parative amounts of peptone present are roughly indicated by the
intensity of the color reactions obtained.

The term “‘acidity’’ here indicates the number of cubic centi-
meters of tenth-normal sodium hydroxide required to neutralize
to cc. of milk using phenolphthalein as indicator.

For convenience in comparing the intensity of the biuret re-
action in the different tests, those in which no reaction was ob-
tained are marked o; those showing a faint reaction, 1; those
with distinct reaction, 2; those with strong reaction, 3.

In order to compare the offensiveness of the odor, the samples
having either no odor or a clean, sour odor are marked 0; those
with a slightly musty or very slightly putrid odor, 1; those
with a musty or slightly putrid odor, 2; those with a distinctly
putrid odor, 3.

We are fully aware that no great significance can be attached
tonumerical expressions of the apparent intensity of a biuret color
and still less to that of an odor, and have used these numerals
only for convenience of comparison and summation.

It is also recognized that the conditions of our experiments
differ from those of commercial pasteurization. On the one
hand the temperature at which and time during which our sam-
ples were allowed to stand after pasteurization may be somewhat
excessive, although the ordinary dealer tends to be less careful of
pasteurized than of raw milk. In so far as our conditions are
unusual in this respect they would tend to give a more pro-
nounced result, but there is no reason to suppose that they would
change it otherwise. On the other hand our conditions are more
favorable than those of commercial pasteurization in that any

250 Peptonization in Milk

subsequent contamination is more effectively precluded. The
work here reported should be only preliminary to a much more
extended study including both chemical and bacteriological ob-
servations on a larger number of samples. Lacking the time
necessary to make such a study we record the results obtained
with the inferences which they suggest in the hope that others
may continue the investigation.

The results obtained from these experiments are tabulated
below.

While the individual samples show wide variations the general
bearing of the results may be summarized as follows:

Pasteurization at 60° appears to have restrained peptoniza-
tion to about the same extent that it restrained souring. It
apparently had no constant effect in rendering the milk either
more or less liable than raw milk to the development of offensive
odors.

Pasteurization at the higher temperatures (75° and go®°) de-
layed souring to a much greater extent and had less restraining
effect upon peptonization. The samples pasteurized at 75° and
go° developed much more offensive odors than those of the same
origin pasteurized at 60° or not at all.

While the development of an offensive putrid odor does not
necessarily run parallel either with the ammonia content or with
the intensity of the biuret reaction, and while it is evidently
incapable of exact measurement it must be regarded as of
considerable sanitary significance, since it indicates the pre-
dominance of an objectionable type of change.

The results of this investigation together with that of the effect
of pasteurization upon the development of ammonia in milk tend
to emphasize from the standpoint of the subsequent chemical
changes the desirability of low temperatures as recommended by
Rosenau and others, in pasteurizing milk when necessary as a
safeguard against infectious diseases, and the objectionableness
of depending upon pasteurization as a preservative measure.

The importance of keeping milk cold and consuming it quickly
are apparently not diminished by its pasteurization even under
conditions so favorable as to preclude subsequent contamination.

OFFENSIVENESS OF ODOR.

INTENSITY OF BIURET REACTION.

ACIDITY.

Rachel H. Colwell and H. C. Sherman
o
= OmArA tN Ont Ks KB HFN TNT OO Loan
a re re N
Cs}
se]
o
Ah dN
i Mee NrAnATANONODCOCONANAON AAO oe)
re
Gs
a
es
=) ONFHOCCOCOCOH OOOH OHON Jo)
3
3 | sooscenconccconnse | «
=
BIL NNFHOCHANACHAHANANANNOCO =
ss re - N
|| 3
Coil
8] °°
Ori e ANANTNOCHNHANNANMANAYD >
rallies 4 4 qi
2 | 3
n
S .
Py lo
= SCHANATNOHNOHAAANAANHOS =)
= = = nN
s
5 HANANAMOHANDHAHAANANAANAA!] &
a mo N
[ox
S| rrr MNHORMONKDNRAO oy
= Se I ee ee es oS (op)
- 3 c = N
le -
Bl | MOMHMOKRMHKROCOnRDAHNOM ron
2 iS ee Ro)
Z| 3 a nN oD
ales |
i]
Ay
SL HMONMNHHHRMOMOWDOOHWD Yo)
SIL NANANARR RH HONK AOoOKRR AH | ©
3 oD = i
& HMMOAHROODODHANAHANHOHOH HS To
Ss ODr~rMODEFNNHAHAOrFArFALr S =)
io To) wD AQ
re
a e
ES AND HDON DSR ~_AAM Hor w FE
am | TR
= TD
s a :
we) ‘o: 4
é S oS 3
is a a
S a et
ae S
nN _

251

+
#
lf -
‘
it 7
t FS
| a’

A PRELIMINARY STUDY OF THE SENSITIZING PORTION OF
EGG-WHITE.

By MARY F. LEACH.
(Western College, Oxford, Ohio.)
(Received for publication, July 23, 1908.)

This work was undertaken to supplement some work recently
published upon the corresponding portion of the cell substance of
the colon bacillus,t and upon the effects of egg-white and its split
products on animals.” For some years past researches have
been in progress in the Hygienic Laboratory of the University
of Michigan upon the bacterial cell, its chemical and physio-
logical nature.’

It has been found that by prolonged treatment of the bacterial
cell substance with an alcoholic solution of caustic soda, a viru-
lent poison is dissolved out by the alcohol, while in the case of the
colon and typhoid bacillus, the part insoluble in alcohol confers
immunity to its homologous bacterium. Subjecting egg-white
to the same process, it is similarly separated into a soluble por-
tion, which though harmless administered per os, is promptly fatal
when injected intraperitoneally or intravenously, and an insoluble
portion which is not poisonous. It is certainly interesting that
a given protein taken into the circulation by way of the ordinary
processes of digestion serves to build up the body of the animal
and to enable it to carry on its vital processes; while that same
protein introduced directly into the circulation can not be thus
utilized. A single injection of egg-white has no apparent effect,
a second dose within a few days is also well borne, but a second
dose after an interval of ten or more days is promptly fatal; thus
showing that the first dose has produced sensitization. Now

1 Leach: On the Chemistry of Bacillus Coli Communis, This Journal,
iii, P. 443, 1907.

2 Vaughan and Wheeler: Journ. of Infect. Diseases, iv, p. 476, 1907.

3 For bibliography, see preceding reference.

253

254 The Sensitizing Portion of Egg-white

the non-poisonous portion of egg-white sensitizes an animal to
egg-white, but not to itself. Thus bacterial cell substance may
be separated into poison and immunizing portion, and egg-white
by the same method yields poison and sensitizing portion.

Material. The whites of fresh eggs were poured into 96 per
cent alcohol, the precipitate repeatedly extracted with alcohol,
dried upon filter paper, pulverized, thoroughly extracted with
ether, again dried and pulverized. After standing in the labora-
tory for some months, the egg-white thus prepared was boiled
with absolute alcohol containing sodium hydroxide according to
the method used for bacterial and other proteins.!. This treat-
ment was repeated three times, experience with other proteins
having shown that this is necessary to remove allthe poison. The
insoluble portion was extracted with alcohol in Soxhlets for
three and a half days to thoroughly wash it free from poison and
other alcohol-soluble substances. After drying and pulveriz-
ing, itis a cream colored powder very similar in appearance to the
egg-white from which it was obtained. The poisonous portion,
obtained from the alcoholic solution as described in the preceding
reference, is a dark brown solid, difficult to dry sufficiently to pul-
verize it. Both poisonous and non-poisonous fractions give the
various protein color reactions, but show striking chemical differ-
ences without further purification. The poisonous portian gives
a very pronounced Millon test, gives no evidence of carbohydrate,
and contains only traces of phosphorus. The non-poisonous frac-
tion gives only a faint Millon reaction, but responds to carbohy-
drate tests, and contains a considerable amount of phosphorus.
These same distinctions have been found in splitting up other
proteins by this method.

This preparation of egg-white is markedly hygroscopic, and the
split products are still more so. However two to four weeks’
drying in an oven not above 103° sufficed to bring them to
constant weight. Two samples were cooled in the same desic-
cator, weighed, returned to the desiccator and weighed again in
the same order. Both had gained more at the second weighing
than they had lost since the preceding day. Onarainy day sam-

1 Wheeler: The Extraction of the Intracellular Toxin of the Colon
Bacillus, Journ. of the Amer. Med. Assoc., xliv, p. 1271, 1905.

Mary F. Leach 255

ples would absorb more water during the process of weighing than
they had lost in the previous twenty-four hours.

Ash, mtrogen, phosphorus, sulphur. For purposes of compari-
son, ash, nitrogen, phosphorus and sulphur were determined in
this preparation of egg-white and the two products obtained
therefrom. All determinations were made in duplicate, and the
figures given are the averages of closely agreeing determinations.
Ash was determined by incinerating to low redness in a platinum
crucible, taking pains to avoid volatilizing inorganic matter.
Since the phosphorus is originally in organic combination, and is
present in the ash as orthophosphate, an amount of PO,isdeducted
corresponding to the amount of phosphorus present, giving results
tabulated as inorganic ash in the second column of the following
table, and these results were used in calculating percentages
“‘ash free.’’ Nitrogen was determined by the Kjeldahl-Gunning
method, and phosphorus by the Neumann method with modifi-
cations. Sulphur determinations were made by fusion with
sodium hydroxide and nitrate, and precipitation by addition of
hot barium chloride drop by drop. Ofcourse the figures obtained
do not correspond with analyses of purified products, such as
ovalbumin, ovimucoid, or even of untreated egg-white; but they
are of more value in tracing the physiological units than if sub-
jected to further treatment. However the results for egg-white
show nitrogen content not far removed from figures for other
preparations, but the sulphur is somewhat higher than has been
found even in ovimucoid (2.2 per cent).? Still it is not unreason-
able to suppose that the long continued action of alcohol and
ether would give rise to changes aside from those incident to the
process of denaturalization. While many observers seem to have
ignored the phosphorus in egg-white, Osborne and Campbell’
found in their four purest preparations of crystallized ovalbumin
amounts of phosphorus varying from 0.126 per cent to 0.131 per
cent. Certain data in regard to the colon bacillus are appended
to the table.

' For further details, see paper previously cited On the Chemistry of
Bacillus Coli Communis.

2M6rner: Zeztschr. f. physiol. Chem., xviii, p. 525, 1893.

* Protein Constituents of Egg-white, Journ. of the Amer. Chem. Soc.,
Xxii, p. 422, 1900.

256 The Sensitizing Portion of Egg-white

TABLE I. PERCENTAGES OF ASH, NITROGEN, PHOSPHORUS AND SULFUR.

| Inor- s

fae.| a | x |e |e el anal

ash. free.
Heo-whitesares oe eit 2.48) 2.066)14.48} 0.135 |2.66|14.7 |0.138/2.73
Poisonous fraction...... ie We! 13.74] traces |2.19|13.9 Deas
Non-poisonous fraction . ./13.57/12.8 |12.67| 0.253 |2.79|14.53/0.29 |3.2

Bacillus coli communis

Gellsubstances. sone 8.61 110.65] 2.87
Poisonous fraction...... 2.33 |11.15) traces
Non-poisonous fraction _. 133 25/26 1088 5256) 25345 4 7.52|3 .99

SENSITIZING PORTION.

Comparison with the corresponding fraction of some other
proteins. Edestin, casein, egg-white and the ceil substance of
Bacillus coli communis were all subjected to the action of alkali
and alcohol as previously described. The insoluble portion of
each, containing the sensitizing or immunizing elements, in all cases
gave the various protein color tests, Millon’s reaction less mark-
edly than the others, perhaps because of the alkalinity of the
preparation. On stirring with water, the edestin preparation
was entirely soluble, there was a slight flocculence with the casein
preparation, the other two were mainly but not entirely dissolved,
but the physiologically active portion of all goes into solution.
Addition of a little sodium hydroxide increases the solubility.
Mineral acids give a precipitate with the casein and egg prepara-
tions.

The most marked difference was found in testing for carbohy-
drates. As edestin contains no carbohydrate, its preparations of
course show no evidence of such groups; however the tests were
applied as controls, always with negative results. Although
casein is said to contain no carbohydrate, it has been found to
respond to the Molisch test? and our preparation from casein does
also. Reactions previously reported point to the presence of
pentose in the bacterial cell preparation.? As was to be expected

1 Mann: Chemistry of the Proteids, p. 403.
*This Journal, iii, p. 453, 1907.

Mary F. Leach 257

the egg preparation gave evidence of hexose and not pentose.
The lead sulfide reaction showed the presence of loosely combined
sulfur in the preparations from egg and from edestin, not in the
ones from casein and the colon germ.

Action of acid. As shown in the preceding paragraphs, quali-
tative tests indicate the presence of protein, nucleo-compounds
and carbohydrates. Mineral acids give a precipitate with aque-
ous solutions. The method of preparation precludes the presence
offffee organic acid, but suggests the possible presence of sodium
salts of such acids. Accordingly 10 grams of the sensitizing por-
tion of egg-white were stirred with 200 cc. of water and filtered.
The filtration was slow, and accompanied by formation of much
foam, but finally a clear golden brown solution was obtained. A
sample of this solution was titrated with 5, sulfuric acid, using
dimethyl-amido-azobenzol as indicator, until red coloration
indicated the presence of free mineral acid. The calculated
amount of sulfuric acid was then added to the original solution.
On standing the precipitate settled as a tough leathery mass,
readily separated from the supernatant liquid by decantation.
The precipitate was washed and dried upon a porous plate. It
was insoluble in both alcohol and ether, and hence is not fatty
acid. It may be added that without further purification it gives
the protein color tests, but awaits additional investigation.

Carbohydrate. Differences between the poisonous and non-
poisonous fractions of a given protein, differences between the
non-poisonous fractions of different proteins, and work upon an
immunizing product obtained from the non-poisonous fraction of
the colon germ, all suggest an investigation of the carbohydrates
present. Accordingly samples of the sensitizing portion of egg-
white were stirred with water, filtered, and attempts made to
separate protein and carbohydrate in the filtrate by means of
uranium acetate. The acetate was added both with and with-
out sufficient alkali to keep the solution alkaline. A copious
precipitate resulted in both cases, which was filtered out with
some difficulty. The slight excess of uranium was removed from
the filtrate by addition of sodium phosphate. The filtrate gave
evidence of carbohydrate, but the separation was not sufficiently
sharp, and the physiological action was lessened, hence that
method was abandoned. Acidifying until there was a slight per-

258 The Sensitizing Portion of Egg-white

manent precipitate, the addition of either methyl or ethyl alcohol
cleared the solution. Phosphotungstic acid precipitated both
protein and carbohydrate. In short no method was found that
would remove the protein from the solution, leaving the carbohy-
drate. It is perhaps a legitimate inference that the two are united.

Samples were subjected to hydrolysis and titrated with Fehling’s
solution. The protein and other compounds present interfered
with the reaction, but by adding the solution all or nearly all at
once it was possible to obtain comparable results. Experiments
with split products from the colon germ gave a maximum reduc-
tion by boiling 24 hours with 2.5 per cent hydrochloric acid.

Three grams of the non-poisonous fraction of egg-white were
mixed with 200 cc. of water and 20 cc. of 25 per cent hydro-
chloric acid. A second sample was prepared in the same way
except that it was filtered before adding the acid. Both were
boiled with reflux condenser. After boiling half an hour, and
then at intervals of three hours, aliquot parts were removed,
neutralized, titrated with Fehling’s solution, and amount of
reducing substance calculated. Other samples were hydrolyzed
with sulfuric acid with less satisfactory results. These prelim-
inary experiments indicated that the reducing substance is all
present in the portion soluble in water, and that the maximum
yield, which if calculated as dextrose is about 9 per cent, is
obtained by boiling 10 to 12 hours until the mixture no longer
gives the biuret test.

Accordingly 25 gram samples of the non-poisonous portion
of egg-white were shaken for two hours in a mechanical shaker
with ten times their weight of water, filtered, 200 cc. more water
added, the solution neutralized with hydrochloric acid, then 50
ce. of 25 per cent hydrochloric acid added, thus making approxi-
mately a 5 per cent solution of material in 2.5 per cent acid.
This was boiled with reflux condenser for ro to 12 hours, until the
solution no longer gave the biuret reaction. It was then filtered,
leaving very little upon the filter. The clear red-brown filtrate
was cooled, neutralized with sodium hydroxide, and benzoylated
by the Schotten-Baumann method. The mixture became very
warm, but was cooled by surrounding the flask with pounded ice
and salt. When the reaction ceased the compound settled nicely
and was filtered with suction after standing two or three hours.

Mary F. Leach 259

The precipitate was washed with water containing a little am-
monia, and treated with boiling alcohol, in which a large portion
was freely soluble. On cooling and concentrating the alcoholic
solution, a fine yield of crystals was obtained. The crystals from
several samples were united and recrystallized from hot absolute
alcohol until the solution was water clear and colorless. Macro-
scopic bundles of needles were thus obtained showing very charac-
teristic grouping. They were washed in alcohol and in ether,
dried upon porous plate, the operations repeated until samples
from two recrystallizations melted side by side within one or one
and a half degrees. The crystals are pure white, readily soluble
in benzol, in chloroform and in glacial acetic acid, as well as in
alcohol, and melt at 203°. When boiled with sodium hydroxide,
ammonia is given off; after removing benzoic acid by boiling with
hydrochloric acid, the resulting solution reduces Fehling’s solu-
tion.

0.4150 gm. gave 0.00891 gm. N, corresponding to
eA2z20gm. “ 0.00962 “ “ ‘
Average = 2.213 per cent N.

.147 per cent N.
dz fo | “ “

bp bo

~ These characteristics suffice to identify the compound as glu-
cosamin benzoate which Pumm! reports as melting at 203°.
Kueny? prepared different benzoates from glucosamin by vary-
ing the conditions of the experiment. The one most readily
formed was tetrabenzoate, melting at 199° when recrystallized
from alcohol, and at 207° when recrystallized from glacial acetic
acid. He tried by various methods to prepare a pentabenzoate,
but without success. Langstein* prepared glucosamin benzoate
from egg-white, which after once recrystallizing from hot alcohol
melted at 201° to 202°, and gave 1.95 per cent of nitrogen. The
theoretical amount of nitrogen in the tetrabenzoate is 2.35 per
cent. Thus this benzoate prepared from the sensitizing fraction
of egg-white agrees with glucosamin benzoate prepared from glu-
cosamin and from egg-white at least as well as those prepara-
tions do with each other. Numerous observers have found glucos-

1 Monatsh. f. Chem., xii, p. 435-

2 Zeitsch. f. physiol. Chem., xiv, p. 355, 1890.

’Ueber die gerinnbaren Stoffe des Eiklars, Beitr. z. chem. Physiol. u.
Pathol., i, p. 83, 1902.

260 The Sensitizing Portion of Egg-white

amin in egg-white, and this work shows that it has remained in
the sensitizing portion. It is hoped later to investigate the carbo-
hydrate in the corresponding fraction of the colon bacillus.

SUMMARY.

Recent articles have described a method for separating bacterial
cell substance into a poisonous fraction and an immunizing frac-
tion. Applying the same method to egg-white, it is separated
into a poisonous fraction and a sensitizing fraction. The sensitiz-
ing fraction of egg-white contains about the same per cent of
nitrogen as does the egg-white, and a little higher percentage of
sulfur. Evidence of protein, nucleo-compounds and carbohy-
drate were found; mineral acid gives a precipitate with aqueous
solutions, but there was no evidence of fatty acid. Phosphorus,
frequently overlooked in analyses of egg-white, remains in the
sensitizing fraction, while only traces are found in the poisonous
fraction. The carbohydrate and nucleo-compounds of the egg-
white remain in the sensitizing fraction, the poisonous portion
giving no evidence of such groups. Carbohydrates and pro-
tein are apparently united, as different methods of separation
fail. After long hydrolysis glucosamin was separated as benzo-
ate. As glucosamin is present in egg-white, it may be inferred
that the carbohydrate passes unchanged from the egg into the
sensitizing portion.

The experimental work herein described was carried on in the
Hygienic Laboratory of the University of Michigan, and I wish
to express my indebtedness to Dr. V. C. Vaughan for turning over
this problem to me, and for his kindly interest and helpful sugges-
tions during the progress of the work.

MILK PROTEINS.

By GEORGE A. OLSON.

(Contribution from the Agricultural Chemical Laboratory, University of
Wisconsin.)

(Received for publication, July 14, 1908.)

Our present knowledge concerning the protein components
of milk is limited to casein, lactalbumin,’ lactoglobulin,? and
probably mucoid? and opalesin.* A number of components, such
as fibrin,® galactozymase,® albumose,’ gélatine,® galactin,® leci-
thin! and lactoprotein," have likewise been found. Aside from
the proteins mentioned above the writer has isolated a new pro-
tein from milk which is characteristically different from any of
those alluded to.

Casein. When skim milk is treated with dilute acetic acid or
on saturation with a salt like sodium chloride a coagulum is
formed, which when removed is made up largely of casein, cal-
cium phosphate, water and some fat. On repeatedly dissolving
this coagulum with dilute alkali, precipitating with dilute acid
and washing the precipitate thus formed with alcohol, ether and
alcohol, a product free from calcium phoshate and fat is finally
obtained which is known as casein. Casein is a white amorphous
body, insoluble in alcohol, ether or dilute acids, but on treatment
with alkalies, alkaline carbonates, phosphates or strong acids it

1J. Sebelien: Zeitschr. f. physiol. Chem., ix, p. 445, 1885.

2 J. Sebelien: Ibid.; Journ. of Physiol., xii, p. 95, 1891.

$V. Storch: Monatsh. f. Chem., xviii, p. 244, 1897; XX, p. 837, 1899.
* A. Wroblewski: Zeitschr. f. physiol. Chem., xxvi, p. 308, 1898.
5S. M. Babcock: Soc. for the Prom. of Agric. Sct., p. 13, 1898.

$A. Béchamp: Compt. rend. de lV Acad. d. sct., 1xxvi, p. 836, 1873.
™M. R. Palm: Jahresbericht —. Thierchem., xvi, p. 143, 1887.

8 Morin: Journ. de phys. et de chim., xxv, pp. 428, 431 and 433, 1854.
®Selmi: Gazetta chim. ttal., iv, p. 482.

1° Koch: Zeitschr. f. physiol. Chem., pp. 47, 327-330, 1906.

4 Millon and Comaille: Jahresber. d. Chem., p. 622, 1864.

261

262 Milk Proteins

readily passes into solution. Casein unites with alkaline earths
and particularly with calcium phosphate for the association
which is present in milk. Rennet splits casein into two proteins
one of which (paracasein) in the presence of calcium salts is coagu-
lated forming with the entangled fat of the milk the curd which
ultimately becomes the commerical product known as cheese.
The other protein, a soluble albumose body (whey albumin)
remains in solution. According to Lehmann and Hempel! the
composition of casein is as follows:

(Cp eee Jal, 7/02 ING lGuGe Se Oeil IPs OMeviye OBL le

Lactoglobulin. A few milligrams of lactoglobulin per liter of
milk can be obtained by saturating the filtrate resulting from
the removal of casein by means of sodium chloride with magne-
sium sulfate. In a moist condition lactoglobulin is neither tough
nor elastic and possesses a snow white flaky appearance. Little
is known in regard to this protein component of milk.

Lactalbumin. On further saturating the lactoglobulin filtrate
with ammonium sulfate, Sebelien obtained a protein, lactalbu-
min, which comprises about one-tenth of the entire protein com-
ponents of milk. Lactalbumin when dry is a transparent,
brittle, hygroscopic mass or a white powder. It is precipitated
by hydrochloric acid but dissolves in an excess of the same.
Lactalbumin has the following composition:

Ch S219. Hy 7ALS3s NI eS ele 1a; One

It differs in composition from casein in that it has no phos-
phorus, twice as much sulfur, less carbon and more oxygen.

Opalesin. Wroblewski has observed an albumin in milk,
which is rendered opalescent by the addition of acetic acid. This
albumin which he termed opalesin can be salted out with sodium
chloride or magnesium sulfate. It is soluble in water and non-
coagulable. It gives the color test of albumins, also those not
given by casein and is further remarkable for its low carbon and
high sulfur content. The composition of opalesin is as follows:

C, 45:0; '.H,, 7:3) N, 15.07; POs Ss: 4974/0) ae

1°W. Hempel: Arch. f. d. ges. Physiol., lvi, p. 558, 1894.

George A. Olson 263

Mucoid. Storch isolated from fat- and salt-free butter serum
a protein which after air drying formed a fine loose, hygroscopic
pOWier of grayish white appearance. It is insoluble in water,
alcohol, ether and dilute mineral acids. It dissolves in weak and
strong sodium hydroxide with difficulty. On treatment with
dilute hydrochloric acid this mucoid reduced Fehling’s solution.
The ash-free mucoid contained 14.2 per cent nitrogen and 2.2
per cent sulfur.

New protein. The new protein was first isolated from centri-
fugal slimes, which were obtained from the inner walls of the
bowls of cream separators. Water-free, centrifugal slimes are
composed largely of proteins, among which casein is by far the
most abundant constituent. The following method was used to
isolate the new protein from the slime.

Fresh slime was triturated in the cold with large quantities of
0.5 to 1.0 per cent sodium hydroxide solution. The alkaline ex-
tracts thus obtained were filtered and then cautiously treated
with 1 per cent hydrochloric acid until precipitation commenced.
This precipitate first formed was removed and the filtrate ob-
tained was treated with more hydrochloric acid and the process
continued until all of the acid precipitates were removed. When
no further precipitation occurred with the acid the process was
reversed by adding enough 0.5 per cent sodium hydroxide solu-
tion to the acid filtrate until slightly alkaline, removing any pre-
cipitate that formed during the process. To every five parts of
the slightly alkaline filtrate one part of concentrated hydro-
chloric acid (sp. gr., 1.20) was added and the entire contents were
allowed to stand until the precipitation of the new protein had
completely settled to the bottom. This requires from twenty-
four to forty-eight hours. The supernatant liquid was then de-
canted and the remainder filtered. The precipitate was purified
by repeatedly dissolving in water and precipitating with concen-
trated hydrochloric acid, after which it was finally dialyzed,
either in parchment tubes or dialyzers, washed in alcohol, ether
and finally dried over sulfuric acid. Approximately 1.5 gram
of this precipitate was obtained from about five kilos of slime.
Sufficient chloroform was used to inhibit bacterial action in the
preparation of the protein from slime.

New protein found in milk, cream and butter. The presence of

264 Milk Proteins

this new protein in whole and skim milk, cream and butter was
also confirmed. The casein in the milk and cream was pre-
cipitated in the cold by weak acetic acid and removed. The
albumin and calcium phosphate were precipitated from the acetic
acid filtrate with weak sodium hydroxide solutions and removed.
The resulting albumin filtrate was then treated with concen-
trated hydrochloric acid (sp. gr., 1.20) and allowed to stand over
night for the precipitate to settle. Cream appeared to contain
a much larger amount than milk. In working with butter, the
fat was removed with gasoline and the resulting residue was
treated as in the case of the centrifugal slime with positive
results.

Physical and chemical characteristics. The new protein when
desiccated has a brown, varnish-like luster. When dried in the
water oven at 95° C. its color does not change. When pulverized
it appears as a yellowish white powder. In water it swells and
becomes white in appearance. The protein dissolves in weak
sodium hydroxide solutions and is precipitated from such solu-
tions by phosphotungstic acid, tannic acid, ferrocyanide of pot-
ash, Millon’s reagent, concentrated hydrochloric acid, etc., but
not with acetic acid. When the protein is added toconcentrated
sulfuric acid (sp. gr., 1.84) it takes on the white appearance in the
cold, but on heating it rapidly chars. It is insoluble in ether or
alcohol and gives the biuret reaction. It does not reduce Feh-
ling’s solution. Hydrogen peroxide is decomposed by a solution
of the protein dissolved in sodium hydroxide solution. Different
preparations have been made which before being dialyzed tested
from 14, 14.57, 14.28 per cent of nitrogen and after dialyzation
tested from 18.43, 18.62, 18.70, 18.81 and 18.93 per cent of nitro-
gen. The dialyzed protein which contained 18.93 per cent of
nitrogen also contained 1.53 per cent of sulfur and 0.811 per cent
of ash. The high nitrogen content, together with the other men-
tioned facts, indicates that this compound is a protein of high
complexity. What is more remarkable, the protein when added
to sterile milk produces proteolysis.

George A. Olson 265

MILK ENZYMES.

Among the numerous enzymes which are claimed to be natural
constituents of milk may be mentioned galactase,’ trypsin,’
amylase,* lipase,‘ lactokinase,’ peroxidase and catalase.®

In this paper only proteolytic ferments, such as galactase,
trypsin, pepsin or erepsin, are considered, since the specific func-
tion (digestion) of any of these has a direct bearing on the work
presented in the following pages.

Three distinct methods have been employed in studying pro-
teolysis, namely, the digestive action on milk, the digestive action
on milk agar cubes and tests to determine the optimum tempera-
ture for its action.

Proteolytic tests with the new protein. Twenty series of experi-
ments were conducted with fresh preparations of the new protein
in order to learn the nature of the same with reference to pro-
teolysis.

In one series of experiments eleven bottles containing 180 cc.
of freshly skimmed milk in each were boiled, cooled, and treated
with 5 cc. of chloroform per bottle. Two were kept for control
and three lots of three bottles each were treated as follows: In
every lot one bottle was made neutral, another bottle, after it
had been neutralized, was made acid to the amount of 0.2 per
cent of hydrochloric acid and the remaining bottle was made
alkaline to the amount of 0.2 per cent of sodium carbonate. In
Lots I and III two preparations of the new protein from slime
were added. In Lot I] a hydrochloric acid precipitate from whey
was added. All bottles were made up to equal volumes with ster-
ilized water, and allowed to digest for periods 131 and of 1099
days at room temperature. The bottles were frequently shaken
during the earlier part of the digestion period. The total amount
of nitrogen in this milk was 0.473 per cent, having a soluble nitro-
gen content of 0.0283 per cent or equal to 6.026 per cent of the

1S. M. Babcock and H. L. Russell: rgth Ann. Rpt. Wis., p. 179, 1897,
15th Ann. Rpt. Wis., p. 85,1898, 16th Ann. Rpt. Wis., p. 157, 1899.

* Moro: Jahrbuch. f. Kinderheilkunde, N. F., lvi.

3 Béchamp: Compt. rend. de l’ Acad. des. sci., xcvi, pp. 1508-1509.

* Marfan and Gillet: La presse medicale, pp. 13-16, got.

> Hougardy: Bull. acad. roy. belg., pp. 888—-goo, 1906.

® Wender: Oesterr. Chem. Zeit., vi, p. 13.

266

Milk Proteins

total nitrogen of the milk. All bottles were strong with chioro-
form at the close of the experiment.

TABLEIL.

The results are as follows:

Proteolytic test of the new protein on milk.

New protein Lot I.

| 131 DAYS DIGESTION. | 1099 DAYS DIGESTION.
|
Per cent Per cent Per cent Per cent
scluble total soluble total
nitrogen. nitrogen. nitrogen. nitrogen.
Controlnnsiece seco 0.0285 6.026 0.062 13.13
INGuibrall Porn arae aes serene 0.0473 9.990 0.340 71.88
Oi 0.2 percentrner eres 0.0550 11.620 0.109 23.04
Na: COls0s2 pericentree.. cae. a 0.0547 | 11.560 0.077 16.28
Hydrochloric acid precipitate from whey Lot II.

: | |
Control ie cae tenes 0.0285 | 6.026 | 0.062 | 13.18
INewtral ss :c esas see earn ine | 0.0360 | 7.610 | 0.058 12.27
HIGH 02 perectite..W.n eee 2: 0.0295 | 6.240 | 0.061 12.89
Nas CO} O22) pericent. 22. a. - = 0.0495 | 10.450 | 0.078 16.39

|

New protein Lot IIT. .
Cprigral sis 0h aos amin Ware | 0.0285 | 6.026 | 0.062 | 13.13
INewitinall Sh mtrey se teac nie eis | 0.1275 26.950 0.297 62.79
HCl, 0.2 per cent .............| 0.1225 | 25.900 | 0.234 | 49.47
| 25.790 | 0.171 | 36.15

Nae COs Ok) per Celine ete | 0.1220

From the data obtained in the above series of experiments it
will be noted that at 131 days the largest increase of soluble nitro-
gen was found in Lot III. While at 1ogg days, including the
increased soluble nitrogen in the control, it will be observed that
the greatest amount of soluble nitrogen was found in the neutral
milks and the least in the milks containing sodium carbonate,
Lots I and III. It is probable that the increase in soluble nitro-
gen in the milks made alkaline with sodium carbonate was partly
due to the alkali present.

The most important fact developed in this and similar experi- —
ments was that the hydrochloric acid precipitates obtained from

Y

George A. Olson 267

whey, which were treated under similar conditions as the precipi-
tates obtained from slimes, did not show any perceptible digestive
action, indicating that all external factors that might have been
introduced in these preparations and caused proteolysis were
removed. Whatever the changes are that took place in Lot II,
they are undoubtedly due to some other causes than proteolysis.
There is a similarity in the amount of digestion in Lots I and III
for 1099 days and in similar experiments where the period was
carried on for 1536 days.

The influence of acid and alkali on proteolytic action. Accord-
ing to Nasse’ and Schmiedeberg? albumins denaturalize with
great rapidity when brought in contact with alkalies and acids.

For this reason it was desired to learn to what extent the acid
and alkali to the amount used influenced an increase in the solu-
ble nitrogen and whether or not this power of dissolution was
progressive. In this experiment 180 cc. of neutralized and
boiled skim milk was added to each bottle. The bottles and
contents were then reboiled, cooled and 5 cc. of chloroform was
added to each bottle. Two bottles were kept for control, five
were made alkaline with sodium carbonate to equal o.2 per
cent solution, and two bottles were made alkaline with suf-
ficient sodium carbonate to make a o.4 per cent solution. Two
bottles were kept neutral and four were made acid with hydro-
chloric acid to equal a 0.2 percent solution. A hydrochloric acid
precipitate from slime was added to two bottles of milk contain-
ing 0.2 per cent of sodium carbonate, two bottles containing 0.2
per cent hydrochloric acid and two bottles of neutral milk. All
bottles were made up to equal volumes and then allowed to digest.
The total nitrogen in the skim milk was 0.49 per cent with 0.038
per cent soluble nitrogen or equal to 7.75 per cent of the total
nitrogen. The results of this experiment are as follows:

10. Nasse: Arch. f. d. ges. Physiol., vii, p. 139, 1872.
70. Schmiedeberg: Arch. f. exp. Path. u. Pharm, xxxix, p. 1, 1897.

268 Milk Proteins

TABLE II.

Influence of acid and alkali on boiled milk.

PER CENT SOLUBLE PER CENT TOTAL
NITROGEN. NITROGEN.
108days | 192days | 108 days | 192 days
digestien. | digestion. | digestion. digestion.
Controls s sistas ee ee 0.038 | 7.75
HCl On2spericent ae ase er |} 0.040 | 8.16
HCL Os2.percentecase a pea | 0.0435 | 8.88
Na, CO;,.0.2 percent... betes | 020405) 4 8.26
NanCO)022.percenteoea.ceeor 0.0410 | S23%
Na, CO;, 0.4 per cent............} 0.0410 | 8.37
Na, CO;) 022 per Genta. cose =i 0.0460 | | 9.40

It will be seen from the above data that neither the hydro-
chloric acid nor sodium carbonate altered the composition of the
milk materially. The above results further show that neither
the action of hydrochloric acid nor sodium carbonate are pro-
gressive.

TABLE III.

Influence of new protein on same milk as used in Table II.

108 DAYS DIGESTION. , 1050 DAYS DIGESTION.

| =,
Per cent Per cent Per cent Per cent
nitrogen. | nitrogen. | nitrogen. nitrogen.

|
soluble | total | soluble | total
|

Controlehe ete vn ae eae ee i OF038 715 VOROTS 15.92
Weitbrallize: ta: 2 as ate oes oat) apo: 25.32 | 0.286 58.37
HCW O;27per-cent o..0 45. ee MOM: a, Darke | 0.150 | 30.61
Na; CO; OF2 pericente-cee--e aac) 1On LS 24.08 | 0.107 | 21.84

From the data given in the preceding table it will be seen that
the digestive action is similar to that found in Table I, p. 266.
Comparing the above data with those found in Table II, p. 268,
for the 108 day period it will be seen that the new protein has
exercised a pronounced digestive action, while the milk brought
in contact with the acid or alkali was practically the same as in
the beginning of the experiment as far as the soluble nitrogen is
concerned.

a ae oT

George A. Olson 269

The new protein in relation to its physiological condition. ‘The
protein was heated at different temperatures for periods of ten
minutes each. All were cooled to uniform temperature and
added to tubes containing milk agar prisms.1. They were then
allowed to stand for three days, observing the condition of the
cubes at twenty-four hour periods. In the tubes where digestion
took place the action commenced on the outer surface of the cube,
dissolving the casein of the milk suspended in the agar inwardly.
In a similar manner the filtrate obtained from the separation of
a hydrochloric acid precipitate from slime was studied.

TABLE IV.

Physiological tests with new protein and filtrates from which it was
removed.

New protein. Hydrochloric acid

| filtrate.
Heated ab 35° OC... 0.25.6. ...005.2.% digestion |
beatedati40° Coe cece es vanes oes digestion
Pleated a6 45° C.... 25... ee ees digestion
enema OOO... ee crs ac digestion
Peerage Oo. 2 e d seed | digestion
Piece ab GOO.) 5. oes eee | digestion
Bee PeOnaNO De Cons Anse lsd a, sae aks | most marked | digestion
| digestion
= Ect C028 Ce digestion
Bteeiiedrat SU" CO... sk ee es | none. none.
isl Qniech ah estan Opp | none.
Seemeeet LOO? Oro. ce oes ek none. none.
“LA, See | none. none.

It is apparent from the results of the physiological tests given
above that the new protein, as well as the filtrate from the same,
contains at least one enzyme.

Proteolytic test with filtrate obtained from new protein. In the
preceding experiment it was learned that the filtrates from the
new protein had the function of proteolysis. Owing to the strong
hydrochloric acid which was present in the filtrate it was neces-
sary to neutralize the same before adding to a lot of boiled milk
which was treated under the same conditions as in previously

1E. G. Hastings: 21st Ann. Rpt. Wis., p. 170, 1904.

270 Milk Proteins

described experiments. The total nitrogen of the milk was 0.56
per cent, containing 0.063 per cent soluble nitrogen. The results
of a thirty day incubation at 39° C. are as follows:

TABLE V.

Proteolytic test of hydrochloric acid filtrate.

| 30 DAY DIGESTION aT 39°C.

| Per cent soluble orca

| nitrogen. nitrogen.
Waittrol aes: Re eee eee ae / 0.063 11.25
INfe an trest 8 ies, ech eraekan hae Winn Oe aie 2 OM on a | 0.102 18-21
FICY OMI per Gente eae setae ted ee Me onl ase eae | 0.1015 18.12

ir nuOlOy DL nee ben pers | 0.1015 18.12

It will be seen from the above data that hydrochloric acid does
not completely separate from milk all of the enzyme or enzymes
capable of proteolysis.

Character of decomposition products formed. It has been shown
from the results given in the preceding pages that whenever the
new protein was added to milk digestion took place. In the
following pages the decomposition products formed by the action
of the new protein on milk are compared with those formed by
animal ferments and bacteria. The analytical methods followed
are the same as those used by Vivian! while studying the soluble
nitro en compounds formed by galactase.

Influence of sterilization and chemicals. In the following table
is given the per cents of the various soluble nitrogen compounds
present in fresh skim milk, both before and after sterilization,
together with the results of the influence of acid and alkali on
this milk for periods of 93 and 224 days, respectively. For the
study of the acid and alkali influence each bottle containing 200
cc. of sterilized milk plus 5 cc. of chloroform, plus either 1o cc. of
* hydrochloric acid and ro cc. of normal sodium hydroxide, or
to cc. of water. After corking and sealing, the bottles were
placed in the incubator and kept at a temperature of 39.5° C. for
the above stated periods with the following results.

1A. Vivian: 16th Ann. Rpt. Wis., p. 179, 1899.

George A. Olson 271

TABLE VI.

Soluble nitrogen compounds in milk as influenced by sterilization and

chemicals.
| | PER CENT OF TOTAL NITROGEN EXPRESSED
| IN THE FORM OF-——
| g Peptones. |
eas ww bs |e
co} S| os cas | o fe)
A hist = ae | < §
Before sterilization.............. | 2.01 | 1.24 | 4.29
After sterilization...............| | 4.43 23 | 0.40 | 4.16 | 0.00
Ajter sterilization: |
eer so! OS | 202"). 1.44. | 0.42) | 3.485 eOn0n
Hydrochloric acid.......... | 93 | 1.48 | 1.88 | 0.30 | 4.14 | 0.60
Sodium hydroxide.......... 93 | 2.78 | 2.25 | 0.48 | 3.23 | 0.75
ere... | 224)| 0.50) 3.41 | O74 |) eee coer
Hydrochloric acid.......... | 224 | 0.45 | 2.89 | 0.72 | 3.48 | 1.08
Sodium hydroxide..........| 224 | 1.45 | 3.30 | 0.72 | 3.13 | 1.08

The most evident feature brought out by this table is the de-
velopment of ammonia in milk containing acid and alkali at the
93 day and also in the watered milk at the 224 day period. While
there was a fluctuation in the albumose content, a decided
increase in anti and hemi-peptones was observed in the milk
containing acid and alkali. The failure to obtain peptones by
phosphotungstic acid before sterilization indicates that a mod1-
fication had taken place during sterilization.

Influence of proteolytic bacteria. According to Effront* bac-
terial spores can retain their secreting power although their ability
to germinate has been destroyed. In accordance with this view
a series of the same milk under similar conditions as the preceding
experiment were inoculated with the Bacillus subtilis and allowed
to digest for 224 days at 39.5° C. with the following results.

1 J. Effront: Monit. Sct., 4 Ser., xxi, No. 782, p. 81, 1907.

2712 Milk Proteins

TABLE VII.

Influence of Bacillus subtilis in the presence of chloroform.

PER CENT OF TOTAL NITROGEN EXPRESSED
IN THE FORM OF—
Peptones. |

: | Ls | g

fo) li eae rae Ss

| 22 | Bee F

2 | Bo 9) ae g

4 |) 2 | 2 <
Control (amphoteric) .................| 0.50 | 3.41 | 0.74 | 1.45 | 0.91
Control (amphoteric) inoculated.......| 0.56 5.98 | 3.20 13.93 | 0.90
Hydrochloric acid.. Sige, ete | O45 212980") ane 3.48 | 1.08
Hydrochloric acid eed aed ae A Re |.O.81 |.5.627\ 267 8.24 | 0.90
Sodium hydroxides. 2. esee 5. nace ae: |. 1.45 | 3.30 | 0:72 | (Seis eee
Sodium hydroxide inoculated. -....... | 0.72 | 4.40 | 1.26 8.53 | 0.90

In this case it is evident that in spite of the chloroform present,
modifications took place in the presence of Bacillus subtilis, be-
ing most marked in the amphoteric milk. With the exception
of the albumose and ammonia contents marked increases in solu-
ble nitrogen compounds were observed in all separations from
the inoculated milk.

TABLE VIII.

Decomposition products formed in the presence of the new protein.

PER CENT TOTAL NITROGEN EXPRESSED IN
THE FORM OF—

Peptones.

: | as | 3

¢ | a3 | oo | oe

2 oe eors EB 5

2 a) oak et £ S|

< = By < <
Controls sect 20k bia eee 1.69| 4.86| 0.85| 4.99 | 0.74
Neutral. S226 Ms ae ie Bena te eee 4.02 | 14.16 | 12.05 41.21 | 0.44
Hydrochlozie seid: ..i-..¢2)..5.08:->| 2750 8.67 | 2.32 | 9.15 (Opme
Sodium. bydroxide.. ores eee 2.32 | 5.49 | 1.90 | 6.44 | 0.13

It will be seen from the above data that the nature of the de-
composition products are not due to pepsin digestion, owing to

ee a eo ee

George A. Olson 272

the comparatively small amounts of albumose present. With the
exception of the decomposition products in acid, this digestion
agrees fairly well with trypsin which dissociates the albumins
into peptones, peptids and finally into amino acids and according
to Kutscher,! Hirschler,? and Stadelmann,? into ammonia. It
can not be erepsin since this ferment is most active in alkaline
solution. The decomposition products closely resemble those
formed by bacterial enzymes in the presence of chloroform,
which in turn are similar to those obtained with galactase.

Relation of new protein to galactase and bacterial enzymes. In
the following table the nature of the decomposition products formed
from the action of the new protein and Bacillus subtilis in the
presence of chloroform are compared with Babcock and Russell’s
results on cheese, galactase, and bacteria and their enzymes in
the absence of chloroform, which have been recalculated by the
writer on the basis of per cent of total nitrogen. The galactase
which these men used in studying the nature of the decomposition
products was an extract of fresh centrifugal slime added to steril-
ized skim milk.

From the data given in the table (p. 274) it will be noted that
the decomposition products formed in cheese, by galactase, new
protein, Bacillus subtilis in the presence of chloroform, and B.
299 at the 28th day are similar. The most marked difference in
the decomposition products formed is in the transformations of
the more complex albumose and peptones to the formation of
less complex end products as amides and ammonia. In regard
to galactase digestion the results further indicate that the meta-
bolic processes ceased somewhere between the 28th and the 56th
day (4.92 to 6.56 per cent of ammonia) owing to the repressing
action of the chloroform present, while on the other hand in the
absence of chloroform the metabolic processes were allowed to go
on with Bacillus subtilis, B. 299, and B. 83, increasing in these
cases in ammonia from the 28th to the 112th day. The remark-
able, close agreement of B. 299 at the 28th day with galactase at
112th day, and B. 299 with B. 83 at the 112th day, are of note-

1F. Kutscher: Marburger Habilitationschrift, Strassburg, 1899.
2A. Hirschler: Zeitschr. f. physiol. Chem., xxxi, p. 165, 1900.
3E. Stadelmann: Zeitschr. f. Biol., xxiv, p. 261, 1888.

274 Milk Proteins

TABLE IX.

Comparison of decomposition products.

F PER CENT OF TOTAL Nee Ge EXPRESSED
IN THE FORM OF——
°
% ; Peptones. =
= 3 Tannic |Phospho-| Amide. g
a = acid. |tungstic. |
A <a <
CHEESE ounce sitree eT ey ces lake 4.47 2.52 9.49 | 17.04 4.75
Galactases ie cain peeomen 28 | 16.39 O83" | se es aloe an 4.92
LG Wl Sacre Vacientec tata | 56 | 19.67] 13.11 | 9.83] 16.40] 6.56
ES Ln git a Hk 8 RR Rie ee LO Ty 9.84 | 18.03 | 24.59 | 16.40 6.56
New protein (a)............| 1826 3.93 | 18.75 | E1079: || 22R59R Sees
INewaproteimn(b) seer acess 1096 A402) 14. T6512 (05) ae 0.44
Bacillus subtilis in presence
onchloroformi nse een eee 0.56 5.98 3.200 eeds 0.90
Bacillus subtilis I in absence |
OMmehloOroronnms aac | 56 3.63 | 45.45 SiGoil eevee
Bacillus subtilis Il in ab-
sence of chloroform........| 56 5.45 | 45.45 1.82 | 20.91
Bacillus subtilis I in absence |
Olachlororonmeusees ree ee eel a2 3.63 | 38.88 1.82 | 37.04
Bacillus subtilis IL in ab-
sence of chloroform....... .| 1 TO Zon 27a 0.00 | 39.38
B. 299 in abs. of chloroform.| 28 | 10.00 | 17.50 | 10.00 | 47.50 7.50
HY rf 56 0.00 2:25 | 15.00 | 60.00 | 15.00
of . 112} 0.00'| O00 | 7.25 |\{5Seamaeaees
B. 88 in abs. of chloroform. 28 0.00 7.32 | 12.19 | 65.85 | 12.19
er i 56 0.00 0.00,| 14.63 | 53.66 | 56.13
af 6 ii 0.00 0.00 7.35 | S613) 30nee

worthy consideration. The new protein (a) contained one spore
former per cubic centimeter of milk at the 1826th day. From
these deductions it would follow that the decomposition products
formed in cheese, by the new protein, Baczllus subtilis in the pres-
ence of chloroform, galactase, B. 299 at the 28th day, which is
nearly quantitatively the same as B. 83 at the 112th day, are col-
lectively speaking the results of bacterial enzymes with the pos-
sibilities of more or less metabolic changes.

That galactase is not inherent, but of bacterial origin is con-
trary to the views held by Babcock and Russell and in conformity

———"

George A. Olson 275

with the views held by earlier investigators, such as Duclaux,
Adametz, Weigmann and others who have concluded that bac-
terial cultures with the aid of their enzymes (casease, Duclaux)
can digest casein and form cheese-like odors which are in some
cases better flavored than those made under ordinary factory
conditions. Adametz and Winkler even went so far as to place
upon the market pure cultures under the name of tyrogene! for
the purpose of ripening cheese.

The conflicting results obtained by Freudenreich and in earlier
researches by Babcock and Russell were due to the fact that
they believed lactic acid organisms or lactic acid played the im-
portant role in the ripening of cheese. Chodat and Hoffman-
Bang,” however, have clearly shown that lactic acid organisms
do not digest coagulated casein even in the absence of sugar.

Contrary to the earlier accepted biological views that digestive
bacteria are fundamentally important in the process of cheese
ripening, Babcock and Russell followed the generally accepted
idea that all living cells secrete enzymes and assumed that this
phenomenon took place in the milk glands and ascribed the pro-
teolytic changes in milk to this source and named the ferment
characteristic in the changes produced in milk and cheese galac-
tase. To support their theory they treated milk with different
protoplasmic poisons, suchas chloroform, ether, benzol and toluol,
assuming that these agents were destructive to bacterial life. In
spite of these agents the milk always curdled and finally digestion
resulted. It was found, however, that these milks were not
sterile ‘‘as cultures made in various culture media after the anti-
septic had been removed quite often showed bacterial growth.”
Van Slyke, Harding and Hart’ in Tables I, II and III of their
bulletin in a study on this problem, have clearly demonstrated
that the presence of 2.5 to 30 per cent of chloroform had very
little influence upon spore formers, there being 25 to 42 germs
per cubic centimeter for the first 21 days with a decrease from 6 to
14 germs per cubic centimeter at 192 days in spite of the fact
that these samples were shaken daily. They also concluded

1 Molkeret Ztg., xiv, p. 817, 1900.
2 Ann. inst. Pasteur, xv, p. 36, 1901.
° Bull. No. 203, N. Y. Exp. Sta.

276 Milk Proteins

from their experiments that chloroform had very little influence
upon the enzymes present.

Numerous investigations! have been conducted in order to
test the germicidal effect of chloroform. The results of these
investigations clearly show that the vegetative forms of bacteria
are destroyed in the presence of chloroform while the spores are
able to survive for long periods of time.

Owing to the danger of including bacteria from external sources
Babcock and Russell? next endeavored to obtain milk under
aseptic conditions, and after being placed under antiseptics these
milks also showed an increase in the soluble nitrogen content
with each succeeding analysis. That these milks were obtained
under aseptic conditions does not imply that they were sterile
for as they have stated* ‘‘after being exposed to the action of
antiseptics for a time these milks were found to contain bacteria,
but as they were spore bearing forms entirely, the presumption
is that they existed in the milk in a latent condition.’’ Accord-
ing to Van Slyke, Harding and Hart* only one sample of milk
was obtained by them which was germ-free and this one was not
studied for proteolytic changes. Boekhout and De Vries’ drew
200 samples of milk containing 15 cc. in each from one cow and
120 samples of milk containing 15 cc. in each from another. It
is a noteworthy fact that out of the 320 samples of milk obtained
not one of them was germ free. Some of the bacteria present
were able to coagulate the milk ina longer or shorter time, while
others were able to digest casein. These samples of milk, to-
gether with those obtained at the New York Experiment Station
were drawn under aseptic conditions. Ward* found in nineteen
udders examined ‘‘that the lactiferous ducts harbor bacteria
throughout their whole extent.’’ Granting that sufficient quan-
tities of germ-free milk can be obtained for proteolytic studies,

1M. Kirchner: Zettschr. f. Hyg., viii, pp. 465-489, 1890; A. B. Green:
Centrabl. f—. Bakt., I Abt., Ref., xxxv, p. 804, 1905; Corini: Ibzd., I Abt.,
Ref jXEKVI1; p..47, 1905; A. i. Niland: Arch. Hyg., lvi, p. 361, i

4 Rpt. Was.’ Exp. Sia., p. 174, 1897.

5 [btd., p. 175, 1897.

* Bull. No. 203 N. Y. Exp. Sta.

5 Centralbl. —. Bakt., II Abt., vii, p. 826.

® Bull. No. 178, Cornell Exp. Sta.

George A. Olson 297

the possibilities of bacterial enzymes will always confront us for
(if the writer is not mistaken on this point) no milk has been ob-
tained from the entire quarter of an udder which is absolutely
germ free. Further it is reasonable to believe that a bacterial
enzyme once formed would be distributed throughout the entire
mass.

Many of the bacterial forms present in the udder are of a pro-
teolytic type and in consequence we should expect bacterial
growth, resulting in the production of enzymes before the milk
was ever drawn. At most dairy farms cows are milked at inter-
vals running from ten to twelve hours apart, leaving ample time
for the production of sufficient quantities of bacterial enzymes to
bring about the necessary transformations of the casein in milk.

In Babcock and Russell’s experiments the milk was placed in
2.5 liter bottles, leaving approximately one-half liter air space in
every bottle. The specific gravity of the chlorofrom (1.4) which
was added to the milk, or vice versa, is heavier than milk(sp. gr.,
1.032) and nearly all settles to the bottom. Under this’ condi-
tion the upper strata of milk, together with what milk adheres
to the walls of the container above the milk, contain very little
chloroform, if any, and therefore permit of ocean beds for the
bacteria to grow in. Bacterial growth means the formation of
more enzymes, and if the bottles are shaken at twenty-four hour
intervals an additional amount of enzyme results with each shak-
ing until the quantity of ferment has been increased to an amount
that will hasten the curdling of the milk and digest the casein.
The by-products formed, together with the chloroform present
in the milk, tend to devitalize the spore formers and they finally
become extinct. This is undoubtedly what has taken place in
the investigations conducted by Babcock and Russell, who went
on the assumption that 2 to 3 per cent of chloroform would keep
a milk sterile.!

In reviewing their work on the antiseptic effect of various
substances on milk,? we learn that their means of estimating the
efficiency of various agents is based upon the agent which is most
suitable to allow spore formers to survive and at the same time

*Rpt. Wis. Exp. Sta., p. 103, 1808.
* Ibid., p. 99, 1898.

278 Milk Proteins

be destructive to the lactic acid organisms present, for all milks
which soured with certain antiseptics were discarded for as they
have said! ‘“‘2 to 3 per cent of chloroform was sufficient to keep
the milk from souring.”’

Method of isolating galactase. In several periodicals and text
books issued during the past few years the statement has fre-
quently occurred that Babcock and Russell isolated the enzyme
galactase. Asa matter of fact the enzyme galactase (meaning by
this an inherent enzyme) has never been isolated. In 1901 the
writer, then a student of Dr. Babcock, studied the protein com-
position of centrifugal slimes with a special effort to isolate the
ferments present therein, which resulted in the finding of the new
protein. In the Fourteenth Annual Report, p. 178, 1897, is given
the method used by Babcock and Russell for isolating galactase
and in short is as follows: Equal quantities of slime and 4o per
cent alcohol were worked into a paste and finally the coarser
particles of suspended matter were removed. To the resulting
extract thymol or benzol was added in order to allay any possible
fermentation. This extract, after standing twenty-four hours,
was then concentrated to one-tenth its volume at a temperature
not exceeding 25° C. and finally neutralized with sodium car-
bonate and added to milk with the results of very rapid digestion.”
It is needless to say that every step undertaken in their proce-
dure of isolating galactase, with the single exception of the anti-
septic added, was favorable to bacterial growth to say nothing
of the number of bacteria plus their enzymes previous to such an
undertaking, In the experiments where these concentrated
bacterial extracts plus their enzymes were added to milk we have
data’ which are in accord with the results obtained by the writer
(see Table IX), with the exception that the writer worked with
a specific digestive organism rather than a possible number of
digestive organisms.

Referring to Table IX it will be seen that the decomposition
products formed by Bacillus subtilis and its enzyme in the pres-
ence of chloroform are sharply contrasted from those resulting

1 Rpt. Wis. Exp. Sta., p. 86, 1898.
2Tbid., Pp. T79,, 1807
® Tbid., p. 165, 1899.

George A. Olson 279

from the action of the Bacillus subtilis and its enzyme in the
absence of chloroform, where 76 per cent of the soluble nitrogen
was in the form of peptones (by phosphotungstic acid) and am-
monia. Owing to the repressive action of chloroform upon the
vital processes of the Bacillus subtilis, there are practically, under
this condition, only specific chemical changes which take place,
viz: the digestive action of the bacterial enzyme, while on the
other hand in addition to these changes there are also chemical
changes taking place in the organic body.

Importance of enzymes in cheese ripening. That enzymes play
an important role in the ripening of cheese is admitted. Du-
claux’ as long ago as 1880 advanced a theory believing that
the enzyme casease derived from the Tyrothrix organism (allied
to Bacillus subtilis) brought about the characteristic changes in
the process of cheese ripening. Duclaux, Adametz, Weigmann
and their pupils have shown that pure cultures of organisms with
the aid of their enzymes can digest casein, sometimes resulting in
flavors superior to those found in the usual make of cheese.
Freudenreich found in Emmenthaler the lactic acid organism,
Micrococcus caset liquefaciens. According to Peterson? this
organism is frequently met with and increases rapidly during the
first ten days of curing. Other microérganisms have been found
which play an important réle in the ripening of cheese, viz:
Bacillus casei limburgensis in Backsteinkase, Micrococcus casei
liquefaciens and the Bacillus casei limburgensis combined have a
powerful action in Camembert cheese,? Penicillium album in
Briekase, Penicillium glaucum in Roquefort and a kopfschimmel
in gamelost. Some forms of mold‘ are capable of dissolving a
large part of the casein and according to Rahn’* there is present
in the hard type of cheese a yeast besides lactic acid bacteria.

In earlier researches investigators failed to find digestive organ-
isms in the hard type of cheese. It is now known, that digestive
bacteria are present in cheddar cheese as well as in other forms
and are generally all found in colonies.

1 Journ. Amer. Chem. Soc., p. 436, 1882.
2 Trois Peterson s. Ann. 3 8., 487.
Swepstein: Arch. f. Hyg., xliii, p. 1.
4Tiechert: Milch. Zig., p. 786, 1903.

5 Rahn: Centralbl. f. Bakt., II, xv, p. 786.

280 Milk Proteins

The conformity of results obtained at the Wisconsin and New
York Experiment Stations in experiments with normal and
chloroformed cheese is in favor of bacterial action. The presence
of chloroform in cheese as in the case with milk, represses bac-
terial action, thus practically perventing chemical changes within
the organic body, while the bacterial enzyme is left free to act.

It will be seen from the discussions presented in the preceding
pages that the question rests upon the following thought, i.e.,
whether or not the enzymes present in milk or cheese and carry
on the important réle in the processes of ripening have originated
from an inherent Source or from bacteria, yeast or molds. The
fact that digestion is most marked in milk and cheese in the
presence of bacteria, the finding of digestive bacteria in udders,
the resistance of spores to chloroform, the ability of Bacillus sub-
tilis enzyme to form similar changes in the milk as in the case
of galactase, the failure of Salkowski to find demonstrable evi-
dence of digestion in two samples of normal milk in the presence
of 0.3 per cent of chloroform (undoubtedly due to the absence of
proteolytic enzyme secreting organisms) all indicate that the
proteolytic enyzme or enzymes present in milk have originated
from one source, namely, bacteria.

Since centrifugal slime, milk, cream, etc., generally contain
large numbers of bacteria it may be that the new protein is.a
resultant of bacterial action. The fact that the new protein
digested sterilized milk, as well as the fact that enzymes attach
themselves to materials in suspension (as in the case of the for-
mation of the new protein) together with the fact that the bac-
terial action is certain before, if not during, preparation, thereby
forming large quantities of bacterial enzymes, all tend to indicate
that the characteristic digestion of the new protein is one of
incorporation.

The failure to obtain digestion with precipitates prepared
from cheddar cheese whey indicates one of two things, that either
the ferment is incorporated with the curd or the amount present
in whey was so diluted that what little was present was injured
by the action of strong hydrochloric acid.

BERS (ORF = wen ape ess

George A. Olson 281

CONCLUSIONS.

(1) By the careful removal of the casein and albumin in slime
with weak acid and alkali, a filtrate is obtained in which a new
protein can be isolated by the addition of one part of concen-
trated hydrochloric acid (sp. gr., 1.20) to every five parts of the
filtrate.

(2) This new protein is found in milk, cream and butter.

(3) Chemically the new protein is very rich in nitrogen, con-
taining 18.93 per cent of the same. It gives the biuret reaction
and is dissolved in weak sodium hydroxide solutions.

(4) Physically the new protein possesses a brown varnish-like
luster which when pulverized is changed into a whitish appear-
ance. In water it swells and takes on a whitish appearance.

(5) When added to milk a part of the casein is digested in the
same. Digestion is most favorable in neutral milks on prolonged
standing.

(6) Physiologically the enzymic properties are most active at
a temperature of 65° C. At 80° C. the ferment was destroyed.

(7) The filtrate obtained after the removal of the new protein
also has digestive properties.

(8) The influence of chemicals and sterilization tend to slightly
modify the soluble nitrogen compounds of the milk.

(9) The addition of digestive bacterial cultures to sterilized
milk in the presence of chloroform caused proteolysis.

(10) The decomposition products formed in the presence of the
new protein are similar to those formed in the presence of galac-
tase and bacterial enzymes under the same conditions. From
these facts it is believed that the characteristic digestion of the
new protein and galactase are of bacterial origin.

(11) The enzymic property of the new protein is one of incor-
poration.

lg

THE USE OF THE FERMENTATION TUBE IN INTESTINAL
BACTERIOLOGY.

By C. A. HERTER, M.D., ano A. I. KENDALL, Pu.D.
(Received for publication, July 5, 1908.)

During the progress of a series of investigations bearing upon
the bacterial flora of the intestinal tract of infants and adults in
health and disease, the writers have obtained valuable informa-
tion from the routine use of the fermentation tube, by using
methods similar to those of Dr. Theobald Smith,! who has
repeatedly called attention to the fundamental importance of this
apparatus in the study of fermentative bacteria. His researches
have shown conclusively that the gas volume, gas ratio (propor-
tion of gas soluble in caustic alkali to the insoluble portion) and
the length of time necessary to produce the maximum amount of
gas are characteristics of prime importance in the study of this

group of organisms. Furthermore, by the simple addition of

bits of fresh sterile animal tissue* he has succeeded in cultivating
in these tubes a number of anaérobes which would not grow under
ordinary conditions.* Finally, by introducing the use of milk in

' Das Gahrungskélbchen in der Bakteriologie, Centralbl. f. Bakt., vii,
PP- 502-506, 1890; Einige Bemerkungen tiber Sdure- und Alkalibildung
bei Bakterien, [bid., viii, p. 389, 1890; The Fermentation Tube, etc.,
Wilder Quarter-Century Book, pp. 187-234, 1893.

2 Centralbl, f. Bakt., vii, pp. 502-506, 1890.

* There seems to be considerable question about the priority of this
procedure. According to Marino (Méthod pour isoler les anaérobies,
Amn. de Vinst. Pasteur, xxi, p. 1005, 1907) the credit belongs to Duen-
schmann (Etude experimentale sur le charbon symptomatique et ses
relations avec l’oedéme malin, Amn. de l’inst. Pasteur, p. 482, 1895) who
mentions the fact that Roux, under whose direction this work was carried
out, had previously used serum (beef) to cultivate certain anaérobes.
As a matter of fact Theobald Smith (Das Gahrungskélbchen in der Bak-
teriologie, Centralbl. 7. Bakt., vii, p. 502, 1890) very distinctly mentions
the fact that sterile animal tissue may be employed to advantage inthe
cultivation of certain obligate anaérobes and he actually employed fer-
Mentation tubes enriched in this manner at that time.

283

284 Fermentation Tube in Intestinal Bacteriology

the fermentation tube, he has added much to the value of this
medium for bacterial research.!

Smith’s published results have been obtained chiefly with
pure cultures. Herter and Ward” have studied the gas volume
with fermentation tubes inoculated with mixed intestinal flora
and Herter® has studied the sediments derived from the inocu-
lation of such tubes.

TECHNIQUE.

(1) Preparation of sample. It is necessary at the start to
have fresh specimens of stools for investigation—samples that
have stood for some time or have been freely exposed to the air
or to high temperatures are liable to change rapidly in their
bacterial composition. Nonspore-forming anaérobes may rapidly
succumb while, coincidently, hardy forms multiply until they
exist in abnormally large numbers, thus influencing very markedly
the bacterial aspect of the result.

It is necessary in this particular line of investigation for several
reasons to inoculate appropriate amounts of fecal material; the
portion added must not be too great or the acids and other antag-
onistic products formed by the more rapidly growing faculta-
tive organisms will seriously inhibit the growth of the more strictly
parasitic types; too small an amount, on the other hand, may
fail to furnish a representative growth of significant microérgan-
isms. The less numerous but perhaps extremely important
forms may, without these precautions, be overlooked and missed
entirely.

For routine purposes one gram of feces thoroughly emulsified
in ten cubic centimeters of physiological saline solution is an

1 Milk intended for the fermentation tube must be sterilized at least
four successive times at appropriate intervals to insure the absence of
resistant spore-forming bacteria which ordinarily escape observation.
After sterilization and before inoculation milk fermentation tubes must
be incubated several days at body temperature to test their sterility.

See Smith, Brown and Walker: The Fermentation Tube in the Study of ©

Anaérobic Bacteria with Special Reference to Gas Production and the
Use of Milk as a Culture Medium, Journ. Med. Research, xiv, pp. 19 3-206,
QOS.

> This Journal, i, p. 415, 19006.

3’ The Common Bacterial Infections of the Digestive Tract, 1907.

ee oe. oe

C. A. Herter and A. I. Kendall 285

excellent dilution. Probably all types present that are capable of
development in fecal media are thus represented.’ If mucus is
present in the stool it should be washed in sterile water and inoc-
ulated separately. One cubic centimeter of the suspension is
placed in each tube.

Certain forms, as for example, the B. bulgaricus described by
Metchnikoff, and many alkali-producing bacteria, will not grow
in the ordinary fermentation media but usually develop rapidly
in milk fermentation tubes.

The period of incubation is a very important factor. Experi-
ence has shown that during the first eighteen to twenty hours
(rarely longer than this) the majority of the vegetative cells will
be at their maximum growth; after this time, owing partly to
antagonism, partly perhaps to the fact that the nutrient material
that was carried over with the suspension is exhausted, many
forms die out, while the more saprophytic organisms increase
enormously.

The actual gas volume is rather less at twenty hours than at
subsequent periods as a rule, but the relative value of this feature
is at the maximum. The bacteria derived from ordinary stools,
particularly of the colon type, tend to attain a more or less con-
stant gas volume after forty-eight hours.

In plain bouillon without the addition of carbohydrates (par-
ticularly with media made from meat juice instead of meat
extract as a basis, gas is sometimes liberated after acidification
of the culture with hydrochloric acid, although no gas was
present during the incubation period. Also the addition of cystin
to the medium tends to increase the gas volume. This gas
is hydrogen sulphide, and its total amount may be estimated with
a considerable degree of accuracy through the absorption effected
by the addition of a soluble salt of a heavy metal.

The gas ratio is not an especially important characteristic in
mixed fecal flora, much less so in fact than is the case with pure
cultures. Frequently the volume is too small to measure, and

1The emulsion should be rapidly prepared and inoculated. Too long

an exposure in the relatively aérobic saline solution may, and frequently

does, eliminate many vegetative and nonspore-forming anaérobic forms,

particularly those of infant flora, while the more vigorous aérobic bacteria
increase rapidly.

286 Fermentation Tube in Intestinal] Bacteriology

the inability to decide offhand which organisms are concerned
in its elaboration makes the determination at best of doubtful
value.

Of the greatest importance, on the contrary, is the character
of the sediment. Smith’ has shown that the deposit at the base
of the closed arm is an excellent source of material for reinocu-
lation with anaérobic bacteria (using pure cultures) and our obser-
vations have demonstrated in addition the fact that certain
organisms assume characteristic appearances which in some
instances are even diagnostic.

For making suitable smears the material is best removed by
a finely drawn out capillary pipette, then spread upon slides,
using the tip of the pipette as a spreader. The excess of solu-
tion runs back into the pipette by capillarity, leaving a thin,
uniform smear which dries quickly and stains readily.

Gram’s staining method followed by dilute carbol fuchsin or
anilin-oil safranin furnishes the most distinctive preparations.

2. Fermentative media. For ordinary purposes, the regular
dextrose, lactose and saccharose bouillons are employed. Freshly
sterilized media are desirable because the anaérobic condition is
much better developed in the closed arm in such media.

For special investigations, where it is necessary to be absolutely
certain of the reaction of an organism upon different sugars,
one of us (A. I. K.) has succeeded in eliminating the chief sources
of error due to heating carbohydrate media, namely, the caramel-
ization and inversion. This is conveniently accomplished by
sterilizing the sugar separately (as a 10 per cent solution) by
two passages through appropriate Berkefeld filters. The sugar
solution is aded to the nutrient bouillon (freed from fermentable
substances by Smith’s method) in such amounts thata1 per cent
solution of fermentable carbohydrate results.” The bouillon is

1 Das Gé&hrungskélbchen in der Bakteriologie, Centralbl. f. Bakt., vii,
PP. 502-506, 1r89go.

? Liborius (Zettschr. f. Hyg., i, p. 116, 165, 1886) was the first to recog-
nize the importance of sugars and particularly dextrose, as a nutritive and
reductive medium. He used 2 percent, and his example has been followed
without question by the majority of English, French and German investi-
gators. Smith (Centralbl. f. Bakt., xviii, p. 7, 1895; Xxii, p. 49, 1897) also
finds dextrose very important for the development of anaérobes but his

C. A. Herter and A. I. Kendall 287

sterilized in fermentation tubes in the usual manner, adding rather
less than usual to provide space for the carbohydrate which is
introduced later.

Media made up in this way have one disadvantage, namely, the
lack of absolute anaérobiosis that obtains when they are sterilized
as one solution. This objection is more imaginary than real,
however, because experience has shown that intestinal organisms
capable of growing in pure culture in the ordinary single solution
media will grow quite as well in the double solution medium.

RESULTS.

Herter and Ward!’ found, using dextrose, Schering’s diabetin
(dextrose-levulose mixture), lactose and saccharose, the follow-
ing average amounts of gas produced by normal stools (sixteen in
all):

Dextrose. Dextrose-levulose. Lactose. Saccharose.

26,75 27.50 29.9 19.5 mm.

These fermentation tubes had anaérobic arms approximately
95 mm.long. Asa rule the lactose tube showed the most gas;
saccharose the least. In conditions of disease the gas volume
varied greatly—in certain instances 50 per cent of the vertical
limb was filled; in other cases no gas was found.

Our own series indicate in addition that feces derived from
many individuals harboring Bact. Welchii (the gas bacillus) may
form extremely large volumes of gas—even yo to 100 per cent of
the whole tube—although the average is much less, usually about
45mm. Pure cultures of B. coli form about 30 mm. of gas under
similar conditions.

Coccal forms when present in numbers, generally inhibit gas
formation: gas may even not be formed at all in certain cases,
although gas-forming organisms are present. Normal infants,
particularly those that are exclusively breast-fed, form as a rule
tather less gas than adults. The volume varies with age and con-

results show that 2 per cent is too great a quantity and that 1 per cent is
far preferable. Marino (loc. cit.) finds from 0.3 to 0.5 per cent even better
and our experience indicates that the lesser amounts are preferable in
many instances.

1 Loc. cit.

288 Fermentation Tube in Intestinal Bacteriology

dition but fifteen to twenty millimeters represents fairly the aver-
age. The amount furthermore depends upon the relative pro-
portion of B. bifidus present; if the latter organism be abundant,
less gas is formed, because this species produces sufficient acid
to inhibit the growth of the ordinary gas-formers.

Diarrhoeal stools vary in their gas production. In those cases

where large numbers of cocci are brought down from higher levels’

of the intestine, relatively small gas volumes are the rule, while
in similar movements associated with large numbers of colon
bacilli, or with organisms of the lactis aérogenes type, a much
greater production takes place.

Mention has already been made of the fact that the fermenta-
tion tube is a particularly simple and efficient apparatus for culti-
vating anaérobic intestinal bacteria. Of these organisms a few
will not grow in pure culture under the same conditions, although
they usually thrive symbiotically with facultative anaérobes.

There can be no doubt that certain substances, particularly
favorable for the growth of these more or less strictly parasitic
forms are carried into the fermentation tube as a part of the fecal
suspension, and during the first eighteen or twenty hours furnish
a suitable pabulum for their growth—material, furthermore, which
is not present in the fermentation tube as it is ordinarily made
up. It is extremely probable that their growth is further aided

by the presence of more readily growing bacteria which frequently

render the tubes extremely anaérobic by the removal of the last
traces of dissolved oxygen.

In the fermentation tube every transition from almost com-
plete anaérobiosis to aérobiosis obtains and it is possible for bac-
teria to find almost any tension of oxygen from more or less com-

plete saturation in the bulb to practically its entire absence in

the closed arm.
With such favorable conditions—proper food supply and gase-
ous environment—the growths are very varied and in a meas-

ure representative of the organisms originally present in the ~

feces. This fact is best appreciated after one examines the sedi-
ments, particularly those stained by Gram’s method followed by
the counter stain mentioned above.

The organisms are as a rule much more characteristic morpho-

logically in the deposit at the bottom of the closed arm than is the ~

un «> pre = tine

C. A. Herter and A. I. Kendall 289

case in the feces from which they were derived, because the major-
ity are in what may be termed the “active” vegetative stage.
Bacteria in this condition are larger and more nearly typical than
under conditions where they have become attenuated and degen-
erate in their morphology, as frequently happens in the case
of constipated stools. The staining reactions also are much
sharper and more distinctive at this period.

Perhaps the most striking example of the differentiation one
may ordinarily meet with in a sediment from a fermentation tube
is that shown by a common bacterium in infants’ stools called by
Tissier, its discoverer, B. bifidus commums. This organism is an
anaérobic Gram-positive bacillus, frequently occurring with
rather pointed ends in normal infants’ stools; not readily recog-
nized and not especially characteristic. Furthermore it is not an
easy organism to cultivate in ordinary media. In fermentation
tubes, however, it grows rapidly and at the end of eighteen hours
shows the peculiarly striking bifid ends to which it owes its name.
This fact, judging from the literature published upon the subject
so far, has hitherto been unrecognized, but it appears to be char-
acteristic, of great constancy, and a unique example of the value
of such examinations. This organism will not grow, or at least
only slightly, in fermentation media in pure culture. If, how-
ever, one adds a bit of sterile animal tissue, or inoculates directly
from a stool, together with other bacteria, the growth is marked.

It is advisable, and frequently necessary, to use fermentation
tubes containing plain bouillon instead of the regular fermenta-
tion media. Certain bacteria will not grow well where ferment-
able sugars are present, while others are rapidly eliminated as the
medium becomes acid. B. putrificus isa good example of such an
organism. Sediments derived from plain bouillon in fermenta-
tion tubes, particularly those which are rendered more suitable
by the addition of sterile animal tissue, frequently show anaé-
robic growths that would not be included in carbohydrate solu-
tions.

One point in connection with the fermentation tubes deserves
special mention—gas volumes are frequently variable with the
Same individual and it is necessary to cover considerable periods
of time before assigning a special value to this factor in indi-

vidual cases or attaching much importance to deviations from
the average.

290 Fermentation Tube in Intestinal Bacteriology

Among the organisms ordinarily met with in the feces, a few
of the more important may be mentioned:

(1) Bact. Welchit, a thick, rather large, strongly Gram-posi-
tive bacillus. In large numbers they give rise to a considerable
augmentation of the normal gas volume, so that the amount is
frequently twice the normal.

An organism described by Herter’ resembles the gas bacillus
morphologically but does not form gas.

(2) B. cok. Short bacilli, about one micron in diameter,
Gram-negative. These organisms usually determine the gas vol-
ume and it is chiefly to their action that the normal gas volume
is due.

(3) B. lactis aérogenes, somewhat more oval than the colon
bacillus and like that organism, Gram-negative. This form is
not particularly common in the stools of adults? but is present
usually in the excreta of bottle-fed infants and tends to increase
the gas volume, if numerous.

(4) Coccal forms, usually Gram-positive. These bacteria pro-
duce as a rule considerable amounts of acid but no gas, and
inhibit to a considerable degree the fermentative action of the
above mentioned forms.

(5) A Gram-positive bacillus with bifid ends (B. bifidus). It
is very common in the stools of breast-fed infants. When this
organism is present in numbers, the amount of gas is usually con-
siderably reduced. Its inhibitory action is due, as is the case
with the coccal forms, to its excessive acid production.

In several instances, B. bifidus has been isolated from mucus,
while the remainder of the stool was almost devoid of these forms.

The full significance of the fermentation-tube sediments cannot
be regarded as completely worked out. It is a striking peculiar-
ity of the growths in the sediments that they frequently do not
show a multiplication of microérganisms closely representative
of the varieties which are seen in the Gram-stained fields of the
feces. This failure in correspondence between the characters of
the dominant organisms in the fermentation tubes on the one

IEOGMGLE:

*MacConkey (Lactose-fermenting Bacteria in Feces, Journ. of Hygiene,
1905, pp. 333-379) found it in 4 out of 625 lactose-fermenting cultures
from normal stools, both animal and human.

C. A. Herter and A. I. Kendall 291

hand and the feces themselves on the other, depends largely
upon the fact that the nutrient conditions are ordinarily radically
altered by the transfer to the fermentation media. This altera-
tion in medium makes it possible for types of bacteria not obvi-
ously dominant in the feces, or, indeed, clearly in the minority,
to gain a relatively prominent position under these conditions.

The fact that such a readjustment of types is liable to occur
has important advantages and equally significant drawbacks.
' Without recognizing the disproportionate growth one might
erroneously assume that a much larger portion of a certain flora
is present than is the case; for example fecal fields may contain
small numbers of B. bifidus, yet in the fermentation tubes they
may be prominent. Again, the gas bacilli may undergo exten-
sive multiplication in the fermentation tube despite the fact that
the fecal material from which they were obtained contains them
in moderate numbers only. Here again one sees the necessity for
controlling the appearances obtained from the sediments of the
fermentation tubes by means of cultures from the stools as well
as by close examination of the Gram-stained fecal fields. Similar
overgrowths occur with the coccal forms, Mic. ovalis (entero-
coque), streptococci and staphylococci.

The disproportionate growth has its advantages as well as its
disadvantages. Certain types which are significant although orig-
inally occurring in small numbers are thus brought to notice when
otherwise they would be overlooked.

Experience has shown that overgrowths occurring during the
first eighteen to twenty hours of incubation are due to the pres-
ence of significant numbers in the stools, capable of asserting
themselves in the higher levels of the digestive tract, and capable
of enormous proliferation under suitable nutrient conditions.
An excellent example is again furnished by Bact. Welchii. In
patients who have an infection of the intestinal tract with this
organism there may be times of improvement when the numbers
of this particular type in the feces is small, asshown by the micro-
scopic examination of the fecal fields—so small, in fact, that if
the observation were confined to the patient at this time, no
suspicion would be excited of the existing tendency of over-
growth of these organisms in the intestine. Yet upon inocula-
tion of the feces into fermentation tubes a prominent, active

292 Fermentation Tube in Intestinal Bacteriology

growth of these organisms is very liable to occur under these con-
ditions.

In contrast with this is the following observation. The fecal
fields from normal nurslings and bottle-fed children commonly
show a few organisms having the morphology of the gas baciilus.
That these bacteria belong in the class of the gas bacilli is
made probable through the fact that by inoculating relatively
large amounts of the feces into rabbits’ ear-veins, with subse-
quent incubation (Welch-Nuttall test) the typical gas-liver will be
developed. Inoculations into fermentation tubes made from
feces of this type of case have, in our experience, failed uniformly
to show overgrowths of this bacillus.

It should be clearly understood that the presence of moderate
or even considerable numbers of Bact. Welchiz does not necessarily
lead to overgrowth in the fermentation tube.

AN OBSERVATION ON THE FATE OF B. BULGARICUS (IN
BACILLAC) IN THE DIGESTIVE TRACT OF A MONKEY.

‘

(Plates I-III).

BN

C. A. HERTER
AND
A. I. KENDALL.

(Fellow of the Rockefeller Institute.)

(Received for publication, August 12, 1908.)

The incentive to make the observations here recorded came
from certain problems suggested by the rapidly extending use
of artificially soured milk which is following closely upon the
publication of Metchnikoff’s highly speculative volume, entitled
“The Prolongation of Life.”” Really decisive experiments bearing
upon the value of what may be termed ‘“‘sour milk prophylaxis”
are not numerous and even Scriptural quotations have been in-
voked to lend color to, and enhance the meager supply of litera-
ture. One of us' has studied the effect of introducing largenum-
bers of B. coli, B. proteus vulgaris and B. acidt lactici into the in-
testinal tract of dogs to determine their action upon indol pro-
duction. Living cultures of B. coli and B. proteus, for example,
caused an increase of indican and of ethereal sulphates in the
urine.? Killed cultures prepared in the same manner gave little
or no increase in these putrefactive constituents.

1C. A. Herter: On Certain Relations Between Bacterial Activity in
the Intestinal Tract andthe Indican of the Urine. Brit. Med. Jour., 1897,
di, p. 1847.
_ ? Precautions were taken to prevent the introduction of putrefactive
products derived from the media upon which the bacteria were grown.
The organisms were cultivated upon slanted agar (removed by careful
washing with salt solution, avoiding the inclusion of any portion of the
agar) and then injected.

293

294 B. Bulgaricus in the Digestive Tract

Lactic acid bacteria injected directly into the small intestine
in similarly conducted experiments showed a tendency to cause
a reduction in the output of indican and ethereal sulphates.

Following similar lines, Metchnikoff directed the performance
of experiments which lead him to believe that lactic acid-pro-
ducing bacilli once successfully established in the intestinal tract
decrease or even prevent the multiplication of putrefactive
organisms. This inhibiting effect he attributes to the lactic
acid which is produced by microérganisms introduced in these
experiments.

Where putrefactive disturbances are already present the lactic
acid bacilli appear to.reenforce the enfeebled action of the normal
intestinal lactic acid bacilli and increase the amount of acid pro-
duced where it is insufficient, or reintroduce it where it is absent,
thus assisting the host to throw off the ‘‘wild races” of bacteria
that may have become habituated to the intestinal tract. This,
at least, is the assumption.

Not all lactic acid-producing bacteria are suitable for this
purpose and the choice of a culture should be based upon the fol-
lowing criteria: First, the organisms should be able immediately
or in time to become habituated to the intestinal tract; second,
they should produce no toxins or putrefactive substances or
other injurious products, detrimental to the host; third, they
should be able to make sufficient lactic acid to accomplish the
purpose for which they are introduced.

The organism which Metchnikoff selected for his researches
was originally obtained from the Bulgarian ferment called “‘ You-
gourt”’ and the bacillus which apparently is the most potent
lactic acid producer in this ferment has been named B. bulgaricus.’
This organism, as we have studied it,is rather long (1 X 4-6 mic-
rons), large, with rounded ends, growing singly or in pairs, rarely
in chains. It stains well with the ordinary anilin dyes particu-
larly in young cultures, and is Gram-positive. In old milk cul-
tures some of the rods may be Gram-negative while others pre-
sent a punctate appearance, due apparently to the concentration
of protoplasm in certain portions of the cell which are Gram-
positive, and suggest Ernst-Babes granules, the remaining por-

1 Cohendy: Compt. rend. de la soc. de biol., Ix, 1906.

>.

C. A. Herter and A. I. Kendall 295

tions undergoing rarefication of protoplasm and becoming Gram-
negative. On media containing no carbohydrates there is no
growth. On dextrose or lactose agar the growth is slight and
usually appears as small, stellate colonies rarely exceeding one
to one and one-half millimeters in diameter. In stab cultures
on the same media a slight growth appears after three to five
days but always remains limited to the line of inoculation. In
dextrose and lactose bouillon there is usually a feeble growth
after several days, appearing as a fine sediment visible only after
agitating the tube. In milk cultures, on the contrary, the growth
is very vigorous, resulting in a rather finely flocculent coagulum
with a minimal separation of fluid. In this medium considerable
amounts of lactic acid are formed. Bertrand and Weissweillert
studied the chemical action of B. bulgaricus on milk and found
that a slight amount (usually less than 10 per cent), of the casein
is peptonized and apparently utilized as food by the bacterial
cells. A small amount of the fat is saponified while practically
all of the lactose is changed to dextro- and levo-lactic acid, the
dextro variety predominating. Twenty-five grams per liter of
lactic acid are easily formed and at the same time small amounts
of acetic, formic and succinic acids are produced, usually not
more than half a gram per liter. Inoculations into milk are
active, even after fourteen days, although at the end of three
weeks the bacilli are usually dead. In our experience the amount
of lactic acid produced by B. bulgaricus is sufficient at the end of
forty-eight hours to render the milk unpleasant to the taste, par-
ticularly if previous to inoculation the culture has been fre-
quently transplanted, so that the organisms are in an active
vegetative state. Soured milk prepared according to Metchni-
koff’s directions is said to contain about ten grams of lactic acid
per liter.”

Several investigations have been made by various observers to
determine the effects of the Bulgarian bacillus upon the intestinal
flora and various putrefactive products. Cohendy,’ experiment-
ing upon himself, found that pure cultures of lactic acid bacill
had a tendency to reduce intestinal putrefaction and that the

1 Ann. de Vinst. Past., xx, pp. 977-990, 1906.

7 Metchnikoff: The Prolongation of Lije, p. 180.
ERILOGL CIE:

296 B. Bulgaricus in the Digestive Tract

organisms could be recovered from the feces without difficulty
several weeks after the experiment was stopped. He took B.
bulgaricus for about two and one-half months. Pochon! con-
sumed cultures of lactic acid bacilli in milk and noted the dimi-
nution of indol and phenol in his feces. Leva? investigated the
effect of Lactobacilline, milk, and milk plus Lactobacilline upon
the excretion of ethereal sulphates, volatile fatty acids, aromatic
oxyacids, phenol and indican. The experiment was divided
into four periods: (1) a uniform daily diet; (2) diet + Lac-
tobacilline;? (3) diet + Lactobacilline + one liter of milk; (4)
diet + one liter of milk. His conclusions are as follows:

(1) The excretion of ethereal sulphates during the experiment
was practically unchanged.

(2) The excretion of volatile fatty acids with Lactobacilline
alone, or milk alone, as well as with Lactobacilline and milk com-
bined, showed a considerable decrease.

(3) The excretion of aromatic oxyacids and hippuric acid was
uninfluenced by milk, decreased distinctly in amount with Lacto-
bacilline, decreased greatly with Lactobacilline + milk.

(4) The phenol excretion decreased somewhat under the in-
fluence of Lactobacilline alone, as well as with milk alone; there
was a much greater decrease with a combination of Lactobacilline
+ milk.

(5) The indican excretion was very slight at the beginning of
the experiment (too small an amount to measure accurately) and
remained practically unchanged throughout the entire period.

Belonowsky? studied the influence of these organisms upon the
intestinal flora of mice. His method was to contaminate the food
(usually grain previously sterilized by heat) with B. bulgaricus,
allowing the animals to eat sufficient quantities to make certain
that the organisms were actually introduced in large numbers.
His results may be summarized as follows: First, the Bulgarian

1 Cited by Combe: L’autointoxication intestinale, Paris, 1906.

* Leva, J.: Zur Beurteilung der Wirkung des Lactobacillins und der
Yoghurthmilch, Berl. klin. Wochenschr., xlv, pp. 922-924, 1908.

’ The Lactobacilline was obtained from ‘‘Le Ferment’? Company of
Paris. Leva’'s observation that a yeast was present in the Lactobacilline
agrees with our own observation on this point. (See Fig. 1.)

4 Ann. de l’inst. Past., xxi, p. 991, 1907.

-

C. A. Herter and A. I. Kendall 297

ferment modifies the normal intestinal flora of mice by a general
alteration in their character and by elimination of putrefactive
forms. There is a diminution in the total number of bacteria as
well as a lessened virulence of the feces when these are introduced
intraperitoneally or subcutaneously into other animals. Second,
the action is not attributable to the formation of lactic acid alone,
but also to certain products inhibitory in nature, formed by the
bacilli themselves. Third, the organisms become more or less
established in the intestine about the tenth day in adult mice
and persist without further reinoculation for a considerable but
variable interval of time. Fourth, the cultures seem to have
exerted a beneficial action upon the mice, particularly on those
infected with the organisms of mouse typhus; in this case the
results are due exclusively to the lactic acid.

From these investigations it would appear that many of the
animals fed upon grain contaminated with the Bulgarian ferment
gained in weight, that the feces contained fewer organisms cap-
able of growing upon ordinary culture media, that putrefactive
organisms tended to disappear and that this beneficial action was
due in part to the lactic acid, in part to the products of metabo-
lism of the bacteria themselves. The fact that relatively few
of the Bulgarian bacilli are microscopically discernible in the
feces raises the question, In what portion of the intestinal tract
do these bacilli find their most favorable environment? If they
occur in the upper (duodenal or jejunal) regions of the small in-
testine and only a few gain a foothold in the large intestine (usu-
ally considered the chief site of putrefaction) the organisms must
act from a distance and their products, theoretically at least,
must be less effective than if they were generated at the focus of
infection. No experiments so far have been recorded which an-
swer this question and the present investigation has approached
the problem from this point of view. For our work, which was
undertaken specifically to study the distribution of B. bulgaricus in
the intestinal tract, the preparation called Bacillac was employed.
This is said to be made according to Metchnikoff’s personal direc-
tions, cultures of the organism described above (B. bulgaricus)
being employed for this purpose. This organism is stated to
have been isolated by Metchnikoff. The Bacillac is obtainable
in pint bottles and contains a moderately large, Gram-positive

298 B. Bulgaricus in the Digestive Tract

bacillus and (by accident or design) a large, oval, Gram-positive
yeast as well (see Fig. 1). The bacillus isolated from specimens
of Bacillac grows slowly upon ordinary dextrose and lactose agar,
very poorly in corresponding bouillon media, but luxuriantly in
milk. It produces in the latter medium a soft coagulum which
after standing for a few days becomes massed into more or less
permanent lumps with a moderate separation of fluid. The
acidity increases rapidly and after forty-eight hours becomes
decidedly unpleasant to the taste.

Partly because of its rapid growth, but chiefly because of the
considerable amount of acid which this bacillus produces, it is
easy to obtain cultures of the organism grown in milk at 37° C.,
even if it be originally associated with other organisms. Careful
sub-culturing gives a differential enrichment of the Metchnikoff
bacillus, so that one may obtain it in pure culture, as may be
demonstrated by plating on slightly acid Bierwort agar. In the
present investigation this method of enrichment has been used
successfully for the isolation of the organism from the mixed
intestinal flora.

For experimental purposes a moderate sized Rhesus monkey
was used. The animal received daily half a liter of sweet milk
for a period of three days. The feces were examined for lactic
acid! as well as for organisms resembling the Metchnikoff bacillus

1 Fletcher and Hopkins: Journ oj Physiol., xxxv, pp. 247-309, 1908.
Reagents:

(rt) Very dilute alcoholic solution of thiophene (10 to 25 drops
in I00 cc.

(2) Saturated aqueous solution of copper sulphate.

(3) Concentrated sulphuric acid.

Procedure: 5 cc. strong sulphuric acid, 1 drop copper sulphate and a
few drops of the suspected mixture are well shaken, then heated for 2 to 5
minutes in a water-bath in a test tube. Cool the solution, add 2 to 3 drops
thiophene solution, replace in water-bath and again heat, watching con-
stantly. Lactic acid rapidly and characteristically gives a bright cherry-
red color under these conditions. The lactic acid must be as nearly free
from water and organic matter as possible (ma lic acid and probably other
oxy-acids give the reaction).

The lactic acid mav be employed as an alcoholic solution or as a syrupy
residue. Acetaldehyde and glyoxylic acid give color reactions with thio-
phene and sulphuric acid, but the copper sulphate used as above destroys

ee ee

C. Ay Herter and A. |. Kendall 299

either culturally or morphologically. In no instance was lactic
acid found.

The experiment with Bacillac was carried out as follows. The
amount of soured milk given was the same as in the control ob-
servations, namely, five hundred cubic centimeters daily. Atthe
end of the second day, chemical and bacteriological examina-
tions were commenced, but until the sixth day no lactic acid
bacilli were isolated nor could lactic acid be detected in the ani-
mal’s feces. Two days later the first positive test for lactic acid
was obtained, using the thiophene reaction of Fletcher and Hop-
kins. The feces were slightly but distinctly acid at this time—
more acid than upon previous occasions. It was difficult to ob-
tain a satisfactory test for lactic acid from the feces owing to the
presence of a brownish-yellow coloring matter soluble in ether.
But it was possible by removing the ether through evaporation,
taking up the oily residue in water, boiling with animal charcoal,
filtering, washing, and again evaporating to a syrup to obtain a
slightly yellow solution which gave a good reaction with the
thiophene. Controls made from normal feces of the same animal
(free from lactic acid), which were treated in the same manner,
but to which were added known minute amounts of lactic acid,
gave in every instance the same color reaction. After fourteen
days the monkey was given the usual meal of five hundred cubic
centimeters of Bacillac. Then, after allowing three and one-
half hours for digestion (the whole portion of milk having been
consumed at this time) the animal was killed by chloroform and
examined. Samples taken from the stomach and from various
levels of the small and large intestines were removed with appro-
priate precautions for microscopical, bacteriological and chemical
examination. The material for chemical examination was placed
in ether slightly but distinctly acidified with sulphuric acid. The
bacterial material was inoculated into milk and fermentation
tubes containing dextrose, lactose and saccharose bouillon. The
specimens for microscopical study were smeared on slides and
stained by the Gram method for identification of forms resem-

them. If ether is employed for extracting lactic acid, it should be first
washed with water to remove aldehyde bodies. The color produced by
lactic acid is transitory unless the tube be cooled immediately after its
appearance.

300 B. Bulgaricus in the Digestive Tract

bling morphologically the acid bacillus fed. The animal was in
good health before and during the experiment. At autopsy
about one hundred cubic centimeters of partially digested milk
were found in the stomach. The small intestine contained a
moderate amount of semi-fluid, yellowish, gelatinous substance.
At the ileo-czecal valve the contents were more abundant and
about the consistence of thick paste. The color was a deeper
brown. The color and consistence increased progressively to
the anus, where the feces were solid and fairly dark. The
mucous membrane throughout the gastro-intestinal tract ap-
peared normal, although the reaction of the contents from the
stomach to the anus was distinctly acid to litmus.

The Bulgarian bacilli were present and easily demonstrated by
smears and by cultures in relatively large numbers in the stomach
contents but associated with yeasts. In some instances, particu-
larly where the milk was obviously undergoing digestion, the or-
ganisms showed undoubted signs of degeneration. The staining
was very irregular and faint in such bacilli, whereas the yeasts,
so far as could be determined by microscopical examination,
showed no such changes. In the duodenum and ileum the Bul-
garian organisms were encountered in almost pure culture,
although inoculations into fermentation tubes showed a few
gas-forming bacilli of the colon type. In the region of the ileo-
cecal valve there was a rather abrupt change in the nature
of the bacterial smears. Not only was the amount of fecal ma-
terial greater but the character of the microédrganisms was dif-
ferent. B. bulgaricus ceased to be the dominant organism,
although it was still present in moderate numbers. Gas-forming
bacteria were, on the other hand, increased enormously. Gram-
positive rods and cocci were also present. No attempt was made
to identify the latter. From the ileo-cecal valve, progressively
down the large intestine towards the anus, the number of the Bul-
garian bacilli decreased while the number of other bacteria
increased, until at the rectum there were very few Metchnikoff
bacilli but enormous numbers of bacteria of the colon type and
many Gram-positive rods.

Lactic acid was demonstrated throughout the gastro-intestinal
tract as well asin the feces. Although no attempt was made to
determine quantitatively the amount of acid at any level of the

ntsc Rt gle

Gr Av ierter and A: 1: Kendall 301

intestine, the results indicated that much less lactic acid was
present in the large intestine than in the stomach and small intes-
tine. This diminution (shown by the decidedly lessened color
developed by the thiophene, using approximately equal amounts
of intestinal contents) began rather abruptly at the region of the
ileo-ceecal valve, and progressively increased to the anus. This
phenomenon was particularly marked in the lower portions of
the large intestine, where the contents were more desiccated.
The amount of material obtained from this region was greater
than was the case in higher levels, while at the same time the
volume of lactic acid was decidedly less, although the ether
extraction was prolonged. This coincides with the relatively
smaller number of lactic acid bacilli found.

CONCLUSIONS.

(1) By feeding a Rhesus monkey for two weeks exclusively
on milk fermented with B. bulgaricus (but containing also some
yeasts) it was possible to maintain an acid reaction throughout
the digestive tract. The acid reaction was more pronounced
above the ileo-cecal region than at this region or below it. The
acidity decreased progressively from the ileo-cecal region to the
anus. Lactic acid was detectable at every point in the digestive
tract that was tested, the reaction growing less marked below
the ileo-cecal region.

(2) Exclusive feeding for two weeks with milk fermented
with B. bulgaricus failed to establish the predominance of this
organism in the ileo-cecal region or in the large intestine. In
the latter situation the number of bacilli of this type was rela-
tively small and decreased towards the anus. Thus in the regions
characterized by most active putrefaction the lactic acid bacilli
failed to establish themselves in relatively large numbers.

302 B. Bulgaricus in the Digestive Tract

DESCRIPTION OF THE PLATES.

Fig. 1. Smear from stomach contents of monkey, three and one-half
hours after feeding Bacillac. Practically pure culture of B. bulgaricus and
few yeast cells. Bacteria show ‘‘punctate’’ staining due, apparently, to
combined effect of partial digestion and excessive acidity.

Fig. I]. Contents of small intestine in the region of the duodenum.
The Bulgarian bacilli predominate. Afew punctate formsare seen. The
bacteria are multiplying at this point. The normal appearance of the
bacteria is in accordance with this observation.

Fig III. Contents of the small intestine near the ileo-czecal valve.
Uniformly staining Bulgarian bacteria are present, but not in predominat-
ing numbers as in Fig. II. Gram-negative forms begin to predominate.

Fig. IV. Contents of the large intestine about two feet from the ileo-
cecal valve. <A few typical Bulgarian bacilli still persist. One of these
shows the punctate staining seen in the stomach. Yeast cells are also
present. Gram-negative bacilli are the dominant forms.

Fig. V. Feces of monkey. Bulgarian bacilli have practically disap-
peared. They were, however, readily demonstrated by the milk-enrichment
method described in the text.

Fig. VI. Fermentation sediment obtained from a lactose fermentation-
tube inoculated with intestinal contents obtained from the level of the
ileo-czecal valve (cf. Fig. III). No Bulgarian bacilli are present in this
sediment, although they were readily demonstrated in the smear made
at this level as well as culturally in milk. (This figure is introduced to
show that the Bulgarian bacilli do not grow in fermentation media.)

The slides from which these plates are reproduced were stained
by Gram’s method. Magnification, 1000 diameters. Zeiss 2
mm. homo. imm. apochromatic lens.

Our thanks are due to Dr. Leaming of the Rockefeller Insti-
tute for Medical Research for the above photographs.

THE JOURNAL oF BIOLOGICAL CHEMISTRY, Prate I,
VOL. V, NOS. 2 AND 3.

aa £ -
“eo 4
ee ™ .

Fig. I. Smear from stomach contents of monkey, showing great pre-
dominance of B. Bulgaricus and a few yeast cells.

—

i i

_. Fig. II. Smear from contents of duodenum: B. Bulgaricus predominating.

= $
5 a
: E
3 da)
= ~
S|
oO
BS
g
fe)
H
Qy
“a
=
S
ws
~~
3
a
S
isa}
Yo e
[4 o
; 5
a $
= o
fea) »
rE z
S) 35 =|
26 E
OZ 5
PAI <tt n
o
fom E
BA 9
co os
ong a
jap ey oO
ome Bs
HA wn
20 aE
Z> moe
2 Hs
P HS
io) Peers
= on O
ES
H
[or

4

VIL.
ube inoculated with intestinal contents from the level of the

Puare III.

Smear from fermentation-tube sediment from lactose fermen-

FURTHER STUDIES OF THE MODE OF OXIDATION OF
PHENYL DERIVATIVES OF FATTY ACIDS IN THE
ANIMAL ORGANISM. III.

(SYNTHESIS OF SOME DERIVATIVES OF PHENYLPROPIONIC ACID.)
By H. D. DAKIN.
(From the Laboratory of Dr. C. A. Herter, New York.)
(Received for publication, October 20, 1908.)

In a former communication’ it was stated that cinnamoyl-
glycocoll had been identified as an intermediary product in the
catabolism of phenylpropionic acid and phenylvaleric acid.
This result suggested much that is of interest not only with
regard to the mode of catabolism of fatty acids, but also as to
the origin of the unsaturated fatty acids in the body and in
addition, the réle which glycocoll and other amino-acids play
in fatty acid metabolism.

Knoop’s observation’ of the excretion of hippuric acid follow-
ing the administration of phenylvaleric acid to dogs left the
question open as to whether the side-chain had been oxidized
directly in the 0-position or not, but the observation of the forma-
tion of cinnamoylglycocoll at once proves that, at least in part,
the oxidation takes place at the @-carbon atom with removal of
two carbon atoms and that a second oxidation in the {-position
results in the formation of benzoic or rather hippuric acid.
These results furnish a complete proof of the accuracy of the
hypothesis of f-oxidation advanced in the first instance by
Knoop.

Excluding the intermediary steps in the oxidation, and with-
out representing the glycocoll grouping, the change may be
expressed as follows: |

C.H,.CH,.CH,.CH,.CH,.COOH—>C,H,.CH,.CH,.COOH—»C,H,.COOH

1 This Journal, v, p. 185 (Note added during proof correction).
? Beitr. z. chem. Physiol. u. Pathol., vi, p. 150, 1904.

3°23

304 Derivatives of Phenylpropionic Acid

The formation of cinnamoylglycocoll following the administra-
tion of phenylpropionic acid can apparently only be explained
on the assumption of a prior formation of phenyl-$-oxypropionic
acid or a derivative of it, which subsequently parts with a mole-
cule of water to give cinnamic acid or a derivative of cinnamic
acid:

C,H;.CH,.CH,.COOH->C,H,.CHOH.CH,.COOH->C,H,.CH :CH.COOH

It is a curious fact, however, that phenyl-§-oxypropionic acid
itself, administered to an animal, proves to be far more difficult of
combustion than phenylpropionic acid, the acid being excreted
largely unchanged and not combined with glycocoll. It would
therefore appear as if combination with glycocoll might be a
necessary preliminary to oxidation. To test this hypothesis it
was necessary to synthesize the glycocoll derivative of phenyl-
$-oxypropionic acid and to determine its behavior in the body.
The following paper contains an account of the synthesis of this
substance together with the synthesis of cinnamoylglycocoll and
some of its derivatives. The synthetic cinnamoylglycocoll was
found to be identical in every respect with the substance isolated
from the urine of animals that had received injections of sodium
phenylpropionate or sodium phenylvalerate.!

Cinnamoylglycocoll was prepared from cinnamic acid which
was converted in the acid chloride and the latter substance was
allowed to interact with glycocoll in the presence of caustic soda,
at a low temperature:

C,H,.CH :CH.CO.C1+ NH,.CH,.COOH =C,H;.CH :CH.CO.NHCH,.COOH
+HCl

Apart from its behavior as an unsaturated substance, cinnamoyl-
glycocoll resembles hippuric acid closely. On reduction with
sodium amalgam it is converted into phenylpropionylglycocoll.
Cinnamoylglycocoll on bromination in acetic acid solution is
converted into phenyl-a,@-dibromopropionylglycocoll which on
boiling with water is converted into phenyl-a-bromo-f§-oxy-pro-
pionylglycocoll and this substance on reduction with sodium
amalgam yields the desired phenyl-G-oxy-propionylglycocoll.

‘An account of the animal experiments will be published shortly.

Hed: Dakin 305

C.,H,.CH :CH.CO.NH.CH,.COOH—>C,H,.CHBr.CHBr.CONH.CH;.COOH
(cinnamoylglycocoll) (phenyl-a,3-dibromopropionylglycocoll) *
=? O,t.. CHOH.CHBr.CO.NHCH,.COOH—C,H,..CH(OH).CH,.CO.NH.
CH,COOH.
(Phenyl-a-bromo-f-oxy-propionylglycocoll) ,(phenyl-$-oxy-propionyl-
glycocoll)

Phenyl-8-oxypropionylglycocoll on heating with concentrated
hydrochloric acid, parts with the element of water to yield cinna-
moylglycocoll.! On further heating the latter undergoes hydro-
lysis with formation of cinnamic acid and glycocoll.

C,H,..CHOH.CH,.CO.NH.CH,.COOH->C,H,.CH :CH.CO.NH.CH,.COOH
+H,O—C,H,.CH :CH.COOH + NH,.CH,.COOH

Phenyl-a-bromo-f-oxypropionylglycocoll on boiling with con-
centrated hydrochloric acid is converted primarily into an
almost insoluble substance, which proved to be phenyl-a-bromo-
$-chloro-propionylglycocoll. On further heating with more dilute
acid phenyl-a-bromo-f-oxypropionic acid is obtained. These
reactions afford convincing evidence of the correctness of the
constitutions assigned to the various substances.

C,H,.CHOH.CHBr.CONH.CH,. COOH->C,H,.CHC1.CHBr.CO.NH.CH,,.
COOH-—>C,H,.CHOH.CHBr.COOH + NH,.CH,.COOH

EXPERIMENTAL PART.
Synthesis of Cinnamoylglycocoll.

Cinnamoylchloride was prepared in the usual way by acting
upon cinnamic acid with phosphorus pentachloride, distilling off
the phosphorus oxychloride and fractionating the residue in
vacuo. ‘The acid chloride quickly solidifies and melts at 35° to
36°. The cinnamoylchloride (16.6 grams) was melted and slowly
dropped into a flask containing 8.0 grams of glycocoll dissolved in
80 cc. of 10 per cent caustic soda solution. The flask was kept in
a freezing mixture and a few cubic centimeters of ether were
added in order to assist in dissolving the cinnamoylchloride.
The contents of the flask were shaken vigorously until all the

' This reaction is of interest since formation of cinnamoylglycocoll in
the body probably results from a similar change.

306 Derivatives of Phenylpropionic Acid

cinnamoylchloride had disappeared. The clear solution was then
acidified with sulphuric acid and the precipitate of crude cinna-
moylglycocoll was filtered off, washed with cold water, dried
and then washed with a little ether to remove a trace of cinnamic
acid. The cinnamoylglycocoll was recrystallized from boiling
water and separated in the form of long shining needles melting
sharply at 192° to 193°. The yield of recrystallized substance was
I5 grams, equivalent to about 75 per cent of the theoretical
amount.

ANALYSIS.

0.2005 gm. gave 0.01358 gm. N. = 6.77 percent N.
C,,H,,NO, requires 6.83 per cent N.

Cinnamoylglycocoll is sparingly soluble in cold water and
moderately soluble in boiling water. It is easily soluble in
alcohol and ethyl acetate but almost insoluble in dry ether, chlo-
roform and petroleum ether. It is a fairly strong acid and
readily forms salts. When dissolved in a little sodium carbonate
solution, it instantly reduces dilute potassium permanganate
and an odor of benzaldehyde is at once noticeable. On boiling
with strong hydrochloric acid it is reconverted into cinnamic
acid and glycocoll.

The synthetic cinnamoylglycocoll was identical in every way
with that isolated from the urine of animals that had received
injections of sodium phenylpropionate or sodium phenylvalerate
and the melting point of a mixture of the substances of different
origin was unchanged.

Reduction of Cinnamoylglycocoll.

Cinnamoylglycocoll, o.5 gram, was suspended in ro cc. of water
and treated with 20 grams of 2 per cent sodium amalgam.
After standing for a couple of hours in a warm place the solution
was precipitated with hydrochloric acid. 0.35 gram of phenyl-
propionylglycocoll, m.p. 114°, was obtained. The substance was
identical with that previously prepared from phenylpropionyl-
chloride and glycocoll.!

’ This Journal, iv, p. 431, 1908.

aD. Paki 307

ANALYSIS.

0.2865 gm. gave 0.01932 gm. N(Kjeldahl) = 6.74 percent N.
C,,H,,NO, requires 6.76 per cent N. a)

‘Synthesis of Phenyl-§-oxy-propionylglycocoli.i . ;

PHENYL-a@,3-DIBROMOPROPIONYLGLYCOCOLL. Cinnamoylglycocoll
(10.25 grams) was dissolved in 60 cc. of warm glacial acetic acid.
The solution was cooled to a point just short of crystallization and
then bromine (8.0 grams) dissolved in 15 cc. of glacial acetic acid
was added fairly rapidly. While the bromine was being added,
the liquid was well shaken and cooled under the tap. The bromine
was rapidly absorbed and after a few moments the solution was
diluted with ice water and the precipitated phenyl-a,@-dibromo-
propionylglycocoll filtered off and washed with cold water. The
yield is practically quantitative. The substance when heated
rapidly melts with complete decomposition at 190° to 191°. If
heated slowly it decomposes at a slightly lower temperature.
It is very sparingly soluble in ether and in cold water and insolu-
ble in carbon bisulphide, chloroform and petroleum ether. It is
readily soluble in glacial acetic acid and crystallizes in hard, shin-
ing prisms.

ANALYSIS.
0.2045 gm. gave 0.2134 gm. AgBr = 44.4 per cent Br.

0.2621 gm. gave 0.00987 gm. N (Kjeldahl) = 3.77 percent N.
C,,H,,Br,0,N requires 3.84 per cent N and 43.7 per cent Br.

PHENYL-a-BROMO-§-OXYPROPIONYLGLYCOCOLL. Phenyl-a, (-di-
bromopropionylglycocoll readily parts with one atom of brom-
ine when boiled with water.1 Five grams of the substance were
boiled with 75 cc. of water under a reflux condenser until all the
acid had dissolved. The solution was then extracted with ether
in a continuous extraction apparatus for three hours. On eva-
poration of the ether an almost quantitative yield of phenyl-a-
bromo-f-oxypropionylglycocoll was obtained. It crystallizes
in needles and is readily soluble in water, alcohol and ether. It
may conveniently be recrystallized from water and melts at
87° to 88°.

1 An estimation of the amount of bromine liberated as hybromic acid on
boiling with water gave 22.7 percent Br, theory demanding 21.92 per cent.

308 Derivatives of Phenylpropionic Acid

ANALYSIS.

0.2274 gm. gave 0.1380 gm. AgBr = 25.82 per cent Br.

0.1672 gm. substance dried at 70° gave 0.00749 gm. N (Kjeldahl) = 4.49
per cent.

0.2194 gm. substance dried at 70° gave 0.01008 gm. N (Kjeldahl) = 4.59
per cent.

C,,H,,0, NBr requires 4.63 per cent N and 26.48 per cent Br.

Phenyl-a-bromo-f-oxypropionylglycocoll when covered with
concentrated hydrochloric acid and gently warmed under a
reflux quickly dissolves and after a few moments a copious separa-
tion of a fine crystalline substance takes place. This substance
is very sparingly soluble in water and crystallizes in hard, highly
refractive hexagonal prisms which melt at 203° to 204° with
complete decomposition. On analysis the substance proved
to be phenyl-a-bromo-{-chloropropionylglycocoll, the alcoholic
hydroxyl group of the oxy-acid having been replaced by chlorine.

ANALYSIS,

0.1565 gm. gave 0.1602 gm. AgCl, AgBr.

Calculated for C,,H,,O,NCIBr = 0.1622 gm.

0.1896 gm. gave 0.0084 gm. N (Kjeldahl) = 4.43 per cent.
C,,H,,0O,NC1Br requires 4.38 per cent.

On boiling with water the substance slowly dissolves with libera-
tion of hydrochloric acid and phenyl-a-bromo-$-oxypropionyl-
glycocoll, m.p. 87° to 88°, crystallizes out on cooling. The sub-
stance is almost insolublein ether but readily dissolvesin alkaliand
its aqueous solution does not decolorize alkaline permanganate in
the cold.

If instead of filtering off the insoluble precipitate, of phenyl-a-
bromo-f-chloropropionylglycocoll the boiling with hydrochloric
acid is continued, the precipitate slowly dissolves, especially if the
acid be diluted, and eventually oily drops appear. On cooling the
oil drops solidify and on recrystallization from a mixture of
chloroform and petroleum, phenyl-a-bromo-f-oxypropionic acid
is obtained in the form of crystals melting at 125° to 126°.

PHENYL-§-OXYPROPIONYLGLYCOCOLL. The remaining bromine
atom in phenyl-a-bromo-f-oxypropionylglycocoll is readily re-
placed by hydrogen when treated with sodium amalgam in

ie DS Dakin 309

faintly acid solution. The substance (5.0 grams) was dissolved in
twenty parts of water and treated with three times the theoret-
ical amount of 2.5 per cent sodium amalgam. The solution was
kept acid by the occasional addition of a little sulphuric acid.
After some hours the solution was strongly acidified with phos-
phoric acid and extracted with ether containing 10 percent alcohol
in a continuous extraction apparatus. On distilling off the ether
an almost quantitative yield of phenyl-$-oxypropionylglycocoll
was obtained. The substance is easily soluble in water and
alcohol and crystallizes from a mixture of alcohol and chloroform
or from water in star-shaped aggregations of needles, m.p. 146°
(0p be Oy a

ANALYSIS.

0.1402 gm. gave 0.00875 gm. N (Kjeldahl) = 6.23 per cent N,
C,,H,,0,N requires 6.28 per cent N.

Phenyl-$-oxypropionylglycocoll dissolved in sodium carbonate
solution gives benzaldehyde when gently warmed with potas-
sium permanganate. On heating to boiling with strong hydro-
chloric acid and at once cooling, a precipitate of cinnamoylglyco-
coll is obtained, which crystallizes from water in long needles and
melts at 192° to 193°. On prolonged heating with hydrochloric
acid cinnamic acid, m.p. 133° is obtained.

ie
;
UR) i
+ es
J ‘
. +
\ a)
\
ye" 1
s
;
.
\
ry
i Pat!
t) 4%
fay
i 1 1g
ibm
\
} Dk ¥ } }
a) } 1
haya Wi
; bent.
Rabe | A
Pi ; el
: 7 : ha By
: Uv ' Pres fie or.
mY i pf aay se oes
Dar eee
gered

CHEMICAL STUDIES IN CYTOLYSIS.
By ALONZO ENGLEBERT TAYLOR.

(From the Hearst Laboratory of Pathology, U niversity of California.)

(Received for publication, October 10, 1908.)

The phenomenon of spermatolysis, first described by Metchni-
koff, offers in many respects favorable opportunities for the study
of the details and variables of the reaction of cytolysis. The
following report represents the results of an investigation designed
to elucidate the nature of the active factor or constituent in the
spermatozoén which, on inoculation into an appropriate animal,
gives rise to the formation there of the antibody which acts as
the spermolytic agent when the blood of the immunized animal
is mixed with the spermatozoa under consideration. When the
living spermatozoa (or fresh dead sperm) of the salmon (Oncor-
hynchus Quinnat) are injected into a rabbit, the blood of the
animal acquires the power of causing cytolysis of salmon Sper-
matozoa. The reaction of cytolysis, under the microscope, is
resolved into two stages. First the spermatozoa swell; later
the cell membrane seems to dissolve or burst, and the proto-
plasm becomes converted into a detritus. The first step, swel-
ling, is not essential to the cytolysis. Whenever these cells are
placed in a salt solution of the concentration of the serum of
the rabbit (that of the spermatic fluid is much lower in osmotic
pressure) they swell. If the serum be an immune one, in addi-
tion to the swelling, the cell wall dissolves or bursts. This
swelling of these cells when placed in salt solution is not a
phenomenon peculiar to these cells, but is one of widespread
occurrence. It can be demonstrated in a striking manner,
macroscopically, by placing ripe salmon eggs in a bottle in a
Strong solution of sodium chloride. It might at first be supposed
that under such circumstances salt would pass into the eggs and
water pass out, that a shriveling of the eggs would occur; in
fact, the eggs swell. This swelling is in all probability due to

311

B12 Chemical Studies in Cytolysis

the passage of both salt and water into the eggs, the membrane
does not act as the dividing membrane in an osmosis cell. The
swelling of the spermatozoa in a serum, normal or immune, has
therefore no relation to the cytolysis. This can be proved
directly by mixing the active serum with three parts of distilled
water, when the suspended spermatozoa will undergo cytolysis
without swelling.

To what is this cytolysis of the spermatozoa due? There
can be no purpose in a review here of the now enormous literature
on the subject of specific cytolysis. Two generalized views are
current; one that it is due to a disturbance of the complex
lipoid-protein in the cell membrane, whereby the integrity of
the same is destroyed; the second is that it is due to a disturbance
of the complex lipoid-protein in the protoplasm. That the
lipoid fraction is essential to the physical constitution of the
protoplasm as well as of the cell membrane seems quite certain.
It would be easy to say that in cytolysis the colloidality of the
protoplasm or membrane is reduced or destroyed; but little
would be gained in the concrete sense by such an expression.
As is well known, a fractionation of red corpuscles has led to the
conclusion that it is the lipoid fraction which on injection gives
rise to the activation of the serum of the animal receiving the
injections. Since salmon spermatozoa may be obtained at the
hatcheries in large quantities, one may attempt upon a large
scale the fractionation of these cells in the chemical sense; and
attempt to identify and isolate the substance or fraction that
on injection into the rabbit causes the formation of the anti-
body there.

Something over two kilos of salmon sperm was submitted
to such a fractionation. In view of the known chemical compo-
sition of the spermatozoa, due largely to the researches of
Miescher and Kossel, the cells were divided into four fractions:
(a) the protamin; (b) the nucleinic acid; (c) the lipoids soluble
in ether by simple extraction; and (d) the lipoids soluble in

ether after the digestion of the residue of (c) with trypsin. ©
Rabbits were inoculated for three weeks with ascending doses —
of a, b, cand d and with mixtures of c and d suspended in isotonic ~

salt solution; also with fresh spermatozoa. Protamin is quite
toxic to rabbits; the nucleinic acid causes a leucocytosis, but is

AE he

Alonzo Englebert Taylor a3

otherwise not toxic in ordinary doses; the lipoids are not toxic
in ordinary doses. Fresh spermatozoa of the salmon are sur-
prisingly toxic to rabbits, and several deaths were due to these
injections. This toxicity is not a matter of salts, isotonicity
or mass of the injection fluid; it is resident inthesperm. The
sera of the several inoculated animals were collected in the
usual manner, conserved upon ice and taken to the hatchery
and tested upon matured freshly recovered spermatozoa.

Except in the case of the sera from rabbits inoculated with
the spermatozoa the results were entirely negative. The active
sera obtained from the animals inoculated with the sperm were
not, however, highly active; a dilution of 10 : 1 was sufficient to
obliterate the cytolytic action. Curious is the fact that unless
the spermatozoa be fresh, no cytolysis occurs. If the sperm
has been out of the body of the fish for an hour, the spermatozoa
can no longer be cytolyzed by a serum that is effective with the
freshly removed spermatozoa of the same fish. One may infer
that within this hour a coagulation-necrosis has occurred within
the cells. Entirely inactive, as stated, were the sera obtained
from the animals inoculated with the lipoidal, protamin and
nucleinic acid fractions; swelling of the spermatozoa occurred,
but no cytolysis.

It cannot be taken as proved by these experiments that the
activation of a serum is not due to a chemically definable sub-
stance in the cells employed; that it is, to use a poor but expres-
sive phrase, not a chemical but a biological fraction that pro-
duces the antibody in the animal receiving it. It is possible
that the period of inoculation was not long enough. This is,
however, very unlikely, since the period of inoculation with the
sperm, which was effective, was much shorter and concerned
much smaller quantities. It might be that the materials were
chemically alteredin the course of the manipulations of isolation.
This could surely not hold for the protamin and the nucleinic
acids; it might, however, have been true for the lipoids. Lastly it
might be assumed that not one substance alone is needed, but
the interaction or combined action of a second substance is
necessary. Thus, it might possibly be not the lipoid but the
complex lipoid-protein that is necessary. Speculation upon this
matter can lead to no result. This experiment proves only one

314 Chemical Studies in Cytolysis

thing: that the ether-soluble fraction of the spermatozoa of
the salmon, like the protamin and nucleinic acids, has not the
power of causing in the inoculated animal the formation of an
anti-body capable of producing cytolysis of the spermatozoa
of the same fish.

To Mr. E. M. Ball, director of the Battle Creek Hatchery,
California, I express my gratitude for permission to carry on
the work and for the many acts of courtesy and assistance
extended to me.

ON THE CONVERSION OF GLYCOGEN INTO SUGAR IN THE
LIVER.

By ALONZO ENGLEBERT TAYLOR.

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

(Received for publication, October 10, 1908.)

The older literature of physiology contains many papers deal-
ing with the relation of the liver to the conversion of glycogen
into sugar. The facts that sugar is known as a simple chemical
reaction to be formed from glycogen, that the carbohydrates of
the digestion pass on resorbtion into the portal circulation and
not into the lacteal system, and that the liver contains glycogen
in amounts that are proportionately related to the carbohydrates
of the diet were sufficient to convince Claude Bernard that glycoly-
sis is a function residing in the liver. The fact that analytically
it has been difficult to demonstrate a greater sugar content in
venous than in the arterial or portal blood (of the liver) proves
nothing when it is realized that if in a dog of 10 kilos weight
Ioo grams of glycogen were daily converted into sugar in the
liver, the amount of sugar passing out of the liver through the
hepatic vein in each second of time would be a little over 1
milligram. The chemical demonstration of the relations is
impossible in the higher animals. The reaction may, however,
be easily studied upon invertebrate material, as the experiment
reported in the following lines will illustrate. An obvious
assumption in the argument is that facts determined upon such
simple material are held to be validin mammalian physiology.

The large clam of the Pacific Coast (Schizotherus Nutallit)
presents in the liver amounts of glycogen that may run as high
as 8 per cent of the dried residue. The amount of sugar that is
to be found in the same fresh material is low. These facts are
known to hold only for the material collected at the very low
tides, after the period of feeding. The livers of some twenty of
these large bivalves were carefully freed of the other structures,

325

216 Conversion of Glycogen into Sugar

ground to a pulp, and an adequate amount of toluol added to
prevent bacterial action. A known amount of the total mass
was then placed in strong alcohol in the proportions of 1 : 3,
and later subjected to analyses for sugar, glycogen and total
nitrogen. From these figures the original substrate of the
reaction experiment (glycogen) and the product (sugar) could
be calculated for the total reaction system. The nitrogen was
determined as a basis for the calculation of the amounts removed
for the analyses of sugar at the stated times. The reaction
material was placed in a jar and suspended into the water of a
deep well whose temperature (12.4°) was nearly constant day
and night, as determined by the thermometer introduced with
the jar containing the material. From time to time known
quantities were removed, mixed with three parts of alcohol,
and later analyzed for total nitrogen, sugar and glycogen. With
a thick mass of this kind, it is not possible to measure accurately
the quantities removed, but by the determination of the total
nitrogen and the comparison of this figure with the figure for
the total determined in the beginning of the experiment, the
exact amount of material removed for analysis could be estimated.
The sugar was estimated by the gravimetric method of Pfli-
ger, the copper being weighed as the reduced metal; the glycogen
was estimated by the last method of Pfluger; the nitrogen
was estimated by the method of Kjeldahl. The data contain
obviously the desiderata of a’ quantitative experiment. The
original total mass was known, the original total nitrogen,
glycogen and sugar. The several parts removed for analyses
at the stated times were known and the sugar in them deter-
mined by an accurate method. The total concentration of the
system was constant, constant also the temperature. Culture
tests done at the close of the experiment showed the material
to be sterile. From the analytical data the following constants
were obtained by the use of the common equation representing
the progress of a monomolecular reaction:

dz
—=C(A—2z
a ( )

in which A and x stand respectively for the substrate and the

a

Alonzo Englebert Taylor 207

product of the reaction—the glycogen and the sugar. The times

given are in hours.

T= 4 8 12 16 26 28 32 38
C(X 10“) = 70 79 74 48 66 58 58 55

From these constants it is obvious that corresponding to the
gradually decreasing figures for the constants, the curve of reac-
tions falls away somewhat from that representing accurately a
monomolecular reaction. This fact, however, is exactly what
is found in the digestion of starch or glycogen by amylase im
vitro; and the figures obtained in this experiment deviate no
more than these from the normal curve of a monomolecular
reaction. This slowing of the reaction is in all probability due
to a gradual inactivation of the ferment through hydrolysis;
in terms of physical chemistry, the constants fall because the
mass of the catalysor is reduced.

This experiment represents what is commonly called a post-
mortem digestion. Since glycogen diffuses with the greatest
difficulty, it is certain that the reaction occurs in the liver cells
and not in the intercellular fluids. A postmortem digestion
differs from the reaction zm vivo, so far as known, simply in this:
in the living, fasting animal, the sugar is removed from the
liver as fast as formed, the mass of the product of the reaction
is therefore practically nil; the mass of ferment is subject to
variations through new formation by the liver cells. In the
experiment 7m vitro, it can be shown that the mass of sugar
present under the conditions of the experiment is of practically
no action in retarding the progress of the reaction. To what
extent variations in the mass of glycolytic ferment in the liver occur
within the same period of time in a fasting animal, is not known.
Under the circumstances, however, we are warranted in assuming
a close analogy between the postmortem and the living hydrolysis
of glycogen in the liver; and the conclusion is thus reached, that
in the living animal the conversion of glycogen into sugar in the
liver is a function of two variables and proportional directly
to them, namely, the masses of glycogen and of glycolytic ferment
in the liver cells.

ON THE ANTAGONISM OF ALCOHOL TO CARBOLIC ACID.

By ALONZO ENGLEBERT TAYLOR.

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

(Received for publication, October ro, 1908.)

It is a common practice with surgeons to apply ethyl alcohol
to wounds to which carbolic acid has been previously applied in
order to check the prolonged action of the carbolic acid. Alcohol
has also been employed as an antidote to the internal action of
carbolic acid. A review of the chemical properties and relations
of the two substances gives no indications of the possible nature
of the assumed antagonism. It is obviously possible that the
relationship may be either chemical or physical. The suggestive
investigations of Sollmann, published a year ago, tended to
indicate that physical factors were largely concerned and might
possibly explain the facts entirely. It is possible, however, to
test experimentally the hypothesis of a chemical antagonism,
and such an investigation forms the subject of the present com-
munication.

Yeasts are all more or less sensitive to the presence of alcohol
in their culture media. For this reason fermentation ceases in
many wines before the sugar contained in the grape-juice has
been wholly converted into alcohol and carbon dioxide. Not
only are the living yeast cells affected by higher concentrations of
alcohol; the activity of the isolated ferment, the zymase, is also
depressed by the presence of more than minimal amounts of
alcohol. In this respect, however, yeasts vary. Through the
kindness of Dr. Arthur Lachmann I have come into the posses-
sion of a wine yeast that ferments sugar in the presence of 15
percent of alcohol. In a solution of 10 per cent of alcohol this
yeast ferments sugar quite as actively as in the absence of alco-
hol. Since this yeast is very sensitive to the antiseptic action of
carbolic acid, it presents a material for direct experimentation
on the relations between alcohol and carbolic acid. In the ulti-

319

320 Antagonism of Alcohol to Carbolic Acid

mate interpretation of the experimental results, the argument
obviously assumes a complete analogy in this regard between
the yeast cell and the animal cells. If alcohol be held to detoxi-
cate carbolic acid in its action upon animal cells, we assume that
it should diminish the toxic, that is, the antiseptic action upon
yeast cells. If the carbolic acid were in any way chemically
detoxicated by the alcohol, its antiseptic property ought to
display a corresponding reduction. Itis, of course, well known
that plant and animal cells display widely varying reactions to
chemical substances, dependent in many instances upon varia-
tions in oxidation—and reduction—ferments. In the system-
atic investigations of the reactions of chemical substances
in plant and animal cells, any parallelism could in no wise be
assumed. In the present instance, however, dealing with a
substance like carbolic acid, possessing highly toxic action upon
both plant and animal cells, it may be fairly assumed that the
modus operandi of this toxic action (and ipso facto of its de-
toxication) is the same in both instances.

The experiment calls for a series of fermentation tests with
varying concentrations of carbolic acid (from 1 : 1000 to 1 : 500,-
ooo) mixed with standard solutions containing to per cent of
alcohol, 2 per cent of glucose, 0.5 per cent of peptone and the
salts commonly added to the culture media employed with yeasts.
The velocities of the fermentations, measured in terms of carbon
dioxide, were compared under constant conditions of control
with those displayed in exactly similar tests in the absence of
carbolic acid. The experiments were repeated with variations
in the temperature and in the contents of sugar, alcohol and
yeast—always with the same results.

The results were entirely negative to the idea of a chemical
detoxication of carbolic acid by alcohol. Alcohol does not
reduce in the least the antiseptic action of carbolic acid. The
yeast used in these experiments was sensitive to as high dilutions
of carbolic acid as 1 : 100,000; at 1 :10,0o00 fermentation was
practically inhibited. This marked action of carbolic acid was
in no wise retarded or inhibited by the presence of 10 per cent
of alcohol. The presence in the media of soluble sulphates
reduced appreciably the toxicity of the carbolic acid; but alcohol
in no concentration had any such effect.

Alonzo Englebert Taylor 321

On the contrary, with high concentrations of alcohol and low
concentrations of carbolic acid, the alcohol seemed to increase
to some extent the toxicity of the carbolic acid. A 1 : 100,000
concentration of carbolic acid was more toxic in the presence
of 10 per cent of alcohol than without it. This is in all prob-
ability an expression of the action of the alcohol upon the outer
membrane of the yeast cells, rendering it more permeable to the
entrance of the carbolic acid. The presence of as much sodium
chloride as the yeast is known to tolerate alone, heightened the
toxic action of the carbolic acid. This is, however, an expres-
sion of the law of partition; the presence of the sodium chloride
renders the culture medium less of a solvent for carbolic acid,
which has the effect of increasing the relative solubility of car-
bolic acid in the protoplasm of the cells. The presence of
alcohol in the culture medium renders it a better solvent for
carbolic acid, which would tend to reduce its relative solubility
in the cell protoplasm, thus lowering its toxicity; since the con-
trary result in toxicity is observed, we are warranted in inter-
preting this as the effect of the alcohol upon the membrane of
the yeast cells, whereby it is rendered more permeable to car-
bolic acid.

These experiments tend therefore to prove that there is no
chemical detoxication of carbolic acid by ethyl alcohol, and
that the effect observed in therapeutic practice must rest upon
some physical basis. With this conclusion the more recent
investigations of Sollmann are in full accord.

ary
t .
bd

THE BLOOD CLOT OF LIMULUS POLYPHEMUS.

By C. L. ALSBERG anp E. D. CLARK.

(From the United States Bureau of Fisheries Laboratory at Woods Hole,
and the Department of Biological Chemistry of the Harvard Medical School.)

(Received for publication, August 22, 1908.)

As far as we are aware no analytical investigation of the com-
position of invertebrate blood cells has been published; and we
know of no efforts to isolate the substances of which such cells
are composed. The present research was begun as an attempt
to undertake such studies upon favorable material. It was
thought that Limulus would prove excellent for the purpose
because of the ease with which considerable quantities of blood
may be procured from this animal, and also because, as has been
shown by L. Loeb,! there is but one kind of cell in the blood.
However, we soon abandoned the attempt to study unchanged
Limulus blood cells because of the difficulty of preventing large
quantities of blood from clotting. Coagulation may be prevented
by the methods of both Halliburton”? and L. Loeb. These meth-
ods necessitate the working up separately of small quantities
of blood and involve more labor than we were able to give. In
the light of our present knowledge such a study of unaltered cells
seems worth this labor, and, it is hoped, will be attempted. In
the present investigation only the clot was studied. Halli-
burton! believed that material from the plasma entered into the
clot in addition to that derived from the cells. Howell,’ on the

1 Folia Hematologica, iv, May, 1907.

? Journ. of Physiol., vi, pp. 300-335, 1885.

3L. Loeb: Ueber die Koagulation des Blutes einiger Arthropoden,
Beitr. z. chem. Physiol. u. Pathol., v, p. 192; Vergleichende Untersuch-
ungen tiber die Thrombose, Arch. f. path. Anat. u. Physiol., clxxxv,
1906; Studies on Cell Granules and Amceboid Movements of the Blood
Cells of Limulus, Univ. of Pa. Med. Bull., May, 1905.

# Loc. ‘cit.

5 Johns Hopkins University Circ., v, p. 4, 1885.

323

324 Blood Clot of Limulus

other hand, believed, upon purely morphological evidence, that
the cells alone form the clot. Microscopical appearances, how-
ever, can not be regarded as capable of settling such a question
as this. Experimental evidence is needed, and this has been
furnished by L. Loeb? who, by showing that fibrinogen cannot
be prepared from Limulus blood and that the plasma deprived
of its cells cannot be made to clot, has established the fact that
the clot of Limulus is formed of the agglutinated cells. We
have not specially concerned ourselves with this question. We
have made no attempt to prepare fibrinogen; but numerous
incidental observations as well as our inability to obtain unmis-
takable fibrin from the clot bring our results in harmony with
Loeb’s finding that fibrinogen is absent.

We were also ableto observe the apparent secondary coagula-
tion described by Loeb” We have not yet completed our inves-
tigation of the material involved in this phenomenon. Our
studies, as far as they have at present proceeded, give no indica-
tion that a formation of fibrin takes place, and therefore support
Loeb’s interpretation of this phenomenon.

Our present paper deals merely with the chemical composition
of the insoluble material of the clot. Following Loeb, we propose
to call it cell-fibrin.

The clot is composed of tough elastic fibrous masses, which,
when freed from adherent serum, are white or slightly yellowish.
The clot itself does not reduce Fehling’s solution, but does so
powerfully after hydrolysis with mineral acids. The serum does
not behave in this way. L. Loeb’ made a simple chemical study
of the clot, with which our findings are consistent. He found
that the clot swellsinalkalies but not in acids. It is coagulated
by heat. After long boiling with 5 per cent potassium hydroxide
he was unable to get a reduction of cupric oxide. This is all
quite consistent with our findings that reduction takes place
only after inversion. Loeb concludes that it isnot mucin. This,
as will appear, is also our own conclusion.

We prepared the insoluble protein of the clot, the cell-fibrin,

1 Beitr. 2. chem. Physiol. u. Pathol., vi, p. 279; Arch. f. path. Anat. u.
Physiol., clxxiii, p. 48.

? Beitr. z. chem. Physiol. u. Pathol., v, p. 194.

8 Arch. f. path. Anat. u. Physiol., clxxiii, p. 37.

C. L. Alsberg and E. D2 Clark 325

as follows: The pieces of clot after thoroughly washing with 5
per cent sodium chloride were ground to a fine pulp in a porce-
lain ball-mill with 5 per cent sodium chloride. The salt solution
was removed by straining through raw silk, and the extraction
repeated. The material was then ground up repeatedly with
distilled water. It was quite insoluble in distilled water, salt
solutions and dilute acids; but slightly soluble in alkaline car-
bonates, and readily soluble only in caustic alkalies. It was,
therefore, ground up in the mill with 2 per cent sodium hydrox-
ide, filtered, and the clear filtrate precipitated by means of an
excess of acetic acid. This precipitation was repeated. The
acetic acid did not completely precipitate the protein. After
removal of the acetic acid precipitate, a little more could be
precipitated from the clear filtrate by hydrochloric acid. Whether
this precipitate is identical with the acetic acid precipitate could
not be decided for lack of material.

The cell-fibrin prepared in this way still gave a good reduction
with Fehling’s solution after hydrolysis with mineral acid. It
also gave the Molisch-Udransky and the orcin reactions. We
therefore assumed it to be a mucoid. On analysis, however, it
proved to be poor in sulphur (0.62 per cent). In a mucoid con-
taining chondroitin-sulphuric or glucothionic acid more sulphur
is to be expected. Moreover the small sulphur content pointed
to a very large molecular weight. Inasmuch as this protein
does not exist preformed in the blood, but must be formed in the
process of clotting, it is probably a decomposition product of a
more complex substance of even greater molecular weight. These
considerations led us to suspect the purity of our preparations
and to endeavor to purify them further. After repeated solu-
tions in dilute sodium hydroxide and precipitation with acetic
acid, the power to reduce Fehling’s solution, or to give the Molisch-
Udransky and orcin reactions disappeared. The substance
which is responsible for these reactions will form the subject of
a future communication. The preparations were then washed
salt-free, treated with alcohol and ether, and dried in the
Schmiedeberg drying apparatus 7m vacuo over sulphuric acid at
ma °C.

326 Blood Clot of Limulus

Preparation I gave the following analytical results:

0.2097 gm. of substance gave 26.8 cc. dry N at 17° and 768.5 mm.:
N = 15.21 per cent.

0.2163 gm. of substance gave 27.4 cc. dry N at 17° and 768.5 mm.:
N = 15.08 per cent.

0.2466 gm. of substance gave 0.1565 gm. H,O : H = 7.10 per cent.

0.2129 gm. of substance gave 0.3771 gm. CQ, and 0.1336 gm. H,O :
C = 48.31 per cent; H = 7.02 per cent.

0.2288 gm. substance gave 0.4039 gm. CO,; C = 48.14 per cent.

0.4692 gm. substance fused with Na,O, according to the method of
Folin, gave 0.0213 gm. BaSO, :S = 0.61 per cent.

These figures do not represent the true percentage composition
because unfortunately no ash determination was made and they
can not be recalculated for the ash-free substance. It was not
supposed that after the frequent reprecipitations and thorough
washing appreciable amounts of ash could still be present. Such
is, however, the case as an ash determination on Preparation II,
which had been dissolved in sodium hydroxide and precipitated
oftener than Preparation I showed.

0.2429 gm. of substance yielded o.o107 gm. ash or 4.3 per cent. The
ashing was done in platinum, and after the weight had been recorded the
ash was boiled out with hydrochloric acid. Very little went into solution,
the bulk of the ash consisting of silica. There was present also a little
calcium but no magnesium, iron, copper or phosphorus. Whether the
silica is present in organic combination with the protein or is merely an
accidental contamination from the mill is a question which will be sub-
jected to further investigation.

The analytical figures obtained for Preparation II checked
fairly well with those obtained for No. 1, though they were a
little lower, pointing to a little greater ash-content for No. II.
They are as follows:

0.1981 gm. substance gave 24.9 cc. dry N at 18° and 770 mm.: N =

14.94 per cent.
0.1985 gm. substance gave 0.3495 gm. CO, and 0.1229 gm. H,O: C =

48.02 per cent; H = 6.9 per cent.
Recalculated for the ash-free substance these figures become: C, 50.15;

Hy. 7.2; N,. 15.60. per cent.

In order to verify the sulphur determination a third Prepara-
tion was used which had been precipitated once more than Pre-

C. L. Alsberg and E. D. Clark 2907

paration II. As the sulphur content was so low a great deal of
material was used to minimize the errors of the method.

1.2666 gm. substance yielded 0.0516 gm. BaSO,: S = 0.55 per cent.
This preparation contained 4.2 per cent of ash. The sulphur content of
the ash-free substance then becomes 0.57 per cent.

These results were controlled by a determination in which the
oxidation was carried out in silver by fusion with a mixture of
sodium hydroxide and potassium nitrate. The result did not
check up very well, but was quite a little lower than in either of
the others.

To summarize, the substance has the following characteristics:
It is soluble readily only in caustic alkalies, and precipitated
from its solutions by acids, not being soluble in an excess. It
gives the biuret reaction well; Millon’s very faintly, if at all;
Hopkins-Cole weakly but distinctly. Molisch’s reaction and the
orcin reaction are negative. Boiled with an alkaline lead solu-
tion it fails to give more than a faint trace of black color. Hence
there is no evidence that it is in any way related to fibrin. It
contains little or no tyrosin (negative Millon reaction) and is
poorer in carbon, nitrogen and sulphur than fibrin, though
richer in oxygen. It is of course possible that the low nitrogen
and sulphur content is due to the treatment with alkali. To
test this question we subjected some fibrin, prepared by defibri-
nating ox blood and washing, to exactly the same treatment. It
was dissolved in alkali of the same strength in a ball-mill and

the alkali allowed to act for the same length of time. It was
precipitated with acetic acid and redissolved in alkali. At this
point its behavior was quite different. The Limulus material
Went into resolution with difficulty, the fibrin with ease. The
fibrin was reprecipitated and washed salt-free and analyzed. It
contained 1.22 per cent of sulphur and 15.85 per cent of nitro-
gen (Kjeldahl). Evidently, therefore, if the Limulus material
Owes its low sulphur content to the treatment with alkali, it
must be derived from some other material than ordinary mamma-
lian fibrin. Whether the fibrin of invertebrates would behave
in the same way on solution in alkaliis, of course, unknown. We
ought really to have used such a fibrin, but as nothing is known
about it, it would have involved an independent research such

328 Blood Clot of Limulus

as we hope to undertake at another time. Finally, it must be 7

borne in mind that we have no other guarantee than the fairly
constant percentage composition that we are dealing with a pure
substance.

Our substance is clearly an albuminoid and its low sulphur
content as well as the negative or low tyrosin content point as
clearly to a relationship with the glutin or elastin group. It
differs from both in having a larger oxygen content, about 26 per
cent. Elastin has a little less than 22 per cent: collagen nearly
25 per cent; and the collagen derivative, gelatin, about 25.25
per cent. As far as the oxygen content is concerned, our sub-
stance, therefore, is nearer to the glutin group. Its sulphur
content, which is higher than that of elastin, also points to its
belonging among the glutins. Its solubilities also fit in better
with the glutins. It is not so insoluble as elastin. It is, to be
sure, more soluble than coliagen itself, and far less soluble than
gelatin; but there is one member of the glutin group which has
somewhat similar solubilities. This is the glutolin discovered by
Faust! in the blood serum of the horse. This substance has
about the same sulphur and hydrogen content, though about 1
per cent more carbon and 2 per cent more nitrogen. Only in its
low nitrogen content does our substance differ very profoundly
from glutolin. In spite of the fact that it contains 4 per cent
less carbon than elastin, the C:N ratio of our substance (37.5 :10)
is almost the same as that of. elastin and quite different from
that of glutolin (34:10). In this respect our substance is near
to fibrinoglobulin or fibrinogen, though there are no other signs”
of relationship, and many totally dissimilar properties (among
others C:S ratio).

The C:N ratio of glutolin itself is, as Faust pointed out, nearer
to that of the ordinary proteins than to that of the other members
of the glutin group; and he, therefore, gave it an intermediary
position between the other glutins and the ordinary proteins.’

The glutins can be arranged in a series according to their C: » | |

ratios beginning with gelatin, with the highest nitrogen content,
and ending with glutolin, with the lowest nitrogen content;

1 Arch. f. exp. Path u. Pharm., xli, p. 309ff.
2 Loc. Cit: Pp. gene

CE. Alsberg and ‘E> Di'Clark 329

collagen and conchiolin occupying an intermediary position.!
Our substance would be still more extreme even than glutolin,
having both a lower nitrogen content and a higher oxygen con-
tent. In this respect it is nearer the ordinary proteins. It,
nevertheless, seems to belong rather to the glutin (or elastin)
group, as its insolubility, its low sulphur content, its negative

Millon reaction, clearly indicate. Definitely to settle its position
will, however, require a study of its component amino-acids, a
task that will be undertaken with more material in the future.
It is hoped that other properties not hitherto studied because of
lack of material will then be investigated.

If we are in fact dealing with a member of the glutin group,
it is not many times before that such a substance has been iso-
lated from an invertebrate and characterized chemically. Ac-
cording to v. Furth’ collagen has never been convincingly demon-
strated in invertebrates. It is interesting to note that an animal
which, like Limulus, has the matrix of its skeleton composed of
chitin instead of collagen, perhaps contains a glutin substance.
It will be interesting to investigate, as we propose to do, whether
the tendons, etc., of this animal contain a glutin, andifso, whether
it resembles the cell-fibrin we have described.

It is, moreover, interesting that the cells of the blood of inverte-
brates have many points in common with the platelets of higher
forms. Whether the similarity extends to the chemical com-
position is a question that demands investigation. Perhaps
agglutination thrombi are composed of a material like’ Limulus
cell-fibrin. It is not beyond the range of possibilities, judging
by the points of resemblance between Limulus cell-fibrin and
Faust’s glutolin, that the latter may be derived from the plate-
lets. We, therefore, propose to study the platelets from this
point of view.

"Ch Faust: Loc. cit., p. 319.

* Chemische Physiologie der niederen Tiere.

$ Deckhuyzen: Anat. Anz., xix, 1906; L. Loeb: Bettr.z. chem. Physiol.
u. Pathol., v, p. 197; Folia hematologica, iv, May, 1907.

THE EFFECT OF DIET ON THE MALTOSE-SPLITTING
POWER OF THE SALIVA.

By CHARLES HUGH NEILSON anp M. H. SCHEELE.

(From the Physiological Department of St. Louis University.)
(Received for publication, July’ 31, 1908.)

The most important enzyme of the saliva is ptyalin. Its
action according to the best authorities is the splitting of starch
into maltose. Some investigators say it also acts on maltose,
splitting it into dextrose. Not much stress is laid on the sup-
posed presence of a second enzyme in the saliva, namely, maltase.
In several of the best text-books on physiology, however, it is
mentioned as being present in the saliva.

It occurred to us that the maltase in the saliva might be
increased on a carbohydrate diet and decreased on a protein
diet. Neilson and Lewis! have shown an increase in the amylo-
lytic power of human saliva on a carbohydrate diet and a decrease
on a protein diet. The object of this paper is to determine
whether a similar change may be found in the action of saliva
on splitting maltose.

The ‘‘adaptation” of the digestive glands to the character of
the food has received considerable attention in recent years.
Wassilief? and Lintwarew? have shown that in dogs the quality
of the pancreatic juice as well as the quantity of the enzymes is
dependent on the food of the animal. Ellinger and Cohn* have
shown that the human pancreatic juice is affected by diet as
Walter® has shown for the dog. Weinland® has shown that the
pancreatic juice of the dog has lactase present on a diet con-

1 Neilson and Lewis: This Journal, iv, p. 501, 1908.

? Wassilief: Arch. des biol., St. Petersburg, ii, p. 219, 1893.

3 Lintwarew: Biochem. Centralbl., i, p. 201, 1903.

*Ellinger and Cohn: Zeitschr. f. physiol. Chem., xlv, p. 201, 1908.

5 Walter: Jbid., vii, p. 1, 1899.

®Weinland: Zeitschr. f. Biol., xxxviii, pp. 16 and 607, 1899; xl, p. 386,
1900.

Siu

332 Effect of Diet on Saliva

taining an excess of lactose. Neilson and Terry! have shown
that the amylolytic action of dog’s saliva is increased on a
carbohydrate diet.

Method. The subjects for these experiments were students.
No change was made in their manner of living other than the
change indiet. This diet consisted of two kinds; one an excess of
carbohydrate food; the other an excess of protein food. Coffee,
tea, milk, or beer were allowed if the subjects were accustomed
to use them. Since the taking away of liquids might increase
the concentration of tlfe saliva it was thought best to allow the
usual liquids.

The saliva was first tested for three consecutive days while
the subjects were on a mixed diet. In testing this saliva the
following experiments were done:

A definite quantity of saliva (25 cc.) was collected. The
amylolytic power of the saliva was first tested. Two and one-
half cubic centimeters of saliva were diluted to 100 cc. with dis-
tilled water Ten cubic centimeters of this diluted saliva were
placed in a 250 cc. Erlenmeyer flask containing 75 cc. of 2: per
cent arrow root starch and 15 cc. of water. The total contents
of the flask were warmed to 37.5° C., the temperature of the
incubator. The flasks were thoroughly shaken, placed in the
incubator and left for 20 minutes. At.the end of this time the
contents were boiled, cooled to room temperature and the
amount of maltose produced determined by Haines’ titration
method. A control was always made with boiled saliva.

The experiments to show the action of the maltase in the
saliva were made as follows:

In a 250 cc. Erlenmeyer flask 1 gram of maltose in 45 cc. of
distilled water was placed. The contents were warmed to 37.5°
C. To one flask 5 cc. of undiluted saliva were added and the
contents of the flask at once boiled To the other flask 5 cc. of
undiluted unboiled saliva were added. These flasks were placed
in the incubator and left for one hour. At the end of this time
the contents were boiled and made up to 100 cc. with distilled
water. This dilution made a solution of 1 per cent maltose.
The reducing power was determined by the Haines titration

‘Neilson and Terry: Amer. Journ. of Physiol., xv, p. 406, 1906.

Charles Hugh Neilson and M. H. Scheele 333

method. The result from the flask with the boiled saliva was
foracontrol. The result from the flask with the unboiled saliva
showed the action of the maltase in the saliva by the increase in
reducing power over the control. Maltose has approximately
one-third less reducing power than dextrose. If maltase were
present we would expect to find the reducing power of the solu-
tion with unboiled saliva to be greater than the boiled saliva.
This increase in the reducing power of the solution with the
unboiled saliva is due to the action of the maltase in the saliva
splitting the maltose into dextrose. This increase was found
in every case without exception.

A polarimeter determination was also made with the contents
of each flask. The solution with the unboiled saliva always
showed a smaller rotation than the solution with the boiled
saliva. This result one would expect as the power of rotation
of dextrose is much less than that of maltose. In many of the
experiments it was necessary to filter through animal charcoal
to clear the solution before the polarimeter readings were made.

The amylolytic power of the saliva of the subjects on a mixed
diet and the power of this saliva in splitting maltose were used
as controls for the saliva when the subjects were placed on a
carbohydrate or protein diet. Twenty-four hours after the
protein or carbohydrate diet was started, the amylolytic power
of the saliva and its maltose-splitting power was determined.

The amylolytic power of the saliva was estimated in order to
determine whether the change in maltose-splitting power was
parallel with the change in the amylolytic power of the saliva
from the subjects on the different diets.

In Table 1 the increase of the maltose-splitting power of the
saliva from those on a carbohydrate diet and the decrease on a
protein diet is seen. This change is calculated in per cent using
as a basis the maltose-splitting power of the boiled saliva. Of
course, the boiled saliva had no splitting power and its reducing
power was that of pure maltose as none of the maltose had been
split into dextrose. The reducing power of the boiled saliva
Was considered as roo per cent. We now have the maltose-
splitting power of the saliva from the subjects with stated diet.
To show the effect of this diet these results are compared with
the maltose-splitting power of the saliva from a mixed diet.

ure

Effect of Diet on Saliva

ve

OOT
OOT
OOT
OOT
OOT
OOL
OOT

” ”
here Tah One = gap waren ; fs ¢ es
Jokers else ao) Mie Eger ABD Ip Oly
SP By ches Me aac eho ehal ot eae *
i sey epee oy SORTA MRy gave exe orate s .
pp Ci ENO POLO Senn tea Untie” 4orp ayeapAyoqse),
rete sielts RPE aE Ne nna: oye 8? Paes Rea cf

do1Y} JO} asvIVAR Sz[NSOY “JorIp poxip

G LNGWINad xo

$8 pi
0¢ got
OIT 91+
09T get
IST eer
821 ez+
O1T 8+
cIT Sait
StI Siete
OST 6€ +
OST 8e+
IIt Lt
TOL icy
eOr oo
COT Gor
Tél €6+
‘asojjpuL “But *quao sad
“BAlles Jo raMod oy Ajo[AMYy ‘1aMod Sulonpel ul esueyyD
*BAT[BS 93NIIp “09 OT “eAl[es pe[loqun ‘00 ¢
‘948M Pos][YSIp “90 CT ‘10J@M pal[ysIp °09 CF
‘ajysed Yoie4s Jueo Jed Z “00 Gy ‘9SO4[BUl WEIS T
€ G

OOT
OOT
OOT
OOT
OOT
OOT
OOT
OOT
OOT

‘quay sad

*gua0 19d QOT
perepisuod 1aMod Zulonpey

“BAI[BS pa[log ‘00 G
‘TOYEM POT[\9SIP “00 CH
‘Q80}/8UL WGI T

‘T INGWINadxXa

” ”

a) -S:\ppie i! ies (ee) ws lle) ecnkansae Nee tie male, tale yeIp ul9}O1g

” ”

” ”

” ”

” ”

dd1Y} IO} ODVIOAB S}[NSeY “QIp paxil]

"yarp Jo Auq

“LaIad

Charles Hugh Neilson and M. H. Scheele 335

In every experiment without exception a carbohydrate diet
increased the maltose-splitting power of the saliva and a protein
diet decreased it. This change is parallel to the change in the
amylolytic power of the saliva on the same diet.

TABLE 2.
EFFECT OF VARIATION IN CONCENTRATION OF MALTOSE.
The time and amount of saliva are constant.

5 grams lgram /|1.5 grams} 2 grams 4 grams
SOLUTION. | maltose. | maltose. | maltose. | maltose. | maltose.

per cent. | per cent. per cent, per cent. per cent.

The reducing power of each
amount of maltose plus

45 cc. water and \

5ce. boiled saliva {

45 ce. distilled water |

5ce. unboiled saliva {

100 100 100 100 100

126 132 130 120 116

EFFECT OF VARIATION IN AMOUNT OF SALIVA.
The time and amount of maltose are constant.

| 2.5¢ee. | 5. ec. 10 ce.
saliva. saliva. saliva.

| per cent. per cent. per cent.
The reducing power of each solution containing g
1 gram of maltose and the different!
amounts of boiled saliva plus enough 100 100 100

water to make 50 cc . ae |

1 gram. of maltose ae cas aifecaat
amounts of unboiled saliva plus eaeuel

MADE COIMAKC OOICC tac: = 2 nese. o | 116 133 132

EFFECT OF VARIATION IN TIME.

The amount of maltose and saliva are constant.

REDUCING POWER AND TIME.

SOLUTION.
4 hour. 1 hour. 14 nour.

per cent. | percent. | per cent.

1 gram maltose |

45 ec. water rene eine tineeeaine Bae 100 100 100
5 ce. boiled saliva (control) |
1 gram maltose ]
45 cc. distilled water \
J

5 ec. unboiled saliva

Ss tet race 2 116 130 134

336 Effect of Diet on Saliva

ExeZ

Mixeol Diet
Protein, Diet

CareoujyoRATE

PUR} OSE AND BOILED SALIVA:
NG POWER = 100%

In all the experiments polarimeter determinations were made
and they always agreed with the results obtained by the reduc-
tion method. The polarimeter readings are not given as they
would make the tabulated results too complex.

In Table 2 the effect of variation in time, amount of maltose
and the amount of saliva is seen. These experiments show that
in the higher concentrations of maltose, for instance, a 4 per cent
and 8 per cent solution (2 grams and 4 grams in so cc. of solution)
the maltose-splitting power of the saliva is lessened. The opti-
mum concentration is 2 to 3 per cent maltose. The amount of
splitting is not proportional to the amount of saliva used in the
experiments. It is proportional for 2.5 cc. and 5 cc. but not
when to cc. were used. The amount of splitting is not directly
proportional to the length of time of the experiment.

In diagram above a curve to show the change in the maltose
splitting power of the saliva on different diets is given.

Charles Hugh Neilson and M. H. Scheele 337

We see from these experiments that a change of diet produces
a change in the maltose-splitting power of the saliva. The
change is parallel to the change in the amylolytic power of the
saliva on the same diets. As the amylolytic power of the saliva
‘increases on a carbodydrate diet, the maltose-splitting power
increases in approximately the same ratio. On a protein diet
the amylolytic power and the maltose-splitting power decrease
in the same ratio. These experiments seem to add further proof
in favor of adaptation to diet.

If the action of ptyalin and maltase be specific, then the
explanation lies in the increase of the amount of maltase or the
secretion of a maltase with greater splitting power. The increase
of the concentration of the saliva on a carbohydrate diet with a
relative increase in the amount of maltase would also offer an
explanation.

If the ptyalin has both an amylolytic and maltose-splitting
power the increase in the quantity or action of the ptyalin would
explain these results. No attempts were made to determine
whether the result is due to the action of ptyalin or to the action
of the maltase.

:

t

.

Bh
fa ane
\
iw
i] ie ‘i

SOME NOTES ON THE CHEMICAL COMPOSITION AND
TOXICITY OF IBERVILLEA SONORAE.

(PLaTEs IV anp V)

By JULIA T. EMERSON anp WILLIAM H. WELKER:!
(From the Laboratories of the New York Botanical Garden and the Labora-
tory of Biological Chemistry of Columbia University, at the College of
Physicians and Surgeons, New York.)

(Received for publication, September 2, 1908.)

Page

PE TU PCEIOT FFL ges ee a Sia wh Deen dw aie sts gd eRe! dud a Seam 339
AMD TS GORE CUM ons uci sy SUS. 5 cate ns 5), al aoe cieelic. Haha ats Sys! Salleh Suche om eee eC 340
Pel General description of the species Sonorz...........0ase-+« 340
MPeeencral quantitative composition..........662..0.000 ees 341
Seme@rencral qualitative camposition. .... 06.262 60s asensgeeees 342
SE RIERAPCHODICAL TOCES es 0... 2d c's en eis sce Sins 2 2 & low apsl 5 walle malasis 345
NESS -EGHELISIONS, . 5 i).i0<.5 > Pe oo 26 oe 2b 2.0 ois a Bao ele 350

I. INTRODUCTION.

Among the many plants about which we know little or nothing
chemically is Ibervillea Sonore. Some of the largest tubers of
this species were collected by Dr. D. T. MacDougal in Sonora,
N.M.,in 1902. At the suggestion of Dr. MacDougal, and under
the direction and with the assistance of Dr. William J. Gies, this
study of the stem? (‘‘root’’) was undertaken. An effort was
made to ascertain its general chemical composition. We also
endeavored to determine the nature of the substance or sub-
stances in it which account for the poisonous qualities attributed
to it by the inhabitants of the regions in which itis found.

1 The work on chemical composition was done by Miss Emerson at the
New York Botanical Garden. The experiments on the toxicity of the root
were carried out by Mr. Welker in the Laboratory of Biological Chemistry
of Columbia University, at the College of Physicians and Surgeons, New
York.

? Knox: Bulletin of the Torrey Botanical Club, xxxiv, p. 329, 1907. See
the second footnote on p. 341

339

340 Ibervillea Sonorae

II. HISTORICAL.

(1) A genus under the name of Sicydium was described by
Schlechtendal in 1832 (Linnea, vii, p. 388), based upon Sicydium
Schiedeanum, a Mexican plant.

(2) Gray described in 1850 Sicydium Lindheimert (Bost. Jour.
Nat. Hist., vi, p. 194) from Texas.

(3) Sicydium Lindheimeri was made the type of a new genus,
Maximowiczia, by Cogniaux, in 1881 (D. C. Monog. Phan., iii, p.
720).

(4) Maximowiczia Sonore was described by Sereno Watson
in 1889 (Proc. Amer. Acad. Arts and Sct., xxiv, p. 51).

(s) As the name Maximowiczia had been used before 1881 for
another genus (by Ruprecht in 1859, for a genus of Magnoliacee),
Greene (Eryithea, iii, p. 75, 1895) substituted for the cucurbita-
ceous genus the name [bervillea, and Maximowiczia Sonore thus
became Ibervillea Sonore.

III. GENERAL DESCRIPTION OF THE SPECIES SONORS.

Ibervillea belongs to the Cucurbitacee or Gourd family, and
has been found in Mexico, New Mexico, Texas, and California.
As Watson’s (4) is the only full description of the species Sonore,
it is quoted below.

Climbing, glabrous throughout, the large root (stem) projecting above
ground; leaves four inches broad or less, twice three-cleft nearly to base,
with broad sinuses, the lobes coarsely sinuate-toothed; male flowers race-
mose, short-pedicellate, the calyx-tube cylindrical, 3 lines long; petals
pubescent, villous within, bifid; fertile flowers on peduncles 2 to 3 lines
long; ovary ovate, long-attenuate above; fruit ovate, abruptly stout
beaked, amber colored 1} to 13 inch long, smooth with a thick fleshy rind;
placentz many (about 14) seeded; seeds covered with a red pulp, com-
pressed, rough coated excepting the smooth margin, obovate or oblong-
obovate with a broad base, 3 lines long or more.

This species much resembles M. Lindheimert, but the leaves are more
dissected and the ovary and fruit more attenuate above, and the seeds
though decidedly turgid in the young fruit, become compressed and are
peculiar in their generally very rugose-tuberculate surface. The green
fruit is described as having about ten longitudinal rows of white dots.

The upper side of the swollen underground portion whereit is
exposed to the light is of a green-gray color owing to chlorophyll
just beneath the periderm, while the upper side is thicker, brown

Julia T. Emerson and William H. Welker 341

and without the green. The substance of the ‘“‘root”’ is much the
same throughout; firm, fibrous near the stem or a bud, more dif-
ficult to cut and tougher than a potato, and of a light yellow color.
Its general shape is round, like a gourd, with a neck leading up to
the shoots and several thick roots extending downwards, and it
may become as large as two feet in diameter and weigh many
pounds. There is a characteristic odor somewhat like that of an
earthy potato, and after working with fresh material some time
the nose and throat feel dry and irritated, while a bitter tasting
substance remains on the fingers, even after several washings.

Figs. A and B show the organographic characters of the species, both at
the adult stage in its native habitat, and as a seedling from four to five
years old. The large projecting root of [bervillea in the specimens in the
greenhouses of the Botanical Garden reaches a diameter of from 25 to 30
cm. In the desert this enormous tuberous growth lying in the dry sand
looks like a gray, dust-covered boulder. Frequently irregularities of
shape give it still more the effect of stone, and it is only when the cortex
is flecked off that one discovers the healthy green color beneath the super-
ficial layer. From the tubers arise yearly long flexible liana-like shoots
which reach a length of three or more meters. The shoots are round,
smooth and green above, brown-gray and gray-spotted or streaked below.
The flowers are dioecious, the tendrils branched, and the leaves bright-
green and twice three-cleft as is so frequently the case throughout the
family. The fruit is said to be ‘‘amber colored” and 1} to 14 inch long;
none has ripened in the greenhouse, as the flowers there are staminate
only. The plant is able to persist in its arid habitat with remarkable
vitality. In fact, so provident is it of water and nutritive substances
that one in the museum case at the Garden which has been lying on a
board since 1902 is, in 1908, still sending up yearly shoots bearing leaves
and tendrils. Every fall the shoots die back and sprout again early the
next spring.!

IV. GENERAL QUANTITATIVE COMPOSITION.

Water, solids, organic and inorganic matter in the stem were
determined by the usual methods. After removal of the bark,
fresh portions of the stem? were quickly cut into small pieces and

1 Knox: Loc. cit. This paper gives the results of a thorough anatomical
study of the stem (‘‘root’’) of [bervillea Sonore.

* This tuber is ordinarily referredto asaroot * * * The picture of
the old plant (Fig. A) with its shoots rising from the tuber, shows the
gradual enlargement of the stem, though the appearance of the seedling

342 Ibervillea Sonorae

weighed in porcelain crucibles as rapidly as possible. Each
sample was dried to constant weight in an air bath at 100° C., in
which process all water was removed. The weight of the dry
solid matter was taken and, after complete incineration at a low
temperature, the weight of the ash was also noted. The follow-
ing weights and percentages were recorded in this connection:

RECORDED ANALYTIC DATA,

Tuber Fresh substance. Dry substance. Ash.
Sample. grams. grams. gram.
As i xs Sate one ore 12.0060 1.6690 0.1710
SATO EEE AT ECS AK CR 15.4403 1.9223 0.2203
Oe Ai ae Pie dag C nk eee 19.3512 2.3222 0.3822
Be We iste hethe SNPS age 14.9534 1.9834 0.2144
IVOTARE ois 5.2 ie Sn 15.4377 1.9742 0.2219

GENERAL PERCENTAGE COMPOSITION.

Tuber Organic Inorganic
Sample. Water. Solids. matter. matter.
We yesStvde Gis witete nome 86.10 13.90 12.49 1.41
Di eer ich teste Austen ee 87.56 12.44 11.02 1.42
Be ie ete, nok Pucnaseeie 88.00 12.00 10.60 1.40
A Bcy ag hottie una tele 2: 2 86.74 13.26 11.82 1.44
Average ....... 87.10 12.90 11.48 1.42

V. GENERAL QUALITATIVE COMPOSITION..

In an endeavor to ascertain the character of the more conspicu-
ous constituents of the tuber, grouped under the general term
‘solids,’ various extracts were made and tested. Between the
arrival of the tubers from Sonora and the time when they were
used for this study (almost a year), they remained in the labora-
tory on saucers, without earth or water, and seemed to keep per-
fectly, one even putting out several shoots during the summer.’

Preparation of material. In the preparation of the tuber for
these tests, dirt and dust were brushed off and the tuber was

(Fig. B) would indicate that the swollen portion includes root, hypocotyl
andstem. * * * The formation is stemand at least half of the swollen
portion may claim that distinction Knox, loc. cit., p. 340.

1 The quantitative data recorded above were obtained from tubers in the
same general condition.

EH tage ag

Julia T. Emerson and William H. Welker 343

washed. Sometimes the bark was removed; at other times the
entire root, after removal of injured or abnormal portions, was
finely divided ina meat chopper. Portions of the chopped tuber
were at once put in bottles with one of the solvents named below,
a few drops of toluol or chloroform being added as preservatives.
After standing generally a week, the extract was filtered and this
filtrate used in the experiments.

Extracts were made with the following solvents: Water, 95
per cent alcohol, ether, glycerol, 10 per cent sodium chlorid, or
acidified water (very dilute hydrochloric acid). The aqueous
and saline extracts were slightly acid in reaction (litmus).

The common qualitative reactions on these extracts made it
evident that the root contains representatives of the following
groups of substances: proteins, carbohydrates, fats, cholesterins,
lecithins, basic substances, salts of organic acids, and inorganic
salts.

Proteins. The simple proteins present in the tuber were of the
coagulable type only. Proteose and peptone could not be de-
tected in any of the extracts after removal of the coagulable pro-
teins. Nucleoprotein was present in small proportions in the
extracts and could be precipitated by acid. On decomposition
in hot acid the protein yielded purin bases, simple protein and
phosphate.

It was hoped by finding the temperature at which coagulations
resulted in the extracts, to gain some idea of the number and
types of the coagulable proteins present. The results were not
very definite, however. Ten per cent salt extract, for example,
showed turbidity at the low point of 25° C. The temperature
was gradually raised and at 61° C. a fine flocculent precipitate
appeared. On filtering, a clear yellow filtrate was obtained.
On reheating, initial turbidity became apparent above the tem-
perature of former turbidity but below that of former flocculation;
in this way six to eight flocculations were obtained at tempera-
tures ranging between 25° C. and 80° C.

On adding to an aqueous or saline extract an equal volume of
saturated ammonium sulfate solution, a fairly bulky precipitate
of ‘‘globulin’’ was obtained. The half saturated filtrate gave an
abundant yield of albumin when ammonium sulfate was added
to complete saturation. Saturation of the saline extract with

344 Ibervillea Sonorae

magnesium sulfate precipitated globulin. Solutions of both the
globulin and the albumin precipitates obtained in this way gave
no more satisfactory differential results with the coagulation
method than the original extracts. The number of globulins
and albumins present was not, therefore, ascertained. It appeared
probable that these proteins coagulated through ranges of degrees
of temperature that overlapped each other, so to speak, thus
making differentiation by this process impossible, or else that other
substances were precipitated in such ways as to interfere in the
manner indicated.

The precipitates obtained at some of the lower temperatures
may have consisted largely of earthy phosphate, for they were
not appreciably soluble in 10 per cent sodium chlorid solution or
in 0.5 per cent sodium carbonate solution, but readily dissolved
in 0.2 per cent hydrochloric acid.

The precipitates obtained at the higher temperatures, on the
other hand, were true coagula, for they were practically insoluble
in all of the three reagents named.

Carbohydrates. The carbohydrates present in the tuber were
chiefly woody fiber and small amounts of reducing sugar and
starch. The nature of the small amount of sugar was not deter-
mined.

Fat, cholesterin and lecithin were extracted and separated by
Zuelzer’s ether-acetone method.t No attempt was made to
ascertain the characters of either the fatty matter or the lecithin.
Their quantities appeared to be relatively slight.

Saline matters, especially chlorid, sulfate and phosphate, so-
dium, potassium, calcium, magnesium and iron could be readily
detected. The ash contained carbonate, most of which was
formed by the combustion of the organic matter. Salts of or-
ganic acids were frequently indicated, in the various reactions
tried, but no attempt was made to ascertain the characters of the
organic radicals. Tri-basic calcium phosphate was precipitated
in relative abundance from aqueous and saline extracts on warm-
ing. Earthy phosphate was subsequently found to occur in the
root, and in such extracts.

Enzymes were looked for but were not detected. Proteolytic,

1 Zuelzer: Zeitschr. fur physiol. Chem., xxvii, p. 255, 1899.

Julia T. Emerson and William H. Welker 345

diastatic and oxidative enzymes appeared to be absent. It is
possible that the conditions of the tests were not favorable to
digestive results, although ordinary methods were used.

Basic substances. A few tests for alkaloids, with the so-called
alkaloidal reagents, in extracts free from coagulable protein,
gave indications of the presence of basic substances (see p. 346).

Chlorophyll was conspicuous in the subcortial tissue.

VI. TOXICOLOGICAL NOTES.

In Mexico where I[bervillea Sonore grows in abundance, the
natives consider it poisonous. Watson! has stated that ‘‘a
decoction of the root is used as a cathartic.”

Preliminary toxicological experiments on frogs. Several pre-
liminary tests of the toxicity of the tuber were made on frogs, by
injecting under the skin or into the stomachs small volumes of
concentrated aqueous solutions of dry residues from alcohol or
ether extracts, with the following results:

(a) Residue from alcohol extracts. First frog: Introduction
per os. No visible effects. Second frog: Injection subcutane-
ously. Spasms occurred in a few minutes; later there was retch-
ing. Death ensued in three hours.

(b) Residue from ether extracts. Third frog: Introduction
per os,about 3 p.m. There were signs of nausea during the after-
noon. Death occurred during the night. Fourth frog: Injec-
tion subcutaneously. Convulsive movements preceded death in
24 hours.

The results of these tests led us to look for alkaloids. For the
preliminary isolation of any of the latter that might have been
present in the tuber we proceeded as is indicated below:

Tests for alkaloids. The Stas-Otto method was employed for
the preparation of the extracts to be tested. The interior por-
tions of two tubers were finely minced and subjected to extraction
separately. Two alcoholic extracts were made of each. The
residues from each yellow alcoholic extract, which were obtained
by evaporation im vacuo at 40° C., were treated with absolute
alcohol, filtered and again evaporated to dryness. The residues
from the absolute alcohol were treated with water, the filtered

"Dee p. 340.

346 Ibervillea Sonorae

solutions rendered alkaline with sodium carbonate’ and any
alkaloids extracted with ether. Each residue was treated thrice
with fresh ether. The ether extracts were allowed to evaporate
to dryness spontaneously. The resultant dry residues, which
were yellowish in color and very bitter, were used in the tests
referred to below:

Chemical tests. Samples of the various residues obtained by
the above described method were dissolved in dilute hydrochloric
acid or alcohol and tested with common reagents for alkaloids.
White, amorphous precipitates were obtained with potassium
hydroxid, phosphomolybdic acid, phosphotungstic acid and
tannic acid.

Precipitates failed to form with solutions of picric acid, platinic
chlorid, iodin in potassium iodid solution, bromin in hydro-
bromic acid solution, potassium sulphocyanid, potassium chro-
mate, potassium dichromate, potassium iodid, mercuric chlorid,
potassium ferrocyanid, potassium ferricyanid.*

The material in the extracts gave color reactions with the fol-
lowing reagents: Sulfuric acid, Erdmann’s reagent, sulphovana-
dic acid, and Froéhde’s reagent. None of these results was char-
acteristic of any particular alkaloid, however, nor were mixtures
of alkaloids suggested.

The reaction with sulfuric acid and potassium permanganate
in its first stages resembled the coloration obtained with strych-
nin in the oxidation test, but it failed to exhibit the successive
changes of coloration characteristic of the alkaloid in that reac-
tion. The color was not given with any other oxidizing reagents
in sulfuric acid. According to Wormley, this reaction, if potas-
sium permanganate is employed as the oxidizing agent, is given

1The yellow color changed to orange.

2 While we were removing residue from one of the evaporation dishes,
an associate who was fully ten feet away and who knew nothing about the
peculiar qualities of the residue, expressed surprise at a bitter taste he
suddenly experienced. The very minute amount of dust that was formed
as we carefully scraped the residue from the dish accounted for the experi-
ence referred to, which gives a good idea of the extreme bitterness of the
material.

3 Precipitates formed by reaction between the solvent and the reagent
are here ignored.

Julia T. Emerson and William H. Welker 347

by aqueous extracts of certain plants and even by shreds of filter
paper.*

Physiological tests. Toxicity of the residues obtained by the
method described above was determined by dosing a 2.7 kilo
female pup as follows:

Administration per os: 1st day. The dog was given in one meal daily
an ordinary quantity of food consisting of meat, cracker meal, lard, bone-
ash and water. The weight of the residue given the first day was 0.0523
gram. It was placed inside of a ball of meat and in this condition swal-
lowed without mastication. Except very slight diarrhea during the fol-
lowing night, no effects were observed.

2d day. On the second day of the experiment 0.1046 gram of residue
was administered in the food as before. Slight diarrhea, about nine hours
after feeding, was the only apparent effect of the material.

8d day. On the third day the animal was apparently perfectly normal,
and accordingly a study of the effects of subcutaneous injections of the
material was begun on the following day.

Subcutaneous injection: 4th day. The animal appeared to be normal
in all respects. About 0.05 gm. of residue was dissolved in 3 to 4 cc. of
alcohol, and injected on the right side. No general effects whatever were
observed during the following twenty-four hours.

5th day. On this day about o.1 gram of residue was injected in 3 to 4
ec. of alcoholic solution. No symptoms of any toxic action were observed
during ten hours following the injection. A small abscess formed at the
point of injection; otherwise the animal apparently remained normal
during the next two days, when observation was discontinued.

These tests failed to show the presence in the extracted mate-
tial of any known alkaloid or alkaloids. The physiological tests
indicated further that, if any alkaloid was present, its propor-
tionate content was too slight to enable it to exhibit distinct or
characteristic effects.

Tests of the effects of hashed tuber given by mouth to dogs. The
results of the foregoing toxicological tests suggested either that
our extracts did not contain the toxic principle of the tuber or
that the reputed poisonous action of the tuber is due to some
compound other than an alkaloid. To decide this question, a
number of special experiments on toxicity were conducted.

Two tubers were rapidly peeled, sectioned and hashed. The interior
of one of them was found to be full of small brown masses. The other

1Wormley: Micro-chemistry of Poisons, 2d ed., p. 589, 1885; Sedgwick:
Amer. Chem. Journ., p. 369, 1879.

348 Ibervillea Sonorae

tuber was apparently normal in all respects. The mass of each tuber was
kept separate on the assumption that possibly the abnormal specimen
would show different toxic results from those of the normal tuber; also
with the idea that perhaps the toxicity ascribed to Jbervillea Sonore in
general was due to special poisonous principles in such abnormal tubers
and that particular effort should be made to determine this possibility.
The rinds were also hashed. Each product was bottled and placed in a
cold room which was kept constantly at a temperature of about — 5° C.

The following quotations from our records of this series of experiments
will suffice to show the results obtained.

Sixth experiment with normal tuber. Weight of dog, 25.7 kilos.!

1st day. Sixteen grams of tuber were administered in balls of meat at
1ra.m. In the afternoon a feverish and restless condition was observed.
This was very much less marked during the evening. Pronounced diar-
rhea occurred between midnight and8 a.m. 2dday. By 10a.m. the con-
dition of the dog was apparently very close to normal. Twenty-four grams
of tuber were given in meat at 10.30 a.m, The feverish condition reap-
peared unmistakably at 1:45 p.m. Vomiting occurred at intervals be-
tween 5 and 9:45 p.m. No defecation occurred during the day, at the
end of which the dog was apparently normal. 3dday. Thirty-two grams
of tuber were administered with the regular diet at ro:50 a.m. At 12:30
p.m. the animal became feverish and uncomfortable. Three hours later
very marked diarrhea occurred. About 6 p.m. a considerable portion of
the food given in the morning, including a fairly large proportion of meat,
was vomited. Between 7 and 8 p.m. marked diarrhea occurred. From
this time until the next morning, recovery wasrapid. The dog was appar-
ently normal at the end of that period. 4th day. The dose of tuber was
again increased and 40 grams were given with the usual diet between
Ira.m.andrp.m. At 1:40 p.m. the first signs of sickness appeared.
Between 4:15 p.m. and 5:20 p.m. very marked diarrhea occurred. This
was followed by rapid recovery and the dog was nearly normal four hours
later. 5th day. The animal refused to eat the meat balls containing the
tuber and so was allowed to fast until the sixth day. 6th day. The dog
again refused the meat balls. They were given forcibly. He received
'50 grams of tuber at noon. The feeding was followed by a five-hour
period of feverishness and restlessness. This was succeeded by very
marked diarrhea. By midnight the dog had practically recovered. He
remained well. The same animal was used in the seventh experiment.

Seventh experiment; with abnormal, i. e., ‘‘spotted’”’ tuber, and with the
rind. About a month and a half after the completion of the preceding

experiment, the dog of that experiment, which showed no subsequent ill —

effects of the treatment described above, was subjected to dosage with the
abnormal tuber. Diet and method of dosage were the same as in the
sixth experiment.

1 Daily diet—hashed meat, cracker meal, lard and water, with bone-ash.
A preliminary period of five days established normal conditions.

Julia T. Emerson and William H. Welker 349

A. mstday. Five gramsof ‘spotted’ tuber were given without effect.
The feces passed this day were hard and dry. 2d day. Ten grams of
tuber were given. There were no apparent effects. 3dday. Twenty
grams of tuber were administered in gelatin capsules. This dose also
failed to produce any toxic symptoms. The feces remained hard and dry.
4thday. Forty grams of tuber were given in capsules as before. Appar-
ently no effects were produced. sth day. Fifty grams of tuber were
given at noon in capsules as before. Slight diarrhea occurred during
the afternoon and evening, while marked diarrhea ensued between mid-
night and 8 a.m. The dog was apparently normal at 9 a.m. 6th day.
The regular diet was fed without the tuber.

B. 7th day. Twenty-five grams of rind were given in gelatin capsules
with the regular diet. Very marked diarrhea occurred about six hours
later. The dog seemed normal on the following morning. 8th, oth and
roth days. The dog was fed the usual diet without anyrind. rth day.
Twenty-five grams of rind were given in capsules without producing any
apparent effect. sr2thday. Twenty-five grams of rind were given. Three
hours later marked diarrhea ensued. This was followed by the vomiting
of undigested meat. 13th day. The animal fasted. 14th day. A small
mealwasfedintheafternoon. r5thday. Twenty-five grams of rind were
given in capsules with the regular diet at 10 a.m. Diarrhea occurred
between 8 and 8:45 p.m. By the following morning the dog had appar-
ently recovered.

These experiments and the remaining ones of the series made
it evident that the toxicity and even the cathartic effects of
Ibervillea Sonore, as measured by results on our dogs, have per-
haps been overrated. In some instances, comparatively large
quantities of the hashed tuber or the rind were entirely devoid of
Obvious effects. The ‘‘spotted”’ tuber was no more (but perhaps
less) toxic than the normal tuber. These results were also in
accord with our previous conclusion that there is little if any
alkaloid present in the tuber.

Although the toxic and cathartic effects were relatively slight,
it was thought that possibly oxalic acid was present in sufficient
proportion to share in producing the effects noted. Analysis of
a fresh portion of tuber made it evident, however, that the
proportion of oxalic acid in the tuber was very slight.

It also seemed possible that magnesium salts might have par-
ticipated in the cathartic effects observed. Partial quantitative
analysis of the ash gave the following percentage results: CaO,
23.76; MgO, 15.62; CO,, 13.92; Cl, 11.96; P,O,, 8.1. Sodium and
potassium were present in abundant proportions. The propor-
tion of ash in the fresh tuber was only about 1.4 per cent.

350 Ibervillea Sonorae

Although it did not seem probable that the observed cathartic
action could be due solely to the small amounts of compounds of
the alkali earth metals that were present, we tested the effects of
the ash in some experiments of which the following, as described
in our records, is an illustration.

Tests of the effects of tuber ash given by mouth to dogs. Tenth experi-
ment. Weight of the dog, 7.22 kilos.1 On the first day of the experi-
ment the animal was given 0.6 gram of ash corresponding to the inorganic
matter in about 45grams of tuber. Thiswas administered in several small
balls of meat. No other food was given that day. There were no ap-
parent effects. On the following day the fast was continued and an equal
portion of ash, i. e., 0.6 gram (which had been largely converted into chlorid
by treatment with hydrochloric acid, with subsequent removal of the acid
by evaporation to dryness) was dissolved in distilled water and admin-
istered by means of a stomach tube. This also failedto produce any
noticeable effects. The dog was apparently normal in all respects during
the three succeeding days, after which he was no longer kept under
observation.

The results of the experiments in this series also fail to explain
the effects caused by the tuber. Unfortunately, our supply of
tubers was exhausted at this point and further study had to be
given up.

VII. GENERAL CONCLUSIONS.

Besides yielding the data for composition of [bervillea Sonore,
our study indicates that the tuber does not contain alkaloidal
material, or at least not in sufficient proportion to cause particular
toxicity. We have not learned the reason for the cathartic efiects
exhibited by the tuber or its decoctions.

We wish to thank Dr. Gies for his kind guidance and advice in
this study, and also for his assistance in carrying out the work.

1 Daily diet—hashed meat, cracker meal, lard and water, with macer-
ated filter paper to make the feces bulky and approximately normal in
consistence. For this purpose bone ash was avoided in order to give the

tuber ash full opportunity to display its greatest possible toxicity, without

unnecessary interference. A preliminary period of five days established
normal conditions.

+

Pirate IV

Tue JourNat oF BroLocicaL CHEMISTRY,

Vv, NO. 4.

VOL.

Ibervillea Sonore in its native habitat.

THE JOURNAL OF BIoLoGICAL CHEMISTRY, PLaTE V.
Wal V,-NO= 4.

Fig. B. IJbervillea Sonore, four or five years old.

ty F

THE INHIBITING EFFECT OF POTASSIUM CHLORIDE IN
| SODIUM CHLORIDE GLYCOSURIA.

By THEO. C. BURNETT.

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

(Received for publication, October 15, 1908.)

While making experiments on the production of glycosuria
by sea-water' I became aware that a mixture of sodium chloride
and potassium chloride in the proportion of 100 mols. of the for-
mer to 2.2 mols. of the latter, was much less effective than pure
sodium chloride in inducing glycosuria, and the fact seemed of
sufficient interest to warrant further investigation. It has been
known for years that the injection of large quantities of sodium
chloride will cause glycosuria in rabbits, and Fischer? has shown
that this glycosuria can be inhibited by calcium chloride. Brown’
arrived at thesame conclusion independently. It seemed im-
portant therefore, to ascertain if potassium chloride had also an
inhibitory action on the glycosuria caused by the injection of a
pure sodium chloride solution.

The technique was the same as formerly, but the well known
poisonous effects of potassium precluded the possibility of carry-
ing on the experiments for longer than about three hours, 1 liter
(about) of the solution (NaCl, 100 mols. + KCl, 2.2 mols.) being
injected during that time.

1 Burnett, Theo. C.: This Journal, iv, p. 57, 1908.

Note: Since the publication of the paper just referred to, my attention
has been called to an article by Underhill and Closson, Amer. Journ. of
Physiol,, xv, p. 321, 1906; also to one by Hedon and Fleig: Arch. internat.
de physiol., p. 95, July, 1905. Reference should have been made to
these in the original article, and I take this first opportunity to correct
the omission.

2 Fischer, M. H.: Univ. of Calif. Publ. (Physiol.), i, p. 87, 1904, where
teferences to earlier authors will be found.

3 Brown, O. H.: Amer. Journ. of Physiol., x, p. 378, 1904.

35%

252 Inhibition of Salt Glycosuria

When such a solution as the above (sodium chloride plus potas-
sium chloride) is injected into the marginal vein of the ear of a
rabbit in large quantities, after about one and a half hours sugar
makes its appearance in the urine. The amount varies from a
trace to about 0.03 percent. At first sight this seems like a very
marked variation, but by comparing with a pure sodium chloride
solution it will be seen that the maximum does not nearly equal
the maximum given by Fischer for a pure sodium chloride solu-
tion, his tables giving o.25 per cent.1_ It must be borne in mind
that rabbits vary very widely in their susceptibility, and hence
the impossibility of having adequate controls in these experi-
ments. I have the records of two experiments in which abso-
lutely no sugar appeared during three hours injection with a
mixture of sodium chloride and potassium chloride. On the
other hand, one rabbit gave a maximum of 0.04 percent. The
bulk of the experiments, however, gave a figure around o.o1 per
cent. When it is said, therefore, that a given solution produces
no glycosuria, the fact must not be lost sight of that now and
then a rabbit will be found that develops glycosuria, even with
Ringer’s solution; but in that case it would probably be found
that with a mixture of sodium chloride and potassium chloride
the amount would be considerably larger, and with a pure sodium
chloride solution, a profound glycosuria would result. Table I
is selected as fairly typical.

TABLE I.

December 6, 1907. Belgian hare, female, weight 3000 grams.
Injection fluid, M NaCl + KCl.

Time. Amount injected.| Amount of urine.

a.m. cc. cc.

8.45 Began injection.
10.00 450 199 Trace sugar.

10.30 150 154 Sugar, .01 per cent.
11.00 200 154 Sugar, .007

ILS %0) | 150 160 Sugar, .01

1 Fischer, M. H.: loc. ctt., p. 111.

Theo. C. Burnett 353

The experiment was then tried of inducing glycosuria by a
pure sodium chloride solution, and then substituting a mixture
of sodium chloride and potassium chloride. The result was a
diminution in the amount of sugar, as will be seen from Table
II, which is condensed from the originalrecord. It willbenoticed
that the sodium chloride solution was ¥, and that sugar appeared
in one hour. It seems evident, therefore, that potassium antag-
-onizes, in part at least, the poisonous effects of a pure sodium
chloride solution.

TABLE II.
December 5, 1907. Belgium hare, male, weight 3000 grams.
Injection fluid, “@ NaCl: changed later to = NaCl + KCl.

°

Time. Amount injected. | Amount of urine.

a.m. ce. ce. |

9.00 | Began injection “ NaCl.
10.00 375 | 107 | Sugar.
10.30 200 156 | Sugar, .027 per cent.

Changed to ™ NaCl + KCl

11.00 200 160 Sugar, .016 is
11.30 175 110 | Sugar, 014 “
12.00 300 155 | Sugar, .007 a

What, now, is the action of calcium chloride under similar
conditions? What has been said of potassium applies as well to
calcium. When a solution of sodium chloride, 100 mols., plus
calcium chloride, 2 mols., is injected in large quantities, a small
amount of sugar usually appears in the urine. The maximum
obtained was 0.026 per cent—a figure slightly better than for
potassium chloride. The minimum, as in potassium chloride,
is zero, no sugar appearing in three hours. If, however, the injec-
tion is begun with,sodium chloride plus calcium chloride and
after sugar appears the injection fluid is changed to sodium chlor-
ide plus potassium chloride the amount of sugar is increased.
The same holds good if the order of the injections is reversed.
Table III (p. 354) illustrates this very well.

It will be recalled that Loeb! has shown that a pure NaCl solu-
tion causes two forms of destructive process in the eggs of the sea

1 Loeb, J.: Biochem. Zeitschr., ii, p. 81, 1906.

354 Inhibition of Salt Glycosuria

urchin. One of these is the ordinary cytolysis, and the other,
a process which he calls ‘“‘black disintegration.’’ He found that
the addition of calcium chloride inhibited the cytolysis, while the
addition of potassium chloride inhibited the ‘‘black disintegra-
tion.’’ In view of this the above results are clear. The addition
of one salt, for example, potassium chloride to the injection fluid,
prevents one of the destructive processes, with a consequent
diminution of the amount of sugar, but not a complete disappear-
ance of it, for there is the other destructive process still active.
When sodium chloride plus potassium chloride is substituted for
sodium chloride plus calcium chloride, or vice versa, there is an
additive effect ; either the cells which are partially injured by the

TABLE III.

August 19, 1908. Belgian hare, female, weight, 1500 grams.
Injection fluid, M NaCl + CaCl,: changed later to NaCl + KCl.

Time. Amount injected, | Amount of urine.
a.m. cc. t cc.
9.05 Began injection, NaCl +
CaCl,
10.05 350 103 Sugar.
11.20 375 363 Sugar, .004 per cent.
Changed to NaCl + KCl.
12.20 300 279 Sugar, .009 3

one combination, are further injured by the second, or a new lot
of cells may be injured by the substituted combination of salts’
In either case the result would be an increase in the amount of
sugar in the urine. McGuigan and Brooks' conclude that “the
pathology of experimental glycosuria is very probably due to
changes in the protoplasmic activity of the cell.”” The above
results seem to indicate the nature of the change which occurs in
the cells involved, be they in the central nervous system, or
in the kidney, or elsewhere. Microscopic examination of the
kidney in several of these cases shows interesting changes; but the
work has not been carried very far, and it will be left for a future
communication should results warrant it.

1 McGuigan and Brooks: Amer. Journ. of Physiol., xviii, p. 256, 1907;
ef. also Underhill and Kleiner: This Journal, iv, p. 395, 1908.

THE URIC ACID EXCRETION OF NORMAL MEN.
By PAUL J. HANZLIK ann P. B. HAWK.

(From the Laboratory of Physiological Chemistry of the Department of
(Animal Husbandry of the University of [llinots.)

(Received for publication, October 26, 1908.)

The purpose of this investigation was to observe the course of
the excretion of uric acid in normal men living on an ordinary
mixed diet. Each subject was allowed to select his own diet,
and then was required to ingest the diet selected during the
course of six periods of four days each. The variation in the
excretion of each individual as well as the average excretion
of all the subjects was noted.

Ten university students served as subjects. Quarters were
provided where the men could easily be observed as to certain
regulations of sleep and diet. The body weights of the sub-
jects ranged from 53.1 to 76.7 kilograms and their ages varied
from 19 to 29 years. There were no athletes among the sub-
jects so that no individual took excessive or violent exercise,
but all lived the life of the average normal university student.

As was previously stated the diet was a mixed one for all
periods. Table I, p. 360, shows the diet for period 1, November
29 to December 2. The diet for the remaining periods was
approximately the same in all particulars.

The fruits consisted of red and white cherries, pineapples, apples,
Oranges, plums, peaches and pears. Of the meats, which were all fresh,
there were boiled beef, roast beef, veal, mutton, beefsteak, chicken, pork
chops, pork sausage, and boiled and roast ham. The soups consisted of
pea, consommé, bouillon and vegetable soup. Cornmeal, rice, cream of
wheat, oatmeal, and cracked wheat, composed the cereals. The puddings
were custard, starch cream and rice. Ice cream was also served every
four days. The vegetables included potatoes, peas, beans, corn and
tomatoes. The other constituents were wheat bread, coffee, milk, cocoa,
butter and water. Meat was served twice a day, morning and evening.
Soups were served at the noon meal only and puddings instead of fruits
at night, otherwise the same constituents were served at each meal. In

355

356 Uric Acid Excretion

some instances the men selected no coffee, while in other cases a subject
would, for example, drink coffee for the morning meal and cocoa for the
evening meal.

For the sake of uniformity in the mode of normal living, each subject
was required to carry out a routine daily schedule. They all rose at
6:30 a.m., were immediately weighed and then breakfasted at 7:15 a.m.
Lunch was served at 12:15 p.m. At 5:00 p.m. each individual was sub-
mitted to a clinical examination. Dinner was served at 6:30 p.m., and
all subjects retired between 10:30 and 11:00 p.m. Blood tests, includ-
ing blood pressure, were made upon each subject at intervals of two
weeks. If at any time a subject was found to be abnormal for any reason,
the experiment upon him was discontinued until he again became nor-
mal. This will be seen to be the case with subjects I and K, Table II,
p-. 361, where one period for each of these subjects is missing.

The experiment was divided into periods of four days each.
Urine was collected in 24-hour periods and preserved with
powdered thymol, each bottle being kept in a refrigerator until
the entire 24-hour sample had been collected. The urine bottles
were then removed, one-half of each sample carefully measured
and placed in cold storage. At the end of the four day period,
the four one-half volumes were combined to form a single sample
for each subject, thus making ten composite samples for analysis.
These composite samples were then analyzed for uric acid.

The Folin-Shaffer procedure was employed for the determina-
tion of the uric acid.

r50 cc. of the urine was introduced into a beaker, 37.5 ec. of the Folin-
Shaffer reagent (consisting of 500 grams of ammonium sulphate, 5 grams
of uranium acetate and 60 cc. of 10 per cent acetic acid in 650 ce. of dis-
tilled water) added and the mixture filtered. 150 cc. of the filtrate was
immediately taken, 7.5 cc. of concentrated ammonium hydroxide added
and the mixture allowed to stand for 48 hours. The precipitated am-
monium urate was then transferred quantitatively to a hardened filter
paper and washed with a ro per cent solution of ammonium sulphate to
free it from chlorides. The paper was then removed from the funnel,
opened, and by means of hot water the precipitate was rinsed back into
the beaker in which the urate was originally precipitated. The volume
of the fluid was, at this point, about 150 cc. The solution was allowed
to cool to room temperature, 15 cc. of concentrated sulphuric acid was
added and the acid solution immediately titrated with potassium
permanganate.

By an examination of Table II, p. 361, it will be observed that —
there exists considerable uniformity in the daily excretion of :

Paul }.c;Hanzlik and P. B. Hawk 357

uric acid. Subjects E, F, G, I and J especially exhibited a
very constant output. The average daily excretion of subject
F, 0.475 gram, was the lowest of any of the subjects. J excreted
an average of 0.520 gram per day; I, 0.619 gram and G, 0.644
gram. It might be supposed that differences in the quantity
and kind of food taken would influence the daily excretion of
uric acid to a marked degree, but it will be seen that this was not
altogether the case. Subject F excreted 0.475 gram of uric
acid per 24 hours, while subject G excreted 0.644 gram, a much
greater quantity than that of subject F, notwithstanding the
fact that Table I shows that the diets of subjects F and G were
very similar in many respects, the main difference being that G
ingested somewhat more meat, potatoes and milk. This shows
that in this case the variation in the excretion of the uric acid
was probably due entirely to the influence of the diet. The
average daily excretion of subject E was 0.660 gram, which was
the maximum output of uric acid observed during the experi-
ment, yet the diet of this subject varied but little from that of
subjects G and I, the main difference being that E ingested
considerably more bread. Subject J differed more from subject
E in the matter of diet than either subjects F or G, yet subject J
excreted a greater average quantity of uric acid than subject
F, but less than subjects E and G. It will be noticed then that
individuality in these cases played an important part in the
variation of the excretion of uric acid. This variation was
probably due to a variation in the endogenous uric acid excre-
tion, which, according to general belief, is governed largely by
the individuality of the subject.

Subjects A, B, C, H and K showed less uniformity in the daily
excretion of uric acid than the other subjects. Subject A
excreted on the average 0.615 gram per day. This subject also
showed a gradual increase in uric acid output up to the last
period at which time there was a slight decrease. Subject B
excreted on the average 0.588 gram with considerable irregu-
lar variation. The average excretion of subject C was 0.597
gram with a gradual decrease for the first four periods fol-
lowed by two periods of higher excretion. Subject H exhib-
ited a pecularity in that the daily excretion averaged 0.608
gram but varied alternately. In the first period there was a

358 Uric Acid Excretion

high excretion, i1.e., 0.621 gram, while period 2 was lower, i.e.,
0.597 gram. Then again period 3 returned to an excretion of
0.633 gram almost the same as for period 1, and period 4 showed
an excretion of 0.583 gram, similar to that of period 2. Begin-
ning with period 1, every alternate period exhibited a striking
uniformity in the excretion of uric acid, for example, 0.621,
0.633 and 0.627 gram and 0.597, 0.583, and 0.586 gram for the
alternate periods. Subject A showed an average excretion of
0.651 gram with a general tendency for the daily elimination of
uric acid to increase. This subject excreted the second largest
average quantity of uric acid in 24 hours, although the diet did
not differ from that of some of the other subjects except in the
large quantity of milk ingested and from the fact that no cocoa
nor cereals were taken. The diets of subjects A, B, C and H
also were very uniform in quantity as compared with the diets
of the other subjects, yet there were variations in the excretions
of uric acid, some of the quantities being greater than those of
subjects F, G, I and j. Here again, as in subjects E, F, G, I
and J, the same reason for this individual variation in excretion
may be given. There is probably no other factor which influences
these variations to such an extent as does the individuality of
the subject.

Concerning the average daily quantity of uric acid excreted
by all subjects, it can be readily seen by referring to Table III,
p. 364, that it does not amount to o.7 gram as is generally stated
to be the average 24 hour output of uric acid for a normal man.
The lowest quantity excreted, 0.475 gram, was by subject F, while
the greatest quantity, 0.660 gram, was by subjectE. Theaverage
for all subjects was 0.597 gram.

The evident tendency of the majority of the subjects to exhibit,
for no apparent cause, a relative high uric acid output during the
fifth period of the investigation is interesting.

Apart from the uric acid viewpoint our investigation fur-
nished some interesting data relative to the protein ingestion of
normal men. Facts regarding this feature may be found sum-
marized in Table IV, p. 365. It will be noted by referring to
this table that these ten normal men, ranging in age from 19 to
29 years, when permitted to select their own mixed ration,
ingested a diet which contained from 70.3 grams to 109.6 grams

SS

ee

Paul J. Hanzlik and P. B. Hawk 359

of protein per 24 hours. The average daily ingestion of pro-
tein as 91.2 grams. The protein ingested per kilogram of body
weight varied from 1.13 gram to 1.51 gram with an average
of 1.33 gram.

CONCLUSIONS.

1. The average daily excretion of uric acid for ten men, rang-
ing in age from 19 to 29 years, and fed a normal mixed diet was
0.597 gram, a value somewhat lower than the generally accepted
average of 0.7 gram for such a period.

2. The average daily protein ingestion for these same sub-
_ jects, when permitted to select their diet, was 91.2 grams or 1.33
gram per kilogram of body weight.

360

TABLE I.
Diet for Period 1, November 29 to December 2, 1907. Subjects A to K.

Uric Acid Excretion

S DanNDNOWDGrH wo N
. ie) oaoom ono Oo _
id DBD DCOMHORN 20
(=>) wd oD rt OD CO = NX

1
Se oqomoooo 08 © © 19
BS Neve ay (os) fs) 1) (SS) (08) iF TeX (ey 68) i!
Oorernrtoxowd oe - oo oO
SEED, Co) We) Geire) er COGS) ss) Ee)
SMW~MOCTCTONNOC AS
i MWwoOoocoeo noon oN
OoOonwtrordand 0 6 &

O % © 0% 09 Oo O O HH

= N
i Ve aia aii) Nel ere) Sei im
Owontwtorwndn baal
Ce) (Ce Selec Tee N
rh No of T=) a oO
oO LgeS CO SC S 19-0 rH oS N 0
OoOonrt oro Ho Oo ©
& 0) CO CO MH % O 19 19 Oo oO

eel N
SsyenS SS SS e>) CO sH
g IDS 1S SS 19) cd) Soo Icuad
aon wm ok © Ow oO
ODmOo OO HO OTN On

re re
oS SS iS) Ser) | oO
fea SSO 1538) 160) 7 Gy N
oon st Oo = Cc wm Oo ol te)
oO CO oD Oo oD © oD

— ON |
SS) (SYS SS 1s! ~ey
Oo Vay (ooh |i (a) KA) (0 Slap) NAN oD
a ots OWOON DO oO ty
Omimononmnror © on

ce N
Socojwe aS eS I~ &
faa] Vey (Sy) iS) 1S (et Ne) (eo) =) 198) NAN
on (ST (68 ent (Sy ee) ie'9} CO 10
DO OO MD HM © 19 © ON

al N
SS) ee Sy =) Sy Sy tS) fe lo oe}
< ao co S © 1819 O OS N DD
Co ONS =O Osco Oo Dd
OO WOON DO WO CO Oo re

me N
: : hea AONE RW RAG 9S ie fal
= Sil ele Sa Gorse ee oe ar sa ae
< bre) 9 8 Bi en ek

AS .
Pl SS esos aay © ob

na =) ~~ WTS OS ©

Bo, 8 0 0 5S 6 # = 6 oO 2
Notes eS yray (eh pa (eB oon 28) = Sai Sulee|

ad

Paul J. Hanzlik and P. B. Hawk 361
TABLE II.
Subject A.
Uri id
Period: Date Volume | Specific ais tues Ris Sy nitrogen
& ; of urine. | gravity. | in 24
96 hours. | 24 hours. | hours.
ce. | gram.
l. 3g eee Nov. 29 to
Dec. 2 5,660 | 1,018 2.260 | 0.565 | 0.189
1 Dees 3— 6) 4.1635 | 1,022 | 2.216 | 0.554] 0.185
LCL Dec. 7-10] 4,473 | 1,021. | 2.324) 0.581 | 0.194
LW 3p ose Dec. 11-14 4,050 | 1,0275 | 2.524 | 0.631) 0.211
Rs iS Dec. 15-18 3,/05 | 1,0275 | 2.864 | 0.716 | 0.239
“I Dec. 19-22 |} 5,579 | 1,018 2.576 | 0.644 | 0.215
GHAI 3 ood Ghee 27,630 | 6,1340 | 14.764] 3.691 1233
A STOIGTE » 6 ie ee 4,605 | 1,0223 | 2.461 0.615 | 0.206
Subject B.
lobe eee Nov. 29 to
Dees 2 | 3,797 | 1,028 2.268 | 0.567 | 0.189
“lly Ae Dee. 3-6 3,604 | 1,024 | 2.268] 0.567) 0.189
IN Dec. 7-10 3,389 | 1,027 2.456 | 0.614) 0.205
OW ot | Dec. 11-14 3,926 | 1,026 | 2.336] 0.584) 0.195
oo Oe Dec. 19-22 3,012 | 1,029 2.400 | 0.600 | 0.200
ei eee | Dec. 23-26 3,498 | 1,028 2-371 | 0593 0n198
ILGUa) add ee eee le 21,226 | 6,157 14.099 | 3.525 1.176
ES GTRETS ee 3,536 | 1,026 2.350 | 0.588 | 0.196
Subject C.
ota | Nov. 29 to |
Dee: 2 4,939 | 1,019 | 2.644! 0.661] 0.221
ee ieee 2-6 4,468 | 1,018 2.304 | 0.576 | 0.192
Dl) | Dec. 7-10 4,574 | 1,018 2 300h|) (0255 0.192
Ll) 26. Dec. 11-14 A223) 120235) | 22220) eOrono | Oates
3 eee Dec. 15-18 4,623 | 1,021 2.440 0.610 0.204
an Dec. 19-22 4,845 | 1,022 2.416 0.604 0.202
Total 2612 | Olds) da S24 WS. DST | 1.196
MAES os Phas 4,912 | 1,022 | 2.387| 0.597] 0.199
|

362 Uric Acid Excretion

TABLE II—continued.

Subject E.
Period Date. Volume Speci ae URIC ACID IN GRAMS.
of urine. | gravity.
96 hours. | 24 hours.
cc.
1 gO A ep ie Nov. 29 to
Dec. 2 | 8622] 1,010 2.700
18 PN oe Re Dee. 3— 642—6:72041-1,017 | 27288
Hiss alee Dee 71089 5 0408 «Ors 2.644
TVs ee eee Dec. 11-14 7,130 | 1,0155| 2.868
VER et ne wee: Dec. 15-18 | 6,752 | 1,016 2.476
1) Rae Meek aca Dee. 519-22" |= 6:419) | “L 0175 |) a2
otal ay LO. teens Petes 41,585 | 6,0910 | 15.828
Averacees! (td were ee 6,931 | 1,015 2.638
Subject F.
N ieeastt Bae es eae Nov. 29 to
Dec. 2 4,482 | 1,017 1.912
Tie 8.) eect ee. Sa" 60) 45,043.) ison 1.824
AUS rv erred | Dee. 11-14] 6,374 | 1,015 1.884
Vitale os. eee Dec. 15-18 | 6,517 | 1,014 2.068
Wit oe. gt Ae Dec. 19-22 | 7,535 | 1,0125| 1.804
5] hee nee Ge Dec. 23-26 | 6,790 | 1,013 1.904
UR ctles. Seen. Oo eae ee Poe 36,741 | 6,0835 | 11.396
Aggeraror <%. etaer Es a 6,124 | 1,0139 | 1.899 |
Subject G.
|S Lo ty bee Nov. 29 to
Dec. 2 3,249 | 1,027 2.824
TRA SY ony elt Dec. 3-6] 3,258 | 1,025 2.440
UE a ae eo Dec. 7-10] 2,944 |. 1,029 2.584
LVS op eon Dec. 11-14 | 3,160] 1,031 2.572
Vices in eisens Dec. 15-18 | 2,806 | 1,030 2.640
Vile a ae Dec. 19-22} 3,364] 1,029 2.388
Total yx foc. Ge dlec, sha] PASI Ca iada betas
= ee |

ea
<
7)
ie)
=)
og
@
ue
=
ise)
oO
—"
i=)
bo
19/9)
i)
or
~I
on

Paul J. Hanzlik and P. B. Hawk

TABLE II—continued.

363

Subject H.
Uri id
Period Date. Volume | Specific Pe A? Ee nittozen
f Of LITINe; |) HEaVv1G yo. || nea EE |e:
96 hours. | 24 hours. | hours.
cc. gram.
0. 3 Nov. 29 to
Dec. 2 7,004 | 1,011 2.484 0.621 0.207
aise ss Dee. 3- 6 6,409 | 1,012 2.388 0.597 0.199
LU See Dec. 7-10 6,033 | 1,014 PRT 2D 0.633 O-2E
1 eee Dec. 11-14 6,090 | 1,0195 2.002 0.583 0.195
Oe Dec. 15-18 5,455 | 1,0205 2.508 | 0.627 0.209
en ee ee Dec. 19-22 o,7a2 | L020 2.344 0.586 0.196
Total 37,223 | 6,0970 | 14.588 3.647 We2len
ENSIGRRYUE Co ee ea 6,204 | 1,0161 | 2.431 | 0.608} 0.2038
Subject I.
U4 oe Nov. 29 to
Dec. 2 5,701 | 1,013 2.476 0.619 0.207
hoo Dec. 3-6 5,083 | 1,012 2452 0.613 0.205
Ut he Dee. 7-10 6,373 | LoL2 2.536 0.634 0.212
UW ola se eee Dec. 11-14 D203) L023 2.456 0.614 0.205
Wo hat aoe Dec. 15-18 5,206 | 1,0175 2.464 0.616 0.206
Total 27,566 | 5,0775 | 12.384 3.096 1.035
AORN, es ea 5,513 | 1,0155 2.477 0.619 0.207
Subject J.
ll. 20,6 Nov. 29 to
Dec. 2 6,819 | 1,013 2.060 ORs OSRTZ
Me eas - Dee. 3-6 5,360 | 1,016 1-912 0.478 0.160
I. ai Dee. 7-10 5,059 | 1,018 2.144 0.536 0.179
LV Dec. 11-14 5,179 | 1,022 PAA? 0.528 0.176
noe) oe Dee. 15-18 4,446 | 1,023 2.204 0.551 0.184
Who ee Dec. 19-22 5,658 | 1,020 2.048 0.512 0.171
Total 32,521 |) 16,102 12.480 3.120 1.042
POL AUIG arene ee ohare: She che 2 5,420 | 1,002 2.080 | 0.520] 0.174

364

Uric Acid Excretion

TABLE It—continued.

Subject K.

Uric acid
nitrogen

n
hours.

gram.

0.180
0.209
0.228
0.212
0.259

1.088

excretion (grams

; URIC ACID IN GRAMS.
Period. Date. Volume | Specific
of urine. | gravity.
96 hours. | 24 hours,
Oe
Cte Be oS Nov. 29 to
Dec. 2 4,709 | 1,017 2). 1152 0.538
TD eisai ee Dee. 3- 6 3,412 | 1,025 2.590 0.625
JOE CIES, Mewes eet, Dec. 7-10 4,432 | 1,020 2.728 0.682
DWV ne ae ici ae Dec. 11-14 3,632 | 1,0265 2.536 0.634
A ard lan Pe Dec. 19-22 4,042 | 1,023 3.104 0.776
Ro talen rete gets Segoe 2022/9 Oo, Addon alo O20 3205
INV CT AG Gin tetas hore ecm 4,045 | 1,0223 | 2.604| 0.651] 0.218
TABLE III.
Average daily uric acid excretion.
Average uric acid
SuBseEct. per 24 hours).
IKEA Paneer ie he Je ee ae. NR EO ied ee 0.475
Vists'h hesteetye Bis thar, ween eRe bo Negi mn oan Pt 0.520
SS SRR Hansa ei nee ore cain att nen do ttcnt Atoid Asc cane he 0.588
Gy ean eee ark aa. eas De 5 hacer Ree ede teat Ta ee 0.597
ls epee A ee ae ee ees RL ah Nee ee Me ERR 0.608
1s RTE Cee PH ELE ON MERCR AS ol Eee ee ESAiGia Oo A nly -c 0.615
AN Stra tenicae Sot tN oA ae Reker Hao dprA RARE Rote TARA E EES er 0.619
( CREE RaereRn Dare Fay hetinarte ne nGRF cre Cua ee RRL SS 2 0.644
1 CAE aE A SN ey IPOS ere ete el feo 0.651
Ty Ph had, 8 cecaires ee Beto las Oye SORE eee Rc 0.660
Grand :AVeTage ) oct: Slice Oe nc ees eee 0.597

365

TABLE Iv.
Subjects A to K.

Protein ingestion.

Paul J. Hanzlik and P. B. Hawk

"suBIs

sal oD ¢ ae) mr ©
w aysiem Apoq BHARRBSRAA
‘oly sted ureyo1d sh pave Cis hae tate ae
TOMUOITSO SUN eA [LI (fie tare tte ters eet tod red A rest ioe SLits

o = > oD re
“sUIBIs UI UOTS0S LR ae Ae ae ee
“wud sma FS RZSLALSS

col
mw ~— ri

“SUIBISO[D] LAS ena tt aia RS
Pree aS eee eee sk

Sussect.

dg 0h RB OMHKM

91.2 1.33

68.8

OO ook

4“

aa

STUDIES ON ENZYMES: I.

THE ADSORPTION OF DIASTASE AND CATALASE BY COLLOIDAL
PROTEIN AND BY NORMAL LEAD PHOSPHATE.

By AMOS W. PETERS
(From the Zodlogical Laboratory of the University of Illinois.)

(Received for publication, September 18, 1908.)

INTRODUCTORY.

Most of the experimental work with enzymes and other sim-
ilar obscurely known substances is conducted with preparations
which contain the desired body in admixture with other, in
most cases, undesirable substances. The nature and condition
of the extraneous matter is so little known that it is not possible
to reproduce with assurance and accuracy the same conditions in
the repetition of experiments. Thevalidity and range of applica-
tion of the results obtained thus remains unknown. Progress
of afundamental nature depends largely upon making some ad-
vance in the separation of these substances. The present paper
describes the initial steps in an attempt to apply the method
of adsorption for this purpose. The phenomena described are
usually classed under the term adsorption which is here used,
however, only conventionally. In this paper no attempt will be
made to explain the real nature of this process which is so typ-
ically shown by these experiments.

I take pleasure in acknowledging my obligation to Mr. Bruce
M. Harrison, M.S., of the Graduate School of this University,
for valuable assistance in obtaining many of the numerical data
in the part on the Results of Experiments.

SOURCES OF DIASTASE AND CATALASE.

The diastases upon which these experiments were made were
purposely produced from materials of different origin. This was
done in order that the methods developed might be of general

367

368 Adsorption of Diastase and Catalase

application, as well as to provide means for a possible future
comparison of preparations of diastase of different origins. Ger-
minated wheat, liquid bacterial cultures, and extracts of auto-
lytic liver constituted the principal materials which served as
sources of both diastase and catalase.

It is well known that the wheat grain upon germination pro-
duces as much or more diastase than barley.!. The grain was
washed with tap water until the wash water was clean. It was
then washed with saturated thymol water. It was soaked in
distilled water for a day and then placed upon filter paper and
kept moist in a closed copper germinating pan. The germination
was conducted at room temperature and continued for about
three days, after which the cover was removed and the germi-
nated grain permitted to become air-dry for keeping or it was used
while still moist. Just before extraction with a liquid as subse-
quently described the grain was ground in a mill or a mortar.

The liver of cow or pig was purchased fresh in the market,
superficially washed and then cut or hacked into small pieces.
To one weight of liver there were added three weights of distilled
water and a layer of several centimeters in depth of toluol con-
taining o.5 per cent of thymol was poured on and the whole was
then thoroughly agitated. The agitation was repeated on several
successive days. For use as a source of diastase and catalase the
required quantity of liquid was withdrawn from beneath the
toluol layer with a pipette. The content of diastasein the extract
of liver so prepared first increased and subsequently disappeared.
This material always supplied catalase in great abundance.
The liver thus treated was easily kept for months practically free
from bacterial action.

The bacterial cultures were grown in a liquid medium of the
following composition. Distilled water, 1 liter; monopotassium
phosphate, 1 gram; Witte’s peptone, 2 to 5 grams; Kahlbaum’s
soluble starch, 2 to 5 grams. The whole was sterilized by steam
at 100° C. ina Jena glass Erlenmeyer flask in which the culture
was subsequently raised. The natural reaction of the liquid was
permitted to remain. In this medium numerous kinds of bac-
teria grew abundantly both at room temperature and especially

1 Maerker-Delbriick: Spiritusfabrikation, pp. 106, 287, 1903.

>
id
a

Amos W. Peters 369

upon incubation at 37° C. to 40°C. The first material for inocu-
lation was obtained from the waterand from the decaying veget-
able matter of a small, much polluted stream. The bacteria so
obtained were found to be diastatic but the great variety of
species was undesirable. By plating upon agar, pure cultures
of a diastatic organism were obtained which were used in the
experiments to bedescribed. Thecatalase content of these cul-
tures would develop gradually and reach its maximum after the
amount of diastase had become too small to be of good service
experimentally.

q IDENTIFICATION OF DIASTASE.

The presence and concentration of diastase were always tested
by its effect upon a freshly-prepared 1 per cent solution of Kahl-
baum’s soluble starch. The starch was dissolved in hot water and
a few crystals of thymol added before cooling. Sucha solution as
a rule contained no perceptible amount of reducing sugar. In
testing for diastase one volume of starch solution was thoroughly
mixed with an equal volume of the fluid to be tested, toluol con-
taining o.5 per cent thymol added to a depth of several cen-
timeters, and the mixture incubated at 50°C. for periods vary-
ing from one to twenty-four hours. Sugar was estimated both
at the beginning and at the end of the fermentation period and
frequent controls showed the absence under these conditions of
hydrolysis due to other influences than the diastase.

Sugar was estimated by Bang’s method,’ chosen on account of
its convenience and accuracy. The method is based upon the
reduction of a copper solution in the presence of potassium sul-
phocyanide, the excess of unreduced copper being estimated by
titration with standard hydroxylamine solution. In my deter-
minations, only 2 cc. of the sugar solution were used and hence
only 1o cc. of the copper solution. All of the precautions sug-
gested by Bang were observed. The results are expressed as
volumes of hydroxylamine solution used in the titrations, and
as their equivalents in milligrams of dextrose per cubic centi-

1Bang: Zur Methodik der Zuckerbestimmung, Biochem. Zeitschr., ii,
PP. 271-290, 1907.

370 Adsorption of Diastase and Catalase

meter of the fermenting solution. The nature of the reducing
substances produced by the enzymes studied was not investigated.

As a unit for the measurement of diastase, the production of
o.1 milligram of sugar during the first hour of digestion at 50° C.
was arbitrarily chosen. This amount is equivalent to nearly
0.3 cc. of the standard hydroxylamine solution and lies outside
of the limits of experimental error.

In testing a bacterial culture for its content of diastase the mode
of procedure was as follows: To a measured volume of the cul-
ture fluid (10 to 100 cc.) contained in a tall cylinder, about one-
fifth volume of toluol was added and the mixture thoroughly
shaken. Further bacterial action was thus inhibited. A vol-
ume of 1 per cent soluble starch equal to the volume of cuiture
fluid used was then addec. After shaking and withdrawing a
portion in which the original sugar content of the mixture was
estimated, the mixture was incubated at 50° C. Portions of the
mixture were withdrawn at intervals in order to test the progress
of digestion. The amount of culture fluid used should be such
that an appreciable amount of sugar (by Bang’s method) is
formed during the first hour. A control test was made in which
boiled culture was used. It may here be stated that in the
absence of toluol, digestion with the bacterial culture resulted in
the disappearance of the starch as measured by the iodine reac-
tion but no sugar could be detected.

IDENTIFICATION OF CATALASE.

The term catalase has been used to denote a substance of pro-
toplasmic origin which produces evolution of oxygen from hydro-
gen peroxide, whose activity is inhibited at or below a boiling
temperature, and which when free from peroxidase does not give
the guaiacum and similar color reactions. The unity of this
substance or the reality of its enzymic nature were not subjects
of investigation. The presence of catalase was determined by
treating the liquid under examination with an equal volume
of approximately 3 per cent neutralized hydrogen peroxide.
To facilitate detection of small volumes of gas formed under
these conditions it was often found desirable to pour strong alco-
hol on the surface of the mixture. Bubbles of gas which may

Amos W. Peters 371

otherwise be obscured by the turbidity of the fluid become easily
visible in passing through the clear, alcoholic layer.

SEPARATION OF ENZYMES.

Isolation and purification of enzymes employed in my experi-
ments were accomplished by the use of two procedures, viz:
extraction by acetone and byadsorption. By means of extraction
with 50 per cent acetone, i.e., the addition of one volume of pure
acetone, it is possible to free the fluids which have served as
our sources of enzymes from contaminating proteins, carbohy-
drates and other extraneous matter without affecting the activity
of the enzymes diastase and catalase. The resulting solution
usually gives no reaction with tannic acid. By this means, how-
ever, solublecrystalline material may fail to be removed and hence
adsorption of the enzymes by a number of substances has been
made use of in their isolation in a state of still greater purity.

Experiments have been made to ascertain the influence of 50
per cent acetone upon the diastase and catalase contained in
germinated wheat, autolyzed liver tissue and bacterial cultures.

Germinated wheat. A clear aqueous extract of dried germi-
nated wheat contained both diastase and catalase. A similar
extract made with 50 per cent acetone was found to contain only
diastase. Throughout these tests it was uniformly found that
catalase was insoluble in this fluid. Residues from the extrac-
tion with acetone possessed catalytic activity, hence the ferment
is not destroyed by this reagent. Furthermore, on mixing some
of the catalytically active residue with a mixture of equal vol-
umes of hydrogen peroxide and acetone, oxygen is evolved in
large amounts. Failure to detect catalase in 50 per cent acetone
extracts must therefore have been due to its insolubility in this
medium.

Autolyzed liver tissue. When liver tissue cut in small pieces is
allowed to autolyze in the presence of water and toluol containing
0.5 percent of thymol, a solution results which possesses marked
catalytic power. During the earlier stages of self digestion it also
contains diastase. If such an extract is mixed with an equal
volume of acetone a heavy precipitate results, the filtrate from
which contains no catalase. It does however contain appreciable

372 Adsorption of Diastase and Catalase

amounts of diastase, sometimes a considerable quantity. Asa
tule the diastatic power of such a filtrate is not equivalent to
that of the original solution. The experiments on germinated
wheat showed that diastase was soluble in 50 per cent acetone
and it seems evident, in view of the adsorption experiments to be
described later, that diastase is carried down with the precipi-
tate caused by the acetone. Diastase will therefore be contained
in the filtrate only when this precipitate is not sufficient in bulk
to adsorb all of it. These facts have suggested the possibility of
separating enzymes from their solutions by the addition of a pro-
tein such as Witte’s peptone after treatment with acetone. Such
a method will be described later.

Bacterial cultures. In the separation of the enzymes from
bacterial cultures, filtration is ineffective because of the fact that
the enzymes are apt to be retained by the filter. Centrifuga-
tion is difficult and at best affords incomplete separation. It has
been found however that when an equal volume of acetone is
added to a bacterial culture, and allowed to act on the bacterial
cells for some time before it is removed in great part by evapora-
tion under an air blast at room temperature or at 40° C. the result-
ing unfiltered fluid possessed greater diastatic activity than the
original culture. This may have been due to the action of the
acetone upon the bacterial cells in killing them and rendering
them more permeable to their own enzymes. From a fluid so
obtained more of the ferment could be obtained by the adsorp-
tion methods to be described than from the original fluid. This
method may have extended application in the extraction of
bacterial cells for other purposes.

ADSORPTION OF THE FERMENTS.

Witte’s peptone. By the addition of Witte’s peptone in ro per
cent watery solution to a solution containing diastase or cata-
lase and subsequently adding an equal volume of acetone a preci-
cipitate will be obtained in which the ferments are concentrated.
I have made use of this observation in the methods employed
in these experiments. As a reagent 20 grams of Witte’s peptone
were dissolved in 200 cc. of hot water. It should not be boiled
but the heat should be great enough to destroy any ferments

ae

Amos W. Peters 373

which might be present. After cooling an equal volume of ace-
tone is added, the abundant precipitate allowed to subside and
the supernatant fluid removed by decantation. To the precipi-
tate is added enough 50 per cent acetone to make the mixture up
to the original volume, i.e., 200 cc. This reagent, a 10 per cent
suspension of precipitated peptone in 50 per cent acetone, is
applied for the adsorption of diastase and catalase from liquids
containing them. Such liquids are treated with equal volumes
of acetone and to cc. of the peptone suspension, containing 1
gram of peptone, are added. The precipitate with the adsorbed
enzyme, is removed by decantation or by the centrifuge. It may
be washed with so per cent acetone though this is usually unneces-
sary. If the separation of the enzyme is not quantitative the
process can be repeated until all of the enzyme is so removed.
The precipitate can then be digested in water and the action of
the adsorbed enzymes tested.

This method is of advantage not only because it concentrates
the enzymes but especially because it removes such impurities as
may be precipitated by 50 per cent acetone and such soluble
crystalline substances as are not absorbed by the peptone. The
enzyme separated by this method, is however associated with
the peptone from which it is indistinguishable.

Lead phosphate. The lead phosphate reagent used in this
method was prepared by precipitating 100 cc. of 0.3 M solution
of lead acetate with 100 cc. of 0.2 M solution of sodium phosphate,
washing the precipitate thoroughly with water and finally sus-
pending itin soocc.of water. According to Dammer! normal
lead phosphate does not hydrolyze in cold water and only slowly
in water at 50° C.

For adsorbing the enzymes studied in these experiments, one
volume of this suspension of lead phosphate was added to the
fluid containing diastase, either aqueous or containing 50 per
cent of acetone. A more concentrated suspension of lead phos-
phate may be used if dilution is to be avoided. The advantage
which this reagent possesses over the Witte’s peptone is that it
favors more complete purification of the enzyme. The insolu-
bility, real or supposed, of the enzymes in condition of adsorption
does not prevent their action upon their appropriate substrata.

1 Handb. d. anorg. Chem., iv., p.560, 1903.

374 Adsorption of Diastase and Catalase

The methods described in this paper for obaining diastase
and catalase from different types of materials may be briefly
summarized as follows:

Solid materials, e.g., germinated wheat or autolyzed liver, are
extracted with water, toluol-thymol being used as an antiseptic.
An equal volume of concentrated acetone is added to the extract
and the mixture warmed at 40° C. for a short time. A heavy
precipitate formed under these conditions is apt to contain most of
the enzymes, diastase and catalase. If no precipitate forms, the
enzymes are absorbed from the fluid by means of Witte’s pep-
tone suspended in 50 per cent acetone or by normal lead phosphate
suspended in water.

Liquid bacterial cultures are treated with equal volumes of
acetone. If lead phosphate is used as the adsorbent, it is added
to the mixture in watery suspension after the bulk of acetone has
been removed by evaporation. For the peptone method ace-
tone is again added in the same quantity (or it is not removed at
all) and the mixture finally treated with a suspension of Witte’s
peptone in 50 per cent acetone.

RESULTS OF EXPERIMENTS.

The experiments described in this section illustrate the applica-
tion of these methods in a variety of conditions.

Adsorption by Lead Phosphate.

Ten cc. of an extract of germinated wheat made with 50 per cent ace-
tone were thoroughly mixed with r1occ. of lead phosphate suspension.
The lead phosphate (with its adsorbed enzymes) was removed by the cen-
trifuge and made up to to ce. with distilled water. This suspension,
designated as Precipitate 1, was mixed with 1o cc. of a 1 percent soluble
Starch solution, and incubated at 50°C. after estimating sugar in the mix-
ture by Bang’s method.

The centrifugate obtained above was treated with the lead phosphate
obtained by centrifuging 10 cc. of the watery suspension of lead phosphate
used as adsorption reagent, and the mixture made up to 20 cc. with dis-
tilled water. After thorough shaking, the lead phosphate (with its
adsorbed enzyme) designated as Precipitate 2, was removed by the cen-
trifuge, made up to ro cc. with water, mixed with an equal volume of a
I per cent solution of soluble starch and, after estimation of its sugar con-
tent, incubated at 50° C.

Amos W. Peters 375

Sugar estimations were made at the end of 18 hours and at the end of
24 hours incubation.

The results are given in the following table:

TABLE 1.

VOLUME OF HYDROXYLAMINE.

Initial After18 | After 24 | Hydroxyl-| SUGAR.
reading hours of | hoursof | _ amine
incubation. incubation. | difference.
ce. ce. ce. cc. mg. / cc.
Precipitate No.1....... rise 4.7 4.7 2.4 1.0
Precipitate No.2....... Cao ies ek 0.45 0.2
irate NO.2.......... 6.4 6.4 6.4

These results show that the first treatment with lead phosphate
resulted in the adsorption of about five-sixths of the diastase, the
remainder being removed by the second addition of lead phos-
phate.

It is probable that a given quantity of lead phosphate has its
characteristic saturation capacity and hence to remove all of the
enzyme from a fluid it may be necessary to treat it more than once
with the adsorbent. The relation between quantity of adsorb-
ent and quantity of enzyme requires further investigation than
has here been given it. Similar results to those given above
have frequently been obtained and show the capacity of lead
phosphate to adsorb diastase completely and in an active condi-
tion.

The next series of experiments showed the power of lead phos-
phate to adsorb diastase from extremely dilute solutions.

One cc. of the acetone extract of germinated wheat used above was
diluted to 100 cc. with water, to cc. of lead phosphate suspension were
added, the mixture thoroughly shaken, and after settling, the supernatant
fluid was decanted. The precipitate was made up to ro cc. with water,
mixed with 1occ. of 1 per cent soluble starch, sugar content estimated and
incubated at 50°. The hydroxylamine values were as follows:

Before incubation, 8.3 cc.; after 3 hours, 6.4 cc.; after 54 hours, 6.1 cc.;
after 23 hours, 6.1 cc. The hydroxylamine difference is thus2.2 cc. which
represents very nearly 1 mg. of sugar per cc. that was produced by the

376 Adsorption of Diastase and Catalase

adsorbed diastase. Negative results were obtained when the filtrate from
the lead phosphate was similarly tested.

A series of experiments to determine the possible effect of lead
phosphate upon diastatic action was made as follows:

Six cc. of lead phosphate suspension were added to 4 cc. of a 50
per cent acetone extract of malted wheat. After thorough shaking,
to cc. of 1 per cent soluble starch solution was added after estimation of
sugar in a portion of the mixture, the rest was incubated at 50°C. for 48
hours. Asa control, a mixture of 4 cc. of the same extract, with 6 cc. of
water and ro cc. of soluble starch was similarly incubated. The results
follow:

TABLE 2.

VOLUME OF HYDROXYLAMINE.

ppm vias ae Initial After 48 |Hydroxyl-| Accelera- | 57°4*
reading. hours. ,amine. tion.
difference.
ce. ce. cc. ce. mg. /ce.
4 cc. extract of wheat }
10 “ percent starch 4.7 0.9 3.8 ey, 0.7

f

6 “ Pb; (PO,). J

4 “ extract of wheat |

10 “ 1 per cent ee 4.2
6 ‘ distilled water

Phe

i)
—

The results show that the action of diastase is markedly accel-
erated by the presence of lead phosphate. The possible causes of
this acceleration have not been studied.

The diastase contained in an extract of autolyzed liver tissue
may similarly be adsorbed by lead phosphate as the following
experiment indicates:

Fifty cc. of aqueous extract of liver tissue which had been allowed to
undergo ro days autolysis was mixed with 50 cc. of lead phosphate sus-
pension. After shaking thoroughly and allowing the precipitate to
settle 75 cc. of supernatant fluid were removed by decantation and the
residue (25 cc.) made up to 50 cc. with a 10 per cent solution of soluble
starch. This was digested at 50°, portions being withdrawn for sugar
estimation at the end of 14, 3 and 21 hours. The hydroxylamine differ-
ences were 0.4, 0.7, and 1.4 cc., respectively. The last value is equivalent —
to about 0.6 mg. of sugar per cc.

—

<h—s

Amos W. Peters a7

The following experiment indicates the applicability of the
method in the removal of diastase from bacterial cultures.

One hundred cc. of the bacterial culture were thoroughly shaken with
20 cc. of toluol and 83 cc. of the aqueous fluid removed by a siphon. To
this was added an equal volume of lead phosphate suspension. On separa-
tion of the precipitate from the supernatant fluid by decantation, both
were found to contain diastase. If however the lead phosphate precipi-
tate is collected upon a filter paper by means of a Buchner funnel and suc-
tion and the supernatant fluid passed through the layer of lead phosphate
on the paper, the latter takes up practically all of the diastase. The
results given below show that this is the case. The precipitate with the
filter paper was suspended in water and 1 per cent soluble starch added in
such proportion that the volumeof the mixture was the same as thatof the
original culture and contained 0.5 percent of soluble starch. The filtrate
was mixed with an equal volume of 1 per cent soluble starch and hence the
fluid is twice as dilute as the original culture.

TABLE 3.
Precipitate.
Duration of Hydroxylamine
digestion, hours. difference.
ae See AES Se nA AS Oe Bee aan 0
EY Ce lS yrer.c; 0 '< arava Sista ia isis Wlodowa ele EROS aS
ae IEE PN SS cc chal ai ol sche: aemvevers 0s caleueneekeke, a 2.0
Filtrate.
Duration of Hydroxylamine
digestion, hours. difference.
Ee a Si cra gta ed jad Dee ad ais wees 0
SER ey 8 2p e os. a) ey ater ele 2 oheje ape sila ard ayabe aye Stave 0
a PI rs a 52. o Tans gaia “speedy Saueree oe 0.2

These results show that this procedure was successful in
removing practically all of the diastase from the culture fluid.

The power of some other substances to adsorb diastase and
catalase in a similar fashion has beentested. Zinc phosphate was
found to be less efficient in adsorbing diastase and possessed no
power of increasing its activity as did lead phosphate. Precipi-
tated gelatinous aluminum phosphate was less active than zinc

phosphate. Zinc oxide used in the same manner gave precipi-

tates and filtrates which were wholly inactive.

The following experiments represent attempts to remove the

) enzyme from the lead phosphate after adsorption.

378 Adsorption of Diastase and Catalase

The diastase from to cc. of a strongly diastatic extract of
germinated wheat was adsorbed by means of lead phosphate,
the latter was thoroughly washed with water, made up to 20 cc.
with water and 20 cc. of 0.002 per cent phosphoric acid added.
The mixture was allowed to stand for two days at room tempera-
ture: at the end of this time, 1o cc. of the clear, acid superna-
tant fluid was withdrawn and tested for diastase with negative
results. The lead phosphate in this mixture when washed free
of acid was found to be strongly diastatic. The failure of acidi-
fied water to dissolve the diastase may possibly be connected
with the inhibitive influence which the acid of the medium would
exert upon the very small hydrolysis of normal lead phosphate
at room temperature.

The power of alkalies to remove the enzyme from the adsorb-
ent was not tried on account of the well-known power of alka-
lies to inactivate diastase.

Distilled water and so per cent acetone were tested in this
regard. The results showed that appreciable amounts of dias-
tase could be extracted from the lead phosphate precipitate by
distilled water; 50 per cent acetone was found to be about one-
half as active as water in this respect. Fifty per cent glycerine
was also found to be an excellent solvent and preservative. It
should be noted in this connection that the purer an aqueous
solution of diastase is, the greater is its instability and this fact
may have a bearing upon the above experiments. Since lead
phosphate adsorbs the enzyme in a very active condition there
is for many purposes no need of its solution. The preparation
of the thus far imaginary pure diastase was not sought in these
experiments.

Adsorption by the Protein Method.

The following experiments illustrate the application of the
method of adsorption of diastase by peptone.

An active aqueous extract of germinated wheat was prepared and fil-
tered. 25 cc. of acetone were added to 25 cc. of the extract, no precipitate
being formed. rocc. of the peptone reagent was added and after standing
for an hour with occasional shaking the precipitated peptone was removed
by the centrifuge. Both precipitate and supernatant fluid (designated
filtrate) were tested as follows: The precipitate, treated with 25 cc. of

~

—<

Amos W. Peters 379

water and mixed with 25 cc. of 1 per cent soluble starch, was incubated
at 50° for 1 hour. At the end of this time the sugar estimation showed
an hydroxylamine difference of 2.9 cc., equivalent to 1.2 mg. of sugar per
ce. 20 Cc. of the filtrate was mixed with 20 cc. of 1 per cent soluble starch
which made the fluid containing the enzyme about 5 times as dilute as the
original extract. Sugar estimation in this digestion mixture gave an
hydroxylamine difference of 2.8 cc. in the first hour, showing that a large
portion of the diastase had not been adsorbed. By repeated application of
the peptone reagent practically all of the diastase could have been removed
from the original extract.

Another experiment illustrates the removal of diastase by this
method from an extract of autolyzed liver tissue.

Fifty cc. of an extract of liver tissue autolyzed for 7 days was mixed
with socc. of acetone and warmed at 40° for 15 minutes. The precipitate
so obtained was filtered off, suspended in distilled water and warmed at 40°
for 3hours. At the end of this time the insoluble matter, designated below
“Residue” was filtered off. The filtrate from the residue is designated
the ‘‘ Digest.” The ‘‘ Residue’ was mixed with distilled water and starch
solution so that the volume of the mixture was the same as the volume of
the original liver extract. A portion of the ‘‘ Digest’? was mixed with
starch solution in proportion to make the dilution 3 compared with the
original liver extract. Another portion of the ‘‘ Digest’? was mixed with
lead phosphate reagent, the final dilution of the diastase in this mixture
also being 3. All were incubated at 50° with the following results:

TABLE 4.
Residue.
Hydroxylamine
Duration of difference.
digestion, hours. ce.
ny 5 Ste aa ee 0
I eee ts ica a wine or ajocere everson © Oepaale 0.7
a PR Pe a a4) oF) oy dy gnro's athe os 0: Sau oeoge Lars
Digest
Hydroxylamine
Duration of difference.
digestion, hours. ce.
nals ells g SEARS ae 6 2 0
IE Fn Me 2 SN Bios Lae a) asinl a! 0cky S4ah pupttopanss & hea OF2
Las lies EE Bae A gn ee ae on 0.4
Lead Phosphate.
Hydroxylamine
Duration of difference.
digestion, hours. ce.
SE CONT Ree So Oe cic ag ie Bare petal 0
Se eae ee te ey sic ot ehhotiel at ck ovate the 0.2
PM Oi 8 oes, sl dM eer eees 1.0

\

380 Adsorption of Diastase and Catalase

These figures show that much diastase remained unextracted
in the residue and that such quantity as is extracted is easily
concentrated by lead phosphate, thus showing that the two
processes can be used together. The accelerating effect of the
lead phosphate is of course represented in the above figures.

The method as applied to a bacterial culture is illustrated by
the following experiment:

Twenty-two cc. of a perfectly clear fluid very rich in diastase was removed
by decantation from a 7 days’ old bacterialculture. 22 cc. of acetone were
added, without forming a precipitate. 6 cc. of peptone reagent were then
added, and the precipitated peptone removed by the centrifuge. The
fluid was again treated with 6 cc. of peptone reagent, and the precipitated
peptone removed by the centrifuge. Each of these precipitates was
digested in 22 cc. of water and 22 cc. of 1 per cent soluble starch solution.
Sugar estimations after incubation at 50° gave the following results:

TABLE 5.
Precipitate 1.
Duration of Hydroxylamine
digestion, hours. difference.
Og eT Re AE ht BE Eh cS a) Se 0
UE Gia Sage Sete ts iMate Mente cuenta Mang Me imo RRS OS a (Ne 7
PR ia ee Fo eM LT alg he Ge ST a a 0.9
ELS goa eR ig a eck ape at Sie Go vae Weis caer ne bet Choy se one
Precipitate 2. °
Duration of Hydroxylamine
digestion, hours. : difference.
Ee eer SIS 2e ol oe es 2 Vid SANS ie ieee 0
Tye Se eat eek oe en RAGE cae on a A ae Om
Sn) NE RR? et IORI AIC ENS edt | Sent MEO NaS eno i i tae

QUESTIONS FOR FURTHER INVESTIGATION.

There are important aspects of the method of adsorption and ©
also of uses to which it might be applied that have not been ©
developed in this paper. Among these may be mentioned its ©
quantitative relations, such as thesaturation capacity of a given —
amount of adsorbent, also the quantitative separation of the
enzymes. Further questions pertain to the purity of the en-~
zymes adsorbed and the feasibility of studying their properties
as they are exhibited by the adsorbed substance. There also re-—
mains the fundamental question of the real nature of the process
here only conventionally denominated as adsorption.

A
+
5
G Pee! :
ving |
© -
al
a] ” -
ny Ne
i
‘ :
f ui
> -
-
mal al . ce
£
* /
A
y
a
7 . =
= ie -
~ y A a
-
ei
whe
- i
“ an
p Dd
of
A °
4 - ied 4 MS ;
. “ae i petri
; = | Pend a
Fa : 4 4
. , aie ane ag

i ab

ms - ' i Va
rt Daa

ON THE SYNTHESIS OF PROTAMIN THROUGH FERMENT
ACTION.

By ALONZO ENGLEBERT TAYLOR.
(From the Hearst Laboratory of Pathology, Untversity of California.)
(Received for publication, December 12, 1908.)

In the third volume of this Journal! I reported the results of
a successful experiment in the synthesis of protamin from the
products of its digestion, through the agency of a trypsin obtained
from the liver cf the large clam of the Pacific Coast, the Schzz-
otherus Nuttallit. The controls were negative, no combination
of the products of the digestion of the protamin occurred on
standing, in the simple mixture of the products or in the presence
of the ferment inactivated by heating. In the strict application
of the laws of organic chemistry, however, the experiment ought
to be carried out with the use of the isolated and purified amido-
acids themselves. This experiment I have since performed, in
the winter of 1907-1908. The publication of the results has been
delayed by illness.

Four hundred and fifty grams of salmin sulphate (calculated
as the base) were digested with a commercial preparation of
trypsin. The material was dissolved in 20 liters of warm water
the solution made faintly alkaline and the digestion maintained
until one part of the solution could be mixed with three parts of
absolute alcohol, acidulated with sulphuric acid, without the

_ production of an opacity. The solution of the digestion products

was then concentrated, the sulphuric acid carefully removed with
barium hydroxide, the slight excess of barium removed with
carbon dioxide, and the resultant mixture of amido-acids sub-
mitted to the Fischer method for the isolation of the esters of the
monamido-acids. Thedistillation of the esters was carried out

Taylor: This Journal, iii, p. 87, 1907.
381

382 Synthesis of Protamin

in two fractions; one at 100° and 10 mm. pressure, the other at
200° and 10 mm. pressure. ‘The distillate obtained in the first
fraction was saponified by prolonged boilingin water. The dis-
tillate secured in the second fraction was saponified by heating
in baryta water, the barium removed with sulphuric acid. These
two solutions of amido-acids were then concentrated at 50° and
a pressure of ro mm. until the first signs of crystallization appeared
at room temperature; that is, concentrated to saturated solutions.
Three hundred and seventy-five grams of arginin, prepared from
previous lots of salmin, were then converted into the carbonate
according to the method of the Kossel school, and a saturated
solution formed therefrom. In such a solution are represented
three states of arginin,in equilibrium; the carbonate, bicarbonate
and hydroxide. These three solutions were then mixed. From
this solution, which formed the substrate of the experiment, two
small controls were prepared: 2ne consisting of the plain solution;
the second consisting of an equal amount, to which had been
added some of the glycerin extract that had been inactivated by
heating. The remaining bulk of the solution was divided into two
parts. To one, representing half of the material, was added
roo cc. of a fresh glycerin extract of the livers of the variety of
clam used in the previous experiment. To the other lot was
added 2 grams of pancreatin (Grtibler’s- Pancreatin nach Spate-
holz). To each, toluol was added in large amount, and the four
tests then set aside for nearly four months at room temperature,
sealed.

As time passed cloudiness developed in the two larger tests;
the two controls remained clear. At the end of neariy four
months, the containers were opened and the contents tested
bacteriologically, with negative results. All the four solutions
were then diluted with four volumes of water, acidulated with
sulphuric acid and mixed with three volumes of absolute alcohol.
Heavy white precipitates occurred in the two larger tests; none
in the controls.

The precipitate in the test with the glycerin extract of the clam
liver was washed with alcohol, redissolved in water, precipitated
and washed with alcohol, again dissolved in water and a third
time precipitated and washed with alcohol. Finally it was dis-
solved in water, precipitated as the picrate and again recovered

pg yen a dhe

Alonzo Englebert Taylor 383

according to the Kossel method for the purification of salmin.
It was finally again precipitated from solution in water by alco-
hol, washed with alcohol, then absolute alcohol, then ether, the
ether removed with a stream of dry air, and the powder dried
im vacuo over sulphuric acid for a week, then dried for several
days im vacuo at 105°. It was a white, rather fluffy powder,
and weighed 5.3 grams. This powder was soluble in thirty parts
of water at room temperature, was salted out of solution by satu-
ration with ammonium sulphate and by ro per cent concentration
of sodium chloride. On mixing with blood serum, a heavy
colloidal precipitate was produced. It was easily digested with
trypsin, but was entirely resistant to pepsin. One gram of the
powder, on hydrolysis with hydrochioric acid, yielded 0.89 gram
of arginin. The elementary analysis of the powder gave the
following figures:

0.3402 gm. of substance yielded 0.5028 gm. CO, and 0.2051 gm. H,O.

0.3024 gm. of substance yielded 0.079 gm. of nitrogen.

0.5400 gm. of substance yielded 0.2659 gm. barium sulphate.

There are two slightly varying formulae ascribed to salmin,
based upon different analyses by different men. I tabulate them
as follows:

SS Fe a a ces bin sho a boned age pop aes Ci 6H oN ,O,.-
PULLS. 050.8: 6d Se See ee eee ears See Conti O,
at hae aoe yess. Sete oon he 5 al Oa oS CU,H aN 0:
IS ELI, | VE. sic a is ce bie sos ls Wee's oA C,,HoNO-.
IE 5 2 FS hy ao vlwiabena 6: a)a-h ewiwaed Ci-HN,O;
ee og sy wn, ipa a -mimlars CU, .
a cal es Se Cipla Os:
ee Oe sc vw.a wich: ss es vse wn eae CLE Oe

The figures of Goto! agree with none of these fully, best with the
last figure of Kossel. The variations in the hydrogen and oxygen
may be disregarded, the real difference is in theratio of carbon to
nitrogen. This ratio is for the formula C,,H,,N,O,, 1777: 1000;
for the formula C,,H;,N,,0,, 1765: 1000. If the figures for the ele-
mentary analysis are to be contrasted with the figures demanded
by these tw equations, the larger formula must be written with
the same relations in the hydrogen and oxygen that obtain in the
smaller. Then we have:

* Goto: Zeitschr. f. physiol. Chem., xxxvii, p. 94.

384 Synthesis of Protamin

Calculated for Found: Calculated for
CigH 29N 902-H2SO,: CoH g3N1704.2H2SO,;
Ci ki eee ee eee 40.25 40.3 39.52
Lc errr ane Cer or 6.50 6.7 6.25
bee kee weten ree oreraae 26.41 26.2 26.12
ELSO M8 teers ee 20.55 20.7 21.51

The figures present a satisfactory agreement with those de-
manded by the first equation. During the course of the experi-
ment therefore there has been formed from the amido-acids a
substance having the percentage composition of salmin. This
substance has the common reactions of salmin, though these
are of little value as criteria for the question under consideration.
A specific reaction for salmin is not known. From the stand-
point of the physical chemist there can be no question that in
view of the theory of reversed reactions, the substance might
without further investigation be denominated as salmin. From
the point of view of organic chemistry, however, the qualitative
demonstration in the strictest sense has not been accomplished;
it has not been shown that the substance obtained by synthesis
is the same in molecular construction as that natively residing in
the spermatazoa of the salmon.

This aspect of the question demands particular attention in
view of the fact that in the reported syntheses of disaccharides
from hexoses under the action of ferments, the iso-sugars have
been obtained, not the natural disaccharides from which the
primary sugars were derived. There can be no question that in
these several investigations disaccharides have been synthesized
from monosaccharides, but there has been a qualitative devia-
tion. In the language of Wegschneider, there has been not only
a “‘Folgewirkung”’ but also a ‘‘Nebenwirkung.” In view of
the little we know of the iso-sugars, these reversion experiments
urgently require repetition and amplification. But there can
be no question, in view of the results of the elementary analyses
that disaccharides have been formed from monosaccharides
through the action of inversion-ferments.

Salmin is composed of four amido-acids, as described by Kos-
sel: arginin, serin, prolin, and amidovalerianic acid. If the
molecule of salmin contained but one molecule each of the four
named amido-acids (disregarding the stereoisomerism of the

amido-acids) it is evident that twelve isomeric molecular forms ~

Alonzo Englebert Taylor 385

would be possible. Each molecule of salmin, however, contains
not one molecule each of the component amido-acids, but several.
Kossel and Dakin! have suggested two formule:

to mol. arginin + 1 mol. amidovalerianic acid + 2 mol. prolin + 2 mol.
serin;

12 mol. arginin + 1 mol. amidovalerianic acid + 3 mol. prolin + 2 mol.
serin.

In another paper? I suggest another equation, very like those
of Kossel and Dakin:

12 mol. arginin + 1 mol. amidovalerianic acid + 2 mol. prolin + 3 mol.
serin.

The number of isomers in such an equation would run into
the hundreds. The investigations of the Fischer school have
taught us that in the study of the polypeptides, the condensation-
products of amido-acids that so much resemble peptones, it is
possible to determine to some extent the relations in the linkings
in the chains of amido-acids. These methods are not at present
highly enough developed to be applied to the study of protamins;
we know nothing of the linkings of the amido-acids in their
molecules. Since we do not know the internal structure of the
molecule »f salmin as it occurs in nature, we cannot determine
whether the substance of the same percentage cumposition and
generally known reactions obtained in the experiment herein re-
ported is identical with the salmin that occurs in the spermatazoa
of the salmin or an isomer. There can be no question that a
protamin has been synthesized; thequalitativestudy of the prod-
uct must however await new and exact knowledge of the molec-
ular structure of natural salmin. I believe it to be the same
protamin, there are no apparent reasons to doubt it; but it must
be admitted that from the standpoint of the strict organic chem-
ist it is not now possible to prove that in this molecule exist the
same linkings of amido-acids that exist in the natural protamin
in the spermatozoa of the salmon. The velocity of the reaction
is low, the yield very small—as has been the fact with all ferment
reversions. The remainder of the material will be preserved for
future investigations.

1 Kossel and Dakin: Zeitschr. f. physiol. Chem., xli, p. 407, 1904.
* Taylor: This Journal, v, p. 394, 1909.

386 Synthesis of Protamin

The material precipitated by alcohol in the test with bovine
pancreatin is not a protamin. The amount, was 3.8 grams. It
is salted out of watery solution by saturation with ammonium
sulphate, but not by a 1o per cent concentration of sodium
chloride. It is not only digestible with trypsin, but also by pep-
sin as well. The analysis for nitrogen indicates that the sub-
stance cannot be a protamin.

0.3229 gram of substance yielded 0.0716 gram of nitrogen = 22.2 per
cent.

This figure is far too low for salmin sulphate, being much nearer
that of arginin sulphate. This substance is not, however, to be
derived from the preparation of pancreatin used in the experiment.
One and a half grams of this pancreatin dissolved in 60 parts of
water and mixed with 3 volumes of alcohol will not give a heavy
precipitate. Nevertheless, in the future, fresh pancreatic juice
alone is to be employed. I have done nothing further with this
material. Rather than consume it in analytical studies that
would in all probability reveal nothing, I shall reserve it for the
following future experiment: the reintroduction of the material
into a saturated solution of the amido-acids concerned, in the
presence of fresh trypsin. This experiment oughtin my opinion
to be done with the iso-maltose and iso-lactose; they should be
placed in strong solutions of the appropriate hexoses, with fresh
portions of the inversion-ferments to determine whether further
changes would occur.

Additional investigations with the ferments in the liver of the
clam have yielded interesting results. The enzymic strength
of different glycerin extracts varies greatly at different times,
with different collections, and at different seasons of the year.
The habits of the clams are so little understood that it is not pos-
sible to attempt to relate the variations to any variables in their
habits of life. Certain is one fact: that at the times when the
glycerin extract is quite inert, examination of the digestive appa-
ratus will indicate that the animal is fasting. There are great
seasonal variations in the condition of the clams; in the summer
they are usually fat and large, in the winter often poor and thin.
Investigations of the glycerin extract have also elicited one other
fact of importance: it contains (apart from the amylase that does

Alonzo Englebert Taylor 387

not here concern us) not only a trypsin, but also an erepsin.
This is made probable by the fact that at neutral reaction the
extract will digest peptone with rapidity, but is able to make
very little impression upon native albumin. The glycerin ex-
tract contains but traces of protein. With three parts of alcohol
but a slight yellowish precipitate is produced. Tannicacid, how-
_ ever, brings down a much heavier precipitate.

It is my intention to repeat these experiments next season
upon a very large scale, with the use of fresh ferments from dif-
ferent sources, and with the experiments extended over longer
intervals of time, up to a year at least. With proper technique,
it ought to be possible, without violating the conditions of bac-
terial sterility in the tests to introduce at regular intervals fresh
portions of active ferments, in order that the enzymic action
may be maintained at the maximum, and the inactivation of the
ferments circumvented. So far as I can determine by compara-
tive tests, the trypsin from the clam differs from the trypsin of
higher mammalians only in an increased resistance to hydroly-
sis; in general the ferments from the cold-blooded animals, espe-
cially the marine animals, are much more resistant than those
derived from higher animals. If the fresh pancreatic juice of the
dog were added each week to such a reversion experiment, the
result ought to be the same. In time I shall have such experi-
ments to report. Some of the ferments of vegetable origin de-
serve also a trial.

Some of the analyses herein reported were done in the labor-
atory for physical chemistry of the Nobel Institute, Stockholm,
and in the chemical laboratory of the Stockholm High School.
For these facilities, I take pleasure in expressing my appreciation
to Prof. Svante Arrhenius, and to Prof. H. v. Euler.

ON THE COMPOSITION AND DERIVATION OF PROTAMIN.
By ALONZO ENGLEBERT TAYLOR.

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

(Received for publication, December 12, 1908.)

According to the investigations of Kossel and Dakin! the
protamin of the salmon yields on hydrolysis arginin, an amino-
valerianic acid, serin and prolin. Abderhalden? analyzed a
preparation of salmin and reported further the positive finding
of leucin, alanin, with probably phenylalanin and aspartic
acid. Having in hand material of exceptional purity, I have
been able to reinvestigate the matter, and here report the
results of this study, together with some general considerations
bearing upon the question of the derivation of protamin in
the metabolism of the salmon. The results of my analyses
show that the statements of Kossel and Dakin are correct, those
of Abderhalden in error.

To obtain a pure protamin, much more is necessary than to
follow the methods of Miescher. The desiderata are of two
kinds: the proper collection of the raw material, and the meth-
ods of Kossel for the isolation of the base. The ripening of
the salmon testicle is a process lasting weeks. Careful studies
upon the part of those engaged in the extensive salmon hatch-
eries in California have shown that from four to six weeks are
consumed in the ascent of the fish to the mouths of the Battle
Creek and the McCloud river, where the hatcheries are situated.
Salmon caught in the brackish water of the bays of San Fran-
cisco and San Pablo and in the tide waters of the lower Sacra-
mento river (where the fish still feed), display in their testicles
not a little protamin. Some of the male fish are ripe on their

1 Kossel and Dakin: Zeitschr. f. physiol. Chem., xl, pp. 311, 565, 1904;
xli, p. 407, 1904.
* Abderhalden: Zettschr. 7. physiol. Chem., xli, p. 55, 1904.
389

390 Composition and Derivation of Protamin

arrival at the waters of the hatcheries; others remain for weeks
after their arrival in the impounding basin before the sperm has
ripened. The process of ripening has two main stages: the
maturation of the spermatazoa, which requires weeks, probably
months; and the liquefaction of the intercellular connective
tissue of the testicle, whereby the testicular tissue becomes con-
verted into fluid sperm. This process of liquefaction does not
occur in the testicle en Lloc, but extends through some ten days;
in other words, the entire testicle does not. become ripe at once.
In the practical operation of the hatchery utilization is made
of this fact. As the salmon ascend the stream on which the
hatchery is located, they are trapped and impounded in a large
basin in the stream, salmon-runs being installed at the lower
end of the pond, while above the further ascent of the fish is
barred with wooden gratings. This pond is seined several times
daily. All the ‘“‘ripe females” caught are retained and placed
in crates; the ‘‘unripe females’ are returned to the pond. Only so
many of the “‘ripe males”’ caught are retained as may be neces-
sary to keep the pens designed for them filled, the rest arereturned
to the pond. For the purpose of the hatchery the males are
always in large excess. At a certain time of each day, the
‘‘ripe females’? caught upon the previous day are killed and
their eggs placed in containers. The .male salmon are taken
from their pens, allowed to lie upon a platform out of water
until they are so far stupefied by lack of respiration that they
may be handled, when they are held over the dishes containing
the eggs and the sperm ejected therein by gently milking the
ventral surface of the fish. When no further flow of sperm is
to be obtained, the male fish are returned to their pens, where
recovery soon occurs. The male fish will be stripped in this
manner daily until the entire testicle is liquefied and the fish
exhausted. A healthy male salmon can be stripped daily for
from six to ten days. Protamin prepared from unripe testicles
will be mixed with protein, histone, and amido-acid in amounts
proportional to the immaturity of the material, due to the fact
that the unripe glands contain blood, tissue protein, connective
tissue and histone. Protamin prepared from ripe glands will
contain but traces of these, since no blood, cellular protein or
histone remain in the glands, while there will be traces of nucleinic

“ie

Alonzo Englebert Taylor 391

acid, and of amido-acids derived from the liquefaction of the
intercellular connective tissue. The nucleinic acid and the
amido-acids tend to cling to the protamin sulphate when the
latter is precipitated with alcohol from aqueous solution. If
the salmin sulphate be separated from a hot saturated solution
in the oil-like state, most of the salts and a part of the organic
impurities, especially the proteins and histone, will be removed.
If later the combination of the picrate be prepared, the remainder
of the organic impurities will be removed if the initial material
wasmature. The proteins cannot be entirely removed in this
way. If they be present in considerable amount they will pass
into the oil-like state in a condition of adsorption, and will
remain with the protamin when precipitated with picric acid;
when the protamin is finally recovered, part of the protein will
be found with it. If one is working with a protamin sulphate
obtained from unripe testicles, where the admixture of protein
and histone is large, one carrying-through of these two manipu-
lations is not sufficient to insure a pure product, as the protein
clings tenaciously; if one is working with a fully ripened material,
containing the least traces of protein, one will secure a pure
product, as the nucleinic acid and amido-acids will be completely
removed by the two procedures of purification. In other words,
the method of Kossel is adapted to yield a pure product when
applied to a ripened initial material, for which it was in fact
devised; it is not suited to the manipulation of unripe initial
material, nor is any known method. These facts I have stated
in detail, because to judge from the work of Abderhalden, they
do not seem to be generally known. In his publication’ he
says: ‘‘Den Einwand, den man iwbrigens gegen jede Unter-
suchung an Proteinstoffen erheben kann, dass das untersuchte
Material keine Guarantie fiir chemische Einheitlichkeit biete,
suchte ich durch méglichst sorfaltige Darstellung einzuschranken.”
He does not state in what the special precautions in the prep-
aration consisted. In his later text-book,? following the appear-
ance of the article of Kossel and Dakin, he states in reference
to this preparation of salmin, that his own results * * *

1 Abderhalden: loc. cit., p. 56.
2 Abderhalden: Lehrbuch der physiologischen Chemie, pp. 191, 192,
1906.

392 Composition and Derivation of Protamin

‘‘welcher von einem sorgfaltig gereinigten Salmin zur Beobach-
tung kam, durfte vielleicht in der nicht vdlligen Reife der Hoden,
aus denen das betreffende Produkt gewonnen war, seine Erk-
larung finden.” It is to be inferred that Abderhalden obtained
his material from a distance and at second hand, and was not
therefore in a position to control the collection of the material.

Sixty grams (calculated as the base) of most carefully purified
protamin from the quinnat salmon df the Pacific coast, obtained
from fully matured material, were hydrolyzed with hydrochloric
acid and the resultant solution ot amido-acids submitted to the
Fischer ester method. In the distillation of the esters but two
fractions were made: the first contained the esters that would
distill over at less than 100° on a water bath at a pressure of ro
mm.; the second fraction contained all that would distill over
up to 200° on an oil bath, at a pressureof 1omm. The first
fraction was saponified by prolonged boiling in water, the
resultant solution carefully concentrated to a syrup, and this
then dried to a gummy mass in a stream of hot air at 50°. It
was then extracted with warm absolute alcohol three times.
The combined alcoholic extracts were evaporated to dryness
and the residue again extracted with absolute alcohol, this
time with a large volume, over night at room temperature. A
slight residue remained which was returned to the original residue
of the alcoholic extraction. ‘The amido-acid contained in the
alcoholic extract, presumably prolin, was secured by crystalli-
zation on the evaporation of the alcohol. It weighed 6.5 grams.
A portion of this was twice recrystallized from absolute alcohol, and
submitted to analyses for carbon, hydrogen and nitrogen, the
latter by the Kjeldahl method.

The results were:

Calculated for

CsHoNOo: Found:
Cie cay LR ne ea oe che 52.2 per cent. 52.0 per cent.
1a gh AAA a oe IR OO OOSS 0 as , 8.0 %
Ni Succyehecehens a Geaictel Scie vetns teaeeenen 12.2 Pye il 7

The figures agree well with those for prolin.

The residue after the extraction with alcohol was dissolved in
the minimum amount of hot water, filtered to remove a slight
flocculation, and hot alcohol added until the first traces of cloudi-

ede

ee ae

e

era

Alonzo Englebert Taylor 393

ness appeared. On cooling and standing over night, a heavy
crystallization occurred. The crystals were collected by filtra-
tion and the filtrate evaporated to dryness; but a slight trace of a
yellowish gummy material remained, weighing but a few milli-
grams. Thecrystals obviously included the entire solute. These
crystals were again dissolved in hot water and hot alcohol added
until cloudiness appeared. On cooling and standing over night,
the crystals had again separated out. They were collected,
washed with dilute, then absolute alcohol, finally with ether; the
weight was 3.21 grams. The crystals were twice recrystallized
from alcohol. On analysis the substance, presumably amido-
valerianic acid, gave the following figures:

Calculated for

CsH1,NOp: Found:
Os oe oe 51.3 per cent. 51.1 per cent.
1 ee 9.4 = eS
Le 11.9 a 221 e

The figures agree well with those for amidovalerianic acid.

The fraction »f ester secured at the higher temperature was
saponified by heating with baryta water, the barium carefully
removed with sulphuric acid, filtered hot, the filtrate concen
trated to a syrup, to which hot alcohol was added until a cloudi-
ness appeared. On cooling and standing over several days a
heavy crystallization appeared. The crystals were filtered off,
and the filtrate evaporated to dryness; but a trace of gummy
material remained. The crystals again obviously represented
the entire solute. The crystals were three times recrystallized
from water-alcohol, and weighed 5.22 grams. ‘The substance,
presumably serin, gave on analysis the following figures:

Calculated for

7NOz: Found:
A 34.3 per cent. 34.2 per cent.
8 ae 6.6 > 6.8 es
. oc eS 13.3 S 13.5 cs

The figures agree well with those forserin. In all these figures
I have used the rounded figure for the first decimal, since it can
be easily shown that the second decimal has no certainty in an
elementary analysis.

394 Composition and Derivation of Protamin

Further identification of these substances I have not carried
out, on account of enforced absence from the laboratory. Were
the subject new, further identification in the qualitative sense
would be necessary ; in view of the work of Kossel and Dakin, how-
ever, I do not consider that it can be doubted that the substances
of the denominated percentage compositions were prolin, serin
and amidovalerianic acid. It is also clear from the results of the
evaporation of the filtrates that the three amido-acids are all
the monamido-acids that were obtained from the salmin by the
Fischer ester method.

A further lot of salmin sulphate (30 grams calculated as salmin)
was hydrolyzed with hydrochloric acid and the arginin isolated
according to the method of Kossel. The amount was 27.52
grams. From all the figures it is evident that for the prepara-
tion of salmin here under analysis:

One hundred parts protamin + 1 water = 91.73 parts arginin
+ 10.83 parts prolin + 5.35 parts amidovalerianic acid + 8.70
parts serin. Kossel and Dakin! have attempted, from the quan-
titative data of a similar analysis, to determine the composition
and molecular weight of salmin. They suggested two formule:
10 mol. arginin, 2 mol. serin, 1 mol. amidovalerianic acid, and 2
mol. prolin, leading to the formula, C,,H,,,N,;0,,; and 12 mol.
arginin, 2 mol. serin, 1 mol. amidovalerianic acid and 3 mol. prolin
leading to the formula, C,,H,,,N;,0.,. Neither of these formule
gives the carbon: nitrogen ratio observed in the analyzed prep-
arations of salmin —C,,H,,N,O, and C,,H,,N,,O,. The later anal-
yses by Goto of the salts with platinum chloride agree exactly
with neither of these. An equation that does give exactly the
carbon: nitrogen ratio of the first of the above two equations,
and which also, with one exception, presents a good agreement
with the analytical results herein reported is the following:

12 arginin (C,H,,N,O,) + 3 serin (C;H,NO, + 1 amidovalerianic acid
(C,H,,NO,) + 2 prolin (C,H,NO,) = 1 salmin (C,,H,,,N,,0,) + H,0.

Comparing now the amounts of arginin, serin, prolin and amido-
valerianic acid which such a molecule would yield on hydrolysis —

with the figures for the amounts obtained in the analyses herein ©

reported, we have in 100 parts:

1 Kossel and Dakin: loc. cit., xli, p. 414, 1904.

Alonzo Englebert Taylor 395

a
se oS ee 94 .07 91.73
2 a A ee ere ae 10.16 10.83
0 Se a oe 5.46 Sa
fs oh aie Aer 13.91 8.70

The agreement in the case of arginin, valin and prolin is good, in
the case of serin very poor. It is, however, known that serin is
isolated with great difficulty, being apparently in the state of the
ester not well extracted with ether, and it is possible to account
for the deficit in some such way. Upon the basis of this equation,
the molecular weight of salmin is 2274, a figure vastly lower than
those known approximately for the higher forms of protein. In
his earlier investigation Kossel was led to define the protamins
as the simplest forms of protein. Their general behavior was so
different from that of the common forms of protein, that at that
time objections were energetically raised against the application
of the term protein to protamin. The studies of the Fischer
school upon the amido-acids contained in the proteins have fully
established the correctness of Kossel’s position that protamin
is as typical a protein as serum albumin. Most of the higher
forms of protein contain from ten to fifteen different amido-acids
while salmin contains but four. A consideration of the relations
of spacial isomerism as applied to the linkings of the amido-
acids in a molecule of protein will indicate that a substance con-
taining four such components is certain to be more simple in its
chemical structure than one containing a dozen components.
The figures given above have naturally only a provisional value,
though from the theoretical point of view, an undoubted impor-
tance does attach to them.

Kossel* has sketched an interesting scheme for the derivation
of the protamins. Protamin he regards as being derived from
histone, and this to have been built up in the salmon from the
fragments (amido-acids) of the disintegration of protein of muscle
since the testicle is in part formed during the period during which
the salmon ascend the streams t2 spawn, at which time they fast
and their muscles waste; in the case of the other fish, the histones
are naturally to be derived from the amido-acids (respt. protein)

Kossel: Zeitschr. f. physiol. Chem., xliv, p. 347, 1905.

396 Composition and Derivation of Protamin

of the diet. This general scheme, which has in its favor every
known fact, can in the opinion of the writer be made much more
definable if it be interpreted from the standpoint of reversible
reactions. The facts are few and certain. In the testicle of the
salmon of the average size are formed some 30 grams of protamin.
To provide the necessary amido-acids, less than half a kilo of
muscle would be required. The migrating salmon does not feed
after leaving salt or brackish water. At this time the testicles
are rich in histone and contain not a little protamin, derivable
from the proteins of the diet; the amount that must be later
formed during the period of fasting is thus much less than the
total, and the figure of a half kilo of muscle is largely excessive.
The observations of the experts in charge of our fish hatcheries
tend to show that the mate salmon lose from 2 to 5 kilosin weight
during the period of migration and spawning. This is partly
fat, partly muscle. During this period the fish are actively
engaged in ascending streams against the current, in driving
marauders from the nests and in combats with each other, all
of which require maintenance, in part without question by mus-
cular tissue. Nevertheless, it is certain that the amido-acids
required for the formation of the histone are easily available in
the muscle known to be disintegrated during this period, and
there is no necessity to consider any other source. The hypothet-
ical question concerns the modus operandi of the formation of
histone from the amido-acids of the muscle.

When proteins are ingested, they reappear in the circulation
in the form of serum albuminand globulins. No matter what the
protein in the diet, in the stage of resorption there is an intra-
molecular rearrangement, so that the circulating protein is in
the state of serum albumin and globulins. These blood proteins ;
are carried in the circulation to the various tissues, and from them —
the cells form myosin, reticulin, gelatin, etc. These transform- —
ations concern not only intramolecular rearrangements, with the ©
maintenance of the same molecular mass; but include also the
introduction or extrusion of particular groups of amido-acids. A
In general, the common proteins contain the same amido-acids,
but in different amounts and inferentially in different linkings; —
but some proteins contain certain amido-acids, such as cystin or
phenyl-amido-acid, that are not present in others, or but to —

Alonzo Englebert Taylor 397

slight degree. The faculty of the body to effect such intramolec-
ular rearrangements is well illustrated in the sugar metabolism.
No matter in what form sugar is ingested, in the blood we find
but one form, d-glucose; in other words, the levulose and galac-
tose are converted into d-glucose by intramolecular rearrange-
ment. This glucose in the blood passes into the active mammary
gland, where it is in part reconverted into d-galactose, which
combined with d-glucose forms milk sugar. D-galactose is also
formed from d-glucose in the central nervous system. That the
different hexoses tend thus to pass into each other was first
shown by Lobry du Bruyn and van Erenstein;} it has been exper-
imentally demonstrated for the mannoses by Neuberg and Mayer.”
In this same manner we assume that the protein of muscle is
formed from the proteins of the blood serum in the salmon. For
the formation of the histone of the testicle in the salmon, we
may now choose one of two conceptions. According to the one
apparently adopted by Kossel, the muscle protein is disinte-
grated 7m situ, and the amido-acids thus set free in the circulation
are carried to the testicle, there to be converted into histone.
The second possible conception is to assume that the protein of
the blood serum is taken into the testicular tissue, there hydro-
lyzed by enzymic action, and the histone then built up from the
amido-acids. The reduction in the proteins of the blood serum
thus occasioned is then made good by the reconversion of muscle
protein into serum protein, the reversal of the reaction whereby
during the period of feeding the muscle protein is formed from the
serum protein. In the same manner, as the protein of the blood
serum is burned to maintain the movements of the fish, the loss
is made good by the reconversion of muscle protein into serum
protein. Just as the blood furnishes d-glucose to the mammary
gland, which forms from it d-galactose, so the blood furnishes
to the testicle of the fasting salmon the common protein, from
which histone is formed. In the hypothesis of Kossel, the his-
tone is formed fr»m amido-acids derived from the cleavage of
muscle protein in the muscles and transported to the testicle by

1 Lobry de Bruyn and van Erenstein: Ber. d. deutsch. chem. Gessellsch.,
XXViii, pp. 3078, 3085, 1898; Rev. Chim. d. Pays-Bas., xiv, pp. 116, 203,
1898; xvi, p. 263.

? Neuberg and Mayer: Zeitschr. f. physiol. Chem., xxxii, p. 256, 1901.

398 Composition and Derivation of Protamin

the circulation. In the hypothesis here suggested, the histone
is formed from the proteins of the blood serum, the cleavage of —
which is assumed to occur in the testicle; since the blood proteins
are in the fasting fish held to be derived from the muscle protein
the histone is of course derived from the muscle, though indirectly.
In favor of the hypothesis here suggested is the fact that the pro-
teins of the blood serum constitute the substratum from which
are formed all the specialized proteins of the body, and it is
natural to believe that this modus applies also to the formation
of histone. In this entire argument it is assumed that histone
is constructed from preformed amido-acids; that these might
be synthesized in the testicle is not impossible, but there is every
reason to believe that the building-stones of the histone, and thus
of protamin, are derived from the hydrolysis of the body protein.

ON THE QUESTION OF THE IDENTITY OF PEPSIN AND
CHYMOSIN.

By ALONZO ENGLEBERT TAYLOR.
(From the Hearst Laboratory of Pathology, University of California.)

(Received for publication, December 12, 1908.)

Since the earliest days in the study of the digestive ferments,
pepsin and chymosin have been considered as distinct substances
endowed with properties of very different nature, though both
the secretion-products of one organ. Within recent years, how-
ever, through the work of Nencki and Sieber,! Schoumow-Siman-
owski,? Sawjalow,? Gewin,* and especially of Pawlow,® the idea
has been advanced that they are in reality but one substance.
The older conception has been defended by Schmidt-Nielsen,®
Bang,’ Hemmeter,*® Jacoby® and especially by Hammarsten,’° to
whose work so much of our present knowledge of the coagulation
of milk is due. Owing to the indefiniteness of our conceptions
of the chemical nature of the coagulation of milk, the proponents
of the unistic theory cannot be expected to put into concrete
terms just what they conceive the action of pepsin upon the milk
to be, but in general terms we may say that the conversion of
casein into paracasein is held to be the result of the proteolytic
action of the pepsin upon the casein (probably a reaction of
hydrolysis), the actual coagulation of the paracasein being, as

1 Nencki and Sieber: Zeitschr. f. physiol. Chem., xxiii, p. 291, 1901.
? Schoumow-Simanowski: Arch. f. exper. Path. u. Pharm., XXXiil,
P. 336, 1899.
3Sawjalow: Zeitschr. f. physiol. Chem., xlvi, p. 307, 1905.
*Gewin: Ibid., liv, p. 32, 1907.
5 Pawlow and Parastschuk: Jbid., xlii, p. 415, 1904.
® Schmidt-Nielson: Ibid., xlviii, p. 92, 1906.
7 Bang: Ibid., xliii, p. 358, 1904.
8 Hemmeter: Berl. klin. Wochenschr., Ewald Festnummer, 44, 1905.
* Jacoby: Biochem. Zeitschr., i, p. 53, 1905.
4°Hammarsten: Zeitschr. f. physiol. Chem., lvi, p. 18, 1908.

399

400 Identity of Pepsin and Chymosin.

generally admitted, a purely secondary and extraneous process.
The unistic conception is with strict consistency applied to the
pancreatic juice, which contains also a curdling ferment; this
is held to be identical with the trypsin. The proposition of the
unistic school is in chemical terms to be stated concretely as
follows: the coagulating and the proteolytic properties of the
molecule of pepsin reside in different groups within the molecule,
just as an organic molecule may contain a carboxyl and an hy-
droxyl group. There is unquestionably a legitimate application
of the well-known facts of organic chemistry to this problem. In
innumerable instances is it known that reactions of different
types may be carried on independently with the different groups
contained in a large organic molecule or even in relatively small
ones. Upon the basis of the current conception of fermentative
acceleration as consisting in the establishment of intermediary
reactions, the chemical properties of the different groups within
the molecule would determine whether a certain substance could
be an accelerator for a certain reaction; and a single molecule
could in all correctness be assumed to contain different groups
that would qualify it to act as the accelerator for different
reactions. The digestion of protein is an act of hydrolysis; the
presence of pepsin introduces into the reaction between the pro-
tein and water new intermediary steps whereby the internal
resistance in the molecule of pepsin is reduced and the velocity of
the cleavage increased. There is no theoretical reason why the
molecule of pepsin should not contain another group that bears
the same relation to the reaction between casein and water,
whatever the nature of that reaction may be. A simple scheme
will make the matter clear.

AR-PEPSIN-RB
Proteose (slow reaction).
AR-Pepsin-RB = Proteose (rapid reaction).
Paracasein (slow reaction).
BR-Pepsin-RA = Paracasein (rapid reaction).

Protein + water
Protein + water
Casein + water
Casein + water

steal ote all

There can be no theoretical reason why the two reactions should
not be carried on simultaneously. Whether this scheme is true
or not, is a matter to be determined by investigation. Whether
the assumption be a natural and obvious one or a distant and

Alonzo Englebert Taylor. 401

vague conjecture, is also of no consequence so far as the theory
of fermentation is concerned; upon the theory of fermentation as
representing the acceleration of a reaction through the reduction
in the internal resistance by the installation of new intermediary
reactions, the hypothesis of Pawlow is logical and permissible.
It is possible to go further; there is no reason why all the enzymic
activities of the pancreatic juice (lipolytic, proteolytic, amylytic,
inversion and coagulation) should not be ascribed to the differ-
ent groups of one organic molecule. It is not necessary to in-
voke the Ehrlich hypothesis; the theory of catalytic reactions
and the facts of organic chemistry are broad enough to support
the proposition. For the future investigation of the question
it will be necessary to determine in every way the relations of
coagulation and proteolysis.

Gastric secretions occur in man, under conditions of disease,
in which the proteolytic property remains and the coagulating
property is absent. It has been long known by clinicians that
in individuals suffering from carcinoma of the stomach, the dis-
appearance of chymosin from the gastric contents was apt to
occur early, and a certain diagnostic value has been attached
to this finding. I have observed, in the examination of many
gastric contents from cases of pyloric carcinoma, five instances
in which the gastric contents would digest protein but could not
coagulate milk. In order to determine the actual state of secre-
tion in the subjects of this disease, certain preliminary procedures
are necessary. The stomach should be carefully cleaned by lav-
age on several successive days, and the diet carefully regulated.
Then the patient is to be given an Ewald test-meal and a dose of
hydrochloric acid, say 10 cc., well diluted. After an hour anda
half the gastric contents are removed by expression, and the
tests for the different secretions carried out. In the cases here
reported, the proteolytic activity was tested upon coagulated
egg-albumin. The coagulating activity was tested by adding
the neutralized juice to fresh neutral milk in regularly varying
amounts, and the tubes held in the thermostat; for controls plain
milk and milk to which the neutralized boiled gastric contents
had been added, were employed. In five instances in my expe-
rience, the gastric juice displayed proteolytic activity, but would
not coagulate milk.

402 Identity of Pepsin and Chymosin.

Such facts would be interpreted directly and naturally by
Hammarsten as follows: Pepsin is one substance, chymosin is
another substance; in these cases of carcinoma of the stomach,
the secretion of the one substance, pepsin, is not affected, the secre-
tion of the other substance, chymosin, is inhibited. According
to Pawlow, the interpretation would be something as follows:
Pepsin contains two reaction-groups, one accelerating the hydrol-
ysis of protein, the other accelerating the conversion of casein
into paracasein; in these cases of carcinoma of the stomach, this
substance pepsin has been so modified in the cells in which it was
formed or after its formation there, that the group in which the
proteolytic function resides has remained unaltered, while the
group in which resides the coagulating property has suffered such
alteration as to deprive it of its reaction qualities. A decision
of the question, such facts as those here recorded cannot yield.

In favor of the unistic interpretation is the parallelism that is
usually found in gastric contents between the proteolytic and
the coagulating activities. Against the unistic theory are the
following facts; it is possible by thermic and chemical manipula-
tion to affect the one activity and not the other. It is possible
to prepare pepsin, free from chymosin, and chymosin, free from
pepsin. Jacoby’ prepared an antiserum by the injection of
gastric ferments into animals; the antiserum contained an anti-
chymosin, but no antipepsin. In the case of the pancreatic
juice, Vernon? and Wohlgemuth*? have shown that the action
and the activation of the trypsin and the coagulating ferment
run parallel. The antiserum obtained by Jacoby after injections
with pancreatic juice was both anti-tr\ ptic and anti-coagulating.

To the students of ferments 1t must seem that thedata at pres-
ent at hand are best interpreted as indicating that pepsin and
chymosin are different substances. This provisional conclusion
can be combated by showing that in all the chymosin-free
pepsins and pepsin-free chymosins the inactivity of the one func-
tion is the result of inhibiting substances (as was shown for
Glassner’s chymosin-free pepsin by Pekelharing*), and by the

1 Jacoby: loc. cit.

? Vernon: Journ. of Physiol., xxix, p. 302, 1903.

3 Wohlgemuth: Biochem. Zettschr., ii, p. 350, 1907.

‘ Pekelharing: Arch. d. sci. biol., xi, suppl., p. 36, 1906.

al
*

Alonzo Englebert Taylor. 403

demonstration that the coagulating activity can be restored to
chymosin-free pepsin and the proteolytic activity to pepsin-free
chymosin. Naturally a chemical group can be so completely
removed by substitution that its reaction-qualities cannot be
restored; but this must be demonstrated to occur in the prepara-
tion of pepsin-free chymosin and chymosin-free pepsin, unless
the above stated requirements can be fulfilled. The theory that
pepsin and chymosin are one substance cannot be proved by
quoting the lateral chain hypothesis of Ehrlich; the application
of this hypothesis to the phenomena of fermentation in the pres-
ent state of our knowledge is arbitrary and devoid of experimental
basis.

ON THE INVERSION OF CANE SUGAR AND MALTOSE
BY FERMENTS.

By ALONZO ENGLEBERT TAYLOR.
(From the Hearst Laboratory of Pathology, University of California.)

(Received for publication, December 12, 1908.)

The recent papers of Hudson! induce me to publish some
figures that I have obtained in work upon the same subject,
but which I have for some time held back pending a larger
investigation of the subject. In my work I obtained the same
results as those obtained years ago by O’Sullivan and Thompson
for cane sugar; in the individual series, the rate of inver-
sion was proportional to the concentration of the substrate,
and when the values were calculated according to the equation
for a monomolecular reaction, constants in good agreement were
obtained. With different initial concentrations of the substrate,
however, all other conditions being identical, the constants in
the different series did not agree. The results have been held
back pending the investigation of this lack of agreement in series
of different initial concentrations. Since the publication of the
results of the investigations of Hudson, which have confirmed
the results of O’Sullivan and Thompson, there is no longer any
purpose in withholding the figures, pending the study of the un-
solved question now being undertaken by Professor Arrhenius and
myself. In my work with cane sugar and maltose, I followed
the methods of O’Sullivan and Thompson, the solutions were
made alkaline before being placed in the polariscope. For mal-
tase I employed a taka-diastase, which is rich in maltose; for
sacchrase, I used a preparation from brewer’s yeast. I have
always used low concentrations. Temperature 140°. As illus-
trations I give a series each for maltose and saccharose, the con-

1 Journ. of the Amer. Chem. Soc., xxx, pp. 1160 and 1564, 1908.

405

406 Inversion of Cane Sugar and Maltose

stants being calculated according to the equation for a mono-

molecular reaction: = = C (A-x). t are hours.
t

Saccharose: concentrations 4, 1 and 2 percent. Ferment con-
centration constant in all. Constants are X 107.

| 4 |) @ lee tha | pe 1° 6
per cent | | | |
4 | 420 410 416 | 430° | 426 | 428 | 434 423
1 | 320 330 | 312 | 322 | 338 332 | 326 326
2 | 160 1650 |e OOS) a2 LG 168 | 170 167
| | | |

Maltose: concentrations 1, 2 and 3 per cent. Ferment con-
centration constant. Constants are X Io °.

Gee a b= el tiple: 4 5 6 7 | Mean.
per cent
1 360 | 362 | 3 | 370° |. 376 -} 348 354 363
2 | 220 268) | 222 | 242 238. | 242 238 | 239
3 | 140 132 | 136 | 160 | 140 138 | 145 | 142
| | | |

To these I will add, from work already published, the con-
stants for a similar experiment in the digestion of starch with
salivary amylase. The substrate concentrations were 4, 4 and
# percent. Time is minutes. The temperature was 35°. The
constants are x 710°:

(t) 30 45 | 60 75 90 | 120 | 150 | 180 | Mean.
per cent | | }

4 490 465 455 470 465 | 455 560 45 Seale ei

4 | 480] 420} 390] 415] 405} 395| 430| 410 | 412

3 | 390 370 | 385 | 390 480 | 370 | 365 3¢0) 1" 390

In all these series it is clear that the conformity in the con-
stants in each series is satisfactory. Between the constants in
the several series, however, the conformity is not present. On
the contrary the lower the substrate concentration the higher the
figure for the constant. This behavior of the figures for the con-

+ tae Qual RR

74

Alonzo Englebert Taylor 407

stants is to be seen in the results of many investigators in this
field, though the matter has been given little discussion. In the
experiments whose constants are given above, the total conver-
sions were about 60 per cent for the lower concentrations, less
of course for the higher concentrations. The behavior of the
constants does not lie in this, as it persists if the conversions be
carried on so that the several series overlap. Under the strict
application of the law, the constants should be the same for the
tests with 4,1 and 2 percent substrate concentrations. In fact,
they are not, though they are in conformity in the individual
series. This state of affairs is, however, not peculiar to the reac-
tion of fermentation, but is seen, though to a less pronounced
extent, in other catalytic reactions. For example, the constants
for the acid inversion of cane sugar are not identical in Io and
30 per cent substrate concentrations; the difference here observed
is, however, much less than is observed in the case of the action of
invertase, and is in any event proportional to the osmotic pres-
sure.

My experiments, therefore, confirm the results recently pub-
lished by Hudson for cane sugar—that the fermentation follows
the course of a monomolecular reaction—and indicates that the
same holds true for the fermentation of maltose. In this last,
my results are contradictory to those recently published by Mlle.
Piloche,! according to which the constants are regular only when
calculated according to the equation of Henri. A careful review
and recalculation of the data in her paper, however, show that
the results are not regular; in some experiments the normal law
is followed, and in other cases there is a maximum in the curve,
the cause of which cannot be discerned from the publication and
will need to be determined and elucidated by additional experi-
mental work.

1Piloche: Journal de chemie physique, vi, p. 229, 1908.

44

NOTE ON THE OXIDATION OF GLUTAMIC AND ASPARTIC
ACIDS BY MEANS OF HYDROGEN PEROXIDE.

By H. D. DAKIN.
(From the Laboratory of Dr. C. A. Herter, New York.)
(Received for publication, December 21, 1908.)

In previous communications! it has been shown that a number
of amino-acids of the typical formula R.CH.NH,.COOH, under-
go oxidation with hydrogen peroxide so as to yield ammonia,
carbon dioxide and an aldehyde R.CHO, the latter being more
or less oxidized to the corresponding acid, according to the con-
ditions of the experiment:

R.CHNH,.COOH + O = R.CHO + NH, + CO,
R.CHNH,.COOH + 0, = R.COOH + NH, + CO,

This reaction was found to occur in the case of glycocoll, ala-
nin, a-amidovaleric acid, a-amido-isovaleric acid, leucin and
phenylalanin, and was of especial interest on account of the sim-
ilarity of the products of oxidation with those obtained through
some biochemical oxidative changes.

The following experiments were undertaken in order to ascer-
tain whether the reaction could be extended to the dibasic amino-
acids. This was found tobethecase. Glutamic acid proved to
be readily oxidized when the ammonium salt was gently warmed
with hydrogen peroxide, with liberation of ammonia and carbon
dioxide and formation of large quantities of succinic acid. The
reaction may be expressed as follows:

CH,.CHNH,.COOH CH,.COOH
CH,.COOH CH,.COOH

1 This Journal, i, pp. 171 and 271; iv, p. 63.

409

410 Oxidation of Glutamic and Aspartic Acids

The reaction is clearly analogous to the oxidation of the mono-
amino-acids under similar conditions. Aldehyde-butyric acid,*
COH.CH,.CH,.COOH, may be considered as an intermediary
product, but no indications of the presence of this substance were
obtained.

The formation of succinic acid from glutamic acid through
oxidation with hydrogen peroxide is of interest since a precisely
similar reaction is brought about by the action of yeast (Ehrlich)
and also through the agency of putrefactive organisms.”

The oxidation of aspartic acid with hydrogen peroxide showed
that essentially the same reaction took place. The change is
complicated, however, by the fact that the primary product of
the reaction, namely the half aldehyde of malonic acid is very
unstable and breaks up with extreme ease into acetaldehyde and
carbon dioxide, both of which products were readily identified.
A small amount of malonic acid was isolated, however, it doubt-
less being formed from the further oxidation of the aldehyde
acid.

In addition, minute quantities of acetic acid and formic acid
resulting from the further oxidation of the acetaldehyde were
identified.

The changes may be represented as follows:

CH.NH,.COOH CHO COOH

+ O— NH, + CO,+ | +0— |
CH; COOH CH,.cCOOH CH,.cCOOE

CO, + CH,.COH + 0 > CH,.COOH.

EXPERIMENTAL.

Oxidation of glutamic acid. One-tenth gram mol. of glutamic
acid was neutralized by the addition of a slight excess of ammonia
and then gently warmed to about 70° with 0.3 gram mol. of
2.5 per cent hydrogen peroxide. After about an hour the liquid

was distilled. The residue did not reduce Fehling’s solution, ~

aldehyde-butyric acid was therefore absent. It was acidified
with phosphoric acid and extracted with ether in a continuous

‘Perkin and Sprankling: Trans. Chem. Soc., 1xxv, p. 16.
?'W. Brasch and C. Neuberg: Biochem. Zeitschr., xiii, p. 299.

He Ds Dakin A4II

extraction apparatus. On distilling off the ether a crystalline
residue of succinic acid was obtained amounting to 47 per cent
of the theoretical amount. It was recrystallized from water and
melted at 180-181°. A considerable amount of unoxidized glu-
tamic acid was present in the solution after extraction with ether,
so that by repeating the oxidation it would be possible to still
further improve the yield of succinic acid.

Oxidation of aspartic acid. Sodium aspartate, 0.1 gram mol.,
was oxidized with 0.3 gram mol. of neutral hydrogen peroxide.
A trace of ferrous sulphate was added as a catalyst. After
standing for a few minutes at the ordinary temperature of the
room, the liquid became warm and a vigorous reaction took
place with evolution of much carbon dioxide. The liquid was
acidified and distilled. The distillate contained acetic and for-
mic acids equivalent to approximately 0.25 gram of acetic acid.
In addition all the usual reactions for acetaldehyde were obtained
including Rimini’s reaction with sodium nitroprusside and pipi-
ridine. The distillate was redistilled and the first portion to come
over was treated with para-nitrophenylhydrazine acetate in the
usual way. The small amount of precipitated hydrazone was
recrystallized from alcohol and melted at 127-128.° The para-
nitrophenylhydrazone of acetaldehyde melts at 128-128.5°.

The residue from the first distillation was acidified with phos-
phoric acid and extracted with ether. On evaporation of the
ether a crystalline residue of malonic acid was obtained. It was
identified by its melting point (131-132°) and by its rapid decom-
position with liberation of carbon dioxide at the same tempera-
ture. The yield of malonic acid was considerably less than the
corresponding yield of succinic acid from the oxidation of glu-
tamic acid. ;

¢ a te
aie en oa ae

om

THE ACTION OF GLYCOCOLL AS A DETOXICATING AGENT.!
By Ho Ds DAKTN:.
(From the Laboratory of Dr. C. A. Herter, New York City.)

(Received for publication, December 21, 1908.)

The power of the living organism to convert poisonous sub-
stances into easily soluble, relatively non-toxic, compounds which
are excreted by the kidneys is commonly spoken of as a protec-
tive mechanism. The detoxication of the poisonous substance
is commonly effected through oxidation, reduction, hydrolysis or
union with some other substance such as glycocoll, sulphuric
acid, glucuronic acid, etc., or a combination of these processes.
But the fact that a substance when introduced into the body
undergoes one or the other of the reactions just mentioned is not
in itself sufficient proof that the original substance was toxic,
although this appears not infrequently to be tacitly assumed.
Thus, for example, benzoic acid when introduced into the body
is promptly converted into hippuric acid and eliminated by the
kidneys, but it is by no means clear that this is an actual protec-
tive mechanism for we know nothing of the action of benzoic
acid itself upon the animal organism. By administering very
large doses of benzoic acid to some animals, it is true that a small
portion of it may be excreted unchanged, but observations under
these conditions when the experiment is complicated by the simul-
taneous formation and excretion of large amounts of hippuric
acid can teach us but little. As regards the lower organisms
such as bacteria, Dr. Herter informs me that he has observed but
little difference in the toxicity of equivalent sodium benzoate and
sodium hippurate. It is clear therefore that the coupling of a
substance with glycocoll in its passage through the organism

1 The term ‘‘ detoxicating agent”’ is used in this paper as an equivalent
of the widely used German word ‘‘Schutzmittel.”

413

414 Detoxicating Action of Glycocoll

is in many cases not definitely proven to be a protective mechan-
ism.

The present paper contains definite examples of the lowering
of the toxicity of an aromatic acid resulting from its combina-
tion with glycocoll and also contains an account of experiments
upon the relative ease of oxidation of three aromatic acids and of
the products formed by their union with glycocoll. Theseexperi-
ments show that the resistance of the glycocoll derivatives to
oxidation is enormously greater’ than that of the free unpaired
acids. It might be thought that union with an amino-acid such
as glycocoll might bring the substance more readily within the
sphere of the oxidizing ferments of the cell, for it is known that
the amino-acids are readily withdrawn from the general circu-
lation. This appears, however, not to be the case.

The substances examined were as follows:

Substance, Formula. Result,
Phenylpropionic
acid .........C,H,.CH,-CH;:COOH. .).... diame ena
dose = about
0.8 gm. per
kilo. Easily
oxidized.
Phenylpropionyl-
glycocoll-.... C,H,.CH,.CH,.cCO.NH.CH,.COOH Non-toxic in
doses of 1.5
gm. per kilo.

Not so easily
oxidized as
phenylpropi-
onic acid.

Cinnamic acids cs 66 25). ee ee ab bl vie, ge em 0 ee
phenylpro-
picnic acid.
Readily oxi-

I dized.

Cinnamoylglycocoll ...ccc06s cone sone ce +s on « e NOM=TOmIE eee
sistant to oxi-
dation.

Hy DP. Dakin 415

Substance. Result.
{ Phenyl-6-oxypropionic acid....................Lesstoxic than
phenylpro-
pionic acid.
Moderately
easily oxi-
dized.

Phenyl-f-oxypropionylglycocoll.................Non-toxiec.
Very resistant
to oxidation.

The free acids were prepared by the usual methods. The gly-
cocoll derivatives were synthesized by the methods recently de-
scribed in this Journal.1 The substances were administered by
subcutaneous injection of the sterile solutions of the neutral
sodium salts, with the exception of cinnamic acid which on
account of the insolubility of the sodium salts, was neutralized
with ammonia.

The results were as follows:

(1) Phenylpropionic acid? administered to cats in doses of 1
gram per kilo was invariably followed by death; the animal
usually dying in from forty to sixty hours. Much acetophenone
is found in the urine,? and it is possible that the toxicity of the
acid is largely due to the formation of this intermediary product
of metabolism. But little, if any, unchanged phenylpropionic
acid can be detected in the urine, oxidation of the acid is therefore
probably fairly complete. Half a gram per kilo given to cats
or dogs is not a fatal dose and but slight poisoning symptoms
are observable. The composition of the urine secreted under
these conditions is qualitatively similar to that obtained when
fatal doses are administered.

(2) Phenylpropionylglycocoll administered to cats in even
larger quantities, 1.5 gram per kilo, than the corresponding

aiWwoep: 431; V, Pp. 303.

2 The doses are all recorded in terms of the free acids although they
were administered in the form of their salts.

3 The practical details of the analysis of the urines will be recorded in
a later paper on the fate of phenylpropionic acid derivatives in the animal
body.

416 Detoxicating Action of Glycocoll

lethal dose of phenylpropionic acid, produces no toxic symptoms
whatever. The substance is excreted almost entirely unchanged
and either no acetophenone or only a minute trace can be de-
tected in the urine. When administered to dogs in dose of 0.5
gram per kilo, more than half the amount is excreted unchanged
but the urine contains notable amounts of phenyl-@-oxypropi-
onic acid, acetophenone and hippuric acid. When given to large
dogs in doses of about 0.5 gram per kilo, oxidation is more nearly
complete.?

Cinnamic acid. Ammonium cinnamate given to cats anddogs
in doses equivalent to from 0.25-0.5 gram per kilo of the free
acid produced no noteworthy effects. No unchanged cinnamic
acid could be detected in the urine, but a small amount of a sub-
stance which appeared tu be cinnamoylglycocoll was excreted.
A little phenyl--oxypropionic acid and acetophenone were also
detected in the urine together with much hippuric acid. Oxida-
tion of the cinnamic acid was probably almost complete and when
smaller doses than the foregoing are given theintermediary prod-
ucts of metabolism are completely oxidized also.

Cinnamoylglycocoll. The sodium salt of this acid was prac-
tically unattacked when given to cats in doses corresponding to
0.25 to1.o gram per kilo. Aboutthree-fourths of substance given
was recovered from the urine. No signs of any toxic action were
observed.

Phenyl-8-oxypropionic acid. This acid was found to be more
resistant to oxidation than either cinnamic acid or phenylpro-
pionic acid. Administered to cats in doses of 0.6 gram per kilo,
it was found that three-fourths could be recovered from the urine
unchanged. Only minute traces of acetophenone and hippuric
acid were found in the urine. When given to dogs in doses of
0.25 gram per kilo only a minute amount of unchanged acid
could be detected. Acetophenone was present in fair amount
and much hippuric acid. Oxidation of the oxy-acid was fairly
complete.

‘In a previous paper in which the fate of phenylpropionylglycocoll was
recorded (this Journal, iv, p. 432) no evidence of an increased resistance
to oxidation, as compared with phenylpropionic acid, was obtained. This
was owing to the employment of insufficiently large doses.

. H. D. Dakin 417

Phenyl-8-oxypropionylglycocoll. This substance was given to
cats and dogs in doses of from 0.4-1.0 gram per kilo. The
substance proved to be practically unattacked as almost the whole
was recovered unchanged from the urine. Traces of cinna-
moylglycocoll were apparently formed.

BACILLUS INFANTILIS (n.s.) AND ITS RELATION TO
INFANTILISM.

(Plates VI and VII.)

By ARTHUR I. KENDALL.
(Fellow of the Rockefeller Institute for Medical Research.)

(From the Laboratory of Dr. C. A. Herter, New York.)
(Received for publication, January 1, 1909.)
Introduction,

The bacilli described in this paper, together with several vari-
ants were isolated from a series of chronic intestinal infections of
obscure origin and etiology. The cases present a definite symp-
tom-complex which has recently been described by Dr. C. A.
Herter! and termed by him, ‘‘Infantilism from chronic intestinal
infection.” Clinically the pronounced symptoms are: an arrest
of body development associated with only a slight retardation of
cerebral development; marked abdominal distension (without,
however, a noticeable enlargement of the spleen); slight to mod-
erate secondary anemia; marked bodily and mental fatigue,
brought on by relatively slight exertion, and disturbances of
intestinal function manifested usually by repeated diarrheal
attacks and impaired powers of absorption.

Chemically, the salient features are: the presence of excessive
amounts of indican and phenolic bodies in the urine; indolacetic
acid occurs in some cases and may be present in greatly excessive
amounts. The aromatic oxyacids may also occur inexcessive
quantities.

In patients receiving considerable fat in the diet, the feces
contain a large excess of neutral fats, fatty acids and soaps, the
excretion of the latter leading to considerable losses of calcium
and magnesium to the body. This loss of calcium and magne-

1 Herter: Injantilism, The Macmillan Company, 1908.
419

420 Bacillus Infantilis

sium furnishes a logical explanation of the retardation of skeletal
development.

Part I: Microscopical Examination of Infantilism Stools.

Representative fields of stools from typical cases of infantilism,
stained by the Gram-Weigert method (Gram-stain followed by
dilute carbol fuchsin, as a counter-stain) are strongly Gram-posi-
tive, and recall similar pictures described by Escherich’ in his
cases of Blaue Bactllose with respect to the great diminution or
even absence of Gram-negative cocco-bacilli of the coli-aérogenes
type, and the dominantly Gram-positive character of the fields.
Certain differences are detectable, however, upon closer scrutiny.
The Gram-positive curved bacilli which are characteristic of
Blaue Bacillose cases are few in number or even absent in infan-
tilism stools, while the most distinctive organisms in the latter
are straight rods measuring from 0.50 to 0.75 microns in diameter,
and from 2.5 to 3.75 micronsinlength. These bacteria are Gram-
positive and eccur singly or in pairs, rarely in short chains of
from four to six elements. They present no distinctive peculiar-
ities of arrangement, although they are frequently collected into
rather sharply defined masses.

It is not uncommon to find signs of degeneration among these
bacilli, particularly if they are derived from constipated stools.”
In such instances the bacilli are more or less irregular in outline
and do not stain uniformly, some portions of the cytoplasm
remaining Gram-positive, while other portions stain faintly with
the counter-stain, or not at all.

Cultural methods demonstrate these bacilli to be B. infanitls,
B. bifidus? or B. acidophilus. One type may predominate or all
may be represented at the same time.

1 Escherich: Jahrb. f. Kinderheilk., lii, pp. 1, et seq., 1900.

2? Herter and Kendall: This Journal, v, p. 289, 1908.

’ Tissier: La flore intestinale du nourrisson, Paris, 1900. ‘Tissier gave
this organism the name B. bifidus communis, a trinomial. This name,
being a trinomial, violates the law of botanical nomenclature and is
incorrect. Inasmuch as the term bifidus expresses tersely the most
prominent characteristic of the bacillus, it should be retained as the
specific name, eliminating the superfluous term, communis.

4Moro: Jahrb. f. Kinderheilk., lii, p. 38, 1900; Wien. klin. Wochenschr.,
no. 5 1900.

Arthur I. Kendall 421

It should be possible, theoretically, to make a morphological diagnosis
between B. bifidus and B. acidophilus on the one hand, and B. injantilts
on the other. Tissier claims that B. bifidus occurs typically with pointed
ends and not infrequently one sees pairs of bacilli lying at an angle with
each other—the so-called geniculate arrangement. B. acidophilus resem-
bles B. bifidus somewhat morphologically, but the bacilli occur in sharply
circumscribed masses, instead of geniculate pairs in stools. B. injantilts,
on the contrary, has definitely rounded ends; clostridial forms are uncom-
mon, and the bacilli do not occur in regular alignment as a rule.

The various manipulations to which stools are subjected prior to
microscopical examination are usually sufficient to disturb the original
orientation of the bacilli, while their morphology varies somewhat with
the consistency of the stool, so that the final criteria of their differentia-
tion and recognition must be based upon cultural methods.

Branched forms are not uncommon in the stools of infantil-
ism patients, and rarely one sees a peculiar modification of the
branched form to which the term “‘antennate”’ has been applied.
The branched forms may be referable to either B. bifidus or B.
infantilis; each organism produces under certain undetermined
conditions branched forms in stools. It is not impossible that
some development has taken place after the stool was passed
in such instances, because many of the bacteria show signs of
degenerative changes which are known to occur when they are
exposed to unfavorable conditions, outside of the body.

The “‘antennate’’ form is unusual in the feces and so far as
the writer is aware, no form similar in appearance has been
described for any known bacillus. The cell body is slightly
spindle-shaped (clostridial) in outline and possesses a deeply
Staining metachromatic granule at one or both ends of the rod.
The cytoplasm between these granules takes the stain feebly or
even not at all, and suggests both by its position and general
appearance a presporogenic body. From one, rarely both, of
the metachromatic granules processes arise, which may be one,
two or three in number; two is the most usual arrangement.
These outgrowths are slightly curved, slender and elongated;
they do not project in the same general direction as that of the
long axis of the cell proper, but rather at an obtuse angle, resem-
bling in a striking manner the antenne of certain beetles. This
resemblance is further accentuated by the irregular manner in
which they take the stain, which gives them a jointed or articu-

422 Bacillus Infantilis

lated appearance. The same forms have been encountered in
old broth cultures of this organism, and they are probably to be
regarded as involution forms. If this supposition be correct,
the antennate forms have no particular significance in the life
history of the bacterium.

Besides B. infantilis, B. bifidus and B. acidophilus one notices
frequently in the feces of infantilism patients large, Gram-posi-
tive cocci which occur typically in pairs, less commonly in short
chains. They measure about one micron in diameter and their
ends are more or less pointed. Their elongated appearance has
led the French writers to refer to them as “‘Flamme de Bougie”’
—an epithet which is very expressive of their peculiar mor-
phology. These organisms are known as the “Enterocoque”’
of Thiercelin,’ or the ‘‘ Mcrococcus ovalis’’ of Escherich,? and,
according to Kruse* are identical with the Streptococcus of
Hirsch-Libmann. Kruse proposed the name, “Streptococcus
lacticus.””

These organisms are the Gram-positive types most frequently
found in the dejecta of infantilism cases, and B. bifidus, B. acido-
philus and Micrococcus ovalis represent the acidophilic bacteria
which are characteristic of well marked clinical states of this
kind. It is noteworthy, as Herter* has pointed out, that the
latter organisms are the dominant ones present in the stools
of breast-fed infants—a fact that is of great importance in
this connection because the bodily development and intestinal
flora seem to be related to each other.

Part II: Methods for Isolating Bacillus Infantils.

The first culture of Bacillus infantis was obtained from a
stool in which they were present in unusual numbers, and no dif-
ficulty was experienced in obtaining fresh strains for nearly a
month. At the end of that time, however, the bacterial flora
changed and it became increasingly difficult to obtain cultures
by the usual method of fishing plates.

1 Compt. rend. de la soc. de biol., April 15 and June 24, 1889.
2 Darmbakterien des Saiiglings, Stuttgart, p. 89, 1886.

8 Centralbl. —. Bact., Xxxiv, Pp. 737, 1903.

ST OGHa Cite

Arthur I. Kendall 423

The colonies produced by B. tnjfantilis are reasonably distinc-
tive when they occur in pure culture, or when there are only a
few other types of organisms in the same plate. As the number
of colonies, either of the same type of organism or different or-
ganisms, increases, they lose their characteristic appearance, and
it becomes increasingly difficult to make even a probable diag-
nosis by simple observation alone.

Again, contrary to generally accepted ideas, it has been found
that not only do colonies of the same organisms vary in appear-
ance in the same plate, but also that radically different types
of bacteria produce colonies which are indistinguishable.

These differences are explainable upon two hypotheses. In the first
place the rigidity of the medium has an extremely important réle. In
low percentage agar, those bacteria which are strongly motile, together
with the forms that produce chains readily, will grow in colonies that are
characterized by considerable lateral expansion. This is particularly
the case in surface colonies. Submerged growths to a lesser extent
obey the same general laws. On the other hand, particularly among the
intestinal bacteria, the metabolic products of one type of organism may
react by diffusion upon adjacent bacteria, resulting in a mutual inhibi-
tion of growth. The strongly acidogenic bacteria, for example, particu-
larly when grown in media which contain fermentable sugars, will seriously
interfere with the growth of alkali-producing types.

A consideration of the facts presented in the preceding pages
indicates clearly the necessity of examining relatively large numbers
of colonies, particularly in the study of unusual cases of intestinal
disease. Obviously this is impracticable in actual work, where
several cases are under observation at the same time. What is
needed is a method whereby one may study the morphology
of the bacteria from a representative majority of all the colonies
which have developed in the various media, and to be able to
return to specified colonies which through this preliminary exam-
ination have been demonstrated to consist of unusual types of
bacteria.

This may be accomplished in the following manner. Plates
showing from 20 to 50 well-separated colonies are marked with
a wax pencil in such a way that the projections of these colonies
upon the bottom of the plate are enclosed in a smallring. This
procedure is kept up until a sufficient number have been selected

424 Bacillus Infantilis

for examination. The colonies are then numbered in sequence,
beginning with 1 and continuing until all are numbered, remem-
bering to place a line under the figure 6 to distinguish it from
the 9.

An ordinary microscopical slide is now ruled into squares, the
size and number depending upon the number of colonies which. it
is proposed to study. A small amount of growth from each
colony is emulsified in a drop of sterile water in its appropriate
square, the films are dried, stained by the Gram-Weigert method
and examined microscopically. This examination will indicate
the general type of organism in each instance, and not infre-
quently demonstrate the fact that certain colonies are composed
of two distinct kinds of bacteria. Having thus not only recog-
nized the morphological peculiarities of the organisms, but also
having a fair idea of the relative purity of each colony, one may
select those organisms which it is desirable to study in detail,
resorting to the colonies from which they were derived and sub-
culturing them at convenience. This procedure has been very
helpful and it is extremely probable that the rather large number
of apparently undescribed forms which have been met with in
this work, have been successfully isolated because attention has
been paid to the study of large numbers of colonies which upon
superficial examination show no macros¢opical differences. By
adopting this procedure, one may be reasonably certain that at
least the dominant forms which will grow in artificial media, will
be detected. In order to make a systematic investigation, how-
ever, it is absolutely necessary to use media containing either
dextrose (lactose is usually better for infants’ stools) or other
easily fermentable carbohydrate for the fermenting and acid
producing organisms, and carbohydrate-free media for the alkali
producers. One may thus eliminate to a considerable extent
the effects of antagonism between the acidogens and the alkali

producers.
Part III: The Biology of Bacillus Infantilts.

This organism is a motile, aérobic, facultative anaérobic, spore-
forming, Gram-positive, non-capsulated bacillus. It produces
acid in dextrose and saccharose, but no gas.

Arthur I. Kendall 425

Morphology. The bacillus measures from 0.50 to 0.75 micron
in diameter, and from 2.5 to 3.5 microns in length.

Upon solid media the bacilli appear as rather short, plump
rods with rounded ends. They occur singly or in pairs, rarely
in short chains. Spores are readily formed in such media, in the
presence of free oxygen. In fluid media, on the other hand,
- spores are not formed as a rule, and in anaérobic cultures spore-
formation has never been observed. This need for oxygen prob-
ably explains the inability of B. infantilis to form spores in the
intestinal tract. Repeated attempts to obtain cultures of this
organism after destruction of the vegetative cells by heating
suspensions of the stool to 80° C. for 10 minutes have failed except
in two instances. It should be stated that it was possible to
isolate these organisms in unheated suspensions made from the
same stool. The two cases in which spores were present do not
necessarily invalidate this view. The stools were 24 hours old
at the time they were examined and it is perfectly evident that
the necessary conditions for spore formation were present.

The spores are central, oval, and cause a slight bulging of
the cell; this enlargement is not great enough, however, to give
the organisms the appearance of clostridia:

The bacteria found in the condensation water of agar, as well
as in fluid media in general, differ conspicuously in appearance
and staining’ reaction from those grown upon solid media. In
the latter instance the bacilli are relatively short and thick; in

1 At first the difference in staining reaction between organisms derived
from the slanted surface of agar, on the one hand, and the condensation
water of the same culture, on the other, led to the suspicion that a
contamination had occurred. Experiments definitely reproducing the
phenomena were made, however, and it was discovered that this same
property is common to a number of Gram-positive bacteria of intestinal
origin. The differences observable are very striking in some instances.
When the bacteria are kept upon artificial media for several successive
transfers the variations tend to disappear and the organisms assume a
uniform appearance whether they are obtained from the slanted surface
or from the condensation water. It is probable that the sudden change
from the intestinal environment to that of ordinary media is responsible
for the lack of stability, and as the bacteria are in a weakened state,
changes in their environment as slight as those existing in slanted surface
and condensation water will produce noticeable responses in bacterial
growth.

426 Bacillus Infantilis

the former they are longer, thinner, and conspicuously Gram-
negative. The staining reaction has been described previously
in this paper, and it will suffice to say that the ‘“‘punctate”’
appearance which one sees in stools can be reproduced exactly
in fluid media.!

Temperature relations. B. infantilis grows most character-
istically and luxuriantly at the body temperature; at lower tem-
peratures the growth is scanty, and at 18° C. it practically ceases.

Anaérobiosis. The organism is an aérobe, although it can de-
velop in hydrogen, carbon dioxide and in the depth of anaérobic
stab cultures. Certain peculiarities concerning its oxygen needs
will be described later on, in the section upon fermentation.

Agar slant. Upon dextrose agar (slanted) there is produced
a spreading, gray, smooth, opaque, shining layer, In the con-
densation water growth is abundant and usually associated with
the production of a pellicle. The medium becomes brown in
older cultures and exhibits a tendency to the production of a well
marked opacity. These two phenomena are distinct and appar-
ently have no relation to each other.

Three variants have been noticed. The first is characterized
by the production of a viscidity due to the ability of tne bacteria
to form adhesive, easily drawn out threads. The second variant
grows poorly upon the slanted surface, and appears as a delicate,
translucent, shining filiform layer which never produces the
browning of the medium referred to above. Organisms of this
type are chemically and morphologically identical with the
luxuriantly growing varieties. They may be regarded as strains
which have not become thoroughly habituated to artificial media.
It should be stated that the latter type grows better in hydrogen
than in aérobic conditions. The third variant is a chromogen.
The pigment, which is soluble in the medium, and occurs in the
upper layers, where oxygen is present, is red-brown, recalling
the pigment produced by an organism erroneously named B.
lactis erythrogenes.

The chromogenic variant may arise spontaneously. In fact

‘ Prolonged cultivation in artificial media causes the organism to lose
its ability to take the Gram-stain; in this respect it agrees with several
well known bacteria of similar origin.

Arthur I. Kendall 427

a chromogenic strain has never been isolated from infantilism
stools. Culturally, morphologically and chemically these strains
are identical with typical infantilis cultures.

Gelatin stab. Growth in gelatin stabs is scanty. In dextrose
gelatin it is more vigorous. At temperatures ranging from 18°
to 22° C. development takes place very slowly, and there is no
appreciable liquefaction. As the temperature rises, however,
the organisms develop more rapidly, and at 24° liquefaction may
take place at the end of six days. The liquefaction is very slow
and frequently is manifested merely by the presence of a slight
infundibuliform, dry depression. Evaporation usually progresses
parallel to peptonization, so that no fluid remains in the cavity.
Rarely a small amount of liquid remains. In cultures which
have been kept at room temperature for several months lique-
faction may progress to such an extent that a funnel-shaped
depression measuring half a centimeter in depth may result.
The medium is not favorable for the development of the bacillus,
however, and certain strains do not liquefy at all, but produce
merelyasofteningof themedium. Thereaction becomesalkaline.

Milk. Freshly isolated strains produce a transient acidity
followed by a return to the neutral point. In older cultures
the reaction becomes alkaline. The primary acidity is doubt-
less due to the fermentation of carbohydrates (hexoses) present
in milk, due to the hydrolysis of lactose. Prolonged cultiva-
tion of the bacteria in milk is associated with more marked
changes in the medium. A gelatinous coagulum is formed which
is not marked and which usually requires a boiling temperature
to demonstrate. This stage is followed by a gradual solution of
the coagulum, an increased alkalinity and a gradual thinning
of the whole medium. The thin, alkaline fluid resembles that
produced by some varieties of the paratyphoid bacillus.

1 Beijerinck (Kon. Akad. von Wetenschappen te Amsterdam, Oct. 27,
1900, and Rodet (De la variabilité dans les microbes, Paris, 1894) have
studied the question of bacterial variation, and define three types, Degener-
ation, Transformation and Variation, in the narrow sense. The produc-
tion de novo of pigment in Bacillus injantilis seems to correspond with
the latter type of variation, which is defined as ‘‘the assumption of a
_ new characteristic (by a single individual of a single strain) which remains
constant. Insufficient or improper media, or the excess of excretory
products in old cultures may be the cause of this phenomenon.”’

428 Bacillus Infantilis

Milk is not, however, a favorable medium in which to cultivate
Bacillus infantilis and the changes produced by this organism
in milk are not distinctive.

Potato. The growth is luxuriant, raised, shining, smooth,
brownish and is attended with a darkening of the potato. The
reaction is not appreciably changed.

Fermentation media. Dextrose and saccharose are fermented
with the production of acid, but no gas. Lactose is only slightly
attacked. The acidity reaches its maximum about the fourth
day, and after that time there is a gradual return to the neutral
point, or even in extreme cases to alkalinity. The explanation
is probably to be correlated with the ability of this organism to
produce ammonia and primary amines. During the first few
days there is a marked growth in the closed arm (in dextrose
and saccharose fermentation tubes) and it is during this period
that the acidity rapidly increases At the end of the time speci-
fied the bacteria grow more abundantly in the open arm of the
tube and this aérobic growth is associated with a large ammonia
production.

Lactose fermentation media are not favorable to anaérobic
growth of B. infantilis and it is a noteworthy fact that in the
closed arm growth is practically wanting. There is, however,
evidence that the organisms are proliferating.! A pellicle makes
its appearance within 30 hours after inoculation, which suggests
that the organism cannot derive its oxygen through the combus-
tion of lactose, although it can thrive in an atmosphere devoid
of oxygen if either dextrose or saccharose is present. Confirma-
tion of this view is indicated by the formation of a similar pellicle
in plain broth. Here again there is no substance from which the
organism can derive its oxygen. Hence it is forced to obtain it
from the air. The wrinkling of the pellicle, and the aimost total
absence of turbidity (the organism is actively motile) point
strongly to the correctness of this explanation.

Plate cultures. Although B. infantilis is an aérobic organism,

1 This inability to develop in lactose is significant when one considers
that this sugar rather than dextrose or saccharose is the important, car-
bohydrate in the dietary of young children. The normal intestinal
bacteria on the other hand are able to grow luxuriantly in media contain-
ing lactose.

Arthur I. Kendall 429

it usually grows beneath the surface in dextrose agar plates.
(Plain agar plates, on the contrary, show relatively more surface
colonies.) The submerged colonies are lenticular, oval or even
round. They are opaque, yellowish or gray (depending upon the
depth below the surface) with delicately ciliate edges. The colo-
nies are surrounded by an opaque area which resembles a halo.
Microscopically the edges are seen to be filamentous and the gen-
eral look of the colony suggests an aggregation of filaments,
presenting a floccose appearance. Klatsch preparations, how-
ever, do not show the presence of long filaments; the individual
bacteria, on the contrary, have a disjointed appearance and even
short chains are uncommon. Surface colonies are translucent
to opaque, and round, with irregular edges. Their appearance
depends largely upon the relative density of the medium and par-
ticularly upon the amount of moisture present. If the latter is
excessive, the colonies spread and resemble delicate films. The
colonies are not characteristic, particularly when there are sev-
eral different kinds of bacteria present in the same plate, and it
is only by examining individual colonies microscopically that one
can safely make the diagnosis.

Biochemistry. The biochemistry will be discussed in detail
in a subsequent paper. A summary of the principal products,
however, is here appended for completeness.

Bacillus infantilis is a powerful alkali-producer. Ammonia
and an unidentified primary amine are the chief basic bodies
found in cultures. The organism does not produce indol, skatol,
phenolic bodies, mercaptan, hydrogen sulphide, alcohol, acetone
oraldehyde. It produces lactic and succinic acids from dextrose,
together with small amounts of volatile acids of unknown composi-
tions. It gives the Voges-Proskauer reaction’ in peptone media.

Certain strains of Bacillus infantilis are, at the time of their
isolation, strongly acidophilic, or, better, have acquired the abil-
ity to withstand unusual amounts of acid. A sharp distinction
should be drawn between bacteria which grow well in moder-
ate amounts of acid, e. g., B. acidophilus, and those organ-
isms which—originally not acidophilic—become by changes

1 Harden and Walpole (Proc. Roy. Soc., 1xxxvii, p. 424, 1906) have
shown that this reaction is due to the presence of methylacetylcarbinol.

430 Bacillus Infantilis

in their environment able to withstand and to develop in great
amounts of acid. Several cultures were obtained from a case
of infantilism in association with Bacillus acidophilus and Micro-
coccus ovalis. All of these organisms were able to develop in
a broth medium containing acetic acid of such a strength that
three cubic centimeters of normal sodium hydroxide were required
to neutralize one hundred cubic centimeters of the medium. The
organisms were obtained in the following manner. A small
amount of feces was emulsified in a tube of the acetic acid broth,
and incubated two days. A loopful of the first broth culture
was introduced into a second tube; after two days a third cul-
ture was prepared from the second. At the end of the last
period of two days, the organisms were plated out and cultures
were obtained in the usual manner. There can be no doubt
that there was a decided development of B. iujfantilis during
the progress of the experiment and considerable numbers of
colonies were obtained on the plates. Subsequent cultural and
chemical studies showed that they were identical with the first
strains of this organism which had been isolated in the usual
manner, in neutral media. The striking fact is that two trans-
fers in ordinary media so changed the acid-resisting ability of the
“‘acid infantilis’’ that it would not develop in media having an
acidity greater than that corresponding to 1.25 cc. normal acid
per 100 cc.

1 This experiment has a distinct value in relation to the question of
bacterial conditions obtaining in the intestinal tract. Innoother patho-
logical conditions are the bacterial flora so complex and dependent in
their ensemble upon obscure changes in environment. These environ-
mental changes may consist of altered intestinal secretions, changes in
diet or of bacterial antagonism and symbiosis or of combinations of these
factors. Apparently slight nutritional alterations are frequently accom-
panied by surprising modifications in the bacterial behavior, the bacterial
response seeming to be out of proportion to the intensity of the stimuli.

Although we are ignorant of the fundamental principles which domi-
nate the bacterial flora of the intestinal tract, evidence is slowly accumu-
lating which points to a symbiotic relationship between the host on the
one hand, and the dominant types of bacteria on the other, at least in

normal, healthy individuals. Undoubtedly diet plays a prominent part’

in determining which bacteria shall increase and which shall be inhibited.
When one stops to consider the extraordinary number of bacteria,
dead and living, that are voided every day in the feces, and compares

Arthur I. Kendall 431

The discovery of the acidophoric strain of B. infantilis brought
up the question of antagonistic and symbiotic relations between
this organism and bacteria with which it is commonly associated
in the intestinal tract. Naturally it is impossible to reproduce
conditions comparable to those existing in the intestines, but
attempts have been made to compare the behavior of pure cul-
tures of a few representative bacteria with combinations of the
same bacteria grown together.

Bacillus colt and Micrococcus ovalis (enterococcus )have been
chosen as the types to compare with B. infantilis, and all observa-
tions have been made in fermentation tubes, where every tran-
sition can be obtained from almost absolute anaérobiosis to
complete saturation with free oxygen.

this figure with the relatively few organisms which are ingested during
the same period, some idea of the enormous proliferation which takes
place in the intestinal tract will be obtained.

Escherich, Tissier and Moro have studied the kinds of bacteria present
in the dejecta of normal nurslings, and find that certain well defined
types of organisms are regularly present to the more or less complete
exclusion of other types. Tissier has noticed that the vast majority of
these bacteria are acidophilic, and he believes that the high degree of
acidity which they can resist while still thriving acts as a deterrent to
the growth of foreign organisms. When aninfant suffers from an intesti-
nal upset, the normal bacterial conditions are disturbed, and new varieties
make their appearance. As conditions return to the normal, there is a
gradual corresponding reappearance of the customary flora, associated
with the disappearance of the abnormal types.

If, however, the abnormal conditions persist, the bacteria associated
with the change may become habituated in the intestine, and eventually
may replace, in part at least, the normal inhabitants. This invasion,
as the writer has noticed in a number of instances, is accompanied by
reciprocal modifications in both the normal and the invading organisms.
This change is manifested by the ability of the two types of organisms to
grow in the presence of each other, even if originally this was impossible.
One may demonstrate this fact by comparing corresponding strains from
normal intestines with the ‘‘ modified’’ varieties. Usually one finds that
the ‘‘normal’’ organisms will not thrive in such cultures.

This fact points to deep seated modifications in the biology of both the
invading bacteria and the normal obligate organisms and undoubtedly
these mutual adaptations explain in part the persistence of unusual types
of bacteria in chronic intestinal diseases.

The facts brought forward indicate the necessity of studying with
great thoroughness those cases of intestinal derangement accompanied
by evidences of abnormal bacterial development.

432 Bacillus Infantilis

Protocols of two experiments are reproduced. The experi-
ments were carried on dur ng four days, both in dextrose and
actose. All of the bacteria used in these experiments were given
preliminary cultivation in dextrose broth to insure a high degree
of reproductive growth.

These experiments emphasize once again the inability of B.
infantilis to ferment lactose. This is indicated not only by the
absence of turbidity in the closed arm of the fermentation tube,
but by the production of a thick pellicle upon the free surface
of the mediuminthe bulb. Therve is a corresponding absence of
inhibitory action of this organism upon B. coli and Micrococcus
ovals, although in dextrose under similar conditions the restra n-
ing activity is marked.

It is obvious, then, that the marked decrease in gas formation
which is observed in cases of infantilism is due to some factor
other than the mere presence of B. injantilis because this reduc-
tion is as marked in lactose as in dextrose. It is extremely
probable that B. bifidus is instrumental in preventing the devel-
opment of bacilli of the colon-aérogenes types, and inasmuch as
the latter organisms are the dominant gas-formers of the intesti-
nal tracts of young children, it is logical to associate the compara-
tive absence of the gas-formers and the simultaneous presence of
B. bifidus with the non-appearance of gas by the fecal bacteria
from well marked infantilism cases.

Unfortunately the réle of B. bifidus in this connection must, for
the present at least, rest upon purely circumstantial evidence.
It is not possible with our present methods to prepare media in
which B. bifidus and B. coli (or other facultative organisms) shall
under the same conditions grow with the same relative intensity.
Obviously, if one organism develops more rapidly than the others
it is out of the question to draw correct conclusions with reference
to their antagonistic properties.

The systematic position of B. infantilis. B. infantilis differs
in essential characteristics from any previously described organ-
ism known to the writer. Particular attention has been paid to
the previously published descriptions of bacteria found in nurs-
lings’ stools, both in health and disease, and the works of Tissier,'

1 Loc cit., p. 263, 1900.

Arthur I. Kendall 433

EXPERIMENT I.

DEXTROSE. } LACTOSE.
gia | ee sa
* HC eA, ace
ulture. : 5 ee) = | «3 -g +
Slog \eel 8 pe pa |say @ =
A|o|s ea | Hci = aed) Ra py
1A a NI a + ec
| “
ipamialis.......... Waar Aa ae a ae ei
ra a +]|—|-—]| alk. | ++
4}—|+]41.2 |+]—|-—|- 0.5) ++
| | | | |
| | } | | |
1 ei is acid — |15|}+ | aeid =
es... He | ee peter SS Sha a ere tea =
aa hee nad gece eed ee ioc co =
4|22| + | +2.25| — a5| + | +1.5 ~
|
1) —| + acid =— |= | + acid _
eae hyraee |tnesets | te , se all Agaca a ke ‘;
8 eae ea etl tee =
| Aah | is | — |") ee) elo
| | |
ie oe) d24 | acid |) = | =e acid -
Metre...) 2)227|+| © jo peepee =
SAD ee | 1 [ee i eae
} 4/12) + | +2.2 | — | 32] + L,6 CN
| | |
}1}—|+| aed |—|—| +] acid =
“ pom ae “ a
Infantilis and Ovalis “alae me |
BO ae | te | “ ee — ley | “ —
cig Sega ea +|+3.0 | —
1| 22/4 | acia | — | 16 + | acid | —
| | «“ = “ =
Coli and Ovalis.... Na a | 28 aN
= a a ian a pe (el -
AV | eB |e | BB Vahl retells
. C
1/12) +] acid eg vn ie | acid | =
Infantilis, coli and 2 | fae) es) Bae Ir _
Sees... || 3 | 11 |.+ PA ON cae Nee ees -
| 4|10| + |} 43.5 | — |2r| + | 43.8) —

|
! |

* Trials with a freshly isolated culture of B. infantilis showed a much greater inhibition
of gas production. The total volume of gas produced in the preliminary experiments in
those tubes containing B. coli and B. infantilis only showed-5 mm. of gas, although B. colt
alone produced 26 mm. in the same media.

a

A34 Bacillus Infantilis

EXPERIMENT II.

DEXTROSE. LACTOSE.
g |g q |g
A | ae
Culture. 2 5 rs a e Be 5 |.3 E
5 =
Als |é eo eS ates
tei) act) (ey cece 1 lesa
ee 7 eet ie . iy | a eee
Infants .ony0% 6 ae
nfantilis Si ee a « rig, awe Ne
4) —| +) 41.3 ap || =) Se
1|26|+ | aeid — | 35) + acid -
Coli 1 2) 29} + : alot Weare i =
Sink Mee wea a net, Sie bales iy Veg lin ” st
4 | 28| +-| +2.45 | — | 34| + | +3.0 _
iL |) =) ae acid —}—-—)}+ acid -
Bho 2 met ges oa" F z
TS cies
Ovalis 3 mi, a “ a 2S 3b “ —
4} —| +] +1.05| — | — | + | 41.50} —
1 2 acid — | 29) + acid -
2 | 23 | + AP Retell | i =
Infantilis and Coli. . 3|22| + a aaa tee « mail
4 | 22) + | +2.20; — | 31] + | +2.65 -
| = |e acid —{|—-|]+ acid ==
2 a a “ Sod oe + “ =
li
Infantilis and Ovalis Be teil « > eae « as
4/—|+)42.7 | —|=—9 4 | oe
( 1) 250) = acid — | 28) + acid -
2| 26) + . ath eee r =
Coliand Ovalis..... 3192| 4 fp ee ey ee « =
4); 22) + | +3.8 — | 30; + | +3.0 _
( 1 iar Bl Ea isc = acid — | 30; + acid =
Infantilis, Coli QTL) 6 hg al lee 3
Dalle. cures Ve SS ch 90a : ar lPccaulinc ‘ a
t 4/10/ + | +8.45 | — | 28| + | +3.9 _

SS

Arthur I. Kendall 435

Escherich,’ Finkelstein,? Moro* and Salge,* have been freely con-
sulted.

An organism described by Salge attracted especial attention.
He found it in cases of catarrh of the small intestine and dis-
covered that it had the property of breaking up sodium oleate into
lower fatty acids, and that the presence of fats increased its
fermentative powers. Inasmuch as an abnormal excretion of
fatty acids and fats is a feature of infantilism, a possible relation-
ship between infantilis and Salge’s bacillus was suspected. Sub-
sequent investigation, however, showed that there is very little
resemblance between the two organisms.

Finkelstein described an acidophilic organism which he ob-
tained from cases of Blaue Bacillose, but it is evident that his
bacillus is closely related to B. acidophilus. (This organism, as
already pointed out, occurs in infantilism stools together with B.
infantilis.)

B. tnfantilis belongs to the B. subtilis group. It produces
resistant spores, forms alkali in milk and non-saccharine media,
and does not produce gas. It is, however, smaller than any
hitherto described subtiloid bacillus, and its ability to liquefy
gelatin is much less marked than is the case with other members
of this group. It should be mentioned in this connection that
an organism having the morphological and cultural characters
of B. infantilis has been isolated by the writer from canned plums.
It differs from B. infantilis, however, by its relatively rapid
peptonizing action in gelatin and its inability to form primary
amines. This organism seems to be a connecting link which
not only indicates the relation of B. infantilis to B. subtilis group
but emphasizes the lines along which the latter organism tends
to deviate from the typical organisms of the group.

PartIV. The Relation of B. Infanttlis to Infantilism.

B. ‘njantilis is a spore-forming organism, and as Theobald
Smith® has pointed out, spore-forming bacteria are not, as a rule,

Loc. cit.

* Deutsch. med. Wochenschr., p. 263, 1900.

BRIEOG NGUb

4 Jahrb. —. Kinderhetlk., lix, p. 399, 1904.

5 Theobald Smith: Some Problems in the Life History of Pathogenic
Microérganisms, Amer. Med., viii, pp. 711-718, 1904.

436 Bacillus Infantilis

obligate parasites, at least in man. The organism has not only
been found in the stools of all the typical cases of infantilism so
far examined! but also in the dejecta of normal nurslings derived
from various sources, although in the latter instances the bacilli
were very few in number and were nct obtained from all the feces
examined.

The fact that these cases, both pathological and normal, rep-
resent a fairly wide geographical area seems to indicate that B.
infantilis may be well distributed in nature. In some instances
it apparently finds a favorable environment in the intestinal
tract of young children, obtains a foothold and proliferates there.
In spite of this proliferation, however, specific agglutinins are
apparently not produced, and although it certainly occurs in
large numbers at certain peviods of the disease, evidence is strongly
in favor of the view that it is non-invasive.

One is not justified, however, upon these grounds in conclud ng
that there is no etiological relationship. Feeding experiments
upon a dog resulted in the establishment of a well-marked diar-
rhoea” associated with the appearance of B. imfantzlis in the stools.
A monkey, whose diet was carefully regulated, reacted even more
strikingly. The movements became soft, pale in color, strongly
Gram-positive, and there was simultaneously a marked increase
of acidogenic bacteria (first of the Micrococcus ovalis type, then
a rapid rise in the proportion of B. acidophilus, associated with
a moderate number of B. bifidus). There was a corresponding
diminution in the Gram-negative coli-aérogenes type of bacilli.
That is, there was bacterially a decided tendency toward the
development of the infantilism type of stool.* These effects are

1 For list of cases, see Herter, loc. cit.

2 It is interesting to note in this connection that Ardoin (Thése de Paris,
p. 78, 1898) and Spielgelberg (Jahrb. f. Kinderheilk., xlix, p. 194, 1895)
have isolated and described bacteria belonging to the subtilis group which
cause decided diarrhceal disturbances, particularly in young children.
Certain cases quoted by these investigators were characterized by the
relative abundance of subtiloid organisms in the stools.

’ The appearance of acidophilic bacteria following so closely upon that
of B. infantilis is perhaps the most noteworthy feature of these feeding
experiments, because cases of infantilism usually run the same course,
bacterially. The explanation of this bacterial sequence is not known
and the data available at the present time do not justify more than the
bare statement of the fact.

Arthur I. Kendall 437

transient and tend to disappear after a few days, but they may
be reproduced by fresh infection with B. infantilis. These
diahrroeal disturbances are conceivably due to the irritant action
of the products produced in situ by B. infantilis, and the experi-
ences of Ardoin'and Spiegelberg’ certainly arein favor of this view.

SUMMARY.

(1) A spore-forming organism, B. infantilis, described above,
has been isolated from each of a series of cases of infantilism.

(2) It has also been found in limited numbers in the feces of
some although not all normal infants.

(3) B. infantilis is not an obligate intestinal bacillus, but a
saprophytic organism which under certain undetermined con-
ditions finds a suitable environment in the intestinal tract and
proliferates there.

(4) It produces no agglutinins and there is no direct evidence
indicative of its etiological relationship to infantilism.

(5) B. infantilism fed to a dog and a monkey caused in each
animal a pronounced softening of the stools and diarrhcea. In
the monkey, this diarrhcea was followed by a decided diminu-
tion in the Gram-negative gas-producing bacilli of the coli-aéro-
genes type, and a noteworthy increase in the Gram-positive
acidophilic flora. B. bifidus in moderate numbers and B. acido-
philus in excessive numbers were the dominant organisms.
There was a gradual return to the normal type of stool, both
macroscopically and microscopically.

(6) These experiments furn sh evidence in favor of the view
that the diarrhoea observed in cases of infantilism may be caused
by irritant metabolic products resulting from the proliferation of
B. injfantilts in the intestinal tract.

In conclusion the writer wishes to express his indebtedness to
Dr. C. A. Herter not only for the opportunity of studying these
cases but also for many helpful suggestions and advice during
the progress of the work. The photographs which accompany
this paper were made by Dr. Leaming of the Rockefeller Insti-
tute, to whom the writer is indebted.

1 Loc. ctt.
2 ioc cit.

438 Bacillus Infantilis

EXPLANATION OF THE PLATES.

1. Pureculture of B. infantilis, showing Gram-positive and “‘punctate’’
forms, the latter staining faintly. a shows an antennate form (X 1000).
2. Submerged colonies of B. infantilis, showing the floccose structure.

3. Atypical field from an infantilism stool. The principal organisms
represented are: B. infantilis, B. acidophilus, B. bifidus, Mic. ovalis, B.
coli. The doubly contoured bodies, resembling large capsulated cocci
are relatively common in typical infantilism stools. Their significance
is unknown (X 1000).

4. A typical dextrose fermentation-tube sediment from a case of
infantilism. I-IV, VI, B. bifidus; II, pseudo-branched form of Mic.
ovalis; III, club-shaped form of B. bifidus; V, ‘antennate form of B. infan-
talis (X 1000).

The diagnosis of these organisms is based upon the cultural auiey of
the stool from which the photograph was prepared.

PraTE VI

Tue JouRNAL oF BIoLo

>
io
fei
n
a
ica]
en)
ry.
AS.
aA
Oo-4
Oo
Ww
n
[o)
Zz
S
4
eo}
S

‘ 4 2 a. a
/\ rs aah
— ie /
*~. eel}
i h - es
; ‘5 * ae /
el te, ae tod
e* . QV
\ Vite fe
i als / “eo” Vw >
ag, OF one
y \r oe of ail
/ a7] 4 por fs 's
* ae ‘
Se ee vA irae
i : Pak a! sF
<> Oop ®, \ are : Vg,
se ans es Roe |
, ' . -
\ ts ‘
i 4 “
’ wT) “eee | ee
a }

-
ud
‘ ‘

!
mF
~ Fa < ‘
pe

}. «Ra

4

4

/ VY,
’ o

s ”

« Times
Spe Os

>
Vv

~ .
-™

oe,

?

me
hee

IG.

Ries

THE JourRNAL oF BIoLocicaL CHEMISTRY,
; VOL. V, NOS. 5 AND 6.

PiaTe VII

ay

|

NOTE ON THE PRODUCTS OF BACILLUS INFANTILIS
GROWN IN ARTIFICIAL MEDIA.

BY

C. A. HERTER
AND
A. I. KENDALL
(Fellow of the Rockefeller Institute.)

(Received for publication, December 12, 1908.)

Bacteriological studies of the feces in certain cases of arrested
development in infancy have shown this disorder to be associated
with the presence of large numbers of the microérganisms to
which we have given the name B. injfantilis. Although the
relation of this microdrganism to the derangement of intestinal
function is not yet clear it is evident that it is so prominent
among the intestinal bacteria in this disease, even in its early
stage, as to deserve careful study from every standpoint. The
morphological and cultural characters of B. znfantilis have been
studied by one of us (Kendall) and we now desire to record
here some observations which have been made on the bio-chem-
ical properties of the bacillus.

The most noteworthy fact relating to the biological chemistry
of B. tnjantilis is its ability to form volatile alkali. When
grown for two weeks in plain broth at the body temperature
this organism was found to have given rise to considerable
quantities of volatile bases, consisting chiefly of ammonia. The
quantity of volatile base formed in three different experiments
carried on under approximately the same conditions was equiva-
lent to 3.8 per cent, 4.2 per cent and 4.1 per cent normal sodium
hydroxide, using alizarin as an indicator. It is interesting to
observe that the formation of volatile alkali by B. infantzlis
was from three to four times greater than the amount made in
plain broth by B. colt growing under essentially the same con-
ditions for the same length of time. While the greater part

1See Infantilism from Chronic Intestinal Infection, by C. A. Herter,
The Macmillan Company, 1908.
439

440 Products of B. Infantilis

of the volatile bases formed by B. infantilis is neutralized by
acids simultaneously formed (lactic and succinic, together with
volatile acids, probably for the most part propionic and butyric)
the bases after a few days predominate sufficiently to impart a
decided alkaline reaction to the broth culture.

We are certainly justified in classing B. imfantilis as one of
the very active producers of ammonia. The differences just
mentioned respecting the ability of B. colt and B. infantilis to
make volatile bases are not attributable to differences in the
growth of the bacteria in the two sets of culture. For while it
is not possible to gauge with accuracy the inequalities in rapidity
of growth, a comparison of the turbidities, as well as of the
microscopical appearances, indicated that we were dealing with
fairly comparable rates ot reproduction, as shown by the con-
ditions at the end of the incubation period of two weeks.

The volatile alkali obtainable from broth cultures of B. znjan-
tilis does not consist entirely of ammonia. The use of Hoff-
mann’s carbylamine reaction showed clearly that a primary
amine is formed early in the course of the decomposition. The
development of an unmistakable carbylamine reaction has been
a feature of all our alkaline distillates obtained from broth
cultures of B. injfantilis. We are disposed to attribute this
reaction to the presence of an alkylamine but it cannot be
denied that the presence of diamines such as putrescine and
cadaverine is not impossible in our distillates. Methylamine
and ethylamine probably do not occur separately among decom-
position mixtures and it is not unlikely that both are present
in our distillates, the former perhaps preponderating. We have
made no observations with a view to determining the amount of
primary amines formed by B. infantzlis.

In order to determine whether B. injantilis causes the putre-
factive decomposition of proteids it was grown in broth for a
period of three weeks or longer. The cultivation was carried
on in flasks containing one liter of the culture medium and
under such conditions as were likely to secure both aérobic and
anaérobic development. No attempt was made, however, to
secure such strict anaérobic conditions as are obtainable under
hydrogen or under carbon dioxide. As there is a strong tendency
for the organisms to collect on the surface of the culture medium

hint

C. A. Herter and A. I. Kendall 441

there may be some difficulty in securing good growth in the
lowest part of the flask where the conditions are relatively anae-
robic, but this difficulty was in a degree overcome by frequently
shaking the receptacles. Under these experimental conditions
we were unable to detect the presence, of indol, skatol, phenol,
aromatic oxyacids, hydrogen sulphide or mercaptans. From
acidified concentrated broth cultures and from milk cultures it
was possible to obtain ethereal extracts containing material
which gave the color reactions for indolacetic acid, but this
derivative of tryptophan was not thus obtainable in amounts
sufficient for identification. Assuming that we are justified
from the color reactions in considering that indolacetic acid

'was formed, it is certain that it was present in only very slight

concentration.

The addition of tryptophan to the broth medium did not
yield indol or skatol from the action of B. ¢mfantzlis nor did this
addition lead to the formation of an increased amount of indol-
acetic acid. Similarly, the addition of tyrosin to the broth did
not lead to the development of phenolic derivatives of tyrosin.
Finally, the addition of cystin to the broth was not followed by
the liberation of hydrogen sulphide or methyl mercaptan. From
these experiments we have reached the conclusion that our
organism does not possess putrefactive activities, at least under
ordinary conditions of growth. But it is proper to say that
different results may conceivably be obtained under strict
anaérobic conditions. It is also possible that under states of
symbiotic action with other bacteria B. infantilis may develop
powers different from those which we have described. It 1s
apparently a characteristic of infantilism from intestinal infec-

_tion that the urine gives very strong reactions for aromatic

oxyacids and we do not consider it impossible that our micro-
organism has a part in their formation, although the indications
are at present opposed to this view.

When boiled with caustic potash or caustic soda the broth
cultures of B. injfantilis yield a strong reddish-brown color which
corresponds to the characters of the Voges-Proskauer reaction.
This reaction is only obtainable from cultures grown on media
containing peptones or albumoses.

On media containing dextrose B. injantilis forms lactic acid,

A442 Products of B. Infantilis

succinic acid and volatile fatty acids. We have not detected
the presence of alcohols, ketones or aldehydes.

It is a question of some interest whether so strong an alkali
producer as B. tinfantilis may, by virtue of its production of
ammonia, give rise to significant irritant action on the intestinal
mucous membrane in those cases where the organism is in process
of becoming parasitic and is present in very large numbers.

Finally it should be observed that the ether extract of old
cultures of B. infantilis in broth, yields an abundance of an
apparently fatty substance which we deem worthy of further
study.

A COMPARISON OF THE METHODS OF REID AND SCHENCK
FOR QUANTITATIVE ESTIMATION OF THE REDUCING
SUBSTANCES IN BLOOD.

By J. J. R. MACLEOD.

(From the Physiological Laboratory, Western Reserve University,
Cleveland, O.)

(Received for publication, November 20, 1908.)

A critical examination of the various methods which, up to
that time, had been suggested for the estimation of the reduc-
ing power of blood was made by Seegen! in 1892. In 1894,
Schenck,? partly in reply to Seegen’s criticisms of a method
suggested by him, published a detailed account of this method
and supplied evidence of its accuracy. Two years later, Way-
mouth Reid? published another method which he showed to
be of great accuracy. The prime difficulty in making an esti-
mation of the reducing substances in blood resides in the removal
of the proteid, and it is in the exact process employed for this
that the methods essentially differ. Since 1896 various other
ways have been suggested for removal of the proteid, among
which may be mentioned that of colloidal precipitation by
means of dialyzed iron (Michaelis and Rona‘ and Oppler and
Rona’) and that of dialysis of the blood against sterile isotonic
saline at ice cold temperature (Edie and Spence’).

In connection with researches carried on during the past two
years in this laboratory, the method of Waymouth Reid has been
that employed, but more recently parallel determinations have
been made by the method of Schenck. The object of making
this comparison was to see whether Schenck’s method is as
reliable as that of Reid for, if so, considerable expense and time

1J. Seegen: Zentralbl. f. Physiol., vi, p. 501, 1892; Vii, p. 604, 1893.
7 Schenck: Arch. f. d. ges. Physiol., lv, p. 191, 1894.

*Waymouth Reid: Journ. of Physiol., xx, p. 316, 1896.

*L. Michaelis and P. Rona: Biochem. Zeitschr., viii, p. 356, 1908.

5 Oppler and Rona: Jbid., xiii, p. 121, 1908.

*Edie and Spence: Biochem. Journ., ii, p. 103, 1907.

443

A44 Reducing Substances in the Blood

could be saved. Another reason for making the comparison
lay in the fact that in a recent research by Liefmann and Stern,}
in which the reducing power of the normal blood of man was
determined by Schenck’s method, much lower values were
obtained than by the older investigators who used one or other
of the older methods. Thus, the average percentage reducing
power for human blood found by Liefmann and Stern is 0.08,
that given by Seegen 0.17, by Frerichs 0.12 to 0.33, by von
Mering 0.1 to 0.15, by von Noorden 0.05 to 0.15 and by Naunyn
0.08 tO 0.09.

In dog’s blood, removed immediately after etherizing, Vosburgh
and Richards? found by the use of Waymouth Reid’s method, the
normal percentage of reducing substance to be 0.17 (16 results).
(in this average, unusually high figures are omitted.) Under-
hill? by the same method, found an average of 0.16 (16 results),
which is about the same as that obtained by me. By the use
of Schenck’s method, Embden, Lithje and Liefmann® found
the normal percentage of reducing substance in dog’s blood
removed without anethesia from the jugular vein to vary con-
siderably with the temperature of the room in which the dog
had been kept. The average percentage at a temperature
between 5° and 10° C. was about 0.098, and for higher tempera-
tures it was less. It must be remembered, firstly, that this is
for venous blood which, as is well known, contains less reducing
substance than does arterial, and, second, that no ether was used.
On account of the discrepancies it was thought advisable to
make a comparison of the amounts of reducing substance found
in the same specimen of blood by the two methods.

A brief description of these two newer methods, as employed
by me, may not be out of place.

Waymouth Reid’s method. The blood is delivered directly from the
artery intoa weighed beaker containing 250 cc. of a solution of 7 per cent

‘Liefmann and Stern: Biochem. Zeitschr., i, p. 291, 1906.

? Vosburgh and Richards: Amer. Journ. of Physiol., ix, p. 38, 1903.

3F. P. Underhill: This Journal, i, p. 113, 1905.

4J.J.R. Macleod: Amer. Journ. of Physiol., xix, p. 388, 1907; XxXii,
P- 373, 1908.

5 Embden, Liithje and Liefmann: Beitr. z. chem. Physiol. u. Pathol.,
X, p. 265, 1907.

J. J. R. Macleod 445

phosphotungstic acid and 2 per cent hydrochloric acid. After weighing,
to ascertain the amount of blood taken, the beaker is placed on a sand
bath and its contents heated to boiling, and the boiling continued. (two
or three minutes) until the proteid precipitate collects into a hard crumbly
mass. After cooling, the clear supernatant fluid is filtered off, nearly
neutralized to litmus (left just acid) and placed in an evaporating dish
on a water bath. The precipitate is transferred to a mortar and ground
into a chocolate-like paste to which is then added, with constant rubbing,
a large amount of cold water, the resulting suspension being then thrown
on to asuction filter and sucked dry. The washings are nearly neutral-
ized to litmus (but left just acid) and added to the evaporating dish
containing the original filtrate. The washed precipitate is transferred
(along with the filter paper) to the mortar and again rubbed up with a
large amount of water and filtered under suction, the filtrate after partial
neutralization being added to the contents of the evaporating dish.
The same process is repeated a third time. The contents of the evaporat-
ing dish are then evaporated to a bulk of about 40 cc. exactly neutralized
towards litmus paper and filtered through ash-free thick filter paper into
a I0o cc. graduate, the evaporating dish being washed on to the filter.
The total volume of the filtrate and washings is made up to 85 cc. which
is then mixed in a beaker with 30 cc. each of the two constituent solutions
of Allihn’s reagent. The beaker, covered with an inverted Petrie’s
capsule, is passed through a retort ring and suspended by its rim into a
briskly boiling water bath for exactly half an hour, after which the
beaker is removed from the water bath and 130 cc. of water are added
to its contents. After standing for a few minutes so as to allow the
cuprous oxide to settle the clear blue supernatant fluid is filtered through
an asbestos mat in a weighed Gooch crucible, the precipitate collected
on the mat, washed with water alcohol and ether, then dried at about
100° C., cooled and weighed.

The relative amounts of Allihn’s solution and sugar solution, the
length of time of heating and all other details being exactly same as
recommended by Pfliiger, the tables published by him! are then employed
in calculating from the weight of cuprous oxide precipitate, the amount
of sugar (reducing substance) present. The 85 ce. of final filtrate will
of course contain variable amounts of dextrose in the different experi-
ments. The smallest amount of blood taken in any experiment is 30
gtams; assuming this blood to contain o.1 per cent of dextrose, the 85
ce. of final filtrate will contain only 0.0 3 gram of dextrose. The largest
amount of blood taken is 50 grams; assuming this to contain as much
as 0.4 per cent of dextrose, the final filtrate will contain 0.2 gram of
dextrose. The variation in percentage of dextrose in this solution is
therefore 0.03~0.20 per cent which is well within the range of Pfitiger’s
table.

*E. F. W. Pfitiger: Arch. j. d. ges. Physiol., xevi, P- I, 1902.

A46 Reducing Substances in the Blood

Schenck’s method. Fifty cubic centimeters of blood are delivered into
250 cc. of a solution containing 0.8 per cent hydrochloric acid and 2 per
cent mercuric chloride. Schenck originally recommended that 50 cc.
of blood should first of all be delivered into 50 cc. of water, then mixed
with 100 cc. of 2 per cent hydrochloric acid and then with 1roocc. of a 5
per cent solution mercuric chloride. It is, however, much more con-
venient to receive the blood directly into the made up reagent, it having
been first of all shown that to do so does not yield results which differ
from those obtained when Schenck’s process is exactly followed. The
mixture is well shaken and then allowed to stand for some time! until
the precipitate has settled after which it is filtered without suction through
a dry folded filter. Washed sulphuretted hydrogen gas is then passed
through the filtrate until all the mercury is precipitated, the mercuric
sulphide filtered off, an aliquot part—z5occ. (corresponding to half the
blood taken)—of the filtrate removed to a flask and air bubbled through
it to remove all traces of sulphuretted hydrogen, then nearly neutralized
but left slightly acid, evaporated over a water bath to a small bulk,
carefully neutralized, filtered and the filtrate made up to a suitable vol-
ume for the estimation of reducing power. Schenck and those who have
strictly followed his directions have employed for this last purpose, the
titration method of Knapp, which, however, according to Sutton is not
veryaccurate. Ina fewofourexperiments, Pavy’s titration was employed
but was not found very trustworthy for this purpose.

The chief difficulty in using Pavy’s method occurs when small traces
of proteid remain in the final filtrate, thus giving, with the alkaline copper
solution. a biuret reaction and masking the end point of the reaction.
In all the other estimations we have adopted the gravimetric method, as
recommended by Pfliiger, and discussed above.

To ascertain the accuracy of their respective methods, both
Reid and Schenck have employed the test of adding a certain
quantity of dextrose to blood and determining the amount of
reducing substance (in terms of dextrose) before and after the
addition of the dextrose. The difference between the two
determinations should equal the amount of dextrose added.
It has been found that both methods give entirely satisfactory
results by this test,a fact which we have confirmed in this labora-
tory. Thus, by Reid’s method, the percentage amount of re-
ducing substance in a specimen of dog’s defibrinated blood was
found to be 0.0905. In another portion of the same blood to
which 0.1005 per cent of dextrose had been added the amount

1Embden has found that prolonged standing of Schenck’s reagent
and blood causes some of the reducing substances to be precipitated
(quoted by Liefmann and Stern, loc. cit.).

J. J. R. Macleod 447

found was 0.191 giving a difference of 0.1002. A similar test
by Schenck’s method (using Pavy titration) gave a less sat-
isfactory result, viz: to a blood containing 0.211 per cent of
dextrose, 1.264 per cent dextrose was added. A sample of this
contained 1.52 per cent dextrose, giving a difference of 1.310
gram. Duplicate determinations of the same sample of blood
by the same method give closely agreeing results when Reid’s
method is employed and the same is true for Schenck’s method
when the gravimetric method is employed for determining the
amount of reduction. The following are duplicates by Reid’s
method, taken at random from our protocols: 0.246 and 0.230;
0.306 and 0.312; 0.139 and 0.136; 0.223 and 0.228; 0.342 and
0.353; 0-259 and 0.260; 0.253 and 0.260; 0.155 and 0.157; 0.144
and 0.152; 0.139 and o.121. The following are duplicates by
Schenck’s method: 0.196 and 0.166; 0.079 and 0.075; 0.117
and 0.123; 0.139 and 0.134; 0.123 and 0.135; 0.137 and 0.133.
According to both these tests of accuracy either method is satis-
factory although that of Reid gives the more constant results,
but when the amount of reducing substance determined by the
one method is compared with that determined by the other
in the same sample of blood a very considerable discrepancy
is found to exist. By Schenck’s method less reducing substance
is found than by Reid’s. Thus by Schenck’s method as above
described and by Reid’s method, the following amounts of
reducing substance in the same sample of blood were found.

TABLE I.
REDUCING SUBSTANCE IN 100 GRAMS OF
= alta Difference ee cent Ga
c Difference.
Reid’s method. Schenck’s method.
gm. gm. gm.
0.255 0.182 0.073 28.6
0.151 0.129 0.022 14
0.266 0.193 0.073 27
(Pavy titration)
0.165 0.135 0.030 18
0.231 0.207 0.024 10
0.1438 0.129 0.014 10

448 Reducing Substances in the Blood

It is seen that the values obtained by Schenck’s method are
always much lower than those obtained by Reid’s, and more-
over, that the difference varies for different bloods. These
results show that either one or other method is inaccurate, or
that the mercuric chloride employed by Schenck precipitates
certain of the reducing substances in blood which phospho-
tungstic acid does not.

Let us see first of all whether any errors of accuracy can be
detected in either method. There are two stages at which
errors are most likely to be made. The first of these is in the
method of treatment of the proteid precipitate and the second,
in the process employed for the estimation of the reducing power
of the final filtrate.

Concerning the estimation of the reducing power of the final
filtrate, we have as before stated, adopted, for both Schenck’s
and Reid’s methods, the method of Allihn-Pfliger. In using this
in the case of Reid’s method, there is a possibility of too high
a value being sometimes obtained on account of the presence
in the final filtrate of a slight opalescence.1 As far as can be
seen, however, this opalescence disappears when the filtrate is
mixed with Allihn’s solution. To ascertain whether this opacity
or some other possible impurity in the.cuprous oxide precipitate
does materially affect the accuracy of the methods, we have
made determinations of the copper in the precipitates by the
method of Volhard with the following results?

‘In Reid’s process during the evaporation to small bulk of the original
filtrate and washings a considerable precipitate of a white or paie blue
color usually separates out. Sometimes when filtering off this precipitate,
some of it passes through the filter paper with the wash water and so
renders the final filtrate opalescent.

? These controls by Volhard’s method were made in most cases where
the results as obtained by the gravimetric process were unexpected, and
the table on the following page is composed of selections from the proto-
cols, good as well as bad results being included.

J. J. R. Macleod 449

TABLE II.
copper ealeu- | Amount of |
SOE Mebar || lated from | Co Ees Soned | Petereice ae ee
weight of Cug0 by Volhard’s grams.* | difference.
ron method.

(a | 0.1046 0.1040 0.0006 0.6
GC) 0.2323 0.2397 | +0.0074 +3.0
OO | 0.0816 | 0.0820 | +0.0004 +0.5
I 0.1764 0.1672 0.0092 5.0
ef 0.1098 0.1090 | 0.0008 ORT
Ch ae | 0.1676 0.1682 | +0.0006 +0.3
Se | 0.0735 0.0732 0.0063 0.5
ee Oe 0.0413 0.0470 | +0.0057 +7.0
° G5)\ 2 | 0.0675 0.0686 | +0.0005 | +0.6
(Cs 0.1285 0.1308 +0.0023 +0.1

penenck (4) -........ | 0.1437 O23 74. 9) 0.0063 4.0

(6) | 0.1600 0.1602 +0.0002 +0.15

Dextrose solution.... 0.2978 0.2960 | 0.0018 0.6

a ae 0.2997 | 0.2964 0.0033 Oat

* In those with + sign the estimations by Volhard’s method were higher than the gravi-
metric. Those with no sign were smaller by Volhard’s than by the gravimetric method.

It will be seen that there is usually very little difference in the
amount of copper found by titration as compared with that
calculated from the weight of the cuprous oxide precipitates.

It is evident, therefore, that the difference in results obtained
by the two methods must arise at some earlier stage in the proc-
ess, viz: in the method of treatment of the proteid precipitate.
In Reid’s method this is thoroughly macerated with cold water
and in Schenck’s method it is assumed to contain the same
percentage of reducing substance as the fluid and is therefore
disregarded, an aliquot portion of the fluid being employed for
the further stages of the process. That by treating the precipi-
tate obtained by Reid’s method, as above described, all traces
of reducing substance extractable by cold water are certainly
removed, we have shown by collecting several of the exhausted
precipitates and again triturating them in a mortar and extract-
ing with cold water. On evaporating the filtrate from this
and heating with Allihn’s solution we have been unable to
detect any reduction. There is no doubt that by Reid’s method

450 Reducing Substances in the Blood

of treating the precipitate all reducing substance is removed
from it.

Regarding Schenck’s method, the only means by which it can
be shown whether or not fluid and precipitate contain the same
percentage of reducing substance are indirect, since, on account of
the gummy nature of the precipitates, it is impossible to extract
them with water. To test this point, Schenck added a known
quantity of dextrose to defibrinated blood and found, by his
method as above described, the same percentage reducing power
as when an equivalent amount of dextrose was added to the
proteid-free filtrate just prior to titration. It, therefore, made
no difference in the results whether the dextrose was added to
the blood before or after precipitation of the proteids. This
seems to be the only result from which Schenck concludes that
the distribution of dextrose is uniform in precipitate and solu-
tion. In two papers, published previously to that in which
the above method is described, Schenck states that when the
proteid of serum or of blood is coagulated by acidifying and
boiling, only a portion of the added dextrose can be accounted
for in the filtrate even when the coagula are most thoroughly
extracted with water... When, however, the thoroughly washed
coagula are boiled with 5 per cent hydrochloric acid, reducing
substance is removed from them, and indeed in almost sufficient
amounts to make good the loss, when the filtrate aloneis em-
ployed.

These observations led us to the next step, viz: to see whether
by boiling the blood with Schenck’s reagent reducing substances
would become dislodged from the precipitate and appear in the
filtrate, thus bringing the results up to the same level as those
obtained by Reid’s method. In doing this, we have as before
compared the percentage reducing power of the same blood
by Reid’s method with that found by using Schenck’s method
combined with heat.?

1 Fr. Schenck: Arch. f. d. ges. Physiol., lv, p. 203, 1894; xlvii, p. 621,
1890.

?In employing heat it was necessary to weigh the solution before
boiling and make up to this weight after boiling. An aliquot portion
of the HgS-free filtrate was then employed for the further stages of the
process.

J. J. R. Macleod 451

The following table gives the results:

TABLE III.

Sa Modified Schenck. | Difference. Per cent difference.
0.191 0.055 28.7
0.214 0.037 UY fine
0.259 0.078 30
0.256 0.063 DAD
0.174 0.054 31
0.114 0.037 32.4

It is evident that boiling does not increase the amount of
reducing substance as found by Schenck’s method. Since
both methods give satisfactory results when tested reyarding
their accuracy by adding a known amount of dextrose to blood,
the deficiency by Schenck’s method must be due to the reagent
precipitating along with the proteids some of the reducing sub-
stances of blood which phosphotungstic acid does not precipi-
tate. This substance cannot be dextrose for the reason just
stated, and must therefore be some other reducing body such
as glycuronic acid or pentoses. Before considering this possi-
bility, it ought to be pointed out that Embden has found pro-
longed contact of the mercuric chloride with blood to cause
some of the dextrose to be precipitated. Liefmann and Stern?
allowed the mixture of blood and reagent to stand six hours
before filtering, but Embden himself in a previousresearch waited
twenty-four hours.’

The presence of glycuronic acid in blood (of the ox) has been
established by Paul Meyer. The same author has found that
the jecorin of dog’s blood, unlike that of horse’s blood, contains
a dextrose group, but that not more than 2.5 to 5 per cent of the
total reducing power of dog’s blood can be accounted for by the
presence of this substance. The amount of dextrose thus com-
bined varies with the diet. The presence of pentose is uncertain.

1Liefmann and Stern: Biochem. Zeitschr., i, p. 297, 1906.
? Quoted by Liefmann and Stern: loc. cit., p. 300.
3 Paul Meyer: Zeztschr. f. physiol. Chem., xxxii, p. 518, 1901.

452 Reducing Substances in the Blood

It has been claimed by Asher and Rosenfeld’ that all the sugar
in blood is in a free state, because they have found, when fresh
unclotted blood containing sodium fluoride is dialyzed against
blood which has been fermented with yeast, and also contains
sodium fluoride, that all the dextrose disappears from the fresh
blood, having dialyzed into the blood containing yeast and been
thus destroyed. The fallacies of this experiment are pointed
out by Pfluger? as well as by Meyer.*

There are therefore, besides free dextrose, possibly four sub-
stances in dog’s blood which have reducing properties, viz: gly-
curonic acid, jecorin, pentose (?) and combined dextrose.

It may be that some or all of these are precipitated by Schenck’s
reagent, but not by that of Reid. This is a question requiring
further investigation.

In conclusion it may be stated as an outcome of this investi-
gation that:

(x1) Schenck’s method gives considerably lower values for
the total percentage reducing power of dog’s blood than does
that of Waymouth Reid.

(2) This deficit is probably due to the mercuric chloride
precipitating some reducing substances which are not precipi-
tated by phosphotungstic acid.

(3) Until it is shown what these substances are, it is unsafe
to employ Schenck’s method, Reid’s method being, therefore,
recommended as the more serviceable and accurate.

1 Asher and Rosenfeld: Biochem. Zettschr., iii, p. 351, 1907.
2 Pfliiger: Arch. f. d. ges. Physiol., cxvii, p. 217, 1907.
3Paul Meyer: Biochem. Zettschr., iv, Pp. 543, 1907.

Yn ee ee

THE RELATIONSHIP BETWEEN THE IONIC POTENTIALS
OF SALTS AND THEIR POWER OF INHIBITING
LIPOLYSIS.

By R. H. NICHOLL.

(From the Laboratory of Biological Chemistry and Pharmacology of the
University of Chicago.)

(Received for publication, November 14, 1908.)

That a relationship exists between the toxic action of any ion
and the ease with which it parts with its electrical charge, has
been shown by Mathews,! who proved that the amount of action
any ion can exert on protoplasm depends primarily upon the
amount of available potential energy it contains. As a measure
of the amount of energy in an ion, or the ease with which it
parted from its charge a property to which was given the name
“ionic potential,’’ he provisionally adopted the solution tension
of the ion, although it was clear that the solution tension involved
not only the ionic potential but the factor of concentration as
well. However, as no method was known at that time of com-
puting the ionic potential, he had arbitrarily to adopt for the
purpose of comparison of different ions, their solution tension
in normal ionic concentrations. Since both negative and posi-
tive ions are toxic and have opposite actions, he provisionally
concluded that the toxicity of any salt must be a function of the
sum of the toxicities of the ions; and that there was a relation-
ship between the sum of the solution tensions of the ions of a
salt in normal solutions and the toxicity of the salt. The sum
of the solution tensions constituted the decomposition tension
of the salt and the relationship was anticipated that the toxicity
of salts ought to be an inverse function of their decomposition
tensions in normal solutions. That is, salts with a high decom-
position tension should be less toxic than those with a low decom-

1 Mathews: Amer. Journ. of Physiol, x, p. 291, 1904.
453

454 Ionic Potential a Factor in Toxicity

position tension. The somewhat arbitrary relationship thus
looked for was found to exist. The toxicities of salts towards
certain fish eggs was studied and a large number of these salts
arranged themselves, with a few exceptions, in the inverse order
of their decomposition tensions. The general result was thus
established and a physical characteristic of salts was discovered
which permitted us to arrange the elements in the order of
their toxicity, independent of their position in the periodic
system and largely so of their atomic weight or valence.

This result was further confirmed by studies of the action of
salts on motor nerves, where again the relationship appeared
that the action of any ion was a function of its ionic potential
or energy. Subsequently, the same factor was proved by Ma-
thews' and by Koch’ to be of great importance in determin-
ing the power of any ion to precipitate a colloid of the opposite
sign and to hold in solution a colloid of the same sign. There
were a few exceptions to the general rule, but these exceptions
were to be anticipated, since ionic potential is but one of the
factors in determining precipitation, valence also being of great
importance. R. S. Lillie working on the toxic and antitoxic
action of salts toward cilia, arrived at the same conclusion of
the importance of ionic potential. McGuigan* found that this
factor was of importance in determining the toxicity of ions
towards diastase. Subseqent to the publication of these papers,
a way was found‘ to compute the ionic potential of the ions
directly and thereafter these values were used instead of the
decomposition tensions of the salt as a measure of toxicity.

The fact, therefore, of a direct relationship between the
toxicity of any ion and its content of available potential energy
(i.e., its valence x its ionic potential) is now well established by
experiment and was anticipated by theory. Such a relation-
ship is practically self-evident.

Recently, however, Pond® has brought forward certain obser-
vations which have led him to dispute the validity of Mathews’

1 Mathews: Amer. Journ. of Phystol., xiv, p. 203, 1905.
? Koch: This Journal, iii, p. 1, 1908.

§’ McGuigan: Jbid., x, p. 444, 1904.

4 Mathews: Journ. of Infect. Diseases, iii, p. 572, 1906.
5 Pond: Amer. Journ. of Physiol., xix, p. 258, 1907.

Se 2, ee

R. H. Nicholl 455

conclusions. Pond attempted to find the weakest concentra-
tions of various salts which would prevent the interaction of
ethyl butyrate and lipase. His results were found to be at
variance with the ionic potential theory in several particulars.
He found that sodium, potassium and lithium were isotoxic, as
were also magnesim, barium and strontium, although there were
considerable differences in the energy content of these different
ions. Zinc was nearly as toxic as lead, although its ionic poten-
tial was far lower and mercury was enormously more toxic than
silver, although the energy content of the mercury ion was a
little lower than that of silver. In general, however, his results
showed that the elements with a low energy content, such as
sodium, potassium and strontium, were little toxic; that those
of medium energy, i.e., cobalt, cadmium and manganese were
more toxic; those of still greater energy, lead and copper, were
still more toxic, and silver and mercury, the most powerful in
energy, were the most toxic. In spite of this general confirma-
tion of Mathews’ view, the exceptions were apparently so numer-
ous that Pond threw doubt on the whole hypothesis, although
he had to admit that ionic potential must be one of the factors
of toxicity.'

1 Pond: Botanical Gazette, xlv, p. 232, 1908.

Footnote BY Pror. MatHews: Mr. Nicholl has kindly allowed me
to append a footnote to his paper. Pond criticises my conclusions
because the computed results deviate so widely in some instances from
those actually found. The computation from the empirical formula gave,
for example, the minimum fatal dose for KCl as about 0.9 N, whereas
0.5 N, was actually found to be the fatal dose. This Pond states means
that the computed is greater than the found (value) by 87 per cent.
He cites ferrous chloride as 531 per cent less poisonous; zinc chloride as
266 per cent more toxic; and manganese chloride as 133 per cent less
toxic than the computed values from their respective solution tensions.
Pond cites these figures in order to throw doubt on my conclusion, for
he says at once thereafter, ‘‘It is merely a matter of judgment as to the
conclusion to be drawn * * * _ but I think it would be perfectly
safe to say that such figures leave solution tension as a determining factor
in toxicity in considerable doubt.’’ The absurdity of this criticism of
Pond’s will be apparent by the following example. Suppose it was
predicted that a man who might be anywhere along a path 100,000 miles
long, was predicted to be o.9 of a mile from one end of the path. He was
actually found to be o.5 of a mile from that end. According to Pond,

456 Ionic Potential a Factor in Toxicity

As Pond had evidently failed to take into account the very
important fact that he was dealing with a heterogenous, not a
homogenous system, it was suggested to me by Professor Ma-
thews that I repeat his work to determtine the concentration
of the salts in the ethyl butyrate at the toxic limits. Pond had
only considered the concentration of the salts in the water, but
as lipase was probably soluble in ethyl butyrate, and ethyl
butyrate does not mix with water, it is evident that unless the
salt gets into the butyrate the lipase will escape from the action
of the salt. The first point was to determine whether lipase
was soluble in ethyl butyrate; the second to determine the dis-
tribution coefficient of the salts between water and ethyl buty-
rate; the third to determine with more accurate methods the
toxic limits of the salts in water; and the fourth to calculate
the concentration of the least toxic dose in the ethyl butyrate.

I. THE SOLUBILITY OF LIPASE IN ETHYL BUTYRATE.

To determine whether lipase is soluble in ethyl butyrate 10
grams of water-free, powdered pancreas, from a commercial
preparation called Holadin, were placed in a tube containing
20 cc. of neutral, water-free, ethyl butyrate and were shaken in
a shaking machine for two hours at 24°C. After filtering, the

this would be an error of prediction of some 80 per cent and the method
by which his position was calculated entirely untrustworthy.

As a matter of fact the prediction was actually 0.000004 out of the
way, or an error in real units of 0.0004 per cent. Any of these salts
might be anywhere in their toxic dilutions between the least toxic and
the most toxic, or between a dilution of approximately normal and that
of silver, or 0.ocooor normal. Potassium chloride might have been down
near silver in its toxicity. The prediction was that it would lie o.1 of a
mile from one end of a street 100,000 miles long; it was found to lie 0.5
of a mile from that end; that is, it was o.4 of one mile of the 100,000
miles out of the way, a real error of only 0.0004 per cent. Manganese
chloride while 133 per cent apparently less toxic than it should be is
actually only 0.09 per cent out of position. Zinc chloride is 0.5 per cent
instead of 266 percent and ferrous chloride nominally 531 per cent less
toxic than it should be is only 0.78 per cent out of its predicted position.
In my opinion this remarkable correspondence between the theoretical
and actually found values is best interpreted in the way [ have indicated.
[A. P. MarHews.]

Vibe? 4 eas 20

R. H. Nicholl 457

excess of ethyl butyrate was evaporated in a drying machine
at a temperature of 35° C. A yellowish residue remained which
possessed marked lipolytic powers. At times this residue ap-
peared to the naked eye as crystalline, but microscopic inspec-
tion failed to reveal any crystals. The residue was fatty in
nature.?

On comparing the lipase solution made from a tenth of a
gram of this residue with a solution made from a tenth of a gram
of the dried pancreas, the latter was always found to be several
(about three) times as strong, irrespective of the length of time
incubated, showing that while the lipase is extracted by the
ethyl butyrate, it is either less soluble than some other con-
stituents, or that it is partially destroyed in the process of extrac-
tion. The aqueous solution of the residue, although markedly
lipolytic failed to give any of the tests for proteins. This
method of extraction may render it possible to eliminate the
proteins from lipolytic solutions, and a further investigation of
the possibility of preparing lipase by this method will be made,

In Table I, the lipolytic activity of a normal lipase solution
made from 0.1 gram Holadin, and one made of the same amount
of the ethyl butyrate residue, as described above, are compared.
Each tube contained 5 cc. of solution as shown in the table.

Table I, Tube (a), shows that lipase is soluble in ethyl buty-
trate. We cannot, however, say whether some fat is necessary
for its solution or not.

Il. THE SOLUBILITY OF SALTS IN ETHYL BUTYRATE.

The solubilities were determined in the following way: Ten
cubic centimeters of the salt solutions of a known strength,
generally about a molecular solution, were placed in a test tube
to which 1o cc. of neutral ethyl butyrate were added. The
tubes were then stoppered and placed in a shaking machine for
two hours. The temperature varied between 20°C. and 21°C.
After two hours’ vigorous shaking, the tubes were withdrawn
and, after settling, 6 cc. of the ethyl butyrate were pipetted

Taylor: This Journal, ii, p. 87, 1906. Taylor states that lipase is
soluble in ether containing fat and that the solubility of the ferment
is directly proportional to the amount of fat in the ether.

458 Ionic Potential a Factor in Toxicity

TABLE I.

Temperature 40° C. Time, 17 hours.

N
Enzymic Ce. of
Water. solution Ethyl NaOH to
(a) or (b). | butyrate. | neutralize
after incu-
bation.

cc. cc. cc. ce.

(a) Lipase extracted by ethyl buty-
rate, 0.01 percent solution....} 2.8 2 0.2 0.66

(b) Lipase aqueous solution, 0.01
per cent solution,fromHoladin| 2.8

Control 4 etme bun whee ae is 4.8 0

oo
Lo)
o

TABLE MII.

Concentration of salts in ethyl butyrate when shaken in contact with M. aqueous

solutions.

HeCl M_ | NELNO =

(mW nasa oedobonuos oooo mod 21.6 AN Bose c ese e eres ecrwees 7080
M z M

CuCNOnaee ates et ete te 556g | LINOs....---.2-00eeeeees ae

ED ONOD gear ty cot oni: seis Tn(NOj)5... sac 56
M M

Ba(NO,), Sisi®ls Wie. ehe\« ais /e)'s sn 6) « 4000 Mg(NO,). or OS 8132
M M

CANO ie eek cee Fa5p | BNOs srs 5i63
M M

NaNO, SOO I einen icin iC aciCnC inci n 4860 Co(NOs,)>. cece eee 11,632
M

SaUNOa asa ceases nausea si9 AgNO ct ons sea 3355

R. H. Nicholl 459

and after evaporating the ester on the steam bath, the salts were
determined gravimetrically as sulphates with the exception of
copper and silver which were determined as oxides. Before
pipetting off the ester, ample time was allowed the water to sink
from the lighter fluid, so the results are not due to admixture of
the water. As the following tables show, the solubilities of the
salts may vary within wide limits. The most remarkable is the
solubility of mercuric chloride, which is almost half its solubility
in water.

Table II gives the solubilities in the ethyl butyrate of the
various nitrates of the metals except mercury which was the
chloride, in terms of molecular solutions determined in the
manner described above. The figures given in Table II are
calculated to be those had the aqueous solutions with which
the ester was shaken been tenth molecular in strength. As
a matter of fact the solutions were stronger than this.

The distribution coefficients, that is, the ratio of the mole-
cular concentration in the water to the molecular concentration
in the ethyl butyrate under the conditions of the experiment,
are given in Table III.

TABLE III.

Concentration in water + Concentration in ethyl butyrate.

i ee ee FIG, MetNO:)o.s-eneu eee 813.2
aS... i... -...-- D368 Ca(NO;).... + scheme neee Gees
1 ee Spd tacit Want Acca 286.0
Le AD -0° |" INELNO:, 0-022, f eee 708.0
0 44a5".0% We LiNOs.), -)1 eee 714.0
fe Gs 2) |) NOs -2)s caren ees 946.0
ae 77010. | SaNO, 0). eee eee 1393.5

Ill. THE CONCENTRATIONS OF THE VARIOUS SALTS NECESSARY
TO INHIBIT LIPOLYSIS.

The salt solution, sufficient water when necessary to make
up to 2.8 cc. and 0.2 cc. of ethyl butyrate were mixed in a test
tube, corked and shaken in the machine for two hours at room
temperature which varied from 21° C. to 24°C. This preliminary
shaking was in order to get the salt dissolved in the butyrate

460 Ionic Potential a Factor in Toxicity

before the addition of the lipase. The tubes were then with-
drawn and 2 cc. of lipase solution added to each. They were
then shaken for two more hours after which they were with-
drawn and titrated with ,“, sodium hydroxide, using phenol-
phthalein as an indicator, except with ammonium salts when
litmus was used. The saponification in a shaking machine at
room temperature is naturally not as great as that occurring at
40° C. but is sufficient when a dilute solution of alkali is used.
By this method, the heterogenous system is rendered as homog-
enous as possible. The surface of the ethyl butyrate is greatly
increased.

The following controls were run: (1) A tube containing the
same amount of boiled lipase as the corresponding normal tube;
(2) a tube containing ethyl butyrate and water, in order to
ascertain whether the ester decomposed spontaneously when
in the shaking machine; and (3) a tube with ethyl butyrate,
water and lipase, but no salt. Two sets of tubes were used for
each concentration of the salt and as the experiments were per-
formed twice the results following are checked by four tubes.

The salts used were the nitrates, except mercury which was
the chloride; all but lead were Kahibaum’s best chemicals; their
concentration was determined by quantitative methods after
being made up. Mallinckrodt’s absolute ethyl butyrate was
used. This we found was a good preparation, water-free, fail-
ing to react with P,O,; it was also neutral. o.2cc. were added
to each tube.

Holadin was employed to furnish the enzymatic solution.
It was made up to a theoretical strength of o.o1 per cent, but in
reality was less than that figure, since but a part dissolved. It
was always filtered. Its reaction was slightly acid and in the
tables following this acidity, and that of the distilled water, were
always deducted. All reagents were measured from pipettes
graduated to hundredths of a centimeter, except the alkali,
which was measured from a burette graduated to fiftieths of a
cubic centimeter.

Column I, Table IV, is self-explanatory; column 2 gives the
cubic centimeters of ,*, NaOH necessary to neutralize 5 cc. of
the salt solution of the concentration of column 1. Column 3
gives the acidity of the control of salt and ethyl butyrate in the

461

R. H. Nicholl

TABLE Iv.

‘p puw ¢ suuUINjoo usem}

-9q 9ouUaIa}I . e 90) ~ S ie.) me Y 1D mt ©
ook jo Sages ie = i) oS Gt Las © Oo 4 S Sa o 3
7 I I 6 : : A : é were 4 :
0} onp AjIplow jo esvesouy = Sd 4 Se ° So os {SiS SSS ° °
"00 G
eZI[BiyneU 0} HOVN aoe 0 i oO (ore) m4 i 2D AO So. 3 AQ
jo‘00 =‘uolyeqnoul 109;8 Te ° ° ae) N tl >) oO o MUN ° ©
eyerdgnq [Aq pues 4[eg fr) —) —) i) i) i) >) y=) (=) Sa i) o
‘asedy] pajfoqun jo AyIproy
‘00 G azI[B1yneU
0} paiMmber FOeN cee
JO ‘00 ‘oInjxIur ouIAz —) o o o Oo i) oo o o o
“ua pe[loq pue aed
“nq [Aqye ‘47es yo AyIploy
*T uuinjoo
UI UOleIyU90N09 Jo U0
“NOS 38S JO “99 g ozt[er} °O ° ° ro) °o °O oo oO ° o
“neu 03 Aressaoou FOUN
ares jo °00 4[¥s jo AyIpoy
| 2
5
Ki 5
: E :
E = :
o by .
ce 4
aR ® :
~ ~
3 b " .
a a | a | = o ., = | a | = a ial | = | ;
Ye) Ye) iY) Ye) Ore Ye) Ve)
‘6 a eee ee lg 3 Qt! ee ea <4 Is 00 4 st :
; =e B xe 3 = 5
8 2 : ee ;
[e} 9 = = st _
S 3 Su os c= a g
A Bie ete z fog z
ol } ef 3 3
4 oO i oO A Oo -4 S)

Ionic Potential a Factor in Toxicity

462

“p puw g suUIN[OO ua0M}
-od eoueIeyIq “HON

_ jo ‘00 “sIsAjodtiy

oyenp AjIplowe jo osvas0U
[D2 ORNS I

"29 G
azipexynou 07 FON ow“

jo‘00 ‘“UOTeqnoUl 109j;8
ajyerAynq [Aye pue 4[Bs |
‘ese dy] pajfoqun yo Ayppoy

0.14

0.14

00 G azi[v1ynou
0} permbar HORN on
jo ‘00 ‘aInyxtul eutAz
-ua po[loq pue oyB14y
-nq [Ayqjo ‘4[8s yu AZIpIy

*[ wuinyoo
Ul UONBIZUGDUOD yO UOT}
-N[OS 48S JO “00 G ozITBI4
-nou 04 Aresso00uU FTOBN
Ore JO -09 4[Bs JO APIPlIOy

Conc. of salt in the water.

0.000

0.000

0.68

Control eerie te ee

0.11

0.00

0.00

0.56 ©

0.000

0.000

Ba(NO,).

0.000

0.000

Controlasensee

0.00

0.000

0.000

.000

0.49

Controls... -

0.38 0.49 0.11

0.3

0.0

0.5

0.5

0.5

0.0

1.2

1.2

1a

0.6

0.6

Control....

ise)
\O
+

R. H. Nicholl

‘Pp pus g suUIN]Oo u90A\4
-aq sousleyIq =“HOPN
oH JO 700 “sIsKody

0} onp AjzIpOB jo osvatoUy

URS |
0.0

09g
ezt[eiyneu 07 Horn’ 2!
jo°00 “uOTyBqnoUl 19448
eyerAynqg [Ajo pue 4yes
‘asedl] pojioqun jo A4yIpoy

0.3
0.35

0.5

0.5

0.38

0.66

2.1

0.66

0.21
0.0

0.91
1.70

‘20 G azt[Biynou
0} peimnber Foun “2%

jo (09 ="eiInyxtul euUIAZ
-ua pelloq pue o4yB1A4
“nq [4y3e ‘9[8s jo AUIpIOYy

0.2
0.35

1.72

4.015

0.7
LY 4

*T uuinjoo
Ul UOI}B1}U99MOD JO UOT,
“NIOS 4[B8S JO "00 g OZT[BI,
-nou 0} Aresse00u FTOUN

O02 Jo ‘00 48s Jo AJIPILY

0.17
0.34

1.6

3.65

0.5
1.68

Conc. of salt in the water.

Control....

MU STIETINMEI Ie eon ow laces

2500
M of
1250

0.0

3.4

3.4

3.3

0.58

0.58

0.37 0.037

1.02

0.00 0.00

1.2

Control...

0.55

_M
1250

Pb(NO,).

1.17

0.65

0.65

Control.....

464 Ionic Potential a Factor in Toxicity

essa | S43 2 52S 2 2yeen
"Rege [2° oq |dtsa | @ ee
a Pay is ab5 8 3 Og
Seu | Bee e | Se 2 ee
epee ($3.2. | 25g 2 Sige
O Ru q 4 e®oo Sts 3 (+) FI
Cone. of salt in the water. ce Uae S58 w2 | see 3 “BE
€353, | 298 ¢ | 305 es
Send | Cotte | Seas jeomme
Sots q zee 5 PF te F 2255
geass | SEER | See288 | SBee
BaSss oR T=] saa Po) St Zs
< < < if
Heel eee eee 0.0 0.0 0.2 0.2
2 180,000
CAA ME ag Ways Cy aaa 0.0 0.0 0.0 0.0
150,000
WARE as) Neti eee 0.0 0.0 0.0 0.0
75,000
Controls 3o ee Eee 0.56 0.56
M
ec Ue pe Wek, ate eS 0.00 0.0 0.36 0.36
AgNO; 3000
M
PEI REL Bh HY 0.0 0.0 0.16 0.16
4000
is IRA OD) UO cm ak | 0.0 0.0 0.0 0.0
2666
Controle Mier wep eae ee 0.6 0.6

same terms as column 2. Column 4 gives the acidity of the lipase
salt mixture after incubation. Column 5 shows the increase
of acidity, when any exists, due to the lipolysis of the ethyl buty-
rate. There was always a control tube of water, ethyl butyrate
and lipase.

Table V summarizes the results obtained. In this table
column 2 gives the minimum concentrations in the water of the
salts mentioned in column 1, which were just capable of inhibit-
ing lipolysis for two hours under the conditions of the experi-
ment; column 3 gives the concentration of the salt in the ethyl
butyrate when it was of the concentration specified in column

R. H. Nicholl

TABLE

Vv.

Inhibiting concentrations of salts in the water and ethyl butyrate.

goa

pe
caus ie
Se8s

=

1. - =
eee ="
rere... Besiine si) =
Men ....... =
1b) re =
0) aa
SO) =
er eee ss
NA Cs
Me. we... ss... =
Dr a5
a =a
ec. va. aut
1 i ae

in ethyl buty-
fe at point
of inhibition

Molecular cone.
rate at

2044
M

2000
M

55,250

M
145,400

nee
221,562

M
296,000

M
385,450

M
324,000

M
3,715,071

465
ae are
SS S24 .
215 Peta
Ou ese)
5° 2aaZ
ae Eten
Ela z #5

N

— 35s on

N

= 9'39 on

N

— 2.92 406

N

885
N
— 2.26 564
9 ay
— 2.49 1022
N
— 2.54 1000
N
N
N
+ 0.179 | 77 0781
N
N
— 0.434 | 199 795
N
+ 1.080 | 162000
N
+ 1.163 | 3775071

466 Ionic Potential a Factor in Toxicity

2inthe water. The figures of column 3 are computed from those
of column 2 by the distribution coefficients of p. 459; column 4
is the ionic potential of the cations of the solution in volts and
column 5 gives the equivalent concentration corresponding to
the molecular concentration of column 3.

As regards column 2, the minimum fatal doses of the salts in
water show marked divergence from those of Pond’s first paper.
Pond measured only the concentration in the water. He found
the minimum fatal dose for sodium, lithium and potassium
nitrates to be the same, i.e., an ¥ solution. A reference to
Table V shows that the limiting concentration is far higher than
this and indeed with potassium even a saturated solution will
not entirely inhibit. The differences between lithium and
sodium are very slight. Pond’s statement of the isotoxicity of
the salts is certainly not correct, and he, himself, has corrected
this statement as regards their inhibition of the action of lipase
upon ethyl acetate in a later paper. His oversight of this point
and of the actual toxic limits is owing probably to the use of
too strong sodium hydroxide solution in titrating which masked
small differences of acidity. The same criticism applies also
to magnesium, strontium and barium nitrates. They are also
not isotoxic; as regards the water concentration strontium is
least toxic, barium most toxic. Different relationships hold,
however, for the concentration in the ethyl butyrate. Various
modifications are made in the fatal doses of the other salts, but
the order of the salts is the same n general as that found by
Pond. Zinc and mercury are more toxic than their ionic poten-
tials would indicate. Zinc, however, requires twice as concen-
trated a solution to inhibit as copper; mercury is extremely
toxic.

From the water concentrations alone, therefore, the general
law is seen to hold;i.e., that those ions of low energy or, in other
words, of low ionic potential, such as sodium, lithium, potassium
and strontium, are little toxic, and those of high ionic potential
are very toxic, such as mercury and silver. The intermediate
ions take in a general way their proper positions, but as Pond
points out, there are many exceptions. Many of these exceptions,
however, clear up if we turn to column 3 and consider the fatal
dose in the ethyl butyrate. By a reference to columns 3 and

R. H. Nicholl 467

4, it will be apparent that the salts follow, with one marked
exception, the same order if one compares the potential of the
cations of the solution and that of the minimum fatal dose.
This result which was anticipated, constitutes a strong confirma-
tion of the truth of Mathews’ hypothesis, that ionic potential is
an important determining cause of toxicity.

The exceptional position of mercury is cleared up. Its
greater solubility in ethyl butyrate really makes its fatal dose
more concentrated than that of silver. The exceptional position
of zinc still remains obscure. It is possible that the zinc lipase
compound, if one exist, is more soluble in ethyl butyrate than
zinc nitrate; further work will have to be done on the zinc to
clear up its exceptional position. Mercury, to be sure, is some-
what less toxic than it ought to be, but its low ionization makes
it not impossible that if we had the ionic concentration its posi-
tion might be more nearly what it ought to be.

TABLE VI.

Comparison of logarithms of minimum fatal dose with the ionic potential of the
cations referred to the potential of sodium as zero.

fous

SAA 55 ie i s of ratio

gett | ae
Na=0 of other salt in

column 1.

WIE. 2 270 oo bE Dc BOO one ee eee 0.0 0.0

as eee ae dice wc) dat nee 0.22 0.16
0.28 |

MEE oc - ia) Aistw wks Sete we oe Site ‘ 1.38 { 0.49

lee -s20 68 SMS GED eee nee 2.45 2.44

Wtic acoo dane OR AROeL EE eee 2.65 2.93

ae oA 5- 2) 2), si scys over dlass wird) eheaaiel > 2.42, 3.10

ES ha oes os nan Oe ad wie wide Srl! 3.28

VOT - oo on Sloe Sete rece ee ee (2.11) 3.41

SAEED a dias ote Bi2 a aud ahaa RS 3.62 3.02

eee I sc) of 5SVax5 5 Aye eens Pie eo wrelag welolavess 3.70 4.89

In Table VI I have made a comparison of the logarithms of
the minimum fatal dose and of the ionic potentials. It will be
observed that a close relationship exists between the difference

468 Ionic Potential a Factor in Toxicity

of ionic potential and the logarithms of the ratios of the minimum
fataldoses. Or expressed in a formula

M.F.D. of salt 5

Hoe:
Fe Oe: a Dae alae

= ionic potential of cation a@ — ionic potential of cation of b.

SUMMARY.

The power of the nitrate salts of various metals to inhibit the
action of lipase on ethyl butyrate is a function chiefly of the
energy content or ionic potential of the cations. This result
confirms the results already obtained by Mathews, McGuigan,
R. S. Lillie, and others, demonstrating that toxicity is a function
of the energy content of ions, and thus proof is added of the
truth of Mathews’ conclusions.

The contrary conclusions drawn by Pond were incorrect,
owing to the fact that he neglected to consider that the ethyl
butyrate-water-lipase system was a heterogenous system and
that he accordingly neglected to consider the concentration of
the salt in the ethyl butyrate. Zinc alone occupies a markedly
abnormal position. This may be due either to the fact that zinc
has a spec-fic affinity for lipase, or ethyl butyrate or else that
the zinc lipase compound is more soluble in ethyl butyrate than
the zinc nitrate or the other metal lipase compounds. Itis sug-
gested that some of the exceptions noted by various observers
in the toxic order of the salts toward various cells may very
possibly be due in part to the fact that only the concentration
of the salts in the water, not that in the protoplasm, has been
considered, whereas the concentration in the protoplasm is the
effectual concentration.

I wish in conclusion to express my thanks to Professor Mathews
for suggesting this work and for his aid during its progress.

Pah -- ~~

ON THE REDUCING COMPONENT OF YEAST NUCLEIC ACID.

By WILLIAM F. BOOS, M.D., Ph.D.

(From the Laboratory of Physiological Chemistry of the Massachusetts
General Hospital.)

(Received for publication, November 5, 1908.)

When yeast nucleic acid is boiled with concentrated hydro-
chloric acid a product is obtained which strongly reduces Fehling’s
solution. Kossel' ascribed the reducing action to glucose and
a pentose, both of which he found in the decomposition fluid.
Pure nucleic acid obtained as the copper compound by the
method I have described? does not yield glucose as one of its
decomposition products; only one reducing substance is present
in the solution, a body giving the orcéin-hydrochloric acid and
the phloroglucin-hydrochloric acid reactions. Impure prepara-
tions of yeast nucleic acid are very apt to contain traces of yeast
dextrin which is converted into glucose by boiling with hydro-
chloric acid.

The isolation of the reducing component of yeast nucleic acid
in a pure state has been variously attempted, but up to the pres-
ent with negative results. Perhaps the material used in former
researches had not been obtained in a sufficiently pure state or
the slight stability of the reducing body made its isolation diffi-
cult. Using the copper compound of yeast nucleic acid as my
material I succeeded in obtaining a small quantity of the reduc-
ing body in form of asyrup. I was also enabled to prepare well-
characterized phenylhydrazine derivatives of the body.

The benzylphenylhydrazone is very stable and may be easily
purified; on this account I chose it formy analyses. The carbon,
hydrogen and nitrogen determinations yielded most unexpected
values which do not in the least suggest a pentose nature for the

1A. Kossel: Zettschr. jf. physiol. Chem., iii, p. 284, 1879.
2 Arch. f. exp. Path. u. Pharmakol, lv, p. 16, 1906.

469

470 Reducing Component of Yeast Nucleic Acid

reducing substance. The chemical identity of the reducing body
remains undecided for the present.

EXPERIMENTAL PART.

The preparation used as the material (Preparation A) gave
the following analytical values:

0.2725 gm. subst. gave 0.3169 gm. CO, and 0.0946 gm. H,O.

022082 ~*~ A es 0.c300 “ N.
052443, 35 eS z 0.0423 “ CuOand 0.0777 gm. Mg,P,0O,.
Calculated for
CzgH52N 14014P 205: Found :
Oe Vegi EAE, ee An eg aan oe mee Near caret 30.00 36.36
1S Dae aRe cee Net, eae LA a acai 4.41 4.93
Tl erate h wae ne tne ote ah “ace es ON, eo Ae Re Le 16.53 16.95
PSOE yeu: Sea nar eeens tae eke ee core me aiectese 23.88 23.86

Twenty-five grams of copper nucleate (Preparation A) are
intimately triturated with a little 1 per cent sulphuric acid,
then more sulphuric acid of the same strength is added gradually
until the volume is1 liter. The mixture is allowed to stand for
several days in a covered beaker on a closed water-bath, care
being taken that the temperature of the latter is not too high.
Under these conditions the copper nucleate, which at first forms
a doughy mass covering the bottom of the beaker, is slowly dis-
solved with the formation of a clear greenish-blue fluid. From
time to time samples of the latter are taken and tested for unde-
composed nucleic acid in the following manner:

Ammoniacal silver nitrate solution is added in excess to
precipitate the bases which have already been set free. The
filtrate is freed from silver with hydrochloric acid and the filtered
solution is repeatedly boiled down to a small volume in a test
tube with an excess of concentrated hydrochloric acid. This
process decomposes any nucleic acid which is still present and
the resulting free purin bases give the characteristic precipitate
with an excess of ammoniacal silver solution.

When all the nucleic acid is decomposed, the copper is pre-
cipitated by hydrogen sulphide and the excess of the latter is
removed by warming the filtered solution on the water-bath.
The precipitate on the filter is thoroughly washed with water

ete

William F. Boos 471

containing a little hydrogen sulphide, the wash waters being
united with the main filtrate on the water-bath. The solution
is then cooled and shaken in a flask with successive small por-
tions of freshly prepared silver oxide until a filtered sample of
the supernatant fluid is found to contain a trace of silver. Care
must be taken to have the solution constantly weakly acid with
sulphuric acid during the precipitation. The voluminous pre-
cipitate of the bases in form of their silver compounds is removed
by filtration and the precipitate on the filter is washed with
distilled water containing a trace of sulphuric acid until the fil-
trate no longer reduces Fehling’s solution. The doughy con-
sistency of the precipitate makes a long continued washing
necessary.

The filtrate from the precipitate of bases is treated with
hydrogen sulphide to remove the excess of silver, filtered and
carefully concentrated to a small volume on a moderately heated
water-bath. Barium oxide in substance is added to the cooled
solution until the reaction is strongly alkaline, to precipitate the
sulphuric acid and the phosphoric acid which was set free as a
decomposition product of nucleic acid. In the alkaline filtrate
an excess of alcohol causes a voluminous precipitate of a sub-
stance which has not as yet been studied, while the reducing
body remains in solution. The alcoholic filtrate is treated with
dilute sulphuric acid in very slight excess and is then carefully
heated on the water-bath to remove the alcohol. The filtrate
freed from alcohol and barium is treated with freshly washed
moist lead oxide to very slightly alkaline reaction, in order to
Ttemove the excess of sulphuric acid, and filtered. The trace
of lead present in the filtrate is precipitated with hydrogen sul-
phide. The solution is carefully warmed on the water-bath to
remove the excess of hydrogen sulphide, and filtered. The result-
ing yellowish filtrate is decolorized by warming with animal
charcoal and the filtrate concentrated to a syrup in vacuo over
sulphuric acid.

In order further to purify the reducing body the syrup is
treated with 75 per cent alcohol, in which the reducing substance
is easily soluble. A gelatinous, non-reducing residue which re-
mains probably contains nucleotin and its decomposition prod-
ucts. The alcoholic solution is evaporated in vacuo over sulphuric

472 Reducing Component of Yeast Nucleic Acid

acid to syrupy consistency and is again treated with 75 per cent
alcohol. This entire process is repeated until the syrup dissolves
without leaving atrace of residue. The alcohol is then removed
once more by evaporation im vacuo.

It was not possible to obtain a colorless product, although the
aqueous solution of the syrup could be easily decolorized by
warming it with animal charcoal; nor could the syrup be made to
crystallize. For these reasons the substance itself was not
analyzed.

The coloriess aqueous solution of the syrup was used to
determine the reducing power and the optical properties of
the new body.

Reduction of Fehling’s Solution.

Freshly prepared Fehling’s solution was standardized with a
solution of pure glucose containing I gram in 200 cc.

1. Fehling’s solution 10cc. = 9.90 cc. glucose solution.
2: “ “ 10 “ = 10.00 “ “ “
3. “ “ 10 “ — 9.95 “ “ “
4. “ “ 10 “ = 9.90 “ ‘ “ “
Taking the average of the four determinations:
Fehling’s solution 10 ec. = 9.94 cc. glucose solution.

The solution of the unknown reducing substance (Preparation
1) contained 1.428 gram in 100 cc.

1. Fehling’s solution 10ce. = 15.75 cc. solution of new body.
2 “ “ 10 “ = 15 . SO “ “ “ “
a: “ “ 10 “ = 15 : 70 “ “ “ “
4. “ “ 10 “ = 15 : 75 “ “ “ “

Average (Preparation 1): Fehling’s solution 10 cc. = 15.75 ce. solution
of new body.

Therefore 0.0225 gram of the new body have the same reduc-
ing power for Fehling’s solution as 0.00494 gram glucose.

A second solution of the reducing substance (Preparation 2),
obtained from a different sample of copper nucleate, contained
0.9600 gram in 100 cc.

William F. Boos © 473

1. Fehling’s solution 10 cc. = 23.35 cc. solution of new body.
Ax “ “ 10 “ — 23 ‘ 30 “ “ “ “
+ “ “ 10 “ = 23 A 35 “ “ “ “
4, “ “ 10 “ — 23 E 40 “ “ “ “
Average (Preparation 2): Fehling’s solution 10 ce. = 23.35 cc. solution

of new body.

Therefore 0.02242 gram of the new body has in this case the
same reducing power for Fehling’s solution as 0.00494 gram
glucose. Average of the two preparations: Reducing power of
0.2246 gram new body = reducing power of 0.0495 gram glucose.

The reducing power of the new substance taken as 1 that of
glucose = 4.5373. In other words, pure glucose has a reduc-
ing power for Fehling’s solution roughly 44 times that of the
new reducing body.

Specific Rotation of the New Body.

The rotation was determined at 28° C. in aqueous solution.

Observed angle = —0.95°. Quantity of subst, in solution per cc. =
0.0096 gram, Length of tube = 2 decimeters.

[a] 7° = — 49.47°
Phenylhydrazine Derivative.

The phenylhydrazine derivative of the reducing body was
obtained from the colorless aqueous solution of the syrup accord-
ing to the usual method. The derivative crystallizes in form
of rosettes composed of fine yellow needles; it is practically insol-
uble in cold water and only very slightly soluble in hot water.
It is easily soluble in hot dilute alcohol, when the solution cools
the compound crystallizes out. After repeated recrystalliza-
tion from hot dilute alcohol the phenylphydrazine derivative
melted sharply at 164° C. The phenylhydrazine derivative is
very unstable; on exposure to the air it is gradually transformed
into a resin.

Benzylphenylhydrazine Derivative.
One and a half gram of the syrup are dissolved in 200 cc. of 70

per cent alcohol and the calculated quantity of benzylphenyl-
hydrazine, 1.2 gram, dissolved in 70 per cent alcohol is added.

474 Reducing Component of Yeast Nucleic Acid

The mixture is heated in a soo cc. flask for about a half hour on
the water-bath and is then allowed to stand. After 24 hours’
standing the mixture is heated on the water-bath and hot water
is added little by little until the turbidity which forms on the
addition of water is no longer dispelled by further heating. The
contents of the flask are allowed to cool. A gradual separation
of fine, slightly yellowish-green needles occurs. After repeated
recrystallization from hot dilute alcohol the compound melts
promptly at 114° C.

The benzylphenylhydrazine derivative is insoluble in hot and
cold water; in cold alcohol it is slightly, in hot alcohol easily
soluble. When an alcoholic solution of the compound is allowed
to stand for days in a stoppered flask a part of the compound
separates out in form of large needles, even if the solution is
very dilute. The benzylphenylhydrazone of the reducing body
is slightly soluble in ether; this is an exception to the rule since
most benzylphenylhydrazones are quite insoluble in ether.

The benzylphenylhydrazine derivative is very stable; on this
account it was examined as to its action on the plane of polar-
ized light and was used for carbon, hydrogen and nitrogen de-
terminations. ,

Specific Rotation of the Benzylphenylhydrazone.

The rotation was determined at 28° C. in alcoholic solution.

Observed angle = — 0.50°. Quantity of subst. in solution per ce.
= 0.004 gram. Length of tube = 2 decimeters.

[a]? = — 62.50°.

The analysis of the benzylphenylhydrazine derivative yielded
the following values:

0.2132 gm. subst. gave 0.6537 gm. CO, and 0.1301 gm. H,O.

0.2316 * ¥ € 0.6778, © GCOsand 0.12367 -9) BLO:
0.1886 “ ¢ MGS ec. Nat 18° and 728mm. = 0.0185 gm.N.
0.1273 “ : ee “« Nat 21° and 724mm. = 0.01235 gm. N.
I. II. Average.
Qe ins 2 RRL ee 83.66 83.50 83.58
AG ete 3k Sainte me 6.83 6.76 6.78
1 OS SA ei ie Aes «Lee 9.78 9.70 9.74

Slt actin i

William F. Boos 475

The formula corresponding to these figures is that of a body
free from oxygen:

Calculated for
ist .

20 2: Found:
Oo aes of teen 6-2 aide ee eee 83 .34 83.58
PAT ois oo ress 3. 3 cavah pa aahe tay ta era 6.94 6.78
I. . 55.00 5 Sa eee ae en ee ee 9.72 9.74

The benzylphenylhydrazone of a pentose has the following
empirical formula: C,,H.,N.,O,.

Calculated for

CisHo3N204: Found :

Ni dN has a he. ore sre Sa. cab ete otoveilene aN 65.26 83.58

RR occ ovo s cash oho Seda rin suse Ales 6.95 6.78

NNN oes estes ls ca. dle-olgaeJabinre Dele eMR or 8.46 9.74
(Vases 9.63 rere erie ie 16.31

99.98 100.10

It is evident from these figures that the reducing body can
hardly be a pentose. If the values for the benzylphenylhy-
drazine radical are subtracted from C,,H,,N, and one atom of

- oxygen is substituted for them, the resulting simplest empirical
formula for the reducing body will be C,H,O.

I have not succeeded thus far in identifying the body but I
hope that the analyses of other derivatives will furnish valuable
information.

COMPARATIVE TESTS OF SPIRO’S AND FOLIN’S METHODS
FOR THE DETERMINATION OF AMMONIA AND UREA.

By PAUL E. HOWE anp P. B. HAWK.

(From the Laboratory of Physiological Chemistry of the Department of
Animal Husbandry of the University of Illinois.)

(Received for publication, December 15, 1908.)

In the course of extended metabolism studies which include
the collection of detailed data regarding the ‘‘nitrogen partition”
the determination of urea, either by the standard method of
Folin or the Morner modification of this method, entails the con-
sumption of a vast amount of time. A method embracing the
accuracy of those just mentioned and at the same time being of
such a nature as to permit of rapid manipulation would be of
great assistance to all those engaged in metabolism work upon a
large scale. Appreciating this fact, at the outset of our metab-
olism studies, we were most willing to devote the time necessary
to make a series of comparative tests between the methods of
Folin for the determination of urea and ammonia and the com-
bination method of Spiro for the determination of these forms
of nitrogen, a description of which method had recently been
published.” Our attention was called to this combination method
by Professor Mendel, who suggested that a series of check tests
might perhaps show the method to be worthy of adoption.

The detailed methods used by us in making this comparison
are as follows:

Folin’s method for the determination of ammonia. The method is based
upon the liberation of ammonia by the addition of Na,CO,, the removal
of the ammonia thus liberated by means of an air current and its collec-
tion in acid of known strength.’

Folin’s method for the determination of urea. Place 5 cc. of urine in a
200 cc. Erlenmeyer flask and add to it 5 cc. of concentrated HCl, 20
grams of crystallized MgCl,, a piece of paraffin the size of a hazelnut and
a few drops of 1 per cent aqueous solution of alizarin red. Insert a Folin

1 Spiro: Beitr. z. chem. Physiol. u. Pathol., ix, p. 481, 1907.
* Zeitschrift fiir phystologische Chemie, xxxvii, p. 161.

477

478 Determination of Urea and Ammonia

safety tube into the neck of the flask and boil the mixture until each
drop of reflow from the safety tube produces a very perceptible bump;
the heat is then reduced somewhat! and continued an hour and a half.
The contents of the flask must not remain alkaline, and “to obviate this
at the first appearance of a reddish tinge in the contents of the flask a
jew drops, and only a few, of the acid distillate are shaken back into the
flask. When the boiling process is completed (14 hour) the contents of
the vessel are transferred to a 750 cc. Kjeldahl flask with about 500 cc.
of water, about 20 cc. of 10 per cent NaOH solution is added, and the
mixture distilled into a known volume of standard HCl until the contents
of the flask are nearly dry or until the distillate fails to give an alkaline
reaction to litmus, showing the absence of ammonia. The time devoted
to this process is ordinarily from 45 minutes to an hour. Boil the distil-
late a few minutes to free it from CO,, then cool and titrate the mixture
with standard ammonia using alizarin red as indicator.

Spiro’s method for the determination of ammonia and urea. Place 25 cc.
of urine in a high, narrow cylinder, such as is used in the determination of
ammonia by the Folin method, and which is properly graduated at 270 cc.
and 400 cc., and add 1} gram of finely pulverized barium hydroxide and
a thin layer of petroleum. The petroleum serves to decrease the frothing
and may be replaced by toluol or alcohol provided the cylinder is suf-
ficiently high and the air current is strong enough.

Close the cylinder with a two-hole rubber stopper and through one
of the openings pass a current of ammonia-jree air to the bottom of the
cylinder. Through the other opening pass a glass tube which is pro-
vided at the top with a safety-tube containing glass wool and glass
beads. This tube is then connected with a suction pump and the liber-
ated ammonia is caught in an acid of known strength as already described
in connection with the Folin method: In this connection use is made
of the Folin absorption tube! in order to secure complete absorption of
the ammonia by the acid even when an extremely rapid air current is
employed.

After all the ammonia has been removed from the urine under exami-
nation the glass tube, etc., are washed out with 95 per cent alcohol, the
solution diluted to the 270 cc. mark with alcohol, then to the 400 ce.
mark with ether, the cylinder tightly stoppered, shaken thoroughly and
allowed to stand over night. The nitrogen in the whole or an aliquot
portion of the solution may now be determined. Spiro advises the making
of duplicate determinations of the nitrogen content of roo cc. portions of
the solution, i.e., the portion equivalent to 6.25 cc. of urine. This pro-
cedure of making duplicate determinations as indicated we have found
to be entirely satisfactory. The solution is simply transferred to a
Kjeldahl flask, acidified with sulphuric acid, and, after the evaporation
of the alcohol-ether mixture, the nitrogen determined by the* Kjeldahl
method.

' Hawk’s Practical Physiological Chemistry, p. 381.

Paul E. Howe and P. B. Hawk 479

It will be noted that the combination method of Spiro embraces
essentially the Folin technique for the determination of ammonia
coupled with the Moérner-Sjoqvist principle for the estimation of
urea. The determination of both of these forms of nitrogen in
the same specimen of urine is obviously a most desirable feature
provided accuracy in neither form of determination is sacrificed.
Another factor in favor of the Spiro combination method is that
less urine suffices for the determinations involved than when they
are made upon individual specimens. This feature is of con-
siderable importance particularly in connection with experi-
ments upon fasting animals in which the daily urine flow, at
certain stages in the experiment, may be exceedingly small.
But the argument in favor of a combination method of this sort,
which, in importance to the metabolism worker, overshadows
all others, is the vast amount of time saved in the course of a
long series of experiments.

In our comparative study of the methods of Folin and of Spiro,
as just described, we made a careful examination of four differ-
ent types of solutions as follows: (1) Urea solution; (2) Ammo-
nium chloride-urea solution, which was prepared by adding dilute
hydrochloric acid to a given volume of an ammonia solution
of known strength until the solution was slightly acid and then
adding a weighed amount of re-crystallized urea; (3) Normal
urine; (4) Normal urine + two volumes of water. In Table I
Pp. 480 will be found the analytical data from the analyses of the
urea solution. An examination of the results here tabulated
indicates that the methods of both Spiro and Folin yielded very
close to the theoretical percentage, the average value obtained
by the Spiro procedure being 100.02 per cent whereas the method
of Folin yielded 99.17 per cent as an average. The data from
the analyses of the ammonium chloride-urea solutions are given,
in Table II, p.481. As was the case with the solution of pure urea
the analyses of these solutions indicate that there is practically
no choice to be made between the methods of Folin and of Spiro
as regards the determination of ammonia and urea in solutions
containing a mixture of pure ammonium chloride and pure urea
alone. The results are, in both cases, sufficiently near the theo-
retical value to be conclusive. In the ammonia determinations
an average of 99.38 per cent was recovered by means of the Spiro

480 Determination of Urea and Ammonia

technique, as against 99.46 per cent by means of the Folin tech-
nique; whereas the Spiro method gave an average value of 99.98
per cent for the urea and the Folin method a value of 99.46 per
cent.

TABLE I.

Check determinations on urea solution.

UREA KECOVERED.
Number of | Grams of urea in the = =
deeermine- volume of ee solution | Grams. | Per cent.
10n. used. | s5 i ze .
Spiro. Folin. | Spiro. Folin.
1 J| Spiro = 0.1991 | 9 yog2 | 0.1961 99.55 | 99.04
|| Folin = 0.1980
2 J| Spiro = 0.1991 | “9 i999 | 0.1967 | 99.95 | 99.34
\| Folin = 0.1980
Spiro = 0.1991
3 J| Sp 0.1998 | 0.1965 | 100.35 99.24
\| Folin = 0.1980
4 ae = Paes 0.2011 | 0.1950 | 101.00 | 98.48
8) => y
= ‘ Ee sg Re 0.1981 | 0.1973 99.50 99.65
oun = b
6 Sey = ee 0.1981 | 0.1973 | 99.50 | 99.65
1 aS 7
{| Spiro = 0.1991
7 (pee 0.1997 | 0.1956 | 100.30 98.79
Average { a = Rep 0.1991 | 0.1963 | 100.02 | 99.17
Oo => rn

Judging alone by the data included in Tables I and II as just
mentioned, one would seemingly be justified in substituting the
Spiro combination method for the Folin methods for the deter-
mination of ammonia and urea in metabolism work. However,
the accurate determination of urea, in particular, in the presence
of other nitrogenous substances, such for example, as creatinine,
hippuric acid, uric acid, etc., is an entirely different proposition

481

Paul E. Howe and P. B. Hawk

9F' 66 ge°ag [secon
L0°O0T | $9° 66 Thr10'0 | SEFTO'O
00°O0T | T€°66 PFIO'O &F10°0
1G OOT | 00°OOT | &FFIO'O PrIO'O
00°O00T | $9° 66 FIO’ 0 | SEFIO'O
P8°86 | IL°66 | 22c600°0 | 8L2600°0
60°66 | 18°86 | 692600°0 | 02600°0
0°66 | €0°O0T | 692600°0 | F9E600°0
8PF'86 | 8h°86 | 612600'0 | 612600°0
“uNoy _ ‘o11dg “UlOWy F “o1idg
*4ue0 10g *SsuUIBID
*paisA000z

*VINOWAY

FFIO'O

FFL0'O

FFIO'O

bPFIO'O

T9€600'0

19€600°0

19£600°0

T9€600°0

“UOT}N[OS
yooyo oy
JO ‘00 68°FS
UI SUIBID)

‘II pue [ suomnjog 10; o#v10ae puswry

TI Boros
10j adBIDAY

J WorNos
10} OSBIOAY

=
Le

“uly
-njos Yoey JO “ON,

JEP JO “ONT
*u0ol}

“eu Tua

9F'66 | 86°66
{| aio = a
76°46 | 89°66 | S8ITT'O | EFFIT'O)| Gry-9 ag
are =r
00° 001 SFIT'0 | gsit-o = g
J| @It'o = a
11°26 | 19°86 | G6OIT'O | ZITO \/ geit-9 = g
ZrIl'0 = A
41°86 | POOL | SZIT'O | ESIT'O 1] oit-g = g
_ {| ¢8ht'0 = a
SOOT | L2°O0T | Z6FT'O | L6FT'O \| eger-g = g
{| $800 = a
60°ZOT | €1°66 | 9IST'O | O8FT°O 1] ogst-g = g
\ cstl'0 = a
SF O0I FL°00T | Z6FT'O | FOST'O | sest-g = g
_ {| Ssh 0 = a
98°86 | F6°00T | 89FT'O | L08T°0 | ggst-9 = g

‘uyog | ‘ondg “uyoy ‘ordg

*posn
“4u90 Iq “SsuUrBl ‘) ets Ha esa
a ites

“poetaA00ay
‘vauo
*uwoUnjos

DALN—APLLOJYI-UNLUOWULD UO SUOYDUULIAIAP Yay?

"il DIAvVL : —

482 Determination of Urea and Ammonia

from the determination of this substance when unaccompanied
by the other nitrogen-containing substances above mentioned.
Appreciating this well recognized fact, after obtaining the data
already indicated as to the comparative accuracy of these two
methods when the analysis of solutions containing only urea or
urea and ammonium chloride was involved, we at once set about
to obtain a comparison of the methods from the standpoint of
direct urinary analyses. In this connection we used normal
urines, making our analyses upcn the undiluted specimens as
well as upon the normal urine after the addition of two volumes

TABLE III.

Determinations on urine.

URINE.

Ammonia. Urea.
No. of = . J Sea z
Urine. A :
Grams in 48-hour urine Grams in 48-hour urine
sample. sample.
Spiro. Folin. Spiro. Folin.
1 1.399 ¥5392:- - | -33 602 30.405
2 1.423 1.394 : 32.547 30.622
3 2.088 2.021 38.103. 34.041
4 1.351 1.335 37.142 33.910
5 1.346 1.329 37.376 33.191
6 1.800 1.785 44 .847 38 .430
Z 1.851 1.781 44 .839 40.584
8 2.115 2.085 44.879 41.235

of water. The data from this series of tests showed the forty-
eight hour output as calculated upon the basis of the results
from the Spiro technique to be slightly higher in every instance
than the analogous value obtained by means of the Folin method.
The differences between the forty-eight hour output of ammonia
as determined by the two methods is, however, very slight in
most instances, in fact, for the most part the value obtained in
a given case by one method is an acceptable duplicate for the
value obtained by the other method. When we come to consider
the data for the urea content of the urine specimens, however,

Paul E. Howe and P. B. Hawk 483

the comparative uniformity in the results obtained by the two
methods was no longer apparent. As in the case of ammonia,
the values for the urea were in all cases higher when determined
by the Spiro method than when secured as a result of the Folin
technique. The excess of urea as determined by the Spiro pro-
cedure was, however, not trifling as was the case with ammonia.
The conditions were practically the same with the diluted as
with the undiluted urines. There being no uniformity in the
amount of urea indicated by the Spiro method as present in a
given urine sample above that shown to be present by the Folin
method, it is not possible to apply a correction which will yield
results even approximately correct.

Considering it of sufficient interest and importance to compen-
sate for the time spent in its investigation we concluded to make
an attempt to learn the origin of this excess of nitrogen which
was determined in the method of Spiro and calculated as urea.
In the study of this problem, several check solutions of creatinine,
hippuric acid and uric acid in individual solutions and in com-
bination were analyzed. In nocase was any of the nitrogen from
these nitrogenous urinary constituents determined by the Folin
procedure whereas all the nitrogen present as creatinine and hip-
puric acid as well as a portion of that present as uric acid was
determined by the Spiro technique and consequently calculated
asurea. The facts just stated account for the apparent excess of
urea as determined by the Spiro method. Inasmuch as the crea-
tinine, hippuric acid, and uric acid content of urine samples must
vary according to the individual conditions it is apparent that
it is not feasible to apply any correction to the value for the urea
content as obtained by the Spiro procedure, which shall in any
way assist in the obtaining of anything like an approximately
accurate idea of the urea present in such urine.

CONCLUSIONS.

(1) Both the Spiro and Folin methods for the determination
of urea and ammonia yield practically theoretical results when
applied to the examination of solutions of urea, as well as to
the examination of mixtures of urea and ammonium chloride.

(2) The values for ammonia as determined by the Spiro

484 Determination of Urea and Ammonia

method are always slightly higher than those determined by the
Folin procedure. In most instances, however, the two values
are essentially duplicates.

(3) The values for urea as determined by the Spiro method
are always decidedly higher than the values determined by means
of Folin’s method. The excess may amount to 6—20 per cent
of the total urea output.

(4) All of the nitrogenof the creatinineand the hippuric acid
as well as a portion of the nitrogen of the uric acid present in
any urine sample is determined by the Spiro method as urea
nitrogen. None of these forms of nitrogen is determined by the
Folin procedure. These facts account for the high urea values
obtained by the Spiro method as compared with the values
obtained by the Folin metiiod.

(5) No correction can be applied to the urea values as ob-
tained by the Spiro procedure which will permit of the obtaining
of even approximately accurate data as to the actual urea con-
tent of a given urine.

(6) Spiro’s combination method for the determination of
ammonia and urea cannot be used to advantage in metabolism
work inasmuch as the urea values are inaccurate and do not per-
mit of the application of a correcting factor.

a

A REAGENT FOR THE DETECTION OF REDUCING SUGARS.
By STANLEY R. BENEDICT.
(From the Sheffield Laboratory of Phystological Chemistry, Yale University.)

(Received for publication, December 18, 1908.)

It has already been shown that the hydroxides of the aikali
metals have a greater destructive action upon dextrose and
various other carbohydrates than have the carbonates,’ and in
accordance with this fact, a copper-containing solution in which
the alkalinity is secured by sodium carbonate makes a more deli-
cate and specific test for the detection of dextrose than does a
copper solution which contains sodium hydroxide. A reagent
of this nature, containing copper sulphate, Rochelle salt, and
sodium carbonate, was suggested in a previous paper.” This
reagent affords a delicate test solution for dextrose, but it has
the disadvantage common to so many of the alkaline copper
solutions, viz: that after mixing, it rapidly deteriorates and soon
becomes useless for detecting small quantities of sugar. For this
reason it seemed desirable to obtain a solution in which the alka-
linity is secured by carbonate, and which shall at the same time
be permanent after mixing.

Rochelle salt is the constituent of the alkaline copper solutions
which undergoes change upon standing, and forms products which
cause a spontaneous reduction of the solution. As a substitute
for the tartrate we may (theoretically) use any aliphatic com-
pound which carries two or more OH radicals, and which is in
itself incapable of reducing the copper at a boiling temperature.
Substances in great number and variety have been proposed in
the literature as substitutes for the Rochelle salt in Fehling’s
solution. In the present case, where the alkalinity is to be secured
by carbonate, it was found that none of the substances usually
employed could be used with advantage. Thus glycerol and
mannite almost always carry sufficient reducing substance as

1 Benedict: This Journal iii, p. 101, 1907.
? Benedict: loc. cit.
485

486 Detection of Reducing Sugars

impurity to affect the carbonate reagent, even where there is
not enough present to reduce Fehling’s fluid.

Citric acid (in the form of its salts) should, from the theoretical
point of view, be capable of holding cupric hydroxide in solution
in an alkaline medium. Upon practical test it hasbeen found to
be a most satisfactory substance for this purpose. The follow-
ing formula yields a satisfactory reagent:

Copper Sulphate (pure crystallized)............... 17.3 grams.
Sodium Citrated 2. darctyem niet allel aa ee ee 173.0 3
Sodium Carbonate (anhydrous).................-- 100.0 <
Distilled: Water! cee. casera hee ic Se ee ue oe ae to 1000.0 ec.

With the aid of heat dissolve the sodium citrate and carbonate
in about 600 cc. of water. Pour (through a folded filter if nec-
essary) into a graduate and make up to 850 cc. Dissolve the
copper sulphate in about 100 cc. of water and make up to 150
cc. Pour the carbonate-citrate solution into a large beaker or
casserole and add the copper sulphate solution slowly, with con-
stant stirring. The mixture is ready for use.

This reagent is more sensitive to dextrose either in pure solu-
tion or in urine than is Fehling’s fluid, is not reduced by uric
acid (or appreciably by chloroform, chloral, or formaldehyde),
and appears to suffer no deterioration on standing. The solu-
tion is not caustic and may be kept in cork or glass stoppered
bottles. Samples of this solution prepared somewhat over a
year ago appear to be in as good condition, in every respect,
as when freshly prepared. These were kept in partially filled,
uncolored glass bottles, exposed to light, heat, etc. A recent
examination of these samples showed that not only had they
undergone no spontaneous reduction, but that no sign of reduc-
tion or other alteration occurred upon heating for twenty-four
hours in a bath of boiling water. (The heating was not con-
tinued longer.) Fehling’s solution, freshly prepared and mixed,
subjected to this treatment showed a marked precipitation of
cuprous oxide after three hours heating, and this precipitate
increased continuously during the subsequent heating.

1 The ordinary sodium citrate of the drug trade appears to be sufficiently
pure for use in this reagent. An examination of several samples pur-
chased in the open market failed to reveal the presence of objectionable
substances.

t ¥ 2) vey

Stanley R. Benedict 487

The following points may be mentioned in connection with the
use of this reagent. No strongly dehydrating substance (such
as potassium hydroxide) is present; hence upon reduction this
solution is more apt to yield the hydrated oxides than is Fehling’s
solution. Thus the reduction product is frequently yellow or
green, rather than red, as in Fehling’s test. The reagent is not
dark colored, like the hydroxide-containing solutions, and even
the slightest precipitates may readily be observed without wait-
ing for them to settle. For general work the solution is used
just as is Fehling’s fluid, save that it is desirable to continue the
boiling for from one to two minutes, and then let the tube cool
spontaneously.

The following is the procedure for the detection of dextrose
in the urine. To about 5 cc. of the reagent in a test tube are
added 8 (not more) drops of the urine to be examined. The
fluid is then heated to boiling, kept at this temperature for from
one to two minutes, and allowed to cool spontaneously. In the
presence of dextrose the entire body of the solution will be filled with
a precipitate, which may be red, yellow,or green. If theamount
of dextrose is small, the precipitate forms only upon cooling. If
no dextrose is present the solution either remains absolutely
clear, or a very faint turbidity, due to precipitated urates, may
be apparent. Even small quantities of dextrose in urine (0.1
per cent) yield precipitates of surprising bulk with this reagent,
and the positive reaction consists in the filling of the entire body
of the solution with a precipitate so that the solution becomes
opaque. Since bulk, rather than color, of the precipitate is
made the basis of the reaction, this test may be applied, even
for the detection of small quantities of dextrose, as readily in
artificial,as in day light. Urines containing 0.08 per cent dex-
trose give a very positive reaction with this test. Fehling’s
solution requires the presence of about 0.12 per cent of dextrose
in urine to yield an equally positive test.

It is hoped to present a later paper in which certain other appli-
cations of this reagent will be discussed, including its employ-
ment in quantitative processes.

“

NOTE ON THE OCCURRENCE OF SKATOL AND INDOL IN
THE WOOD OF CELTIS RETICULOSA (MIQUEL).

By CHRISTIAN A. HERTER.
(Received for publication, November 13, 1908.)

The occurrence of putrefaction products derived from the
breakdown of tryptophan has now become one of the common-
places of physiological chemistry, but their appearance in the
course of the normal metabolism of the higher plants has been
as yet so rarely noted as to invest their formation by plants with
the interest of novelty. In 1889 Dunstan’ noted the occurrence
of skatol in the wood of a large tropical tree growing in Java
and known as Celtis reticulosa.2, The intense odor of this wood
suggested the presence of a-naphthylamine. As this substance
had never been observed as a plant constituent, Dunstan deter-
mined to investigate the question of its presence. Working with
less than 200 grams of the wood from Celtis reticulosa he was
unable to obtain naphthylamine. He succeeded, however, in
crystallizing from water a substance possessing an intolerable
fecal odor. The scaly crystals were soluble in ether, alcohol and
benzene. Dunstan succeeded in determining the presence of
skatol, which he obtained as a picrate and analyzed for its nitro-
gen. The crystals melted at 93.5°, the correct melting point
for skatol. No attempt was made to determine the quantity
of skatol in the wood and Dunstan remarks, with what seems to

1 Dunstan, Wyndham R.: ‘‘On the Occurrence of Skatol in the Vege-
table Kingdom,”’ Proc. Roy. Soc., London, xlvi, p. 211, 1889.

2The name, Celtis reticulosa, was given by Miquel to a tree growing
in Java. Planchon subsequently identified the species with Celtis
cinnamomea, of Lindley (in De Candolle’s Prodromus) which occurs
throughout Eastern India and Ceylon. Thwaites regards the Ceylon form
as a distinct species under the name of Celtis dysodoxylon, but botanists
generally group it with the Java species in Celtis cinnamomea. Thwaites
says: ‘‘The freshly cut timber of the tree possesses a powerful and very
disgusting odor.” Mr. Dyer states that he has not met with other instances
of this singular property. The Celtis belongs to the Urticacec.

489

490 Skatol and Indol in Celtis Reticulosa

us almost exaggerated conservatism, that it was considerably
less than 1 per cent. It was observed by Dunstan that the
. fecal odor is most marked when the substance is present in minute
quantities. He states that when larger quantities of the sub-
stance are smelled the odor is penetrating and aromatic rather
than fecalin quantity. Dunstan furthermore noted the peculiar-
ity that skatol accumulates at a late period in the growth of the
tree and comments on the absence of indol. He suggests that
the superior stability of skatol may possibly be the explanation
of its survival in the plant. He remarks that it would be of
interest to determine whether indol is present in Celtis reticulosa
at an earlier stage of its growth as, for example, at the period
when the skatol first makes its appearance.

The exact source of the skatol in Celtis reticulosa was not
determined by Dunstan. He, however, calls attention to two
possibilities; first, that the proteid constituents of the wood may
yield it, and secondly, that it may be formed from some inter-
mediate substance, as an amino-acid. He considers that the
synthetical production of skatol from nitrocumic acid furnishes
us with a clue to one possible mode of formation.

In 1898 Greshoff! stated that he found a tree in Pulu Wei
in Batavia corresponding to the Celtis reticulosa of Koorders.
Koorders is said to have made the interesting observation that
there is an especially strong development of skatol in trees whose
branches have been wounded some time previously. On the
other hand Greshoff states that fresh Celtis reticulosa wood from
the Buitenzorg Gardens is odorless.

On learning of Dunstan’s publication I felt sufficiently inter-
ested in the occurrence in vegetable tissues of heterocyclic
aromatic derivatives of putrefaction to seek and obtain from
Dr. Dunstan an introduction to Professor Treub, Director of the
Department of Agriculture in Buitenzorg, Java. In response
to my request Dr. Treub was so kind as to forward a considerable
quantity of the wood, bark, branches, twigs, leaves and roots of
Celtis reticulosa Miquel. In view of recent developments in
regard to the origin of skatol from tryptophan, it seemed of
special interest to determine if possible whether the skatol

* Greshoff: Mededeelingen uit’s Lands Plantetuin, xxv, p. 175, 1898.

es

Christian A. Herter AQI

present in Celtis reticulosa could be regarded as derived from
tryptophan through the action of some special enzyme. A
possiblity which particularly interested me was that of finding.
indolacetic acid in the wood of Celtis reticulosa, since the occur-
rence of this substance could almost certainly be regarded as
evidence of the origin of skatol from tryptophan in this wood.

I had no difficulty whatever in confirming the chief statements
of Dunstan with regard to the presence of skatol in the wood of
the Celtis reticulosa. The odor of the wood sent me was very
intense and disagreeable, like that of an illy cared for privy. It
tichly deserves the names under which it is known in Java,
namely kaju tai, which is translated as Dreckholz or filth-wood,
by W. G. Boorzma.' I made no effort to determine the percent-
age of skatol in the specimen sent me by Dr. Treub, butit can
safely be said that it was present in little more than traces. I
may safely say that judging from the intensity of the color reac-
tions with paradimethyl-amido-benzaldehyde the skatol was
present in less than o.o1 per cent. It wasof interest to observe
that skatol was present only in the wood of the trunk of the
specimens sent me and that it was obtainable neither from the
bark nor from the wood of the branches. I was also unable to
obtain it from the roots, from the twigs or from the leaves. A
feature which further interested me was the fact that the skatol
was unevenly distributed in the wood, some portions yielding
more than others. No definite relation could be made out
between any peculiarities of structure in the wood and the con-
centration of skatol present. It is, however, perhaps worthy
of note that the specimens of wood sent to me were irregularly
channeled by a soft woody material suggesting the action of
some insect of microérganism. But no evidence of the presence
of insects could be obtained and the material itself was sterile
as regards microédrganisms. Nevertheless this channeling of the
wood is probably to be regarded as a pathological condition and
should be remembered in connection with the statement of
Koorders, already quoted, that the strongest development of
skatol occurs in connection with wounded branches.

1 Boorsma, W. G.: ‘‘Ueber Aloéholz und andere Riechhdlzer,’’ Bull.

du Département del Agriculture aux Indes Néerlamdaises, no. vii (Pharma-
cologie, iii).

392 Skatol and Indol in Celtis Reticulosa

In reply to an inquiry addressed to Dr. Treub he says: ‘‘I
beg to inform you that only the old wood of Celtis reticulosa has
the odor of skatol. As the tree has never been specially grown
in the Gardens here I cannot say at what period the odor begins
to be detected.”

In addition to detecting skatol in the wood of Celtis reticulosa
I was able to obtain unmistakable traces of indol. I did not
attempt to isolate indol, but the distillates after the removal of
skatol gave reactions with dimethyl-amido-benzaldehyde, with
G-naphthaquinone sodium monosulphonate and with nitrous
acid pointing, I believe, unmistakably toindol. These reactions
for indol were only obtained from material from which skatol
was also obtainable. As above mentioned, Dunstan was unable
to obtain evidence of the presence of indol.

The attempt to detect the presence of indolacetic acid in the
wood of Celtis reticulosa proved unsuccessful. There was indeed
obtainable from 1000 grams of the wood pulp, from which skatol
and indol had been distilled by means of a very thorough steam
distillation, an extract giving a reaction which suggested in-
dolacetic acid. But it was impossible to thoroughly free the
extract from an interfering yellow coloring matter present in
the wood. It is possible that the failure to obtain more satis-
factory reactions was due to the association of this coloring
matter. It may be said, therefore, that while no positive evi-
dence of indolacetic acid could be obtained, I feel unable to
absolutely exclude its presence in traces. In spite of being
unable to demonstrate the presence of indolacetic acid in the
wood of Celtis reticulosa I am disposed to think its origin from
tryptophan in this instance more likely than its synthesis from
simpler substances. Further studies based on very large quanti-
ties of the material are probably necessary to settle the question
of the presence of indolacetic acid.

ON THE SYNTHESIS OF PARANUCLEIN THROUGH THE
AGENCY OF PEPSIN AND THE CHEMICAL MECHAN-
ICS OF THE HYDROLYSIS AND SYNTHESIS OF
PROTEINS THROUGH THE AGENCY OF
ENZYMES.

By T. BRAILSFORD ROBERTSON.

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

(Received for publication, December 15, 1908.)

EXPERIMENTAL.

I. Introduction.

It is evident that in all true instances of catalysis the catalysor
must accelerate the reaction in both directions. The definition
of a catalysor’ implies that no work is done by the catalysor and
that hence the point of equilibrium between the substrate and its
products cannot be shifted through its agency;? now the point
of equilibrium is reached when the velocity of the forward is
equal to that of the reverse reaction, hence the velocity-constants
of each of the reactions must be multiplied, through the agency
of the catalysor, by the same factor. In other words, since the
point of equilibrium is the same whether the catalysor be present
or absent (since none of the catalysor is used up in the reaction,
and, consequently, no work is performed on the substrate by the

1Cf: Nernst: Theoretical Chemistry, English trans. of the 4th German
ed., London, p. 566, 1904.

2;W. Ostwald: Lehrbuch, 2, ii, pp. 248, 262; 1896-1902; Ueber Katalyse,
Liepzig, 1902; Zettschr. f. Elektrochem., vii, p. 995, 1901; Die Schule der
Chemie, Leipzig, i, p. 88, 1903.

3 For literature containing experimental confirmation of this deduction
in the field of inorganic catalysors, see Mellor, Chemical Statics and Dyna-
mics, London, p. 250, 1904. Taylor has found that the equilibrium con-
stant in the hydrolysis of esters by lipase is not altered in value by the
presence of the enzyme, cf. ‘‘On Fermentation,’ Univ. of Calif. Publ.
Pathol., i, p. 269, 1907.

493

494 Synthesis of Paranuclein

catalysor), and since the point of equilibrium depends upon the
ratio of the velocity-constants of the opposing reactions, it fol-
lows that this ratio cannot be altered through the agency of the
catalysor; hence if the velocity-constant of one of the reactions
is multiplied through the agency of the catalysor that of the
reverse reaction must be multiplied by the same factor. This
deduction has been experimentally verified in a great many cases.
In many instances where a catalysor has been known to accelerate
One reaction it has also been found to accelerate the reverse
reaction. The reversion of the action of yeast upon maltose,’
of Kephir-lactose upon lactose,? of diatase upon glycogen,’ of
emulsin upon amygdalin,* and of lipase upon esters® afford
examples, in the field of fermentation, of the experimental veri-
fication of this deduction. It should be observed, however, that
the mere observation that both the forward and reverse reaction
are capable of being accelerated by an enzyme is not sufficient
to demonstrate that the enzyme acts as a true catalysor unless it
be also shown that the station of equilibrium is not shifted by the
presence of the catalysor.

In recent papers Taylor has described the synthesis of a protein
(protamin) through the action of trypsin and I have described
the synthesis of a protein (paranuciein) through the action of
pepsin. In Taylor’s experiments the concentrated products of
the tryptic digestion of 400 grams of protamin were converted
into their carbonates and subjected to the action of a consider-
able quantity of the trypsin (obtained from the liver of the soft-
shelled California clam); at the end of five months about 2 grams

1 Croft Hill: Journ. Chem. Soc., \xxiii, p. 643, 1898; Ber.d. deutsch. chem.
Gesellsch., xxxiv, p. 1380, 1901.

? Fischer and Armstrong: Ber. d. deutsch. chem. Gessellsch., xxxv, p.3144,
1902.

3 Cremer: Ibid., xxxii, p. 2062, 1899.

4Emmerling: [bid., xxxiv, 3810, Igor.

5 Berninzone: Atti del soc. ligi. di scion. nat. e. geograph., Genoa, Xi,
P. 327, 1900; Kastle and Loevenhart: Amer. Chem. Journ. xxiv, p. 491,
1900; Hanriot: Compt. rend. de l’ Acad. d. Sci., cxxxii, p. 212, 1901. A. E.
Taylor: Univ. of Calif. Publ. Pathol., i, p. 33, 1904.

6A. E. Taylor: Univ. of Calif. Publ. Pathol., i, p. 343, 1907; this
Journal, iii, p. 87, 1907. T. Brailsford Robertson: Univ. of Calif. Publ.
Physiol., iii, p. 59, 1907; this Journal, iii, p. 95, 1907.

ee

T. Brailsford Robertson 495

of protamin (weighed as sulphate) were recovered from the solu-
tion. In my experiments 400 cc. of ;4, potassium hydroxide
“‘saturated’’ with casein,’ were, after complete digestion, con-
centrated to 70 cc. This concentrated solution of the products
of the peptic digestion of casein is a clear yellow-brown syrup
which gives no precipitate or opalescence upon the addition of
acetic acid either before or after neutralization by the addition
of potassium hydroxide. To 70 cc. of this solution were added
30 cc. of a Io per cent solution of Gribler’s pepsin; within two
hours a precipitate had formed which was shown to be one of
the constituents (‘‘Paranuclein A’’) of the mixture of substances
to which the collective name “‘paranuclein’”’ has been applied.

The reaction of the hydrolysis of a protein is, to all intents and
purposes, a monomolecular reaction,? but the reverse reaction,
the condensation, must be at least bimolecular, almost certainly
of a much higher order. Hence the velocity of hydrolysis varies
as the first power of the concentration of the substrate while the
velocity of the condensation must vary as the second or higher
power of the products. Hence, by concentration of the solution
the velocity of condensation will be increased at a much greater
degree than that of hydrolysis and the station of equilibrium
will be shifted in such a sense that the ratio of substrate to prod-
ucts at equilibrium will increase; so that although hydrolysis
may be practically complete in dilute solution yet, on concentrat-
ing the solution of the products, the point of equilibrium may
conceivably be shifted so far as to lead, in the presence of a cataly-
sor, to the re-formation of a considerable quantity of the original
substrate. Hence it was natural to suppose, as Taylor did, that
these syntheses, accomplished through the action of proteolytic
enzymes upon concentrated solutions of the products of protein
hydrolysis, are examples of the reversion of a catalysed reaction
in which the catalysor plays no part in determining the final
equilibrium, and that the hypothesis of van’t Hoff, that the
natural syntheses of proteins, fats and carbohydrates in the liv-

1 Cf: T. Brailsford Robertson: this Journal, ti, p. 337, 1907.

2Cf: A. E. Taylor, ‘‘On Fermentation,” Univ. of Calif. Publ. Pathol.,
i, pp. 220, etc., 1907. Hans Euler: Zeitschr. f. physiol. Chem., xlv, p. 420,
1905. Svante Arrhenius: Immunochemistry, New York, chapter 3, 1907.
T. Brailsford Robertson: loc. cit.

496 Synthesis of Paranuclein

ing organism are examples of the reversion of the catalytic action
of enzymes, obtained in these syntheses, to some extent at least,
its experimental verification for the case of proteins.

There are several groups of facts, however, which speak against
the correctness of this conclusion, and among these we may par-
ticularly draw attention to the following:

II. The Influence of the Initial Substrate-concentration upon the
Velocity-constant of Hydrolysis.

In the initial stages of protein hydrolysis the products of the
reaction, as can be experimentally demonstrated, exert but little
depressing effect upon the velocity of its progress, since the
point of equilibrium for the reaction is, under the usual experi-
mental conditions of ferment and substrate concentration, so
close to that of complete hydrolysis. Hence, were the reaction

of hydrolysis monomolecular, the equation, log. 20 = kt, where
a—x

a is the mass of the substrate, x the amount hydrolysed at time
t and k is the velocity-constant of the reaction, should hold good.
As a matter of fact the majority of the investigators who have
carried out quantitative experiments upon protein hydrolysis
under conditions of adequate control are agreed that, for a given
initial substrate-concentration the above equation does hold
good.’ If, however, we use varying initial substrate-concentra-
tions we find that, for a constant ferment-concentration, k is not
identical throughout the series; in other words k, in the above
formula, is not only a linear function of the ferment-concentra-
tion, as might be expected, but is also a function of the substrate-
concentration a. Thus Weis has found that in the peptic diges-
tion of the protein of wheat k diminished as a increases.” Similar
results have been obtained by Taylor.2 Taylor has suggested in
explanation of this phenomenon that ‘“‘there may be different pro-
portions of combination between ferment and substrate, different
valencies, so to speak, and thatwhen the two aremixed thereaction

1 Cf. footnote 2 to p. 495.

? Quoted after Euler: Zeitschr. f. physiol. Chem., xlv, p. 420, 1905.

§A.E.Taylor: ‘(On Fermentation,”’ Univ. of Calif. Publ., Pathol., i, p. 239,
1907.

T. Brailsford Robertson 497

proceeds according to the particular complex adjusted in that par-
ticular system.’”’ It is to be observed, however, that it is not the
proportion between substrate and ferment concentrations which
determines the velocity of the reaction since, for a given concen-
tration of substrate, the velocity of hydrolysis is directly pro-
portional to the concentration of the ferment, it would appear
as if the absolute concentration of the substrate determines, in
some manner, the degree of activity of the ferment.

III. The Influence of the Concentration of the Ferment upon the
Velocity of Synthesis.

We have seen that the velocity of protein hydrolysis is directly
proportional to the concentration of the ferment.' If, therefore,
the ferment acts as a simple catalysor and does not in any way
affect the equilibrium in the system, then the velocity of synthesis
must also vary directly as the concentration of the ferment, for
otherwise, the ratio between the two velocity-constants would
be a function of the ferment-concentration and, since the station
of equilibrium is determined by this ratio, the equilibrium of the
system would also be a function of the ferment-concentration.
The synthesis of ‘‘Paranuclein A” through the agency of pepsin
affords us an opportunity of determining the influence of the
ferment-concentration upon the velocity of protein-synthesis,
since the product is rapidly formed and can readily be determined
quantitatively; the following was the experimental procedure.

The products of the complete peptic hydrolysis of .*, potassium
hydrate, “‘saturated’”’ with casein? were evaporated to one-sixth

1Cf. Sjéqvist: Skand. Arch. f. Physiol., v, p. 277, 1895. Sawjalow:
Zeitschr. f. physiol. Chem., xlvi, p. 307, 1905. V. Henri and Larguier
des Bancels: Compt. rend. de l’ Acad. des Sct., cxxxvi, pp. 1099 and 1581,
1903. A. E. Taylor: Univ. of Calif. Publ. Pathol., i, p. 7, 1904; p. 242,
1907. Vernon: Journ. of Physiol., xxx, p. 334, 1903. Bayliss: Arch. des
sct. biol., St. Petersburg, xi, Suppl., p. 261, 1904, reprinted in the Collected
Papers of the Physiological Laboratory, University College, London, vol.
xiii. Euler: Arkiv. fur Kemi, ii, p. 31, 1907. Madsen and Walbum:
quoted after Arrhenius, Immunochemistry, New York, p. 86, 1907. T.
Brailsford Robertson: this Journal, ii, p. 346, 1907.

2 Wherever, in this paper, the ‘‘products of the complete peptic hydrol-

ysis of X, alkali, ‘‘saturated’”’ with ‘“‘casein’’ are referred to they were

498 Synthesis of Paranuclein

of their volume and filtered. Seventy-five cc. of the clear deep
yellow filtrate were placed in each of six flasks and to each, respec-
tively, 0, 5, 10, 15, 20 and 25 cc. of 10 per cent pepsin (puriss.
sicc. Grtibler) were added and the total volume on each flask was
made up to 100 cc. by the addition of distilled water. After the
addition of toluol, the tightly-stoppered flasks were set aside
at 36° for 22 hours. At the end of this time the flasks containing
25 and 20 cc. of 10 per cent pepsin, respectively, contained heavy
precipitates, while that containing 15 cc. contained a precipitate
and those containing 5 and 1o cc. had undergone no change
beyond a slight increase in opalescence. The contents of the
flasks were now filtered through S. & S. No. 589, ‘“‘black band”’
papers and the precipitates were washed with distilled water
until colorless filtrates were obtained; in all cases the filtrates
gave no precipitate or increase in opalescence upon the addition
of acetic acid. The filters were then washed with ro cc. of +
potassium hydrate and the filtrate was collected in water con-
taining 20 cc. of 54 acetic acid. The filter papers were then
thoroughly macerated in dilute potassium hydrate and the mag-
ma thus prepared was filtered and washed with distilled water,
the filtrate and washings being collected in the beaker which
received the first washings with 5 potassium hydrate. Care
was taken to prevent more than a slight excess of acetic acid from
being finally present in this beaker. The precipitate which
formed in the beaker settled rapidly in small flocculi, was col-
lected upon anS. &S. No. 589, ‘‘black band”’ paper and thoroughly

prepared as follows: To 6 liters of X sodium or potassium caseinate
were added 2 grams of Griibler’s pepsin puriss. sicc. which had previously
been dissolved in a little water; this solution, after thoroughly mixing,
was allowed to stand at 36° for 10 days, 2 more grams of pepsin being
added after the first 4 days (in the presence of toluol) and was then
sterilized by steam at 100° and filtered through hardened filter-paper.
To the filtrate were then added 2 more grams of pepsin, dissolved, as
in the previous cases, in a little water, toluol introduced, and the solution
was again allowed to stand at 36° for 7 to 8 days; it was then again sterilized
by steam at roo® and filtered through hardened filter-paper. The fil-
trate thus obtained is of a clear yellow color with little or no opalescence
and gives no trace of a precipitate or opalescence upon the addition of
acetic acid either before or after neutralization with alkali; hence both
casein and paranucleins are completely absent from the solution.

T. Brailsford Robertson 499

washed with distilled water until the washings were neutral to
litmus; in all cases the filtrates and washings were perfectly clear
and free from protein. The papers and precipitates thus ob-
tained were macerated in water containing a known quantity
(ro cc.) of + potassium hydrate, the magma thus obtained was
diluted to Bbadt 200 cc., phenolphthalein (4 drops of 2 per cent
alcoholic solution) was added and the solutions thus prepared
were titrated to neutrality with 5, hydrochloric acid. A weighed
amount (227 milligrams) of apetariuclerd A” prepared in the
manner described in my previous paper! was dissolved in about
roo cc. of distilled water containing exactly ro cc. of 5 Tp Poesssiaas
hydrate and the solution was titrated to Hederatey with +,
hydrochloric acid and phenolphthalein: indicator; in this way aE
was found that 1 gram of ‘‘Paranuclein A” ca 4:8 Ce.
of *, potassium hydroxide? hence 1 cc. of * alkali = 0.208
eae of “‘Paranuclein A,” and we can estimate from the deter-
minations described above the amount of ‘‘Paranuclein A”’ in
each of the solutions containing varying amounts of pepsin. The
following were the results obtained:

TABLE I.
Amount at pepsin in 100 ce. Milligrams of ‘‘Paranuclein A’’ produced
of solution. at end of 22 hours.

25 cc. of Io per cent. 296
20 = & 210
15 3 € 162
Io e “4 °
5 “ a“ °
oO “ “ °

‘T. Brailsford Robertson: this Journal, iii, p. 95, 1907.

71 have found that upon standing in the presence of excess of alkali
the amount of alkali neutralized by a given quantity of ‘‘Paranuclein A”
increases slightly. I have been unable to determine whether this phenome-
non is due to the slow dissolving of microscopic suspended particles or
whether it is due to hydrolysis (cf. T. Brailsford Robertson and C. L. A.
Schmidt: this Journal, v, p. 31, 1908) the number of cubic centimeters
neutralized by 1 gram given above is the lower figure obtained directly
after complete solution judged by the disappearance of obvious particles
within the solution. I have obtained figures as high as 5.2 after stand-
ing for about an hour in a warm temperature. The titrations in the
experiment described were performed immediately after the complete
disappearance of obvious particles.

500 Synthesis of Paranuclein —

Hence we see that the velocity of reversion by no means stands
in direct proportionality to the concentration of the ferment;
while the velocity of synthesis in the most concentrated solu-
tions (25-15 cc. of 10 per cent pepsin in 100 cc.) roughly approx-
imates to direct proportionality to the concentration of ferment,
in the more dilute solutions (10-5 cc. of 10 per cent pepsin in 100
cc.) the velocity of synthesis falls off with extraordinary rapid-
ity as the concentration of the ferment diminishes. Making
every possible allowance for experimental error arising out of
loss of material during the estimation, an increase in pepsin-con-
centration from 1 per cent to 1.5 per cent multiplies the velocity
of the synthesis over ten times, while an increase in pepsin con-
centration from 1.5 per cent to 2.5 per cent only doubles it;
these facts are obviously irreconcilable alike with direct propor-
tionality between the velocity of synthesis and the concentration
of ferment and with the Schiitz rule of proportionality to the
square root of the concentration of the ferment. The velocity
of reversion is not directly proportional to the concentration
of pepsin while the velocity of hydrolysis is; hence the ratio of
the velocity constants of hydrolysis and reversion must be
dependent upon the concentration of the ferment, or, in other
words, the equilibrium between protein and its products must, to
some extent, be altered by pepsin.

IV. The Influence of the Relative Proportion between the Con-
centration of the Ferment and that of the Products of Hydrolysis
upon the Velocity of Reversion.

The following mixtures of ee products (concentrated 6 times)
of the peptic hydrolysis of *. potassium caseinate and 10 per
cent pepsin solution were fade up and kept at 36° (in the pres-
ence of toluol) in tightly-stoppered flasks:

(a) 75 cc. of prods + 20cc.of ropercent pepsin + 5 cc. H,O.

(b) 5 fo) “ + 5 ° “ 2 “ “
(c) 2 5 “ “ ae 7 5 “ 2 “ “
(d) I “ “ + 8 “ 2 “ “
(e) I 2 “ “ + a “ 2 “ “

After 24 hours an abundant precipitate was produced in (a)
while in (b), (c), (d) and (e) no trace of precipitate had appeared
after 48 hours.

T. Brailsford Robertson 501

The following mixtures (of the products of the complete peptic
hydrolysis of * potassium caseinate, concentrated 6 times, and
pepsin solution) were then made up and kept (in the presence of
excess of toluol) at 36°.

(a) 75 cc. of products + 25cc.of 10 percent pepsin.
(b) 60 “ “ st 4 ro) “ Io “ “
(c) 45 “ “ + 5 “ Io “« “
(d) 35 “ “ a5 65 “ Io “ “
(e) 2 5 “ “ + 7 5 “ Io “ “
(f) I 5 “ “ + 85 “ Io “ “

After 18 hours bulky, flocculent precipitates resembling the
usual precipitate of ‘‘ Paranuclein A”’ which is obtained in rever-
sion, had formed in the mixtures (a), (b), (c), (d) and (e), but no
trace of precipitate had formed in (f); even after 84 hours no
precipitate had formed in (f).

It is evident, therefore, that the solution of the products of
the peptic hydrolysis of .* potassium caseinate which has been
concentrated 6 times can be diluted at least 4 times without
losing the power of yielding ‘‘Paranuclein A” upon treatment
with pepsin provided the pepsin be sufficiently concentrated.

These experiments indicated the possibility that a reversion
of hydrolysis in the unconcentrated solution of the products of
the peptic digestion of * alkali caseinate might be brought
about provided a sufficient concentration of pepsin could be
introduced. Accordingly, to 75 cc. of the unconcentrated prod-
ucts of the complete peptic hydrolysis of .*, potassium hydrate
““saturated”’ with casein were added 25 cc. of 10 per cent pepsin
and excess of toluol. No precipitate whatever was produced
even after the lapse of a month. It is to be observed, however,
that in this experiment the products had actually been diluted
one-fourth. Accordingly 5 grams of Gritbler’s pepsin puriss.
sicc. were dissolved in 100 cc. of the unconcentrated products
and the mixture was allowed to stand (in the presence of excess
of toluol) at 36°, being shaken at intervals. The pepsin took
some hours to fully dissolve and meanwhile the solution had be-
come very cloudy. At the end of 24 hours the pepsin had com-
pletely dissolved, but a flocculent white precipitate, resembling
the usual precipitate of ‘‘Paranuclein A”’ obtained in reversion,

502 Synthesis of Paranuclein

had settled at the bottom of the flask, while the supernatant
fluid was still cloudy. After 48 hours the supernatant fluid was
nearly clear and a bulky flocculent precipitate had formed.
After allowing the system to remain at 36° for 7 days it was
filtered, the precipitate was thoroughly washed with distilled
water and dissolved in a minimal quantity of = sodium hydrate,
the filtrate being caught in excess of dilute acetic acid. The
precipitate thus procured was separated from the faintly acid
fluid by filtration, thoroughly washed in water, alcohol and ether
and dried over calcium chloride. Twenty-seven milligrams of
a grayish-white, friable, highly hygroscopic powder was thus
obtained, resembling, in the properties described, ‘‘ Paranuclein
Jee

It appeared possible that a precipitate might be obtainable
with less pepsin. Accordingly, to 300 cc. of the unconcentrated
products of the complete peptic hydrolysis of ;*, sodium caseinate
were added 6 grams of Grtibler’s pepsin and toluol and the mix-
ture was kept at 36°. Phenomena similar to those described in
connection with the previous experiment were observed and
after 48 hours a fairly bulky precipitate had settled to the bottom
of the flask, leaving the supernatant fluid clear. After 6 days
the precipitate was separated by filtration and purified and
dried in exactly the same manner as that obtained from the
previous experiment. The product obtained (approximately
100 mg.) resembled in its physical properties that obtained from
the solution containing 5 per centof pepsin. The mixed prod-
ucts from the two experiments were analyzed for phosphorus
by Neumann’s method!' with the following result:

0.1305 gm. of substance yielded 0.00299 gm. P,O;:
Hence P,O, = 2.3 percent.

‘“‘Paranuclein A,’’ obtained by digesting paranuclein obtained
from casein with calcium hydrate for 12 hours and that obtained
by the action of concentrated pepsin upon the 6 times concen-
trated solution of the products of the complete peptic hydrolysis
of ,\, alkali caseinate both contain about 1.6 per cent of P,O,;?

1 Neumann: Arch. f. Anat. und Physiol., p. 159, 1900.
27. Brailsford Robertson: loc. cit.

T. Brailsford Robertson 503

hence the product obtained when solid pepsin in considerable
amount is introduced into unconcentrated solutions of the prod-
ucts of the complete peptic hydrolysis of .* sodium caseinate
contains a considerably higher percentage of P,O, than that
obtained by the action of concentrated solutions of pepsin upon
concentrated solutions of the products... I have pointed out,
however, that the paranuclein ordinarily obtained in the initial
stages of the peptic digestion of casein contains a much higher
percentage (4.2) of P,O, than that contained in “‘Paranuclein A’’
and that the paranucleins which have been examined by many
observers have frequently been mixtures of different paranucleins
containing varying percentages of P,O;.2, Under ordinary cir-
cumstances, when the reaction of synthesis occurs in a homoge-
neous system, ‘‘Paranuclein A”’ is the first insoluble product
formed and as, in its formation, it is thrown out of the sphere of
action, no synthesis of paranucleins of higher phosphorus-content
occurs. It is possible that when solid pepsin is introduced into
a solution of the unconcentrated products of the peptic hydrolysis
of 5y alkali caseinate synthesis occurs, in the main, at the sur-
face of the undissolved pepsin or at the surface of suspended
particles of pepsin, as yet incompletely dissolved, since these
would be regions of high pepsin-concentration. It is possible
that under these conditions of intense action some of the higher
members of the paranuclein group are formed and that an admix-
ture of these accounts for the high P,O,; content of the product.

However this may be, the high phosphorus-content of the
product, together with its physical properties and its solubility
in alkalies and precipitability by acetic acid indicate sufficiently
clearly that it is a member of the paranuclein group and that a
true reversion of hydrolysis can be brought about in an uncon-
centrated solution of the products of the complete hydrolysis of
sy alkali caseinate, provided only that the concentration of pep-
sin in the system be sufficiently high and the experimental con-
ditions be chosen aright. Now, reversion cannot take place in the
system at equilibrium unless the concentration be altered or the
equilibrium otherwise shifted; hence it is evident that, in this

1Tt is to be noted, however, that only sufficient material has been as

yet obtained to make one analysis.
?T. Brailsford Robertson: loc. cit.

504 Synthesis of Paranuclein

case, the equilibrium between paranuclein and its products can
be shifted by the introduction into the system of a sufficient quan-
tity of pepsin.

V. The Influence of Temperature upon the Synthesis of
‘“Paranuclein A” through the Agency of Pepsin.

It has been observed by Schwarz! that if solutions of pepsin
be heated to 80° for some time and then added to peptic digests
the digestion is greatly retarded, while Pollack had previously
obtained, by heating pancreas extracts to 70°, a substance which
greatly retards tryptic hydrolysis of proteins;? he further ob-
served that this substance is a colloid, since it does not pass
through the membrane of a dialysor.. Hensel? has further ob-
served that if the mucous membrane of the stomach be treated
with acidulated water at 50°, the watery extract thus obtained
contains an organic substance which greatly retards peptic
hydrolysis of proteins, which is not precipitated by lead salts or
by phosphotungstic acid nor bysix to seven volumes of alcohol;
the author further states that the substance thus obtained does
not appreciably hinder tryptic hydrolysis of proteins nor the action
of ptyalin nor that of rennin.

Bayliss, apparently unaware of the observations quoted above,
for he does not refer to them, has observed that when kept for
some time at a warm temperature trypsin, added to solutions of
gelatin, no longer causes progressive increase in the conductiv-
ity of the solution, as normal trypsin does, but, on the contrary,
causes a marked decrease in conductivity, amounting, in some
instances, to over 100 gemmbhos (1 cc. 2 per cent heated trypsin
added to rocc. of 5 per cent gelatin), followed by a slow increase.‘
The anti-tryptic actions of such substances as egg-white,° nor-

1 Schwarz: Beitr. z. chem. Physiol. u. Pathol., vi, p. 524, 1905.

? Pollack: [bid., iv, p. 95, 1903.

3 Hensel: Sitz. d. Volks-gesundheits ges. zu St. Petersburg, Jan., 1903,
quoted after Biochem. Centralbl., i, p. 404.

4 Bayliss: Arch. d. Sct. Biol., St. Petersburg, xi, Suppl., p. 261, 1904;
reprinted in the Collected Papers of the Physiological Laboratory, Uni-
versity College, London, xiii.

6 Cf. Vernon: Journ. of Physiol., xxxi, p. 346, 1904. Bayliss: loc. cit.

T. Brailsford Robertson 505

mal blood-serum! and extracts of intestinal worms, and that
the anti-peptic and anti-tryptic actions of the bodies produced
in the circulation by the injection of pepsin and trypsin into liv-
ing animals* have usually been attributed to the formation of
amore or lessstable combination between the ferment and the
anti-ferment. One would be inclined to similarly attribute the
action of heated solutions of pepsin and trypsin in inhibiting
these ferments to the formation of compounds between the
heated and the normal ferment! were it not that the heated
ferment, upon addition to a solution of protein, produces a dim-
inution in the conductivity of the solution, while the normal fer-
ment produces an increase in conductivity from the beginning.

‘Hahn; Berl. klin. Wochenschr., xxxiv, p. 499, 1897; quoted after
Euler: Zeitschr. f. physiol. Chem., lii, 152, 1907. Pugliese and Coggi:
(1897), quoted after Euler, loc. cit. Landsteiner: Zentralbl. f. Bakt.,
XXVii, Pp. 357, 1900. Cathcart: Journ. of Physiol., xxxii, P- 390, 1905.
Hedin: Compt. rend. soc. biol., lv, p. 132, 1903.

* Dastre: Compt. rend. soc. biol., p. 634, 1903.

3 Sachs: Fortschritte d. Medicin, xx, P- 425, 1902. Achalme: Ann.
del Institut Pasteur, xv, p.737, 1901. Weinland: Zeitschr. f. Biol., xliv,
Pp. 46, 1902.

*The hypothesis of Schwarz (loc. cit.) that a ‘‘negative catalysor”’ is
developed by heating need only be mentioned to be rejected. In the
first place no one has ever brought forward indubitable proof of the
existence of any negative catalysor whatever, all so-called instances of
“negative catalysis’”’ being, in all probability, due to the removal from
the system of positive catalysors. In the second place, even if a true
“negative catalysor’’ were introduced into a reacting system it would
not retard the main or the positively catalysed reaction, for, on the prin-
ciple of the independence of parallel reactions (cf. Coppadoro: Gazz. chim.
tal., xxxii, p. 425, 1901; V. Henri and Larguier des Bancels: Compt. rend.
soc. biol., liii, p. 784, 1901; lv, p. 864, 1903; Mellor: Chemical Statics and
Dynamics, London, p. 70, 1904), the main and the ‘“‘negatively catalysed”’
reactions would proceed side by side and the more rapid main (uncatalysed
or positively catalysed) reaction would ‘‘mask’’ the slower, negatively
catalysed reaction; since, in a series of parallel reactions, the one which
determines the velocity of transformation is that which takes place most
rapidly; just as in a series of catenary reactions the one which determines
the velocity of transformation is that which takes place most slowly.
The only effect which a true negative catalysor can have, therefore, is
to add to the observed velocity of transformation an amount less than
the velocity of the main (uncatalysed) reaction; no retardation of the
total transformation can occur.

506 Synthesis of Paranuclein

The inhibitory action of the heated ferment is thus clearly seen
to consist in its bringing about a change which is opposite in
sense to that which is brought about by the normal ferment.
Bayliss, regarding the phenomenon from the standpoint of
Ehrlich’s ‘‘side-chain’’ hypothesis, has advanced the opinion
that the decrease in conductivity which is observed upon
the addition of heated ferment to a solution of protein is to be
attributed to the destruction of the ‘“‘zymophore”’ or digesting
group of the ferment, while the ‘“‘haptophore”’ group, by which
the enzyme attaches itself to the protein molecule, is unaffected.
The heated ferment he terms ‘“‘zymoid,” and he believes that
the decrease in conductivity is to be attributed to the formation
of a compound between the ‘“‘zymoid’’ and the protein. It is
improbable, however, that so great a decrease in conductivity
could be produced by the formation of such a compound, since
the enzyme, and therefore the ‘‘zymoid”’ is probably present in
great dilution in all solutions commonly employed and the com-
pound between ferment and substrate is in all cases certainly
very small in amount. Moreover, the hypothesis developed by
Bayliss fails utterly to explain the strong inhibition of proteolytic
activity which heated solutions of pepsin or trypsin bring about
when added to peptic or tryptic digests; since the diminution in
substrate-concentration, due to the formation of a compound
with the ‘‘zymoid’’ would certainly be too minute to appreciably
affect the velocity of the hydrolysis.

The facts point, therefore, to an action of heated pepsin and
trypsin which is reverse in sense to that of the normal enzymes.
It therefore occurred to me that heated solutions of pepsin might
be found to be more efficient in bringing about the synthesis of
paranuclein from the products of the peptic hydrolysis of casein
than normal pepsin. I have made a number of experiments in
which pepsin solutions were heated to various temperatures for
varying periods and added in varying proportions at 36° to con-
centrated solutions of the products of the peptic hydrolysis of
zy alkali caseinates, without having obtained results capable
of any very definite interpretation. I have obtained the syn-
thesis with pepsin solutions which had been heated to 70°, but
have not been able to secure the synthesis at 36°, with smaller
quantities of the heated than of the normal enzyme. It occurred

T. Brailsford Robertson 507

to me, however, that, since the reversion takes from 2 to 12 hours
to become apparent, any effect of heating the ferment-solution
might have been to a great extent lost ere the reversion had had
time to show itself (since, as Bayliss has shown, the proteolytic
activity of the enzyme is slowly regained). I therefore made up
the following mixtures in duplicate, having first ascertained that
the unconcentrated products of the peptic hydrolysis of 3,
alkali caseinates and 10 per cent pepsin can be kept separately
for weeks at 65° without a trace of precipitate forming in either
solution:

(a) rocc. of unconcentrated products + 0.5cc.of 15 percent pepsin.
(b) Io “ “ “ + I “ I 5 «“ “
(c) Io “ “ “« = re 5 “ I 5 “ “
(d) Io “ “ “ ey 2: “ I 5 “ “
(e) Io “ “ “ + 3 “ I 5 “ “

The one set was kept at 65°, while the other was kept at 36°,
both in tightly-stoppered vessels containing excess of toluol.
After 24 hours there was no sign of any precipitate or opalescence
in the mixtures which had been kept at 36° while in the duplicate
set, which had been kept at 65°, (e) contained a heavy precipi-
tate which left the supernatant fluid clear, (c) and (d) contained,
also, heavy precipitates which, however, left the supernatant
fluid strongly opalescent, and (a) and (b) both contained slight
precipitates. After 24 hours more no change had occurred in
any of the solutions and those which had been kept at 65° were
now returned to 36°. After a lapse of three weeks no trace of
precipitate had appeared in any of those solutions which had been
kept at 36° throughout, while no further change had occurred
in those which had been kept at 65° for 48 hours.

The identity between the precipitate thus produced in solu-
tions of the unconcentrated products of the complete peptic
hydrolysis of 5. sodium hydrate ‘‘saturated’’ with casein by the
action of pepsin at 65° and that which is produced by the action
of pepsin upon the concentrated (= 6 times) products at 36°
was shown by the following experiments:

Thirty cc. of 15 per cent pepsin (Grtbler’s puriss. sicc.) were
added to 150 cc. of the unconcentrated products of the complete
peptic hydrolysis of + sodium caseinate and the mixture was

508 Synthesis of Paranuclein

kept at 65° for 48 hours in the presence of excess of toluol. The
resulting precipitate was collected on a hardened filter-paper
and washed with distilled water until the washings were color-
less; it was then dissolved by allowing dilute sodium hydrate
to pass through the filter and reprecipitated by allowing this
filtrate to pass into a beaker containing excess of dilute acetic
acid. The precipitate thus obtained was collected on a hard-
ened filter-paper, well washed with water, alcohol and ether,
and dried over calcium chloride. The product was a grayish-
white, friable, highly hygroscopic powder resembling in its phy-
sical properties and precipitation reactions ‘‘Paranuclein A.”
It was analyzed for phosphorus by Neumann’s method, with the
following result:

0.109 gm. of substance yielded 0.001795 gm. P,O;:
Hence P.O; = 1:65 per cent.

‘“‘Paranuclein A’’ contains 1.6 per cent of P,O,; there can be
little doubt, therefore, regarding the identity of the two products.

A reversion of hydrolysis can be brought about, therefore,
even in the diluted products of the complete peptic hydrolysis
of ,%, solutions of the alkali caseinates (diluted, since varying
amounts of pepsin solutions were added to the solution of prod-
ucts) by the addition of 0.5 cc. of 15 pet cent pepsin to 10 cc. of
products (final concentration of pepsin, 0.75 percent) and keep-
ing the mixture for 24 hours at 65°, while it requires 15 cc. of
Io per cent pepsin in 100 cc. of mixture (final concentration of
pepsin 1.5 per cent) to bring about, in 24 hours, reversion of the
hydrolysis in from four to five times concentrated products at
36°. There can be little doubt that a shift in the station of equi-
librium between ‘‘Paranuclein A’’ and its products occurs as a
result of the addition of pepsin and that this shift in equilibrium
is favored by a rise in temperature.

It is a noteworthy fact that the synthesis which occurs at 65°
does so at a temperature from 10° to 15° in excess of the temper-
ature at which, according to the majority of authors, pepsin is
rapidly and completely deprived of its proteolytic activity.’

Cf. Oppenheimer: Ferments and their Actions, trans. by Ainsworth
Mitchell, London, p. 92, r901. A. E. Taylor: ‘‘On Fermentation,” Univ. of

T. Brailsford Robertson 509

True, the destruction, even at this temperature, must be a matter
of time, and one might be inclined to believe that a short period of
very intense action at 65° produced, in the above experiments, a
similar result to the much more prolonged but weaker action of
pepsin at 36°. The facts are not in favor of this view, however,
since the appearance of the precipitate which marks reversion
does not occur until the solution has been standing at 65° for
two or three hours and it progressively increases in amount for
over 24 hours. It appears that the active agent in reversion 1s
not tdentical with the active agent in hydrolysis.

An experiment which indicates very clearly that a shift in
equilibrium between ‘‘Paranuclein A’ and its products is in-
volved in the synthesis of ‘‘Paranuclein A”’ through the agency
of pepsin is the following:

To 300 cc. of the unconcentrated products of the peptic hydrol-
ysis of .*. sodium caseinate were added 6 grams of dry pepsin.
After 48 hours at 36° (in the presence of excess of toluol) a pre-
cipitate had formed, while the supernatant fluid remained some-
what opalescent; after 6 days the supernatant fluid was quite
clear and a bulky precipitate lay on the bottom of the flask;
the clear fluid was now decanted from the precipitate and divided
into two parts; the one was kept at 36° and the other at 65°;
within 8 hours a precipitate had formed in the latter, the super-
natant fluid being strongly opalescent, while the part of the solu-
tion which remained at 36° had developed no trace of precipitate
or opalescence after a period of two weeks. It is clear, therefore,
that the system had arrived at equilibrium at 36° before the fluid
was decanted, since this fluid must have been ‘‘saturated’’ with
‘“‘Paranuclein A’’ (soluble in these acid solutions only to an unde-
tectable extent) and any further formation of ‘‘Paranuclein A”
would have resulted in an increase in opalescence if not in actual
precipitation. This did not occur, however, even during a period
of two weeks. Yet at 65° a fairly abundant precipitate was pro-
duced within 8 hours.

Calif. Publ., Pathol., i, p. 252, 1907. From the determinations of Schwarz
(loc. cit.) it appears that concentrated (10 per cent) solutions of pepsin
are deprived of their power of accelerating protein hydrolysis and at the
same time acquire considerable power of inhibiting the activity of unal-
tered pepsin, after having heen heated to 60° for five minutes.

510 Synthesis of Paranuclein

THEORETICAL.

The hypothesis which I venture to put forward in explanation
of the above facts is an extension and modification of a somewhat
similar hypothesis which, for somewhat different reasons, has
recently been put forward by Euler.

This investigator has suggested that two varieties of every
enzyme exist, the one accelerating, primarily, hydrolysis, the
other accelerating, primarily, synthesis and that for every definite
proportion of substrate to products there exists a definite equi-
librium between the two forms of enzyme. He bases his view,
in the main, upon the non-constancy of the velocity-constant of
certain enzyme-accelerated reactions with varying substrate-con-
centration and upon the non-identity of the synthesized product
(isomaltose) derived by the action of malt€se upon concentrated
glucose with the maltose from which the glucose was derived by
hydrolysis? and of the product (isolactose) obtained by the action
of lactase upon a concentrated solution of galactose and glucose
with the lactose from which the galactose and glucose are derived.’
Euler believes that the ‘“‘antitrypsins’’ found in serum and in
egg-white are simply the synthesising form of the enzyme and
that the anti-pepsin and anti-trypsin which appear in the circu-
lation after the injection of the enzymes do so as a result of the

1H. Euler: Zeitschr. f. physiol. Chem., lii, p.146, 1907. The essential
features of the hypothesis which is put forward in this paper were already
present in my mind two years ago, when I wrote my paper on the hydrol-
ysis of casein (this Journal, ii, p. 317, 1907) but the facts in my possession
at that time were not of a sufficiently specific nature to warrant the pub-
lication of the hypothesis. I state this, not with the remotest idea of
claiming priority for such features of the hypothesis as to some extent
resemble that put forward by Euler, but merely to point out how two
different observers, approaching the question from very different sides,
have been independently led to somewhat similar conclusions.

2 Croft Hill: Journ. Chem. Soc., London, 1xxiii, p. 634, 1898; Ber.
d. deutsch. chem. Gessellsch., xxxiv, p. 1380, 1901. Emmerling: Ber.
d. deutsch. chem. Gessellsch., xxxiv, pp. 600 and 2206, rgor.

3 Fischer and Armstrong: Jbid., xxxv, p. 3144, 1902. This latter
argument of Euler’s, however, loses its force when we recollect that even
sulphuric acid, acting upon a concentrated solution of glucose, gives rise
not to maltose, but to isomaltose (Wohland Fischer, quoted after A. E.
Taylor: ‘‘On Fermentation,’’ Univ. of Calif. Publ. Pathol., p. 185, 1907).

T. Brailsford Robertson 511

production of the synthesising form in restoration of the equi-
librium between the synthesising enzyme and the products of
hydrolysis to which the hydrolysing enzyme and the products of
hydrolysis to which the hydrolysing enzyme gives rise in the
blood. He further believes that the ‘‘plastein formation”’
observed by many investigators upon adding concentrated solu-
tions of enzymes to digests is to be attributed to the action of the
synthesising form of the enzyme.

The hypothesis which I have to suggest is one which I venture
to term an hypothesis of ‘‘ Reciprocal Catalysis.”’ Let us con-
sider, for a moment, the possible chemical mechanics of the
hydrolysis of proteins through the agency of enzymes. In a
previous paper! I have suggested that, during the hydrolysis of
proteins by a proteolytic enzyme, a product of theenzyme may be
formed which regenerates the enzyme as rapidly as it is formed.
We may throw this suggestion into a somewhat more concrete
form as follows: We know that during or preceding the hydrol-
ysis of proteins by proteolytic enzymes the ferment combines with
the substrate;? it is furthermore probable that since pepsin and
trypsin will digest both proteins which are predominantly acid
and those which are predominantly basic these enzymes are able
to form combinations with each type of protein and are therefore
themselves amphoteric electrolytes; hence we may, with reason-
able assurance, represent the first stage of the reaction between
enzyme and substrate, schematically, as follows:

HXX: OH + H;FFOH = HXXFFOH + H,O (T]

When the ferment-substrate compound breaks down the reaction
may, not improbably, be represented as follows:

HXXFFOH + H,O = 2HXOH + FF (11)
while, subsequently, the enzyme-product FF reacts with water

FF + H,O = HFFOH {IIT}

1T. Brailsford Robertson: this Journal, ii, p. 380, 1907.

2Cf. Vernon: Journ. of Physiol., xxxi, p. 346, 1904. V. Henri: Compt.
rend, soc. biol., lviii, p. 610. Dauwe: Beitr. z. chem. Physiol. u. Path.,
vi, p. 426, 1905. Ibid., vii, p. 151, 1905. T. Brailsford Robertson: loc.
cit. Bayliss: loc. cit. S.G. Hedin, Biochem. Journ., ii, p. 81, 1907.

512 Synthesis of Paranuclein

the net result of the first two reactions being the transference
of the elements of water from the ferment to the substrate mole-
cule, while in the third reaction, the ferment recoups itself from
the medium. Provided the station of equilibrium in the reaction
FF + H,O @ HFFOH lay far enough to the right and the vel-
ocity of this reaction measured from left to right, were great
compared with that of either of the reactions I and II measured
from right lo lejt, the enzyme would, through the greater part of
the reaction, simply accelerate the hydrolysis through multi-
plying the velocity-constant of the hydrolysis by a factor pro-
portional to its concentration. Under these circumstances, also,
as I pointed out in the paper referred to above, the monomolecu-
lar formula, which is the one experimentally obtained for the
hydrolysis of protein by pepsin and by trypsin,’ would hold
good, provided reaction II, in the above scheme, proceeded at an
infinite velocity compared with reaction I. I neglected to ob-
serve, however, that the experimental equation can also be
obtained provided ezther of these two reactions occurs at an infi-
nite velocity as compared with the other; the following consider-
ations will make this clear. Suppose that at any moment the
concentration of the ferment-substrate compound is x while that
of the products of hydrolysis is y, then we have, at that moment
provided the retarding influence of the small concentration of the
ferment substrate compound can be neglected:
On
cy a ey)

which represents the velocity with which reaction I proceeds
from left to right, where a is the initial concentration of the sub-
strate and f is the concentration of the ferment (which is practi-
cally constant under the conditions assumed above regarding
reaction III).
Similarly we have:
dy
i

for the velocity of reaction II from left to right, provided, as
assumed, the concentration of the enzyme-modification FF is

* Cf. literature quoted in footnote 2, p. 495.

T. Brailsford Robertson Bes

vanishingly small, so that the velocity of reaction II from right
to left is negligiblé.
Let x + vy =2, hence:
Ge dz - dy
dt \di

If, now, the velocity of reaction II be so great compared with that
of reaction I that at any moment equilibrium may be supposed
to have been attained and

dy A dz S

pe Fer, then ae as 7(a — 2),
which is the equation experimentally obtained.

If, however, the velocity of reaction I be so great conipared
with that of reaction II that, atany moment, the condition of
equilibrium in reaction I may be supposed to have been attained
and

dx
Gp 7 280
then, from equation I,
K gis RE ae
x =Kf(a—z)an aE rae re f (a — 2)

which is also the equation experimentally obtained.

It is evident that the latter supposition (velocity of reaction
I great compared with that of reaction II) is the one which rep-
resents the true state of affairs, and not, as suggested in my
former paper, the former (velocity of reaction II great compared
with that of reaction I); since the ferment-substrate compound
can be and has been obtained in a fairly stable form,’ which
would be impossible were the velocity of its decomposition great
compared with the velocity of its formation. The chemical

1 Cf. footnote 2, on p. 511. Thus trypsin can be extracted from its
solution by coagulated egg-white or by fibrin and the ferment can be
regained from the compound by prolonged washing with water. This
latter fact does not in the least militate against the view that the com-
bination between the ferment and the substrate is chemical in character
(cf. T. Brailsford Robertson: this Journal, ii, p. 378, 1907; Zeitschr. jf.
Chemie und Industrie der Kolloide, iii, p. 26, 1908).

514 Synthesis of Paranuclein

reaction-velocity actually measured, therefore, in estimations of
the rate of proteolysis by enzymes, is that of the decomposition
of the ferment-substrate compound (reaction II).!

So much for the kinetics of protein hydrolysis under the con-
dition, during the initial stages of the reaction, of negligibility of
the concentration of FF compared with that of the hydrated form
of the enzyme, HFFOH. As regards the statics of the hydrol-
ysis, however, it is evident that, if the above scheme be correct,
the enzyme cannot be regarded asa simple catalysor but that its
presence must result in a greater or smaller shifting of the point
of equilibrium between the protein and its products; this is evi-
dent, for the hydrated form HFFOH only accelerates the hydrol-
ysis, while the anhydrous form FF only accelerates the synthesis,
and since these are, in general, present in unequal concentrations,
the forward and reverse reactions of protein hydrolysis must be
unequally accelerated, and hence equilibrium must be shifted.
Since, however, for every shift in equilibrium, there must be a
corresponding expenditure of energy, the equilibrium between
the anhydrous and hydrated forms of the enzyme must also be
shifted by the protein; just as the enzyme accelerates the hydrol-
ysis of the protein more than its synthesis because the hydrate
form of the enzyme is initially present in considerable excess of
its anhydrous form, so the protein accelerates the dehydration
of the enzyme more than its hydration because it is initially
present in great excess of its products of hydrolysis. This latter
fact, however, will itself lead to a slowing of the hydrolysis of the
protein, since the hydrated (and hydrolysis-accelerating) form
of the enzyme is thereby diminished in concentration; as the
hydrolysis proceeds, however, this effect will diminish, the prod-
ucts of the protein hydrolysis will tend to increase the propor-
tion of the hydrated form of the enzyme to the anhydrous form,
and the rate of hydrolysis will increase at the expense of the rate
of synthesis.2. Ultimately, it is evident that a condition of equi-

1T am indebted to Dr. F. G. Cottrell for having pointed out to me
this alternative to my former conclusion.

? Hence the velocity-constant of hydrolysis would increase with time;
this effect is, however, usually masked by the auto-digestion of the
enzyme which occurs simultaneously, and also by the retardation of
reactions I and II through the mass-action of the products of hydrol-
ysis.

———————

T. Brailsford Robertson 515

librium must be reached in which the station of equilibrium be-
tween the protein and its products is shifted further in the direc-
tion Protein — Products than its position in the absence of the
enzyme, while the station of equilibrium between the hydrated
and anhydrous forms of the enzyme is shifted further in the direc-
tion Hydrated form — Anhydrous form than its position in the
absence of the protein. This position of equilibrium will de-
pend, obviously, upon the total concentration of the enzyme
and of the substrate, respectively, and, once attained, a further
addition of substrate would re-inaugurate the hydrolysis of pro-
tein, it is true, because the active mass of hydrolyzable material
would be thus increased, but it would shift the point of equ-
librium in the direction Products— Protein; addition of enzyme
would shift the station of equilibrium in the direction Protein
— Products, as has been found by Bayliss! and others.2 This
latter statement, however, holds good only while the water in the
system is in great excess of the enzyme, so that varying the con-
centration of the enzyme does not appreciably affect the propor-
tion subsisting between the hydrated and anhydrous forms at
equilibrium in the absence of proteins; only while thiscondition
holds, also, can direct proportionality between the velocity of
hydrolysis and the concentration of the enzyme be observed.
If, however, the enzyme be very concentrated, then the propor-
tion of water to enzyme will appreciably affect the equilibrium
in the equation FF + H,O @ HFFOH and there will be present
a relatively greater proportion of the anhydrous (synthesis-
accelerating) form FF; hence, under these conditions, a further
addition of enzyme will shift the equilibrium of the protein in the
direction Products — Protein.

A remarkable feature of the syntheses of protein through
enzyme-agency which have been accomplished both by Taylor
and myself is the high concentration of enzyme which has to be
employed; the reason for this is now clear; the highly concen-
trated enzyme actually shifts the equilibrium of the protein in the
direction of synthesis. It is now clear, also, why a sufficiently

1 Bayliss: loc. cit.
2A. E. Taylor: ‘‘On Fermentation,’’ Univ. of Calif. Publ. Pathol., p.

133, 1907.

516 Synthesis of Paranuclein

high concentration of enzyme will actually bring about synthesis
of protein in the wnconcentrated products of the complete hydrol-
ysis of a solution of protein, brought about by the agency of
dilute enzyme (cf. section 4, experimental part). The depend-
ence of the velocity-constant of hydrolysis upon the initial con-
centration of the substrate (cf. section 2, experimental part) is
also readily comprehended since, as pointed out above, increase
in the concentration of substrate must, in the initial stages of
hydrolysis, lead to an increase in the relative proportions of the
anhydrous to the hydrated form of the enzyme (cf. equations
I and II) and hence to a diminution of the concentration of the
hydrolysis-accelerating form of the enzyme; hence the velocity-
constant of hydrolysis diminishes with increasing substrate-con-
centration (cf.: section 3, experimental part).

The relation between the anhydrous and hydrated forms of the
enzyme is, not improbably, analogous to that subsisting between
the ‘‘internal salt,’’?

of an amphoteric electrolyte (amino-acid) and its hydrated form,

NH,OH
Reo

\cooH

I have pointed out elsewhere? that the action of heat upon a
solution of a protein must be interpreted as a shifting of the
equilibrium among complexes of the type HXOH in the direc-
tion of higher complexes of the types HXX....0H; doubtless
the equilibrium between the anhydrous and the hydrated form of
the protein (amphoteric electrolyte) molecule is quite different
among the higher complexes to that which obtains among the
lower complexes and coagulation itself not improbably indicates

1 Cf. Winkelblech: Zeitschr. f. phys. Chem., xxxvi, p. 546, 1901. T.
Brailsford Robertson: Journ. of Physical Chem., v, p. 524, 1906.
2 T. Brailsford Robertson: this Journal, v, p. 147, 1908.

T. Brailsford Robertson 517

an increase in internal-salt formation... The action of high
temperatures in enhancing the synthesizing power of the enzyme
while diminishing its power of accelerating hydrolysis (section
5, experimental part) may be interpreted in either of two ways.
Either the high temperature destroys the hydrated form by accel-
erating its hydrolysis,? while leaving the anhydrous form unaf-
fected, so that during the period that the anhydrous form is
changing into the hydrated form only the synthesis of the protein
is being accelerated and not its hydrolysis, or, more probably, in
view of the rapidity with which the conversion of the anhydrous
form into the hydrated form probably takes place, the high
temperature actually shifts the equilibrium of reaction III in the
direction HFFOH — FF + H,0O.

As one approaches the iso-electric point (usually near the
neutral point) in solutions of amphoteric electrolytes, the amount
of internal salt formation increases and, at the iso-electric point,
when the numbers of hydrogen and hydroxyl ions which are split
off from the molecule are equal to one another, the amount of
internal salt formation is a maximum,* hence the precipitation
which frequently occurs at or near the neutral point. This pos-
sibly explains the observations of Robertson and Schmidt,*

1Cf. T. Brailsford Robertson: loc, cit, Gustav Mann: Chemistry of
the Proteids, London, p. 318, 1906. Wm. Sutherland: Proc, Roy. Soc.
London, p. 130, 1906.

* The fact that heat shifts the equilibrium of the protein in the direc-
tion of higher complexes is not, of course, inconsistent with the fact
that it may also accelerate the hydrolysis of the lower complexes. The
action of heat upona chemical reaction istwofold. It invariably acceler-
ates the reaction in whichever direction it occurs, but it accelerates the
forward and reverse reactions unequally, so that the equilibrium of the
reacting substances is shifted in a direction and to a degree which can
be anticipated from thermochemical data. Now although the applica-
tion of heat shifts the equilibrium of the protein towards the higher
complexes, it must also accelerate the hydrolysis of the lower complexes
(and thus, indirectly of the whole of the protein), since, as hydrolyses with
steam demonstrate, these are not in equilibrium with the products of
their hydrolysis.

3T. Brailsford Robertson: Journ. of Physical Chem., x, p. 524, 1906.

This, indeed, obvious, since when two parts of the same mass are equal
in magnitude their algebraical product is a maximum.

4T. Brailsford Robertson and C. L. A. Schmidt: this Journal, v, p. 31,
1908.

518 Synthesis of Paranuclein

who measured the decrease in the alkalinity of protein digests
(tryptic) as digestion proceeds, using very low alkalinities and
employing the gas-chain as a means of estimation. They found
that for alkalinities exceeding 10 °OH™ the decrease in OH con-
centration with time follows the monomolecular law; as the alka-
linity approaches 10 *OH_, however, the character of the hydrol-
ysis somewhat abruptly changes over and it henceforth obeys,
fairly accurately, the bimolecular law, indicating a more rapid
progressive decrease in velocity with time than that at higher
alkalinities. Ifthe OH™ concentration is initially below the limit
10 °OH-, the progressive decrease in OH™ concentration obeys
the bimolecular law throughout; if alkali be added to a solution
undergoing diminution in alkalinity according to the bimolecular
law, the monomolecular law is regained. As the authors point
out these phenomena are such as to indicate a progressive dimi-
nution of the active trypsin as the alkalinity falls below a certain
limit.1. It is possible that this phenomenon is to be attributed
to a progressive increase in internal salt formation in the enzyme
as neutrality is approached, and a consequent progressive increase
in the anhydrous (synthesis-accelerating) form of the enzyme at
the expense of the hydrated (hydrolysis-accelerating) form of the
enzyme. In support of this view may also be cited the fact
observed by Robertson and Schmidt, that the station of equi-
librium attainedin a tryptic digest varies with the initial allkalin-
ity, a greater amount of alkali being neutralized the greater the
initial concentration, since the onset of neutral-salt formation
would thereby be delayed. This also serves to explain why an
addition of trypsin failed to reinaugurate the hydrolysis, since
the added trypsin would, after addition in small amount contain
nearly the same relative proportion of anhydrous to hydrated
form as that already present in the solution.’

1 These observations of Robertson and Schmidt’s are, so far as I am
aware, the only accurately controlled measurements which have been
conducted upon the tryptic hydrolysis of proteins at low and varying alka-
linities. Other observers have either used higher alkalinities, at which
the monomolecular law holds good, or else neutral digests, in which the
alkalinity does not vary as the digestion proceeds.

? This fact also serves to differentiate many of the equilibria attained
in protein digests from the ‘‘false equilibria’’ attained in certain catalytic

.

bd
;

~*~

|

|

T. Brailsford Robertson 519

The thermodynamical aspect of the hypothesis developed above
is of interest. As I have pointed out, the shift in equilib-
rium of the system Protein = Products towards the right which
results from the introduction into the system of the enzyme must
result in a corresponding shift in the equilibrium of the system
Anhydrous enzyme @ Hydrated enzyme towards the left and vice
versa. Since the enzyme is usually present in small concentra-
tion compared with the protein, the shift in the equilibrium be-
tween the two forms of the enzyme must be large compared with
that of the equilibrium between the protein and its products;
or else the energy expended in a shift of the enzyme-equilibrinm
must be great compared with the energy expended in a shift of the
protein-equilibrium. The latter appears to me the more probable
view; since, otherwise, the shift in the protein-equilibrium would
probably be in all cases too small to be observed, and, moreover,
we know that the reaction of protein hydrolysis is only very
slightly exothermic? so that the energy change involved in a shift
in the equilibrium between protein and the products of its hydrol-
ysis is probably comparatively small.

It may here be noted that the hypothesis outlined above is not
inconsistent with the generally expressed view that the enzyme
may be recovered without appreciable loss of activity from a
protein digest which has reached equilibrium. Even if, in con-
sequence of the presence of protein, the station of equilibrium
between the two modifications of the enzyme has been shifted in
sucha direction as to diminish its power of accelerating hydrolysis,
yet upon removal from the system, the enzyme would regain its
normal equilibrium rapidly (since, as we have assumed through-
out, reaction III proceeds rapidly in comparison with the reac-
tions Iand II)? and the energy thus apparently gained would be

reactions. (Cf. Mellor: Chemical Statics and Dynamics, London, p. 297,
1904; G. Tammann: Zeittschr. f. physiol. Chem., xvi, p. 285, 1892; xviii,
Pp. 428, 1895.

1Cf. Tangl: Arch. f. d. ges. Physiol., cxv, p. 1, 1906. v. Lengyel:
Ibid., p. 7. Hari: Ibid., p. 11; ibid., cxxi, p. 459, 1908. Henderson and
Ryder: Proc. of the Amer. Soc. of Biol. Chemists, i, p. 26, 1907.

* The justification for this assumption lies in the fact that otherwise
the monomolecular equation for the hydrolysis of protein could not be
obtained even in the presence of dilute enzyme; unless, indeed, reactions
I and II proceeded at a velocity infinitely great compared with that of

520 Synthesis of Paranuclein

slowly lost through the slow (because of the absence of enzyme)
regaining of equilibrium between the protein and its products.
The Protein @ Products system would, in other words, be left
‘‘supersaturated’’ in respect to products, or, were the enzyme
in the system highly concentrated, or did it, for any other reason,
initially contain a high proportion of the anhydrous (synthesis-
accelerating) form so that it gained power of accelerating hydrol-
yses as a result of the presence of the protein, then the Protein
= Products system would be left, by removal of the enzyme,
‘‘supersaturated’”’ with respect to protein. By mechanical sepa-
ration of the enzyme and substrate and through the slowness
with which, in the absence of enzymes, the majority of proteins
regain their equilibrium, it is obvious that if such reciprocal
relations as that outlined above find a place in living material,
the organism may be enabled thereby to temporarily and locally
store up large quantities of energy. The significance of this
possibility in the general interpretation of life-phenomena is
patent.

Finally, it will be admitted, I think, that the term “‘Recip-
rocal Catalysis” suggested for the sum of the processes described
above is a suitable one, since the relation between the protein
and the enzyme is assumed to be a reciprocal one, bearing
superficial resemblances to the relation between catalysor and
substrate. The enzyme being assumed to carry water into the
protein molecule and, parting with water, to recoup itself from
the medium, while the protein, upon the introduction of the
water into the molecule, splits up into the products of its hydrol-
ysis; part passu, the products of the protein hydrolysis are as-
sumed to part with water to the anhydrous form of the enzyme,
whereby protein is regenerated and the hydrated form of the
enzyme is set free. For every substrate and enzyme concentra-
tion and for every temperature and, probably, for every alka-
linity or acidity, a definite ratio between the velocities of these
two reciprocal processes exists and this ratio determines the final
equilibrium of the system, and, at this equilibrium, the propor-
tion subsisting between the concentrations of the anhydrous

equation III, but in that case combinations of the substrate with the
enzyme could never be obtained in an even approximately stable con-
dition.

T. Brailsford Robertson 521

(synthesis-accelerating) form of the enzyme and its hydrated
(hydrolysis-accelerating) form, no less than that subsisting be-
tween the protein and the products of its hydrolysis.

CONCLUSIONS.

(1) It is pointed out that the hypothesis that the synthesis
of protein which may be brought about in concentrated solutions
of the products of the hydrolysis of protamin by the addition of
large quantities of trypsin (Taylor) or in concentrated solutions
of the products of the complete hydrolysis of casein (and, there-
fore, of paranuclein) by the addition of large quantities of pepsin
(Robertson) are examples of the action of pure catalysors in
accelerating both forward and reverse reactions, is incompatible
with the following facts:

(2) That the velocity-constant of hydrolysis diminishes with
increasing substrate-concentration.

(3) That the velocity of the synthesis of ‘‘Paranuclein A”
which is brought about in concentrated solutions of the products
of the complete peptic hydrolysis of §, sodium or potassium case-
inates by the addition of a large quantity of pepsin does not vary
directly as the concentration of the pepsin, but falls off abruptly
as the concentration of the pepsin falls below a certain value.

(4) That upon the introduction of a sufficient quantity of solid
pepsin into a solution of the unconcentrated products of the com-
plete peptic hydrolysis of X sodium or potassium caseinates a
precipitate is produced which, upon isolation and purification,
is found to closely resemble the paranucleins in its physical prop-
erties and precipitation-reactions and high phosphorus-content
(2.3 per cent P,O;). It is pointed out, in this connection, that
since reversion cannot take place in a system at equilibrium
unless the concentration be altered or the equilibrium otherwise
shifted, in this case, the equilibrium between paranuclein and its
products must be shifted by the introduction into the system
of excess of pepsin.

(s) That at atemperature (65°), 10° to 15° above the “thermal
death-point”’ of pepsin a rapid synthesis of a substance identified
as ‘“‘Paranuclein A” (1.65 per cent of P,O,; ‘‘Paranuclein A”’ con-
tains 1.6 per cent P,O,) occurs in mixtures of concentrated pepsin

522 Synthesis of Paranuclein

solution and the unconcentrated products of the complete peptic
hydrolysis of 4, sodium or potassium caseinate. From this, and
for other reasons mentioned in the body of the paper, it is con-
cluded that the active agent in reversion is not identical with
the active agent in hydrolysis.

(6) I have suggested, in explanation of these phenomena,
an hypothesis of “‘ Reciprocal Catalysis,’’ as follows:

Representing the protein by the schematic formula HXXOH
and the enzyme by the similar formula HFFOH, for various
reasons dwelt upon in the body of the paper it is suggested that
the various steps in the hydrolysis of proteins through the agency
of enzymes may be represented as follows:

HXXOH + HFFOH = HXXFFOH +H,0 (1]
HXXFFOH + H,O = 2HXOH + FF [11]
FF + H,0 = HFFOH [IIT]

while the synthesis may be represented as the reverse of these
reactions. It is pointed out that all the above-mentioned facts
(Conclusions 2-5), as well as others to which reference is made in
the body of the paper, admit of simple interpretation upon the
basis of this hypothesis. It is concluded that the velocity of
protein hydrolysis, as experimentally measured, is, in the pres-
ence of dilute enzyme, determined by the reaction II. The
thermodynamical aspect of the hypothesis is briefly considered.
It is also pointed out that this hypothesis is not inconsistent with
the generally expressed view that the enzymes may be recovered
without appreciable loss of activity from a protein digest which
has reached equilibrium.

(7) The essential features and theoretical consequences of
this hypothesis may be summarized as follows: The relation
between the protein and the enzyme is assumed to be a reciprocal
one, bearing superficial resemblances to the relation between
catalysor and substrate; the enzyme being assumed to carry
water into the protein molecule and, parting with the water, to
recoup itself from the medium, while the protein, upon the intro-
duction of water into the molecule, splits up into the products of
its hydrolysis; part passu, the products of the protein hydrolysis

T. Brailsford Robertson 523

are assumed to part with water to the anhydrous form of the
enzyme, whereby protein is regenerated and the hydrated form
of the enzyme is set free. For every substrate and enzyme con-
centration and for every temperature and, probably, for every
alkalinity or acidity, a definite ratio between the velocities of
these two reciprocal processes exists, and this ratio determines
the final equilibrium of the system, and, at this equilibrium, the
proportion subsisting between the concentrations of the anhy-
drous (synthesis-accelerating) form of the enzyme and its hy-
drated (hydrolysis-accelerating) form no less than that subsist-
ing between the protein and the products of its hydrolysis.

(8) It is suggested that the relation between the anhydrous
and the hydrated forms of the enzyme may be similar to that
between the ‘‘Internal salt’’ and the hydrated forms of an amido-
acid.

Further experiments of the same general nature as those
described in this paper, together with the preparation in bulk of
the various products obtained, with a view to their complete
analysis, are in progress.

INDEX TO VOLUME V.

Acid, aspartic, oxidation of, 409

Acid, glutamic, oxidation of, 409.

Acid, -oxybutyric, detection and
quantitative determination of in
the urine, 207

Acid, @-oxybutyric, method for the
quantitative determination of in
the urine, 211

Acid, phenylpropionic, synthesis of
some derivatives of, 303

Acid, uric, excretion of in normal
men, 355

Acid, yeast nucleic, reducing com-
ponent of, 469

Acids, fatty, mode of oxidation of
phenyl! derivatives of in the ani-
mal organism, 173, 303

Acids, nucleic, identity of, 1

Adsorption of diastase and cata-
jase by colloidal protein and by
normal lead phosphate, 367

Alcohol, antagonism of to carbolic
acid, 319

ALsBERG, C. L. and E. D. Ciarx:
On a globulin from the egg yolk
of the spiny dog fish, Squalus
Acanthias, 243

AMSBERG. ©, lL. and BE. D. CLARK:
The blood clot of Limulus Poly-
phemus, 323

Ammonia, comparative tests of
Spiro’s and Folin’s methods for
determination of, 477

Ammonia, notes on the efficiency of
the Folin method, 71

Amphoteric electrolytes, applicabil-
ity of laws of to serum globulin,

155

Antagonism of alcohol to carbolic
acid, 319

Aspartic Acid, oxidation of, 409

Bacillus bulgaricus, fate of in diges-
tive tract of a monkey, 293

Bacillus injantilis (n.s.) and its
relation to infantilism, 419

Bacillus injantilis, products of
grown on artificial media, 439

Bacteriology, intestinal, use of fer-
mentation tube in, 283

BENEDICT, STANLEY R.: A reagent
for the detection of reducing
sugars, 485

Buack, O. F.: The detection and
quantitative determination of
f-oxybutyric acid in the urine,

207
Blood clot of Limulus Polyphemus,

2
Blood: quantitative estimation of
reducing substances in, 443
Boos, WiLtiaM F.: On the reducing
component of yeast nucleic acid,
46
ee Tueo. C.: The inhibit-
ing effect of potassium chloride in
sodium chloride glycosuria, 351

Cane sugar, inversion of, 405

Carbolic acid, antagonism of alcohol
to, 319

Casein, influence of temperature
upon the solubility of, 147

Catalase, adsorption of by colloidal
protein and by normal lead phos-
phate, 367

Celtis Reticulosa (Miquel), occur-
rence of skatol and indol in the
wood of, 489

Chloroform necrosis (delayed chloro-
form poisoning) chemistry of the
liver in, 129

Chymosin, question of the identity
of with pepsin, 399

CLapp, SAMUEL H.: see Johnson and
Clapp, 49, 163

Cea D. : see Alsberg and Clark
243, 323 :

CoLwWELL ,RacuEL H. and H. C.
SHERMAN: Chemical evidence of
peptonization in raw and pas-
teurized milk, 247

Cytolysis, chemical studies in, 311

Cytosin, action of diazobenzene sul-
furic acid on, 163 ;

Cytosin, nitrogen-alkyl derivatives
of, 49

525

526

Daxin, H. D.: Further studies of
the mode of oxidation of phenyl
derivatives of fatty acids in the
animal organism, 173

Dakin, H. D.: Further studies of
the mode of oxidation of phenyl
derivatives of fatty acids in the
animal organism; III. Synthesis
of some derivatives of phenyl-
propionic acid, 303

Dakin, H. D.: Note on the oxida-
tion of glutamic and aspartic
acids by means of hydrogen per-
oxide, 409

Dakin, H. D.: The action of glyco-
coll as a detoxicating agent, 413

Delayed chloroform poisoning,
chemistry of the liver in, 129

Detoxicating action of glycocoll,
413

Diastase, adsorption of by colloidal
protein and by normal lead phos-
phate, 367

Diazobenzene sulfuric acid, action
of on thymin, uracil and cytosin,
163

Diet, effect of on the maltose- split-
ting power of the saliva, 331

Digestive tract, fate of B. bulgari-
cus in, 293

Diphtheria toxin, concentration of,
27

Egg-white, sensitizing portion of,

253

Egg-yolk of Squalus Acanthias,glob-
ulin from, 243

Electrolytes, amphoteric, applica-
bility of laws of to serum globu-
Lint 5

EMERSON, JuLia T. and WILLIAM
H. WELKER: Some notes on the
chemical composition and toxic-
ity of [bervillea Sonore, 339

Enzymes, chemical mechanics of
the hydrolysis and synthesis of
proteins through the agency of,
493

Enzymes, studies on, 367

Fatty acids, mode of oxidation of
phenyl derivatives of in the ani-
mal organism, 173, 303

Fermentation tube, use of in intes-
tinal bacteriology, 283

Folin method for determination of
urinary ammonia, 71, 477

Index

Gres, WILLIAM J., see Steel and
Gies, 71

Globulin from egg yolk of Squalus
Acanthias, 243

Globulin, serum, applicability of
laws of amphoteric electrolytes
to, 155

Glutamic acid, oxidation of, 409

Glycocoll, action of as a detoxicat-
ing agent, 413

Glycogen, conversion of into sugar
in the liver, 315

Glycosuria, sodium chloride, inhibit-
ing effect of potassium chloride
bay gist

Hanz_ik, Paut J. and P. B. Hawk:
The uric acid excretion of normal

men, 355

Hawk, P. B.: see Hanzlik and
Hawk, 355

Hawk, P. B.: see Howe and Hawk,
477

HEINEMANN, P. G.: Note on the
concentration of diphtheria toxin,

27],

‘ Herter, C. A.: Note on the occur-

rence of skatol and indol in the
wood of Celtis Reticulosa (Miquel),
489

Herter, C. A. and A. I. KENDALL:
An observation on the fate of B.
bulgaricus (in Bacillac) in the
digestive tract of a monkey, 293

Herter, C. A. and A. I. KENDALL:
Note on the products of B.
injantilis grown on artificial
media, 439

Herter, C. A. and A. I. KENDALL:
The use of the fermentation tube
in intestinal bacteriology, 283

HeYL, FREDERICK W.: see Osborne
and Heyl, 187, 197

Howe, Paut E. and P. B. Hawk:
Comparative tests of Spiro’s
and Folin’s methods for the
determination of ammonia and
urea, 477

Hydrogen peroxide, oxidation of
glutamic and aspartic acids by
means of, 409

Hydrolysis of legumelin, 197 :

Hydrolysis of proteins by trypsin,
part played by the alkali in, 31

Hydrolysis of proteins through the
agency of enzymes, chemical
mechanics of, 493

Hydrolysis of vicilin, 187

Index

Ibervillea Sonore, chemical com-
position and toxicity of, 339

Identity of pepsin and chymosin,
question of, 399

Indol, occurrence of in the wood of
Celtis Reticulosa (Miquel), 489

Infantilism, relation of B. zufantilis
to, 419

Inhibiting effect of potassium chlo-
ride in sodium chloride glycosu-
Tia, 351

Intestinal bacteriology, use of fer-
mentation tube in, 283

Inversion of cane sugar and mal-
tose, 405

Ionic potentials of salts and power
of inhibiting lipolysis, relation be-
tween, 453

Jounson, Treat B. and SAMUEL
H. Crapp: IX. Researches on
pytimidins: Synthesis of some
nitrogen-alkyl derivatives of cy-
tosin, thymin and uracil, 49

Jounson, Treat B. and SAMUEL
H. Ciapp: X. Researches on
pyrimidins: The action of diazo-
benzene sulfuric acid on thymin,
uracil and cytosin, 163

Jones, WALTER: On the identity
of the nucleic acids of the thy-
mus, spleen and pancreas, 1

KENDALL, ARTHUR I.: Bacillus
tnjantilis (n.s.) and its relation
to infantilism, 419

KENDALL, A. I.: see Herter and
Kendall, 283, 293

Leacu, Mary F.-.: A preliminary
study of the sensitizing portion of
egg-white, 253

Lead phosphate, normal, absorp-
tion of diastase and catalase by,

367

Legumelin, hydrolysis of, 197

Limulus Polyphemus, blood clot of,
323

Lipolysis, relationship between ionic
potentials of salts and their power
of inhibiting, 453

Liver, chemistry of, in chloroform
necrosis, 129

Liver of reptiles, note on the
chemistry of, 125

Lyman, Joun F.: A note on the
chemistry of the muscle and liver
of reptiles, 125

527

Mac.Leop, J. J. R.: A comparison
of the methods of Reid and
Schenck for the quantitative
estimation of the reducing sub-
stances of the blood, 443

Magnesium sulfate, influence of on
metabolism, 85

Maltose, inversion of, 405

Maltose-splitting power of the sali-
va, effect of diet on, 331

Mechanics, chemical, of the hydrol-
ysis and synthesis of proteins
through the agency of enzymes,

493

Metabolism,i nfluence of magnesi-
um sulfate on, 85

Metabolism, intermediary, influ-
ence of thyroidectomy and thy-
roid feeding upon, 225

Milk, peptonization in, 247

Milk proteins, 261

Muscle of reptiles, note on the
chemistry of, 125

NEILSON, CHARLES HucuH and M.
H. ScHEE Le: The effect of diet
on the maltose-splitting power
of the saliva, 331

NicHo.u, R. H.: The relationship
between the ionic potentials of
salts and their power of inhibit-
ing lipolysis, 453

Nucleic acid, yeast, reducing com-
ponent of, 469

Nucleic acids, identity of, 1

OusEN, GeorcE A.: Milk proteins,
261

OsBoRNE, THomas B. and FREpD-
ERICK W. Hey: Hydrolysis
of legumelin from the pea (P1-
sum sativum), 197

OsBoRNE, TuHoMAS B. and FRED-
ERICK W. Hey: Hydrolysis
of vicilin from the pea (Pisum
sativum), 187

Oxidation of glutamic and aspar-
tic acids, 409

Oxidation of phenyl derivatives of
fatty acids in the animal organ-
ism, 173, 303

fB-Oxybutyric acid, detection and
quantitative determination of in
the urine, 207

f-Oxybutyric acid, method for the

uantitative determination of in

the urine, 211

528

Paranuclein, synthesis of through

the agency of pepsin, 493

Pepsin, question of the identity of
with chymosin, 399

Pepsin, synthesis of paranuclein
through the agency of, 493

Peptonization in raw and pasteur-
ized milk, 247

Peters, Amos W.: Studies on
enzymes: I. Adsorption of dias-
tase and catalase by colloidal
protein and by normal lead phos-
phate, 367

Phenyl derivatives of fatty acids,
mode of oxidation of in the ani-
mal organism, 173, 303

Phenylpropionic} acid, synthesis of
some derivatives of, 303

Potassium chloride, inhibiting ef-
fect of in sodium chloride glyco-
suria, 351.

Protamin,:.composition and deriva-
tion of, 389

Protamin, synthesis of through fer-
ment action, 381

Protein, adsorption of diastase and
catalase by, 367

Proteins, chemical mechanics of
the hydrolysis and synthesis of
through the agency of enzymes,

493

Proteins, hydrolysis of by trypsin,
31

Proteins, of milk, 261

Pyrimidins, researches on, 49, 163

Reagent for the detection of reduc-
ing sugars, 485

Reducing component of yeast nu-
cleic acid, 469

Reducing substances in the blood,
quantitative estimation of, 443

Reducing sugars, reagent for the
detection of, 485

Reid’s method for estimation of
reducing substances in the blood,
443

Reptiles, note on the chemistry of
muscle and liver of, 125

ROBERTSON T. BrartLtsForD: Note
on the applicability of the laws
of amphoteric electrolytes to se-
serum globulin, 155

ROBERTSON, T. BRAILSFORD: On
the influence of temperature upon
the solubility of casein in alka-
line solutions, 147

Index

ROBERTSON, T. BRAILSFORD: On
the synthesis of paranuclein
through the agency of pepsin and
the chemical mechanics of the
hydrolysis and synthesis of pro-
teins through the agency of en-
ZYMES, 493

RoBERTSON, T. BRAILSFORD and
C. L. A. Scumipt: On the part
played by the alkali in the hydrol-
ysis of proteins by trypsin, 31

Saiki, Tapasu: see Underhill and
Saiki, 225

Saliva, effect of diet on the maltose-
splitting power of, 331

SCHEELE, M. H.: see Neilson and
Scheele, 331

Schenck’s method for estimation of
reducing substances in the blood,

443

ScumipT, C. L. A., see Robertson
and Schmidt, 31

Serum globulin, applicability of
laws of amphoteric electrolytes
to, 155

SHAFFER, PuHitip A.: A method
for the quantitative determina-
tions of /-oxybutyric acid in the
urine, 211

SHERMAN, H. C.; see Colwell and
Sherman, 247

Skatol, occurrence of in the wood
of Celtis Reticulosa (Miquel), 489

Sodium chloride glycosuria, inhibit-
ing effect of potassium chloride

in, 351

Spiro’s method for estimation of
urea and ammonia, 477

Squalus Acanthias, globulin from
the egg-yolk of, 243

STEEL, MatTHEew: A study of the
influence of magnesium sulfate
on metabolism, 85

STEEL, MATTHEW, and WILLIAM J.
Gigs: Some notes on the effi-
ciency of the Folin method for
quantitative determination of the
urinary ammonia, 71

Sugar, conversion of glycogen into
in the liver, 315

Sugars, reducing, reagent for the
detection of, 485

Synthesis of paranuclein through
the agency of pepsin, 493

Synthesis of protamin through fer-
ment action, 381

Index

Synthesis of proteins through the
agency of enzymes, chemical me-
chanics of, 493

Synthesis of some derivatives of
phenylpropionic acid, 303

Synthesis of some nitrogen-alkyl
derivatives of cytosin, thymin
and uracil, 49

TayYLor, ALONZO ENGLEBERT:
Chemical studies in cytolysis, 311

Taytor, ALONZO ENGLEBERT: On
the antagonism of alcohol to car-
bolic acid, 319

TAYLOR, ALONZO ENGLEBERT: On
the composition and derivation of
protamin, 389

TayLor, ALONZO ENGLEBERT; On
the conversion of glycogen into
sugar in the liver, 315

TayLor, ALONZO ENGLEBERT: On
the inversion of cane sugar and
maltose by ferments, 405

TayLor, ALONZO ENGLEBERT: On
the question of the identity of
pepsin and chymosin, 399

TayLor, ALoNzo ENGLEBERT: On
the synthesis of protamin through
ferment action, 381

Temperature, influence of upon the
solubility of casein, 147

Thymin, action of diazobenzene sul-
furic acid on, 163

Thymin, nitrogen-alkyl derivatives
of, 49

529

Thyroidectomy, influence of upon
intermediary metabolism, 225

Thyroid feeding, influence of upon
intermediary metabolism, 225

Trypsin, hydrolysis of proteins by,
31

UNDERHILL, FRANK P. and TaDAsu
SaIki: The influence of complete
thyroidectomy and of thyroid
feeding upon certain phases of
intermediary metalbolism, 225

Uracil, action of diazobenzene sul-
furic acid on, 163

Uracil, nitrogen-alky]l derivatives of,
49

Urea, comparative tests of Spiro’s
and Folin’s methods for deter-
mination of, 477

Uric acid excretion in normal men,

355
Vicilin, hydrolysis of, 187

WELKER, WILLIAM H.: see Emer-
son and Welker, 339

WELLs, H. GipEon: The chemistry
of the liver in chloroform necro-
sis (delayed chloroform poison-

ing), 129

Yeast nucleic acid, reducing com-
ponent of, 469

eae 4). ip :

I vi di | ii
r i

z |
f
>
}
‘
a
|
—]
a
«©
|
ond,

QP The Journal of biological
501 chemistry

J77

wo

cop.2

Biological
& Medical
Serials

PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET

$e
UNIVERSITY OF TORONTO LIBRARY

aS Res:
See

pn
Neary
i
Siu
:

la
€ us

RRP
a
ure

eS EE
pattie idee ae
SSF: 4
=e

%
rts
kis

es

Fave

rages

Oy) carat
ff
i

Vt Pes eet
BN ay \

4! Sip ay
Wy hess te ‘
RA Rt
Ss 7 eal
Me part at
: Cs
\ i a) Plies ¥
it Boy i! E z

a Ni

Vale
Oa
Hay

VATA