be
Las

i

pa hn
Riad

Digitized by the Internet Archive
In 2009 with funding from
Ontario Council of University Libraries

http://www.archive.org/details/biochemicaljournO6bioc

THE

BIO-CHEMICAL JOURNAL

EDITED BY

BENJAMIN MOORE, M.A., D.Sc., F.R.S.

EDWARD WHITLEY, M.A.

VOLUME VI

£LQf2
(
po? ee
5 !
ae
COPYRIGHT

BIO-CHEMICAL DEPARTMENT, JOHNSTON LABORATORIES
UNIVERSITY OF LIVERPOOL

\ te ee Re .
HANNS oye
71 .
ce
AAV pe ATOONE 7?
Coe | |

ora

a Seoad sucuir ) Tle

CONTENTS OF VOLUME VI

The Carbohydrates of the foliage Leaff the Snowdrop (Galanthus nivalis, L.), and
their bearing on the first Sugar of Photosynthe ‘sis. By John Parkin, M.A., F.L.S. ;

On Deamidization. By Gertrude D. Bostock, B.Sc., M.B., Ch.B., Carnegie ea

Note on the Estimation of Potassium in Urine. By H. H. Green, B.Sc.

On the Freezing Point of the Unhaemolysed Corpuscles during Haemolysis of Blood,
and on the efiects of Evaporation on the Resistance of Erythrocytes to o paee By
U. N.. Brahmachari, M.A., M.D., Ph.D. é t ;

A Contribution to the Study of the Protein igetsbolintt of the Foetus The Dis-
tribution of Nitrogen in the Maternal Urine and in the Foetal Fluids throughout Pregnancy.
By Dorothy E. “Lindsay, B.Se., Carnegie Scholar ,

A Preliminary Study of the Bio-Chemical Relations of orang Lipoid Sh Ae
in the Liver. By Frederick P. Wilson, M.Sc., M.D., Beit Memorial Research Fellow

The Action of Selenium Salts on the Red Blood Sonus. By Charles O. Jones,
M.D., M.Sc., M.R.C.S. é

Direct ‘Measurement of the Praiotio See of Capein in Alkaline Salation.
Experimental Proof that apparent impermeability of a Membrane to Ions is not due to
the Properties of the Membrane but to the Colloid contained within the Membrane. By
Benjamin Moore, Herbert E. Roaf and T. Arthur Webster, University of Liverpool

The Sodium Phosphate Standards of Acidity. By E. B. R. Prideaux ;

Note on the Synonymy and Histological Characters of East London Boxwood
(Gonioma Kamassi, BE. Mey.). By R. J. Harvey Gibson, M.A., F.L.S., Professor of reer
in the University of Liverpool

Cryoscopic Determinations of the Osmotic Besee ‘of the Blood ina Body Fluids
of some Australian Animals. By Judah Leon Jonah, D.Sc., M.B., B.S. . :

On the Influence Exercised by certain Acids on the Inversion of Saccharose by Sucrase.
By Frederick Stoward, D.Sc. : : ; : i c d

The Physiological Action of Indolethylamine. By P. P. Laidlaw : :

The Effects of Caffeine upon the Germination and Growth of Seeds. By Fred
Ransom, M.D., Beit Research Fellow : : ; é

The Effects of Caffeine upon the Germination and Gide of Sesiy (Second com-
munication). By Fred Ransome, M.D., Beit Research Fellow - : :

Cell Stimulation by means of i ii Ingestion of Alkaline Salts. By The Jessup
Committee : : ; : :

_ The Proteins of Rice. By S. Kajiura : : ,

Some Further Experiments on the Physiology of the Allyl Compounds By E.
Wace Carlier, M.D., F.R.S.E. : :

The Influence of Protoplasmic Poisons on ‘ Radtcte By D. Tease Har ris, M. D.,
D.Se., Professor of Physiology and Histology in the Dalhousie University, Halifax, N. s.

Note on Hydrolysis of Vegetable Oils by Emulsion of Ricinus Communis. By ‘David
Sommerville, B.A., M.D., M. R. C.P., Lecturer in Public Health, King’s College, Univ ersity
of London :

A Clinical Method of Estimating as poe ‘of Calei ‘ium in the Sa and me
Physiological Fluids. By W. Blair Bell, B.S., M.D. poo Assistant Gynaecological
Surgeon, Royal Infirmary, Liverpool : :

Observations on the Secretion and Somes of nee Bile. By J. ra Menzies,
M.D., Edinburgh : : : : : : .

Notes on some New Substitutes for the Galenical Preparation of Digitalis, By
Douglas Cow, M.D. :

The Anti-neuritic Bases of vegans Origin in a Hbtationshis to Beri- Beri, with a
Method of Isolation of Torulin the Anti-Neuritic base of Yeast. = EK. 8S. ~ iW. H.
Evans, B. Moore, G. C. E. Simpson and T. A. Webster

The Creatin Content of Muscle in Malignant Disease and Se Pagialepical Condi-
tions. By R. A. Chisolm, M.A., D.M., M.R.C.P., Greville Research Student

On the Nature of Animal Lactase. By i Stephenson

\3

76

79

100

122

130

: CONTENTS

Hi

¢
The Nutrition and Metabolism of Marine Animals in Relationship to (a) Dissolved
Organic Matter and (b) Particulate Organic Matter of Sea-Water. By Benjamin Moore,
M. iV , D.Sc., F.R.S., Fdward S. Edie, M. EA. Edward Whitley, M.A., .and W. J. Dakin, D.Sc.

The Effects of Atmospheres enriched with Oxygen upon Living Organisms, (a) Effects
upon Micro-Organisms, (b) Effects upon Mammals experimentally inoculated with Tuber-
culosis, (c) Effects upon the Lungs of Animals, or Oxy gen Pneumonia. By Alfred

Adams, M.B., Ch.B. (Liverpool)

Observations on Electro-Osmosis. 7 J. O. W AMS sheet i a B. ce

The Phosphatides of the in ee — Hugh MacLean

On the Purification of Phosphatides. By Hugh MacLean : i

On the Relations of Phenol and Meta-Cresol to Proteins: A Contribution to our
knowledge of the Mechanism of Disinfection. By E. A. ae B.Se., Beit- Memorial
Fellow, Uter Institute of Preventive Medicine : : ¢ : :

The Distribution of Nitrogen in Autolysis, with er ett to Deaminization.
By Gertrude D. Bostock, B.Se., M.B., Ch. B., Carnegie Scholar

The Effects of Purine derivatives and other Organic Compounds on Growl an
Cell-Division in Plants. By N. G. 8. Coppin, M.Sc. : : ; : :

On Glycolysis in Blood. By G. Spencer Melvin, M.D.

The Reduction of Ferric Chloride by Surviving Organs. By David Frazer Harris,
M.D., D.Se., and Henry Jermain Maude Creighton, M. ine M.Se., D.Sc.

The Action of Opium Alkaloids. By F. W. Watkyn-Thomas, B.A., Trinity College

The Detection of Aceto-Acetic Acid by Sodium es Paar and Ammonia. By
Victor John Harding and Robert Fulford Ruttan

The Fatty Acids of Butter. By Ida Smedley, Beit Meio el Reseach Tele

PAGE

255

297

315
333

355

362

388

416
422
429
433

445
451

}
/

THE CARBOHYDRATES OF THE FOLIAGE LEAF OF
THE SNOWDROP (GALANTHUS NIVALIS, L.), AND THEIR
BEARING ON THE FIRST SUGAR OF PHOTOSYN-
THESIS'

By JOHN PARKIN, M.A., F.U.8S., Trinity College, Cambridge.
(Received May 30th, 1917)

CONTENTS
Introduction F i
Analytical Procedure i Ls 8 fe eH 4% 4
Experiments and Results’... a re Ae es ee 22
Summary of Results 39
General Discussion ze a co i Ss ee 10
INTRODUCTION

Since the publication of Brown and Morris’s important memoir
entitled ‘ A Contribution to the Chemistry and Physiology of Foliage
Leaves ’,? no extensive researches by chemical methods have apparently
been carried out on the carbohydrates, which arise in the green leaf as
the result of photosynthesis. With the exception of the fairly recent
work of Strakosch,? which will receive attention later, little has been done
either in the way of confirming or refuting the results obtained and the
conclusions drawn by Brown and Morris from their study of the sugars
of the foliage leaf.

The investigations about to be described were commenced with the
primary purpose of testing Brown and Morris’s novel view, namely, that
cane sugar is the first carbohydrate to be synthesised in carbon-
assimilation.

The plant used for their experiments was Tropaeolum majus, the
common nasturtium of our gardens. It forms starch plentifully in its
leaves. The carbohydrates they recognised and attempted to estimate
quantitively were five in number, viz.:—glucose (dextrose), fructose
(levulose), sucrose (cane sugar), maltose, and starch. From analyses
made of leaves picked at various times of the day and under diverse

1. A brief account of this work, as far as it had then progressed, was given before the

Botanical Section of the British Association, Dublin Meeting, 1908. (Sce, for abstract, the
Annual Report for that year, p. 907).

2. Brown and Morris, Journ. Chemical Soc., LXITII, p. 604, 1893.

3. Strakosch, ‘ Sitz. K. Akad. d. Wissen., Wien,’ Wath.-Nat. Cl., CXVI Bd., H. VI, Ab. I,
p. 855, 1907.

2 BIO-CHEMICAL JOURNAL

conditions, they came to the somewhat unexpected conclusion that suerose
is the first of the five carbohydrates to appear in photosynthesis. The
others they derive from it as follows :—The starch arises directly from the
cane sugar. They consider that when the latter exceeds a certain
concentration, the excess is condensed to starch, thus relieving the osmotic
pressure of the cell-sap. The glucose and fructose come also directly from
part of the sucrose through inversion. The maltose proceeds from the
starch through the dissolution of the latter by the action of the leaf-
diastase. Further, finding that in their analyses the fructose was
invariably in greater quantity than the glucose, they put forward the
view that the latter is more readily consumed in respiration. —

In the discussion,! which followed the reading of this paper before
the Chemical Society, the President of the year, Professor H. KE.
Armstrong, suggested in his comments the hkelihood of the cane sugar
arising directly from the maltose, thus opposing the view of its being the
first carbohydrate of photosynthesis. Such a criticism appeared then
justifiable, especially since Brown and Morris themselves had shown three
years earlier? that the barley embryo is capable of changing maltose into
sucrose. =

Irom the time of Sachs several plants, chiefly Monocotyledons, have
been known, which never under natural conditions form starch in their
green tissue (mesophyll). The idea struck the writer that such a type
of leaf would be a profitable one upon which to experiment on the lines of
Brown and Morris’s work. Without starch, maltose would most likely
be absent from such a leaf; consequently any sucrose found to be present
could not have this sugar for its origin. The analysis would also be
simplified, as in all probability there would only be three instead of four
sugars for estimation, viz.:—sucrose, glucose and fructose.

The Snowdrop (Galanthus nivalis) was chosen, partly as a pleasanter
subject with which to work than the Onion (Alliwm cepa), the best known
of the non-starch forming Monocotyledons, and partly because some years
previously? the writer had made an extensive microchemical study of the
distribution of starch and inulin in this plant. In no case was starch
ever to be recognised in the mesophyll of the leaf, nor could it be induced

1. Brown and Morris, loc. cit., p. 677.

9 y P. "10 ~ ., + Mo +
2. Brown and Morris, ‘Researches on the Germination of some of the Gramineae,’ Journ.

Chemical Soc., LVII, pp. 513-520, 1890.

3. Parkin, ‘Contributions to our knowledge of the Formation, Storage and Depletion
of Carbohydrates in Monocotyledons,’ Phil. Trans. Roy. Soc., London, Ser. B., Vol. CXCI,
pp. 40, 49, 53, 58, 61 and 66, 1899. P

14

CARBOHYDRATES OF FOLIAGE LEAF OF SNOWDROP 3

to form here by feeding with sugar solutions. My trials on this point did
not meet with the success obtained apparently by Boelhm,! who mentions
the Snowdrop, among others, as capable of producing starch in its
mesophyll, when the leaf is supplied with sugar solution.

Starch, however, occurs in the guard cells of the stomata, but the
quantity is too small to be taken into account in chemical analysis. The
bulb contains starch and inulin in abundance, but neither of these carbo-
hydrates ever appears, in the author’s experience, under natural con-
ditions in the mesophyll of the leaf.

Thus by using a plant lke the Snowdrop, the investigation of the
carbohydrates of the foliage leaf becomes simplified. There are no
products of starch dissolution to complicate the issue. The sugars found
to be present, viz. :—sucrose, glucose and fructose, must in all probability
be closely connected with the photosynthetic process. The main problem
seems to centre round the origin of the cane sugar in the leaf. Does its
formation precede that of the hexoses, or is the opposite the case? The
results set forth in this paper favour the former idea, viz., that sucrose
is the first sugar to appear in quantity in a free state in photosynthesis,
and that the two hexoses, glucose and fructose, arise from it by inversion.

This research has been in progress for the last five or six years. As
the leaf material can only be collected at one period of the year, chiefly
in the months of March and April, the work has necessarily been somewhat
protracted. The results obtained from the leaves collected one spring
naturally suggest the work to be carried out when more material is at
hand. This, however, cannot be obtained till the following spring, as
the Snowdrop is not a plant which can be induced to produce leaves at
any time of the year; and even if it could, it might not be advisable to use
leaves grown under artificial conditions.

Though the author is fully aware that this investigation is by no
means complete, other lines, both for purposes of verification and attack,
suggesting themselves, yet the research appears to have progressed
sufficiently far under present methods to warrant publication, and invite
criticism. Further repetition on the same lines seems hardly necessary.
Then again, opportunities for this class of practical work may cease in
the near future, so that it is well to publish the results and draw the con-
clusions, while the experimental details are fresh in the memory.

Kven though the results may be proved wrong in some respects, or

1. Boehm, * Ueber Starkebildung aus Zucker,’ Bot. Zeit, XLI, p. 49, 1883.

4 BIO-CHEMICAL JOURNAL

the interpretation put upon them unacceptable, yet I shall feel repaid if
they arouse interest in this somewhat neglected branch of physiological
botany.

Throughout the practical work I have had the valued and cheerful
assistance of my friend Lieut.-Col. Dixon, M.A., Trinity College,
Cambridge, to whom I wish to express my best thanks. As far as the
experimental part of this research goes, our names might well appear
jointly in the title, but he has not seen his way to consent to this, as he
considers himself not competent to judge of the botanical aspect of this
investigation, and thus could not hold himself responsible for the con-
clusions drawn or views expressed. i

My thanks are also specially due to Mr. J. H. Millar, B.Se., I’.1.C.,
for much helpful correspondence in regard to methods for sugar estima-
tions, particularly respecting those requiring the use of yeasts.

To Professor T. B. Wood, we are much indebted for kindly allowing
us the use of the saccharimeter belonging to his department, the
Cambridge School of Agriculture.

The analytical work has been carried out wholly in the Chemical
Laboratory of the University of Cambridge, and we desire to express our
thanks to the former and present Professors of this Department for the
privilege of working there, and for the opportunity it afforded us of
employing liquid air in some of our experiments.

ANALYTICAL PROCEDURE

It will be necessary to deal at considerable length with the details of
the general method which has been followed in this investigation for the
estimation of the sugars in the Snowdrop leaf. This branch of
physiological chemistry is as yet in the tentative stage. Every step
requires close study and checking, in order that some degree of accuracy
may be attained. No great advance, except with respect to the
introduction of discriminating yeasts, has been made since the
publication of Brown and Morris’s work.! The method they used, in the
light of work since carried out on pure sugar mixtures, is open to some
objections. Being aware of these sources of error we have endeavoured
either to overcome them or to allow for them, hoping that this research
may thereby gain in reliability, though we are quite prepared for the
possibility of further sources of error being brought to light in the
quantitative estimations of sugars in plant extracts.

1. Loe. cit., p. 663, 1893.

CARBOHYDRATES OF FOLIAGE LEAF OF SNOWDROP 5

Analytical work of this nature is at the best slow. Our object has
been to dispense with operations which we deemed to be unnecessary
refinements with respect to the degree of accuracy at present obtainable in
this line of investigation. At the same time our main purpose has been
to carry out the analyses of the leaf extracts as uniformly as possible, so
that the relative value of the results may be as great as possible.

In a research of this kind a large number of analyses with a moderate
degree of accuracy seems preferable to a few with many precautions. In
the latter case, on account of the small number of experiments, the effort
to eliminate small errors may allow large ones to be passed over un-
observed, whereas in the former, though slight variations may continually
be arising, yet, owing to the number of analyses, large errors can hardly
escape detection. Our conclusions are drawn from wide differences in
sugar contents. Small fluctuations do not affect the generalisations.
Ditterences in grams of sugar, and not fractions of a gram, per 100 grams
of dry leaf, form our basis for drawing conclusions.

Drying the Leaves. This research has been conducted in the main
with air-dried material. The leaves were desiccated quickly at a
temperature sufficiently low to prevent browning or discoloration. They
were then reduced to a powder for extraction.

In the process of drying, changes may take place in the sugars
present in the leaf, altering their relative proportions. The cane sugar,
for example, may undergo some inversion either through enzymic action
or the acidity of the cell-sap. The effect, if appreciable, will be to
increase the reducing sugars at the expense of the sucrose. This
possibility has been put to the test in the following way.

Experiments with Liquid Air. A batch of fresh leaves was divided
into two equal portions by weight. From one part the sugars were
extracted at once, and from the other after drying. In order to reduce
fresh leaves to a pulp for extraction, and at the same time to prevent
any enzymic action taking place, recourse was had to liquid air. The
leaves were immersed in this reagent, then withdrawn, and while frozen
and brittle, were quickly ground up and the whole mass at once thrown
into hot water to kill the enzymes, a few drops of ammonia being added
to neutralise any acidity which might have an invertive action. The
sugars so extracted were in much the same proportions, as in the control
extracts made from the dried leaves.

The above fears then appear somewhat groundless. The air-dried

6 BIO-CHEMICAL JOURNAL

material may be assumed to retain the sugars in somewhat the same pro-
portions as in the fresh leat.

Details of two such comparative analyses are given below. The
leaves were picked at two different periods of the spring—for the first
experiment on March llth, and for the second on April 30th. The
amounts of the sugars are calculated for 100 grams of dry leaf.

Fresh leaf. Dried ieaf.
(Ground up by means of Control
liquid air)
Experiment Speen ee ee de
I

Sucrose oF a sss 306 ise 12-84 10-46 12-74 10-42
Hexoses (reducing sugars) aap sah 5-94 12-87 5-67 12-38
Total sugar ... Bat ess mee 233 18-78 23:33 18-4] 22-8

Ratio of sucrose to hexose... eis ... 1: 0:46 Den 23 1: 0-44 1: 1-19

In both cases, instead of the proportion of hexose to sucrose being
greater in the air-dried (control) material than in the freshly extracted
leaf, it is slightly less.

The drying, then, has not resulted in any appreciable inversion of
the cane sugar.

It is to be noticed, also, that the total sugar extracted from the fresh
leaves in both instances is rather greater than from the dried leaves.
That is to say, the extraction is most likely more complete from the fresh
pulp; the relative proportions of the two classes of sugars, however,
remain very nearly the same.

For each separate analysis five grams of the leaf powder were taken,
corresponding to about 150 fresh Snowdrop leaves.1 As the analyses were
carried out in duplicate, double this number of leaves were required.
For a comparative experiment, such, for example, as the comparison of
early morning with late afternoon leaves, fully 600 were needed; and to
allow for a margin of safety and the possibility of having to repeat one or
more of the analyses, 1,000 leaves are not too many.

Thus a research of this kind entails the growing of a considerable
area of the plants and the manipulation of a large bulk of fresh material.

For the most part the Snowdrops were grown, and the leaves collected
and dried at my home near Carlisle, in Cumberland. Through the kind-

1. The number required not only varies with individual differences in the size—a minor
matter—but also with the season, as the leaves continue to grow throughout the spring by basal
growth. Fewer would be required, e.g., late in April than at the beginning of March to make up
5 grams of dry material.

CARBOHYDRATES OF FOLIAGE LEAF OF SNOWDROP t

ness of Mr. Lynch, the Curator, a plot of Snowdrops was grown in the
Cambridge Botanic Gardens for my use; but the bulbs do not seem to
grow so well there, so I had to rely chiefly on my Cumberland stock.
Colonel Dixon also collected and dried leaves for me at St. Bees, which
came in useful for preliminary and general purposes.

The Snowdrop leaf-powder is hygroscopic, like all organic air-dried
material. In order to obtain absolute dryness in a hygroscopic substance,
recourse must be had to desiccation in an atmosphere devoid of moisture.
This research hardly warrants the employment of this refinement, but in
order to have the powdered leaf samples of approximately the same degree
of dryness, exactly half a gram of each sample at the time it was being
used for analysis was taken and placed in a steam oven for six hours,
cooled in a desiccator, and then weighed. The equivalent of the loss of
weight (10-20 milligrams) suffered by the half gram is then to be deducted
from the amount of air-dried leaf powder taken for extraction.
Experiments on this point showed that after six hours’ heating in the
steam oven, most of the moisture had left the half gram of leaf powder,
i.e., the loss of weight on further heating was very small.

In future work of this kind the above operation might perhaps be
dispensed with, especially if the powdered leaf sample be kept for some
time in a desiccator before being weighed for extraction.

Extraction. Moderately strong alcohol is commonly used for dis-
solving out sugars from dried and ground up plant tissues. Brown and
Morris! in their investigation of the sugars of Tropaeolum leaves employed
80-85 per cent. alcohol.

The use of spirit for dissolving out the sugars necessitates the extrac-
tion of the leaf-powder previously with ether or some such solvent to
remove the chlorophyll. Then further, after obtaining the alcoholic
extract, the spirit has to be distilled off, in order to obtain the sugars in
aqueous solution for analysis. Hence this mode of sugar extraction is
somewhat tedious.

As no starch occurs in the Snowdrop leaf, the feasibility of directly
dissolving out the sugars from the leaf-powders by distilled water pre-
sented itself. Both cold and hot water were tried, and the former method
was finally adopted as the solvent. Thus the necessity of removing the
chlorophyll was dispensed with, as this is insoluble in water, and hence
the time taken over the extraction was considerably shortened.

1. Brown and Morris, p. 662, 1893.

8 BIO-CHEMICAL JOURNAL

The method of procedure was as follows : —The five grams of powdered
leaf taken were well shaken and allowed te soak some time in a flask with
50 c.c. of distilled water. The extract was then filtered into a 100 c.c.
flask. The residue was afterwards washed into the original vessel, well
shaken and again filtered into the 100 c.c. flask. This operation was
repeated twice more, but these filtrates were first concentrated
down on the water bath before being added to the measured flask, other-
wise the volume of the aqueous extract would have greatly exceeded
100 c.c. Thus the sample of leaf-powder taken was extracted four times
with cold water, and by this means most of the sugar was obtained in
solution. 'The aqueous extract thus procured is rich brown in colour, too
deep for reading in the polarimeter; and so to remove the pigment, as
well as any tannin or glucoside present, basic lead acetate solution had
to be added. Five cubic centimetres of this reagent were always used.
Upon filtering, a clear faintly yellow extract was obtained, containing
the sugars.

The removal of the sugars by means of a water Soxhlet apparatus
was tried, but abandoned in favour of cold water extraction, as less sugar
was dissolved out by the former method, and some of the sucrose appeared
to have undergone inversion, even though precautions were taken to
neutralise any acidity. The Soxhlet extraction was continued for 8-9
hours. Below are the results of two comparative experiments on different
leaf samples.

Total sugar in Ratio of

Mode of extraction Sucrose Hexoses 100 grams sucrose to
dry leaf hexoses
I. Cold water... is cee 9:29 16-35 25-64 1:1-8
Water Soxhlet... awe a 7:08 17-86 24-94 2b
II. Cold Water... Les 7 6-17 13-72 19-89 1: 2-2
Water Soxhlet... a ash 5-04 14-32 19-36 1: 2:8

Not only has the Soxhlet water extraction been less complete than
the cold water one, but also the sucrose is considerably less in amount,
with a corresponding increase in the reducing sugars; thus this somewhat
drastic method of extraction is to be avoided, as it conduces to sucrose
inversion.

To what extent are the sugars removed by the cold water method of
extraction? This point has been submitted to investigation. Though
the removal of the total sugar is not quite completed by the four successive
extractions, as Judged by the amount of cupric reduction, yet for the
purpose of this research it may be deemed near enough. The continua-

SS

CARBOHYDRATES OF FOLTAGE LEAF OF SNOWDROP 9

tion of the extraction beyond the fourth time would result in a little
more sugar and also greater accuracy, but an inconvenient volume of
liquid for concentration would be obtained, and the whole operation
would be difficult to carry out in a single day.

Details of two estimations on different leaf samples are given below,
showing separately in each successive extraction, after inversion, the
total amount of copper reduced.

Copper in grams. Copper in grams,
I IL I 5 i
Ist extraction .. 0-9815 0-7105 5th extraction wwe 0:0333 0-0342
2nd me we 0°7035 0-591L5 6th Be ... 00-0180 0-OL12
3rd Pe we 02800 0-2425 Tth or ... 00034 0-0014
4th eS we OOD: 0-097 8th a ... O-O014 0.0007
Total for the first ——. —_———. Total for the second four —_——— ————
four extractions 2-0703 1-6415 extractions 0-0561 0-0475

Thus the first two extractions remove two-thirds to three-quarters of
the total sugar. After the fourth extraction only about one-fortieth of
the sugar originally present remains in the residue, viz.: the amount
corresponding to about 0°05 grams of copper, i.e., roughly speaking,
0025 grams of hexose sugar. Hence we may consider that in the cold
water method of extraction adopted, about this amount of sugar is left
in the residue of the original 5 grams of leaf taken. Tor 100 grams this
will be equivalent to 0°5 gram of sugar, and so to be nearer the absolute
percentage of total sugars in 100 grams of dry leaf, about half a gram
should perhaps be added to the figures given in this paper.

Does any of the cane sugar undergo inversion during the cold water
extraction? As the leaf material was dried finally in a steam oven, it
might be expected that the enzymes present would all be destroyed and
that none of the sucrose would be inverted during extraction with cold
water, provided micro-organisms were prevented from growing. This,
however, we have found by experiment to be hardly the case. An extract
made with cold water when compared with a control carried out with
boiling water shows a slightly lower percentage of sucrose and a
correspondingly higher one of hexose. The longer the cold water
extraction takes, the greater the difference becomes. The following are
some of the results obtained on this point :—

10 BIO-CHEMICAL JOURNAL

Conp AnD Hort WaTER EXTRACTIONS COMPARED

Experiment Sucrose Hexose Total sagar Sucrose
difference

ee Cold*-... Stic “05 Sng ae 9-25 10-97
20-22 0-52

Hot sn ae ois oe atin PHT) 10-45

Mey Cold yy sc fac sine 500 ioc 15-72 9-04
24-76 0-35

Hot... a0 BOC Sot 500 15-37 9:39

COMPARISON OF USUAL COLD WATER EXTRACTION WITH ONES LEFT LONGER

Experiment Sucrose Hexose Total sugar Sucrose
difference
TI. Usual... Be wae dee ae 14:79 5-35 :
20-14 0:97
Allowed to soak extra 24 hours... 13:82 6°32
Il. Usual... ee big oh oe 15-30 10-68
7 25:98 1-12
Extra 24 hours Mie See aoe 14:18 11-80
III. Usual ... a se sbe as 4-53 9-45
13:98 0-84
Extra 24 hours 3:69 10-29
IW, (WEEN GG. ais 50 Sac oe 8:43 9-53 2-12. (or
17:96
Extra 3 days “ide 500 500 571 12:25 0-9 per 24 hrs.

The amount of inversion which has taken place through prolonged
soaking comes out very similarly in the above four experiments, viz.,
about 0°9 grams for twenty-four hours.

The diminution in the sucrose and corresponding increase in the
hexoses was equally observable whether the extraction was made with or
without the addition of an antiseptic, it seems, therefore, due to surviving
invertase in the dried leaf material. It is a recognised fact that enzymes
in a dry state can resist a higher temperature than when moist.

To prevent the above error arising, the obvious course would have
been the substitution of hot water extraction for cold, thus destroying any
active enzyme present. There is, unfortunately, a drawback to this
procedure. Once the leaf material is treated with hot water, then the
decoction ceases to filter easily, and the whole extraction becomes
difficult and tedious. The cold water method has, therefore, been adhered
to, though aware of the error arising through the inversion of a little of
the sucrose, which amounts approximately to 0°5 gram per 100 grams of
dry leaf. In the leaf analysis given, the cane sugar should be, therefore,
increased by about half a gram and the hexoses correspondingly
diminished.

CARBOHYDRATES OF FOLIAGE LEAF OF SNOWDROP 11

The effect of basic lead acetate. The cold water extract, though clear,
is of too deep a colour for its angle of rotation to be read in the polari-
meter, consequently a decolourising agent has to be employed. Both
animal charcoal and basic lead acetate remove the colour. The latter has
been generally used in our analyses, but experiments for comparison have
also been made with the former. Basic lead acetate has the advantage
over animal charcoal of removing tannin and allied matters which reduce
Fehling’s solution. | Now basic lead acetate is a somewhat disturbing
factor to introduce into a sugar mixture, but unfortunately it can hardly
be dispensed with, as no adequate substitute could be found. A number
of experiments have been made to test its influence on the sugar
estimations, and to find out the possible errors it introduces.

As already pointed out, 5 c.c. of the standard basic lead acetate
solution has, as a rule, been added to each aqueous extract made from the
five grams of leaf, the whole volume being then accurately made up to
100 c.c. By a series of tests it has been found that this quantity of the
reagent just suffices to remove all matter from the extract capable of
being precipitated by basic lead acetate. A further addition of the lead
solution causes no more precipitate to come down.

The extract is then filtered to remove the lead precipitate, and its
angle of rotation at once read in the polarimeter, and its cupric reducing
power estimated, without going to the trouble of removing any excess of
lead that may remain in solution. The removal of the lead by
sulphuretted hydrogen is a somewhat lengthy process, and the presence of
this slight amount has no appreciable effect on the rotation or the cupric
reduction, as the experiments described below will show.

Now it is well known that basic lead acetate affects the rotation of
fructose, forming a compound with it, which has a less rotation than this
sugar itself.

We have ascertained the effect on the rotation and cupric reduction
of adding 5 c.c. of our standard basic lead acetate solution to one per cent.
solutions of the three sugars—sucrose, glucose and fructose, respectively.
The only immediate appreciable effect is a lessening of the rotation in the

case of fructose and to some extent of glucose, as the following figures
show : —

Without With basic lead acetate
Rotation Copper in grams Rotation Copper
Glucose ~ + 3-10 1-828 + 2-47 1-822
Fructose 3 — 5:53 1-857 — 2-46 1-844

Sucrose Sef + 3-82 —_ + 3:98 —

12 BIO-CHEMICAL JOURNAL

A shght decrease in the cupric reducing power of the ‘ leaded” sugar
solutions (glucose and fructose) is also noticeable.

The next point to be investigated was to see if the addition of tannin
to the sugar solutions and its removal by basic lead acetate altered the
percentages of the carbohydrates. ‘Tannin was added to the one per cent.
solutions of the sugars just to the extent that it was all carried down in
the precipitate, caused by the addition of 5 c.c. of the standard basic lead
acetate solution. Any excess of lead was not removed. The rotation and
cupric reducing power were found to correspond very closely, so that
neither the lead precipitation nor the slight amount of lead salt left in
the solution altered the percentage composition of the separate sugar
solutions. Below is shown the close agreement in the figures :—

Tannin added and then

precipitated by 5 c.c. Control in distilled

basic lead acetate water

Sucrose... ... Rotation ... = ea + 3°76 + 3-717
Copper after inversion... 2-011 2-011

Glucose... ...Rotation ... er a + 2-74 + 2-75
Copper Age 5b ca 1-742 1-757

Fructose ... ..kotation ... ane acts — 4:75 — 4-75
Copper Ss one ee 1-709 1-737

A similar experiment carried out with a mixture of the three sugars
showed little difference between the one containing tannin and the
control. The mixture consisted approvimately of 0°5 grams of sucrose,
02 glucose, and 0°25 fructose—total 0°95 grams of mixed sugars in 100 e.e.
The results obtained are set forth below :—

Tannin added and precipitated

by 5 c.c. basie lead acetate Control in distilled water
Sucrose ae ee “a 0-487 0-489
Glucose a. a ana 0-204. 0-201
Fructose HEC ‘ie ae 0-244 0-244
‘Total sugar 0-935 0-934

Thus the two estimations are in very close agreement. The addition
of tannin and its removal from solution by the necessary amount of basic
lead solution has hardly affected the composition of the sugar mixture;
and further, the slight amount of lead left in the filtrate after the removal
of the tannin-lead precipitate has had no influence on the estimation.

It thus appears that if the basic lead acetate solution is not added in
great excess, but only in sufficient amount to precipitate tannin and
allied matters, it has no appreciable effect on the sugar estimations, even

CARBOHYDRATES OF FOLIAGE LEAF OF SNOWDROP 13

when no attempt is made to remove any lead remaining by sulphuretted
hydrogen. Let us see if this is borne out by comparing a leaf extract
which has been treated with sulphuretted hydrogen to remove the lead

with one which has not.

Excess of lead removed by Lead not removed
sulphuretted hydrogen
Sucrose ar if 7 10-29 10-56
Glucose +t er is 3°88 3:79
Fructose Re one ~~ 44 4-80
Total sugar from 100 grams Fr)
of dry leaf... ef ari 18-61 19-15

By the extra operations required through treatment with sulphuretted
hydrogen, the total sugar has undergone some reduction, but the relative
proportions of the sugars remain much the same.

A number of trials have been made to see if the addition of 5 c.c. of
the basic lead acetate solution per 100 c.c. of distilled water has any effect
on the slight spontaneous reduction shown by Fehling’s solution. By
careful manipulation to make the comparisons as parallel as possible, the
data show that the lead solution has a slightly lowering effect on the
cupric reduction, the mean difference being 05 milligram. This small
amount will be negligible in the sugar estimations of the leaf extracts,
as much less lead will be left in the saccharine solutions.

The aqueous leaf extract reduces Fehling’s reagent considerably
more before treatment with basic lead acetate than after.

The following table taken from four experiments demonstrates this
difference :—

AMOUNT OF COPPER REDUCED BY THE EXTRACTS MADE FROM 5 GRAMS OF FOUR DIFFERENT LEAF
'
SAMPLES

Before treatment with After treatment with Diiference

basic lead acetate 5 c.c. basic lead acetate
I po t; 1-374 1-205 0-169
WE Pee ve 0-926 0-718 0-208
ILI. Se ere 0-820 0-656 0-164
IV. Bas fin 0-734 0-509 0-225

The effect of basic lead acetate is, however, without influence on the
amount of sucrose estimated by Clerget’s acid method.

AMOUNT OF SUCROSE FOUND IN 5 GRAMS OF Dry LEAF

Without lead With lead Dilference
I. Bae es 0-629 0-628 0-001

lV, Re oS 0-576 0-569 0-007

14 BIO-CHEMICAL JOURNAL

The sucrose in the ‘ leaded’ extracts is in both cases slightly less, due
perhaps to a slight retention of sugar by the lead precipitate.

The analysis of the extract after filtering off the basic lead acetate
precipitate. Yor the estimation of the three sugars—sucrose, glucose and
fructose, in the leaf extracts, the cupric-reducing power and angle of
rotation before and after inversion have provided the data, checked and
supplemented by yeast methods.

Fehling’s reagent has been employed throughout for the copper
estimations under the uniform conditions laid down by Brown and Morris.'
The cuprous oxide produced by the sugar was collected in a Soxhlet tube
plugged with asbestos, and reduced to copper by heating in a current of
hydrogen, and weighed as such. These estimations were always made in
duplicate, and a maximum difference of two milligrams allowed. With
extra care the duplicates can without great difficulty be brought to agree
within the milligram, but for the sake of speed, as well as for the general
degree of accuracy of this work, a difference of two milligrams seemed
permissible.

For the calculation of the amount of the three sugars from the weight
of copper reduced, the table prepared by Brown and Morris has been
employed.2 A different factor has thus been used for fructose than for
clucose.

lor the estimation of the sucrose, inversion by means of hydrochloric
acid under Clerget conditions? has been the method followed, being
quicker than by using the invertase of yeast. These two ways were,
however, compared, and found to agree closely.

Clerget Yeast

I. Rotation... sere ae 40% aiate 2-28 P33
Copper ... Soe ae ae or 1-882 1-879

Il. Rotation... Bae ae Ne ALE 2°16 2:1
III. Rotation... roe er oe 50 1-04 1-06
Copper ... se Sb ant 506 U9 1-69

Thus the two methods give practically the same data, and it may be
assumed that treatment with hydrochloric acid under Clerget conditions
results merely in the inversion of the cane sugar, and further, that the
fructose is unaffected by the hydrolysis. Therefore the loss of rotation

1. Brown, Morris and Miller, Journ. Chem. Soc., LX XI, pp. 95 and 278, 1897.
2. Brown, Morris and Miller, ibid., p. 281.

3. ‘To 50 c.c, of the sugar solution 5 c.c, of concentrated hydrochloric acid are added,
and the whole heated on the water bath to 68° C. for fifteen minutes.

CARBOHYDRATES OF FOLIAGE LEAF OF SNOWDROP 15

and the inerease in cupric reduction can be considered as being wholly
due to the inversion of the sucrose.

The analyses have been conducted on the assumption that the only
sugars present in the leaf extracts are sucrose, glucose and fructose. The
data obtained have invariably worked out correctly on this assumption,
which could hardly have been the case if appreciable quantities of other
sugars or soluble carbohydrates or additional substances capable of
reducing Fehling’s reagent or of rotating the plane of polarised light
were present. But, naturally, we have not been content without
additional investigations supporting this assumption.

Experiments with Yeast. A series of experiments with ordinary
brewer's yeast have been carried out on the aqueous extracts of the Snow-
drop leaves.

On submitting an extract, which has been precipitated with basic
lead acetate and which has had the excess of lead removed, to yeast
fermentation, the rotation and cupric-reducing power become negligible -
quantities. Hence it is to be inferred that these two properties possessed
by the extract before fermentation are due to the presence of fermentable
sugars only. No extraneous bodies capable of rotating the plane of
polarised light or of reducing Fehling’s solution to any appreciable
extent appear to be present. We have, therefore, to deal solely with
plastic sugars and presumably glucose, fructose and sucrose.

An estimation of the different sugars in a mixed solution can be made
by taking the specific gravity and opticity before and after fermentation,
coupled with inversion by means of invertase. Such yeast methods can
be employed solely for sugar estimations, but the extracts require to have
a total strength of sugar at least five times that possessed by the aqueous
solution made from 5 grams of dry Snowdrop leaf. Hence for continuous
work on these lines a much greater quantity of material would be required
than was at our disposal. The yeast method, however, has been used as a
check on the copper one, and has given concordant results, thus
confirming the view that the three sugars—glucose, fructose and sucrose—
are alone present in the extracts.

Erperiment I,—
Total percentage of sugar in 100 grams
of dry leaf
By yeast and specific gravity sen we 21-52
By copper method i re Nae op 21-71

16 BIO-CHEMICAL JOURNAL
Experiment [1,— :
Yeast method decolourised Copper method
by animal charcoal
Sucrose... vit He ne as 11-4 12-2
Glucose Weis sais ae he Se 3-33 3-33
Fructose... he wes ae 68 4-67 4:0
Total sugar in 100 grams dry leaf ane 19-4 19-53

Results obtained after decoloration with animal charcoal. As already
mentioned the aqueous extract of Snowdrop leaves is of too dark a colour
to allow its angle of rotation to be read at all clearly in the saccharimeter.
Basic lead acetate, in addition to precipitating tannin matters, removes
the colour, Besides this reagent animal charcoal will also abstract the
pigment. On boiling with this substance nearly all the colour is absorbed,
leaving the solution almost colourless.

Unfortunately animal charcoal also removes some of the sugar. Its
absorbent capacity for saccharine matter has first to be satisfied before it
can be used in quantitative work. Hence it is a tedious substance to
employ in estimations.

We have tested the absorptive capacity of animal charcoal for the
three sugars, glucose, fructose and sucrose, separately, using the charcoal
at the rate of 10 grams for every 100 c.c. of the one per cent. sugar

solutions.
The sugar solution was first boiled with the requisite amount of
charcoal, filtered, and its rotation taken when cool. Another quantity

of the sugar solution was then boiled with the same charcoal, filtered and
rotated, and the operation was repeated twice more, with the same
charcoal.

The angle of rotation of each sugar solution after the first treatment
was cousiderably less than before, showing that some sugar had been
absorbed by the animal charcoal.

Rotation of sugar solution after first treatment

previous to treatment with animal charcoal
Glucose... re ae see i + 2°84 + 2:36
Fructose... =a ae ie 5 — 4:88 — 4:01
Sucrose wie ee ee ae aCe + 3:8 + 2-59

After the second treatment the chareoal is about satisfied as regards
the two hexose sugars respectively, but not quite for the sucrose. This
latter sugar seems to be absorbed somewhat more greedily than the former
two. Thus in abstracting colour from a solution of a mixture of these
three sugars, the animal charcoal must be boiled successively with two

—

CARBOHYDRATES OF FOLIAGE LEAF OF SNOWDROP 17

separate portions of the sugar solution, and these must be rejected. The
third portion after decolourising with the same charcoal may be regarded
as unaltered in respect to its sugar percentages.

This is borne out by our experiments with the Snowdrop leaf extract
—the same charcoal, 5 grams being used consecutively with six different

portions (50 c.c. each) of the original leaf extract.

Grams of copper reduced per 100 c.c. totation
Before Treatment ae Sg 0-883 too dark to read
Ist * We he not taken + 0:88
2nd ue ae ee 0-753 + 1-05
3rd $3 ie wes 0-865 + 1:5
4th 4 Ay. eee 0-883 + 1-45
5th “ ers ee 0-900 + 1:55
6th 3 ee Je 0-879 + 1-34

The rotation is practically constant after the third treatment, and the
cupric reduction nearly so too. An interesting point also is the fact that
the amount of copper reduced after decoloration (provided the animal
charcoal has previously been satisfied as regards its capacity for absorbing
sugar) is the same as that in the original coloured extract, showing that
no substances which reduce Fehling’s reagent have been abstracted with
the colouring matter. As regards the opticity it is impossible to state, as
before decoloration the rotation could not be read with any accuracy.

Analysis of the Extract after decoloration with animal charcoal

In the first place, as might be expected, estimations of cane sugar
in the original coloured extract and in a portion of the same extract
decolourised by ‘satisfied’ charcoal agree closely, as revealed by the
increase in cupric reducing power brought about either by Clerget’s
method of inversion or by invertase. Also the quantity of sucrose agrees
fairly closely with the amount found in the extract treated in the usual
way with basic lead acetate and hydrolysed by Clerget’s method.

Coloured extract gave wee 0-567 grams sucrose

Decolourised extract gave ... “0-574 ns From 5 grams

Treated with basic lead acetate in the. dry leaf
usual way gave 0-555 A

Some of the decolourised extract was treated with yeast, and after
fermentation its cupric reduction taken. The weight of copper obtained
amounted to 0'126 grams per 100 c.c. of extract, i.e., per 5 grams of dry
leaf. Before fermentation this came to 0'883 grams of copper. The

18 BIO-CHEMICAL JOURNAL

difference, 0-757 grams, represents the weight of copper due to fermentable
sugars.

Not only has the decolourised extract after fermentation still
considerable cupric reducing power as shown above, but also a shght
levorotation. Particular attention has been paid to this latter point, and
there seems little doubt that the leaf extract, when untreated with basic
lead acetate, has after fermentation a slight left-handed rotation, due
probably to the substance or substances which catse the residual cupric
reduction. An experiment carried out very carefully afforded the
following result :—

Before fermentation After fermentation
Rotation in the 200 mm. tube 5th “+ 1:3 alee — 0:2

>

Thus we have gleaned the following experimental facts : —

Extract decolourised by Extract treated with the
‘ satisfied ’ animal charcoal requisite amount of the basic
lead acetate solution

After yeast fermentation ... Some cupric reduction No cupric reduction
Slight levo-rotation No rotation

Hence before fermentation in the case of decoloration with animal
charcoal a small part of the cupric reduction is due to other causes than
fermentable sugars, and the rotation is in a small measure likewise
influenced. Whereas in the case of colour-removal by basic lead acetate
all the cupric reduction and rotation may be taken as arising from fer-
mentable sugars alone (presumably from glucose, fructose and sucrose).

When the fermented ‘ unleaded’ extract is boiled for some time with
a little mineral acid, the cupric reduction is clearly raised, indicating
the presence of a glucoside. Further investigation on this point has not,

however, been pursued.

The cupric reducing power of sucrose in the presence of hexoses

Sucrose alone has no reducing power on Fehling’s reagent, but if a
little hexose sugar be present, then it does reduce slightly. The cuprous
oxide produced is greater than can be accounted for by the hexose solely.
This, then, is a disturbing influence which must be taken into account
when using Fehling for the estimation of sugar mixtures containing
sucrose.

We have tested various mixtures of the three sugars, glucose, fructose
and sucrose, to see what influence the various proportions of cane sugar
may have on the cupric reducing power, and we can confirm previous

CARBOHYDRATES OF FOLIAGE LEAF OF SNOWDROP 19

work in finding that the increase due to sucrose is very small, unless it
be present in a preponderating ratio. Our leaf-analyses strengthen this
conclusion also, since the amount of sucrose calculated from the increase
of copper due to inversion was found to differ little from the sucrose
obtained from the loss of rotation after inversion. Our results, however,
show with very few exceptions, that the cane sugar calculated from the
loss of rotation is slightly greater than that obtained from the increase
in copper

a result to be expected.

The error arising from the slight cupric reducing power of the
sucrose can be easily avoided by first estimating the sucrose from the loss
of rotation through hydrolysis. The amount of copper reduced by this
quantity of sucrose on inversion can then be calculated and deducted from
the total copper obtained after hydrolysis, the remainder will represent
the copper due to the glucose and fructose jointly. The rotation exerted
by the sucrose present in the extract before hydrolysis, can be calculated
and deducted from the original rotation of the extract. Two equations can
then be formed, one from the copper and the other from the opticity.
By solving these the separate amounts of glucose and fructose can be
ascertained. :

Below is given, by way of illustration, a typical example of how the
percentages of sugars in a leaf-analysis are calculated.

Data obtained from an examination of the extract prepared from
5 grams of dried leaf.

(1) Angle of rotation before inversion in 200 mm. tube at 19° C.

=+1-37 divisions.

(2) Angle of rotation after inversion in 200 mm. tube at 19° C.

= —1-84 divisions.

(3) Total copper reduced after inversion =2-302 grams.

From (1) and (2) we obtain the total fall in the rotation due to the
inversion of the sucrose. This comes to —3-21 divisions, which is
equivalent to 0°6331 grams of sucrose. [In the polarimeter used it was
calculated that a 1 per cent. solution of this sugar loses —5-07 divisions
at 19° C.]

Thus we find that 5 grams of this leaf sample contain 0°6331 grams
of cane sugar.

Further data can now be obtained for calculating the respective
amounts of glucose and fructose.

(i.) 0-6331 grams of sucrose will give a rotation of + 2-44 divisions.

20 BIO-CHEMICAL JOURNAL

Therefore the rotation due to glucose and fructose jointly originally
present in the extract = + 1.37 —2-44= — 1°07 divisions.

(ii.) 0°6331 grams of sucrose when inverted will reduce 1°283 grams
of copper.

Therefore the cupric reduction due to the glucose and fructose =2-302
—1-283=1'019 grams of copper.

From (i.) and (ii.) two equations can be framed, the solution of which
will give the separate quantities of glucose and fructose originally
present.

Let « = glucose.

y = fructose.
Then (1) 3:°032 — 5'34y = — 1:07 (from the rotation).
(2) 2:0llz + 1:842y =1°019 (from the cupric reduction).

The constants used in the first equation are the angles of rotation
given respectively by one per cent. solutions of these sugars in the 200 mm.
tube at the temperature observed.

Those used in the second are the grams of copper reduced by one gram
respectively of these sugars under the conditions observed, the constants
being calculated from Brown, Morris and Miller’s Table!.

From the two equations we obtain respectively 0-2128 and 0-3209
grams of glucose and fructose.

Therefore we arrive at the following analysis of the sugar-mixture
in the 5 grams of leaf taken—

Sucrose
Glucose ...
Fructose

0-6331 grams,
O-21285;
0:3209 __,,

tl

Total sugar ane Soc 000 i 16G3iee

Multiply by twenty and we get the quantities in 100 grams of this
sample of dried leaf, the manner in which results are stated in this

paper.

Sucrose 12-662 grams.

Glucose .. a, 4-256 ,,
Fructose = 6-418 ,,
Total sugar = PBYBBO op

The duphecate analysis gave the following percentages :

Sucrose wide ate an = 12-618 grams.
Glucose ... see BG HOG = 4-498 PS
Fructose a6 ae cae = 6-508 =)

23-624 5

Total sugar

The mean of these two analyses gives the final result.
1. Journ. Chem. Soc., LXXI, p. 281, 1897.

CARBOHYDRATES OF FOLIAGE LEAF OF SNOWDROP 21

Evidence from the Osazone. The aqueous extract of the Snowdrop
leaf on heating with phenyl hydrazine and acetic acid produces an
abundance of yellow crystals characteristic of glucosazone. A further
deposit of crystals suggesting those of maltosazone were never obtained
on further heating and allowing the liquid afterwards to stand for some

time.

The whole of the osazone resulting from representative samples of
the leaf material was put together and the melting point and percentage
of nitrogen of the mixture ascertained. The melting point was taken after
re-crystallising the osazone from pyridine according to the method
described by Tutin!. This came out between 216-217", the same figure
practically as he obtained for glucosazone.

The percentage of nitrogen found in this mixed sample of osazone
was 15-7, which agrees very closely with the theoretical number, 15-64,
required for glucosazone. If an appreciable quantity of maltosazone had
been present in it, then the nitrogen would have been lower, for it
contains only 10°77 per cent. of this element.

No evidence of maltose could therefore be obtained by this means,
nor of other sugars such as mannose or galactose. The osazone resulting
is in agreement with what might be expected from an extract containing
only the sugars glucose, fructose and sucrose.

A Summary of Errors likely to occur in the Leaf-analyses

(1) The total sugar estimated is probably a little low—approximately
about 0°5 per cent—owing to incomplete extraction.

(2) The figures for the cane sugar err rather on the low side on
account of some slight inversion during the process of extrac-
tion. The amounts found in the various analyses of the leaf
should perhaps be raised by nearly 0°5 per cent. and a propor-
tional quantity deducted from the total hexose sugar.

(3) The fructose is probably somewhat underestimated owing to the
action of the basic lead acetate on this sugar, though precau-
tions were taken to avoid this as far as possible.

Fortunately the errors likely to arise in the estimations would not
favour but rather oppose the main conclusions arrived at in this paper.

1, Proc. Chem. Soc., Vol. XXIII, p. 250, 1907.

99 BIO-CHEMICAL JOURNAL

EXPERIMENTS AND RESULTS

Having dealt at some length with the methods employed for the
estimation of the sugars in the foliage leaf of the Snowdrop, we now turn
to their application and to the results thereby gained. These are set forth
under six headings. Under each are given the character of the experiment
and the results of the analyses, concluding with remarks.

I. The Quantity of Sugar in the Leaf is Considerable.

Twenty to thirty per cent. of the weight of the dried leaf consists of
sugar. The amount varies naturally according to the time of day the
leaves are picked for analysis. Allowing 80 per cent. for moisture, an
average quantity, the fresh leaf will therefore contain 4 to 6 per cent. of
mixed sugars.

Brown and Morris’s figures! for sugars in the leaves of Tropaeolum
vary from about 10 to 14 per cent. in the dry material, just about half the
quantity. Even on adding the starch, which is absent, as we have seen,
in the Snowdrop leaf, the total carbohydrate-content of the Tropaeolum
leaf is considerably less than that of the Snowdrop, being in round
numbers 14 to 18 per cent.

It is fortunate that the Snowdrop leaf is so rich in sugar, as it has
allowed us to select five instead of ten grams (the amount taken by Brown
a great saving of material and so of

and Morris) for each analysis
labour. ;
The abundance of sugar may partly be explained by the fact that the
Snowdrop is a spring plant, performing its assimilation when nights are
cold, which will retard translocation and thus tend to carbohydrate
accumulation in the leaf. The ratio, therefore, of the rate of
photosynthesis to that of translocation is probably greater in a plant like
the Snowdrop than in a summer-growing one like Tropaeolum; and so
the carbohydrate formed in assimilation will tend to accumulate in the
leaf more in the one than in the other.

Further, a bulbous plant like the Snowdrop has little available tissue
other than the mesophyll for temporarily storing the carbohydrate
produced in carbon assimilation, hence its leaf will be apt to contain a
higher percentage than that found in a plant with a considerable extent
of stem.

Whatever reasons may be adduced, the Snowdrop leaf contains a
surprising amount of sugar under ordinary conditions.

1. Brown and Morris, loc. cit., pp. 669 and 671.

CARBOHYDRATES OF FOLIAGE LEAF OF SNOWDROP 23

Il. The amount of Sugar increases from above downwards in a Single
Leaf, and at the same time the ratio of the Sucrose to the Hexoses
diminishes.

By comparing the amount of sugar in the upper green parts of the
leaves with that in the lower green parts, it soon became evident that care
must be taken to pluck the same relative length of leaf, in order that the
results may be strictly comparable.

The foliage leaves of bulbous Monocotyledons commence to assimilate
in their apical portions long before they attain their full length. As soon
as the uppermost part of the Snowdrop leaf has emerged from the basal
protective sheath, it is in a state of maturity and doubtless capable of full
functional activity provided the external conditions are favourable. The
upper green assimilating part of the leaf is continually being added to by
basal growth. From personal observation the leaves of the Snowdrop
continue to grow in length till nearly the close of their functional period,
the increment of growth ina given time gradually decreasing as the spring
advances. The increase in length ceases about the third week in April,!
and early in May the leaves begin to turn yellow and then gradually die
away. Thus most of the carbon assimilation in the case of the Snowdrop
occurs in the months of March and April.

The Snowdrop plant is a simple or rather simplified structure.
Disregarding the floral part, it consists of an outer sheath enclosing two
foliage leaves, joined to a short axis bearing below a tangle of rootlets.
The upper part of the sheath remains membraneous and is protective in
function; the lower part acts as a storage organ and forms the outermost
fleshy scale of the ‘resting’ bulb. Lach foliage leaf is sharply divided
morphologically into two regions, a basal sheathing part functioning as
an organ of storage and an upper assimilating and conducting part—the
foliage leaf proper. After the plant has died down, the basal parts of the
leaves persist as the middle and inner fleshy scales of the bulb. The leaf
proper is further physiologically divided into two regions, an upper and a
lower, through the embrace of the external sheath. The upper mem-
braneous part of this covers the lower part of the foliage leaf and prevents
this portion turning green. The leaf thus consists of an upper green part
and a lower colourless part, in addition to its sheathing base (the
immature bulb scale). The accompanying diagrams will help to explain
the morphology of the plant, as it bears upon this research.

1. From observations made in the vicinity “of Carlisle.

24 BIO-CHEMICAL JOURNAL

Diagram To EXPLAIN THE MorPHOLOGY OF THE SNOWDROP PLANT
AS IT AFFECTS THIS PAPER.

Fig. 1. A single plant (somewhat reduced in size) :—/f.l., the two foliage leaves (shaded).
sh. protective sheath covering the lower parts (‘‘ stalks ’’) of the foliage leaves. b., bulb.
r., rootlets. fl.s., flower stalk terminating in the solitary flower,
Fic. 2. A complete leaf detached from the plant :--The upper shaded portion (f) represents the
green part (the foliage leaf proper) : the middle narrow unshaded portion (s) the colourless
stalk part, which in the intact plant is covered by the protective sheath : the basal portion
(b) one of the new (immature) bulb-scales, which gradually become filled with starch and
inulin, elaborated from the sugar formed in the chlorophyll cells. The line, U—L,
arbitrarily separates the foliar part into an upper and lower portion. Leaves were so

divided for the analyses in this section of the paper. Otherwise, as a rule, the whole of
the green part of the leaf was taken for analytical purposes.

CARBOHYDRATES OF FOLIAGE LEAF OF SNOWDROP

25

The sugars, therefore, formed in the upper green part of the leaf
through photosynthesis, will have to travel through the colourless region
enclosed by the protective sheath, then across the meristematic zone into
the young bulb-scale, to be transformed there into starch and inulin—the
two reserve carbohydrates of the ‘ resting’ bulb.

The colourless part of the leaf enclosed by the membraneous, external
covering is very rich in sugar—30 to 40 per cent. of its dry weight
consisting of such. It evidently functions as a temporary storing place
for the sugars formed above in the green part through photosynthesis.
The sugar is most likely produced in the green tissue at a quicker rate
than it can be stored away as starch and inulin in the cells of the growing
bulb-scale, and so accumulates in the non-functional, colourless cells.

In this paper the sugar contents of this lower colourless part of the
leaf will not be considered further. Attention will be focussed only on the
sugars of the green part, that is the portion of the leaf which appears
above the protective sheath.

The leaves for this series of experiments were first cut from the
plants just above the protective sheath, and then each one was divided
either at its middle or two-thirds of the way up or otherwise, according
to the analytical comparison desired.

The sugar was always more abundant in the lower than the upper
regions. The difference, however, was by no means uniform, and seemed
to depend on whether the plants used for the experiment were growing
in close clumps or separately.
former than in the latter.

The following are the details of these experiments : —

The difference was more marked in the

GROWING IN CLOSE CLUMPS

Portion of leaf taken Total Ratio of sucrose Date and time
sugar to hexose of picking
Experiment I. Upper two-thirds 25-95 1:0-6 Mar. 30, 1905
Lower ie 28-72 1:11 5.30 p.m.
Experiment II. Upper two-thirds 21-98 1:1-05 Mar. 31, 1905
Lower of 26-48 1:1-38 7.30 a.m.
Experiment III. Upper halves 22-48 131-39 Mar. 31, 1905
Lower 53 30-59 1:4-0 3.30 p.m.
GROWING SEPARATELY
Experiment I. Upper halves 22-93 1:0-57 April 11, 1906
Lower 9 24-46 1:0-6 7 p.m.
Experiment If. Upper halves 19-26 1:0-5 April 12, 1907
Lower = 21-40 1:0-75 noon
Experiment III. Upper halves 24-03 1:0-65 Mar. 30, 1910
Lower e 24-36 1:0-82 2.30 p.m.

BIO-CHEMICAL JOURNAL

bo
mS

In the above six experiments the total sugar, which is calculated for
100 grams of dry leaf material, is invariably greater in the lower part of
the foliage leaves as compared with the upper; but this difference is very
much greater in the leaves taken from plants growing thickly together than
in those growing separately. In the one case the mean difference of the
three analyses works out to 513 grams, and in the other to 1-33 grams.

Secondly, it is to be noted that in every case the proportion of hexose
sugar (glucose and fructose) to sucrose is greater in the lower portions
of the leaves; and similarly this difference in the ratio is much more
marked in the leaves taken from close clumps than in those from plants
growing separately, the mean differences being respectively 1-15 and 0°15.

The leaves of Snowdrops growing in clumps tend to shade one another.
The upper portions may be fairly well illuminated, but the lower green
parts of many will be much in the shade, and so capable of performing
little assimilation. It will be otherwise in plants growing separately.
The light will be able to reach in the main the whole length of the green
part of the leaf, and so the full extent will share in the assimilation. The
results of the analyses rather suggest that if the leaves were similarly
illuminated for their whole length, the disparity in sugar contents
between the upper and lower parts would disappear. Special care was
taken in No. III of the second series of experiments to gather only leaves
well exposed to the light for their full length, and the figures show a very
close approximation between the two different halves. Further research.
on the point would seem desirable.

It looks as if the lower green part of the leaf when obscured functions
somewhat as a storage-tissue accumulating a larger amount of sugar than
it would otherwise have done if exposed to good hght and assimilating
on its own account. An actively assimilating chlorophyll cell perhaps
resists a high percentage of sugar arising within it. The act of assimi-
lating may be antagonistic to that of storing.

For most of the investigations, therefore, the full length of the green
leaf has been taken from plants grown in special plots. The bulbs were
placed about two inches apart, so as to keep the individual leaves from
shading each other for the most part; consequently the whole length of
the green part of the leaf was fairly well exposed to the light.

CARBOHYDRATES OF FOLIAGE LEAF OF SNOWDROP 27

Ill. The proportion of Sucrose to Hevose Sugar decreases as the

season advances.

If, for example, a batch of leaves be taken about the beginning of
March and compared with a similar quantity plucked at the commence-
ment of April, then, judging from the results here obtained, it will be
found that the proportion of sucrose to the hexoses (glucose and fructose)
will be considerably greater in the first set of leaves than it is in the
second. That is to say, there is in the dried leaf material more cane
sugar in proportion to reducing sugar during the early part of the season
than later. Thus it seems that as the spring advances the hexose sugars
increase at the expense of the sucrose. A comparison, of course, must be
made between leaves picked about the same period of the day. Leaves
gathered in the morning must not, for instance, be compared with those
taken in the afternoon; for if this be done, then the effect of
photosynthesis will mask the seasonal change occurring in the ratio of
these two classes of sugars.

The following analyses supply the data upon which the above
deduction is based :—

Maximum shade In 100 grams. dry leaf.
Date and time of plucking temperature, in Sucrose, Hexose,
degrees Fahrenheit grams. grams.
February 16, 1906, 3 p.m. ... Sct 48-9 19-8 3°56
February 26, 1907, 4—5 p.m. Ste 45 15-07 2-53
March 7, 1906, 3.30—4 p.m. Sue 66-9 14-55 5:69
March 30, 1905,5—6 p.m. ... ee 49-6 15-5 11-4
April 5, 1906, 4—4.30 p.m. ... BH 60 14-64 HEI 7
April 5, 1907, 4—4.30 p.m. ... mee 58 14-64 11-61
April 24, 1905, 4—4.30 p.m. or 51 14-84 17-29
May 4, 1905, 3—3.30 p.m. ... ee 53 10°3 12-78

From the above table the following approximate ratios are obtained
for the two classes of sugars. The sucrose in each case is represented by
unity and the hexose calculated in proportion to this.

1905 1906 1907
February 16th ae ie aa -— 1:0-2
February 26th Scr of a -- = 1:0-2
March 7th “fc EE ae aes — L:0-4 -—
March 30th... ae Ke ee 1:0-7 — —
April 4th xe nt vee eB = 1:0-8 1.0-8
April 24th wi ok ie ie 1:1-2 -- --
May 4th a os sae exe 1:1-2 -- —

It is thus seen that the proportion of the two hexose sugars in the
foliage leaf gradually rises from February to May, when the leaves begin

Pe

28 BIO-CHEMICAL JOURNAL

to change and wither. With respect to leaves taken in the late afternoon,
in February the hexose only represents a sixth of the total sugar; in early
March rather less than a third; in late March and the beginning of April
nearly a half; and towards the end of the assimilating season more than
a half.

Another striking instance of this feature in the ratios of the two
classes of sugars is afforded by a couple of analyses carried out in the
spring of last year. The leaves for this comparison were picked in the
morning about 10 o’clock—one set on March 11th and the other on

April 30th, thus seven weeks apart.

10 a.m., March 11th ; 10 a.m., April 30th
Sucrose a0 5o¢ a0 12-74 grams. 10-42 grams.
CXC Beam Pci, ans 567 12:38) ee
Ratio 1:0-44 1:1-19

Thus early in the season the quantity of cane sugar present in the
leaves was more than double that of the reducing sugars; whereas seven
weeks later the two types of sugars approximate closely in quantity.

These two analyses are especially interesting, as the leaves were
primarily collected for another purpose, viz., the comparison of the sugars
in air-dried leaves with those killed by immersion in liquid air. On
glancing back to page 6 it will be seen that both the quantities and
ratios of the two kinds of sugars, sucrose and hexose, were nearly the same
in the two parallel cases. The liquid air method of preparing the extracts
for analysis was so carried out that there was no time for any appreciable
invertive action taking place. Consequently it seems very unlikely that
the gradual rise noticed in the above analyses in the proportion of hexose
sugars as the spring advances is due to any variation in the action of the
invertase during the extraction of the leaf powder. The analytical results,
then, in all probability represent fairly closely the actual proportions of
the two kinds of sugars occurring in the living leaf at the various dates
of examination. Thus this gradual proportional increase in the hexose
sugars as the season advances appears a fact requiring consideration and,
if possible, an explanation.

The mean daily temperature between February and May gradually
rises, and so a possible explanation may be sought here. A higher tem-
perature would favour greater enzymic action. More cane sugar might
therefore be inverted, and so the reducing sugars would tend to increase
at the expense of the sucrose.

CARBOHYDRATES OF FOLIAGKHE LEAF OF SNOWDROP 29

Ilowever, the seasonal variation observed in the relative proportions
of the two classes of sugars canshardly be due to the direct effect of
temperature. [For example, the maximum shade temperature, as shown in
the Table, on March 7th was 66°9° I'.,! while on March 30th it was only
49-69 F., yet the proportion of hexose to sucrose in the first case is much
lower than in the second. Again on April 4th the maximum shade
temperature was 60° I., while on April 24th it was only 51° I., yet this
disparity in temperature has not influenced the seasonal change in the
ratios of the sugars. Further the analyses of February 16th and April
24th, when the maximum shade temperatures were fairly similar, viz.,
48°9° F. and 51° F. respectively, show the proportions of the sugars very
different.

In some previous work on the carbohydrates of Monocotyledons?
certain observations suggested the probability that, as the leaf increases
in age, the starch in the mesophyll diminishes. Now sucrose, as a rule,
favours the formation of starch in the chloroplasts more than any other
sugar. If its percentage falls, then starch will most likely appear less
readily and in smaller quantity. Consequently the reason why starch-
forming Monocotyledons produce less of this carbohydrate in their foliage
leaves, as the season advances, may be due to a general disposition on
the part of such leaves as they grow older to accumulate less sucrose and
more reducing sugar; at any rate, the behaviour of the Snowdrop leaf in
this respect suggests this explanation.

IV. During any single day of the spring the percentage of hexose sugars
in the leaf remains fairly constant, no matter what hour out of the
twenty-four the leaves may be evamined. The sucrose, on the other
hand, fluctuates greatly. It increases during the day and
diminishes during the night. Further, leaves detached and

insolated contain decidedly more sucrose than their controls.

The above deduction must not be confused with preceding one. Here
we are dealing with the daily variations in the percentage of the two
classes of sugars; not with the seasonal change in the relative pro-
portions observed in leaves taken at different dates throughout the spring,
and at the same period of the day in each case. The experiments leading
to the above generalization arrange themselves under four sub-headings.

1. This was an exceptionally warm day for the time of the year.
2. Parkin, loc. cit, p. 46.

30 BIO-CHEMICAL JOURNAL

1. The comparison of leaves picked in the early morning with those
gathered in the late afternoon.

This series of experiments will show the increase in the total sugar
in the leaf due to a certain number of hours of photosynthesis, and also
reveal what kind of sugar is especially affected. For the sake of brevity
the glucose and fructose are put together as hexose. The ratio of these
two sugars to one another is dealt with in a separate section of this paper.
The results are stated in grams of sugar per 100 grams of dry. leaf—the
mode adopted throughout this paper.

Experiment I. March 7th, 1906. Cambridge.

Maximum shade temperature 66-9° F., and minimum temperature previous night 43° F.

: 9 a.m. 3.30 p.m.
Sucrose ee Ae 11-22 14-65
Hexose ae a6 6:35 5:48
Total sugar oie 17-57 20-13

Remarks. The sucrose has increased during the 63 hours of sunlight
by about 5-4 grams, while the hexose shows somewhat of a decrease.

This was an exceptionally warm day for the time of year. The total
increase in sugar is not as marked as in the two following experiments.
Perhaps the temperature, or rather strength of sunlight, retarded assimi-
lation in such a cold-season plant as the Snowdrop; or the small increase
may be explained on the grounds that increased temperature favours
translocation and so less temporary storing of the sugar formed in
photosynthesis.

Experiment II.—April 5th, 1907. Carlisle.
Maximum shade temperature 58° F.

9 a.m. 4.30 p.m.
Sucrose ays foc 9-80 14-65
Hexose siz Bie 12-38 11-61
Total sugar 400 22-18 26-26

Remarks. The total increase in sugar is here about 4 grams, and is
due solely to the sucrose, the hexose having decreased somewhat.
Lxperiment III,—April 8th, 1908. Carlisle.

Maximum shade temperature 58° F., and minimum temperature previous night 32° F.

8.15 a.m. 4.15 p.m.
Sucrose 55 wee 8-88 12-92
Hexose 5Se an 9-4 10-74
Total sugar “0d 18-28 23-66.

CARBOHYDRATES OF FOLIAGE LEAF OF SNOWDROP 31

Remarks. ‘The total increase is nearly 5} grams, and most of it is due

to cane sugar.

The leaves for the above three experiments were picked on fine days,
which might be considered favourable for photosynthesis. A considerable
increase in the total sugar is noticeable.

The following experiment shows the lack of sugar-increase in leaves
gathered on an exceptionally cold day for the time of year :—

Experiment IV.—April 26th, 1908. Carlisle.
CoMPARISON OF MORNING AND AFTERNOON Leaves PIcKED ON A CoLp Sprinc Day

Maximum shade temperature 46° F., and minimum temperature previous night 32° F.

8.30 a.m. 4.30 p.m.
Sucrose oa Bees 9-74 10-10
Hexose as Ace 11-61 11-94
Total sugar on 21-35 22-04

Remarks. The increase is here very small, only 0-7 of agram. The
temperature has most likely been too low for much assimilation. The
sucrose has augmented slightly more than the hexose.

Experiment V.—A fifth investigation was carried out to see the effect on the sugar contents
of a day’s illumination on leaves of plants kept previously in the dark for four days.

The plants were uncovered and exposed to the light at 8.30 a.m., April 12th, 1908; part
of the leaves being taken then and dried, and the rest at 4.30 p.m. after 8 hours’ illumination.
The day was a favourable one for assimilation, the temperature being about normal for the time
of the year.

8.30 a.m. 4.30 p.m.
Sucrose nae oe 5-09 12-55
Hexose ae nae 2-717 3-61
Total sugar on 7-86 16-16

Remarks. This is an interesting experiment. It shows in the first
place the great fall in the percentage of sugar in the leaf, as might be
expected from four days’ obscurity. The effect of darkening on the
diminution of the sugar is, however, referred to in detail in a special
section of this paper. Eight hours of sunlight has caused a great increase
in sugar, over § grams, and most of this is due to sucrose.

This experiment seems to be strongly in favour of the view that
sucrose arises in the leaf before free hexose sugar, such as glucose. If the
sucrose results from condensation of hexose in order to lower the osmotic
pressure, it would be rather expected in the foregoing experiment that the
hexose would be increased more than the sucrose by the exposure of the
leaves to light, as the total sugar percentage in the leaves is so low and
presumably also the osmotic pressure; but the reverse is seen to be the
case.

52 BIO-CHEMICAL JOURNAL

2. The sugars in leaves picked in the evening compared with those
gathered the following morning.

This investigation gives some indication of the quantity of sugar
which passes out of the green leaf towards the bulb during the night.
The total loss of sugar sustained can, however, hardly be due solely to
translocation. The respiration of the leaf will consume a small amount
most likely.

Experiment I. April 2nd and 3rd, 1905. Carlisle. ;

The upper halves only of the leaves were used in this experiment. One batch was gathered
at 4.30 p.m. and the other at 8 a.m., the next morning. The plants in this instance were not
covered, so there might be a little assimilation proceeding after 4.30 p.m., and also before 8 a.m.
The minimum temperature during the night was 32-9° F., and the maximum shade temperature
the previous day 52-6° F.

4.30 p.m. 8 a.m. next morning
Sucrose ses na 14-01 10-36
Hexose rs 53¢ 10-76 10-45
Total sugar oer 24-77 £0-81

temarks. The sugar has diminished by about 4 grams, and this is
due almost wholly to the loss in the sucrose. The hexose has remained
nearly constant.
Experiment II. March 30th and 31st, 1905. Carlisle.
PLANTS NOT COVERED

Maximum shade temperature 49-5° F., and minimum temperature 38° F.

5.30 p.m. 8 a.m. next morning
Sucrose de a0 15-46 10-84
Hexose soc 302 11-41 12-64
Total sugar See 26-87 23-48

Remarks. The sugar has diminished during the night by 3-4 grams,
and this is due solely toa fall in the sucrose, the hexose having increased
somewhat.

Experiment IIT, March 7th and 8th, 1906. Cambridge.

The plants were covered, when the first lot of leaves were gathered from them, in the late
afternoon, and kept in darkness till the time the second batch of leaves were taken the following
morning ; thus no assimilation could take place in the interval, even if the light and temperature
were sufficient.

Maximum shade temperature 66-9° F., and minimum temperature during the night 43° F.

4 p.m. 9.30 a.m. next morning
Sucrose ee ia 14-65 9-64
Hexose a ite 5:48 5:67
Total sugar at 20-13 15-31

CARBOHYDRATES OF FOLIAGE LEAF OF SNOWDROP 33

Remarks. Over 5 grams of sugar have disappeared, due solely to a
reduction in the sucrose. The hexose has remained almost constant in
amount. The conditions were distinctly favourable for the translocation
of sugar.

Experiment IV.—April 5th and 6th, 1907. Carlisle.

The plants were covered when the late afternoon batch of leaves were taken and kept in
darkness till noon the following day—a longer period. Hence a greater loss of sugar might be
expected, which was the case.

Maximum shade temperature previous day, 58° F.; minimum temperature during the
night, 39° F.; maximum shade temperature following day, 52° F.

4.30 p.m. Noon next day
Sucrose ae a 14-65 7-80
Hexose sae ry. 13-66 13°29
Total sugar see 28-31 21:09

Remarks. The sucrose has been nearly halved, while the hexose has
kept about constant. The total decrease is over 7 grams in 17} hours of
darkness.

3. The comparison of sugar contents of leaves detached from the plants
with those attached, both being kept in the dark overnight.

This investigation is supplementary to the foregoing. An indication
will thus be given as to what amount of sugar disappears from the leaf
otherwise than by translocation. In the late afternoon or early evening a
batch of leaves was cut from a plot of plants and their ends placed in
water. These detached leaves were put alongside of the remainder of the
plants with their leaves intact, and the whole covered till some hour the
following morning, when both sets of leaves were taken and dried for
analysis. Also at the time when the frame was covered the previous
evening, a batch of leaves was picked and dried at once.

Experiment I.—March 7th and 8th, 1906. Cambridge.
Minimum temperature in the frame during the night, 46° F.

4 p.m., March 7th 9.30 a.m., March 8th
Attached Detached leaves
Sucrose ... pa 5 14-65 9-64 12-51
Hexose ... aoa “ee 5:48 5:67 5-65
Total sugar oa 20-13 15-31 18-16

Remarks. The percentage of hexose has practically remained
constant in the three sets of leaves. The fluctuation, as usual, is in the
amount of sucrose. This has diminished by 5 grams in the attached

34 BIO-CHEMICAL JOURNAL

leaves, and by 2 grams in the detached. The difference, 3 grams, must
be due to translocation. Whether as much as 2 grams of sucrose has been
consumed in respiratory processes or not, cannot be definitely asserted.
At any rate, this amount has disappeared. Some may have diffused into
the water from the cut ends of the leaves, but any loss in this way must
be very small. The injury due to severance of the leaf from the plant
may lead to pathological complications so that more sugar may be
consumed than normally would be utilised in metabolic activity, hence
we cannot conclude for certain that in this experiment the 3 out of
the 5 grams of sucrose which have disappeared from the attached
leaves during 174 hours of darkness are due to translocation, and the

remainder, 2 grams, to consumption in respiration.

Experiment II.—Apvril 5th and_6th,°1907. Carlisle.

Minimum temperature in the frame during the night, 39° F.
The plants and the detached leaves were kept covered till noon.

6:30 p.m., April 5th Noon, April 6th
Attached Detached
Sucrose ... 568 aoe 14-65 7:80 12-12
Hexose ... 5a cas 13-66 13-29 13-39
Total sugar ce 28-31 21-09 25-51

Remarks. The hexose, as in the previous experiment, has kept
remarkably constant throughout. The extra 2} hours of darkness has
resulted in the additional loss of 2 grams of sucrose. Of course the two
experiments are not strictly comparable, as both the time of year and
temperature differ. The detached leaves have lost much less sugar
than the attached ones, and the relative proportions of the losses in the
two sets of leaves are much the same for the two experiments.

4. The comparison of sugar contents in leaves detached and insolated
with those left attached to the plants.

The leaves were severed from the plants in the morning and their
cut ends placed in water contained in a large flat tray. The individual
leaves were both kept free from one another, and in a_ position
approximating to the vertical by means of a device formed of wire
netting. The tray containing the severed leaves was placed alongside of
the control plants. Thus care was taken to submit the two sets of leaves
to similar conditions of illumination and temperature.

The object of this experiment was to see to what extent the sugar

CARBOHYDRATES OF POLIAGE LEAP OF SNOWDROP 35
accumulated in the leaf when translocation was prevented, and which
sugar was most affected.

Experiment I.—April 5th, 1906. Carlisle.

Maximum shade temperature, 60° F,; a fairly bright day, but hardly any direct sunshine.

Leaves were cut from the plants at 9.30 a.m.
4.30 p.m.

Both sets of leaves were taken and dried at

Attached Detached leaves
Sucrose ons ee 15:83 19-02
Hexose er nee 9-65 8:85
Total Sugar... 25°48 27:87

Remarks. Owing to the translocation of the carbohydrates formed
in assimilation being stopped by detaching the leaves from the plants,
the total sugar has increased by about 24 grams, and this increment is
wholly due to the sucrose, the hexose having diminished by 0°S gram.

Pxperiment IT.—April 5th, 1907. Carlisle.

Maximum shade temperature, 58° F. Sunny morning and dull afternoon.

Leaves treated as for Experiment I.

Attached Detached leaves
Sucrose a ae 14-65 20-42
Hexose ane wits 11-66 10-5
Total sugar... 26-31 30-92

Remarks. These results correspond with those of Experiment I,
except that the total increase in sugar due to prevention of translocation
is considerably greater—nearly twice as much. The sucrose, as usual,
is the sugar chiefly affected.

Bxperiment I1I.—April 21st, 1910. Carlisle.

Maximum shade temperature, 56° F. ; minimum temperature previous night, 45° F. Fairly
bright, windy day.

Leaves treated as in the two foregoing experiments. Insolated 9-9.30 a.m., cut 4-4.20 p.m.

Attached Detached
Sucrose ae Boe 18-73 20-66
Hexose ... es ses 7:78 741
Total sugar at 26-51 28-07

Remarks. The difterences are here less marked than in the two
foregoing experiments, but of the same nature. Detachment has caused
an increase in total sugar of about 1} grams, wholly due to sucrose, the
hexoses having slightly diminished. The windy day, with direct
suushine, may have been less favourable for assimilation in the detached

leaves, as these are apt to flag somewhat under such conditions.

36 BIO-CHEMICAL JOURNAL

V. The fructose, as a rule, is in excess of the glucose.

In the great majority of the analyses, not only have the amounts of
sucrose and hexose been estimated separately, but also those of the two
sugars composing the latter, viz., the glucose and fructose.

With very few. exceptions the fructose has been found to be present
in greater quantity than the glucose. In these exceptions the two
sugars were in nearly equal proportions, the difference being slightly in
favour of the glucose. We are inclined to trace these deviations to errors
in analysis, arising most likely from using an excess of basic lead
acetate. It has already been pointed out (p. 11) that this reagent is a
dangerous one to use, without proper precautions, with sugars, and
especially with fructose. If the presence of this acetate does affect the
sugars, then ‘the fructose will be influenced much more and come out
relatively lower in the calculations than the other two. The tendency
in our mode of estimation, as already shown (p. 21), is for the fructose
to be lower in proportion than it actually should be, hence the value of
the above deduction—viz., that the fructose is in excess of the glucose—
is enhanced.

Out of fifty-two duplicate leaf analyses made, forty-seven had the
fructose in excess of the glucose, and only seven the reverse.
Representing fructose as unity, in the former cases the ratio varied from

1:04 to 1:0°76, and in the latter from 1:1°01 to 1:106. The average

ratio for the whole number of the analyses was 1:0°76. This compares
very favourably with the ratio 1:0°7, obtained for fructose and glucose in
a miscellaneous sample! of leaf, in which the sugars were estimated in
such a way as to dispense with the use of basic lead acetate.

The calculation of the separate amounts of glucose and fructose
depend chiefly upon one observation, viz., the optical angle of rotation.
A slight error in the reading of this will affect the results considerably ;
consequently it might hardly be expected that any further conclusion,
beyond the bare fact of the excess of one hexose sugar over the other,
could be reached.

However, the results do suggest the following two probabilities,
viz., that the proportion of fructose to glucose tends to rise during the
night, i.e., when photosynthesis is in abeyance; and secondly, that this
ratio appears to increase from above downwards in the leaf, the excess of

1, This quantity of dried leaf material was accumulated by putting together the small

portions that remained from all the other samples of leaves used. Hence its analysis gives a
good indication of the average ratio of the sugars.

CARBOHYDRATES OF FOLIAGE LEAF OF SNOWDROP 37

fructose being especially noticeable in the lowest (colourless) part of the
leaf. These two tentative conclusions practically resolve themselves into
one, viz., that, as the total hexose (reducing) sugar increases in
proportion to the sucrose, so does the fructose rise in ratio to the glucose.

On the assumption that the two hexoses arise from the inversion of
part of the sucrose, it would seem that the glucose disappears in some
fashion at a quicker rate than the fructose.

The preponderance of fructose over glucose in the Snowdrop leaf is
a condition somewhat similar to what Brown and Morris! discovered in
the case of the leaf of Tropaeolum. In their researches on this plant
they found the fructose, as a rule,? in much greater abundance than the
glucose. In fact, in leaves gathered after several hours of sunshine, no
glucose at all could be recognised, though fructose was present in
considerable amount.

These authors account for the difference in the amounts of these
two hexoses by considering that ‘dextrose is more readily put under
contribution for the respiratory processes of the cell than is levulose.’3

Further discussion on this point is reserved for the concluding
section of this paper.

VI. Leaves darkened for some days still contain a moderate quantity of
sugar. The percentage falls rapidly during the first forty-eight
of obscurity and then remains nearly constant.

The experiments to be described under this heading were begun with
the purpose of seeing to what extent the green leaf could be depleted of
its sugar by prolonged darkening. Is there a limit to the reduction, or
does the percentage of sugar fall to nearly zero? And further, how are
the relative proportions of the individual sugars affected ?

This part of the research is the least satisfactory and complete. The
results obtained for different years are not as concordant as one would
have liked. This defect may be partly attributed to the imperfect
exclusion of light in the first two experiments. Greater precautions were
taken later to have the plot containing the plants completely darkened.

The chief feature shown by these darkening experiments is that the
sugar depletion rapidly reaches a point at which it stops; the percentage

1. Brown and Morris, loc. cit., pp. 669 and 671, 1893.

2. On page 666 of their memoir an analysis is given showing dextrose in excess of the
levulose. This is an exception to the other estimations they record.

3. Brown and Morris, loc. cit., p. 672.

38 BIO-CHEMICAL JOURNAL

then remains nearly constant, though the darkening may be continued
for some time.

For the four experiments the time of year in each instance was very
similar. The full length of the green part of the leaf was taken, except
in the case of Experiment II, in which only the upper three-quarters of
the length was gathered. The leaves appeared to the outward eye quite
fresh and green, even after two weeks in obscurity.

Experiment I.—Plants darkened at 6 p.m., April 6th, 1906.

Time of plucking of leaves Sucrose Hexose Total sugar in 100
for analysis grams of dry leaf
At the time of darkening... 15-83 9-65 25-48
After 24 hours x68 Sea 8-13 9-83 17-96
After 48 hours oo oe 6:39 7-76 14-15
After 4 days ... 00 0. une 4-79 9-15 13-94
After one week oe = 3-41 11-39 14-80

Remarks. The total percentage of sugar has fallen in two days
from about 25°5 to 14. On further darkening for another two days, the
sugar-contents of the leaf hardly show any additional diminution, and
after a full week of obscurity the percentage has, in fact, risen slightly
instead of fallen. The hexose during the period that the total sugar
remains nearly constant, i.e., from the second day onwards, appears to
rise in quantity at the expense of the sucrose. This latter, then, would
seem to undergo considerable inversion during prolonged darkness.

Experiment I7.—Plants darkened on April 5th, 1907.

Time of plucking of leaves Sucrose Hexose Total sugar in 100
grams of dry leaf
At the time of darkening... 14:65 11-66 26-31
After eleven days ... 300 2:27 9-54 11-81
After two weeks aa ies 5:05 6-54 11-59

Remarks. ‘This experiment was tried with the view of seeing what
effect darkening beyond one week (the limit in Experiment I) would
have on the sugar-contents. The percentage, instead of remaining about
14, has dropped below 12, and apparently stays nearly constant at this
lower figure, even though the darkening be continued for a fortnight.
The relative proportions of the sucrose and hexose have, however, changed
considerably, but in the opposite direction to what took place in the first
experiment.

The two experiments are only in agreement in the fact that the sugar
percentage falls to a certain point and then remains constant.

At this stage it was observed that the method of darkening the
plants was somewhat imperfect. Here and there it seemed possible that

—

CARBOHYDRATES OF FOLIAGE LEAL OF SNOWDROP 39

light sufficient for photosynthesis might have penetrated. Hence for the
next two experiments precautions were taken to ensure complete darkness.

This has resulted in the percentage of sugar dropping considerably

lower.
Experiment [1/,—Plants darkened at 5 p.m., April 8th, 1908.
Time of plucking of leaves Sucrose Hexose Total sugar in 100
grams of dry leaf
At the time of darkening... 12-92 10-74 23-66
After 34 days ... “es dee 5-09 3°35 8-44

Remarks. 'The total percentage of sugar has dropped from over
23 grams to below 8 grams during the darkening period.

Experiment IV.—Plants darkened at 8 a.m., April 11th, 1910.

In this instance a sample of the leaves was not taken for analysis at the time of covering,

but presumably the amount of sugar then would be somewhere near 20 per cent.

Time of plucking of leaves Sucrose Hexose Total sugar in 100
grams of dry leaf
After 3} days darkening... 3°66 554 9-2

Remarks. The fall in total sugar for the same period of darkening
as in Experiment III is not so great, but the difference between the two
analyses comes almost within the limit of experimental error. The
ratio, however, between the sucrose and hexose is the reverse of what it

is in the preceding experiment.

SUMMARY OF RESULTS

1. Only three carbohydrates appear to be present in recognisable
quantities in the Snowdrop leaf, viz., the sugars sucrose (cane sugar),
glucose (dextrose), and fructose (levulose). Starch is always absent,
except for a small amount in the guard cells of the stomata, nor has
inulin been found, though both these polysaccharides are stored
plentifully in the bulb-scales. Maltose has been searched for, but in vain.

(2) The quantity of total sugars in the leaf is considerable—20 to
30 per cent. of the dry weight, as a rule, in leaves actively assimilating.

(5) The amount of sugar increases from above downwards in a
single leaf, and at the same time the ratio of the sucrose to the hexoses
(glucose and fructose) diminishes. These differences are much more
marked in leaves taken from plants growing in close clumps than in those
from plants growing separately.

(4) The proportion of sucrose to the hexoses decreases as the season
advances. That is to say, there is in the leaf more cane sugar in

proportion to reducing sugar during the early part of the season than

.

|
|

40 BIO-CHEMICAL JOURNAL

later. Thus it seems that as the spring advances the hexose sugars
increase at the expense of the sucrose. The comparisons were naturally
made between leaves gathered about the same period of the day.

(5) During any single day of the spring the percentage of hexose
sugars in the leaf remains fairly constant, no matter at what hour out
of the twenty-four the leaves may be examined. That of the sucrose, on
the other hand, fluctuates greatly. It increases during the day and
diminishes during the night. Further, leaves detached and insolated
contain decidedly more sucrose than their controls, but the quantity of
hexose sugar remains much the same.

(6) The fructose, as a rule, is in excess of the glucose, irrespective
of the period of the spring or time of day the leaves are picked for
analysis. The results further suggest that, as the total hexose sugar
increases in proportion to the sucrose, so does the fructose rise in ratio to
the glucose.

(7) Leaves darkened for some days still contain a moderate quantity
of sugar. The percentage falls rapidly during the first forty-eight hours
of obscurity, and then remains fairly constant.

GENERAL DiscuSSION

Our knowledge of the carbohydrates of foliage leaves dates from the
year 1862, when Sachs published the first of his classical memoirs on the ©
significance of starch in carbon-assimilation. By his many and beautiful
researches, covering a period of twenty years, he proved beyond doubt
that the appearance of starch in the chloroplast is the direct outcome of
the fixation of carbon under the influence of sunlight. This carbo-
hydrate was therefore regarded by Sachs as ‘ the first visible product of
assimilation,’ though at the same time he was quite open to the view
that the formation of sugar preceded that of starch. He, however,
considered that all the carbohydrate synthesised in the leaf passed
through the starch stage. Respecting the disappearance of the starch
from the leaf he was strongly of the opinion that it was conducted away
in the form of sugar,

The connection between starch and sugar received considerable
attention from Schimper, who, in his paper of 1885,! held that starch is
not only converted into sugar (glucose) for the purpose of translocation,
but that it also arises in assimilation from this sugar. According to

1. Schimper, Bot. Zeit., p. 738, 1885.

CARBOHYDRATES OF FOLIAGE LEAF OF SNOWDROP 41

Schimper’s work, then, glucose rather than starch should be looked upon
as the first recognisable product of photosynthesis.

Taking into account the formaldehyde hypothesis of Baeyer
advanced in 1870, a general theory of carbohydrate synthesis in the green
leaf could now be framed. The carbonic acid (CH,O,) is reduced by the
light and chlorophyll to formaldehyde (CH,O) with the evolution of
oxygen. The aldehyde as soon as formed is polymerised to glucose
(C,H,,0,); this sugar is then condensed to starch, and temporarily
stored as such in the chloroplasts. The starch, when required to be
moved from the leaf, is re-converted into glucose, and travels in this
form. Some such theory of carbohydrate photosynthesis may be said to
have been the general one held up to the year 1893.

Practically no attempt had been made to discriminate between
glucose (dextrose) and fructose (levulose). Maltose, a sugar then known
in connection with the germination of the barley grain, had not even
been suggested as a possible leaf-carbohydrate. Even cane-sugar had
received scant attention, and its presence in the leaf had not been taken
into serious account with regard to photosynthesis. It was looked upon
more as a reserve than as a circulating carbohydrate.

A fact known at this period, namely, that several Monocotyledons
never under ordinary conditions form starch in their foliage leaves, may
be regarded as a feature not wholly in accordance with the above simple
theory of carbohydrate assimilation. Arthur Meyer,’ paying especial
attention to such plants, comes to the conclusion that the reason why such
leaves fail to form starch is not due to a too rapid translocation. He
favours (erroneously, as it would appear now) the view that the
synthesised carbohydrate is stored in a different form, and cites Yucca
filamentosa as a probable case in point. Sinistrin (inulin) he discovered
in its leaf, and this he considers may be in place of starch. By estimating
the amounts of soluble carbohydrates, he discloses also the interesting
fact that the leaves of non-starch formers contain much more sugar than
those which normally produce starch.

Meyer’s work, though distinctly valuable and suggestive, just failed
to advance the subject of carbohydrate assimilation a real step beyond
Sachs. This was reserved for Brown and Morris. Their work published
in 1893? marks the commencement of the second stage in our knowledge
of the carbohydrates of foliage leaves.

1. Meyer, Bot. Zeit., Nos. 27-32, 1885,
2. Brown and Morris, loc. cit.

42 BIO-CHEMICAL JOURNAL

In their study of the Tropaeolum leaf, they show strong reasons for
rejecting Sachs’ supposition that all the carbohydrate formed in
photosynthesis passes through the starch stage. It is only the excess of
sugar produced that is so transformed and temporarily stored. Further,
they prove beyond doubt that the leaf starch is dissolved in a somewhat
similar fashion as in the germinating barley grain. The enzyme,
diastase, acts upon it and converts it into maltose. Thus this sugar 1s
demonstrated in the foliage leaf for the first time.

But their most novel conclusion is with respect to the cane-sugar.
This they found more abundantly present than even starch, and the
fluctuations in its amount suggested strongly to them that it is the first
sugar to be formed in carbon-assimilation. The presence of the two
hexoses are nore readily accounted for as the products of the hydrolysis
of some of the cane sugar than as originally preceding the sucrose. Since
levulose was, as a rule, in greater abundance than the dextrose, the
authors conclude that the latter contributes more readily to the
respiratory needs of the leaf.

The general theory respecting the carbohydrates of the foliage leaf
to be deduced from their experiments and conclusions thus differs
considerably from, and is more complhcated than, that resulting from
the work of Sachs and his immediate successors.

At this point it may be well to bring into line the results of the
work on the Snowdrop leaf described in this paper. Only three carbo-
hydrates (sugars), viz., sucrose, glucose and _ fructose, require
consideration. Starch is absent, and so is its hydrolytic product maltose.
From the analyses the main daily fluctuation is shown to occur in the
amount of the cane sugar, the quantity of the hexoses remaining
remarkably constant, both under natural and experimental conditions.
Even in leaves previously depleted largely of their sugars through
remaining in darkness, the glucose or total hexose is not increased by
exposure to sunlight, the augmentation occurring almost solely in the
sucrose.

Thus our analyses suggest strongly that sucrose is the first
recognisable sugar to appear in the Snowdrop, and that the two hexoses
arise from it through inversion. Our results further support those of
Brown and Morris in the fact that the quantity of the fructose is almost
invariably in excess of that of the glucose, pointing to the latter sugar
contributing more readily to the needs of the leaf.

Let us now turn to the consideration of a piece of research which

CARBOHYDRATES OF FOLIAGE LEAF OF SNOWDROP 43

supports the older view that glucose (hexose) is the first sugar to arise
in photosynthesis. Strakosch! in 1907 published an important paper on
the carbohydrates of the sugar-beet leaf, employing a new microchemical
method for sugar identification based on the production of osazones in
the tissues. By this means of localisation dextrose was found to be the
only sugar recognisable in the mesophyll, levulose and sucrose appearing
first in the lateral veins of the lamina and increasing in the midrib and
petiole. These results were confirmed by direct estimations of the sugars
in extracts made from the mesophyll (ground tissue) and the veins of the
leaf-lamina, respectively.

Strakosch considers, then, that dextrose is the first sugar to arise in
photosynthesis. It travels from the mesophyll to the small veins, and a
part is there transformed into levulose. The two hexoses then combine to
produce sucrose, and the carbohydrate in this form travels to the root to
be stored. In leaves previously darkened and then exposed to sunlight,
the formation of starch in the chloroplasts commences later than the
increase in the sucrose. Maltose seems only to have been detected in the
petiole.

A striking point of agreement in this research with the work on the
Snowdrop is the constancy observed in the amount of the hexose sugar.
There was no appreciable diminution in its quantity during prolonged
darkening, and no increase during many hours of illumination. The
author, however, has noticed a slight diminution in the hexose, with a
corresponding increase in the sucrose, on the first exposure of a darkened
leaf to light. He concludes, therefore, that the transformation of hexose
into sucrose is dependent upon lght, and ceases when the leaf is
darkened.

Though Strakosch’s theory regarding the sequence of sugars
apparently fits the facts, yet his interpretation fails to carry complete
conviction. The rapid rise of sucrose in the leaf on its first exposure to
light suggests its. very early appearance in photosynthesis. Then
sucrose, as a rule, seems to immediately precede, that is, gives rise to
starch in plant tissues, but here in the chloroplasts of the beet leaf,
according to Strakosch’s results, it would appear that glucose was the
precursor of starch, as cane-sugar was not discovered in the mesophyll.

Robertson, Irvine and Dobson,? in a paper published in 1909 on the

1. Strakosch, Sitz. K. Akad. d. Wissen. Math.-Nat. Cl., Bd. CXVI, p- 855, 1907.
2. Bio-Chemical Journal, IV, p. 258, 1909.

44 BIO-CHEMICAL JOURNAL

sucroclastic enzymes in Beta vulgaris, favour the view of Strakosch in
regarding both the formation of glucose and fructose as preceding that
of the sucrose in the leaf. These two hexoses they consider are
condensed to cane-sugar by enzymic means either through the reversible
action of the invertase or by a special enzyme.

The problem affecting the first sugar to arise in a free state in
photosynthesis thus centres around the origin of the sucrose in the foliage
leaf.

Three different views may be presented. The first, that glucose
appears primarily. Part of this is converted into fructose; then the two
hexoses are condensed to cane sugar. This theory harmonises with the
formaldehyde hypothesis. A weak point is the transformation of glucose
into fructose.” This can be brought about 7m vitro by the action of weak
alkali, but then mannose appears also. There is no evidence that this
third hexose occurs in the foliage leaf.

The second view, viz., that the cane sugar is derived from the
maltose, and so indirectly from the leaf starch, is ruled out of court by
the fact that the Snowdrop leaf contains neither starch nor maltose,
and yet holds an abundance of sucrose.

The third view, first advanced by Brown and Morris, states that
sucrose is the first sugar to arise in photosynthesis, and that the two
hexoses, glucose and fructose, as well as the starch, are derived from it.
The results with the Snowdrop leaf seem more in accordance with this
theory, though at the same time they cannot be considered wholly at
variance with the view first put forward.

Granting that sucrose is the first sugar to arise in the free oie
how, it may be asked, is this to be reconciled with the aldehyde
hypothesis? In the first place this theory still awaits complete proof.
As crudely put, the hypothesis may be incorrect. Free formaldehyde,
as well as free hexose directly polymerised from it, may never occur
normally in the chloroplast. The photosynthetical process connected
with the reduction of the carbonic acid may be more complicated than
has generally been supposed. The carbohydrates resulting from it may
none of them be truly up-grade products. The earliest to appear in the
free state may be spht off from some complex molecule, and might quite
easily be sucrose in many, if not most, of the higher plants.

There is much evidence to show that this sugar is very widely if
not universally distributed in the Flowering Plants.! It occurs also in

1. Schulze and Frankfurt, Zedtschrijt fiir physiol. Chemie, XX, p. 511, 1895, and XXVII,
p. 267, 1899; Bourquelot, J. Pharm. et Chimie, Sér. 6, XVIII, p. 241, 1903.

CARBOHYDRATES OF FOLIAGE LEAF OF SNOWDROP 45

the Vascular Cryptogams,! but has been less studied in these. Its presence
in the Mosses has been asserted. As yet it has not been recognised in the
Algae. It is apparently absent in the Fungi, its place being taken most
likely by the disaccharide sugar, trehalose.

Though it can function as a reserve carbohydrate, yet there are
considerable grounds for believing that it is largely the form in which
sugar travels in the tissues. The plant apparently converts indirectly
its reserve carbohydrate into this form of sugar for purposes of
circulation. This has been observed in several germinating seeds,
tubers and bulbs. Though in each case the sugar immediately arising
from the stored carbohydrate through enzymic action is different from
sucrose, yet on its way to the growing parts of the seedling or young
sprout it seems to have been transformed into sucrose, as this latter
sugar now appears. To give a few instances. Brown and Morris?
showed that the maltose arising from the starch in the endosperm of the
sprouting barley grain becomes changed into cane-sugar on its passage
into the growing embryo. Further, excised embryos nourished with a
solution of maltose accumulate not this sugar but sucrose. Griiss? has
shown, in addition, that such embryos, when fed with glucose, can change
this sugar into sucrose. This author* has also found that in the
germination of the date palm seed, the mannose and galactose resulting
from the reserve cellulose appear as cane-sugar in the embryo. Schultze
and Frankfurt® find sucrose quickly following the dissolution (i) of
starch in the germinating seed of Vicia sativa, (11) of galactan in that of
the lupine, and (ii1) of oil in those of Helianthus and Ricinus. Also in
the sprouting potato, where the carbohydrate is stored as starch, sucrose
soon makes its appearance. We have found, likewise, in the bulb-scales
of the growing Snowdrop, sucrose the most abundant of the sugars instead
of maltose and fructose, which might have been expected from the nature
of the reserve materials, starch and inulin. As far as the writer is aware,
in no case has it been definitely shown that the sucrose arises directly
from the hydrolysis of any reserve carbohydrate. It thus bears an
analogy to asparagine in nitrogenous metabolism. This amide is of
common occurrence in the higher plants, and is probably the form in

Anderssen, Zeitsch. f. physiol. Chemie, XXIX, p. 423, 1900.
Brown and Morris, J. Chem. Soc., LVII, p. 458, 1890.
Griiss, Ber. deut. Bot. Ges., XVI, p. 17, 1898.

Griiss, Ber. deut. Bot. Ges., XX, p. 36, 1902.

Schultze and Frankfurt, loc. cit.

Sv Ghee

46 BIO-CHEMICAL JOURNAL

which nitrogen largely travels; yet in germinating seeds, on a parallel
with cane-sugar, it does not appear directly as a result of protein
hydrolysis, but arises somewhat later in an indirect fashion.! It may
thus be regarded as the ultimate product of protein dissolution, just as
sucrose is in the hydrolysis of reserve carbohydrates.

Sucrose has been found in many cases to be the sugar of most
nutritive value to the plant. It is also the one which in most cases
favours starch deposition. It can therefore be regarded as the precursor
of starch.

Its special physical and chemical properties are also .of interest.
It is very soluble, and readily crystallises—more so than the other sugars
occurring in plants. It is very easily hydrolysed by acids and by
invertase. It shares with trehalose alone among the disaccharides in
having no reducing properties. Maltose, lactose, etc., do reduce, and so
may be said to have the aldehyde group in their molecule functional.

Sucrose may thus have been selected by the higher plants as the
chief circulating sugar, partly on account of its non-reducing properties
and soluble (mobile) nature, and partly on account of the ease with which
it can be hydrolysed into its two components, glucose and fructose. These
hexoses may, as a rule, play distinct parts in metabolism—the glucose
more readily lending itself to the respiratory needs, and the fructose to
constructive work, such as the building up of the plant’s framework.?
It is also within the bounds of probability that cane-sugar itself may
take a direct part in the formation of cell-walls. Just as it appears able
to be condensed to starch without previous inversion, so it may be
transformed directly to cellulose in the construction of cell-walls.
Fenton’s work? is interesting in this connection. He has shown that
various kinds of cellulose respond markedly to a special ketose test; and
thus concludes that this substance may contain one or more groups
identical with that present in fructose.

To sum up, sucrose then seems to be of paramount importance to the
higher plants. It probably functions largely as a circulating sugar,
travelling as such to the growing regions, to be there partly inverted to
supply hexose sugar for respiration. To draw a rough analogy, it may be

1. Brown, H. T., Trans. Guinness Research Laboratory, Vol. I, pt. 2, p. 296, 1906; Schulze,

Landv, Jahrb., XXXYV, p. 621, 1906; Scurti and Parrozzani, Gazzetta, XX XVIII, i, p. 216, 1908 ;
Wassilieff, Ber. deut. Bot. Ges., XXVla, p. 454, 1909.

2. See Lindet’s recent paper, C.R. Acad. Se. Paris, CLII, p. 775, 1911.

3. Fenton and Gostling, J. Chem. Soe., LXXIX, p. 361, 1901. See also Cross and Bevan,
J. Chem. Soc., UX XIX, p.. 366, 1901.

ARBOTYDRATES OF FOLIAGE LEAK OF SNOWDROP 47

likened to the sovereign in our currency. This coin can circulate over the
whole world, and at any time or place be changed for local purposes, so
sucrose in the plant may be able to travel anywhere in the tissues, and is
perhaps continually being called upon to change itself into hexose sugar
for local needs.

Is sucrose, then, not intimately bound up with the photosynthetical
process? If this sugar be of so much importance to the flowering plant,
has not the chloroplast devised a means of rapidly forming it without the
previous production of glucose and fructose in quantity? Much further
investigation with this possibility in view seems desirable.

48
ON DEAMIDIZATION*

By GERTRUDE D. BOSTOCK, B.Sc., M.B., Cu.B., Carnegie Scholar,
From the Laboratory of Physiology, University of Glasgow
Communicated by Dr. E. P. Cathcart

(Received May 22nd, 1911)

PRELIMINARY.

When we consider the fate of ingested protein in the organism, we
find that its cleavage products, consisting of polypeptides and amino
acids, are present in the intestine as the result of digestion. In the
course of absorption we lose sight of them, nor have attempts to find them
in the blood proved very successful, though it must be remembered that
Howell! claims, by the use of the most modern methods, to have
demonstrated their presence. When the nitrogen of ingested protein
again appears, it is almost wholly in the form of urea in the urine. The
reappearance of the ingested nitrogen does not necessarily imply the
complete metabolism of the protein itself. Indeed, Voit has shown that
the nitrogen of the protein is rapidly excreted, while the metabolism of
the nitrogen free moiety, as estimated by the CO, output, is spread over
many hours. As a yielder of energy, nitrogen cannot be regarded as
being of much importance. The dynamic value of an amino acid is only
shghtly lessened by the removal of the amino group. Further, the
organism has great difficulty in storing nitrogen; an increased intake of
nitrogen leads only to a temporary retention, and nitrogenous
equilibrium is quickly re-established.

In attempting to establish nitrogenous equilibrium by feeding dogs
with the minimum amount of protein, Michaud? emphasised the
importance of the chemical composition of the protein as indicated by the
relative amounts of the various amino acids it contains. He found that
the protein minimum was lowest when the composition of the protein
fed corresponded most closely with that of the animal’s own tissues,
proving apparently that only appropriate ‘ Bausteine’ were used in

“In all that follows the term deamidization is used to express the fact that ammonia
makes its appearance as a result of the action of living tissue on an amino acid or the amide of

such an amino acid; strictly speaking the latter is a deamidization and the former a
desamination.

ON DEAMIDIZATION 49

building up the living flesh. In the conditions of his experiments the
protein was manifestly all needed for synthesis. In the well-fed animal,
however, protein is probably used largely for purely dynamic purposes,
and in this case deamidization may well play an important part in
protein metabolism.

Whatever the immediate precursors of urea may be, it has been
abundantly proved that ammonia is converted into urea in the liver. The
ammonia liberated from the amino acids produced in the course of
digestion will share a like fate, leaving the non-nitrogenous remainder
available for the energy demands of the organism.

Investigation of the literature in this field has shown that there is
much indirect evidence available in support of this deamidization theory.

Abderhalden and Terruchi? found that subcutaneous injection of the
dipeptides, glycyl-glycin and alanyl-alanin in dogs led to a large increase
in the excretion of urea in the urine. Stollte+ injected amino acids into
the blood stream of rabbits, and found that glycocoll and leucin were
practically all excreted as urea. Asparagin, elutaminic acid and alanin
were partly excreted as urea and partly as unchanged monamino acid.

Salaskin® perfused the liver with blood to which amino acids had
been added, and found that glycocoll, leucin and asparagin could be
largely recovered from the blood as urea or an amide-like body. Various
observers have carried out feeding experiments on animals with aromatic
amino acids, and have succeeded in isolating the deamidized aromatic
acid from the urine. Flatow® found in rabbits that o. tyrosin was
converted into 0. oxyphenyl acetic acid, and m. tyrosin was converted
into m. oxyphenylpyruvic acid. Chlorphenylalanin also appeared
abundantly in the urine as chlorphenylpyruvie acid. Friedmann?
fed dogs with p. chlorphenylalanin, which, after conversion into
p. chlorphenylacetic acid, paired with glycocoll, and reappeared in the
urine as chlorphenaceturic acid. Finally, Neubauer and _ Fischer®
perfused a dog’s liver with phenylamino acetic acid, and found that
phenylglyoxylec acid had taken its place in the blood at the end of the
experiment.

In alcaptonuria interesting evidence is also obtained of a deami-
dization occurring in the side chain of the aromatic amino acids. Tyrosin
and phenylalanin are both excreted in the urine as homogentisinic acid,
or 1, 4 dioxyphenylacetic acid, the presence of the aromatic nucleus in
this condition apparently interfering with a further oxidation of the
side-chain, It is interesting to note that there has been a loss of one

50 BIO-CHEMICAL JOURNAL

carbon atom in the side chain in addition to that of the amino group.
Abderhalden® fed alcaptonuric patients with polypeptides containing
tyrosin and phenylalanin, and found an increase in homogentisinic acid
corresponding in amount to the aromatic amino acid present in the
polypeptides. Neubauer and Falta! found that phenylalanin in
alcaptonuria was first converted into phenyl a lactic acid.

Further evidence in favour of deamidization of amino acids is found
in the capacity of the liver to form acetone and diacetic acid on perfusion
with certain amino acids.

As regards direct im wtro evidence, there is not much available.
Experiments which have been carried out have consisted as a rule of
digesting tissue to which amino acids have been added. Jacoby",
experimenting with autolytic enzymes, found an increase in ammonia
on digesting glycocoll with liver extract. During the course of autolysis
he found that a loosening of the tightly-bound nitrogen of the amino
acids occurred, with a resulting increase in amide nitrogen (urea ?).

Loewi!? on repeating Jacoby’s work also found that glycocoll added
to an alcohol liver extract was converted into a urea-like body. He
concluded that amino acids were readily attacked by ferments, and gave
rise to bodies containing loosely bound nitrogen which were closely
related to urea.

Gonnermann!® incubated certain amides—formamide, acetamide and _
succinamide—with liver and kidney emulsions, and found that saponifica-
tion occurred, the free acid being liberated. The organs sometimes
differed in their action, e.g., acetamide and succinamide are split by
liver, but not by kidney. Lastly, Lang!4 claimed that deamidization
occurred in most of the organs and tissues of the body, and Furth and
Friedmann! carried out a few experiments to investigate the extent of
asparagin splitting by organ ferments. They used Lang’s method, and
found that liver, spleen, muscle, kidney, lung, brain and intestinal
mucous membrane all split amide nitrogen from asparagin in autolytic
experiments. The intestinal mucosa evinced the greatest activity.

Lang’s (loc. cit.) work was received practically without criticism,
owing no doubt to the fact that it fitted in admirably with many known
facts in the metabolism of protein, such as the rapid appearance in the
form of urea of the nitrogen of protein or amino acid, when introduced
by mouth into the organism. In view of the importance of such a link in
the downward metabolism of the amino acids, it seemed well worth while
to repeat and extend some of Lang’s work, especially as his claim that

—_——

ON DEAMIDIZATION 51

the process is carried on to a significant extent in intermediate protein
metabolism, is based on rather slender evidence.

To test the capacity of the tissues to liberate ammonia from amino
acids in vitro he carried out a large number of experiments with various
organ emulsions, incubating them at 38-40° C. for longer or shorter
intervals with certain amino acids and other bodies containing the
amide group. The organs examined were liver, pancreas, spleen, kidney,
testis, adrenal, lymph gland and muscle tissue. Glycocoll, tyrosin,
aspartic acid, asparagin, cystin, leucin, phenylalanin, acetamide,
glutamin, glycosamin, urea and uric acid were the _ substances
investigated. Lang’s more important results were as follows :—

The total amide nitrogen of asparagin and glutamin was split off
by all the organs; a small portion of the amino nitrogen as well was
apparently liberated by the liver.

Glycocoll was split to a less extent by kidney, liver, testis and
adrenal, and abundantly by pancreas and intestine.

He further noted that fresh aseptic tissue was more active than tissue
incubated under antiseptic precautions, e.g., the fresh intestinal mucous
membrane of a dog sphi an amount of glycocoll in one to two hours,
equal to that split in twelve days by the incubated mucous membrane
incubated in presence of an antiseptic.

The figures quoted from Lang in the following table (Table I) are
concerned only with liver and intestine, as those are the organs which
have been used in the present series of experiments.

It is largely on the value obtained for the ammonia liberated from
glycocoll in Test 21 that Lang bases his claim for the importance of
deamidization in intermediate proteid metabolism, i.e., in one and a half
hours as much ammonia is liberated in the aseptic test as is liberated at
the end of twelve days’ incubation in presence of an antiseptic.

PRESENT INVESTIGATION
I. Liver

The first question to be answered was as to what extent deamidization
could be demonstrated in an organ emulsion to which an amino acid or
an amide of an amino acid had been added.

The method followed in carrying out the present series of experiments
closely resembles that used by Lang. The organ, unless obtained from

-- eo oS = 96-¢ tI
€1 €8-6 1¢-S1 oe 89-¢ §&%
Cl ras Z1-S1 ae 26-8 sal
= — GOST - 2-06 — cal
"NL quds ‘3 Og Ul “3 Og Ur ‘3 Og Ut poeyeqnout
esryueoleg N*HN ‘SIV NPN “SIT N°HN “31 H®HN ‘3 sinoy
= TIONDODATY) NIDVUVdSYV TOULNOD
2, -- = = — 0 ZE-29 — — _— ZE-29
ae — — — €Z SLT 8-09 — — — 80:69
5 —_— — pass £3 8-8 6L L-8h 8-6L F-9E
S L83 FI 99:67 — — — — — — 9:8
hs — — — — _- — — —- —_ IF-FE
— = = = =3 3 Le IL-8¢ 86 6G: FF
= — — — EG € 67 — — — oF
= — = —_ GG LUE 6-1F = — — EL-8E
oD — — — G21 L G68 eg 8-1¢ +8 raya
a — — — — g-g— 8-8 — — — 9:16
= = — = ee rat? 83-6F — — — 91-SF
eal = = a ae 26-0— Z-9¢ — — — ZI-LG
a8)
'S) ( poy omardsy
3 — = aa — 26° — 9G-Eh 8-1 SPI 96:L¥ ) SL-9F
= = a= — = — rE F9-18 9-19 96:62
fea) a = — G-F1 9¢-81 FZ-0¢ = — _ 89-98
peppe
utono'y utono'y ay 9) , OAT uisvivdsy ‘dsy wody urs10

JO'N'L JO woszgyds “Bog ur jo nL jo Wow qyds “Bog ur jo NL JO poywoqy “Sogur “Bogur
advqueoied N8HN ‘SIN N°HN ‘SIV odvqueoI0g N*HN “SIX N°HN “SIN ostquooieg N*HN “SIT N°HN’SIN N°HN SI
NIOOYT TIOOOOATY) NIOVUVdISY IOULNOY)

J aiavy,

52

poofq og + s0Aq 30S ‘%

poojq OF + AOATT “BLZ “1S
pooyqg OOT + “3 09 Jeary Zoq “OZ

96
6
a

aory
-uqnouy
sAvq

“ec

eee 3 0g 7
So, Soe

eee 3 GP “ec

rel

V ‘3 0g “é

“BD OGT wary xO

3 OO oe

‘399 yn3 xO
ourqsoyUl XO
SYIVWUOY

.
_

‘OT
“ST
FI

ON DEAMIDIZATION 53

an animal killed in the laboratory, was brought as quickly as possible
from the slaughter-house (cizea half an hour) and was immediately
minced finely by a mincing machine.

A number of tests each weighing 50 grams were then weighed into a
series of glass-stoppered jars; 100 c.c. of 0°9 per cent. sodium chloride
solution was added to each test, with 2 ¢.c. toluol. To every test, save
those serving as controls, an addition of 0°3 to 1 gram of an amino acid
or amide was made. All the tests were well shaken, and after being
overlaid with toluol were incubated at 38° C. for periods varying from
days to weeks, and even months.

To ensure antisepsis, Lang used a shaking machine; this was
frequently omitted by me, as Meltzer!® has stated that enzymes could be
destroyed by continued shaking. The tests were shaken by hand for five
to fifteen minutes, when the shaking machine was omitted, and were then
overlaid with toluol and incubated as usual.

At the end of the incubation period the tests were acidified with
acetic acid, and after the addition of 40 c.c. of a 5 per cent. tannic acid
solution they were heated, with continual stirring, until coagulation
occurred. On cooling, the precipitate and fluid were made up to 400 or
500 c.c.; half of the filtrate for this was taken for the estimation of the
ammonia, which was carried out in vacuo at a temperature of 38-40° C.
Lang used magnesia for driving off the ammonia, but Kruger and
Reich!’ point out that magnesia causes a gradual liberation of ammonia
from certain nitrogenous bodies. Grafe’s!® recommendation was therefore
adopted, and a mixture of sodium carbonate and sodium chloride was
substituted for the magnesium oxide.

A number of tests prepared according to this method were incubated
in the presence of an antiseptic (toluol) for periods varying from a few
days to three or four months. Liver was the organ first investigated, as
previous work, such as that of Salaskin, Stolte, and others, had shown
that the liver was the site of some kind of deamidization. The amino
acids used were glycocoll, leucin, alanin, and the amide asparagin. A
glance at the table below (Table IL) will show that in every case ammonia
has been liberated from asparagin in fair amount, but in the case of
glycocoll the liberation has been much less marked, and in some cases
has even been absent.

The next question was how far the antiseptic used inhibited the
action of the deamidizing enzyme; in other words, in a fresh organ
emulsion incubated for several hours without an antiseptic, is the amount

BIO-CHEMICAL JOURNAL

54

Ge G
— 0
GE G
— 0
€-¢ &
= I-
GE &
e+ 8
€-F 8
9-F1 LG
“oATS) “oATO
JO'N'L jo wmoay qryds
esvqueoi0gy NHN ‘SIV
TIOOOOATY

“3 Og UL
N°HN ‘ON

£-06 6-3
L-09 Ll
L->L 6-02
L-09 LI
8L GS
0Z €-L
€-FF F-El
GS L
FO LI
1-L8 9&
GS ¥G
1-Z& cT
‘dsy ‘dsy
jo "N°, Jo amoas yids
esvyuod1eg

NIDVUVdSV

67
6E

‘BD Qg ul
N°HN ‘3 N®HN “SIN

Il 214V.L

€-96 99 se
cE 6 ahs
66 61
IZ L
9% L ‘
€-F1 SI
sheq
9¢ lac ;
F-09 $¢ ao
syyUOW,
LE rat 4
GG Gl ie.
GZ £9 eee
66 LT
O& L ‘a
VG 1G
86 6 oe
8P 61
GZ Il
¥ f¢ ;
‘3 gg ut peyeqnoutr
NHN “3IN skeq
IOULNOD

‘B QG 10AIT XO
“dig tOAIn |
see 3 0g oe “<
see 3 Og 4nx) “ce

‘3 Og (teary + yn) dooyg

ia ‘BOG JoaryT 1qq"a

‘3 oc “
see + +3 9g Jeary xO
eis ‘3 Oe Jeary uvuny,
‘3 Q¢ JoAryT [ejoog ueuINA
“BOG Joary sod [e}04

‘3 og sesry deayg [e}90F
eee ++ +B Og JOAVT goq

sc ‘3 Og sary doayg
"3 OG Jeary XO

tsi

ON DEAMIDIZATION 55

of ammonia liberated from added amino acids greater or less than in an
antiseptic test incubated for days instead of hours? Lang laid very great
stress on the rapidity of deamidization in the ‘aseptic’ tests; he was
satisfied that bacterial action could be disregarded if proper aseptic
precautions were observed, and if the incubation period were not
prolonged (1} to 24 hours). It may be remarked that there is no record
of any cultures being taken to test the success of his precautions. As the
result of these series of tests with the fresh organ incubated aseptically,
Lang maintains that the deamidizing enzyme is retarded in its action by
toluol, and that deamidization in the tissues of the organism is much more
extensive than a consideration of the amount of liberated ammonia in the
antiseptic experiments would indicate.

Some tests were prepared in the usual way, but without the addition
of any antiseptic; the results are tabulated below in Tables III and IV.
The instruments and vessels were sterilized, and the tests were incubated
from 1} to 5 hours; they were not shaken continuously for reasons given
above. In the serial tests (No. 3, Table III), in which the first tests
were incubated for 1} hours, and the others for 3} hours, an addition of
toluol was made to the latter at the end of incubation, as it was
impossible to work up all tests on the same day at the end of incubation.

These tests (b and c) were kept at laboratory temperature (7° C. to
14° C.) until they could be dealt with: As a result it was found that not
only did the ammonia liberated increase in the latter tests, but the
ammonia in the controls increased also; evidently autolytic and
deamidizing processes were still active at a temperature much lower than
38° C., and in the presence of an antiseptic. In No. 6, Table III, the
same thing is evident, although in this case the tests were placed in the
refrigerator immediately after incubation. In order to stop further
ferment action the tests in all subsequent experiments were always heated
immediately after incubation till the protein was coagulated.

TABLE III
CoNnTROL ASPARAGIN
: Hours Mg. NH,;N Mg.NH3sN NH;3N split Percentage
incubated in 50g. in 50 g. from Asp. of T.N. of
Asp.
3. 6.5.10 Ox Liver 50g. (a) 1} 3-6 5-2 1-6 3-4
7.5.10 si (b) 3h 12 19 7 15
9.5.10 a (c) 34 21 40 19 47-2
4. 13.5.10 Dog Liver (a) 3 14 15 1 1
15.5.10 z (d) 3 14 20 6 6-4
6, 28.5.10 Dog Liver (a) 4} 15 16 1 2
30.5.10 = (d) 43 16 21 5 10-7
33.5.10 a (c) 4h 30 49 19 38-5

56 BIO-CHEMICAL JOURNAL

A comparison of Lang’s table (Table I, Experiments 20, 21 and 22,
with Experiments 3, 4 and 6, Table III) shows that Lang obtained
higher values for the liberated ammonia than I was able to get. As he
used defibrinated blood in these ‘aseptic’ tests, it was thought that the
absence of this in my experiments might explain the lower figures
obtained.

In Experiments 11 and 15, Table IV, an addition of 50 c.c.
defibrinated blood to each test in 11, and of 35 ¢.c. in 15, was made.

These tests incubated with defibrinated blood do show on the whole
a more active deamidization, probably because of a closer relation to
vital conditions. In most of the subsequent experiments, therefore, an
addition of 25 c.c. defibrinated blood was made to each test.

Part played by Bacteria. As gelatine cultures taken from the
‘aseptic’ tests at the end of incubation invariably gave positive results,
it was of the greatest importance to ascertain the possible part played by
bacteria in deamidization. It has been proved that a liberation of
ammonia does occur when putrefaction organisms are incubated with
amino acids. In a recent publication, Brasch!® showed that the
B. putrificus deamidizes glycocoll and alanin, yielding acetic and
propionic acid in their place, respectively. In the more complex amino
acid, aspartic acid, formic, propionic and succinic acid appeared, and not
merely the products of a simple deamidization.

It might be urged that the presence of an antiseptic in the tests is
sufficient to prevent bacterial action. Gelatine stab cultures were made in
Experiments 12, 16, 17 and 19, Table II, with negative result, but it
must be remembered that a negative culture does not prove the absence of
bacteria.

The simplest way of investigating the part played by bacteria in the
deamidization of asparagin in an organ emulsion is to compare the
amount of ammonia liberated at the end of incubation in a test containing
both enzyme and bacteria with the amount liberated in a test containing
bacteria only. The absence of the enzyme in the latter case can be
assured by heating the tests to 100° C. Subsequent inoculation provides
the bacteria.

In accordance with this, in Experiments 21, 22, 23 and 24, Table V,
a series of tests, A, B, C and D, each consisting of 50 grams liver tissue,
were prepared in the usual way. A and C were used as controls, no
addition of any kind being made. To B, 0°3 gram of asparagin was added ;
C and D were immediately boiled to destroy the enzyme, and were then

57

ON DEAMIDIZATION

Suryeys snonuyuoo gg G3 9-ST 6-E GS 9-ST OF GL 9-06 Fel
== — = == — GI I L:Zl = a = Ll
== = - = =< 9-0 6€-0 LE-L1 a am =F 86-91
= _ — — = oa = 73 Or $9-G $961 6-§1
G0 FI a ae a — = os a is eel GPL
¥-0 Gl re rs a —— = = ar ae a 9-81
— — a9 G tI oa =. = 9 GE G-CT al
= == = — = as a ae g-9 9° F:9G 8°3S
as — L- _ Gl 0% = ae == L-FI BL 9-93 8°81

$380} Yonoyy pelqqnq are —_— a) Z81 — = — 8-LI

$3893 YFnosyy poyqqna are = aad a 6-11 L-9 9°36 6-°C1

= — a = = — G-T TF L-LY ¥6°6 $S°GS 9-1

— = = = a a FG L-€1 6-11 L-9 8I eI

— — = a = ac 0 61 8°LZ SI (eo 61

= = = a a ae I 9T FG Or GG cI

a = cart. x aa = 5 <= T-L1 8 81 Or

— aa am :y iz: = fae a G-8 ¥ 91 al

aaely ulone’y jo ulone"T “OAD “OAT ‘dsy ‘dsy
mou gyds “oc uw ‘NOL yO woysqyds “Bog¢ur jon EL jJo woyqyds “Boo wu JO'N'L Jo wowygyds “3 oe at “3 Og ut

N°HN “1¢ N°HN “SIV 93¥}U9010d N°HN 3 N°HN “SIV 93 vqu0000g N*HN “SW N*HN “SIX oBvquooeg N°HN “3 N°HN ‘SI NHN “SIV
NINVIYV NIoaa’'T TIOOODATYY NIOVUVASY TOULNOD

Al 1aVJ,

“AVAL [VNSN OY} UL poyeqnour $4804
UL PUNOF osoyy jo OMOS ULYZ J0}VOIS JOU ov VIMOUIUIR poPeIogIT OY IO poureyqo
SON[VA oY} | UOLqnoUL SuLNp A[snonuyUoo UOYLYS O10M QP UL 84809 OYT.—"A'N

are

Cano

HOO oD <H CI OD OD SH OS ON OD OD

rts

W%

it

$z
z
z

aor
eq LOUL
sinoy{

(4)
(p) o1Q0R, + JOATT XO

(2)

JOA] XO
JOATT XO
a7 ‘é
“ee “e
JOA] Bop Buno x
eee .

“e «

“* goaty daayg

plow

IOATT xO
poorq
“qyop *o°0 GZ
(1) TAT, XO
(Z)
poorq “qyep

(1) ‘0°70 0g + JaaTT XO

IOATT XO
JOATL HQqLY

‘TI
‘Or

58 BIO-CHEMICAL JOURNAL

allowed to cool; A and B were incubated three hours. From each test
2 c.c. fluid were removed for the inoculation of C and D. A and B were
then immediately worked up in the usual way.

After inoculation, C and D were incubated for 14 hours to give the
bacteria a start; 0°3 gram asparagin was then added to D, and a further
incubation of three hours followed, at the end of which time the tests
were worked up. |

Here there is evidence of a deamidization of asparagin due to
bacterial action alone, the largest amount of ammonia liberated being
2 mgm. in Experiment 22; the smallest amount set free, 0°2 mgm., is
found in Experiment 24. On the other hand, the combined action of
enzyme and bacteria liberated 6 mgm. ammonia nitrogen in 23, but only
17 mgm. in 24.

Possibly the conditions were made too favourable for the bacteria in
22, they had 13 hours’ incubation before the addition of asparagin was
made. As a matter of fact, in Experiments 20 and 24, gelatine cultures
taken from each test at the end of incubation showed colonies on the third
day in the case of the C and D tests and not until the fifth day in that
of the A and B tests.

In Experiments 23 and 24, two additional tests, E and F, were
prepared. They were treated in exactly the same way as C and D, with
this exception, that they were not heated to 100° C., i.e., they were
allowed to stand at laboratory temperature, and were inoculated at the
same time and in the same way as C and D, the result being that they
contained active enzyme plus a double dose of bacteria. The addition of
asparagin to F was made immediately before incubation—FE served as a
control. Positive and abundant gelatine growths were obtained from each
test of the four series at the end of this incubation period.

The results of these experiments show that the bacteria present are
only responsible for a minor part of the ammonia liberated from
asparagin.

In connection with bacterial deamidization, it is interesting to note
that when two gelatine cultures of three days’ growth (obtained at the end
of incubation from the control and asparagin test, respectively, in
Experiment 15, Table IV) were each incubated for two hours with
0°15 gram asparagin; 3°S mgm. ammonia nitrogen were liberated from the
first, and 3°6 mgm. from the second culture. Unfortunately the absence of
a third culture as a control, to which no addition of asparagin had been
made, leaves it uncertain whether the whole of the ammonia nitrogen

ON DEAMIDIZATION 59

liberated came from the asparagin. The technique was the same as 1n the
case of an organ emulsion.
TasLe V

CONTROL | ASPARAGIN

Hours Mg. NH,N Mg. NHsN Mg. NHN Percentage
incubation in50g. | in 50g. split from of T.N. of
As

Asp. Asp.
liberated
21. Sheep liver, 50 g., Enzyme intact, 3 8-4 13-7 5-3 18-9
A and B
Bacteria only present, Enzyme 3} 9-5 10-6 ll 3°8
destroyed, C and D
24. Sheep liver, Enzyme intact, A and B 3 15-8 17-9 2:1 75
Bacteria present, Enzyme destroyed, 3 11-2 il-4 0-2 0-7
C and D
Bacteria present, Enzyme intact, 3 148 | 165 1-7 6
E and F
22. Ox liver, Enzyme intact, A and B 3 11-7 | 148 3-1 il
ae ra destroyed, C and D 3 10 ie b 2 7
23. Ox liver, Enzyme intact, A and B 3 99 | 15-9 6 21-4
pa = destroyed, C and D 3 10-9 12-3 1-4 5
3 152 € 18:9 3-7 13-2

we - intact, E and F

In view of the facts shown in Table V, bacterial deamidization has
been ignored in the subsequent investigations, which have been almost
entirely carried out with fresh organ emulsion incubated without the
presence of any antiseptic for two to four hours. The results detailed in
Table IV obtained in this manner are, in my opinion, of greater value in
considering intravital metabolic processes than those obtained from
antiseptic organ emulsions at the end of one or more weeks.

Other Factors

Oxygen Supply. It was thought that a better supply of oxygen to the
tissues might increase deamidization activity, and in tests 26 and 28
(Table LV) air was bubbled through the emulsion during incubation (the
air leaving the digest flask was passed then through decinormal acid to
collect any ammonia liberated). No appreciable increase of deamidization
was observed.

Presence of Lactic Acid. In view of the results of Schryver,?? who
demonstrated, when working on the autolysis of the liver, that there was
an increase of rapidity of breakdown (increased activity of ferment) if the
tests were acidified with lactic acid, Experiment 38, Table IV was carried
out using the usual methods except that the tests were acidified to the
extent of 0°1 per cent. with lactic acid. There was no increased
deamidization.

60 BIO-CHEMICAL JOURNAL °*

Optical Activity. As Abderhalden?! found that organ extracts and
pancreatic juice only liberated from certain racemic polypeptides the
optically active amino acids which occur naturally in protein, the
question of optical activity had to be considered. In this connection it
is interesting to note that the expressed juice of liver, placenta and
kidney, according to Bergell and Brugsch,?? can cleave dl leucinamid
and dl alanyl-alanin into the active components. They find that this
is not due to autolytic activity, for the fresher the juice the more active
is the cleavage.

The material I used was an abiuret pancreatic digest, very kindly
given to me by Dr. Cathcart. As a considerable amount of ammonia was
present in this digest, it was neutralised with acetic acid and to obviate
the necessity of making an ammonia estimation in it, an amount of this
digest, containing 0°296 gram nitrogen (Kjeldahl), was added to Test 39
(Table VI) before digestion, and an equal amount to the control test
at the end of incubation. The result obtained showed but a very small
increase of ammonia.

A second experiment (46) was carried out with a freshly prepared
casein digest, which had been digested with pancreatin for a month.
Before use the digest was precipitated by means of phosphotungstie acid,
the excess of the precipitant being removed with barium carbonate.
100 c.c. of the filtrate thus obtained, containing 10 mgm. ammonia
nitrogen and 170 mgm. T.N. was then added to 50 gram ox liver. As
this test contained at the end of incubation only 9:1 mgm. of ammonia
nitrogen more than the amount of ammonia nitrogen in the control test,
it is evident that there has been no deamidization of the digest.

TaBLe VI
CoNnTROL DIGEST
Hours of Mg. NH,N in Mg. NH3N in Mg. NH3N
incubation 50 g. 50 g. liberated
39. Ox Liver... ee 4 91 96-2 5-2
46. Ox Liver ... as 3 11-7 20:8 9-1
3 11:6 lost —

Il. Intestinal Mucosa

A consideration of the following facts led to an investigation of the
deamidizing action of the intestinal mucous membrane. Nencki, Pawlow
and Zaleski found the ammonia content of the portal blood constantly
higher than that of arterial blood. They found also an increased amount

.
|

ON DEAMIDIZATION 61

of ammonia in the intestinal mucosa after flesh feeding compared with
the amount present at the end of two days’ starvation. These results
were confirmed by Horodynski, Salaskin and Zaleski.2? Cohnheim*4
found an increase of ammonia at the end of six hours’ digestion in the
blood bathing a fish’s intestine, which he had previously filled with a
solution of an amino acid (aspartic acid), proving that a deamidization
had occurred. He obtained the following figures :—

9mgm. ammonia were liberated from 0°25 gram aspartic acid; and

15 mgm. ammonia were liberated from 0°23 gram lysin chloride.

This would suggest that an active deamidization had taken place in
the intestinal mucosa.

Lang carried out three experiments with ox intestinal mucosa, with
the results shown in Table I.

I carried out a series of experiments detailed in Table VII. In
Test 12, 50 grams sheep intestine, incubated antiseptically for seven
days, liberated 17 mgm. ammonia nitrogen from asparagin, and 2 mgm.
ammonia nitrogen from glycocoll. 50 grams of sheep liver hberated a
similar amount of ammonia nitrogen from asparagin and glycocoll in
nine days.

In the aseptic tests with mucous membrane there is always a small
amount of ammonia nitrogen liberated from asparagin, the maximum
amount in three hours’ incubation being 2°8 mgm. Contrasting this
with the amount liberated by 50 grams liver in the same time, we find, in
Experiment 15, Table IV, a maximum of 9:94 mgm. The deamidization
of asparagin in the aseptic tests carried out in the manner already
described is more active apparently in the case of liver than in that of
intestinal mucosa. ct

Possible co-ferment. The possibility of a co-ferment action was
considered in the same series, 25 grams liver and 25 grams mucous
membrane being incubated together. There is, perhaps, evidence of a
slightly increased activity in the mixed test with asparagin, but it is not
enough to postulate the existence of a co-ferment.

It was thought that a more active deamidization might be found at
the height of digestion, and for this reason, in Experiment 40,
Table VII, a young dog was killed seven hours after a large meat meal.
The intestinal mucous membrane was scraped off the muscular coat and
finely chopped. It was then incubated with a mixture of 50 c.c.
defibrinated blood and 100 c¢.c. Locke’s solution, and air was
bubbled through the tests during the four hours of incubation; 4°7 mgm.

une'T
moi; qyds -“So¢ ut

€-0 1G-U 16°C]
T-¥ LF 6E
ae 8-0- €-11
G-€ G £6
*oAPD
JO "NL Jo rds 3 OG Ul

PL

L-09
‘dsy

6-06

LI

‘dsy

L:96
G 1G
&<l
Gs
L-¥1

6-OP

8t

L-Gt
G96
6-61
EVE
66

PGI
1-él

66

IG

Ce eae aD “+ buuspog
£ burjsv] “3 Gz “

1g paf “
5 gees: ** paf—fddng
Cre ‘8 gg dasyg
€ (@)
€ (1) ‘og a3 x0
sino —ondosy
61 (3)
ondosyuy

L (1) ** ‘3 og desgg

“EP
“GP
“IP
‘OF
“LE

66

“GI

BIO-CHEMICAL JOURNAL

62

c JO'N YL JO wmoyyyds ‘s Qg ur “6 Og aI sivgq
N°HN “20 N®HN “SI 95R}0U00I0g N®HN “SW N®HN SIV esvyue.eg N®HN “SIT N°HN “SIA N®FIN “SIT

NIOONWT TIODOOA TY) NIOVUVdSYV TOULNOD VsOO0] IVNLLSAYLN]

IIA @1avy,

phy |

ON DEAMIDIZATION 63

Tasre VIII .
CONTROL ASPARAGIN GLyYCcocoLL
Sheep Incuba- Mg. NH,;N Mg. NHsN Mg.NHsN Mg. NHsN Mg. NH3N
ted in50g. ind50g. splitfrom in 50g. split from
Aap. Glyc.
Gut + Liver aa 25¢. ... 7 26 48 22 26 —
Antiseptic—
Gut 50 g. EOS 7 21 38 17 23 2
Liver 50 g. ee ee 9 35 52 17 37 2
Gut 50 g. 7 ae 19 29 49-9 20-0 29-4 0
Hours
Sheep Tneuba- LEUCIN
ed
Aseptic—
(37) Gut + Liver aa 25 g.... + 22-4 28 5-4 23-2 0-8
Gut 34¢. 2 9-9 12:3 2-4 _— =
Liver 59 g. 2 14-2 — — 19-4 6-2

ammonia nitrogen were liberated from glycocoll. Unfortunately the
amount of mucous membrane was only sufficient for two tests—the control
and the glycocoll test.

In Experiment 41 two young dogs were killed five hours after a meat
meal, whilst Experiments 42 and 43 were carried out with similar
material obtained twenty-five hours and twenty-seven hours, respectively,
after a meat meal.

As a result of these experiments it was found that under the
conditions present in one case (Experiment 40) deamidization of glycocoll
is certainly greater at the height of digestion, but the experiments are
too few in number to dogmatise upon. The extraordinarily large amount
of ammonia present in the control of Experiment 40 is in accordance
with the known facts as to the large ammonia content of the intestinal
mucosa at the height of digestion.

Deamidization in various Species and in Foetal States. As metabolism

~ varies in certain particulars in different species of animals, the question

may be asked whether any marked differences are found in the
deamidizing capacity of the liver of the various animals used, namely,
dog, rabbit, sheep and ox. The experiments were not carried out with
this definite point in view, but on looking through the tables it will be
seen that the most active splitting of glycocoll occurs in the sheep
(Experiments I and VII); no other definite variations are noticed.

It is interesting to note that in the foetal livers (Experiments 5, 7
and 21, Table IL) of varying ages examined, quite an active splitting of
asparagin has occurred. The presence of tissue enzymes in the embryo

64 BIO-CHEMICAL JOURNAL

is a well-established fact. Amongst other workers in this field may be
mentioned Mendell and Leavenworth,2> who carried out a series of
investigations on embryo pigs. They found an active autolytic enzyme in
the liver, and demonstrated the early presence of lipase in liver and
intestine, although the activity of the latter was less pronounced than in
the full-grown animal.

In considering the question of how far the amount of ammonia
indicates the extent of deamidization, there are three possibilities as to
the fate of the liberated ammonia : —

(1) It may be converted into urea in the liver tests.

(2) It may be fixed in some way by the tissue.

(5) It appears as such.

The conversion of ammonia into urea by the liver in vivo is a well-
established fact; in the case of organ emulsion, however, added ammonia
only disappears to a limited extent. For example the following :—
Table IX shows that in Experiments 44, 45 and 47, 3°54, 6°52 and 4 mgm.
ammonia nitrogen (in percentage amount 12, 23 and 8), respectively,
disappeared during incubation for three hours with 50 grams liver tissue.

TABLE IX

ConTROL AMMONIA

Hours Mg. NH3;3N Mg. NH3N Mg. NH; Percentage
‘ncubated in 50 g. in 50 g. not of T.N. not

recovered recovered

44. 50g. Ox Liver 25 c.c. 3 12-4 36-68 3°54 12
Ammon. Acet.
Solution added to one test
containing 27-82 mg. N.

45. 50g. Ox Liver + 25 c.c. 3 12-2 33°5 6:52 23
defibrinated blood
Ammon. Acet. = 27:82
mg. N.
47. 50g. Ox Liver, Ammon. 3 13 59 4 8

Acet. = 50 mg. N.

Further insight into the fate of ammonia in these digest experiments
could only be gained by quantitative estimations of urea or of nitrogen
content of an alcohol extract carried out at the end of incubation.
Difficulty was found in extracting the evaporated filtrate obtained with
the alcohol-ether mixture or amyl alcohol, as the addition of these at
once made the residue very sticky, rendering satisfactory extraction
impossible. Folin’s method could not be used owing to the presence of
considerable amounts of sugar in the organ extract. The same objection
applies to Pfluger-Schéndorff’s method. The second possibility as to the

|

ON DEAMIDIZATION 65

fate of ammonia is rendered not unlikely by the extremely interesting
work of Kowalevsky and Marchewicz?® on the capacity of the tissue to
remove an excess of ammonia from the blood. Both points are of extreme
interest and importance in connection with the present research, and are
now being further investigated. The results will be given in a later
communication.

With regard to the amino acids investigated in the present
experiments, it is quite evident that more ammonia is liberated from
asparagin than from any of the others. Glycocoll is split to a
considerable extent in the antiseptic tests, but only to a small extent in
the aseptic tests. Leucin and alanin were used only in the aseptic tests;
no ammonia was ever found to have been liberated from alanin, but small
amounts were set free in some of the leucin tests. It is quite evident
that the amide nitrogen attached to the carboxyl group of asparagin is
more readily liberated than the amide nitrogen occupying the a position
to the carboxyl group of the other amino acids and of asparagin itself.

Having considered the effect of various organ emulsions on amino
acids with regard to deamidization, it is of interest to observe the
behaviour of these amino acids in the living organism itself. How soon
and to what extent does their nitrogen appear in the urine as urea?

Levene and Kober?’ fed a dog on a standard diet of low nitrogen
value. When they added glycocoll to this diet they found that its
nitrogen was removed from the body in the form of urea within twenty-
four hours. Dr. Cathcart, in an investigation (not yet published) on the
rate of absorption of digest products and simple amino acids from the
small intestine, found also a very rapid excretion of the nitrogen, mostly
as urea.

To carry out a somewhat similar experiment, a dog of 21'900 kilos
was fed for some days on a diet consisting of 300 grs. meal and 500 e.c.
milk, representing a caloric intake of 50 calories per kilo. On the four test
days, which were separated by intervals of a few days, the bladder was
emptied by catheter at 9.30 a.m. The urine was next collected by
catheter at 10.30 a.m., and thereafter at intervals of two hours until
6.30 p.m.

On two test days 200 c.c. of water were given alone at 10.30 a.m.,
by means of the stomach tube. On the other two test days 10 grams of
glycocoll and 10 grams of asparagin, respectively, dissolved in 200 c.c.
of water, were given in a similar manner at 10.30 a.m. On all four days

66 BIO-CHEMICAL JOURNAL

the animal fasted. The distribution of total nitrogen ammonia and urea
in the samples of urine thus obtained was investigated. The total
nitrogen was estimated by Kjeldahl’s method, the ammonia and urea by
Folin’s methods.

The results are detailed in Tables X and XI.

Table XI gives the total amount of urea nitrogen excreted in the
eight hours following the ingestion of the 200 c.c. of water alone. The
mean of this amount for the two fast days is subtracted from the amount
of urea similarly excreted on the glycocoll and asparagin day
respectively. In this way it is found that 96 per cent. of the ingested
glycocoll nitrogen in the present experiment is excreted within eight
hours in the form of urea, while 63 per cent. of the ingested asparagin
is similarly excreted within eight hours. ‘This result is in marked
contrast with the results of the 7m vitro experiments, as in the latter
ammonia is liberated more rapidly from asparagin than from glycocoll,
whereas in the living organism glycocoll is apparently metabolised more
rapidly than asparagin.

TABLE X
Vol. of NH;N Urea N Total N
Urine gms. gms. gms.
Fast Day I—
10.30—12.30 ... 95/100 0:06 0-386 0-578
12.30— 2.30 ... 40/60 0-032 0-199 0-231
2.30— 4.30 ... 70/80 0-033 0-419 0-532
4.30— 6.30 ... 20/50 0-028 0-291 0-353
0-150 1-295 1-694 T.N. excreted within 8 hours
— after ingestion of 200 c.c.
Glycocoll Day— water at 10.30 a-m.
10.30—12.30 ... 38/110 0-032 0-747 0-856
12.30— 2.30 ... 40/100 0-026 0-910 0-988
2.30— 4.30 ... 28/100 0-029 0-766 0-855
4.30— 6.30 ... 22/100 0-029 0-709 0-781

0-116 3°132 3-480 T.N. excreted within 8 hours
SS after ingestion of 10g.
Glycocoll in 200 ¢.c. water

Asparagin Day— at 10.30 a.m.
10.30—12.30 ... 60/100 0-031 0-453 0-606
12-30— 12:30) <.. 62/100 0-026 0-589 0-758
2.30— 4.30 ... 28/100 0-022 0-519 0-627
4.30— 6.30 ... 26/100 0-021 0-501 0-562

0-100 2-062 2-553 T.N. excreted within 8 hours
SS after ingestion of
Asparagin in 200 c.c. water

Fast Day Il— at 10.30 a.m.
10.30—12.30 ... 85/100 0-022 0-389 0-434
12.30— 2.30 ... 45/100 0-017 0-344 0-383
2.30— 4.30 ... 44/100 0-016 0-335 0-399
4.30— 6.30 ... 23/100 0-007 0-280 0-305
0-062 1:348 1521 ‘T.N. excreted within 8 hours

a atnnnEEE EERSTE EEE EEE after ingestion of 200 c.c.
water at 10.30 a.m,

ON DEAMIDIZATION 67

TaBLE XI

Nitrogen eliminated in 8 hours following ingestion of water
Urea in Percentage Gms. N in

NH;N Urea N Resid. N Total N in excess of 10 gms.
offast ingested aminoacid
days N added
Fast Day 1 w+ 0150 1-295) 1 99) (0-249 1-694 a = =
Fast Day 2 . 0062 1-348) °°" (O-111 1-521 = a fae
Glycocoll Day... O11G 3-132 0-232 3-480 1-811 96 1-86
Asparagin Day ... 0-100 2-062 0-391 2-553 0-741 63 1-2

CONCLUSIONS

In agreement with Lang, it has been found that emulsions of liver
and intestinal mucosa, when incubated with asparagin, glycocoll, leucin,
or alanin, liberate ammonia except in the case of alanin, and that the
total amount of ammonia so liberated is much greater in the case of
asparagin than in that of glycocoll or leucin.

On the other hand, the presence of an antiseptic does not exert the
marked inhibition of the deamidizing enzyme claimed by Lang, 1.e., the
fresh organ, incubated without an antiseptic, shows no striking increase
in its deamidization of glycocoll.

There is a marked contrast between the fate of the nitrogen of
asparagin and glycocoll, respectively, 7m vivo and in vitro. In the living
organism the nitrogen of glycocoll appears practically quantitatively at
the end of eight hours in the urine as urea, while only about 63 per cent.
of the nitrogen of asparagin so appears in the same time.

In vitro, however, ammonia is liberated more quickly and in larger
amount from asparagin than from glycocoll, i.e., in vitro the amide
nitrogen attached to the carboxyl group is more readily liberated than the
amide group in the a position.

Altogether the results of the investigation indicate that the method
of studying the deamidizing action of the tissues adopted by Lang is by
no means representative of the intravital changes in the body, and that
far-reaching conclusions based upon them are not justified.

In conclusion, I wish to express my best thanks to Professor Noél
Paton for much helpful advice, and to Dr. E. P. Cathcart, at whose
instigation and under whose direction the research has been carried out.
The work was begun during my tenure of a McCunn Scholarship.

rf

68

—_— —
FP Sewmy ane wo

—
hme GS)

rp po bt
bo

—_
.

BIO-CHEMICAL JOURNAL

REFERENCES

Howell, Amer. Journ. Physiol., Vol. XVII, p. 273, 1906.

Michaud, Zeit. f. physiol. Chem., Vol. LIX, p. 405, 1909.

Abderhalden and Terruchi, Zeit. f. physiol. Chem., Vol. XLVII, p. 159, 1906.
Stollte, Hof. Beit., Vol. V, p. 15, 1904.

Salaskin, Zeit. f. physiol. Chem., Vol. XXV, p. 125, 1898.

Flatow, Zeit. f. physiol. Chem., Vol. LXIV, p. 267, 1910.

Friedmann, Bioch. Zeit., Vol. XXVII, p. 97, 1910.

Neubauer and Fischer, Zeit. f. physiol. Chem., Vol. LXVII, p. 230, 1910.
Abderhalden, Bloch and Rona, Zeit. f. physiol. Chem., Vol. LII, p. 434, 1907.
Neubauer, Otto and Falta, Zeit. f. physiol. Chem., Vol. XLII, p. 81, 1904.
Jacoby, Ziet. f. physiol. Chem., Vol. XXX, p. 149, 1900.

Loewi, Zeit. f. physiol. Chem., Vol. XXV, p. 511, 1898.

Gonnermann, Pfl. Archiv., Vol. UXXXIX, p. 493, 1902; Vol. XCV, p. 278, 1903.
Lang, Hof. Beit., Viol eet VA

Furth and Friedmann, Bioch. Zeit. Vol. XXVI, p. 435, 1910.

Shaklee and Meltzer, Amer. Jour. Physiol., Vol. XXIII, 1908; Vol. XXIX.
Kruger and Reich, Zeit f. physiol. Chem., Vol. XX XIX, p. 165, 1903.

Grafe, Zeit. f. physiol. Chem., Vol. XLVIII, p. 300, 1906.

Brasch, Bioch. Zeit., Vol. XXII, p. 402, 1909.

Schryver, Bioch. Jour., Vol. I, p. 123, 1906.

Fischer and Abderhalden, Zeit. f. physiol. Chem., Vol. XLVI, p. 52, 1905.

Bergell and Brugsch, Zeit. f. physiol. Chem., Vol. LXVII, p. 97, 1910.
Horodynski, Salaskin and Zaleski, Zeit. f. physiol. Chem., Vol. XXXYV, p. 246, 1902.
Cohnheim, Zeit. f. physiol. Chem., Vol. LIX, p. 239, 1909.

Mende!l and Leavenworth, Amer. Jour. Physiol., Vol. X XI, pp. 67 and 95, 1908.
Kowalevsky and Marchewicz, Bioch. Zeit., Vol. IV, p. 196, 1907.

Levene and Kober, Amer. Jour. Physiol., Vol. XXIII, p. 325, 1909.

Levene and Meyer, Amer. Jour. Physiol., Vol. XXV, p. 214, 1909.

69

NOTE ON THE ESTIMATION OF POTASSIUM IN URINE
By H. H. GREEN, B.Sc., Physiological Laboratory, Glasgow University.
Communicated by Dr. EB. P. Cathcart

(Received June 9th, 1911)

The purpose of this note is to direct attention to the method of
estimating potassium by precipitation as the double cobalti-nitrite of
sodium and potassium, K,NaCo(NO,), ’ H,O—a method useful, but
apparently little used, especially in physiological work. Its advantages
over the platinic chloride process are the greater rapidity, the shorter
working time involved, and the very much lower cost. In point of
accuracy it may, in ordinary practice, be made equal.to the platinum
method. Its special feature lies in the fact that the presence of sodium
does not interfere, since the precipitant is itself a sodium salt. Salts of
other metals do not in general interfere, and hence a troublesome
preliminary separation is avoided. Further, the difficulty of separating
the mixed chloroplatinates of sodium and potassium by washing with
alcohol is well known, and the results obtained by the platinum method
as usually conducted may well be high, owing to occlusion of Na,PtCl,,
which cannot be separated by extraction with alcohol, however long
continued. The method of obviating this error by re-dissolying the
K,PtCl, in water and re-precipitating by evaporation is troublesome.

The first reference to the cobalti-nitrite reaction appears to be the
discovery by Fischer! in 1849 that on adding a solution of potassium
nitrite to a cobalt salt, an orange coloured double compound is precipitated.
Fischer found that a cobalt salt of concentration so dilute as 1 in 3,000
was apparently completely precipitated on the addition of potassium
nitrite. Amongst other workers who have investigated the reaction may
be mentioned Erdmann?, Sadtler*, de Koninck*, Curtman5, and Biilman®.
Macallum’ in 1905, in a paper upon ‘ Potassium in cells,’ gives a full
Poggen, Annalen, LX XIV, p. 124, 1849.

J. pr. Chem., XCVII, p. 385, 1866.

Amer. J. Sci., 2, XLIX, p. 189, 1870.

Zeit. fiir annal. Chem., XX, p. 390, 1881.
Berichte d. ch. G. Jahrg., XIV, p. 1951, 1881.

6. Zeit. fur annal. Chem., XXXIX, p. 284, 1900.
7. J. Physiol., XXXII, p. 95, 1905.

oe eh

70 BIO-CHEMICAL JOURNAL

discussion of the reagent, and uses the method with satisfactory
results for histological purposes. The fact that ammonium compounds
react with sodium cobalti-nitrite suggested the possibility of amino and
amido compounds reacting similarly. On investigation, however,
Macallum finds that neither glycin, leucine, tyrosin, aspartic acid,
eglutaminic acid, sarcosin, nor glucosamine, precipitate and that therefore
their presence in tissues presents no difficulty. Neither do urea, asparagin,
alloxan, allantoin, guanidin, and the purin bodies react. Although
creatinine does not react, creatine does so readily, giving a precipitate
similar to that of potassium.

The first quantitative method of importance appears to be that of
Adie and Wood’ in 1900. The reagent as prepared by them, and now in
general use, is made as follows :—113 grams of cobalt acetate are dissolved
in 400 c.c. of water, and 100 c.c. of glacial acetic acid added. 220 grams
of sodium nitrite are separately dissolved in 400 c.c. of water, and the
two solutions are then mixed with stirring, the last of the NO, evolved
being removed by evacuation over night or by blowing air through the
mixture. After standing at least twenty-four hours any sediment is filtered
off and the solution made up to a litre. It will be found that the bulk is
approximately a litre without the addition of water, and it is perhaps
best not to dilute further. The reagent prepared in this manner is of
a dark plum colour. It should be stored in the dark and decanted for
use as required.

Adie and Wood describe both volumetric and gravimetric methods
of determining potassium in soils and manures.

The precipitation is accomplished by adding 10 c.c. of the reagent
and 1 c.c. of glacial acetic aid to 10 c¢.c. of the solution to be determined,
which must be so arranged as to have a concentration of 0°5 to 1 per
cent. K,0. Adie and Wood found the most important factor in the
precipitation to be the state of dilution—at 0:1 per cent. concentration
of K.O, only two-thirds of the theoretical amount of potassium was
recoverable. They determine the formula of the precipitate as obtained
by them under standard conditions as K,NaCo(NO,),°H,O, and its
solubility in water as 1 in 20,000.

Autenreith and Bernheim? in 1902 described a method for the
determination of potassium in urine, and so far as I have ascertained,
this represents the only instance in which the use of the cobalt method

8. J.C.S., IL, p. 1076, 1900.
9. Zeit. physiol. Chem., XX XVII, p. 29, 1902.

_——_

ESTIMATION OF POTASSIUM IN URINE 71

in the analysis of physiological products 1s recorded. ‘Their method
consisted in treating 50 c.c. of the filtered urine direct with 6 to 10 c.e:
of the reagent, shaking, allowing to stand, and determining the potassium
in the precipitate by subsequent conversion into the perchlorate. This
method, however, is tedious, and it was determined to seek a shorter
method. A direct treatment of the precipitate without conversion into
the perchlorate does not give satisfactory results.

Drushel!® in 1907 published a modification of Adie and Wood's
method in which the reagent was added to the solution to be estimated,
and the mixture evaporated almost to dryness. The precipitate was then
extracted, washed with water, and decomposed directly with standard
potassium permanganate. The reaction involved in treating the
precipitate, K,NaCo(NO,), ‘ H,O, with permanganate, is one of oxidation
of the nitrite to nitrate and concomitant reduction of trivalent cobalt
to the divalent form.

The method of dealing with the precipitate which I have adopted
for physiological purposes is in essence that of Drushel. The process as
carried out on urine is as follows :—

25 c.c. of the urine is evaporated to dryness in a platinum dish, and
gently ignited at a dull red heat. The ash is then moistened with nitric
acid and reignited until free from organic matter and ammonium salts.
The use of nitric acid allows of easy ignition at a temperature sufficiently
low to prevent any loss of potassium by volatilisation. The ash is then
dissolved up in water containing a few drops of hydrochloric acid and
washed into a porcelain basin (200 ¢.c. capacity). The solution is then
neutralised or rendered slightly alkaline with a few drops of caustic
soda, acidified with acetic acid, and evaporated to between 5 c.c. and
10 c.c. 1le.c. of glacial acetic acid, and 10 c.c. of the reagent (prepared
according to the recipe of Adie and Wood, Joc. cit.) are then added, and the
evaporation is continued until the resulting syrup is of such consistency
as to set to a stiff hardish mass on cooling. The time taken for evaporation
is rather less than an hour, but the stage of evaporation is readily fixed
after a little practice.

The cold mass, consisting of the yellow precipitate, K.,NaCo(NO,),
‘HO, imbedded in a dark purplish-brown matrix, is then worked up
with about 50 ¢.c. of 10 per cent. acetic acid and allowed to soak (stirring
up occasionally with « glass rod) until the yellow precipitate is

10. Amer. J. Sci., XXIV, p. 433, 1907.

72 BIO-CHEMICAL JOURNAL

~

disentangled and the rest of the mass completely dissolved up to give the
clear tawny colour of the dilute reagent. The yellow precipitate is then
washed by decantation with dilute acetic acid, the washings being poured
through an asbestos filter. Finally, the precipitate is transferred completely
to the filter, washed with cold water, and sucked dry. In general, six
washings by decantation are necessary, using in all about 300 c.c. of
wash-liquid. From a mechanical point of view, the precipitate is easily
washed and filtered, but it is somewhat difficult to completely wash free
from the excess of reagent used. A funnel with a Gooch disc well
packed with clean coarse Gooch-asbestos, makes a convenient filter.
If any trouble is experienced by sucking through of the precipitate
taking place, a second packed disc above the first ensures complete
retention, _

While the filtering of the precipitate is in process, about 300 c.c. of
water is raised to boiling and the requisite excess (30 c.c. to 50 e.c. N/5)
of potassium permanganate run in from a burette. The pad of asbestos
with the precipitate is picked off the disc and dropped into the hot
permanganate, any precipitate adhering to the funnel being also carefully
washed in. The precipitate is stirred up, and after a few minutes dilute
sulphuric acid, in amount required to react with the permanganate, 1s
added (20 c.c. of 25 per cent. sulphuric acid). The contents are well
stirred, replaced over the bunsen and kept at, or just below, the boiling
point until the yellow precipitate is completely decomposed (about -
five minutes). Darkening of the liquid with formation of manganese
hydroxide begins before the addition of sulphuric acid and continues
on heating. Standard oxalic acid is then run in from a burette, with
stirring, until the manganese hydroxide is dissolved and the hot liquid
is clear and of a very faint rose pink colour. Care must be taken that any
manganese hydroxide adhering to the glass and asbestos is completely
dissolved up. The excess of oxalic acid is then titrated back to colour
with permanganate, the end-point being sharply marked. The difference
between the total permanganate and oxalic acid burette readings gives
the amount of permanganate used, from which the amount of K,O is
calculated. The theoretical factor for N/5 potassium permanganate is
1 cc. = 0001714 grams K,O. It is, however, preferable to standardise
the permanganate accurately against a solution of pure fused potassium
chloride, using the equivalent so found—by this means any slight
constant errors are embodied in the standard, if the procedure adopted in
standardisation be rigidly adhered to in all subsequent determinations.

ESTIMATION OF POTASSIUM IN URINE 73

Calculation is obviously saved if the standard permanganate and oxalic
acid be made up so that 1 ¢.c. = 0-001 grara K,0; or, if more convenient,
Le.c. = 0:002 gram KO. This latter concentration is a little over N/5 and
10 cc. of a 1 per cent. solution of KC] used in standardisation then
requires 316 c.c. of permanganate—a convenient burette reading on a
convenient working quantity of precipitate.

In order to investigate the influence of other factors which have been
said to interfere with the accuracy of the results, a series of a hundred
determinations was made in which various foreign salts were added to
the same quantity of potassium chloride solution. Quantities of sodium
chloride, sodium phosphate, calcium chloride, calcium _ nitrate,
magnesium chloride and magnesium sulphate, in amounts varying
between wide limits were found not to seriously influence the results
—in most cases the error was below 0'4 c.c. in a titrate of 32°6 c.c. of
permanganate, i.e., the determination was accurate to about 1 per cent.
The general conclusion is that it is unnecessary to first remove such
foreign salts unless they are present in amount many times greater than
the potassium itself. For practical purposes, the amount of salts other
than those of potassium likely to be present in the ash of urine, may
be ignored. Potassium was found to be estimable with equal accuracy
in the form of chloride, sulphate, nitrate and acetate, provided that
no free acid other than acetic was present during evaporation with the
sodium cobalti-nitrite reagent. The quantity of reagent used was found
to be of little consequence provided that large excess was taken.
Theoretically 2 c.c. of the reagent is rather more than sufficient to
precipitate the 10 c.c. of 1 per cent. KCl (= 0:0632 gram K,O) used in
standardisation. The 10 c.c. of reagent used therefore represents the
necessary large excess. The colour of the filtrate from the washed
precipitate should be the characteristic colour of the dilute reagent—
a pink filtrate indicates decomposition, and estimations showing such
should be repeated, using a larger excess of the reagent. The concentrated
reagent itself, if stored in the dark, may be kept for a year, but the
fresh reagent is always preferable. In sunlight, decomposition is more
rapid. If diluted and exposed to light it quickly goes pink, indicating
rapid decomposition. After the ‘thick paste’ stage of evaporation
in the estimation, further heating should be avoided as the precipitate
then becomes more difficult to extract, and results may be accordingly
vitiated.

To test the accuracy of the method exactly as it is conducted upon

74 BIO-CHEMICAL JOURNAL

urine, an artificial urine (free from potassium) was made up of the

following composition permictire——

Urea 30°0 grams

NaCl 12:0 ss

iu 3 beg ©) eens nS

MgSO, O13 es

CaCl, eater Ve

Na. HPO; oi);

Glucose ORO oy (to simulate difficulty of ignition)

—the solution being acidified te keep the phosphates. in solution.

20 c.c. of this solution plus 10 c.c. of a one per cent. solution of
KCl were evaporated to dryness, ignited until free from ammonium
salts, and treated exactly as in the process described for urines. 10 c.e.
of the KCl solution alone required 524 c.c. of the standard permanganate
used. The following table shows, to the first decimal place, six successive
determinations in which the artificial mixture was present :—

No. KoMnsOx Oxalic acid Difference c.c. error Per cent. KCl
burette burette = KoMnoOxg used recovered
] 42-6 c.c. 10-0 e.e. 32-6 c.c. + 0-2 e.c. 100-6
2 41-6 9-3 32:3 —0-1 99-7
3 42-3 10-0 32:3 — 0-1 99:7
4 42-9 11-0 31-9 — 0-5 98-5
5 42-3 10-0 32:3 — 0-1 99:7
6 42-5 10-0 32°5 + 0-1 100-3

Two estimations, adopting the method of wet combustion with
sulphuric acid, removing the excess of sulphuric acid by evaporation
over a low flame to avoid spattering, and subsequently removing the
ammonium sulphate by igniticn, gave 100°3 per cent. and 99:4 per cent.
of potash recovered. ‘Tie cobalti-nitrite method is obviously applicable
to all physiological products after careful preliminary ashing to remove
organic matter and ammonium salts.

Where sodium as well as potassium is to be estimated, the ash is
dissolved up in dilute hydrochloric acid, and the foreign metals along
with sulphate and phosphate removed by treatment with barium chloride
and ammonium carbonate as in the ordinary preliminary treatment in
the platinic chloride process—the ‘ NaCl+KCl’ being weighed as such,
and the sodium found by difference after the estimation of potassium.

t

ESTIMATION OF POTASSIUM IN URINE Ti

~

Even at this stage great economy in time, labour and cost of material
is effected.

The cobalt method, as here described, has been used with satisfactory
analytical results (checked by comparison with the platinum method)
in a series of experiments upon the mineral metabolism of the dog—a
discussion of the results of which is reserved for a future date.

Norr.—Since the work in this paper was completed, I have discovered
a third paper by Drushel (Amer J. Sei., December, 1908) which had
unfortunately been previously overlooked. He adapted his method for
use on physiological fluids, and found it satisfactory.

The expenses of this investigation were defrayed by a grant made to
Dr. Catheart by the Carnegie Trustees.

76

ON THE FREEZING POINT OF THE UNHAEMOLYSED
CORPUSCLES DURING HAEMOLYSIS OF BLOOD
AND ON THE EFFECTS OF EVAPORATION ON THE
RESISTANCE OF ERYTHROCYTES TO HAEMOLYSIS

By U. N. BRAHMACHARI, M.A., M.D., Ph.D., Lecturer in Medicine
at the Campbell Medical School, Calcutta.

(Received June 10th, 1911)

In the Bio-Chemical Journal, volume IV, page 284, I have pointed
out that the resistant corpuscles during haemolysis of blood are either less
permeable to water or can bear the tension of distension better than those
that haemolyse.

In order to determine whether the unhaemolysed corpuscles are less
permeable to water or not, I have estimated the freezing point of the
unhaemolysed corpuscles as well as that of the supernatant fluid after
the separation of the unhaemolysed corpuscles by centrifugalisation.

In all of the experiments the blood was allowed to clot, the clotted
blood squeezed through thin muslin and mixed with two parts of distilled
water, and the mixture thoroughly centrifuged after one hour.

The following results have been obtained : —

(1) Homan Broop. (Blood taken from the jugular veins four hours after death.)

1. A for the haemolysed corpuscles sit oe = 7: es a» O85
2. A for the unhaemolysed corpuscles Sie dss sine So 5a -.. O-195

(2) Fowr’s BrLoop

1. A for the haemolysed corpuscles she ane ae a Son -. 0-224

2. <A for the unhaemolysed corpuscles sie a age a ee 277) 10:285
(3) Fow.’s Broop

1. A for the haemolysed corpuscles aes age ope ase one s+ 0-210

2. A for the unhaemolysed corpuscles ra se Bie sat ss ... 0-210
(4) Fowu’s Bioop

1. A for the haemolysed corpuscles see ae ane Ape 508 ... 0-210

2. A for the unhaemolysed corpuscles a ant ade = ae ... 0-200
(5) Fow1’s Boop

1. A for the haemolysed corpuscles 3s ae ie ae Ses -.« 0:235

2. A for the unhaemolysed corpuscles sas Sn0 “its aH “ae ... 0-208
(6) Fow.’s Broop

1. A for the haemolysed corpuscles oh 50H oe Sie ne ... 0-230

2. A for the unhaemolysed corpuscles Sa tes Bs vee ee ao ROSESO
(7) Fow1’s Bioop

1. A for the haemolysed corpuscles see Se5 B05 a aS = 0-250

2. A for the unhaemolysed corpuscles 0-250

PREEZING POINT OF UNHAEMOLYSED CORPUSCLES 77

It will be seen from the above that the freezing points of the
haemolysed corpuscles were nearly the same as those of the unhaemolysed
cerpuscles, and therefore osmosis must have taken place to the same
extent in the haemolysed as well as the unhaemolysed corpuscles. In
other words, the unhaemolysed corpuscles are nearly as much permeable
to water as the haemoylsed ones. The fact that some of the corpuscles do
not haemolyse must therefore be due to their being able to bear the tension
of distension better than those that haemolyse. We may describe this
resistance of the erythrocytes to rupture as their specific resistance. This
specific resistance to rupture varies in different animals, and may be
expressed by the relative haemoglobin value of their resistant erythrocytes.

In the following tables the erythrocytes were invariably suspended in
0°85 per cent. saline, and the amount of suspension taken was always half
that of the dissolving fluid, which in the present case was distilled water.

Specific resistance of erythrocytes to rupture in different animals

1. Human blood at Be Ey: ae soe sie Pe ae ... 0°2934
2. Fowl’s blood sf eee ae ace an af Se ee se OSS
3. Dog’s blood B52 sis ses ée ae a ao bac ... 03333
4. Frog's blood ae ee ise Sie sae ace _ 53 eee OTOL
5. Sheep’s blood a Ses Sor ae Sic ode 6c 50 ston MY
6. Rabbit’s blood aon Rae fee pee ot i ae i Sch al

Lffect of evaporation on the resistance of erythrocytes to haemolysis

If we spread the erythrocytes, froma suspension of these in N/10 sodium
chloride solution, over a slide and then allow them to dry gently, we find
that these dried corpuscles when treated with N/10 sodium chloride solution
completely dissolve. Let us consider what happens when the erythrocytes
are drying on the slide. They tend to stick to the slide and, as evaporation
goes on, they tend to contract. As a result of these two antagonistic
processes there is, perhaps, rupture of their membraneous structure, and
they dissolve when they come in contact with what would be an isotonic
solution in the case of undried erythrocytes. Their complete solution
indicates complete disruption of their cellular framework in the process
of drying up.

This complete disruption is, as just now stated, either purely
mechanical due to rupture of all the membraneous portions inside the
erythrocytes that become adherent to the slide, or the removal of water
during evaporation brings about a complete change in the chemical

78 BIO-CHEMICAL JOURNAL

constitution of the erythrocytes, converting them into particles soluble
in a saline of any strength, just as a lump of sugar dissolves in water.
This latter view best explains the complete solution of erythrocytes when
treated with saturated solution of sodium chloride in water. In the process
of desiccation of the erythrocytes by the saturated sodium chloride solution,
water is removed from them. <A portion of this water is perhaps in chemical
combination with the erythrocytes, and this compound is broken up when
they are allowed to dry up over a slide or are treated with saturated
sodium chloride solution. Water, therefore, as it exists inside
erythrocytes, is partly in a chemical combination with its structure, and
H
this compound may be expressed as X

: OH.

A CONTRIBUTION TO THE STUDY OF THE PROTEIN
METABOLISM OF THE FOETUS. THE DISTRIBU-
TION OF NITROGEN IN THE MATERNAL URINE
AND IN THE FOETAL FLUIDS THROUGHOUT
PREGNANCY

By DOROTHY E. LINDSAY, B.Se., Carnegie Scholar.

(From the Physiological Laboratory, University of Glasgow)

Communicated by Professor D. Noél Paton
(Received June 12th, 19/1)

CONTENTS
I. Preliminary.

If. Methods.
Ill. Distribution of nitrogen in the urine of adult herbivora.

IV. Foetal fluids.
A. Relationship of total non-protein nitrogen content to growth of foetus.
B. Variations in the nitrogenous constituents throughout pregnancy.

C. Comparison of the early allantoic fluid with the adult urine and the indications
afforded as to foetal metabolism.

I. PRELIMINARY

That the foetal fluids of the allantoic and amniotic sacs are mainly
foetal urine may now be considered as definitely proved. Structurally
the allantoic sac develops as part of the urinary bladder, while the urethra
of the foetus early opens into the amniotic sac. This anatomical evidence
is confirmed by the experimental work of Gusserow!, Doderlein?, D. Noél
Paton*, and Bruno Wolff.

Since chemical examination of the adult urine throws so much light
on the processes of metabolism, it was hoped that the examination of the
foetal fluids might help to reveal any characteristic differences between
adult and foetal protein metabolism.

Il. Meruops

The total nitrogen was determined by Kjeldahl’s method; non-protein
nitrogen by Kjeldahl’s method after precipitating with an equal volume
of trichloracetic acid, the filtrate being tested and found not to give

Mgrs. nitrogen per cent.

80 BIO-CHEMICAL JOURNAL

the biuret reaction; ammonia by Folin’s method; urea by the Morner-
Folin method; amino acids (including hippuric acid) and allantoin by
an indirect method described by me®; amino acids (not including
hippuric acid) by the formol titration method, hippuric acid being thus
got by difference; creatinine and creatine by Folin’s colorimetric method,
using a Dubosq colorimeter.

In dealing with the amino acids in the adult urine, the large
proportion of hippuric acid which herbivorous urines contain has to be
considered. By the method for the estimation of amine acids described
by me® and used throughout this work, hippuric acid is estimated along
with the other amino acids. It was therefore considered necessary to
ascertain what proportion of this amino acid nitrogen was due to the
presence of*hippuric acid, what to the free monamino acids. To actually
isolate hippuric acid is a long and troublesome process, and it is difficult to
obtain satisfactory quantitative results. An indirect method was therefore
sought for which might prove simpler. This was found in the formol
titration method first proposed by Sérensen®, and later elaborated by
Henriques and Sérensen’. This method gives the amino acids not
including hippuric acid, as the glycin being linked to benzoic acid does
not react with formol. It is clearly possible from the difference in the
amino acid nitrogen determined by my method and that determined by
titration with formol to ascertain the amount of hippuric acid nitrogen
present, and this procedure I have adopted. More recently, Henriques
and Sérensen® have modified their method so that by it hippuric acid
may be directly determined. The following table gives a comparison
of the results got by the two methods :—

Hippuric acid Hippuric acid
difference method formol titration direct
Per cent. total nitrogen Mgrs. nitrogen per cent. Per cent. total nitrogen

20 - 7 21 = 74
36 = 15:7 30 = 13

99 = 34-9 99 = 34-9
15 = 2-1 12 = 1-6

Henriques and Sérensen also show that heating with hydrochloric
acid and subsequent titration with formol provides a convenient method
for the estimation of the nitrogen present as polypeptides and in the

more complex compounds. ™

STUDY OF THE PROTEIN METABOLISM OF THE FOETUS 81
III. Disrriserion or NrrvroGen 1x tHe Urine or Aputr HErsivora

As a preliminary to the study of the distribution of nitrogen in
the foetal fluids, an investigation of the urine of the adult was essential.
I have been unable to find any complete examination of the distribution
of nitrogen in the urine of herbivora, though several workers have
investigated the occurrence of specific constituents. Salkowski? made
no detailed examination of nitrogen in herbivorous urines, but was
concerned chiefly with the presence of oxalic acid in urines which had
been kept for a long time, and thence with the occurrence of allantoin
and hippuric acid. He found that hippuric acid was present in very
considerable quantities. In one case, from a twenty-four hours’ urine,
he recovered 11°85 grams hippuric acid, containing about 12 per cent. of
the total nitrogen present. By evaporation he obtained crystals of
allantoin, which previous to this had not been found in cows’ urines.
Salkowski, however, says it is a normal constituent, the presence of
which had probably been overlooked, as was the case with allantoin in
the dog’s urine.

The urines examined by me were got from sheep, oxen and cows
at the slaughter-house. These animals are kept for a few days in the
slaughter-house before being killed, so that the food given previous to
the collection of the urines was approximately the same in each case.
This food consists mostly of hay.

These cxaminations show that urea averages 83 per cent. of the
total nitrogen; that allantoin is almost entirely absent; that the amino
acids vary im amount, the variation being apparently mostly in the
hippuric acid and probably depends on the amount of benzoic compounds
in the food. They average 6°5 per cent. of the total nitrogen, of which
more than half is hippuric acid. Creatinine averages 3 per cent. of the
total nitrogen, creatine 1°9 per cent. The amount of creatinine is almost
constant, but that of creatine varies considerably. The sex of the
animals was not determined in the sheep.

These tables for the ox show a very marked difference between the urine
of the male and of the female. In the urine of the bullock there is a
much smaller proportion of amino acids, only 56 per cent. of which is
hippuric acid, while in the cow nearly 90 per cent. is in this form.
In the urine of the cow allantoin is present in considerable amounts,
while in that of the bullock there is at most only a trace. Creatinine
constitutes about 5 per cent. of the total nitrogen aud creatine, which

BIO-CHEMICAL J OURNAL

82

9F0°0
GL0-0
S100
¥00:0

uosorpu
autyeedy

6°G
L-&
6:0
G0

uono1}yTu
aur} ear

1¢0°0
810-0
F700
8¢0:0

ues01}tU
QUTUT}VOT)

uos01}1U
auTuryeody

980-0
0
Sé1-0
FIT-0

esoy} Ul
gou_ues01}IN

og)
0

8-L
L:9

9soyy Ul
you ues014IN

pioe
oranddry

prov
oranddry

usso0rztU [e40} JO onvque010g

proe
orunddry
Surpnpour you
‘ploe OULULY

aur *9°0 OOT Ut wosoIpIU “SIX)
aur s,deeyg—"T T1aViL

Pies
oranddry

Surpnpour you
‘prove OUTULY

¢80:0
FILO
890-0
OFT-0
plow
oranddry
Surpnpout

‘prow OUI

das

9G
F8
9b
¥8
por

ormddry

Surpnypout

‘pov OUTTLY

0

90-0

uloJURT LV

uroqur TLV

GLG'I
989-1
Ter
GIFl

VOL)

6:18
¢-F8
T-€8
G68

vod)

etUOmMuLYy

eUOWuy

OST
F961
60L°1
SOL:T

uos0I} IU
[e491

‘ON

‘ON

83

rUsS

vr
4

IN METABOLISM OF THE FOE

PROTE

’
4

STUDY OF THE

0
00-0
090-0
670-0

£10-0
6F0-0
O10-0
820-0

uas01310
out} ReID

a

ver
ueds01}1U
our} val

160-0
900-0
80-0
Z80-0

FFO-0
980-0
£00-0
480-0

uado0rq1u
aururyrad9

WodOI}IN
OUIUT} VAT)

060-0

£90:0
6FL-O

860-0
£68-0
0
661-0

esoyy Ur
you UdSOIJINN

0O-¢g

oF
68

ret
G-LT
0
GF

osoyy ul
you uas0191Ny

600-0
190-0
20:0

6FL-O
I8T-0
£60-0

poe

ond dry

pror

ornddrzy

wasoI}TU [e404 JO oFvqUd.I0g

800-0
£60:0
620-0

60:0
110-0
660-0
£P0-0
prior
orinddry

Surpnpour you
‘ploe ourury

I-
prow
ornddry

Surpnpour you
‘plov ourmy

8L0:0
L100
9L0-0
901-0

€L1-0
Z61-0
LIT-0
Pros
ouinddry

Surpnypour
‘plow ouruty

OULIN *9°O OOT UI ueSorqIH “sxx

proe
ornddry
Surpnypoutr
‘plo ourmy

£100
0

00-0

610:0

TE1-0
616-0
LE0-0
60:0

urozuryTy

oul eUuTAOg—*| J] aTavy,

XO GXV MOD

OFS:
OST-O
EIT
CEE-T

F60'T
6081
O6F-0
166-0

Vol a)

0
900-0
0

FE0-0

110-0
600-0
0

£00

vow

0
GE
0
ST

OT
FO
0

LT

emowuny

S18-1
9LT-0
16F-T
LEL‘T

F8S'T
LOGE
LOL-O
LOFT

uod01}1U
[MOL

xO

MOD

xO

MOO

a

‘

84 BIO-CHEMICAL JOURNAL

is generally present, is rather more abundant in the female than in the
male.

Salkowski, who found allantoin in the urine of the cows he examined,
maintained that it is a normal constituent, but suggested the possibility
that it is present in the urine only of milch cows and not in that of
bullocks. This occurrence of allantoin in the urime of the female—
presumably in calf, as are most cows killed in the slaughter-house—
suggested that there may be an absorption from the foetus through
the maternal! placenta of metabolic products which are excreted by the
mother. But two or three observations which I have made as to the
occurrence of allantoin in the urines of goats during the period of
lactation, and after lactation had stopped, do not support this explanation.

The following are the results :—

Grs. nitrogen

I. Goat in milk-— in 100 c.c.
Urea nitrogen fee Ee ats 1-654 == O-Oho/ len
Allantoin nitrogen... Seis sis 0-136 ==) 64-910 Ne
II. Goat in milk—
Urea nitrogen Bite ae 56 3-136 = 19:29, 4oNe
Allantoin nitrogen ... one Yee 0-152 = 139 Seen
Ill. Goat not in milk and not pregnant—
Urea nitrogen 56 boc sie O32 = 70°95 ee
Allantoin nitrogen... 568 st 0:063 a OA ALIN

These observations show that the amount of allantoin present is very
considerable, the proportion of nitrogen in this form being almost the
same as that found in the urine of cows.

It must therefore be concluded that the presence of allantoin in the
urine of pregnant cows is not due to the excretion of foetal products
by the maternal kidney but is a characteristic of the maternal metabolism.
The larger amount of hippuric acid in the urine of the cow is probably
due to the greater intake of benzoic compounds in hay, &c., during
pregnancy.

In considering the relation of the foetal fluids te the adult urine
the urine of the female is of the most direct importance.

These tables show that in the cow urea constitutes, on an average,
66 per cent. of the total nitrogen; that allantoin averages 6°3 per cent.;
that the amino acids vary somewhat, but on an average amount to
9°8 per cent., and of these hippuric acid constitutes about 87°6 per cent. ;

Pin as

STUDY OF THE PROTEIN METABOLISM OF THE FOETUS — 85

that creatinine averages 5 per cent. of the total nitrogen, and that
creatine is generally present in, considerable proportions.

The impossibility of fixing an absolute standard of normal urine
renders it more difficult to determine the significance of variations in
the distribution of nitrogen in the foetal fluids since, although these
are primarily produced by the foetal kidney, their character may also
be modified by the conditions of maternal metabolism.

IV. THe Foerat Fuvips

While both fluids are largely of the nature of foetal urine and the
allantoic in its early stages entirely so, it is possible that some of the
amniotic fluid may be formed from the membranes before the urethra
opens, although the structure of the epithelium does not suggest a
secreting function. It is further probable that exchanges take place
between the two fluids during pregnancy. This question is considered
by Noél Paton, who indicated the factors which may determine
the passage of water from the allantoin to the amnion and the diffusion
of sugar and non-protein nitrogen in the same direction. The intermittent
addition of foetal urine along with these would account for the increased
volume of the amniotic fluid and also the increase in its non-protein
nitrogen and sugar content.

The possibility of such exchanges between the two fluids complicates
the interpretation of any changes which may occur in their chemical
composition. But undoubtedly in its early stages the allantoic fluid is
urine alone, and its composition may be safely accepted as giving an
indication of purely foetal metabolism.

A. Revarionsuip or tHE Tora Non-Prorery NrtroGEN 1x THE Forrar
FLuips to tHE GrowrH or THE Fortus

Mere exchanges between the two fluids will not modify the sum of
the metabolic products, and it is of interest therefore to determine
whether there is any relationship between the waste nitrogen accumulated
in these sacs and the growth of the foetus. Taking Noél Paton’s figures
for the sheep, as being more complete on this point than those of the
present investigation, the following table shows that the waste nitrogen
accumulates as the foetus grows, but that the nitrogen per unit of weight
markedly decreases. The present observations show the same thing in
the cow.

86 BIO-CHEMICAL JOURNAL

SHEEP
Non-protein
Weight of foetus Mean weight nitrogen per 1000 grs.
Below 15 wos 8 9
30— 100 a 50 4
100— 300 a6 230 2
300—1,000 Safe 420 1-T
Over 1,000 se 1,700 0-5
Cow
Non-protein
Weight of foetus Mean weight nitrogen per 1,000 grs.
100—1,000 500 400 1-7
1,000—2,000 Be 1,731 1-0
2,000—4,000 ne 3,475 0-9
4,000—6,000 oc 4,800 0-7

This may be explained by the more active metabolism of the young
and rapidly growing foetus in which all the tissues are protoplasmic
and undergoing the active metabolism of growth, whereas later the
living tissue of the foetus constitutes a smaller proportion of the body.

This decrease in the nitrogen excreted per unit of weight continues
throughout infancy. Camerer shows that from the age of 1} years
this decrease is constant, gradually diminishing till at the age of
15 years the metabolism is the same per unit of weight as in the adult.

Another factor which may influence this decrease is the fact that
as pregnancy advances, and the vascular connections between the mother
and foetus through the placenta become more fully established, the
placenta may more and more play the part of the foetal kidney.

The following table and diagrams show the relationship which exists
between the sum of the products accumulated in beth sacs and the
weight of the foetus :—

TABLE _V.—SHEEP

Nitrogen per 1,000 grs. of foetus

Weight of foetus Non-protein nitrogen Urea nitrogen _—_Allantoin nitrogen Amino acid
nitrogen

235 1-9 0-96 0-4 0-18
660 0-47 0-27 0:07 0-05
860 0-49 0-21 — 0-17
1,100 0:83 0-15 0-15 0-14

1,230 1:3 0-4 0-17 0-5
1,340 0-41 0-18 0-08 —
Full time 0-52 0:07 0-04 0-06

Cow
Nitrogen per 1,000 grs. of foetus
Weight of foetus Non-protein nitrogen Urea nitrogen Allantoin nitrogen Amino acid
nitrogen

100 1-54 0-52 — tas
a 1:5 0:5 0-12 —
1,93 1:3 0-52 0-09 0-14
3,200 1-0 0-35 0-16 0-33
3,870 1-1 0-28 0-12 0-22
4,050 0:7 0-13 os 0-22
4,500 0-9 0-13 0-22 0:27
5,600 0-5 0-08 0-11 0-13

STUDY OF THE PROTEIN METABOLISM OF THE FOETUS 87

10

4

i \ oy

paar? ee ed
a

/
po

= /
& — pee ee ee ee fn
A otate eee eee 4 VPS wpoessecoeecseee
ore 4 20e@ e 7 ant
eng A eee ere
235 660 860 1115 1230 1340 3000

Weight of Sheep foetus in grams.

Non-protein nitrogen —————
Urea = —— ee oe oe
Amino acid ” ee ee ee

Allantoin 9 eee eee ewe eee

88 BIO-CHEMICAL JOURNAL

45
40
30
20
as
Pd a aN a ‘,
10 _—~ g Te
a 7 Ns
/ Swe ox ; ra \
7 4 N ‘ Lea
5 Z 4 N= woe =~
Y. / Fe =
° s . ¢ 406 F ns
Be) a / Co .
GZ vie : A Te m5
oo aioe
sees? pa
100 900 1930 3200 3875 4050 4500 5600

Weight of Cow foetus in grams.

Non-protein nitrogen — ——_—_—________
Urea. A Sei
Amino acid 99 — ee eee

Allantoin ss BAO oOo OOO Ee

It will be seen that the nitrogen of the fluids varies directly with
the weight of the foetus till about the middle of pregnancy. Allantoin
shows a small but steady rise, but urea after about the middle of
pregnancy decreases, its place being taken by the amino acids.

89)

FOETUS

=
{i

TABOLISM OF THI

~
4

IN M]

’
4

STUDY OF THE PROTE

HOUT

*

CONSTITUENTS ‘THROUC

NOUS

NITROGE?

4

THI

VARIATIONS IN

B.

Ts Og Os [8 9° 0°88 SFI ounl} [IQ
&1 £81 9:9 £6 8S L-19 GF1 0¢&°T—000°T
0 91% TL 6-9 £-01 £9 0-6 { 000°I—008
0 6-81 g-¢ 81 OL 8-9 L:9 F 00& —O09
WedOAQIU BUT}PRAIO ornddry ornddry u9d019TU [2404
+ asoyy Ur ou Surpnjpour you Sutpnyour ULlOJURTTY val jo “quoo aod pouruexo snqooy
uoes01gTU ues0.4IN ploe oury prov oury ues014 10 ‘ON JO FDIO A
auluT} RAL ulojO.Lg
usso1q1 urojo1d-uoM Jo oF vyUa0IVg
= Qs Ol 0-9 OG GL OFT 08 O-F6 F oul} [[OT
I OL O-g OL £8 LE O-<T gg O-LY 0°8*T—000'T
0 €-L GG GG 9-€ GS O-€ FE OLE £ 000°I—008
0 0-9 OG 0-9 OG 0G GG Ig G-E& i 00& —O09
UdaDOAZIU BUTYZRALO oumnddry onnddry pourar
os asoyy UI Surpnypour you Surpnyour uLlojURT Ty voIQ, u9s50141U uas01q1u ues01410 -BX0 snqooy
a uss0I4Tu jou Was04IN = ploe our poe ouruy ul9jO1 urajo1d-u0 Nn 190], ‘ON JO qYDIO AY
LS ouTUT, RAID
S pmy jo quoo wed uedoiq1u ‘sadyy
a aint OMOINWy
A &-OF L-FI L-9% 31 9-1 9-F F onary TY
a L-ct 8-01 L-1& Lat GST 8:8 ib 0£&T—000'T
FIE GL S81 £-06 GSE © FI F 000°T—00z
oumnddry orandd ry uaso1zTU 12407 pourmex9 snjooy
esoy} Ul jou Surpnjpour you Surpnypour ulojUueTy val) jo “‘quoo aed ‘ON JO qYDIO A
UaDOIQIN prov ouruy plo oum1y uoes0.141U
ul0jO1g
woso141t Utez01d-uoU 1%409 Jo onRquadleg
OT 0-0% O-9L 9€ 9-8¢ GS-01 68E 666 Ours [AT
9-96 G61 £99 86 G-O0& GLE 606 9EG OS°T—000'T
OFF O-L OSE oT O-SF G-LT SEI OFT 000°IT—00z
asoyy Ut jou oranddry otanddry u9d0.191U ues0.141U uas0.41U pourmexo snqooy
UuIsTOAIN Surpnypour you Surpnyou. Oloz,URITy VOI) Ul94O1g urejoid-uoN [v90q, ‘ON JO JYBIO Ay

poe ourmy

ploe ouray

ping jo -quoo aed usSoagra “sadyy
aqInaty OlOINVITY
didKQ—'TA Wavy,

JOURNAT,

BIO-CHEMICATL

|

a0

9.
6:

u950141U
uIyraIg

F-0
F-0
WoDO0IJIN
autzRaIp

weso1qTU
our}VaIO,

0-0T

GG
woso1q1u
oulgvady

eat §°ST aie P96 €-9T 8-GP L:LG
0 6-96 rE 6-6 LP ¢-6¢ TL |
0 g-61 acl SGI GG 0-9 0g
ormddry ornddry ueso01zIU [v4.04
weDsO1} IU asoyy Ul you Surpnypour you SuUTpNypoUtr Ul0JULT TW vol yo ‘quoo sod
oUTUTZRIAD uas01}IN plow ourmy prov oulmy Lerner ayes
ulo}O1g
uaso1jtu uTejy01d-UoT [e404 Jo VdvyUEDINg
or SF a; 6-9 9°& SPL 9-6 0:92 9-Cc€
0 8-¢ i FG OT O-FT 6-1 0-6 6-46
0 OF ai GG Ol O-GT O-1 ¢-61 £06
uss01jIU =. 9S9T]} UT Jou oranddiy otanddiy Lecersonanint ueso1ytu = WeD04TU
euIUIyvaIg, §«=»- UOSOAJIN §=Surpnpour you = Surpnyour UlOJULTTV v1.) uleqjoIg ureyoid-uoN —s[RYOF,
ploe ourmy plore ourmly
pimp jo “quoo aed uedorqiu ‘sadyy
aIniyq OILOINWY
9-9 Gor a T-L6 O:8T 9-CI G1
O-F 0-0F &°oG 6-96 G-O1 L-GG FG
mae 9-FG oe L3G Gg F-8E G-ST
oranddiy oranddry WdSOAZTU [240
uoDs0aq1U esoyy UI jou Surpnyour you Surpnjour UlOJURTTW RoI jo ‘quoo aod
OUTUTLOID uds0I19IN plore our y ploe ourmly ues01}10
uloqO1g
uweso14tU UTeJOId-UOU [e404 JO odvyUDd.IOg
C11 ¢-€6 = 3 0'8¢ c-LE LLG O-€ 91G ¢-616
G6 9-99 GLP PLP P-L1 O-LE GF 891 9-GLI
= ¢-F1 O-FT &¢ 9-61 OTT L9 O-8L
uoesoujztu = osoygurgou — orand dy ormnddi1y uesoijItu = WanOaqTU ues01910
OUIUTZ VOL) UssONIN Surpnpour you Surpnpour urlojueypy voi ulejolg utoyord-u0NN [eqOy,
plot ournmy prov ourmy
ping jo ‘queo zed uodoayIu “sis

GIANT OOLNVITY
MOO— ‘TIA IId¥L

QI oF

pourmexo
‘ON

18 oD

pourmrexo
‘ON

pourmexa
‘ON

HN =H

pourmexa
‘ON

000°9—000°F
000°F—000°T
000°T—00I

snqaoy |
JO IGSI9A\

000°9—000‘F
000‘F—000'T
000‘I—00T
snqooy
JO FYSIOA\

000°9—000F
000—000'T

000‘I—00I
snqooy
JO WYSIOA\
000°9—000°F
000°#—000‘T
000°I—001
snqaoy

JO JYSIOAL

STUDY OF THE PROTEIN METABOLISM OF THE FOETUS 91

Urea nitrogen.In the amniotic fluid the amount of urea present
does not undergo any very marked variation with the advance of
pregnancy, and the proportion it bears to the non-protein nitrogen shows
a correspondingly small change, a slight decrease taking place towards
the end of pregnancy. In the allantoic fluid the amount of urea is very
much greater, but the proportion of non-protein nitrogen in this form
is very much less than in the amniotic fluid and decreases markedly
as the foetus increases in weight. This is true for both the cow and
sheep, but the decrease in the proportion of urea nitrogen is much more
marked in the allantoic fluid of the cow than in that of the sheep.

Allantoin nitrogen.—The amount of allantoin present in both
amniotic and allantoic fluids is very considerable, but, as one would
expect, is greater in the allantoic and shows a very marked increase
towards the end of pregnancy. In both fluids of the cow the proportion
is about the same, and in the later stages of development about three
times what it was at the beginning. This increase in the proportion is
less well marked in the amniotic fluid of the sheep, and is not found in
the allantoic fluid of the sheep.

Amino acid nitrogen (including hippuric acid)—In the allantoic
fluids, both of the sheep and cow, the amino acids are abundant, especially
towards the middle of pregnancy. The increase in the amount is very
rapid in the first half of pregnancy, and becomes slower as the foetus
increases in weight. The allantoic fluid of the cow shows little change
in the proportion of amino acid nitrogen with the advance of pregnancy,
but in the amniotic there is a distinct rise in the later stages. The fluids
of the sheep, on the contrary, show a slight increase in the proportion
of amino acid nitrogen in the allantoic and in the amniotic very little
change.

Hippuric acid nitrogen.—The presence of large amounts of hippuric
acid in the adult urine renders it interesting to note the evidences of its
presence in the foetal fluids. Gusserow! found that the injection of
benzoic acid into the maternal circulation caused the appearance in the
fluids of hippuric acid, not of benzoic acid, and from this argued that
the foetal fluids were formed in the foetal kidneys. In the allantoic
fluid of the sheep there is a considerable increase of hippuric acid as
pregnancy advances, both absolute and proportionate, but in the amniotic
fluid the reverse appears to be the tendency.

Creatine and Creatinine nitrogen.—In the allantoic fluid of the cow,
in which most determinations have been made, both creatine and

92 BIO-CHEMICAL JOURNAL

creatinine are present in very nearly equal amounts, and they apparently
increase with the weight of the foetus. In the amniotic fluid creatinine
is absent, but creatine is present in small amounts. In the amniotic and
allantoic fluids of the sheep no creatinine could be found, but a trace of
creatine is present in the later stages of pregnancy.

Ammonia nitrogen.—This was absent in every case examined.

The chief points to be noted in regard to these fluids are—the large
amount of amino acids, the presence of allantoin in considerable amount,
and the large amount of undetermined nitrogen, 1.e., nitrogen not in
any of the above substances. In all these points the fluids differ from
the adult urine. .

Nitrogen not in.these substances.—Perhaps the most interesting point
is the very high proportion of nitrogen not in any of the ordinary urinary
constituents of the adult. In the adult urine, 88 per cent. in the cow,
6-0 per cent. in the sheep, of the total nitrogen is in compounds other than
those determined. In the early allantoic fluid the proportion in the cow is
24-6 per cent., in the sheep 314 per cent., figures which correspond
approximately to the difference between the total non-protein nitrogen
and that in the phosphotungstic acid filtrate. This large amount of
undetermined nitrogen was also observed by Noél Paton, who used
either the Morner Sjéqvist method for urea determination or the Bohland
methods, which gave quite comparable results. The nature of this
I have been able only partially to elucidate. It occurs altogether in the
phosphotungstic acid precipitate, and is therefore probably mainly of the
nature of polypeptides and diamino acids. ‘'Trichloracetic acid was used
to remove the protein, and lower proteins and peptones may therefore
be excluded since the filtrate from the trichloracetic acid gave no biuret
reaction.

The possibility that the protein metabolism of the foetus had not
gone beyond the stage of chains of amino acids was considered, and the
amount of peptide nitrogen was estimated. The method used was that
of Henriques and Sérensen ®.

SHEEP
Polypeptide nitrogen Still undetermined
per cent. per cent.
Allantoic XIIt 7-9 non-protein nitrogen 35-9 non-protein nitrogen
XV 0-6 38-9
XVII 13-9 23-6
Amniotic XIIT Doubtful 17-7
XVII 0 4-6
XVIII 7-6 — 2-1
‘ Cow
Allantoic XV 12-3 15-8
Amniotic XIV 6-9 24-2

STUDY OF THE PROTEIN METABOLISM OF THE FOERTUS — 98

The peptide nitrogen is quite insufficient to account for more than
a small proportion of the undetermined nitrogen. I therefore investigated
the nitrogen in the phosphotungstic acid precipitate. The precipitate
was decomposed by means of barium carbonate and hydrate, the excess
of barium removed with carbou dioxide, and silver nitrate added. The
precipitate thus obtained probably consists mainly of the silver salts
of the diamino acids.

Cow
Silver precipitate nitrogen Still undetermined
per cent. non-protein nitrogen per cent. non-protein nitrogen
Allantoic Xx she Fc 17-9 14:3
VII a Sed 19-0 26-0
Amniotic ace “re 9-4 24-0
35 0-1

Diamino acids are thus seen to be present. The purin nitrogen is
quite negligible in amount, and hence the conclusion seems to be that
nitrogenous compounds which are not products of metabolism in the
adult are present in the early urine of the foetus. Their nature will

require further investigation.

C. Comparison or tHE Karty ALLANTOIC FLUID with THE ADULT URINE

A comparison of the early allantoic fluid with the adult urime is
of especial importance, since the former gives undoubted evidence as
to the nature of purely foetal metabolism.

Cow
Amino acid,

Amino acid, not
Urea Allantoin including including Creatinine Creatine Nitrogen

hippuric hippuric not in these
Early allantoic ... 38-4 5:2 22-7 _— -= —_ 24-6
26°9 25-3 4-6 2-6
Adult urine ope | wliler: 6° 12-7 1-9 5-0 3:2 8-8
SHEEP
Jarly allantoic ... 33-3 20:3 18-5 75 — — 31-4
Adult urine .. 83:0 ? 6-5 3-0 3:0 1-9 6:7

These results for the allantoic fluids of the cow and sheep are fairly
concordant and show :—

1. A small proportion of urea nitrogen in both.
2. A large proportion of allantoin in the sheep, but not in the cow,

94 BIO-CHEMICAL JOURNAL

in which the increase of allantoin oceurs later and coincides with an
increase in the allantoin of the amniotic fluid.

3. A high proportion of amino acids other than hippuric acid. The
figures for the earliest stage in the cow are unfortunately wanting, but
the fact that in the next period the total amino acids were almost the
same, and consisted almost entirely of acids other than hippuric—269
of amino acids including hippuric corresponding to 25°3 per cent. of
amino acids not including hippuric acid—justifies the conclusion that im
the early stages, too, a high proportion of amino acids is present.

4. In the cow creatine and creatinine were present in considerable
amounts in the second stage of gestation—amounts very similar to those
of the adult urine, In the sbeep the result is more doubtful.

5. The undetermined nitrogen was high for both sheep and cow,
and consisted in part probably of diamino acids and polypeptides.

6. The picture of foetal metabolism thus shown by the chemical
composition of the early allantoic fluid is one of low deamidizing power
as indicated by the low urea and ammonia output and by the high
proportion of amino acids. Since the purines are practically absent,
the allantoin may be held to represent the purin metabolism. Its presence
in the early urine of the cow in amounts similar to that of the adult
probably indicates that the nuclear metabolism is already established
in its normal course.

The amniotic fluid throughout resembles the adult urine more closely
than does the allantoic. Possibly this is because it is a later foetal
urine, and occurs when fuller placental development is masking the
purely foetal metabolism by the passage of maternal excretory products.

SUMMARY

1. Certain differences are shown to exist between the distribution
of nitrogen in the urine of the bullock and of the cow. ‘The urine of
the cow contains a large amount of allantoin and of hippuric acid.
In the bullock’s urine the amino acid content is much smaller and
contains a much smaller proportion of hippuric acid, while allantoin is
almost entirely absent.

*. There is an increase in the amount of non-protein nitrogen in
the foetal fluids throughout the first half of pregnancy. But the amount
of nitrogen per unit of weight of the foetus decreases regularly.

STUDY OF THE PROTEIN METABOLISM OF THE FOETUS 95

?

3. The foetal fluids are shown to contain the ordinary urinary
constituents of adult urine—urea, allantoin, monamino acids, creatinine,
creatine, with, in addition, small amounts of polypeptide nitrogen,
nitrogen in diamino acids, and nitrogen in compounds which are not
found in the adult urine, and the nature of which has not so far been
elucidated.

4. Variations in the distribution of nitrogen in the foetal fluids
throughout the course of pregnancy exist, which, in the main, consist of
a decrease in the proportion of urea nitrogen with a corresponding
increase in the proportion of allantoin and amino acid nitrogen.

5. A study of the early allantoic fluid—the urine of the early
foetus—shows, as compared with the adult urine, a low urea content, a
high proportion of allantoin and amino acids, and a large amount of
undetermined nitrogen.

G. It is therefore concluded that the foetal metabolism differs from
that of the adult in the less complete catabolism of the protein and in
the greater activity of nuclear metabolism as indicated by the amount
of allantoin.

I take this opportunity of acknowledging my deep indebtedness to
Professor Noél Paton, under whose guidance this work has been done,
for much help and criticism.

REFERENCES
1. Gusserow, Arch. f. Gyn., II, 1872.
2. Doderlein, Arch. f. Gyn., XX XVII, p. 141, 1890.
3. D. Noél Paton, Trans. Roy. Soc. Edin., XLVI, p. 71, 1907.
4. Bruno Wolff, Arch. f. Gyn., LXXXIX, p. 177, 1909.
5. D. E. Lindsay, Bio-Chem. Journ., IV, p. 448, 1909.
6. Ssrensen, Biochem. Zeitsch., VIL, p. 47, 1908.
7. Henriques and Sérensen, Ztsch. f. phys. Chem., LX, p. 1, 1909.
8. Henriques and Sérensen, Zésch. f. phys. Chem., UXIII, p. 28, and LXIV, p. 120.
9. Salkowski, Ztsch. f. phys. Chem., XLII, p. 212.

= = GPP are 1-06 9:81 Tél 0 FE8-G FE8-G
L-€ 8-9 O:SP cas 6-¥G T-1G 0-6 | 0 £h& aang
G8 ie) GCE = $-66 0-96 &-é1 | ) 9E6°E 9E6°E
a >i L:LP = 8-66 tL (ets | Os L88°G TPO"
8-1 L-€ O-LP ms £-06 £-01 |, i) &L-P el
GE She) G-OF 8-06 G-9& FOL 9-GI | O-€ 96-1 TLO&T
=a 1:86 8:66 PSE 9-91 6-16 | 8-7 PPLE 606-6
=e = L8G = 6-18 6-1 G-LE 0) L¥8-0 L¥8-0
= a L-GG a 8-6 0-61 G-6T a SPT PIS-1
= —s = mas cae L8 L-91 6-41 G10-T IStT
a aot 8-& ro 8-E 8-9 9-46 (BS LOV-O T&P-0
a ae VST = 9-11 ) O-&L S16 690-0 680-0
es aa = = = a = ae 8:06 170-0 €90-0
= asoyy = plovoranddry ~—s prow ueD01}TU ping [e307 Ur = prnyy [vyz0} Ur
7, wosouy = WaBOAyTU ur jou Surpnypour = ornddmy uroyuryy vory) [e40} “JUoo «= UaOAYTU “sad UO HOIYTU "S15
mS (OUTZROI = OUTUT}VOA/) WoDOTJIN, jou Sutpnyout tod use50191U ‘uan01}1U uos01}1U
[> ‘prov oul plov oulMly uroyorg utoyo1d -W0 NY Teq07,
= uoso0aqa utoyoad-uou JO “quad IOg aInig OlOLNVITY
MOO—IT
4
<q + ra 8cI = OL #¢ sé 0
= L &1 G8 a LP OF LT a)
— SI Ol Og = OF 66 1G 0
= oe = 11 = 69 LI Ge ral
a € 9 9L ae 6 LI wg 0
eal 8 €1 £6 SP $8 FG 6G L
sa aa cP Lv “¢ Ke ag 8
= = GP 7 og € 6¢ 0
= = tL = LI 9T 9G 9
Fe = 2 — — — I it fe
i alk GG rr GG g 61 i
7; aa 1 aa 3 0) 61 L
= = = ee = a ad &1
uoDOIpIU u0501}1U asoy} ur jou —s prow: orand diy otanddry uasoaytu
OUTPRIAID OULUT]VOA/) Wwos01ZIN, Sutpnypout you Ssulpnyout uloJULTTV VOt() Ulop,OIg

96

‘plore oul ‘plow OULU
pp jo ‘quoo zed uoSoaytU ‘sid ]
dIOT OIOLNVITY
MOOD

SNOILVAUASAQ ANIN—SLTIASAY GATIVEaC, PNIAT) XIGNUddy

S86
G18‘T
Oss
OSS‘ T
oga‘s
OS
OLO'T
OFS
060'T
008
ogg
OG
06

*0°0
UE OUINTO A

066
68T
9ST
1&@
ZO
O&E
O9T
LST
tel
961
tL

9G

OF

wa5019310
u1eyo1d-uo NY

009°
oole
00¢*F
0S0'F
CLS‘
oge'’s
004°
e10%
O&6'T
006
C&t
OST
OO

snqooy

JO FYSIO AM

uasoIqu
190,

97

STUDY OF THE PROTEIN METABOLISM OF THE FOETUS

= Cra = Git ae GGG 6-8E L-L@ PEE
a <= = &1E — 6-16 L-é L-€P OST
GE = = 9-€ = GSE TL T-L¢ S16
= = aa 6-GF ae C6 1 SE L:8
6-9 = =i LIE FE Late 0 g-G9 8h
= 9S 0 P+ = 1-96 L-8 8-09 OG
=o = as 6-LT == L-OT 0 FIL GE
= on = FEE == 0 0 9-99 L-9
ie = = 9-¢ om 9° 9° &°€8 0-01
¥-6 61 0 FEE a 0-61 8-F 8-GP 0
pro prior ues0141U
esoya ur = oranddryq = oranddry 1@309
uesoy4tu §«ueso4Mm = uaZo1q1U you Sulpnput Surpnyour urozueypy ROI) ‘quad aod
epydeddjog ounvoiy ourunvoery uoS0.4IN jou prow uan0rqtU
prow ourmy Ulo}O1g
ourlulry
uaso1j1U ureyo1d-uoU Jo “qUeD Jog
GInTy OLLOINNY
MOD—AT
= — G y ¥ L G
a3 a. Or = L T FI
a = I aes e & OT
¥ a 6 — jG & L
= 0 6 i I 0 61
F-0 0 I as 9 Gj FI
T oe ¢ = & 0) 0G
= = ¢ 4 0 0 OT
a Pa I = T T CT
F-0 0 L = t T 6
9soqy ul plore £ proe
uas01}1U ueso.1q1U you orunddry oumddry uloqueypy v1)
aul} Va suluTy RAID UdsOM4IY Surpnyour Sutpnyour
you plore ourmy

ploe ourmy
ping yo ‘ques sod uasoaqru ‘sadyy

atmo Ty OLLOINWY
MOOD—TIT

£860
106-0
OGT-0
168-0
90L-0
99-0
996-0
6SP-0
L¥E-0
611-0

CEF-O
9PZ-0
8LT-0
6GF-0
918-0
6L9-0
FOOT
SOS-0
1OF-0
6IT-0

ping [e307 pny [eyo

UI Ues01}1U Ul UeSOI41U

"S15

"sa3

‘uos01q1U ues01q1a

uroqoid

-uON

IPOD

_

Yen)
SCNHHOMAMEO

ues0191U

ulaJOLg

06¢°T
069

O&P

006'1
OOF
068°S
OOF'E
O8Z'¢
S661
O&S

‘O°O

009°¢
00°F
0SO'F
GL8"S
002°
£10°%
O£6'T
OSS'T
006

OOT

snqooy

UT OUINTOA JO FYSIO MA

u950.141U
uroqoud
-uON

ues0141U
TP90.L

IA
x
IIx
it
AIX
I
AI
Ix
Wx
XI

‘ON

aa o'6E L61 9-63 Fare yy FG 9F9-6 FILS S18
aa GOP LL G8E 1:8 GET 8-h LLL-O 918-0 OFZ
£-0 GLE L:9 resi eit GE 0 POUT FOL OP
ms L-&P 8°66 G-LI 0-S1 S11 867-0 699-0 OST
= aa a == OE S-61 ) 899-0 89¢-0 O&t
i$ 0-0F 1-91 PSE 9-7 0:06 PEL GZE-0 PLE-O OSG
OP ©-6P 9-G EGG L&I 9-F1 8-16 SP 9SF-T gE
ae I-LE = GLE C81 6:9 0 OOL-0 OOL-0 O&&
ry 6-66 Fe S16 me £86 GL SGP-0 6SP-0 O8T
oe 8-SP 6-L OT GST LT& 9:91 S1&-0 8LE-0 OSG
= 0:6 oa Pas a = O-8T LIT-0 &PT-0 O9T
_ a 0:06 GL SGI GLE 0:0F PGT PPE-0 SOF-0 ST9
—{ = uaSoaqu prov poe ussoIyIU = pINFJ[vyOF UL pNP [eq09 utr
A auryraao esoyy urgou — otand diy ormnddry yvq04 uIsoIqIU ‘s13 UESOA}TU "s18 “00 UL
on ar UdSOMIN SUTpnpour you surpnypout uloJULTTY voIQ ‘queo aod ues0.191U uoes01}1U OUIN]OA |
bs) uoes0141u plov oulwty proe our y uss0jtu = utoyoad -UoN [e940],
© ouuryroig uI0}01g
ere
3 ugsorqtu Uteyord-uwoU fo “Gud 10g
<q aIaTy OlOLNVTTY
= daadHQ—TA
—
| = OT 8¢ L8 = Fa L F6G
an ai FOI oF 66 13 cg €1 69
oO I ¥6 LT St 8% 16 0 1G
=) = ial = 6L 8g 0g GP GEE
= = Fe = Be GG && 0 69T
= = ree 1é oF 9 9 0G O&T
91 8ST SI GL FP LY 68 1Gé
oe 08 raat , 18 OF ST 0 91Z
re $6 a GL = L9 61 IES
= gg Ol Fl LT Ov GG 961
— 9] — = = — 9T €L
c 1a 1 L cI GG Ol gg
uos01}1U pre ounddry prov ormnddry :
QUTPRAIO asoyy Ul surpnyout SurIpnypout ULlOJURTTV TOL) ues0.141U u950141U
4 you uasoO.14IN you ‘plow ouTULY uld}O1g u1oyo1d-aoNy
u050.191U ‘plov oulmly
fea) OUTUTZvOd-) pi jO ‘queso 19d UoHOigTU “SADT
jor)

dint OIOLNVITY
LAGHS—A

000°9

omy Te G AX
OOL’é

omy TA 1 IAX
009%

eee | Ure I TIAX

oulTy [UY I A
OFE'T I IITs
OFT I IIx
O&S'T I x
SIlT G Ale
098 I II
099 I Wx
O0& I AIX
G&S I IX

snqooy snqooj

jOsyqsiom jo ‘ON ‘ON
jo wing

106 AX
GLE IAX
ISG IIAX
PLE A
69T IIe
OST IIx
OIF ».¢
91G Als
Gos II
[ST IIx
68 AIX
G9 Ix

uoes0141U ‘ON

[On

99

STUDY OF THE PROTEIN METABOLISM OF THE FOETUS

000°9
§ OT == 0 9-4 ar = 6-16 oLi SEP-0 S6S-0 00¢ oat, [TNT é AX
009°
) = 9+ 0 L-¢ a L-68 6-1 606-0 bES-0 O&% ours TO I AX
0S0'S
OL == g-¢ 0 oF = 6-06 9-11 99T-0 L81-0 O8T Ourry TIM I WIAX
a = 81 = & FI 9° £-08 L-LI &LG-0 O&E-0 OSPF ours Tw I A
= << = = ss GOT L:9¢ — oa ChS-0 ots OPE T i IIx
= €-1 L-9 9-9 £6 OSI OGL L:GG 9LF-0 L19-0 0&9 O&S'T I x
ad ric 0-08 0 PET 9-99 £9 816-0 FESO O&L SIT'T G Als
= = £36 = 6-81 G-G &-8¢ a may £80-0 O&% 098 I II
Cowl], =a LL 0 0 6S FOL = zs OIT-0 OE 099 I Tx
= 0 0-3 3 0 S61 LSE) LL S6G-0 S1&-0 S69 GLE G IIA
a = = GI a8 —— oa £01 L&S-0 OOF-0 OGO'T O0& I AIX
= 0 c-9T G-¢ S-61 a OSL LL 916-0 E60 009 06Z I XI
a? re O-GT aa a 0:8 0:08 0 FOL O FOL-0 SIP S&S I Ix
= 0 L-9T = £85 0-01 0:09 0-61 ST-0 {81-0 00¢ Ost G IIA
— 9-0 F-08 7; Ll O-& GPS 0 16-0 16-0 Soo g-c9 G IA
ues0.41U prloe prow uesoajtu =p [e4o7 UL pny e404 ur snjooy = sesnqaoj
uarO1jtu auIZvedo = asayg ut = oranddry ~— oranddry 1909 uesO1qIU ‘sid © UeH0I9TU ‘S15 "0°0 JO 4YDIOM jo ‘ON
epydedsjog + you Surpnpoul Surpnpour urojyurypy wvorg | ‘quao aod ‘uss01qTu ueso4tu =u emNjOA =: JO ang ‘ON
uasoijia = uasOWIN 401 plov prov ouTUITy ues04ytu = UTegord-u0Ny 1890,
aulutqvaL ) our y uloqO1g
uasoij1U uto}01d-uOU JO “quUa0 Jog
OILOINWY daaHgQ—]ITA
0 F = — 08 8T L8 SOT AX
=o Ld 0 i — 6L &1 88 I0r IIAX
=> g 0 F = £8 él G6 POL TIAX
= I — 8 G cP él 9g 89 A
i = = oa él cP a cr FL Ills
i ¢ ¢ L 6 $S GG GL L6 pg
ae 6 = 0 t LT G Of G& Ale
a 8 = g G 1G oe => 9& Il
a 9 0 0 G 9G a ae ve Ix
0 6 = 0 L 0G & 9& 6E IIIA
= = ¢ =3 = car 4 ce 6& AIX
i) 9 ra L = LZ & 9& 6& xI
= £ = ma G 0% 0 GG GG ibe
0 ¢ == L & ST L Of LE ITA
oO Ol = ia I 81 0 e€ § IA
uan041410 prior proe
aulRalo aseyi ul orndd ry orandd ry U10qUeTTy vai uas01}10 uos0141U ues0a91U ‘ON
+ you GaDs014iN Sutpnyout you Surpnyout ul9j0O1g ureyo1d-uoNy [®99,,
uaso1410 ploe ourmy prow ourmy
ourUIyRad

ping jo ‘yueo aed uesoagru ‘sadyy
dainty OLLOINKYy

daidHS—TIA

100

A PRELIMINARY STUDY OF THE BIO-CHEMICAE
RELATIONS OF VARIOUS LIPOID SUBSTANCES IN
THE LIVER

By FREDERICK P. WILSON, M.Se., M.D., Beit Memorial Research
Fellow.

(Received June 28th, 1911)

The discovery of Landsteiner, Miller and Pétz] that an alcoholic
extract of human or animal liver could be used as an antigen, shed quite
a new light on the Wassermann reaction. The idea that this reaction
was of the same nature as the typical Bordet-Gengou phenomenon had
to be abandoned. An enormous amount of work on the subject at once
followed, and it was shown that the test could be performed with extracts
of other organs, as well as with substances such as lecithin, cholesterin,
bile salts and phosphatides from different organs.

Further, it was found that the antigenic properties of these substances
varied considerably, and their suitability for use was still further modified
by other factors, such as their haemolytic properties. While most workers
with the Wassermann reaction have been content to use alcoholic
extracts of liver or heart of varying composition, which were rejected
if they possessed undesirable properties, others have made efforts to
apply substances of more or less definite nature and free from the most
objectionable features such as the presence of protein or haemolytic
bodies. These have chiefly been lecithides obtained from various organs.
Browning used a lecithin obtained by digesting a heart with ethyl
acetate at 60° C., precipitating with acetone, and further purifying by
repeatedly taking it up with ether and precipitating with acetone.
He found that the antigenic powers of such a lecithin are increased by
the addition of cholesterin.

Noguchi has advocated the use of a lecithin obtained by treating
a liver or heart with warm alcohol, evaporating the extract to dryness by
means of a current of air at 37° C., and treating this repeatedly with ether
to separate the ether-soluble substances. The ether-soluble portion is
evaporated, then dissolved with a little ether and precipitated with
water-free acetone. It is this acetone-insoluble fraction which Noguchi
has recommended for use as antigen.

BIO-CHEMICAL RELATIONS OF LIPOID SUBSTANCES 101

Recently, Noguchi and Bronfenbrenner have published the results
of an investigation into the aftigenic, anti-complementary, and haemolytic
properties of the different ether and acetone soluble and insoluble
fractions obtained in the above process. The livers were chiefly derived
from human beings who had suffered from various disorders, but, in
addition, the livers of animals were also examined. The iodine values
of the acetone-insoluble fractions were determined, and an attempt made
to associate these with the antigenic properties and the clinical conditions.
These investigators found that the most satisfactory antigenic properties
were shown by the acetone-insoluble fractions and that, speaking
generally, a high antigenic value in these fractions was associated with
a high iodine value. This group consists chiefly of phosphatides.

Some recent workers on lipoids, such as Erlandsen and Maclean,
have thought that under certain conditions part of these bodies might
be in a free state in the tissues, and the rest combined with other tissue
elements from which they might be separated by suitable treatment.
The present investigation was undertaken with the object of ascertaining
the haemolytic and anti-complementary properties of extracts obtained
by the method used by these workers.

The tissue was first extracted with ether, then with cold alcohol,
and finally with hot alcohol. From these three main extracts further
fractions were separated by evaporation and precipitation with acetone.
If the amount of material in each fraction sufficed, the saponification
and iodine values were worked out.

Merruop or PREPARATION

A pig’s liver, obtained immediately after the slaughter of the animal,
was passed through a mincing machine, and then spread out on glass
plates and dried under a fan at 30°C. The dried mass was passed through
a coffee mill and the resulting fine powder thoroughly dried in vacuo
over H,SO,.

The tissue was then extracted with ether in a shaking machine for
three or four hours and rapidly filtered. This was repeated four times
till the addition of acetone caused only a very faint cloudiness. This
extract was carefully preserved in a stoppered bottle in the ice-chest.

The residue of the tissue was then repeatedly treated with cold
absolute alcohol in a similar manner, and the extract preserved. Finally,

102 BIO-CHEMICAL JOURNAL

the tissue was extracted with hot absolute alcohol until practically no
more acetone-insoluble material could be obtained.

Three different extracts were thus obtained, which were treated in
the following manner : —

Ether-soluble Extract—The ether was distilled off at 40° C. until
the extract had a syrupy consistency. Excess of acetone was now added,
when a brownish-yellow mass separated out. The residual fluid was
filtered and a clear yellow fluid obtained which, on evaporation under
the fan, gave a greasy-looking yellow substance. These two fractions
were preserved over H,SO, in vacuo.

Cold Alcohol Extract——Evaporation under the fan yielded a grayish
greasy-looking substance. This was thoroughly extracted with ether
and two fractions thus obtained—one ether-soluble and the other ether-
insoluble. The former was further separated into an acetone-soluble
portion of white colour, and a brownish acetone-insoluble one. Only
very small amounts of these two fractions were obtained, as most of
the cold alcohol extract seemed to be insoluble in ether, and was obtained
as a light brown tenacious substance.

Hot Alcohol Extract—On leaving this in the ice-chest over night,
a small quantity of a brownish-yellow substance was deposited. This was
insoluble in hot alcohol, but partly soluble in ether, from which a white
precipitate was thrown down by acetone. This acetone precipitate was
found to contain a fair amount of phosphorus. The rest of the hot-
alcohol extract was evaporated until a dark brown sticky mass was left.
This was treated in a similar manner to the cold alcohol extract. The
ether-soluble fraction thus obtained was very small in amount and was
practically all soluble in acetone. The various fractions thus obtained
were preserved over H,SO, in exhausted amber-coloured desiccators,
and their haemolytic and anti-complementary properties tested as soon
as possible .

Preparation of Emulsions.—In each case the emulsion was prepared
by dissolving 0°025 gram of the substance in a little warm absolute
alcohol, and slowly adding 5 c.c. of a 0-9 per cent. solution of sodium
chloride.

Determination of the Haemolytic Property.—Increasing dilutions of
the emulsions were made with 0-9 per cent. saline solution. Four parts
of each of these dilutions were mixed in a series of test-tubes with
one part of a 20 per cent. suspension of washed human red_ blood
corpuscles. The tubes were placed in a water bath at 37° C., and the

BIO-CHEMICAL RELATIONS OF LIPOID SUBSTANCES 103

result noted after incubation for two hours. The figures in the
accompanying table denote the weakest emulsion which produced
haemolysis.

Determination of the Anti-complementary Property.—The technique
used for this purpose was that advocated by D’Este Emery. This worker
has pointed out in his Hunterian Lecture that all sera will give some
absorption of complement if the extract be strong enough, and that the
difference in the powers of absorption between normal and_ syphilitic
sera is only one of degree. He claimed that for standardizing an extract
it was better to use normal sera, as these are very much more constant
than syphilitic ones, which vary over a wide range. This plan has
been adopted in the present investigation, and the absorptive powers
of the extracts tested against normal sera.

Progressing dilutions of each emulsion were placed in a series of
small test-tubes and mixed with fresh normal human serum in the
proportion of-four parts of the former to one of the latter. The tubes
were then placed in a water-bath at 37° C. for half an hour. To each
tube was now added one part of a 20 per cent. suspension of sensitized
human red blood corpuscles, which had previously been saturated with
an excess of an appropriate dilution of anti-human rabbit serum.
By using this technique of D’Este Emery a very minute amount of
unabsorbed complement may be detected. Moreover, the time taken up
is very much lessened, as haemolysis occurs very rapidly when there is
any free complement, and when absorption has been complete the
agglutination of the corpuscles enables the result to be read off in a
very short time. The figures in the chart denote the weakest dilution

which gave complete absorption.

Anti-
Fraction Saponification Jodine Haemolytic complementary
value value property value
Ether-soluble—
1. Acetone-soluble ais wate 190-5 106-5 Pe 2
2. Acetone-insoluble wee an 5p. 92-4 4 72
Cold aleohol—
l. Entire ... ed a wee 246-4 28-4 9 22
2. Ether-insoluble oe Ses 285-5 16-9 8 5
3. Ether-soluble—
(a) Acetone-soluble ... Ree — 47-5 5 38
(b) Acetone-insoluble “ne Amount insufficient.
Hot aleohol—
]. Entire ... Ee aC eo 266 54 8 6
2. Ether-insoluble Son ees 257 91 il 31
3. Ether-soluble—
(a) Acetone-soluble ... ae — 32-2 5 20

(b) Acetone-insoluble “cr Amount insufficient.

104 BIO-CHEMICAL JOURNAL

CoNSIDERATION OF RESULTS

Saponification Values——In the cases where sufficient material was
obtained to estimate them, there was not much variation in the
saponification values, except in the acetone-soluble fraction of the ether
extract which was somewhat lower than the others. This may have been
due to its containing fats of a higher molecular weight or to the presence
of cholesterin. :

Iodine Values..The ether extract showed the greatest degree of
unsaturation; the acetone-soluble fraction, as might be expected, giving
the highest iodine value. The other extracts geve considerably lower
values; that of the cold alcohol being about half the hot alcohol figure.
No relation between the iodine value and the haemolytic property could
be established. The fact that the acetone-soluble fraction of the ether
extract, with its high iodine value, was practically non-haemolytic may
have been due to the presence of cholesterin.

Haemolytic Property.—Vhe fractions derived from the etner extract
had a much weaker haemolytic action than those of the other extracts,
which did not differ much from each other.

Anti-complementary Property—Great differences were manifested
by the various fractions in this respect. The acetone-soluble portion
of the ether extract had practically no power of absorbing complement.
On the other hand, the acetone-insoluble fraction possessed this property
to a very high degree, in fact, was two and a half times stronger than
the next highest fraction. A curious difference was shown by the cold
and hot alcohol extracts, and the fractions obtained from them. While
the entire cold alcohol extract gave a value of 22, and the hot alcohol
a value of 6, the ether-insoluble fraction of the latter yielded a value
of 31 against a value of only 5 by the ether-insoluble fraction of the
former. The acetone-soluble fractions of the ether-soluble portions of
both cold and hot alcohol extracts possessed very fair powers of absorbing
complement, but, as mentioned above, were much inferior to the acetone-
insoluble portion of the ether extract.

It was unfortunate that the amount of liver originally used was
insufficient to yield enough acetone-insoluble portions of the cold and
hot alcohol extracts to perform the tests. The greatest amount of material
was obtained from the ether extract, which also contained a large quantity

BIO-CHEMICAL RELATIONS OF LIPOID SUBSTANCES 105

of acetone-insoluble substances. Very small amounts of the cold and
hot alcohol extracts were soluble in ether, and the greater part of what

was soluble was also soluble in acetone.

CONCLUSIONS

1. The acetone-insoluble fraction of the ether extract yields the
most suitable antigen.

2. The cold and hot aleohol extracts yield substances differing from
each other and from the ether extract in physical and bio-chemical

properties,

3. These differences are apparently not dependent on the

saponification or iodine values.

106

THE ACTION OF SELENIUM SALTS ON RED BLOOD
CORPUSCLES

By CHARLES O. JONES, M.D., M.Sc., M.R.C.S.
From the Bio-Chemical Laboratory, University of Liverpool
(Received June 29th, 1911)

While investigating the action of selenium salts, it was noticed
that there was produced a marked effect on the red blood corpuscles, and
as the action appeared to be an haemolytic one, this work was undertaken
to investigate its action.

As was shown in the paper referred to, the soluble salts are reduced
in the body by some reducing substance, in this case probably glucose.
and the insoluble amorphous selenium becomes deposited in the cells,
and can be demonstrated as a golden-brown deposit in the cells. Part
of the selenium is excreted as soluble salts by the kidneys, part is
excreted by the bile and large intestine, and part as methyl-selenide is
excreted in the breath.

In these experiments rats were used. Owing to the extreme ease with
which sodium selenite is reduced, all doses were given hypodermically and
were freshly-prepared. Blood films were prepared in the usual way,
dried without heating and stained with Leishmann’s stain. Each film
being prepared and stained in the same manner.

The first rat was given 0°3 cubic centimetre of 0125 per cent.
solution of sodium selenite. The blood was examined and found to be
normal.

Very shortly afterwards there appeared an increase in lymphocytes;
the polymorphonuclear leucocytes remained normal, so that there
appeared twice as many lymphocytes as other leucocytes. The red
blood cells, at the same time, had developed a cloudy white centre
(fig. 1).

This was followed by the almost complete disappearance of leucocytes
and the development of the cloudy centres described above into vacuoles.
Some have one central vacuole of varied shape, while others have
several small ones, and it is probable, as will be seen later, that most of
these vacuoles are at first central and gradually move towards the
periphery (fig. 2).

ACTION OF SELENIUM SALTS 107

The rat was now evidently very ill; it was left till next day to
see if the blood cells would recover. Next day its blood was normal
again.

It was again given an hypodermic injection, and the blood again
went through the same changes, namely increase of lymphocytes,
development of a cloudy white centre in many of the red blood corpuscles,
disappearance of leucocytes, and vacuolation of many of the red blood
cells. It was given another hypodermic injection of the same amount.

The vacuolation became more marked, the vacuoles increasing in
number and working towards the periphery. In one or two instances
the vacuole was seen opening into the blood stream, and occasional
corpuscles were seen that had extruded the vacuole (fig. 3).

The movement of the vacuoles to the periphery appeared to be
accelerated, a large number of red cells having numbers of vacuoles all
round the periphery of the cell (fig. 4).

The films now commenced to stain very badly. Many of the red
blood cells, as will be seen in fig. 5, showed the extrusion of buds all
round the periphery.

The next films stained hardly at all, indeed, it was difficult to see
that any staining had been attempted. The rat was now extremely ill.
The next few films showed little difference from fig. 5, but an hour later
there was a distinct improvement: the films commenced to take the
stain again, the red blood cells began to lose their vacuoles, and the
rat was evidently recovering. Large numbers of polymorphonuclear
leucocytes and lymphocytes were present. An hour later, except for
occasional red blood cells with vacuoles, the blood appeared normal, and
next day the blood was normal.

This recovery is interesting, for if this is a condition of haemolysis,
as will be shown in the next experiment, the quick recovery must be due
to the re-absorption of haemoglobin into the red blood cells that are not
too injured to recover. For it is evident, that when the conditions that
cause the haemolysis are removed, the red blood cells are able to prevent

any further loss, and also to make good the loss which they have
previously gone through.

These experiments were repeated with the same results. A second
rat was given 0°6 cubic centimetre of 0°125 per cent. solution of sodium

selenite hypodermically; the time from the injection to the death of the
animal was about two hours.

108 BIO-CHEMICAL JOURNAL

The blood when first examined was normal. It was seen to go
through the same changes as the previous one, the increase of
lymphocytes with the white filmy centres in many of the red blood cells,
followed by the presence of vacuoles which moved to the edge of the
cells and were extruded. The polymorphonuclear leucocytes became
scarce, and were swollen, and in some cases disintegrated. The film
stained badly and, finally, the corpuscles would not take the stain at all.

The rat now developed profound dyspnoea, convulsions appeared and
death very quickly followed.

The blood was removed from the animal and was found to have
undergone haemolysis.

No haemolysis was produced by sodium — selenite in test-tube

>

experiments.

CONCLUSIONS

Although sodium selenite does not produce haemolysis outside the
body, yet it has that effect when injected hypodermically into the
tissues of the animal.

The cause of this haemolysis is obscure. Sodium selenite is
reduced to selenium in the portal circulation, chiefly in the spleen and
liver unless a very large dose is given, in which case it floods the whole
system and selenium is found in all the organs; but im the doses given
in the previous experiments the reduction took place in the spleen and
liver. This is confirmed by the fact that only a proportion of the red
cells undergo these changes. In a film one part will be found almost
entirely changed, another portion almost entirely normal, while_a third
portion will show clumps of changed cells and clumps of normal ones,
just as one would expect if the portal red blood cells alone became
changed and then became mixed with the normal systemic blood. The
only other factor which might have some action is the remarkable
disappearance of glycogen and sugar which the soluble selenium salts
cause, but what effect this would produce it is impossible to say.

My thanks are due to Professor Benjamin Moore for his assistance
and many valuable suggestions.

e —- a

ACTION OF SELENIUM SALTS 109

110

DIRECT MEASUREMENTS OF THE OSMOTIC PRESSURE
OF CASEIN IN ALKALINE SOLUTION. EXPERIMEN-
TAL PROOF THAT APPARENT IMPERMEABILITY
OF A MEMBRANE TO IONS IS NOT DUE TO THE
PROPERTIES OF THE MEMBRANE BUT TO THE
COLLOID CONTAINED WITHIN THE MEMBRANE

By BENJAMIN MOORE, M.A., D.Sc.; HERBERT E. ROAF, M.D.,
D.Sc.; anp ARTHUR WEBSTER, University of Liverpool.

From the Laboratories of Bio-Chemistry and Physiology, University of
Inverpool

(Received June 30th, 1911)

The most important experimental points settled by this research are
the unequal distribution of alkali on the two sides of the membrane at
the conclusion of each experiment, the high pressures observed, the flow
of crystalloid towards the region of high pressure, and the proof that
crystalloid, although associated with colloid, still exercises osmotic

pressure.

This combination of observations appears to us to finally settle
several outstanding questions as to osmotic pressure effects of colloids,
and of the behaviour of living cells towards apparently impermeable ions
which are, nevertheless, found in high concentration within the cells.

These experiments finally dispose of the older view, that osmotic
pressures of colloids are not due to the colloids but to slowly diffusible
crystalloids present as impurities (the ash of the colloid), which never
get into equilibrium on account of the slowness of their diffusion through
the membrane from the region of high pressure to that of low pressure.

For, the direction of movement of the crystalloid associating with
the colloid, and causing the osmotic pressure as a crystallo-colloidal
complex, is from the side of low pressure to the side of high pressure,
and this to such a great extent as to leave no question about the matter.
Until the colloid is almost saturated by crystalloid, practically no alkali
remains on the non-colloid side of the membrane.

OSMOTIC PRESSURE OF CASEIN 111

The alkali moves against osmotic pressure, into the colloid side,
there to unite in some form with the colloid, and it is this union which
causes the osmotic pressure. ‘The purpose of the membrane is purely
the mechanical one of holding together the colloidal aggregates, and its
apparent impermeability to the ions is quite fictitious. Free ions pass
through quite readily, ions anchored to colloid are retained on the
colloid side and exercise pressure, and the membrane has no influence

on permeabilty of free inorganic ions.

This observation throws a clear light on the behaviour of living cells
towards ions: the varying concentrations of sodium, potassium, chlorine
and phosphatic ions within and without the cell are an expression of
specific affinities of the definite colloids of each particular cell-type for
these ions, and do not mean that there is a membrane acting as a closed

gate to these ions.

The osmometer containing the casein solution may be regarded as a
cell, and the adsorbed or combined alkali is the analogue of the
potassium ion of the living cell; the alkali has no appreciable
concentration outside until the concentration within has become 0°03 N,
and as alkali concentration is increased there persists always a higher
concentration within and a lower without, just as the cell always has an
almost negligible concentration of potassium ion without and a high
concentration within the cell, and this concentration within the cell only
slowly rises when the concentration outside is increased. On _ the
contrary, if the crystalloid pressure outside is diminished, the specific
affinity of colloid and crystalloid holds them together, so that there is
only a very slow passage of crystalloid away from cell or osmometer.
In this way a fictitious appearance of impermeability is produced.

There is no experimental evidence that the cell-membrane is
impermeable to potassium, the actual presence of that ion within it
shows that it must be permeable, and the analogous action of the
osmometer membrane shows that the membrane is easily permeable to
the free crystalloid, and that the high concentration of crystalloid
within and low one without is due to a specific affinity of the colloid for

the crystalloid.
In these experiments we can actually follow the process of the

crystalloid being drawn through the membrane, against the gradient of
osmotic pressure, the rise in osmotic pressure as a result of this, and the

112 BIO-CHEMICAL JOURNAL

establishment of a different concentration of the crystalloid on the two
sides leading to the simulation of an impermeable membrane, such as
is presented by the picture of the living cell.

Reasons have previously been assigned at greater length for the view
that living cells are in reality permeable freely to ions which appear to
pass with the greatest difficulty into or out of them.' In the first place,
the heaped-up concentration of potassium and phosphate within the
majority of living cells, as contrasted with the concentration of these
ions in the outer plasma or other bathing fluids, is to us clear evidence
that the cell, or cell membrane, is easily permeable to the potassium ion.
Otherwise, in the process of reproduction of new cells as the organism
grows, the initial concentration in potassium ion within must have
rapidly fallen to zero, instead of remaining heaped-up high above the
outer level, as one cell gave rise to many. This heaping-up of certain
ions within the cell expresses for us a certain molecular affinity between
the colloid contents of the cell and these ions, and it is that same affinity
which prevents such ions being washed out when the cell is subjected to
dialysis, and so causes the fictitious appearance of impermeability of a
membrane upon which those observers still rely who believe in the
existence of a cell-membrane impermeable, or only permeable with great
difficulty, for these ions. Also, when the affinity of the cell colloids for
an ion, such, say, as potassium, is satisfied, further uptake of potassium
will be very sight and on a purely physical basis, so that when cells,
such as red blood corpuscles, are bathed with salines rich in potassium,
they take up but little more than their normal quantity. This, however,
is no proof of impermeability, any more than is the fact that appreciable
uptake of oxygen ceases when the haemoglobin is saturated with oxygen
is a proof that the red blood corpuscle is impermeable or difficultly
permeable to oxygen. All these experimental findings will present
themselves in any living cell possessing a colloid with a molecular
affinity for ions or other dissolved crystalloid substances, and such
phenomena have nothing whatever to do with permeability, or
impermeability, of a membrane to the ions or other crystalloids.

First, when the cell is washed with a saline solution free from the

ion in question or containing that ion in very low concentration, the
amount of the ion washed out of the cell may be practically inappreciable,

1. See Moore, ‘Recent advances in Bio-Chemistry and Physiology,’ edited by L. Hill,
Arnold, London, 1907 ; Moore and Roaf, previous papers this Journal and elsewhere.

OSMOTIC PRESSURE OF CASEIN 113

for it will only wash out till the concentration of free ion balances against
the amount of adsorbed ion, and, depending on the molecular affinities,
this concentration may be a very low one indeed, falling almost to zero
in the case of the potassium ion. Thus, sodium ions can readily be
washed out of cells because the cell proteins have no affinity for them,
while potassium ions are retained to the last. This shows also how
fresh-water organisms can saturate themselves with potassium and
phosphatic ions from waters in which all tests for potassium would fail
even after considerable concentration. A similar affinity of the colloidal
constituents of the soil for potassium ions and phosphates retains these
valuable constituents in the soil, while the sodium and chlorine ions are
carried seawards by the rivers. No membrane retention hypothesis ever
put forward can explain such facts as these.

Secondly, when a cell is washed with a saline containing a hyper-
tonic quantity of one of these apparently impermeable ions, since the
cell already contains, on account of its molecular affinities, a large
amount of the particular ion, the further amount which it takes up
physically is comparatively so small as to simulate impermeability and
appear to he within the limits of experimental error.

Physiological tests, which are always more delicately balanced than
volumetric or gravimetric chemical tests with purely inorganic agents,
clearly demonstrate, however, that the balance in the cell is being upset
under such conditions. While the cell, to ordinary chemical tests, as
pointed out above, appears to be at least difficultly permeable to
potassium ions, physiological tests show that it is being profoundly
affected. This is shown by the beautifully contrasted effects of perfusion
with ordinary saline on the one hand, and with Ringer’s solution on the
other, in the case of the excised heart, as also in the effects of potassium
starvation and potassium excess for practically all types of living cells.
Witness also the profound effects of two or three grammes of potassium
administered to a man of sixty to seventy kilograms, in depressing the
whole nervous activities, and it must be admitted that the infinitesimal
extra amount which has entered the cells has produced an enormous
result. Such an effect cannot be explained by membranes. It is a result
produced within the cell amongst the constituent cell-proteins affecting
their aggregations, and fitness for chemical change and exhibition of
chemical energy. Apart from living cells, we know experimentally that
ious or crystalloids produce just such effects upon colloids in forming

114 BIO-CHEMICAL JOURNAL

their crystallo-colloidal aggregations. The solution-aggregate of the
colloid varies in size with the nature and amount of crystalloids present,
even far short of actual precipitation.! It is, therefore, easy to see that
variations in amount and character of adsorbed crystalloids must le at
the very citadel of cell activities.

Now, what is claimed as the chief point of beauty and interest about
the present series of experiments is that in an experimental and artificial
system this march of events is clearly demonstrated.

At first there is the colloidal casein in the osmometer inert and
showing no osmotic pressure, just as a living cell would be if deprived of
all its potassiym ion, then as alkali, which takes the place of the
potassium ion of the living cell, is admitted slowly in the earlier
experiments, and in increasing amounts in the later ones, the system, so
to speak, becomes alive and active, and the osmotic pressure, which
corresponds to the index of life, rises all the time till an optimum limit
is reached, beyond which there is no further effect but rather a
diminution in effect. It is not, of course, pretended that we have in this
simple system of casein and alkah anything so labile and delicately
balanced as a living cell. But in this simple type we see the same
primary factors as in the cell, and the important point is that the
chemical play and energy-discharge is within the osmometer as we hold
it is within the cell. The function of the osmometric membrane is a
mechanical one of holding crystalloid and colloid together, just as we
hold that of the membrane of the living cell to be. The osmometer
membrane is freely permeable to the alkali alone, it merely prevents back
passage of the crystallo-colloidal mass, and this produces a fictitious
appearance of impermeability, which we know, from having watched the
experiment from the beginning, to be erroneous. We cannot watch the
experiment from the start in the case of the cell, we take it up when the
cell is already charged with potassium ion, and hence our first impression
is that the cell possesses a membrane which is impermeable to the
potassium ion. A moment of reflection convinces us, however, that the
cell is now full of potassium, that it has arisen from a thousand previous
generations of cells, all containing a like percentage of potassium to
itself, and since in the thousand-fold repeated cell-division the volume
has increased a thousand-fold of the entire mass of cells, it is irresistible

1. Moore and Roaf.

OSMOTIC PRESSURE OF CASEIN 115

to conclude that potassium must have been going in all the time. Here
the osmometric experiment shows us how the uptake or out-put of the
crystalloid is regulated and how the fictitious and merely apparent
impermeability is produced and maintained,

The thirst of the casein for alkali, demonstrated in the three
experiments at the upper end of the table where the concentration of
alkali is lower, illustrates the manner in which the colloids of living
cells can extract and concentrate crystalloids for their purposes from
infinitesimally low amounts in the fluids bathing them. Such, for
example, as bone formation from the excessively low concentration of
calcium ion in the blood, the formation of calcareous and siliceous shells
in freshwater and marine organisms, and many similar cases. Such
concentrations arise from affinities of a molecular type between colloids
and crystalloids, which vary from time to time, so causing periods of
uptake and deposition in a rhythmic manner.

The osmometer used was of metal lined by platinum and with a
platinum grid supporting a parchment paper membrane. The two
instruments used were those described by us in previous papers. The
percentage of casein present was determined from the nitrogen of a
Kjeldahl carried out in each case at the conclusion of the experiment,
since the high pressure registered caused considerable dilution of the
colloid solution during the experiment. Readings were recorded at
intervals as shown in the protocols of experiments, and great care was
always exercised that equilibrium was established before the experiment
was concluded, Titration for free alkali was made, using phenol-phthalein
as an indicator, when the end point is reached at the point where alkali
and casein are combined. Of course no casein is present on the colloid-
free side, and hence the titrations there are quite absolute. It is to be
observed that in the beginning there is no free alkali on either side, and
that the free alkali on the colloid-free side remains very low until about
0:07 to 0°08 N concentration has been reached, when the casein begins to
be saturated at the levels used in the experiments.

It may be noted that, although the table seems but a short one
involving little labour, that each line represents a prolonged experiment,
with many observations and accompanying analyses.

SiregeF
116 BIO-CHEMICAL JOURNAL
Concentration Concentration
Pressure Initial | genes pice Soden
i z ntrat odium rate in 2 3 ie
Detection Observed one nea cent. bie S = ae Hydrate ae Crystaloid Temporary noone days
sur f Sodium y id ch A
oan ese ata Gaseindcen Hydrate of Oanonicior Ounipinciee Experiment lasted
Ezpt. I — mm. mm.
82% 233 28 0:02 N 0-031 N Nil 11-5° C. 14
Acid
Expt. I1—
§:4% 310 57 0-033 N 0:02 N Nil 12:5° C. 10
Acid
Expt. III— ;
8:0 % 334 42 0-04 N 0-004 N 0-006 N 11-0° C. ‘ 14
Expt. IV—
56% 460 82 0-06 N 0:025 N 0-013 N 15:0° C. 6
Expt. V—
Gor 488 85 0-065 N 0024 N 0-018 N 19-0° C. 13
BLapt. VI—
6-08 % 610 | 101 0-07 N 0:022 N 0-015 N 15:0° C: 10
Expt. VII—
B16 686 129 0-08 N 0-047 N 0-025 N 15-0° C. 14
Expt. VILI— :
4-7 % 588 125 0-09 N 0-048 N 0-034 N 13:0° C. 14
Expt. IX—
5:04 % 586 7 0-095 N 0-056 N 0-032 N 145° C. 14
Expt. X—
516% 538 105 0-1 N 0-066 N 0-044 N Hapa (Cr 9
Expt. XI—
51% 508 100 0-1N 0-06 N 0-:042N 13-0°C. 22

Notr.—The concentration of alkali at the commencement was the same exactly
on both sides of the membrane, and varied from one-fiftieth normal in the first to one-tenth
normal in the last experiment. Observe the passage of alkali through the membrane to join
casein, and the rising osmotic pressure demonstrating that the osmotic pressure is due to the
combination of crystallo-colloid, and that the diffusion of crystalloid is against the gradient of
osmotic pressure. Observe also that a maximum of osmotic pressure per one per cent. of
colliod is reached at about 0-047 N of free alkali, and further increase of alkali lowers osmotic

pressure.

Experiment I—Commenced December 6th, 1909. Ten per cent. caseinogen in 0-02 N
sodium hydrate against 0-02 N sodium hydrate.

Time from commencement Osmotic pressure in Temperature
Days Hours mm. of mercury in °C.
0 0 ae 0 ann 14°
be 19 NY 6 a 14
1 19 of 82 os: 11-5
2 19 ae 152 don 12:5
3 11 a 172 a 15-0
4 3 iss 196 Boe 14-0
4 6 a 233 ie 16-0
6 3 wf 233 xe 15-0
7 3 oe 233 je 14-5
8 3 fers 233 ae 13-0

Analysis of contents of osmometer at end of experiment :—

Colloid Side : Clear solution, slightly opalescent. Titrated with N/10 NaOH and phenol-
phthalein, 5 c.c. required 1-55 cc. N/10 NaOH + Nitrogen by Kjeldahl gave 1:33 per cent. nitrogen,
equal to 8-2 per cent. caseinogen.

Crystalloid Side : Clear neutral solution.

OSMOTIC PRESSURE OF CASEIN 117

Pxperiment 11--Commenced February Ist, 1910. Ten per cent. caseinogen in 0-033 N
sodium hydrate against 0-033 N sodium hydrate.

Time from commencement Osmotic pressure in Temperature
Days Hours mm. of mercury in °C,
—- 18 is 116 a 13
l 2 a 166 és 12
l 17 ans 216 ae 14
2 17 ie 256 oF 11-5
3 U7 we 288 phe 13-5
4 20 “a 306 = 16
6 1 ape 310 a8 15
7 0 mee 310 vie 16

Analysis of contents of osmometer at end of experiment :—

Colloid Side: Opalescent fluid. Titrated with N/10 NaOH and phenolphthalein, 5 c.c.
required 1 c.c. N/LO NaOH. Nitrogen by Kjeldahl gave 0-86 per cent. N, minal to 5-4 per cent.
caseinogen.

Crystalloid Side : Clear neutral fluid.

Experiment I[1]—Commenced December 6th, 1909. Ten per cent. caseinogen in 0-04 N
sodium hydrate against 0-04 sodium hydrate.

Time from commencement Osmotic pressure in Temperature
Days Hours mm. of mercury in °C,
1 0 a 226 fey ll

2 0 ie 242 OP 11-5
3 0 om 348 a 12-5
3 8 re 348 ~ 15
3 23 ass 338 aa 14:5
5 2 as 336 ewe 16
6 23 ae 332 a 15
il 23 re 332 cae 14-5
9 2 a 332 es 13

10 23 oa 334 =f 12
12 23 ae 334 wae ll

Analysis of contents of osmometer at end of experiment :—

Colloid Side: Opalescent fluid, alkaline; 5 c.c. required 0-2 c.c. N/10 sulphuric acid.
Nitrogen by Kjeldahl gave 1-28 per cent. N, equal to 8-0 per cent. caseinogen.

Crystalloid Side : Clear alkaline fluid, no protein present ; 5 c.c. took 0-3 c.c. N/10 H,SO,.

Experiment IV—Commenced December 19th, 1909 Ten per cent. caseinogen in 0-06 N
sodium hydrate against 0-06 N sodium hydrate.

Time from commencement Osmotic pressure in Temperature
Days Hours mm. of mercury in; °C;
0 17 Si 160 5: 14:5
1 17 ae 356 ee 14
2 13 es 396 cee 11
3 13 ane 430 ae 12-5
4 1] = 460 ioe 17-5
6 0 ast 460 “=p 15-5

Analysis of contents of osmometer at end of experiment :—

Colloid Side :; Pale yellow alkaline fluid; 5 c.c. took 1-45 c.c. N/10 sulphurie acid. Nitrogen
by Kjeldahl gave 0-88 per cent. N, equal to 5-6 per cent. caseinogen.

Crystalloid Side : Clear alkaline fluid, free from protein ; 5 c.c. took 0:65 c.c. N/10 sulphuric
acid.

118 BIO-CHEMICAL JOURNAL

Experiment V—Commenced April 14th, 1910. Ten per cent. caseinogen in 0-065 N sodium
hydrate against 0-065 per cent. sodium hydrate.

Time from commencement Osmotic pressure in Temperature
Days Hours mm. of mercury in °C.
0 18 as 112 ee 16
1 1 a 166 Beh 17-5
3 16 cise 236 Ais 14-5
4 0 322 08 18
5 0 366 sate 16,
6 0 500 406 506 15-5
7 0 a 454 ate 16
8 0) 480 ado 16
9 0 ae 484 BR 12
11 Oo - 08 488 O08 18-5
12 23 Be 488 ae 19-5

Analysis of contents of osmometer at end of experiment :—

Colloid Side: Opalescent alkaline fluid; 5 c.c. required 1:2 c.c. N/10 H,SO,. Nitrogen
by Kjeldahl gave 0-91 per cent. N, equal to 5-7 per cent. caseinogen.

Crystalloid Side: Clear fluid, alkaline, no protein present ; 5 c.c. required 0-9 c.c. N/10
sulphuric acid.

Experiment VI—Commenced January 31st, 1910. Ten per cent. caseinogen in 0-07 N
sodium hydrate against 0-07 sodium hydrate.

Time from commencement Osmotic pressure in Temperature
Days Hours mm. of mercury in’ gC:
0 17 cee 202 AOS 14
1 17 ae 432 ae 13:5
2 1 Jac 484 ae 13-0
2 Mi a 542 oie 12-0
3 17 340 584 oe 14-0
4 17 a 604 set 13-5
5 20 mee 612 500 16-0
6 16 386 614 500 16-0
7 1 509 614 wigs 15-0
8 1 <n 610 ie 11-5
9 1 ook 610 ee 14-0
10 ] 610 Bos 15-0

Analysis of contents of osmometer at end of experiment :—

Colloid Side : Clear alkaline fluid ; 5 ¢.c. required 1-1 c.c. N/10 sulphuric acid. Nitrogen
by Kjeldahl gave 0:96 per cent. N, equal to 6-08 per cent. caseinogen.

Crystalloid Side : Clear alkaline fluid containing no protein; 5 c.c. required 0-75 e.c. N/10

sulphuric acid.

OSMOTIC PRESSURE OF CASEIN 119

Experiment VII—Commenced April lth, 1910. Ten per cent. caseinogen in 0-08 N
sodium hydrate against 0-08 N sodium hydrate.

Time from commencement Osmotic pressure in Temperature
Days Hours mm. of mercury in °C.
— 16 ie 98 aie 17
1 16 ie 284 aa 17
2 16 ae 412 Se 15
3 0 s 452 ee 15

3 17 ay 512 ene 17-5
4 0 te 536 So 19
4 19 is 576 a 17
6 16 as 632 ce 14-5
7 0 os 644 ber 18
7 16 UE 652 Ae 15-5
8 0 ete 660 “ee 16-0
8 16 he 668 sets 15-5
9 0 Oe 680 Ae 16-0

10 0 an 686 ae 15

Analysis of contents of osmometer at end of experiment :—
Colloid Side : Clear alkaline fluid ; 5 ¢.c. required 2-35 c.c. N/10 sulphuric acid. Nitrogen

by Kjeldahl gave 0-85 per cent. N, equal to 5-37 per cent. caseinogen.

Crystalloid Side: Clear alkaline fluid, free from protein; 5 c.c. required 1-25 ¢.c. N/10
sulphuric acid.

Experiment VIII—Commenced February 26th, 1910. Ten per cent. caseinogen in 0:09 N
sodium hydrate against 0-09 N sodium hydrate.

Time from commencement Osmotic pressure in Temperature
Days Hours mm. of mercury in °C,
1 22 ae 332 ae 12
2 6 os 392 = 14
2 22 oa 458 ox 10
3 6 ers 490 ACC 15
4 0 ee 520 Sea 14

5 i) pet 542 ae 13-5
6 0 ae 998 “he 13-5
6 8 53 564 ee 16
7 0 es 568 ea 13
8 3 ore 576 ae 15
9 5 iy 580 Ae 16-5
9 22 Sos 580 ee 14-0

10 22 oe 582 aa 14
11 22 a 584 ss 13
12 6 re 588 eae 16
13 6 8 588 <F 16
14 6 “ns 588 rie 13

Analysis of contents of osmometer at end of experiment :—
Colloid Side : Clear alkaline fluid ; 5 c.c. required 2-4 c.c. N/10 sulphuric acid. Nitrogen
by Kjeldahl gave 0-76 per cent. N, equal to 4-7 per cent. caseinogen.

Crystalloid Side : Clear alkaline fluid free from protein; 5 c.c. required 1-7 c.c. N/10
sulphuric acid.

120 é BIO-CHEMICAL JOURNAL

Experiment IX—Commenced March Sth, 1910. Ten per cent. caseinogen in 0-095 N
sodium hydrate against 0-095 N sodium hydrate.

Time from commencement Osmotic pressure in Temperature
Days Hours mm. of mercury in °C.
iE 19 Bes 254 si 14
1 19 sae 422 aha 13
2 3 wt 456 ob 16
2 19 se 500 seis 14
3 3 eas 508 ae 16
3 19 S06 516 wa 13
5 19 cs 550 aa 13
6 19 oe 560 wis 13
7 3 boc 566 Hs : 16
7 19 50c 568 ie 14
8 Igy 506 570 oe 13
9 19 ee 571 he 13
10 21 ahs 576 are 13
11 23 ane 576 Bc 16°5
12 19 36 579 Ene 14:5
13 3 ae 584 zoe 165
13 19 sis 586 ae 14°5
14 19 Re 586 sin 14

Analysis of contents of osmometer :—

Colloid Side: Clear yellow alkaline fluid; 5 c.c. required 2°8 c.c. N/10 sulphuric acid.
Nitrogen by Kjeldahl gave 0-8 per cent. N, equal to 5-04 per cent. caseinogen.

Crystalloid Side: Clear alkaline fluid, free from protein; 5 c.c. required 1°6 c.c. N/10
sulphuric acid.

Experiment X—Commenced April 11th, 1910. Ten per cent. caseinogen in 0°1 N sodium
hydrate against 0-1 N sodium hydrate.

Time from commencement Osmotic pressure in Temperature
Days Hours mm. of mercury in °C.
1 0 i 284 300 17
2 0 363 334 50° 17
3 0 oe 484 596 15
3 8 508 494 a 16
4 0 ana 510 ee 17:5

4 8 not 514 Sot 19
5 3 36 524 Sc 17
7 0 aoe 534 “a 14:5
Z 8 536 538 Soe 18
8 8 ee 538 300 15-5
9 8 sie 538 a 15-5

Analysis of contents of osmometer :—

Colloid Side : Clear alkaline fluid; 5 c.c. required 3-3 c.c. N/10 sulphuric acid. Nitrogen
by Kjeldahl gave 0-82 per cent. N, equal to 5-16 per cent. caseinogen.

Crystalloid Side: Clear alkaline fluid, free from protein; 5 c.c. required 2-2 c.c. N/10
sulphuric acid.

OSMOTIC PRESSURE OF CASEIN 121

Experiment XI—Commenced November 28th, 1910. Ten per cent. caseinogen in 0-1 N
sodium hydrate against 0-1 N sodinm hydrate.

Time from commencement Osmotic pressure in Temperature
Days Hours mm. of mereury nia iehel OF
~- 16 ee 4 ax 11
2 16 ee 130 a 11-5
3 19 oe 268 anf 14:0
4 16 340 12-0
5 16 400 13
6 16 430 11
7 16 450 12-5
8 16 462 15
9 16 468 14
LO 16 472 14:5
Il 16 476 14
12 16 480 14
13 16 484 16
14 16 foc 486 ne 14
15 Ge a 490 AG i4
16 16 vie 500 At 16
17 12 sor 504 ee 14
18 12 “sf 508 ate 15
20 12 Bee 508 oct 18
22 12 aes 508 es 13

Analysis of contents of osmometer :—

Colloid Side: Clear yellow alkaline fluid; 5 c.c. required 3 ¢.c. N/10 sulphuric acid.
Nitrogen by Kjeldahl gave 0-81 per cent. N, equal to 5-1 per cent. caseinogen.

Crystalloid Side ; Clear alkaline fluid, free from protein; 5 c.c. required 2-1 c.c. N/10
sulphuric acid.

THE SODIUM PHOSPHATE STANDARDS OF ACIDITY
By BB. PREDRAUX

(Received July 13th, 1911)

During a recent investigation into the hydrogen ion concentration
[H’] of phosphate solutions by the hydrogen electrode method!, certain
relations were discovered between these and the second and_ third
dissociation constants of phosphoric acid. These constants, K, and K,,
allow of an easy calculation of [H'] over the range most required in the
case of N/10 sodium phosphate solutions. The experience gained in the
experimental work has led to certain conclusions as to the advantages and
disadvantages of making up such solutions from standard NaOH and
H,PO,, rather than from mixtures of mono- di- and_ tri-sodium
phosphate?, or from mixtures of other salts®.

To eliminate the diffusion potential between the solutions, saturated
potassium chloride was used in most of the measurements. A few
comparisons with the liquid potential eliminator used by Salm, i.e., the
addition of N/8 or N/10 sodium chloride to all solutions, showed that the
same E.M.F. is obtained with either method, a fact which considerably
strengthens the probability that both methods do really attain the desired
object.

The normal H,PO, was made from Merck’s purest acid (density
= 1°75) which was found free from nitrate, chloride, sulphate, and HPO,,
and carefully analysed by conversion into Mg,P,0,.

The NaOH was made from the sticks (pure by alcohol) dissolved in
CO, free water, and analysed as sodium chloride.

All the decinormal solutions referred to can, of course, be rapidly
prepared from these two stocks.

As will be seen from the enclosed table, all solutions up to the neutral
point give values of [H’] closely agreeing with those obtained by Salm
from various mixtures of N/10 NaH,PO, and N/10 Na,HPO,, and with
those calculated by the formula mentioned below. No special care need
be taken even to secure alkali or water free from CO,, since the solutions
weighed in Jena glass flasks can be boiled for a few minutes after making

1. Journ, Chem. Soc.,"p."2224, June, 1911.

2. Salm, Zeitsch. physikal. Chem., LVI, p. 492, 1907; Friedenthal, Zeitsch. Elektrochem.,
X, p. 113, 1904.

3. See Walpole.

THE SODIUM PHOSPHATE STANDARDS OF ACIDITY 1238

up, then brought to the original weight with the required volume of
water,

As will be seen from the enclosed curve the [Il°| = 1 x 10-4 to
[H'] = 1 x 10° varies rather rapidly with the relative proportions of
NaOH and H,PO,. In this part, therefore, the ratios given in the table
should be accurately observed, a difference of 1 per cent. in amount of
NaOH added making a difference of about 0°5 in the exponent of [H°].
Since, however, the normal solutions can easily be standardised within
one per cent., and excess of CO, can be driven off as described
above, the standards are easily reproducible in practice.

Even less accuracy is required in the preparation of solutions from
{[H-] = 1 x 10-® to the neutral point, since here the curve is quite flat,
and the unavoidable errors of the hydrogen electrode will be far larger
than any accruing from possible excess or defect of alkali. And from
[H:] =1 x 10-* to [H'] =1 x 10-® the solutions are still easy to
reproduce with sufficient accuracy, although, since they are on the
alkaline side of the neutral point, CO, must be more carefully guarded
against.

From this point onwards, however, the third dissociation constant
begins to have an increasing effect, and the curve shows a rapid decrease
of [H'] with added NaOH, until at [H'] = 1 x 10-9, which corresponds
almost exactly to the composition Na,HPO,, an error of | per cent. in
the composition of the mixture makes a difference of 0°25 in the exponent
of [H'] on the acid side of 1 x 10-°,-and 0°37 on the alkaline side. It
was actually found that deviations of about this order occurred in the
measured [H°| of synthetic solutions of this composition, the cause
being almost certainly the presence of small and varying quantities of
CO,. The effect of CO, may easily be observed by leaving a solution
coloured with phenolphthalein exposed to the air. It was measured in
the following way. A solution giving by the hydrogen electrode a value
of [H'] = 16 x 10-1! was exposed for an hour to a current of CO,. The
[H'] again measured was found to have risen to 70 x 10-%. This is an
extreme case, of course, but serves to illustrate the care required.

The reproducibility of Na,HPO, standard from crystalline salt was
next studied. The salt must, of course, be re-crystallized, since crude
specimens may give [H’| as far from the true value as 1 x 10-!!. In

H,PO,

crystallisation would appear not unimportant, in order to ensure that

view of the required accuracy of the ratio the conditions of

124 BIO-CHEMICAL JOURNAL

traces of Na,PO,, or Nall,PO, are absent. A solution of the crystalline
salt is made slightly acid to phenolphthalein with phosphoric acid, and
evaporated down to beyond the point where crystallisation should
commence, a high degree of supersaturation being possible. The solution
may be slightly pink when hot. To this, normal NaOH is added, drop
by drop, from a burette, until, at a certain point crystallisation takes
place, the mass of fine crystals being then rapidly sucked dry and
washed on a Buchner funnel.

This preparation gave a value of [H'] = 13 x 10-° approximating
very closely to that of Salm (1°5 x 10-°) and to the value depending on
the dissociation constants (1 x 1079).

Slightly on’ the alkaline side of [H'] =1 x 10-!° the errors in
stoichiometrical proportions of NaOH and H,PO, and the presence of CO,
again become relatively unimportant compared to those of the hydrogen
electrode, and this statement holds true up to Na,PO,.

In conclusion it may be stated that solutions from [H'] = 1 x 10-4
to [H'] = 1 x 10-8 and 1 x 10-! to 1 x 10-12 may be prepared, either
from mixtures of the NaOH and H,PO, or solutions prepared from the
crystalline sodium phosphates, the former method being perhaps
preferable as requiring only two standard solutions and analytical
controls. For solutions in the neighbourhood of [H’] = 1 x 10-9,
however, it seems better to use the re-crystallised salt Na,HPO,.

As far as an easily attainable analytical accuracy 1s concerned, the
exponent of [ H’| is actually determined for each solution to about 0°1 of
a unit in the flat parts of the curve. Unfortunately the [H'] cannot be
measured with anything like this accuracy. By using, however, a large
number of measurements and finding an equation to the whole curve, it
is considered that the numerical value [H*] has been determined with a
much greater probability than could be attained by any individual
measurement. The constants K, K; of this equation have been found to
be 2°0 x 10-7 and 3°0 x 10-12, the former being practically identical with
the second dissociation constant of H,PO, found by Abbott and Bray!,
using a different method; that of distribution ratios and conductivities.

A simplified equation which permits of a rapid calculation of the

NaOH
HsPOs

always 54, molecular with respect to the latter, is as follows? :~

ratio required for any given value of [H’]; the solution being

1. Journ. Amer. Chem. Soc., XX XI, p. 730,"1909.

2. The deduction of this may be found in the preceding communication already referred
to—-Journ. Chem. Soc., p. 2224, June, 1911.

THE SODIUM PHOSPHATE STANDARDS OF ACIDITY 125

2 a, [H] Bay K,
NaOH ar eG
HPO |, w{H] | aK,

eek iar |

%; % a, the degrees of dissociation of Na,HPO,, NaH,PO, and
Na,PO, were found to be 0:06, 0:077, 0°05 by Abbott and Bray!,
Putting in K,, K, as above :—

2+ 0°39 x 1071 H'} +

NaOH *" LHI 4

HED,

10'8 <x 10-"

: REC
aes 3°6 x 10-7
ih OEE ee
+ [H'j 4 ca
This calculation may be used from [H']=1x 10-5 to
[H-] = 1%x10-!!. From 1 x 10-5 to 1 x 10-5 the K,, term is negligible

and the equation reduces to : —

7.Cc. ——_—-—

VS NaOH _ 2+ 0:39 x 107[H']
oN 1 + 0°39 x 10’[H']

C.C. HPO,

The table gives some observed values for comparison with those
ealeulated by this equation.
Molecules NaOH to 10

Molecules H,PO, [H-] found [H:] calculated
10-0 (NaH,PO,) 9-3 x 10-> (Salm) 12) 10-4 *
I 15 x 10-4 (Prideaux)

10-6 1:0 x 10-5
11-05 ef, <1 0>S)(B:)
11-5 9:7 x 10-7 (S.)
12-05 10 x 10-6
14-45 2-6 x 10-7(P.)
14-45 3-2 x 10-
15-50 15 x 10-1(8.)
15-50 20) % 10-1 (P-)
17-2 1:0 x 10-
17-7 (Neutral point) 0:73 x 10-7
17-6 0-76 x 10-7 (P.)
19-5 1-0 x 10-8 (S.)
19-6 1:0 x 10-8

20-0 (Na,HPO,) 1:3 x 10-9(§.) 1:0 x 10-9
z 1-5 x 10-9(P.)

20-04 1:0 x 10-10
20:3 9-8 x 10-11 (§.)

20-5 5-9 x 10-4 (P.)

21-4 13 x 10-11 (8).

21-6 2:0 x 10-1 (P.)

22-2 1-2 x 10-1 (P.)

22-6 10 x 10-

| * Directly from K,, a (NaH,PO,) and C. = 0-1.
1. Loe. cit.

BIO-CHEMICAL JOURNAL

126

0-16

0:06

0-61

fied) eS ee eA ee OR, ee ee ee Be eee gl oo Uh A dt ly Da Veer | ene Maw IN

§ OTE O-LT 0-9T 0-ST O0-FL 0-8T 0-81

Peer

“OL

WOT] RAJUGOUC
uol jos

S
2
+H

00-9

00-4

00-8.

00-6.

00-01

00-LL

(0a

.
|

127

NOTE ON THE SYNONYMY AND HISTOLOGICAL
CHARACTERS OF EAST LONDON BOXWOOD
(GONIOMA KAMASSI, E. MEY.)

By R. J. HARVEY-GIBSON, M.A., F.L.8., Professor of Botany in the

University of Liverpool.
(Received August 2nd, 1911)

In this Journal! some observations were published on the
physiological properties of a timber used in the manufacture of shuttles
in Lancashire and Yorkshire, and believed to have an injurious effect
on the health of the workmen employed. The piece of wood then
submitted to me was described as ‘ West African Boxwood.’ I was at first
led to regard it as the wood of a plant described by De Wildeman as
Sarcocephalus diderrichii, but on further examination I found that the
wood was exported, not from West Africa, but from South Africa, and
also went by the name of ‘ Knysna Wood’ or ‘ Kast London Boxwood.’

I obtained a large quantity of the sawdust produced in the process
of shuttle manufacture when *‘ Kast London Boxwood’ was employed for
that purpose, and Dr. A. W. Titherley, Lecturer on Organic Chemistry
in the University, was kind enough to make an extract of this sawdust,
obtaining from it an alkaloid or mixture of alkaloids, prepared in
accordance with the method described in detail in the paper above
referred to.

I then asked Miss S. C. M. Sowton, a research worker in the
Physiological Laboratories to determine, if possible, whether the
alkaloid extracted had any effect on the mammalian heart, as I was
informed that the injurious effect on the health of those employed took
the form of so-called ‘ cardiac asthma.’ The experiments made, and the
results arrived at, all pointed to the fact that the alkaloid was a marked
cardiac depressant.

I had occasion recently to re-investigate the whole question, and as
there seemed to be some doubt as to the exact botanical nature of the
wood I thought it worth while, in view of the economic importance of the
subject, to attempt to settle this problem of botanical identity. The
present note, therefore, may be taken as supplementing the work already
carried out in 1905.

1. Vol. I, No. 1, 1905.

128 BIO-CHEMICAL JOURNAL

Gonioma kamasst is generally spoken of as ‘ Knysna wood’ or
‘Knysna Boxwood.’ In Stone’s ‘Timbers of Commerce’ Knysna
Boxwood is given as the trade name of Gonioma kamassi and
synonymous with ‘Cape Boxwood,’ * Kamassihout,’ and ‘ East London
Boxwood.’ Gonioma is a genus of Apocynaceae and native of South
Africa. Unfortunately, Burus macowanit, a member of the
Kuphorbiaceae, and also a South African plant, is known as well in
commerce as ‘Cape Boxwood’ or ‘ Hast London Boxwood.’ I submitted
the questions of synonyms to the Director of the Royal Gardens, Kew,
and was informed by him that Gonioma hamassi was only partly
synonymous with ‘ Kast London Boxwood,’ because the latter name also
included Bueus, macowant. The problem to be solved, therefore, was
which of these two timbers was the one used in shuttle manufacture,
whose deleterious effects on the workmen had been the subject of
physiological investigation. To settle this question I examined,
microscopically, preparations of the wood used in shuttle manufacture,
and described as ‘Knysna wood’ and compared them with sections
prepared from an authentic specimen of Gonioma kamassi, obtained from
the Conservator of Forests, Knysna, and with sections prepared from a
specimen of the timber of Burus macowani, for which I am indebted to
the Director of the Royal Gardens, Kew.

The wood of Gonioma is considerably lighter than that of Buxus, a
cubic foot of the former, when dry, weighing (according to Stone) 58 lbs.,
while a cubic foot of Buxus weighs 74 lbs. Gonioma yields a pale
yellow solution in water, while Buxus gives no coloration. When
transverse sections are compared certain marked differences make their
appearance.

In Gonioma the main mass of the xylem is composed of lgnified
fibres, whose walls are channelled by narrow canals and whose lumina are
almost occluded. These fibres vary from 12 to 15 in diameter, while
those of Buzus macowani rarely exceed 104. The fibres of Gonioma are
irregularly arranged while those of Buxus are in fairly regular radial
rows, not unlike those of a Gymnosperm. Mixed with these fibres are
tracheae which in Gonioma average 35m” in diameter; the tracheae of
Buxus seldom exceed 25” in diameter. The medullary rays of Buxus
are very narrow, seldom being more than two cells in breadth, and
composed of thin, radially extended, pitted plates. The cells of the
medullary rays of Gonioma are much larger and broader, and the rays

HISTOLOGICAL CHARACTERS OF EAST LONDON BOXWOOD 129

themselves may be from one to four cells broad. The wood of Buus
macowant is said by Stone’ to be ‘nearly identical’ in anatomical
characters with Bucus sempervirens, the European Boxwood. This is not
so. The fibres as seen in transverse section are much more irregularly
arranged than in B. macowani, and the tracheae are much larger, being
seldom less than 354 in diameter. An examination of many sections
shows that the histological characters afford a reliable means of diagnosis
of these three woods. Certainly, with the aid of an eye-piece micrometer,
there is no difficulty in distinguishing ‘ East London Boxwood’ derived
from Gonioma kamassi, from ‘ East London Boxwood’ derived from
Burus macowant. A histological comparison of the wood used in shuttle
manufacture, and called * East London Boxwood,’ leaves no doubt that
that timber is to be referred to Gonioma kamassi and not to Buaus

macowant.

l. Loc, cit., p. 194.

130

CRYOSCOPIC DETERMINATIONS OF THE OSMOTIC
PRESSURE OF THE BLOOD AND BODY FLUIDS OF
SOME AUSTRALIAN ANIMALS

By JUDAH LEON JONA, DiScy a8. io.
From the Physiological Department of the University of Melbourne

(Received August 22nd, 1911)

The determination of the osmotic pressure of the body fluids of
various animals has been carried out by a number of investigators. To
the data thus acquired I am able to add the results of some experiments
made with the blood, etc., of various Australian animals. The method
employed was the determination of the depression of freezing point by
the Beckmann apparatus (Centigrade scale being used throughout).

Lanp MamMMats :— A
Sheep (defibrinated blood obtained from slaughter yards) ... 0-59
Rabbit St aid a8 a ue ee nae sa8 0-59
Ecutpna Hystrix :—Hibernating (A) ... (1) 0-624
(September) : (2) 0-622
Awake (5) ee (1) 0-600
(May) (2) 0-600
LAND REPTILES :—
Tortoise. LEmydura macquariae A (1) 0-559
(2) 0-561
(3) 0-560
Bien um 5a ae (1) 0-550
(2) 0-545
(3) 0-550
Lizard. Mixed blood of Lygernia cunninghami and a species
of Tiliqua (1) 0-639
(2) 0-637
Fisu :— Teleosts (Sea water)—Barracouta aH 556 Sc (1) 02979
(Thyrsites atun) SG (2) 0-978
183 doe (1) 0-976
(2) 0-980
(Fresh water)—Murray Cod (Oligorus macquariensis) 0-642
0-660
0-650
CRUSTACEAN (Fresh water)—A stacopsis bicarinatus “100 “oc ass (Ll) 0761S
(2) 0-6)1

(3) 0-618

131

ON THE INFLUENCE EXERCISED BY CERTAIN ACIDS
ON THE INVERSION OF SACCHAROSE BY SUCRASE

By FREDERICK STOWARD, D.Sc., Vepartment of Agriculture,
Perth, West Australia.

From the Laboratorte de Chimie Biologique, Institut Pasteur, Paris
(Received August 25rd, 1911)

Among the chemical conditions which influence enzymic action is
the degree of alkalinity or acidity of the medium in which the action
takes place; that is to say, the rapidity with which a given enzymatic
splitting proceeds is influenced by the addition to the reaction medium
of small quantities of an acid or base. In general, enzyme action, if
certain exceptional cases are excluded, is greatly retarded and ultimately
arrested by the addition of comparatively small amounts of a base. The
addition of acids to the reaction medium, on the other hand, in small
but variable amounts, according to the particular acid employed, appears
to favour enzyme activity im certaim cases. Most enzymes are, however,
remarkably sensitive to changes in the chemical reaction of the medium
in which they ordinarily manifest their normal activity, and examples
illustrative of this characteristic are not difficult to find.

Malt amylase, for example, as the investigations of Ford!
demonstrate, evidently function best in a medium which is approximately
neutral. Departure from this condition of relative neutrality on either
side, provided by the addition of quite minute amounts of either an
acid or alkali, leads, as his results indicate, to perceptible diminution
of the activity of this enzyme.

The researches of G. Bertrand? aftord ample evidence of the extreme
sensitiveness of the oxidizing enzyme laccase towards a number of
different acids, a very active yet dilute preparation of this enzyme being
completely paralysed by quite minute amounts of such acids as sulphuric,
hydrochloric, acetic, oxalic, etc.

While a relatively neutral medium favours amylolytic activity, as
exhibited by malt amylase, and while the oxidising action of laccase is
enormously retarded and eventually completely arrested by minute
amounts of different acids, the inversion of cane sugar by sucrase is

1. Journ. Soc. Chem Ind., XXIII, 8, 1904.

2. Ann. Institut Pasteur, XXI p. 673, 1907.

132 BIO-CHEMICAL JOURNAL

favourably influenced by the presence of small but variable amounts of
different acids. Thus, Kjeldahl’ states that inversion of cane sugar by
sucrase is favoured by the addition of very small quantities of sulphuric
acid, while larger amounts exercise a marked retarding influence on the
action.

The correctness of Kjeldahl’s observations was confirmed by
O’Sullivan and Tompson,? who state that the most favourable amount
of acid varies with the preparation of enzyme employed and the
temperature at which the experiment is made. They were, however,
unable to explain the cause of the variations observed.

Fernbach’s® researches include a comparative study of the influence
exercised by various acids, chiefly organic, on the inversive activity of
sucrase. The fact is established that for each acid there exists a certain
range of optimal concentration, varying considerably according to the
acid employed, within the limits of which the enzyme manifests its
maximal activity. The addition of still smaller amounts of acid than
those indicated by these limits leads to a slight diminution of the action
in most instances, while addition of progressively larger amounts produces
great retardation, and ultimately complete arrest of the action, as shown
in the case of sulphuric and oxalic acids.

The more recent work of Sorensen* directs attention to the importance
of the concentration of the hydrogen ions in enzymatic reactions. Results
of experiments on the inversion of cane sugar by sucrase in the presence
of sulphuric acid and of mixtures of citrates and phosphates in which
solutions of sucrase differing in strength and mode of preparation were
employed, tend to prove the statement that, in otherwise equal conditions,
the optimal concentration of hydrogen ions in the inversion of cane sugar
is almost independent of the kind and quantity of sucrase employed, and
likewise of the acidifying reagent utilised.

Although the fact has been established, that in general the presence
of small amounts of different acids exercises a favourable influence on the
inversive function of sucrase, it nevertheless appeared to be of interest
to investigate the subject anew by the employment of different acids
either in molecular concentrations or in amounts proportional to these
latter, and further to endeavour to ascertain whether highly ionised acids

C. R. des travaux du Lab. de Carlsberg, 1879 and 1881.
Journ. Chem. S c., LVIT, p. 854, 1890.

These doct. Sciences Paris, 1890.

C.R. des Travaux du Lab, de Carlsberg, 1909.

Pete

INVERSION OF SACCHAROSE BY SUCRASE 133

(such as hydrochloric, ete.) markedly differentiate themselves from the
less highly ionised acids (sueh as phosphoric and acetic acids) in the
influence they exercise on the activity of this enzyme.

The present paper, therefore, contains a short account of the results
furnished by this method of enquiry.

PREPARATION OF SUCRASE

While it is impossible to prepare a solution of sucrase completely
exempt from substances which are capable of interacting with acids or
bases, yet by repeatedly adhering to an identical method of procedure,
preparations may be obtained which differ among themselves in a minor
degree only, and, therefore, probably possess a sufficient degree of purity
to permit of their being employed in experiments of a comparative
nature.

Following the method indicated by Duclaux, and utilised by
Fernbach!, cultures of Aspergillus niger were established on Raulin’s
solution and continued for four to five days at a temperature of 30° C.
The mycelium having attained maturity, generally indicated by the
commencement of blackening over its entire upper surface, the culture
liquid after careful decantation and thorough washing of its under
surface with distilled water? (the operation being repeated several times),
was replaced by a definite volume of distilled water. The plasmolysis
of the mycelium was thus established, and shielded from the light was
allowed to proceed for forty-eight to fifty hours at 30° C., and invariably
yielded an almost colourless solution of sucrase possessing a satisfactory
degree of activity. In this connection it is to be noted that during
plasmolysis of the mycelium it must not be submerged; to obtain a
satisfactory preparation of the enzyme by this means it is desirable that
only the under surface of the tough mycelial growth shall be in contact
with the plasmolysing liquid. In the preparation of the culture and the
subsequent operations of decantation, washing and maceration of the
mycelium, Fernbach’s flasks are specially serviceable. Each of these
flat-bottomed flasks, in which a volume of culture liquid having a depth
of about 2 to 5 cm. is placed, has a lateral tubulure by means of which
aeration of the mould during its growth may be conveniently carried

I. Loe. cit.

2. Throughout, in these cultural and also the inversion experiments, the water employed
was prepared by re-distillation, under diminished pressure, of ordinary distilled water with the
aid of a glass apparatus,

134 BIO-CHEMICAL JOURNAL

out. Under these conditions Aspergillus makes regular and vigorous
growth. Repeated experience of this, and trial of other methods of
culture, showed distinctly the superior advantages afforded by the
employment of the Fernbach flasks in these cultural experiments.
Under the above conditions it is possible to prepare an active
solution of sucrase entirely free from other disturbing organisms, but
often containing minute mycelial fragments which may still possess tardy
or limited powers of development. Such a solution exposed to free access
or even a limited volume of air, as experience of such preparations
retained under these conditions showed, leads more or less speedily to
diminution either of the quantity of active enzyme or of its activity.

The experiments, to be presently described, were carried out in
sections, each comprising a certain range of concentrations of the acid
under investigation, and the period of time involved in passing from
the initial to the final series of concentrations of acid covered several
days. Hence, in order to render the results of one portion of an
experiment comparable with those of another, a necessary condition
imposed was that the enzyme solution should be conserved in such a
way as to retain, as far as possible, its original activity. As it is almost
impossible rapidly to prepare solutions of enzymes which are quite
identical, the following method of procedure was followed. A number
of pipettes were made, each of such a capacity that its contents sufficed
for a single sectional series of experiments. After sterilisation, these
pipettes were filled with the sucrase solution and at once hermetically
sealed. The enzyme solution thus aseptically collected and protected
from the action of diffuse daylight, was conserved without undergoing
any very marked alteration even during considerable periods of time.

Mone or OPERATION

To study the influence of the different acids investigated on the
inversion of sucrose by sucrase, after a number of preliminary trials the
following method of experiment was adopted :—

Into each of a series of tubes exactly 5 c.c. of a 20 per cent. solution
of saccharose and definite volumes of acid and water were measured,
which, after the addition of the sucrase solution, brought the total
volume of the reaction medium to 10 c.c. Corresponding series of

INVERSION OF SACCHAROSE BY SUCRASE 135

controls were similarly prepared, with the addition of a solution of
passive suerase prepared by ®aising the enzyme solution to the boiling
point and retaining it at that temperature for some minutes. The
contents of each tube containing the solution of saccharose, together with
the requisite volume of water, were first allowed to acquire the
temperature of the latter. The requisite volume of acid was then added,
well mixed, and a definite and equal interval of time permitted to elapse
before the addition of the enzyme-solution, in order that inversion due to
the acid itself might be identical in the experimental tube and in its
corresponding control, and then, after addition of the enzyme solution,
the experiment was continued for one hour at 56°C. and terminated by
adding a very moderate excess of strong soda solution.

Finally, the contents of each tube were diluted to 50 c.c., and the
amount of inversion measured by determining the copper-reducing power
of a definite volume of this dilution.

Throughout the course of this work these determinations, carried
out in duplicate, were effected by means of the method of Mohr, as
modified by Bertrand.}

Carefully standardised solution of acids have been employed, and in
preparing the more dilute solutions for the weaker concentrations,
quantities of not less than 200 ¢.c. of each dilution have been freshly
prepared.

The volume of sucrase solution employed in each sectional series of
experiments has been so chosen as to produce under the given conditions
of time, temperature, etc., an amount of inversion of approximately the
same order of magnitude.

In the tables which follow, the quantities of saccharose inverted are
given in milligrams; these values representing the results obtained after
deducting the amounts furnished by the corresponding control

experiments with passive enzyme in the presence of similar quantities
of acid.

Experiments were first undertaken with representatives of the
‘strong’ acids in the physico-chemical sense, namely, sulphuric,
hydrochloric and nitric acids.

1. Bull. Soc. Chim., XXXV, p. 1285, 1906.

136 BIO-CHEMICAL JOURNAL

Concentration of acid H.SO, HCl HNO.
Sucrose inverted
(i) (i1) (i) (ii) (i) =~ (Gx)
0:00 (Control)! .. ae 1150 109-2 111°6 142.0 151-0 151-0
N°50 $2 6 = 9°8 9°3 0:00 0:00 6:6 6°6
60 se Be ee 21°8 20°7 9:3 14:0 10°3 8:7
70 Vs dts a 23:2 22:0 19°5 27-2 17°6 14:6
80 be a de 29-0 27:5 380 40°3 29°5 24-9
0:00 (Control)? .. te 120:0 114:0 — See 151-0 151-0
N:90 ae sis es 36°3 34:5 38°7 51-4 49:7 38°5
100 rh oe et. 40°5 38°7 52°7 76°9 61:7 61:7
200 Be SF os 80:0 76:0 81:8 135°8 1448 149-6
500 oy. oe sic 117°5 111°6 105°9 150-0 183°8 183°8
0:00 (Control) .. vs 117°5 hile) 1116 139°2 150°5 1510
N:'1000 eS, a8 5 119-0 113°5 117-5 155°5 181-9 1762
2000 st: Ss i 118-0 112-0 122°5 143°7 160°5 163°4
0°00 (Control) .. sh — oe a — — 171°9
N:3000 oe a S45 108°7 118°7 145°8 165°7 180°5
4000 Be A? a: 110-0 104°5 118°7 145°8 166°2 7e5
5000 Br = 115-0 109°2 97°8 142°5 159-6 1720

These experimental results show that these acids in minute amounts
favour, in a varying degree, the inversion of saccharose by sucrase. For
each acid there exists a range of concentrations within the limits of which
the action proceeds most rapidly, while the addition of gradually
increasing quantities of acid leads to its progressive retardation, until
ultimately when the concentration of the acid attains a value of about
N/50 practically complete arrest of the action ensues. In general, amounts
of these acids corresponding to concentrations ranging from N/1000 to
N /5000 exercise a favourable influence on the action of the enzyme.

The method of measuring the amount of inversion by determining
the copper-reducing power of the products of change, evidently does not
permit of the defining of any one particular concentration as the optimal
one, because the change in the amount of saccharose inverted in the

_l. Control in this and succeeding experiments containing sucrose and enzyme without
addition of acid.

2. Although solution of enzyme employed here perceptibly stronger, results are not quite
comparable with those of first experiment.

INVERSION OF SACCHAROSE BY SUCRASE 137
passage from one dilute concentration of acid to the other of these
optimal ranges of concentrations is relatively small. Moreover, it is
probable that small amounts of substances capable of fixing a relatively
large proportion of the added acid now make their presence felt, and thus

ge pro} I

somewhat obscure the actual influence exercised by the acid on the action
of the enzyme.

The three acids so far examined are typically strong acids, and the
experimental results adduced show that their behaviour is very similar.

The influence of phosphoric acid on the progress of the action affords

pAOS] prog

a marked contrast to that exercised by either of the foregoing acids, as

the following experiment shows : —
Phosphoric acid

Concentration of acid Saccharose inverted
Milligrams
0:00 (Control) ee a 135°8
M/10 HO ai 8:0
M/20 ae e 44°3
M 40 Pr Sie 91°2
0:00 (Control) ve "r 164°3
M/60 Bie a 124-9
M/80 ‘ ne 140°6
M/100 i ire 152°4
M/200 af a 166°3
0:00 (Control) A a 144-4
M/500 me ‘iF 169-1
M/1000 we OG 163-4
M/2000 Fe ae 164°3
M/4000 as bic 158°6

To effect virtual arrest of the enzymatic action, a concentration of
acid equivalent to that of an M/10 solution of the acid is required. The
range of optimal concentrations evidently lies within the limits of
M/200 and M/4000. The acid behaves somewhat similarly to a
monobasic acid, apparently containing only one active hydrogen atom, a
point already indicated by the work of G. Bertrand.!

The following experiment with KH,PO, lends support to this
supposition, for in the experimental condition employed this salt is

without any influence on the action of the enzyme.

lL. Loc. ctt.

138 BIO-CHEMICAL JOURNAL

Potassium dihydric phosphate

Concentration of salt Saccharose inverted
Milligrams
0:00 (Control) ar ee 159°6
M/5 me ae 160°5
M/10 a a 159°6
M/20 ae BE 159°6
M/40 a Bin 159°6

The influence exercised by acetic acid on the action is peculiar.
When comparison is made with the other acids examined, the principal
feature exhibited is that relatively large amounts of this acid, as already
evidenced by the investigations of Fernbach, + favour inversion instead of
retarding it. This favourable influence is manifested, as the following
experiment shows, over a wide range of concentrations. In the presence of
amounts of acid corresponding to that of N/500 to N/2000 solutions, this
favouring influence, however, declines somewhat.

Acetic acid

Concentration of acid Saccharose inverted
Milligrams
0:00 (Control) ae ae 121°6
N/1 Be Be ie?
N/2°5 ss ait 119°7
N/5 “i ea 128-2
N/10 ee Se: 133°9
N/20 ee i 137°7
0:00 (Control) = Sys 120°6
N/40 ote Ag 130°1
N/60 es Be 130°0
N/80 oe se 132°0
N/100 a a 133°0
0:00 (Control) ie ac 151°0
N/200 os ms 167°2
N/500 oe 2 159°6
N/1000 ae fe 156°7
N/2000 f x, 155°8

ii COC mCe:

INVERSION OF SACCHAROSE BY SUCRASE 139

SUMMARY

The results described in this paper show that within the limits
represented by certain concentrations, each of the acids examined
exercises a favourable influence on the inversion of saccharose by
sucrase. The amounts of acid which exercise this favouring influence,
as well as the extent of the latter, exhibit considerable variation, and
apart from other considerations are apparently dependent on the nature
of the acid and the proportion of enzyme employed.! Thus, while the
action proceeds most rapidly in the presence of comparatively small
quantities of sulphuric, hydrochloric, nitric and phosphoric acids, a
similar result also ensues even where relatively much larger amounts of
acetic acid are included in the reaction medium.

The retarding influence exhibited by these various acids also differs
in a similar sense. It is most marked in the case of sulphuric,
hydrochloric and nitric acids; amounts of these acids in the reaction
medium corresponding to that of an N/50 solution being sufficient to
almost completely arrest enzymatic action. Phosphoric acid behaves
similarly when the quantity present is equivalent to that of an M/10
solution of this acid. The behaviour of acetic acid in this respect is
singular; even when the amount present corresponds to that of an N/1
solution the action, instead of being completely arrested, is merely
retarded to a pronounced extent.

In general the results demonstrate that the retarding and arresting
influence exercised by the acids investigated on saccharose-inverting
action of sucrase differs in a marked degree from that observed by

1. In view of the work of S6rensen, it appears to be not improbable that the observed
changes in the progress of inversion of saccharose by sucrase in presence of varying quantities
of the different acids examined actually turns, not on the degree of the acidity of the reaction
medium, but on the concentration of the hydrogen ions present at each stage of the experiment.

On such an assumption the differences exhibited in the optimal range of concentrations

possessed, for example, by acetic and phosphoric acids, as compared with the other acids investi-
gated might perhaps be brought into line.

While the results of this enquiry confirm the favourable influence exercised, it is equally
evident that the difficulty of determining the optimal”range of concentrations for a given acid
becomes a matter of difficulty on account of the disturbing influence of minute amounts of
foreign substances unavoidably present in the enzyme solution, the influence of these being
progressively more marked with diminution of the quantity of acid employed.

140 BIO-CHEMICAL JOURNAL

G. Bertrand! with these reagents in the case of laccase, the oxidising
action of this enzyme being completely arrested by very minute amounts
of sulphuric, hydrochloric, phosphoric and acetic acids.

In conclusion the writer desires to record his acknowledgements to
“M. Gabriel Bertrand for the laboratory facilities he so kindly placed at
his disposal, and also for much critical and useful advice generally

accorded during the progress of this enquiry.

Ll. Loe: crt.

141

THE PHYSIOLOGICAL ACTION OF INDOLETHYLAMINE
By P. P. LAIDLAW.

From the Wellcome Physiological Research Laboratories, Brockwell

Hall, Herne Hill, S.E.

(Received September 7th, 1911)

The indolethylamine with which this paper deals is 3-f-amino-
ethylindole. It is the amine corresponding to the amino-acid
tryptophane. A number of amines formed from the native amino-acids
by the elimination of CO, have been shown to be physiologically active
substances.!: 2: 3, 4.

It appeared probable that indolethylamine would be physiologically
active, and a brief investigation of its action on the normal animal be of
some interest. The synthesis of this amine was accordingly undertaken
by Mr. A. J. Ewins,° to whom I am indebted for a supply of material for
physiological experiments. At the same time I prepared a small quantity
of the base by putrefaction from tryptophane (see Note on p. 150).

The Intact Animal. Subcutaneous administration of the
hydrochloride of indolethylamine to cats and rabbits gives rise to
practically no symptoms, a rapid heart is the only certain symptom
observable after 100 mgm. doses. (In the cat, symptoms of nausea,
uneasiness, salivation, etc., may occur.) It will be seen later that
indolethylamine produces vaso-constriction; its absorption after this
method of administration will therefore probably be somewhat slow; the
effects it produces are of short duration, and hence symptoms will not be
well marked unless very large doses are given by this means. Ten
milligram doses were given to rabbits by the marginal ear vein. In
each case, immediately after the injection, a spastic condition of the
limbs developed, on which was superimposed a fine, rapid tremor of fore
and hind limbs; this condition persisted for about one minute and then
rapidly disappeared. The respiration became slow, temporarily, and then
recovered. The heart one minute after the injection was accelerated. A

Dale and Dixon, Journ. of Phys., XX XIX, p. 25, 1909.
Barger and Dale, Journ. of Phys., XLI, p. 19, 910.
Ackermann and Kutscher, Zeit. fiir Biol., LIV, p. 387, 1910.
Dale and Laidlaw, Journ. of Phys., XLI, p. 318, 1910.
Ewins, Trans. Chem. Soc., XCIX, p. 270, 1911.

et

142 BIO-CHEMICAL JOURNAL

small cat was given 20 mgm. of indolethylamine hydrochloride
intravenously (long saphena vein), and a remarkable series of symptoms
ensued, Within thirty seconds of the administration violent convulsive
movements of limbs and body occurred, quite suddenly these clonic
spasms became tonic: the fore limbs were spread out straight in front,
paws off the ground and claws protruded; hind limbs stiff, flexed on the
trunk, and claws extended. <A fine tremor was present in all limbs. The
pupils were dilated, salivation was marked, and although the animal had
purred contentedly during the injection it now exhibited a high degree of
excitement. These symptoms were maximal in one minute, and passed
off in about three minutes. The cat sat up, all signs of muscular spasm
had disappeared, it was once more docile and purred when stroked.
Salivation continued for three or four minutes more. The pupils became
small slits about three minutes after the injection and remained so even
in dull illumination. Fifteen minutes after the first symptoms the cat
appeared to be perfectly normal. The heart beat was good throughout
the experiment, but respiration was severely interfered with during the
convulsive stage, being jerky and irregular. It is very probable that a
slightly larger dose would have caused death from respiratory failure.

The convulsive movements seen in the intact animal on intravenous
administration of indolethylamine appear to be due to a transient
stimulation of the central nervous system. <A very similar series of
muscular movements is seen when a spinal cat, under artificial respiration,
is given a small dose intravenously. On complete destruction of the cord
these symptoms disappear. It is noteworthy that the effect is peculiar
to warm-blooded animals. I*rogs receiving doses of 10 mgm. showed no
sign of convulsions. The symptoms shown by the latter animal are a
gradually increasing depression culminating in coma.

The Vascular System. Indolethylamine produces a large rise of

blood pressure in the spinal cat when administered intravenously (see |

fig. 2). The pressor effect has a superficial resemblance to that produced
by

accompanied by an increased rate of heart beat and the return to normal

a small dose of adrenine. ‘The rise of pressure is very rapid, and is

follows quickly upon the attainment of the maximum. Tonic contraction
and tremors of trunk and limb muscles also occur, which may help in
producing the rise of pressure through raising the intra-abdominal
pressure. This, however, is quite a subsidiary factor since large rises of
pressure are obtainable with the abdomen laid open. Direct comparison
of rises of pressure produced by equal doses before and after opening

PHYSIOLOGICAL ACTION OF INDOLETHYLAMINE 148

abdomen is untrustworthy, since a second dose of the base rarely
produces quite as large an effect as the first. In the animal anaesthetised
with volatile anaesthetic the rise of blood pressure is never as great as
in a spinal animal. The normal blood pressure is higher, the cardio-
inhibitor mechanism prevents any large rise, the anaesthetic depresses the
heart’s response to most drugs, and, therefore, a large rise of blood
pressure is impossible. Moreover, large doses cannot be given in this
condition since the respiratory centre, already somewhat depressed by
the anaesthetic, may be paralysed by comparatively small doses of the
base. Five milligrams may cause respiratory arrest in an anaesthetised
animal, while an intact cat, as shown above, survived 20 mgm. The
difference can only be due to the anaesthetic depression of the respiratory
centre.

The rise of blood pressure is due to vaso-constriction and increased
cardiac activity. Fig. 1 shows a tracing of the isolated heart of a rabbit,
together with a drop record of the coronary outflow; between the arrows
3 mgm. of indolethylamine were injected into the perfusion cannula. It
will be observed that there is a considerable rise in the tone of the heart
muscle, accompanied by a greatly increased rate of heart beat; a count
of the heart beat shows that the rate is increased by one-half (from 18 to
27 in equal time intervals). The coronary outflow is quicker. It is
probable that the last-mentioned feature of the tracing is due to the
increased rate of heart beat, and not to an actual dilatation of the coronary
vessels. The heart beat exerts such a profound influence on the coronary
circulation that the increase seen in this tracing could easily be explained
in this way. Moreover, as will be seen later, all forms of plain muscle,
elsewhere in the body, which respond to the new base, show motor effects
irrespective of innervation, and it is improbable that the coronary vessels
behave in an exceptional manner. ‘The vaso-constriction may be observed
on inspection of the viscera after intravenous administration, and also
after local application of a 1 in 1000 solution to a highly vascular surface.
The effect is not nearly so striking as that produced by adrenine. <A
diminution of viscus volume is not readily demonstrated in oncometer
experiments. As a rule, the first effect is constriction, but this is often
followed by a secondary dilatation. This diphasic response is probably
due to the vessels in the plethysmograph giving an abnormal response.
The exposure, cooling, and manipulation necessary in the introduction
of a viscus into a plethysmograph frequently destroys the sensitiveness
of the preparation. The initial constriction is the true result, and the

JOURNAL

MICAL

7
4

BIO-CHI

144

a0

VECO RCL EL OL CRE CULE LE EEEE LEER EEE EE ER EER CCE SEER EERE DR CUCCEL CORLL AAL

mie i wae
Feet BRS

ace

AVA U OUR UCR CALERA CURLER CCR COR CURA RUUAL DANY \\SAANUN ADAG YASALSGAALUALI CAUEUREAAU UAE BE AUER TULLE LUNE EEL ELUVER EEUU U ERLE URN ERLE EELS
DVLA VAS AS VATA AMARA DARD UU ULV OED ALAN VADLD RAND URULDA LADD OD DOE RVDRDVDDANDANIMVDAND( rDUDAUDNNDLLNARLDAALAMDDANDLDDDVU)VUDAVAANAVEVLLLG KUED EM | AVNSUVYNVVALDVN MD 4 4. VVVANNNANSNVANVAVAA IVAN NS

M6) havuoso)

dd

|

PHYSIOLOGICAL ACTION OF INDOLETHYLAMINE 145

dilatation following is due to the large rise of blood-pressure dilating
arterioles which, though stimulated to constrict, owing to abnormal
conditions are unable to do so against a large rise of blood pressure.
Occasionally a pure diminution in volume is observed. There can be no
doubt that considerable vaso-constriction does occur, but the exact
method by which this vaso-constriction is brought about is not so certain.
The vaso-constriction is undoubtedly due in part to a peripheral effect
of the amine upon the plain muscle of the arterioles, because after large
paralytic doses of nicotine or curare have been given to a spinal cat, a rise
of blood-pressure is still obtainable with 3 or 4 mgm. indolethylamine.
This rise is, however, diminished by about half by the administration of
nicotine. In this experiment the ganglion cells and the paths from cord
to periphery are paralysed completely. It is possible that the diminution
in response to the new base brought about by paralytic dose of nicotine
is an expression of the cutting-out of vaso-constrictor impulses of central
origin. A similar effect is observed after paralytic doses of ergotoxine.
In cats which had received ergotoxine in sufficient quantity to reverse
the adrenine response (Dale!) a pressor effect is still elicited by
indolethylamine, but this pressor effect is not nearly so striking as that
observed previous to the administration of ergotoxine. If one could
regard nicotine and ergotoxine as having a selective paralytic action on
the nervous structures, a central factor in the pressor effect induced by
indolethylamine would be proved. Unfortunately nicotine and ergotoxine
have a decided effect on plain muscle itself. As far as these experiments
are trustworthy they suggest a central factor of some importance. This
suggestion is borne out by the following further evidence. In a spinal cat
preparation, the rise of blood pressure brought about by the intravenous
administration of 2 mgm. indolethylamine was recorded. The cord was
then completely destroyed, and after a short interval the effect of the same
dose of indolethylamine was recorded once more. The effect of the second
dose was about half the first. The evidence of a general stimulant effect
on the cord, as illustrated by the generalised muscular spasms, has
already been described. The balance of evidence is decidedly in favour of
a mixed origin of the vaso-constriction: (1) central from vasomotor
centres in the cord, and (2) peripheral upon the plain muscle of the
arterioles.

One or two features of interest are met with in the responses of the

1. Dale, Journ. of Phys., XXXIV, p. 163,51906, and Bio-Chem. Journ., II, p. 240, 1907.

146 BIO-CHEMICAL JOURNAL

several organs of the body containing plain muscle; of these the uterus
and the eye are the most interesting.

The Uterus. Fig. 2 shows the typical action of indolethylamine
upon the virgin cat’s uterus 7m sitw, and upon the blood pressure. Just

after the rise in blood pressure has started to develop, the uterus

Fia. Il.—Spinal cat. Virgin. Effect of 2 mgm. indolethylamine.
Uterine Contractions. Blood Pressure.

commences to relax. This effect is not abolished by section of the
hypogastrics. This combination of effects is suggestive of an adrenine
type of action. But this is not borne out by further experiment, for if
the virgin cat’s uterus be isolated, suspended in Ringer solution, and

its movements recorded graphically, the opposite result is obtained. The

eo rc el lO el rt

PHYSIOLOGICAL ACTION OF INDOLETHYLAMINE 147

new amine always produces motor responses from a virgin cat’s uterus in
this condition. Fig. 3: In®the pregnant cat a well-marked motor
response is obtained on administration of indolethylamine. Fig. 4: The
rabbit’s uterus and the guinea-pig’s do not respond well to the new base.
In two rabbits, where adrenine and other drugs produced well-marked
contractions of the uterus, indolethylamine was without action. This
difference in action on the virgin cat’s uterus in situ and when isolated,
is one that is not readily explained. If it were demonstrable that the new
amine had an action resembling that of nicotine, the explanation is
obvious. The result in any given experiment would be the sum of two
variables: (1) stimulant action on plain muscle, hence motor response in
the isolated organ; (2) stimulation of sympathetic ganglion cells or
peripheral neurone supplying the uterus. The dominant supply in the
case of the virgin cat is inhibitor, hence relaxation 7m situ. The dominant
supply in the pregnant cat is motor, and thus the nervous influence aids
the muscular one. Experiments with a view to demonstrating a
nicotine-like action of indolethylamine on the superior cervical ganglion
or on the splanchnic system all failed. It is quite possible that this
substance has an action on the peripheral neurone supplying the uterus
as well as on the muscle of the organ, but not on the peripheral
sympathetic neurones elsewhere. Very similar effects have been described
by myself with a number of other substances (hydrastinine,
cotarnine, and 6:7-dimethoxy-2-methyl-5:4-dihydro isoquinolinium
chloride), and the same explanation was tendered. There is some
peripheral structure which does not survive excision, and determines the
action of many substances upon the cat’s uterus in situ. From the
manner in which the response varies with the dominant nerve supply to
the uterus, and the close parallelism which exists between the action of
these substances and that of nicotine on this organ, it is suggested that
the unknown peripheral structure is nervous, probably the peripheral
neurone. It is, unfortunately, impossible to submit the problem to the
crucial experiment of excision of these ganglion cells, owing to their
scattered peripheral distribution.

The Iris. Upon the plain muscle of the iris indolethylamine has a
well-marked effect in doses of 10 to 20 mgm. If a dose of this size be
given intravenously to a cat with pithed brain, the pupil becomes

‘constricted to a fine slit (fig. 5). This contracted pupil persists for

several minutes. Smaller doses produce smaller effects which are less
lasting. A further dose of 20 mgm. produces a dilatation of the pupil,
and the elongated oval opening tends to become circular. This dilatation

BIO-CHEMICAL JOURNAL

D
fe)

=
-—r
= |

+}
QO.8
AY
o

P|

Ml fr fi i) ih |

ii Mii | W

| Fic. I11.—Isolated uterus. Virgin cat. Effect of 10 mgm. indolethylamine in 250 c.c. Ringer’s
solution.

PHYSIOLOGICAL ACTION OF INDOLETHYLAMINE 149

gradually gives way again to profound constriction, and once more a thin
black slit is all that is visible. Stimulation of the cervical sympathetic
can still produce some dilatation at this stage. lull doses of atropine do not
abolish the slit-like pupil or prevent its appearance. It must, therefore,
be regarded as of direct muscular origin. The tendency to form a circular
pupil after the second injection indicates a simultaneous motor effect
upon both sphincter and dilator muscles, the sphincter muscle being
the more powerful, ultimately overcomes the dilator muscle and produces
the slit-like pupil. The effect is not observable in cats anaesthetised with

chloroform or ether, nor in an intact cat which received 100 mgm.

aN

Fic. I1V.—Cat. Ether. Pregnant. Effect of 10 mgm. indolethylamine. Uterine Contractions.
Blood Pressure.

Fic. V.—Cat. Pithed brain. Circulation to left eye impaired by cannula in left common
carotid. Effect of 20 mgm. indolethylamine on pupil of right and left.

hypodermically. It can only be produced by intravenous administration
to an intact animal or one which is anatomically anaesthetised.
Instillation, like hypodermic administration, is without effect.

Other plain muscle. The plain muscle of the intestine is mildly
stimulated by indolethylamine. The effect is better shown in the isolated
organ, but even in this condition the effect of 20 mgm. of the base in
250 c.c. Ringer solution is very small. The retractor penis of the dog
also gives weak motor responses to the amine when isolated from the
body. The plain muscle of the bladder is thrown into contraction on
administration of 2 or 4 mem. of indolethylamine; and a bladder, which
before administration of the base was quiescent, may develop a slight
rhythm. The amount of contraction of the bladder volume under the
influence of indolethylamine is increased on destruction of the cord.

Lhe salivation observed in cats receiving hypodermic or intravenous

150 BIO-CHEMICAL JOURNAL

doses of the new base must be central in origin, for no evidence of
salivary secretion could be obtained in the anaesthetised animal. The
pancreatic secretion in the dog is unaffected by indolethylamine. Ten
milligrams of indolethylamine were given intravenously to an
anaesthetised cat in which the urinary secretion was being recorded. A
marked slowing of the urine flow was observed during the rise of blood
pressure, which, however, soon passed off. The effect is in all probability
the expression of a transient constriction of the arterioles of the kidney.
Experiments in conjunction with Mr. A. J. Hwins, are in progress with
regard to the metabolism of the new base.

Nore oN THE ForMATION OF INDOLETHYLAMINE FROM 'T'RYPTOPHANE

It has been shown by several observers that putrefactive miucro-
organisms will remove CO, from amino-acids and form the corresponding
amine. It was thought possible that indolethylamine might be made in
this manner from tryptophane. A number of experiments were carried
out with different mixtures of bacteria: only one of them proved
successful. 0°5 gram tryptophane was dissolved in 260 c.c. of a simple
culture medium.!' This was then infected from a putrid pancreas
subculture which had been shown to be capable of forming f-iminazolyl-
ethylamine from histidine and p.hydroxyphenylethylamine from
tyrosine. After one fortnight’s incubation the mixture was found to
produce a good rise of blood pressure, and indolethylamine could be
isolated from it. The mixture was boiled with charcoal and filtered. It
was then evaporated to about 100 c.c. and excess of picric acid added. On
cooling, the highly insoluble indolethylamine picrate separated out as a .
deep orange-red crystalline precipitate. The crude product was then
recrystallised from alcohol and again from aqueous acetone, when
crystalline form, colour reactions, melting point, and physiological
action were found to be identical with those of the synthetic product.
The yield from the one experiment which went in the desired direction
was poor, about 140 mgm. of pure picrate being obtained from 0°5 gram
tryptophane.

SUMMARY

(1) Indolethylamine produces a transient stimulant effect upon the
central nervous system, causing clonic and tonic convulsions, tremors of
limbs, and vaso-constriction.

(II) It has a direct stimulant action on plain muscle, which is most
marked in the arterioles, the iris, and the uterus.

(III) The formation of indolethylamine from tryptophane by
bacterial action is described.

1. Peptone 2 grm., dextrose 8 grm., trace sodium phosphate, trace magnesium sulphate,
precipitated calcium carbonate 5 grm., and tap water to the litre.

151

THE EFFECTS OF CAFFEINE UPON THE GERMINATION
AND GROWTH OF SEEDS

By FRED RANSOM, M.D., Ldinburgh, Beit Research Fellow.

From the Pharmacological Laboratory, Cambridge
(Received November 7th, 1911)

The following report contains the results of some preliminary
experiments made to ascertain the effect of subjecting seeds to the action
of caffeine.

The method adopted was to take two equal quantities of each sort of
seed and place one lot in 1 per cent. solution of caffeine in tapwater,
the other in tapwater alone. Both sets were then left for twenty-four
hours in a thermostat at 25°C. At the end of this period each portion of
seed was washed on a muslin filter with tapwater and sown in ordinary
coarse garden sand. Very small seeds were first mixed with a little dry
sand, so as to allow of equal distribution. The sowings were made in
red flowerpot saucers, placed in the laboratory windows and watered as
required. Care was taken that the conditions of light and temperature
were as far as possible equal. The results observed are given in tabular
form at the end of this report, Table I.

Although the list of seeds is comparatively short, still the results
obtained are so uniform that they can hardly be attributed to chance. In
every case except one (nasturtium) the seeds which had been treated with
caffeine were later in germinating than the tapwater controls, and in
a great majority of cases fewer seeds germinated in the caffeine groups.
In the course of growth the caffeine seeds were at first relatively less
vigorous than those from tapwater, and in many cases they remained so
till the end of the experiment. In three caffeine groups no seeds
germinated. In no instance except the nasturtium had the caffeine
groups any advantage over the controls.

The well-known action of caffeine upon striped muscle and its effect
on plant cells, as described by Bokorny,! show that it has a disintegrating
effect upon certain vegetable and animal proteins. In the case af
muscular tissue theré is considerable resemblance between the action o}
caffeine and that of chloroform, toluol, ether, aleohol, and similar bodies.
With plants, on the contrary, there would appear to be a notanje
difference. This may be shown by an adaptation of the method

152 BIO-CHEMICAL JOURNAL

described by Armstrong? for the rapid detection of emulsine, and by
Waller? for the colorimetric estimation of hydrocyanic acid. The
following, Table II, demonstrates the difference in action of two members
of the chloroform group on the one hand and caffeine and formaldehyde
on the other.

TABLE IT
Condition of Condition of
solution after solution after
3 hours at 37°C. 24 hours at 37°C.
Laurel leaf Na picrate solution 5 c.c. Unchanged Unchanged
immersed in Aqua 35 ¢.c.
s Na picrate solution 5c.c. Dark-red-brown ~ Dark-red- brown
Aqua 35 ©.¢.
-Chloroform 12 drops
3 Na picrate solution 5c.c. Dark-red-brown Dark-red- brown
Aqua aDIC.C:
Toluol Sec:
“6 Na picrate solution 5¢,c. Unchanged Unchanged
1% Caffeine solution 35 c.c.
# Na picrate solution 5c.c. Unchanged Unchanged
1% Formaldehyde 35c.c.
Bs Na picrate solution 5c.c. Unchanged Unchanged

10% Formaldehyde 35 c.c.

Even when the immersion in caffeine solution was prolonged to
forty-eight hours there was no production of hydrocyanic acid; but the
same leaf, placed afterwards in chloroform water with sodium picrate,
quickly turned the solution brown. The two protoplasmic poisons,
caffeine and formaldehyde, obviously act upon the laurel leaves in a
manner different from that of chloroform.

It seems not unlikely that caffeine may produce its effects both on
muscle and on seeds by increasing the activity of a protease. If this were
the case it might be expected that in small doses it would stimulate
germination and growth. Caffeine in dilute solution apphed to muscle
increases the response to stimulation, whereas in greater concentration it
induces rigor and death. I am inclined to ascribe the exceptional
position of the nasturtium seeds in these experiments to the density of the
seed-covering limiting the access of caffeine.

In the meantime the present paper brings evidence that caffeine in
1 per cent. solution may retard or even prevent the germination and
srowth of seeds. In a paper which has just appeared Bokorny* gives
some instances of the deleterious effects of caffeine upon seedlings.

Name of seed

Endive

Lettuce

9

Cress

Carrot

”

Radis

2

Spinach

>

Sweet peas

”

Peas

Previous
treatment
24 hours at

25° C. in

Tap water
1% Caffeine

Tap water
1% Caffeine

Tap water
1% Caffeine

Tap water
1% Caffeine

Tap water
1% Caffeine

Tap water
1% Caffeine

Tap water
1%, Caffeine

Tap water
1% Caffeine

Tap water
1% Caffeine

Tap water
1% Caffeine

TABLE I

~ Quantity

ee
Or Or
RP

O°
Crcr
08 8

=
bo te

10
10

Date of
planting

29/4/11

”

30/4/11

9°

99

”

Sprouts
first
visible

7/5/11
11/5/11

11/5/11
3/6/11

8 5/11
12/5/11

5/5/11
7/5/11

5/5/11
7/5/11

12/5/11
22/5/11

4/5/11
6/5/11

6/5/11
8/5/11

9/5/11
30/5/11

6/5/11
12/5/11

THE EFFECTS OF CAFFEINE UPON SEEDS 153

Condition 4 weeks
after planting

Fewer seeds have ger-
minated in Caffeine
group. The growth
in both groups’ is
about equal.

Fewer seeds have ger-
minated in the
caffeine group, and
they are shorter and
thinner than those of
the tap-water group.

Fewer seeds have ger-
minated in the
caffeine group. The
growth in both
groups about equal.

Fewer seeds have ger-
minated in the caffeine
group. Growth about
equal,

About the same number
have germinated in
each group, and the
seedlings are about
equally vigorous.

Fewer seeds __—ihave
germinated in the
caffeine group, and
they are less vigorous
than those of the tap-
water group.

About the same number
of seeds have ger-
minated in each croup
and the seedlings are
about equally vigor-
ous.

About equal in number
and vigour.

Nine seeds of tap-water
group have germi-
nated, average length
12 inches; four seeds
of caffeine group, with
average length 4
inches.

Of tap-water group six
seeds have germi-
nated, average length
22 inches ; of caffeine
group five seeds have
germinated, average
length 18 inches,

154

Name of seed

Broad beans

29

Nasturtium

9°

Viscaria

99

Helianthus

99

Nemophila ...

99

Wheat

39?

Barley

Oats

Canariensis

9°

Eschscholtzia

99

Zinnia

9

BIO-CHEMICAL JOURNAL

Previous
treatment
24 hours at

25°'C: in

Tap water

1% Caffeine

Tap water
1% Caffeine

Tap water
1% Caffeine

Tap water
1% Caffeine

Tap water
1% Caffeine

Tap water
1%, Caffeine

Tap water
1% Caffeine

Tap water
1% Caffeine

Tap water
1% Caffeine

Tap water

1% Caffeine

Tap water
1% Caffeine

Or or

NA
oe 09

rn
09 08

—
5 35
ww

Sprouts
first
visible

12/5/11
Nil.

30/5/11
21/5/11

16/5/11
18/5/11]

13/5/11
18/5/11

10/5/11
16/5/11

30/5/11
4/6/11

30/5/11
Nil

11/5/11
31/5/11

16/5/11
Nil

18/5/11
25/5/11

15/5/11
17/5/11

Condition 4 weeks
after planting

Six seeds of tap-water
group have germi-
nated, vigorous plants.

Six seeds have germi-
nated in the tap-water
group. nine in the
caffeine. Both are
about equally vigor-
ous.

Fewer seeds have
germinated in the
caffeine group. Both
groups about equally
vigorous.

Fewer seeds have
germinated in the
caffeine group. Both
groups about equally
vigorous.

Many less seeds have
germinated in the
caffeine group. Both
groups equally vigor-
ous.

Ten seeds have germi-
nated in tap-water
group. One in the
caffeine group.

Six seeds have germi-
nated in the tap-water
group.

Eleven seeds have ger-
minated in the tap-
water group, two in
the caffeine group ;
the former are the
stronger.

Six seeds have germi-
nated in the tap-
water group.

Many more seeds have
germinated in the
tap-water group, and
they are much the
stronger.

In the tap-water group
thirty-one seeds have
germinated, in the
caffeine group nine-
teen; the latter are
the weaker.

Name of seed

Marigold

”

Gaillardia

Chrysanthemum

eo

Linum rubrum

”

Convolvulus

’

Sweet Rocket

Coreopsis

be

Lavatera

2?

Nasturtium

”

ia ae

Previous
treatment
24 hours at

25° C. in

Tap water
1%, Caffeine

Tap water
1°, Caffeine

Tap water
1%, Caffeine

Tap water
1%, Caffeine

Tap water
1%, Caffeine

Tap water
1%, Caffeine

Tap water
1% Caffeine

Tap water
1%, Caffeine

Tap water
1%, Caffeine

Bokorny, Pfléger’s Archiv
Armstrong, Journal of Physiology, XL, p. 22, 1910.

Quantity Date of
' planting
0-2 er, 7/5/11
0-2 g, ad
0-25 g. 2
0-25 g. ry
0-5 g. .
0-5 &, ;
0-5 g. ;
0-5 g. “
0-5 g. 20/6/11
0-5 g. 33
0-5 g. es
0-5 ¢. ae
0-3 g. 3
0-3 g. ~
0-5 g :
0-5 g ‘
16 :
16 :

REFERENCES

.» CXXXVII, p. 470, 1911.

3. Waller, Journal of Physiology, XL, p. 49, 1910.
4. Bokorny, Biochem. Zeitschr., XXXVI, p. 83, 1911.

THE EFFECTS OF CAFFEINE UPON SEEDS

Sprouts

first
visible
17/5/11
23/5/11

16/5/11
23/5/11

16/5/11
21/5/11

23/6/11
25/6/11

29/6/11
10/7/11

2/7/11
5/7/11

23/6/11
25/6/11

29/6/11
27/6/11

155

Condition 4 weeks
after planting

The tap-water group is
more numerous and
better grown than the
caffeine group.

The number of seeds
which have germi-
nated is about equal.
The tap-water group
is much more vigor-
ous.

Many more seeds have
germinated in the
tap-water group. and
the seedlings are
much stronger than
in the caffeine group.

Fewer seeds have
germinated in the
caffeine group, and
the seedlings are
weaker than in the
tap-water group.

The seedlings in the
tap-water group are
stronger and more
numerous than in
the caffeine group.

The seedlings in the
tap-water group are
stronger and more
numerous than in the
caffeine group.

The seedlings in the
tap-water group are
stronger and more
numerous than in the
caffeine group.

The seedlings in the
tap-water group are
stronger and more
numerous than in the
caffeine group.

The seedlings in the
caffeine group are
more numerous, both
groups about equally
vigorous.

156

THE ACTION OF CAFFEINE UPON THE GERMINATION
AND GROWTH OF SEEDS

By FRED RANSOM, M.D., Ep., Bert Research Fellow.

From the Pharmacological Laboratory, Cambridge
(Second Communication)

(Received January 17th, 1912)

In the following communication an endeavour is made to throw
light upon the function which caffeine exercises in plant economy. At
present the reason for the storage of caffeine in certain parts of some
plants is quite as unexplained as is the appearance of its allies, xanthine
and hypoxanthine, in animal tissues. On the other hand, the fact that
it is found chiefly in the seeds and in young and growing shoots suggests
that it may have something to do with germination or with growth.

In a previous paper I have shown that 1 per cent. solution of
caffeine has a marked effect in delaying or preventing germination in a
considerable variety of seeds. The following experiments unreservedly
confirm those results. There was, however, the possibility that by
treating seeds with very dilute solutions of cafteie a stimulating effect
might be produced on germination or on growth or on both. The
settlement of this point was one object of the present investigation.

The method adopted was the following:—On the bottom of glass
needle boxes a thin layer of absorbent cotton-wool was placed, over this
a piece of fine muslin was laid and the whole moistened with 50 c.c. of
caffeine solution or tapwater. The seeds were then scattered thinly upon
the muslin. The needle glasses were closed with loosely-fitting glass
lids and stood either in the laboratory at about 17° C. or in a thermostat
regulated to 22°C. Usually a series of glasses were charged with
caffeine solutions of varying strength, with tapwater controls. At
certain intervals the growths were compared by examination with the
eye and, finally, the seedlings were carefully removed, either with or
without the ungerminated seeds, superficially dried on blotting paper
and weighed at once.

As regards germination it was found that caffeine solutions of
10, O'1, 0°01 per cent. caused very distinct delay. With 0-001 per cent.
that was not so obvious, and seeds sown in 0:0001 per cent. germinated
as quickly as the tapwater controls (Table I).

Besides delay, the stronger caffeine solutions had a very marked

THE ACTION OF CAFFEINE UPON SEEDS 157

inhibitory effect, so that of the seeds sown in 10, 0°1, and 0°01 per cent.
fewer germinated than in the weaker solutions or in tapwater (Table IT).

Very dilute solutions of caffeine, 0°0001 per cent. and less, were
without any cbvious effect at all, either on germination or on growth. As
will be seen from the table, 0°001 per cent. appears to have a slight effect
in hindering growth (Table III).

The effect of caffeine in solutions of about 0°01 per cent. or more, was
such that the seedlings, weighed after two or three days’ growth, were
distinctly lighter than the tapwater controls (Table IV). In most of the
following tables the same effect can be observed.

It being thus evident that caffeine added to tapwater does not
stimulate germination or growth; further experiments were made to
ascertain whether a different effect would follow when a salt or other
substance was added to the caffeine solution.

Nirrate or PorasH—SaccuaroseE. (All solutions in tap water)
Cress 0-5 g. in each glass and 50 c.c. fluid. Sown on 6/12/1911. Room temperature

Medium ~ Condition on Condition on Weight on
8/12/1911 9/12/1911 12/12/1911
(1) KNO, 0:05 % s=«. ... Land 4 are the best Relatively about the 2-620
same
(2) KNO, 0:05 % side ... 2 and 3 about equal 2-385
Caffeine 001% ... sat but smaller than
the others
(3) Caffeine 001% ... Son 2-590
(4) Tap water ... 20 50 2-830
(5) Saccharose 2% ... ... 5 and 6 are about 6 is smaller than 5. 2-620
equal and less Both are less
than 4 than 4

(6) Saccharose 2 %
Caffeine 0:01 °%

The table shows very distinctly the influence of caffeine in retarding
growth, but there was no stimulation.

2-350

Dextrose. (All solutions in tap water)

Mustard 1-0 g. seed in each glass and 50 c.c. fluid. Sown on 12/12/1911. Room temperature

Medium Condition on 14/12/1911 Weight on 16/12/1911
(1) Dextrose 0:8% ... ... 5 and 6 are the best grown and 3-100

are about equal
(2) Dextrose 0:2 %

ie Bae 3-730

(3) See WEA re ... 1 and 3 are the smallest 3-040
affeine 0-001 % ... or

(4) aia Oo eee ... 2 and 4 are about equal 3-670
affeine 0-001 % ... Ay.

(5) Tap water ... 2 Stic “er 4-420

(6) Caffeine 0-001 % ... re 4-320

There is here no evidence whatever that the caffeine combined with
dextrose stimulates, on the contrary what little difference there is
indicates retardation. The best growth took place in tapwater, followed
closely by the 0°001 per cent. caffeine solution,

158 BIO-CHEMICAL JOURNAL

Auanin. (All solutions in tap water)
Mustard 1-0 g. seed in each glass and 50 c.c. fluid. Sown on 16/12/1911. Room temperature

Condition on Number of not Weight, 20/12/1911
Medium 19/12/1911 germinated without not
in order of merit 20/12/1911 germinated
Alanin 0°5 % eta CBest 39 2-750
Tap water
‘Alanin 0-1 % | 0-0001 °% caffeine 27 3-465
Alanin 0:5 % 0-001 ,, caffeine 62 2-265
Caffeine 0-01 % | .
Alanin 0-5 % ae ae {0-1 » alanin 30 3-050
Caffeine 0-:001% ... at (0-0001 ,, caffeine
Alanin 0:5 % poe {0-1 , alanin 46 2-610
Caffeine 0:0001 % ... | (0-001 ,, caffeine
c 0-1 », alanin
Alanin 0-1 % s | (0-1 » alanin 27 3-190
Caffeine 0-01 % o {0-01 ,», caffeine
°{ 0-01 ,, caffenie
Alanin 0-1 % | 34 3-550
Caffeine 0-001 % |
| = {0-5 , alanin
Alanin 0-1 % a ae | = } (0-0001 ,, caffeine 25 3-380
Caffeine 0-0001 % ... 31/05 ,, alanin
| (0-001 ,. caffeine
Caffeine 0-01 % 22 3-850
Caffeine 0-001 °% 4 (0-5 5, alanin 24 3°850
ene 0:01 4, caffeine
Capoeo oles... Vuelos © 2 ann 25 4-355
Tap water ... ane ... Worst 19 4-420

The table shows that more seeds germinated and the seedlings
developed best in tapwater. There were no signs of stimulation, on the
contrary the combination alanin-caffeine was more deleterious than
caffeine alone.

Witte’s Perrone. (All solutions in tap water)
Mustard 1-0 g. in each glass and 50 c.c. fluid. Sown on 19/12/1911. 22° C.

Medium Condition on Condition on Weight on
20/12/1911 21/12/1911 21/12/1911
(1) Peptone 05% .... ... 9 and 6 have begun 6 is much the best 3-410
to sprout
(2) Peptone 0-5 % ae aes Then | and 2 3-280
(3) Peptone 05% ... ola srome aie Then 5 2-830
Caffeine 0:02 % ... soe hardly any
sprouts
(4) Piers aoe oe oe 3 and 4 are the least 2-700
affeine 0:02 % ... as grown
(5) Tap water ... nes be : 2 900
Caffenie 0:02 % ... oH

(6) Tap water ... =. Ses 4-800
The seeds in tapwater developed best, and the combination of peptone

with caffeine was more inhibitory than either peptone or caffeine alone.

THE ACTION OF CAFFEINE UPON SEEDS 159

Wrrre’s Perrone. (All solutions in tap water)

Mustard 10 ¢. in each glass and 50 c.c. fluid. Sown on 20/12/1911. 22°C.

Medium Condition on Condition on Weighed on Arranged in order
21/12/1911 22/12/1911 22/12/1911 of merit
(1) Peptone 071% . Ll and 12 have 12 is the best 3°220 Tap water, best
germinated ;
best. 00001 ©, caffeine
(2) Peptone 0:5 % Then 9, 10, 11 2°960 '
0-01 caffeine
(3) Peptone 0°5% . Then9 and 10 Of the rest, 1, 2, 2-820 i
Catfeine 0-01 ©, 8, 7,6 are 0-001 °%, caffeine
better than
3, 4,5 01% P.+-0-0001%,
caffeine
(4) Peptone 0°59 — ... The rest are 2°930
Caffeine 0°001°) ... fairly equal, 0-1 % P.+0°01%
5 being the caffeine
least grown
(O'1 Yee
(5) Peptone 0-59)... 2-690 + 0-1 % P.+0°G01%
Caffeine 0-0001 %... caffeine
(6) Peptone 0-1 % 3:370 Ori e/cakes
Caffeine 0-01 °%
0-5 P.+0-001 %
(7) Peptone 0:1% ... 3:220 caffeine
Caffeine 0-001 % ... .
0:5 % P.+0-01 %
caffeine
(8) Peptone 0:1% __... 3-410 ok
Caffeine 0-0001 °%... 0-5 % P.+0-0001%
caffeine
(@) Caffeine 0-01 °% 3-650
(10) Caffeine 0-001 % ... 3-600
(11) Caffeine 0-0001 °%... 3-840
(12) Tap water os 3-890

The best results were obtained with tapwater followed closely by the
weakest caffeine solution. The combination of caffeine with peptone was
more inhibitory than either caffeine or peptone alone.

CONCLUSIONS

The foregoing experiments afford ample evidence that caffeine added
in the proportion of from 1 per cent. to 0°01 per cent. to water, in which
seeds are then sown, exercises a powerful effect in retarding germination
and growth. If as much as 1 per cent. caffeine is present there may even
be complete inhibition of germination.

On the other hand there is nothing at present to show that caffeine,
either alone or in combination with other substances can act as a
stimulant to plant life. Nevertheless, the subject is not exhausted and
the investigations are being continued.

BIO-CHEMICAL JOURNAL

TABLE [
Sown on 31/10/1911, at Room temperature

Cress, 0-5 g. in each ~—- Radish, 18 seeds in

glass each glass
Medium Condition on Condition on
3/11/1911 4/11/1911

1 % caffeine in water . None germinated

0-1 % caffeine in water . A very few small
shoots

0-01 % caffeine in water . The greater part
sprouted 2 mm.

0-001 °% caffeine in water . The greater part
sprouted 3 mm.

Equal. Almost all

0-0001 °% caffeine in water
sprouted 3-4 mm.

|
Tap water only awe }

Mustard, 0-5 g. in each glass. Sown on
10/11/1911, at room temperature

Medium Condition on Medium

13/11/1911
0-1 % caffeine A few very small
shoots

0-1 % caffeine

0-02 % caffeine Mostly sprouted, 0-01 % caffeine
but small

0-004 °% caffeine Mostly sprouted,

longer

Mostly sprouted.
The best
grown. Equal

0-0008 % caffeine

Tap water only | Tap water only

TaBLe II

Sown
Room Temperature

Mustard, 0-5 g. in each glass.

on 31/10/1911. glass. Sown on

0-001 % caffeine

0-0001 % caffeine

Radish, 18 seeds in each

None germinated

None germinated

A few very small
shoots

The greater part
sprouted 2 mm.

Equal. Mostly
sprouted 3 mm.

Lettuce, 0-3 g. in each glass. Sown on
25/11/1911, at room temperature

Condition on

28/11/1911

None germinated

Few and small

best grown

5/12/1911.

Room temperature

Number not

Condition of seeds on

More and larger

About equal. The

Mustard, 0-5 g. in
each glass
Condition on
3/11/1911

None germinated

None germinated
A few just showing

The greater part
sprouted

Equal. Almost all
sprouted. The
best grown

Radish, 18 seeds in
each glass. Sown
on 5/12/1911.
22° C.

Condition on
7/12/1911

A few very small
shoots

More and stronger

More still and
stronger

About equal.

Distinctly the
best grown

Radish, 18 seeds in each
glass.

Sown on 5/12/1911.
22°C:

Condition of seeds on

Medium germinated 12/12/1911 12/12/1911
Germinated Not Germinated Not
germinated germinated

1 % caffeine 103 =all sown 5 13 8 10
0-1 caffeine 35 i 11 10 8
0-01 % caffeine 26 6 12 10 8
0-001 % caffeine 15 12 6 13 5
0-0001 % caffeine 17 14 + 13 5
Tap water 16 13 5 14 4

THE ACTION OF CAFFEINE UPON SEEDS 161

TABLE ITI

Mustard seed, 1-0 g. in each glass. Sown on 17/12/1911. 22°C.

Medium Weight on 19/12/1911
0-001 °% caffeine ... “a5 ee 4-520 g.
0-O001 °%, caffeine ... ee are 4-750
0-00001 ©; caffeine is tense 4-610
000001 °,, caffeine aa das 4-770
Tap water ... ae oo ar 4°830
Tap water ... rey ac aie 4-710

TABLE IV

Turnip seed, 0-5 g. in each glass. Sown on 15/11/1911. Room Cress, 0-5 g. in each glass.
temperature Sown on 15/11/1911. Room
temperature
Medium Weight on 22/11/1911 Weight on 22/11/1911

0-1 % caffeine ms a 2-400 g. ae 2-530 g.

0-02 °% caffeine see ee 3-070 56 3-830

0-004 °% caffeine =a ae 3-530 ane 4-320

0-0008 °6 caffeine Ne . 3-760 ae 5-370

0-00016 °% caffeine Due 265 3-740 ao 6-320

Tap water <2. Ape Sk 3-730 rc 6-430

CELL STIMULATION BY MEANS OF PROLONGED
INGESTION OF ALKALINE SALTS!

(From the Bio-Chemical Department, University of Liverpool)

(Received November 14th, 1911)

The research here recorded was undertaken originally in the hope
of ascertaiming the relationship of cell stimulation to cancer. The
original scheme of work included the driving in of alkaline ions daily
into the same spot by means of weak electric currents, using both normal
rabbits and those which were receiving daily an alkaline salt in their
food; the ordinary and differential counts of leucocytes in rabbits under
alkaline treatment; the effects of keeping mice for several weeks in an
atmosphere containing an excess of carbon dioxide; and the daily
ingestion of increasing doses of alkaline salts by mice, rabbits and dogs.
As negative results were obtained in all except the last class of
experiments, these are the only ones which will be described in detail.

INGESTION EXPERIMENTS ON MICE

Two sets of mice were taken; one received sodium bicarbonate, the
other alkaline sodium phosphate in the food. The salts, containing their
water of crystallisation, were thoroughly mixed up with the food which
consisted of bread or dog biscuit soaked in water, with an occasional
change to dry oats. It was thought important to vary the food in this
way, so as to avoid the effects of prolonged maintenance on bread alone
or meat alone, such as have been described by Chalmers Watson. The
amount of food was so regulated that there was none left over from the
previous day. In each case the dose to begin with was 0°'1 gram per
head; this was gradually increased till some mice were receiving 0°7 gram
per head. Some of the mice died within a day or two of the commencement
of the experiment, probably from extraneous causes, and so have not been
included in the results detailed below. As regards the more resistant
ones it was noticed that after a time they became very thin and their
coats dull, but up to the time of their death they seemed to feed well.

1, This research has been carried out by Dr. F. P. Wilson, working under the direction of
a Committee consisting of E. C. C. Baly, F. W. Goodbody, A. C. Jessup, A. E. Jessup and
B. Moore. The expenses of the research have been defrayed by Mr. A. E. Jessup. In the
histological part of the research the Committee has been much assisted by Mr. George Arnold.

CELL STIMULATION 163

The period of treatment in the different cases varied from two weeks to
five mouths. Some of the mice died naturally, others were killed for
purposes of examination. All cases which had been on alkalies for any
length of time showed, post mortem, marked wasting with an entire
absence of body fat, in this respect coinciding with the experience of
Moore, Roaf and Knowles in the case of guinea-pigs. Beyond this
extreme wasting no reason for death could be found.

Pieces of the various organs were fixed in Zenker’s fluid as soon after
death as possible, cut in paraffin, stained by Breinl’s basic fuchsin
methylene blue and orange method, and examined microscopically.

No obvious changes were evident in the liver, kidneys, lungs, spleen
and intestines, but in the testes some extraordinary alterations were
found. These results are of especial interest as the cells of the testis,
except the basal cells, are regarded by many cytologists, as out of
‘co-ordination ’ with the somatic cells. As a result of these experiments
it would seem that they are more susceptible to changes in the
reactivity of the surrounding plasma. On the other hand, no changes
were observed in the ovaries of the female mice subjected to similar
conditions.

Below are set out in detail the kind of salt administered, the length
of time given for, and the histological changes found in the seminiferous
tubules. The photographs appended show a normal tubule and part of
a degenerated tubule in Mouse 1. This was one of the cases which
showed the most profound changes.

CHANGES IN THE SEMINIFEROUS TUBULES
Mice fed with Alkalies

Mouse 1. Na,HPO,. March 9, 1909 to May 6, 1909.

Fig. 1 is a portion of a tubule from a healthy untreated mouse.

The lumen is occupied by spermatozoa and spermatids, with one or
two spermatocytes of the second layer (homotype). The layer of cells
between these and the outermost layer are all spermatocytes of the first
order (heterotypes). The outermost layer contains early heterotype
prophases with leptotene nuclei, a few spermatogonia (spy.) and foot cells
(fe.)

A section of a tubule from a mouse which has been treated for

eight weeks, with the phosphate salts, presents a startlingly different
picture. Indeed, the alterations produced in the tissue are so profound,

164 BIO-CHEMICAL JOURNAL

that at first sight the section would not be recognised as a portion of
the testis. (Fig. II.)

The cells of the tubule, Fig. II, have been reduced to a single layer,
adjacent to the external investing membrane.

=

ERY,

.

(ace }
>
a
an
an
@
o

The interior of the tubule is filled with a stringy mass of

cytoplasm. Some of it appears to be connected with the cytoplasm of the
®

cells surrounding the periphery of the tubule, and it is possible that

CELL STIMULATION 165

these portions represent protoplasmic extensions of those cells, since
there are some observers who claim that the foot cells, or cells of Sertoli,
nossess a cytoplasm of an uneven outline with long pseudopodial
extensions connecting one cell with another. Such a structure cannot be
demonstrated definitely in the normal testis, but possibly the hypertrophy
of the cell under these abnormal conditions may make these protoplasmic
strands show up more clearly. However, some of the cytoplasm which
occupies the lumen of the tubules encloses traces of broken-down
nuclear material (nur.).

The nuclei of the cells are slightly larger than the nuclei of the
normal spermatogenic cells, except cells in the later heterotype
prophase, and they appear to be even larger than these. This, however,
is not really the case, but is brought about by the fact that the
cytoplasm of these altered cells is considerably less in proportion to the
nucleus, than in healthy cells. The cytoplasm is also more coarsely
reticulated, or perhaps, the reticular mesh-work takes the stain more
readily. The nuclear membranes take the basic stain very deeply. In
each nucleus there is at least one nucleolus, sometimes two or three.
Generally, in the neighbourhood of the nucleoli are other large round
bodies, which represent nearly all the chromatin of nucleus massed
together. (Fig. IL and Fig. IIT.)

These cells, by the distinctive appearance, can be recognised as
Sertoli or foot cells, and difter from the foot cells in the normal testis
only by their size, in the manner already mentioned.

Except by the irregularity of the margins, and the fact that they
generally take the blue stain very strongly (whereas the nucleolus is
distinctly red), these bodies are very similar to the nucleoli. Rarely some
division figures are seen (b), and also the intermediate stages leading up
to those divisions (a and d). In the latter a very coarse and irregular
spireme is formed, obviously at the expense of the nucleolus-like masses
of chromatin just mentioned, which get smaller and smaller as the
spireme is completed. The chromosomes of these divisions also have an
abnormal stamp. In Fig. LV these chromosomes (abn.) can be compared
with typical heterotype (#7) and somatic (S) chromosomes. It will be
seen that they are unlike either.

Judging by the position of these cells, and the general character
of the spireme stages, it is certain that they are spermatogonia, and their
irregular mitoses indicate degeneration, and compared with the foot
cells they are few in number, much below the normal found in

166 BIO-CHEMICAL JOURNAL

transverse section of a healthy tubule; and probably even these would
disappear at an early date, leaving the tubules lined with foot cells only,
as is indeed the case in a large number of the tubules in this slide.

There can be no doubt that some of the stringy masses of
cytoplasm in the interior of the tubules. represent the remains of the
cells which in a healthy tubule would normally occupy that position.
These cells, then, have completely broken down, and only here and there
in the cytoplasmic threads can traces of their nuclei be seen.
(nur., Fig. IT and Fie. V.)

The nucleolar activity in these cells is very noticeable, a very large
number showing the nucleoli dividing. Some of these daughter nucleoli
are extruded through the nucleus and cytoplasm into the lumen of the
tubule. (Fig:-II, n.n.)

The striking features in the altered testis are these :—(a) Only those
cells which le on the investing membrane of the tubule are living and
active. (b) The volume of their nuclei is out of all proportion to that
of the cytoplasm. (c) The chromatin is massed together in each nucleus
into a few round lumps.

Mouse 2. NaHCO,. April 27, 1909, to June 23, 1909.

Normal.

Mouse 3. NaHCO,. April 27, 1909, to July 27, 1909.

Almost as bad as No. 1, except that spermatozoa are still present in
some numbers. Like No. 1, marginal cells with nuclei considerably
altered, and abnormal in appearance, line all the tubules. Spermatocytes
completely absent. The central part of the tubules filled with
cytoplasmic remains and spermatozoa, and some concentrations of
chromatin, derived from broken-down nuclei.

Mouse 4. Na,HPO,. March 15, 1909, to August 3, 1909.
Normal.

Mouse 5. NaHCO,. September 1, 1909, to September 10, 1909.
Double nucleoli-marginal cells very marked. Spermatid degeneration
extensive. Nuclei of some spermatocytes considerably altered, having
a mossy appearance, due to the chromatin and linin breaking up into
irregular lumps and specks, especially noticeable in spireme stages.

Mouse 6. NaHCO,. September 1, 1909, to September 15, 1909.
Like No. 5.

CELL STIMULATION 167

Mouse 7. Na,H PO,. September 1, 1909, to September 15, 1909.
Multipolar mitoses rather plentiful. Chromosomes unrecognisable,

merely shapeless lumps of chromatin.

Mouse 8. NafiPo,. September 1, 1909, to September 26, 1909.
Normal.

Mouse 9. NaHCO... September 15, 1909, to September 28, 1909.
Normal.

Mouse 10. NaHCO... September [pe LOG 60 September 28, 1909.

Like No. 5. The degeneration of the spermatocytes takes the
following course. The spireme threads lose their staining reaction and
the nucleolus enlarges and attaches to it a large mass of chromatin
derived from the spireme, in fact the latter stains more faintly owing

to the absence of chromatin.

Mouse 11. Na,#PO,,. September 1, 1909, to September 28, 1909.
Normal.

Mouse 12. NaHCO... September 15, 1909, to September 29, 1909.
Normal.

Mouse 13. Normal control.

Normal.

Mouse 14. Normal control.

Normal.

Mouse 15. Normal control.
Normal.

Mouse 16. Normal control.
Normal.
Mouse 17. Normal control.

Normal.

Mouse 18. NaHCO,,. September 15, 1909, to October 28, 1909.
Mitoses normal.

Mouse 19. NaHCO,,. September 15, 1909, to October 28, 1909.
Normal.

Mouse 20. NaHCO,. September 15, 1909, to October 28, 1909.
Normal.

168 BIO-CHEMICAL JOURNAL

Mouse 21. NaHCO,. September 1, 1909, to October 28, 1909.
Normal.

Mouse 22. NaHCO,. September 1, 1909, to October 28, 1909.

Normal.

Mouse 23. NaHCO,. September 1, 1909, to October 30, 1909.
A large mass of degenerated chromatin, spermatocytes present.

Mouse 24. Normal control.
Normal.

Mouse 25. Normal control.
Normal.

Mouse 26. Normal control.
Normal.

Mouse 27. Normal control.
Normal.

Mouse 28. Normal control.
Ovary, normal.

Mouse 29. NaHCO... September 15, 1909, to November 30, 1909.
Normal.

Mouse 30. Na,HPO,. August 9, 1909, to January 14, 1910.
Ovary, normal.

Mouse 31. NaHCO,. September 15, 1909, to January 14, 1910.
Ovary, normal.

Mouse 32. Na,HPO,. September 9, 1909, to February 22, 1910.
Ovary, normal

Mouse 33. Normal control.

Normal.

CONSIDERATION OF THE ABOVE RESULTS

Out of twenty-two treated mice there were ten normal testes and
three normal ovaries, the remaining nine (all testes) showed deviations,
more or less marked, from the normal. The eleven controls were all
normal.

The fact that 41 per cent. of the treated mice presented unusual

CELL STIMULATION 169

degenerative changes in the testes is good evidence that these results
were due to the ingestion of the alkaline salts.
In addition to the above tame mice, the testes of six wild mice

were examined and found to be quite normal.

InGestion OF Povasstum BICARBONATE

A third series of ten mice were fed on increasing doses of
potassium bicarbonate. The general effects of the potassium ion
resembled those of the sodium, but seemed to be produced more quickly.
Changes in the liver and the testes were found similar to the sodium-fed
mice.

K. I. In a very small number of the tubules of this testis only
marginal cells, ¢.e., Sertoli or foot cells remained intact. The rest of
the cells are in a state of degeneration, as evidenced by the cytoplasm
taking the basic stain, and the nuclei fragmenting. In other tubules
the percentage of foot cells is above the normal. This slide probably
indicates the earliest steps on the road to degeneration leading to the
condition found in sodium bicarbonate mice, 1, 3 and 14.

Ky 2. Resembles K 1.
K 3 and K 4. Normal.
AK 5. Normal.

KX 6. Resembles K 1.
K 7. Resembles K 1.
AK 8. Normal.

K 9. Ovary, normal.

K 10. Ovary, normal

INGESTION EXPERIMENTS WitH RABBITS

Altogether five rabbits (Nos. 3, 15, 18, 20 and 21) have been fed
with increasing doses of sodium bicarbonate for varying periods. The
daily dose to begin with was in each case 0°5 gram per head. This was
increased by 0°5 gram each week until the animal was taking 7 grams
of salt in its food daily. The quantity of food given was adjusted so
that as far as possible all was eaten,

170 BIO-CHEMICAL JOURNAL

The animals seemed to stand 5 grams per head daily without much
change, but beyond this dose they rapidly began to lose flesh and to
become poor in coat.

At post mortem examination the most marked feature was the
wasting and the entire absence of body fat, even round the kidneys. The
lungs were congested, and in the case of rabbit No. 21, which had
developed a purulent cyst of the neck, the lungs were tuberculous.
Otherwise there were no obvious naked-eye changes met with, and it
seemed probable that the animals would have died of extreme inenition.
As regards this, it is worthy to note that the animals continued to feed
well to the end.

Rabbits 3, 13 and 20 were females. Treated with NaHCO, for ten
weeks. The ovaries presented no pathological changes. In Rabbits 5
and 20 there were changes in the liver cells, which stained very badly.
The cytoplasm was in a slightly necrotic condition in parts, and the
nuclei lying free in clear spaces. The cells of the renal tubules were in
a desquamative condition. Rabbit 15 presented no evident histological
changes. Rabbits 18 and 21 were males. Treated with NaHCO, for
eight weeks. The testes appeared normal. The liver cells had a hyaline
appearance and took the stains badly. The renal epithelium presented a
similar appearance. The other organs seemed normal. In the rabbit,
therefore, it would appear as though the sexual glands were more
resistant to the action of the alkali than was the case in mice. On the
other hand, the salt seems to irritate the liver and renal cells in the
rabbit more than in the mice. It is possible that this susceptibility or
otherwise is in some way dependent on the natural habits of the animal
in respect to the kind of food usually eaten.

171
THE PROTEINS OF RICE

By S. KAJIURA, /mpertal Japanese Navy.

From the Physiological Laboratory, King’s College, London
(Received November 24th, 1911)

Rice forms the staple food of many Oriental races, and a knowledge
of the composition of its proteins is therefore important from = an
economic and hygienic point of view, especially in relation to * Beri-beri,’
the disease of rice-eating nations. It appears, however, from a review
of the literature that the nature and composition of the rice proteins
has, up till now, not been studied with anything like the thoroughness
which has been bestowed on the proteis of European cereals.

In a preliminary communication, Rosenheim and Kajiura! showed
that the distribution and nature of the rice proteins differ characteristically
from what is found in other cereals. They found that the principal
protein, to which they gave the name Oryzenin, belongs to the class of
glutelins (proteins insoluble in water and neutral salt solutions, but
soluble in dilute alkali); the amount of albumin and globulin is very
small, and an important result noted was the almost complete absence
of a protein soluble in alcohol. This latter observation sharply
differentiates the proteins of rice from those of other cereals.2 The
present communication gives a fuller account of the experiments which
led to these results.

The material used was procured directly from Japan, and consisted
of the ordinary rice used there for dietary purposes, namely, peeled and
polished ‘white rice.’ Portions of this were freshly ground, when
required, to a fine flour in an ‘ Excelsior’ mill. The ground rice
contained 12°92 per cent. of water. The dried material contained
1:24 per cent. of nitrogen. If all the nitrogen were present in the form
of protein, the percentage of protein would be according to the usual
calculation 7°75 (N. x 6°25).

The methods generally used for the preparation of vegetable proteins
are based on their different behaviour towards solvents, and they are
classified according to their solubilities into four main groups.®

1. Proc. Physiological Society, January, 1908. Journal of Physiology, Vol. XXXVI.
Suzuki, Yoshimura and Fuji shortly afterwards obtained practically the same results (Journ.
Tokyo Chem. Soc., XXIX, No. 3, March, 1908).

2. See S. Kajiura and O. Rosenheim. ‘A Contribution to the Etiology of Beri-Beri.’
Journal of Hygiene, Vol. X, p. 49, 1910.

3. O. Rosenheim, ‘Science Progress, No. 8, 1908. ‘TI. B. Osborne, ‘The Vegetable
Proteins, published in Longmans’ Monographs on Bio-Chemistry, 1909.

172 BIO-CHEMICAL JOURNAL

I. Phyto-albumins; soluble in water.
Il. Phyto-globulins; insoluble in water, soluble in saline solution.
III. Prolamins (Gliadins) ; insoluble in water, soluble in alcohol.
IV. Glutelins (phyto-caseins) ; insoluble in water, saline solutions and
aleohol, soluble in dilute alkah.

The first step in the investigation, in the absence of any data from
previous literature, consisted in examining systematically the effects
of these various solvents on rice, and thus to classify the proteins which
are present.

I. Tur ALBUMINS AND GLOBULINS OF RICE

Preliminary experiments showed that distilled water extracts from
rice contained exceedingly small quantities of protein. The aqueous
extract must be assumed to contain the albumins, together with proteoses,
which may also possibly be present (see later). But it is well known that
under these conditions, a considerable quantity of globulin also goes into
solution owing to the presence of salts in the seeds. The total quantity
of protein so obtained was, however, small, and it was therefore
considered advisable to dissolve out the albumins together with the
globulins by means of salt solution, and to separate the two classes of
protein subsequently. The methods employed for separating them were :
(1) dialysis; (2) saturation by ammonium sulphate followed by dialysis;
and (3) fractional heat-coagulation.

Before dealing with the extraction of large quantities of rice with
neutral salt solution, a number of preliminary quantitative experiments
were carried out on a small scale. For this purpose 10 to 12 grams of
finely ground rice were extracted with 40 c.c. of a 10 per cent. solution
of sodium chloride with frequent shaking. The mixture was then allowed
to stand, and the supernatant fluid was decanted off. The residue was
extracted twice more with 30 c.c. of the salt solution; chloroform
was used throughout as an antiseptic. The combined extracts were
filtered under pressure through asbestos or paper pulp, and made up
to 100 c.c. Nitrogen was then estimated in two portions of 50 ¢.c. each.
The results of four such experiments were as follow :—Nitrogen present
in the extract of 100 grams of rice made with 10 per cent. salt solution.

1) O14 ¢ram.

(a MOM Yn
(im) (Oi4
(iv) 010

+)

+)

THE PROTEINS OF RICE 173

T'wo nitrogen estimations were made in each case, except in (iv), and
the above figures represent the amount calculated from the average.
The mean result of the four experiments is that 100 grams of dry rice
contain 0°12 gram of nitrogen, corresponding to 0°75 gram of protein
soluble in 10 per vent. sodium chloride solution. This assumes that all
the nitrogen was combined in protein form, and on this assumption only
about 10 per cent. of the total proteins of rice consist of albumins and
globulins. When working on a larger scale, however, with a view to
preparing these proteins in bulk, it will be seen that even this small
theoretical yield was not reached.

(a) Preparation of Rice Globulin by Dialysis. Five kilograms ot
finely ground rice were extracted at room temperature with 8 litres of
10 per cent. sodium chloride solution; the extract was filtered off and
the residue was pressed in a filter press and again extracted with 4 litres
of the salt solution. The extracts were filtered through paper and paper
pulp. Nine litres of a perfectly clear extract were obtained, which were
subjected to dialysis in parchment tubes in a special apparatus. Toluene
was used as an antiseptic. Dialysis was continued for 190 hours, after
which time the contents of the dialyser were practically free from
chloride. The small precipitate which had formed in the dialysing tubes
was separated by decantation and finally by the centrifuge. After
washing and drying im vacuo it weighed 7°34 grams.

The product was a yellowish, amorphous powder, which gave the
usual protein colour reactions, and showed the characteristic solubilities
of a globulin. It contained phosphorus and is, therefore, possibly a
globulin nucleate. On dissolving it in 10 per cent. sodium chloride
solution, a small amount remained insoluble. The filtered solution, when
dialysed, gave again an amorphous precipitate. The quantity was,
however, too small for further purification or study. All attempts to
obtain it in crystalline form by dialysis or by cooling its warm solution
failed.

The main bulk of the extract thus freed from globulin by dialysis
still contained a very small amount of protein coagulable by heat. Owing
to the large volume of the solution and the extremely small amount of
protein contained in it, no attempt was made to isolate it. Its isolation
was, therefore, carried out in the experiment to be described next, in
which the bulk of fluid to be dealt with was much reduced.

(6) Preparation of Rice Albumins and Globulins from the saline
extract by saturation with ammonium sulphate. Five kilograms of finely

174 BIO-CHEMICAL JOURNAL

evound rice were extracted with 10 per cent. salt solution as deseribed
under (a). Hight litres of clear filtrate were obtained; this was
saturated with ammonium sulphate; the flocculent precipitate so
produced was filtered; the filtrate was protein-free; the precipitate was
washed with saturated ammonium sulphate solution, and redissolved in
15 litres of 10 per cent. sodium chloride solution. This was filtered
and dialysed against running water, as before, for 120 hours, putrefaction
being prevented by the addition of toluene. At the end of this time the
contents of the dialyser were free from sulphates and chlorides. The
precipitated globulin was separated and dried as before. When
dissolved in 10 per cent. salt solution, a sample became shghtly opaque
at 55° C., but a floceulent precipitate did not appear until the
temperature was raised to 87° C.

After the removal of the globulin the solution contaimed a
considerable amount of protein which coagulated at 65°C. The whole
bulk of the solution was, therefore, left at this temperature for some
time, and the coagulated protein was filtered off, dried and weighed.
It is difficult to decide whether this protein can be considered as a true
albumin, or whether it represents a globulin held in solution by minute
quantities of acids'. On heating the remaining solution to boiling
point, a further very small amount of protein was coagulated, and the
solution thus freed from all coagulable protein gave a faint biuret
reaction; it was concentrated to a small bulk on the water-bath and
finally poured into absolute alcohol. The white precipitate so produced
was filtered off and dried as before. The last precipitate presumably
consisted of proteoses, but whether these are preformed in the rice grain,
or are produced hydrolytically by the process of boiling, it is not possible
to say.

The weights of the different fractions obtained in this experiment
will be found in the following table :—

(1) Globulin, precipitated by dialysis ss se .. 140 gram ) From

(ii) Albumin (?) coagulated at 65° = a a gaan lari zs 5,000
(iii) Residual protein coagulated at 100 =e Pex. seer (Pato 3. ' grams
(iv) Proteoses (?) precipitated by alcohol a es soo HEP) a | of rice.

4:99 = 0-1 per cent.
The figures in the foregoing table must not be taken as absolute,
for there was probably some loss owing to the operations being performed

with large volumes of fluid and owing to incomplete extraction. They

1. A similar difficulty was experienced by Osborne (Journ. Amer. Chem. Soc., XIX, p. 525,
1897) in his examination of the proteins of maize.

THE PROTEINS OF RICE 17d

are useful as an indication of the relative amounts of the different
fractions. At any rate the total quantity of protein in saline extracts 1s
very small, and on the assumption that rice contains 775 per cent. of
total protein, and that the foregoing estimations are approximately
correct, the albumin and globulin account for but little more than 15 per
cent. of the total.

In this particular, rice resembles maize! and differs from other
cereals.

(ce) Fractional heat coagulation. ‘he following results” were
obtained by heating a clear extract of rice made with 10 per cent.
solution of sodium chloride.
49°-50° C. Opalescence.
52°-53° C. Slight flocculent coagulum ; the fluid was left at this temperature for two minutes.

57°-58° C. Slight increase of coagulum ; the fluid was left at this temperature for five minutes,
and then filtered, and the filtrate was heated further.

64°-65° C. Opalescence ; the fluid was left at this temperature for ten minutes, but no coagulum
formed. ;

70°-71°C. Bulky coagulum; the fluid was kept at this temperature for five minutes, and
then filtered. The clear filtrate was heated further.

75°C. Slight opalescence.

$95°-86° C. Distinct coagulum ; the fluid was kept at 87° C. for five minutes, and then filtered ;
the filtrate was further heated.

90°-99° C. Slight opalescence.

It will be seen that the two principal coagula were those obtained
at 70° to 71° C., and at 85° to 86° C.; of these the former was the more
bulky. The great limitations of this method for the separation of vegetable
proteins are well known.2 As the method, however, seemed to encourage
the hope that it would provide a larger supply of these proteins for
further study, some experiments were carried out on a larger scale,

Five kilograms of rice were extracted with 10 per cent. salt
solution as before; three extractions were made and 15 litres of a clear
solution were obtained. ‘This was worked up in two portions. Hach
portion was gradually brought to 70°C. in a large water-bath, and kept
at that temperature for half an hour. After cooling, the clear
supernatant fluid was decanted, and the coagulated protein was collected
on a filter, washed with water, aleohol and ether, dried in vacuo and
weighed.

The clear filtrate, after the separation of the 70°C. coagulum, was

1. Chittenden and Osborne, Amer. Chem. Journ., XU, p. 453, 1891; XLV, p. 20, 1892 ;
Osborne, Journ. Amer. Chem. Soc., XIX, p. 525, 1897.

2. See Osborne, ‘The Vegetable Proteins, Longmans’ Monographs on Bio-Chemistry,
pp. 15, 44-45, 1909.

176 BIO-CHEMICAL JOURNAL

heated to 95°C. in a boiling water-bath, and the coagulum collected,
washed, dried and weighed as before.

The experiment was repeated in another sample of 5 kilograms of
rice in exactly the same way. The yields obtained are given in the
following table. Unfortunately in the second experiment an accidental
loss occurred during the filtration of the second protein, so the figure
‘there has to be omitted.

Amount of proteins extracted from 5,000 grams of rice, with 10 per

cent. sodium chloride solution :—

Coagulated at 70°C. Coagulated at 95° C.
Haperiment I a6 abe ... 9:58 grams 50% 1-26 grams
x Il a oer e <9:09.;, ae —

Taking the first of these two experiments, the only complete one,
as a fair sample, it will be seen that the total yield of protein coagulable
by heat is 6°84 grams: 100 grams of rice will, therefore, yield 0°13 grams,
a figure approximately equal to that obtained in Experiment (b)
(ammonium sulphate method).

The products obtained by heat-coagulation were obviously of no use
for further purification, and an elementary analysis would have been of
no value and was therefore not carried out. It also seemed doubtful
whether a study of their cleavage products would serve any useful
purpose, and so no attempt was made to obtain them in the large
quantities which would have been necessary; the amount of raw material
which would have had to be dealt with exceeds the usual laboratory
conveniences.

The general conclusion to be drawn from this part of the work is that
the amount of albumin and globulin forms only a small fraction of the
total proteins of ‘ white rice,’ and in this respect rice differs from other
cereals (maize alone excepted), the proteins of which have been
investigated.

Il. Tim Atconor-SoutusBLte Prorerns or Ricr

Preliminary quantitative experiments were first carried out on a
small scale in order to estimate the amount of nitrogenous material
soluble in 75 per cent. alcohol at various temperatures before and after
the extraction of the albumins and globulins. Two of these may be

given in detail.

THE PROTEINS OF RICE 177

(i) 11-504 grams of finely ground rice were extracted, with frequent shaking, at room
temperature with 40 c.c. of 75 per cent. alcohol. After eighteen hours the fluid was decanted
off, and the residue again eXtracted with 5 c.c. of the aleohol. This was repeated three times,
and the combined extracts were filtered through asbestos under pressure. The clear filtrate
was made up to 100 c.c., and divided into two portions of 50 c.c, each, in which nitrogen estima-
tions were made. The residue was then treated in a similar way with alcohol at 50°-60° C.,
and nitrogen estimations were made also in the warm alcohol extract. ‘The results caleulated

from these estimations for 100 grams of dry rice were :

Cram
Soluble in cold 75 per cent. alcohol ad eS Dae ... 004 Nitrogen
warm x ¥r Ree nae ges ... 0-04 =
Total amount soluble in 75 per cent. aleohol ... cae 008

(ii) 10-925 grams of ground rice were extracted with 10 per cent. salt solution to remove
albumins and globulins, and subsequently with 75 per cent. alcohol at 60°C. The result
caleulated for 100 grams of rice came out very near that of the first experiment, namely, 0-085

per cent

Assuming that the whole of the nitrogenous material soluble in
aleohol consists of protein (which is probably not the case), its amount
is only about 0°5 per cent. of the whole rice, or approximately 6°5 per
cent. of the total protein present.

As it seemed advisable to control this result by working on a larger
seale, 3 kilograms of the finely ground * white rice” were allowed to
soak in 4 litres of 75 per cent. alcohol. Afterwards the temperature was
raised to 50°C. and the mixture kept at this temperature for some time.
After filtration the residue was again extracted with 2,500 c.c. of the
alcohol, and the clear extracts were concentrated im vacuo. As
concentration progressed and the alcohol evaporated, a precipitate formed
in the residual aqueous solution; more water was added and_ the
precipitate was separated by the centrifuge; this was again suspended
in water and again centrifugalised. On warming it in 75 per cent.
alcohol to 50°C. most of the precipitate went into solution. The
solution was filtered warm, and poured into a large volume of absolute
alcohol. A white precipitate formed, which was collected on a filter,
well washed with absolute alcohol, and dried in vacuo. A granular,
whitish powder was thus obtained which weighed 0°45 gram (0°014 per
cent. of the whole rice).

This experiment was repeated several times with different samples
of ‘white rice’ with practically the same results; and the average
percentage worked out at 0015.

The product appeared to be of protein nature, as it gave the usual
protein colour reactions, but it is doubtful if it can be considered a true

178 BIO-CHEMICAL JOURNAL

alcohol-soluble protein. If we compare its amount (0°018 per cent.) with
the amounts easily obtainable from other cereals (for instance, zein from
maize is present to the extent of 5 per cent. of the whole grain), the
conclusion one arrives at is that ‘white rice,’ as used for dietary
purposes in Japan, contains only a negligible trace of alcohol-soluble
protein, a feature which distinguishes rice sharply from all other cereals
-used as food. This result has since been confirmed by Suzuki,
Yoshimura and Fuji and by A. Fraser and A. T. Stanton’.

ILL. Oryzenrtn, THE ALKALI-SOLUBLE PROTEIN OF RICE

It is clear from the preceding that the proteins of rice are only to
the smallest extent soluble in salt solution and in 75 per cent. of alcohol.
A preliminary experiment on a small scale showed that rice treated
with 0°2 per cent. sodium hydroxide solution gave an extract in which
a stream of carbon dioxide, or dilute acids produced an abundant white
precipitate of protein nature, and from quantitative experiments carried
out as before, it appears that approximately 70 per cent. of the total
proteins of rice consist of this alkali-soluble protem, to which
Rosenheim and Kajiura® gave the name Oryzenin.

In order to prepare oryzenin on a larger scale, the ground rice
was first thoroughly extracted with 10 per cent. sodium chloride
solution, and the residue thus freed from albumins and globulins was
then subjected to the alkali treatment.

It was found that the crude preparation always contained a
considerable amount of phosphorus, evidently owing to admixture with
nucleic acid. When dilute mineral acids were used for the precipitation
of the protein from its alkaline solution, the percentage of phosphorus
Was in every case nearly double of that found when the protein was
precipitated by carbon dioxide. Although the amount of phosphorus
decreased after repeated precipitation, the last traces were difficult to
remove when mineral acid had once been used in its original
preparation. The following table shows strikingly the variation in the
phosphorus percentage of the crude products. Two consecutive extracts
of rice with 02 per cent. sodium hyroxide were prepared, and each

6. Loc. cit.

7. Studies from the Institute of Medical Research, Federated Malay States, No. 12,
1911. These authors also found that the ‘ polishings’ removed from ‘ white rice’ contain a
rather larger amount of alcohol-soluble protein, which they, however, were unable to isolate.

8. Loc. cit.

THE PROTEINS OF RICE 179

extract divided into two parts; one part was precipitated with dilute

sulphuric acid, and the other by a current of carbon dioxide.

P
Crude oryzenin from first extract Precipitated with sulphuric acid... 1-71 per cent.
= ., carbon dioxide ... 0-88
From second extract a ree es » sulphuric acid ... 1:58
e .. carbon dioxide ... 0-78

In the following preparations, therefore, the alkaline extracts of
rice were precipitated by carbon dioxide, and purified by redissolving in
02 per cent. alkali and reprecipitation by the same gas.

Preparation of Oryzenin. Five kilograms of finely ground rice were
extracted with 10 litres of 10 per cent. salt solution, and after filtration
the residue was again extracted with 5 litres of the salt solution. This
process was repeated a third time. The residue was then freed as far
as possible from the salt solution by pressure in a strong hand-press
and then treated with 8 litres of O° per cent. solution of sodium
hydroxide, the extract was decanted off, and the extraction repeated four
times more with fresh alkali solution, 6 litres being employed each time’.
The extracts were obtained quite clear by filtration through several
layers of thick filter paper, and then precipitated with carbon dioxide ;
an abundant white amorphous precipitate formed. The precipitates
obtained from the first two and the last three extracts were combined and
purified separately. They were washed repeatedly by decantation with
distilled water, dissolved in 0°2 per cent. alkali and reprecipitated by
varbon dioxide. ‘They were washed again by decantation with
distilled water, followed by 50 per cent. alcohol, and finally filtered,
washed with absolute alcohol and ether, and dried in vacuo. The total
average yield of crude oryzenin so obtained from 5 kilograms of rice
was 62 grams.

For still further purification the product was redissolved in 0°2 per
cent. alkali and reprecipitated, the process being repeated many times.
As oryzenin, like the glutenin of wheat, is insoluble in neutral salt
solutions, its purification by fractional precipitation is impossible, and
the only available criterion of its purity consists in the absence of
phosphorus after many reprecipitations.

Properties and composition of Oryzenin. he product finally
obtained was a perfectly white powder, insoluble in water, neutral salt

9. In some cases the extraction was continued still further, until no precipitate was
obtained on neutralisation. This is usually the case after the seventh extraction.

180 BIO-CHEMICAL JOURNAL

solutions alcohol and ether. It dissolves easily in dilute alkalies and
acids, and is precipitated on neutralisation. On boiling its suspension
in water, oryzenin is coagulated and is thus rendered insoluble in dilute
alkali. It gives all the usual protein colour reactions.

Although elementary analysis is of secondary value in the case of
high molecular substances, the results obtained by the analysis of two
_of the purest preparations may not be without interest. The specimens
had been redissolved and reprecipitated five times and were practically
free from phosphorus. ‘The combustions were carried out by Dennstedt’s
method; nitrogen was estimated by Kjeldahl’s and sulphur by
Pringsheim’s method.

As oryzenin represents the only glutelin, except glutenin from
wheat, which can be easily obtained from grain, its elementary
composition 1s ‘compared in the following table with that of glutenin.
The figures in the table are percentages.

C H N > O
Oryzenin preparation (1) a OLedO 7-05 16-27 0-90 24-48
+ - (2) soe OL-34 7-24 16-13 0-93 24-36
Glutenin from wheat10 ... Bee DD od: 6-83 17-89 1-08 22-26

The figures for oryzenin show a remarkable agreement, considering
the nature of the substance, and speak in favour of the uniformity of
the product.

The Partition of Nitrogen in Oryzenin. WHausmann’s method is now
generally accepted as a convenient means for comparing the different
forms in which the nitrogen of proteins is present after complete
hydrolysis with hydrochloric acid. In the following table the results
are given which were obtained from oryzenin by this method, and those
furnished by wheat glutenin!! are added for the sake of comparison.

In per cent. of Protein In per cent. of Nitrogen
Total Nas Nin MgO Basic Non-Basic WN as Basic Non-Basic
= Ammonia precipitate N N Ammonia N N
Oryzenin (1) ... =e 16:20 1-43 0-11 4-53 10-13 | . Bate 2
6 6Q)un - 1620 142 O11 402 9-75) SP =i Sue
Glutenin aa wos ©=17°49 3°30 0-19 2:05 11-95 18-8 11-7 68-3

It will be seen that the partition of nitrogen in oryzenin differs
considerably from that in glutenin, the nitrogen present as ammonia
being much lower and the basic nitrogen much higher in the former
protein. These features also distinguish oryzenin from the alcohol-
soluble proteins. In this respect oryzenin most nearly approaches what
is found in the phyto-globulins.

10. Osborne and Vorhees, Journ. Amer. Chem. Soc., XV, p. 392, 1893,
11. Osborne and Harris, Jbid., XXV, p. 323, 1903.

THE PROTEINS OF RICE 181

LV. Genxerat CONCLUSIONS

1. The proteins of ‘ white rice,’ as used for dietary purposes in
Japan consist only to the smallest extent of albumins and globulins.

2. The amount of alcohol-soluble protein is practically negligible
thus distinguishing rice sharply from all other cereals hitherto
investigated.

4. The main protein of rice, oryzenin, belongs to the glutelin
class (proteins soluble in dilute alkali). In its nitrogen partition it
differs from wheat glutenin (the only other glutelin so far studied) very

considerably.

This research was carried out under the direction of Dr. Otto Rosenheim. Unfortunately
it was left in an unfinished state owing to the author's return to Japan. It has, however, been
thought advisable to publish the results, and the paper has been compiled by Dr. Rosenheim
from the notes which Dr. Kajiura left behind him.

The expenses of the research were defrayed from grants made by the Government Grant
Committee of the Royal Society, to whom the thanks of the Laboratory are due.—
W. D. Hatiipurton, Professor.

182

SOME FURTHER EXPERIMENTS ON THE PHYSIOLOGY OF
THE ALLYL COMPOUNDS

By E- WACE CARLIER, M.D., F.RIS.E., Ere.
From the Physiological Laboratories of the University of Birmingham

(Received December 9th, 1911)

Owing to the difficulty of determining whether the effects of
allyl-isothiocyanate upon respiration and blood pressure, previously
recorded by me,! were due to the allyl or the isothiocyanate moiety of
the compound, it appeared advisable to perform similar experiments with
at least one other isothiocyanate as well as with various allyl compounds.

With this end in view it was determined to investigate the action
of phenyl-isothiocyanate C,H,NCS upon deeply-anaesthetised rabbits.

As in the case of the allyl compound, 10 per cent. and 20 per. cent.
solutions of the substance in olive oil were used, but produced so little
effect that im subsequent experiments undiluted phenyl-isothiocyanate
was alone employed.

Phenol-isothiocyanate

Only one experiment with this substance needs to be recorded in
detail. Fig. 1.

1 minim (0°06 c.c.) of the drug was injected into the jugular vein of
a deeply-etherised rabbit, weighing 1950 grams and registering a blood
pressure of 75 mm. mercury.

Kleven seconds after the commencement of the injection the blood
pressure began to rise, and attained to 82 mm. in 20 seconds; this was
followed by a depressor effect to 67 mm. at the 60th second. Seven
seconds later it rose again to 70 mm.; this shght recovery was succeeded
by a slow fall, which reached as low as 31 mm. at the 480th second, after
which a slow rise commenced that had attained to 79 mm. at the 1070th
second.

No marked effect upon the respiratory movements was observed until
the 63rd second after injection, when a deep gasp occurred followed by
respirations rapidly diminishing in amplitude to the 95th second, after

1. Bio-Chemical Journal, Vol. 1V, p. 107,

EXPERIMENTS ON PHYSIOLOGY OF ALLYL COMPOUNDS — 188

which they quickly increased, almost but not quite reaching their initial
amplitude at the 120th second. At the 170th second a slight decrease set

in. followed at the 195rd second by a sudden fall to one-fourth the initial

Respiration.

Fic. 1. The effect of 1 minim (0:06 c.c.) Phenyl-isothiocyanate injected into the

Jugular Vein of a Rabbit.

\ll Figures to be read from right to left.

amplitude accompanied by a slowing of rhythm; from this recovery was
slow, reaching only one-third the original size at the 1076th second. <A

marked feature of this recovery was the occurrence, at fairly regulai

184 BIO-CHEMICAL JOURNAL

intervals, of a gasping inspiration, the first of which took place at the
210th second, the second at the 274th, followed by others at the 340th,
A03rd, 471st, 530th, 580th, 636th, 675th, 717th, 753rd, 801st, 836th,
873rd, 910th, 950th, 980th, 1012th, 1047th, . . . . second.

Very little change in the force and rate of the heart beat appears in
the tracing.

The temperature of the animal, as ascertained by placing the bulb
of a thermometer in the vagina from the 210th to the 274th second, was
30°C.

Some time later, when the respiratory movements had become equal
to half their initial size, another minim (0°06 c¢.c.) was injected into
the vein, with similar though less-marked results, except for respiratory
gasps which were repeated every twenty to thirty seconds. The animal
was killed twenty-two minutes after the second injection.

In other experiments 2 minims (0°132 ¢.c.) and 3 minims (0'198 c.c.)
doses were injected into the jugular vein, and produced similar results
on the pressure, but caused more marked and more lasting respiratory
spasms. The 3 minims (0°198 c.c.) dose proved fatal in 190 seconds, at
which moment the blood pressure fell to zero, the heart having ceased
beating ten seconds earlier. Nevertheless the respiratory movements
continued, increasing in amplitude and rate till the 240th second when
they suddenly failed. The heart, therefore, stopped quite fifty seconds
sooner than the respiration.

Post mortem examination showed that intra-vascular clotting had
taken place.

Effect of small doses.— 5 minim (0°006 c.c.) in olive oil when injected
into the circulation produces a temporary slight fall in the blood
pressure. 1/5 minim (0012 ¢.c.) dose causes an initial slight rise followed
by a fall, which is still more marked by a } minim (0°03 c.c.) dose.

To affect the respiration a 1 minim (0°06 ¢.c.) dose is necessary, and
a dose of 1} minims (0°09 c.c.) to alter the heart beat. Three minims
(0'198 c.c.) may be taken as a fatal dose for a full sized rabbit weighing
2200 grams. This is approximately 14 minims (0:09 c.c.) per kilogram
ot body weight.

Phenyl-isothiocyanate, though a poison, is much less fatal than
allyl-isothiocyanate, the lethal dose of which is 4 minim (0°012 c.c.) per
kilogram of body weight.

The phenyl compound acts upon the circulatory system more
powerfully than upon the respiratory; probably, therefore, the very

EXPERIMENTS ON PHYSIOLOGY OF ALLYL COMPOUNDS = 185

poisonous action of allyl-isothiocyanate upon the respiratory centre 1s
due to the allyl moiety; its action on the circulation may or may not

be due in some measure to the isothiocyanate moiety.

‘Two substances analogous to allyl sulphide were experimented with :
ethyl sulphide and ethyl hydrogen sulphide. Ethyl sulphide (C,H,),S
differs from allyl sulphide (C,H.),S only in containing two atoms of
carbon less in its molecule. Ethyl hydrogen sulphide or mercaptan is

C,H.HS.
Ethyl Sulphide (C,H.),S

Ethyl sulphide in saturated solution has but slight effect upon the
frog’s heart.

Five minims (0°33 ¢.c.) injected into the ventricle of a pithed frog
diminished the force of the heart to half the normal for a beat or two
only, the heart completely recovering from this dose in five beats. A
similar dose injected into the right auricle produced no alteration in the
heart’s action. A 10 minim (0°66 c¢.c.) dose, slowly injected into the
ventricle produced a more marked effect, arresting its contractions for
forty seconds, after which the ventricle began to beat again, feebly and
slowly at first, but rapidly increasing in force and frequency, with very
strong contractions at intervals; it regained first its normal rhythm and
some time later, at the 390th second, its initial force. The auricles
continued beating throughout the experiment, though with diminished
rapidity during the ventricular stasis.

Fifteen minims (0°99 c.c.) injected into the ventricle produced a
similar change in the heart’s action lasting 410 seconds, but its effect
was rather one of duration than of amount.

Owing to the feeble action of this drug on the frog’s heart no

b]
S
Do
experiments upon rabbits were made with it.

Nthyl mercaptan, CH HS

Two minims (0°12 ¢.c.) injected into the right auricle of a pithed
frog resulted at once in the loss of one heart beat, followed by
considerable increase in the force of the contractions for eleven beats.
This was followed by a sudden diminution in force lasting for twelve
beats, from which it returned again to normal eleven beats later. The
effect of the dose lasted in all through thirty-five heart beats, occupying
seventy-four seconds. The injection of 2 minims (0°12 c.c.) (Fig. 2a)

186 BIO-CHEMICAL JOURNAL

into the ventricle immediately produced a diminution in the rate and
force of its contractions, which gradually became smaller during
twenty-four beats. Recovery began at the 39th beat and was achieved at
the 74th. From the moment of injection to the completion of recovery
125 seconds had elapsed. At no time was there complete cessation of either
auricular or ventricular movement and the diminution in rate was trivial
and transitory, the beats succeeding each other more rapidly as the
heart’s contractions diminished.

2 eam “wwe
Neat

(E0CtF C1 LTEE ES SEE

Fic. 2a. Injection of 2 minims (0°12 c.c.) Ethyl mercaptan_into Frog’s Ventricle.

An injection into the ventricle of 3 minims (0°18 c.c.) (Fig. 2b)
resulted in a gradual diminution of the heart beat, until the auricles
alone remained in action. Complete recovery eventually occurred, but
not until the 789th second after injection.

Five minims (0°30 c.c.) introduced into the ventricle stopped that
chamber in sixteen beats, but succeeded only in slowing the auricular
rhythm. After a time the ventricle started to beat again, and quickly
not only regained but surpassed its normal force, but with much slower
rhythm.

When 10 minims (0°60 c.c.) were slowly injected into the ventricle,
spasms of the skeletal muscles occurred with arrest of the heart for
thirty-six seconds, both auricles and ventricle ceasing to beat. At the
36th second a feeble contraction took place, followed at intervals by
several stronger ones, then, after a pause, a number of rapid beats were
registered in succession, followed by a pause of a few seconds’ duration;
after which the heart rapidly recovered its strength but not its speed,
which remained diminished as long as the experiment lasted.

These experiments prove that in the frog ethyl mercaptan acts more
powerfully upon the ventricle than upon the auricles; that in some cases
it can produce heart block, but the dose required to effect this is large

EXPERIMENTS ON PHYSIOLOGY OF ALLYL COMPOUNDS = 187

and the duration of the blocking short. Small doses accelerate the
rhythm to some extent, whilst large doses retard it for a considerable
time. In no case Was the dose administered sufficiently large to

permanently arrest the heart.

Fic. 2. Injection of 3 minims (0°18 c.c.) Ethyl mercaptan into Frog’s Ventricle.

Read from right to left

A number of experiments with ethyl mercaptan were performed on
deeply etherised rabbits; the solutions used varying in strength in
different cases. As with allyl sulphide the best diluent was found to be
olive oil.

The injection of 5 minims (0°30 c.c.) of a 20 per cent. solution into
the jugular vein of a large rabbit weighing 2250 grams produced a slight
fall in blood pressure, followed 40 seconds later by a slight rise above
the normal. The highest pressure was recorded 150 seconds later, and was
maintained for 90 seconds; then a gradual fall set in which reached the

normal in 260 seconds.

188 BIO-CHEMICAL JOURNAL

An injection of 15 minims (0°90 c.c.) of the same solution was then
made, producing a sudden rise followed by a continuous slow rise lasting
310 seconds.

When 20 minims (1°20 c.c.) of a 20 per cent. solution were introduced
into the jugular vein a slight fall from 86 to 81 mm. mercury at once
occurred, followed by a rapid rise to 93 mm. in 25 seconds. Then. there
was a sudden fall to 85 mm. in 2 seconds, after which a gradual rise to
89:2 mm. set in and occupied 163 seconds. At this moment the pressure
suddenly fell to 89 mm. and continued to fall slowly to normal, which
was reached in 135 seconds.

These small doses produced no visible alteration either in the heart
beats or in the respiratory movements.

One minim (0°06 c.c.) of undiluted mercaptan introduced into the
jugular vein of a rabbit (Fig. 3) produced, 10 seconds after its injection,
a rapid fall in blood pressure from 79 mm. mercury to 74 mm. in
13 seconds, succeeded by a further fall to 50 mm. in the next 29 seconds.
Thereafter the pressure began to rise and attained to 76 mm. 423 seconds
after injection. Coincident with the beginning of the pressure fall a great
respiratory effort was made, followed by increased respiration, lasting
60 seconds. The amplitude of these movements then began to diminish,
but at the 130th second another deep breath occurred, followed by
increased breathing, which continued so to the end of the experiment.
At the 310th second a very deep inspiration was registered, after which
shght Traube-Herring curves appeared in the pressure trace.

At the time of the pressure fall, the heart became weakened without
change of rhythm, and remained affected for some time though it
completely recovered its strength towards the close of the experiment.

The injection into the jugular vein of a rabbit of 3 minims (0°18 c.c.)
of undiluted drug caused at the end of 10 seconds a deep inspiration,
and at the end of 29 seconds a rapid fall in blood pressure. The pressure
fell from 69 mm. mercury to 48 mm. in 22 seconds, followed by a sudden
drop to 28 mm. in I second. At the 80th second recovery began, but
after reaching 88 mm. at the 103rd second began again to fall rapidly.
At the 150th second the pressure was only 10 mm., and at the 293rd it
had died out altogether.

The respiratory mechanism was much affected by this dose. The
deep inspiration at the LOth second was followed by normal breathing
until the 40th second, when suddenly the movements trebled in size for
four seconds, after which they grew rapidly more and more shallow,

— —

a

Pm J

MXPERIMENTS ON PHYSIOLOGY OF ALLYL COMPOUNDS © 189

without alteration in speed, until the 7Oth second. After that their

frequency as well as their force became less, until they failed entirely at

winmnin
ATT dda dull Re:

I] Hi ne HH)

fil

Fite ia B.P.

een lO a hee Lys Artem eMC ram anenaileionn tetsu we OOO ay /
e tert ny yygaantO ee rept enn

Seconds

Fic. 3. 1 minim Ethyl mercaptan injected into the Jugular Vein of a Rabbit.

the 124th second. Artificial respiration proved of no avail, though it was

continued to the end of the experiment.

190 BIO-CHEMICAL JOURNAL

The force of the heart began to diminish at the same time as the
blood pressure began to fall and the heart almost stopped when the
sudden fall in pressure was registered; it, however, recovered for a time,
but again became feeble just before the respiration ceased. Thereafter
the beats were minute, and the heart finally stopped at the 293rd second
after the injection.

Three minims (0°18 c.c.) of ethyl mercaptan introduced into the
circulation of a rabbit appears to cause death by respiratory paralysis,
whether it is administered in one dose or injected in two doses with a
short interval between.

The fatal dose for the rabbit works out approximately at 1} minims
(0°075 c.c.) per kilogram of body weight.

“ Injection into Serous Cavities

Three minims (0°18 c.c.) of ethyl mercaptan were injected into the
pericardial cavity of a deeply-etherised rabbit weighing 2800 grams. Its
blood pressure, which registered 98 mm. mercury, rose in 260 seconds to
126 mm., at which it was maintained for 90 seconds, then muscular
tremors set in, producing a slight fall, but when they ceased the pressure
increased to 146 mm. This high pressure was not long maintained, but
was followed by a fall, rapid at first, then gradually becoming slower,
it reached to 124 mm. Next a gradual rise set in to 158 mm., followed
by a slower fall to 108 mm.

In this experiment the respirations increased in amplitude
somewhat, especially just after the muscular tremors ceased, but
returned again to their initial size long before the blood pressure had
fallen to 108 mm. The heart’s action was not appreciably affected.

When the blood pressure had remained stationary at 108 mm. for
some time, 1 minim (0°06 ¢.c.) of allyl sulphide was introduced into the
pericardial sac, with the result that in 13 seconds the blood pressure
began to fall and reached 96 mm. in 17 seconds. Fifty-five seconds later
it had again returned to 108 mm.

No change was observed in either respiratory or cardiac movement.

Two minims (0°18 ¢.c.) of allyl sulphide were now slowly injected
into the pericardial cavity (Fig. 4) and produced considerable variation
in blood pressure, due no doubt to the fluid pressing on the heart. Ten
seconds after the injection a rapid fall in pressure occurred to 84 mm.
succeeded by a slower fall to 44 mm., which was reached at the 35th
second and at which it remained until the 50th second. A slight pressor

19]

“WQQUY VT JO APABH [LIpPABoLag 917] oqut poqoofur opryding [ATT Jo (-9° ZT.o) SUTIUIM GZ ‘p “DIG

SPUO 0S Ta ee area rere FETT TTT ATE TATTLE ET ETAT STL ATATTS ATTN DT VOTE TTD TO BRTTOTT VOSTATENT FUETOVTOT SUOTVNIY SETOTOOONVTOOTTOTY VSTVITETY TOTTATETY MTVSSII HTSTTIFTFFIVTAT TTT TTT TT TIT TE TA BTTIE VT ITaas edt

TNDS

Balboa WTO mee ae
rps Yap Mery a

dp eporepnagy wee PP [yc edpmannactee | Aad “Amen penne et ines
TT
HH MS ha l | | |p ae
, VAlderuaad yn nA
\?

' Ui

ALLYL COMPOT

Ol

_
ch
of
Lo

ia
LT fae

Vif tan

ie

EXPERIMENTS ON

192 BIO-CHEMICAL JOURNAL

effect then set in which reached 50 mm. at the 70th second, followed
by a depressor movement which ended at 0 mm. at the 420th second. The
respiratory mechanism was at first somewhat depressed, and from the
31st second began to fail rapidly. At the olst second hyperpnoea
supervened and lasted thirty seconds, after which respiratory movements
became gradually feebler and failed altogether at the 270th second,
thus failing much sooner than the heart. This is a typical action of
ally! sulphide.

One minim (0°06 c¢.c.) of ethyl mercaptan was injected into the
pericardial cavity of a deeply-etherised rabbit weighing 1850 grams. The
initial blood pressure was 85 mm. mercury. Eight seconds after injection
the pressure began to fall, reaching 56 mm. at the 24th second, a rise then
began and reached 68 mm. at the Tith second. This was succeeded by
another fall which attained to 63 mm. at the 107th second, at which level
it remained until the 217th second, when a new pressor movement set in
and reached 78 mm. by the 517th second.

Respiratory movements were but slightly diminished in amplitude,
without change of rhythm, and the heart remained quite unaffected by
this dose.

Two more minims (0°18 c.c.) were then slowly introduced into the
pericardium. Nine seconds later the blood pressure began slowly to. fall
and reached 60 nim. in 175 seconds, this was followed by a gradual rise
to 68 mm., which was reached 150 seconds later. At this time the
respiratory movements were becoming weaker though the heart was
beating normally. Four minims (0°24 ¢.c.) were now slowly injected into
the pericardial sac and produced uo effect for 126 seconds, when
respiratory spasms occurred and lasted 20 seconds. They produced a
shght rise in blood pressure, but it returned again almost immediately
to 68 mm. The animal was then killed.

Allyl Acetate

One minim (0°06 ¢.c.) (Fig. 5) of allyl acetate was injected into the
jugular vein of a deeply-etherised rabbit weighing 2300 grams.~ Ten
seconds after injection the blood pressure began to fall from 98 mm.
to 70 mm. in 16 seconds. The pressure then began to rise again, slowly
at first, then more rapidly and attained to 100 mm. at the 51st second.
A U-shaped depression was thus produced in the pressure trace. This
dose produced considerable alteration in the respiratory movements,
reducing their amplitude by one-third in the first twelve seconds. Slight

KXPERIMENTS ON PHYSIOLOGY OF ALLYL COMPOUNDS

L
a A
a “n F

i
;
:

19

1 minim (0:06 c.c.) Allyl Acetate injected ii

5 .

ria.

injected into

2 minims (0°12 c.c,) Allyl Acetate
Jugular Vein of

6.

Fic.

)

ular Vein of a Rabbit.

Jug

a Rabbit.

5

the

194 BIO-CHEMICAL JOURNAL

recovery then set in, but at the 19th second they again became smaller
and remained so to the 58th second; from this moment the movements
increased until they finally slightly surpassed their initial size.
At the 88th second a deep inspiration was recorded. During the whole
experiment no change in the respiratory rhythm or alteration in the
cardiac movements occurred.

Two minims (0°18 c.c.) (Fig. 6) were next injected into the vein,
producing similar results, beginning from the 8th second and being more
intense and prolonged than with the previous dose. Another 2 minims
(0°18 c.c.) were introduced into the vein, with the result that the blood
pressure began to fall from the 8th second, but did not fall so low as on
the previous occasion, and on again rising did not attain a higher level
than 93 mm., at which it remained to the end of the experiment.

These results show that allyl acetate is very feeble when compared
with ally] sulphide, but that its action is similar in that it affects the
respiratory mechanism first, leaving the heart practically unaffected by
the small doses exhibited. No doubt larger doses would have altered
the cardiac action also, but the supply of the drug failed.

The lethal dose of allyl acetate was not determined.

Allyl Alcohol, C,H,OH

To ascertain the action of this substance upon the vascular and
respivatory mechanisms it was deemed advisable to dilute it, otherwise.
intravascular clotting was to be feared soon after its introduction into the
circulation. One minim (0°06 c.c.) of a ten per cent. solution in normal
saline was injected into the jugular vein of a deeply-etherised rabbit
weighing 2750 grams, with the result that the blood pressure fell from
105 mm. mercury to 96 mm. in 79 seconds. The heart beat and
respiration remained unchanged, and therefore 4 minims (0°24 c.c.) of
the same solution were introduced, which sent the blood pressure up to
98 mm. in 31 seconds. A fall to 90 mm. at the 42nd second followed
and was succeeded by a rise that attained to 97 mm. at the 66th second.
This was not maintained, the pressure falling to 90 mm. at the 82nd
second, after which it again started to rise. This alternate rise and fall
of the pressure continued until the 222nd second, when another injection
was made. During the whole of this time the heart beats and respiratory
movements remained unaltered.

Five minims (0°30 c.c.) were introduced, as just stated, at the 222nd
second, and produced a fall of pressure to 87 mm. 19 seconds later,

MXPERIMENTS ON PHYSIOLOGY OF ALLYL COMPOUNDS = 195

followed by a steady rise to 108 mm., which was reached in another
131 seconds. As before, the respiration remained unaffected, but the
force of the heart was increased by one-half, without alteration in rhythm.
The introduction of another 5 minims (0°30 c.c.) produced a similar
result.

When this dose was doubled the blood pressure rose to 122 mm. in
190 seconds, but the respiratory and cardiac mechanisms were not further
altered by it.

Three minims (0°18 c¢.c.) of undiluted allyl alcohol were next
injected into the vein, with the result that the blood pressure fell to
79 mm. in three seconds, and was followed by a rise to 84 mm. at the
21st second. Then another fall began, reaching 69 mm. at the 52nd
second followed by a second rise to 88 mm. at the 47th second. Thereafter
a slower fall commenced, reaching as low as 65 mm. by the 95rd second
and remained at that level until a new injection was attempted at the
300th second, which was found impossible owing to a long clot extending
from the syringe to the thoracic veins. The only effect produced by this
dose upon the respiration was to cause two deep inspiratory movements,
but the force of the heart beat was increased by half its initial amount, its
rhythm remaining the same.

Forty minims (2°4 ¢.c.) of a twenty per cent. solution in normal
saline, equivalent to 8 minims (0°48 ¢.c.) pure allyl alcohol (Fig. 7), were
introduced into the jugular vein of a deeply-etherised rabbit weighing
2270 grams, the injection taking 86 seconds.

From the very commencement of the injection the blood pressure
began to rise from 108 mm., and reached 116 mm. in 11 seconds; then
it began to fall more gradually to 100 mm., taking 14 seconds to reach
that point, only to again rise gradually to 110 at the 51st second; followed
by another fall to 98 mm. at the 66th second. Again it rose to 108 mm.
at the 174 second only to fall to 46 mm. at the 1100th second. This was
followed by a rise to 66 mm. at the 1900th second, when suddenly violent
convulsions occurred that put a stop to further experiments.

This dose had an effect upon the respiration, causing at first a
temporary decrease in the amplitude of the respiratory movements,
followed by an increase with sundry fluctuations until the movements
had nearly doubled at the onset of the convulsions. A marked feature
of its action was the occurrence of a number of sudden very deep
inspirations, beginning at the 258th second and recurring at intervals
of about 70 seconds until the end of the experiment. The heart was but
little affected.

196 BIO-CHEMICAL JOURNAL

Res

B.I

40 rim 41:5 wugecbed

l

AL AUUUUEEEE GEASEE EEE GESEEE TEL COEKEE EEE GOPETETES CEEEEE EE CEEt Eo REE PpppPLMAL AAMECLEL MA ASCCEE LSALSSES SECSEEEEE SESE EASE CESESSEEE SSSSESSES SSSSESSES FESESENES SEUSESEES Ghareteee C4t: Se

Fic. 7. 40 minims (2:4 c.c.) of a 20 % solution of Ally] Alcohol in normal Saline injected

into the Jugular Vein of a Rabbit.

EXPERIMENTS ON PHYSIOLOGY OF ALLYL COMPOUNDS = 197

In the rabbit, therefore, allyl alcohol acts chiefly upon the blood
pressure, but in large doses affects the respiration also, resembling
phenyl-isothiocyanate in producing deep inspirations at intervals. — Its
effect upon the heart is slight, merely increasing its force.

The lethal dose was not ascertained.

Though allyl alcohol appears to have little action on the
mammalian heart when injected into the circulation, the case is very
different when it is applied directly to the amphibian heart, as will be
apparent from the following experiments : —

A drop of pure allyl alcohol] allowed to fall upon the surface of the
ventricle of a pithed frog at once diminished the force of the contraction
by one-third. This depressed condition gradually passed off, the beat
attaining to three-fourths of its initial amplitude in 140 seconds.

Two drops allowed to fall on the ventricle diminished the heart beats
by one-half; from which they recovered somewhat, but attained only
two-third their original amplitude in 230 seconds.

One minim (0°06 c.c.) of a twenty per cent. solution in normal saline
injected into the left auricle (Fig. 8) produced cessation of the ventricular
contraction in 15 seconds, though the auricles continued to beat as before.
Ninety seconds later the ventricle started to beat again, but not with
normal force; 320 seconds later the beats had regained only three-fourths
of their normal size.

When one minim (0°06 c.c.) of the same solution was injected into
the cavity of the ventricle (Fig. 9) the whole heart ceased contracting
in 10 seconds and remained motionless for 120 seconds. The auricles then
resumed beating, feebly at first, but gaining in strength they continued
to contract for 750 seconds, their contractions succeeding each other more
slowly than before the injection. The ventricle never contracted again.

From a number of such experiments performed upon large healthy
frogs it may be concluded that allyl alcohol does act upon the heart,
both auricles and ventricles being affected by it; but its action is much
more profound upon the latter chamber than upon the auricles.

Some experiments, by way of contrast, were performed upon the
hearts of pithed frogs with ethyl alcohol (C,H.OH) with the object of
ascertaining whether or no the results obtained with alcohol were due
to dehydration or to coagulation of the muscular substance.

One minim (0°06 c.c.) ethyl alcohol placed on the surface of the
ventricle, at once brought the heart to a standstill, but this lasted only
ten seconds, after which the whole heart started beating again with
normal force and rhythm.

JOURNAL

BIO-CHEMICAL

198

‘QPOIGUOA $,Ho1yf OFT poqoolur [Oqoo[y Lqaop jo uorgnqos 9% 0% ¥ Jo (70° BT.O) SUE gE "OT “OL

TOMI wn
S31 seme sy

|

\ Oe ee Ase aneaentine Sea |
~ os snag

“OPPLATTU9 $00.1 O77 poqoolut [OUOOLV [ALLY jo UOT}N[OS Ye OZ ¥ Fo ("0"0 90-0) UUTLUTUL OU) > OLA

RRA

‘OOLINW § Soa UL poqor 1 y 7 TATLY é (
PLNY $,Soaq oqut poqoolar ourpeg [VULION ut [OYooTY [ATTY Jo Worynyos % % jo (100 90-0) WAtUII 9UQ “8 “DIY

___— mA

EXPERIMENTS ON PHYSIOLOGY OF ALLYL COMPOUNDS = 199

A single drop placed on the auricles caused the heart to stop at once
for 17 seconds, then, after slowly contracting twice, it resumed its
normal beat.

In both cases the effect was suddenly produced and as suddenly
recovered from.

Three minims (0°18 c¢.c.) of a twenty per cent. solution in normal
saline injected into the ventricle (ig. 10) stopped the heart in diastole
for 120 seconds, the auricles then resumed their beating, somewhat
feebly at first, but steadily gaining in strength. The ventricle resumed
beating 140 seconds later, at first feebly with numerous interruptions, but
gradually increasing in strength, the beats had attained to one-half their
original amplitude when a second injection was made. This only
increased the duration of the period of rest and delayed recovery.

This is again an example of the lethal effect of allyl as compared
with ordinary alcohol. One minim (0°06 ¢.c.) of a twenty per cent.
solution of the former was sufficient to permanently arrest the ventricle
whilst 5 minims (0°18 c.c.) of the same strength of ethyl alcohol did not
prove fatal to the heart’s action. Both alcohols have this in common,
that they affect the ventricle more profoundly than the auricles.

CONCLUSION

From all these experiments it may, I think, be concluded that the
poisonous action of the allyl compounds is due to the allyl they contain,
though in some cases, no doubt, its activity may be increased, and in
others diminished by the substances with which it happens to be
combined.

My thanks are due to Mr. C. Lovatt Evans, B.Sc., for -the
preparation in my laboratories of most of the compounds used in these
experiments.

200

THE INFLUENCE OF PROTOPLASMIC POISONS ON
‘REDUCTASE ’

By D. FRASER HARRIS, M.D., D.Sc., Professor of Physiology and
Histology in the Dalhousie University, Halifax, N.S.

(Received December 16th, 1911)

Continuing my investigations! into the reducing endo-enzyme, I
have examined a number of ‘ protoplasmic poisons’ to see whether they
inhibited the activity of tissue reductase.

Waller? lately emphasised the necessity of distinguishing between
the action of a poison upon protoplasm and its action on enzymes. He
pointed to chloroform killing the laurel leaf but liberating instead of
restraining an enzyme in the leaf which evolves hydrocyanic acid. The
poison studied must not, of course, by itself bleach any of the pigments
used in this research, viz., soluble Prussian blue, methylene blue and
sodium indigo disulphonate. This condition eliminated such a poison
as hydrocaynic acid, for instance, so that I need only record results with
those as follow :—

Chloroform. ive drops of chloroform added to 5 c.c. of sheep's
liver juice in presence of 5 c.c. of 0'06 per cent. of Prussian blue exerted
only a very shght inhibitory action.

Chloroform has, however, the property of altering the physical state
of the press-juice as to make it much less miscible with watery liquids so

that it thus removes the pigment from the sphere of action of the
reductase.

To vary the conditions, I perfused a sheep’s kidney, previously
washed till blood-free, with 1 per cent. sodium chloride having 10 per
cent. chloroform in it. As soon as the vessels had been traversed by
this mixture, the kidney became of a soft, ‘ dead white’ appearance, and
was also more friable.

1. This Journal, Vol. V, p. 148, 1910.
2. Proc. Phys. Soc., June 18th, 1910.

INFLUENCE OF PROTOPLASMIC POISONS 201

The glycerine extract of this chloroform-killed kidney reduced
Prussian blue, but much more slowly than did the juice of a control of
normal kidney.

The general conclusion come to was that chloroform did not inhibit
the enzyme, but tended to coagulate the cell-proteins in which the enzyme
was fixed and immobilised.

Sodium fluoride. A O05 per cent. solution was used. Employed
against a control, this substance certainly did not seem to act as an
inhibitant of the reducing enzyme. The solution was noticed to be
distinctly alkaline, so that the bleaching might have been due to that
condition.

Accordingly some sodium fluoride solution was made very slightly
acid, and added to a mixture of glycerine extract of sheep’s liver and
Prussian blue. This mixture was well reduced, aud the leuco compound
readily restored by treatment with hydrogen peroxide. Sodium fluoride
does not per se inhibit reductase.

Nitrobenzene. This is a powerful protoplasmic poison without
reducing action on the three pigments. A glycerine extract of sheep
liver juice was used with this poison. -

After twelve hours in the water, bath at 40°C. the contents of the
tube with the poison and the control tube were both well reduced; the
laboratory notebook has the entry: ‘if anything, the tube with the
nitrobenzene is the whiter.’ This observation was repeated several times,
so that I think we may infer that nitrobenzene is not a poison to
‘reductase.’

Formalin (10 per cent. solution). This poison does not reduce
Prussian blue although it precipitates it. Some kidney press-juice and
Prussian blue were mixed with 5 ¢c.c. of the formalin solution and, along
with a control tube, were kept in the water bath at 40°C. overnight.
Next morning the control tube was almost white, while the tube with
the poison was still blue. Suspecting acidity, I found that this formalin
was acid both to litmus and to phenolphthalein. When a little formalin
was the very slightest over-neutralised by some N/10 sodium hydrate,
and added to kidney juice in presence of Prussian blue, the pigment was
bleached. It would appear that the reductase was previously inhibited by
the acidity of the formalin and not by its poisonous property, whatever
that may be.

202 BIO-CHEMICAL JOURNAL

From these experiments the general impression seems to be
that acidity is. the inhibitant factor rather than the toxicity.
Hydroxylamine sulphate, for instance, a powerful protoplasmic poison,
is so acid in reaction that the fact that it prevents any fading of
Prussian blue in presence of tissue-juice is not to be taken as a proof
that hydroxylamine sulphate inhibits the reducing endo-enzyme.

The expenses involved in this research were met by a grant from the
Government Grant Committee of the Royal Society.

208

NOTE ON HYDROLYSIS OF VEGETABLE OILS BY
EMULSION OF RICINUS COMMUNIS

By DAVID SOMMERVILLE, B.A., M.D., M.R.C-P., Lecturer im
Public Health, King’s College, University of London,

(Received October 7th, L9/1)

In 1902 Connstein Hoyer and Wartenberg described a series of
experiments carried out at Charlottenburg, by which it was sought to
determine the modus operand: otf fat-splitting in vegetable oils by an
emulsion of the castor bean. Hoyer continued the work in 1904 aud
made an attempt to isolate the enzyme. He used as activators of his
enzyme acetic acid and sulphate of manganese.

In 1906 Nicloux, after several years’ work on the fat-splitting
properties of the castor oil bean, came to the conclusion that the
cytoplasm of the seed contains the active agent, and that this substance
acts Im every way as an enzyme, but is not an enzyme.

It is customary to-day to assume that hydrolysis is preceded by a
combination of the hydrolyte with the enzyme

Emil Fischer’s lock and
key relationship between bodies of asymmetric configuration.

It is known that esters of the lower terms are more difficult to
hydrolyse than those of the higher; in other words those esters most
readily soluble in water are most difficult of hydrolysis. Is this due to
a previous hydrolysis of some portion of the hydrolyte, or to a previous
hydrolysis of some portion of the so-called enzyme?

In reviewing the experiments of Nicloux and Hoyer | worked with
the castor oil bean acting on its own oil and on cottonseed oil. The results
were equally good in the two cases, so that no specificity can be claimed
for the enzyme of the castor seed towards its own oil.

I prepared an emulsion of ground castor beans and incubated it at
25°C. until hydrolysis of the oil had been definitely established. A
measured quantity of the incubated emulsion was then added to a
mixture of measured quantities of oil (castor or cotton, as the case might
be) and water. The mixture was shaken for a few minutes at intervals
of a quarter of an hour for some hours and then left. The percentage of
free fatty acids formed in three days at laboratory temperature ranged
from 80 to 85.

On repeating Hoyer’s experiments, in which he used acetic acid
and manganese sulphate, I did not find that the addition of these bodies

produced any increase in the amount of free fatty acids liberated.

204 BIO-CHEMICAL JOURNAL

I noticed early in the experiments that if hydrolysis (fatty acid
liberation) were not properly established in the emulsion (enzyme) the
yield of free fatty acids was correspondingly low. Success in the
experiment largely depends on the intimate mixture of oil and enzyme,
The preparation of the enzyme demands the use of a quantity of water
from which the enzyme can later be separated. This water when applied
to oil is wholly inactive. Further, the active enzyme, when mixed with
two or three volumes of water, alcohol, or acetone, rapidly loses its
activity; whereas when mixed with two or three volumes of ether or
benzene no loss of activity occurs. On the lock and key hypothesis the
meaning of these contrasts is difficult to seek.

Nicloux deseribed his enzyme as refractory to heat when protected
by oil; my emulsion, when heated to 60°C., rapidly loses its activity.
But when the’ bean from which the emulsion is prepared is heated to
100° C. for twenty-four hours in a dry bath little or none of its activity
is lost.

Pancreatic lipase has been separated into two portions by ordinary
filtration, and it has been found that the activity of the filtrate -is not
diminished by boiling, whereas heating the residue destroys its potential
energy. I have not been able to effect any such separation in the castor
bean enzyme.

That the activity is related to nitrogen-containing matter can be
demonstrated; but whether the nitrogen is active or not it 1s impossible
to say. The total organic nitrogen of my active emulsions, as estimated
by Kjeldahl’s method, always amounted to at least 0°2 per cent.

Cotton oil and castor oil, when carefully neutralised and freed from
proteins fails to undergo hydrolysis. If to this oi] an enzymic emulsion,
in which hydrolysis of fat has not yet commenced, be added, nothing
results. But let hydrolysis be established in the emulsion, and
hydrolysis proceeds in the added oil, irrespective of whether this oil be
neutral or contain within wide limits (25 per cent.) free fatty acids.

Perhaps the most striking of all these features is the failure of the
prolonged heating at 100°C.

a temperature which kills practically all
proteins—to interfere with the activities of the bean. Is it that in the
case of the dry bean, when heated, hydrolysis fails to commence, whereas
in the mixture of enzymes and water hydrolysis rapidly exhausts itself?

205

A CLINICAL METHOD OF ESTIMATING THE AMOUNT OF
CALCIUM IN THE URINE AND OTHER PHYSIOLOGICAL
FLUIDS

By W. BLAIR BELL, B.S., M.D. (Lonn.), Assistant Gynaecological
Surgeon, Royal Infirmary, Liverpool, ete.

From the Bio-Chemical Laboratory, University of Liverpool
(Received March 15th, 1912)

INTRODUCTION

To gauge the condition of the general metabolism of the calcium salts,
an accurate knowledge of the amount at the disposal of the tissues is
necessary, and this knowledge has been difficult to obtain.

Tor instance, some workers have estimated—chiefly experimentally—
the amount of calcium taken in with the food, and have compared this
with the amount recovered from the faeces.

This method cannot give reliable results. For although we know the
amount taken in by the mouth, we are unable to find out how much of this
quantity is absorbed and how much passes through the alimentary canal
without being absorbed. Consequently, it is evident that although we
may kuow the amount taken into the alimentary canal and the amount
passed out, calculations based on such data must err, because we do not
know how much has simply passed through. Recognising this difficulty,
in 1907 I published a method of estimating the quantity of ionizable
calcium in the blood!. By this method one is able to determine at any
given time the relative amount of ionizable calcium in the blood, and so
to get a good indication of the amount at the disposal of the tissues. But
as I then, and have since, emphasised, even this method requires careful
literpretation. For instance, a low calcium content in the blood may
mean either that very little was being absorbed or that too much was being
excreted. I devised the method for a specific experimental purpose in
regard to menstruation, and it served that purpose; but when it was used
for routine clinical work, unless careful consideration were given to the
findings, wrong deductions—as I have just pointed out—might be drawn
from them.

If, however, the blood estimation method be used in conjunction with
a method which will indicate the amount excreted, a perfectly reliable

1. Brit. Med. Journ., April 20, 1907.

206 BIO-CHEMICAL JOURNAL

knowledge of the condition of the general metabolism in“regard to the
calcium salts may be obtained.

Since the excretions from the intestinal canal are, as I have just
shown, useless for the purpose, we must fall back on the urine, in which
it is safe to say that, provided the urogenital apparatus be healthy, the
calcium found must represent a true index of the excretory ratio. If the
- blood index be low, and the amount excreted in the urine also be small, it
is fair to assume that not enough calcium is being absorbed. If, on the
other hand, the blood index be low and the urinary excretion high, then
probably too much is being excreted; and conversely a high blood index
and low excretory ratio might indicate that an excessive amount was being
retained in the tissues.

>

PRELIMINARY INVESTIGATIONS

1. The precipitates obtained with oxalic and acetic acid from many
specimens of urine were pooled, and on chemical examination were found
to consist of pure calcium oxalate. It may be noted here that practically
all the calcium present in urine is in an inorganic form, consequently it is
precipitated. When the proper measures, to be described directly, are
adopted, no phosphates were found in the many precipitates examined,

2. It was found that a rough calculation of the amount of precipitate
was not accurate, except in so far as a very large quantity could be
distinguished from a small one. Our attention was therefore directed to
devising some clinical method of estimating accurately the amount of the
precipitate which we had proved to consist of pure oxalate of calcium.
Our experiments consisted in centrifuging and measuring the deposit in
calibrated tubes.

Our first difficulty was to prevent the deposit sticking to the sides of
the tube, and to get it to lie fairly evenly to the calibrated portion. After
trying many things, we finally overcame what threatened seriously to
interfere with the perfection of our method by adding alcohol or methy-
lated spirit to our urine and reagent.

Next, we found, as we anticipated, that with different rates of
revolution of the centrifuge,and with difterent times spent on the operation
we obtained different results; that is to say, urines and solutions containing
the same quantity of calcium gave deposits which stood at different levels
according to the circumstances mentioned.

However, we eventually got over this difficulty by centrifuging a

THE AMOUNT OF CALCIUM IN THE URINE 207

standard solution of artificial urine containing a known quantity of
calcium in one tube and the urine to be examined in the opposite one. In
this way we were able always to get an absolutely accurate relation between
the two deposits whatever the rate of revolution or the time employed. So
that if the urine showed a deposit of calcium oxalate standing at half the
height of the precipitate in the tube containing the standard solution, that
urine was found on chemical analysis to contain a quantity of calcium just
half in amount of the known quantity in the standard solution. All our
results were, of course, verified chemically.

DESCRIPTION OF THE METHOD

A sample from a twenty-four hours’ specimen of urine is made faintly
acid with hydrochloric acid to dissolve any insoluble phosphate present.
It is then made faintly alkaline with ammonia, and filtered. Next
5 c.cm. of the filtrate are placed with a pipette in the special centrifuge
tube, which is of the usual size and shape in the upper portion, but tapers
at the lower end into a cylindrical extremity of even bore (1°25 mm.) and
calibrated into 1 mm. divisions (Fig. 1). A line, with ‘ urine’ marked
below it, encircles the upper part of the tube at the 5c.cm. level. Any air
bubbles which may collect in the calibrated portion are got rid of with a
fine wire or strand of silkworm gut. Then 1 c.cm. of the reagent,
consisting of a saturated solution of oxalic acid in a 5 per cent. solution of
acetic acid, is added. The correct quantity of reagent (1 c.cm.) is also
indicated by a line round the tube which is marked ‘reagent.’ Finally
2 c.cm. of alcohol or methylated spirit, as indicated by the line marked
‘alcohol,’ are added, and the contents of the tube are thoroughly mixed
by shaking.

The second tube is then taken, and 5 c.cm. of the standard solution!
is run into it with a pipette that is up to the line marked ‘ solution’
(Fig. 2), and any air bubbles removed as before. Next the reagent and
alcohol are added as in the case of the first tube, and the whole
is thoroughly shaken. Both tubes with their calibrated ends packed in
wool are then carefully placed in the opposite buckets of a centrifuge, and
are centrifuged for about a quarter of an hour. On removing the tubes
the precipitate will be found to stand at a certain height, say 10 mm. in
the ‘standard solution’ tube, while it may stand at 7 mm. in the other

1. Standard Solutions.—0-05 gram. of calcium phosphate (Cas(PO,4)o] is dissolved in a little
hydrochloric acid. This is made alkaline with ammonia, and finally acid with acetic acid.
Finally, 2 grams of urea are added to the solution, and the whole is diluted up to 100 c.cm,
with distilled water. Specific gravity =1015.

208 BIO-CHEMICAL JOURNAL

which contains the urine to be examined. As a rule, there is a slight
slant on the surface of the deposit. This can be obviated by stopping the
machine at the end of one or two minutes and turning the tubes through
half a circle.

But if this be not done—and it is not absolutely necessary, although it
makes for greater accuracy—the height of the mzddle of the meniscus or

slant in each tube should be read off and compared. In the above instance
the deposit in the tube containing the standard solution was stated
to stand at 10 mm., and that in tube containing the urine at 7 mm.,

WT

nw

— rw
HA ©

~
°

o

SEL e

Jehwei, al Fig. 2

consequently the urine contains 7-10th or 0'7 of the quantity of calcium
in the standard solution, which is known to be 0°02 per cent; in other
words, the urine contains 0°02 x 0:7 = 0°014 per cent. of calcium. This

es are U 1 ‘ : ;
gives a general equation g x 50 =percentage of calcium in the urine

examined, in which U = height in millimetres of precipitate in the urine

THE AMOUNT OF CALCIUM IN THE URINE 209

to be examined, and S = height in millimetres of precipitate in the
standard solution.

If any specimen of urine be found to contain an unusually large
amount of calcium, so that the precipitate more than fills the calibrated
portion of the tube, it should first be diluted with an equal quantity of
distilled water, and the final result obtained in regard to the calcium
content doubled.

In our experiments to test the accuracy of the method we found that
the difference between the percentage of calcium obtained in this way and
that obtained by chemical analysis never amounted to more than 1 per
cent. of the quantity present.

210

OBSERVATIONS ON THE SECRETION AND COMPOSITION
OF HUMAN BILE

By J. A. MENZIES, M.D. (Epr.)

From the Physiological Laboratory, University of Durham College
of Medicine, Newcastle-upon-Tyne.

(Received March 16th, 1912)

An opportunity having offered for making some observations on the
secretion of bile in the human subject in a case of biliary fistula, it has
appeared advisable to record the results obtained, though it is recognised
that in themselves these observations do not lead to any definite
conclusions, and are only of value as confirming, or otherwise, accepted
teaching or as offering suggestions for further work.

The patient was a woman, aged 24, who had been operated on in
October, 1911, for acute suppurative cholecystitis with gall-stones. The
gall-stones and half of the gall bladder were removed. The wound was
drained for a time, but had healed five weeks after the operation. It
remained healed for two days, then broke down with the formation of a
fistula. Jaundice was present before the operation, but after operation,
in the opinion of the surgeons, the common bile duet was not completely
blocked. The faeces were said to be ght coloured. Apart from the
fistula the woman’s health was good.

The research was undertaken on Professor Bainbridge’s suggestion,
the primary object being to ascertain the immediate effect, if any, on bile
secretion of the administration of protein, carbohydrate, fat or hydro-
chloric acid. It was, of course, impracticable to introduce these substances
into an absolutely empty stomach, and therefore pure results could not be
anticipated. Moreover, inasmuch as the common bile duct was pre
sumably patent, though probably constricted, the further difficulty had to
be allowed for that the results obtained by the collection of the fistula bile
would be unreliable in so far as there wasa passage of an unknown quantity
of the secretion into the intestine. With a view to minimising this source
of error, arrangements were made for the examination of the faeces, but
unfortunately owing to a series of misunderstandings the stools were never
obtained. The pale colour of the faeces and the probability that the
common bile duct was thickened as a result of the original disease make it
reasonable to suppose that only a small quantity of bile, if any, found its

BILE 211
at employed for bladder

It was measured at intervals of
With each is given the total

ow
=a
el ie
23
—
a &
fae Secessesecusecssscscsusesesccscasssscse oamma sesssbesctseseeiss 20 m
——————— ooo ounecesa Z, ead
— gasseeues th
re Secsesesootsuesceceucgonuueana gag a
Sotsecuecarscueceeus ccqescesasGocscsqune ‘sacuoses cw
Gucereezeuceuas zuseee Beas.
AG susescsscssuene seaeeees: iS On
Sesecnensccesee uuesesa 4 pete)
aqueucuecsoness com:@ 2-5
Tose osenoses| gpaue: : A's iB
LAA NUR Soeede co + Li om
peagugescsesuscsss 2.
tere auauengees o) Oo
agunecaeso eusneas Pw
ro esunseeceso seané Le =
hig egage saegseneee ees!) euue. oO .f
seaceseuesser- - Oacec
EI quewmeecascosse Sn
Bar Sesugdesaeegaga m~ad
2Se8se000 cuses > 8
= sucmmasuuaceses o 4
gquagageuss sens: Chea yeee)
SS 5 or
gusngeesea D> ro
eeanascens to]
poses geusua’. cas ooo
SEO 727s... =| a
suvenac- “su nN :
oo ooo rr S
ISq Gees oeeeeneeee ios om
asageesesessess PS NS
egaen snuaeeuess a Gs q
cade engarenaao a
egeausesceseuss oI =| =
Aa! Segue enone cones: ry a .
jagdeeae seRESseses sane co O o
aacae sueeacuesa law} ~
SSH seatoasseccanss Co
Ade weees sesso seess o
aaenpecces onsen ne

TION AND COMPOSITION OF HUMAN

“
v)

SECRI

Making full allowance for these difficulties, there

way into the intestine.

are still some points of interest to be found in the results obtained.

The bile was collected by a tube similar to th

“

drainage after suprapubic cystotomy.

five, ten or fifteen minutes, but it was found that the results were most

instructive if the fifteen-minute interval was adopted throughout, and

joined charts each second ordinate represents a

.

5

accordingly in the sub

The patient was in bed throughout the whole

fifteen-minute interval.

The charts show the rate of flow in relation to the

period of observation.

diet, each abscissa representing 0°5 cubic centimetre of bile and every

second ordinate a quarter-hour period.

quantity collected and a note of the nature of the pigment.

¢

2

Bile in 11 hours

darker in tint than day bile.

Fia I.

39 o.c. (8 p.m. to 8

In this and the sueccee

0°5 c.c., and each second ox
Fig. I shows the rate of flow during one day on ordinary hospital diet.

The curve divides itself into three periods, the first of three and a half
hours from 9-30 to 1 o’clock, ending in a sharp fall within half an hour

after the midday meal; the second of three hours from 1 to 4, ending in

a sharp fall immediately after tea; and the third of four hours from 4 to 8,

showing a slight fall after supper at 6 o’clock and a final fall at the end

of the curve, The average flow for each quarter-hour during the first

212 BIO-CHEMICAL JOURNAL

period was 83 c.c., during the second period 6°5 c.c., and during the third
23 ¢.c. The flow from 8 p.m. to 8 a.m. the next day totalled 99 c.c., or an
average for each fifteen minutes of 2 c.c. There are thus two main features
of interest in this curve, (1) the progressive fall in the output from morning
till evening, and (2) the sudden diminution after the midday and afternoon
meals. ‘The suddenness of the fall within at most half an hour of the
meal suggests four possibilities, (1) a mechanical interference with the
flow by pressure of the distended stomach on the portal fissure,
(2) inhibition of the secretion, (3) diversion of the flow to the common
bile duct, or (4) removal of a normal stimulus. Mechanical interference
is a priort improbable, and is made more so by the fact that there was no
increase in the flow from the fistula as the stomach contents diminished,
but, on the contrary, the rate of flow remained at the lower level.
Inhibition either by chemical or nervous means would be contrary to the
known facts with regard to other secretions in connection with the
digestive system. With regard to the possibility of a diversion of the
flow into the duodenum, it is common knowledge that no flow of bile into
the intestine occurs normally so soon after food is taken, and the question
remains whether this fall has any association with the previous meal.
Normally, of course, the gall bladder empties itself a few hours after a
meal is taken, and it may be assumed that the contraction of the gall
bladder is accompanied by some degree of dilatation of the bile duct.
Mr. R. J. Willan informs me that in a case of duodenal fistula in the Royal
Victoria Infirmary, Newcastle, a gush of partially digested food mixed
with bile escaped with great regularity from three to four hours after each
meal. If this is the normal time for the outflow of bile in the human
subject, it is obvious that the diminished flow in this case coincides with
the normal period for escape of bile into the duodenum in relation to the
previous meal, and the occurrence immediately after food may be a pure
coincidence. On the other hand, the meal at 9 a.m. was a very light
one, consisting of soup; and again on the 22nd November, when the midday
meal was given an hour earlier than usual, the diminution of flow also
occurred earlier. Professor Bainbridge has suggested the fourth possible
explanation which appears to be the most probable, viz., that as a result
of the entrance of the meal into the stomach the pyloric orifice is closed for
a time, and the normal stimulant for the production of secretion is thus
withdrawn, leading to a diminished flow until digestion has proceeded for
a time and chyme again begins to pass into the duodenum.

I had no opportunity of repeating this full-day collection, but it is

SECRETION AND COMPOSITION OF HUMAN BILE _— .218

noteworthy that the quantity of bile collected during the midday hours in
the subsequent experiments, with two exceptions to be discussed later,
agreed closely with the quantity during the same hours on the normal
day. The quantities obtained between the hours of 11 a.m. and 1 p.m.
on the seven days were as follows (11.10 to 1.10 on the 28th) :—

21st ee x 1 SEE Ge.
eend © 3%. +5 wet -OS-G6;
rs) re oak ide pe ES
24th oe = Seg anes
25th ie $22 .. 85be.e.
e6th ... a wis WOTGSES
2oth a... file c: Sapeie,

On the 22nd a solid meal was taken an hour before the usual time,
between 11 and 12 instead of between 12 and 1, and the corresponding
fall in the rate of flow took place before 1 o’clock. The small quantity
obtained on the 23rd will be discussed in relation to the diet. Omitting
these abnormal days, it will be seen that the rate of flow at this period of
the day is remarkably constant, and gives some support to the possibility
that Fig. I represents a normal rate of secretion.

As will be seen from Fig. II, a carbohydrate meal with a negligible
amount of protein was given on November 22nd instead of the customary
midday meal, and about an hour before the usual time for that meal. It
was followed by a diminution in the rate of flow similar to that observed
in connection with the ordinary meals on the 22nd, but the curve shows a
sharp rise followed by an equally sharp fall two hours after the
meal. This rise appears to be due to a sudden gush, possibly from some
mechanical cause such as a change of position, and the average rate of
flow for two and a quarter hours after food is just under 7 c.c¢ per quarter-
hour.

The absence of protein in this day’s food appears to be the only
explanation available for two peculiarities which appeared on the
following days. On November 23rd the flow was markedly diminished
(Fig. III), only 28 c.c. being obtained in six hours, or an average of just
over 1 c.c. per quarter-hour. This came mainly in three gushes, two of
which followed the administration of cream and soda, but, in the light
of the other experiments, were probably independent of this. The third
gush occurred just before tea without any perceptible cause.

The second peculiarity which may possibly be due to the carbohydrate
diet was that on November 24th, the second day after the non-protein food,

214 BIO-CHEMICAL JOURNAL

the bile was markedly green in colour. This green colour persisted for
three days, till on the night of the 26th the bile showed a preponderance of
brown, and on the 28th (no bile was collected on the 27th) the colour was
again the normal brown of human bile. Kunkel? states that in dogs the
sulphur taken in protein is excreted partly in the bile, and that whereas
the excretion of protein sulphur in the urine takes place on the same day
on which the protein food is taken, the increase in the biliary sulphur does

sat

a
aner

ry Ty

a

o

i
bial
Aare ara
ri
aa

a

oF
a
a

2
;

ne

eee
geceill

Fic. II. Bile in 2? hours 81 c.c. Fia. III. Bile in‘6 hours 28 c.c.
Colour, brown. Colour, brown.

not appear till two or three days later. In view of that fact, it appears
possible that the green colour of the bile, appearing in this case from the
second to the fourth day after that on which the carbohydrate food was
taken, may be traceable to the nature of the food. This is especially
possible when one remembers that the bile of herbivorous animals is
habitually green in colour.

Hydrochloric acid (half a pint of 0°2 per cent. solution) was admini-
stered with a view to stimulating the production of secretin. As Hertz? has
shown that food begins to leave the stomach very shortly after a meal, and
as the present observations appear to indicate that no marked constriction

1. Quoted by Noel Paton in Schifer’s Physiology.
2, Brit. Med. Journal, 1912,

VS ee ee eee

SECRETION AND COMPOSITION OF HUMAN BILE 215

of the pylorus follows the ingestion of fluid, it may be assumed that any
increase in the flow of bile will be due to the influence of the acid in
stimulating the production of secretin, the fact that secretin does excite
the flow of bile having been demonstrated by Bayliss and Starling.' The
chart shows that such an increase did occur (Fig. IV). The flow averaged
over 12 c.c. per fifteen minutes during two and a half hours after the
administration of the acid, in spite of the fact that some bile was lost by
leakage. A repetition of the experiment on November 28th confirmed
this result (Fig. VII), the fifteen-minute average during one and three-
quarter hours after the acid was given being 11°4 c.c.

ae nh
ty
HAH 1
ogee coosecae:
seas
oe nee tte
ra
44
Ht

Fic IV. Bile in 3} hours 129 c.c. Fic. V. Bile in 2}hours 98 c.c.
Colour, markedly green Colour, distinctly green.

Peptone (Fig. V) and bovril (Fig. VI) did not appear to have any marked
immediate effect on the flow of bile, the average flow for fifteen minutes
during one and three-quarters hours after the administration being 10 c.c.
and 9 c.c. respectively, or about the average rate for the period of the day
at which these substances were given. It is worthy of note, however, that
the amount collected on the 26th and during the night of the 26-27th was
above the average (see below), and the question arises whether this increase

1, Journal of Physiology, Vol. XXVIII, 1902,

216 BIO-CHEMICAL JOURNAL

is in any way connected with the increased nitrogenous intake. The

peptone was administered on the 25th and the bovril on the 26th, in each

case during the forenoon and in addition to the ordinary meals. The
increased flow occurring on the day after the addition of the peptone to the

diet is in sharp contrast to the great diminution which occurred on the day

after a protein-free diet, and supports the suggestion that the remote
metabolism of protein food is a factor in bile production.

(
&

fa wauewesenee.
Sage eaape so

aL
He

1
eane

s
Boot

EERE EEE EEE Ee
Sooeceoe casseeceuueueccen
Sal eal ee ratte

Fia. VI. Bile in 3 hours 114 c.c. (8 p.m. to Fig. VII. Bile in 2? hours 110c.c.
8a.m., 27th, 266 c.c.) Colour—in day bile, Colour, brown.
ereen preponderates ; in night bile, brown,

ToTaL DaIny QUANTITY

The total quantity obtained on the 21st-22nd in twenty-three hours
was 338 c.c., and the quantity for twenty-four hours may therefore be
estimated as 350 c.c. This was on ordinary diet. On the 26th-27th
380 c.c. were collected in fifteen hours, giving an estimated total for
twenty-four hours of about 600 c.c. This was, as stated above, after an
increased nitrogenous intake during the previous two days. Previous
observations! of the daily quantity of bile secreted in the human subject

1. Quoted in Gamgee’s Physiological Chemistry of the Animal Body, Vol. II, 1893,

SECRETION AND COMPOSITION OF HUMAN BILE 217

show results varying from 374 c.c. (Yeo and Herroun) to 850 c.c. (Mayo
Robson). Ranke obtained 636 ¢.c., von Wittich 542 ¢.c., Copeman and
Winston 779 c¢.c., and Paton and Balfour 590 c.c. These figures would
indicate that in the case under investigation a very considerable proportion
of the daily secretion was obtained through the fistula.

COMPOSITION

Several analyses were made, the charcoal method being chiefly used,
but these only confirmed the accepted figures for fistula bile. It was
found that the bile secreted by night had virtually the same composition
as that obtained during the day. As a rule the night bile had a shghtly
higher specific gravity, approximating to 1,011 when the day bile was
1,010, and the night bile of the 21st-22nd was distinctly darker in colour
than the day bile. The results of analysis were almost identical with those
of Jacobsen!, who found the solids in his case (also fistula bile) amounted
to 2°26 per cent. In the present case 9°275 grams of day bile (November
26th) yielded 0°209 gram or 2°253 per cent. of solids. 10074 grams of
the corresponding night bile yielded 0°217 gram or 2°2535 per cent. of
solids. The ash of the day bile equalled 0°582 per cent. of the total bile,
and that of the night bile 0°516 per cent. But whereas Jacobsen found
over 1 per cent. of glycocholate of sodium, I found the bile salts always
under 0°5 per cent. A typical result is the analysis of the night bile of
November 26th-27th :—

Water sag: ae 97-747
Total solids .% ee 2293
Bile Salts <4 ~ 0416
Cholesterol... sa 0094
Lipoids 33 ae 0°298
Agph ... ee ae 0°516

This leaves 0°929 for mucin and pigments.

REMARKS

The observations recorded in this paper are much diminished in value
by the uncertainty as to the condition of the common bile duct and the
absence of information as to the quantity of bile which may have found
its way into the intestine. They indicate, however, that the composition

1. Quoted in Gamgee’s Physiological Chemistry of the Animal Body, Vol. I, 1893.

218 BIO-CHEMICAL JOURNAL

of day and night bile is the same, and they suggest that the eftect of food
substances on bile secretion may show themselves from the second to the
fourth day after ingestion, in agreement with Kunkel’s results! on sulphur
excretion. They further appear to support Starling’s statement? that
‘As an excretion the production of bile must be continuous, and related,
not to the processes of digestion, but to the intensity of the destruction of
the red corpuscles.’ They raise the question as to whether the more
remote metabolism of the ingested food may not have an equal share with
the destruction of red corpuscles in determining the amount of secretion,
thus throwing doubt on what Moore? has said, that * No connection exists
between the amount of proteid metabolism and the amount of cholates
produced, such as would be found if the cholates were a channel for the
excretion of the nitrogen and sulphur of proteid decomposition products.’
They accord rather with Spiro’s statement‘ that nitrogen and sulphur are
increased in bile by a protein diet, and with that of Rosenberg and
Barbera® that in dogs both bile and its solid constituents are increased
after protein food.

I have to express my thanks to Professor Rutherford Morison for
permission to make these observations on his patient. I am very greatly
indebted to Professor Bainbridge who initiated the work and has helped
me throughout with his advice. I would also especially thank Mr. R. J.
Willan for arranging for the special diet and for organising the collection
of the bile and applying the collecting apparatus. Finally, Iam indebted
to the nurses and dressers who willingly undertook the task of keeping
the necessary log.

I. Loc: cit:

2. Recent Advances in the Physiology of Digestion, 1906.
3. Sohafer’s Physiology, Vol. I.
4 and 5. Quoted in Schafer’s Physiology, Vol. I, by Noel Paton.

219

NOTES ON SOME NEW _ SUBSTITUTES FOR’ THE
GALENICAL PREPARATIONS OF DIGITALIS

By DOUGLAS COW, M.D.
From the Pharmacological Laboratory, Cambridge
(Received March 25th, 1912)

The experiments, which form the basis of this paper, were undertaken
with the object of comparing the actions of Digalen*, Digipuratumt and
Digitalone}, all recently put on the market as substitutes for Digitalis
preparations in therapeutics.

Digalen is an aqueous solution containing 25 per cent. of glycerine:
it was first prepared by Cloetta?, and is an amorphous powder, soluble in
water. Its exact chemical composition appears to be somewhat doubtful.
Dixon? states that it bears a certain chemical and physiological resem-
blance to Digitoxin; and he points out that it is not identical with the
product of Schmiedeberg and Killiani, which is crystalline and insoluble
in water. Killiani!® believes that it consists chiefly of Digitaléin.
Cloetta states that, whilst it is quite as good a therapeutic agent as
Digitoxin, it is less toxic. The latter statement is confirmed by Neavel',
Hale’, Hatcher’, Symes! and Moore !2. There is great variation in the
reports on the clinical value of this product.

It has been stated that Digalen is more rapid in its action than the
Galenical preparations of Digitalis, though the results of Frankel®,
Miiller and others show that it is not absorbed any more rapidly than
Digitalis. Digalen, whilst it is probably less irritant to the stomach
than Digitoxin, appears to be a violent irritant when used subcutaneously,
and not devoid of irritant power when injected intramuscularly (Dixon*).

Digipuratum is prepared by treating Digitalis leaves with alcohol
and ether, the preparation being in the form of a powder, which is
insoluble in water, though soluble in dilute alkaline solutions. — It is
said to contain about 95 per cent. of the active principles of Digitalis
leaves, though 85 per cent. of the saponins and inert matter are excluded
(Gottlieb)®. Schiittler!> states that Digipuratum is about 50 per cent. less
toxic than Digitalis leaves: it is also said to be free from irritant effect
on the stomach.

* Digalen, Hoffmann La Roche.

7 Digipuratum, Knoll.
{ Digitalone, Parke Davis & Co.

Norr.—The tincture of digitalis used as a control was the product of Ferris & Co. physio-
logically standardised.

220 BIO-CHEMICAL JOURNAL

Hopffner? reports favourably on its action: Miller!’ states that
gastric disturbance is rarely caused by it, though Szinnyei’’ records
nausea and vomiting as following its use.

Digitalone is an alcohol-free solution containing 06 per cent. of
chloretone, and is said to contain all the active principles of Digitalis
leaves. It is said to be less irritant than Digitoxin or Digitalin, though
subcutaneous injections are by no means free from pain.

My experiments divide themselves naturally into two parts, the
first part consisting of a series of perfusion experiments on the isolated
mammalian heart, in an endeavour to ascertain the relative stimulating
powers and toxicities of these preparations, whilst the second part is
confined to an attempt to establish their relative rates of absorption from
the intestinal tract and their irritant properties.

PERFUSION [7XPERIMENTS

Kquivalent doses of these three preparations were taken, aud of these,
solutions of 1 in 2,500 were prepared as required in freshly oxygenated
Ringer’s solution made up with special glass-distilled water. Rabbits
of medium size (averaging 18 kilograms) were used in all these experi-
ments: these were decerebrated, their hearts immediately excised and
perfused by the method of Langendorft, using the apparatus described
by Brodie. By this means the hearts were perfused through the coronary
vessels, and maintained at a steady temperature of 37°C. A silk thread
was attached by an entomological hook to the apex of the heart, and
connected with a suitably placed counterbalanced recording lever, the
writing point of which recorded the cardiac movements on a revolving
drum. The perfusion fluid was placed in two glass jars at a suitable
height above the suspended heart, one jar containing Ringer’s solution
and the other a 1 in 2,500 solution of the drug in Ringer’s solution. By
means of a Y-tube junction-piece either solution could be made to supply
the perfusion fluid at will.

In all, twenty experiments were performed, the results of which may

be given under six heads :—

1. Alteration in rate of heart-beat.

2. Alteration in amplitude of heart-beat.
3 Losceruy.

4. Alteration in rate of coronary flow.

5. Alteration in height of mean tonus.

6. Ineidental remarks.

GALENICAL PREPARATIONS OF DIGITALIS 421

1. Alteration in rate of heart-beat.

In this particular Digipuratum had by far the strongest action, the
average alteration being a slowing of 33 per cent. in 30 minutes. In one
experiment the beat was slowed from 90 to 48 per minute, nearly 50 per
cent. Between Digalen and Digitalone there was but little difference,
both being far less potent than Digipuratum: Digalen produced a slowing
of 76 per cent. and Digitalone of 62 per cent. The greatest degree of
slowing produced by these drugs in any experiment was from 81 to 69
per minute by Digalen, and from 96 to 82 by Digitalone. These results

are shown graphically in Fig. 1.

A.—Digipuratum
1. Normal
2. After drug

B.—Tinct. Digitalis
1. Normal
2. After drug

Fie. 1. Rate of Heart
Beat showing per-

centage slowing pro- C.—Digalen

1. Normal
2. After drug

duced.

po +—}-
|

D.—Digitalone

1. Normal
2. After drug

2. Alteration in amplitude of heart-beat.

Digipuratum again showed the most powerful action, very closely
followed in this case by Digalen; the average figures being an increase
of 97 per cent. for Digipuratam, 90°9 per cent. for Digalen and 22°7 per
cent. for Digitalone; but, whereas the Digipuratum effect was produced
in 10 minutes, that of Digitalone did not reach its maximum under
30 minutes. Digitalone also required 30 minutes for the production of
its maximum effect. Corresponding experiments with Tincture of
Digitalis showed an increase in amplitude of 88°9 per cent., produced in
30 minutes. It should be stated, however, that the increase in amplitude
produced by Digipuratum is not so well maintained as is that produced
by Digalen and Digitalone, as shown in Fig. 2.

999 BIO-CHEMICAL JOURNAL

3. Toxicity.

The relative toxicities of these preparations were measured by the
times elapsing between the moment at which the drugs were first put into
the perfusion fluid, and the time at which the heart finally ceased beating
in systole. The average results are as follows :—

Digitalone ba ele sf 260 minutes.
Digalen ... = mS: ae 235 minutes.
Tincture of Digitalis ... ce 165 minutes.
Digipuratum .... oak ne: 160 minutes.

Their ratios of toxicity may then be put at: Digitalone 0°63,
Digalen 0°70, Tincture of Digitalis 1:00 and Digipuratum 1:03. These
results are shown graphically in Fig. 3.

Professor W. E. Dixon very kindly undertook to perform a series of
injection experiments on frogs, and by his courtesy I am able to publish
the results of these in confirmation of the above toxicity figures. The
frogs were injected with varying doses diluted with an equal quantity of
saline solution, the injections being made with a fine hypodermic needle
into the dorsal lymph-sac.

Briefly stated, these four drugs showed exactly the same order of
toxicity as in the perfusion experiments: 2°5 minims of Digipuratum
killed a frog of 20 grams in 1} hours, a frog of the same weight was killed
by 2°6 minims of Tincture of Digitalis in 3 hours, whilst 45 minims of
Digalen were necessary to kill a 20 grams frog in 3 hours, and 5 minims
of Digitalone failed to kill a frog of 22 grams in 3 hours.

If we compare the ratios of toxicity given above with similar figures
obtained from the results of alteration in amplitude of heart-beat, in each
case putting Tincture of Digitalis at 1:00, we obtain the following table :—

Ratio of increase Ratio of toxicity
in amplitude
Tincture of digitalis... 506 560 Soe 1-00 506 1-00
Digipuratum .... ane gia sal a6 1-09 x 1-03
Digalen ae wee 50h 8 ob 1-02 0c 0-70
Digitalone as sop Ace Le aS: 0-25 ae 0-63

From these figures it is evident that, taking Tincture of Digitalis as
a standard, both Digipuratum and Digalen are pharmacologically
relatively more potent than toxic, whilst Digitalone is both less potent,
pharmacologically, and also less toxic than the Tincture; but, whereas
its toxicity is 387 per cent. less than that of Tincture of Digitalis, its
pharmacological activity, as judged by its power of increasing the
amplitude of the heart-beat, is as much as 75 per cent. less.

‘ep
>
&
St
Z
—_
©
e—
ce
ro
rd
=
rg
>
8
>
>
—
2
—

TP
©
=
—)
=)
ep,
_
a}
>
=
pf,
TR
bo
te

co

50

% Variations
a

28
Epetaers
ale

CCC ted et Digalen
BASRA.
ph || emmee ‘Tincture of

Digitalis

eee ishsl 2 Digittleng

Fie.3. Relative Toxicities to
the Isolated Heart.

Time elapsing before death in minutes

——— Digipuratum

60

A.—Digitalone
B.—Digalen
C.—Tinct. Digitalis

D.—Digipuratum

224 BIO-CHEMICAL JOURNAL

4. Alteration in Rate of Coronary Flow.

Digipuratum has a far more powerful action than the other
preparations, the rate of coronary flow sinking from 5 c.c in 60 seconds to
5 ¢.c. in 201 seconds. Digalen comes next with a fall of from 6 c.c. in
5 seconds to 5 c.c. in 90 seconds. Digitalone produced a fall of from
c.c. in 42 seconds to 5 c.c. in 65 seconds. This gives the following table

of percentage diminution in rate of coronary flow :—

Digipuratum Bee cis aes act) 2330 %
Digalen.... 500 909 ec 500 63-6 %
Digitalone ... 30 508 30 on 54:8 %
Tincture of digitalis adh a. 900 40-0 %

The results are shown graphically in Fig. 4. It will be noticed that
Digipuratum produces its maximum effect in 10 minutes, whilst that of
Digalen is not reached till 45 minutes after the drug was first admitted

to the perfusing fluid.

= eAelutol | AR Rave eRoee se eoee
Te ——>S= amuse )igitalis
ECOCCONS ACCC ieee eee) oom a
EXER HeS WRRP2. ARERR Sao USS eee pc 8S
EES SSRSBA. > 4ER2 SERRE See! 2UhOsSleoee

SIRF ARGER rt itt rr tH Bee eye: Digalen
HS Ff Hm Ne ol eel a om a

He (ea Sap fs Fs TST a ee ai

eet ttt PE |

Ss Wa oN Ta TANT ATP [a a

Eb? hn Uo eee ee REEL

Babb eee eR ALESLeeee

Bt -|-44 nN HE tt BE

ERR BeBe HEBER SERRE Eee

nae eee BREESE REE e eR DARE

Siar FH ORs Pee Ee PeEReR Oe eekseeAeo

BP Be eee eeeeSseeconmes Se eEee

PECEEEC CREE ere &
ERS RP RESTR RRR ARERR CELL ANS —— Digipuratum
ERPS RRReE CEP Cees Seo seesaw

SEF ARRES WARS 4B RiRle_EeBekeensaes

ef Meet Mee [os a [Ta 7 a a a a a

FFT ao oa a a me |p a a

Sah TNE a aA a an a ef 8 a i
eee EEE

B81 ee eee eee Le Re eeeseescoo

Dad ee ef ed a P cafay a Pa aaa a as a

Fia. 4. Variation in Rate of Coronary Flow in the Isolated Heart.

5. Variation in Height of Mean Tonus.

These results were obtained by measuring the height of the middle
point of the curve above the base-line at certain times. It was found
that, starting from an average height of 23 mm., Digalen produced an
increase in tonus to 28 mm. in 10 minutes, after which followed a gradual

|

GALENICAL PREPARATIONS OF DIGITALIS 925
fall. Digipuratum, from an initial average height of 49 mm. above the
base-line, caused a rise to a maximum of 53 mm., 60 minutes after the
introduction of the drug. Digitalone and the Tincture of Digitalis

produced no increase in mean tonus until the final rise of approaching
death in systole. Reducing these results to percentages we can express
them graphically, as in Fig. 5.

| SRR L dee ps bods aes | os tee
BN222748U 20884 ee) eee 2 ee
| | jj SERRE REE eRe
_ JD DSSS SS Be aes ee Se eae Ses
Peet beep he eee Pe pete
__ SEER RE SERRE ASR Ree eee
| ASR ERRERE EE Eee eee
eerste tie tebe ee Tt POPE PSE |
ARERR PRR
isn azy seep Pe eee ees ee speed Fe eee P|
ARRESTED RA eee
1 eae ae

& | he a

je eRe Ae Digalen

J ISDE aa eee as Digipuratum
Raa 2A Raseee

pepe S aa
Dot Et ete tL SS Digitalone

S Seeeee._ “2 eeexmes ‘Tincture ol
(SSS 5p. Saay dee eee ees ipeitalis
Recep eee parse

Fie. 5. Variation in Height of Mean Tonus.

Reading these results in conjunction with the results of variation in
total amplitude of heart-beat, it is noted that Digipuratum in the first

10 minutes produces an increase of nearly 100 per cent. in amplitude
without altering the tonus, so that when the amplitude of the beat is
greatest the tonus is unaltered. At the end of an hour after the

introduction of this drug, when the total amplitude of the beat is nearly
15 per cent. less than the normal, the mean tonus is nearly 10 per cent.

226 BIO-CHEMICAL JOURNAL

higher than the initial figure. In the case of Digipuratum the heart
has by this time commenced its final rise of tonus towards its approaching
death in extreme systole. Digalen, on the other hand, shows two almost
parallel curves, in that, 20 minutes after the introduction of the drug,
the amplitude of heart-beat has increased by 80 per cent., and the mean
tonus has risen 20 percent. At this time each curve often shows a slight
fall.

Thus, in the therapeutic stage, Digipuratum increases the amplitude
of the beat without increase of tonus, whilst Digalen produces an increase

in tonus coincident with the increase in amplitude of heart-beat.

6. Incidental Remarks.

Under this heading I have grouped such occurrences as variation in
rhythm, partial heart-block and extra-systole, merely wishing to direct
attention thereto, owing to the prominence given to such features of
cardiac action by Mackenzie!! and others.

Variation in rhythm was noticed once only, and that in a heart under
the influence of Tincture of Digitalis, the change being toa 3 : 1 rhythm,
and occurring 105 minutes after the commencement of perfusion with that
drug.

Partial heart-block was produced once by Digipuratum, once by
Digalen, twice by Digitalone, and not at all in this series of experiments

by Tincture of Digitalis. (Fig. 6.)

nn eae bee EN INEDN. ae c X

Fic.6. Shows partial Heart Block Fic. 7. Shows extra-Systole produced
produced by Digalen 190 minutes by Digipuratum 30 minutes after
after introduction of drug. introduction of drug.

Extra-systole was noticed but once, Digipuratum being the drug
which produced it: it occurred 30 minutes after the introduction of the
drug into the perfusion fluid. (Fig. 7.)

Figs, 8, 9, 10, 11 show typical tracings of the action of Digalen,
Digipuratum, Digitalone and Tincture of Digitalis, respectively, when

perfused through the isolated heart.

GALENICAL PREPARATIONS OF

DIGITALIS 22

Ve)
7
ro |
2 ro
on =
. 3
oe | <=
+
di
qi
e)
mn
L
=,
—
~ uM
=) D
et Ry
a q
ao
2
°
ad
=
a
S
a uo]
: 7)
— —
——-
=
g
=|
te)
<}
oc
n
x
>) E
oD
5 =
rt em)
=
Et
Jo)
Oo
—
=) a)
ri gq -
—
a NS
ro}
Some
aime)
ri wm
ri
=) a ©
re) asm
nN Te
= s
wets
= srt
oH
.
=
=
2 S
te C2)
fo | F
2 Dn
re ig
ate
7
q
‘aut
_
=) a
— =
fos =
aa
zc
D
3
G
° oF
iD
= Sal
_— od
SS 3
ned
Hy
nm
~
= 5
= rom
aS "a
re a2
cD
en) <
4 >
—

11.0

BIO-CHEMICAL JOURNAL

los}

CTT

“Spuooes Ul SOUL],
‘pny uorsnzrzed oy} OFUT poyyIUIpe sem WANQeINdIsIq QOG‘S UIT ‘OTT IV

cG’Ol CECl GS GI OLGTI cG'IT

‘UOI4NIOS S,Josury IA pesnjrzod qavay{ 8,y1qqry

OF TT GO IL eye CTT

"6 “Dlg

GALENICAL PREPARATIONS OF DIGITALIS 229

6.0

ret
S
—
wD
ae =
7. Oo
Von op
a
ar
hi
for
a oat
1D no
ne
2
1D a
er) n
wD ss
>
=
=,
i)
OL
= B=
en! fm
~
uD
N
oom Nn
so 4
wD ee toe
sH 1d
= io
es ne
~ az
<q
a

4.10

3.40

Rabbit’s Heart perfused with Ringer’s Solution.

ID
an
=
= ~
cy
Ve
Ve

JOURNAL

BIO-CHEMICAL

230

*spuooes UI OUT],

“pring uorsnysed oy} OFUT paygTUIpe Sem STTRIIBIC, JO emMgouLL OOS UL T “CS LT IV

OLG Gs 0% GPT CTT cGGL Se'GL

i

“UOTINOY 8 esUIYyY WM posnjsod yavoH] 8.41qQqQVy

OGGI Sl Og TT OF TT

TT “SIH

ae

p*

GALENICAL PREPARATIONS OF DIGITALIS 231
Rate OF ABSORPTION FROM THE ALIMENTARY CANAL

Cats of medium size were used in these experiments (weighing circa
2 kilo.). They were anaesthetised with ether, followed by urethane
intraperitoneally; tracheotomy was performed and a respiration tube tied
in the trachea. The animals were then decerebrated, and the carotid
artery of one side put into connection with a mercurial manometer. The
abdomen was then opened in the middle line, and the duodenum brought
forward in readiness to receive the drug. When the blood-pressure had
reached a constant level—usually in about 15 minutes—the drug was
injected into the lumen of the duodenum by means of a fine hypodermic
needle; the bowel was immediately replaced in the abdomen, which was
stitched up. Doses equivalent to 16 c.c of the Tincture of Digitalis were
used in all cases, and diluted up to a total volume of 20 c.c

The average results of these experiments may be given with advantage

in tabular form :—

Tincture of
Digipuratum Digalen Digitalone Digitalis

Variation in blood Slight rise after Slight rise after No appreciable No appreciable

pressure 70 minutes 50 minutes rise in 5 hours rise
Condition of heart Still beating after Death 2 hours 20 Still beating Death 23 hours
at end of 64 hours minutes after 5 hours after after introduc-
experiments introduction of introduction of tion of drug
drug drug

Condition of duo- Very slight injec- Slight injection No sign of injec- Considerable

denal mucous tion tion inflammation
membrane

Evidence of Yes Yes No Yes
absorption

In order to obtain further information as to the irritant properties of
these drugs a further experiment was performed: a rabbit, as large as
could be obtained, was anaesthetised with ether, followed by urethane
intraperitoneally ; the abdominal wall was then cleaned of fur by means
of a depilatory, and washed. Four skin areas was then mapped out,
and a hypodermic injection of one drug was made with due aseptic
precautions into each area. Doses equivalent to 1 minim of the Tincture
of Digitalis were used. The animal was then wrapped up and kept on a
warm stage and examined from time to time.

232 BIO-CHEMICAL JOURNAL

Little or no change was noticed in the condition of the skin around
the areas of injection for more than 4 hours—after which time the
following changes were noted : —

5 hours after 63 hours after
injection injection

Digalen ... Sa ae ... Distinct redness Oedema, definite

inflammation,

hot
Digipuratum .... fee ... Slight injection As before
Digitalone sas ae ... No change No change
Tincture of digitalis... .... Discoloration and As before

slight injection

CONCLUSIONS

In comparing these three preparations with the Tincture of Digitalis,
Digitalone may be at once discarded as a useful therapeutic agent, since
its cardiac action is in every respect inferior to that of the Tincture of
Digitalis; and when this drug was administered by way of the alimentary
canal no evidence was obtained that it exerted any physiological action
on the circulatory system.

Digalen has the advantage over the Tincture of Digitalis that it 1s
less toxic when directly perfused through the heart, though, when it is
administered by way of the alimentary canal, this difference is not
evident: it increases the amplitude of the heart-beat to about the same
extent as the Tincture of Digitalis. It is, however, inferior to the
Tincture, in that it does not slow the beat so much; indeed this effect
on the vagus is extremely small. The mean tonus of the heart is raised,
producing a condition which involves a large expenditure of energy
(Barcroft and Dixon!) it may be of a useless kind. It is conceivable
that, in certain conditions such as acute dilatation of the heart, this might
be advantageous, but if the mean tonus of the heart is normal it could
hardly be other than a harmful effect. Injected subcutaneously, Digalen
produces more irritation than any other preparation. Moreover, Digalen
has not the merit of cheapness.

Digipuratum is, if anything, more toxic than Tincture of Digitalis,
when perfused through the isolated heart; but it has the disadvantage of

GALENICAL PREPARATIONS OF DIGITALIS 2933

very markedly diminishing the rate of coronary flow. It slows the heart
rather more than does Tincture of Digitalis: between it and the Tincture
there is but little to choose in the matter of increase in amplitude. — It
acts quite well when administered by the alimentary canal; and is much
less irritant than Digalen when injected subcutaneously, though quite as
irritant, when given in this way, as the Tincture.

All things considered, it appears that not one of these new preparations
has made a successful bid for superiority to the Tincture of Digitalis.

It gives me great pleasure to record my thanks to Professor W. KE.
Dixon for his continued help and advice.

BIBLIOGRAPHY

1. Bareroft and Dixon, Journ. of Physiol., XX XV, pp. 3, 182. 1907
2. Cloetta, Mzinch. med. Woch., LI, p. 1446, 1904.

3. Dixon, Quart. Journ. of Medicine, V, pp. 18, 297, 1912.

4. Dixon, loc. cit.

5. Frankel, Arch. f. Exp. Path. u. Pharm., Leipzig, LVII, p. 131, 1907.
6. Gottlieb, Mzinch. med. Woch., LVIII, p. 10, 1911.

7. Hale, U.S. Hygien. Lab. Bull., p. 74, 1911.

8 Hatcher, Amer. Journ. Pharm., LX XXIII, p. 360, 1910.

9. Hodpfiner, Mzinch. med. Woch., LV, p. 1774, 1908.

10. Killiani, Minch. med. Woch., LIV, pp. 886, 1112, 1907.

11. Mackenzie, Heart, II, p. 4, Aug., 1911.

12. Moore, Brit. Med. Journ., p. 60, Jan. 13th, 1912.
13. Miller, Minch. med. Woch., LV, p. 2651, 1908.

14. Neave, Scot. Med. and Surg. Journ., Edinb., XX, p. 390, 1907.
15. Schiittler, Dissertation, Berlin, 1909.

16. Symes, Brit. Med. Journ., I, p. 1346, 1911.

17. Szinnyei, Therap. Monatsh., XXIV, p. 427, 1910.

234

THE ANTI-NEURITIC BASES OF VEGETABLE ORIGIN IN
RELATIONSHIP TO BERI-BERI, WITH A METHOD OF
ISOLATION OF TORULIN THE ANTI-NEURITIC BASE
OF YEAST

By E. S. EDIE, W.-H. EVANS, B. MOORE, G. C.-E. SIMPSON,
anp A. WEBSTER.

From the laboratories of Bio-Chemistry and Tropical Medicine of the
University of Liverpool

‘(Received March 29th, 1912)

In a recent paper from these laboratories an account was given of the
chief investigations into the causation of Beri-beri up to 1910; this
account was largely an abstract of the masterly monograph in which
Schaumann? described the work of other investigators, added much
important work of his own, and brought evidence to suggest that other
diseases might have a similar etiology to Beri-ber1.

Some of our earlier results were briefly mentioned in that paper, and
the account of our further researches will be prefaced by a brief summary
of some of the numerous papers on this subject which have appeared in
the interval.

One of the most interesting papers gives an account of the later
researches of Fraser and Stanton? who have already contributed so largely
to the knowledge of this subject.

They demonstrate that the active anti-neuritic substance of rice meal
is soluble in water and in alcohol, show that it is stable in acid, unstable
in alkaline solution, and that its thermolability varies with varying
physical factors. They show that it is not a phytin or a fat, and
that it probably is not a protein nor does it contain phosphorus.

They further confirm, however, the fact that the phosphorus content
of a rice is an indicator of its safety as an article of diet, and, with a view
to the prevention of Beri-beri, recommend administrative measures to
prevent the sale in the Malay States of rice with less than 0°4 per cent.
phosphorus pentoxide.

Eijkman‘* recapitulates his earlier work on the subject, and all must
regret that he and his colleagues were prevented from continuing their
work in Batavia and Java where, in the years 1889 to 1897, they had so far
advanced the study of Beri-beri and Polyneuritis on experimental lines,

THE ANTI-NEURITIC BASES 235

and had demonstrated that Beri-beri could be cured and its occurrence
prevented by the use of hand-milled rice.

His further researches tovisolate the active substance from rice meal
are recorded. Cure of polyneuritis in fowls was effected by two or three
doses of extracts from rice meal; less than five grams of extract containing
0085 per cent. P,O. and 0°012 per cent. N, sufficed to restore to activity
birds severely lamed. He strongly opposes Schaumann’s theory (1910)
that the active substance is a phosphorus-containing compound, though
awaiting fuller accounts of his later communications®.

Shiga and Kusama® report extensive investigations disproving the
bacterial and toxic theories of Beri-beri, and confirm,much of Eijkman’s,
and Fraser and Stanton’s work.

Kilbourne’ shows that the potassium content of rice meal is of almost
equal value to the phosphorus content as an indicator of its safety.

Chamberlain and Vedder’, following Fraser and Stanton, showed
that an extract of rice meal in 70 per cent. alcohol, concentrated at a low
temperature till alcohol free, maintains its activity to cure polyneuritis.
They further showed that the active substance is able to dialyze through
parchment. A daily dose of these extracts, containing 0°16 mg. PO, and
4°06 mg. Nitrogen, cured, in a few days, fowls severely lamed.

In another contribution? they confirm these results, having kept fowls
for 100 days on polished rice and the extract without development of
neuritis. They find that the sucrose and ash in the extract are inactive,
and so exclude 0°91 of the 1°34 per cent. solids in the extract.

In addition, they find that the active substance is absorbed by bone-
black, and are now proceeding to attempt to isolate and analyse it.

Funk" has isolated from rice meal a crystalline nitrate of an organic
base which is extremely active in reviving pigeons with polyneuritis from
feeding on polished rice.

The necessary dose contains about 4 mgs. of Nitrogen, corresponding
to 0°05 gram. of the nitrate of the base, to which he allots the provisional
formula C,,H,,0,N(HNQ,).

The crystals were in the form of microscopic needles, melting at
233° C., insoluble in cold water or alcohol, soluble, with difficulty, in
hot water. They were free from ash and from chlorine and sulphuric
acid.

This is the first record of the isolation and analysis of an active substance, and the method
by which it was obtained will be briefly described. One and a half kilograms of rice
meal was extracted with four litres of acid alcohol; separation of the filtrate was
completed by the hydraulic press ; about three and a half litres of extract were obtained, and
evaporated in vacuo at 30°, leaving a fat-like residue. This was melted and treated with water,

236 BIO-CHEMICAL JOURNAL

and filtered while warm. The aqueous part was treated with ether to remove all fatty
substances; it cured pigeons in doses corresponding to 20 grams of the original polishings.

The total aqueous extracts from 54 kilos rice meal amounted to 17 litres, which was treated
with sulphuric acid and phosphotungstic acid throwing down 900 grams of precipitate. The
precipitate was dried, washed with 5 per cent. H,SO4, ground with baryta and shaken three
hours with water. he precipitate was filtered off, the filtrate smelt of ammonia and methyl-
amine. ‘The baryta was precipitated with sulphuric acid, and the filtrate neutralised with
hydrochloric acid, and evaporated in vacuo at room temperature. The residue was extracted
with alcohol, and the alcoholic solution was active in doses = 40 grams of rice polishings. The
solution was free from proteins, phosphorus, and carbohydrates.

The alcoholic solution gave a crystalline precipitate with mercuric chloride, which was
- separated, washed, and recrystallised from water; this consisted mostly of cholin, but some
active substance was also present. Active substance was present in both the alcoholic and
aqueous filtrates.

Aqueous filtrate. The mercury was removed, and the filtrate evaporated and taken up
in alcohol was treated with platinic chloride to remove cholin. After removing the platinum
from the filtrate it was treated with phosphotungstic acid, giving a crystalline precipitate which
yielded an active substance when freed from phosphotungstate with baryta and carbon
dioxide. ,

Alcoholic filtrate evaporated and dissolved in water: mercury removed by sulphuretted
hydrogen; chlorine, &c., were removed by successive treatment with silver sulphate,
sulphuretted hydrogen, and baryta. The alkaline solution was precipitated with silver nitrate
and baryta, the precipitate decomposed with sulphuretted hydrogen and freed from silver
and barium. It proved active, and, after evaporation in vacuo, crystals were with difficulty
obtained from alcohol, with the composition, &c., given above.

Bro-CHEMISTRY oF ExtTrRAcTS oF Rick MEAL anp YEAST

Among other points, we find that nearly twice as much of the
phosphorus of rice meal goes into solution in water after denaturisation
at 120°C. Of the soluble phosphorus of rice meal nearly five-sixths
dialyses: of the soluble phosphorus of denaturised rice meal only two-
thirds dialyses. The protective properties of the fractions separated were
not tried, as the quantities were insufficient for continued feeding
experiments.

More than twice as much of the phosphorus of dried yeast, after
denaturisation at 120° C., appeared as so-called phosphatide phosphorus.

The pentosan content was also investigated, and it was found that :—

100 grams of rice meal yielded 3°93 grams. of phloro-glucide.

100 grams of potato scrapings yielded 0°43 grams. of phloro-glucide.

There appeared to be no pentosan in Chamberlain-Vedder extract of
rice meal.

Aqueous and alcoholic extracts of rice meal have a marked reducing
power after hydrolysis with dilute acids: the reducing materials were
not soluble in alcohol-ether mixture.

Alcoholic (90 per cent.) extract of rice meal was found to be active
(as stated by Fraser and Stanton) in protecting birds from the onset of
neuritis, and in curing them, but concentration on the water bath rendered
the extracts inactive.

THE ANTI-NEURITIC BASES 237

The extracts were concentrated under a fan at room temperature to
small bulk till all smell of alcohol had disappeared. These extracts
preserved some activity (see Chamberlain and Vedder, loc. cit.). Four
birds, which were very weak and disabled with polyneuritis, were each
given the extract from 25 grams of rice meal, daily: in the first week they
showed a decided improvement; became more active; three gained
5 per cent., 15 percent. and 8 per cent. in weight, respectively, in one week
(loss previous to treatment, 32 per cent., 30 per cent. and 36 per cent.).
The other, though it became more active at first, lost a further 2 per cent.
in 7 days, and 14 per cent. in ten days, being then 47 per cent. below its
original weight. It died on the 15th day of feeding, having lost a further
7 per cent.

Of the other three birds, two fell in weight (5 per cent. and 2 per
cent.) between the 7th and 10th days, the other one gained a further
2 per cent: two fell again in weight, one losing a further 5 per cent.,
the other falling 5 per cent. more to nearly its weight at the commencement
of treatment. The last just maintained its increase. All were now
extremely weak again and could scarcely survive more than a day: they
were now put on yeast, and rapidly improved; walking and flying well
in a few days, and gaining, respectively, 9°6 and 6 per cent. in the week.

Attempts were made: to precipitate the active principle from these
extracts by the lead acetate method, which will be described in more
detail in dealing with yeast.

Neither the normal or basic lead acetate precipitates proved active :
the filtrate, however, was active: the precipitate from this by
phosphotungstic acid did not, however, prove active.

Funk’s method was now tried directly on the original extract of the
meal. A strong odour of ammonia and methylamine was noticed in the
treatment with baryta. In spite of continued treatment with this extract,
however, the birds died in a few days.

We considered that other foodstuffs might give more favourable
results, and so, for the present, abandoned the investigation of rice meal.

On account of the resemblance of the active substance in solubility,
etc., to some of the peculiar lecithins and bases described by Winterstein
in wheat meal, we tried the lead acetate precipitable portions of Katjang
beans, but found them inactive. Feeding with fresh brain also failed to
preserve birds from death when incapacitated with neuritis.

An attempt to isolate the active substance from Katjang beans proved

238 BIO-CHEMICAL JOURNAL

unsuccessful; neither the lead acetate filtrate or precipitate proving active.
Further experiments will be made.

Natural yeast had been previously found to possess marked preventive
and curative properties, and extracts from yeast were next investigated.

INVESTIGATION OF ANTI-NEURITIC POWERS OF EXTRACTS OF YEAST

1.—Cold Alcohol. Extract of yeast with 90 per cent. alcohol in the
cold takes out the active substances.

The residue had lost its power to cure neuritis.

The extract rapidly revived birds suffering from neuritis, both of the
convulsive and lame types. The weights improved markedly under
treatment, but occasionally relapsed on continued treatment.

Bird a2.—Loss of weight 25 per cent., and marked lameness: 1 c.c. extract (6 grams yeast) given
daily. “Bird able to walk and fly next day. Improved in weight 5 per cent. in 7
days, but then began to lose again. Lived 4 weeks on this dose of extract, and
re-developed lameness slowly on cessation of the extract.

Bird d5.—Lameness, loss 32 per cent. Improved 6 per cent. in weight in 3 days ; 11 per cent.
in 21 days on 1 ¢.c. extract daily.

Bird el.—Lameness, moribund, loss 23 per cent. Improved 4 per cent. in weight in 3 days;
14 per cent. in 14 days on 1 ¢.c. extract daily.

Bird /2.—33 per cent. loss, severe convulsions. Active in 24 hours. Gained 3 per cent. in
3 days.

The first extracts were made by standing the yeast under alcohol,
shaking at intervals. The later extracts were made by passing the
alcohol successively over fresh portions of yeast. Finally, the principle
of the ‘ Gegenstrom’ was adopted: the alcohol first passing over partly
exhausted yeast and then over successive fractions, and finally over fresh
yeast. These extracts proved much more potent than the earlier ones.

_The alcohol was removed under the fan at room temperature before
use.

All extracts were thus made alcohol free before further precipitations
or feeding experiments.

Il.—Hot Alcohol. Hot alcoholic extracts of yeast, or extracts
concentrated on the water bath, proved inactive.

II].—Acetone. The precipitate thrown down by acetone at 0°C. from
cold alcoholic extract proved inactive.

The filtrate after removal of the acetone proved active.

Bird e3.—Loss 40 per cent., lameness very marked, moribund. 2 c.c. daily acetone filtrate.
Flying and walking well in 24 hours. Gain in weight 14 per cent. in 3 days; 18 per
cent. in 10 days. Lived on this extract for 15 days, though losing a little weight in
the last 4 days, and was then put on another extract.

IV.—Platinic Chloride. Cold alcoholic yeast extract throws down a
golden yellow precipitate with platinum chloride.

od

THE ANTI-NEURITIC BASES 239

Precipitate (freed from platinum) dissolved in very dilute hydzo-
chloric acid. Proved inactive.

Filtrate (freed from platinum) seemed largely inactivated.

V.—Phosphotungstic Acid. A bulky greyish white precipitate was
obtained by treating cold alcoholic yeast extract free from alcohol with
phosphotungstic acid. This was separated and allowed to stand with
excess of baryta at 57° C., being frequently shaken. The colour changed
to bright yellow. On filtering, a bright yellow liquid was obtained, from
which the baryta was precipitated with carbon dioxide. Tremethylamine
and ammonia are liberated by the baryta.

The filtrate from the phosphotungstic acid was similarly treated.

The active substance is separated in the phosphotungstic precipitate.
The filtrate is inactive.

The solution of the phosphotungstic precipitate, prepared as above
described, was fed to several birds with neuritis.

The convulsive form of neuritis is rapidly cured by it, but the improve-
ment in lameness is not so marked, and the lameness sometimes progresses

in spite of the treatment being continued.

Bird p4.—Loss 17 per cent., severe convulsions, lame. Bird flying in 3 days, but lost 10 per cent.
more weight in a week.

Bird 03.—Loss 17 per cent., very severe convulsions. Convulsions cured. Bird became
markedly lame, and lost 3 per cent. more weight in 3 days.

Bird 01.—Loss 25 per cent., convulsions, lame. Convulsions cured. Lameness improved in
2 days. Gain 5 per cent. in 3 days.

The solution of the precipitate freed from baryta was concentrated to
a syrup under the fan: crystals separated from the syrup. A solution of
these crystals was active.

V.—The active filtrate after lead acetate precipitation was similarly
treated with phosphotungstic acid: the baryta was added in the solid
form (thoroughly ground into the precipitate) to avoid great dilution.
A large amount of amine is hberated.

The active substance was precipitated by phosphotungstic acid from
the lead acetate filtrate.

Bird 03.—20 per cent. down, very lame. Continuous treatment with extract. Lameness cured
in 7 days, but bird still very thin and weak. Gained 3 per cent. in 10 days,

VI.—Benzoylation of lead acetate filtrate.

No satisfactory benzoyl compound isolated.

VIl.—Precipitation of lead acetate soluble, phosphotungstic
precipitable, fraction.

Active substance was removed by silver nitrate and baryta, but the
amount obtained was too small to be tested.

240 BIO-CHEMICAL JOURNAL

VIII.—Ammonia throws down a scanty fine white precipitate from
yeast extract: the precipitate dissolves in dilute acid; it is inactive.
The filtrate (neutralised) remains active.

Bird i2.—Loss 33 per cent., very lame, severe convulsions; 2 c.c. ammonia filtrate daily.
Running about, no convulsions, 24 hours. Quite lively, slightly lame, 48 hours.

IX.—Lead Acetate (normal and basic).

Cold alcoholic yeast extract was precipitated with slight excess of
normal lead acetate, giving a fine white precipitate.

(i) Normal lead acetate. Precipitate was freed from lead by
sulphuretted hydrogen, and from the latter under the fan and
concentrated.

The solution of the precipitate proved inactive.

(11) Normal lead acetate. The filtrate remained active when freed
from lead and’ sulphuretted hydrogen.

This was now treated with basic lead acetate in slight excess.

A yellow precipitate in small amount was obtained.

Basic lead acetate precipitate (freed from lead and snlpharegeee
hydrogen) inactive.

Basic lead acetate filtrate (freed from lead and sulphuretted hydrogen)
active.

X.—Yeast extract—portion non-precipitable by lead acetates.

A small portion of the filtrate, after removal of lead and sulphuretted
hydrogen, was concentrated in vacuo to a thick syrup.

From this a deposit of fine feathery crystals takes place (also a few
other crystals). One decigram of the crystals was separated as well as
possible, dissolved in water, and given to a bird lame with neuritis.
The bird recovered and was able to walk and fly in forty-eight hours.
It lived for a week without further treatment, and then again become
lame, and died fourteen days after the one dose.

Metuop or [souatTion FINALLY ADOPTED

Twenty pounds of commercial fresh pressed yeast were extracted in
the cold with successive quantities of methylated spirits, using in all,
about twenty litres of spirit; the yeast was filtered through thick calico, and
the alcoholic filtrate was freed from alcohol at room temperature by means
of an electric fan. There remained about 7 litres of watery fluid, dark
yellow in colour, smelling strongly of beer, and with an intense bitter
taste.

THE ANTI-NEURITIC BASES 241

This water extract was mixed with sufficient plaster of Paris to make
it ‘set.’ The plaster matrix, after standing overnight, was ground to a
fine powder, and extracted in the shaking machine with successive small
quantities of methylated spirits made faintly acid with hydrochloric acid.
These extracts were freed from alcohol, as before, and the watery fluid
obtained, amounting to 3-4 litres, was precipitated with excess of basic
lead acetate. The lead precipitate, having previously been found to be
inactive, was discarded. The filtrate was freed from lead with
sulphuretted hydrogen, and then concentrated to a syrup in vacuo at 38°C.
This syrup was treated with absolute alcohol, and the sticky hygroscopic
yellow precipitate (creatinin, etc.) was filtered off. The alcoholic filtrate
was again freed from alcohol and then precipitated with baryta and silver
nitrate. This precipitate was decomposed with sulphuretted hydrogen,
filtered, excess of sulphuretted hydrogen removed by the fan, and then
taken to dryness in vacuo at 38°C. A small quantity of a brown, sticky,
hygroscopic mass was obtained in this way, easily soluble in cold water,
and intensely active.

A dose of 0°006 gram administered to a bird with severe convulsions
and lameness, improved the convulsions in four hours: the bird was flying
strongly in twenty hours, and the lameness disappeared in forty-eight
hours. Two further doses of 0°003 gram were given on the third and
eighth day; the bird appeared normal, and gained weight on polished rice
diet, but died on the 15th day without return of lameness or convulsions.
Other results were equally favourable, the dose (3 mgs.) corresponds to
15 grams of yeast.

The substance was further purified by treatment with alcohol: it
was insoluble in ether and acetone, and on standing yielded feathery
crystals identical with those found in Experiment X.

The ash consisted principally of barium nitrate and a small amount
of phosphate.

Pending further investigations into the exact nature of the ash and its
relationship to the organic compound, it is assumed to consist entirely of
impurity, and the composition of the residue is approximately :—

C= 40°5

H= 8:07
N= 13°32
O= 3811

100°00

242 BIO-CHEMICAL JOURNAL

This corresponds to the formula C,H,,N,O, or C,H,,NO,(HNO,).
As the action of baryta splits off trimethylamine we may assume
further the presence of this group, and write the formula as :—

N(CHs)3 y C4H7O2 5 (HNO,).

The substance isolated we propose to call Torulin, and we hope to
prepare a larger amount with a view to further investigations into its
exact composition and relationships, and also into its physiological action ;
among other questions to be determined may be mentioned :—

(i) Whether pigeons, etc., can fully maintain their weight and
activity on a diet of polished rice with the addition of small doses of
Torulin; or whether it will only prevent the onset of convulsions or
nervous changes without being able to maintain full nutrition.

(11) Whether it is active in itself or only serves as an ‘ activator’
for some other substance (cp. Schaumann).

(iii) In what state does it exist in the food stuffs? Has each food
stuff a special base of this class, these bases being interchangeable in
animal metabolism ?

(iv) What is the cause of the convulsive form of polyneuritis? Is it
an early neuritis of the labyrinth or of the cerebral or cerebellar cortex ?
How does Torulin cure it so rapidly?

(v) Why does the phosphorus content of a food stuff serve as an
indicator of its richness in antineuritic bases? Was the phosphate in
our purified product merely an accidental contamination ?

REFERENCES

1. Simpson and Edie, Annals of Tropical Medicine, Vol. V, 1911.
2. Schaumann, Archiv. f. Schiffs- und Tropen- Hygiene, Vol. XIV, Beiheft 8, 1910.

3. Fraser and Stanton, ‘ The Etiology of Beri-beri,’ Studies from Institute for Med. Research,
Malay States, No. 12; also Philippine Journal of Science, Vol. V, p. 55, 1910.

4. Eijkman, ‘ Polyneuritis Gallinarum und Beriberi,’ Archiv. f. Schiffs- und Tropen- Hygiene,
Bd. XV, p. 698, 1911.

5. Schaumann, Transactions of Society of Tropical Medicine, 1911.

6. Shiga and Kusama, Archiv. f. Schiffs- und Tropen- Hygiene, Vol. XV, 1911.

7. Kilbourne, Philippine Journal of Science, Vol. V, p. 127, 1910.

8. Chamberlain and Vedder, Philippine Journal of Science, Vol. VI, p. 251, 1911.
9. Chamberlain and Vedder, Philippine Journal of Science, Vol, VI, p. 395, 1911.
10. Cooper and Funk, Lancet, No. 4601, 1911.

ll. Funk, Journal of Physiology, XLIII, p. 395, 1911,

12. Kutscher, Zeits. f. physiol. Chem., Bd. LI, 1907.

~

243

THE CREATIN CONTENT OF MUSCLE IN MALIGNANT
DISEASE AND OTHER PATHOLOGICAL CONDITIONS

By R A. CHISOLM, M.A., D.M., M.R.C.P., Greville Rescarch Student.

From the Gordon Pathological Department, Guy's Hospital
(Received April 24th, 1912)

Since Folin! in 1904 published his adaptation of Jaffé’s picric acid
test to the quantitative estimation of creatin and creatinin, much work*
has been done on the circumstances which govern the appearance of
creatin in the urine; but, so far as I am aware, no figures have yet been
published showing the alterations that occur in the amount of creatin in
the muscles of man in various conditions as estimated by this method.
The present communication shews the results of some work undertaken to
fill this gap in our knowledge.

The method adopted was that given by Mellanby?, and is as
follows:—At the time the body was opened for the post-mortem
examination, part of the pectoral muscles, usually about 40 grams, was
removed and freed from fat and connective tissue. It was then finely
divided with scissors, and ground up in a mortar with 97 per cent. alcohol.
After standing for twenty-four hours the alcohol was strained off through
muslin, and the residue extracted four times with distilled water, the last
drops of the extract being squeezed out in a meat-press. Control experi-
ments showed that all the creatin is extracted by this process. The
extracts were then added together and evaporated to dryness on a water
bath: the dry residue was then extracted with 75 per cent. alcohol,
filtered, and made up to a definite volume, usually 200 c.c. An aliquot
part of this was then evaporated to dryness and extracted with a measured
amount of distilled water. Finally 20 c.c. of the extract was taken and
heated on a water bath for five hours with a 5 c.c. of normal HCl, the
volume being kept constant. After cooling, the acid was exactly
neutralised with normal KOH solution; and the creatin, which by this
process is entirely converted into creatinin, estimated by Folin’s method.
The figure arrived at expresses the amount of creatinin present, and the
creatin figures shewn in the tables are obtained by multiplying the
creatinin figures by 116 (Folin).

*¥For references to the literature, see Mendel and Rose, Journ. of Biol. Chem., Vol. X,
p. 213, 1911.

244 BIO-CHEMICAL JOURNAL

Mellanby and others* have shown that creatin alone is present in
muscle, to the exclusion of creatinin. In several cases the mixed alcoholic
and watery extracts, which had undergone no treatment capable of
converting creatin into creatinin, were tested for the presence of the
fatter body, which was however always absent.

At Guy’s the bodies awaiting post-mortem examination are kept in
a cold chamber at a temperature of about 40°F. In view of the
| possibility that during the period of waiting some alteration might be
produced in the creatin content of the muscles, control experiments were
undertaken with rabbit’s muscle. In two cases rabbits were killed with
chloroform, and the creatin estimated at once in the shoulder-girdle
muscles, thigh muscles, and rectus abdominis (Rabbit A), or erector spinae
(Rabbit B) of one side. The rabbit was then wrapped in a cloth and
placed in the cold chamber for twenty-four hours. At the end of that
time the creatin was estimated in the corresponding muscles of the
opposite side of the body. The results are shewn in the following table : —

TaBLE I.—Rabbit A, male, 2,500 grams

Creatin as percentage of fresh muscle.

At once After 24 hours Difference
Thigh muscles... si Ae 0-296 0-279 57%
Shoulder girdle... as ae 0-306 0-290 55%
Rectus abdominis a ses 0-253 0-217 14-:2%
Average ... sis 0-285 0-262 8:1%

Rabbit B, male, 1,270 grams.

Thigh muscles... ses ss 0-384 0-349 9-1%
Shoulder girdle... ae ie 0-305 0-204 33-2%
Erector spinea__... Re a: 0-364 0-352 33%

Averages Sg6 0-351 0-302 14:0%

The table indicates that a stay of twenty-four hours in the cold room
lessens the amount of creatin in the muscles. The number of hours
elapsing between death and removal of the muscle from the body has
therefore been noted in all cases. Further, there appears to be a
difference in the creatin content of muscles from different parts of the
same body; in all cases, therefore, the same muscle (the pectoral) has

been used.

* The first observers to make this statement were Toppelius and Pommerehne. Archiv. de
Pharm., Vol. CCX XXIV, p. 380, 1896. Their work has been confirmed by Worner, Zezts. f.
Physiol. Chemie, XXVII, p. 1, 1899, and by Grindley and Woods, Journ, of Biol, Chem., II,
p. 309, 1906-7, ;

j

THE CREATIN CONTENT OF MUSCLE 245

The Creatin Content of Normal Muscle

In order to arrive at the creatin content of normal muscle, the cases
selected were all young adults who had met with an accident, and were
either brought in dead to the hospital or died shortly after admission,
and in whom no abnormality, apart from their injuries, could be found.
As permission has to be obtained from the coroner in cases of accident
before the post-mortem can be made, the time elapsing between death and
the post-mortem examination is considerably larger than in any of the
other groups of cases. As will be seen, the results are remarkably
uniform, but rather lower than those given by Hofmann? (0°282—0°316
per cent.), who worked with Neubauer’s method. This may be partly due
to the destruction of creatin caused by the prolonged stay in the cold
room, and also in part to the imperfections of the method employed by
Hofmann. If we estimate the loss of creatin during the stay in the cold
room at 10 per cent., we arrive at 0°300 per cent. as the creatin content of
the muscles of a healthy young adult.

TasBie I].—Normal Adults

Creatin as
Sex Age Injury Hours after percentage of
death fresh muscle
1. Male te 27 Multiple fractures 32 0-257
2. Male ae 31 Fractured spine 30 0-290
3. Male i 27 Ruptured lung and spleen, 39 0-271
fractured ribs
4, Male ae 22 Multiple fractures 36 0-280
5. Male oF 35 Fractured ribs 46 0-251
6. Male ay 38 Fractured tibia and 20 0-268
fibula
Averages... 30 34 0-270

Tue CreATIN ConTENT IN PATHOLOGICAL CoNDITIONS

The pathological conditions may be separated into the following
groups :—
1. Cases of carcinoma and sarcoma.
2. Cases of acute disease of short duration.

3. Cases of subacute and chronic disease : —
(a) Accompanied by marked loss of weight.
(6) Not accompanied by marked loss of weight,

246 BIO-CHEMICAL JOURNAL

1. Cases of Carcinoma and Sarcoma

It has been shown by Mellanby in the paper already quoted that
patients suffering from malignant disease frequently excrete creatin in
their urine, and Shaffer* has never failed to find it under these conditions.
This excretion is specially marked in patients suffering from growths in
the liver. Indeed in Mellanby’s cases the amount of creatin in the urine
in such cases was greater than the amount of creatinin, and the same was
true on many occasions of a case of secondary carcinoma of the liver that
was under my own observation for some time. The cases of carcinoma
and sarcoma have, therefore, been grouped together in Table III.

TasLeE III.—Cases of Sarcoma and Carcinoma

Creatin as
Sex > Age Disease Hours after percentage of
death fresh muscle
1. Female ... 62 Carcinoma of colon 12 0-186
and liver
2. Male aes 71 Carcinoma of cheek 28 0-201
abscess of lung
3. Female ... 38 Carcinoma of pylorus 25 0-199
4. Female ... ae Sarcoma of uterus 16 0-232
5. Male oe 67 Carcinoma of stomach 24 0-315
6. Female ... 68 Carcinoma of rectum, 27 0-167
deposits in liver and
lung
7. Male ao 50 Carcinoma of gall bladder 27 0-212
and liver
8. KEemale ... 58 Carcinoma of pancreas 21 0-228
9. Male ies 65 Carcinoma of hepatic 3 0-226
duct, jaundice
10. Male aCe 52 Carcinoma of rectum 8 0-210
Averages... 57 19 0-218

In spite of the fact that the average interval between death and the
post-mortem examination is 44 per cent. less than in the normal cases,
there is a decrease in the average creatin content of the muscles of
19 per cent., and this decrease is shewn by all the ten cases, with the
exception of Case 5, where the creatin is for some undiscovered reason
abnormally high. It will be noticed that the amount of creatin is lowest
of all in Cases 1 and 6, where there were large masses of growth on the
liver; in Case 7, where the content of creatin is somewhat larger, the
growths on the liver were both smaller and fewer. The appearance of
creatin in the urine of patients suffering from malignant disease is
apparently due to a larger muscle breakdown, as suggested by Mellanby,*

* In support of this Shaffer’s results may be quoted. This observer found the largest

amount of creatin in the urine in the case of puerperal involution of the uterus, where muscle
breakdown is at its highest.

THE CREATIN CONTENT OF MUSCLE 247

following Benedict and others;5 and this muscle breakdown is most
marked in cases of carcinoma of the liver, that is to say, in cases marked
clinically by excessive loss of body weight.*

This excessive muscle breakdown may not, however, be the sole cause
of the low creatin content of the muscles in this class of case. Mellanby
has shewn that there are reasons for supposing that muscle creatin is
formed in the liver, and that one of the functions of that organ is to keep
the amount of creatin in the muscles up to a certain standard. It is,
therefore, quite possible that the low creatin content is not due merely
to excess of muscle breakdown, but is also due in part to failure on the
part of the liver to make good the usual loss. In both Cases 1 and 6 the
amount of liver substance remaining was obviously much less than
normal, and it is a point of some interest in this connection that the
lowest creatin content in Table V, which consists of cases where no wasting
was observed clinically, is found in Case 6, where much of the
parenchyma of the liver was replaced by connective tissue: diminished
creatin production as well as increased creatin loss may therefore both be
factors in the causation of the low creatin content of the muscles in cases
of hepatic disease.

2. Cases of Acute Disease of Short Duration

The cases in the following table showed no muscular wasting, and the

creatin content is that of the normal adult, and requires no comment.

TABLE IV.—Cases of Acute Disease
Creatin as

Sex Age Disease Hours after percentage of
death fresh muscle
1. Male es 10 Intus-susception. 11 0.275
Operation—general
peritonitis
2. Male nae 65 Acute obstruction by 22 0-285
band. Operation
3. Female ... 7 Intus-susception. 19 0-280
Operation—general
peritonitis
4. Male os 59 Appendicitis. Operation— 15 0-328
general peritonitis
5. Male gee 21 Lateral sinus thrombosis, 12 0-215
septicaemia
6. Male Bre 59 Acute ptomaine poisoning 13 0-240
Average... 37 15 0-270

* A possible factor in the origin of the large amount of creatin found in the urine in cases
of carcinoma of the liver is the production of creatin by the new growth itself. That this factor
is comparatively unimportant is shown by the fact that in one case the percentage of creatin
in a mass of new growth taken from the liver was only 0-014, while in two other cases the amount
could not be measured accurately, but was even less than this. Even supposing the growth
to weigh 1,000 grams, this only represents 0-14 grams of creatin, while the total creatin excretion
per diem in Mellanby’s cases averaged 1-18 grams. Cf. Saiki, J, Biol. Chem., p. 23, VII,
1909-10.

248 - BIO-CHEMICAL JOURNAL

3. Cases of Subacute and Chronic Disease
(2) Not accompanied by marked loss of body weight.
(b) Accompanied by marked loss of body weight.

TABLE V.—Subacute and Chronic Disease, without marked loss of weight

Creatin as
Sex Age Disease Hours after percentage of
death fresh muscle
1. Male ae 48 Chronic interstitial 48 0-268
nephritis, pericarditis
2. Male sige 36 Aortic disease 12 0-232
3. Male ee 56 Fibroid lung 14 0-278
4. Male oe 59 Myocardial degeneration 27 0-224
5. Female ... 76 Diabetes 50 0-257
6. Female .... 58 Gall stones, cirrhosis 9 0-202
Averages ... _ 55 27 0-244
TasBLE VI.—Subacute and Chronic Disease, with marked loss of weight
Creatin as
Sex Age Disease Hours after —_ percentage of
death fresh muscle
1. Male See 10 Appendicitis, 6 0-259
subdiaphragmatic
abscess
2. Female ... 13 Phthisis 9 0-184
3. Female... 17 Infective endocarditis, 20 0-184
pericarditis
4. Male see 31 Laceration of liver 22 0-219
Fae and kidney, general
peritonitis
5. Female ... 15 Chronic peritonitis, 17 0-203
intestinal obstruction
6. Male ws 32 Phthisis 10 0-266
7. Male a 9 Adherent pericardium 23 0-147
8. Male oes 58 Hodgkins’ disease, 24 0-260
a few nodules in liver
Averages ... 23 16 0-215

The cases in Table V showed no obvious signs of loss of weight during
life, and no mention is made of loss of weight in the reports. The state-
ment, however, rests usually on the patient’s own impression, and in no
case were the actual weights given. There may, therefore, easily have
been some slight unnoticed loss, which would account for the fact that
the creatin content is somewhat below the normal for healthy adults.
All the cases in Table VI were accompanied by loss of flesh so marked as
to attract the patient’s own attention, and here the average creatin content
is the same as that found in cases of malignant disease.

THE CREATIN CONTENT OF MUSCLE 249
SUMMARY

1. The average creatin content of the muscles of a healthy adult is

0°300 per cent. of the fresh weight.
9

2. The creatin content is reduced in cases of malignant disease,
especially in cases of growth in the liver, and there is reason to suppose
that in these latter cases the low content is due to diminished production
as well as to increased loss.

3. The content is not reduced in cases of rapidly fatal acute disease,
but is lowered in cases of subacute and chronic disease, where there has

been marked loss of body weight.

REFERENCES
1. Folin, Zeitschrift f. Physiol. Chemie, XLI, p. 223, 1904.
2. E. Mellanby, Journ. of Physiol., XXXVI, p. 447, 1908.
3. Hofmann (quoted by Voit), Zeitschrift f. Biologie, 1V, p. 82, 1868.
4. Shaffer, American Journ. of Physiol., XXIII, p. 1, 1908.

on

Benedict, Carnegie Institute of Washington, Publication No. 77, 1907.
Benedict and Myers, Amer. Journ. of Physiol., Vol. XVIII, p. 406, 1907.

250

ON THE NATURE OF ANIMAL LACTASE

By MARJORY STEPHENSON.

From the Bio-Chemical Laboratory, Institute of Physiology, University
College, London

(Received May 15th, 1912)

The rate of hydrolysis of disaccharides by enzymes has been found
to be considerably influenced by the presence of the products of
hydrolysis. The cases of invertase and emulsin by Tammann in 1892,!
maltase by Croft Hill? in 1898, and lactase by Armstrong? in 1904.
V. Henri, in 1901, showed that the retarding effect on invertase was
mainly due to the fructose, an observation which was confirmed in 1902
by Adrian Brown,® who proved that it was not due to the increased
viscosity of the solution. E. F. Armstrong found that the lactase present
in Kefir grains was retarded in its action by galactose, but that the lactase
present in emulsin was retarded by glucose. He thus showed the
existence of two kinds of lactase—a galacto-lactase, inhibited only by
galactose and a gluco-lactase inhibited only by glucose. As no data as
to the nature of animal lactase at present exist, it was of some interest
to ascertain whether the lactase in the intestines of animals was of the
type present in Kefir grains or of the type present in emulsin.

The procedure for examining for the existence of animal lactase
which was adopted in these experiments was essentially the same as that
of Plimmer.® The enzyme solution was prepared by extracting the
ground-up mucous membrane of the intestine of the dog for forty-eight
hours with water containing toluene as antiseptic, and then filtering
through cambric.

A definite volume of this extract was mixed with an equal volume of
(a) 5 per cent. lactose solution, (b) 5 per cent. lactose solution to which
glucose or fructose, in varying concentrations were added, to both
mixtures 5 per cent. toluene was then added as antiseptic.* Samples of
50 c.c. were withdrawn immediately after mixing; the remainder was
kept at 37° C.; and at intervals of one to four days, so as to give ample
time for the enzyme to act, further samples of 50 c.c. were withdrawn.
Each sample was treated with 5 c.c. mercuric nitrate solution to

* In all the experiments the same volume of extract was added to 5 per cent. solutions of.

glucose and galactose, to ascertain if there were any disappearance of reducing carbohydrate.
No disappearance was obtained in any of the experiments here recorded.

ON THE NATURE OF ANIMAL LACTASE 251

precipitate protein; 30 c.c. of the filtrate from this precipitate were
treated with 2 ¢.c. of caustic soda; after again filtering 20 c.c. of the
solution were treated with hydrogen sulphide, excess of the gas was
removed with copper sulphate, and the filtrate and washings from the
sulphides made up to 100 c.c.

The reducing carbohydrate was then estimated by Bertrand’s
method, instead of gravimetrically. This method is as accurate as the
gravimetric method and more rapid; the precipitate of cuprous oxide is
filtered off, dissolved in acid ferric sulphate solution, and the ferrous
salt titrated with permanganate. A solution of glucose was found to
contain by the gravimetric method 4°63 per cent., by Bertrand’s method
4°67 per cent.

The reducing power of completely hydrolysed lactose calculated from
Bertrand’s table is raised from 67 to 10. These figures were verified.

Hydrolysed
Lactose Solution Lactose Solution
79 c.c. KMn0, Bec. KMn0, 19 = 0-670

With the concentration of lactose employed in these experiments,
complete hydrolysis corresponds to a difference of 2°8 c.c. of perman-
ganate. Differences in the titration of two solutions are generally less
than 0°2 c.c.; a difference of 0°5 c.c. for experimental error has been
allowed for. This is equivalent to 10 per cent. hydrolysis. Less than
10 per cent. hydrolysis of the lactose is therefore neglected in the results.
The following data have been obtained; for comparison a similar
experiment was made with the lactase in emulsin.

Experiment I. ANIMAL LACTASE
No. of c.c.
Concentration of Time of action KMnO, for Difference in Percentage of
carbohydrate solution of enzyme 10 c.c. of sugar —_e.c. of KMnO,4 hydrolysis
solution
2-5 per cent. lactose 0 a 57 0 0
od )
36 hours i 6-9 1-2 43
4 days 7-5 ) 7-5 1-8 64
55 1 Naa
2-5 per cent. lactose 0 15-0 } 15.9 0 0
15-0 j
2-5 per cent. glucose 36 hours oe Lip1 0-1 0
5-0 |
4 days 15-2 ) 1. ‘
15-0 f 15-1 0-1 0

- 4
ee ee ae ¢

252 BIO-CHEMICAL JOURNAL
No. of c.c.
Concentration of Time of action KMn0O, for Difference in Percentage of
carbohydrate solution of enzyme 10 c.c. of sugar c.c. of KMnO, hydrolysis
solution
2-5 per cent. lactose 0 PB 2e yi
13-4 | 13:3 0 0
2-5 per cent. galactose - 36 hours et) se :
14-4 14-4 1-1 39
4 days 15-2
deve a L151 1-8 7
Experiment II. ANIMAL LACTASE
2-5 pe t. lactose 0 5:9
) per cent. lactose De | 5-9 0 0
26 hours en)
; 7-7 | 7 1:8 62
96 hours 8-9 |
8-9 | 8-9 3-0 84
2-5 per cent. lactose 0 14-2 ) 1 ,.
14-6 | 14-4 0 0
5A oO .
2-5 per cent. glucose 26 hours we | 15-05 0-65 29
96 hours 16-9 )
16-8 | 16-85 2-45 84
2-5 per cent. lactose 0 10-3
at 10-4 | 10°35 0 0
1:25 per cent. glucose 26h 11:5
e ors aude 11-5 1-15 40
pens oe f 12-55 2-2 76
2-5 per cent. lactose 0 13:8 s
tae 13-75 0 0
5 RCHOE 4
2-5 per cent. galactose 26 hours De 15-65 1-9 65
96 hours 16-2
ee } 16-2 2-45 84
Experiment ITI, ANIMAL LACTASE
2-5 per cent. lactose 0 6-1
| He 6-05 0 0
89 hours 8-5 |
8-5 f 8-5 2-45 84
168 8-7
naoe Sa} 86 2-55 88
2-5 per cent. lactose 0 15-0 )
& - ) 15.2 ; 15 0 0
-5 per cent. glucose 89 hours 16-0 | 1 «. ;
15-6 | 15:8 0-7 24
168 hours 16:8 ) .
16-8 | 16-8 1-7 58

aul

ON THE NATURE

Time of action
of enzyme

Concentration of
carbohydrate solution

OF ANIMAL LACTASE

No. of c.c.
KMn0O, for

10 c.c. of sugar

Difference in
c.c. of KMnO,

Percentage of
hydrolysis

solution
2-5 per cent. lactose 0 10-5 )
| 10-4 j 10-45 0 0
1-25 per cent. glucose 89 hours Gate zB
11:8 plea 1-25 43
2:5 per cent. lactose 0 Ae 13-9 0 0
2-5 per cent. galactose 89 hours a 15-8 1-9 65
168 hours 16-4 :
ae 16-3 2-4 83
Experiment IV, EMULSIN
2-5 per cent. lactose 0 58) -
5-6 | 5-7 0 0)
1 day 6-6 )
6-7 7 56 0-9 32
3 days Te
73) 73 1-6 57
11 days Ssonrae ie
8-2 J 8-2 2-5 89
2-5 per cent. lactose 0 14-8 ;
- Aa } 14:8 0 0
2-5 per cent. glucose 1 day 14-8) 1).
15-0 | 14-9 0-1 0
3 days 16-2 ; :
58 | 16:0 1-2 42
11 days 18-0 | ,-.~ ;
17-4 17-7 2-9 100
2-5 per cent. lactose 0 14:0) ,,.
1 14-0 14-0 0 0
2-5 per cent. galactose 1 day 15-0 ) 14:8 0-8 28
14-6 j
3 days 15-8) a. ‘
16-0 ) 15-9 1-9 68
11 days 17-0 ) 32. :
16-8 | 16-9 2-9 100

In all the experiments the inhibition by glucose is very marked,

whereas galactose has little or no influence.

Animal lactase appears te

be extraordinarily sensitive to the presence of glucose: in the first

experiment 2°5 per cent. glucose entirely inhibited the action, in the

second and third experiments this quantity of glucose reduced the
hydrolysis from 62 and 84 per cent. to 22 and 24 per cent.

amounts of glucose produced

inhibition.

Even smaller
In the experiment with

254 BIO-CHEMICAL JOURNAL

emulsin, glucose caused total inhibition at first, but after forty hours
hydrolysis occurred. The curve represents the hydrolysis of Experiment 3.
lt may be concluded that animal lactase is a gluco-lactase like the

lactase in emulsin.

I wish to thank Dr. Plimmer for suggesting these experiments and
for assistance in the methods employed.

A. 2°5 per cent. lactose

B. 2°5 per cent. lactose + 2°5 per cent. glucose.
C. 2-5 per cent. lactose + 1-25 per cent. glucose.
D. 2°5 per cent. lactose + 2°5 per cent. galactose.

REFERENCES

Tammann, Zeit. physiol. Chem., XVI, p. 271, 1892.

Croit Hill, Trans. Chem. Soc., LVII, p. 919, 1898.
Armstrong, Proc. Roy. Soc., LX XIII, p. 516, 1904.

V. Henri, Compt. Rend., CX XXIII, p. 891, 1901.

A. Brown, Trans. Chem. Soc., UX XXI, p. 373, 1902.
Plimmer, J. Physiol., XXXIV, p. 93; XXXV, p. 20, 1906.

See bs

SDM i

Limit of experimental error,

255
[PERCY SLADEN TRUST RESEARCH]

THE NUTRITION AND METABOLISM OF MARINE ANIMALS
IN RELATIONSHIP TO (a) DISSOLVED ORGANIC
MATTER AND (b) PARTICULATE ORGANIC MATTER OF
SEA-WATER

By BENJAMIN MOORE, M.A., D.Sc., F.R.S.; EDWARD 8S. EDIE,
M.A.; EDWARD WHITLEY, M.A.; ann W. J. DAKIN, D.Sc.

From the Marine Biological Station, Port Erin, Isle of Man, and the
Bio-Chemical Department, The University, Liverpool

(Received May 15th, 1912)

A fundamental problem in the economics of life in the ocean is that
of the manner in which different types of marine organisms take up their
food from the sea. There is no question that bacteria in the sea-water
are nourished by food taken up in solution. The distribution of such
bacteria is not uniform, they flourish in abundance near any decomposing
organic matter of either vegetable or animal origin, and, although
invariably present everywhere, they occur in much smaller numbers in
purer water some distance from shore, where decomposing organic matter
in quantity does not exist. Just as in the case of terrestrial bacteria, it
is obvious that bacteria in the sea, growing on or near dead matter from
plants or animals, throw out secretions which aid in the solution of such
organic matter, and increase in this way the amount of their available
pabulum in dissolved form. Such dissolved organic matter is carried
away in part by the motion of the sea-water, and along with similar
organic matter, such as sewage and humus, from rivers and smaller
streams, must give a certain amount of dissolved organic matter to the
general bulk of the ocean water.

The amount of such organic matter undoubtedly remains balanced
or constant, at a certain value, on account of its being broken down into
inorganic matter as it is oxidised by bacteria and other agencies.

It is hence a very probable view, prima facie, that many forms of
protozoa and other minute organisms possessing no special alimentary
system, or free movement of their cytoplasm for enveloping and
digesting such particulate food as bacteria, algae or minute débris, may
exist and grow and maintain an oxidizing metabolism on the dissolved

256 BIO-CHEMICAL JOURNAL

organic material so added to the sea, either from the land or from the
débris and excreta of larger and higher denizens of the sea.

From this point of view it is of the highest importance that our
knowledge of the amount of dissolved organic matter in the sea should
be enriched by more quantitative determinations than have hitherto been
carried out, so that some idea may be obtained of the value of the sea as

-a nutrient medium apart from the particulate organic matter in living
condition and otherwise which it contains.

Such experimentation upon the dissolved organic matter is made still
more essential for two reasons: first, there exist at present most marked
divergencies in the experimental findings of observers as to the amount
of dissolved organic matter in the sea-water, the figures obtained ranging
from quantitatively indeterminate traces of less than 1 mgm.
per litre to over 90 mgms. per litre; and, secondly, it has been
prominently urged by one investigator in recent years that not only
bacteria and protozoa, but many classes of higher invertebrates, and
even fishes, obtain the greater part of their nutrition from the dissolved
organic matter, rather than, as has hitherto been held, from organisms
forming the floating life or plankton of the sea, or from other organisms
accumulated along the shore, which form their prey.

From the bio-chemical or physiological point of view, it is quite
well known that organic matter destined to form an intrinsic part of the
body of an animal must, first of all, be broken down into smaller
molecules and dissolved. These soluble molecules are then, in the specific
metabolism of the particular animal, built up into the definite colloidal
molecules or aggregates which exist in the cell substance of that type of
animal.

Even in the simplest and most rudimentary type of digestion, such
as that shown by an amoeboid protozoon or a leucocyte, the food particle
has not become part of the animal when it is engulfed and surrounded
by the cytoplasm. It is afterwards gradually dissolved in so far as the
digestive cell-fluids can act upon it, and itis the soluble molecules only
which are then aggregated and built up in the single-celled organism
to give those specific colloidal structures which lie at the basis of the
differentiation of species from one another, and yield those delicate
differences of chemical structure which, in turn, give the morphological
substratum for functional differences.

The same is true for the metazoa, only the special types of cell
fabricate the secretions which act upon the colloidal aggregates of the

— eS

NUTRITION AND METABOLISM OF MARINE ANIMALS 257

food and break them down into soluble molecules to form the building
stones, for constructing on different lines the new colloidal aggregates
necessary for the animal’s own cells with their accompanying peculiar
functions. Accordingly, all food of protozoa and metazoa alike must
pass through a stage of hydrolysis and disintegration and solution, and
there is nothing inherently erroneous in the suggestion as a working
hypothesis that the dissolved organic matter of water in which aquatic
animals are living may play a large or even a predominating part as a
food supply.

On the other hand, in all such preliminary considerations of a
general theoretical nature, the fact that a large number of species of
protozoa possess means of taking up particulate matter such as organic
débris, diatoms and bacteria, must not be lost sight of, as also that with
the exception of certain parasitic forms, all of the higher metazoa possess
a specially developed system of cells with both mechanical and chemical
devices (e.g., intestinal canal and digestive glands) for taking up solid
food and rendering it soluble afterwards. In addition, as the scale of life
is ascended, there arise special systems for absorption of the material
prepared for the nutrition by the solvent action of the digestive secretions,
and other systems again for the intermediate metabolism which renders
the absorbed material suitable by specialized chemical syntheses for entry
into the circulating nutrient fluid which bathes or irrigates the living
cells.

This process has proceeded so far in the case of most terrestrial
animals that there remains no possibility of their nutrition in any other
way than by organic matter taken in by the alimentary canal and passed
through the processes outlined above. A great part of the existence of such
animals consists either in the pursuit of nutriment or in its ingestion and
assimilation, and it is a general law, even up to and including the human
species, that the distribution and migration of life follows the distribution
in comparative abundance or scarcity of food.

Now, in nearly all the aquatic invertebrata and in the fishes, similar
developments for the capture and elaboration of particulate food are to be
found, and the same variations in distribution in relationship to
abundance or scarcity of other plant and animal life occur, as also the
same migrations of motile animals after food. Therefore a purposeful
relationship of these factors becomes obvious, and it would appear
probable that with the more extensive development of organs designed
for the capture, digestion and metabolism of solid food, the direct

258 BIO-CHEMICAL JOURNAL

absorption of soluble organic food in unchanged condition by organs
evidently intended for other physiological purposes, such as the gills of
molluscs or fishes, would lessen in degree and become comparatively
unimportant as a source of energy, if, indeed, such an intake persisted
at all.

Organs specially developed for the performance of such an important
function as the intake of nutrition, it can safely be said, would never have

evolved to such a high degree of perfection and elaboration if they were

chly used occasionally as a luxury or an accessory to a main mode of
nutrition, in which the food was ready prepared and soluble, and capable
of entering the main stream of nutrition at once without injury to the
animal. If an abundance of dissolved nutrient matter already existed
in the sea, the saving in work by direct utilization of this supply of energy
would be so gréat that digestive and metabolic organs would not develop,
and if already developed would atrophy from disuse.

Dissolved organic matter means practically no struggle for existence,
and no reason for the development of many organs, such as the sense
organs, which have developed in connection with such a struggle.

This is clearly seen in the case of many parasites such as the cestodes,
which utilize dissolved nutriment from their hosts.

If the sea-water were a nutrient medium of so _ considerable
concentration in dissolved organic matter as to play any important part
in the nutrition of such an animal as a fish, it must not be forgotten that
the sea would undoubtedly be swarming with creatures so designed as to
be capable of flourishing without solid aliment, and, in addition, there
would be a host of other creatures making their prey of these. Life in
the ocean away from land would be far more abundant than it is at
present. If the sea-water, for example, contained 90 mgms. per
litre of organic matter of a type suitable for supporting life, instead of
being, as at present, comparatively free from bacteria and containing
only 0°83 mgm. per 1,000 litres,! it would be a most excellent culture
medium, swarming with bacteria and infusoria like the stagnant pools
containing decomposing sea-weed found at higher levels along the shore.

Finally, it may be pointed out that as the size of an organism

1, This is the moist weight of bacteria found by Lohmann working at Syracuse in the
Mediterranean, as computed by Piitter (Zeitsch. f. allgem. Physiol., Bd. VII, p. 290, 1908) in so
large a quantity as 1,000 litres of sea-water. In the 1,000 litres there were only by volume
53°63 cubic millimetres of Plankton made up as follows :—Protophytes No. 2,082,850, vol. 17
c.mm.; Protozoa 325,510, vol. 1:13c.mm.; Metazoa 17,325, vol. 34-7 c.mm.; Bacteria 785,000,000,

vol. 0'80c.mm. Allowing a specific gravity of 1,080 this gives a total moist weight of Plankton
of 55°6 milligrams.

ee eS

NUTRITION AND METABOLISM OF MARINE ANIMALS 259

increases, the surface decreases relatively to the weight, and hence a
concentration of dissolved organic matter which would be ample for the
support of a bacterium or protozoan, would be altogether useless for the
support of a fish. When solid organic food is swallowed and digested by
a fish, it is taken up in higher concentration from the intestinal tract,
and a smaller area of absorptive tract is capable of carrying on the work.
But if such a creature were compelled to derive its nutrition from a
solution of only one or two parts in a million, the necessary uptake could
only be obtained by providing a large surface, such as the lung of an
animal, or the green leaf of a plant, or a fish gill, over which the water
was very rapidly moved. No such mechanism designed for absorption
of soluble food and connected up with a metabolic gland, such as a liver,
has ever developed.

The gill of the fish has been supposed by Piitter to carry on such a
function in addition to its more obvious function of respiration, but all
experimental proof of this supposed function of the gill is lacking. The
gill epithelium is not such as to warrant any belief that it can modify
any organic matter possibly absorbed, and there is no provision anywhere
for a metabolic change of any material absorbed by the gill before it is
thrown into the general circulation and carried as nutriment to the tissue
cells. All analogy teaches us that promiscuous organic matters so
absorbed, and distributed without appropriate intermediate changes,
instead of acting as nutriment would produce the effects of a violent
poison. The products of digestion removed from the intestine and
injected into the blood stream, so that they may reach the tissue cells
without passing through the liver, act in man as exceedingly toxic agents
and cause death when administered in this way, even in small quantities.

Again, for an adequate uptake of dissolved matter from a very dilute
solution, it becomes necessary to have some substance in the circulating
fluid with a specific affinity for it. The haemocyanin of crustacean blood
and the haemoglobin of fishes is an example of such a body, related to
uptake of oxygen in dissolved form by the gills. Yet sea-water contains
8 mgms. of oxygen in solution per litre, it is hence exceedingly difficult
to believe that, without any such chemical absorbent, the blood of a fish
in process of circulation through the gill could take up anything
approaching the amount required for a nutritional balance. Especially
is this the case in view of the low limit of organic matter shown by the
present experiments as the maximum which can possibly be present in

sea-water.

260 BIO-CHEMICAL JOURNAL

The figures which have been obtained in the present research for the
rate of oxidation of organic matter by the species experimented upon,
taken in relationship with the low values for dissolved organic matter,
demonstrates clearly that the views put forward by Pitter,! that marine
animals obtain a large amount of their food in soluble form from the
sea-water must be regarded as erroneous.

On the other hand, from our own observations, as well as from a
consideration of the results of plankton studies of other observers, it
becomes evident that many marine animals could not possibly maintain
themselves in nutritive equilibrium by any process of consuming
plankton as uniformly distributed in the sea without some special local
concentration, for the organic material contained in the plankton is also
far too small for such a purpose. On the basis of the higher animal
acting even as a complete filter to the quantity of water passing through
its body, the volumes of water which musi necessarily be filtered to yield
the daily quota of food are far too large to be credible. There is no
mechanism in most species at all comparable to a Chamberland filter for
such purposes, since this filter refuses passage even to bacteria, but we find
that the combined amounts of particulate organic matter held back by (a) a
fine silk netting of 20 gauge, which is the finest used for tow-netting, and
(b) that afterwards retained by the Chamberland filter amount together
to not more than 1 mgm. per litre, that is to say, to one part in one
million of organic matter oxidizable by potassium permanganate. Ifa
calculation is made of the volume of water necessary, as shown in Section
B of this paper, to yield the amounts of organic matter experimentally
shown to be utilized by those animals with an active metabolism, it
becomes at once obvious that the uniformly distributed plankton in this
sense cannot play any considerable part in the nutrition of such animals.

It is only in the case of fixed animals with a slow metabolism, such
as sponges and ascidians, that the demand for organic nutrition can be
satisfied by the minute forms present in plankton by such a reasonable
volume as may be supposed to be passed through the animals. (See
Section B, p. 281.)

To this extent, therefore, we confirm the view of Pitter, but we
disagree with him in that we do not find an available source of supply of
adequate value in the soluble organic matter, and believe his results are

1. Die Ernahrung der Wassertiere, Zeitsch. f. allg. Physiol., Bd. VII, S. 283-320; Der

Stoffhaushalt des Meeres. Ibid. S. 321-868; Studien z. vergleichenden Physiologie des Stoffwechsels,

ss os kgl. Gesellsch. d. Wissensch. zu Gottingen, Math.-Physiol. Klasse, Neue Folge, Bd.
, 8. 1-79.

aE

ee

NUTRITION AND METABOLISM OF MARINE ANIMALS 261

due to lack of appreciation of the defects of the ordinary methods of
analysis of fresh water when applied to sea-water.

All the wealth of argument so ably applied by Piitter to prove that
the organic energy supply of the plankton is inadequate for the support
of life, apples with equal force to the dissolved organic matter, for
neither source yields more than one part per million.

But if the plankton figures, on the basis of uniform distribution, fail
to furnish an explanation, and the organic matter in solution also is
inadequate, what, it may be asked, is the food-supply? In our opinion
the answer les in the unequal distribution of the plankton, similar in
nature and cause to the unequal distribution in space of the higher
marine animals and of nearly all terrestrial animals.

The terrestrial animals have to seek their food, and, as a rule,
congregate in flocks or herds for that purpose where food is most abundant
and follow the food. Other purposes of a biological nature are also served
and fostered by the communities so formed for the primal necessity of
obtaining food. Fish in the sea live together in shoals for the same
reason, and anyone who has walked along a sea-cliff in summer must have
been struck by the fact that birds, such as sea-gulls, follow the movements
of the young fish, and do not spread themselves uniformly over the sea-
surface regardless of the concentration of the food.

Where there is a good supply of pasturage and water, under natural
conditions there will be a greater wealth of herbivorous animals, and
where these are found in abundance there will be more of the carnivorous
animals which prey upon them. Thus some factor in the environment
of a permanent physical nature in soil or water, or of seasonal variation,
such as light or heat or moisture, produces abundance or scarcity of one
species of animal or plant, and this precipitates a whole train of events,
the causal relationship of which can only be unravelled by patient study
and experimentation.

There is no prevalence of monotony and uniformity in Nature, all is
cyclical and undergoing constant variation, and so also there exists no
uniform distribution of plankton in the sea.

The postulate of a uniform distribution of plankton would appear to
be assumed by many workers on the floating hfe of the sea and its
variations in season and in space. ‘This postulate is taken as a basis by
Piitter when he deduces the inefficiency of the plankton as a tood-supply.
It underlies all his arguments that the fish, or large invertebrate, is
engaged all the time in exhausting a certain volume of water either

262 BIO-CHEMICAL JOURNAL

(a) of plankton on the views of the orthodox school of plankton workers,
or (b) of dissolved organic matter on his own views.

This is the fundamentally wrong conception in the whole argument,
because it leaves out of account anything in the nature of a taxis towards
the food in the lower species or of what we term instinct in the higher
species, and in the case of non-motile organisms any selective action of
the organism in choosing a site for its operations more favourable for the
capture of the particular kind of food upon which it lives. Yet every
observer of distribution of marine animals knows that this is precisely
what happens, and that at a given time of year he will find particular
species by dredging or fishing in particular spots.

The knowledge of the floating lfe of the sea and its variations in
season, and in locality, is most important, and although the experimental
methods available may be imperfect and may not show all variations in
the content of the water in living organisms, yet much as to sequence and
interdependence of life may be found out by such study.

It is equally important that the limitations of such study should be
appreciated, and that figures so obtained should not be applied with a
direct mathematical rigour for which there is no warrant.

Tow-nettings taken with care throughout a season may show quite
clearly the relative quantities of a certain organism available as food for
another, but such figures taken and multiplied into volumes of water to
show how many litres of water the latter organism must filter through its
body before it can make a sufficient meal from the former constitutes
a drifting into ridiculous absurdity, such as usually overtakes the biologist
when he applies mathematics in a blindfold fashion without thinking
of the many natural limitations unrepresented in his simple formula.

A herring, which swam about aimlessly with its mouth open in a
sea containing 0°5 parts in one million cf organic matter as plankton,
would have to filter a good deal of water before it obtained an adequate
supply of food, and might grow very lean in the process; another herring
in the same sea making proper use of its sensory and nervous systems might
be able to detect enormous variations in the distribution of food and grow
fat where the other was starving.

It is, accordingly, the factor of the animal’s one action which appears
to us to have been disregarded too much, and this explains how a balance
of nutrition can be obtained, in spite of the low figures for dissolved
organic matter and for plankton on the supposition of a uniform
distribution.

NUTRITION AND METABOLISM OF MARINE ANIMALS — 263

These theoretical considerations illustrate the importance of
determining such questions as the organic content in solution and
suspension of the sea-water, and of the amounts of organic matter
required by different types of animal.

In regard to the first part, viz., the organic content of the sea-water,
we believe that we have obtained definite results which place a maximum
at such a low level as to show that the organic content in solution is
almost negligible and lies within the limits of error of determination,
and also that the plankton, whether as ordinarily fished with a fine silk
netting or as separated by the Chamberland filter, is represented by a
very low figure. But in respect to the amounts of organic matter oxidized
by the species of animal with which we have worked, we are only able
to put preliminary figures forward as yet, and intend to supplement
these later by fuller study of each species. This latter portion of the
work has opened up certain new problems which will be mentioned in
describing the experiments.

The work was carried out at the Marine Biological Station, Port
Erin, Isle of Man, during April, 1912, and our thanks are due to Professor
Herdman, for the invitation to work at the problem, for the use of the
steam yacht ‘Runa’ in collecting water and specimens, as also for
many most valuable suggestions in the course of the work. The expenses
of the research have been defrayed by a grant from the Percy Sladen
Memorial Trust, and we desire to take this opportunity of expressing
our indebtedness to the Trustees.

The water used was collected purposely in different situations,
sometimes four miles out at sea, sometimes between one and two hundred
yards, sometimes close to shore from beside Port Erin Breakwater, and
sometimes in the Aquarium at the Biological Station, so as to follow all
variations which might be possible, but all water used was clean fresh
sea-water.

The plankton was completely separated from the water in a measured
volume passed through the Chamberland pump, in two portions, viz.,
(a) that separated by a No. 20 silk-netting, which is the finest silk gauze
made and used for plankton work, and (b) that separated by the material
of the Chamberland filter candle after passing through the silk.

Determinations were then made (a) of the organic matter oxidizable
by potassium permanganate, (1) in acid, (2) in alkaline solution; and
(6) of (1) free and saline ammonia, (2) organic or albuminoid ammonia,
in the following portions: (i) The unfiltered sea-water, (ii) The sea-water

264 BIO-CHEMICAL JOURNAL

filtered through the silk-netting only, (111) the sea-water filtered through
both silk and filter, (iv) the residue caught on the silk, (v) the residue
left on the candle of the Chamberland filter.

As the total number of litres filtered for each experiment was known,
it follows that such a method of procedure gave (a) total amounts of
organic matter in unfiltered, partially filtered, and completely filtered
water, and (b) amount of organic matter removed by ordinary silk tow-
netting and by complete filtration of all organic matter, including
bacteria.

It may at once be indicated that the organic matter as determined by
the direct method was exceedingly small, and fell well within the hmits
of experimental error. The residues gave reliable titrations, as the
suspended organic matter from several litres could be taken for analysis
and so experimental errors could be discovered, but when worked out to
amount per litre the figures give very small amounts and amply confirm
those obtained directly in the analyses in the water, since they fall within
one part per million. The fraction stopped by the Chamberland is much
greater than that stopped by the silk-netting.

In the second portion of our enquiries, estimations were made of the
rate at which organic matter was oxidized by various species of marine
animals, by making determinations of the rate of disappearance of
dissolved oxygen from a known large volume of filtered sea-water in
wide-mouthed stoppered bottles in which the animals were enclosed
without the presence of air for a known period of some hours’ duration.
In addition, the carbon dioxide produced in the same period by the
respiration of the animals was estimated by simple titration of a known
volume of the water with centinormal acid at the end of the experiment,
and using phenol-phthalein as indicator. This simple method was found
to give results of considerable value. Also the amount of oxidizable
matter in the water was determined at the beginning and end of each set
of experiments by the permanganate methods, alongside controls in
which no animals were present in the water, and the amounts similarly
of free and saline, and of albuminoid or organic ammonia. So far from
showing any diminution in dissolved organic matter, as might have been
expected if dissolved organic matter formed the main source of nutrition,
as supposed by Piitter, it was always found that both organic carbon and
organic nitrogen were largely increased, and this also even when the
animals had been kept without food for twenty-four hours previously,

and no obvious solid dejecta were present.

cme

NUTRITION AND METABOLISM OF MARINE ANIMALS 265

In such experiments as these the animals are, of course, living on
the reserve food of their tissues and metabolic organs. The total wastage
from the animals is rightly represented by the sum of the oxygen
disappearing in oxidizing organic matter (and so yielding energy to the
animal for its life-processes) and of the unoxidized organic matter added
as waste of excretions and secretions to the water. Both these quantities
would be increased in an animal which was fed during an experiment, so
that the sum of these two quantities of organic matter must be regarded
as the minimal demand. The amounts are given in Section B, and it
will there be shown that the amounts of organic matter in solution, or
as separated by a Chamberland pump, sufficient to satisfy such demands,
would be distributed in too inordinately large voiumes of water for the
animals to obtain them by fishing out or filtering average sea-water.

It would appear from these experiments as if the higher animals,
instead of living upon dissolved organic matter, or bacteria or protozoa,
on the contrary contributed energy in the form of organic compounds to
sea-water for the use of these lowlier organisms. The source of this
organic matter, in addition to that required for their own use, being
grosser food-supply, or plankton occurring in shoals.

The results of the investigations will be described in two separate
sections, the first dealing with the organic matter of the sea-water and
the second with the metabolism of the marine animals.

Section A. THE ORGANIC MATTER OF THE SEA IN SOLUTION AND
SUSPENSION

Any accurate determination of the amount of organic matter in
sea-water is rendered almost impossible for two reasons: first, the amount
of organic matter is excessively small; secondly, the amount of inorganic
matter as chlorides and carbonates is relatively enormous, and the
chlorides especially interfere with all known methods of oxidation by
yielding free hydrochloric acid, and, in presence of oxidising agents,
free chlorine which interferes with the titrations. For these reasons all
that can at present be given is a very low figure as a maximum, which
it can safely be said that the organic matter does not exceed.

Strange as it may seem, in view of the immense superstructures which
have been built upon the supposed presence of appreciable amounts of
organic matter, we have only been able to find one observer, E. Raben,}

1. Wissensch. Meeresuntersuch.herausg. v.d. Kommission z. Wissens. Untersuch. d. deut.
Meere in Kiel u. d. biol. Anstalt auf Heligoland, N.F., Bd. XI, Leipzig, Lipsius u. Tischer, 1910.

266 BIO-CHEMICAL JOURNAL

who has carried out a series of analyses in filtered sea-water, and the
greater part of these are vitiated by being made on polluted water from
Kiel Ford, the great naval base of the German Navy.

In regard to unfiltered water the amount of quantitative work is also
most meagre. Not only has Piitter failed to make any determinations of
the amount of organic matter in water completely cleared of suspended
matter by Chamberland filtration, there is no note in his paper that the
water was filtered from plankton by silk-netting or filter paper, and one
is left from the description to understand that the water analysed was
entirely unfiltered water from the Gulf of Naples. Twelve analyses are
given, and, as Henze points out, although these were taken on days close
together, the values obtained show enormous variations from 68 mgms.
of organic carbon per litre to 134 mgms. Such variations might have
caused Piitter to suspect the accuracy of the method used, but he draws
no attention to them. In discussing the nature of the supposed organic
carbon, he concludes from the simple process of distillation in acid, and
catching the distillate in decinormal alkali, that no less than 36 mgms.
per litre are present as volatile organic acids. One would have thought
that this large quantity of volatile organic acids would have afforded an
opportunity of finding out the nature of such acids, but Piitter replaces
this experimental work by a theoretical speculation, as a result of which
he places the average molecular weight of the volatile fatty acids at 128.
Inspired by this result, he proceeds to fix on a theoretical basis, the
composition and oxygen value of the remaining part of the supposed
organic matter amounting to two-thirds of the whole.

There is little doubt that the supposed volatile organic acids of
Piitter consist of hydrochloric acid distilled off from his mixture of sea-
water and sulphuric acid, and the same type of explanation holds for his
supposed 68 to 134 mgms. of organic carbon per ltre of sea-water, upon
which the whole superstructure of his solution-nutrition theory is based.
This organic carbon in Piitter’s experiments probably arose, as Raben?
points out, from acetic acid set free from lead acetate, which Pitter
interposed to take up the evolved hydrochloric acid. The volatile acetic
acid was carried over to the glowing tube containing lead chromate and
copper oxide, and there combusted. So that Piitter’s figures for organic
carbon represent a variable amount of organic matter arising from one
of the reagents used in the analyses.

The method used was that known as Messinger’s wet combustion

1. Loc: cit.

NUTRITION AND METABOLISM OF MARINE ANIMALS $267

method, in which the organic matter is oxidised by a mixture of potassium
bichromate and strong sulphuric acid, or by chromic acid. This method
was originally designed for use in potable waters where the amount of
chlorides is low. When transferred to use in the case of sea-water, the
utmost care must be taken that hydrochloric acid or chlorine bodies
are not allowed to pass over and become absorbed in the weighed potash
bulbs intended for absorption of the carbon dioxide.

Shortly afterwards Henze! repeated Piitter’s work, taking care by
the interposition of a tube of metallic antimony and of a glowing tube
of lead chromate and copper oxide to prevent any chlorine compounds
or acid reaching the carbon-dioxide absorption apparatus. With such
precautions, Henze found that while organic matter added to sea-water
was readily oxidizable and re-obtained with ease, the amount of organic
matter in sea-water under natural conditions was such as to fall within
the region of experimental errer. Henze concludes that so tar as Piitter’s
views are based on presence of appreciable amounts of organic matter in
the sea-water they are devoid of all experimental proof.

In his later monograph,’ Pitter refers to this work of Henze’s, and
appears to accept it as proving that his former work was based on much
too high figures, and without any further experimentation discards his
earlier average amount of 90 mgms. of organic carbon, and replaces it
by a new figure of 4 to 5 mgms.

In this later monograph on the subject, Piitter admits that the
Messinger method of determining organic carbon cannot be utilized,
and suggests its replacement by determinations of the amount of oxygen
necessary for oxidation of the organic matter as completely as may be
done by potassium permanganate in acid or alkaline solution.

This method is used by Piitter in his respiratory experiments, and
he appears to be unaware of the fact that it also is upset by the action
of the hydrochloric acid and chlorine set free by the interaction of the
acid and permanganate upon the chlorides present in such preponderating
quantity.

As will be shown later, the method can only be used with the
greatest caution, in the cold, and using as a zero a blank, in which no
organic matter but an equal amount of chlorides is present, under
comparable conditions throughout.

When this is done, the difference in readings between sea-water and

1. Pfltiger’s Archiv. f. d. ges. Physiol., Ba. CX XIII, S. 487, 1908.

2. Die Ernahrung der Wasserthiere, und der Stoffhaushalt der Gewisser, Fischer, Jena,
1909, pp. 49 and 99 to 101.

268 BIO-CHEMICAL JOURNAL

the inorganic control fall practically to zero, showing that organic matter
is scarcely detectable in the sea-water.

But Piitter recommends that the mixture of sea-water, acid, and
permanganate shall be boiled for ten minutes.

In an experiment carried out by us to test the effects of such a
procedure, six grams of pure fused sodium chloride were taken and
dissolved in 250 c.c. of distilled water, so as to represent approximately
the concentration in chlorides of sea-water, and the usual amount taken
in actual experiments. To this, 10 c.c. of 25 per cent. sulphuric acid
and 10 c.c. of N/20 KMnO, solution were added, the same quantities as
might be used in an experiment.' The whole was then boiled for only
five minutes, instead of the ten recommended by Piitter, cooled and
titrated with N/10 Na,S,O,, after adding excess of potassium iodide
to release iodine, when only 5°8 ¢.c. were required instead of 10 c.e.
A parallel experiment with 250 c.c. of distilled water and no sodium
chloride, treated otherwise similarly, gave no loss of chlorine, requiring
at the end 10 c.c. of decinormal thiosulphate for neutralization.

This shows clearly that any kind of result as to content of oxidizable
organic matter may be obtained by following the method described by
Piitter in his moncgraph. For example, the above experiment would
fallaciously show 13°4 mgms. per litre of oxygen required for oxidation
of organic matter, this in an artificial sea-water containing not a trace
of organic matter but only pure sodium chloride fused before use.

The experiments recorded below with ‘ Falk’ salt,? and the salts of
sea-water after evaporation and ignition over a bunsen-burner, show
the same effects, but in lessened degree, because the experiments were
carried out in the cold and in stoppered bottles. These figures form
the blanks to the sea-water. Where organic matter is actually present
as in the residues on silk-netting and Chamberland, the change in the
figures shows that such true organic matter is readily oxidizable by the
permanganate, and results can be obtained when proper precautions
are taken.

For these reasons, the results obtained by Piitter by the permanganate
method resemble those with the Messenger method in being vitiated by
experimental errors, and all proof is lacking of any appreciable amount
of oxidizable organic matter in normal filtered sea-water.

1. See below.

2. This is a pure table salt manufactured by the Salt Union, Ltd., Liverpool. It contains
no appreciable amount of organic matter, but in any case it was always heated before use.

NUTRITION AND METABOLISM OF MARINE ANIMALS 269

More recently, an important contribution to the subject of the amount
of organic matter in sea-water has been made by Raben, who has, like
Henze, demonstrated the illusory character of Piitter’s analyses, and
expresses his complete concurrence with Henze’s concluding statement
quoted above.

Four out of the six complete and careful analyses made by Raben
were carried out upon water from the ‘ Kieler Féhrde,’ and two only
with water from the open sea. The results show a wide difference, the
four from Kiel yield 13°9, 12°8, 11°4, and 10°9 mgms. per litre
respectively, while those from the open sea yield only 3'0 and 3°2 mgms.
respectively. Raben ascribes this difference to three causes, viz., nearness
to land, presence of the German fleet in the harbour, and the access of
sewage (Schmutzwasser).

An amount of 3 mgms. per litre is equivalent to 0°3 mgm in the
quantity of water taken for analysis, viz., 100 ¢.c., which would yield
1‘l mgm. of carbon dioxide or approximately 0°5 c.c. This is a
quantity, as Henze states, which is well within the limits of experimental
error with such a complicated apparatus, working for a period of two
hours in a stream of air.

The exacter experiments of Henze and of Raben have accordingly
shown that the organic carbon dissolved in sea-water is an exceedingly
minute quantity lying at the limits of detectability by the best known
methods.

There are one or two figures given by earlier observers indicating a
higher value, but the methods of analysis are so questionable in the light
of more recent knowledge as to rob them of all value. In connection
with the Challenger Expedition, J. Y. Buchanan made a few estimates
of organic carbon by the rather drastic procedure of distilling to dryness
with potassium permanganate, and found in different samples from 2°3 to
77 mgms. of organic carbon per ltre. But in summarizing, he himself

‘

states: ‘The few determinations of ‘‘ organic carbon’’ by means of
permanganate of potash were experimental, and were entered in my
journal in course. I was, however, so dissatisfied with the experiments
that I soon gave them up, and I attach no value to the results.’

The early observations of Natterer in 1892 need only be mentioned
because Piitter seems to attach some importance to them at the present
day. They were made long before the difficulties of the problem were
appreciated, and gave results varying in different samples from 10 to
50 mgms. per litre. But when it is explained that they were obtained

270 BIO-CHEMICAL JOURNAL

by heating an alcoholic filtrate, still containing a large residue of
inorganic salts, in a test-tube until the test-tube began to soften, the
difference in weight of the tube before and after being taken as the amount
of organic matter, chemists of to-day will know how much credence to give
them.

Piitter in his monograph! also lays considerable stress on an organic
substance obtained from sea-water resembling stearin or palmitin. This
substance was obtained at the end distillation of 200 litres of water,
thrown together from all sources at the end of the expedition, as Natterer
himself states it came over ‘in ganz geringer Menge’ only a few
centigrams from the whole 200 litres, of a miscellaneous collection of
water samples at the end of the trip, and the only evidence of palmitin
or stearin is that in heating in an open tube it gave a smell resembling
overheated palmitic or stearic acid, and that it melted at approximately
the same temperature as stearic acid.

It is certainly remarkable to find such an isolated experiment of such
a type as this quoted fifteen years later, without any further experimental
work by the person quoting it, as evidence of the presence of higher fatty
acids in sea-water.

The most reliable method for obtaining the organic carbon in well
and river water consists in evaporating down to dryness in presence of
sulphurous acid to drive off the inorganic carbonate, and then subjecting
the residue to combustion by either the ordinary dry method or a wet
method, and estimating the carbon dioxide produced either volumetrically
by gas analysis, or by absorption in potash balls and weighing.

This method is rendered useless in the case of sea-water by the
enormous amount of chlorides present compared to the small amount of
organic matter. When sea-water is evaporated to dryness in a porcelain
basin and then heated over a Bunsen burner, there is just a faint yellowing
of the mass and then it turns white. To obtain enough carbon to give
any appreciable reading in a combustion, would require enough sea-salt to
fill a combustion tube.

The method used throughout was wet oxidation with potassium
permanganate in the cold for a number of hours stated in each case.
As soon as we appreciated the precautions required we employed a
control, carried out under like conditions, of ignited sea-salt or Falk table
salt, also ignited before use.

For the greater number of cases oxidation in acid solution was

fey 0c. Cit. po LOzs

NUTRITION AND METABOLISM OF MARINE ANIMALS 271

employed, but as we thought part of the difficulty might be avoided by
oxidizing in alkaline solution, in later experiments we also used alkali.
It would appear, however, that the greater part of the difficulty with the
chlorides is attached to the actual titration at the end, which must be
done in both cases in acid solution, and when allowance is made for the
fact that in alkaline reduction only 3, in acid solution 5 atoms of
oxygen are set free, there is practically no difference in the two sets of
results.

It has been suggested by Dupré! that the method can be used in
presence of chlorides, and even in sea-water, if glacial phosphoric acid
be used instead of sulphuric acid. Accordingly, we experimented with
this modification, but without change in the result. In presence of
sodium chloride the titration figure showing remaining excess of
permanganate at the end is always lower than with distilled water, so
falsely indicating a small amount of organic matter.

The nearest we were able to approach the problem, accordingly,
consisted in carrying out a control under as like conditions as possible to
give a zero, and when this is done the organic matter found in the sea-
water is negligibly low.

The method is too well known to require detailed description.2 ‘The
following is an outline:—250 c.c. of the water were taken, alongside
other samples and controls, in wide mouthed glass-stoppered bottles of
about 400 c.c. capacity, 10 ¢.c. of one in four sulphuric (or phosphoric
acid) was added, and the appropriate amount of N/20 or N/80 KMn0O,
solution, such as 10 or 20 c.c. of N/80, or 5, 10, or 15 c.c. of N/20, as
detailed below. The ‘bottles were left at laboratory temperature, except
where otherwise stated, for a number of hours, but all for some time. <A
few drops of a 5 per cent. solution of potassium iodide are then added, so
as to give excess and set free iodine corresponding to the unused excess
of potassium permanganate, and the iodine was then titrated in the usual
way with a N/40 solution of sodium thiosulphate, using starch indicator.

Experiment I.—The water used for this experiment was taken on
the steam yacht ‘ Runa,’ about four miles west of Port Erin, in the open
sea. It was filtered on board the yacht, and the potassium permanganate
and sulphuric acid were at once added on board, and it was then taken to
the laboratory for analysis. In addition, about five litres of the completely
filtered water were retained for determinations of the free and saline

1. Analyst, Vol. X, 1885.
2. See Sutton, Volwmetric Analysis, tenth edition, 1911, pp. 484-488.

272 BIO-CHEMICAL JOURNAL

ammonia and of the organic or albuminoid ammonia. The removal of
the non-dissolved constituents took place in two stages. The water
coming up from the sea in a rubber hose pipe of about one inch bore
(2°5 centimetres), which was sunk about 4 metres in the sea, before passing
to the Chamberland pump was attached to a piece of wide glass tubing
about two feet long. This glass tube passed through a cork about four
- inches from one end. The cork fitted into the neck of an ordinary filter
flask, such as is used with a suction pump in ordinary laboratory work.
A piece of the finest silk netting (20 mesh gauze) was used to make a tube
about five inches long, somewhat narrower at the closed end, and fitting
by its wider open end around the cork, so that it was enclosed airtight
between the cork and the neck of the filter flask, and all the water filtered
had to pass through it. The side outlet of the filter flask was connected
by a second short piece of hose with the Chamberland pump. The silk-
netting accordingly retained the amount of plankton separated from the
quantity of water filtered in any such experiment, the filter flask contained
water which had been silk filtered only, the deposit on the Chamberland
candle at the end gave the non-soluble matter in the volume of water
filtered, which was too small to be removed by the silk netting, and the
completely filtered water was obtained and measured as 1t came from the
pump. At the time of the experiment a sample of unfiltered water was
taken by the closing water bottle from the same depth as the open end
of the inlet hose. In this way five things, viz., two deposits and three ~
waters, were taken at one time from the same spot for analysis.

The quantity of water filtered in this particular experiment was
forty litres. Two samples of the Chamberland filtered water, and one
sample each of the unfiltered and silk-filtered waters, were analysed, 10 c.c.
of N/80 KMnO, (1 c.c.=0°0001 gram of oxygen) being added to each, and
all being then kept in the incubator at 25° C. for four hours. As the first
amounts added were decolourized, 10 c.c. of N/20 KMnO, were added later
in each case. The results obtained are shown in the following table in
milligrams of oxygen required for oxidation of organic matter per litre,
but as explained above these results really show that no organic matter
is present, forthe amounts are no greater than those shown in Experiment
VY, p, where no organic matter is present, and merely represent a small
amount of hydrochloric acid decomposed with liberation of chlorine and its
partial escape. This loss is higher than in subsequent experiments,
because the samples were heated in an incubator to 252 C. instead of
being kept at air temperature.

NUTRITION AND METABOLISM OF MARINE ANIMALS = 273

No. 1. Chamberland filtered water =3'16 mgms. per litre.

No. la. ” 9 =—3'16 ”) ”
No. 2. Silk filtered water =2°48 ,, 3
No. 3. Unfiltered water — 3°84 - fe

The fact that the completely filtered samples appear to contain more
organic matter than that only filtered through silk, is confirmatory
evidence that it is not really organic matter which is determined. But
even making no allowance for experimental errors, it is seen that the
amount of organic matter possibly present taken at a maximum is very
small.

As at this period the disturbing influence of the chlorides was not
fully appreciated, the two deposits on silk and Chamberland candle,
respectively, were taken up in sea-water in this experiment and the next
one. In later experiments the deposits were taken up in distilled water.
Allowances based on the above data were made as deductions for the
amounts of filtered sea-water so used in taking up the residue. The
residue on the silk-netting, which was very small in amount, required
permanganate representing 11°6 mgms, and as this came from 40 litres,
the amount of plankton, separable from the sea-water by such silk,
corresponded to 0°29 mgms. of oxygen! per litre. Similarly, the
Chamberland deposit, which was more copious, was taken, mixed
thoroughly with 400 c.c. of filtered sea-water, and 50 c.c. were taken for
the analysis. This portion was equivalent to 4°55 mems., or 36°4 in the
whole amount of 40 litres, which works out at 0°91 megms. of oxygen per
litre. The total amount of matter in suspension is therefore equivalent
in oxygen to:—silk=0°29, Chamberland 0°91—1°2 mgms. of total
plankton. This figure is free from the chloride error, and as the titration
figures are quite large, it probably fairly accurately represents the
plankton on that day, when fairly heavy takes were being recorded by the
ordinary tow-nettings of the Biologists on board. It is to be noted that
the Chamberland here retains about three times the quantity removed by
the silk.

The free and saline, and albuminoid ammonia as determined by the
usual water analysis method, and testing by Nessler’s reagent against
standard dilutions of ammonium chloride solution, gave 0°04 parts per
million of free and saline ammonia, i.e., 0°04 mgms. per litre, and of
albuminoid ammonia 0°2 parts per million or 0°2 mgm. per litre. If all

1. One milligram of oxygen may be roughly taken as equivalent to 2°5 milligrams of organic
matter, and this factor must be used throughout, see Table at end of this section of the paper.

O74 BIO-CHEMICAL JOURNAL

this albuminoid ammonia be taken as protein it must be multiphed by
the factor 5 to get protein, which would give 1 mgm. per litre of protein.

Experiment I]. In this experiment the oxidation with perman-
ganate and sulphuric acid was carried out at laboratory temperature for
a period of about fifteen hours. The volume of water filtered was 43
litres. It was taken from three to four metres below the surface, on the
yacht ‘ Runa,’ on the afternoon of the same day as that of the preceding
experiment, the situation from which it was taken was about three miles
north-west of Port Erin in the open sea.

The analyses concern one sample of water completely filtered through
Chamberland, two samples filtered through silk only, one of unfiltered
water, one the deposit retained by the silk, and one of the deposit on the
Chamberland candle.

The water analyses gave the following figures, to which the same
remarks apply as in the preceding experiment : —

No. 1. Chamberland filtered water=2°24 mgms. per litre

No. 2. Silk filtered water = e890) es x
No. 2a. 9 9 = 2°24 ” ”
No. 3. Unfiltered water =Ii3 .,; -

The deposit on the silk-netting, taken up in 240 c.c. of filtered sea-
water, required 6°55 mems. of oxygen for oxidation. On making allowance
for the sea-water, this gave 5°79 mgms. in 43 litres, or 0°13 mgms. of
oxygen per litre.

The deposit on the Chamberland candle was taken up in 600 c.c. of
filtered sea-water, and 50 c.c. of this, or »,th, with 50 c.c. additional
of filtered sea-water were taken for analysis. Its organic matter
required 2°75 mgms. of oxygen, which amounts to 2°64 mgms.
for the fraction of the residue only, multiplied by 12, this gives
31°68 mgms. from the 43 litres, or 0°74 mgm. per litre. The total
suspended matter per litre is accordingly represented by 0°87 mgm. of
oxygen, of which about } is removed by the silk and 8 by the
Chamberland.

The ammonia determinations were made in this case on the unfiltered
water, and gave:—TFree and saline ammonia=0°l mgm. per litre;
albuminoid ammonia=0°16 mgm. per litre. These do not differ greatly
from the results of the previous experiment with filtered water, and
illustrate the excessively low organic content of pure sea-water whether
filtered or unfiltered.

.*

NUTRITION AND METABOLISM OF MARINE ANIMALS = 275

Exprertment IIT. The oxidation in this experiment was also carried
out in the cold for a period of fifteen hours, and phosphoric acid (ortho-
phosphoric prepared by acid combustion and oxidation of phosphorus in
the laboratory) was substituted for sulphuric acid to attempt to minimize
loss of chlorine from the chlorides as recommended by Dupré (loc cit.).
No difference, however, was found, as will be seen by comparing
the results with sea-salt and ‘Falk’ salt after ignition as shown in
{xperiment V, p.

The water was filtered and treated on the yacht, taken on Saturday
morning, April 13th.

The volume of water filtered was 53°2 litres. The silk netting
deposit was so meagre that it was not regarded as worth doing, and it
may be remarked that the ordinary tow-nettings taken on that occasion
for biological purposes were very meagre. There was no visible deposit on
the silk-netting in this experiment at the end of filtration of the 55°2 litres.

One sample each of the three conditions of the water were analysed
for oxidizable organic matter, using in all 25 ¢.c. of N/20 KMnO, and
10 c.c. of 25 per cent. phosphoric acid to 250 c.c. of sea-water, the
amounts used up of the standard KMnO, were only 0°7, 2°2, and 2°5 ¢.c.
respectively in the three samples, and the results per litre work out as

follows:—
1. Chamberland filtered water=1°'12 mgms. per litre.
2. Silk filtered water =3'52 ,, -
3. Unfiltered water —3'°68 ,, is

The residue in the Chamberland candle showed 9°58 mgms. of
oxygen required for 53°2 litres, corresponding to 0°18 mgms. per litre.

In this experiment, both filtered and unfiltered water were analysed
for ammonia with the following results :—

Unfiltered water :—Free and saline ammonia = 0°06 per million or
0°06 mgm. per litre; albuminoid ammonia=0'17 per millon or O17
mgm. per litre.

Chamberland filtered water:—Free and saline ammoma='G2 per
million, or 0°02 mgm. per litre; albuminoid ammonia =O0'l per million
or 0'l mgm. per litre.

Such a water would be put down as excellent amongst potable waters
as regards organic matter. The figures again show what an infinitesimal
trace of organic matter is present, especially in the filtered water, which
contains less than the unfiltered, but all figures lie at the limit.

ExprertMent IV. The water used in this experiment was taken, on

276 BIO-CHEMICAL JOURNAL

the yacht, about 200 yards off Sugar Loaf Rock, outside Calf Island
Sound in deep water, about two metres from surface, on Monday
afternoon, April 15th. The volume of water filtered was 25°2 litres.
The oxidation was carried out in the cold for fifteen hours, using 20 c.e.
of N/80 KMnO, and 10 c.c. of 25 per cent. glacial phosphoric acid to the
250 c.c. sample of water. The reagents, as usual, were added at once on
the yacht.
| The results were as follows :—
No. 1. Chamberland filtered water=0°82 mgms. per litre.
No. 2. Silk filtered water —0°82
No. 3. Unfiltered water —(0°89

In regard to the silk-netting and Chamberland candle deposits,
these, instead of being taken up in sea-water, were washed off by distilled
water, so as to get rid of adverse influence of chlorides. The entire
residue was taken in each case, in 250 c.c. of distilled water, and 10 c.c.
of N/20 KMnO, and 10 c.c. of 25 per cent. glacial phosphoric acid were
used for oxidation.

The silk-netting deposit was oxidised by 0°664 mgm. of oxygen,
which represents 0°03 mgm. of oxygen per litre.

The Chamberland candle deposit required 2°33 mgms., which gives
0°09 mgm. of oxygen per litre.

The total plankton is accordingly represented by 0°12 mgm. of
oxygen per litre.

The unfiltered water gave the following low figures in the ammonia
determinations:—Free and saline ammonia=0'035 per million, and
albuminoid ammonia=0'08 per million.

EXPERIMENT V. The sea-water used in this experiment was pumped
from the beach at the shore end of Port Erin Breakwater, when the
water was about eight or nine feet deep, and the inlet hose was about
five or six feet below the surface. A volume of 140 litres was pumped,
and in addition to the analyses this water was utilized for Experiment VII
of Section B for the oxygen exchanges of the animals.

Twelve analyses of oxidizable organic matter with permanganate
were made in all in this experiment. Six of these were with the three
conditions of the water, viz., Chamberland, and silk filtered, and
unfiltered (a) in acid solution, and (b) in alkaline solution. Four were
made on the two deposits, on silk and on Chamberland, (a) in acid and
(6) in alkaline solution.

Two analyses were made upon 8 grams of ‘ Falk’ salt made up, after

"-*

<)) Gee lee eos -y “a

7

NUTRITION AND METABOLISM OF MARINE ANIMALS = 277

ignition to destroy any possible organic matter, to 250 c.c. with distilled
water, so as to give an artificial sea-water free from any organic matter
to serve as a control. All were carried out alongside of one another in
similar bottles of equal size, and all conditions made as closely identical
as possible.

The results of these two controls may also be used for the previous
experiments, and show that the apparent small amounts of organic matter
can all be accounted for on the basis of chloride decomposition.

The results were as follows :—

(A) WATERS IN ACID SOLUTION
No. 1. Unfiltered water

250 c.c. + 20 c.c. N/80 KMnO, + 10 c.c. of 25 per cent. glacial

phosphoric acid, used up 1°4 ¢.c. = 0°56 mgm. per litre.
No. 2. Silk-filtered water

200 c.c. + 20 c.c. N/80 KMnO, + 10 c.c. of 25 per cent. glacial

phosphoric acid, used up 1°16 ¢.c. = 0°46 mgm. per litre.

No. 3. Chamberland filtered water
200 c.c. + 20 c.c. N/80 KMnO, + 10 c.c. of 25 per cent. glacial
phosphoric acid, used up 0°96 c.c. = 0°38 mgm. per litre.
(B) WATERS IN ALKALINE SOLUTION

No. 4. Unfiltered water
250 c.c. + 20 c.c. N/80 KMnO, + 5 c.c. N NaOH, at the end of
20 hours made acid (with glacial phosphoric acid 25 per cent., 11 c.c.)
and titrated used up 2 ¢.c. = 0°48 mgm. per litre.
No. 5. Silk-filtered water
250 c.c. + 20 c.c. N/80 KMnO, + 5c.c. N NaOH, titrated as above
in acid, used up 1°16 c.c. = 0°28 mgm. per litre.
No. 6. Chamberland filtered water
250 c.c. + 20 c.c. N/80 KMn0O, = 5 c.c. N NaOH, titrated in acid,
used up 1:16 c.c. = 0°28 mgm. per litre.
(c) DEposirs IN ACID AND ALKALINE OXIDATION
No. 7. Silk-netting Deposit ovidized in acid solution
One half of the deposit + 250 c.c. distilled water + 10 c.c. N/20
KMn0O, + 10 c.c. of 25 per cent. glacial phosphoric acid, after 20 hours
at air temperature, used up 1°66 c.c. of N/20 KMnO,=0°664 mgm. = 1°33
mgm. in whole deposit, this in 140 litres = 0°009 mgm. of oxygen per

litre.

278 BIO-CHEMICAL JOURNAL

No. 8. Silk-netting Deposit oxidized in alkaline solution
One half of the deposit + 250 c.c. distilled water + 10 c.c. N/20
KMnO, + 5 c.c. N NaOH, after 20 hours at air temperature, made acid
with 11 c.c. of 25 per cent. glacial phosphoric acid, used up 3°16 of N/20
KMnO, = 0°758 mgm., so in whole deposit = 1516 mgm., as this is in
140 litres, the amount is 0°011 mgm. of oxygen per litre.

No. 9. Chamberland deposit in acid solution

One-tenth of Chamberland deposit + 250 c.c. of distilled water
+ 10 c.c. N/20 KMnO, + 10 c.c. of 25 per cent. glacial phosphoric acid,
left 20 hours at air temperature, used up 2°4 c.c. of N/20 KMnO, =
0°96 mgm., therefore whole deposit = 9°6 mgms., this in 140 litres =
0°069 mgm. of oxygen per litre.

No. 10., Chamberland deposit in alkaline solution

One-tenth of Chamberland deposit + 250 c.c. of distilled water + 10
e.c. of N/20 KMnO,+ 5 c.c. N NaOH., at end cf 20 hours at am
temperature made acid with 11 ¢.c. of 25 per cent. glacial phosphoric acid,
used up 3°8 c.c. of N/20 KMnO, = 0°92 mgm., so in whole deposit there
are 9°2 mgm., as this is in 140 litres the amount per litre is 0°066 mgm. of
oxygen.

The total suspended matter removable both by silk and Chamberland
is when oxidized in acid solution=0'078 mgm., in alkaline solution 0°077
mem. In working out the calculation it is to be remembered that the
oxidizing power of the permanganate in alkaline solution is only three-
fifths of its power in acid solution. Also the end titration in both cases is
in acid solution. These concordant results demonstrate clearly that the
oxidizing value of the organic matter in suspension, or total plankton in
this water, is less than 0°1 mgm. per litre. The factor for conversion of
oxygen used into total organic matter, as shown by combustion, ranges
between two and six for different waters, taking the highest value, the
maximum amount of organic matter in the total plankton hes at 0°4 mgm.
per litre.'| The amount of material available in the deposits from 140
litres was quite sufficient to give good figures, there were practically no
chlorides present, and there is, therefore, little question as to the accuracy
of these figures.

1. The remaining eight-tenths of the Chamberland deposit was dried and weighed and
found to be 0:195 gram, this gives 0-247 gram for the entire deposit. This deposit was found by
careful incineration to contain 77°5 per cent. of ash and 22-5 per cent. of organic matter. The
amount of organic matter in the total 140 litres stopped by the Chamberland filter was accordingly

56 milligrams or 0-4 milligram per litre. On comparison with the result by the permanganate
method in the text, this yields a multiplier of 6.

NUTRITION AND METABOLISM OF MARINE ANIMALS = 279

(D) * FALK’ SALT AFTER INCINERATION
No. 11. ‘Falk’ salt (ignited), oxidized in acid solution

Hight grams ‘ Falk’ salt + 250 cc. distilled water + 20 c.c. N/80
KMnO, + 10 c.c. of 25 per cent. glacial phosphoric acid left for 20 hours
at air temperature, used up 3°52 ¢.c. of N/80 KMnO, = 1°40 mgms. per
litre.

No. 12. ‘Falk’ salt (tgnited), oxidized in alkaline solution

Fight grams of ‘ Falk’ salt + 250 ¢.c. of distilled water + 20 c.c. of
N/80 KMnO, + 5c.c. N NaOH, left at air temperature for 20 hours, then
made acid with 11 c.c. of 25 per cent. glacial phosphoric acid and
titrated, used up 4°72 c.c. of N/80 KMnO,, equivalent to 1°15 mgms. per
litre.

These two samples of artificial sea-water, quite free from organic
matter, hence show more organic matter per litre than the natural water,
and prove that the slight amounts of organic matter there indicated lie
within the limits of experimental error, and are probably due to slight
chlorine escape.

The ammonia figures were also obtained for this filtered sea-water
and confirm the above results in showing the usual amounts of organic
matter present. The results are as follows:—Free and saline ammonia
=0°04 per million; albuminoid ammonia=0°21 per million.

ExrertMeNnt VI. This experiment was carried out to provide more
controls than those mentioned in the previous experiment.

No. 1. Sea-water titrated at once without allowing time for
oatdation.

A volume of 250 ¢.c. of Chamberland filtered sea-water was taken,
10 c.c. of N/20 KMnO,, and 10 c.c. of 25 per cent. sulphuric acid, then
excess of potassium iodide solution was added at once without waiting and
it was at once titrated. It appeared to have used up 0°8 c.c. of N/40
thiosulphate = 1°6 ¢.c. of N/80 KMnO, = 0°16 mgm. of oxygen or 0°64
mgm. per litre.

No. 2. A volume of 250 c.c. of sea-water was taken to dryness in a
porcelain evaporating dish, and incinerated over a Bunsen burner. The
salt turned slightly yellowish and then white. The residue in which
organic matter had been thus destroyed was taken up in 250 c.c. of
distilled water + 15 e.c. of N/20 KMnO,+10 c.c. of 25 per cent. sulphuric

280 BIO-CHEMICAL JOURNAL

acid, after standing 15 hours this had used up 2°1 c.c. of N/20 KMnO,=
0°84 mem. of oxygen or 3°36 mgm. per Litre.

No. 3. A similar experiment but kept for 15 hours in the oven at
25° C., used up 44 c.c. of N/20 KMnO, = 1°76 mgm. or 7:04 mgms.
per Litre.

No. 4. Hight grams ‘Falk’ salt (unincinerated) + 250 c.c.
distilled water + 20 c.c. N/80 KMnO, + 10 c.c. of 25 per cent. glacial
phosphoric, for 24 hours at air temperature, uses up 1°8 c.c. = 0°18 mgm.
of oxygen or 0°72 mgm. per litre.

No. 5. Same as No. 4, but ‘ Falk’ salt, incinerated beforehand,
uses up 2'1 c.c. = 0°21 mgm. oxygen or 0°84 mem. per litre.

No. 6. Same amounts as Nos. 4 and 5, but in alkali instead of acid
and unincinerated, uses up 4°6 c.c., which with factor of 3 instead of 5
oxygen, amounts to 0°27 mgm or 1:08 mgm. per litre.

No.7. Sameas No. 6, also in alkali, but the ‘ Falk’ salt incinerated,
used up 4°3 c.c. = 0°26 mgm., or 1°04 mgm. per litre.

No. 8. Sea-water, 250 c.c. evaporated down and incinerated, gave
8°2 grams of salt, this made up again to 250 c.c. with distilled water,
20 c.c. of N/80 KMnO, and 10 c.c. of 25 per cent. glacial phosphoric acid
added, and left 24 hours at room temperature alongside Nos. 4 to 7 of this
series, used up 2°7 c.c. = 0°27 mgm., or 1°04 mgm. in one litre.

It appears, accordingly, from the results of these controls that an
amount of 1 to 3 mgms. per litre must be deducted from the results
obtained above with the sea-water and the various filtrates, in order to
show any organic matter, and when this is done all organic matter
disappears; in other words, dissolved organic matter in pure sea-water is
so small in amount as to fall within experimental error, and in the average
certainly can be safely said not to exceed one mgm. per litre.

In other words, the titration figure of the permanganate solution in
presence of so much chlorides is slightly lower than in distilled water, so
falsely showing a trace of organic matter if compared with the titration in
distilled water; when allowance is made for this, no organic matter is
shown by this method, which would certainly detect it if any appreciable
amount were present.

For this reason, the organic ammonia in the case of sea-water gives
a more satisfactory guide. Even when all the organic ammonia is taken
as protein, there is less than 1 mgm. per litre present, and it is probable
that only a fraction is present as protein.

OL ee ae

‘pe

NUTRITION AND METABOLISM OF MARINE ANIMALS 281

In all cases analysed, the amount of organic matter as plankton is
surprisingly low. Although the Chamberland filtration removes about
three times as much as silk-net filtration, the two together only form a
fraction of a milligram per litre.

It may be concluded that at the season and under the conditions of
our experiments, the combined dissolved and suspended organic matter
does not on the average much exceed 1 to 2 mgms. per litre.

The various results of Experiments I to V on amounts of plankton
are shown in the accompanying Table, in which the last column gives
organic matter obtained by multiplying the oxygen consumed by 2°5,
the average factor! :—

ToraL PLANKTON IN MILLIGRAMS PER LITRE
Silk Netting. Chamberland. Total. Organic Mat ter.

Expt. I ae mer tao 0°91 V2 3°00
Hxpt. fl... a meBh | ta 1: 0°74 0°87 Pal by
Pept. FED >*... oe) Net 0°18 0:18 0°45
Beept. LV... -« 0F05 0°09 0°12 0°30
Expt. V,inacid ... 0°009 0°069 0078 0°19

ne in alkali 0011 0°066 0077 0°19

Section B. THE RATE OF OXIDATION OF ORGANIC MATTER IN CERTAIN
MARINE ANIMALS

The amount of oxidizable organic matter in solution and suspension
in sea-water, under conditions of average distribution, having been
determined as nearly as possible in the preceding experiments, we passed
on to obtain some idea of the rate at which certain marine animals
consumed organic matter, as measured by the rate at which they
consumed oxygen dissolved in the water. Three series of experiments
were made, in the first two series the animals were used as soon as taken
from the sea, and the considerable amount of excrement in some cases
showed that digestion was actively proceeding; in the third, the animals
were kept without feeding, in filtered sea-water for 24 hours before
commencing the experiment, in order to obtain results in the analyses of
the water afterwards as free as might be from the disturbing influence of
excretion of solid matter.

In the latter case the period of deprivation of food was not of

1. The actual value of this factor cannot be determined for the sea-water. The estimation
of the organic matter in the Chamberland deposit of Expt. V leads as already indicated to a
value of 6, but this is probably too high.

282 BIO-CHEMICAL JOURNAL

sufficient duration so far to reduce the reserve store of metabolized
material as to decrease the rate of using up of food by oxidation. The
volume of sea-water allowed was in most cases large compared to the
volume of the animals, and under such conditions it might have been
expected on the hypothesis of Piitter that the amount of dissolved organic
matter would have ‘decreased, but, on the contrary, it was always
_considerably augmented.

Both this increased amount of dissolved and suspended organic
matter, and the amount which had disappeared, corresponding to the
amount of oxygen used up by the animal in its respiratory exchanges,
must be added together to indicate the amount which under natural
conditions would have to be replaced by the food of the animal.

These two quantities of organic matter are the index of the demand
of the animal for a food supply, the amount of the oxygen used up gives
‘a measure of the energy utilized by the animal, and the organic matter
added to the water shows the amount passed through without utilization.

Each type of animal was kept separately for an observed time-interval
in a large wide-mouthed bottle, the average volume of the bottle being
2,800 c.c. A control bottle was kept under like conditions throughout
the experiment without any animals, so as to give a deduction for
any consumption of oxygen by bacteria or small plankton during the
experiment, but the loss of oxygen in this way was found to be very
small. .

In starting the experiment the animal was transferred to the bottle
in which it was to be kept, and well washed with filtered sea-water coming
directly from the Chamberland pump. The bottle was then filled in a large
clean bucket, which was also full of the filtered water, so that the lid or
stopper of the bottle could be shipped on under water without including any
air. Hach experiment was carried out in two stages. Samples were taken
out at the expiry of the first period mentioned in each case for the
determinations to be indicated presently, and then a sufficient amount of
filtered sea-water was added to replace that removed, and the experiment
allowed to proceed overnight till the following day, when it was finally
interrupted and the determinations made.

The most important determination in this series was (a) the amount
of dissolved oxygen in the measured volume of water at the beginning and
end of the experiment, but in addition () the output of carbon dioxide
was determined by titration, as described below, of the water at
beginning and end with centinormal caustic alkali using phenol-phthalein

a

lin’ Seer 4g.

NUTRITION AND METABOLISM OF MARINE ANIMALS — 283

as indicator, (c) the oxidizable organic matter by the permanganate
method as in Section A, (d) the free and saline and organic ammonia,
(¢) the moist weight of the animals, the total dried weight, the weight of
organic matter when dried, the weight of skeleton, and the ash in the
skeleton.

The method used for determining the dissolved oxygen was that of
Winckler, which has been so often described in connection with similar
work and in text-books of water analyses,! that it need not be given in
detail. In principle it consists in utilizing the free dissolved oxygen to
convert manganous to manganic oxide, then by acting upon this with
hydrochloric acid im presence of hydriodic acid to set free the
corresponding amount of iodine, which is then titrated in the usual
fashion by standard sodium thiosulphate in presence of starch indicator.

The work is carried out in an ordinary narrow-mouthed bottle of
about 300 c.c. capacity, such as is used for bench reagents in the
laboratory. This is completely filled with water, so that the stopper is
insertable without any air-bubbles, then the stopper is removed, and the
volume it contains is measured and noted. Such a bottle of measured
volume is completely filled with the sample of sea-water to be estimated,
taking care this is not aerated in the process. By means of a pipette 1 c.c.
of a 10 per cent. solution of manganous chloride is introduced to the
bottom of the bottle, and followed by 5 c.c. of a mixture of saturated
caustic potash containing 10 per cent. of potassium iodide, also passed in
at the bottom. The stopper of the bottle is then inserted without
introducing any air-bubbles and the contents mixed by rotation. The
bottle is left for some time for the manganous hydrate to take up all
oxygen and settle to the bottom. In about half an hour, the bottle may
be opened and 5 c.c. of concentrated hydrochloric acid passed to the
bottom by a pipette, the stopper is again inserted and the contents
thoroughly mixed, when the acid re-dissolves the mixed hydrates of
manganese, and an amount of iodine is set free corresponding to the
amount of oxygen in the water. The contents are transferred to a flask
and titrated with thiosulphate.

ExrertMENt VII. The water used in this experiment was
Chamberland filtered, being a portion of the 140 litres pumped from
beside the Breakwater as described in Experiment V, Section A. Three
large wide-mouthed bottles of 2,800 c.c. capacity were completely filled,

1. See Sutton’s Volumetric Analysis, tenth edition, 1911.

284 BIO-CHEMICAL JOURNAL

the lids being fitted on under the sea-water, so as to exclude all air,
and contained as follows:—(a) Control containing the filtered sea-water
only, (b) a large-sized sponge, (c) a group of ascidians.

The experiment was commenced at 11 a.m., April 16th, and was
interrupted in 8} hours, viz., at 7.30 p.m., to take samples for
determination of dissolved oxygen, a small volume of fresh sea-water was
then added to replace that removed for analysis (about 300 c.c.), and the
experiment was allowed to continue till 11 a.m. on April 18th; that is,
for 48 hours in all. The results of the two intervals are shown in Table I,
Part 1 and Part 2, as under :—

TAsEy .—-vank Lf

Dissolved Oxygen after 85 hours
Milligrams Total mg. in 2,800 c.c. Milligrams of Oxygen

per litre. used by Animals.
(a) Control aa 7:93 22°20 —
(6) Sponge ae 7:42 PAU ETUC 1°43
(c) Ascidians ae 7°00 19°60 2°60

Taste I.—Part 2
Determinations at end of 48 hours

1. Dissolved Oxygen used up
Milligrams Milligrams in 2,800 c.c. Milligrams used by

per litre. Animals in 48 hours.
(a) Control rs 702 19°66 —
(6) Sponge ie 6°46 18°09 1°57
(c) Ascidians oy 6°82 LS 0°56
2. Oxygen given out in form of Carbon dioxide
Milligrams Milligrams Milligrams excreted Respiratory
per litre. in 2,800 c.c. by Animals in 48 Quotient.
hours.
(a) Control oe 1°28 3°58 — —
(b) Sponge ses 2°40 6°72 o14 x1
(c) Ascidians ... 1°12 j14 0°44 Indeterminate

3. Oxygen required for oxidation of organic matter
Milligrams Milligrams in 2,800 c.c. Added to water

per litre. by Animals.
(a) Control ne 1°04 2°91 =
(b) Sponge os 1:84 4°15 1:24

(c) Ascidians ah 1°16 O20 0°34

i ll

zz

NUTRITION AND METABOLISM OF MARINE ANIMALS — 285

4. Ammonia determinations (a) ree and Saline, (6) Albuminoid

FREE AND SALINE. ALBUMINOID.
Mg. per Mg. Mg. excreted Meg. per Mg. Mg. excreted
litre. Total. by Animals. litre. Total. by Animals.
(a) Control ... O07 0°20 — 0°22 0°62 —
(b) Sponge... 0°38 0°92 0°72 1°20 3°36 2°74

(c) Ascidians ... 0°03 0°08 (12 ()°42 118 0°56

5. Constants for Animals (Weights, ete.)

Moist Weight Total Dry
in grms. Weight.
(a) Control] ae Nil —
(6) Sponge x 43°9 9°345
(ec) Ascidians... 11°3 1092

Remarks.—The chief points to be noted about this experiment
are:—(1) That there was an ample supply of water, because when
allowance is made for specific gravity, the volume of the Sponge would be
about 40 c.c., and that of the Ascidians 10 c.c., and as the volume of
water was 2,800 c.c., the Sponge had about 70 times its volume of water
and the Ascidians about 280 times their own volume of water. Also the
supply of oxygen was ample, as shown by the slight fall in the 48 hours.
With larger volumes it would have been impossible to estimate the
gaseous exchanges. (2) The very small amounts of oxygen required by
these two types of animals in the 48 hours, less than two milligrams
by the Sponge, and about half a milligram by the Ascidians, and the
very much greater output of carbon dioxide by the Sponge, the
amount for the Ascidians, although it is negative apparently, probably
hes within the mits of experimental error. (3) So far from using up
any dissolved organic matter, both types distinctly increase the
amount of organic matter in the water, and this is confirmed
by two quite independent methods, viz., oxidation by permanganate
and = determination of albuminoid ammonia. Permanganate
oxidation is not complete, and, on an average, multiplication of
the weight of oxygen used by 2°5 gives the weight of organic matter
present. In the case of the Sponge this gives 5°1 mgms., and for the
Ascidians 0°85 mgm. of organic matter excreted and added to the water
in the 48 hours. (4) In the complete oxidations of respiration, the
weight of oxygen required may approximately be taken as equal to the
weight of organic matter combusted. If the higher figure of the oxygen
in the carbon dioxide be taken, the Sponge shows about 3 mgms.

286 BIO-CHEMICAL JOURNAL

combusted. Similarly, taking the higher figure of oxygen consumed for
the Ascidians, the amount combusted is about 0°6 mgm. (5) Adding
together the organic matter under sources (3) and (4), the probable
amounts used in the 48 hours are: Sponge, 6 mgms.; Ascidians,
1-5 mgms. (6) The determinations given under Experiment V, for the
total plankton in the water used for the experiment, are very concordant
for the acid and alkaline oxidation, giving for oxygen used, 0°078 in
acid, and 0°077 mgm. in alkali per litre, taking 0°08 mgm. as a round
number and multiplying by the factor 2°5, the amount of organic
matter in the plankton becomes 0°2 mgm. per litre. (7) From these
data it can be calculated what volume of sea-water would require
to pass through each type. The amount works out at 30 litres
for the Sponge, and 7'5 litres for the Ascidians, in 48 hours, or
approximately 600 c.c. per hour for the Sponge and 150 c.c. per hour
for the Ascidians. Taking the volume of organism used as the unit, viz.,
40 c.c. and 10 ¢.c., this may be expressed by saying that in each case the
organism requires fifteen times its own volume of sea-water per hour.
It may be pointed out that this sample of water was the lowest of the
series in plankton, and on the other hand, that in the sea the animals

would possibly have been more active and had a higher metabolism.

Exrertmenr VIII. This experiment was also carried out with
Chamberland filtered water obtained from the same spot by the
Breakwater on April 19th. The details were as described in the previous.
experiment. The experiment was commenced at 4 p.m., April 19th, was
interrupted to take oxygen samples at 8 p.m. (i.e., in 4 hours), a small
volume of water was added to replace that removed, and the experiment
allowed to go on till next morning at 11 a.m. (i.e., total duration was
19 hours). Volume of filtered sea-water=2,800 c.c. in each case. The
results are shown in Table II.

TaBLE I1.—Part 1
Time-interval =4 hours
Dissolved Oxygen used up

Milligrams per litre Milligrams in Milligrams used
present at end. 2,800 c.c. by Animals
(a) Control ao T° 42 20°78 —
(b) Aplysia (6 specimens) 7°48 20°94 0°16
(c) Fusus ve 6°63 18°56 228
(d) Echinus ate 733 20°52 0°26
(e) Cancer o 1:07 2°99 17°79

NUTRITION AND METABOLISM OF MARINE ANIMALS — 287

Taste II1.—Vart 2
Determinations at end of 19 hours

1. Dissolved ¢ xygeu

Milligrams per litre Milligrams in Milligrams used by
at end. 2,800 c.c. Animals in 19 hours.
(a) Control at 7°94 - 22°23 —
(b) Aplysia a 6°37 17°84 “39
(c) Fusus cA 275 770 14°53
(d) Echinus “ae 6°76 18°95 v0
(e) Cancer as 0°98 2°74 19°49
2. Oxygen given out in form of Carbon dioxide
Milligrams Milligrams in Amount formed Respiratory
per litre. 2,800 c.c. by Animals. Quotient.
(a) Control ian kere 6°16 -- --
(b) Aplysia saa Oe 17°92 11°76 2°68
(c) Fusus Pen ia, 33°60 27°44 1°88
(d) Echinus Dae mS ko, 8°40 2°24 0°68
(e) Cancer HON eG 49°28 45°12 221
3. Ammonia excreted (a) Free and Saline, (4) Albuminoid
FREE AND SALINE ALBUMINOID
Mg. per Mg. Mg. excreted Mg. per Mg. Mg. excreted
litre. Total. by Animals. litre. Total. by Animals.
(a) Control .... O07 0°20 — 0°22 0°62 =
(b) Aplysia... 0°16 0°45 0°25 0°54 151 0°89
(c) Fusus MEST 2°44 2°24 1°10 3°08 2°46
(qd) Echinus ... O14 0°39 0°19 0°35 0°98 0°36
fe) Cancer»... 0-11 O31, O11 0°49 132 0-70
4. Constants of Animals (weights, etc.) in grams
Moist Weights. Total Dry Weights.
(a) Control —
(b) Aplysia ... 47°98 3°935
Soft Parts + Skeleton.
(c) Fusus ... 249°80 (115 grams soft parts) 20°448 + 134°317
(qd) Echinus ... 277°15 11°447 + 35°54]
(e) Cancer... 236°20 37-364 + 63°748

Remarxs.—(1) The most remarkable feature in this experiment is

the great consumption of oxygen by the crustacean, as compared with the

other types. Thus in the first four hours nearly 18 mgms. of oxygen are

288 BIO-CHEMICAL JOURNAL

used up, and the dissolved oxygen has been reduced to a very low level.
In the remaining 15 hours, the available oxygen was so little that only
2°30 mems. was used in this longer period, while the total output of oxygen
as carbon dioxide for the nineteen hours was 43°12 mgms., or more than
double the oxygen taken in during the same period. Similar results with
another species of crab of smaller size are seen in the next experiment,
and these findings seem to us of such importance that it is our intention
to make a special and more detailed examination of the matter. Both
animals, although very quiescent, were quite healthy at the end, and
recovered and became quite active when placed in ordinary fresh
oxygenated water. On account of the rapid fall in dissolved oxygen, it
is probable that the 17°79 grams used up in the first four hours do not
represent the full oxidation if the dissolved oxygen had not fallen, but
taking them as a basis for calculation, the hourly requirement would be
4°45 mems., and if the total plankton be again taken at 0°08 mgm. per
litre, this would require 55 litres of water completely cleared per hour,
or approximately 250 times the animal’s own volume per hour. This
would appear to demonstrate that such a type of animal can only live by
the capture of food in other ways. (2) It is again to be noted that the
animals add dissolved organic matter to the water in all cases, and that
there is accordingly no evidence of consumption of dissolved organic
matter by any of these types. (3) With the exception of Hchinus
esculentus, where the production of carbon dioxide and uptake of oxygen —
were both so small as to be hable to considerable experimental error, the
oxygen given out as carbon dioxide is much higher than the oxygen intake
from the water, so leading to such respiratory quotients as 1°88, 2°21, 2°68,
the possible meaning of this peculiar respiratory phenomenon is discussed
at the end of this section. (4) The greater output of albuminoid
nitrogen in the case of Fusus is in all probability due to the large amount
of mucin excreted by this type. (5) Calculations of the volumes of water
containing the amount of organic matter as plankton required, on the
basis of 0°2 mgm. per litre, yield the following results :—Aplysia,
3°7 litres per hour or 82 volumes per hour; Fusus, 10 litres per hour or
90 times volume of soft parts per hour; Cancer has already been given
at 250 times its volume per hour. These values would appear too great
to support purely plankton feeding.

EixperRIMENT IX. This experiment was carried out with Chamber-
land filtered water pumped from the supply to the Fish Hatchery. The
details were as described under Experiments VII and VIII. The

NUTRITION AND METABOLISM OF MARINE ANIMALS — 289

experiment was commenced at 12 noon, April 2Ist, the first part of the

experiment lasted 7} hours, the second 16} hours, making 24 hours in all.
All the animals were alive at the end, at noon, April 22nd, The results

are given in Table ITT,

TabiLe ITT.—VPart J

Time-interval =71 hours

Dissolved Oxygen used up

Milligrams

per litre.
(a) Control : 7°64
(6) Fish (Blennius) TTR
(c) Eehinoderm (Asterias) 72
(2) Molluse (Buecinum) Tol
(e) Crustacean (Eupagurus) .... 9 2°75

Milligrams in

9
ay

Taste IIT.—FPart 2

Determinations at end of 24 hours

1. Dissolved Oxygen used up

800 c.c.
21°3
21°62
16°02
21°67
770

Milligrams used
by Animals.

0°25
37
0°28
15°69

Milligrams Milligrams in Milligrams used
per litre. 2,800 c.c. by Animals.
(a) Control se we) Ak Oe MeO 20°55 —
(6) Fish (Blennius) ... eo: 11°14 9°41
(c) Echinoderm (Asterias) ... 1°22 3°42 1 Wee bs
(d) Molluse (Buccinum) ee ene, 16°21 4°34
(e) Crustacean (Eupagurus) ... 0°97 he 17°83
2. Oxygen given out as Carbon-dioxide
Milligrams Milligrams Milligrams Respiratory
per litre. in 2,800 c.c. excreted by Quotient.
Animals.
(a) Control Re. Da eo nleae 5°38 —
(6) Fish (Blennius) ... 2a dh OF 19°71 14°33 1°52
(c) Echinoderm (Asterias) ... 9°60 26°88 21°50 1°25
(d) Molluse (Buccinum) sais EOE: 10°75 87 1°24
25°09 1934 tig

(e) Crustacean (Hupagurus) ... 8°96

3. Oxygen required to oxidize organic matter in water at end

Milligrams

per litre.
(a) Control a ae ‘os ja Ores
(6) Fish (Blennius) —... ram Oa kee
(c) Echinoderm (Asterias) .... 1°52
(2) Molluse (Buecinum) <iny OO

(e) Crustacean (Hupagurus) ... 164

Mi

lligrams in

2,800 c.c.

1°06

oO.~

0 |

4°26
5°60
4°59

Milligrams added
by Animals.

290 BIO-CHEMICAL JOURNAL

4. Ammonia excreted (a) Free and Saline, (6) Albuminoid

FREE AND SALINE

Milligrams Mg. in Mg. excreted
per litre. 2,800 c.c. by Animals.
(a) Control ate ok 5. ING Nil —
(b) Fish (Blennius)_ ... OWT 0°20 0°20
(c) Echinoderm (Asterias) ... 0°20 0°56 0°56
-(d) Molluse (Buccinum) soa, 0202 0°06 0°06
(e) Crustacean (Kupagurus) ... 0°27 0°76 0°76
ALBUMINOID
Milligrams M.g.in Mg. excreted
per litre. 2,800 c.c. by Animals.
(a) Control °... ok ene 005 0°14 —
(b) Fish (Blennius) _... ay ee Ope 0°87 073
(c) Echinoderm (Asterias) ... 0°41 1:15 101
(d) Molluse (Buccinum) ees O65 0°98 0°84
(e) Crustacean (Kupagurus) ... 0°37 1°04 0°90

5. Constants of animals (weights, etc.) in grams

Moist Weights. Dry Weights of Soft Parts.
(a) Control on ane as a =
(6) Fish (Blennius) —... a Ogom 1918
(c) Echinoderm (Asterias) ... 80°60 5031
(7) Molluse (Buccinum) ... 49°68* 16°565
(e) Crustacean (Hupagurus) ... 34°34* 4°046

Wt. of Dried Organie Matter Ash in Skeleton.

Skeleton. in Skeleton.
(a) Control ee Bey: Bee — — _-
(6) Fish (Blennius)~ ... cg eae 1151 0°335
(¢) Echinoderm (Asterias) ae. LOPES 4°024 6°105
(d) Molluse (Buccinum) .. db'dd2
(e) Crustacean (Hupagurus) ... 5°757

*In each case without shell, the Mollusc shell weighed 56:88 grams, the shell inhabited by
the Hermit crab was not weighed. :

a

Orde

NUTRITION AND METABOLISM OF MARINE ANIMALS 291

TABLE LY

TRMPERATURE OF Ska ann Atr puRING EXPERIMENTS or Section B.

Sea. Air, Sea. Air.

April 16th, 1912 April 18th, 1912—

Morning ...... ou. OU, 8°35" C, Morning... 80°C. 86°C.

Afternoon ... 8°6°C. 11°4°C. Afternoon ... 8°6°C. 9°4°C.
April 17th, 1912 April 19th, 1912—

Morning ... 8°0°C. 80°C. Morning’ ©... "8:370,.. O02 0;

Afternoon ... 8°6°C. 10°0° C. Afternoon ... 8°3°C. 10°8° C.

Sea. Air.

April 20th, 1912—
Mornme ....8:0°C. 9:4°C.
APEINOOH...2o00 C. 11-6-'C:

TABLE Y

CoMPOSITION OF THE Sorr PARTS OF CERTAIN OF THE ANIMALS
USED IN SECTION B.

After removal of the skeleton the soft parts were dried and the
percentage of mineral matter and of organic matter, and in the latter
of fat and protein were determined. It was found impossible without a
great expenditure of labour to estimate carbohydrates on account of the
interference of the proteins. All figures are in percentages of total dried
weights, it may be especially emphasized that this is so in the case of fat
and protein which are not percentages of the organic matter. Attention
may be called to the high percentage of inorganic matter even in the soft
parts of these animals. Fat means total ethereal extract, and proteins
were obtained by estimating the total nitrogen by Kjeldahl’s method and
multiplying by the factor 6°25.

Ash, Organic Matter. Fat. Protein.
Sponge ae ae 99°54 40°46 2°75 18°8
Ascidians ... be, wa 46°43 93°57 2°35 28'8
Aplysia nce cM aia Soon 72°42 2°87 370
Fusus yt an ene Zag 87°51 8°47 52°6
Echinus ... ee OUP to 39°25 1°95 78

Cancer ee Pris ne 10°06 89°94 14°06 50°7

292 BIO-CHEMICAL JOURNAL

Tur RESPIRATORY QUOTIENT IN INVERTEBRATES

In nearly all the cases shown in the tables above, the respiratory
quotient is greater than unity; that is to say, the volume of carbon-
dioxide produced is greater than that of the oxygen used up. In many
cases the over-production of carbon-dioxide is very marked, and in the
case of the two crustaceans in the later stages of the experiment, scarcely
_ any oxygen is taken up, while a considerable evolution of carbon-dioxide
continues.

Such a process is also found in the ripening processes of seeds and
fruits, and has been observed in animals well fed with carbohydrates
which are in process of fattening.

A respiratory quotient greater than unity is impossible if complete
oxidation of foodstuffs takes place, and such a quotient is an indication
of the linking together of a reducing and an oxidizing agency, whereby
a smaller quantity of a substance of higher chemical energy potential is
formed, so setting free oxygen which is used in oxidizing more material,
as a result of which energy for the use of the organism is set free without
any intake of oxygen.

If carbohydrate be completely oxidized, since hydrogen and oxygen
are present in the proportions to form water, a volume of carbon-dioxide
is set free exactly equal to that of the oxygen used for combustion, so
that the respiratory quotient is unity.

If fat be completely oxidized, since there is a large excess of
hydrogen, this uses up a great deal of oxygen which appears as the
end-product, water, and as a result the carbon-dioxide formed is
considerably less than the oxygen used. As a result, the respiratory
quotient falls to 0°5, or even less. Proteins lie intermediately, but much
nearer to the carbohydrate, as there is very little hydrogen excess.

As the energy yielded weight for weight by hydrogen in undergoing
combustion is much higher than that given by carbon, it follows that the
heat value of the fats is much greater than that of the carbohydrates or
proteins for equal weights.

In the building up of fats, proteins or carbohydrates in the organism,
these processes are reversed, and, if, as commonly happens, fat is
synthesized from carbohydrate, twice the weight of carbohydrate at least
must be used up to yield a given weight of fat. The process is one of
reduction, and if the end members of the molecular groupings be
neglected, the carbon atom of the carbohydrate may be represented as

H—C—OH, while that of the fat may be written as H—C—H, there is

|

ee

ee

NUTRITION AND METABOLISM OF MARINE ANIMALS 293

hence an atom of oxygen set free from each carbon atom, which may be
utilized to oxidize carbon down to carbon-dioxide in another carbohydrate
molecule, and so yield energy for the formation of the fat. If this process
be aided by a little oxygen from outside, still more carbohydrate may be
oxidized, and the net result is that the exo-thermic or carbohydrate
oxidizing reaction runs the endo-thermic or fat synthesizing reaction,
and a surplus of energy is left over for the uses of the animal.

When such a linkage of reactions has occurred, the respiratory
quotient may much exceed unity in value, and there is no doubt that,
even in mammalia, when fat is being produced from carbohydrate, such
a condition of affairs does arise.

It was observed by Gautier,! in 1881, that even warm-blooded animals
lived in part in an anaerobic fashion, producing carbon-dioxide without
equivalent uptake of oxygen, so obtaining a portion of oxygen not from
the air but by fermentative processes from oxygen-containing organic
bodies. Gautier also indicated that a portion of such oxygen might arise
from fat formation from carbohydrate.

Hanriot,? in 1892, showed that after heavy meals of carbohydrate the
respiratory quotient might rise as high as 1°30, and attributed this to
conversion of carbohydrate into fat. Hanriot illustrates how such an
effect might arise by the equation : —

18C,H,,0, = C,H

1 ee)

O, + 23C0, + 26H,0.

1046

In this equation the formula C..H,,,O, represents a neutral fat
containing one molecule each of oleic, palmitic and stearic acid.

Thus Hanriot was the first observer clearly to put forward the view
mentioned above in explanation of respiratory quotients greater than
unity.

In the course of work somewhat similar to our own, upon the
respiration of invertebrates, Vernon? found respiratory quotients greater
than unity, and also, like us, found that certain of these animals could
continue living with a very minimal oxygen supply, and during this
period produced large amounts of carbon-dioxide, so that the respiratory
quotient in one case rose to over four.

Vernon attempted to drive all the carbon-dioxide out of the water
by heating in a vacuum, and as completeness here is almost impossible,

1. Comptes rendus CXLV, p. 374, 1892.

ae Lod... po dt1.
3, Journ, of Physiol., XIX, p. 18, 1895,

294 BIO-CHEMICAL JOURNAL

it is probable that his results ought to be even higher than they are.
Vernon ascribes his results to the animals being in a _ pathological
condition and moribund, but since, like our own, most of them recovered
when fresh oxygen was admitted we scarcely can agree in this view.
Also, in our case we have obtained similar results long before there was
any real dearth of oxygen in the water.

Athanasiu,! in 1900, found a mean respiratory quotient of 1°24 in
frogs, which were not feeding during the experiments, and concluded
that frogs stored up oxygen in their tissues for use during such a period of
hibernation.

Pembrey,? on the other hand, found an exceedingly low respiratory
quotient in the hibernating marmot, but a quotient of well over unity
(viz., 1°24, 1:28, and 1°39) during the feeding and fat-storing period in the
same animal. —

Pembrey agrees with Hanriot as to the explanation being the
conversion of carbohydrate into fat.

It is somewhat difficult to believe that fat storage can occur during
starvation, except for a short period, and it may be that the real
explanation in our own experiments may be as suggested by Athanasiu,
some oxygen storage in the tisues in some form at present unknown to us.
Certainly the amounts of fat present at the end, as shown by Table VI,
are not very high; and this, so far as it goes, indicates that either some
reduced substances other than fat were formed, or that there is normally
a storage of oxygen in some organic form which was utilized during the
period of experiment.

Similar results were obtained by Piitter,? by keeping leeches in an
atmosphere free from oxygen. This observer traces the results to the
substitution of hydrolytic cleavages for oxidations, but the amount of
energy obtainable in this way is very small.

Granting, however, that such a linkage of reactions can occur, as is
premised above, then the effect might be shown by such an equation as
the following :—

/C,,H,,COO
190,H,,0, + 8C,H,,0, = 2C,H; ~C,,H,,C00 + 48C0, + 54H,0.
\C,,H,,COO

1. Journ. de Physiol. et de Path. gen., Il, p. 243, 1900.
2. Journ. of Physiol., Vol. XXVII, p. 407, 1902.
3. Zeittsch. f. allgem. Physiol., Bd. VII.

NUTRITION AND METABOLISM OF MARINE ANIMALS = 295

The two amounts of carbohydrate on the left-hand side are written
separately to indicate that nineteen molecules go to form the two molecules
of neutral fat shown on the right-hand side, while the eight additional
molecules are completely oxidized in combination with the oxygen set free
in the fat formation to form carbon-dioxide and water, so yielding more
than enough energy to run the reaction exo-thermically.

The weight of carbohydrate expressed in gramme-molecular weights
is 4,860 grammes, the weight of fat formed is 1°776 grammes, carbon-
dioxide is 2,112 grammes, and water 972 grammes.

Taking the relative caloric values of fat and carbohydrate as 9 to 4,
the 1,776 grammes of fat are equivalent in energy to 3,996 grammes
of carbohydrate, and subtracting this from the original weight there is
the energy of 864 grammes of carbohydrate to run the reaction exo-
thermically, or about 32,000 large calories.

This may be stated otherwise thus:—Without any oxygen from
without, carbohydrate can yield about one-third of its weight of fat, and
at the same time yield a supply of energy equivalent to about one-sixth of
that total amount which it would yield if directly and completely
combusted to carbon-dioxide and water.

A parallel case of utilization of carbohydrate energy by partial
reduction and partial oxidation, linked together by the agency of a living
cell, is the formation of alcohol from glucose by the yeast cell.

All degrees between these two cases (where the respiratory quotient
is, of course, infinity) and those exemplified above in invertebrate animals
with respiratory quotients lying between 1 and 2, can be obtained by
varying intakes of oxygen.

If, for example, in the above equation eight gramme molecules more
of carbohydrate were taken in on the left-hand side, and forty-eight
gramme molecules of oxygen for their complete oxidation, there would
then appear a total of ninety-six gramme molecules of carbon-dioxide on
the right-hand side, and the carbohydrate would now yield only about
one-fourth of its weight for storage as fat. Under such conditions the
respiratory quotient would be 2 exactly, and there would be a large amount
of energy for heat and external work.

It is most interesting to note that these considerations indicate fat
formation in the tissues under conditions of diminished oxidation, which
is in accord with physiological facts.

290" - BIO-CHEMICAL JOURNAL

An animal with a rich reserve of carbohydrate might also actually
increase its fat content during inanition and inactivity, for a short period,
until carbohydrate became exhausted. Such a process as this actually
does occur in the ripening of many fatty seeds.

Our thanks are due to Mr. A. Webster for much valuable assistance |
in the practical work, and especially for the determinations contained in

Table V.

297

THE EFFECTS OF ATMOSPHERES ENRICHED WITH OXYGEN
UPON LIVING ORGANISMS, (a) EFFECTS UPON MICRO-
ORGANISMS, (b) EFFECTS UPON MAMMALS EXPERI-
MENTALLY INOCULATED WITH TUBERCULOSIS, (c)
EFFECTS UPON THE LUNGS OF MAMMALS, OR OXYGEN
PNEUMONIA

By ALFRED ADAMS, M.B., Ch.B. (Liverpool).
From the Bio-Chemical Laboratory, University of Liverpool
(Received May 22nd, 1912)

Section A. Tur Action or OXYGEN UPON MICRO-ORGANISMS

In the course of two previous investigations, published by Moore
and Williams!, in 1909 and 1910, it was shown that while certain micro-
organisms are indifferent to amounts of oxygen lying between the
atmospheric value and a full atmosphere of oxygen at ordinary barometric
pressure, others, notably Bacillus tuberculosis and Bacillus pestis, are
very sensitive to oxygen, and cease to grow when the percentage of oxygen
exceeds about seventy per cent. of 760 millimetres.

It is remarkable how few organisms show this oxyphobic property,
and how specific and definite the effect is when the property is present.

Out of twenty-six various organisms tested im a partial pressure of
oxygen of 500 millimetres of mercury and over, ouly two showed complete
stoppage of growth, viz., the bacilli of plague and tuberculosis, while
those of the staphylococcic group were adversely aftected and grew less
rapidly than normally, and on some occasions failed to grow entirely,
viz., Staphylococcus pyogenes aureus, and albus. All the remaining
twenty-two organisms grew as well, so far as could be observed, in the
increased oxygen as in air.

In the present investigation, these experiments have been repeated
and amplified, and two additional and interesting members have been
added to the oxyphobic group, namely, actinomyces and mycetoma from

the streptothrix family.

1. Btio-Chemical Jowrnal, Vol IV, p. 177, 1909; Vol. V, p. 181, 1910,

298 BIO-CHEMICAL JOURNAL

It was noticed by Moore and Willams that B. pestis was only
arrested but not killed by exposure to the oxygen, for although no growth
developed after some days’ keeping in the enriched atmosphere, on
afterwards transferring to ordinary air a good growth soon appeared.
On the other hand, these observers found it impossible to revivify or
sub-culture from B. tuberculosis which had been kept for a period of
from two to three weeks in an atmosphere of 70 per cent. of oxygen, or
over that amount.

In the present work it has been found that B. tuberculosis, while
arrested completely in oxygen as they described, was found to grow
again when transferred to air, as shown below. It was found in
confirmation of their work that B. pestis was completely arrested during
a stay in the oxygen-enriched air, but commenced to grow again as soon
as the culture tubes were replaced in ordinary air.

It was otherwise, however, with the streptothrix organisms, which,
in both instances, failed to grow again after exposure to the higher
percentage of oxygen.

All experiments were carried out at atmospheric pressure of the total
gases, the partial pressure of the oxygen being increased by complete or
partial evacuation of the bell-jar under which the tubes containing the
culture tubes were placed, so as to get varying percentages of oxygen.

The oxygen used was prepared most carefully in the laboratory from
potassium permanganate, as the gas sold in the usual cylinders is
contaminated with a good deal of nitrogen. It was most carefully
washed, and stored until required in a large gas holder capable of holding
about 100 htres.

The tubes with the cultures to be exposed to additional oxygen were
placed in a large bell-jar fitting on to a ground glass-plate, and it was
found necessary, on account of the long duration of each experiment, to
prevent slow leakage by a mercury trap outside the flange of the bell-jar.
The mercury was held in position by a circular mound of putty, and no
mercury penetrated under the bell-jar. The bell-jar was evacuated after
placing the desired inoculated culture tubes within, and then oxygen to
the desired amount was allowed to enter, and analyses of the percentages
of oxygen at beginning and end of the experiment were carried out by the
usual gas analysis methods.

The control tubes were kept under similar conditions under a lke
bell-jar but in ordinary air, and both bell-jars were placed alongside of
each other in the same incubator at 37° C.

THE EFFECTS OF ATMOSPHERES 299

The enriched oxygen atmosphere was entirely without effect on the

following organisms :—

Bacillus coli communis Bacillus pyocyaneus
‘5 typhosus 55 of hog cholera
* dysenteriae (Flexner) i diphtheriae
bs A (Kruse ) a anthracis
4 mallei Vibrio cholerae
: gaertner Sarcina lutea
~ prodigiosus .: aurentiaca
- pheumoniae Oidium albicans

Proteus vulgaris

The staphylococcic group showed very fair growth, but not so
marked as in the controls. The results with B. pestis were as stated
above. B. tuberculosis showed a very small but definite growth at the
end of a prolonged experiment lasting seven weeks. The organism was
not killed but only inhibited by its treatment, and was able on further
incubation in air to overcome to a slight degree the adverse circumstances
under which it had been placed, and sub-cultures then made from it
developed normally.

Actinomyces and mycetoma, on the other hand, were not only
inhibited but killed, no growths appearing after the oxygen treatment.

These results are shown in the following protocols : —

Experiments I, II and 1/1.—Six tubes of Glycerin-acid-agar were
inoculated with a laboratory strain.of Bacillus Tuberculosis (Iluman),
and three were placed in an atmosphere of 90 per cent. of oxygen and the
controls in the air chamber.

Results :—At the end of three weeks’ incubation of the two sets
under similar conditions, the three controls showed good growths and the
tubes in oxygen, in which the percentage had diminished to 70 per cent.,
showed no appreciable growth.

The tubes were replaced in the oxygen chamber with oxygen of
85 per cent. strength and incubated for a further four weeks, when the
oxygen had diminished to 68 per cent., and there was an appreciable
amount of growth. The oxygen was renewed and incubation continued
for a further two weeks, when the growth was about equal in amount
to the growth at the end of three weeks only in the control tubes, but
it was of a dark brown colour and consisted of raised discrete masses
differing from the uniform growth in the controls.

300 BIO-CHEMICAL JOURNAL

One of the tubes was continued in the oxygen and three weeks later
showed a very slight increase im growth of the same character.

Another was placed in the air bell-jar, and after three weeks showed
a much more profuse growth, but still of the dark brown colour.

The remaining tube was sub-cultured, and incubated in the ordinary
manner, and three weeks later, growths, normal in colour and other
respects, were obtained.

Specimens from the various tubes were now stained and submitted to
microscopical examination, but no appreciable difference could be
discovered in the smears.

These results were confirmed later in all respects by two additional
similar experiments.

Experiments LV and V.—Three agar tubes were inoculated with each
of two different strains of actinomyces, and two tubes of each strain were
placed in the oxygen chamber, which was then filled with oxygen and the
percentage determined, and found to be 87 per cent. The remaining
tubes were placed as controls in the air chamber.

After four days’ incubation the control tubes showed a shght growth
as minute white specks, whilst the tubes in oxygen showed no growth.
The tubes were replaced in their respective chambers, and seven days
later again examined, when the controls showed a good growth. There
was no trace of growth in the tubes kept in oxygen.

A further seven days’ incubation in the oxygen chamber was tried,
without any trace of growth, and the tubes were then transferred to the
ordinary incubator. Fourteen days subsequently there was no trace of
growth in either of them.

This experiment was again repeated, with similar results
throughout.

Experiments VI and V//.—Three agar tubes were inoculated with
mycetoma from the original strain, and two of these were placed in the
oxygen chamber, and one in the air chamber for control. Seven days
later, on, careful examination, the control tubes showed a growth, but
the tubes in oxygen showed no trace of growth. After incubation for
an additional seven days, under the same conditions, they were again
examined, and the controls showed a good growth, while again the
tubes in oxygen gave no growth.

This was repeated for a third week with a like result, and the oxygen
tubes were then transferred to the ordinary incubator and examined
fourteen days later, when no evidence of growth could be found in the
tubes.

as

THE EFFECTS OF ATMOSPHERES 301

This experiment was then repeated twice with similar results.

It having been suggested to me by Professor B. Moore that the
inhibiting effect of the oxygen on the Bacillus tuberculosis in vitro, as
well as the well-known effects of fresh air on tubercular patients, might
be due to the presence of traces of oxides of nitrogen in the atmosphere,
the following experiments were carried out.

Heperiment VI1I,—Suitable culture media, in tubes as for the

previous experiments, were inoculated with the following organisms :—

Bacillus tuberculosis Actinomyces
‘ pestis Mycetoma
"4 typhosus Pneumococcus
» diphtheriae Staphylococcus aureus
» coli communis Streptococcus pyogenes
E anthracis Staphylococcus albus

The tubes to be tested were placed in a similar bell-jar to that used
in the previous experiments, and this was exhausted to a pressure of
ten millimetres of mercury. <A sufficient quantity of nitric oxide gas,

ras prepared by action of nitric acid on copper, and washed and found
to contain 90 per cent. of nitric oxide, was added so as to form 0°5 per
cent. of the total capacity of the jar. Atmospheric air was now admitted,
thus ensuring the entrance of the nitric oxide dilution into the tubes,
and the jar was placed in the incubator, accompanied by the controls
in a similar jar, but containing only atmospheric air.

Twenty-four hours later there was no appreciable difference in the
results obtained from the two sets of tubes.

The incubation was repeated for one week, when there was no
apparent difference between the controls and those tested, and at the
expiration of three weeks the tubes containing the more slowly growing
organisms, viz., Bacillus tuberculosis, Actinomyces, and Mycetoma, were
identical in amount of growth.

Experiment IX.—The following organisms were inoculated on

suitable media, viz. : —

Bacillus coli communis Bacillus diphtheriae
», typhosus », anthracis
if dysenteriae (Flexner) 7. pestis
_ = (Kruse) Staphylococcus pyogenes aureus
a mallei ss 5 albus
«s gaertner 9 a citreus
»» prodigiosus Vibrio cholera
a pheumoniae Sarcina lutea
» pyocyaneus 5, aurentiaca

Proteus vulgaris Oidium albicans

302 BIO-CHEMICAL JOURNAL

One set of the inoculated culture tubes was placed in the oxygen
chamber, and by evacuation and introduction of oxygen the oxygen
concentration was raised to 90 per cent. <A series of controls in the
ordinary air bell-jar in the same incubator were compared at the
expiration of twenty-four hours with the tubes being tested in the
oxygen. On examination, Bacillus pestis showed no growth on the
oxygen tube, while the control tube showed a commencing growth, and
two days later the control showed a good growth but the oxygen tube
showed nothing.

Staphylococcus pyogenes aureus on the oxygen tube gave a very
minute growth, while the S. albus and S. citreus showed no growth. All
three of the controls of the staphylococci gave good growths.

All the remaining organisms gave practically identical growths on
the oxygen and control tubes.

The tubes of the Bacillus pestis and the staphylococci were replaced,
and again examined in a_ further forty-eight hours, when the
staphylococcus tubes all showed growth, but definitely less in those
exposed to the oxygen than in the controls. The B. pestis was quite
free of growth in the oxygen tube, the control showed good growth.

CONCLUSIONS

1. That the action of oxygen in increased percentages on the
Bacillus tuberculosis is to arrest its growth, but the action is not
bactericidal but inhibitory, as in the case of those other oxyphobic
micro-organisms, Bacillus pestis, and in a lesser degree the members of
the staphylococci group.

2. That an atmosphere enriched with oxygen not only arrests the
growth of actinomyces and mycetoma but kills these organisms. This
may be the explanation of the differences in the literature as to the
description of actinomyces, which some authors describe as aerobic,
while others state that it is anaerobic. The probability is that it hes on
the border line; a small increase in oxygen turns the scale and prevents
its growth.

3. That the action of the oxygen is not due to oxides of nitrogen

contained as an atmospheric impurity is shown by Experinient VIII.

a

THE EFFECTS OF ATMOSPHERES 303

Secrion B. Experiments upon MamMMAuLs EXprerIMENTALLY INOCULATED
with TUBERCULOSIS

The effects of oxygen in stopping the growth of the Bacillus
tuberculosis obviously suggests an attempt at treatment of the disease
by respiration of oxygen, and with this object in view Moore administered
oxygen of about 80 per cent. at ordinary atmospheric pressure to three
patients with pulmonary tuberculosis for one hour daily over a period
of some weeks. The oxygen inhalation was enjoyed by the patients, but
did not show any clear effect upon the course of the disease.

It was thought that an extension of the period of oxygen
administration, by keeping animals for a prolonged period in oxygen in
a respiratory chamber, might have an effect, and accordingly guinea-
pigs were introduced into the chamber described below after
subcutaneous inoculation with a strain of Bacillus tuberculosis, and
retained there for some days.

Unfortunately it was found that with the more prolonged periods in
oxygen, that peculiar form of simple exudative pneumonia, which has
been termed oxygen pneumonia, made its appearance when over 70 per
cent. of oxygen was employed, and below this level no action on the
development of the bacilli as shown by arrest of the tubercles was
obtainable.

The lung effects are described in Section C. For the sake of
completeness, the history of the inoculation experiments is here recorded.

A suitable dose for inoculation which would show definite
pathological lesions in a guinea-pig in a reasonable time was determined
after a considerable number of experiments with various strains. This
dose ranged from 0'1 to 10 milligrams, according to the virulence of the
particular strain, and was so selected that definite and fairly advanced
lesions appeared in about six weeks. The injections were made
subcutaneously in the thigh. A number of animals were inoculated at
the same time, and of these a portion were kept in the animal house as
controls, while the rest were placed in the oxygen respiratory chamber.

The oxygen used was prepared by heating permanganate of
potassium in an iron retort. After all the air had been expelled, and a
steady stream of oxygen was coming off, this was conveyed through wash-
bottles into a gasometer of a capacity of 100 litres, with suspension weights
so adjusted that there was a constant pressure sufficient to cause the gas
to bubble through a wash-bottle placed between the gasometer and the

804 BIO-CHEMICAL JOURNAL

respiratory chamber, where it was stored for use, a sufficient quantity
for four or five days being made at once.

By this means an oxygen of a strength of 95 to 96 per cent. was
obtained, the remaining 4 to 5 per cent. consisting of nitrogen. The
solid and soluble impurities were removed by the washing process.

The respiratory chamber used consisted of a cylindrical vessel
_(see fig. 1) made of galvanized iron, with a conical bottom ending in a

A Respiration Chamber. B WaterSeal. C Soda LimeTray. D Excreta Outlet. E Gasometer.
F Wash Bottle. G Indicator Wash Bottle. H Outlet to Vacuum Pump.

rubber tube which dipped into a vessel of water for about three centi-
metres, and served to convey away the urine. Its roof was formed by a
glass dome,! made airtight by a circular water-seal in the tinware
under-portion. This dome was about fifteen centimetres in diameter,
and through it the animals could be observed during the experiments.
The capacity of the chamber was about forty litres.

The animals rested on a floor formed by a perforated zine plate, and
suspended round the chamber was an annular gallery filled with soda
lime, of which it contained about 500 grams when loosely charged. The
inlet and outlet for the gases to the chamber were situated just above
the floor, the former attached by indiarubber tubing to the wash-bottle
from the gasometer, and the latter to a small vacuum pump, with a
thumb-screw on the tube by which the flow could be easily regulated,
and a small wash-bottle served as an indicator of the rate of flow.

1. The glass-cover was such as are made for protecting memorial flower wreaths.

een 9 te rhe wet.

AL AMP

THE EFFECTS OF ATMOSPHERES 305

The inoculated guinea-pigs were placed in this chamber, with an
adequate supply of food for three days. Three pigs were used at each
test, it being found impossible to store food for more for the required time
in the chamber, owing to their increased appetites, due possibly to the
raised oxygen percentage. The outlet was then attached to the vacuum
pump and the inlet to the oxygen container and the chamber flooded with
oxygen until the required percentage was obtained, as shown by analysis.

Specimens of the gas were obtained by disconnecting between the
vacuum pump and the wash-bottle indicator, and tested in Hempel’s
bulbs for carbon-dioxide and oxygen. Once the required concentration was
reached, it was found that a flow of about twenty bubbles per minute from
a drawn-out tube of about one millimetre diameter through the wash-
bottle was sufficient to preserve the standard. This outflow is necessary
because it is impossible to prepare oxygen quite free from nitrogen in
such large quantities.

Specimens of the gas were taken morning and evening for the first
week, and later daily, and by regulating the thumbscrew on the exhaust,
it was found that fairly constant oxygen percentage could be retained,
the variation not exceeding five to ten per cent. over the periods between
the changing of the food supply. The carbon-dioxide percentage with
fresh soda-lime did not rise above three per cent. and was usually much
lower. Any excess over this showed the necessity of a fresh supply of
the absorbent.

At the end of each period of three days, the chamber was cleaned out
and the food supply renewed. It was found advisable on each occasion
to wash it out thoroughly with a dilute solution of Lysol, and the use of
this preparation on the glass dome was most advantageous in preventing
condensation of moisture in drops, thereby obscuring vision.

The controls and the animals undergoing the tests received alike the
same food. Bran moistened with boiling water, carrots, and straw were
the staple articles of diet.

The percentage of oxygen used in the tests was regulated by the
amount the animals were able to bear. It was found, as already stated,
that a continuous exposure for prolonged periods to a concentration greater
than 70 per cent. caused death by pneumonia. Below this strength the
animals seemed to be very little affected by their surroundings; the only
observable difference being that they ate much more greedily, and on
increasing the concentration they were for a time much more lively, but
soon settled down to their normal condition again. The sudden increase

306 BIO-CHEMICAL JOURNAL

from the normal atmospheric percentage of oxygen to the increased
concentration used, showed no ill effects, and on opening the chamber
for cleansing purposes, the guinea-pigs, after their three days in the
stronger oxygen, seemed unaffected bv the change back to atmospheric
air, and greedily attacked the fresh food supplied to them.

Experiment X.—A number of guinea-pigs were inoculated with a
virulent strain of Bacillus tuberculosis (Human) under the skin of the
thigh. Three, weighing together 1250 grams, were placed in the
respiration apparatus, and the oxygen passed in until a sample of the
gas withdrawn at the exhaust showed a strength of oxygen of 64 per cent.
The thumbscrew on the exhaust was then screwed down so as to allow
only a small bubble at a time to pass. The suspension weight on the
gasometer was so regulated for this purpose, that the water-seal round
the glass dome showed the water to be about one centimetre lower inside
the dome than outside.

The remaining guinea-pigs were placed as controls in the ordinary
animal house.

Next morning an analysis of the gas gave Carbon dioxide, nil;
Oxygen, 60 per cent.

The current from the exhaust was slightly increased, and two days
later an analysis gave Carbon-dioxide, 1°4 per cent.; Oxygen, 59 per
cent.

The weight of the guinea-pigs on taking out showed an increase to
1,345 grams. They were again placed in the oxygen chamber, and the
oxygen percentage raised to 66 per cent. The gas analysis next day gave
Carbon-dioxide, 0°8 per cent.; Oxygen, 60 per cent.

The succeeding day, the percentages were Carbon-dioxide, 1°2 per
cent.; Oxygen, 61°0 per cent.

The chamber was again cleansed next morning and the experiment
restarted. This course of treatment was continued during the whole
period with short intermissions for cleansing, the percentage of oxygen
not rising above 68 per cent. or falling below 50 per cent., and the
Carbon dioxide not rising above 2°8 per cent. on any occasion.

During the third week one of the guinea-pigs in the chamber died
of natural causes. Small glands in the groin were found, post mortem,
also Bacillus tuberculosis at the site of inoculation.

The lungs were healthy. This animal was replaced by one of the
controls in the latter part of the experiment.

At the end of the fifth week, the guinea-pigs were showing signs of

thang

THE EFFECTS OF ATMOSPHERES 307

loss of weight, and were killed and examined and compared with the
controls kept in the animal house.

All three animals kept in the oxygen chamber showed definite
tubercular lesions, consisting, in two cases, of sinuses discharging pus,
enlarged caseating glands in the groins and axillae, and general
emaciation. ‘The controls also showed definite lesions of a similar
character.

No definite difference could be observed between the stages of the
lesions in the controls and treated guinea-pigs, and it was impossible to
say that either was the more advanced or serious.

A number of other experiments were carried out under the same
conditions, but none were continued for so prolonged a period, owing,
usually, to accidental death of the tested animals, but all the results
pointed to the same conclusions.

CONCLUSIONS

1. That the inhibiting effect of the increased oxygen percentage on
the growth of the Bacillus tuberculosis in vitro cannot be extended to
treatment in vivo.

2. That the beneficial effects of fresh air and ventilation on strumous
diseases are not due to the small increase of oxygen over that of the
interiors of dwellings, or to any chemical condition of the air, but rather
to some physical stimulus.

Section C. ACTION UPON THE MAMMALIAN LUNG oR OXYGEN PNEUMONIA

As stated in the previous section, it was found that while oxygen up
to a partial pressure of about 500 millimetres of mercury could be readily
tolerated for days together without any observable ill effects, a slight
increase to about 550 millimetres always led to the animals perishing
from a condition resembling lobar pneumonia. The limit bearable
without injury for prolonged periods thus lies just below seventy per cent.
of oxygen in an atmosphere at 760 millimetres of total pressure.

This pneumonic effect has been observed by several previous
observers, but in most cases it has been studied in cases where the total
pressure has been raised to many atmospheres, and the duration of the
experiment has been a short one,

308 BIO-CHEMICAL JOURNAL

As the subject has some importance now that oxygen is becoming
popular in the treatment of disease, and also has relationship to caisson
work, rescue work with oxygen apparatus where oxygen may be breathed
for long intervals and similar uses of the gas for respiration in confined or
vitiated spaces, it was determined to experiment more fully and find the
limits of causation, show freedom from any bacterial causation, and
obtain micro-photographs showing the nature of the lesions and the
minute pathology. It was found by several observers in early days after
the isolation of oxygen that respiration of the gas produced inflammatory
effects. The respiration and circulation were both found to be increased,
in time the lungs became congested, and in certain cases inflammation
and death occurred. The adherents of the school of Lavoisier set these
effects down to the effects of increased pulmonary combustion, but this
was controverted by Regnault and Reiset, who showed that no increased
oxidation took place. It was Bert who introduced the fact that the
pathological effects due to high pressures were dependent on the partial
pressure of the oxygen only, by demonstrating that as much effect was
obtained with four atmospheres of pure oxygen as with twenty
atmospheres of compressed atmospheric air.

Bert worked chiefly with very high pressures, and found that these
induced a status epilepticus in birds, and also in mammals at a higher
pressure. He compared the effect to that of tetanus or strychnine
poisoning, and thought it due to a toxic action on the nervous system
either directly by the oxygen or indirectly by some nerve intoxicant
formed by the action of the oxygen on the blood. The symptoms persisted
for hours after brief exposures to high oxygen pressures, the animals
remaining in an epileptic condition and having fits of tetanic nature at
short intervals. Blood transfused from such animals did not, however,
induce the condition in normal animals.

The animals in Bert’s experiments succumbed to the nervous effects
before the lungs became pneumonic, and he does not appear to have
investigated the effects of lower concentrations of oxygen acting for
longer periods. After a considerable interval, this was taken up by
Lorrain Smith!', who also worked with a pressure chamber, but used
much lower pressures than Bert. In Bert’s experiments the convulsions
occurred at about 3 atmospheres of oxygen with birds, and 4 atmospheres
with dogs; in Lorrain Smith’s work while some experiments were carried

1. Journal of Physiology, Vol. XXIV, p. 19, 1899. We are also indebted to this paper for
the materials for the brief resumé of the literature given above.

—————————————E—————————— ee eee eee

ee

Wie iieas

THE EFFECTS OF ATMOSPHERES 309

out at 3 atmospheres or slightly higher, the greater number were carried
out at a level lying between 1 and 2 atmospheres of oxygen, and two were
performed at pressures between 70 and 80 per cent. of an atmosphere of
oxygen.

An important point made out by Lorrain Smith was that previous
exposure to an intermediate pressure (1°4 atmosphere) acted to a certain
extent as a preservative against the onset of convulsions, which did not
occur at the later stage until a higher concentration for a longer
interval. This is ascribed by Lorrain Smith to the previous lower
exposure reducing or upsetting the normal secretory activity of the lung
epithelium, which when uninjured causes the blood to be more highly
and rapidly charged with oxygen.

In the first experiment recorded by Lorrain Smith, the percentage
of oxygen was only 41°6 of an atmosphere, and this caused no injury
within eight days, our own results confirm this finding. In his second
and third experiments the percentages were 73°6 and 79°9 respectively.
In each two mice were employed, and in each one succumbed in four
days with more or less marked congestion of the lungs, while two of the
animals survived for a week.

All the other experiments recorded in Lorrain Smith’s paper are
with amounts of over an atmosphere of oxygen, and more results seemed
to us desirable between 70 per cent. and one atmosphere of oxygen.

The fact of working with a small pressure chamber precluded also the
use of large animals, and with the exception of one short experiment at a
fairly high pressure on a guinea-pig, and a brief one on a rat, all the
experiments recorded by Lorrain Smith were made upon mice or small
birds.

The pneumonia produced in these experiments with over one
atmosphere of oxygen pressure, as described by Lorrain Smith, was
identical with that described in this paper. Bacteria in many cases
were entirely absent. The alveoli were filled with an exudate which was
granular and fibrillated in appearance, but did not give the fibrin test by
Weigert’s method. There were no leucocytes in the exudate. The
pneumonic condition was universal, and could be compared only with
the earliest stages of catarrbal pneumonia.

The onset of the pneumonia and time of death, varied with the
percentage of oxygen, as stated above in his two experiments with
between 70 and 80 per cent., death occurred, in two cases out of four only,
in 4 days; at 120 to 130 per cent., sluggish in 48 hours and found dead

310 BIO-CHEMICAL JOURNAL

at 90 hours, lungs pneumonic; at 128°9 per cent., sluggish in 48 hours,
died in about 69, lungs congested and consolidated; at 129°7 per cent.,
dead in 40 hours, similar appearances; at 114 per cent., dead in 60 hours,
consolidated lungs; at 182°9 per cent., one death in 25 and one in 27 hours;
at 176°6 per cent., deaths within 24 hours; at 188°5 per cent., death in
7 hours; at 285 per cent. death in 8} hours; at 357 per cent., death in
5 hours with the characteristic consolidation and congestion of the lungs.

The filling of the lungs with exudate in such a short interval as five
hours appears to us to be a most remarkable and interesting result, and
seems to point to the exudate being a kind of pathological secretion by
the lung epithelium, rather than an irritative exudation due to injury of
that epithelium. .

Observations have also been made by Leonard Hill! and J. J. R.
Macleod showing that prolonged exposure of mice to a partial pressure of
oxygen amounting to 70 to 80 per cent. of an atmosphere produces
inflammation of the lungs, but 80 to 90 per cent. of oxygen used for two
hours or more with the Fleuss rescue apparatus by miners in training
for rescue work produced no ill results whatever. This has also been our
own experience in this laboratory, in breathing oxygen of 80 to 90 per
cent. from a large gas-holder before giving it to patients as mentioned
above. The carbon-dioxide was fully taken up by a series of four annular
trays in the lower part of the gas-holder, into and out of which the patient
breathed by a very wide tube without the interposition of any valves,
and no inconvenience whatever was ever experienced.

In the experiments here to be recorded with guinea-pigs, no total
pressure greater than atmospheric was used throughout, the animals had
plenty of food in a large respiration chamber sufficiently roomy for them
to move about, the carbon-dioxide was efficiently removed by the soda-
hme, and as much oxygen as required was sucked in as the animals used
it, and enough extra supply was drawn in as described in a previous
section by the suction pump to keep up the desired oxygen percentage.
The oxygen appeared to increase appetite in the earlier stages, and the
animals ate fheir food greedily and seemed to suffer from no
inconvenience.

EXPERIMENTS XI, XII, XIII

Three guinea-pigs were placed at the same time in the respiratory
chamber under the usual conditions described already. The history of
each animal is separately given in the following account :—

1. See British Medical Journal, Vol. I, p. 71, 1912. ; Proc. Roy. Soc., Vol. LXX, p. 455,
1902; Journal of Physiology, Vol. XXIX, p. 492, 1903,

THE EFFECTS OF ATMOSPHERES 311

Beperiment XI.—The guinea-pig was placed in the chamber and the
oxygen was slowly raised to 79 per cent. Next day the oxygen stood at
88 per cent., the carbon-dioxide at 1°5 per cent. The animal seemed
quite well, and was eating and lively. On the third morning the oxygen
was 79 per cent., and carbon-dioxide 1°6 per cent. The chamber was
cleansed out, the guinea-pig was in the air outside for thirty minutes
during this operation, and the oxygen was raised to 70 per cent. On the
fourth morning the animal was found dead. The oxygen stood at 75 per
cent., and carbon-dioxide at 0°76 per cent.

A post-mortem examination was at once made with the following
findings : —

(a) Microscopic Examination.—The lungs showed a markedly acute
active congestion, which was in patches in certain places. All the lobes
were so affected, and the tissue sank in water. The heart showed the
distended and engorged condition of an asphyxial death.

(b) Microscopic Examination.—TVhe sections of the lung when cut
and stained showed the extraordinarily acute active congestion demon-
strated by the first micro-photograph accompanying this paper. The
capillaries are seen to be dilated to three or four times their normal
diameter. There is a slight exudate in the alveoli, and a few leucocytes.
The epithelium of the bronchi is somewhat desquamated, and there are
traces of proliferation. By Weigert’s stain traces of fibrin were found in
some of the alveoli.

Bacteriological smears and cultures on various suitable media were
made from the cut lung, as well as from the exudate in the smaller
bronchi, and all these were found to be negative.

The cause of death was an acute lobar or catarrhal pneumonia,
suppressing the respiratory function of the lungs.

Experiment XI/I.—The second of the three guinea-pigs in the
chamber being treated in the same manner, was noticed, when the
preceding one was dead, to be suffering from dyspnoea. At 10 a.m. on
the fourth day from the beginning of the experiment, the respirations
were 120 per minute, the breathing was laboured, and the animal did not
eat and avoided movements. At 2 p.m. this animal succumbed, and it
was found that the oxygen was then 76°3 per cent., and the carbon-
dioxide 0°5 per cent.

The examination gave identical results with those described in the
previous case, the micro-photograph shows the extreme condition of

congestion in the lungs.

312 BIO-CHEMICAL JOURNAL

Experiment XIII.—The third guinea-pig in the chamber with the
preceding two was removed from the oxygen chamber at 2 p.m. on this
same fourth day. It was then very dyspnoeic; respirations, 120 per
minute, with heavy respiration, but not quite so pronounced as in the
last experiment. It was removed to the animal house and placed in a
straw bed. Next morning it was somewhat recovered, but still had not
~ eaten and remained still. In the afternoon it began to eat a little, and
was moving about freely in two days’ time, having made a rapid recovery.
Four days later, this animal was killed in order to examine the condition
of the lungs. These showed, microscopically, a congestion in patches,
but a considerable amount of air-space was present, and pieces of the
lung now floated in water, instead of sinking as in Experiments XI and
DUE, >

Microscopical Examination.—The lungs showed patchy congestion,
alveoli generally full, capillaries distended, epithelium of bronchi
desquamating and proliferating. Catarrhal cells show many fat vacuoles
in the cytoplasm; the nuclei stained well and were displaced to one side.
(See Micro-photograph No. 3.)

Experiment X7V.—Three guinea-pigs were placed in the respiratory
chamber, and the percentage of oxygen was raised to 69. The next
morning the oxygen had risen to 74 per cent., and the carbon-dioxide
0°6 per cent. The guinea-pigs were active and well on the third day,
when analysis showed: oxygen 71 per cent., and carbon-dioxide 1 per
cent.; and on the fourth day, before cleaning out the chamber, it was
found that the oxygen was 68 per cent., and carbon-dioxide 1°3 per cent.

The chamber was now cleansed and the guinea-pigs, which were not
so active or feeding so well as usual, were replaced, and the oxygen
raised to 72 per cent. Next morning (fifth day), one animal was found
dead: the oxygen standing at 69 per cent., and the carbon-dioxide at
0°3 per cent. During this fifth day the two remaining guinea-pigs showed
evident signs of the effects of the oxygen in their respiration and general
appearance, and the second one died next morning (sixth day), and the
third one succumbed in the same afternoon (sixth day.) The oxygen on
analysis showed now a strength of 72 per cent., and the carbon-dioxide
stood at 0°6 per cent. The examinations of the tissues showed a
pneumonia of a slightly more chronic type than that of the previous
experiment (see Micro-photograph No. 4, from the second of these three
animals),

From the results of these experiments it was concluded that

pty.

a.

THE EFFECTS OF ATMOSPHERES 313 |

about 70 per cent. of oxygen was the minimum which was necessary to
cause this pneumonic flooding of the lungs, and this was found in
practically every succeeding experiment to be the result; for if the
strength of the oxygen dilution rose over 70 per cent. for a day, the
guinea-pigs invariably showed symptoms of lung mischief; whereas,
with the lower dilutions, leaving out accidental causes of death, the
guinea-pigs progressed favourably and were free from symptoms. Such
results were obtained in four similar experiments to the above, in which
about 70 per cent. of oxygen was used, while in an equal number with

60 to 65 per cent. of oxygen no pneumoniec effects were observed.

CONCLUSIONS

1. That oxygen in any percentage at atmospheric pressure can be
inhaled for short periods without ill effects.

2. That at percentages below 70 per cent. it can be inhaled for
prolonged periods without giving rise to any symptoms or pathological
signs.

3. That above 70 per cent. its use for a prolonged period is attended
with serious risk of causing an irritative pneumonia, and if persisted in,
produces death.

4. ‘That the pneumonia is of a lobar and catarrhal character and due
to the effects of the oxygen and not to any micro-organism.

In conclusion, I desire to express my thanks to Professor Benjamin
Moore for the suggestion of the research, and his kindness and unfailing
assistance at all times during its progress; also to Professor Ernest
Glynn for much valuable advice on the pathological changes in the lungs,
and to Dr. Stenhouse Williams for his assistance and advice.

d14 BIO-CHEMICAL JOURNAL

EXPERIMENT XI EXPERIMENT XII

}XPERIMENT NITT EXPERIMENT XIV

815

OBSERVATIONS ON ELECTRO-OSMOSIS
By J. 0. WAKELIN BARRATT anv A. B, HARRIS:
From the Cancer Research Laboratory, University of Liverpool
(Received June 13th, 1912)

The present investigation was undertaken with the object of throwing
light upon the enquiry whether it is possible to introduce dissolved
substances into living tissues by means of electro-osmosis. [arly in the
course of this work it became apparent that, so long as the investigation
was confined to living tissues, serious limitations would be encountered
as regards the number and duration of the observations possible, as well
as in consequence of the relatively low strength of current necessarily
employed when it was desired to preserve uninjured the vitality of the
living surface to which the current was apphed. Since living matter Is
largely colloidal it was determined to study as far as possible electro-
osmosis in respect of colloidal membranes. In all of such experiments
the degree of electro-osmosis was ascertained quantitatively, for it 1s
obvious that the practical value of electro-osmosis, if it can be employed
as a means of introducing dissolved substances, would depend upon the
amount of fluid which could be made to pass into the living tissues under
the conditions of experiment obtaining. The selection of the type of
experiment performed is necessarily largely determined by current
hypothesis as to the nature of this phenomenon, and the same applies to
the interpretation of the results obtained.

It will be convenient, in detailing the results of this investigation,
to exhibit first of all the degree of osmosis obtainable through living
tissues, colloidal membranes and porous diaphragms under similar
conditions of experiment. A short reference will then be made to the
theory of electro-osmosis, after which some experiments bearing upon the
possibility of employing electro-osmosis as a means of introducing
dissolved substances into living tissues will be described. The latter are
incomplete, but as the course of this work is temporarily interrupted,
it has been thought best to present it in an incomplete form rather than
delay for an indefinite period publication of the results already obtained.

316 BIO-CHEMICAL JOURNAL

DEGREE OF HLECTRO-OSMOSIS OBTAINABLE

The fluids employed in the experiments about to be described were
alcohol and water. The latter was sometimes used in the form of distilled
water, but more frequently contained dissolved salts.

The media in which electro-osmosis was studied were of three kinds;

porous plugs, contained in glass tubes; colloids in the form of hydrogel
or colloidal solid; and living tissues.
For convenience of comparison the degree of electro-osmosis is always
given per square centimetre sectional area of the medium and in respect
of a potential fall of one volt per centimetre length of the medium. It
may here be observed that the amount of electro-osmosis with a given
sectional area of medium is proportional to the potential fall per
centimetre. So long as the potential fall is maintained constant, the
degree of electro-osmosis per square centimetre is independent of the
length of medium traversed by the current.

Des Ale

The experiments made with porous plugs were of two kinds; those
made with filter paper or cotton wool and those in which powdered
naphthalene or sulphur was employed.

In experiments made with filter paper the apparatus shown in
fig. 1! was used. The description of this apparatus is given on p. 330
In the experiments made with cotton wool, naphthalene, and sulphur,
Perrin’s apparatus shown in fig. 2 was used. This consists of a glass
tube (a), filled with the substance employed as diaphragm, and fitted by
means of ground surfaces into (1) a vertical limb (6), provided above
with a stopcock and connected with a horizontal measuring tube

1. Figs. 1 and 2 are approximately one-quarter, and Figs. 5 and 7 one-half the natural size.
Figs. 3 and 4 are reduced to one-sixth and three-quarters, respectively, of the actual size.

—

OBSERVATIONS ON ELECTRO-OSMOSIS 317

(d), furnished with a bulb at its extremity; and (2) the curved limb (c),
provided below with a stopcock. Two electrodes, marked + and —, are

introduced into the apparatus as shown in the figure.

lance, DY.

Some observations made with the above forms of apparatus are
recorded in Table I, which serves to illustrate the degree of electro-
osmosis observed. The latter ranged, it may be observed, between
0015 c.c. and 0°2 c.c. per hour, being less marked when 70 per cent.
alcohol was employed than when distilled water or a solution of glucose
was used. In all cases the flow of liquid by electro-osmosis was from
anode to kathode, that is to say, in the direction of the current.

TasLe I.—Electro-osmosis through various diaphragms. ‘The amounts of fluid given below
correspond to a sectional area of one square centimetre and a potential fall in the plug
of one volt per centimetre. The first three experiments were made with the apparatus

shown in Fig. 1; the remaining experiments were made with Perrin’s apparatus, Fig. 2.
The + sign indicates flow of fluid from anode to kathode.

No. of Material of which
Experi- Liquid in contact with diaphragm was Amount of electro-osmosis
ment diaphragm composed per hour.
1 5 % solution of glucose in water Filter paper Approximately + 180 c.mm.
2 5 % solution of glucose in water Filter paper Approximately + 130 ¢.mm.
3 5 % solution of glucose in water Filter paper Approximately + 100 c.mm.
4 70 % alcohol Cotton wool (a) Approximately + 384 c¢.mm.
5 70 % alcohol Cotton wool (6) Approximately + 35 ¢c.mm.
6 70 % alcohol Cotton wool (c) Approximately + 35 ¢.mm.
7 70 % alcohol Powdered naphthalene Approximately + 15 ¢.mm.
8 70 % alcohol Powdered sulphur Approximately + 44 ¢.mm.
9 Distilled water Powdered sulphur Approximately + 191 ¢.mm.
10 _— Distilled water Powdered sulphur Approximately + 197 c.mm.

318 BIO-CHEMICAL JOURNAL

When studying electro-osmosis through colloids immersed in
watery liquids the apparatus shown in fig. 3 was employed. In this
apparatus the colloid is used in the form of a diaphragm (a), 1 mm. or less
in thickness, placed between two curved tubes (bb), in the lower part
of which electrodes, marked + and —, are placed. Above are two
graduated measuring tubes (cc), 3 mm. in internal diameter, by means
of which the amount of fluid passing through the diaphragm is
determined, the stopcocks shown in the figure being closed during the
passage of the current. Fluid is introduced into the apparatus through
the narrowed portion of the curved limbs, which are provided with
stopcocks, or by means of a very fine capillary pipette introduced into
the graduated vertical pieces (cc). The electrodes were usually made of
copper immersed in a 10 per cent. solution of CuSO,, above which was
placed the fluid in which the diaphragm was immersed; but in
experiments with gelatine, silver electrodes were employed, immersed in
a N/10 solution of AgNO,, between which and the liquid used for
observing electro-osmosis a solution of NaNO, generally intervened.

The colloids employed for the diaphragm were three in number; a
hydrogel of washed gelatine (10 per cent.); a hydrogel of agar (1°5 per
cent.); and an insoluble colloidal solid consisting of parchment paper.
These diaphragms were held in position by metal or vulcanite flanges,
indicated in fig. 3, which were clamped together and rendered water-

Kia. 3.

OBSERVATIONS ON ELECTRO-OSMOSIS 319

tight by means of shellac. When gelatine or agar was employed, the
diaphragm was formed by painting linen with the dissolved colloid
which was then allowed to set. In some of our earlier experiments,
instead of the diaphragm of gelatine or agar shown in fig. 3, a solid plug
was employed several centimetres in length, but this was given up in
consequence of a difficulty referred to later (p. 823). When a thin
diaphragm is used diffusion readily occurs, so that an equilibrium between
the salts contained in the colloid and those present in the surrounding
liquid is very quickly obtained, but if a solid colloidal plug is employed
it is difficult to secure such equilibrium.

The fluids employed when using the apparatus shown in fig. 3
consisted, in addition to distilled water, of varying concentrations of the
monovalent, divalent and trivalent salts enumerated in Experiments 2 to
8, Table 2, and in Experiments 9 to 12, of a 0°01M to 0°02M solution of
sodium sulphate or nitrate saturated with iodine and containing, in
addition, in Experiments 9 and 10, 33 per cent. of alcohol. The results
obtained in the former series (Experiments 2 to 8) are given in full
elsewhere ;! in the Table the amount of electro-osmosis obtainable with
002M and 0°01M solutions is given. The degree of electro-osmosis
obtained should be compared with that exhibited in the preceding and
succeeding Tables.

Taste II.—Electro-osmosis through colloidal membranes immersed in watery liquids. The
apparatus shown in Fig. 3 was employed. The amounts of fluid given helow correspond
to a sectional area of one square centimetre and a potential fall in the colloid of one volt
per centimetre. The sign + indicates a flow of fluid from anode to kathode ; the sign —
indicates a flow of fluid from kathode to anode.

Amount of fluid passing per hour by electro-osmosis through colloid. The

Concentra— electrolytes employed are given in brackets
No. of tion of
_..,; electrolyte in ati : / an . ms
Experi- liquid in Gelatine (10 %) Agar (1:5 %) Parchment paper
ment contact with
medium
1 Distilled Could not be measured Could not be measured Could net be measured
water
0-02M 365 c.mm. (NagPO,) 5 e.mm. (Na,PO,)

35 ( + 4-15 ¢
002M + 198-3 c.mm. (Na .SO,) + 374 c.mm. (Na,SO,) + 4:31 e.mm. (Na.SQO,4)
002M + 184-2 e.mm. (NaOH) + 370 c.mm. ( + 3-64 e.mm. (NaOH)
002M + § 68-3 e.mm. (NaNOs) + 362 e.mm. (NaCl) + 2-10 c.mm. (NaCl)
— ( + +
— ( + a5
— +

DMI Oke who

0-02M 171-4 e.mm. (HNOs) 103 e.mm. (HCl) 1-08 e.mm. (HCl)
0-02M 286-5 e.mm. (Cu(NQs,)o) 103 e.mm. (CuCls) 1-32 c.mm. (CuCl,)
0-02M 693-0 e.mm. (Al(NOs)s) 64 c.mm. (AICl,) 0-15 ¢mm. (AIC 15)

9 0-01M — 19-6 c.mm. (Na,SO,)*

10 001M — 28-5 c.mm. (NaNO,)*

LT, 001M — 36-6 c.mm. (NaNO,)*

12 0-02M — 54-8 c.mm. (NaNO,)*

Barratt, J. O. Wakelin, and Harris, A. B. ‘ Elektroosemose und Konzentration der
Blekiroiyte” Zeitschr. f. Elektrochemie, Bd. XVIII, 8. 221, 1912.
*In Experiments 9 and 10 the sodium salt employed was dissolved in water containing
33 per cent. of alcohol. In Experiments 1] and 12 a watery solution of sodium nitrate free from
aleohol was used. In these four experiments the solution of sodium salt in contact with the
diaphragm was saturated with iodine.

320 BIO-CHEMICAL JOURNAL

The amount of electro-osmosis observed ranges, it will be seen, from
0:693 c.c. to less than 0°0002 c.c. per hour. The direction of osmosis is
that of the current, except in some observations made with gelatine, in
which the direction is from kathode to anode. In Experiments 2 to 8
the influence of decreasing valency of anien and of increasing valency of
kation in diminishing the rate of flow of liquid from anode to kathode or
of reversing its direction (in the case of gelatine) is shown. When
gelatine is immersed in a solution of sodium sulphate or nitrate, the flow
of liquid is from anode to kathode; if iodine is added the direction of
flow is reversed.

The observations made upon living tissues now remain to be
described. In these experiments the human forearm was used, the
apparatus shown in fig. 4 being employed. This consists of a flanged
glass bowl, the mouth of which is applied to the skin, to which it is
affixed by means of collodion. To the upper part of the bowl is attached
a graduated glass tube, having an internal diameter of 2 mm., which is

ae aoe

———
Fie, 4.

kept in a horizontal position during experiment. Near this tube a
platinum electrode is sealed into the bowl. Before use the electrode is
covered thickly with electrolytically deposited zine. When used for
experiment the apparatus, after being attached to the skin, is filled with
a 4 per cent. solution of ZnSO,, by means of a pipette introduced into
the horizontal tube, all air bubbles being removed. The forearm is
supported below in a dish contaiming a 4 per cent. solution of ZnSO,,
into which an electrode is placed.

When a current is passed through the forearm a passage of fluid
from anode to kathode occurs. So long as the circulation in the forearm
continues this is not, however, recognisable, for oedema of the skin occurs,
in consequence of which the fluid in the horizontal tube is driven towards
the mouth of this tube whatever the direction of the current may be.
This difficulty was removed by applying an indiarubber bandage to the
forearm and elbow so that the tissues were largely emptied of blood, after
which a ligature was placed on the middle of the arm, arresting the
circulation beyond; the indiarubber bandage was then removed below

a
‘

Vv 2th “te *

OBSERVATIONS ON ELECTRO-OSMOSIS 321

the ligature and the apparatus fixed in position. Under these circum-
stances no oedema occurred during the course of the experiment. The
forearm and hand soon became numb, and after a time cold. The
application of the ligature was attended with a certain amount of
discomfort; it could be borne, however, for a period of about three-
quarters of an hour.

The experiments were performed in the following manner. After
the apparatus had been applied as above described, an interval of ten
minutes was allowed to elapse in order to make sure that the level of
fluid in the horizontal tube was nearly constant. A current was then
allowed to pass for fifteen minutes and the movement of the meniscus in
the horizontal tube noted. At the end of this period the current was
passed in the reverse direction for ten to fifteen minutes, during which
time a nearly equal movement of the meniscus in the opposite direction
occurred. The current was then broken, and the position of the meniscus
noted for a further period of ten minutes, during which no current was
passing. Sometimes the current was passed for about twenty-five minutes
in one direction only, the reversal of current being omitted. In these
experiments it is essential that the forearm should be kept as free from
movement as possible, the subject assuming at the beginning of the
experiment a comfortable position, which is subsequently maintained
unchanged. During the periods, at the beginning and end of the
experiment, while no current is passing, the meniscus in the horizontal

Tarie III.—Electro-osmosis through forearm. “The apparatus shown in Fig. 4 was attached
to the skin. The amounts of fluid given below correspond to a sectional area of one
square centimetre and a potential fall of approximately one volt per centimetre. The
sign + indicates that the flow of fluid is from anode to kathode.

Subject of

Experi- Amount of fluid passing per hour by electro-osmosis through forearm.

ment
B + 11-3 e.mm. (kathode inside apparatus) | + 5-0 c.mm. (anode inside apparatus)
B + 10-7 e.mm. (kathode inside apparatus) + 11-5 c.mm. (anode inside apparatus)
H + 9-0 c.mm. (kathode inside apparatus) + 7-9 e.mm. (anode inside apparatus)
A + 6-3 e.mm. (kathode inside apparatus) + 15-0 c.mm. (anode inside apparatus)
A + 26-0 e.mm. (kathode inside apparatus) -+- 17-5 c.mm. (anode inside apparatus)
W + 13-0 c.mm. (kathode inside apparatus) + 9-8 c.mm. (anode inside apparatus)

Mean 12-7 c.mm. Mean 11-1 e.mm.

tube exhibited slight variation in position, but no definite movement in
one direction; during the passage of the current a movement was observed
which was constant, its direction being determined by that of the
current.

The results of a number of experiments of this type are given in
Table III. With different subjects and strengths of current ranging

322 BIO-CHEMICAL JOURNAL

~

from 2 milliamperes to 4 milliamperes per square centimetre, some
differences in the rate of flow of fluid were observed, and again on
changing the direction of the current the reversed flow did not always
take place at the original rate. Nevertheless, the variations observed
were on the whole inconsiderable, the mean value for a potential fall of
approximately one volt per centimetre being 11°9 c.mm. per hour per
_ square centimetre of surface of skin. With currents exceeding 9
milliamperes per square centimetre of surface, superficial sloughing of
the skin occurred. The duration of application of the current was, as
already mentioned, about twenty-five minutes.
The direction of electro-osmosis through the forearm was from anode
to kathode. For convenience of comparison the direction of flow of fluid

in these and the preceding experiments is indicated in Table IV.

Taste LY.—Direction of electro-osmosis (compiled from I, II and III).

Liquid in contact with medium Medium employed and direction taken by fluid.
No. of
Experi- Plug of Parchment
ment filter Gelatine Agar paper Forearm
paper
1 Solution of glucose +—> -
2 Distilled water No move-
ment
recognis-
able
3 Solution of Na,PO, ar i
4 Solution of Na,SO, +——>- 4——->- 4+-->-
5 Solution of NaOH == = a
6 Solution of NaCl or NaNOg +> — +—> = [> =
7 Solution of HCl or HNO , + <———- 4+->- = =
8 Solution of CuCl, or Cu(NOs) 3p SS SESS = —
9 Solution of AICI, or Al(NOs3)3 +< => = SS =
10 Todine in water =2 <=
11 Solution of zinc sulphate SS =

When a current is passed as above described through the forearm,
little or no sensation is ordinarily experienced, so long as the strength of
the current does not exceed 3 milliamperes per square centimetre. With
a current of 6 milliamperes, tingling is frequently produced. If still
stronger currents are employed, a smarting or burning sensation is felt;
such currents can, however, be borne provided the strength of current is
increased very gradually, the passage of the current being after a time
attended with some blunting of sensibility. When the circulation is
re-established in the limb, transient. swelling of the skin, generally
attended by redness, occurs in the area of application of the current.

OBSERVATIONS ON ELECTRO-OSMOSIS 323

If the current has been too st rong, blanching of the skin may take place.
and superficial necrosis is likely to occur. When such injury follows the
application of the current, it is generally irregular in distribution, the
surface of the skin being unequally affected. Commonly necrosis, if it
occurs, takes place around the mouths of the hair follicles, in which
situation the current appears to enter more readily than elsewhere.
The current which can be applied without injury, varies in different
individuals. Any breach of surface which may be present at the time
of application of the current is likely to be the seat of subsequent
sloughing, even if a current of only 2 milliamperes per square centimetre
is employed.

Before concluding this section, reference may be made to an effect of
electro-osmosis, which at first exercised a disturbing influence upon the
course of these experiments. It was sometimes found that when a current

+

Fia. 5.

was passed through a glass tube (a, fig. 5) partly filled with a hydrogel
of gelatine (containing 40 per cent. of gelatine and 0°6 per cent of NaCl),
above which was a solution of an electrolyte, the degree of electro-
osmosis diminished as the experiment proceeded, and after a time its
direction! became reversed. In such cases, the upper surface of the

gelatine (the condition of the lower surface could not usually be
1. To measure the rate and direction of flow the upper electrode was sealed into the

vertical tube a, which was closed above and connected with a horizontal measuring capillary.
Other forms of apparatus were also used,

324 BIO-CHEMICAL JOURNAL

observed) was seen to become cupped, and a few millimetres below the
apex of the cup a crack appeared after a time in the gelatine. This
phenomenon was observed when a 0°1M solution of CuSO,, containing a
, containing a zine anode, was
placed above the gelatine. In fig. 5, b c d represent the appearance

copper anode, or a 0°1M solution of ZnSO

of the gelatine in tube a, two hours, three hours, and four hours,
respectively, after the commencement of the experiment. Below the
surface of the gelatine is a dark-coloured green zone into which Cu** ions
have penetrated; beneath this is a hghter greenish zone indicated by a
double contour in b, c and d, representing the advancing layer of these
ions. Still lower a horizontal reflecting surface is seen (indicated in the
figure by a horizontal line). This appears to be caused by removal of
water by electro-osmosis from the upper portion of the gelatine, the
refractive index of which is in consequence no longer the same as that of
the more watery portion of the gelatine below. This boundary surface
was found to travel, under a potential fall of one volt per centimetre, at
the rate of 0°93 cm. to 1°05 cm. per hour. A crack, placed obliquely and
reaching below to this boundary surface, is seen in c¢ and d.

The explanation of this phenomenon was ultimately found to be as
follows: —When Cu** or Zn** ions passed into the gel, causing it to
become green in the former case, and somewhat opaque, the direction of
electro-osmosis in this portion of the gel, in contact with the solution of

CuSO, or ZuSO,, became reversed, while that occurring in the unchanged -

gel below continued its original direction unaltered. In consequence, at
the junction of the two the gel became rapidly deprived of water and very
marked cupping took place. As soon as the latter became unable to
compensate for the contraction of the gelatine a crack appeared in the
gelatine, attended with the formation of an air space. When the
hydrogel of gelatine, instead of being made up with 0°6 per cent. NaCl
solution or with distilled water, was made up with CuSO, solution of the
same concentration as that used in the liquid above, cupping did not occur
or was very sheht, and no cracks formed.

It may be observed that the effect of adding an electrolyte to a gel
of gelatine or agar is to cause contraction of the gel, which usually swells
up again (though not necessarily to the original extent) when placed once
more in distilled water. Some salts, however, such as HgCl,, cause
permanent shrinking. Irreversibility is also conferred upon gelatine by
He salts, formol and iodine.

At hung

OBSERVATIONS ON ELECTRO-OSMOSIS 325

Nature or ELecrro-OsmMosis

Klectro-osmosis was observed as long ago as 1807 by Reuss.' It was
re-discovered by Porret,? and was further studied by de la Rive,?
Becquerel*, Daniell,® Napier,® and Wiedemann.

Reuss placed two tubes, containing electrodes immersed in water, in
a mass of clay and found that the level of the fluid during the passage of
the current rose in the kathode tube and sank in the anode tube. In
consequence of this passage of fluid the phenomenon was compared to
osmosis, but the resemblance is only a superficial one, since at the
commencement of the experiment the fluids are of the same concentration.
As the current continued to pass, the difference of level in the two sections
of the cell increased up to a certain point and then remained unchanged,
the hydrostatic pressure balancing the action of electro-osmosis. The

a

XX

a b

Fic. 6.

degree of electro-osmosis can therefore be measured in terms of the
equilibrial height (#7), which has been shown by G. Wiedemann,’

1. Reuss, F. ‘Noticesur an nouvel effet de d’électricité galvanique.’ Mémoire dela Soc. imp.
des Naturalistes a Moscou, T. H., 1809, pp. 330-332.

Porret. Thomson, Annals of Phil., VIII, p. 74, July, 1816.

dela Rive. Traité del Electr., 11, p.379; Ann. de Chim. et de Phys., XXVIII, p. 125, 1825.

Becquerel. Traité de U Electr., L11, 102.

Daniell. Ann. d. Physik. u. Chemie, 1835, Ergiinzungsbd., I, 569.

Napier. Phil. Mag., July, 1846. i

Wiedemann, G. Galvanismns, I, S. 377; ‘ Ueber die Bewegung von Fliissigkeiten im Kreise
der geschlossenen galvanischen Siiule.” Ann.d. Physik. u. Chemie, Ba. LXXXVII, 321,
1852; ‘Ueberdie Bewegung der F'liissigkeiten im Kreise der geschlossenen galvanischen
Saule und ihre Beziehungen zur Elektrolyse.’ Ann. d. Physik. u. Chemie, XCIX, p.177,
1856.

TA wt

326 BIO-CHEMICAL JOURNAL

employing the apparatus shown in fig. 6,1 to be directly proportional to
the strength of the current (c), the specific resistance of the solution (7),
and the thickness of the diaphragm (t), and inversely proportional to the
sectional area S, according to the formula :—

H = constant x =

: rt : : :
Since <~ represents the resistance of the diaphragm the expression
hk
rt
[So
(£), so that the formula may be written :—

H = constant x H

CR, i.e., the potential fall between the sides of the diaphragm

Wiedemann concluded that the galvanic current caused a direct
transport of fluid, in virtue of a traction which it exerted upon each
element of volume of fluid which it traversed. This conclusion was
contested by Graham,? v. Quintus Icilius,? and Breda and Logemann,#
who pointed out that it was impossible to demonstrate the transport of

fluid in the absence of a diaphragm.

Further investigations were made by G. Quincke,® who studying
electro-osmosis in capillary glass tubes provided with platinum wire
electrodes, advanced an interpretation of electro-osmosis, which still
remains the accepted explanation of this phenomenon. Quincke observed
that when particles of clay were suspended in the lquid between the
electrodes, a movement towards one or other electrode occurred (in water
towards the anode). He therefore concluded that, at the surface of
contact of solid and liquid, opposite electrical charges appeared, an
electrical double layer resulting. A system of this kind would be affected
by a potential gradient in such a way that the positively charged portion
of the system is moved in one direction and the negatively charged portion
in the other direction. If the solid were fixed, then movement of the

1. This consists ofa porous clay cylinder, closed below and cemented above to a vertical glass
tube provided with a horizontal capillary which delivers fluid into a receiver or, in another
form of the apparatus, is continuous with a manometer tube. Passing through the vertical
glass tube is a platinum wire connected with a cylindrical electrode of platinum foil, (6) placed

in the clay vessel. The latter is contained in a beaker filled with water and is surrounded by
a second cylindrical electrode (a) made of lead.

2. Graham. Phil. Mag., VIII, p. 151, 1854.
3. v. Quintus Icilius. Lehrbuch der Experimental-Physik, 8. 642, 1855.
4. Breda and Logemann. Ann. d. Physik. u. Chemie, Bd. C, 8. 149, 1857.

5. Quincke, G. ‘Ueber die Fortfiihrung materieller Theilchen durch strémende
Hlektricitaét. Ann. d. Physik. u. Chemie, Bd. CXIII, S. 513, 1861.

' aie

OBSERVATIONS ON ELECTRO-OSMOSIS 327

liquid alone would occur. Subsequently Helmholtz! advanced a
mathematical theory of the electrical double layer and showed that
Quincke’s explanation was theoretically tenable.

Hardy? observed that particles of globulin of different size contained
in solutions of varying concentration, the aspect of which ranged from
transparency to opalescence, moved at the same rate. Since the velocity
of the particles was, within the limits of concentration employed,
independent of their size, it follows from Helmholtz’s theory of electro-
osmosis that the density of the charge per unit of surface is constant and
the total quantity of electricity on each particle (() is directly
proportional to the surface according to the equation

Q=a4rr
Helmholtz showed that the potential difference #, at the junction of
solid and fluid in a capillary tube was given by the equation
ne 4 np
; Kr’ H
Where X is the specific inductive coefficient of the liquid,? » the
coefficient of internal friction of the fluid, (7) the radius of the tube,
H the strength of the electric field, and @ the amount of liquid passing
through the tube per second under the influence of the electric field.

Data respecting the amount of fluid carried by electro-osmosis are
scanty. A few determinations of the amounts of fluid transported through
a clay diaphragm immersed in dilute solutions of H,SO,, CuS0O,,
Cu(NO,),, and AgNO, by currents of a known strength were given by
Wiedemann (1856).4 Further determinations of the relative rate of flow
of fluid through hydrogels under a fixed potential gradient are given by
Perrin.® More recently Barratt and Harris® have determined the
absolute rate of flow of fluid by electro-osmosis through gelatine, agar,
and parchment paper, when various concentrations of electrolytes are

employed. From these observations the data given in Experiments 2 to 8,
Table II, are selected.

1. Helmholtz. ‘Studien tiber elecktrische Grenzschichten. Wiedem., Ann. d. Ph. u. Ch.,
Bd. VII, S. 337, 1887.

2. Hardy, W. B. ‘Colloidal solution. The globulins.’ Journ. of Physiol., XXXII, p. 251,
1905-6.

3. Kis introduced into the formula by J. Perrin, ‘ Mécanisme der 1’électrisation de contact
et solutions colloidales.’ Journ. der Chimie Physique, I1, p. 601, 1904; III, p. 50, 1905.

emeroe. Cit... Pp. Lis

5. Loc. cit.

6, Barratt, J. O. Wakelin, and Harris, A. B. Loc. cit.

328 BIO-CHEMICAL JOURNAL

The flow of fluid in electro-osmosis is obviously determined by the
movement of the ions contained in the liquid at the surface of contact
with solid. These ions, as they move under the influence of the potential
gradient, cause a passive movement, in the same direction, of the
molecules of liquid. The narrower the interval between the particles of
which the diaphragm is composed, the more marked this movement of
. liquid will be. Further, the more numerous the ions in the hquid portion
of the double layer, the greater the flow of liquid will at first be, though
with continued increase in the number of ions a diminution of
concentration of fluid will later occur and the flow will diminish; thus
the curve obtained by plotting concentrations, as abscissae, against rates
of flow, as ordinates, under a given potential gradient, will exhibit a
maximum, corresponding in the case of agar to a concentration of
electrolyte about 0°02M, while on each side of this maximum, with
greater and lesser concentrations, the curve will exhibit a diminished
rate of flow.!

Up to the present it has been tacitly assumed that electro-osmosis
has been brought about solely by the movement of unaltered ions. If,
however, the ions in question became centres for the condensation or
combination of water molecules, then an additional factor in the transport
of fluid would come into play. Nernst? showed how the problem of ionic
hydration might be investigated by means of diffusion and migration
experiments in which an indifferent dissolved substance was used as
indicator. Lotmar,? employing one of Nernst’s methods, showed that
hydration of ions occurs. Lobry de Bruyn,* using a solution of silver
nitrate in aqueous methyl alcohol, was unable to obtain evidence of ion
hydrates or alcoholates. Morgan and Kanolt,> however, studying
electrolysis of a solution of copper nitrate in water and alcohol by a
procedure similar to Nernst’s second method, concluded that copper ions
were hydrated.

Obviously if the ions contained in the liquid portion of the double
electrical layer are hydrated, the rate of flow will be much greater than
would be the case if no hydration occurred. Usually the current

1. Barratt, J. O. Wakelin, and Harris, A. B. Loc. cit. Figs. 172-174, pp. 223-224.

Nernst, W. ‘Zur Frage nach der Hydratation geléster Substanzen,’ I, Nachr. d. k. Gesellsch.

d. Wissensch., G6ttingen, Math.-Phys. K1., S. 68, 1900.

3h Lotmar, H. ‘Zur Frage nach der Hydratation geléster Substanzen,’ II, Nachr. d. k.
Gesellsch. d. Wissensch., Gdttingen, Math.-Phys. K1., S. 70, 1900.

4. Lobry de Bruyn. Jahrb. d. Elektrochemie, Bd. X, p. 260, 1904.

5. Morgan and Kanolt. ‘Uber die Verbindung der Lésungsmittel mit den Ionen,’ Zeitschr. f.
physikal. Chemie, Ba. XLVIII, S. 365, 1904.

OBSERVATIONS ON ELECTRO-OSMOSIS 329

would not be conveyed wholly by the ions contained in the liquid portion
of the double layer, though as an exceptional event this might conceivably
occur if the particles of which the diaphragm was composed were in
sufficiently close apposition. More usually the current would be in part
conveyed by ions of the electrolyte employed, which do not form part of
the double electrical layer at the surface of solid and liquid. In the
latter case there would be, if the ions were hydrated, an additional
transport of fluid representing the difference between the hydration of
opposite ions involved. Presumably, also, the degree of ionic hydration
would be different at the surface of liquid in contact with the solid, in
which situation the concentration of water would be different from that
obtaining elsewhere in the liquid. At the present time sufficient data are
not available to enable an estimate to be formed of the extent to which
electro-osmosis is affected by ionic hydration. If, however, an agar
diaphragm is employed in the apparatus shown in fig. 5, the number of
molecules of water passing when electro-osmosis is at a maximum is found
to be from eighteen to three hundred and seventy times greater than the
number of molecules of electrolyte decomposed by the current,! a
circumstance which may be interpreted as indicating that conveyance of

fluid by hydrated ions plays only a subordinate part in electro-osmosis.

THe Inrropuction or DissoLveD SUBSTANCES INTO ,LivinG TISSUES
BY MEANS OF EHLECTRO-OSMOSIS

The conditions under which experiments directed to this end are
performed, are attended with limitations, which render investigation
difficult.

Thus the amount of current which can be employed without
damaging the skin is limited ordinarily to from one to three milliamperes
per square centimetre, and even this current cannot usually be continued
safely for more than twenty minutes. Again, since ions derived from
dissolved electrolytes enter and leave the skin under the action of a
potential gradient, the dissolved substances which can be employed in the
investigation of the problem under consideration are limited to non-
electrolytes of sufficiently simple molecular constitution to be able to pass
through colloidal material. Further, such dissolved substances, after
passing into the skin, must be capable, even in the very small quantities
which in practice can alone be introduced, of producing a marked local
or remote effect upon the organism. Unfortunately, it is in many cases
1. Barratt, J. O. Wakelin, and Harris, A. B. Loc. cit., p. 225.

330 BIO-CHEMICAL JOURNAL

difficult, when employing dissolved substances in experiments on
electro-osmosis, to be quite sure that an ionisation effect is wholly
excluded.

It is obvious that a solution of the problem whether non-electrolytes
can be introduced into the skin by electro-osmosis is best approached by
first ascertaining if dissolved substances can be made to pass by electro-
osmosis through colloidal membranes. This aspect of the enquiry will
alone be dealt with in this section, experiments made with lying tissues
being reserved for a subsequent communication.

This problem has already been studied by Oker Blom,! who introdmeea
I-ions by means of electrolysis into colloidal masses. As a want of
correspondence between the calculated amounts I~ions which should have
been introduced by the strength of current used, and the amounts actually
present in the gel at different distances from the site of entry of the
current occurred, Oker Blom concluded that evidence of a kataphoric
action of the current had been obtained.

Our own experiments directed to the investigation of this problem
have been made with two dissolved substances, namely glucose and
iedine.

In the first series of experiments a plug of filter paper contained in

glass bulb was moistened with a 5 per cent. solution of glucose in
distilled water, and an electric current caused to pass through it for a
period of time sufficient to permit of the passage of so large an amount
of liquid by electro-osmosis that any alteration in composition, such as
would occur if water were carried to the complete or partial exclusion of
glucose, would be readily determined.

The apparatus employed (fig. 1, p.816) consisted of a glass bulb,
tightly packed with filter paper, and placed horizontally in a beaker
containing a 5 per cent. solution of glucose. At one end the bulb was
open, communicating with the liquid in the beaker; a platinum electrode
was placed in the beaker opposite this open end. The other end of the
bulb was sealed into a vertical tube, also filled with glucose solution,
containing another platinum electrode. Connected with the vertical
piece were two measuring capillary tubes of two millimetres internal
diameter, one above, placed vertically, and the other, placed horizontally,
used for measuring the amount of fluid passing by electro-osmosis. The
latter tube terminated in a thistle funnel in which the fluid passing could,
if desired, be allowed to collect.

1. Oker Blom. ‘Beitrag zur Festellung einer physikalisch-chemischen Grundlage der
elektro-medikamentésen Behandlung mit besonderer Beriicksichtigung der Jodsalzlésungen,’
Kuopio, 1896; ‘Experimentelle Untersuchungen iiber das unter Hinwirkung des konstanten
elektrischen Stromes stattfindende Hindringen von medikamentésen Stoffen in den
Thierkérper,’ Willmanstrand, 1898.

OBSERVATIONS ON ELECTRO-OSMOSIS 331

In the experiments made the strength of current passing was
8 milliamperes to 10 milliamperes per square centimetre, the potential
gradient per centimetre being 25 volts. The amount of fluid passing
per square centimetre by electro-osmosis, calculated for a potential fall
of 1 volt per centimetre, was 0°10 ¢.c. to 0°18 ¢.c. per hour (‘Table I,
Experiments 1 to 5). The current was allowed to pass for fourteen to
sixteen hours, during which period the actual amount of fluid which
passed through the plug of filter paper was 60 c.c. to 70 c.c.

The fluid passing by electro-osmosis was found to exhibit the same
specific gravity as the solution of glucose in the beaker. These
experiments, therefore, indicate that the water passing by electro-osmosis
through the plug underwent no change in respect of its content of
dissolved glucose. It is therefore proven that a non-electrolyte dissolved
in water may be carried passively by electro-osmosis through a colloidal
membrane. The actual amount of glucose transported during the period
of experiment by electro-osmosis was, it may be observed, 30 g. to
35 g.

In the second series of experiments the dissolved substance employed
was lodine in 0°01 N NaCl solution. In these experiments the plug of
filter paper was discarded, and instead a hydrogel of gelatine was
employed in the form of a thin diaphragm. Before use the gelatine was,
in the earlier experiments, allowed to remain in contact with the iodine
solution for some time; in some of the later experiments the gelatine
was not previously exposed to the action of iodine. It may here be
observed that iodine changes the condition of gelatine, rendering it after
a time insoluble even in boiling water. It also causes reversal of the
direction of electro-osmosis when Na,SO,; or NaNO, are employed as
electrolytes (Table IT).

The apparatus employed to determine, if the fluid passing by electro-
osmosis contained iodine, is shown in fig. 7. It consisted of two short
glass tubes, a, b, closed below by a membrane of gelatine half to one
millimetre thick. These tubes were partially immersed in a 0°01M
solution of NaCl, containing a small amount of starch solution. The
current was conveyed to the cups by non-polarisable electrodes formed
by two glass tubes, c, d, closed below by a plug consisting of a hydrogel
of gelatine (10 per cent.), and containing above zinc electrodes, marked
+ and — in the figure, immersed in a saturated solution of ZnSO,. The
potential gradient in these experiments was about 10 volts per centimetre,

1. Separate determinations of the direction and amount of fluid and the amount
passing by electro-osmosis were made with the apparatus shown in Fig. 3.

332 BIO-CHEMICAL JOURNAL

the strength of current being about 5 milliamperes per square centimetre

of gelatine membrane.

+

|
NI

idsce, 7

In carrying out an experiment the empty tubes, a, b, were partly
immersed in the liquid contained in the beaker as shown in fig. 7, the
electrodes, connected to the source of the current, being also placed in
position. The iodine solution was then rapidly poured into the cups,
a, b, thus completing the circuit. The object of the latter procedure
was to avoid the passage of iodine, by diffusion through the gelatine
diaphragm, into the liquid contained in the beaker before the circuit was
closed, whereby a blue deposit or coloration would make its appearance
in the liquid below the cups before electro-osmosis was set up.

When an experiment was made, as above described, it was found
that at the kathode cup iodine passed into the liquid contained in the
beaker, a blue deposit or cloud appearing at the under surface of the
gelatine, while at the anode cup no blue coloration appeared. If the
current was reversed, a blue coloration or deposit appeared below the
gelatine diaphragm of the former anode cup, while no further coloration
appeared beneath the former kathode cup. Thus the iodine passed in the
direction of electro-osmosis, from kathode to anode.

It appears to be admissible to conclude that in these experiments the
dissolved iodine’passes in the unionised condition through gelatine under
the action of electro-osmosis.

ieee

333

THE PHOSPHATIDES OF THE KIDNEY
By HUGH MacLEAN.
From the Lister Institute, Bio-Chemical Department
(Received June 20th, 1912)

The researches of Erlandsen! on the phosphatides of cardiac
muscle showed that the older methods of extracting tissues with alcohol
did not suffice to isolate pure substances, and that the so-called lecithin
obtained was really a mixture. According to this investigator the
phosphatides contained in the ethereal extract of dried cardiac muscle
consisted chiefly of two substances—lecithin with a nitrogen : phosphorus
ratio of 1 : 1, and cuorin with a nitrogen : phosphorus ratio of 1 : 2.
Another ether soluble substance, however, was present in large amount,
but was incapable of being extracted by ether until the tissue had under-
gone a preliminary treatment with alcohol. This behaviour towards
ether suggested that this substance was present in the tissue in some
combination which was split up by the alcohol, or that the ether did not
penetrate the dried tissue sufficiently well to extract it. Erlandsen
extracted this substance by means of alcohol after a thorough preliminary
treatment of the tissue with ether, and found that the greater part of
the alcoholic extract consisted of a phosphatide which differed materially
in constitution from lecithin and cuorin, and contained in its molecule
two atoms of nitrogen to one of phosphorus.

These observations of Erlandsen strongly supported the combination
view, for if the ether were incapable of penetration as the result of purely
physical conditions, it is difficult to understand why there should be any
material difference in the nature of the phosphatide obtained in the ether
and alcohol extracts respectively.

These results with cardiac muscle suggested that other tissues
probably behave in a similar manner towards ether and alcohol, and in
order to determine this the following investigation of kidney phosphatides
was carried out.

PREPARATION OF MATERIAL

Horse kidneys were used on account of their cheapness and the ease
with which large supplies could be obtained. In the first experiment
fifteen fresh kidneys were washed with normal saline solution and freed

334 BIO-CHEMICAL JOURNAL

as much as possible from adherent fat and fibrous tissue. They were
then passed through a mincing machine, and the finely divided tissue
dried on a glass plate by means of a fan at a temperature of about 30°.
The dried substance obtained was now dried in a desiccator over sulphuric
acid and finely ground in a mill to a fine powder.

EXTRACTION WITH ETHER

The dried powder, which weighed 2075 grams, was shaken up with
ether in a dark, well-stoppered bottle, which contained CO, in order to
prevent oxidation. After extracting six times, the ethereal extracts were
evaporated to very small bulk under reduced pressure and at room
temperature. The residue, which now formed a syrupy mass, was treated
with a large,excess of acetone and the mixture allowed to stand for
half an hour. The precipitate which formed was separated from the
ether-acetone liquid, dissolved in some pure ether, and again precipitated
by acetone. It was then thoroughly washed with acetone and dried in
vacuo. By this treatment with acetone, most of the ordinary fat and
cholesterol was separated. In separating phosphatides from fat by
means of acetone, it is important to separate the precipitate from the
ether-acetone solution in a short time, for though acetone first precipitates
phosphatides from an ethereal solution, it gradually precipitates such
substances as neutral fat and certain saturated fatty acids. The crude
phosphatide obtained weighed about 29 grams.

TREATMENT OF CRUDE PHOSPHATIDE OF ETHEREAL EXTRACT

The crude extract was dissolved in about 150 c.c. ether, and an
opalescent solution obtained which on centrifuging yielded a fair amount
of a white precipitate and a clear dark amber-coloured fluid. The
phosphatides were precipitated from this solution by the addition of
excess of acetone, and dried in vacuo. This process of dissolving in ether,
centrifuging and treating with acetone was repeated four times. All
the white precipitates obtained on centrifuging were mixed together,
and for purposes of reference may be called ‘ white substance.’

The phosphatide residue was dissolved in 100 c.c. ether and gave an
almost clear solution; to this 500 c¢.c. absolute alcohol were added, and
the mixture left to stand in an ice-room overnight. Immediately on the
addition of the alcohol, a precipitate formed which became more marked
on standing. After sixteen hours, this precipitate was separated by

THE PHOSPHATIDES OF THE KIDNEY 335

filtration and washed thoroughly with alcohol. It was then dissolved in
ether, precipitated by acetone and dried in vacuo.

An ether-alcohol solution (6b) and a precipitate (a) were thus
obtained.

TREATMENT OF ALCOHOL-INSOLUBLE SUBSTANCE (a)

This substance was treated with alcohol at 45° C., and the residue
dried in vacuo, dissolved in ether and precipitated by acetone. After
drying it was re-dissolved in freshly distilled hot ethyl acetate and
allowed to stand for some hours in the ice-room, when a substance of
syrupy consistency fell out. This treatment with ethyl acetate was
repeated and the syrup obtained dried and analysed.

All the above operations were carried out as far as possible in an
atmosphere of CO, to prevent oxidation; at the same time an endeavour
was made to prevent any deleterious action of light by the use of dark
glass desiccators and bottles.

TREATMENT OF ErHEeR-ALcoHOL SoLuTion (b)

This liquid was evaporated under reduced pressure at a temperature
of about 35° C., and a small amount of a substance obtained, which was
dissolved in pure ether and precipitated by acetone. This substance,
which represents ordinary ‘lecithin,’ was dried and analysed.

TREATMENT OF ‘ WHITE SUBSTANCE’ OBTAINED BY CENTRIFUGE

This substance was first treated with cold alcohol, when a small
amount of it dissolved. On filtering, the residue was extracted with hot
alcohol, when a considerable amount easily went into solution, while
part seemed quite insoluble. The solution was filtered hot and the filtrate
left to stand at room temperature, when a white flocculent substance
separated out. This substance was filtered off, again dissolved in hot
alcohol and filtered hot as before. The flocculent substance obtained
was thoroughly washed with cold ether, in which it was practically
insoluble; it was then treated with hot ether and dried. The part
insoluble in hot alcohol was composed chiefly of inorganic salts and was
not further investigated.

From the ethereal extract of kidney, three substances were now
obtained.

(1) A small amount of a substance soluble in hot alcohol but
insoluble in ether. ‘ White substance.’

(2) A substance soluble in alcohol and in ether—lecithin.

(3) A substance insoluble in alcohol but soluble in ether—
Cuorin,

336 BIO-CHEMICAL JOURNAL

It is interesting to note that only a very small amount of ordinary
lecithin was found. The greater part of the ethereal extract was
composed of the alcohol insoluble portion, of which about 13 grams were
isolated, while only about 2°5 grams lecithin were obtained.

(1) ‘ White substance’ separated by centrifuge

The total amount of this substance obtained was only 0°55 gram. On
analysis it gave the following figures. Nitrogen and phosphorus were
estimated by the methods of Kjeldahl and Neumann respectively.

Nitrogen
0°3125 gm. used 6°6 c.c. N/10 H,SO, = 2°9 per cent. Nitrogen.
Phosphorus
0:1120°gm. used 7 c.c. N/2 NaOH = 3°46 per cent. Phosphorus.
No Pea:

While it is probable that this substance was not quite pure, the small
quantity obtained rendered it very difficult to get quite accurate figures;
it is practically a diamino-monophosphatide, and from its general
appearance and properties is probably of the same nature as a similar
substance described under the alcoholic extract of the kidneys.

Thus it was practically insoluble in cold ether and only very slightly
soluble in cold alcohol, in hot alcohol it dissolved with ease and separated
out on cooling as a flocculent substance which on drying formed a white
non-hygroscopic powder. In chloroform and benzene it dissolved easily,
especially on heating; cold glacial acetic acid dissolved it with some
difficulty, but on heating it dissolved readily. From these solutions it
was precipitated by acetone. In general it bears a close resemblance to
a substance described by Stern and Thierfelder,? and isolated by them
from egg yolk.

(2) Alcohol and ether soluble substance—Lecithin

This fraction had all the properties of the mono-amino-mono-
phosphatide—lecithin. It was of somewhat waxy consistence, yellowish
white in appearance and very hygroscopic. In vacuo it dried very quickly
to constant weight, but was sticky to the touch and incapable of being
powdered. In ether, alcohol, benzene, chloroform and petroleum ether it
dissolved with ease and was readily precipitated from these solutions by
acetone. In ether it gave a slightly opalescent solution which cleared up
in a short time to a clear dark liquid.

THE PHOSPHATIDES OF THE KIDNEY 337

Analyses
01021 gm. substance gave 0°2385 gm. CO, = 63°7 per cent. C.
and 01040 gm. H,O = 11°3 per cent. H.
Nitrogen
0°3219 gm. used 4°5 c.c. N/10 H,SO, = 2°0 per cent.
0°6123 gm. used 85 c.c. N/10 H,SO, = 1°94 per cent.
Average = 1°97 per cent. Nitrogen.
Phosphorus
0°3202 gm. used 21°8 N/2 NaOH = 3°77 per cent.
0°2151 gm. used 16 c.c. N/2 NaOH = 3°9 per cent.
Average = 3°83 per cent.
No Pes P14: 1,

It is probable that this lecithin was not quite pure, as the ratio of
nitrogen to phosphorus is somewhat too high. In general, however, it
agrees well with the lecithins described by Stern and Thierfelder,? and by
Erlandsen,! as the following table shows.

ANALYSES OF DIFFERENT LECITHINS

Lecithin from horse Lecithin from heart Lecithin from egg yolk

kidney muscle (Erlandsen) (Stern and Thierfelder)
C i ay. 63-7 66-29 64-63
H fie ane 11-3 10-17 10-96
N As Se 1-97 1-87 2-08
ze aa <5 3°83 3°95 3-97
N:P ratio aoe 1-:14:1 Ll 1-:16:1

On hydrolysis a substance which gave the reactions of glycerophos-
phoric acid was obtained as well as choline, but, as in other lecithins,
the yield of choline obtained was considerably less than the theoretical
amount.

In order to determine the iodine value of the lecithin, a small amount
was prepared specially for the purpose, as it was found that even when
preserved in as complete a vacuum as possible, the iodine number
gradually diminished. This specially prepared substance was isolated
and purified as quickly as possible, and gave an iodine value of 85. The
free fatty acids obtained from this sample by saponification with alcoholic
potash in an atmosphere of hydrogen gave an iodine value of 113. It is
probable, however, despite the precautions taken, that a certain amount
of oxidation had occurred, so that the value obtained is perhaps somewhat
low.

338 BIO-CHEMICAL JOURNAL

No saturated acids were found while bromine derivatives of acids,
having three or more unsaturated bonds, were prepared.

The nature of the acids present in lecithin and other phosphatides
will be discussed in a later paper.

(3) Substance insoluble in alcohol but soluble in ether—Cuorin

While a comparatively small amount of lecithin was found in the
ethereal extract, a considerable amount of cuorin was present.

It was obtained as a yellowish transparent substance, which was
exceedingly hygroscopic, but dried quickly in vacuo to a hard solid mass,
which was easily reduced to a fine powder. It dissolved at room
temperature in ether, petroleum-ether, carbon disulphide, and chloro-
form; while it was practically insoluble in cold ethyl acetate, it
dissolved easily on heating; on cooling it separated out as a syrupy mass.
In cold and-hot alcohol it was quite insoluble. This substance oxidised
with great readiness, so that it was difficult to keep it unchanged for any
length of time; as it oxidised, it gradually changed its solubility, and,
after a time, it became insoluble in ether and somewhat soluble in water.

It was precipitated by acetone out of its ethereal solution; platinum
chloride and cadmium chloride gave insoluble combinations.

The iodine value of the freshly prepared substance was 100, while
that of the fatty acids was 133. Even when kept in an evacuated
desiccator the iodine value gradually diminished, so that after a
fortnight it was only 83; on exposure to the air it decreased in about a
week to 52.

Analyses
0°1216 gm. substance gave 0°2695 gm. CO, = 60°4 per cent. C.
and 0°1120 gm. H,O = 10°2 per cent. H.
Nitrogen
0°5216 gm. used 3°8 N/10 H,SO, = 1°02 per cent.
0°6210 gm. used 4°77 N/10 H,SO, = 1°06 per cent.
Average = 1°04 per cent. N.

Phosphorus
0°3105 gm. used 25°2 c.c. N/2 NaOH = 4°49 per cent.
0°2108 gm. used 16°8 c.c. N/2 NaOH = 4°41 per cent.
Average = 4°45 per cent. P.
Nine (1
A comparison of this substance with the cuorin described by
Erlandsen shows that it agrees exactly with the analysis of cuorim. (2

‘“

THE PHOSPHATIDES OF THE KIDNEY 339

somewhat similar substance has been isolated by Baskoff from the liver,
but though its properties are similar to those of cuorin, it gave somewhat
different figures on analysis. » It is probable that this substance, which
Baskoft called heparphosphatid, was not isolated in a pure form, and the
fact that it contains sulphur strengthens this idea.

A somewhat similar substance which I isolated from egg yolk?
differed in its percentage of nitrogen and phosphorus from this substance,
but had a N : P ratio of exactly 1:2.

COMPARISON OF ANALYSES OF DIFFERENT MONO-AMINO-DIPHOSPHATIDES HitrHerto [ISOLATED

Mono-amino -
Substance from Cuorin from heart Heparphosphatid diphosphatide

horse kidney muscle from liver from egg yolk
(MacLean )t (Erlandsen) (Baskoff) (MacLean)
Cc 60-4 61-33 61-12 59-12
H 10-2 9-02 8-95 9-44
N 1-04 1-01 1-23 0-812
Ls 4-45 4-47 4-0 3-59
I ack | 1M 1:1-5 i ee

These figures indicate that the substance obtained from horse kidney
is the same as that obtained from heart muscle by Erlandsen.

On saponification with alcoholic potash, 60 per cent. of the weight of
cuorin was obtained as fatty acids, which were solid at room temperature
and melted at about 45°.

They were composed of one solid saturated acid and two unsaturated
liquid acids; the solid acid was found to be stearic, while the liquid
acids, as in the case of lecithin, gave bromination products, indicating
the presence of acids having three or more double bonds. They will be
discussed, along with the acids of lecithin, in a later paper.

The other cleavage products consisted of a substance having the
reactions of glycerophosphoric acid, though the amount of barium
obtained on analysis of the barium salt did not quite agree with the
theoretical amount. Only a very small amount of a basic substance
precipitated by platinum chloride could be obtained, so it is certain that
the nitrogen is not represented by choline.

From these results it is clear that the acetone insoluble phosphatides
present in the ethereal extract of horse kidneys are of the same nature
as those described by Erlandsen from heart muscle. In spite of the
observation of Fraenkel that he could not obtain any evidence of the
presence of lecithin in the kidney, there can be no doubt that a small
amount is present in the ethereal extract. It will be shown later that

©

340 BIO-CHEMICAL JOURNAL

quite a considerable amount of lecithin is present in the alcoholic
extract.

The three substances present in the ethereal extract of kidney are
therefore lecithin, cuorin and a diamino-monophosphatide. Cuorin was
found in greatest amount, while only a small quantity of lecithin and a
very small amount of the diamino-monophosphatide was isolated.

Tue ALcoHoLic EXTRACT

Our knowledge of the phosphatides present in the alcoholic extract
of the different tissues is very unsatisfactory. According to Erlandsen,
by far the greater part of the extract consists of a phosphatide having a
Nese ratio OF 2.23 ;

This observation was based on the analyses of cadmium chloride
compounds, but he was unable to isolate the substance in a free state.
An attempt to liberate the phosphatide from the cadmium compound was
unsuccessful, the substance obtained differing in its constitution from the
theoretical composition deduced from the cadmium salt. In order to
explain this result, Erlandsen assumed that the manipulations required
for setting free the phosphatide resulted in a partial decomposition of
this body, but it is interesting to observe that the N : P ratio of the
substance obtained was 1°48 : 1.

Baskoff extracted the liver, first with alcohol, then with ether, and
lastly with 96 per cent. alcohol. This second alcoholic extract was
evaporated to small bulk in vacuo, and the residue dissolved in a small —
amount of alcohol. An insoluble fraction remained which consisted
chiefly of inorganic substances. The alcoholic filtrate was now evaporated
to dryness, and the residue extracted with ether, when a white substance
was obtained and filtered off. This substance contained 4°16 per cent.
nitrogen and 2°52 per cent. phosphorus = N:P 3°68:1, and was
regarded by Baskoff as a jecorin-like substance. The ethereal solution
was concentrated and precipitated with acetone, the precipitate dissolved
in ether and some alcohol added. A precipitate was now obtained which
on analysis gave 4°703 per cent. nitrogen and 1°93 per cent. phosphorus
= JNe Poin eal,

After filtering off this substance the ether-alcohol solution was
evaporated to dryness and the residue dissolved in ether and precipitated
by acetone. This fraction constituted by far the greater part of the
alcohol soluble substance; it was once again purified by ether and
acetone. This substance was now analysed and compared with a
corresponding body obtained by Erlandsen from heart muscle.

“wv

THE PHOSPHATIDES OF THE KIDNEY 341
Baskoff’s substance Erlandsen’s substance
I Il
P 3-39 % 3-28 3-01
N 3-91 % 3-63 3-53
. P:N Si P ay rol ge 1 2-55. ps 1 : 2-50 1: 2-59

These figures show that the substances contain nearly the same
amount of nitrogen and phosphorus, the N : P ratio in each case being
practically the same.

From this substance Erlandsen succeeded in isolating a cadmium
chloride compound which gave a N : P ratio of exactly 2: 1, while on
the contrary Baskoff’s cadmium compound had a N : P ratio of 1°57 : 1.

These results suggest that the substances thus obtained were not quite
pure, and evidence in support of this will be brought forward later.
Baskoft agrees with Erlandsen that the alcohol extract contains phos-
phatides with a higher nitrogen content than those of the ether extract,
though he was unable fully to substantiate Erlandsen’s results. It is
noteworthy that both the jecorin-like substances and the alcohol soluble
phosphatides all contain a high percentage of nitrogen.

TREATMENT OF ETHER EXTRACTED TISSUE

After thorough treatment with ether, the kidney substance was
extracted with alcohol, first at room temperature and then at 40° C.
This was repeated six times, the combined alcohol extracts evaporated off
under reduced pressure and the residue extracted with pure ether. A
considerable amount of substance was obtained which gave a very
opalescent solution with ether. This was left to stand, the supernatant
fluid treated with excess of acetone and the crude phosphatide obtained
dried in vacuo.

TREATMENT OF CRUDE PHOSPHATIDE MIXTURE

The dried crude substance, which weighed about 50 grams, was now
dissolved in ether, and gave a somewhat opalescent solution which soon
cleared up, leaving a small amount of a white insoluble precipitate. This
was filtered off, the ethereal solution precipitated by acetone, the
precipitate dried and extracted with a small amount of absolute alcohol.
A very turbid liquid was obtained. On centrifuging, a white
precipitate (a) separated, and a clear alcoholic solution was obtained.

This solution was evaporated to dryness under reduced pressure at
a low temperature, and the residue taken up in alcohol. The alcoholic

349 BIO-CHEMICAL JOURNAL

solution was at first somewhat turbid, but soon became quite clear. On -
standing in the ice-room overnight a white substance was precipitated
which, on separation by the centrifuge, was washed twice with a small
amount of cold alcohol = white precipitate (6).

This precipitate (b) was now mixed with precipitate (a), and the
whole extracted twice with hot acetone. The residue was only partly
soluble in ether and in alcohol.
| On analysis it was found to contain 6°2 per cent. nitrogen and 1°5
per cent. phosphorus; it was not further examined.

The alcoholic solution was again evaporated to dryness, and the
residue dissolved in a very small amount of alcohol by the aid of gentle
heat. On standing in the ice-room overnight some more white substance
separated out. After filtering, the alcoholic extract was evaporated to
small bulk and about five times its volume of pure ether added, a
precipitate consisting partly of a white powder and partly of a slimy
mass formed and was filtered off. This precipitate was washed with
ether, in which the white substance remained in suspension while the
slimy portion remained at the bottom of the vessel.

The ethereal suspension was syphoned off and left to stand, when the
white particles fell to the bottom of the glass. On separating the ether
a perfectly white substance was obtained which, on drying, formed a
white solid mass which was easily broken up to a white powder = white
powder (d).

The slimy mass obtained was dried in vacuo = substance (f). The
alcohol-ether filtrate was evaporated to dryness and the residue dissolved
in ether, when a perfectly clear solution was obtained after standing a
short time. This was precipitated by acetone, the precipitate again
dissolved in ether and reprecipitated by acetone. The substance obtained
gave a clear solution in alcohol, while in ether it was at first somewhat
cloudy, but on standing soon became quite clear.

This substance was now dissolved in alcohol and three times its
volume of acetone added. <A precipitate at once formed which was filtered
off, dissolved in ether, precipitated by acetone, and dried. This fraction
constituted by far the greater part of the phosphatides of the alcoholic
extract = phosphatide (m). On adding more acetone to the above
acetone-alcohol solution a slight cloudiness was in evidence, but on the
addition of great excess of acetone a white sticky substance separated,
which differed in appearance from substance (m) above. This will be
referred to as phosphatide (p).

THE PHOSPHATIDES OF THE KIDNEY 343

As it was thought that these two fractions might differ in
constitution, they were dried and analysed separately. The fraction (7m),
which was at first easily soluble in alcohol, became partly insoluble after
some time, while precipitate (p) remained soluble for a much longer time.

From the alcoholic extract four fractions were now obtained.

(1) Lecithin-like substance (m) about 16 grams.

(2) Lecithin-like substance (p) ,, 2°5 grams.
(3) White substance ... (d) ,, 3°1 grams.
(4) Shmy mass a | = oo prem.

SUBTANCE (7)

This was obtained as a yellowish white lecithin-like substance
soluble in alcohol, ether, chloroform, benzene and petroleum-ether. It
dried quickly to constant weight, was very hygroscopic, and had all the
general properties of lecithin.

Analyses
Nitrogen—
1°0460 gm. used 36°2 N/10 H,SO, = 4°9 per cent.
0°9027 gm. used 31°5 N/10 H,SO, = 4°9 per cent.
Average = 49 per cent. N.
Phosphorus—
0°4172 gm. used 27°48 c.c. N/2 NaOH = 3°65 per cent.
04691 gm. used 30 c.c. N/2 NaOH = 3°54 per cent.
Average = 3°59 per cent. P.
oP —'s.: I.

T’'rom the nitrogen phosphorus ratio it seemed as if this might
possibly be a pure phosphatide, and to determine this the cadmium
chloride compound was prepared.

Five grams of the substance were dissolved in alcohol, and an
alcoholic solution of cadmium chloride added as long as a precipitate
formed. The alcohol was then filtered off, the precipitate washed
thoroughly with alcohol, dried and analysed,

Analyses of cadmium chloride salt
Nitrogen—
0°7028 gm. used 11°8 c.c. N/10 H,SO, = 2°35 per cent.
0°3504 gm. used 6 c.c. N/10 H,SO, = 2°4 per cent.
Average = 2°3 per cent. N.

344 BIO-CHEMICAL JOURNAL

Phosphorus—

0°3264 gm. used 19°4 c.c. N/2 NaOH = 3'3 per cent.

0°2603 gm. used 15°9 c.c. N/2 NaOH = 3°4 per cent.
Average = 3°35 per cent. P.
N.S ab apa,

From this result it was obvious that the substance (m) was not a
-chemical entity, and that the precipitation with cadmium chloride had
separated off a large amount of the nitrogen. Analysis of 100 c.c. of the
alcoholic filtrate obtained from the cadmium chloride precipitation
showed that a very great excess of nitrogen was present. As cadmium
chloride does not precipitate phosphatides quantitatively out of alcoholic
solution, it is probable that part of the nitrogen was accounted for in this
way. On the other hand, some nitrogenous substance free from
phosphorus must have been present.

Analysis of filtrate

50 c.c filtrate used 31°5 c.c. N/10 H,SO, = 0°88 per cent. N.
50 c.c. filtrate used 12°6 c.c. N/2 NaOH = 0°14 per cent. P.
NP = 14 ei

These results proved that the phosphatide present could not be of
the nature of the diamino-monophosphatide found by Erlandsen in heart
muscle.

An endeavour was made to decompose the cadmium chloride
compound in order to investigate the phosphatide set free, but the
amount of substance at my disposal was too small to give a definite result.

Another sample of cadmium chloride phosphatide prepared from the
substance (m) was washed very thoroughly with alcohol and gave the
following figures : —

0°3590 gm. required 6°55 c.c. N/10 H,SO, = 2°50 per cent. N.
0°3120 gm. required 18°6 c.c. N/2 NaOH = 3°38 per cent. P.
Ne, Be:

It would thus seem that thorough washing of the cadmium compound
does not materially alter the result.

As already mentioned, part of the substance (m) became insoluble
in alcohol after being kept for some time. This was dissolved in ether,
precipitated by acetone and the nitrogen and phosphorus estimated. It
contained 2°35 per cent. nitrogen and 3°46 per cent. phosphorus =
N:P15:1. On adding an alcoholic solution of cadmium chloride to

THE PHOSPHATIDES OF THE KIDNEY 345

an ethereal solution of this substance, a precipitate was obtained which
had a N : P ratio of 1°2 : 1.

The inference drawn from these results was that the substance
present was possibly a lecithin with a N : P ratio of 1: 1, for since the
treatment with cadmium chloride caused such a marked difference in the
nitrogen percentage of the substance it was quite probable that an
impurity was present, all of which was not removed. Further
purification of the substance, however, by dissolving in ether and
precipitating with acetone did not appreciably change the nitrogen
percentage, and it was obvious that a pure substance could not be
obtained by this method.

SUBSTANCE (7)

This was also a lecithin-like body, but much whiter in appearance
than the fraction described above. It had the properties and solubilities
of ordinary lecithin.

Analyses gave the following results :—

Nitrogen

0°6384 gm. used 25°5 c.c. N/10 H,SO, = 5°4 per cent

0°7544 gm. used 28°25 c.c. N/10 H,SO, = 5°2 per cent.
Average = 53 per cent. N.

Phosphorus

0°5717 gm. used 28°35 c.c. N/2 NaOH = 2°7 per cent.

0°4220 gm. used 22°74 c.c. N/2 NaOH = 2°9 per cent.
Average = 2°8 per cent. P.
Nes = Arar F.

This substance differed materially from substance (m), the
percentage of nitrogen being much higher and that of phosphorus being
considerably lower. This result also suggested that the substance was
impure, but owing to lack of material a cadmium chloride combination
could not be investigated.

Some of these experiments were repeated with other samples of
phosphatides obtained by alcohol extraction in the manner described. It
was found, however, that fractions corresponding to (m) and (p) contained
different amounts of nitrogen from these fractions. This was taken as
strong evidence that the phosphatide of the alcoholic extract contained
impurities which had not hitherto been separated by any of the methods
in use at present. After many attempts at purification of this lipoid, a

346 BIO-CHEMICAL JOURNAL

method was at last found which solved all the difficulties and showed
that the different results obtained were really due to the presence of

nitrogenous impurities.

PURIFICATION OF THE PHOSPHATIDES OF THE ALCOHOLIC EXTRACT

In the course of this investigation it was observed that when the
residue of the alcoholic extract was shaken up with water, the fluid
became coloured, suggesting that part of the substance had passed into
solution. Since phosphatides are ordinarily insoluble in water, it was
thought that extraction with water might result in a purer substance
being obtained, since obviously something more than phosphatide was
present.

Part of fraction (m), described above, was therefore treated as
follows :—Several grammes were taken and rubbed up in a mortar with
a small quantity of water. When the substance was thoroughly broken
up, more water was gradually added, and a dilute aqueous emulsion
obtained. To this was added a small amount of pure acetone, when a
white substance at once separated out and floated on the surface of the
fluid. This was easily skimmed off, when the water-acetone fluid was
found to be yellow in colour. The white substance separated was
subjected to the same process several times in order to ensure access of
the water to the fatty lipoid particles. The final product was thoroughly
extracted with pure acetone, dried and analysed. ‘The result was
exceedingly satisfactory, as it showed that a great deal of nitrogen was
got rid of by this method, while at the same time the purified substance
hada WN : P ratio of nearly 1 : 1.

Analyses of substance (m), which had been treated three times with
water and acetone, gave results very similar to those obtained from
ordinary lecithin.

Analyses of purified substance
Nitrogen—

0°3989 gm. used 5°8 c.c. N/10 H,SO, = 2°1 per cent.

0°4770 gm. used 7 c.c. N/10 H,SO, = 2°06 per cent.
Average = 2°08 per cent. N.
Phosphorus—

0°3800 gm. used 26°6 c.c. N/2 NaOH = 3°88 per cent

0°2158 gm. used 15 c.c. N/2 NaOH = 3°9 per cent.
Average = 3°9 per cent. P.
N : P= 118 ; 1.

THE PHOSPHATIDES OF THE KIDNEY 347

Thus a substance having originally a N : P ratio of 3: 1 and
containing 4°9 per cent. nitrogen had, after being three times treated
as above described, only about 2 per cent. nitrogen and a N : P ratio
of 1:18 : 1.

After being emulsified and precipitated six times it gave the same
figures as ordinary lecithin.

Nitrogen
0°6516 gm. used 9°2 e.c. N/10 H,SO, = 1°97 per cent. N.
Phosphorus

0°2900 gm. used 21°3 c.c. N/2 NaOH = 4°07 per cent. P.
We P= Lor > ft.

This proved that the ether soluble lecithin-like substance present in
the alcohol extract of the kidney is really a monamino-monophosphatide
and not a diamino-monophosphatide as described by Erlandsen in heart
muscle. This point is further discussed in the following paper ‘ On the
purification of phosphatides.’

Wuitre Substance (d)

This white substance isolated as described formed on drying a
whitish yellow non-hygroscopic powder. Before making any attempt to
purify it, an analysis was made, which showed that it was probably an
impure substance.

Nitrogen
03076 gm. used 6°6 c.c. N/10 H,SO, = 3 per cent. N.

Phosphorus

0°4020 gm. used 13°45 c.c. N/2 NaOH = 1°85 per cent. P.
NE are 2-8,

This impure substance was partly soluble in benzene, and in
chloroform, from which it was precipitated by acetone; it was slightly
soluble in ether, and gave Lassaigne’s test for nitrogen and the
ammonium molybdate test for phosphorus. After boiling with weak
HCl it reduced Fehling’s solution. No glycogen reaction was obtained,
and Millon’s protein reaction was also negative. In H,O part of it seemed
to dissolve; the aqueous solution contained abundance of chlorides but
no sulphates. The substance was now shaken up with water several
times until the filtrate gave no chloride reaction. On testing the

348 BIO-CHEMICAL JOURNAL

aqueous filtrate it was found to contain a considerable amount of nitrogen.
This treatment of the substance with water was very tedious on account
of the difficulty of filtration. The addition of water to the powder gave
rise to a more or less sticky opalescent mass like thick starch paste,
which quickly closed up the pores of the filter paper. The use of the
centrifuge evaded this difficulty to some extent, but as the washing
proceeded, precipitation was very incomplete, probably due to the absence
of salts, and some loss of substance was unavoidable.

Analyses of the purified substance showed that it differed materially
from the crude material.

Analyses
01190 gm. substance gave 0°2980 gm. CO, = 68°3 per cent. C
and 0°1335 gm. H,O = 12°46 per cent. H.

Nitrogen
0°3586 gm. used 7°7 c.c. N/10 H,SO, = 3 per cent. N.

Phosphorus
0°2346 gm. used 14°55 c.e. N/2 NaOH = 3°44 per cent. P.
NP 1:93 sole
Thus a substance was obtained which had a nitrogen phosphorus
ratio of almost exactly 2 : 1, and was probably the same substance as
that described under the ether extract.

PROPERTIES

This white substance formed a tasteless and cdourless powder, which
was non-hygroscopic and preserved its properties after being exposed to
the air for a considerable time. It was practically insoluble in cold or
hot ether, but easily dissolved in ethyl alcohol, especially on heating;
from concentrated alcoholic solution it separated on cooling to room
temperature. It was insoluble in acetone but soluble in hot methyl
alcohol, from which it crystallised out on cooling, as in the case of ethyl
alcohol. In chloroform, benzene, glacial acetic acid and pyridine, it
dissolved at ordinary temperature ‘but very easily on heating. With
water it formed an opalescent starch-like mixture, from which it was
precipitated by the addition of acetone. On boiling for some time with
weak HCl it reduced Fehling’s solution, decolourised safranin, and gave
all the ordinary sugar reactions. It contained no sulphur.

The alcoholic solution was neutral and gave a precipitate with
cadmium chloride and with platinum chloride, also with lead acetate.

THE PHOSPHATIDES OF THE KIDNEY 349

On boiling with acids it yielded a substance which formed a double
salt with platinum chloride and gave the reactions of choline. The
amount obtained was much less than that represented by the nitrogen
present, amounting to somewhat less than 50 per cent. of the theory.
No glycerine could be obtained, but owing to the relatively small amount
of substance isolated and the difficulties of testing for this substance, no
definite statement as to its absence can be made.

On hydrolysis, about 65 per cent. of fatty acids were obtained. The
existence in the tissues of substances similar to the above was first
suggested by Thudichum, who isolated from the brain three substances
which he named apomyelin, sphingomyelin and amidomyelin.

Substances of this nature were afterwards described by Stern and
Thierfelder, who isolated a diamino-monophosphatide from egg yolk, and
by Erlandsen, who found a somewhat similar substance in heart muscle.
Dunham and Jacobson also obtained a substance from ox kidney, which
had a nitrogen phosphorus ratio of 3 : 1, but in other respects was very
similar to certain of the above substances.

It has already been shown that the so-called diamino-mono-
phosphatide isolated by Erlandsen was probably a mixture, as from a
similar substance prepared from the kidney I succeeded in obtaining
ordinary lecithin with a nitrogen to phosphorus ratio of 1 : 1.
Erlandsen’s substance also differed from the above in its physical
properties, and especially in the fact that it was soluble in ether. The
bodies described by Thudichum as sphingomyelin and amidomyelin were
both in many respects similar to the substance isolated by me. In ether
and cold aleohol they were only very slightly soluble and _ were
precipitated from alcoholic solution by cadmium chloride. Amidomyelin
was further described as a substance which on drying in vacuo formed a
white mass which could be easily powdered, while sphingomyelin did not
form a sticky mass, but was also capable of forming a powder. The
properties of amidomyelin are not given. The diamino-monophosphatide
described by Stern and Thierfelder agrees in analyses and properties
almost exactly with my substance, the most marked difference being that
it did not reduce Fehling’s solution on boiling.

As the reduction with Fehling’s solution is not very marked and is
only obtained distinctly when a comparatively large amount of the
substance is taken, it is not possible to say at present whether this
reaction is due to traces of impurity or is really dependent on a
carbohydrate radicle incorporated in the molecule.

&.

350 BIO-CHEMICAL JOURNAL

The following table shows the results of analyses of these different
diamino-monophosphatides.

Substance
from Substances from brain
Substance heart (Thudichum) Substance
Substance from egg muscle from ox
from yolk (Erlandsen) kidney
horse (Stern and reckoned (Dunham
kidney Thier- from Apo- Sphingo- Amido- and
(MacLean) felder) CdCl, myelin myelin myelin Jacobson)
compound
C 68-19 68-13 59-48 67-01 65-37 62-4 67-12
H Be : 12:37 12°14 9-42 11-35 11-29 — 11-54
N ane ax 3-0 2-77 3°47 3-0 2-96 -— 2-84
12 3-44 3-22 3-84 3-23 3:24 — 2-18
N : P ratio oe 1-93: 1 19:1 P28 II Daal PAI — 29:1
CARNAUBON

This substance, which was described by Dunham and Jacobson as a
triamino-monophosphatide, possesses all the properties of the substance
isolated from the horse kidney. It is the most fully described and
investigated of this class of substances and its composition is given as
similar to lecithin, but having, instead of glycerine, another body—
galactose—to which the fatty acids are attached. The acids isolated were
carnaubic acid, stearic acid, and palmitic acid. Though described as a
triamino-monophosphatide, the authors agree that the N : P ratio was
often about 7 : 3, and consider the possibility of the substance being
impure. In the light of the results of the present investigation, it seems
probable that the substance was a diamino-monophosphatide, which
treatment with water as above described would have purified.

If so, the question of these ether insoluble phosphatides is much
simplified, for the available evidence suggests that all the substances of
this nature hitherto isolated are of the nature of diamino-monophos-
phatides, and are probably identical in structure with pure carnaubon.
In order to make certain that the substance isolated from horse kidney
was really a diamino-monophosphatide, two more samples were prepared
by different methods.

SAMPLE [TI

This sample was prepared in the same way as that described by
Dunham and Jacobson, only it was finally extracted with water. Fresh
kidneys were treated with hot 95 per cent. alcohol and the extract

THE PHOSPHATIDES OF THE KIDNEY 351

discarded. The residue was extracted first with 95 per cent., and then
with 85 per cent. alcohol, and the mixed solutions filtered hot. They
were then left to stand in the ice-room. The precipitate formed was
collected, dried and dissolved in benzene, the solution filtered, and the
benzene evaporated off by distillation under reduced pressure. The syrup
obtained was thoroughly mixed with ether and left to stand, when a white
precipitate separated. The ether was syphoned off and the residue
washed with more ether. It was then dissolved in hot alcohol and left to
stand, when a white substance separated out. This was treated with
ether, dried in vacuo and extracted with several changes of water as
described above. ‘This final treatment with water is important, as
otherwise a pure substance is not obtained. The resulting product gave
the following figures on analysis :—

Analyses
0°1025 gm. substance gave 0°2567 gm. CO, = 68°29 per cent. C.
and 0°1132 gm. H,O = 12°26 per cent. H.
Nitrogen—
0°3806 gm. used 7°8 c.c. N/10 H,SO, = 2°9 per cent. N.
Phosphorus—
0°2380 gm. used 14 c.c. N/2 NaOH = 3°3 per cent. P.
bese Pad es 9 ee

SAMPLE IIT

This substance was prepared as follows:—The crude alcoholic
extract of the kidneys was evaporated to a syrup, excess of acetone added,
and the precipitate rubbed up with ether, when an opalescent mixture
was obtained. On centrifuging, a white precipitate separated. This
white substance was thoroughly extracted with ether and centrifuged
three times. The residue was dissolved in hot alcohol and filtered hot.
The filtrate, however, was not quite clear, so it was again filtered and a
perfectly clear solution obtained; a small quantity of a sticky substance,
insoluble in alcohol and in ether, remained on the bottom of the flask.

This clear solution was allowed to stand, and the white substance
which separated filtered off and dried. It was then emulsified with water,
and after the addition of some acetone, left to stand for twenty-four hours,
when a flocculent white precipitate was obtained. This process was
repeated twice and the final product dried and analysed. It had the same
composition as the two substances formerly described.

e.

352 BIO-CHEMICAL JOURNAL

Analyses

0°1321 gm. substance gave 0°3292 gm. CO, = 68°0 per cent. C

0°1321 gm. substance gave 0°1475 gm. H,O = 12°4 per cent. H.
Nitrogen—

0°1591 gm. used 3°3 c.c. N/10 H,SO, = 2°9 per cent. N.

Phosphorus—
01300 gm. used 8 c.c. N/2 NaOH = 3°4 per cent. P.
Ne IRS ae
Both these substances had the same properties as the one first
described.
A comparison of their analyses shows that this substance was
certainly a diamino-monophosphatide.

COMPARISON OF THREE SAMPLES OF DIAMINO-MONOPHOSPHATIDE

No. 1 No. 2 No. 3 Average
C 68:3 68-29 68-0 68-19
H 12-46 12-26 12-4 12-37
N 3-0 2-9 2-9 2-93
P 3°44 3°3 3-4 3°38
N : P ratio 1:9:1 1-95: 1 1:9:1 1-92: 1

Since the white substance present in the ether extract agrees with
this in composition and properties, it shows that this substance is identical
with that described above.

Consideration of the properties of this substance suggest that part of
it would be present in the ethereal extract.

GENERAL SUMMARY

The ethereal extract of the horse kidney contains three phosphatides
lecithin, cuorin, and a diamino-monophosphatide. The subsequent
alcoholic extract also contains a diamino-monophosphatide, which is the
same as that obtained in very small amount from the ether extract. This
substance is but very slightly soluble in ether, but there is a good deal of
evidence to show that it is fairly soluble in an ethereal solution of
lecithin. Thus, if the crude phosphatide residue of the ethereal extract
of kidney is taken and extracted with ether, a very opalescent mixture is
at first obtained, but as the lipoids present dissolve, the solution becomes
much clearer. On centrifuging, a perfectly clear ethereal solution is
obtained, but if the ether be evaporated off and the residue again extracted
with ether, the solution so obtained is by no means clear, and some more

THE PHOSPHATIDES OF THE KIDNEY 353

white substance can be separated by the centrifuge; if, however, the
solution is allowed to stand for some time before centrifuging, it may
become quite clear. Again, when it seems as if all the white substance
had been got rid of, it often happens that the addition of ether to the
lipoid residue gives a solution which becomes clear only after some little
time. This property of the diamino-monophosphatide suggests the
probability that every sample of lecithin prepared from tissues
containing this substance was contaminated by it, and it is difficult to see
how complete separation can be achieved, though careful treatment with
ether and quick centrifuging undoubtedly gets rid of all but traces. This
substance is entirely different from lecithin and other lipoids in its
physical properties, as it forms a white non-hygroscopic powder, and not
a more or less sticky mass, as is seen in the case of lecithin and cuorin.

Owing to its greater solubility in alcohol, the majority of it is found
in the alcoholic extract, but it is present only in comparatively small
amount, and is not to be confounded with the substance found by
Erlandsen in the alcoholic extract of heart muscle. Erlandsen’s
substance possessed all the properties of ordinary lecithin and was quite
soluble in ether; it constituted also the chief bulk of the phosphatides
present in the alcoholic extract. A corresponding substance is present in
the kidney in comparatively large amount, but its analyses did not quite
compare with that of heart muscle, and on subjecting it to purification
by means of H,O and acetone, it was found to be of the nature of ordinary
lecithin, with a nitrogen to phosphorus rato of 1 : 1.

Thus the lipoids of the alcoholic extract of kidney are the same as the
alcohol soluble ones contained in the ethereal extract; the alcohol insoluble
substance cuorin, very soluble in ether, is naturally present in the ether
extract. The alcohol extract contains other bodies with a high percentage
of nitrogen, but these are soluble in water and may be separated in the
manner described. The water soluble substances contain only a small
amount of phosphorus and are not of the nature of ordinary phosphatides ;
their nitrogen percentage is very high. From these observations it
follows that the subject of the kidney lipoids has been much simplified,
and an extension of these results to other organs and tissues will probably
show that many of the substances hitherto described as containing large
percentages of nitrogen are really mixtures of simpler bodies with these
nitrogenous water-soluble bodies mentioned above. Other tissues such

as heart muscle, liver and egg yolk, are at present being investigated from
this point of view.

354 BIO-CHEMICAL JOURNAL
CoNCLUSIONS

(1) The acetone insoluble phosphatides of the horse kidney are
lecithin, cuorin, and a diamino-monophosphatide—carnaubon. .

All these are contained in the primary ether extract, while the
subsequent alcoholic extract contains lecithin and carnaubon; cuorin,
being insoluble in alcohol, is not present here. Thus, so far as solubility
allows, the lipoids of the ether extract are the same as those present in
the alcoholic extract.

(2) In the kidneys investigated, the ethereal extract contained much
more cuorin than lecithin.

(3) A method is described whereby the complicated alcoholic extract
can be purified from a nitrogenous substance. In all tissues investigated
hitherto, thé phosphatides of the alcoholic extract must have been
contaminated by this substance; this explains many of the divergent
results obtained.

(4) The diamino-monophosphatide isolated has all the properties of
a substance found in ox kidney by Dunham and Jacobson, and called by
them carnaubon. They state that this substance is a_ triamino-
monophosphatide.

(5) It is probable that carnaubon is not a tri-, but a diamino-
monophosphatide, and that the methods used for its isolation by Dunham
and Jacobson were inefficient to obtain a pure substance. On extracting
a substance obtained by the method employed by the above investigators
with water, a diamino-monophosphatide was obtained.

REFERENCES

Erlandsen, Ztsch. f. physiol. Chem., LI, p. 71, 1907.

Stern and Thierfelder, Ztsch. f. physiol. Chem., LIII, p. 379, 1907.
MacLean, Ztsch. f. physiol. Chem., LVII, p. 304, 1908.

Frinkel, Bio-Chem. Ztsch., XIX, p. 262, 1909.

Baskoff, Ztsch. f. physiol. Chem., LYVII, p. 395, 1908.

Thudichum, Physiological Chemistry of the Brain.

Dunham and Jacobson, Zésch. f. physiol. Chem., LXIV, p. 303, 1910,

See ub Se GO

Bd5
ON THE PURIFICATION OF PHOSPHATIDES
By HUGH MacLEAN.
From the Lister Institute, Bio-Chemical Department
(Received June 19th, 1912

In the preceding paper on ‘ The Phosphatides of the Kidney,’ it has
been shown that the chief part of the lipoid occurring in the alcoholic
extract of horse kidney is not a diamino-monophosphatide as described by
Erlandsen! in heart muscle, but a monamino-monophosphatide which
contains nitrogenous impurities. The phosphatide was freed from these
impurities by repeatedly emulsifying with water and adding some acetone
when the lecithin separated out, leaving the contaminating substance in
solution. In order to obtain fairly complete separation of the phosphatide
from the water-acetone mixture, it is necessary to have some salt present,
otherwise acetone fails to precipitate the lipoid from its aqueous emulsion ;
a trace of sodium chloride suffices, and it is only necessary to add a few
drops of a weak solution of this substance to the water used for the
emulsion in order to ensure good separation. It is important to make
the emulsion with water which contains traces of sodium chloride, for the
addition of salt after the emulsion is formed gives very unsatisfactory
results.

The substance can be very easily emulsified by adding at first a very
little water and rubbing up thoroughly in a mortar; more water is
gradually added and after each addition the whole is well rubbed up. In
this way a very fine emulsion approaching to a colloidal solution is
formed, and the best results are obtained when great care is taken to have
this emulsion as fine as possible. Both the nitrogen impurity and lecithin
are precipitated by acetone, but very much less suffices to separate the
lecithin than is required for the precipitation of the other substance;
only the minimum quantity of acetone necessary to separate the
phosphatide is added. The lecithin rises to the top of the liquid and can
be easily skimmed off, while any of the impurity that may be
precipitated remains suspended in the liquid and gradually falls to the
bottom of the vessel. This separation of the phosphatide is by no means
quantitative, and the water-acetone solution always contains a certain
amount of lecithin. With care, however, the loss in this way is
comparatively small.

356 BIO-CHEMICAL JOURNAL

As the method of purification had been tested only on kidneys which
had been already extracted with ether, some experiments were made on -
other tissues. At the same time an extension of the kidney experiments
was carried out. Since the phosphatide present in the ether extract was
the same as that contained in the subsequent alcoholic extract, it is
obvious that the primary ether extract, the secondary alcoholic extract,
and a control alcoholic extract, without primary treatment of the tissues
with ether, should on purification all give a substance with an
approximate nitrogen to phosphorus ratio as 1 : 1. This proved to be
the case.

KipNEY EXPERIMENTS

Some kidneys were minced and dried in the usual way ‘Two
portions of the dried substance were then extracted as follows :—

(a) With ether.

(6) With alcohol.

(c) With alcohol after ether (a extracted subsequently with
alcohol).

(a) Ether extract

The ethereal extract was treated as before described in order to
separate fat, cuorin and cholesterol. The final lipoid obtained was
then subjected to the acetone-water purification, dried and analysed. The
results show that the substance was a lecithin.

Analyses

0°4242 om. used 6 c.c. N/10 H,SO, = 1°9 per cent. Nitrogen.
0°2814 gm. used 21 c.c. N/2 NaOH = 4'1 per cent. Phosphorus.
ING aaa ke

(6) Alcohol extract without ether

The dried tissue was extracted six times, first at room temperature,
then at 40° C. and the different extracts mixed together On standing
some time a white precipitate formed which was filtered off. The clear
extract was now evaporated to small bulk at 40° C. under reduced
pressure. On adding ether to the residue a very muddy liquid was
obtained, which was treated with excess of acetone in order to get rid of
cholesterol and fat. The precipitate obtained from the alcohol-ether
mixture was treated several times with cold acetone and finally dried in
vacuo over H,SO,. This crude substance was dissolved in a small amount

ON THE PURIFICATION OF PHOSPHATIDES 857

of ether and centrifuged, when a white precipitate was obtained. The
precipitate was discarded and the clear supernatant fluid treated with
acetone. The precipitate was again thoroughly extracted with acetone,
dissolved in alcohol and left to stand in the ice-room overnight. <A
precipitate formed which was filtered off. This process was repeated
until a concentrated alcoholic solution produced little or no precipitate
when left to stand in the ice-room. The alcohol was then evaporated off
and the residue extracted with ether. A fair amount of a slimy
substance (K) remained behind at the bottom of the vessel. The ether
solution was now treated with acetone and the precipitate dried. It was
then emulsified and precipitated with acetone eight times, extracted with
acetone, dried and analysed. It gave the following figures.

Analyses
0°4211 gm. used 6°2 c.c. N/10 H,SO, = 2°0 per cent. Nitrogen.
0°2204 gm. used 16°5 c.c. N/2 NaOH = 4'1 per cent. Phosphorus.
Ny s-2 GOS:
This substance was therefore a monamino-monophosphatide and so
far agreed with that obtained from the ether extract.

(c) Extract with alcohol after ether

The different processes utilised for the separation of the phosphatide
are the same as described for the alcohol extract above. The final
product obtained was treated with water and acetone eight times.

Analyses

0°6241 gm. used 9°5 c.c. N/10 H,SO, = 2°71 per cent. Nitrogen.
0°3216 gm. used 24 c.c. N/2 NaOH = 3°9 per cent. Phosphorus.
Nee PS ees
A comparison of these three substances shows that the analytical
figures are practically the same for all three preparations. The slight
differences are not greater than can be accounted for by experimental

error.
CoMPARISON OF THREE PREPARATIONS
Ether extract Alcohol extract Alcohol extract after ether
I be
Nitrogen... 1-9 2-0 2-1 1-97
Phosphorus... 4-1 4-1 3-9 4:07

N:P ratio... Ms | 1-08: 1 1-2:1 1-07: 1

* Sample described in paper on ‘ The phosphatides of the kidney.’

358 BIO-CHEMICAL JOURNAL

The results prove that the chief phosphatide of the kidney is lecithin
or a lecithin-like substance, and that the method of purification adopted
is capable of separating off the impurities associated with this

phosphatide.
MUSCLE

Experiments similar to the above were carried out on horse muscle,
and three substances corresponding to the fractions described above were

obtained.
(a) Ether extract.

(b) Alcohol extract without ether.
(c) Alcohol after ether.

The treatment of the muscle tissue prior to the separation of the
different fractions was almost entirely the same as that described for
kidney. It is interesting to note that while much ‘ white substance ’
was present in the kidney alcoholic extract, little or none of the diamino-
monophosphatide was found in muscle.

(a) Ether extract.

The phosphatide present, after being separated in the usual way and
purified three times with water and acetone, gave the following figures.

Analyses
0°6212 gm. used 8°5 c.c. N/10 H,SO, = 1°9 per cent. Nitrogen.
0°2164 gm. used 15°5 c.c. N/2 NaOH = 4:0 per cent. Phosphorus.
N > P= ho. 1.
(b) Alcohol extract without ether.

The crude alcohol extract here contained a great amount of a slimy
substance insoluble in ether, similar to a corresponding substance
mentioned under kidney. The final impure phosphatide obtained was
purified with acetone and water in the usual way, and samples examined
at different stages in order to determine how many times it was necessary
to emulsify the substance. The crude product, which contained
originally nearly 4°0 per cent. of nitrogen, gave the following results : —

Substance emulsified and precipitated once.

Analyses

0°3562 gm. used 5°7 c.c. N/10 H,SO, = 2°24 per cent. Nitrogen.
0°4245 gm. used 30 c.c. N/2 NaOH = 3°9 per cent. Phosphorus.
NEE eae

Thus a single treatment reduced the nitrogen percentage from

ON THE PURIFICATION OF PHOSPHATIDES 359

4 per cent. to 2°24 per cent., and gave a substance having a N : P ratio
nearly the same as that of lecithin.

Although this sample of crude phosphatide had been prepared in the
usual way, previous to the emulsification treatment it contained over
50 per cent. of its weight of a substance soluble in water. This was
obtained from the water-acetone solution by evaporation = substance (kK).
The following results were obtained after the phosphatide had been
purified four and eight times respectively.

Substance emulsified and precipitated four times.
Analyses
0°3802 gm. used 5°3 c.c. N/10 H,SO, = 1°95 per cent. Nitrogen.
0°2609 gm. used 18°7 c.c. N/2 NaOH = 4°14 per cent. Phosphorus.
Ne Ps 1042,

Substance emulsified and precipitated eight times.
Analyses
0°3750 gm. used 51 c.c. N/10 H,SO, = 1°9 per cent. Nitrogen.
0°1790 gm. used 12°8 c.c. N/2 NaOH = 3°97 per cent. Phosphorus.
i Ee sf.

From these results it appears that the first purification by this
method is sufficient to remove all but a small amount of impurities.
Naturally it becomes much more difficult to get rid of these traces, but
the substance obtained after four precipitations was just as pure, judging
from the N : P ratio as that obtained after eight precipitations. If care
is taken to make a good emulsion, it is probable that only traces of
impurities remain after four successive treatments. A fine emulsion is
essential in order to provide access of the water to impurities enclosed in
the fatty globules.

(c) Alcohol extract after ether.

This was purified six times in the usual way and gave the same
figures as the other samples.

Analyses
06428 gm. used 8 c.c. N/10 H,SO, = 1°8 per cent. Nitrogen.
0'5069 gm. used 3 c.c. N/2 NaOH = 3°9 per cent. Phosphorus.
ee te Qe, A 8,
The chief phosphatide of muscle is therefore a lecithin, and the lipoid

360 BIO-CHEMICAL JOURNAL

of the alcoholic extract is the same as the alcohol soluble one present in
the ether extract

a monamino-diphosphatide. In muscle a great amount
of water soluble substance is present in the alcohol extract, but this does
not appear to be a lipoid, though it is precipitated by excess of acetone
from both its alcohol and aqueous solutions. A comparison of the
analytical figures with those obtained from the purified lecithin of
kidneys suggests that the phosphatide present in both cases is the same.

On the nature of the nitrogenous impurity

As already described, on emulsifying the phosphatide obtained from
the alcohol extract of muscle and precipitating with acetone, a
considerable amount of material remained in solution. .This substance
was dark in colour, sticky to the touch, and exceedingly tenacious; on
drying it formed a hard substance, but could not be powdered. It was
exceedingly soluble in water, and in alcohol containing traces of water,
but much less soluble in absolute alcohol. It appeared to be absolutely
insoluble in ether, by which it was precipitated from solution as a dark,
sticky mass. On adding some alcohol to its clear alcoholic solution, a
faint white precipitate formed, causing an opalescence which soon cleared
up again.

Analyses

Nitrogen—

0°4202 gm. used 18 c.c. N/10 H,SO, = 6°0 per cent. Nitrogen.

0°2231 gm. used 9°8 c.c. N/10 H,SO, = 6°15 per cent. Nitrogen.
Phosphorus—

0°3513 gm. used 9°1 c.c. N/2 NaOH = 1°44 per cent. Phosphorus.

Another sample gave 6 per cent. of nitrogen and only 0°5 per cent.
phosphorus, so that it is probable that the latter is present as an impurity.
On standing, its aqueous solution deposited small minute white round
balls which were but slightly soluble in cold water, but dissolved easily
on heating. ‘The substance was practically insoluble in cold, but
somewhat soluble in hot alcohol. In ether, chloroform and benzene, it
was insoluble. With mercuric chloride it gave a voluminous precipitate.
On recrystallising twice from water and alcohol, a pure white substance
was obtained. This contained no phosphorus. Analysis gave the
following figures : —

0°0934 gm. used 16°6 c.c. N/10 H,SO, = 24°9 per cent. Nitrogen.

Its melting-point was indistinct, but the high nitrogen content and
other reactions suggest that it is of the nature of a purin base. Further

ON THE PURIFICATION OF PHOSPHATIDES 361

experiments on this substance are in progress. After the separation of
the white substance, the residue obtained by evaporation of the mother
liquor was treated as follows:—An alcohol solution was made and
ether added when a _ precipitate was obtained; this was repeated
twice. The substance was now dissolved in water and again
precipitated by acetone; the residue was then extracted thoroughly
with acetone and ether and dried. On analysis it was found
to contain 6 per cent. nitrogen and only about 0°3 per cent. phosphorus.
As it has been found that tissues contain a substance which cures
beri-beri, some of this material was given to pigeons suffering from
polyneuritis in order to test its effect. The result was very marked, as
three pigeons exhibiting very severe symptoms were completely cured in a
short time. Partially purified lecithin had a certain effect on these
pigeons, but a dose of 1 gm. of such lecithin resulted only in an
amelioration of the symptoms and no cure was obtained. As many
observers have placed it on record that lecithin is more or less efficient in
curing beri-beri, it is probable that here we have an explanation of their
results. The curative substance is not lecithin, but is present in ordinary
lecithin as an impurity. It can be separated, however, by the method
above described, and a lecithin so purified has but little effect in curing
beri-beri. The whole subject is being investigated and the results will
be published at a later date.

CONCLUSIONS

The alcohol soluble phosphatide of kidney and muscle is lecithin
with a N : P ratio of 1 : 1.

(1) Though substances containing different amounts of nitrogen are
obtained in the different extracts, the application of the method
described—emulsification and precipitation with acetone—gives a single
product of the nature of an ordinary lecithin. No alcohol soluble
substance having a higher percentage of nitrogen than that of lecithin
has been found.

(2) From an aqueous extract of the nitrogenous impurity of lecithin
a substance of basic nature crystallises out; after the separation of this
substance the mother lquor is very effective in curing polyneuritis
(beri-beri).

(3) This explains the anomalous results of many observers who
have endeavoured to cure pigeons suffering from beri-beri by lecithin.

REFERENCE
1. Erlandsen, Zésch. f. physiol. Chem., LI, p. 71, 1907.

362

ON THE RELATIONS OF PHENOL AND META-CRESOL TO
PROTEINS; A CONTRIBUTION TO OUR KNOWLEDGE
OF THE MECHANISM OF DISINFECTION

By E. A. COOPER, B.Sc., Beit Memorial Fellow, formerly Jenner
Scholar, Lister Institute of Preventive Medicine.

(Received July 2nd, 1912)

CONTENTS
PAGE
Introduction and method of study ... see me B56 as 500 coe OD
I. The influence of concentration upon the distribution of phenol between water
and proteins—
(a) Gelatin ... se ie ace as Pe ass “os << loos
(b) Dissolved egg-albumen =a sie wise aes an ... 9368
(c) Heat-coagulated ego-albumen a SB “ies soe Bo | Ue
(d) Heat-coagulated egg-globulin ie so soe as 5 Ue
(e) Casein... Se a ae sia sid Sc one acclh gO:
II. The influence of temperature upon the distribution of phenol between water and
proteins—
(a) Dissolved egg-albumen ... ssi a an ais sac once OMG
(b) Heat-coagulated egg-albumen ... 300 abe nee 006 mo
(c) Gelatin... sa she zat re sad igs aa sisiet mame TINO
Ill. The influence of alcohol upon the distribution of phenol between water and
heat-coagulated egg-albumen ... 5H s0¢ sec act sec «oa rOA
IV. The influence of hydrochloric acid upon the distribution of plo between
water and heat-coagulated egg-albumen BEE : Raa. ite!
V. The influence of concentration upon the distribution of m-cresol between water
and proteins—
(a) Gelatin... See we se as was sa bar OUD,
(6) Dissolved egg- apanee Soe Ses es ae des wo O80
VI. The solubility of proteins in phenol and m-cresol ... ale sie sha ate) OO

VII. The application of some of these facts to the theory of phenol disinfection ... 382
VIII. Summary and conclusion wes “an ce et oe: 3% net as) oso

In the case of such substances as formaldehyde, halogens, mercuric
chloride, oxidising agents, acids and alkalies, it is possible to understand
why they should behave as disinfectants, as they all form chemical
combinations with proteins. We have, however, very little knowledge of
the relations of phenol and the cresols to proteins, so that it is at
present not possible to understand the nature of the germicidal action of
these common disinfectants.

PHENOL AND META-CRESOL 363

Reichel,! in 1909, showed that heat-coagulated serum and egg-white
absorbed phenol from its aqueous solution in an amount directly
proportional to the concentration and that the process was reversible.
He also showed that Bacillus pyocyaneus absorbed phenol, and that the
addition of sodium chloride increased the uptake by both bacilli and heat
coagula and also increased the bactericidal action of phenol.

Herzog and Betzel,? in 1911, found that yeast reversibly absorbed
phenol in an amount, which relatively decreased with rise in concentra-
tion. A rise in phenol concentration, on the other hand, increased the rate
of disinfection of yeast-cells to a degree disproportionate to the increase in
concentration. They conclude that the absorption of the germicide is the
first phase in the process of disinfection, and that this is followed by the
chemical action of the germicide upon the cells, which is not affected by
concentration in the same way as the absorption.

The object of the investigation described below was to obtain further
knowledge of the relations of phenol and m-cresol to proteins with a view
to understanding the nature of their germicidal action.

Tue Metuop or STUDY

(a) The estimation of phenol and cresol

The method used was that described by Lloyd? in 1905. It consists
in the bromination of phenol in the presence of a large excess of
hydrochloric acid, under which condition three bromine atoms are
quantitatively introduced into the phenol molecule.

Some further experiments showed that it was also possible to employ
this method for the accurate determination of meta-cresol. Accurate
results, however, could not be obtained with ortho- and para-cresol. The
relative positions of the hydroxyl and methyl groups in the cresol
molecule evidently influence the entrance of bromine atoms into the
benzene nucleus.

(6) The materials used

The proteins used were gelatin, precipitated casein, dialysed
crystalline egg-albumen and heat-coagulated egg-globulin (the fraction of
egg-white precipitated by 1/2 saturated ammonium sulphate).

The gelatin (best gold label, Swiss) was first washed in running water
for twenty-four hours. It was then air dried for twenty-four hours, cut
into slips, and dried in a muslin bag at 110° C.

364 BIO-CHEMICAL JOURNAL

Merck’s casein, prepared according to Hammarsten, was used in the
experiments with casein.

The egg-albumen was prepared in the crystallized state from egg-
white by the method of Hopkins and Pinkus, described in 1898,4 and was
dialysed for ten days, first of all, in tap-water and finally in distilled
water. The heat-coagulum was prepared by heating the dialysed solution
of the protein slightly acidified by a few drops of 1 per cent. acetic acid.
The coagulum was well washed and dried at 105° C.

The globulin fraction of egg-white was separated from the albumen
as far as possible by several precipitations with half-saturated ammonium
sulphate. The globulin was then dissolved in very dilute ammonium
sulphate, coagulated by heat, washed free of ammonium sulphate and
dried at 105° C..

(c) The experimental methods

Experiments with gelatin.—A definite amount of gelatin in the form
of slips about one inch by three inches was suspended in a known volume
of an aqueous solution of phenol of estimated strength. When the
maximum amount of phenol had been taken up by the protein—the
equilibrium-time having been determined by some preliminary experi-
ments—the strength of the aqueous phenol solution was again estimated.
By a comparison of the initial and final concentrations the distribution
of phenol between water and protein was determined. New equilibria
with phenol between water and gelatin were obtained by withdrawing a
known volume of the water-phase and replacing this with an equal volume
of water or phenol solution.

Experiments with casein, egg-albumen and egg-globulin.—The
suspended casein was so fine that it could not be completely removed by
filtration. Accordingly, in the experiments with it, and also with egg-
albumen and globulin, the method introduced by Moore and Bigland®
was employed. This consists in restricting the protein to one region of
the water-phase by means of a piece of dialysing paper, which was
crumpled into the shape of a bag and enclosed in a stoppered bottle.
Phenol was found to pass the dialyser readily. When equilibrium was
attained, samples were removed for analysis from the portion of the water
phase outside the dialyser. In other respects the procedure was similar to
that adopted in the experiments with gelatin. All experiments were
carried out at 20° C. in a thermostat, except where otherwise stated.

PHENOL AND META-CRESOL

v2
fo)
or

THE EXPERIMENTAL RESULTS

I

Ture INFLUENCE OF CONCENTRATION UPON THE DISTRIBUTION OF PHENOL
BETWEEN WATER AND PROTEINS

(a) Gelatin

Phenol below concentrations of 2°6 per cent. caused no visible change
in gelatin. When immersed in concentrations between 2°6 and 5 per cent.
the gelatin became white and semi-transparent but did not lose its shape,
and in concentrations above 5 per cent. it contracted, lost its form, and
settled on the bottom of the vessel as a viscous mass. The precipitated
protein was soluble in hot 5 per cent. phenol solutions, but on cooling the
gelatin was reprecipitated, having lost the property of setting. These
changes were reversed when the supernatant liquid was poured off and
replaced by water. In the case of the effects of concentrations less than
5 per cent., the protein quickly assumed its original appearance, except
that it had acquired the property of swelling considerably, just as it does
after immersion in acid. The reversibility was less rapid when the gelatin
had been immersed in phenol concentrations above 5 per cent. The
recovered gelatin was still soluble in warm water, and had regained the
property of setting when the solution was cooled.

The time required for equilibrium.—The maximum amount of
phenol was taken up by gelatin from concentrations less than 2°6 per cent.
in two minutes. The reverse process was complete in fifteen minutes.

From concentrations of 2°6 per cent. to 5 per cent., namely, those
which visibly changed the gelatin but did not prevent its suspension in
water, the uptake of phenol was complete in twenty-four hours, and the
reverse process was finished in the same time.

From concentrations of 5 to 7 per cent., causing the precipitation of
the protein on the bottom of the vessel, the maximum amount of phenol
was absorbed in forty-eight hours. The reverse process, however,
required five weeks for its completion.

In concentrations from 0 to 2°6 per cent., at which point the gelatin
was precipitated, the phenol was distributed between water and protein
according to the partition-law. It was about three times as soluble in
gelatin as in water.

At concentrations of about 2°6 per cent. the gelatin became white and
opaque, and a greatly increased proportional uptake of phenol occurred.

&
366 BIO-CHEMICAL JOURNAL
TABLE [|
EQUILIBRIA

10 grams gelatin ; 250 c.c. phenol solution

Distribution-ratio

Water-phase Amount of phenol Amount of
in grams phenol held by
Initial phenol Final phenol absorbed by Condition of 1 gram gelatin
concentration concentration 1 gram gelatin protein SS SS
(grams per (grams per (by Amount of phenol
100 c.c.) 100 c.c.) difference) held by 1 gram
water
Experiment I—
B 0-990 1-039 0-026 Unprecipitated 2-5
(B) 1-162 1°237 0-039 3 3-2 |
1-813 1-729 0-047 34 2-7 ; Mean
2-167 1-936 0-058 53 3-0 | =2-9
2-365 2-277 0-069 thts 3-0
2-801 2-609 0-117 Precipitated 4-5
3-077 2-847 0-175 bg 6-2
3-255> 3-027 0-232 - ae
(A) 3-399 3-188 0-285 “5 9-0
3-816 3-488 0-383 33 11-0
4-272 3-938 0-484 $5 12-3
4-535 4-260 0-566 5 13-3
5-625 4-962 0-765 eS 15-4
5-976 5-524 0-901 53 16-3
| 8-283 5-910 1-012 . 17-1
6-489 6-154 1-113 95 18-1

Experiment II (Diagram I).—1-02 grams gelatin ; 60 c.c, phenol solution

Water-phase
Amount of phenol

Tnitial phenol Final phenol in grams Condition of
concentration concentration absorbed by protein Distribution-ratio
(grams per (grams per 1 gram gelatin
100 c.c.) 100 c.c.)
(A) 0-410 0-390 0-0123 Unprecipitated 3-1
0-575 0-551 0-014 5 2-5
0-739 0-706 0-019 x 2-7 lee
1-889 1-791 0-057 = 3:2 | =2-9
2-518 2-402 0-068 x 2°8 |
2-665 2-566 0-075 x 2-9
3-778 3-246 0-313 Precipitated 9-6
4-080 3-609 0-352 Sp 9-7
5-037 4-008 0-605 Be 15:1
6-296 4-838 0-858 93 17-7
6-561 5-510 0-970 x 17-6
7-556 5-606 1-150 6 20-5
(B) 2-419 3-244 0-373 S 11-5
0-934 2-707 0-104 Unprecipitated 3:8
0-540 1-122 0-031 x 2-8 | Mean
0-561 0-579 0-020 op 3-4 | =3-6
0-386 0-392 0-017 » 43

N.B.—(A) Equilibria obtained whilst phenol concentration was increased
(B) Equilibria obtained by reverse process.

PHENOL AND META-CRESOL 367
With further increase in concentration the distribution-ratio continued to

rise. No evidence of saturation of the gelatin with phenol was obtained,

although the saturation point imwater was closely approached.

11

10

Pa eee
HERS

@

“1

Fal ae NE
eee
Ee ea Ear as

C.p. in grms. per gm. Gelatin.
ox o

i

ee

lew
ne eS sree

aera

Sie

Cp bed™

A
0

2 03 :
C.p. Water (gms. per c.c.).

(ae a
PEWS SSG eS esas rerty

|] QO
ea A es cae ee a ey a ae

tt

ea ee eee er pe own aeeled
mane ree

i)

2 1 ae
=

anf
7

DracraM J,—Equilibria with phenol between water and gelatin showing influence of phenol-
coagulation upon the distribution of phenol between water and protein. (Hxperiment LI.
Table I).
O = Equilibria obtained during rising concentration of phenol.
+ = Equilibria obtained by reverse process.
A = Commencement of precipitation of protein.

368 BIO-CHEMICAL JOURNAL

The equilibria (A) in Experiments I and II were obtained whilst the
concentration of phenol was successively raised, and equilibria (B) by the
reverse process. The approximate coincidence of the two series of
equilibria, indicated by Diagram I, shows that the absorption of phenol
by gelatin is a reversible process throughout the whole range of
concentration investigated. The precipitation at 2°6 per cent. is also

reversible.
7 It appears that, although unprecipitated gelatin below concentrations
of 2°6 per cent. absorbs phenol according to the partition-law, when the
concentration of phenol in the water reaches about 2°6 per cent. the
phenol enters into a new relationship with the gelatin with the result
that the protein undergoes precipitation and its absorptive capacity for
phenol is enormously increased. The experiments on the distribution of
phenol between water and gelatin did not afford good results when higher
concentrations of phenol were employed, as equilibria were difficult to
obtain. The precipitation of the gelatin in concentrations above 2°6 per
cent. seems to be responsible for the somewhat erratic results.
Nevertheless, the observational points plotted in diagram 1 evidently
indicate again a linear relation to concentration, but in this case the
uptake of phenol by the gelatin is proportional, not to the total
concentration of phenol in the water-phase, but to this less 2°6 per cent.
What the significance of this may be I am unable to suggest, but it is
evidently associated with the sudden change in condition of the gelatin
on precipitation. |

(b) Egg-albumen in solution

Egg-albumen was more sensitive than gelatin to the precipitating
action of phenol. A 10 per cent. suspension of gelatin in water was not
visibly affected until the phenol concentration reached about 2’6 per cent.,
whereas a 10 per cent. solution of egg-albumen was rendered turbid by
1 per cent. phenol, slightly precipitated by 13 per cent., and completely
precipitated by 2 per cent. phenol. Smaller concentrations of
egg-albumen were affected by still lower phenol concentrations. The
precipitates of egg-albumen were stable on the addition of water.

The absorption of phenol by unprecipitated egg-albumen and the
reverse process were both complete within twenty-four hours. In the case
of the protein precipitated by phenol the equilibrium approached from
either side was attained within ninety-six hours.

PHENOL AND META-CRESOL 369

TABLE II
EQUILIBRIA
Distribution-ratio
Water-phase Amount of phenol Amount of

held by 1 gram phenol held by

Initial Final of egg-albumen State of 1 gram protein
concentration concentration in grams protein ——__
(grams per (grams per (by Amount of phenol

100 c.c.) 100 c.c.) difference) held by 1 gram

water

Experiment I (Diagram II).—10 c.c. 16-47 per cent. egg-albumen solution ; 50 c.c. phenol solution

(B) 0-251 0-237 0-005 Unprecipitated 2-1 | Mean
(B) 0-332 0-312 0-007 ; 2-3 / =2-2
(B) 0-805 0-757 0-025 Turbid 3°3
(B) 1-259 1-055 0-074 Slightly precipitated 7-1
(B) 1-666 1-332 0-122 Precipitated 9-1
(B) 4-233 3-229 0-365 ~ 11:3
(A) 0-777 0-874 0-086 <= 9-8
(A) 0-510 0-579 0-0614 =< 10-5

Experiment II.—10 c.c. 9-62 per cent. albumen solution ; 50 c.c. phenol solution

(B) 0-138 0-132 0-004 Unprecipitated 3-0

(B) 0-214 0-203 ~ 0-007 se 3-5 ee
(B) 0-712 0-695 0-015 is 2-1 { =3-1
(B) 0-971 0-938 0-035 Turbid 3-7

(B) 1-666 1-440 0-140 Precipitated 9-7

(A) 0-600 0-701 0-077 2 11-0

(A) 0-292 0-344 0-045 b 13-1

(A) 0-143 0-174 0-025 3 14-3

(A) 0-102 0-114 0-018 4 15-8

N.B.—(B) Equilibra obtained by uptake of phenol.
(A) Equilibria obtained by reverse process.

370 BIO-CHEMICAL JOURNAL

C.p. in gms. per gm. Egg-albumen.
bo

B\

i p. Water (gms. per c.c.)

DracraM IJ.—Equilibria with phenol between water and egg-albumen showing the influence of
phenol-coagulation on the distribution of phenol between water and _ protein.
(Experiment I, Table IJ).

O = Equilibria obtained during rising concentration of phenol.
+ = Equilibria obtained by reverse process.
A = Commencement of precipitation. B = Completion of precipitation.

Experiment III.—10 c.c. 10 2 per cent. egg-albumen solution ; 50 c.c. phenol solution

Water-phase Amount of phenol
in grams
Initial Final taken up by State of
concentration concentration 1 gram protein protein Distribution-ratio
(grams per (grams per (by
100 c.c.) 100 c.c.) difference)

(B) 0-164 0-155 0-005 Unprecipitated 3-2

(B) 0-410 0-390 0-012 % 3-0

(B) 0-639 0-623 0-014 2:3 + Mean
(B) 0-737 0-727 0-018 ae =2-9
(B) 0-877 0-857 0-030 3:5

(B) 2-148 1-807 0-215 Precipitated 11-9

(B) 4-926 3-466 0-376 3 10-8

(B) 4-651 4-235 0-460 10-9

(B) 5-475 5-280 0-575 10-9

(B) 6-056 5-742 0-760 13-2

(A) 3-346 3-812 0-486 12:7

(A) 1-444 1-699 0-226 13-3

(A) 0-708 0-864 0-134 15-5

(A) 0-360 0-441 0-086 19-5

(A) 0-184 0-231 0-059 25:5

(A) 0-096 0-119 0-046 38-7

PHENOL AND META-CRESOL 371

Experiment IV,—10-2 per cent. egg-albumen solution ; 50 ¢.c. phenol solution

(B) 0-400 0-383 0-011 Unprecipitated 2-9
(B) 0-400 0-387 0-010 ‘ 2-6 | Mean
(A) 0-258 0-262 0-009 9 3-4 | =3-0
(A) 0-258 0-262 0-008 s 3-1 |

N.B.—(B) Equilibria obtained whilst phenol concentration was rising,

(A) Equilibria obtained by reverse process,

The results obtained show that the distribution of phenol between
water and unprecipitated egg-albumen followed the partition-law, and
that phenol was on the average 2°8 times as soluble in the protein as in
water. The absorption of phenol by the egg-albumen was reversible.

A comparison of the above results with those of the experiments
with gelatin indicates that the solubilities of phenol in egg-albumen and
gelatin were approximately equal. Since egg-albumen was precipitated
by much lower phenol concentrations than gelatin, there was therefore
no relationship between their susceptibilities to precipitation and their
absorptive capacities for phenol.

In the neighbourhood of 1°0 per cent. final concentration of phenol,
where the precipitation of egg-albumen commenced, as indicated in
Diagram II, there was a greatly increased proportional uptake of phenol,
and with rising concentration the distribution-ratio continued to increase
until precipitation was complete (at about 1} per cent. final concen-
tration). With further rise in concentration the uptake of phenol by
the protein again became proportional to the concentration in the water,
but with a higher partition-coefficient than at the commencement.

The observations on equilibria, made as the concentration of phenol
in the water was being successively diminished, show that the
egg-albumen once having been irreversibly precipitated by phenol
maintains the high absorptive capacity for this substance until the
concentration of phenol in the water is reduced to zero. Although with
diminishing concentration the coagulated albumen gave up some of its
absorbed phenol, so that in low concentrations it only retained a very
small amount, the protein remained permanently denatured. It would
appear that the denatured albumen is not an intimate phenol-protein
complex, but is some modification of the original protein, which possesses
a high solvent. power for phenol.

372 BIO-CHEMICAL JOURNAL

(c) Egg-albumen coagulated by heat

TABLE III
EQUILIBRIA
Distribution-ratio
Water-phase Amount of phenol Amount of
Weight of in grams phenol held by
Initial Final coagulum absorbed by 1 gram coagulum
concentration concentration taken 1 gram coagulum ————
(grams per (grams per in grams (by Amount of phenol
100 c.c.) 100 c.c.) difference) held by 1 gram
water
Experiment I.—25 c.c. phenol solution
4-916 4-054 0-5 0-431 10-6
1-992 1-682 0-5 0-155 9-2 | Mean
0-987 0-822 0-5 0-082 10-0 9-8
0-984 0-703 1-0 0-070 10-0
0-584 0-515 0-35 0-049 9-5
Experiment II (Diagram III).—25 c.c. phenol solution
(A) 6-512" 5-190 0-5 0-661 12-7
(B) 2-076 2-706 0-5 0-346 12-8 | Mean
(A) 0-495 0-399 - 0-048 12-0 [ =129
(B) 0-159 0-198 0-028 14-2
6

C.p. in gms. per gm. Egg-albumen.,

C.p. Water (gms. per c.c.).

DracraM III.—Equilibria with phenol between water and egg-albumen showing the influence
of heat-coagulation on the distribution of phenol between water and protein.

Lower Curve.—Unprecipitated protein. (Haperiment III, Table II).

Upper Curve.—Heat-coagulated protein. (Hzxperiment II, Table III).
O = Equilibria obtained during rising concentration of phenol,
+ = Equilibria obtained by reverse process,

PHENOL AND META-CRESOL 373

The distribution of phenol between heat-coagulated egg-albumen
and water therefore followed the partition-law.

In Experiment II, equilibria (A) were obtained by the uptake of
phenol by the coagulum, and equilibria (B) by the reverse process.
The similarity in the values of the distribution-ratios deduced from the
two series of equilibria showed that the absorption was reversible.

The mean partition-coefficient in Experiment I was 9°8, while in
Experiment II it was 12°9. The coagula used were prepared from
different batches of albumen. This may explain the difference in these
values which, however, may also have its explanation in the varying
conditions existing during the formation of the heat-coagula.

A comparison of the solvent power of the heat-coagulum for phenol
with that of unprecipitated egg-albumen and of the phenol precipitated
protein (Tables II and III) indicates that heat-coagulation was
accompanied by an increase in the absorptive capacity of this protein
for phenol, and that, except in low concentrations, heat- and phenol-
coagulation affected the distribution of phenol between egg-albumen and
water to somewhat similar degrees. Moore and Bigland® showed that
the uptake of alkali by egg-albumen was also increased by heat-

coagulation.
(2d) Heat-coagulated egg-globulin
TABLE IV
EQUILIBRIA
4 gram coagulum ; 25 c.c. phenol solution
Distribution-ratio
Water-phase Amount of phenol
Amount of phenol held by 1 gram
Initial concentration Final concentration in grams protein
(grams per 100 (grams per 100 taken up by 1 gram ——_
c.Cc.) G-c:) coagulum Amount of phenol
held_by 1 gram
water
5-195 4-470 0-524 11-7
4-632 3-871 0-420 10-8
4-572 3-810 0-408 10-7
1-977 1-653 0-162 9-8
0-505 0-416 0-044 10-6
0-373 0-342 0-039 11-4
0-166 0-201 0-027 13-5
0-247 0-199 0-024 12-0

Mean =11-3

374 BIO-CHEMICAL JOURNAL

The distribution of phenol between water and the coagulated
globulin therefore followed the partition-law, and phenol was 11°3 times
as soluble in the protein as in water. By comparing the above results
with those in Table III it is found that the solubilities of phenol in
heat-coagulated egg-albumen and globulin were approximately the same.

(e) Casein

Casein was not precipitated from its solutions in sodium hydroxide
by phenol. The experiments described below were carried out with the
precipitated casein of commerce. The maximum amount of phenol was
absorbed by casein in forty hours, which time was therefore allowed for
equilibrium to be established.

TABLE V

EQUILIBRIA

4 gram casein; 25 c.c. phenol solution

Distribution-ratio

Water-phase Amount of phenol Amount of phenol
in grams held by 1 gram
taken up by 1 gram protein
Initial concentration Final concentration casein a SS
(grams per 100 (grams per 100 (by difference) Amount of phenol
©G,)) c.C.) held by 1 gram
water
Experiment I (Diagram IV)
1 7-485 5-614 0-936 16-6
2 5-233 4-418 0-662 15-0
3 4-491 3-491 0-500 14:3
4 1-396 1-888 0-254 13:5
5 1-497 1-227 0-135 11-0
6 0-491 0-633 0-064 10-1
7 0-253 0-315 0-033 10-4
8 0-126 0-156 0-018 11-5
Experiment II 0-409 0-344 0-033 9-6
2 1-137 0-986 0-108 11-0
3 3-634 3-125 0-363 11-6
4 4-966 4-508 0-592 13-1
5 5-527 5-266 0-722 13-7

C.p. in gms. per gm. Casein.

PHENOL AND META-CRESOL 375

iat

fea les Salles

aed i
‘03 .

04 05

C.p. Water (gms. per c.c.)

DracraM IV.—Equilibria with phenol between water and casein. (Experiment I, Table V.)

O = Equilibria obtained during rising phenol concentration.

+ = Equilibria obtained by reverse process,

The absorptive capacity of casein for phenol was thus considerably
greater than that of water, and was of the same order of magnitude as
the absorptive capacities of other precipitated proteins for this substance.
The results of Experiment I indicated that although below a final phenol
concentration of 1°8 per cent., the distribution of phenol between water
and casein followed the partition-law, at 1°8 per cent. there was a
deviation therefrom, the distribution-ratio increasing with further rise
in concentration. The results of Experiment II also indicated this
deviation, but there were not sufficient equilibria determined in this

376 BIO-CHEMICAL JOURNAL

experiment to show the constancy in magnitude of the distribution-ratio

at low concentrations.

Equilibria 1, 2, 3, 5, in Experiment I were obtained by the uptake of
phenol by casein; the remaining equilibra by the reverse process. The
coincidence of the two series (Diagram IV) showed that the absorption
of phenol by casein was reversible.

II

Tue INFLUENCE OF TEMPERATURE UPON THE DISTRIBUTION OF PHENOL
BETWEEN WATER AND PROTEINS
(a) Egg-albumen in solution

In the following table are given the amounts of phenol absorbed by
a definite amount of protein from the same concentration at 20° C. and

37°5° C.

TABLE VI.—10 c.c. 11-2 per cent. egg-albumen solution ; 50 c.c. phenol solution

Water-phase Amount of phenol
in grams taken up Distribution-

Tempera- Initial concentration Final concentration by 1-0 gram ratio

ture (grams per 100 (grams per 100 protein

c.c.) €:c3)

20° C. 0-248 0-196 0-006 3-0

20° C. 0-248 0-196 0-006 3-0

37-5° C. 0-248 0-197 0-005 2-5

A rise in temperature of 17°5° C. thus had no measurable effect upon -

the uptake of phenol by unprecipitated egg-albumen.
(b) Heat-coagulated egg-albumen

The experiments with the heat-coagulum were also carried out at

20° (C. andor 26:
TABLE VII
EQUILIBRIA

Water-phase = 25 c.c.
Weight of | Amount of phenol

Initial Final coagulum in grams
Tempera- concentration concentration taken taken up by _ Distribution-ratio
ture (grams per (grams per in grams the coagulum
100 c.c.) 100 c.c.)
20° C. 0-584 0-515 0-35 0-017 9-7
20° C. 0-984 0-702 1-0 0-070 10-0
20° C. 0-984 0-705 1-0 0-070 9-9
37-5° C, 0-984 0-731 1-0 0-063 8-6
37-5° C. 0-984 0-732 1-0 0-063 8-6
37-5° C. 0-584 0-524 0-35 0-015 8-2

The results show that the uptake of phenol by the coagulum was
slightly decreased by a rise in temperature from 20° C. to 37°5° C.

PHENOL AND META-CRESOL 377
(c) Gelatin
As gelatin dissolves in water at temperatures above 20° C.,
experiments were carried out at lower temperatures.
TasLe VIII
EQUILIBRIA
Gelatin, 4 grams ; phenol solution, 100 c.c.

' Water-phase
Amount of phenol

Initial Final in grams
Tem perature concentration concentration absorbed by Distribution-ratio
(grams per (grams per protein (4 grams)
100 c.c.) 100 c.c.)
4-4° (Minimum) to 0-994 0-914 0-080 2-2
6-6° C. (Maximum)
=5-5° C. Mean 0-994 0-922 0-072 1-9
0 Ce os) 0-994 0-914 0-080 2-2
Ge, Sen's ais 0-994 0-916 0-078 2-1
let 6 ie a3 0-994 0-917 0-077 2-1

A rise in temperature of about 15° C. therefore caused no change
in the distribution of phenol between water and gelatin.

The results indicate that, although the solubilities of phenol in water,
gelatin and dissolved egg-albumen were increased equally by rise in
temperature, the solubility of phenol in heat-coagulated egg-albumen was
increased to a slightly smaller extent than that of the same substance in

water.

Itl

Tue INFLUENCE OF ALCOHOL UPON THE DISTRIBUTION OF PHENOL BETWEEN
WATER AND HEAT-COAGULATED EGG-ALBUMEN

A definite amount of coagulum was added to each of a number of
phenol solutions containing different percentages by volume of alcohol.
The mixtures were kept at 20° C. for four days, and the filtrates from
the coagula then analysed. Alcohol was found not to interfere with the
process of estimating phenol. The results are tabulated below.

378 BIO-CHEMICAL JOURNAL

Taste [X.—1 gram coagulum ; 100 c.c. phenol solution
Distribution-ratio

Water-phase Amount of phenol Amount of
Per cent. of alcohol in grams phenol held by
Initial Final initially taken up by 1 gram protein
concentration concentration present 1 gram of protein eee
(grams per (grams per (by volume) (by difference) Amount of phenol
100 c.c.) 100 c.c.) held by 1 gram
water
1-992 1-680 0 0-156 9:3
1-992 1-684 0 0-154 9-2
0-993 0-833 0 0-080 9-6
1-992 1-758 25 0-117 6:7
1-992 1-770 25 0-113 6-4
0-993 0-974 50 0-010 1-0

These experiments indicate that alcohol considerably decreased the
uptake of phenol by the protein. The alcohol as well as the phenol would
be absorbed by the coagulum and, since alcohol is a very efficient solvent
for phenol, its distribution between water and protein would lead to an
increased solubility of phenol in both phases. The results of the above
experiments therefore mean that the amount of alcohol absorbed was
not sufficiently great to increase the solvent power of the protein for
phenol to as great an extent as the alcohol unabsorbed increased the
solvent power of the water for this substance.

IV

Tuer INFLUENCE OF HypROCHLORIC ACID UPON THE DISTRIBUTION OF
PHENOL BETWEEN WATER AND HEAT-CoAGULATED HGG-ALBUMEN

Hydrochloric acid decreases the solubility of phenol in water, and
it was therefore of interest to investigate its influence on the distribution
of phenol between water and proteins. Definite amounts of the coagulum
were immersed in a number of phenol solutions containing different
concentrations of hydrochloric acid. After equilibrium was established,
the phenol contents of the filtrates were estimated. The results are
tabulated below.

TaBLE X.—0-5 gram coagulum ; 25 c.c. phenol solution
Distribution-ratio

Water-phase Amount of phenol Amount of phenol
in grams taken up by
Initial Final Concentration taken up by 1 gram protein
concentration concentration of acid (HCl) 1 gram ==
(grams per (grams per coagulum Amount of phenol
100 c.c.) 100 c.c.) (by difference) taken up by

1 gram water
Experiment I

0-996 0-824 0 0-085 10-3

0-996 0-793 N x3 0-101 12:7

0-996 0-805 N x6 0:095 11:8
Experiment II

0-983 0-808 0 0-087 10-7

0-983 0-784 N x3 0-100 12:7

0-983 0-797 N x5 0-093 11-7

0-983 0-801 Nx5 0-091 11:3

PHENOL AND META-CRESOL 379

The presence of hydrochloric acid therefore slightly increased the
uptake of phenol by the coagulum, but the increase was more marked in
lower concentrations of acid than in higher ones.

The hydrochloric acid, as well as the phenol, would be absorbed by
the protein, and it would seem that the amount of acid absorbed was not
sufficient to decrease the solubility of phenol in the protein to as great an
extent as the amount of acid left in aqueous solution decreased the
solubility of phenol in water.

On comparing the results of the experiments with alcohol and acid
with those of Reichel’s! experiments on the influence of sodium chloride
upon the uptake of phenol by proteins (see introduction), it is found that
a substance increasing the solubility of phenol in water diminishes the
uptake of phenol by a protein, while substances decreasing the solubility
have exactly the opposite effect.

¥

Tue INFLUENCE OF CONCENTRATION UPON THE DISTRIBUTION OF
META-CRESOL BETWEEN WATER AND PROTEINS

(a) Gelatin

Saturated aqueous solutions (1°25 per cent.) of ortho-, meta- and
para-cresols had no precipitating action upon gelatin. Below are
tabulated some equilibria with meta-cresol between water and gelatin,
and also some equilibria with phenol determined under the same

conditions.
TasLeE XI.—5 grams gelatin ; 125 c.c. phenol or cresol solution
Distribution-ratio
Water-phase Amount of phenol or Amount of phenol or
cresol in grams cresol absorbed by
Tnitial concentration Final concentration absorbed by 5 grams 1 gram protein
(grams per 100 (grams per 100 gelatin SS
c.c.) c.c.) (by difference) Amount of phenol or
cresol absorbed by
1 gram water
Experiment I Meta-cresol
1-040 0-946 0-140 2-9 }
0-832 0-757 0-122 3:2 +Mean
0-624 0-570 0-082 28 }) =3-0
Experiment II Phenol
0-906 0-840 0-102 2-4
0-725 0-672 0-080 2-4 | Mean
0-544 0-505 0-057 2-2 =2-3

380 BIO-CHEMICAL JOURNAL

Over the small concentration range possible to work with, owing to
the slight solubility of cresol in water, the distribution of meta-cresol
between water and gelatin followed the partition-law, and quantitatively
was very similar to that of phenol between the same solvents.

Since meta-cresol is much less soluble than phenol in water, the
above results indicate that the introduction of a methyl group into the
benzene nucleus of phenol decreased its solubilities in water and gelatin

to approximately the same degree.

(b) Egg-albumen (dialysed solution)

A 0°7 per cent. solution of meta-cresol was sufficient to produce a
turbidity in a 10 per cent. egg-albumen solution, and a 1 per cent.
solution immediately produced a copious precipitate of protein, which
was irreversible. A 1 per cent. solution of phenol, however, only caused
a faint turbidity in an egg-albumen solution of the same strength. The
introduction of a methyl group into the benzene ring of phenol therefore
increased its precipitating action upon egg-albumen.

In the following table are given some equilibria with meta-cresol

between water and egg-albumen.

TABLE XIJ.—10 c.c. 10-2 per cent. egg-albumen solution ; 50 c.c. cresol solution

Distribution-ratio

Water-phase Amount of Amount of meta-
meta-cresol cresol held by
in grams Condition of 1 gram protein —
Initial Final taken up by protein ———_————
concentration concentration 1 gram Amount of meta-
(grams per 100 (grams per 100 protein cresol held by
CzC3) ce) 1 gram water
0-188 0-180 0-005 Unprecipitated 2:8 Mean
0-471 0-445 0-015 Unprecipitated 3-4) =3-1
0-846 0-696 0-104 Precipitated by cresol 14-9
0-950 0-822 0-179 Precipitated by cresol 21-8

Over the small concentration range possible to work with, the
distribution of meta-cresol between water and unprecipitated egg-
albumen followed the partition-law, and the cresol was 3°1 times as
soluble in the protein as in water.

Experiment III in Table II with phenol was carried out under the
same conditions as the above experiment with cresol, and it was found
that phenol was 2°9 times as soluble in egg-albumen as in water.
Egg-albumen, therefore, absorbed about equal amounts of phenol and
cresol from solutions of the same percentage strength, so that the solu-

PHENOL AND META-CRESOL 381

bilities of phenol in water and egg-albumen were equally decreased by
the introduction of a methyl group into its benzene nucleus. Since
egg-albumen was precipitated by lower concentrations of cresol than of
phenol, the above results indicate that the relative precipitating power of
these substances was not entirely determined by their equilibrium
positions between water and the protein.

A comparison of the distribution-ratio of meta-cresol between water
and egg-albumen with that of the same substance between water and
gelatin indicated that the solubilities of cresol in the two proteins were
equal. It was found that while egg-albumen was precipitated by dilute
solutions of meta-cresol, gelatin was not even affected by saturated
solutions (1°25 per cent.). As in the case of phenol, the susceptibilities
of these proteins to precipitation by cresol were therefore not entirely
determined by their absorptive capacities for this substance.

The precipitation of egg-albumen by cresol was accompanied by a
large increase in the absorptive capacity of the protein for this substance.
This increased absorption took place at a lower concentration with cresol
than with phenol owing to the greater precipitating power of the former
substance.

VI

THE SOLUBILITY OF PROTEINS IN PHENOL AND META-CRESOL

Ritthausen (1872),° and Osborne? and his co-workers (1891-93),
found that zein dissolved in warm crystallized phenol. Kjeldahl] (1896)§
showed that gladin was also soluble in para-cresol and was precipitated
from its phenol solution by many organic reagents.

Reichel (1909)! found that when serum was warmed with anhydrous
phenol a clear solution resulted which underwent no apparent change on
boiling.

The high solvent powers of egg-albumen and gelatin for phenol
suggested that these proteins would also be soluble in anhydrous phenols.

When egg-albumen crystals were heated with molten phenol at
45° C. the protein gradually dissolved, and a clear solution was obtained.
On dialysis or the addition of water to this solution the protein was
precipitated. This was also effected by dialysing at 45° C. Similarly
heat-coagulated egg-albumen dissolved in molten phenol and was
reprecipitated when the solution was diluted. Gelatin dissolved in

382 BIO-CHEMICAL JOURNAL

phenol less readily than egg-albumen. It could be recovered in the
unprecipitated state from the solution by dialysis at ordinary
temperatures.

Meta-cresol also dissolved egg-albumen and gelatin at temperatures
of about 40° C. Solution was not detected at room temperatures. By
dialysing these solutions at room temperatures while the egg-albumen
separated as a coagulum, the gelatin was recovered in the unprecipitated
~ condition.

The facts that egg-albumen could only be recovered from these
solutions in the precipitated state, while gelatin could be recovered
in an unaltered condition, corresponded with the known differences in the
stability in the presence of water or dilute phenol. solutions of the
products of the precipitating action of the phenols upon these proteins.

When anhydrous meta-cresol was added to 1/2 its volume of horse-
serum at ordinary temperatures the serum proteins were precipitated,
but on heating to 100° C. these were redissolved and a clear homogeneous
fluid resulted. On cooling the liquid again became turbid owing to
separation of the proteins. Meta-cresol rapidly dissolved Witte’s peptone
at 100° C., forming a brown liquid. Solution was also found to proceed
slowly at ordinary temperatures.

Although meta-cresol dissolved proteins readily, it had no solvent
action, even at high temperatures, upon the amino acids, glycyl-l-tyrosin
and di-alanyl-cystin. These amino-acids also differed from proteins in
not being precipitated from their aqueous solutions by strong solutions of
phenol.

Vi

Tue APPLICATION OF SOME OF THESE Facts TO THE THEORY OF PHENOL
DISINFECTION

I.—The influence of alcohol upon the germicidal power of phenol

Kronig and Paul (1897)!° showed that the germicidal action of
phenol upon anthrax spores was decreased by the presence of alcohol, a
solution of phenol in 98 per cent. alcohol possessing only a very feeble
bactericidal power.

The experiments described in Section III showed that the presence
of alcohol decreased the uptake of phenol by heat-coagulated albumen.
The decrease was of such magnitude that the protein absorbed eight

PHENOL AND META-CRESOL 383

times as much phenol from an aqueous solution as from a solution of
phenol in 50 per cent. alcohol.

The inhibiting influence of alcohol upon the bactericidal action of
phenol is sufficiently explained by a decreased uptake of phenol by the
spore-proteins, i.e., to a decreased amount of phenol available for
disinfection.

It may be pointed out in this connection that, although the addition
of acids considerably increases the germicidal action of phenol and cresol,
the presence of hydrochloric acid only very slightly increased the uptake
of phenol by egg-albumen. This suggests that the effect of acid upon
germicidal power is only to a small extent due to an increased uptake of
phenol by the bacterial proteins. It is probably due chiefly to the
bactericidal action of the acids themselves.

II.—The effect of the presence of organic matter upon the germicidal
action of phenol and cresol

Blyth,!! in 1906, showed that the bactericidal action of phenol and
cresol was decreased by the presence of milk, and to a lesser extent by
separated milk. As the latter contains the proteins and sugar of the
original milk with only a trace of fat and as lactose was found to have
practically no effect upon the germicidal action of phenol, the depreciating
effect of separated milk upon bactericidal action must be due to its
constituent proteins. Some experiments recorded in Section I showed
that casein absorbed phenol from aqueous solution. In a mixture of
separated milk and phenol solution the casein therefore renders a
considerable portion of the germicide unavailable for disinfection, and
in this rests the explanation of the depreciating effect of separated milk
upon bactericidal action.

III.—The influence of temperature upon the germicidal action of phenol

Henle (1889)!? showed that the disinfection of B. typhosus was more
rapidly completed if the temperature of medication were raised.

H. Chick (1908)!3 quantitatively determined the influence of
temperature upon the rate of disinfection of B. paratyphosus by phenol.
It was found that the velocity of disinfection increased seven to eight fold
for a rise in temperature of 10° C.

Since it was found that the uptake of phenol by an unprecipitated
protein (gelatin) proceeded very raptdly, being complete within two
minutes even at low concentrations, the influences of temperature upon

384 BIO-CHEMICAL JOURNAL

the equilibrium positions of phenol between water and proteins and upon
the velocity of disinfection can be compared. It was shown
that a rise in temperature of about 15° C. had no measurable effect
upon the distribution of phenol between water and unprecipitated
proteins. This suggests that the great accelerating influence of rise in
temperature upon the process of disinfection is not due to an increased
uptake of phenol by the bacterial proteins, but must be due to an increased
velocity of a reaction between the phenol and proteins which proceeds
subsequently to the absorption.

IV.—The unequal germicidal powers of phenol and meta-cresol -
Several investigators have shown that the germicidal power of
meta-cresol exceeds that of_phenol. This is indicated in the following
table, in which the number of times meta-cresol exceeds phenol in
germicidal power is represented by the carbolic acid co-efficient.

Carbolic acid co-efficient

Organism of meta-cresol
(Pure phenol = 1:0)
B. typhosus oe: os fe ae 26
Staphylococcus py. aur viol ae 2°0
B. coli cs Se os 21

The experiments described in Section V showed that, although meta-
cresol was superior to phenol in its precipitating action upon egg-albumen,
there was no difference in the amounts absorbed by this protein from equal
concentrations of the two substances. |

The introduction of a methyl group into the benzene nucleus of phenol
therefore led to an increase in both bactericidal power and _ protein-
precipitating power, but to no change in the uptake of the phenol by the
protein. This suggests that the superiority of meta-cresol over phenol as
a germicidal agent is due to the fact that cresol precipitates proteins in
smaller concentrations than phenol.

V.—Selective germicidal action

The germicidal power of a substance depends upon the nature of

the organism employed. This is indicated in the following table :—

Concentration of phenol

Organism killing in 15 mins. under
constant conditions
iD. ty phosus jaa ae dis a te 7 in 1,000
B. diphtheria... ee He ois a 5 in 1,000
Bo colivac: = fd on whe ae 8 in 1,000
Staphyloccus py. aur ... a ae. «w. 10 1m 1000
B. pestis 4 Be aie ey, fe 7 in 1,000

The cresols exhibit similar cases of selective germicidal action.

PHENOL AND META-CRESOL 385

Some experiments described in a previous section showed that
proteins possessing equal absorptive capacities for a phenol were unequally
susceptible to its precipitating action. This suggests that the
phenomenon of selective germicidal action is determined by the phenol

concentration at which the specific proteins are precipitated.

SUMMARY AND CONCLUSION

I.—On the relations of phenol and meta-cresol to proteins

1. Gelatin and egg-albumen absorb phenol and _ meta-cresol
according to the partition-law.

2. When a certain phenol concentration is reached the proteins are
precipitated. This is accompanied by a great increase in their absorptive
capacities for phenol.

3. The precipitation of gelatin by phenol is reversible, and the
absorption of phenol by this protein is also reversible throughout the whole
range of phenol concentration in water. The precipitation of egg-
albumen, on the other hand, is irreversible, and the precipitated protein
retains permanently its high absorptive capacity for phenol.

4. Heat coagulation also increases the solvent power of egg-albumen
for phenol and, except at low concentrations, influences the equilibrium
of phenol between water and this protein to about the same degree as
phenol precipitation.

5. Meta-cresol does not precipitate gelatin, but it precipitates
egg-albumen at lower concentrations than phenol. The precipitation
leads to an increase in the absorptive capacity of the protein for cresol.

6. The solubilities of phenol in gelatin and egg-albumen at 20° C.
are equal. This is also true of meta-cresol. Although this is the case,
egg-albumen is precipitated by lower phenol and cresol concentrations
than gelatin. The susceptibilities of the two proteins to precipitation

are therefore not entirely determined by their absorptive capacities for
the phenols.

~

i. Hgg-albumen absorbs identical amounts of phenol and meta-

cresol from equal percentage concentrations, but is precipitated by lower

concentrations of cresol than of phenol. The relative precipitating power

of these substances is therefore not entirely determined by their
equilibrium positions between water and the protein.

8. Heat-coagulated egg-albumen and egg-globulin absorb phenol

according to the partition-law, Their solvent powers for phenol are
approximately equal.

386 BIO-CHEMICAL JOURNAL

9. Casein at low concentrations absorbs phenol according to the
partition-law. Above 1°8 per cent., however, with rising concentration
there is an increased proportional uptake of phenol. The absorption is
reversible. Casein possesses the high absorptive capacity of precipitated
proteins.

10. Rise in temperature has no measurable effect upon the
distribution of phenol between water and egg-albumen and gelatin; it
leads, however, to a slightly decreased uptake of phenol by heat-
coagulated egg-albumen.

11. Alcohol considerably decreases the uptake of phenol by heat-
coagulated egg-albumen. Hydrochloric acid, on the other hand, slightly
increases the uptake.

12. Egg-albumen (crystallized or heat-coagulated) and gelatin
dissolve in warm anhydrous phenol and meta-cresol. When the
solutions are dialysed, the albumen is irreversibly precipitated, while
the gelatin can be recovered in the unaltered condition.

13. Horse-serum forms a homogeneous fluid with about twice its
volume of meta-cresol at 100° C. On cooling the serum proteins are
precipitated.

14. Although meta-cresol dissolves proteins and also peptone, it does
not dissolve even at high temperatures the polypeptides—glycyl-l-tyrosin
and di-alanyl-cystin. These substances are also not precipitated from
aqueous solution by strong phenol solutions.

II.—The application of some of these facts to the theory of disinfection

1. The depreciating effect of alcohol upon the bactericidal action of
phenol is explained by a reduction in the absorptive capacities of the
bacterial proteins for this substance. The action of hydrochloric acid in
increasing the germicidal power of phenol, on the other hand, is only
partially explained by a redistribution of phenol between water and
proteins.

2. The depressing effect of separated milk upon the germicidal
action of phenol is due to the absorptive capacity of the constituent casein
for this substance.

3. The great increase in the germicidal power of phenol, which
results from rise in temperature, is not explained by the influence of
temperature upon the distribution of phenol between water and proteins.

4. The superiority of meta-cresol to phenol as a germicide appears

to be due to the fact that cresol precipitates proteins in lower concen-
trations than phenol,

PHENOL AND META-CRESOL 387

5. The selective action of phenol as a germicide upon different
organisms seems to be associated with the observed dissimilar
susceptibilities of proteins to its precipitating action.

Matn Conciusion

The absorption of the phenols by bacteria is merely the initial stage
in the process of disinfection.

The germicidal action which follows the absorption does not seem to
be the result of a typical chemical union between the phenols and
bacterial proteins, as is the case, for instance, with formaldehyde, but is
apparently associated with the de-emulsification of the colloidal
suspension as evidenced by the precipitation of proteins when a certain
phenol concentration is attained. In the case of egg-albumen the change
is irreversible, and the precipitated protein is not again dispersed on the
removal of the phenol. The germicidal action of phenol thus appears to
be similar in mechanism to that of heat. This interpretation of the
process of disinfection by phenol may explain the fact that below certain
concentrations (about 0°5 per cent.) this substance exerts but a very
feeble bactericidal action, there being a disproportionate falling off in
germicidal power when the phenol concentration is reduced to this point.

I desire to express my best thanks to Dr. C. J. Martin, F.R.S., for
much helpful advice in the course of this work.

REFERENCES

1. Reichel, ‘ Zur theorie der desinfektion, Biochem. Zeitschrift XXII, p. 149, 1909.

2. Herzog and Betzel, ‘ Zur theorie der desinfektion,’ Hoppe-Seyler’s Zeit. physiol. Chemie,
LXVII, iv and vy, p. 309, 1910 ; LXXIV, iii, p. 221, 1911.

Lloyd, Journ. Amer. Chem. Soc., p. 23, 1905.

Hopkins and Pinkus, Journ. Physiology, XXIII, p. 130, 1898.

Moore and Bigland, Bio-Chem. Journ., Vol. V, p. 32, 1910.

Ritthausen, Die Eiweisskorper der Getreidearten, Bonn, 1872.

Osborne and co-workers, ‘Alcohol soluble proteins, Amer. Chem. Journ., XIII, XIV,
XV, 1891-93.

8. Kjeldahl, ‘Untersuchungen iiber das optische Verhalten einiger vegetabilischer Eiweiss-
korper, Beid. Cent., XXV, pp. 197-199, 1896.

9. Mathewson, ‘ The optical rotation of gliadin in certain organic solvents, Journ. Amer.
Chem. Soc., XXVIII, pp. 1482-1485, 1906.

10. Kronig and Paul, Zeitsch. fur Hygiene, XXV, ii, p. 1, 1897.

11. Blyth, ‘ The standardization of disinfectants,’ Journ. Soc. Chem. Ind., December 31, 1906.

12. Henle, ‘ Ueber Creolin und seine wirksamen Bestandtheile,’ Archiv. f. Hygiene, Vol. IX,
p. 188, 1889.

13. Chick, ‘ An investigation of the laws of disinfection,’ Journ. Hygiene, Vol, VIII, No. 1,
January, 1908.

STs ease

388

THE DISTRIBUTION OF NITROGEN IN AUTOLYSIS, WITH
SPECIAL REFERENCE TO DEAMINIZATION

By GERTRUDE D. BOSTOCK, B.Sc., M.B., Cu.B., Carnegie Scholar.
From the Laboratory of Physiology, University of Glasgow
Communicated by Dr. E. P. Cathcart
(Received July 5th, 1912)

The object of the present investigation was not primarily to establish
a closer connection between autolysis and intravital processes, but merely
to determine whether the ammonia liberated in liver digests was converted
into amide nitrogen, and whether the incubation of an amino acid with
liver tissue led to a similar increase.

In a previous paper! on deaminization of amino acids when incubated
with liver tissue under aseptic and antiseptic conditions, the fate of the
liberated ammonia was discussed. Lang? had previously found that a
considerable deaminization occurred, but in my experiments the amount
of ammonia liberated from the amino acids (alanin, glycocoll and leucin)

was insignificant, while on the other hand a considerable amount of .

ammonia made its appearance after the incubation of an amide—
asparagin—with liver tissue.

The question arose as to whether the apparent absence of
deaminization might not be due to the conversion of the liberated
ammonia into urea.

In 1898, Loewi? had investigated the urea-forming function of the
liver in vitro, and had found an increase in the amount of a urea-like
body in a liver extract after incubation, but the experiments were not
very satisfactory. An absence of a proportionate consecutive increase is
evident in the following figures, for the amount of nitrogen found in a
Mérner Sjéquist estimation on an alcoholic extract of liver after varying
periods of digestion :—

I. After 8 hours’ incubation 0°093 gm. N increase in 1000 gm. liver.

LE ges i i: 0-004 ” »
| Id ares bee i o 0°027 » %3
LW 2 org, eed + i 0°013 ” ”

DISTRIBUTION OF NITROGEN IN AUTOLYSIS 389

Moreover, he was unable to isolate urea from the alcohol-ether
extract, and he left the nature of the substance unsettled.

In view of the fact that only a small liberation of ammonia from
amino acids, when incubated with liver pulp, was to be expected, the
addition of ammonia or an ammonium salt to the liver tissue suggested
itself as being more likely to lead to an increase in the amount of amide
nitrogen at the end of incubation.

The investigation proved less straightforward than was anticipated,
on account of the difficulty in finding a suitable method for the
determination of the amide nitrogen.

The following methods were finally adopted and gave satisfactory
results.

Methods.—The incubation of the tests was carried out in the way
described in the previous paper, the only modification being that in a
majority of cases 100 gm. liver tissues were employed, instead of 50 gm.,
and the tests were shaken with 5 e.c. toluol, instead of 2 c.c. Shaking
varied in the different experiments from thirty to sixty minutes.

A modification of the working up of the tests was introduced when
the attempt was made to determine the amino acids, 1.e., from Experiment
32 onwards. After this 100 ¢.c. rectified spirit, instead of tannic acid,
was used to coagulate the proteins, coagulation being rendered more
complete by boiling. Filtration was carried out after the combined
precipitate and solution had been made up to a definite volume, i.e.,
400 c.c. The ammonia, amide and amino acid nitrogen and the total
nitrogen were then determined in the filtrate, 100 c.c. of which
corresponded to 25 gm. liver tissue.

All the figures in this paper are expressed in c.c. N/10 sulphuric
acid for 100 c.c. of the filtrate.

Ammonia was determined by a modification of the Kruger-Reich-
Schittenhelm* vacuum distillation method, the temperature was kept at
38° C. or just under. Distillation was carried on two hours; 40 c.c. of
ethyl alcohol were added to the filtrate before distillation. The alkaline
mixture used was one volume of a saturated solution of Na,CO,, and two
volumes of a saturated solution of NaCl, 20 to 25 c.c. being added to
each test. Rosolic acid was used as indicator.

The amount of filtrate taken for examination was 100 c.c. in the
majority of cases; in a few instances only 50 c.c. was taken.

The Amide nitrogen was determined by a modification of the Benedict
Gephart method> for estimating urea suggested by Henriques and

390 BIO-CHEMICAL JOURNAL

Gammeltoft® (omitting, however, the phosphotungstic precipitation).
In carrying out the estimation, 5c.c. of a 1 in 20 solution of
hydrochloric acid was added to a 100c.c.* of the filtrate to be
tested, and the mixture was heated in the autoclave to 150°C. for
1 hours, after which the ammonia was distilled off in vacuo by the
Kruger-Reich-Schittenhelm method. The heating of the autoclave was
- most carefully conducted, the pressure only varying between 68 and
72 Ibs. per square inch. The later modification suggested by Benedict?
in 1910, i.e., the addition of MgCl, before heating, was not adopted, as
most of the work had already been carried out. Wolf and Osterberg,®
Levene and Meyer,® and Benedict himself point out that a disadvantage
of the autoclave method is that small amounts of uric acid and creatinine
are decomposed in addition to the urea. In the present work on liver
digests, however, accurate relative values, and not absolute values are
claimed, and the fact that concordant control results were obtained
shows that the method (within its limitations) is at least reliable.
Legitimate conclusions as to the fate of amide nitrogen in these digests
can therefore be drawn from a comparison of the figures obtained.
Amino acid nitrogen was estimated by the Henriques-Sérensen
method.!° The initial neutralisation, however, was carried out with
phenolphthalein!! as an indicator, the first appearance of a pink tint
being regarded as the end point. In the earlier tests the end point after

the addition of the neutral formol was not taken to a deep rose as |

Henriques and Sérensen recommend. In the later tests, however, the
deep rose was always taken as the end point.

The colour of the original filtrate, which varied from a pale amber
to a deep yellow, interfered considerably with the sharpness of the end
point in some cases, and the only safety lay in repeating the estimations.
Here again no claim is made for absolute values, but a comparison of
the relative values is justified by the concordance of the results obtained
from controls. The error, such as it is, is on the side of low rather than
of high values. Various observers (de Jager 1909),!2 (Henriques and
Sérensen 1910),1° have drawn attention to the fact that in the presence of
ammonia in excess the estimation of the amino acids yields too low a
figure. To obviate this difficulty, Henriques and Sérensen recommend a
determination of the formol-titratable nitrogen on the residue of an
ammonia distillation. This was accordingly tried, but unfortunately

* In a few cases 50 c.c. instead of 100 c.c. of the filtrate, were used for the determination of the
Amide nitrogen,

=

DISTRIBUTION OF NITROGEN IN AUTOLYSIS 391

the yellow colour always seemed to be stronger in the case of such a
residue—and -probably as a result of this and the consequent masking
of the end point, a lower value was frequently obtained for the amino
acid nitrogen than in a direct estimation.

The direct estimations have therefore been given unless otherwise
stated. 20 to 25 c.c. of the filtrate were used for the determinations,
which were carried out in duplicate, except in the case of the very earliest

estimations.

Total nitrogen has been estimated by Kjeldahl method, duplicate
analyses being always carried out. 10 to 30 c.c. of the filtrate were
used for each determination.

PRESENT INVESTIGATION

In the initial experiments recorded in Table I, ammonia and amide
nitrogen only were estimated. It was found after the digestion of
ammonia with liver pulp, that the amount of amide nitrogen, so far from
being increased, was actually smaller than in the case of a control test
to which no addition of ammonia had been made.

As the ammonia was added in the form of the hydrate, there was a
possibility that the addition of alkali had something to do with the
result. In Experiment 27, therefore, additions of equal amounts of
NaOH and NH,OH were made to digests, and in each case, i.e., after
two, four and six days’ incubation, respectively, the amide value agreed
fairly well in the two alkaline tests, and was considerably lower than
those of the control tests. Moreover, if the NaOH test were taken as the
control test for the NH,OH test, there had been no disappearance of
ammonia, but a recovery of the whole amount. Evidently, then, the
alkaline reaction had modified the course of digestion.

It was imperative to determine whether the addition of alkali had
produced this effect by diminishing the rate of autolysis. This could
best be ascertained by estimating the total soluble nitrogen at the end of
incubation. If it were equal in the control and the alkali test, it might
be safely assumed that autolysis had proceeded at the same rate in both.
Without such an estimation of the total soluble nitrogen, conclusions
regarding the conversion of ammonia into amide nitrogen in vitro were
not valid.

The action of acid and alkali in accelerating and inhibiting autolysis
respectively is well known, a number of observers having made
contributions on the subject. A further investigation seemed advisable

BIO-CHEMICAL JOURNAL

392

oe 60-21 LSI GIL 6-8 T-LT c¢-01 skep 9 (F)
Be rE C6 CF-Z1 3-6 C8-L CFI 0-6 skep 7 (g) | AJeatjoodsor poppe HO'HN Put HO®N O1/N
8-F LF 8 6-01 COL S19 8-6 8-9 skep @ (Z);*9°9 3-61 $1897 HOTHN pue HO®N OF,
te a ds Bs se a AG eT ae
i ee + ee a se a a HA er Bae ie
HO'HN HO®N
0-¢ 8-Z 16 TSI fs ; 9G-Z1 Fo-S1 sfep ¢ (g)
0-¢ 9° GG-8 SF es ce 4S0] ¢9-11 sep % (Z)| = I ee urs eee DDS
a ae a : ‘ te FIT £8 oe ji HO'HN O1/N ‘9°9 0% $380} vruomUIe Of,
see see see see vee see Ton Lg smmoy 0Z (TL)
8-6 8:6 (Z) 98 9-F% : ("poppe uorynfos % F B JO ‘0°0 OT 8480}
6-F 9% (1) L88 9€-61 68-LF 69-F1 L-GF CL-F1 sXep FT —- | [O00 uy *(Z) 09 “0° 9-68 Pur (T) 0, poppe
(HO'HN OU/N ‘°° 9-61 $389} eruouMe UT
3G 8-€ ELE 6-F1 c9-1E 8-01 C6-F8 CLIT skep ¢ ee ieee "= uomnyos % F
Se oF O-LT 6-81 cL-91 CF-6 8-61 c-6 skep § ei jo ‘00 ‘39803 10000448 ul ‘“peppe
oS PP 61 €-01 SE-ZI 3-9 cE:6 6-¢ sanoy 0Z (1)) HO'HN OT/N °° 8-0Z $}s0} vruomUIe UT
N°HN JO poroaooot NH oprmy N° N N DENY N°HN N opray N°HN uoneqnout
gunowe N*HN jo —— ————~ eo a a y BD
pezeMoyey = unowy VINOWWW TIOONOOATY) TIOULNOD poweg

“IOAT, XO SUIBAS GZ IO ‘oYRIZIY *0°0 YOT JOJ plow OT/N *0°0

Ul ore SoINSYy oy} [TV “[ LAV

"98

"GSS

DISTRIBUTION OF NITROGEN IN AUTOLYSIS 393

in view of the fact that (so far as I know) no attempt had been made to
estimate the increase in ammonia and amide nitrogen in relation to the
total soluble nitrogen. Attention was now directed to the effect of acid
and alkali on autolysis, particularly with regard to the distribution of
ammonia and amide nitrogen. From this time onward the total soluble
nitrogen was always determined, so that the ammonia and amide values
might be expressed as percentages of this. Eventually a further
modification of the method allowed of an amino acid determination, and
an attempt was made to study the nitrogen distribution in autolysis and
the effect upon it of various factors.
The factors investigated were :—
1. Length of incubation.
2. Acid and alkali.
3. Presence of bacteria.
It was felt that a study of these conditions was a necessity in
determining the fate of amino acid or ammonia added to digests.

LENGTH OF INCUBATION

In discussing the influence of these factors on autolysis, it will be
more satisfactory to study first that of time or the length of incubation.
The literature deals chiefly with the effect of acid or alkali on autolysis
and the time factor itself has been almost ignored. In most of the work
incubation has been of short duration, usually three to seven or eight
days. Simon,! using Salkowski’s!® method, fourd an increase in the
soluble nitrogen of a liver digest up to the 22nd day, when the
experiment was stopped.

Dakin!® found that the formation of ammonia in kidney autolysis
extended over two months, during which time it was proportional to the
time of digestion. Drjewezki!’ points out the value of a consideration of
the partition of nitrogen, as he found that the total soluble nitrogen
might remain stationary, while changes occurred in the ratios of the
various nitrogen constituents to the total soluble nitrogen.

The following two questions may now be asked, before reviewing
the experiments tabulated below.

(1) Does autolysis proceed at the same rate from day to day, or does
it rapidly reach 2 maximum and then gradually diminish until
it finally comes to a standstill ?

(2) Does the percentage distribution of nitrogen remain unchanged
throughout the course of autolysis?

394 BIO-CHEMICAL JOURNAL

In Table II all the control tests from the whole series of experiments
have been collected for purposes of comparison.

Experiments 34 and 35 (1), however, were carried out solely to
determine the influence of the length of incubation on the actual and
percentage distribution of the soluble nitrogen in autolysis. In
Experiment 34 eight tests were prepared at the same time, two were
worked up immediately, while two were incubated for two, four, and
fifteen days respectively. .

Total Soluble Nitrogen.—In fresh liver tissue the total soluble
nitrogen is fairly constant, 51 to 57 c.c. N/10 for 25 gm. liver tissue.
Within the first twenty-four hours’ incubation the increase is more rapid
than at any subsequent period; after the third day (see Experiment 34)
the increase is comparatively small, and it gradually becomes less and
less. Unfortunately, the data are not sufficient to determine the precise
time at which it ceases, but it would seem that after nine weeks, at any
rate, no further increase occurred, the figures indeed show that in
Experiment 35 there is actually a diminution after the ninth week until
the twelfth.*

An examination of Table II will show that there is sometimes a
considerable difference in the total soluble nitrogen of different tests,
incubated for the same period, although in serial tests the increase shows
a fairly regular sequence. In Experiment 28, for example, the values
at the end of two, four and seven days are all higher than in some of the
other tests, such as 34. Such differences are probably due to initial
differences in the organs themselves, or are perhaps due to a longer
interval having elapsed between the slaughter of the animal and the
working up of the tissues in some cases than in others.

Ammonia.—The preformed ammonia in fresh liver tissue is strikingly
small, i.e., 1°75 to 2 c.c. N/10 in 25 gm. This result is confirmed
by some unpublished work on the ammonia content of the fresh dog’s
liver, when the following values were obtained, 1°8, 1°8, 1°25, 1°72 and
1:22 c.c. N/10 in 25 gm. This fact is also emphasized by Folin!® in his
most recent work. He holds that both in blood and liver tissue more
ammonia is produced by decomposition in a few hours than is actually
present as preformed ammonia in the fresh blood or tissue itself.

In the serial experiment 30, the ammonia has more than doubled

* In 35 (1) only 50 gm. of liver tissue were put up with 100 c.c. physiol. salt solution, while in
35 (2) and (3) 100 gms. were put up with the same amount of fivid, this might account for the

difference in the total, soluble Nitrogen figures. In 35 (2) and (3) the difference lies within the
limits of experimental error.

DISTRIBUTION OF NITROGEN IN AUTOLYSIS 395

itself in the first twenty-four hours, after this the increase is much more
gradual (see Experiment 34).

Ammonia, expressed in percentage of the total soluble nitrogen, is
also small in fresh tissue and just about doubles itself in twenty-four
hours. In 34 it shows a gradual increase after the second day, being
73, 77 and 81 on the second, fourth, and fifteenth day respectively.
In 35 (1) it is 8°6 at the end of the ninth week, so that little change
occurs in the ammonia percentage after the second week. In some
experiments, as in 37, the percentage is higher than this, e.g., 11°5 to
11°8, and is practically the same after nine days, two weeks, and three
weeks respectively. In different experiments, then, the percentage
ammonia varies within somewhat wide limits (7°8 to 11°9) without any
definite relation to time after the first twenty-four hours.

Amide nitrogen.—The amide content of the fresh tissue is low, 5°85
to 6°4, but it does not show the same rapid increase as the ammonia.
Experiment 34 shows a fairly regular sequence in the absolute amounts,
but there is a less regular sequence in the percentage amounts. This
holds good for most of the experiments, although Table II shows a certain
lack of uniformity about the percentage figures. In 37 the percentage
amount throughout the series is practically the same as that of the
ammonia, but in every other case, it is higher than the latter. The
explanation may be in the initial reaction of the tissue (see later). In
connection with the total soluble nitrogen in this experiment, it was
noted that autolysis had apparently come to a standstill by the second
week. The amide shows the same standstill. In Experiment 35 the
actual amount shows a decrease, if anything, from the ninth to the
twelfth week, while the percentage amount remains practically constant.

Amino acid nitrogen.—The amino acid nitrogen shows the same
general characters as the ammonia and amide nitrogen, 1.e., it is
smallest in the fresh tissue. It more than doubles itself within
forty-eight hours, after which the increase is more gradual.

In 35 there is apparently an increase in the amino acid nitrogen
after nine weeks’ incubation, which might perhaps be explained by the
splitting of polypeptides. As, however, it has already been pointed out
that the method applied to tissue digests is not free from possible error,
no stress is laid on this single apparent increase.

It is interesting with regard to the percentage figures to see that
practically 50 per cent. of the soluble nitrogen at the end of nine to twelve
weeks’ digestion is in the form of amino acid or rather formol titratable
nitrogen, whilst the proportion in fresh tissue is about 17 per cent.

396 BIO-CHEMICAL JOURNAL

In Experiments 36 and 37 the percentage varies from 49 per cent.
to 51 per cent. in the nine days’ digests, and from 47 per cent. to 54 per
cent. in the three weeks’ digests, so that apparently the maximum
percentage is reached at a fairly early stage in autolysis. It would
appear that after the ninth day amino acid nitrogen increases only
parallel with the total soluble nitrogen.

34 (4) gives only 29 per cent. amino acid nitrogen at end of fifteen
days’ incubation. This low value is probably due to the imperfect
technique, as these tests were among the earliest in which the amino acid
nitrogen was estimated.

Undetermined Nitrogen._In Experiment 34 the undetermined
nitrogen shows a slight actual increase in amount with the increasing
duration of incubation, but the percentage amount has diminished from
68 per cent. in the fresh tissue to 48 per cent. at the end of fourteen days’
incubation. In the later tests, 1.e., 35, 36 and 37, in which incubation
lasted from nine days to twelve weeks, the actual amounts of undetermined
nitrogen are low, the percentage amounts are also low; no definite relation
to time can be made out within this period.

The questions formulated above can now be answered as follows :—

(1) Autolysis does not proceed at the same rate from day to day, it
reaches a maximum within the first forty-eight hours, after
which the increase in all the constituents of the soluble nitrogen
eradually ceases.

(2) The percentage distribution of the soluble nitrogen in fresh tissue
is quite distinctive in character; the ammonia and amino acid
nitrogen are much lower, and the undetermined nitrogen is
considerably higher than at any subsequent period of incubation.
The ratio of ammonia to amide nitrogen is less than 1 : 3 in fresh
tissue. Within forty-eight hours the ratio has changed, and is
usually greater, but never less than 1 : 2.

How long proteolytic ferments remain active under the conditions
of antiseptic incubation cannot be answered; there is a lack of hterature
on this point. Vernon!® points out that if tissue be minced, much of the
enzyme goes at once into solution, but that days or weeks elapse before a
filtered test of the extract shows its maximum zymolising power.

DISTRIBUTION OF NITROGEN IN AUTOLYSIS 397
TaBLe II.—All figures are in c.c. N/10 H,SO, for 100 c.e. filtrate or 25 grams ox liver,

Period of Amino Undeter- Total
incubation NH,;N Amide N acid N- mined N_ Sol. N

3-0 11-2
80. Fresh liver... sarin, OL) 1-75 6-4 56-7
1-75 6-25 57-0
7:2 14:3
(2) l 6-0 11-9 83-2
5-7 13-4
. Fresh liver... ss, (1) 5 hours 3-4 7:95 59-0
9-5 13-8
(2) 1 day 6:8 9-9 72-0
7-2 12-9 30-2 49-7
Bomeesh liver... “ee, (Ql) Ue 5:8 10-4 24-4 40-2 81-0
8-0 12-0 32-2 47°8
(2) 2 days 8-0 11-9 32-0 47-5 99-0
6-7 11-0 22-1 60-2
mresmeniver ... <5. (1) day 5:0 8-2 16-6 45-1 75-0
8-7 12-5 31-0 47-8
(2) 4 days 7-6 10-9 27-0 41-5 87-0
10-3 16-5 34-0 39:2
(3) 20 days 10-7 17-0 35-4 40-8 104-0
9-0 18-5
. Fresh liver... ... (1) 2 days 8-2 16-8 90-5
9-4 21-0
(2) 4 days 10-4 23-1 110-0
11:3 20-8
(3) 7 days 12:8 23-5 113-0
9-0 16-0
. Fresh liver... as (2) Sidays 11-9 21-2 131-0
7:3 18-5
(2) 32 days 13-8 32-6 176-0
8-7 14:8 39-8 36-6
. Fresh liver... «. (L) 9 weeks 17-4 28-4 76:8 196-0
16-4 29-2 77:8 71-0 192-0
9-0 15:8 45-5 29-7
(2) 94 weeks 16-5 28:8 83-0 53-7 182-0
11-0 15-1 52-4 21-5
(3) 12 weeks 19-6 26-8 92-9 37-7 177-0
9-9 12-9 49-8 27-4
. Fresh liver... -. (I) 9 days 11-4 14:8 57:3 31-5 115-0
10-2 15-9 54:3 19-6
(2) 3 weeks 15-1 23-5 80-4 29-0 148-0

398 BIO-CHEMICAL JOURNAL

TaBLE I1.—Continued.

Period of Amino Undeter- Total
incubation NH;N Amide N aeid N~ mined N Sol. N

11-8 11-6 51:0 25-6
37. Fresh liver... ... (1) 9 days 11-8 11:7 51-1 25-4 100-0
11-6 12-0 48-8 27-6
(2) 2 weeks 13-5 14-0 56-5 32-0 116-0
11-9 12:4 46-7 29-0
(3) 3 weeks 13-6 14-1] 53-3 33-0 114-0
3-6 11:3 17-0 68-1
34. Fresh tissue ... seme (i) 0 2-0 6-1 9-0 31-2 53-5
1:8 5-85 8-9 51-5
7-4 10-0 28:9 53-2
(2) 2 days 6-0 8-45 22:8 79-0
5:8 8-35 23-4 42-6 81-0
a 8-2 11-9 26-5 53-4
(3) 4 days 7:5 10:3 23-3 89-0
7:3 11-2 24-7 48:3 92-0
8-2 13-5 29-9 48-4
(4) 15 days 8-8 14-4 31-6 106-0
8:7 14:5 32:3 51:8 108-0

Norr.—The upper row of figures in each case represents percentages.

EFFEctT oF AcID AND ALKALI ON AUTOLYSIS

It is on considering the effects of acid and alkali on autolysis that
the value of the percentage relationship is at once apparent. There is
undoubtedly an inhibition of autolysis on addition of alkali and an
acceleration on that of acid. Whether the effect is merely one of
inhibition or acceleration, or whether it is something more, can only be
determined by an investigation of the nitrogen partition.

The earliest record of the action of alkali in diminishing the amount
of total soluble nitrogen in antiseptic autolysis is by Schwiening?® in
1894. Two years later, Biondi?! noted an increase of total soluble
nitrogen when acid was added to the digest. Hildebrandt?? found that
an addition of acid or alkali to the mammary gland caused an increase
and diminution respectively of the total soluble nitrogen.

Hedin and Rowland in 190123 found that the spleen of various
animals contained a proteolytic ferment which acted most strongly in
acid solution, and least strongly in presence of alkali. In the same
year*4 they found similar proteolytic enzymes in the liver, lymph gland
and kidney of calf, dog and horse.

In 1903/04 Hedin?> came to the conclusion that the spleen contained

DISTRIBUTION OF NITROGEN IN AUTOLYSIS 399

two proteolytic ferments, an a protease which only acted in alkaline
medium and a § protease which only acted in an acid medium.

In 1906,°° he found that previous treatment with alkali was enough
to inhibit autolysis in organs, although acid was added subsequently,
while previous treatment with acid sufficed to accelerate it in spite of
subsequent addition of alkali.

Baer and Loeb in 1905? found that small amounts of alkali
diminished autolysis (1.e., the total soluble nitrogen), while large amounts
accelerated it. This latter observation has not been confirmed.

In 1906, Wiener?® found that alkali diminished the total soluble
nitrogen in liver digests, and that alkali present in amount equal to that
of the blood allowed no increase at the end of one week’s incubation.
The same year Drjewezkil!’ worked at the influence of alkali on autolysis
of the liver—examining not only the total soluble nitrogen but also the
monamino, diamino, and albumose nitrogen. He found that autolysis
took place in an alkaline medium of 0°2 per cent. to 0°3 per cent.
concentration, but was much diminished. It may be noted, however,
that in the alkaline tests, monamino nitrogen has a smaller percentage
value than that of the control test, while the opposite condition is found
in the case of the albumose nitrogen.

Shryver,?? following up some work done by J. Lane Claypon and
himself on autolysis in the fasting and fed animal, examined the influence
of acid and alkali on autolysis as indicated by the total soluble nitrogen.
He found that acids markedly accelerated the process, the amount not
the degree of concentration of the acid being the important factor.
Lactic acid was more powerful than H,SO,. Alkali inhibited autolysis.

In 1907, Arinkin*° thoroughly investigated the influence of inorganic
and organic acids on liver autolysis, and also dealt with the question of
the hydrolytic action of the acid on the liver tissue. The conclusion
arrived at was that the protein cleavage was due to ferment action alone,
and not to acid hydrolysis. The action of the acid was to stimulate the
ferments. In his experiments, incubation was carried on for forty-eight
hours. The most interesting conclusions are that:

(1) There is a definite optimum for each acid, beyond which autolysis

diminishes.

(2) In equivalent and percentage relations inorganic and organic

acids behave differently, i.e., some stimulate more than others.

The accelerating action of acid is specially marked in the first stage

of autolysis.

400 BIO-CHEMICAL JOURNAL

Preti3! in 1907 made another contribution to the action of alkali on
autolysis. He found that small amounts of alkali inhibited the autolytic
action. lLaqueur3? investigated the effect of carbon dioxide on liver
autolysis, and found that it produced a marked increase in the total
soluble nitrogen, thus confirming an observation of Yoshimoto??.

Experiments 28, 33 and 35 deal with the influences of acid and alkali
on autolysis. In 28, lactic acid, 20 c.c. N/10, was added to one series of
tests, while a similar amount of N/10 NaOH was added to another series, a
third series being used as controls. Incubation was carried on for two,
four and seven days respectively.

In 33, lactic acid was added to one series and sulphuric acid to
another in amounts equal to 19°75 c.c. N/10; incubation was carried on for
one, four and.twenty days respectively. Finally, to determine whether
prolongation of the incubation for a considerable period made a marked
difference in the relations already found in the first two experiments,
Experiment 35 was carried out; incubation lasted 94 and twelve weeks
respectively, 20 c.c. N/10 sulphuric acid being added to one series and
the same amount of NaOH to another series.

It is apparent from the total soluble nitrogen figures on Table III,
that the effect of the acid is to markedly increase, while alkali has the
opposite effect. If the control test be taken as 100, the following
relations are obtained :—

Control Acid Test Alkali Test

(1) 1 day’s incubation... S| MEOW 130 —
(2) 2°", 3 se ae E00 181 83
(3) 94 weeks’ As oe Be Lil) 136 67
(4) 225, ; " ee BoP aRO0) 151 68

The percentage figures are full of interest, for there are unmistakable
differences in the nitrogen partition in the control, acid and alkaline
tests respectively. The percentage of ammonia is distinctly lower in the
acid than in the control tests, while that of the alkaline test tends to be
higher. These characters are maintained in all the periods of incubation
investigated. It is interesting to note that the ammonia percentage in
the acid test is almost as low at the end of one day’s incubation as that
of fresh tissue. The percentage amide figures on the whole are higher in
the acid and lower in the alkaline tests than in the control. 28 (2) is an
exception, but these tests were unfortunately overheated in the autoclave,

DISTRIBUTION OF NITROGEN IN AUTOLYSIS 401

and only the control was capable of repetition. These characteristics are
also persistent through the whole incubation period.

The percentage of amino agid nitrogen in 33 and 35 shows the same
characters as the amide nitrogen, i.e., a relative increase in the acid tests
and a relative diminution in the alkaline test.

The percentage of undetermined nitrogen is higher in the acid and
lower in the alkaline than in the control tests, and this difference holds
good at the different stages of incubation.

Before discussing the significance of these changes, another relation
may be brought out, namely, that in all the acid tests the percentage
values for ammonia and amides have a ratio between 1 : 2 and 1 : 3,
whereas in the alkaline tests the ammonia and amide percentage value
approximates 1 : 1°6.

The fact that my experiments show a definite type of autolysis in the
presence of acid and alkali respectively, which is persistent throughout
the whole course of digestion, leads to the conclusion that the effect of
acid is not merely to accelerate, neither is that of alkali merely to inhibit
autolysis. Arinkin noted the peculiar cleavage of the protein molecule
in the presence of acid in liver autolysis. The nitrogen of the monamino,
diamino, ammonia, albumose and peptone fractions was usually
increased. This observation was abundantly supported. Yoshimoto, on
the other hand, found that the cleavage of the protein molecule by the
liver in presence of carbon dioxide occurred in exactly the same way as
in normal autolysis. His conclusion carries less weight, as his
investigation was primarily directed to the influence of antiseptics on
autolysis.

In Drjewezki’s work the percentage of monamino acid nitrogen in
autolysis in presence of alkali is smaller than in normal autolysis, which
bears out the contention that there is a characteristic type of autolysis in
presence of alkali.

The nature of this alteration in the type of autolysis is by no means
clear. We are hampered in the first place by our ignorance of the origin
of the ammonia and amide nitrogen in normal autolysis.

One of the sources of this ammonia is doubtless the so-called amide
group in the protein molecule, which is not only readily split off by the
action of acid or alkali, but which can be liberated in even larger amount
by the proteolytic ferments of the gastric and pancreatic juice according
to Driergowski and Salaskin.34

Another possible source is preformed amide nitrogen. Dakin!6

402 BIO-CHEMICAL JOURNAL

found an increase of ammonia nitrogen in kidney autolysis at the expense
of the amide nitrogen, the latter decreasing in proportion as the former
increased. In the light of my subsequent experiments on bacterial
contamination of tissue digests, however, this result may have been due
to bacterial action.

Vernon!® holds that the hydrolysis of urea by liver tissue is much
less doubtful than its formation—but even so the evidence is scanty.
- Jakoby®5 found that liver juice liberated ammonia from urea after
thirty-six hours’ incubation, although his experiments are not numerous
and the results are not very concordant. Gonnerman*® found a liberation
of free acid from certain amides by liver tissue, and thus produced
indirect evidence of a deamidization. Lang? found a small amount of
ammonia after incubating urea, and an abundant liberation after the
incubation of asparagin with liver tissue. Furth and Friedmann*’ found
an asparagin-splitting ferment in the liver, spleen, muscle, kidney, lung,
brain and bowel. I? also found a considerable liberation of ammonia
from the amide asparagin—after incubation with liver emulsion.
Lastly, Lang’s work would suggest the amino acids as a_ possible
source of ammonia in autolysis. As will be shown later, the percentage
of amino acid at the end of incubation shows no diminution, and the
ratio of ammonia to amino acid nitrogen does not increase with prolonged
incubation; this fact certainly suggests that the amino acids do not
furnish appreciable amounts of ammonia during autolysis.

In considering the cause of the characteristic type of alkaline
digestion, the question arises whether the presence of alkali favours an
enzyme which liberates ammonia from amides, for I have shown that the
percentages for ammonia and amide nitrogen approached each other more
closely than in any other type of autolysis yet considered.

The difficulty of answering this question is great, for undoubtedly
in such a liver emulsion we have not one enzyme but many, some of which
at any rate carry on their activities simultaneously. The continual
production of more amide nitrogen would in this way mask the action of
the amide splitting ferments.

The work of previous investigators is sufficient to show that the
difference in the action of acid and alkali is not merely one of acid or
alkaline hydrolysis respectively. Drjewezki and Arinkin both compared
the action of acid and alkali on boiled tests and found no appreciable
effect. This method of putting the ferments out of action is certainly
not ideal, but it is the only available one.

DISTRIBUTION OF NITROGEN IN AUTOLYSIS 403

On the whole, it seems probable that the different types of autolysis
in presence of acid and alkali are due to an initial difference in the
protein cleavage rather than to variation in the subsequent catabolism of
any of the nitrogen fractions of the digest, for there is no evidence of
any alteration in the relative proportions of the latter, i.e., such as might
be expected if an increase of one fraction at the expense of another
occurred. On the contrary, the characteristic type is present at the end
of twenty-four hours’ incubation and persists throughout digestion.

INFLUENCE OF BACTERIA ON AUTOLYSIS

No investigation on the effects of tissue digestion can afford to ignore
the question of bacterial action, as was pointed out in the previous paper.

Samuely,?® in Oppenheimer’s Handbuch, states that toluol does not
inhibit the tissue ferments in a concentration sufficient to stop bacterial
action—this is a very general statement which lacks confirmation. The
difficulty of testing its accuracy in the case of autolysis is obvious, and
in the hope of finding a characteristic partition of nitrogen for the
common bacteria of putrefaction, some experiments were made in a
manner similar to those carried out in the investigation of deaminization.

In Experiments 30 and 31, three series of tests were prepared, each
consisting of 100 gm. liver tissue in the usual way :—

(1) To the controls the usual addition of toluol was made.

(2) The boiled tests were heated to 100° and were then allowed to
cool, after which about 0°5 c.c. of the mixture from the third
series of tests (3) was added to each one—to inoculate them
with the bacteria of putrefaction.

(3) No addition of antiseptic was made to these tests, which simply
stood at room temperature until the boiled tests had cooled.
1 c.c. of the mixture was then removed from each test, and after
mixing was used for the inoculation of (2).

All the tests were then incubated in the usual way after having been

shaken in the machine for not less than half an hour.

(1) contained enzyme and no bacteria or a minimal quantity.

(2) contained bacteria and no enzyme.

(3) contained both enzyme and bacteria.

The naked eye changes were striking after one day’s incubation, the
bacterial tests showing abundant gas formation, together with alterations
in the colour and in the consistence of the fluid. In Experiments 30
and 32 the gas formation was so active in the (3) tests that some of the
liver mixture actually spilled over.

‘soSvquoosad syueseided osvo youe ul soansy jo Aor oddn oy ,[—aLON

0222 GPS 698 6rE LST ° pr Ae 4 “* O-FOL SOF FS OLT LOT SXep 02 (8)
TeceelGeoeler wi = G6E OF GOL. FOL
I ee eke ee ee ese ee ee
0-0 F6S 899 693 FL O6FT G09 OFF 892 LL O18 GIP OL6 GOL IL skep - (zg) °° . + Apoatqoodsor 83899 'OS*H
96€ GLE O81 6F 90F £98 OST TS SLP O18 GSl L8 pue prow oor, 07 'ORS*H pues poe
: opr] OL/N ‘2 GL-6T JO UOINIppy “EE
OSOT 9:89 &1E T9T 6& OTOL GOS SIE SST SS OGL Th 9OT G8 OF kept (1)
T1¢ 862 €SI 8-€ 66F O18 SST Se G09 1:26 OTL L9
= plow o1poe'y Oo aag
S OIZI L9E OLS EB OST 0892 SIF OLST 89F LBS OLLT LLE 666 896 96 sdep ZI ( ars
a GOE LF LOL &3l GG] £80 FLT 8 GIS 6S (TS OTT ob a + poatqoodsor
b ae sjsoy plow pue oulpeyye 03 'OS°H
= O€sl LSP 9-8F SCL GFL ORFS FOF OLL GOP TLI OSI LES O-€8 883 G-9T Ap f6 ( oF epos OT/N ‘29 OZ JO UOIPpy “Ss
a L6E FSS Sst sl 6-81 if eee 661 89 L6G GSh SST 06
OSH
= =
= 0:06 Chem sOolG. = TOF Sl O-8IT GES 8st Sep L
L SSL SIT G8l 0-9 8-06 & IT
—
= O\Ogen a. oS OO HG OCG] 8 : ZIFF LOL O-OIT SSI FO step F ave + Ajoatqgoodser
SSL SIT FIZ Gg StL 6 $980} plow’ pUw OUTTVYTe OF POV O14RT
i pue epos QT/N ‘9'°0 0% JO UOIFIPpY “8s
er Oe: F G6 GhL O-€9T PenOOc. ook 2:00) = SOL sg skepZ
(=) ZZl «66 6SL FF GSI 0-6
aS ploe oorry
Be a > qc > Mae,
ma #2 BE & 42 OH BS Sie ie + ae fee BS gE & 4 2
me ue Te) ee ee ee ee ee Be ee a Oe ee
pan 2 ees, Fe ne BAe A Ah ee ee
a aa oat Y
TIVUIV aloy TOULNOD

‘IOAT] XO SUIBIS GZ IO 0YVAI[F *0°O OOT IOF "OSH OI/N *9°O Ul ore somnsy 04} [TV “IIL S14.

404

——— ee eS me et he a

DISTRIBUTION OF NITROGEN IN AUTOLYSIS 405

The figures are no less striking, the total soluble nitrogen after one
day’s incubation is more than twice as great in the unboiled as in the
antiseptic tests—in the boiled test, on the other hand, it is practically the
same as in the fresh tissue. ‘This may be due to the fact that coagulated
protein is less readily attacked than uncoagulated, or that the bacteria in
the boiled tests had not had enough ‘ start.’

The most interesting point, however, is that in both boiled and
unboiled tests the percentage amount of ammonia is increased in relation
to that of amide, whereas in the control tests the opposite relation is
found. The obvious explanation is that part of the ammonia is being
liberated from amide nitrogen by bacterial action. Although it must be
remembered that Brasch?? has shown that ammonia is readily liberated
from amino acids by bacteria.

The tests worked up at the end of five hours’ digestion in 31 show
ratios for ammonia and amide very like that of the control, the inference
being that in five hours bacterial growth and action had not made itself
felt to any extent.*

In Experiment 32 no boiled tests were prepared, only tests with and
without antiseptic—the reversal of the normal amide and ammonia ratio
is again visible in the bacterial tests, and at the end of eight weeks’
incubation it holds good.—See 35 A.

Undetermined Nitrogen.—In Experiment 32, after one and two
incubations respectively, the actual and percentage amounts of
undetermined nitrogen are larger than those of the controls, but in
Experiment 36 (2), A and B, the percentage amounts at the end of seven
days’ incubation reach the lowest limit in any of the Tables, namely
20°6 per cent. and 20°4 per cent.

In Experiments 35 and 36, the effect of adding varying amounts of
antiseptic was tried, three series of tests being prepared, each of 50 grams.
To A no antiseptic was added, to B 0°5 c.c. toluol, and to C 1 c.c. toluol
was added. All the tests were then shaken by hand for a few minutes
before incubating them.

On looking at the tabulated results, one is at once struck by their
irregularity. Certain general characters prevail for test A, i.e., the
ammonia percentage is always higher than the amide This is true for
B and C in Experiments 35 and 36 (2), but in 36 (1) the normal (i.e., in

antiseptic) relation of ammonia and amide is found.

* The amide tests were overheated in the autoclave and the values are consequently rather
high,

la,

406 BIO-CHEMICAL JOURNAL i

The amino acid nitrogen show a better sequence in 30 (1), (2), and
35, i.e., after three days, seven days, and eight weeks’ incubation
respectively, than any of the other fractions. The amino acid percentage
increases from the third till the seventh days, and diminishes again at
the end of the eighth week. It is highest in the case of the A and B tests.

The odour varied, being sometimes putrid and sometimes yeasty.
In. Experiment 32 the odour was intensely yeasty, in 36 it was noticed
that the A test had an odour of sulphuretted hydrogen after one day’s
incubation; after three days’ incubation the A tests, however, had also
a sour yeasty odour. At the end of seven days’ incubation there is no
note of any odour, but growing on the surface of A a mould was noticed.

In 35, after eight weeks’ digestion, A had a putrid odour and a
strongly alkaline reaction. B had an unpleasant but not putrid odour,
and was also strongly alkaline; while C had an acid odour suggestive of
butyric acid, and had a faintly acid reaction. The reaction of all the
other tests was faintly acid.

In the light of Ellinger’s?® work on the chemistry of putrefaction,
the variations found as the result of the small amounts of antiseptic
added may probably be explained on the basis of variations in the
bacterial flora.

The following factors would lead one to expect bacterial contamination
in an organ digest ; viz., a high ammonia percentage combined with a lower
amide percentage, and a high total soluble nitrogen figure.

It is interesting to note that the bacterial type of nitrogen
distribution approximates more closely to the alkaline type of autolysis,
with regard to the ammonia and amide ratios; the amino acid percentage,
however, is higher than in the case of the alkaline autolysis.

A few words may be said about the influence of antiseptics on
autolysis.

Cruickshank,4! from a study of the histological appearances in
organs undergoing autolysis, found that the action of toluol markedly
delayed the onset and progress of autolysis—the liver showed complete
necrosis on the third day, in aseptic autolysis, while well-stained nuclei
were found after five or six days if autolysis took place in the presence
of toluol vapour. He also points out that toluol vapour is enough to
completely inhibit the growth of Staph. aur. and B. coli; the organisms
were not killed, however, growth occurring on the removal of the toluol.

Navassart4? investigated the action of antiseptics on yeast autolysis

‘sodequooied syuosordod osvo Youe ur soanSy yo mor aoddn oy .—'aLon

2 O-9FT 0-FF 8-6L 0-01 ZZI G-8L 9-08 0-98 OL PF 0)
Z-08 9-FE 6-9 £8 o-6E 8-GF 9-6 PG
| O-8IZ C-FF F681 €-91 SLT G-LOL P&L O-9L O-FI laa a
F:-0 0-9 GL L8 8-8F 6-6F a6 9% |
5S (2) | ¥. .BQe 9-8F 8-921 LOL 9-08 f O-FZ1 0-CF 0-09 O-¢g 0-FI step ¢ v]
~ 9-0% 6-69 Le FFT sep 1 G98 €-8F O-F ran
N}S* N pout Npoe Nopmy NHN wornreqnour (I)°" 98
[80], -lajopuyy ss OUT jo polleg
PD ee
e quosaad jonjoz *9°9 O-T = 9 0-881 EL G-9% G9 LLY a)
eS £98 8-91 LP 8-ZP |
= quosoid fonjoz *0°0 ¢-.9 = G-Z81 F-&% €-£9 F8E FLG dd
raf 0-€1 9-FE 0-1Z F1E
a ondosyque ynoyyT = VW G-91Z LZ 9-98 raya OGF syooan g v|
0-62 0-0F Z-01 8-0 a
7, 9g pue cE sjuowedxg UO syaRuUoyy
—_
; ondasyue oN 0-692 €-FST 0-29 69-01 O-ZE 0:66 G-LF 0-ZE 6-11 0-8 step Z
i> 6-6 O-F% LF €-Z1 8-LF BBE 0-21 0-8
5 ondesiyue ON O-LFI Ze O-eE FI FLT 0-18 Z-OF FF F-01 8-¢ Aep [too hee
= O-8¢ 6-2 LL 8-11 L-6F Z-0 6-ZI GL
= poeqyenooutr
= pue polog 0-¢F eit ¥: SL-F 8-¢
A ¢-O1 O-€I
a ondasiyue oN 0-82 : FLT G-ZS O-ZL 7 6-6 8-9 Avp |
~ FL 9-6 8-E1 G6
4 peqyepnoout
= pue poplog 0-8 = ; C6-F COE
os GIT O-L
= ondastue on O-1L ie ie 9-8 0-F 0-66 : : COL PE sanoy ¢ " 19
4 LZ LG FE LG
x :
a poqzejnoour
Ke pue paplog 0-68 Z “ LT €-8
lan 6% O-F1
ordastzue oN O-F1Z ae : $6 F-9% o-€8 “ae 6-11 0-9 Be a rr O08
F €-ZI €-F1 GL
0-00 a < 8-3 LUL C06 ; as 8-91 a8 step Z ° A> Ae
6-01 ar CRT 06
NPS Npeuwm N poe Nopmy N*HN NPS Npoumt N phe Nopmy N*HN uoNneqnour
1®99.L -dojyopuyQ) =: ou 240, -lojyopuy) = OUT W JO polleg

SS ——— y —
IVINALOVE,

TIOULNOY

408 BIO-CHEMICAL JOURNAL

in 1911—comparing the various antiseptics with chloroform. He showed
that a half saturated solution of toluol did not destroy bacteria.

Laqueur!? stated that sodium fluoride in 0°3 per cent. solution acted
well as an antiseptic and was less injurious to the autolytic ferments than
toluol.

With regard to so-called aseptic autolysis, Magnus Levy,** in 1902,
found that the addition of antiseptic caused delay and a marked
diminution in autolytic changes. At the end of twenty-four hours’
incubation the liver was intensely acid, soft, and floating in a dark froth,
while it was covered with a foam layer.

It must be mentioned that Magnus Levy’s figures show no regular
and in the light of Wolbach and
Saiki’s work,45 it is an open question whether bacteria were not largely

sequence even in the serial tests

responsible for the-above changes. ‘These workers removed the liver from
dogs with elaborate aseptic precautions, and found that only two remained
sterile out of twenty-three. The two sterile livers were firm and elastic
after forty-eight hours’ incubation, while the ones showing the bacillus
were soft, discoloured, showing gas formation and having a_ peculiar
rancid odour.

The organism found was a spore bearing anaerobe, which was
cultivated with considerable difficulty.

Jackson*® found that the total soluble nitrogen at the end of
forty-eight hours in the two sterile livers of Wolbach and Saiki was only
equal to that of the non-sterile livers at the end of ten to twelve hours’
incubation.

Lindemann,*’ 1911, studied the autolysis of aseptically removed
livers of rabbits, cats and dogs, finding gas formation (carbonic acid
and hydrogen) and acid formation if the organs were immediately
incubated at 37°. His results fully confirm those of Magnus Levy. He
mentions the fact that organs other than the liver and heart never
remain sterile, while one-third only of the heart and liver remained
sterile, i.e., negative cultures were obtained from them.

FORMATION OF AMIDE NITROGEN FROM AN AMMONIUM SALT IN VITRO

The question may now be considered as to the fate of a neutral
ammonium salt when incubated with liver tissue under the conditions
already described. We shall find it convenient to look at the results of
the last two tests first, namely Test 37, Table V. The ammonia was
added to the liver tissue in the form of ammonium hydrate, neutralized

DISTRIBUTION OF NITROGEN IN AUTOLYSIS 409

with lactic acid. The amount added to 100 grams liver was equal to
38°0 c.c. N/10, i.e., 9°5 c.c. for every 25 grams of tissue or 100 c.c. of
the final filtrate.

The percentage values in relation to the total soluble nitrogen are
calculated after subtracting 9°5 (the amount of added ammonia) from
the figure for the total soluble nitrogen. The percentage for the
ammonia values are calculated after subtracting 9°5 from the figure for
the ammonia also.

Three series of tests, 1, 2 and 3, were incubated nine days, two weeks,
and three weeks respectively. The ammonia recovered at the end of that
time was 9°7, 9°1, and 10°4 respectively. Although in the ammonia tests
the actual figures for the amide nitrogen all show a slight increase, the
percentage figures do not bear this out, the differences between the
control and the ammonia tests being within the limits of experimental
error. The same holds good for the amino acid figures. The total soluble
nitrogen (after the subtraction of the equivalent of the added ammonia)
is higher than that of the control test in one case, the greatest difference
being in No. (3). From these figures it is evident that there has been
no appreciable change in the ammonium salt added, and there is no
evidence of any of it being converted into amide nitrogen.

In 35, two series of tests were incubated 94 and 12 weeks respectively.
35°6 c.c. N/10 ammonia neutralized with H,SO, was added to the
ammonia tests or 8°9 c.c. for 100 c.c. filtrate. The amounts of ammonia
recovered were 9°6 in (1) and 7°7 in (2)... The amide figures in absolute
and percentage amount are lower than those of the control. They are so
low im (1) as to raise suspicion of error—the amino acid figures are
actually rather higher than in the control, but the percentage figures,
although higher, are almost within the limits of experimental error.
The total soluble nitrogen figures are equal in (1) and slightly higher in
(2). These results confirm those of Experiment 37.

In 29, where ammonia equal to 20°4 c.c. N/10 neutralized with
lactic acid was added to 100 gm. liver, two series of tests were incubated
eight and thirty-two days respectively. The amount of added ammonia
calculated for 100 c.c. filtrate is 5°1 c.c. N/10. Im (1) 7°95 cc is
recovered, and in (2) 2 c.c., but the latter result is doubtful in view of
the fact that the value for ammonia is actually lower in the (2) ammonia
test than in (1). There is a small actual increase of the amide nitrogen
in both tests, but the percentage increase is negligible.

410 BIO-CHEMICAL JOURNAL

Taste V.—All figures in c.c. N/10 H.SO, for 100 c.c. filtrate or 25 grams liver tissue

CoNTROL
Hs ; =
ise: z : 2% a Z
Bs c s ake SF =
4 = v= ino}
ao pI (3) 2.8 ~
& 2 = < ~e De BZ
9-1 15-8 45-6 29-5
35. NH,N added in ammonia (1) 93 weeks 16-5 28-8 83-0 53-7 182-0
test to 100 grams liver =
35-6 c.c. N/10 H2SO, 11-1 15-1 52-5 21-3
(2) 12 weeks 19-6 26-8 92-9 37-7 177-0
11-8 11-6 51-0 25:6
37. NH,N in glycocoll added (1) 9 days 11-8 11:7 51-1 25-4 100-0
to 100 grams liver = 55-6
N/10 HoSO, 11-6 12-0 48-8 27-6
(2) 2 weeks 13-5 14-0 56-5 32-0 116-0
11-9 12-4 46-7 29-0
NH,N added to 100 grams (3) 3 weeks 13-6 14-1 53:3 33-0 114-0
liver in ammonia test = 38
c.c. N/10 HSO,
9-9 12-9 49-8 27-4
36. NH,N in glycocoll added to (1) 9 days 11-4 14-8 57:3 31-5 115-0
100 grams liver = 55:6
N/10 H.SO, 10-2 15-9 54:3 19-6
(2) 3 weeks 15-1 23-5 80-4 29-0 148-0

9-0 16-0
29. NH;N added to 100 grams
liver in ammonia tests = 8 days 11-9 21-2 Bs 432 131-0
20-4 c.c. N/10 H2SO,

7:8 18-5
NH,N in glycocoll added to
100 grams liver = 53-2 c.c. 32 days 13-8 32-6 act ies 176-0
N/10

Norr.—The upper row of figures in each case represents percentages.

DISTRIBUTION OF NITROGEN IN AUTOLYSIS 411

GLYCOCOLL AMMONIA
Z Lz Ss 332 3 Z t24 3 6 4z
A. 2 Tigqeme hema, Ais 5 A 2 gat) ee 18 2 5
bri = gz ge 3 gas a by = <b}: ce
= E Be Sa as 2°53 8 = = an) fa S$,
DPR ea eee ae ‘ = DA «am sed

_— _— = = — — 9-6 26-1 20-5 88-6 55:8 191-0 182-1
10-1 12-6 56-1 21-2
— = = = — — 7:7 27:3 22-9 102-2 38-6 191-0 182-1

11-0 11-2 49-3 28-5

_— — — — — — of 21-5 12-2 53-4 30-9 118-0 108-5
10-8 12-7 51-0 25:5

— — — — —_ — 9-1 22-6 15-2 60-9 303 129-0 119-5
11-1 12-9 48-0 28-0 10-7 12-2 44-1 33-0

15-2 Leer Uys 38-4 151-0 137-1 10-4 23-9 16-4 59-1 44-2 142:0 = 134-3

— = “= = == == == 24-1 16-5 59-1 — 145-7 —
10-2 14-7 54-7 20-3
A 16-8 76:3 23-2 1280 = 114-1 — — — — — — —
10-1 17:3 55-5 Li]
15:3 26-3 98-2 26-2 166-0 152-1 — — — — — — _—
8-1 iS 8-9 16-8
10-2 22-8 — — 139-0 =125-6 4-9 16-85 22-4 — — 138-0 = 132-9
6-2 19-7
2-0 15-8 34-9 — = 1770s: 171-9

Norr.—The upper row of figures in each case represents percentages.

412 BIO-CHEMICAL JOURNAL

It is interesting to note in connection with the above that Loewi was
unable to obtain urea formation in a liver extract to which ammonium

acetate had been added.

ForMATION OF AMIDE NITROGEN FROM GLYCOCOLL IN VITRO

After finding that the formation of amide nitrogen from a neutral
ammonium salt in liver digests is absent or so small in amount as to be
negligible, one would not expect to find evidence of much amide
formation when an amino acid was incubated with the liver tissue. .

As in the ammonia tests, the percentage figures are calculated from
the total soluble nitrogen after subtracting the amount of added glycocoll
nitrogen. The same amount is subtracted from the figure for the amino
acids before calculating its percentage amount of the total soluble
nitrogen.

In Experiment 36, the amount of glycocoll nitrogen calculated for
100 c.c. filtrate was 18°9 c.c. N/10. Two series of tests were incubated
nine days and three weeks respectively. The increase in ammonia nitrogen
is so small as to be negligible, but there is a distinct increase in the
amide nitrogen both in actual and percentage amounts. In (2) the amide
values of control and glycocoll test are both, unfortunately, too high, as
the autoclave was overheated for about fifteen minutes. This does not
gravely interfere with the comparative values, however. The amino
acid shows a very distinct actual increase in the glycocoll tests with a
less marked percentage increase. There is practically no increase in the
total soluble nitrogen, so that apparently autolysis has not been more
rapid in the glycocoll test.

In 37, after three weeks’ digestion, practically the same conditions
are found, the percentage amide increase in the glycocoll test is
considerably smaller than the actual increase, and is within the limits
of experimental error. The same is true for the percentage of the amino
acid. Here the total soluble nitrogen is distinctly larger than in the
control test. '

In 29, after eight days’ incubation, there is a very small actual and
percentage increase in amide nitrogen. Both Jakoby and Loewi state
that they found a small formation of amide or urea nitrogen on incubating
glycocoll with liver extract.

The evidence on the whole is in favour of a small formation of amide
nitrogen from glycocoll.

DISTRIBUTION OF NITROGEN IN AUTOLYSIS 418

CoNCLUSIONS

1. No satisfactory conclusions regarding the fate of ammonium
salts or amino acids when digested with liver tissue can be drawn
without a consideration of the total soluble nitrogen and such fractions
as the following:—i.e., ammonia, amide nitrogen and amino acid
nitrogen. Both absolute and percentage amounts (of the total soluble
nitrogen) must be considered.

2. The influence of certain factors on Autolysis.

(a) Period of incubation.—The percentage distribution of
total soluble nitrogen in fresh liver tissue is characterised by an
extremely low ammonia content, and a low amino acid content.
After forty-eight hours’ incubation the most definite change is an
increase in the amino acid fraction at the expense of the
undetermined nitrogen fraction. The rate of autolysis reaches a
maximum within forty-eight hours, after which it gradually
diminishes.

(b) The effect of reaction of mediwm.—Acid markedly
stimulates and alkali causes a marked depression of autolysis.
Acid and alkali each produce a characteristic nitrogen partition
throughout the whole incubation period. Autolysis in presence
of acid is distinguished by a lower ammonia and undetermined
nitrogen percentage, and a higher amide and amino acid
percentage than that of control normal autolysis. Autolysis in
the presence of alkali is distinguished from the control autolysis by
a higher percentage of ammonia and undetermined nitrogen and
a lower percentage of amide and amino acid.

(c) The presence of putrefactive organisms in liver digests is
characterised by a relatively high ammonia and a relatively low
amide percentage combined with a large amount of total soluble
nitrogen.

3. There is no evidence of an appreciable conversion of ammonium
sulphate or lactate into amide nitrogen when digested with liver pulp.
Such differences as were found were within the limits of experimental
error.

4. There is no evidence of a liberation of ammonia from glycocoll
when incubated with liver tissue. There is a slight actual and
percentage increase in amide nitrogen, however.

In view of the fact that the total amount of ammonium is markedly
increased when putrefactive organisms are present, conclusions regarding

414 BIO-CHEMICAL JOURNAL

deaminization by autolytic ferments should be received with caution
unless accompanied by figures showing the relation of ammonia to the
total soluble nitrogen of the digest. The fact that even at the end of
twelve weeks the ammonia shows no increase at the expense of the amino
acid fraction is itself an indication that deaminization of amino acids by
liver digests can only occur to a very limited extent.

I wish to express my sincerest thanks to Dr. E. P. Cathcart for
supervising this research. I am also indebted to Professor D. Noel Paton
for his helpful criticism.

The expenses of this investigation were in part defrayed by a grant
made to Dr. E. P. Cathcart by the Carnegie Trust.

i at I

Capes (at) BS

DISTRIBUTION OF NITROGEN IN AUTOLYSIS 415

LITERATURE

Bostock, Bio-Chemical Journ., Vol. VI, p. 48, 1911.
Lang, Hof. Beit., Vol. V, p. 321, 1904.
Loewi, Zeit. f. physiol. Chem., Vol. XXV, p. 511, 1898.

Kruger and Reich, Zeit. f. physiol. Chem., Vol. XX XIX, p. 165, 1903 ; Schittenhelm, Zeit. f.
physiol. Chem., Vol. XX XIX, 73, 1903.

Benedict and Gephart, Journ. Am. Chem. Soc., Vol. XXX, p. 1760, 1908.
Henriques and Gammeltoft, Skand. Arch. f. Physiol., Vol. XXV, p. 153, 1911.
Benedict, Journ. Biol. Chem., Vol. VIII, p. 405, 1910.

Wolf and Osterberg, Journ. Amer. Chem. Soc., Vol. XXXI, p. 421, 1909.
Levene and Meyer, Journ. Amer. Chem. Soc., Vol. XX XI, p. 717, 1909.
Henriques and Sédrensen, Zeit. f. physiol. Chem., Vol. LXIII, p. 27, 1909.
Malfatti, Zect. f. physiol. Chem., Vol. LXVI, p. 152, 1910.

De Jager, Zeit. f. physiol. Chem., Vol. LXII, p. 333, 1909.

Henriques and Sérensen, Zeit. f. physiol. Chem., Vol. LXIV, p. 120, 1910.
Simon, Zeit. f. physiol. Chem., Vol. LXX, p. 65, 1910.

Salkowski, Zeit. f. klin. Med., Vol. XVII, Suppl., p. 77, 1890.

Dakin, Journ. Physiol., Vol. XXX, p. 84, 1903-04.

Drjewezki, Bioch. Zeit., Vol. I, p. 229, 1906.

Folin and Denis, Journ. Biol. Chem., Vol. XI, p. 527, 1912.

Vernon, Ergebnisse der Physiol., Vol. IX, p. 138, 1910.

Schwiening, Virch. Arch., Vol. CXXXVI, p. 444, 1894.

Biondi, Virch. Arch., Vol. CXLIV, p. 373, 1896.

Hildebrandt, Hof. Beit., Vol. V, p. 463, 1904.

Hedin and Rowland. Zeit. f. physiol. Chem., Vol. XXXII, p. 340, 1901.
Hedin and Rowland, Zeit. f. physiol. Chem., Vol. XXXII, p. 530, 1901.
Hedin, Jour. Physiol., Vol. XXX, p. 155, 1903-04.

Hedin, Festsch. Hammarsten, Vol. VI, 1906.

Baer and Loeb, Arch. f. Exper. Path., Vol. LIII, p- 1, 1905.

Wiener, Zentbltt. f. Physiol., Vol. XIX, p. 349, 1906.

Shryver, Bioch. Journ., Vol. I, p. 123, 1906.

Arinkin, Zeit. physiol. Chem., Vol. LIII, p. 192, 1907.

Preti, Zeit. physiol. Chem., Vol. LII, p. 485, 1907.

Laqueur, Zeit. f. physiol. Chem., Vol. LX XVII, p. 38, 1912.

Yoshimoto, Zeit. physiol. Chem., Vol. LVIII, p. 341, 1908-09.
Dziergowski and Salaskin, Centrbltt. f. Physiol., Vol. XV, p. 249, 1901.
Jakoby, Zeit. f. physiol. Chem., Vol. XXX, p. 149, 1900.

Gonnermann, Pfl. Archiv., Vol. LXXXIX, p. 493, 1902; Vol. XCV, p. 278, 1903.
Furth and Friedmann, Biochem. Zeit., Vol. XXVI, p. 435, 1910.
Samuely, Oppenheimer’s Handbuch, Vol. IV, p. 522.

Brasch, Bioch. Zeit., Vol. XXII, p. 402, 1909.

Ellinger, Ergeb. Physiol., Vol. VI, p. 292, 1907.

Cruickshank, Journ. Path. and Bact., Vol. XVI, 1911.

Navassart, Zeit. f. physiol. Chem., Vol. LX XII, p. 151, 1911.

Navassart, Zeit. f. physiol. Chem., Vol. LXX, p. 189, 1910-1911.
Laqueur, Zeit. f. physiol. Chem., Vol. LX XVII, p. 82, 1912.

Magnus Levy, Hof. Beit., Vol. II, p. 261, 1902.

Wolbach and Saiki, Journ. Med. Research, Vol. X XI, p. 267, 1909.
Jackson, Journ. Med. Research, Vol. XXI, p. 281, 1909.

Lindemann, Zeit. f. Biolog., Vol. LV, pp. 37-167, 1910.

416

THE EFFECTS OF PURINE DERIVATIVES AND OTHER
ORGANIC COMPOUNDS ON GROWTH AND CELL-
DIVISION IN PLANTS

By N. G. S. COPPIN, M.Sc., University of Liverpool.
From the Bio-Chemical Department, University of Inverpool
(Received July 28rd, 1912)

This investigation was undertaken mainly with the object of studying
the effect of allantoin on cell growth; this compound having been used
with considerable success as a cell proliferant in the treatment of ulcers,
burns, ete., by Dr. C. J. Macalister (Brit. Med. Journ., Jan. 6th, 1912).
In addition to allantoin, other compounds of the same group, and the
sodium salts of some organic acids were used in these experiments, which
were carried out on similar lines to those employed by Moore and Roaf
(Bio-Chem. Journ., 1908, 3, 279) in their work on the effects of inorganic
salts upon the growth and cell-division in plants.

The plants used were a common variety of hyacinth with white
florets. They were placed in hyacinth glasses, which were enclosed in
blue paper bags, in order to prevent the action of light upon the rootlets.
The list of solutions employed was as follows:—Allantoin, sodium
huminate, sodium nucleate, guanine hydrochloride, xanthine, hypo-
xanthine hydrochloride, caffeine, sodium malate, sodium urate, sodium
oxalate, sodium oleate, sodium linoleate, and Bovril. Two controls were
employed, one of ordinary tap-water, and another to which a trace of
sodium hydroxide was added, on account of the fact that certain of the
solutions when being prepared were made slightly alkaline to phenol
phthalein. The solutions were all made up with ordinary tap-water,
and in most cases of three strengths, 0°5 per cent., 0°1 per cent., 0°02 per
cent.

Measurements of both rootlets and leaves were taken from time to
time, and are shown in the accompanying table.

In general terms it may be stated that in the cases of the 0°1 per cent.
and 0°02 per cent. solutions of sodium huminate, 0°02 per cent. sodium
malate, 0°02 per cent. sodium urate, and 0°02 per cent. sodium oxalate,
the rate of growth was distinctly increased, this increase being most

THE EFFECTS OF PURINE DERIVATIVES 417

marked in the sodium huminate solutions; 0°l per cent. Bovril also
stimulated the growth but not to so great an extent as the above solutions,
In every other case, however, the growth of the bulb was either retarded
or completely arrested.

It is also worthy of note that, with the single exception of sodium
nucleate, every other solution of 0°5 per cent. strength which was
employed entirely arrested the growth of the rootlets and greatly retarded
that of the leaves, and in nearly every case the most rapid growth took
place in the weakest solutions.

The rootlets of the bulbs grown in the sodium nucleate, sodium oleate,
sodium linoleate and Bovril solutions, although comparatively short in
length, were very thick.

ALLANTOIN SOLUTIONS
(Lengths in centimetres)

26 Days

54 Days 97 Days 119 Days
Strength of Roots Roots “plleaiee Roots Ee Roots Leaves
Solution (only)
Control (1) 7-5 13-0 3-0 13-0 15°5 14-0 30-0 (flowering)
cs (II) 8-0 13-0 3-0 13-0 14:5 13-0 20:0 (flowering)
01% (I) 3-0 3-5 15 3°5 8-5 4-0 19-0 (flowering)
a o- (il) 1-5 1-5 0-5 1-5 4-5 1-5 10:0 (flowering)
+ (III) 3-0 4:0 1-0 4-0 10-5 5-0 21:0 (flowering)
02% (I) 1-5 1-5 1-0 1-5 4:5 1-5 9-0 (flowering)
% (II) 5:0 5-0 2-5 7-0 8-5 7:0 13-0 (flowering)
is (IIT) 5-0 5-0 2-5 5-0 10-0 6-0 22:0 (flowering)
03% (I) ...No growth Nogrowth 1:0 Nogrowth 3-5 Nogrowth 60 (flowering)
rs (II) . 5-0 6-5 2-5 8-0 4:5 8-0 6-0
bs (IIT) 55 8-0 2-5 8-0 4-0 8-0 4-5
05% (I)... Nogrowth Nogrowth 0-75 Nogrowth 1:5 Nogrowth 3:5 (just com-
, (II) ... Nogrowth Nogrowth 0-5 Nogrowth 2-5 Nogrowth 3-0 mencing
(III) ... Nogrowth Nogrowth 0-5 Nogrowth 20 Nogrowth 3-0 to flower)

418 BIO-CHEMICAL JOURNAL

SOLUTIONS OF PURINE DERIVATIVES AND OTHER ORGANIC COMPOUNDS

25 Days 48 Days 68 Days 90 Days
— *4# ——-——— a es
Solution Strength Roots Leaves Roots Leaves Roots Leaves Roots Leaves

ems cms ems. ems. cms ems, ems. ems

Control (1) ... 3-0 3-5 4-0 5-0 4-5 11-5 6-0 28-0

(Water) (2) ... 2-5 3-5 3-0 6-0 3-5 14:5 4-0 30-0

Control (GD) See 6-0 3:0 8-0 5:0 9-0 14-0 11:0 35-0
(Alkali) (2) ... 5-0 3:5 55 7-0 7-5 15-0 10-0 35:0 (B.)

Sodium Huminate 05% Notgrowing 3-5 Notgrowing 40 Notgrowing 5:0 Notgrowing 5-0
01% 50 2-5 12-0 3-5 16-0 9-0 18-0 30:0 (F.)
0:02 % 7-0 3-5 13-0 4-5 14-0 11-0 14-0 30:0 (F.)
0:02 % 5-0 4-0 8-0 5:5 8-5 13-0 9-0 31:0 (F.)

Sodium Nucleate 05% 1-5 3°5 2-0 4-0 2°5 5-0 2:5 7:0
O1% 1-0 3-0 1:5 4-0 2-0 6-0 2-5 27:0 (F.)
0:02 % 1-5 2-5 2-0 3-5 2-5 6-5 2:5 25:0 (F.)

Guanine 0:07 % Not growing 1:5 Notgrowing 1:5 Notgrowing 2-5 Notgrowing 7-0
Hydrochloride 0:07 % Not growing 2-5 Notgrowing 3-5 Notgrowing 4:0 Notgrowing 60 (B.)
Xanthine »- 003% 15 2-0 3-5 3-0 9-0 9-0 11-0 13:0 (F.)
0:03 % 1-0 2-0 1:5 3-0 6-0 6-0 10-0 22:0 (F.)

Hypoxanthine 0:07 % Not growing 1-5 Notgrowing 2-0 Notgrowing 2-5 Notgrowing 3-0

Hydrochloride 0:07 % Not growing 1:5 Notgrowing 2-0 Notgrowing 2-5 Notgrowing 3-0

Sodium Malate... 05% Notgrowing 2-0 Not growing 2:5 Notgrowing 3-0 Notgrowing 3-0
01% 2-0 1-0 2-5 1:0 3-0 1-0 4:5 (B.)
0-02 % 7:0 2-0 9-0 3-5 13-0 10-0 15-0 30:0 (F.)
0-02 % 7-0 4-0 8-0 6-0 8-5 16-0 9-0 33:0 (F.)

Caffeine ... 05% Notgrowing 1:0 Notgrowing 1-5 Notgrowing 2-0 Notgrowing 2-0

01% Notgrowing 1:0 Notgrowing 2-0 Notgrowing 3-0 Notgrowing 6-0

0:02 % Not growing 1:5 Not growing 3-0 Notgrowing 4:0 Not growing 11-5
0:02 % 1-0 2-0 4-0 4:5 9-5 11-0 10-5 32:0 (B.)

Sodium Urate 05% Notgrowing 2-5 Norgrowing 3-5 Notgrowing 4:5 Notgrowing 4-0

01% Not growing 3-0 Notgrowing 40 Notgrowing 7-0 Notgrowing 8-0

0:02 % 7-0 3°5 9-0 6-0 11:0 16-0 11-0 30-0

0-02 % 6-0 3°5 9-0 6:0 11-0 14:5 13-0 27-5

Sodium Oxalate... 05% Notgrowing 3:0 Notgrowing 3:5 Notgrowing 4-0 Notgrowing 5-0

01% 1-0 3-5 4-0 3-5 4-0 6-5 5-5 16-0

0-02 % 4-0 3-5 5-0 6-0 6-5 16-0 8-0 31-0

0-02 % 8-0 4-0 11-5 6-5 11:5 16-0 13-0 33-0

Sodium Oleate ... 05% Not growing 3.0 Notgrowing 3-5 Notgrowing 4-0 Notgrowing 4-5

01% 6-0 3°5 7:0 4:5 8-5 9-5 9-0 12-5

0:02 % 4-5 4:0 7-0 7-0 8-5 19-5 9-0 32-6

0:02 % 2-0 4-0 3-0 5-0 3-5 9-0 4-0 12-5

Sodium Linoleate 05% Not growing 3-0 Notgrowing 3-0 Notgrowing 5-0 Notgrowing 5-5

O11% 3:0 3:5 6-0 6-0 7:0 180 80 270

0:02 % 2-0 2-5 2:5 3:5 2-5 4-0 3-0 4:5

0:02 % 3-0 2-5 7-0 4:5 9-0 12-5 10-0 34:0

Bovril 01% 4-0 2-5 6-0 4:5 9-0 16-5 9-5 30-0
01% 3-0 3-0 6-0 7-0 7-5 22-5 9-0 35:0 (B.)

F. = Flowering. B, = Budding—i.e., floret just making its appearance.

THE EFFECTS OF PURINE DERIVATIVES 419

HistTo.Loeicat INVESTIGATION OF THE GROWING CELLS AND DIVIDING
NUCLEI UNDER THE INFLUENCE OF THE ABOVE COMPOUNDS

A few of the ends of the growing rootlets were cut off at various
times for the purpose of studying the effects of the solutions on cell-
division and nuclear changes. ‘These ends were immediately fixed in
Flemming’s solution, and sections cut and stained for microscopical
examination, of which the following are brief notes : —

Sections of Rootlets cut after Seventeen Days’ Growth

Control (Water).—Normal condition but not much cell-division.

Control (Alkali).—Frequent cell-division; much more than in the
water control. Cell-division figures very regular.

Sodium Huminate (O'1 per cent. Solution).—Cell-division much darker.
Divisions less frequent than in the alkaline control and irregular in
form. The nuclei often appear in dark rounded masses with clear
spaces surrounding them.

Sodium Huminate (0°02 per cent. Solution).—Cell-division more frequent.
Chromosomes of irregular length and in beaded rows; different from
the continuous rods of the normal preparation.

Sodium Nucleate (0°53 per cent. Solution).—Similar in appearance to the
0°1 per cent. sodium huminate, apparently indicating the presence
of more chromatin. A few cell-divisions which are very irregular,
showing the tendency of the chromosomes to shorten in length.

Sodium Nucleate (O'1 per cent. Solution).—Same dark rounded nuclei as
in the two previous numbers. Only a few cell-divisions, irregular

in form.

Sections of Rootlets cut after Twenty-seven Days’ Growth

Controls.—Large amount of cell-division.

Allantoin (O'5 per cent. Solution).—Fair number of dividing cells;
divisions regular.

Allantoin (0°3 per cent. Solution).—Very small amount of cell-division.

Allantoin (0°2 per cent. Solution).— 5 ff bs

Allantoin (O°1 per cent. Solution.)— Ps * ¥

Sections of Rootlets cut after Thirty-three Days’ Growth
Control (Water).—Not much cell-division occurring. Nuclei elongated

in form.

420 BIO-CHEMICAL JOURNAL

Control (Alkali).—Amount of cell-division considerably greater than in
the previous control. Divisions quite regular.

Sodium Huminate (0°5 per cent. Solution).—Cell-division practically
absent. ‘

Guanine Hydrochloride (0°07 per cent, Solution).—Degenerating.

Hypozxanthine Hydrochloride (0°07 per cent. Solution).—Degenerating.

Sodium Malate (0°5 per cent. Solution).—Degenerating.

Caffeine (0°02 per cent. Solution).—A fair amount of cell-division taking
place; divisions irregular.

Sodium Urate (OL per cent. Solution).—Most of the roots degenerating.

Sodium Ovalate (O'1 per cent. Solution).—A fair amount of cell-
division; divisions remarkably regular.

Sodium Oleate (0°5 per cent. Solution).—Most of the roots degenerating.

Sodium Nucleate (0°02 per cent. Solution).—Large numbers of nucleoli
present. Hardly any dividing cells; such as they are, irregular, and
majority of nuclei in rounded, resting form.

Xanthine (0°03 per cent. Solution).—Much the same as the previous
number, with perhaps a few more cells in dividing condition.

Sodium Malate (O'1 per cent. Solution).—Only a small number of cell-
divisions, regular in form.

Sodium Malate (0°02 per cent. Solution).—A fair number of dividing cells,
irregular in form.

Sodium Linoleate (OL per cent. Solution).—A fair amount of cell-division,
regular in form.

Sodium Linoleate (0°02 per cent. Solution).—Large numbers of dividing
cells, very irregular and uneven in form, sometimes nearly all the
chromatin passing to one of the poles. Chromosomes of very unequal
length.

Sodium Oleate (O'1 per cent. Solution).—Considerable amount of cell-
division; divisions regular. j

Sodium Oleate (0°02 per cent. Solution).—Less cell-division than in the
previous number; divisions regular.

Sodium Oxalate (0°02 per cent, Solution).
divisions, beautifully regular in form.

Large number of cell-

Sodium Urate (0°02 per cent. Solution).—Divisions less frequent and
somewhat irregular. Nucleoli very evident.
The shdes of the 0°1 per cent. and 0°02 per cent. solutions of sodium
oxalate were of particular interest on account of the remarkable regularity
of the cell-divisions.

THE EFFECTS OF PURINE DERIVATIVES 421
CONCLUSIONS

1. The growth of the bulbs is stimulated by dilute solutions of
sodium huminate, sodium malate, sodium urate, and sodium oxalate,
and to a less extent by Bovril. The increase in growth is more marked
in the rootlets than in the leaves.

In most cases 0°02 per cent. appears to be the most favourable
strength; the growth is considerably less in stronger solutions, and is
completely arrested by a 0°56 per cent. solution.

2. Cell-division is stimulated to a considerable extent by dilute
solutions of sodium huminate, sodium oxalate, sodium oleate, and sodium
linoleate, and to a less extent by sodium malate and caffeine.

3. Both growth and cell-division are inhibited by stronger solutions
of the above compounds, and by all solutions of allantoin, guanine
hydrochloride, xanthine, hypoxanthine hydrochloride, and sodium
nucleate.

This work was undertaken at the suggestion of Professor Benjamin
Moore, to whom, as well as to Dr. C. J. Macalister, I am greatly indebted
for advice and assistance on many points in connection with it.

422

ON GLYCOLYSIS IN BLOOD
By G. SPENCER MELVIN, M.D., Senior Assistant in Physiology.

From the Physiological Department, University of Aberdeen

(Received July 26th, 1912)

Despite the work which has been done on glycolysis in blood, there
still exists a great deal of confusion on the subject. It would seem to be
accepted at the present day, however, that after removal of blood from
the body its sugar undergoes destruction, or, at all events, tends to
disappear.

In 1855, Pavy' stated that from blood removed from the body sugar
disappeared during the process of clotting. This disappearance he
considered to be the result of ordinary decomposition.2 Schenck,
experimenting with blood to which known quantities of sugar had been
added, was unable completely to recover the total amount known to be
present. The deficiency he attributed to a combination formed between
the sugar and the blood protein. Doubts were cast by Rohmann,4
Salkowski,®° and Seegen® on the accuracy of these assertions, and eventually
Schenck’ admitted that his previous results had been based on unreliable
methods. Pavy? had determined that the passage through the blood of
such gases as oxygen, carbon dioxide, hydrogen, had no specific effect
quantitatively on the loss of sugar. Seegen,® however, found that the loss
was markedly increased by the passage of air through blood kept at a
temperature of 39° C. He ascribed this loss to ferment activity, but was
unable to detect any products of such action. Lépine? and _ his
collaborators endeavoured to establish that the deficiency or absence of a
glycolytic enzyme was responsible for the development of diabetes
following abiation of the pancreas. These workers considered?!® that the
ferment was found mostly in relation to the corpuscles, and that it was
removed by defibrination.1! In support of this, Sieber! and von
Schroeders!? were able to extract a glycolytic agent from fibrin. Lépine!4
noted in favour of ferment action, that the destruction of sugar in blood
was more marked the higher the temperature at which the blood was kept,
up to a point (54° C.), at which the loss ceased. Similar results were
obtained by Arthus!® who, however, considered that the ferment was
formed post-mortem, and that it was closely associated with thrombin.
These latter views were held also by Colenbrander!® and Rywosch.}?

ON GLYCOLYSIS IN BLOOD 423

Arthus held that the glycolytic enzyme was derived from the leucocytes
and platelets. Lépine,!® on the other hand, claimed to have established
that the precursor of the ferment was a sugar-producing enzyme, which
could be activated artificially by treatment with acid. This extraordinary
suggestion was disproved by the work of Paderi,!9 and Nasse and Framm.?°
As to the products of this ferment action, different results have been
obtained—by Kraus?! who found carbon dioxide; by Slosse** who found
lactic and formic acids; and by Seegen® who found neither carbon
dioxide nor lactic acid. According to Edelmann,?? glycolysis in
blood kept at 37° C. is slight during the first two hours, but well
marked after six hours. It is during the first two hours, however, that
glycolysis due to enzyme action would be most marked; and, moreover,
it is very probable that blood kept at 37° C. would be no longer sterile
at the end of six hours. —

The investigations here published were carried out to determine (1) if
blood, kept at various temperatures after removal from the body, suffers
any loss of its sugar; (2) if there is any destruction of sugar added to the
blood immediately after withdrawal from the body; (3) if any glycolytic
agent is removed from blood along with the protein or the fibrin.

Throughout the research aseptic precautions were rigidly observed.
The blood utilised was derived mostly from the sheep and ox, and in
certain cases from the horse and cat.

(1.) As a rule, the blood was defibrinated immediately after its
removal from the body, and a definite quantity at once run into excess of
alcohol. This portion, along with the remainder of the blood, was
forthwith conveyed to the laboratory, some ten minutes’ walking distance
away. The examination of the first portion was immediately proceeded
with, while the rest of the blood was kept under various conditions to be
examined subsequently.

For the isolation of the sugar the method adopted was substantially
that employed by Bang, Lyttkens and Sandgren.24_ The protein
precipitated by the alcohol was filtered off, repeatedly washed with hot
alcohol, and re-filtered. The filtrate and filtered washings were then
aggregated and evaporated to dryness. The resultant residue was extracted
with boiling water, precipitated by liq. ferri dialysat. (B.P., 1885) and
sodium chloride (saturated solution), and filtered. This filtrate, made up
to a definite point with distilled water, was then utilised for the estimation
of the sugar. For this, one of three methods was employed—(1) Bang’s,
(2) Bertrand’s, (3) gravimetric. In certain experiments the results appear

424, BIO-CHEMICAL JOURNAL

to show that there is a slight increase in the amount of sugar when the
blood is retained outside the body. In these, Bang’s method was
employed, and it was noticed that the end point was not definitely precise ;
so that slight variations in the amount of hydroxylamine solution required,
representing appreciable differences in the quantities of sugar therefrom
calculated, were obtained even in the same sample. These differences
are naturally magnified when the calculation is carried to parts per
hundred. The blood which had been retained, and from which the later
estimations were to be made, was kept at 37° C., at 15° to 16° C., or at
0° ©. In this series of experiments, also, the effect of passing certain

gases through the blood was investigated.

No 1. Blood from sheep. Kept at 16°C. Estimated by Bang’s method.
Estimated immediately 0-04875 gram sugar per 100 c.c.
ar 2 hours later 0-05
ie 54 hours later 0:04875
No. 2. Ox blood. Kept at 15°C. Estimated by Bang’s method.
Estimated immediately 0-042 gram sugar per 100 c.c.
- 44 hours later 0-047
rs 84 hours later 0-048
No. 3. Blood from ox. Kept at 15°C. Estimated by Bang’s method.
Estimated immediately 0-048 gram sugar per 100 c.c.
a 4 hour later 0-048
- 14 hours later 0-05
No. 4. Ox blood. Kept at 37°C. Estimated by Bertrand’s method.
Estimated immediately 0-031 gram sugar per 100 c.c.
55 4 hour later 0-031
3 2 hours later 0-031
No. 5. Blood from ox. Kept at 37°C. Estimated by Bang’s method.
Estimated immediately 0-053 gram sugar per 100 c.c.
5 4 hour later 0-052
No. 6. Blood from ox. Kept at 37°C. Estimated by Bang’s method.
Estimated immediately 0-031 gram sugar per 100 c.c.
eS 1 hour later 0-033
3 hours later 0-033
No. 7. Blood from sheep. Kept on ice in ice chest at 0° C. Bertrand’s method.
Estimated immediately 0-023 gram sugar per 100 c.c.
a 4 hour later 0-023
3 hours later 0-023
as 54 hours later 0-023
No. 8. Blood from sheep. Kept at 0°C. Bertand’s method.

According to the foregoing results, there is no loss of sugar in blood
kept at various temperatures for periods ranging up to 8} hours.

Estimated immediately
A 4 hour later
es 3 hours later
et 4 hours later

0-028 gram sugar per 100 c.c.
0-028
0-028
0-028

ON GLYCOLYSIS IN BLOOD 425

In the following set of experiments the blood was defibrinated and
a quantity of pure dextrose in solution added. The mixture was then
thoroughly shaken, and 25 c.c. of it removed for immediate analysis.
Through the remainder of the sample was passed a stream of air or of
carbon dioxide, the mixture being kept meanwhile at 37° C. or at 16° C.

No. 9. Sheep blood. Kept at 16°C. Stream of purified CO, passed through. Bertrand’s

method.
Estimated immediately ... 0-04 gram dextrose per 25 c.c.
P 4 hour later ... 0-04
No. 10. Sheep blood. Kept at 37°C. CO, passed through. Gravimetric method.
Estimated immediately _... 0-029 gram dextrose per 25 c.c.
5 1 hour later ... 0-026
- 2 hours later ... 0-029
No. 11. As in No. 10.
Estimated immediately ... 0-022 gram dextrose per 25 c.c,
= l hour later ... 0-025
Bs 2 hours later ... 0-022
No. 12. Sheep blood. Kept at 16°C. Current of air aspirated. Bertrand’s method.
Estimated immediately ... 0-016 gram dextrose per 25 c.c.
i> 4 hour later ... 0-016
33 2 hours later ... 0-016
No. 13. Sheep blood. Kept at 37° C. Current of air aspirated. Gravimetric method.
Estimated immediately ... 0-025 gram dextrose per 25 c.c.
3 1 hour later ee Oe
3 2 hours later ... 0-026

The foregoing experiments are selected as representative of the results
obtained. It would thus appear that the sugar is in no way altered
quantitatively by the passage of a gas through the blood.

(2.) In this set of experiments a solution of pure dextrose was added
to the blood immediately after defibrination, and 25 c.c. of the mixture
examined at once, the remainder, as usual, being retained. All the
portions examined were thus drawn from the same stock, so that any
deficiency of sugar in the later specimens could be established
independently of the absolute amount of dextrose known to be present.
A depreciation of the sugar content thus obtained must be due to the
destruction of the sugar since the experimental procedure was the same

in all cases.

No. 14. Sheep blood. Kept at 16°C. Bang’s method.

Estimated immediately _... 0-226 gram sugar per 25 c.c.
* 2 hours later ... 0-221
No. 15. Sheep blood. Kept at 16°C. Bang’s method.
Estimated immediately ... 0-157 gram sugar per 25 c.c.
BS 2 hours later... 0-156
No 16. Sheep blood. Kept at 16°C. Bang’s method,
Estimated immediately ... 0-04 gram sugar per 28 c.c.
3 2 hours later ... 0-036

e 3 hours later ... 0-038

426 BIO-CHEMICAL JOURNAL
No. 17. Ox blood. Kept at 16°C. Gravimetric method.
Estimated immediately ... 0-038 gram sugar per 25 c.c.
es 3 hours later ... 0-036
No. 18. Sheep blood. Kept at 15°C. Gravimetric method.
Estimated immediately ... 0-0355 gram sugar per 25 c.c.
ie 2 hours later ... 0-0355
35 3 hours later ... 0-0351
No. 19. Ox blood. Kept at 37°C. Bertrand’s method.
Estimated immediately _..._ 0-089 gram sugar per 25 c.c.
ss 3 hours later ... 0-088
No, 20. Sheep blood. Kept at 37°C. Bertrand’s method.
Estimated immediately ... 0-02 gram sugar per 25 c.c.
gs 24+ hours later ... 0-0195
No. 21. Ox blood. Kept at 37°C. Bertrand’s method.
Estimated immediately ... 0-136 gram sugar per 25 c.c.
6 4 hour later ee Oslo
x 24 hours later ... 0-137
No. 22. Ox blood. Kept at 37°C. Bertrand’s method.
Estimated immediately ... 0-036 gram sugar per 25 c.c.
33 4 hour later ... 0-032

It is evident that the mixture of blood and dextrose solution has
suffered no loss of sugar on keeping.

(3.) To determine the possibility of a glycolytic agent removed
with the fibrin.

For this purpose experiments were carried out as follows :—fresh
fibrin was broken up and thoroughly extracted on a mechanical shaker
for three hours with 0°75 per cent. solution of sodium chloride. The
fibrin was then removed and to the extract was added a quantity of
pure dextrose in solution. The mixture was next made up to 100 c.c.,
and of this, 25 c.c. were immediately estimated. The remaining
75 c.c. were placed in an incubator at 37° C. for three hours, at the end
of which time 25 c.c. were estimated. In all estimations, the
gravimetric method was employed.

Dextrose gram per 25 c.c.

Estimated immediately After 3 hours at 37° C.
No. 23 5 “op ae 0-09 088
No. 24 506 500 506 0:059 0-058
No. 25 it ie a 0-03 0-029
No. 26 ot ae oe 0:036 0-036

In Experiment No. 27 a different procedure was adopted. Fresh
fibrin was placed in alcohol, and retained there for three weeks. It was
then removed, and washed for twenty-four hours with boiled distilled
water in a sterile jar. These washings were rejected, and to the fibrin
was added boiled distilled water. This extraction was allowed to proceed
for four days. At the end of that time the extracting water was removed

ON GLYCOLYSIS IN BLOOD 427

and replaced by another quantity of distilled water, which was allowed to
extract for twenty-four hours (i.e., a fifth day). To both these extracts
10 e.c. of a 1 per cent. dextrose solution were added and the volumes
of the mixtures made up to a certain point. They were then kept at
37° C. for three and a half hours in the first case, and for four and a
half hours in the second. The quantity of dextrose was then estimated
in each portion.

Amount added . ... O-l gram dextrose
Portion A = 4 ‘days’ “extract of fibrin + dextrose.
Estimated after 34 hours at 37°C.... 0-096

”

Portion B = 5th day’s extract of fibrin + dextrose.
Estimated after 44 hours at 37°C.... 0-099

No. 28. Conducted on lines similar to those of No. 27.
Amount added . .. 0-1 gram dextrose

2 ‘days’ “extract of fibrin 45 “dextrose.
Estimated after 4 hours at 37° C. 0-

”

”

It thus appears that no ferment or glycolytic agent can be extracted
from fibrin by such washing.

Two experiments are selected to show the results obtained by
testing the glycolytic power of extracts of the protein precipitated from
blood by alcohol. The blood was run directly into alcohol and retained
there for a time sufficient to effect coagulation of the protein. The
coagulum was filtered off, dried at a low temperature (37° C.), and then
extracted with distilled water for several hours. The extract was then
filtered and to the filtrate was added a definite amount of dextrose in
solution. The volume of the mixture was then ascertained, and aliquot
portions of it were analysed (1) immediately, (2) after being kept at
37° C. for three hours.

No. 29. Amount actually present es ee ... 0-105 gram dextrose
5 after 3 hours at 37° C. wets aoe .» 0-105 3

No. 30. Amount actually present fc oes nee ... 0-047 gram dextrose
- after 3 hours at 37° C. Ber nee on, 0047

”

The above method of experimentation postulates that any enzyme
will conform to the accepted characters of enzymes, to the extent that
alcohol will determine precipitation. The same method is applied in
the preparation of thrombin which, presumably, has a_ disposition
similar to that of a possible glycolytic ferment. In none of the
experiments done in this line of investigation could there be found any
evidence of the working of a glycolytic agency.

428 BIO-CHEMICAL JOURNAL

From the foregoing results the following conclusions are warranted.

(1.) Blood kept outside the body, at ordinary or body temperature,
suffers no loss of its sugar—proper precautions as regards asepsis being
taken. It is to be noted that the first estimation has been made on blood
in a condition as nearly that of the blood within the body, as is possible.

(2.) If dextrose be added to blood as soon as possible after its
removal from the body, and the sugar in the mixture be estimated
‘immediately, and at intervals later, no loss of sugar can be made out.

(3.) Extracts of fibrin possess no glycolytic power even after
prolonged extraction.

(4.) Blood protein when coagulated by alcohol yields no glycolytic
agent on extraction. .

(5.) No evidence has been obtained that the sugar is altered by the
passage of a gas through the blood.

REFERENCES

Pavy, Proc. Roy. Soc., VII, p. 371.

Pavy, Proc. Roy. Soc., XXVIII, p. 520.

Schenck, Pflug. Archiv, XLVI, p. 607.

Rohmann, Centralb. ¢. Physiol., IV, p. 12.

Salkowski, Centralb. f. d. med. Wissensch., No. 17, 1890.
Seegen, Centralb. f. Physiol., IV, p. 217.

Schenck. Pflzig. Archiv, XLVII, p. 621.

Seegen, Centralb. f. Physiol., V, p. 821.

Lépine, Compt. Rend., CX, p. 742.

10. Lépine, Compt. Rend., CXII, p. 411.

11. Lépine, Compt. Rend., CXII, p. 1181.

12. Sieber, Zeitschr. f. physiol. Chem., XX XIX, p. 484; and XLIV, p. 560.
13. von Schroeders, Inaug. dissert., Berlin, May, 1904.

14. Lépine, Compt. Rend., CXII, p. 146.

15. Arthus, Archiv. de Physiol., III, p. 425, and IV, p. 337; also Compt. Rend., CXIV, p. 605.
16. Colenbrander, Maly’s Jahres-Bericht, p. 137, 1892.

17. Rywosch, Centralb. f. Physiol., XI, p. 495.

18. Lépine, Compt. Rend., CXX, p. 139.

19. Paderi, Maly’s Jahres-Bericht, p. 211, 1896.

20. Nasse and Framm, Archiv. f. Physiol., LXIII, p. 203.
21. Kraus, Zeitschr. f. klin. Med., XXI, p. 315.

22. Slosse, Archiv. Internat. de Physiol., XI, p. 154.

23. Edelmann, Biochemische Zeits., XL, p. 314, April, 1912.
24. Bang, &c., Zeitschr. f. physiol. Chem., LXV, p. 497.

paortanpPp word —

~

429

THE REDUCTION OF FERRIC CHLORIDE BY SURVIVING
ORGANS

By DAVID FRASER HARRIS, M.D., D.Sc., Professor of Physiology, anv
HENRY JERMAIN MAUDE CREIGHTON, M.A., M.Sc., D.Sc.,
Lecturer on Physical Chemistry in Dalhousie University, Halifaz,
Nova Scotia.

(Received August 28th, 1912)

INTRODUCTION

In 1906 one of us showed! that on perfusing the surviving kidney
with the gelatine and Prussian blue injection-mass, it was possible to obtain
artificial urine in which the pigment was completely reduced to the leuco
condition.

We were anxious to repeat this method with the kidney, and to employ
it in the case of the liver in order to obtain, if possible, artificial bile, in
which some reduced material might similarly be expected. In the first
series of experiments we used the surviving liver of the cat just killed by
bleeding; and we employed a solution of ferric chloride for two reasons :—
(1) Its viscosity is so much less than the gelatine mixture; and (2) Because
we had previously shown? that the reductase contained in the press-juice of
liver reduced ferric chloride to the ferrous condition.

EXPERIMENTAL

(a) Cat’s Liver

The technique of one experiment with the liver will suffice to show the
method. A fine glass cannula filled with the solution of ferric chloride
(0°25 per cent.) was inserted into the portal vein the instant the animal was
dead. <A similar cannula was tied into the gall-bladder, which contained
a little viscous yellow bile. The opened abdomen was kept warm with
pieces of cotton wool wrung out of hot water. From a burette, connected
with the cannula by a rubber tube, a quantity of ferric chloride solution was
allowed to run into the liver, which, in a few minutes, became fairly turgid.
Discoloured blood in due time began to flow from the right external jugular
vein, and was collected in a small glass flask. After the ferric chloride

1. Harris, D. Fraser, Bio-Chemical Journal, I, 355, 1906.
2. Harris, D. F., and Creighton, H. J. M., Proc. Roy. Soc., 1912.

430 BIO-CHEMICAL JOURNAL

solution had displaced nearly all the blood from the liver, and had remained
in contact with the liver tissue for twenty minutes, the liquid was squeezed
out of the gall-bladder into the cannula which was removed from the organ,
and the entire liver was rapidly excised while still warm. The liver was
then cut to pieces in a glass funnel, through which a sanguineous, turbid
fluid dripped. While the animal was being bled to death, some of its
blood was collected and defibrinated, to be used in the control observations,
as detailed below.

’ The following liquids were examined :—

(1) Drippings from the cut liver (coloured red by blood), amounting,
with washings, to 18 c.c.

(2) 2 c.c. of bile from the gall-bladder.

(3) 20 c.c of blood (considerably diluted with iron solution) from the
jugular vein.

Liver Drippings.—Before testing for iron, it was necessary to remove
the haemoglobin from the solution, so as to get rid of the red colour.
It was not practicable to precipitate the haemoglobin by heating, owing
to the fact that, at the temperature of precipitation, ferric chloride is readily
reduced by blood to the ferrous condition. In consequence of this, the
haemoglobin was precipitated at room-temperature with dilute hydrochloric
acid, which also served the purpose of breaking up any iron-protein complexes
that might be present.! It was found when hydrochloric acid was slowly
added to a solution containing blood, that, at first, the dark brown, soluble
acid-haematin was formed, and that further addition of acid was necessary
to precipitate the haemoglobin and blood-proteins.

To the 18c.c. of liver drippings, 2c.c. of dilute hydrochloric acid were
added, and the precipitated haemoglobin and blood-proteins separated by
filtration. The colourless filtrate was divided into several parts and tested
for ferrous and ferric iron. With potassium ferricyanide the colourless
filtrate instantaneously turned to a deep green-blue colour and, after a few
minutes, a blue precipitate was noticeable. On the other hand, potassium
ferrocyanide gave only a very faint blue coloration on being added to the
filtrate.

As a control, the following experiment was carried out. To 10 c.c. of
the defibrinated cat’s blood, 1 c.c. of 0°25 per cent. ferric chloride solution was
added. The mixture was then placed in a thermostat and kept at 38° for twenty
minutes, after which the haemoglobin was precipitated with dilute hydrochloric
acid, and removed by filtration. The colourless filtrate gave a deep blue

1. Creighton, H. J. M., Trans. N. S. Inst. Sci., XIII, 61-75, 1912.

THE REDUCTION OF FERRIC CHLORIDE 431

precipitate with potassium ferrocyanide, but with potassium ferricyanide
the colour was yellow, without the slighest trace of blue or green. At the
end of three-quarters of an hour the colour was still unchanged, but at the
end of one and a half hours it had become slightly green, while after three
hours the mixture was distinctly green and a slight precipitate had formed.
This control experiment was repeated with the same results.

No ferrous iron was found to be present in the ferric chloride used in
these experiments.

Bile.—The 2c.c of bright yellow, viscous bile, to which a few drops
of hydrochloric acid had been added, were diluted to 5c.c with distilled
water, and the solution divided into two parts. These were tested with
potassium ferrocyanide and potassium ferricyanide; iron in both the ferric
and the ferrous condition was found to be present.

Blood from Jugular Vein.—The haemoglobin was precipitated with
dilute hydrochloric acid, and, after separation by filtration, potassium
ferrocyanide and potassium ferricyanide were added to different portions
of the clear filtrate. Both ferric and ferrous iron were found present.

(b) Lamb’s Kidney

The technique for the kidney was briefly as follows:—The kidney was
taken from a lamb just killed by bleeding was brought at once from the
slaughter-house to the laboratory in the water-jacket of a calorimeter, which
was filled with water at 42°. Within three-quarters of an hour of death,
we fixed cannulae in the renal artery, the renal vein and the ureter. We
perfused ferric chloride solution (0°25 per cent.) with a considerable head of
pressure, and obtained a little sanguineous liquid from the renal vein and a
large amount of a perfectly clear, pale yellow liquid from the ureter.

The following liquids were examined : —

(1) Blood from the Renal Vein.
(2) Artificial Urine from the Ureter.

Blood from Renal Vein.—The blood collected, which amounted to
about 5-6 c.c., was treated with dilute hydrochloric acid, and the precipitated
haemoglobin and blood-proteins filtered off. Portions of the clear filtrate
were then tested with potassium ferrocyanide and potassium ferricyanide.
The former reagent gave a deep blue precipitate, and the latter a pale green
coloration, thus showing that some of the ferric chloride had undergone
reduction.

432 BIO-CHEMICAL JOURNAL

Artificial Urine.—The artificial urine, which was somewhat lighter in
colour than the ferric chloride solution used for perfusing the kidney, was
acidified with hydrochloric acid and divided into two parts. On adding
potassium ferrocyanide to one part, a deep blue coloration was obtained ;
while the other part gave, with potassium ferricyanide, a light green
coloration, perceptibly deeper than in the case of the venous blood.

CONCLUSIONS

1. The foregoing experiments prove that ferric chloride is reduced to
the ferrous condition, both by the surviving liver and the surviving kidney.
In the light of our previous work! on the reductase of the press-juice of these
organs, we believe that reductase is active in these intact surviving organs.

2. The fluid which runs through the blood-vessels of the organs is less
reduced than that which is excreted into the gall-bladder and ureter,
respectively. This is precisely what, a priort, we would expect, since the
liquid flowing from the emergent vein has not come into contact with the
living cells of the parenchyma, whereas the fluid found in the gall-bladder
and the ureter has passed through the living cells of the hepatic tissue and
renal tissue, respectively. This latter fluid has had greater opportunities of
being brought within the sphere of action of the reductase.

3. The degree of reduction by the living cells is more perfect in the
liver (of cat) than in the kidney (of lamb). This may be partly explained
by an absolute difference in the intensities of the reduction in these two
classes of organs; partly, possibly, by the fact that the liver was that of a
carnivore, the kidney that of a herbivore; and partly by the fact, in the
case of the artificial urine, that the perfusing liquid passed into the ureter
very quickly, in consequence of which it was only for a very short time
within the sphere of action of the reductase. From our previous work on
the reductase of liver and kidney,! we are inclined to think that the
supposition as regards the difference in reduction intensity is, in itself,
correct.

4. The reductions studied are quite distinct from reductions in the
presence of non-lving organic matter.

i. oe. cit.

433

THE ACTION OF THE OPIUM ALKALOIDS
By F. W. WATKYN-THOMAS, B.A., Trinity College.
From the Pharmacological Laboratory, Cambridge
(Received September 9th, 1912)

Previous Work

Although very complete accounts of the action of morphine have
been given by many observers, such as Bernard and Guinard, it is only
recently that the action of morphine has been compared with that of the
total alkaloids of opium. The action of individual alkaloids, especially
narcotine and codeine, has been studied, but the unsuitability of opium
for intravenous injection has prevented accurate work on the action of
the total alkaloidal content.

In 1909, Sahlit prepared a substance which he called ‘ Pantopon,’
and which contained all the alkaloids of opium in a definite concentration
and in soluble form. Since then a great deal of work has been done on
the action of this preparation. Loewy? has correlated the action of
morphine, codeine, and the total alkaloids on the respiratory centre.
Wertheimer and Raffalovitch® have investigated the action of the total
alkaloids on the circulatory system, Rodari,? Bergien,* and Dobeli,®
have compared the toxicity of morphine and the total alkaloids. An
account of its action is also given by Rodolico,® and Cohnheim has
investigated the action on the gut. The analgesic and anaesthetic
actions of the total alkaloids have been fully described by Grey’ and
Trotain,® among others.

EXPERIMENTAL Mrtruops

1. Animals. In most cases rabbits were used, as, according to
Guinard, in behaviour under morphine they resemble man more closely
than do any other animals except the dog. As controls, and in certain
experiments for which rabbits were unsuitable, cats were used (Experi-
ments X, XII, XIX). A few experiments were performed on frogs.
(Experiments XVIII, XXIII, XXVII). A certain number of experi-
ments were performed on pithed animals (Experiments II, IV, V, VI,

Ee, Sxl).

434 BIO-CHEMICAL JOURNAL

A.C.E. alone was used in one experiment (XII), A.C.E. and ether
in one. In all other experiments the anaesthetic was a 25 per cent.
solution of urethane in Ringer, 6 c.c. (15 grams) per kilo body weight
administered at body temperature. Half the dose was injected under the
skin, anaesthesia was induced by ether, and the other half was injected
into the peritoneal cavity.

The ether anaesthesia was usually maintained during the preparatory
operation, but no other drug was administered until the animal was under
the full influence of urethane. It was thus possible to commence all
experiments under an approximately standard depth of anaesthesia, as
indicated by dosage, blood pressure, size of pupil, and absence of
conjunctival reflex.

Tracheotomy was performed, and cannulae tied in the external
jugular vein on one side and the common carotid of the other side. Any
special operations afterwards performed are described in the account of
the experiment. In all experiments injections were given by the cannula
in the external jugular; blood pressures were always taken in the common
carotid.

Drugs. Tincture of opium is not suitable for experimental work, as
the alkaloid content is variable and the contained alcohol is a possible
source of error.®

Throughout these experiments a 2 per cent. solution of Sahli’s
‘Omnopon ’*—in which all the alkaloids freed from meconic acids, gums
and resins are present as hydrochlorides—was used.

The composition of this preparation is as follows :—Morphine 52 per
cent., narcotine 20 per cent., codeine 2 per cent., papaverine 2°5 per cent.,
thebaine 1 per cent., narceine 1°2 per cent., other alkaloids 4 per cent.,
water 8 per cent., hydrochloric acid 9 per cent.

Standard solutions of morphine hydrochloride, narcotine hydro-
chloride, and codeine hydrochloride were used. They were prepared on
the following basis : —

1 c.c. 2 per cent. omnopon contains 0°0166 gram total alkaloid,
0:0104 gram morphine, 0:0040 gram narcotine, 0:0004 gram codeine.

1 c.c. standard morphine solution contained 0°0104 gram morphine.

1 c.c. standard narcotine solution contained 0:0040 gram narcotine.

1 c.c. standard codeine solution contained 0:0004 gram codeine.

Circulatory effects. The effect of the opium alkaloids on the

* * Pantopon’ is known in this country as ‘Omnopon.’ I am indebted to Messrs, Hoffman,
La Roche, the manufacturers, for a supply of the substance,

THE ACTION OF THE OPIUM ALKALOIDS 435

circulation is insignificant. Nevertheless, as there are some pronounced
differences between the different bodies, it will be advisable to consider
them in detail.

Action on vessels. Perfusion experiments were carried out on the
frog, Ringer’s solution being injected through the peripheral end of the
aorta, and the outflow from the great veins estimated. These experiments
showed that morphine in doses of 2-4 mgm., thrown suddenly into the
circulation

a procedure which gives every advantage to the vessels to
show alterations in their calibre—produced practically no effect. If the
vessels had been previously acted on by some peripheral vaso constrictor
a certain degree of vaso-dilatation was produced. Thus the constriction
caused by narcotine could be entirely antagonised by a suitable dose of
morphine.

Protocol of typical experiment : —

Experiment XXII.—Frog B. Average time for outflow of 2 c.c. = 2 mins. 30 secs.
4 mg. Morphine perfused.
2 c.c. outflow in 2 mins. 25 secs.
2 c.c. outflow in 2 mins. 30 secs.
2 c.c. outflow in 2 mins. 35 secs.
Experiment XXVII.—Frog A. Average time for outflow of 2 c.c. = 59 secs.
0-4 mg. Narcotine perfused.
2 c.c. outflow in 2 mins. 1 sec.
2 c.c. outflow in 4 mins. 2 secs.
4 mg. Morphine perfused.
2 c.c. outflow in 1 min. 12 sees.
2 c.c. outflow in 1 min. 30 sees.
2 c.c. outflow in 1 min. 29 secs.

Narcotine in such small doses as 0°4 mgm. is shown by this last
experiment to produce considerable constriction. This constriction lasts
some minutes, and is followed by a very gradual relaxation.

The percentage of morphine in omnopon is, roughly, about the
optimum antagonising dose which will annul the constricting action of
the contained narcotine. One would anticipate from these experiments
that the effect of omnopon on the vessels should not be very pronounced.
This was found to be the case: omnopon producing little effect in either
direction.

Protocol of typical experiment : —

Experiment XXII.—Frog A. Average time for outflow of 4 c.c. = 1 min. 5 sees.
7 mg. Omnopon perfused.

4 c.c. outflow in 1 min. 9 secs.

4 c.c. outflow in 1 min. 6 secs,

4 c.c. outflow in ] min, 5 secs,

436 BIO-CHEMICAL JOURNAL

Action on the heart. None of these alkaloids produce any decided
action on the heart in the intact animal in dosage which may be regarded
as therapeutically possible. In the isolated mammalian heart, where
concentrated solutions can be suddenly thrown into the coronary
circulation, it is possible to make comparative experiments on their
toxicity. By this means it was shown that 2 mgm. morphine produced
little or no effect, whilst 5 mgm. caused decided slowing and diminished
systole lasting for many minutes, in spite of the fact that the morphine
passed straight through in less than one minute and was then replaced
by normal Ringer’s fluid.

Narcotine in doses up to 1 mgm. left the heart entirely uninfluenced.

Omnopon in doses up to 2 mgm. causes some increase in the systole,
but beyond this dose the systole weakens, and if a dose of 4 mgm. is
injected into the-coronary circulation weakening of systole and slowing
of rate are always pronounced, while the beat may be inhibited for
thirty to sixty seconds.

Measurements of the output of the mammalian heart in the intact
and anaesthetised animal show that none of these alkaloids, under such
conditions, exert any appreciable action on the heart, even in poisonous
doses.

It is true that a dose, sufficient to cause a fall of pressure by vaso-
dilatation, causes an increased output from the heart, but this is not the
direct effect of the alkaloids but is associated with the vaso-motor fall,
and may be produced equally well by suitable doses of some of the
nitrites.

Protocol of experiment :—

Experiment X.—Cat 2 kilos. Ether and urethane. Tracheotomy. Cannulae in external
jugular and common carotid. Artificial respiration, thorax opened and heart in

cardiometer.

1 c.c. Morphine ... ais ar 308 ... Slight fall of blood pressure. Some variation
in beat of heart.

1 c.c. Omnopon ... 580 See ee ... Fall of pressure. No effect on heart.

3 c.c. Narcotine ... au: es at .... Fall and rise to previous pressure. No effect
on heart.

3 c.c. Narcotine ... ar ict Bc ... Fall of pressure. No effect on heart.

3 c.c. Narcotine ... = stele sa ... Fall of pressure. Weakening of heart.

4 c.c. 0:25 % Nicotine ... ae ao ... No rise of pressure. Heart growing rapidly
weaker,

Death.

In none of my experiments has the slightest evidence been afforded
that any of these opium alkaloids, or all of them administered together
as omnopon, exert any direct action on the heart.

THE ACTION OF THE OPIUM ALKALOIDS 437

Blood pressure. The action of morphine on the blood pressure is
insignificant. Many writers claim that it induces a rise of pressure, and
it is true that small doses, injected directly into the circulation, may
cause an increase in pressure of a few millimetres of mercury. But I
have never obtained the smallest rise by subcutaneous injection. Larger
doses, such as 5 mgm. per kilo body weight, in rabbits undoubtedly
produce a small degree of vaso-dilatation and some corresponding fall in
blood pressure, but it is only by the administration of enormous doses,
10 mgm. per kilo, that a fall of pressure of any moment can be produced.
It is hardly necessary to point out that in all experiments in which such
bodies as these are injected efficient respiration is an essential for accurate
estimation of the effect produced.

Codeine has more action on blood pressure than morphine, but its
effect is still insignificant. It raises the height of the blood pressure curve
a shade more than morphine, and it is much easier with this alkaloid to
produce a marked fall in pressure associated with vaso-dilation. Codeine
stands midway in its action between morphine and narcotine.

Narcotine, in small doses, causes a distinct rise in arterial blood
pressure, but when the dosage is increased fall of pressure rapidly ensues.
The rise in pressure is associated with vaso-constriction, and it is easy
to show that the vessels of the gut constrict more or less in proportion to
the rise of pressure. In the same way the subsequent fall in pressure is
associated with splanchnic vaso-dilation. Now, as narcotine is repre-
sentative of a number of other alkaloids present in opium which in large
doses produce fall of pressure and in small doses initial constriction, it
will be well to examine the mechanism by which this type of effect is
produced.

A large dose of narcotine injected into the circulation immediately
produces a maximal fall in blood pressure. Further injections produce
no further fall, although it may readily be shown that the blood pressure
is capable of a further fall by the administration of a little nitrite. If,
after the pressure has undergone this maximal fall, which may oceur in
a rabbit after the injection of 15 mgm. per kilo body weight, the
splanchnic nerve be excited by a faradic current, no response—vaso-
constriction and rise of blood pressure—occurs. Nevertheless, such
response may be obtained by applying the electrodes to the post-
ganglionic fibres. The splanchnic nerve is paralysed and the block occurs
in the ganglion cells. This, then, is the explanation of the fall of blood
pressure caused by these alkaloids, The sympathetic ganglion cells in the

438 BIO-CHEMICAL JOURNAL

neck may be paralysed in a similar manner. Using dilatation of the

pupil as a test of sympathetic activity, it was found that narcotine

paralysed this nerve and that the block occurred in the ganglion cells,

the pupil reacting normally to stimulation of the post-ganglionic fibres.
Protocol of experiment : —

Experiment XIIT.—Rabbit 1-7 kilos. Ether and urethane. Tracheotomy. Cervical sympathetic
exposed. Electrodes between mid- and superior ganglia.

1. Stimulate. Coil at 20... se 506 BSE nic ... Pupil widely dilated.
2. 1c.c. Narcotine.

3. Stimulate ... ‘iis 08 300 spe 30 sAC ... Pupil widely dilated.
4. 1c.c. Narcotine.

5. Stimulate. Coil at 20... 400 a6 50 50 ... Pupil widely dilated.
6. 2 c¢.c. Narcotine.

7. Stimulate. Coil at 20... ais 500 Te ae ... Pupil dilated, but less.
8. 3 c.c. Narcotine.

9. Stimulate. Coilat 20... bor oes aS: oe ... Pupil unchanged.
10. Stimulate above Ganglion 50 soc 336 5 ... Pupil dilated.

OA TTT TT TT THOTT TTT

Fic. 1.—Cat, urethane. Intestinal movements recorded by balloon. Blood pressure below.
At A—1 c.c. 0-5 % Nicotine injected.
At B—2 c,c, 0-1 % Adrenalin,

THE ACTION OF THE OPIUM ALKALOIDS 439

In other words, narcotine and its allies should produce their rise, and
more pronounced fall in blood pressure, in the same manner as nicotine.
The first action of nicotine is to stimulate very powerfully the ganghon
cells of the sympathetic system, and, if these are paralysed by narcotine,
nicotine should produce no action, which was found to be the case; whilst
adrenalin, which is known to act on the end organ, produces its action
almost uninfluenced. Two or three points in this type of paralysis are
worthy of note. The first is that the paralysis may occur even whilst the
medulla is still active, as is shown by natural respiration. This is a
remarkable fact, and, so far as I am aware, every other paralysis of this

nature—such as that caused by nicotine or by apocodeine—is preceded by

complete medullary paralysis.
Protocol of experiment : —

Experiment XI—Rabbit 1-9 kilos. Ether and Urethane. Tracheotomy. Cannulae in Ext.
Jugular and Common Carotid.

2 c.c. Narcotine.

2 c.c. Narcotine.

1 c.c. Narcotine:

1 c.c. Narcotine.

5. 2c.c. Adrenalin ... Marked effect.

6. 2 c.c. Nicotine ... No effect. Feeble respiratory movements continuing.

Pee

Secondly, the nerve cells on the course of the splanchnic fibres
supplying the vessels of the gut are paralysed before those conveying
inhibitory fibres to intestine. Thus an injection of nicotine may fail
to cause any rise in blood pressure, and yet inhibit peristalsis (vide
protocol, Expt. XII).

Omnopon certainly raises blood pressure more than either morphine,
codeine, or narcotine, in corresponding doses. In the rabbit, doses of
over 15 mgm. per kilo body weight cause a marked fall of pressure with all
the characteristics of that described for narcotine, and in sufficient doses
will, like narcotine, entirely antagonise the pressor action of nicotine.
This is in contradiction to the results of Bergien who claims that blood
pressure is not materially affected by large doses of omnopon.

The principal fact learnt from these experiments is that the opium
alkaloids exert an initial slight stimulant action on the sympathetic cells,
this being followed by depression and paralysis. This is very marked
in the case of narcotine, much less for codeine, and least of all for
morphine.

Comparison of Respiratory Effects. A requirement in medicine is a
drug which, while diminishing cerebral reflexes like morphine, will do

440 BIO-CHEMICAL JOURNAL

so without marked depression of respiration. This is especially desirable
in the treatment of useless cough. Morphine, unfortunately, profoundly
depresses the respiratory centre, no doubt by preventing afferent impulses
reaching the medulla.

Codeine, heroine, dionine, and other morphine derivatives have been
employed as substitutes, and it is generally admitted that they are less
depressant to respiration than morphine, while it is stated that they are
‘efficient in relieving cough.

Loewy?, Wertheimer and Raffalovich® and Rodari? claim that the
total alkaloids have considerably less effect on the respiratory centre than
has morphine.

Loewy has shown that the reaction of the respiratory centre to carbon
dioxide is less affected by the total alkaloids than by morphine.

All these observers used ‘ omnopon.’

In the ensuing experiments the effect on the respiratory reflexes was
tested by stimulation of the cut central end of the great sciatic, or of the
external popliteal, with induction shocks and a tetanising current. The
respiratory movements were recorded by a lever attached to the severed
ensiform cartilage.

Protocols of typical Experiments :—

Experiment XVII.—Rabbit, 1-30 kilos. Ether and urethane. Tracheotomy. Cannulae in left
external jugular. Right common carotid. ‘ Diaphragm slip’ method of recording great

sciatic on electrodes.

i. Stimulate great sciatic. Coil at 20 ... (a) Great rise of blood pressure. Traube-
Hering curves.
(b\) Respiration greatly increased in force and

frequency.
ii. 0-1 c.c. Omnopon ae “ise ... (a) Slight rise of blood pressure.
(b) Respiration deepened, but not appreciably
slowed.
iii. 0-2 c.c. Omnopon ae aCe .... (a) Further rise of blood pressure.
(b) Respiration slightly slower.
iv. Stimulate great sciatic. Coil at 20... Respiratory reflex very feeble.
v. Stimulate great sciatic. Coil at 15 ... Vomiting movements. Traube-Hering waves.
Rise of blood pressure.
vi. 1lc.c. Morphine ... ae se ... Respiration slowed and abolished. Drum
stopped for artificial respiration.
vii. Stimulate great sciatic. Coil at 15 Vomiting movements followed by increased
frequency.
viii. 0-3 c.c. Narcotine aa ate ... Fall in blood pressure. Return to Normal.
Respiration more rapid.
ix. Stimulate great sciatic. Coil at 15 ... Response good.
x. 0-6 c.c. Narcotine see i ... Fall of blood pressure. Arrest of respiration,

Death,

THE ACTION OF THE OPIUM ALKALOIDS 441

Experiment XVIJI—Rabbit, 1-5 kilos, Anaesthetics and operative procedure, as No. XVII.

i. Stimulate great sciatic. Coil at 835... Rise of blood pressure, Respiration deeper
and more rapid.

ii. 0-2 c.e. Omnopon ve oa ... Rise of blood pressure. Respiration unaffected
; at first, and then more shallow.

iii. Stimulate great sciatic. Coil at 35... No effect on blood pressure or on circulation.

iv. 0-2 c.c. Omnopon ae eh ... Rise of blood pressure. Respiration unaffected.

v. Stimulate great sciatic. Coil at 25 ... Slight rise of blood pressure. No respiratory
response.

v. Wash out—wait three minutes.

vii. 0-2 c.c. Morphine eee re .... Very slight rise in blood pressure. Respiration
unaffected.

viii. Stimulate great sciatic. Coil at 35... Absolutely no response in blood pressure or
circulation,

ix. Stimulate great sciatic. Coil at 25 ... Absolutely no response in blood pressure or
circulation.

x. 0-2 c.c. Narcotine mee ane ... Fall of blood pressure. Rate of respiration
increased.

xi. Stimulate great sciatic. Coil at 25... Rise of blood pressure. Respiration unaffected.

It was found that morphine always slows respiration and depresses
the reflexes. Narcotine at first renders respiration deeper and slower,
and if the drug is pushed the movements become more rapid and more
shallow. But the most remarkable effect of narcotine on respiration is
its depression of the respiratory reflexes.

It becomes necessary to explain the paradox as to how respiration,
which is largely a reflex phenomenon, can be stimulated by narcotine
whilst the sensory reflexes are depressed. Taken in conjunction with the
work of Loewy this suggests that in the respiratory centre there is a
different receptive mechanism inducing respiratory activity for chemical
afferent nervous stimuli. One explanation which might be offered to
account for the difference in the action of morphine and narcotine would
be that, while morphine depresses both mechanisms, narcotine only
depresses the one; or it may be that the medulla is rendered
hypersensitive to carbon dioxide and other chemical stimuli acting
directly on its motor mechanism, whilst afferent stimuli may still be
blocked on the sensory side.

Omnopon invariably increased the depth of respiration without
materially altering its rate. This effect is not very marked, although
quite distinct. Large doses, 15mg. per kilo body weight in a rabbit,

44,2 BIO-CHEMICAL JOURNAL

cause a temporary paralysis of respiration. As compared with morphine,
then, omnopon must be regarded as less toxic, and much less depressant
to respiration.

Nevertheless, it decidedly depresses the reflex response to nervous
stimulation, and, therefore, should be greatly superior to morphine in the
treatment of cough.

1 \

ale Ee ae aS? (| Terit; FEW) Lone ie

Fic. IIl.—Cat. Urethane. Intestinal movements and blood pressure. Showing effect of a late
injection of 2 c.c. Omnopon at A.

Intestinal Movements. It is well known that morphine, probably
during its excretion, influences Auerbach’s plexus in such a way as to
diminish peristaltic movement. If, however, a large dose of morphine
is suddenly thrown into the circulation the first effect is invariably
augmented peristalsis, which in the case of man may lead to diarrhoea
and vomiting. Narcotine shows this effect in a very much more
pronounced way than morphine. This drug exerts no stimulant action
on any form of plain muscle or motor nerve ending, but it has a markedly
paralysing action on sympathetic ganglia; hence, it is not unreasonable
to suppose that this action is due to depression of the ganglia on the
course of the inhibitory nerves to the intestine.

THE ACTION OF THE OPIUM ALKALOIDS 443

Omnopon, like narcotine, at first increases peristalsis. For the time
being, the narcotine overshadows the morphine action. This effect is, of
course, followed by diminished _ peristalsis. But, whether this
explanation be correct or not, the important fact remains that omnopon
causes at every stage of its action much less effect in diminishing
peristalsis than morphine.

Protocol of typical experiment :—

Experiment XII.—Cat, 2-5 kilos. A.C. E. Tracheotomy. Cannulae in left common carotid

and right external jugular. Abdomen opened. Incision made in a loop of small

intestine, and balloon inserted.

(a) 2c.c. ‘Omnopon’ ... oe aa .... Great fall in blood pressure. Slight increase
in peristalsis,

(6) 3.c.c.‘Omnopon’ ... ae a ... Very little effect on gut movements at first,
and then increased.
Artificial respiration.
Wash out.
Wait for respiration to become natural.

(c) 2 c.c. ‘Omnopon’ ... ore oe .... No effect on blood pressure. Gut movements
further increased. Respiration natural.

(d) 6 c.c. 0-5 % Narcotine... se ... Great increase in gut movements. Temporary
cessation of respiration.
Artificial respiration.
Respiration natural.

(e) 1 c.c. 0-5 % Nicotine aos mee ... No effect on blood pressure. Temporary
inhibition of gut movements.

({) 2 ¢.c.0-1-% Adrenaline... ies ... Great rise in blood pressure. Cessation
of peristalsis.

444

—

Se Oo ae oe oe ee ou

BIO-CHEMICAL JOURNAL

REFERENCES
Sahli, Therapeut. Monatshefte, 1X, p. 1, 1909; Munchener Med. W ochen., LVII, p. 1326,
1910.

Loewy, Munchener Med. Wochen., LVIII, p. 2408, 1910.

Rodari, Therapeut. Monatshefte, X, p. 540, 1910.

Bergien, Munchener Med. Wochen., LVIII, p. 2409, 1910.

Dobeli, Monats. {. Kinderheilkunde, 1X, p. 34, 1910.

Rodolico, Giornale internaz. d. Scienze Mediche., Ann. XXXII, 1910.

Grey, Lancet, p. 673, 1911.

Trotain, L’ Etude Therapeutique de opium, Paris, 1911.

Wertheimer and Rattfalovitch, Deut. Med. Wochen., XX XVII, p. 1710, 1910.

445

THE DETECTION OF ACETO-ACETIC ACID BY SODIUM
NITROPRUSSIDE AND AMMONIA

By VICTOR JOHN HARDING ann ROBERT FULFORD RUTTAN.

(Received September 17th, 1912)

The use of sodium nitroprusside and ammonia followed by addition
of an acid insufficient in amount to completely neutralize the ammonia,
was first suggested by Le Nobel!, as a method of detecting small quantities
of acetone. The test, of course, is based on the original test of Legal, but
as the two tests differ in result, it is proposed to call the former the
Le Nobel test and to reserve the term Legal’s test exclusively to the
action of sodium nitroprusside and potassium (or sodium) hydroxide
followed by acidification. The two tests differ in the following points:
(a) The Le Nobel test gives with acetone a much bluer shade of purple
than the Legal test; (b) The Le Nobel test will detect aceto-acetic acid,
as will be shown in this paper.

The two most important variations of the Le Nobel test are those
of Jackson-Taylor? and Rothera’. Jackson-Taylor in detecting
acetonuria dispenses with any previous acidification, and obtains in
presence of acetone a petunia coloured ring. This modification is highly
recommended by Saxe*t for clinical practice, and is in general
use in many hospitals. Rothera recommends the addition of
ammonium sulphate to the acetone solution, either in water or
in urine, followed by the addition of sodium nitroprusside solution
and concentrated aqueous ammonia. Other ammonium salts may
be used, and the authors can confirm Rothera’s experiments
that such a procedure greatly enhances the value and delicacy of the test.
Experiments on the Legal test also confirm Rothera’s conclusions on the
mechanism of the reaction; the condensation of the sodium nitro-
prusside and acetone by means of alkali giving rise to the production of
a compound which acts as an indicator, but which, in the Le Nobel
test, is unaffected by a low concentration of OH ions. The authors
examined the action of a great number of organic acids in this respect.
The method of applying the Le Nobel test was to acidify the acetone
solution with the organic acid, add 0°5c.c.M/10 sodium nitroprusside

446 BIO-CHEMICAL JOURNAL

(1 c.c. = 0°0298 grs. Na,Fe(CN),NO ° 2 H,O) and then overlie the solution
with concentrated aqueous ammonia. When examining a suspected
case of acetonuria, the organic acid invariably used was acetic acid.

When one comes to apply the Le Nobel test to the detection of
acetonuria, one meets several anomalies which, so far, do not seem to
have been put on record. The difficulties met with may be summarized
as follows :—

(1) Solutions of acetone in water and in urine of concentrations
similar to those occurring in cases of acetonuria do not give the test
as distinctly as the natural cases.

(2) If some samples of urine which give a marked response to the
Le Nobel test be distilled with acids, the test given by the distillate,
where the acetone is presumably ten to twenty times more concentrated
than in the original urine, is usually much less marked. It was a
knowledge of these anomalies which attracted the attention ef the authors
to this subject.

In a paper on the output of acetone under physiological and
pathological conditions, Engel® determined the percentages excreted
in various diseases and under varied conditions of diet. The
acetone was determined by the Messenger-Huppert method, thus
the figures given include acetone from aceto-acetic acid as _ well
as free acetone, and are consequently too high. The highest
amount of acetone found was 0076 per cent. The authors, however,
have found a much higher percentage of acetone in one case of severe
‘diabetes mellitus,’ accompanied by ‘acidosis.’ The percentage of
acetone in the twenty-four hour sample, the day preceding death,
was 0134. This was determined by the Messenger-Huppert method.
The percentage of free acetone as determined by the Folin method
was 0°057. If a solution of 0°134 per cent. acetone be made in water or
in urine and the Le Nobel test applied to it, a light purple coloration
results at the end of about three minutes, while a solution of 0°057 per
cent. acetone will only give a very faint reddish-violet ring at the end
of about twenty minutes, a coloration which might easily be overlooked.
The ‘acidosis’ case in question gave a strongly positive Le Nobel test
at the end of about a minute. Thus, in this case, the test as given by
solutions of acetone in water or in urine of a concentration the same as
that occurring in a natural case of acidosis is not as distinct as in the
natural urine.

In cases where the Le Nobel test is doubtful, it is always recom-

;
i,

«
THE DETECTION OF ACETO-ACETIC 447

mended to acidify the urine with either sulphuric or acetic acid and
distil, usually taking about 100c.c. of urine and collecting from 5 to
10 c.c. of distillate, and then to apply the test to the distillate. Such
a concentration of acetone would be expected to lead to an intensification
of the colour reaction with sodium nitroprusside. In the acidosis case
previously mentioned, and also in a second case examined by the authors,
such an intensification did not occur. In the first case, the Le Nobel
reaction was much weaker in the distillate than in the original urine,
while in the second it had disappeared entirely.

It is evident from a consideration of these facts that there must be
present in certain cases of acetonuria some constituent or constituents
which either of themselves respond to the Le Nobel test or act as
intensiliers to the reaction with acetone. Such substances might be
ammonium salts, hydroxy butyric acid and aceto-acetic acid. From
Rothera’s experiments*, it might be expected that the increased
excretion of ammonium salts which occurs in acidosis would account for
the increased delicacy of the Le Nobel test when applied to a natural sample
of acetonuria. That this is not the explanation is easily proved in the
following manner. Fifty cubic centimetres of urine, which gave a strong
Le Nobel reaction, were acidified with oxalic acid, and gently distilled,
the first five cubic centimetres of distillate being collected. The distillate
gave a weaker reaction than the original urie, and the residual urine
gave a very faint reaction. A second specimen of the urine was then
distilled under exactly similar conditions and the distillate added to the
residual urine. This mixture gave a very faint Le Nobel reaction,
much less intense than the original, thus showing that ammonium salts
in the urine were not present in sufficient amount to greatly increase
the delicacy of the Le Nobel reaction with acetone, and also that the
substance to which the intensifying action was due was destroyed by
heat. This was confirmed by boiling a sample of the urine, acidified
with oxalic, acetic or sulphuric acid, under a reflux condenser for fifteen
minutes, when it was found that an application of the Le Nobel reaction
gave a negative result. As the specimen in question contained aceto-
acetic acid, the urine at the end of the experiment would contain more
acetone than at the beginning, and yet the Le Nobel test failed to show
its presence. This pointed to the fact that it was the presence of aceto-
acetic acid in the urine which caused the marked Le Nobel reaction.
Aceto-acetic acid, however, is stated in the literature to give the following
colorations : —

448 BIO-CHEMICAL JOURNAL

Na,Fe (CN),NO Alkali = Reddish-brown, unchanged by organic
acids or metaphosphoric acid.®
Na,Fe (CN),NO Ammonia = Orange-red.
¥ Me Sold ammonium salt.
Orange-red.
iy KOH = Orange-red, unchanged by acidification
by phosphoric acid.?

No mention has been found of a violet or purple coloration such as
it produced in the application of the Le Nobel reaction to acete-aceturia.
That the violet coloration given by the Le Nobel reaction with aceto-
aceturia is due to the aceto-acetic acting independently of the acetone
is evident from the following experiment. A natural sample of urine
containing aceto-acetic acid was acidified with oxalic acid, saturated with
sodium chloride and all the free acetone removed by aspiration by means
of a current of air, as in the Folin method of estimating acetone. This
was continued for an hour. At the end of that time the free acetone
left in the urine was determined by the Folin method and found to be
nil. The acetone-free urine, however, still gave a strong Le Nobel
reaction, almost undiminished in intensity, which reaction became
negative if the acetone-free urine was boiled under a reflux condenser.
A solution of aceto-acetic acid was next prepared in the following
manner: ‘Thirteen (13) grams of ethyl aceto-acetate were added to a
solution of six (6) grams of potassium hydroxide in 240c.c. of water,
the mixture shaken and allowed to remain at room temperature for

twenty-four hours. It was then diluted with water to 1,000 c.c., and
this solution was used as a solution of aceto-acetic acid. It contains also
some unhydrolysed ethyl aceto-acetate and some acetone. If the

Le Nobel reaction be applied to this solution, it will be found to give a
violet coloration, indistinguishable from that given by urine containing
aceto-acetic acid. If the acidification with acetic acid be omitted as in
the Jackson-Taylor modification, the coloration appears at the end of a
few seconds and is a distinctly redder shade of violet. If the solution
is acidified and boiled under a reflux condenser, the test becomes
negative; an acetone-free solution, obtained by removing the acetone by
a current of air as in the Folin method of estimation of acetone, gives the
test with undiminished intensity. Thus there is no doubt in the minds
of the authors that the intense violet colorations obtained by the
application of the Le Nobel test to urine are due to aceto-acetic acid, and
not to acetone. This diminishes still further the value of this test as a

——— oe

THE DETECTION OF ACETO-ACETIC 449

test for acetone alone. The authors, however, are at a loss to explain
the results obtained by Rothera’, who, by the addition of sodium
nitroprusside and ammonia to aceto-acetic acid, obtained an orange-red
coloration, a colour similar to that given by ethyl aceto-acetate. If a
few drops of our solution of aceto-acetic acid be added to normal urine,
such a urine behaves exactly as the urine of a case of aceto-aceturia.

The point next investigated was the influence of other substances
upon the test, which were likely to occur in urine, and the influence of
acids, other than acetic. The following table shows quite clearly that
the test is unaffected by the ordinary constituents of acidosis or diabetic
urine.

Substance Result Remarks

Creatinin ... ais ae — No appreciable effect on coloration
Glucose... dis ae -— No appreciable effect on coloration
NabieeOr =: ae se -- Coloration as with acetic acid

8 Oxybutyric acid 50C — Coloration as with acetic acid

Formic acid 580 aes — Coloration as with acetic acid

Oxalic acid ae ae — Coloration as with acetic acid

Sulphuric acid... ie — Coloration not so intense as with acetic acid
Hydrochloric acid oie — Coloration not so intense as with acetic acid

In the first two cases the substances were added to the solution
of aceto-acetic acid and the Le Nobel test appled as usual. In the other
cases the acids named were used to replace acetic acid in the Le Nobel
test. It next became important to compare the delicacy of this test for
aceto-acetic acid with that of ferric chloride. The tests were carried out
as follows: The urime containing aceto-acetic acid was rendered acetone-
free by the Folin method, and the aceto-acetic acid then determined by
the Messenger-Huppert method. A sample of the acetone-free urine was
then tested by the ferric chloride test and by the Le Nobel at varying
dilutions, and the dilutions noted at which the urines just failed to
respond to the tests. The dilutions were performed with normal urine.
Two cases were examined.

A. Percentage aceto-acetic acid = 0°139.

ms free acetone = nil.
Dilution 1 part aceto-acetie acid in Le Nobel test Fe,Cl, test
1 720 + a5
1/5 3,600 a5 35
1/10 7,200 sf =
1/20 14,400 + =
1/40 28,800 + (very faint) —

1/60 43,200

450 BIO-CHEMICAL JOURNAL

B. Percentage aceto-acetic acid = 0°0273.
free acetone = nil.

5}

Dilution 1 part aceto-acetic acid in Le Nobel test Fe,Cl, test
1 3,663 se a
1/2 7,326 a5 =
1/4 14,652 a =
1/8 29,304. + (very faint) —
1/12 43,956 — =

_ It will thus be seen that in both the cases examined the limit of
detection of aceto-acetic acid in urine by the Le Nobel test is about 1 part
in 30,000, whereas the ferric chloride test fails to show aceto-acetic acid
when present 1 part in 7,000. The Le Nobel test is thus at least four
times as delicate as the ferric chloride test, and the authors feel quite
confident in recommending its use as a test for aceto-acetic acid in urine
in place of the ferric chloride test.

The delicacy of the test in aqueous solution cannot be easily
determined, and the figures obtained would not, in our opinion, be of
sufficient interest to justify the extremely complicated processes which
would have to be undertaken to obtain pure aceto-acetic acid. As a
rough indication, however, a sample of our solution of aceto-acetic acid
was rendered acetone-free and the aceto-acetic acid determined by the
Messenger-Huppert method, and the test applied at varying dilutions.
The experiment indicated that the limit of detection of aceto-acetice acid
in aqueous solution by the Le Nobel test is about 1 part in 80,000. The
error consists in the presence in the solution of ethyl aceto-acetate, which
would make the percentage of aceto-acetic acid appear too high. The

limit of detection is thus greater than the actual experimental figures.

REFERENCES

Le Nobel, Archiv. fir expr. Path., XVIII, p. 10, 1884.
Jackson-Taylor, Lancet, I, p. 805, 1907.

Rothera, Journ. Physiol., XX XVII, p. 491, 1908.
Saxe, Examination of Urine, Sec. Ed., p. 168, 1909.
Engel, Zeit. fr klin. Med., XX, p. 514, 1892.

Bitto, Ann., CCLXVII, p. 867, 1892.

Rothera, Journ. Physiol., XXXVII, p. 492, 1908.

WHS UH Pr wd

OS ——

451

THE FATTY ACIDS OF BUTTER
By IDA SMEDLEY (Beit Memorial Research Fellow).
From the Bio-Chemical Department, Lister Institute
(Received September 19th, 1912)
I. Tue Nature oF THE Caproic ACID PRESENT IN BUTTER

In spite of the large number of analyses of butter fat carried out
for technical purposes, there remains considerable uncertainty as to the
constitution of the fatty acids it contains. Even the nature of the lower
fatty acids present is not definitely established, for whereas the caproic
acid isolated from butter fat is described in most of the standard works
of reference! as iso-caproic acid on the alleged authority of Chevreul,
Lewkowitsch? prefers to regard it as the normal acid from analogy with
the other fatty acids in butter. No work on the constitution of the
hexoic acid present in fats appears to have been carried out since its
isolation by Chevreul®.

In Chevreul’s pages no reference to the molecular structure of the
acid is to be found, nor was the existence of two isomeric caproic acids
established until fifty years after this work was published. The iso-acid
was the first hexoic acid to be synthesised; Frankland and Kolbe4
saponified the cyan amyl obtained from the amyl alcohol of fermentation

and obtained an acid possessing, therefore, the structure

CH,
CH,.CH,.CH,.CH,.COOH.

It was not until twenty years later than the normal acid was
discovered by Lieben and Rossi>, who carefully characterised the normal
and iso-acids®, the solubility of the barium and calcium salts being
especially characteristic. Up to this time, caproic acid had been found
in butter*, in cocoanut oil’, in the flowers of Satyrium hircinum®, in the
fusel oil of beet-molasses®, and in the fruits of Ginko biloba!®. In none
of these papers is any attempt made to identify this acid as either normal
or iso, and, apparently, it is only because the iso-acid was synthesised
twenty years before the normal that this error has crept into the literature.
Franchimont and Zincke ascribe the normal structure to the hexoic acid
obtained by oxidising the hexyl alcohol from Heracleum!!, and the
caproic acid from the butryic fermentation of sugar was also identified
as the normal acid by Lieben!?.

Examination of the data given by Chevreul indicates that the hexoie

452, BIO-CHEMICAL JOURNAL

acid of butter has probably the normal structure; he gives as the
solubilities of the barium and calcium salts :—
100 parts of water at 10°5 dissolve 8°02 parts Ba salt.
RSs co, cee ee eine.) eee Oh G
Lieben and Rossi® give the following values :—
Normal Acid—
100 parts of solution at 18°5 contain 84967 grs. Ba salt.
3 ate «tere mee yee 98 oo ee Ol yes aeons
Iso-acid—
LOO! 68s ee ameeee He 3 OLGS, “ov a,
15%, es a ae 55 dbs a 5 Aas os

Chevreul’s data appear, therefore, to indicate that the acid which he
obtained from butter possessed the normal structure.

The identification of the acid present in natural fats as normal
or as iso-caproic acid is one of considerable importance; the existence

of iso-caproic acid (CH,),.CH.CH,CH,COH would suggest its derivation ;
from leucin by deamidisation and, hence, might indicate a close
connection with protein metabolism. This view was put forward by

Felix Ehrlich!’, who suggested that the caproic acid present in neutral
fats and that formed in fermentation, although hitherto regarded as the
normal acid, might be mixtures of inactive di-methyl butyric with active
methyl ethyl propionic acid, derived, respectively, from leucine and
iso-leucine.

Ehrlich pointed out that fatty acids have been observed in the
bacterial decomposition of proteins, leucine and caproic acid occurring
together!4; whilst the production of dextro-rotatory caproic acid from _
the decomposition of casein and elastin has also been observed by
Neuberg!>; in the former of these papers, however, no evidence is given
as to the structure of the caproic acid present.

On the other hand, if the fatty acids which occur in butter and in
cocoanut oil containing 4, 6, 8, 10, 12, 14, 16, 18 and 20 carbon atoms,
respectively, consist of normal chains and are synthetic products, their
formation from some simple starting product by an analagous series of
reactions in which at each step two carbon atoms are added, appears
probable.

It appeared desirable, therefore, to settle quite definitely the
structure of the hexoic acid present in butter. The differentiation of the
normal and iso series of acids is conveniently based on the examination

THE FATTY ACIDS OF BUTTER 453

of the amides, those formed from the iso-acids melting at considerably
higher temperatures than those which belong to the normal series.
This is shown in the following table :—

Normal M. Point Iso M. Point
Butyric _... nh re teLLoS Dimethyl acetic ar 2S,
Valeric “cf aie .» 114-116 Iso-propyl acetic va coo wiles
Caproic... nan sen LOO Iso-butyl acetic = pia eS
Heptoic... aa gas 96-97
Caprylic.... rr fe 97-98 Ethyl-butyl acetic ae ayn 102
Nonoic SE sie “1 99
Caprin ae oa ss LOS
Undecanic ... ae -» 108

The isolation of the caproic acid was carried out as follows :—

About 3,500 grs. of pure butter fat* having the constants Iodine
value 387, Saponification value 232, Reichert-Meiss] 29°9, were
saponified with alcoholic potash and the fatty acids liberated. The
mixture was then steam-distilled until the odour of the lower fatty acids
was no longer markedly perceptible in the distillate. The distillates,
neutralised with potash, had a distinct odour of the higher fatty alcohols;
sufficient was not, however, obtained for further identification. After
concentrating to a small bulk, the solution of the soaps was acidified and
extracted with ether. 125 grams of acid thus obtained were distilled
from a flask fitted with a Young’s 8-pear fractionating column and
separated into the following fractions :—

Boiling point Weight
Up to 160°... Se a 24 grs.

163—168 SG Ae Sec 4S) 5 Boiling point, butyric acid, 163 (7 mm.)
168—173 ase see ee 25,
173—178 es 586 ee LORS: Boiling point iso-valeric acid, 174
178—183 ar ae aoe ome
183—188 ees Boiling point N. valeric acid, 186
188—193 2 5
193—198 0-5 ”
198—203 7d Wer Boiling point iso- -caproic acid, 200
203—208 Ba Boiling point N. caproic acid, 205
208—213 6

The residue remaining in the flask (16 grs.) was then distilled without
a fractionating column.

213—218° 1-2. grs.

218—228 AS ys

228—238 Pet or

238—248 sey) es

248—258 1- 7 ”

Residue 4:1 ,, Solidified on cooling

Two grams of each ES ‘ee hear boiling at 173-178°, 183-188°, 198-203°,
203-208° and 208-213° were then converted by Aschan’s method!® into the
corresponding amides. From each of the fractions 198-203°, 203-208°

* Pure butter was obtained from Lovegrove’s Dairy, Checkendon.

454 BIO-CHEMICAL JOURNAL

and 208-213° an amide separated, melting at 96-97°. This was
recrystallised from dilute alcohol and, finally, from a mixture of
chloroform and petroleum; the crystals melted at 99-100°.

Amides were then prepared from Kahlbaum’s normal synthetic
caproic and isobutyl acetic acids. After recrystallisation these melted
at 99-100° and at 118°, respectively. The melting point of the former
was not depressed when mixed with the amide of the caproic acid from
butter. The fraction boiling from 198-213° contained, therefore, normal
caproic acid, and gave no indication of the presence of the iso-acid.

The fraction boiling up to 160° was now examined for acetic acid.
It consisted chiefly of ether, and was neutralised by the addition of 1°2 c.c.
N.KOH. On acidifying the solution of the small amount of potassium
salt, drops of an insoluble acid with the characteristic butyric odour at
once separated, so that the quantity of acetic acid, if any were present,
could only have been exceedingly small.

In the above experiment, the steam-distillation was not coutinued
until the whole of the volatile acids had passed over, but was stopped
when 4°5 per cent. of the total weight of acid had been received. The
proportion of acids present in the distillate examined was

approximately : —
Butyrre acid “e.. bia 85 gr. 74°6 per cent.
Caproic acid... ae Loe. aS Ps
Caprylic and Capricacids 12,, 10° s
Solid acids a ane 4 ,, 3°90 5

II. Tue EXIsteNcE oF UNSATURATED ACIDS IN BUTTER OTHER THAN
OxLrIc ACID

A careful fractionation of the methyl esters of the butter acids was
undertaken in order to obtain evidence as to the possible existence of lower
members of the oleic acid series. It is a remarkable fact that whereas
the 18-carbon atom unsaturated acid is commonly identified in fats, the
existence of lower members of the oleic series has only been shown in
cod-liver oil!?, The experiments now carried out appear to indicate
the presence of a small quantity of a lower unsaturated acid
accompanying the decoic acid present.

The butter used for the preparation gave the following values:
Reichert-Meiss] 26°84, Iodine 37, saponification 228.

1,900 grams of dried butter-fat were shaken up with 567 ¢.c. of a
solution of sodium in methyl alcohol (4536.N). The emulsion formed

—

THE FATTY ACIDS OF BUTTER 455

was allowed to stand overnight and the esters next morning extracted
with ether. After drying and evaporating off the ether on a water bath,

1,630 grams of liquid remained.

Examination for unsaturated acids.

The esters boiling above 160° under atmospheric pressure were then
distilled under a pressure of from 10-15 mm., in a flask fitted with a
Young’s eight-pear fractionating column. The fractions were collected
at first over from five to ten and then over every three degrees, and the
iodine values of the different fractions determined; these are shown in

the following table :—

Weight Saponifica-
Pressure __ Boiling of dis- Iodine tion
point tillate value number
15mm.... 80—85° 4oms. 2-9 — Boiling point methyl caprylate, 83° (15 mm,
85—90 Ges —- -
90—95 — -- —
95—105 figgets 5-4 —
105—110 SPs 10-4 —
110—120 OM: 12-5 308-2 Boiling point methyl caprate, 114° (15 mm.)
: Sapon. No. 302
120—127 10 ,, ered 307-3
127—132 Se «55 8-7 282-9
132—147 — —_— —
147—151 10: ,; 4-9 — Boiling point methyl laurate, 141° (15 mm.)
Sapon. No. 262
151—152 \
te2—te0 {28 - [78 ea
160—167 PAD) ee 7:5 256-8
167—172 PAT 7-7 240-7 Boiling point methyl myristate, 167-8° (15 mm.)
Sapon. No. 232
172—175 1605; 7-9 —
10 mm. ... 165—170 PAs ae ore 8-1 -—
170—173 40 ,, 8-8 241
173—176 1G* 5; 8-0 —
12mm. ... 176—180 Boiss 8-1 —
180—183 14! 8-8 —
183—186 PIBt as 9-4 —
186—189 100 ,, 17-2 —
189192 120 ,, fa 291
192—195 89). 5. 20-5 — Boiling point methyl] palmitate {186° (10 mm.)
Sapon. No. 208 | 196° (15 mm.)
195—198 ass 29-9 215
198—201 29 5, 37:2 —
201—204 GBs: 48-5 —
204—207 138 ,, 61-1 196
207—210 215 ,, 66-6 —

205-6°
210-215 114 ,, 67-3 — Boiling point methyl oleate {319.9 (15 mm)
Boiling point methyl stearate 214-5°(15 mm.
Sapon. No., methyl stearate, 188
215—218 26 67-8 —
Residue in flask
boiling above 220 SP — —

BIO-CHEMICAL JOURNAL

456

DISTILLATION oF Meruynt Esters oF BurrerR-F at (obtained by Bull’s method),

under 10-15 mm. pressure.

Curve showing relation of Iodine values to Boiling points.

Todine

values

(‘urut ¢T)
ayervoyg of “4d “gf
oyva[Q ow “4d “g

(urur OT)
aqvelO eo “3d “gq

(‘urut ¢T)
aqyeqrurpeg ey “9d ‘g

(‘uxut OT)
ayeqrmed oy “4d “qT

‘ww 21
2YNSSIud

(‘uur GT)
oyeqssyy oy 4d “g

(“urur gt)

aqyemey oy, 3d ‘g

(-urut Gt)
ayeooop oy “4d ‘g

(‘urut GT)
eyvoqoo op, “9d “Gg

THE FATTY ACIDS OF BUTTER 457

The curve shows a marked rise in the iodine values of the fractions
boiling from 105-127°. The marked maximum at 120° is followed by a
gradual depression until 150° is reached, the iodine value remains
constant whilst the boiling-point rises through 30° C., and then a rapid
increase marks the distillation of the oleic acid. So far, attempts (based
on the crystallisation of the lead, magnesium and barium salts) to isolate
unsaturated esters from these lower boiling fractions have proved
unsuccessful. Were oleic the only unsaturated acid present, a steady
rise in the iodine value, with increase of boiling-point, might have been
expected. The increase in iodine value in the fraction boiling from 110°
to 130° appears to the author to furnish evidence of the presence of an
unsaturated ester containing a less number of carbon atoms than oleic,
possibly a decylenic acid since it passes over with the fraction
containing methyl decoate (caprate). It is possible that traces of other
unsaturated acids may also be present, and that the iodine values found
for the lower fractions may not be due to traces of oleic acid.

Nistillation of the ethyl esters.

A confirmatory experiment was carried out in which the fatty acids
from the 3,500 grams of butter-fat used for the separation of caproic acid
(cp. p. 453) were converted into the ethyl esters by boiling for several
hours with an alcoholic solution of hydrochloric acid. After washing
with a dilute solution of sodium carbonate to remove free fatty acids, and
drying with CaCl,, 2,662 grams of esters were obtained, which were
distilled under a pressure of 20-30 mm. and divided into three fractions:

(1) boiling up to 200° . - 281 grams;
(2) boiling from 200-220° - - 654 grams;
(3) boiling above 220° . - 1,727 grams.

The fraction boiling up to 200° was then distilled under a pressure of
from 20-28 mm. in this case a six-pear fractionating column being used.
The ethyl esters boil about 20° higher than the corresponding methyl
esters. The iodine and saponification values of the different fractions
were determined as in the previous experiment; a distinct increase in
the iodine value was found in the fraction, the saponification value of
which agreed with that required for the decoic ester.

Distillation of ethyl esters of butter fatty acids boiling up to

200° (20-30 mm.); (the more volatile acids having been partially
separated).

458 BIO-CHEMICAL JOURNAL

Weight Saponifica- ;
Pressure Boiling of tion Iodine
point Fraction Number Number

25-230nm: ~—.. 80—85° 12-8 389-8 0-7. Sapon. No. ethyl hexoate = 388
(caproate)
85—90 2:3 = =
90—95 2:0 == 1-5

95-100 0-2 — —
100—105 1-2 — —
105—1lo )

Lip 41a ry I

115—120 4-0 349-6 2-5. Sapon, No. ethyl octoate = 326
(caprylate)

120—125

125130 } 20-0 316-5 3-0

130—135 5 = =

135—140 5:5 = 6-8

140—145 15-5 288-8 7:8 :

145—150 21 282-1 8-7 Sapon. No. ethyl decoate = 280
(caprate)

* 20 mm. Sec 150—155 -: a ==

155—160 18 259-4 7-2

160—165 9 = 6:8

165—170 Th 252-8 —

170—175 6 = 5-5 Sapon. No. ethyl laurate = 245

175—180 18 238-9 6-5

180—185 3 = =

185—190 8 — 7-4

190—195 14 222-6 9-6 Sapon. No. ethyl myristate = 219

The distillation was not further continued.

* Distillation stopped here : continued on the following morning under a pressure of 20 mm.

DIsTILLATION OF Eruyn Esters oF Burrer-Far (obtained by esterification of fatty acid with
ethyl alcohol and HCl) under 20-25 mm. pressure.

Curve showing relation of Iodine values to Boiling points.

Iodine

values Ap
iE)
10
5

Boiling ©

points

vv ee

THE FATTY ACIDS OF BUTTER 459

Three other experiments were carried out, in each of which between
300 and 400 grams of methyl esters were prepared and distilled. In every
case the iodine values of the fractions corresponding to the decoic acid
were higher than those of the succeeding fractions. The increase is not
very large, but was present in five different specimens of butter
examined, and it may be inferred therefore that it is generally present,
and may be regarded as significant. It appears probable that the

presence of a decylenic acid is indicated.

III]. ON THe PRESENCE oF STEARIC ACID IN BUTTER

In view of the statements of Lewkowitsch and others (cep.
Lewkowitsch, loc. cit.) that, when examined by the ordinary methods
used in the examination of butter, only traces of stearic can
be found, the higher boiling fractions of the methyl esters were
examined for stearic acid. From the fractions in two experiments boiling
at 210-215° a solid ester was separated, which, when twice re-crystallised
from alcohol, melted at 37-38°, and on saponification gave an acid
crystallised from alcohol in plates melting at 69°, and which was
therefore identified as stearic acid. The quantity of stearic present was
approximately estimated as about 10-15 per cent. of the total butter acids.
Lewkowitsch has drawn attention to anomalies observed in the estimation
of stearic acid when lower acids than palmitic are present (cp.
Lewkowitsch, loc. cit., Vol. I, p. 453). The presence of considerable
quantities of stearic acid in butter-fat was also found by Caldwell and
Hurtley'S, who state that the oleic acid of butter occurs chiefly in
combination as an oleo stearo palmitin.

Reaction with sodium-nitro-prusside.

The distillate, boiling up to 140° under atmospheric pressure, gave
a very low iodine value of between two and three units. The fraction
boiling from 100-120° was heated with a solution of caustic potash,
acidified with acetic acid and a solution of sodium-nitro-prusside and
strong ammonia then added; a deep violet colour developed. The same
fraction gave negative results when treated with Fehling’s solution and
with ammoniacal silver nitrate. The sodium-nitro-prusside reaction was
obtained in two out of five experiments in which the esters had been
prepared by Bull’s method; but when the esters were prepared by an
acid method of alcoholysis, in no case was the reaction obtained. This

460 BIO-CHEMICAL JOURNAL

reaction is especially characteristic of aceto-acetic acid and of acetone,
and it is possible that these were present in those samples examined
which gave positive results. In view of the inconstancy of the results
obtained, it is possible that these products had been formed as the result
of bacterial action. Further evidence as to the nature and origin of the
substances producing this reaction is desirable, as the certain detection
of aceto-acetic acid in butter would be a point of considerable importance.

CoNCLUSIONS

1. No evidence of the presence of acetic acid was found in the
butter examined. .

2. The hexoic (caproic) acid present in butter possesses the normal
structure; no indication of the presence of the iso-hexoic acid was
obtained. It seems probable that of all the naturally existing specimens
of caproic acid which have been described, only that occurring in the
bacterial decomposition of proteins has a branched structure.

3. The proportion of stearic acid was estimated as from 10-15 per
cent.

4. Evidence was obtained of the existence of lower members of the
oleic acid series; the iodine value of the decoic ester fraction is
appreciably greater than those of the fractions immediately preceding
or following it. This may be regarded as indicating the presence of a
lower unsaturated acid, possibly of a decylenic acid.

5. The sodium-nitro-prusside reaction, characteristic of aceto-acetic
acid and of acetone, was given by the butyric ester fraction obtained
from two out of five samples of butter, esterified by sodium methylate at
ordinary temperature: the remaining three samples gave negative
results. The reaction may possibly have owed its origin to the products
of bacterial action; it was observed in two cases where Haller’s acid
method of esterification had been used.

to
:

% : a = *. P
THE FATTY ACIDS OF BUTTER 461

REFERENCES

Abderhalden, Biochemisches Hand-Lexikon, Vol. 1, p. 989; cf. Beilstein, Handbuch der
Organischen Chemie, Vol. I, p. 482.

Lewkowitsch, Chemical Technology and Analysis of Oils, Fat and Waxes, 4th Edition,
Vol. I, p. 122, 1909.

Chevreul, Recherches sur les corps gras, p. 134, 1823.
Frankland and Kolbe, Annalen, LXV, p. 288, 1848.
Lieben and Rossi, Ann., CLIX, p. 75, 1871.
Lieben and Rossi, Ann., CLXV, p. 113, 1873.
Fehling, Ann., LIII, p. 406.

Chautard, Jahresber., p. 340, 1864.

Miller, Jahresber., p. 499, 1852.

Bechamp, Ann., CXXX, p. 364.

Franchimont and Zincke, Ann., CLXIII, p. 193.
Lieben, Ann., CLXX, p. 89, 1873.

Felix Ehrlich, Ber., XL, p. 2547, 1907.
Emmerling, Ber., XXX, p. 1863, 1897.

Neuberg, Biochem. Ztschr., 1, p. 368, 1906.
Aschan, Ber., XX XI, p. 2348, 1898.

Bull, Ber., XX XIX, p. 3570, 1906.

Caldwell and Hurtley, Trans., XCV, p. 858, 1909.

ae - We

ae

QP ‘The Biochemical journal
501

B47

v.6

cop.4

Biological

& Medica]

Serials

PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET

UNIVERSITY OF TORONTO LIBRARY

“

—

CE PIG UBB AG
hii; yp. f
GEIEE th

gah
PLAS
HiMlon

Bae
Rares

Wel phere,
ide
Wiig

a it

7,
He na
TELESE LS
oe Leas

Dae +e yee iy
pie a
ee ie Feuacits
; Hs

y Dias

PHU
Min
HME ee

OLE,

fe fe Ae $ ’
Ue I ioe
TT ies
Bh

iy
vy
EG

WH
7 Wie
; Mpeg ay

RY
io
Us