GENERA!, G M R ACTERS

*N "*?*'** *v*j vw. '*v"" i •

KINS

SJV

S. fe. SCHRVVER, Ph.D., D;Sc

MONOGRAPHS ON BIOCHEMISTRY

EDITED BY

R. H. ADERS PLIMMER, D.Sc.

AND

F. G. HOPKINS, M.A., M.B., D.Sc., F.R.S.

MONOGRAPHS ON BIOCHEMISTRY.

Royal 8vo, boards.

THE DEVELOPMENT AND PRESENT POSI-
TION OF BIOLOGICAL CHEMISTRY. By
F. GOWLAND HOPKINS, M.A., M.B., D.Sc., F.R.S.

THE NATURE OF ENZYME ACTION. By
W. M. BAYLISS, D.Sc., F.R.S. 35. net.

THE CHEMICAL CONSTITUTION OF THE
PROTEINS. By R. H. ADERS PLIMMER, D.Sc.
In Two Parts. Part I., 33. net; Part II.,
2s. 6d. net.

THE GENERAL CHARACTERS OF THE PRO-
TEINS. By S. B. SCHRYVER, Ph.D., D.Sc.,
2s. 6d. net.

THE VEGETABLE PROTEINS. By THOMAS B.

OSBORNE, Ph.D.

THE POLYSACCHARIDES. By ARTHUR R. LING,
F.I.C.

GLUCOSE AND THE GLUCOSIDES. By E.
FRANKLAND ARMSTRONG, D.Sc., Ph.D.

THE FATS. By J. B. LEATHES, D.Sc.
COLLOIDS. By W. B. HARDY, M.A., F.R.S.

LONGMANS, GREEN, AND CO.

LONDON, NEW YORK, BOMBAY, AND CALCUTTA.

THE

GENERAL CHARACTERS

OF

THE PROTEINS

BY

S. B. SCHRYVER, Ph.D., D.Sc.

LECTURER ON PHYSIOLOGICAL CHEMISTRY, UNIVERSITY COLLEGE, LONDON

LONGMANS, GREEN, AND CO.

39 PATERNOSTER ROW, LONDON

NEW YORK, BOMBAY, AND CALCUTTA

1909

GENERAL PREFACE.

THE subject of Physiological Chemistry, or Biochemistry, is
enlarging its borders to such an extent at the present time,
that no single text -book upon the subject, without being
cumbrous, can adequately deal with it as a whole, so as to
give both a general and a detailed account of its present
position. It is, moreover, difficult, in the case of the larger
text-books, to keep abreast of so rapidly growing a science
by means of new editions, and such volumes are therefore
issued when much of their contents has become obsolete.

For this reason, an attempt is being made to place this
branch of science in a more accessible position by issuing
a series of monographs upon the various chapters of the
subject, each independent of and yet dependent upon the
others, so that from time to time, as new material and
the demand therefor necessitate, a new edition of each mono-
graph can be issued without re-issuing the whole series. In
this way, both the expenses of publication and the expense
to the purchaser will be diminished, and by a moderate
outlay it will be possible to obtain a full account of any
particular subject as nearly current as possible.

The editors of these monographs have kept two objects
in view : firstly, that each author should be himself working
at the subject with which he deals ; and, secondly, that a
Bibliography, as complete as possible, should be included,
in order to avoid cross references, which are apt to be
wrongly cited, and in order that each monograph may yield
full and independent information of the work which has been
done upon the subject.

It has been decided as a general scheme that the volumes
first issued shall deal with the pure chemistry of physiological
products and with certain general aspects of the subject.

vi GENERAL PREFACE

»

Subsequent monographs will be devoted to such questions
as the chemistry of special tissues and particular aspects of
metabolism. So the series, if continued, will proceed from
physiological chemistry to what may be now more properly
termed chemical physiology. This will depend upon the
success which the first series achieves, and upon the divisions
of the subject which may be of interest at the time.

R. H. A. P.
F. G. H.

PREFACE.

IN the following pages an attempt is made to review the
chief properties of the proteins, with the object of determining
how far they are of value for devising methods of isolation
and identification of individual members of the class.

In view of the limited scope of the essay, certain aspects
of the subject have been purposely treated in a somewhat
empirical manner. No attempt, for example, has been made
to explain such processes as " salting out " or heat coagula-
tion ; subjects such as these may be more fittingly discussed
in a monograph dealing with the general chemical physics of
colloids.

In spite of the great advances recently made in the know-
ledge of the physics and of the constitution of proteins, the
methods available for their isolation and identification are
still very unsatisfactory in character. This fact may serve
as an apology for the tentative treatment of certain sections
of this book ; the importance of the subject, however, both
to chemists and physiologists, may warrant the issue of
this monograph, dealing, as it does, with a part of protein
chemistry still in a rudimentary state of development.

S. B. S.

vii

CONTENTS.

PAGE

INTRODUCTION . i

PART I.
THE PHYSICAL PROPERTIES OF THE PROTEINS.

SECTION

I. THE SOLUBILITY OF PROTEINS IN SALT SOLUTIONS ; THE

"SALTING OUT" FROM SOLUTIONS .... 8

II. THE DEGREE OF SOLUBILITY OF PROTEINS IN SALT SOLU-
TIONS 15

III. SOLUBILITY OF PROTEINS IN ORGANIC SOLVENTS . . 16

IV. SEPARATION OF PROTEINS FROM SOLUTION BY PRECIPIT-

ANTS OTHER THAN SALTS . . . . . .17

V. CRYSTALLISATION OF PROTEINS . ,. . . . 18

VI. THE TEMPERATURE OF HEAT COAGULATION OF PROTEIN

SOLUTIONS . . . . . . . . .21

VII. OPTICAL ROTATION OF PROTEIN SOLUTIONS ... 24

VIII. MOLECULAR WEIGHT DETERMINATIONS BY CRYOSCOPIC

METHODS 24

IX. THE ELECTROLYTIC CONDUCTIVITY OF PROTEIN SOLUTIONS 26

X. THE "GOLD NUMBER" 27

XL THE FRACTIONAL FILTRATION OF PROTEINS 28

x CONTENTS

I

PART II.

THE GENERAL CHEMICAL CHARACTERS OF
THE PROTEINS.

SECTION PAGE

XII. THE QUALITATIVE REACTIONS OF THE PROTEINS . . 30

XIII. THE CHEMICAL COMPOSITION OF PROTEINS. THE NITROGEN

CONTENT AND DISTRIBUTION 32

XIV. THE SULPHUR, PHOSPHORUS AND HALOGEN CONTENT OF

PROTEINS 36

XV. THE TYROSINE FACTOR OF PROTEINS . . . .38

XVI. SALT FORMATION BY PROTEINS. COMBINATION WITH ACIDS

AND BASES 39

XVII. THE PRECIPITATION OF PROTEINS BY SALTS OF THE HEAVY

METALS 56

XVIII. THE OXIDATION OF THE PROTEINS 58

XIX. THE ACTION OF HALOGENS ON PROTEINS . . .61

XX. THE ACTION OF NITROUS ACID ON PROTEINS . . .67

XXI. ACTION OF FORMALDEHYDE ON PROTEINS ... 69

PART III.

BIOLOGICAL METHODS FOR THE IDENTIFICATION
AND DIFFERENTIATION OF PROTEINS.

XXII. THE PRECIPITIN REACTION . . . . .71

BIBLIOGRAPHY . . 78

INDEX . . . . ... . . .85

INTRODUCTION.

THE proteins belong for the most part to that class of bodies which
Graham has designated the " colloids," to which the ordinary criteria
available for the identification and differentiation of simpler organic
compounds, such as boiling points, melting points, etc., are generally
inapplicable. The want of knowledge of the physical nature of
colloids has been one of the chief obstacles to advances in the experi-
mental investigations of the proteins. The principal problem
affected thereby is that of the separation of the proteins from one
another and from other substances. The proteins, furthermore, are
substances of large molecular weight, yielding on degradation a
great number of products, the isolation and quantitative estimation of
which have taxed the ingenuity of the chemist to the utmost.

The investigations on the proteins may be consequently divided
into two main classes : (i) those connected with the elucidation of their
physical properties as colloids, with which are associated the names
of Hardy in this country, and of Hofmeister, Spiro, Pauli and many
others on the continent; (ii) those connected with their chemical
constitution, with which are associated the names of Emil Fischer
and of Kossel and their pupils. A short review of certain aspects
of protein chemistry in the light of the more recent advances will be
advisable before proceeding to a more detailed discussion of those
properties which may be utilised in the processes for the isolation
and identification of individual substances belonging to the class.

ISOLATION AND SEPARATION OF THE PROTEINS.

Proteins derived from different sources have markedly different
properties. Some are soluble in water, others are insoluble in pure
water, but soluble in saline solutions ; others, again, are soluble in
alcohol. Advantage was taken of these differences for the separation
of proteins from one another, especially in the earlier work on the
vegetable proteins.

Another property, viz., that of precipitability from aqueous
solutions by the addition of neutral salts, also received early attention,
and the difference in behaviour of solutions of different proteins as
regards precipitability was soon turned to account in devising a
method of separation. It is of interest in this place to quote the
words of Denis, the first investigator who systematically employed
the method of " salting out ". On the title-page of his monograph,
Memoir e sur le sang, published in 1856, but containing results of work
commenced many years before this date, he describes his researches
as " e*tudes faites suivant la methode d'experi mentation par les sels, la

2 THE GENERAL CHARACTERS OF THE PROTEINS

seule qui, dans 1'etat actuel de la science, semble pouvoir etre applique
avec fruits a des reserches sur ces substances".1 Although these
words were written more than forty years ago, when the protein
investigations were still in their infancy, they contain a statement
which is substantially true to-day, for the process of " salting out "
is the only one which is capable of general application for the
separation of the proteins. Like the analogous processes of fractional
distillation and fractional precipitation, the process of fractional
" salting out " seldom leads to a complete separation of mixtures.
The efficiency and the gradual development of the technique is
discussed in that section of this work, which, owing to the actual and
historical importance of the subject, precedes all others dealing with
the general character of the proteins. It has purposely been treated
in a somewhat empirical fashion ; the physical nature of the process
will, it is hoped, be dealt with in greater detail in a later volume of
this series.

THE CHEMICAL NATURE OF THE PROTEINS.

Although it was clearly recognised that the nitrogenous matters
obtainable from plants and animals varied considerably in their
physical properties, such as in solubility, appearance, etc., it was
nevertheless held by Liebig that only one protein existed. The
differences in the nitrogenous substances obtained from various sources
were ascribed entirely to variations in physical conditions. The slight
differences in empirical composition in the preparations obtained
accounted, no doubt, for this view of Liebig. Mulder, to whom many
valuable observations are due, expressed later a similar view in a
somewhat modified form. Referring to " protein " he remarks : "It
exists in various forms, being either soluble or insoluble in water. It
forms different compounds with sulphur or with phosphorus or both
— and hence the differences it presents in appearance and physical
properties. The substance has received the name of ' protein '
because it is the origin of so many dissimilar bodies, and is itself a
primary substance " (Mulder, Vegetable and Physiological Chemistry \
p. 291. English translation, 1845-49).

Subsequent investigations have shown that these early views of
Liebig and Mulder are incorrect, more especially the careful and
reliable experiments of Schiitzenberger, Hlasiwetz and Habermann,
Horbaczewski, Schulze and Barbieri and others, most of which were
carried out thirty years or more after the publication of the above-
quoted words of Mulder. It is unnecessary to enter into detail in this
place as to the results of the experiments on the hydrolysis of the
proteins, due to these and other workers, culminating subsequently in
the elaboration by E. Fischer and his pupils of an approximately
quantitative method for the isolation of the hydrolysis products ; these
are discussed in the volume devoted to this subject. Suffice it here to
say that the individual proteins have been found to differ from one
another both qualitatively and quantitatively as regards the amino-
acids which they yield on hydrolysis.

Modern work tends to confirm the theory due to Hofmeister that

li.e., proteins,

INTRODUCTION 3

the proteins are built up by the condensation of several amino-acids
according1 to the scheme —

Ri Rii Riii

NH2.CH — COJOH H;NH.CH-CobH. HJNH2— CH — Cok)H H;NH3.—

... — CO|OH H;NH2 . CH . COOH

thereby forming a class of products which have been designated
the polypeptides by Fischer. Such polypeptides are held to form the
essential part of the structure of the protein molecule ; the latter may,
however, contain other groups, such as phosphoric acid, and possibly
also carbohydrates ; as to whether such groups form an essential part
of the actual protein molecule, or whether they are held in loose
combination in the form of what Hoppe-Seyler described as
" prosthetic groups," need not be discussed in detail here. There is,
however, a certain amount of evidence that phosphoric acid, in the case
of caseinogen, for example, forms an essential part of the molecule.
Whatever view may be held as to the other groups, there is little doubt
that the polypeptide group forms the essential part of the protein
molecule, and that it is formed by the conjugation of different amino
acids in the different individual members of the protein class.

THE RELATIVE VALUE OF THE VARIOUS PHYSICAL AND
CHEMICAL CHARACTERS FOR FIXING THE IDENTITY OF A
PROTEIN CONSIDERED IN THE LIGHT OF PRESENT
KNOWLEDGE.

As already mentioned, the majority of the proteins differ from one
another but slightly in their empirical chemical constitution ; the
numbers obtained by the ultimate chemical analysis are therefore, as
a rule, of but little value for the characterisation of proteins. The
most obvious method for differentiation of the proteins is that of the
quantitative estimation of hydrolysis products. This process, however,
yields reliable results only when relatively large quantities of material
are available for examination, and it requires, furthermore, considerable
expenditure of time. In actual practice, such as in physiological and
pathological research, or in the investigation of foodstuffs, it is, as a
rule, quite inapplicable ; it affords, moreover, no certain criterion as to
the homogeneity or heterogeneity of the substance under investigation.

On the assumption, then, that the protein is a polypeptide, it
remains to be considered what chemical and physical properties may
most fittingly serve for its characterisation.

The Acid and Basic Functions of the Protein Molecule and their
Relations to the Physical Properties.

The discussion of the above question will be facilitated by the
consideration of the properties of a typical polypeptide, e.g.^
the tetrapeptide :—

4 THE GENERAL CHARACTERS OF THE PROTEINS

»

Ri Ri» Riii Riv

CH — NH — CO — CH — NH — CO — CH — NH — CO — CH
COOH NH2

The presence of both an amino and a carboxyl group would
indicate the possibility of the formation of two kinds of salts, viz.,
salts with acids and salts with bases. The acidity or basicity of a
protein is the property which most obviously requires investigation. I f
all were constituted like the substance of the formula given above, with
R1, R", . . . Riv representing simple radicals, containing only carbon
and hydrogen, each molecule of protein would react chemically with
one molecule of acid or one molecule of base. In practice, however,
the problem of the acidity or basicity is not so simple as might at first
sight appear. In the first place, proteins do not yield on hydrolysis
simple monoamino acids ; other products of a more basic character
are obtainable, such as arginine, lysine (which is a simple diamino
acid) and histidine. Similarly on hydrolysis proteins yield amino
acids containing more than one carboxylic acid group, such as aspartic
acid, glutamic acid, etc. Simple polypeptides, yielding either
strongly acid or basic amino acids on hydrolysis, could be represented
by the typical formula, in which R1, R", etc., instead of representing
simple radicals of carbon and hydrogen, would represent radicals in
which one or more hydrogen atoms are substituted by a carboxyl
or amino group.

The existence of such extra amino or carboxyl groups gives rise
to a further possible complication, for they can condense with
carboxyl and amino groups of other polypeptides, and thus give rise
to substances of highly complex molecular structure with branching
chains of polypeptide nature. The proteins, considered as polypeptides,
which yield on hydrolysis both diamino- and dicarboxylic acids, can
act, therefore, either as basic or acidic bodies, according to whether the
amino or carboxyl groups predominate.1

As a matter of fact, certain proteins, such as the protamines, first
described by Miescher, have very strongly marked basic characters ;
these substances yield on hydrolysis more than 80 per cent, of basic
products, and the molecules contain, therefore, a large predominance
of amino groups. Other proteins are of distinctly acid character,

1 Nothing is known as to the stereochemical configuration of proteins. H. E. and
E. F. Armstrong have suggested that the main polypeptide can assume a spiral form,
and have constructed models to represent it according to this hypothesis. In this case
the more reactive groups might be represented as the freely moving groups external to
the spiral. The stereochemical configuration would probably also affect the reactivity
of the amino and carboxyl groups ; adopting the Armstrong configuration it is conceiv-
able that those which are external to the spiral would be reactive, and those protected
by groups of other atoms inert. Furthermore, it is conceivable that certain stereo-
chemical configurations would be favourable to the formation of imino groups through
the elimination of ammonia from contiguous amino groups, or of acid anhydrides by
elimination of water from carboxyl groups, diminishing the basicity in the former and
the acidity in the latter case. It is therefore not possible to predict the acidic or basic
character of the protein from the mere estimation of the hydrolysis products, ascertaining
thus whether diamino or dicarboxylic groups predominate, as stereochemical configura-
tion may render either sets of groups inactive. Furthermore, a certain amount of
ammonia is obtainable from most proteins on hydrolysis (see p. 33) ; this is assumed to
be derived by the hydrolysis of acid amides containing the group- CO -NH2. The
existence of such groups would diminish the acidity due to the presence of carbpxyl
groups.

INTRODUCTION

5

such as some of the globulins and caseinogen, although this character
may be ascribed in the latter substance to the presence of the
phosphoric acid group in the molecule.

The majority of the proteins, however, have neither marked basic
nor acidic character, and in this respect resemble the typical
polypeptide of the typical formula with the radicals R1, R", etc.,
containing only carbon and hydrogen atoms.

The determination of the acidity and basicity of proteins of this
character has entailed many difficulties. This arises from the capacity
of proteins as colloids to adsorb simpler substances and from the
large molecular weights of this class of substances and relatively small
number of active carboxylic and amino groups ; the equivalent com-
bining weight is therefore large ; in the case of the crude egg-
albumins Sjoqvist has shown (p. 40) that between 800 and 900 grams
of protein combine with I gram mol. equivalent of hydrochloric acid
to form the hydrochloride salt.

Now a solution which contains a relatively large percentage of a
solute of large molecular weight is technically very dilute ; on the
assumption that the protein of egg-white is a mono-acid base, its
molecular weight would be about 850, which, according to the
experiments of Sjoqvist, is the lowest possible ; a 5 per cent, solution
would be therefore only j^ normal. The salts of weak bases in
dilute solution readily undergo hydrolysis according to the equation —

B . HC1 + HOH = B . OH + HC1.

The acidity or basicity of a protein cannot, as a rule, be determined,
therefore, by titration with the use of indicators, since hydrolysis of
salts can take place in solutions containing relatively large amounts
of protein.

Furthermore, Hardy has pointed out in the case of the globulins
the possibility of the formation of basic salts. If serum-globulin be
submitted to dialysis (see p. 54) it can hydrolyse according to the
equation —

x GHAc + y HOH = (GHOH)j, (GHAc)x-y + y HAc.

Where Ac represents an acid ion. As y increases, the protein
becomes more and more basic^ and the negative ion increases in size,
the change being indicated by the alteration in the appearance of the
solution, which becomes gradually more and more opalescent until it
is finally nearly opaque. It has still more or less the properties of a
true solution, the opacity being due to the formation of what Hardy
calls pseudo-ions, which can still take part in the transport of electricity,
and which, on increasing hydrolysis, become so large as to be capable
of diffracting light. A basic salt is finally obtained containing a very
small amount of acid.

The above considerations have an important bearing upon the
choice of constants to be chosen for the characterisation of proteins,
for it will be obviously impossible to directly determine in most cases
whether a given solution contains a free protein or a salt of the same.
Indicators, as already stated, are useless for the purpose, and owing to
the possibility of the formation of acid or basic salts very minute
quantities of an acid or base will very often suffice to entirely alter
the character of a solution. There exists, therefore, very great

6 THE GENERAL CHARACTERS OF THE PROTEINS

»

difficulty in obtaining reliable physical constants for proteins, for the
physical differences between free bases and free acids, and the salts
obtainable therefrom, are, in the case of almost all physical constants,
very considerable. To indicate but one instance : Leucine as a

free base has an optical rotation of a = - 10*42°, whereas the
rotation of its hydrochloride is a, = + 15*3 3° in 20 per cent.

hydrochloric acid.

Electrolytic conductivity, osmotic pressure, and many other
physical properties are also profoundly affected by the above con-
siderations, and attention has been called in the following pages to
the circumstances which so often render physical determinations
unreliable.

There is still one other disturbing influence to be noted. Owing
to the amphoteric character of the proteins, and their capacity to form
salts with acids and bases, they can also conceivably form compounds
with neutral salts. Now colloids have a considerable power of ad-
sorbing other bodies, a circumstance which also, as already mentioned,
considerably affects the determinations of the acidic and basic func-
tions of a protein. It is seldom, if ever, that an ash-free protein
is obtainable, and it is impossible to determine whether the ash
represents inorganic substance in actual combination, e.g., sodium in
the form of a sodium salt, or whether it owes its origin to bodies
which have been physically adsorbed ; the mineral substances present
have, under any circumstances, a great influence on the physical
properties of the protein.

The above considerations render the majority of the physical
properties extremely unreliable for the characterisation of the
proteins ; the precipitability by salts is, alone amongst the properties
generally determined, but slightly affected by them.

Chemical Characteristics.

It is upon the chemical properties, therefore, that reliance must
be chiefly placed for obtaining constants for the characterisation of
proteins.

The most reliable of these are undoubtedly the numbers represent-
ing the distribution of nitrogen in the molecule, and generally known
as the " Hausmann numbers ". These indicate the relative propor-
tions of nitrogen in the molecule combined in the form of amide,
of monoamino acids, and of basic bodies.1 They are discussed in
detail on pp. 33-36.

The number of " active " 2 amino groups also varies in the
different proteins ; to determine these Dr. Horace Brown has
suggested the determination of the " amino-index " (p. 67). Another
suggestion for the determination of this factor is due to Messrs.
Cross, Bevan and Briggs (p. 65). The capacity also for forming
halogen derivatives varies considerably in the different proteins ; the
halogen numbers should serve also as a method of characterisation.
Furthermore, different proteins yield different quantities of hydrolysis.

1 See p. 4, ? Footnote, p. 4.

INTRODUCTION 7

products, which can be readily estimated quantitatively without the
employment of the elaborate esterification method. Amongst these
cystine can be estimated by determination of the sulphur content of
the protein, and tyrosine by the method suggested by Millar (p. 38).
The cystine and tyrosine factors should therefore be of value.

Much work remains to be done in the elaboration of quantitative
reactions which might serve for the characterisation of the proteins.
It is not necessary that these reactions should lead always to the
formation of definite chemical compounds ; it would be sufficient if
standard conditions could be chosen, under which a definite amount
of chemical reaction should take place. Similar methods have been
employed in the case of the fats, as an instance of which may be
cited the so-called Hiibl iodine number. This indicates approxi-
mately the number of double bonds in a given fat, but it is generally
admitted that the addition of iodine at the double bonds is accom-
panied by a certain amount of substitution. Under defined conditions,
however, the amount of iodine absorbed is a constant for each
particular fat.

Similar reactions are required for the identification of proteins,
and the various suggestions which have been made in this respect
have received what will be, it is hoped, adequate consideration in the
following pages.

PART I.
THE PHYSICAL PROPERTIES OF THE PROTEINS.

SECTION I.— THE SOLUBILITY OF PROTEINS IN SALT SOLUTIONS ;
THE " SALTING OUT " FROM SOLUTIONS.

THE method which has been most commonly employed hitherto, for
the separation of proteins from one another and from other bodies
is that commonly known as "salting out". It was noticed as long
ago as 1853 by Panum that dry sodium chloride does not precipitate
egg-white in the cold, whereas blood-serum gives a distinct precipi-
tate which redissolves in distilled water. Claude Bernard had also
observed that pancreatic juice yielded a precipitate with magnesium
sulphate; Robin and Verdeil (1853) made a similar observation
with ascitic fluid. The latter investigators noticed also that mag-
nesium sulphate gave a precipitate with egg-white and with serum,
the filtrate from which coagulated on heating, whereas the filtrate
from a similar precipitate with pancreatic juice did not.

Virchow, in the following year, carried out further investigations
on this subject, and found that, besides magnesium sulphate, other
salts, such as potassium sulphate, sodium sulphate, alum, calcium
chloride, sodium chloride, possess the property of precipitating
proteins ; he also made the fundamental observation that the
precipitates formed by salts redissolved in water, and that the
solutions thus obtained behaved as true protein solutions. He
confirmed also the French observers, in noticing that the salt does
not always precipitate the whole of the protein, and he assumed that
the latter must exist in solution in different conditions. At this
time the existence of several kinds of protein was unknown, and
Virchow did not recognise the fact that many such bodies could
exist, and that their different behaviour towards salt solutions was a
factor which could be employed for a partial separation, at any rate,
of one protein from another.

De"nis l first clearly discriminated between proteins in this respect.
He noticed that some are readily soluble in salt solutions, which are
insoluble in water. He found, for example, that certain proteins are
soluble in 10 per cent sodium chloride, potassium nitrate, and
sodium sulphate solutions at 30° C., from which they could be
reprecipitated on dilution with water ; he studied also the precipita-
tion of proteins from blood-serum by saturation with neutral salts,
employing for this purpose sulphate of sodium, chloride of sodium,
and sulphate of magnesium. Denis must be regarded, therefore, as

1 D^nis had commenced his investigations on the" proteins in 1835. His mono-
graphs on the subject were published in 1856 and 1859.

8

THE PHYSICAL PROPERTIES OF THE PROTEINS 9

the pioneer in the systematic separation of the proteins by the
method of " salting out ".

Denis included in his investigations not only proteins of animal
origin, but also bodies of a similar nature derived from plants. To
those proteins which are soluble in 10 per cent, sodium chloride
solution, from which they can be precipitated on dilution (especially
after saturation of the diluted solution with CO2), Hoppe-Seyler
gave the name of " globulins ". They formed the subject of an
investigation carried out in Hoppe-Seyler's laboratory by his pupil
Weyl (1877), by whom the general reactions were clearly defined.
To this class belong vitellin, the protein derived from egg-yolk, the
protein first isolated from striated muscle by Kiihne, and called by
him myosin, the serum-globulin, and various plant proteins. Weyl
divided the globulins into two classes, vtz.t those which are soluble
in sodium chloride of high concentration (". vitellins ") and those
which can be precipitated from the 10 per cent, solution by saturation .
with more salt (" myosins "J.1

In addition to this work from Hoppe-Seyler's laboratory,
numerous other researches on the separations of proteins by salt
precipitation appeared between the years 1865 and 1885.

Hammarsten investigated the precipitation of serum by sodium
chloride and magnesium sulphate, which only precipitate a fraction
of the proteins ; his investigations were supplemented by those of
Heynsius,2 who some years later (1884) called attention to the great
precipitating power of ammonium sulphate ; in this respect it is far
superior to the corresponding sodium salt, which Starke (in Ham-
marsten's laboratory) and independently, Schafer, had a short time
previously shown possessed the power of precipitating proteins in the
filtrate from the magnesium sulphate precipitate.

As a result of these various investigations, two classes of pre-
cipitating salts were introduced into general use for the separation of
proteins, viz., magnesium sulphate and sodium chloride, which
precipitate the so-called " globulin " fraction, and ammonium sul-
phate, which precipitates from the filtrate the " albumin " fraction.

By the systematic application of these salts by Halliburton to the
separation of natural proteins of serum, muscle and milk, much
light was thrown on the constituents of animal tissues and fluids.

Halliburton showed that serum could be separated into its
constituents by salts other than those already mentioned ; he found,
for example, that sodium nitrate, sodium acetate and sodium carbonate
could also precipitate serum-globulin, whilst potassium acetate and
potassium phosphate precipitated from serum all the heat-coagulable
proteins. He also applied the method of salt precipitation to
investigations on muscle-plasma and milk. He concluded that in
the former several proteins existed, whereas in the latter only two
could be recognised, viz., casein (or, as it is now termed, caseinogen)
and lact-albumin.

Another application of the salt precipitation about this period
deserves to be mentioned. Wenz showed, in 1886, that peptones
were not precipitated by ammonium sulphate, and Kiihne and
Chittenden utilised this observation for separating the proteoses

1 The present classification of the proteins is given in Dr. Plimmer's monograph.

2 Ammonium sulphate had also been previously employed by Me"hu.

2 *

to THE GENERAL CHARACTERS OF THE PROTEINS

from the peptones, both classes of bodies being intermediate
degradation products of natural proteins produced by means of
digestive enzymes.

Owing to the unsatisfactory technique of using several salts,
Kauder, in 1886, for the first time carried out a fractional precipitation
of protein bodies, employing only one salt, viz.) ammonium sul-
phate. This research was undertaken under the direction of Hof-
meister, who, with his pupils, made valuable contributions during the
course of the next few years to the technique and knowledge of salt
precipitation.

Kauder showed that the globulin precipitation in a serum solution
commenced when ammonium sulphate is present to the extent of
24-29 per cent, of complete saturation, and ended when the saturation
amounted to 36-46 per cent.1 On filtering off the precipitate thus
produced and adding fresh quantities of ammonium sulphate, no
further precipitation took place until the degree of saturation
reached about 64 per cent. At this point the albumin fraction
commenced to be precipitated ; this precipitation was completed
when the degree of saturation reached 90 per cent. The point at
which precipitation of the various fractions commences and ends
varies somewhat ; it depends to a slight extent on the concentration
of the serum solution employed and its reaction.

A series of systematic investigations was also carried out in
Hofmeister's laboratory with the object of determining the relative
value of various salt solutions as precipitants. The following
determinations are due to Lewith, who worked with the serum
proteins : —

IV.

Albumin Precipitation.
Begins. Ends.

33-6 47-2

I. II.

III.

Strength of

Salt. Protein

Globulin

Precipitation.

Solution.

Per Cent.

Begins.

Ends.

Na2SO4 .
(NH4)2S04

0-98
0-99

11-4
14-2

2yi

Na Acetate

2-26

14-6

—

M

0-98

15-0

—

MgS04.

0-98

16-9

257

K Acetate

2-26

17-6

35*2

M

0-98

22-8

Not estimate*

NaCl .J~

1-66

21-8

—

KCI ....',:

1-04

25*9

—

NaN03 : .

0-98

46-7

—

NaN03

2-26

43-4

—

64-6
60-8

More than 82*2
88-1

The numbers in Columns III. and IV. express the amount of salt in grams per 100
c.c. of liquid. The strength of the protein indicates the percentage in the mixture after
addition of the salt solution.

Some of the salts in the above table, such as ammonium
sulphate and potassium acetate, produce complete precipitation of
all the proteins in the serum solution; others, such as sodium
chloride, potassium chloride and sodium nitrate, do not produce
complete precipitation of even the globulin fraction, when present to
the extent of complete saturation.

Hofmeister investigated the relative influence of the acid and the
metal of the salt on the power of precipitating proteins. His results
are tabulated below. The numbers indicate the number of grams

1 For experimental methods, see the end of this section.

THE PHYSICAL PROPERTIES OF THE PROTEINS 11

in 100 c.c. of solution when precipitation of globulin commences in
a solution of egg-albumin, where the protein is present to the extent
of 2 grams in 100 c.c. after admixture with the salt solution.

Sulphate

Lithium. Sodium.
8-61 n'39

Potassium
No pp.

Ammonium.
I3'39

Magnesium.
I5-93

Phosphate .

Not investigated 11-69

I3'99

16-57

Slightly soluble

Acetate

13-83

16-38

No pp.

No pp.

Citrate

14*42

17-07

21-99

Not investigated

Tartrate

15-11

17-08

25-05

,,

Bicarbonate .

No pp.

25-37

Not investigated

»i

Chromate

21*22

25-59

No pp.

,,

Chloride

Changes proteins 21*21

26-28

„

No pp.

Nitrate
Chlorate

Not investigated 46*10
58-82

No pp.

ii

Not investigated

ii

Not investigated

It will be noticed from the above table that the capacity for pre-
cipitation depends on both the acid and the metal of a salt.

• r ***«• • /*

The

2-03 N.
(NH4)aS04

2-51-2-72 N.
MgS04
(NH4)2H(P04)3
Amm. citrate
Amm. tartrate
NaHCO3
Na3CrO4
K2CrO4

3-53-3-63 N
NaCl
KC1

5-42-5-62 N.
NaNO3
NaClO3

metals show a decreasing power of precipitation, passing from left to
right of the table, whereas the acids show a decreasing power, passing
from the top to the bottom. The same result is obvious when the
numbers are expressed in terms of factors of normal solutions.

FACTORS OF NORMAL SOLUTIONS NECESSARY TO START PRECIPITATION OF

GLOBULINS.

1-5-1-69 N.

Li2SO4

Na,,SO4
Na.,HPO4
K.,HPO4
CH3CO2Ka
CH3CO2N

K citrate
Na citrate
K tartrate
Na tartrate

Some salts fail entirely to precipitate proteins owing to their small
solubility in water. Thus, according to the above tables, on the as-
sumption that the capacity for precipitation depends both on the
acid and the metal, some concentration between 11*39 and 13*39 Per
cent, of complete saturation by potassium sulphate should precipitate
globulin. Potassium sulphate dissolves, however, only to the extent
of 1074 parts in 100 parts of water at 15'. Similar remarks apply
to potassium nitrate, potassium bicarbonate and potassium chlorate.

The application of the salt precipitation method to the separation
of the products obtained by the digestion of proteins by proteoclastic
ferments by Kuhne and his pupils, Chittenden and Neumeister, has
been mentioned already. They distinguished between "primary"
proteoses, which are precipitable by sodium chloride or magnesium
sulphate, " secondary " proteoses, which are precipitable in the filtrate
from the primary proteoses by ammonium sulphate, and " peptones,"
which cannot be precipitated by salts. A revision of this work was
undertaken in Hofmeister's laboratory by his pupils, Pick and Zunz,
who used for the purpose of separation of fractions varying concen-
trations of one salt (cf. Kauder's researches, p. 10). The former used
ammonium sulphate, and the latter zinc sulphate, a salt which had
been previously employed for protein precipitation by Bomer.

Pick, by half-saturation of Witte's peptone solution with am-
monium sulphate, separated a so-called hetero- and proto-proteose
fraction. From the filtrate containing the " secondary " proteose

12 THE GENERAL CHARACTERS OF THE PROTEINS

three further fractions were obtained, viz. : (i) " A " fraction, which is
completely thrown out by 62 per cent, of complete saturation, and
which is characterised by the relatively large amount of sulphur ; (ii)
" B " fraction, which is characterised by the intensity with which it
gives the furfurol or Molisch-Udransky reaction (see p. 32), and which
is precipitated by complete saturation in neutral solution ; and (iii)
" C " fraction, which is precipitated from the filtrate from " B " by
the addition of sulphuric acid solution saturated with ammonium
sulphate. By repeated solution and reprecipitation, and treatment of
the precipitates with alcohol, Pick succeeded in obtaining fractions
which differed from one another, not only in their solubility and pre-
cipitability, but also in their elementary chemical composition.

Similar results have been obtained by Zunz with the use of zinc
sulphate.

Another method of " salting out " has been suggested by Pinkus,
who recommends the saturation of the solutions with anhydrous
sodium sulphate at 37° C.

The fractionation method, as employed by Pick and Zunz, has
recently been subjected to a critical examination by Haslam.

It can hardly be contended that a process of the nature of frac-
tional precipitation can readily produce a complete separation of
fractions. Two sources of error are manifestly a priori conceivable,
viz. : (i) any fraction produced by a particular degree of saturation is
not absolutely insoluble in a solution of the strength in which it is
precipitated ; (ii) a precipitate may carry down with it certain quanti-
ties of substance which belong to a subsequent fraction. In a separa-
tion of globulin from albumin, for example, the precipitate produced
by half-saturation with ammonium sulphate will consist principally
of globulin, but it will contain certain quantities of albumin ; the fil-
trate, on the other hand, will contain mainly albumin, but it will not
be entirely free from globulin, which is not absolutely insoluble in
half-saturated ammonium sulphate solution. Two methods have been
suggested by Haslam for testing the purity of a fraction.

I. Where it is sought to prove that a protein precipitate is free
from the substances of the filtrate.

The precipitate is dissolved in water and the whole made up to
a given volume ; the amount of salt is added which is necessary to
produce the requisite degree of saturation ; the mixture after stand-
ing for twenty-four hours is filtered. If the filtrate contain no
protein, or other substance from which the precipitate is to be freed,
the requisite proof is furnished. If not, the organic nitrogen is
estimated by Kjehldahl's method. The precipitate is then re-
dissolved, the solution made up to the volume in which the original
precipitation was performed, the same amount of salt is added, and
the mixture is allowed to stand twenty-four hours, and finally filtered.
The organic nitrogen in the filtrate is estimated. This process of
precipitation at constant volume is repeated until the nitrogen in the
filtrate is a constant, i.e., until only that amount of protein is
present which is dissolved from the precipitate by the given volume
of the solution of the degree of salt saturation in which the precipita-
tion is carried out. This process of precipitation at constant volume
must generally be repeated several times before a homogeneous
precipitate is obtained.

THE PHYSICAL PROPERTIES OF THE PROTEINS 13

II. When the substance it is sought to purify is in the filtrate
(e.g., an albumin contaminated with globulin).

In this case a process of fractional precipitation must be per-
formed. By " fractional precipitation " a different process is implied
to that suggested by Pick. The method employed by Haslam may
be best understood by giving a concrete example of its mode of
application, viz., the preparation of secondary or deutero-proteose.
The primary proteoses in Witte's peptone are first separated by half-
saturation with ammonium sulphate ; from the filtrate the secondary
proteoses are precipitated by adding ammonium sulphate to complete
saturation. This precipitate is then dissolved in water, and the
solution half-saturated with the sulphate; a smaller quantity of
primary proteose is precipitated than that obtained in the first pre-
cipitation ; from the filtrate the secondary proteoses are precipitated
again by complete saturation with the salt. These processes are
repeated until half-saturation with the sulphate no longer produces
a precipitate. Even now, the "secondary" proteose is not quite
free from the primary. To the half-saturated solution, which should
contain about 2 per cent, of proteoses, saturated salt solution is added
until a small precipitate appears. Practice will enable the operator
to judge how much substance it is best to precipitate at each " frac-
tionation". The fraction is then filtered off, dissolved in water so
as to make approximately a 2 per cent, solution, and to this is added
an equal volume of saturated salt (i.e., ammonium sulphate) solution.
A precipitate of primary proteoses (z>., the substance which it is
desired to separate off) will be produced ; this is filtered off, and the
filtrate is returned to the main solution. A second " fraction ' ' is then
taken from this by partial precipitation ; this fraction is dissolved in
water (to make 2 per cent, solution approximately), diluted with an
equal volume of ammonium sulphate (another precipitation of primary
proteoses) and the filtrate therefrom returned to the main solution.
This process is repeated, a small quantity of primary proteose being
removed each time from the solution, until a " fraction " no longer
gives a precipitate on half-saturation. The main solution is then
completely saturated with salt, and a precipitate thereby obtained
which consists of a nearly pure deutero-proteose.

It will be seen from the above descriptions that the process of
obtaining by salt precipitation a protein of constant composition is
an extremely tedious one, and it is highly -probable that most of
the proteins obtained by earlier investigators by the method of frac-
tional precipitation have been impure.

It may be remarked here that the methods employed by Haslam
do not apply only to precipitation by salt solutions ; fractional pre-
cipitation by alcohol of different strengths may be carried out in a
quite analogous way.

Little detail has been given above concerning the fractions obtained
by different observers from the products of digestion of proteins, such
as Witte's peptone ; this subject, it is to be hoped, will be treated more
fully in a monograph on digestion. Neither has anything been said
on the physical processes involved in the method of " salting out " ;
this, again, is foreign to the scope of this article, and should be treated
in the monograph which deals with the physics of colloidal solutions.

14 THE GENERAL CHARACTERS OF THE PROTEINS

Method of Fractional Precipitation.

The limits of precipitability may be determined in the following
way. To 2 c.c. of a protein solution in a series of test-tubes are
added 8 c.c. of liquid containing varying quantities of distilled water
and saturated salt solution. The limits of incipient precipitation
can be readily observed. [Pick found in working with Witte's
peptone and ammonium sulphate solution that precipitation com-
menced when a mixture of 2*4 c.c. saturated ammonium sulphate
and 5 '6 c.c. water were added to 2 c.c. of a peptone solution. Such
a solution contains 24 per cent, of the salt necessary for complete
saturation of the whole 10 c.c. of liquid. In this case the initial
limit of precipitation may be expressed by the number 24.] The
contents of the other test-tubes containing a larger quantity of am-
monium sulphate are examined by filtering off the precipitate and
adding one or two drops of the salt solution to the filtrate ; as soon
as this addition no longer produces a precipitate, the higher limit of
precipitation is reached. In dealing with a mixture, a second pre-
cipitation often commences with the addition of larger quantities of
salts. This can be readily observed in the series of tubes which
contain a higher concentration, and the point is noted when the
addition of a drop of saturated salt solution to a filtrate produces a
precipitate again. This point is the lower precipitation limit of a
second fraction. The higher precipitation limit should also m this
case be noted. A third fraction can also be obtained and its precipita-
tion limits determined. [Pick showed that the first fraction, which
commenced to fall out when the liquid was 24 per cent, saturated (see
above), was. completely precipitated with 42 per cent, saturation,
i.e., the filtrate from the mixture 2 c.c. peptone, 4'2 c.c. ammonium
sulphate solution, and 3*8 c.c. water just failed to give a precipitate
when a drop of saturated salt solution was added ; when, however,
the 54 per cent, limit was reached, the filtrate just commenced to
give a precipitate with a drop of saturated salt solution. By the
method three fractions altogether were obtained, viz.y the 24-42 per
cent, fraction, the 54-62 per cent, fraction, and the 70-100 per cent,
fraction.]

A pure protein gives reliable precipitation constants ; attention
must be called, however, to the criticisms of Haslam already outlined
above on the application of the salting-out process to the separation
of mixtures; as already noted, it is a tedious process to obtain a
fraction of constant composition by the method.

Proteins, which are insoluble in water, but soluble in salt solu-
tions, have also their precipitation constants. Osborne, in his work
on the plant globulins, extracted his raw material with 10 per cent,
ammonium sulphate solution ; he then determined the precipitation
limits by increasing the concentration of this solution. The con-
stants were required to discover whether globulins derived from
different plants were identical or not.

In determining the precipitation limits with animal liquids, such
as serum, it must be remembered that these already contain salts ;
in fact, the globulin can be separated from serum by dialysing these
salts away.

THE PHYSICAL PROPERTIES OF THE PROTEINS 15

As an example of a combined method of fractional precipitation
by both alcohol and a salt, the following, for the separation of five
constituents of Witte's peptone according to Haslam, may be quoted.
To a 2 per cent, solution of Witte's peptone add an equal volume of
alcohol and allow the mixture to stand overnight, or at least for
some hours. Filter, collect the precipitate and swim it out on water.
The insoluble portion is crude hetero-proteose and may be filtered
off or obtained by decantation. The soluble portion contains a-
proto- and a-deutero-proteose ; on half-saturation of the aqueous
solution of the mixture with ammonium sulphate the former can
be precipitated ; the latter comes down from the filtrate on complete
saturation. The /3-proto- and /3-deutero-proteose remain in the first
alcoholic filtrate. If the alcohol be evaporated off at a low tempera-
ture and the solution made up to its original volume, and the liquid
thus obtained be half-saturated with ammonium sulphate, /2-proto-
proteose is precipitated ; by completely saturating the filtrate the
/3-deutero-proteose can be obtained. By performing the precipita-
tions twice purer products can be obtained.

SECTION II. — THE DEGREE OF SOLUBILITY OF PROTEINS IN
SALT SOLUTIONS.

In experiments on " salting out," the precipitation limits have
almost always been determined in solutions containing but a small
amount of protein ; even when so much salt has been added that no
further precipitation takes place, the solution may still contain a
certain amount of protein, owing to the fact that the latter is not
absolutely insoluble. But little work has been done so far in the
determination of this degree of solubility ; the most instructive is
that, perhaps, of Osborne and Harris on the solubility of plant
globulins in salts of various concentrations ; in this case the results
are of interest owing to the fact that the protein itself is insoluble in
pure water. These investigators found that there were two classes
of globulin solvents, viz. : (i) those in which a relatively considerable
quantity of salt must be present before notable quantities of globulin
commence to dissolve, and from which it is precipitated by dilution
with much water, or by the addition of small quantities of strong
acids, and (ii) those in which solution is brought about by low
concentrations of salt from which the protein is not precipitated
by dilution with water or by acids. The first class includes solutions
of the salts of strong bases with strong acids, the second the solution
of salts of weak bases.

The degree of solution is considerably influenced in the case
of the globulin edestin by the presence of small quantities of
acids and bases ; this subject will, however, be discussed in greater
detail later in considering the action of acids and bases on pro-
teins. The behaviour of edestin towards acetate solutions is ano-
malous. Although it is insoluble in solutions of potassium, sodium
and ammonium acetate, it is soluble in the acetates of barium,
strontium, calcium and magnesium ; the solubility in these ace-
tates is in the order of their molecular weights. In solutions of
silver, copper and lead acetates it is nearly as soluble as in solutions
of free acids of corresponding concentration ; the positive ion enters

16 THE GENERAL CHARACTERS OF THE PROTEINS

into combination with the edestin and no longer remains a free ion
in solution. Acetates of zinc and mercury have no solvent action.

The solubility of serum-globulin in salt solutions has been in-
vestigated by J. Mellanby. He found that the amount of globulin
dissolved by a neutral salt depends upon the percentage of protein
present and not upon the total quantity of salt. If a suspension of
globulin be treated with a salt solution of given percentage, the
amount of the protein dissolved will depend upon the strength of
the suspension ; more protein, for example, will be dissolved from a
5 per cent, suspension treated with a given quantity of salt solution
than would be dissolved if the same quantity of a 2 per cent, sus-
pension be treated in the same way. With regard to the solution
capacity of various salts, he found that neutral salts with monovalent
ions have the same efficiency ; those composed of a monovalent
positive ion and a divalent negative ion, or of a divalent positive
and monovalent negative ion have also the same efficiency as one
another. The salts of this latter class have a greater efficiency than
those of the former. Mellanby concludes that the solution of glob-
ulins by neutral salts depends upon the forces exerted by the ions ;
that monovalent ions, whether negative or positive, are equally effective
in producing solutions ; that divalent ions, whether negative or posi-
tive, are also equally effective, but more effective than monovalent
ions.

Another view as to the solution of globulins by salts is that a
molecular compound of the type GBS (G = globulin, B = base,
S = acid) is formed, which is readily hydrolysed by water, with the
liberation of insoluble globulin : —

GBS + H2O = GHOH + BS.

Such molecular compounds would be stable, therefore, only in pre-
sence of a large excess of the salt (Hardy).

In the present state of knowledge but little can be said with
certainty as to the nature of the solution of those proteins in salt
solutions, which are insoluble in pure water.

SECTION III. — SOLUBILITY OF PROTEINS IN ORGANIC SOLVENTS.

Some peptones and proteoses, as has been already mentioned,
are soluble in alcohol. Certain proteins of plant origin are insoluble
in water and absolute alcohol, but soluble in dilute alcohol. Bodies
of this class have been subjected to an exhaustive examination
chiefly by Ritthausen, and by Osborne and his co-workers. A
typical protein of this class is the gliadin obtained from wheat.
These bodies are generally soluble in some other organic solvents ;
zein, for example, the alcohol-soluble protein from maize, will dissolve
in glacial acetic acid, in crystallised phenol on warming and in
glycerol. In water and absolute alcohol it is perfectly insoluble ; it
is most soluble in alcohol of 85-95 per cent., and is but little soluble
in alcohol of less than 50 per cent.

Mayer and Terroine have recently made a curious observation
with reference to the alcohol solubility of proteins. They have shown
that certain proteins which have been precipitated from a dialysed
aqueous solution by alcohol acquire the property of redissolving in

THE PHYSICAL PROPERTIES OF THE PROTEINS 17

strengths of even 80-85 per cent, on the addition of small quantities
of salts. The albumin of horse serum can be redissolved under
these circumstances by the addition of acids, bases or salts. If the
horse serum be not dialysed before precipitation with alcohol, re-
solution in alcohol only takes place on the addition of acid or
fairly strong alkalies. The dissolved serum-albumin does not lose
its original properties. Egg-albumin, even when dialysed, acquires
the property of being redissolved in alcohol wholly, by addition of
bases, partially, by addition of acids, and not at all, by addition of
salts. Nothing is known of the causes of these phenomena.

Many proteins are also soluble in urea solutions (Ramsden) and
also in organic bases. This solubility is due probably to their acidic
nature. Globulins, caseinogen, acid- and alkali-albumin and even
heat-coagulated proteins dissolve in a saturated aqueous solution of
pure urea. Dry gelatin dissolves at room temperature till a 40 per
cent, solution is obtained. If the urea be removed by dialysis, the
gelatin sets to a jelly again. The presence of urea, furthermore, pre-
vents coagulation of solutions.

SECTION IV. — SEPARATION OF PROTEINS FROM SOLUTION BY
PRECIPITANTS OTHER THAN SALTS.

Michaelis and Rona have recently suggested a new precipitant for
proteins. When an alcoholic solution of gum-mastic is added to
water an emulsion is formed, which, on addition of salts, is de-emul-
sified with the formation of a flocculent precipitate. If proteins be
present in the emulsion colloidal particles will be obtained, consisting
both of the mastic and the protein. If the former be in sufficient
excess the whole of the latter will be carried down, when sufficient
electrolyte (e.g., hydrochloric acid) has been added to de-emulsify the
solution. The action is irreversible, in that the protein cannot be
dissolved out from the precipitate by water. On treatment, however,
with organic solvents, the mastic can be dissolved ; if ether be used
as a solvent hardly any nitrogenous matter goes into solution ; with
chloroform or alcohol, on the other hand, very appreciable quantities
of nitrogenous matter are dissolved. The method has the disadvan-
tage when applied to the separation of many proteins, in that the
latter, when the attempt is made to recover them by means of ether,
are rendered insoluble (coagulated), and thus cannot be obtained in
their original form. The mastic can also partially precipitate pro-
teose ; in the case of a commercial peptone obtained from fibrin, Rona
and Michaelis found about 88 per cent, of the nitrogen in the filtrate,
the remainder being precipitated by the mastic. They suggest the
employment of the method for the separation of proteoses. In this
case the mastic can be separated from the precipitated proteose
without altering this substance. In solutions containing coagulable
protein, such as serum, the method may be of use in estimating the
non-protein constituents, such as urea, nitrogenous extractives and
sugar, which are not carried down by the mastic precipitate.

This method is, so far, new. Owing to the irreversibility it does
not seem as if it will be capable of such general application as the
salt-precipitation method for the isolation of proteins ; its chief use is
for freeing solutions from proteins for the subsequent examination of

i8 THE GENERAL CHARACTERS OF THE PROTEINS

>

substances not precipitated. Certain inorganic bodies which readily
adsorb proteins have also been suggested for the removal of the latter
from solution. These inorganic adsorbents have been investigated
by Landsteiner and Uhlirz, who found that the most effective sub-
stances were precipitated silicic acid, meerschaum and iron oxide.
Two grams of these powders, shaken with 20 c.c. of 0*25 per cent,
solutions of proteins, can remove the whole of the latter from solution.
The more readily a protein is precipitated by salts the more easily is
it adsorbed by these non-crystalline powders.

Method.

It has already been stated that the mastic emulsion should be in
large excess. In the case of liquids rich in protein, such as serum, a
preliminary precipitation with alcohol can be carried out, or the
mastic can be added in portions at a time. The following example
will illustrate the method : —

One volume of serum is mixed with three volumes of alcohol.
After separating the precipitate, one volume of a 50 per cent, solu-
tion of mastic in absolute alcohol is added and the mixture is then
diluted with water till the alcohol does not form more than 30 per
cent, of the total fluid. The liquid is then, just acidified with acetic
acid, and 10 per cent, magnesium sulphate solution is added, 10-15
c.c. being employed for each litre of the solution. The precipitated
mastic will then carry down the last traces of the protein remaining
in solution.

SECTION V.— CRYSTALLISATION OF PROTEINS.

Proteins, as colloidal bodies, were at one time thought to be
incapable of existing in crystalline form. Later investigations have
shown, however, that this is not the case, and the elaboration of
methods for the crystallisation of certain substances of this class must
be considered as a distinct advance in the chemical technique for the
preparation of pure substances.

Crystalline bodies of a protein character were first observed in
plants where they occur naturally. They were known as " aleurone "
grains, and were first noticed in gluten by Th. Hartig in 1850. Their
protein character was first demonstrated by Radlkofer in 1858. They
occur in the seeds of a large number of plants, such as the pumpkin,
in hemp seeds, in castor-oil seeds, and in Brazil nuts.

In some respects these aleurone grains differ from ordinary crys-
tals. They can, for example, imbibe water and swell, during which
process they lose to some extent their capacity for refracting light ;
they retain, however, their original contours. The increase in size, as
determined by accurate measurement by Schimper, is not the same
in every axis. Another peculiarity is that aleurone grains will par-
tially dissolve on treatment with glycerine, and an undissolved portion
will be obtained which still retains the original crystalline contour,
but which has nearly the same refractive index as water.

These peculiarities have caused the question to be raised as to
whether these bodies are true crystals.

Similar naturally occurring protein crystals have also been de-
scribed in animal organisms, notably in the intestinal epithelium of
meal-worms and in the eggs of certain fish and amphibia. These

THE PHYSICAL PROPERTIES OF THE PROTEINS 19

latter bodies have been designated " yolk-platelets," and have formed
the subject of an exhaustive investigation by Fremy and Valenciennes.
They are possibly lecitho-proteins (Walther). Finally may be men-
tioned the haemoglobin crystals, which are not observed normally
in the animal body, but are occasionally found in preparations, and
are sometimes formed as the result of post-mortem changes.

The first results in the artificial preparation of protein crystals were
obtained by Maschke (1859), wno evaporated a saturated solution of
aleurone grains from Brazil nuts and obtained tabular, hexagonal
crystals. In 1877 Schmiedeberg succeeded in recrystal Using aleu-
rone grains, which were also obtained from Brazil nuts. He dissolved
the bodies in water at 3O°-35° and precipitated them from this
solution by means of a current of carbonic acid gas. The precipitate
was then redissolved in water at 3O°-35° with the addition of an ex-
cess of magnesia ; on allowing this solution to evaporate at the same
temperature a certain number of large crystals were obtained, which
Schmiedeberg regarded as the magnesium compound of the protein.

The preparation of crystalline proteins from plants formed, during
the next two decades, the subject of a large number of memoirs, the
chief of which are due to Ritthausen, and to Osborne and his co-
workers. The plant globulins, to which class most of these crystalline
substances belong, are, it must be recalled, insoluble in water but
soluble in salt solutions ; by allowing the proteins to separate from these
solutions under suitable conditions the crystalline substances may be
obtained. The following method of preparation of edestin, due to
Osborne and modified by Leipziger, may serve as an example : —

One kilogram of hemp seed is ground, or pressed in an oil-press
(a Buchner press can be employed when available). The remainder of
the fat is then removed by extracting with light petroleum. When
free from this solvent the seeds are digested at 60° with I litre of
5 per cent, salt solution, and the mixture is kept continually stirred.
The liquid is then filtered off from the residue through calico and
allowed to cool. A precipitate forms and settles at the bottom of
the vessel. The supernatant liquid is then decanted off and the
precipitate washed by decantation with distilled water. It is then re-
dissolved in 500 c.c. of 5 per cent, salt solution, and the solution filtered
through a warm filter. On cooling beautiful crystals separate. These
are washed with cold 5 per cent, salt solution, distilled water, alcohol
and ether. Yield about 100 grams.

In the cases of plant crystals, substances have been obtained in the
laboratory in a crystalline form, which can exist in such form in
nature.

A further great advance was made in 1889, when Hofmeister
described the preparation of a crystalline albumin from white of egg ;
this was the first instance of the isolation in crystalline form of
a protein which was not known to exist in this form in nature.
Not long afterwards, Giirber and Michel succeeded in obtaining a
crystalline albumin from horse serum. Hofmeister mixed egg-white
with an equal volume of saturated ammonium sulphate ; the globulin
was thereby precipitated ; on allowing the filtrate from this to
evaporate slowly in an open basin, the albumin separated in the form
of so-called globulites or spherolites. These were redissolved in dilute
ammonium sulphate, and the solution was allowed to slowly evaporate

20 THE GENERAL CHARACTERS OF THE PROTEINS

i

as before. On a second recrystallisation, needles were obtained
mixed with the spherolites ; on repeating the crystallisation a suffi-
cient number of times, a product consisting entirely of needles was
obtained.

The original method of Hofmeister has been modified in various
ways. It has been shown that ammonium sulphate solution con-
taining protein becomes alkaline on standing ; Hopkins and Pinkus
have shown that the addition of acid facilitates very considerably
the process of crystallisation. The method as modified by these
investigators is the one now in general use and is carried out in the
following way : —

Egg-white is beaten to a froth (to break up the membranes) with
exactly its own bulk of ammonium sulphate solution. The mixture,
after standing overnight, or at least for a few hours, is filtered from
the precipitated protein. The filtrate is now measured. Ten per
cent, acetic acid (glacial acetic acid diluted to ten times its bulk) is
then very gradually added from a burette, until a well-marked
precipitate forms — a precipitate sufficient to make the mixture
actually milky in appearance, and not a mere opalescence for which
liberated gas bubbles might be mistaken. Trie actual amount of
acid required to produce such a precipitate will vary (chiefly because
of the varying loss of ammonia which occurs when the liquid has
previously stood in open vessels). The point corresponds roughly
to an incipient acidity of the liquid towards litmus, but the formation
of the precipitate forms of itself the best indicator. This stage
being reached, a measured quantity of the acid is now added, over
and above that required to produce the first precipitate, I c.c. being
added for each 100 c.c. of the filtered mixture as originally
measured. The whole contains, therefore, approximately i part per
thousand of free acid. The bulky precipitate thus produced is at
first amorphous, and if the mixture be occasionally shaken the
amorphous precipitate will give place to crystals within four or five
hours. To obtain the full yield, however, the material should stand
for twenty-four hours. The product thus obtained is already nearly
pure. 'On recrystallising once more from ammonium sulphate
(dissolving in water, and then carefully adding half-saturated
ammonium sulphate containing acetic acid in the proportion of i
per thousand, till a permanent precipitate forms, and then about
2 c.c. of ammonium sulphate per litre in excess of this) a perfectly
pure preparation is obtained.

Considerable difficulty has been experienced in obtaining serum-
albumin in a crystalline form. Formerly it was obtained almost
entirely from the blood of the horse, but even here the attempt to
obtain a crystalline preparation does not always meet with success.
In these cases Giirber's method was employed. The serum was
mixed with an equal volume of concentrated ammonium sulphate
solution ; then, to the filtrate from the precipitated globulin, ammo-
nium sulphate solution was added until there was an incipient
turbidity; on allowing the mixture to stand the serum-albumin
crystals separated. Inagaki has recently shown that crystalline
serum-albumin can be readily obtained by the Hopkins and Pinkus'
method, i.e., in the presence of free acid. Crystallisation also takes
place more readily at a somewhat higher temperature (35°-4O°). The

THE PHYSICAL PROPERTIES OF THE PROTEINS 21

crystals obtained are, according to Inagaki, compounds of protein
with sulphuric acid.

It is possible that the other crystalline proteins are also compounds
with acids. This subject will be reconsidered later, in discussing
the action of acids and bases on proteins.

A special case of protein crystallisation is that of haemoglobin.
This is a conjugated protein, i.e., it is a compound of a protein and a
chromatogenic group, and its power of crystallisation depends on the
presence of this group. Crystals can be obtained by the following
method, due chiefly to Zinoffsky and modified by Abderhalden.

The paste of red blood corpuscles (from a horse), after separating
from the serum and washing, is mixed with twice its volume of
water and the mixture is then warmed to 35°. The corpuscles are
by this treatment laked. A very small known quantity of ammonia
is then added to dissolve the stromata, and then hydrochloric acid
in very dilute solution is added in such quantity as to exactly
neutralise the ammonia. The mixture is then cooled to o°, and one-
quarter the volume of absolute alcohol is added. The whole is kept
on ice, and crystals slowly separate out, and settle at the bottom of the
vessel. The supernatant liquid is then poured off, and the crystals
are washed by decantation with a mixture of one part alcohol and
four parts of water previously cooled to o°. They can be recrys-
tallised by dissolving the crystalline paste in twice its volume of
water at 35°, then cooling to o°, adding one-quarter the volume of
alcohol, and allowing the crystals to form slowly at a low temperature.

It must be remembered that only a relatively small number of
proteins have as yet been obtained in crystalline form, and recrystalli-
sation as a mode of purification has, so far, not obtained a very wide
application. It is not known whether any chemical change takes
place during the process, and there is a certain amount of evidence
that in the case of crystallised egg-albumin, for instance, the cry-
stallised product is different from the original protein existing in
the egg-white (see " gold number " of crystallised and non-crystallised
egg-albumin, p. 27).

SECTION VI.— THE TEMPERATURE OF HEAT COAGULATION OF
PROTEIN SOLUTIONS.

In 1854 Kiihne noticed that there were two proteins present in
muscle-plasma which differed in the temperature of heat coagulation,
and since that time the coagulation point has been regarded as an
important factor for distinguishing between proteins, and the process
of separating proteins in mixtures by means of a fractional heat
coagulation has been repeatedly employed. In this operation a
solution of the mixture is heated to a definite temperature until
a coagulum is produced ; this is filtered off, and the filtrate heated
to a still higher temperature, when there is the formation of
another coagulum ; further fractions can be taken until no more
coagulable protein remains in solution. Thus Halliburton, by such a
process, separated in 1887 several proteins from muscle-plasma, which
he designated as follows : paramyosinogen, coagulating at 47° ; myo-
sinogen, coagulating at 56° ; myoglobin, coagulating at 63° ; all of which
can be precipitated from solution on saturation with sodium chloride

22 THE GENERAL CHARACTERS OF THE PROTEINS

and magnesium sulphate ; in addition to these an albumin coagulating
at 73° was obtained, which cannot be so precipitated.

Other coagulation points were determined by Fre"dericq.

It is not in the least probable that such a process as fractional
coagulation can lead to a complete separation of a mixture of pro-
teins ; it is, moreover, irreversible, and in this respect cannot have
such a general application as, e.g., the separation by means of salt
precipitation.

It remains to be considered, however, how far the coagulation
temperature may be regarded as a constant characteristic of any
individual protein. (NOTE. — All proteins do not form heat-coagulable
solutions.) It was long ago observed that the reaction of the liquid
and the presence of neutral salts exercised a marked influence on
the coagulation temperature. In the case of natural fluids, such as
muscle-plasma, serum, etc., which contain salts, it was noticed that
the coagulation took place more readily in a slightly acid solution ;
the presence of larger quantities of acids, however, inhibited the
coagulation, owing to the formation of acid albumin. In the experi-
ments of Halliburton, already referred to, the fractional coagulation
was carried out with solutions having a certain definite acidity.

The influence of salts was recognised by Aronstein (1874), who
showed that a solution of egg-white, from which inorganic matter
had been almost completely removed by dialysis, lost its coagula-
bility. Subsequent researches were undertaken by Alexander
Schmidt and Heynsius, and it was found that the protein solutions
of Aronstein still contained alkali, although the neutral salts had
dialysed away. Heynsius made the further observation that by
very long-continued dialysis, by means of which nearly all the alkali
is removed, a solution is obtained which recovers its coagulability ;
the coagulation in this solution is inhibited, however, by the pre-
sence of minute quantities of acids and alkalis. Similar results were
obtained by Winogradoff and Haas.

The capacity for coagulation of a solution depends, therefore,
both on its reaction and the quantity of neutral salts present; the
smaller the quantity of the latter, the smaller the amount of alkali
necessary to inhibit coagulation.

Careful quantitative studies on the influence of salts on the tem-
perature of coagulation have been undertaken by Starke and by
Pauli, the latter of whom has endeavoured by his researches to throw
some light on the chemico-physical process involved in the coagulation
of colloids.

Starke determined the amount of various neutral salts that were
necessary to restore the coagulability of a protein solution, which had
been deprived of this property by dialysis. He found that in the
case of the salts of alkaline earths and magnesium, a few milligrams
per 100 c.c. of solution sufficed, whereas with the alkaline salts I
gram per 100 c.c. was required. Starke also investigated the in-
fluence of varying quantities of different salts on the temperature of
coagulation. This question was the subject of a more exhaustive
investigation by Pauii. The following table, which is one out of a
large number to be found in Pauli's earlier paper, illustrates the in-
fluence of varying quantities of different salts on a certain solution
of egg-white. In all cases 2 c.c. of the egg-white solution were

THE PHYSICAL PROPERTIES OF THE PROTEINS 23

diluted to 10 c.c. ; the numbers in the first horizontal line indicate
the number of cubic centimetres of normal solutions of various salts
contained in the 10 c.c.

INFLUENCE OF VARIOUS CHLORIDES ON COAGULATION POINT.

c.c.

0'5

ro

i'5

2'O

2-5

3-o

3'5

4-0

4'5

5'0

5'5

6-0

NH4

57'2

58-2

59-i

60*0

60-3

60-5

607

6ro

K

60-9

6r6

61-9

62*2

—

—

—

—

—

—

—

—

Na

61-8

62-6

6V4

64-2

64-5

63-6

62-4

60-0

—

—

—

—

Li

62-0

64-6

68-2

7i'3

72-6

73*2

72*6

71-7

70-8

68-0

62-8

59'8

Ba

60-9

67-8

7r5

7I-5

—

—

—

—

—

Mg

60-9

70-8

75'2

75*2

76-2

77-8

78-2

78-8

*— *

~

~

It will be seen from the above table that the increase in the
quantity of salt raises the coagulation temperature up to a certain
point, after which, in some cases, further additions of salts cause no
increase, and in other cases even a diminution.

It is not necessary here to discuss the physical meaning of such
numbers as these ; enough has been said to indicate that the coagula-
tion temperature is a constant which is markedly influenced by a
variety of factors. The chief ones are, as Starke pointed out : (i) the
reaction of the solution, (ii) the amount of neutral salts present, in a
minor degree (iii) the concentration of the protein, and (iv) the rate at
which the solution is heated. Far less reliance can be placed on co-
agulation temperatures as a constant for identification of proteins than
on such a constant as the melting point of a crystalline substance,
especially when it is remembered that it is often difficult to exactly
regulate the amount of salt, acid or alkali present in any given
solution.1

Method (according to Pauli).

The liquid to be tested is placed in a test-tube, provided with a
stirrer, and in it is immersed a thermometer graduated in -^ or T2o de-
grees. The test-tube is immersed in water in a beaker of I litre
capacity, which is likewise provided with a thermometer and stirrer.
The whole is slowly heated with a small burner, the liquid in both
test-tube and beaker being carefully stirred, so that both vessels attain
the same temperature. When there is a black background and the
apparatus receives light from two sides the smallest turbidity is readily
recognised. The time a solution takes to coagulate should also be
noticed ; this factor bears no relation to the coagulation tempera-
ture.

1 Wolfgang Ostwald has recently represented the relation between the amount of
salt present and the coagulation temperature by the following equation — -^ = kcm where

t = temperature, c salt concentration, and k and m are constants. He draws attention to
the similarity between this and an adsorption equation. Pauli, in some very recent
work, comes to the conclusion that the coagulation point is influenced by the condensa-
tion of ions on the surface of the colloidal particles.

3

24 THE GENERAL CHARACTERS OF THE PROTEINS

TABLE.

COAGULATION TEMPERATURES OF SOME OF THE PRINCIPAL PROTEINS.

Substance.

Temperature.

Observer.

Remarks.

Bence-Jones' protein .

50-530

Magnus-Levy

Passes again into

solution on fur-

ther heating.

Caseinogen .
Crystallin (a)

94-100°
72°

Lacqueur&Sackur
Morner

Heated dry.

>» (ft)

63°

it

Egg-albumin

56°

Starke

Fibrinogen .

56°

Fredericq

Haemoglobin

64°

Preyer

Lact-globulin

72°

Hewlett

( Paramyosinogen

47°

Halliburton

Cf. von Fiirth.

Muscle proteins j Myosinogen

56°

n

(Myoglobulin

63°

|f

t . f Globulin .
Serum proteins {Albumin .

75°
67°

Hammarsten
Fredericq

Cf. Freund and
Joachim.

Vitellin ....

75°

Weyl

Observer.

SECTION VII. — OPTICAL ROTATION OF PROTEIN SOLUTIONS.

Solutions of proteins are optically active. The rotations, like
those of the amino-acids, vary according to the amount of acid
present in solution. The number of reliable determinations of the
optical rotation of protein solutions is small. Of special interest are
those of the vegetable proteins, in which the rotation was measured
under as nearly as possible identical conditions. Some of the chief
determinations are given in the following table : —

Protein.

Animal proteins : —

Egg-albumin
Haemoglobin
Globin

Nucleo-proteins
Crystallin (o-variety)

„ (j8-variety)
Plant proteins : —

Edestin (Hemp-seed)
Globulin (Flax-seed)

„ (Squash-seed)
Excelsin (Brazil-nut)
Amandin (Almond)
Corylin (Filbert)
Globulin (English-walnut)

„ (Black-walnut)
Phaseolin (Kidney-bean)
Legumin (Horse-bean)
Zein (Maize)
Gliadin (Wheat)

SECTION VIII.

Rotation.

MD- 3070
We + io*4\
[«]c-54'2/
Dextrorotatory
[o]D - 46-9
[ab-43'3

MD

-41-3

-43-53

-3873

- 42-94

-56-44

-43-09

-45-21

-44-43
-41-46

-44-09
-28-20
-92-28

Hopkins and Willcock.
Gamgee and Croft Hill.
Gamgee and Jones.

Osborne.

-MOLECULAR WEIGHT DETERMINATIONS BY
CRYOSCOPIC METHODS.

(A) Depression of Freezing Point.

Several data are to be found in the literature referring to the
depression of the freezing point of water caused by the solution of

THE PHYSICAL PROPERTIES OF THE PROTEINS 25

proteins. The determinations were made with a view to the estima-
tion of molecular weights. The data arrived at by this method are,
however, of little value. In the first place it is, as has been re-
peatedly stated, extremely difficult to free proteins from the last
traces of inorganic substances ; very small contaminations of such
substances with low molecular weight would cause relatively very
large errors in the determination of the molecular weight of such
complex substances as the proteins. In the second place, from the
direct determinations of osmotic pressures, discussed in greater detail
below, the depression in the freezing point of a salt-free protein
should be so small as to be well within the limits of experimental
errors. Moore has calculated that 'OOi° C. should correspond to an
osmotic pressure of 9 mm. of mercury. The smallest depression
that can be measured by a Beckmann thermometer is '005° C.,
which corresponds to an osmotic pressure of 45 mm. As, in prac-
tice, pressures of this order have to be measured in the investigation
of proteins, when estimating osmotic pressures directly, it is obvious
that but little value can be attached to determinations of freezing-
point depressions. Nevertheless, some of those determinations
made with peptones and other degradation products are of interest,
as substances of much lower molecular weight than the proteins from
which they have been derived are here being dealt with. It must be
remembered, however, that these peptones are not pure bodies. A
few data from the literature, with these reservations, are given below : —

Substance.
Egg-albumin

Protalbumose

Deutero-albumose

Albumose

Peptone (commercial)

Propeptone

Antipropeptone (albumose)

Gelatin

Peptone (Merck)

„ (Grubler)
Glutin-peptone hydrochloride

Molecular
Weight.

14,270

6400

2467-2

32CO

2400

1504-1754

I20I-I2I5

776-823

878-960

52Q-555

278-704 Paal.

Observer.

Sabanejevv and Alexandrow.
Bugarsky and Liebermann.
Sabanejew.

>•

Bugarsky and Liebermann.
Sabanejew.
Paal.

Ciamician and Zanetti.
ii

(B) The Direct Determination of Osmotic Pressure.

Considerable controversy has arisen as to whether colloids exert
any osmotic pressure in solution. Several investigations on this sub-
ject are, however, recorded in the literature. Starling, in the course
of some researches on the function of the glomeruli of the kidney,
measured the osmotic pressure exerted by a serum solution of
known protein content. A vessel was used the walls of which were
permeable to the salts but not to the proteins contained in the
serum ; for this purpose they were constructed of peritoneal mem-
brane of calf previously soaked in gelatin. On the other side of
the membrane was placed a fluid which possessed approximately
the same salt contents as the serum ; this was generally prepared by
freeing the serum experimented with from its protein by pressing it
through gelatin filters under a pressure of 30-40 atmospheres. By
having on one side of the membrane natural serum, and on the
other serum deprived of proteins by filtration, and by connecting
the protein containing liquid with a manometer, the osmotic pressure

3*

26 THE GENERAL CHARACTERS OF THE PROTEINS

due to the serum protein alone could be directly estimated, and the
errors due to the presence of salts thereby eliminated.

Moore has modified this method for estimating the osmotic pres-
sure of colloids. He employs a special form of osmometer, consist-
ing of two similar platinised capsules each of about 20 c.c. capacity,
each with a flange. By a special arrangement of rubber bands these
capsules can be screwed together. When in this position they are
separated by a thick platinum grid which passes between the flanges
and supports a parchment membrane. One capsule containing the
colloid is connected with the manometer, whilst the other contains
a liquid of which the non-colloid contents are approximately iso-
tonic with those of the solution containing the colloid. In this way
the errors due to the presence of the non-colloidal, dialysable sub-
stances, etc., are readily eliminated.

Various estimations of osmotic pressure have been carried out
with the use of apparatus of this character, chiefly by Weymouth
Reid, who used the Starling form of osmometer, and by Moore and
Parker, and Moore and Roaf.

Reid showed that proteins that have been well washed have
practically no osmotic pressure. He attributes the osmotic pressure
of protein solutions recorded by other observers to the adsorbed
inorganic substances. He found, however, for purified dogs' haemo-
globin an osmotic pressure varying from 3'63-4'35 mm. Hg per
I per cent, concentration. This is in accord with an observation of
Gamgee (published in abstract, Proc. R. S., 1902, 70, 79) that haemo-
globin possesses conductivity and is therefore in true solution.

Moore and his co-workers observed quite appreciable pressure
with different proteins, which varied with the salt contents of the
solution. Moore maintains, in opposition to Reid, that the proteins
exert a definite osmotic pressure, but that their state of aggrega-
tion varies in solutions containing different quantities of salt ; he
conceives that by repeated reprecipitation, washing, etc., the aggre-
gates become so large that they exert no osmotic pressure.

From the above short summary it is obvious that cryoscopic
methods, at any rate in the present state of our knowledge of
colloids, can bear but little value in fixing the character of individual
proteins.

SECTION IX.— THE ELECTROLYTIC CONDUCTIVITY OF PROTEIN

SOLUTIONS.

The difficulty of obtaining protein solutions free from electrolytes
with relatively high conductivity must be taken into account again
when measuring the electrolytic conductivity of protein solutions.

The method has been employed in a large number of cases, as an
adjunct to chemical methods, e.g.> in determining the capacity of
proteins to form salts with acids or bases, in determining the disso-
ciation constants of such salts, and also in investigating the rate of
hydrolysis of proteins by acids or enzymes. It will be necessary to
refer to the conductivity of protein solutions again later when discuss-
ing the action of alkalies and bases (p. 39).

THE PHYSICAL PROPERTIES OF THE PROTEINS 27

SECTION X. — THE " GOLD NUMBER ".

Zsigmondy has shown that colloids possess the property of in-
hibiting the transformation of bright red colloidal gold solutions into
the non-colloidal form, a change which can be brought about by the
addition of salts. The quantity of substance necessary to inhibit this
transformation varies greatly with different classes of colloids, and is
characteristic for each individual. Zsigmondy proposes, therefore,
to regard this quantity as a definite factor for the identification of
individual colloids. The "gold number" he defines as " the number
of milligrams of a colloid which are just insufficient to prevent 10 c.c.
of a bright red gold solution, prepared under certain specified condi-
tions, from changing into violet, or nuances of violet immediately, or
shortly after the addition of I c.c. of a 10 per cent, salt solution ".

He divides the colloids into four classes, viz. : —

Class I. Colloids with the gold number 0*005 to 0*1. This
includes gelatin, caseinogen, isinglass, animal glue.

Class II. Colloids with gold number 0*1 to 10, including crystal-
line egg-albumin, gum-arabic, tragacanth.

Class III. Colloids with gold number 10 to 500, including dextrin,
potato starch.

Class IV. Colloids which are inactive, including silicic acid,
soluble glass and mucin.

In the following table are given the gold numbers of certain
characteristic proteins : —

TABLE.

Substance.
Gelatin
Russian glue
Isinglass
Caseinogen
Glycoprotein
Deutero-albumose
Egg-globulin
Ovomucoid

Crystallised egg-albumin
Amorphous

Gold Number.

0-005-0-01
0-01-0-02

o-oi
0*05-0-1

00

0-02-0*05
0-04-0-08

2-8
0-03-0-06

(After separation of the crystalline.)
Merck's albumin 0-1-0-3

Fresh egg-white 0-08-0-15

It is of interest to note in the above table the high value for
crystallised egg-albumin.

Method.

Preparation of the Gold Solution. — One hundred and twenty c.c.
of water are distilled through a silver condensing tube into a beaker
of Jena glass of 300-500 c.c. capacity. They are then heated, and
during the warming 2-5 c.c. of a 0*6 per cent, solution of hydrogen
gold chloride and 3-3*5 c.c. of a solution of the purest potassium car-
bonate (O'i 8 normal) are added. After boiling, and whilst the mixture
is still hot, 3-5 c.c. of a dilute solution of formaldehyde (0*3 c.c.
commercial formalin in 100 c.c. H2O) are added. Only Jena glass
rods should be used for stirring. After a short time a bright red

28 THE GENERAL CHARACTERS OF THE PROTEINS
|

colour is developed. A solution thus prepared can be kept for a long
time without changing.

Determination of the " Gold Number". — Small quantities of the
colloid solution under investigation are introduced into a series of 50
c.c. beakers. The colloid should be measured out from a 0*2 c.c.
pipette graduated in thousandths of a cubic centimetre. The quan-
tities generally used are 0*005, O'OI, 0*02, 0*05, up to 0*5 c.c. Larger
quantities of solution are to be avoided. Five c.c. of the gold solution
are then introduced into each beaker, and the mixture is then rapidly
stirred. After three to five minutes 0*5 c.c. of sodium chloride solu-
tion (100 grams NaCl to 900 c.c. water) is introduced into each
beaker. By this method a higher limit can be observed in which
no change takes place, and a lower limit in which the red solution
is converted into violet. In this way the limits are determined. The
number of milligrams of colloid in each of these limits, multiplied
by 2, gives the interval which is designated the " gold number ". This
factor is the one generally determined (see table above).

SECTION XI. — THE FRACTIONAL FILTRATION OF PROTEINS.

In 1896 C. J. Martin described a gelatin filter, through which,
under pressure of 40-50 atmospheres, water and simpler substances,
such as sugar and salts, could be made to pass, whereas more complex
colloidal bodies, such as the proteins, were held back. The apparatus
employed consisted essentially of a Pasteur-Chamberland filter candle,
which acted as a support for a membrane of gelatin or silicic acid.
This was mounted in a gun-metal filter case, which was connected
with a steel gas cylinder containing air under the requisite pressure.
The liquid to be filtered was introduced into the filter case, which
was then connected with the compressed air cylinder. The water and
simple bodies commenced to filter through as soon as the requisite
plus-pressure was attained. This process of filtering served as a
means of concentrating protein solutions and at the same time freeing
them from contamination with simpler substances.

It has since been employed in a limited number of cases for
obtaining some information as to the relative sizes of molecular
aggregates in protein solutions. Thus, for example, W. A. Osborne
has shown that the sodium salt of caseinogen will pass through a
Martin filter, whereas the salt of the dibasic calcium, with approxi-
mately double the molecular weight, is held back. Craw has also
employed this filter for the separation of toxins and antitoxins.

The principle involved in the Martin filter has recently been
extended by Bechhold, who has devised a method of fractional
filtration. Instead of employing a filter candle for impregnation
with the filtering membrane he uses filter paper. This is soaked in
the impregnating membrane (either acetic acid solution of collodium
or an aqueous solution of gelatin) in vacuo in a specially devised
apparatus. In this way a relatively thin filter with a large surface
can be obtained. This is introduced into a suitably constructed
filter funnel, in which it is supported on a nickel gauze ; the funnel
is inserted in a metallic cylinder, which can be connected either with
compressed air or a force pump, so as to produce a plus-pressure on
one side of the filter ; a stirrer can also, if necessary, be inserted so

THE PHYSICAL PROPERTIES OF THE PROTEINS 29

that the surface of the filtering liquid in contact with the membrane
can be continually changed. With this apparatus comparatively
small plus-pressures only are necessary (from i to 5 atmospheres).

By employing impregnating membranes of different concentra-
tions Bechhold has succeeded in devising a method of differential
filtration. Thus, for example, a filter paper impregnated with 3 per
cent, gelatin, when tested with a I per cent haemoglobin, allowed
appreciable quantities of the solute to pass ; a similar filter impreg-
nated with 4 per cent, gelatin was not permeable to haemoglobin.

The more concentrated the impregnating solution the higher the
plus-pressure necessary for filtration.

The filters made from more concentrated solutions have neces-
sarily smaller pores ; such filters will allow the passage only of
substances in a state of comparatively simple aggregation ; they can
in this way be separated from the substances in which the state of
aggregation is more complex. An interesting example of the method
of employment of the filter is afforded by Bechhold in his description
of the attempt to separate the constituents of Witte's peptone. Pick
(see p. 14) has described the following fractions obtained with am-
monium sulphate : Protalbumose (24-42 per cent, saturation), deutero-
albumose A (54-62 per cent), deutero-albumose B (70-95 per cent),
deutero-albumose C (100 per cent + acid). A clear 5 per cent,
solution of Witte's peptone solution, which gave a precipitate on 23
per cent, saturation, was submitted to filtration under I atmosphere
plus-pressure through a 3 per cent gelatin filter ; a similar filter
impregnated with 2' 5 per cent, gelatin was just not permeable to
haemoglobin.1 The residue on the filter was twice diluted with water
and twice subjected to filtration with a similar filter. Precipitation
in the filtrate commenced only when the saturation with ammonium
sulphate reached 34 per cent., whereas the residue on the filter com-
menced to precipitate when the liquid was only 23 per cent saturated ;
hence a separation into two fractions, viz., one precipitable at 34 per
cent, saturation and one at 23 per cent. Other experiments of
similar nature gave analogous results. They are summarised below : —

Fraction and Precipitation Limits,

(NH4)2S04.
Protalbumose (24-42 per cent.)

B

Filter Used. Result.

3 per cent. Residue commenced to precipitate at

34 per cent, saturation.
Residue precipitated between 34 and

4 „ 95 per cent, saturation.

Filtrate precipitated between 95 and
, loo per cent.

Deutero-albumoses A

(54-95 per cent.) 4

Deutero-albumose C (100 per

cent. + acid) 4

Fractional filtration gives results, therefore, which are analogous
to those obtained by fractional precipitation with salts and may serve
as a valuable additional adjunct to this process. Here, again, how-
ever, complete separations are not to be expected ; for neither the
size of the aggregates in the solution to be filtered, nor the size of the
filter pores, are of uniform size, as Bechhold has experimentally
demonstrated. The method is still new, and promises to be a
valuable addition to the technique for dealing with proteins. It
might prove of great utility in conjunction with a tedious fractional
precipitation method, such as that employed by Haslam.

1 One per cent, hsemoglobin was used as a test solution for filter membranes ; the
minimum concentration necessary to produce a filter not permeable to haemoglobin in
this solution was determined. To express this the formula 3 per cent. (H 2-5 per cent.)
is employed.

PART II.

THE GENERAL CHEMICAL CHARACTERS OF
THE PROTEINS.

SECTION XII.— THE QUALITATIVE REACTIONS OF THE

PROTEINS.

PROTEINS may be precipitated from solutions by the following re-
agents : —

I. By various mineral acids. Nitric acid is very often employed.
If allowed to flow into a protein solution a white ring forms at the
junction of the acid and the solution. This delicate reaction is
generally known as Hellers test.

Metaphosphoric acid is a precipitant of the proteins, but not the
ortho- or pyro-acids.

II. Ferrocyanic acid is a good precipitant. A mixture of
potassium ferrocyanide and acetic acid is generally employed.

III. The salts of the heavy metals precipitate proteins. This
reaction will be discussed in greater detail later.

IV. The ordinary alkaloidal reagents precipitate proteins, e.g.,
phosphotungstic acid, phosphomolybdic acid, potassium mercuric
iodide, potassium bismuth iodide, tannic acid, picric acid. These
reagents should be added to the slightly acidified solution.

V. Trichloracetic and sulphonylsalicylic acids are good precipi-
tants.

VI. Also uranyl acetate. The precipitate thus produced is
soluble in acids.

VII. Nucleic acid and protamines.

The following are the chief colour reactions for proteins. Not
every protein gives a positive result when treated with the reagents,
as the various colours are due to certain specific groups contained in
the molecule, which are not common to all proteins. Nevertheless,
every protein will give positive reactions in a large number of the
tests.

I. Biuret Reaction. — The protein is first treated with sodium
hydroxide solution, and then copper sulphate in very dilute solution
is added drop by drop ; a reddish violet to violet-blue colour will be
produced. This is due, according to Schiff, to the presence of the
following groups : —

CO.NH2 /CO.NH2 CO-NH2 | |

NH/ CH2/ | NH2_C-CO~NH-C-

\CO.NH2 \CO.NH2 CO-NH2 |

II. The Xanthoproteic Reaction. — Proteins give on boiling with
strong nitric acid yellow flakes or a yellow solution, which, on

30

GENERAL CHEMICAL CHARACTERS OF PROTEINS 31

making alkaline, becomes orange yellow. This is probably due to
the presence of a radical containing the benzene ring.

III. Milloris Reagent. — This consists of a solution of mercury in
nitric acid, which contains nitrous acid. This reaction is due to the
presence of a tyrosine group.

IV. Colour Reactions due to the Presence of the Tryptophane
Group : —

(a) The chief of these is the Hopkins and Cole modification of

the so-called Adamkiewicz reaction. As originally pro-
posed by Adamkiewicz, the reaction consisted in treating
the protein solution with one volume of concentrated sul-
phuric acid and two volumes of glacial acetic acid ; a reddish
violet colour was thereby produced. Hopkins and Cole
have shown that this reaction is due to the presence in the
acetic acid of glyoxylic acid, produced by the oxidation of
the former body, a process which readily takes place on its
exposure to sunlight. They now use a solution of glyoxylic
acid itself, which is produced by the reducing action of
sodium amalgam on a solution of oxalic acid.

(b) ReichPs Reaction. — On mixing a protein with an alcoholic

solution of benzaldehyde and adding dilute sulphuric acid
(one volume of acid to one volume of water) and ferric
sulphate a blue coloration is produced.

(c) Rhode's Reaction. — A weak solution of dimethylaminobenz-

aldehyde is mixed with the protein solution, and concentrated
sulphuric acid is allowed to flow into the mixture. A reddish
violet colour is thereby produced which changes to dark violet.

(d) Liebermanns Reaction. — When a protein is boiled with

alcohol, then treated with ether, and then heated with con-
centrated hydrochloric acid, a blue solution is produced.
This is, according to Cole, due to glyoxylic acid, contained
as an impurity in the ether. If this is the case, the Lieber-
mann reaction is identical with that of Hopkins and Cole.

(e) According to Cole, proteins on treatment with furfurol and

hydrochloric acid yield a purple-red colour, which is also due
to tryptophane. Some proteins containing a carbohydrate
group will yield the reaction directly (cf. Molisch-Udransky
reaction below).
This series of reactions is only given by those proteins which

yield tryptophane on hydrolysis. Gelatin, for example, does not give

these reactions.

V. Diazobenzene sulphonic acid in the presence of potassium
hydroxide yields an orange to brownish red colour, which on treat-
ment with zinc dust changes, owing to reduction, to a fuchsin colour.
The same reaction is given with tyrosine and histidine.

VI. Reactions due to the Presence of a Carbohydrate Group. —
Certain proteins, which contain a carbohydrate grouping (glyco-
proteins, and possibly certain albumins, such as egg-albumin and
serum-albumin J), yield reactions indicating the presence of sugars.
The chief of these are (a) the Molisch-Udransky reaction, and (b)
the orcin reaction.

1 In these cases it is not definitely proved whether the carbohydrate group is
actually contained in the protein molecule, or whether the protein is contaminated with
a sugar or glyco-protein.

32 THE GENERAL CHARACTERS OF THE PROTEINS

(a) The Molisch-Udransky Reaction. — Concentrated sulphuric
acid is added to a solution of protein containing a few drops
of an alcoholic solution of a-naphthol. A violet colour is
produced which turns yellow on addition of alcohol, ether
or sodium hydroxide. If thymol be employed instead of
a-naphthol a carmine-red colour is produced.

(b) BiaPs Modification of the Orcin Reaction. — A small quantity of

dried protein is added to 5 cc. of fuming hydrochloric acid, and
the mixture is then warmed. When the protein is nearly
all dissolved a little solid orcin is added, and then a drop of
ferric chloride solution. After warming for some time a
green coloration is produced, which is soluble in amyl alcohol.

VII. Sulphur Reaction. — On warming a protein solution with-
sodium hydroxide in the presence of a lead salt (lead acetate) a
black coloration is produced owing to the presence of sulphur in
the protein molecule.

As already mentioned, the above tests are not common to all
the proteins, and they serve, therefore, for qualitatively distinguishing
between them in certain cases. Thus, for example, hydroferrocyanic
acid gives only a faint precipitate with gelatin ; with the proteoses
it gives a precipitate which disappears on boiling but reappears on
cooling the solution ; with peptones it gives no precipitate.

Nitric acid also gives a precipitate with the proteoses, which
dissolves on boiling and reappears on cooling; the peptones are not
precipitated.

The alkaloidal reagents precipitate the majority of the proteins
in acid solution only ; the strongly basic protamines, however, can
be precipitated in alkaline solution. The peptones are not pre-
cipitated by picric or trichloracetic acids, or by potassio-mercuric
iodide ; they are precipitated, however, by tannic, phosphomolybdic
and phosphotungstic acids. The colour reactions, as already men-
tioned, are due to certain specific groups, which are not common
to all proteins. All give the biuret reaction, the peptones giving a
characteristic pink coloration.

The Millon reaction, which is due to the presence of tyrosine, is
given only very faintly by gelatin ; the reaction in this case may be due
to an impurity, but according to Morner the reaction occurs normally
if too much reagent be not present. The reaction is not given by
reticulin, nor by the protamines, with the exception of cyclopterine.

The Adamkiewicz (Hopkins-Cole) reaction varies also in in-
tensity with the different proteins. It is not given at all by gelatin,
which does not yield tryptophane as a product of hydrolysis.

An interesting example of the application of the colour reactions
is afforded by Pick, who found considerable differences in the colour
reactions of the various fractions of Witte's peptone, obtained by the
method which has been already discussed (p. 14).

SECTION XIII.— THE CHEMICAL COMPOSITION OF PROTEINS.
THE NITROGEN CONTENT AND DISTRIBUTION.

Proteins sometimes occur in nature combined with other organic
complexes, which have been designated by Hoppe-Seyler as "pros-
thetic" groups, from which, by gentle chemical treatment (e.g., by

GENERAL CHEMICAL CHARACTERS OF PROTEINS 33
v

weak acids at the ordinary temperature), they can be readily freed.
These conjugated proteins give the same general chemical reactions as
the simple proteins. The chief groups are : (i) the nucleo-proteins, or
proteins in combination with a nucleic acid complex ; (2) the glyco-
proteins, or proteins in combination with a complex which can exert
a reducing action on alkaline copper solutions [it has not been
definitely proved that the prosthetic group in this class is always a
carbohydrate]; (3) the chromo-proteins, or proteins containing a
chromatogenic group.

It is not proposed to consider under this section the chemical
composition of the conjugated proteins; only that of the simple
proteins, either those existing as such in nature, or those obtained by
the decomposition of conjugated proteins, will be discussed.

The essential constituents of a protein molecule are carbon,
hydrogen, nitrogen, oxygen and, in nearly all cases, sulphur. Some
proteins contain in addition" "phosphorus, and the halogens are also
found in a limited number of substances of this class.

It is extremely difficult to prepare the majority of proteins free
from ash ; in some cases the inorganic constituents may be in com-
bination with the organic body (the question of salt formation will be
discussed later) and in others simply adsorbed.

The carbon, hydrogen and nitrogen content of a protein varies
in substances of very different origin and character within compara-
tively very narrow limits ; the numbers obtained by an ordinary
elementary analysis of carbon, hydrogen and nitrogen cannot be
regarded as very characteristic. The following analyses will serve as
an example : —

Substance.
Serum-albumin
Albumin
Edestin

c

Per Cent.

52-93
52-75
51-27

5I-03
52-75
52-96

Per Cent.

7-05
7-12
6-85

6-74
6-84
7-05

N
Per Cent.

15-89

15-43
18-76
18-19
17-72
15-65

Observer.
Abderhalden.
Hopkins.
Osborne.

Hammarsten.

Origin.
Horse

Egg

Hemp-seed

Rye

Gliadin „

Caseinogen Cows' milk

Although these bodies yield very different amounts of the typical
hydrolysis products, and are undoubtedly differently constituted
chemically, their contents of carbon and hydrogen vary within com-
paratively narrow limits. The nitrogen shows larger variations, the
plant globulins containing generally between 2 and 3 per cent, more
than the typical animal proteins. In certain other cases also the
nitrogen content is high, viz., in the proteins of more basic character
like the histones and protamines, substances whicri on hydrolysis yield
relatively large quantities of diamino-acids. The thymus histone
contains, for example, 18-35 Per cent, nitrogen, whilst the protamine
salmine, from salmons' testicles, contains no less than 31-69 per cent.
Taken as a whole, the nitrogen does not show, however, very large
variations ; in fact, the nitrogen content of a mixture is often taken
as the index of the amount of protein present.

The Nitrogen Distribution in Proteins.

It has been already stated that the proteins are essentially poly-
peptides formed by the condensation of varying numbers of groups
of diamino- and monoamino-acids, which are obtained from the
proteins by hydrolysis. In addition to these acids, another hydrolysis

34 THE GENERAL CHARACTERS OF THE PROTEINS

i

product is in almost all instances obtainable, viz^ ammonia. The
relative amount of these bodies forms a factor which is characteristic
of each individual protein. As to the origin of the ammonia but
little is known. It is possibly derivable from asparagine or some
allied amide which pre-exists in a conjugated form in the protein
molecule. The nitrogen obtainable in the form of ammonia by the
hydrolysis of proteins is consequently designated " amide-nitrogen ".
The nitrogen of the monoamino-acids is known as " monoamino-
nitrogen," that from the diamino acids, which can be distinguished
from the monoamino-acids by the fact that the former only are
precipitable by phosphotungstic acid, as "diamino or basic nitrogen".
In addition to the nitrogen in the above forms a small quantity of
pigmented bodies are formed during hydrolysis ; the nitrogen con-
tained in these bodies is generally designated as " humin " nitrogen.

Systematic experiments on the determination of the distribution
of nitrogen amongst the various forms of hydrolysis products were
carried out by Hausmann in Hofmeister's laboratory, and the per-
centages in the various forms of amide, monoamino, diamino and
humin nitrogen are often referred to as the " Hausmann numbers ".

In the method originally devised by Hausmann the following
operations were carried out : (i) the hydrolysis of the proteins with
hydrochloric acid ; (ii) the determination of the amide-nitrogen by
distillation of the diluted hydrolysis products with magnesium oxide,
by means of which the ammonium salts are decomposed, and the
estimation of the ammonia evolved by collecting it in a known
amount of standard acid ; (iii) the precipitation, after distilling off
the ammonia, of the residual liquid with phosphotungstic acid, where-
by the diamino-acids are precipitated, and the determination of the
nitrogen in this precipitate ; (iv) the estimation of the nitrogen in
the filtrate from the phosphotungstic acid precipitate. This gives
the nitrogen of the monoamino-acids (also estimated by the subtrac-
tion of the other numbers from the total).

Several objections were raised to the method of Hausmann
shortly after it was published, notably by Henderson, Kutscher, and
by Schulze and Winterstein. It was urged against it by Hender-
son that the amounts of " amide-nitrogen " varied when different
strengths of acid and varying times of hydrolysis were employed.
Kutscher objected to the method on the ground that the diamino-acids
were not entirely insoluble in water and excess of phosphotungstic acid.
Schulze and Winterstein, furthermore, claimed that certain monoamino-
acids such as phenylalanine were precipitable by phosphotungstic
acid. Another objection raised was that by treatment with magnesia
ammonia was evolved from substances other than ammonium salts.

These objections have been subjected to a critical examination by
T. B. Osborne and Harris, and by Gumbel. The former investi-
gators have shown that identical numbers can be obtained where
the conditions of experiment vary even within fairly wide limits,
although they admit with Henderson that different u amide-nitrogen "
numbers, for example, can be obtained by acids of different concen-
trations and by varying periods of hydrolysis. They show, how-
ever, that variations in conditions necessary to produce different
results must be large, and that, provided that certain readily specified
conditions be adhered to, valuable comparative results can be ob-

GENERAL CHEMICAL CHARACTERS OF PROTEINS 35

tained, and it is now generally admitted that, when the conditions
for determination suggested by Osborne and Harris and by Giimbel
are adhered to the " Hausmann numbers " form most valuable factors
for the characterisation of the individual proteins.

In the following table are given some of the chief results obtained
by Osborne and some results by Kossel and Kutscher, rearranged
and recalculated by J. H. Millar, together with some more recent
determinations by Giimbel. The numbers for the proteoses and
peptones are obtained from the reports of the Guinness Laboratory,
vol. i., pt. ii. (1906), pp. 230, 235.

Group.

Protein.

Source.

'c

0

£°

<u

fc

§ .

is

I*

d
§2

1

<

<

K

Salmine

Salmon-roe

o

87-8

Protamine

Clupeine
Sturine

Herring-roe
Sturgeon-roe

—

o
o

83-5
83-7

Cyclopterine

Sea-owl

— —

0

67-7

Histones

Histone
t >

Thymus
Cod-fish roe

—

3-3
7-46

38-7

35-o

Leucosine

Wheat

16-93

6-85

69-87

20-67

2'54

Albumins and

Conalbumin

Egg-white

i6'ii

7-51

65-11

25-82

1-69

phosphopro-

Vitellin

Egg-yoik

16-28

7-67

62-41

28-56

i-35

teins

Ovalbumin

Egg-white

15-51

8-64

68-13

21*27

1-87

Caseinogen

Milk

15-62

10-36

66-00

22-34

i-34

Legumelin

Pea, etc.

16-09

6-46

68-n

23-05

2*36

Globulin

Wheat

18-39

7-72

53-40

37-10

1-52

Legumin

Pea, etc.

17-97

9-40

60-76

28-82

0-94

Globulins

Edestin

Hemp-seed

18-64

10-08

57-83

31-70

0-64

Conglutin

Lupine

18-05

13-18

57-28

28-58

0-88

Amandin

Almond

19-00

16-05

60-79

21-84

0-89

Glutenin

Wheat

17-49

18-86

68-32

11-72

i -08

Zein

Maize

16-13

18-40

77-56

3*03

0-99

Alcohol-soluble
proteins

Ale. sol. protein
Hordein
Bynin

Oat
Barley
Malt

15-67

17-21

16-26

22*46

23-30
23-49

68-09
69-96
68-69

7-84
4-47
4'6i

1-50
1-33
3-I3

Gliadin

Wheat and rye

17-66

23-78

70-27

5'54

0-79

Gluco-proteid

Cartilage

Nasal septum

_

12-97

72-27

12*27

3-27

of oifif

x

Sclero-protein

Gelatin

Commercial

1-61

62-56

35-83

Gluco-proteid

Chondroitin =

Nasal septum

—

35-27

21-57

32-78

9'54

sulphuric acid

of pig

Albumose
Albumose
Polypeptide

Prot-albumose
Hetero-albumose
Glutokyrin

Witte's peptone
Witte's peptone
Gelatin

—

7-14
6-45

O'O

68-17
57-4
33"o

25*42

38-93
66-0

—

Polypeptide

Caseokyrin

Milk

—

O'O

12-0

88-0

—

Plant-albumose

Malt-albumose I.

(

21-3

62-0

7-5

9-2

Plant-albumose

Malt-albumose II.

12-4

70-8

II'O

5-8

Plant-albumose

Malt-albumose III.

4 Malt

8-0

62-0

24-0

6-0

Plant peptose

Malt-peptone I.

)

—

3'5

64-3

30-5

1-64

Malt-peptone II.

1

-~

6-5

48-8

41-0

3-65

36 THE GENERAL CHARACTERS OF THE PROTEINS

Method.

The method adopted is the Osborne- Harris modification of the
original Hausmann method, with a further modification suggested by
Giimbel, viz.) the employment of a vacuum at a temperature of 40° C.,
for the distillation of the ammonia after treatment with magnesia.

About i gram of protein is boiled with 20 per cent, hydro-
chloric acid until the solution no longer gives the biuret reaction,
usually from seven to ten hours. It is then evaporated at 40° under
diminished pressure to 2-3 c.c. and the bulk of the hydrochloric acid
is thereby removed. The residual solution is then transferred to a
flask with about 350 c.c. of water, and a cream of magnesia, which
has been freed from every trace of ammonia by prolonged boiling
is then added until in slight, but distinct, excess. After distilling
and determining the ammonia by warming to 40° in vacua, and
passing the distillate into a known quantity of a standard acid solu-
tion, the solution in the flask is filtered through a nitrogen-free filter
paper, and the residue thus collected washed thoroughly with water
and the nitrogen determined in it together with the paper by
Kjehldahl's method ("humin" nitrogen). The filtered solution is
next concentrated to 100 c.c., cooled to 20° C., 5 grams of sulphuric
acid added, and then 30 c.c. of a solution containing 20 grams of phos-
photungstic acid and 5 grams of sulphuric acid per 100 c.c. After
twenty-four hours the precipitate is filtered off and washed with a solu-
tion containing 2*5 grams of phosphotungstic acid and 5 grams
of sulphuric acid per 100 c.c. The washing is effected by rinsing the
precipitate from the filter into a beaker and returning to the paper
three successive times, each portion of the wash solution being allowed
to run out completely before the next is applied. About 200 c.c. of
washings are thus obtained. The nitrogen contained in the precipi-
tate (" basic " nitrogen) is then determined by transferring it to a
Jena glass flask of about 600 c.c. capacity and digesting with 35 c.c.
concentrated sulphuric acid for seven or eight hours. During diges-
tion potassium permanganate crystals are added three or four times.
In a few cases, when the phosphotungstic acid precipitate is small,
less sulphuric acid is used, enough being taken in each case to
prevent too violent bumping. The remaining nitrogen, belonging
chiefly to monoamino acids, is found by subtracting the sum of the
nitrogen found in the preceding operations from the total nitrogen
contained in the protein under examination.

Skraup has recently shown that two - thirds of the amide-
nitrogen is evolved in a very short time, even with the use of
dilute acids for hydrolysis.

SECTION XIV.— THE SULPHUR, PHOSPHORUS AND HALOGEN
CONTENT OF PROTEINS.

The Sulphur Content.

Far more characteristic of the individual protein than the per-
centage of nitrogen is that of the sulphur. Although the latter element
is not contained in large quantity, yet the variations in its amount
are considerable, and its percentage may be regarded as one of the

GENERAL CHEMICAL CHARACTERS OF PROTEINS 37

characteristic constants of a protein; it indicates the quantity of
the cystine unit present in the molecule.

In the following table the percentage of sulphur present in some
typical proteins is given : —

Sulnhur

Observer.

Goto (i).
Fleroff (2).
Bang (3).
Schulz (4).
Michel (5).

Osborne and Campbell (6).
Osborne and Voorhees (7).
Hammarsten (8).
Osborne (9).

Osborne and Voorhees (10).
Osborne and Campbell (n).
(12).

Chittenden & Osborne (13).
Osborne and Voorhees (14).

(15).
Hammarsten (16).

(i7).

Walter (18).
Levene (19).
Chittenden and Solley (20).

Von Laar (21).

Horbaczewski (22).
Mulder (23).
Schwarz (24).
Siegfried (25).
of pig's intestine
Aorta, Liver, 2-65-2-89 Lubarch (26).

Spleen, etc.
Witte's peptone 1*22 Pick (27).

2-97

0-8

ri-i-6

Substance.

Sulphur
Origin. Per

r

Jent.

Clupeine

Herring testicles

o

Histone

Thymus

0-62

Scomber histone

Mackerel testicles

0-79

,Globin

Haemoglobin

0-42

Albumins

J Albumin

Serum
Egg

1-90

1-62

I Leucosin

Wheat

1-28

/-Globulin

Blood

I'll

Edestin

Hemp-seed

0-91

Globulins -I Globulin

Wheat

0-69

I Conglutin

Lupine seeds

0-40

^Globulin

Egg

0-123

Alcohol-

(Zein

Maize

o'6o

soluble

Gliadin

Wheat kernel

1-14

proteins

Glutenin

11

i -08

Caseinogen

Milk

0-76

Fibrinogen
Ichthulin
it

Blood
Carp's egg
Cod's egg

1-25
0-41
0-92

Commercial

0-7

gelatin

Hair

Human

5 'oo

Nail

»

2-80

Horn

(average)

3'2Q

Tortoise-shell

2'22

Elastin

0-38

Reticulin

Mucous membrane

1-88

Amyloid

Hetero-prot.-

albumose
Deutero-albu-

mose (Thio)
Deutero-albu-

mose (S. poor)
Other albumoses

Phosphorus Content.

Certain proteins, the phosphoproteins, are characterised by the
relatively large amount of phosphorus they contain. They must be
distinguished from the nucleo-proteins, in which the phosphorus is
contained in the prosthetic group, nucleic acid, and not in the protein
part of the molecule. The typical phosphoprotein caseinogen con-
tains, according to Hammarsten, 0*847 per cent, of phosphorus;
ovovitellin contains, according to Plimmer, i'i per cent.; ichthulin
from carps' eggs contains 0*43 per cent. (Walter, loc. cit.^ in the
table), and the corresponding substance from cods' eggs 0*92 per
cent (Levene). Plimmer and Scott have recently shown that the
phosphoproteins are easily decomposed by I per cent, caustic soda, and
are thus readily distinguished from the nucleo-proteins. Ordinary
proteins contain very small quantities, if indeed, any phosphorus.

Halogen Content.

Certain proteins contain very appreciable quantities of halogen.
The chief of these is the thyro-globulin of the thyroid gland, in which

38 THE GENERAL CHARACTERS OF THE PROTEINS

iodine was first detected by Baumann in 1895. The iodine containing
protein has been exhaustively studied by Oswald. The amount of
iodine varies in different preparations ; Oswald found 0*46 per cent,
in the thyro-globulin prepared from pigs' thyroids, O'86 from ox thy-
roid, and 0-39 from sheep's thyroid ; in human thyroid he found from
0*07 to 0*51 per cent, (the latter after potassium iodide administration).

Halogens have also been found in certain marine animals, es-
pecially in the skeletons. The axial system of Gorgonia cavolini
contains, according to Drechsel, about 8 per cent, iodine (calculated
on the amount of dry substance), which is combined with the protein
substance known as gorgonin. Sponges also contain iodised pro-
teins according to Hundeshagen and Harnack. The latter by the
hydrolysis of sponges with mineral acids isolated the protein body
iodospongin containing 9 per cent of iodine.

Various other proteins with halogen content have been recently
isolated from marine organisms by Morner.

SECTION XV.— THE TYROSINE FACTOR OF PROTEINS.

Each individual protein yields on hydrolysis certain definite quan-
tities of amino acids. The method of determining these quantities
with any degree of accuracy is unknown for the majority of the
amino acids ; even by the esterification method of Emil Fischer only
approximate results are obtained.

There are, however, at least one or two hydrolysis products, of
which the yield can be readily determined when only small quantities
of the protein from which they are derived is available. The quan-
tity of the cystine group can be estimated by a determination of the
total sulphur of the molecule 1 (see p. 30). Another hydrolysis pro-
duct, the amount of which can be readily determined, is tyrosine ;
it is probable that the tyrosine factor (i.e.t the percentage of tyrosine
liberated on hydrolysis) will form a definite characteristic for each
protein.

There are several earlier investigations on this subject, which are
summarised in a paper by Reach. In the researches described in
this and the previous papers the tyrosine was obtained in the form
of crystals, which were weighed. A more accurate method is, how-
ever, due to J. H. Millar, who has shown that tyrosine can be readily
estimated in even complex mixtures by titrating a solution of a
protein containing hydrochloric acid and potassium bromide with
a standard solution of potassium bromate, until a potassium iodide-
starch indicator denotes the presence of free bromine. The reaction
which takes place may be represented by the following equations : —

NaBrO3 + sKBr + 6HC1 = NaCl + sKCl + 3Br2 + 3H2O.
C6H4 . (OH). CH2 . CH(NH2) . COOH + 4Br = C6H2Br2 . (OH) . CH2 . CH . (NH2) . COOH.

According to the theory of these equations 1765 grams of bro-
mine would be required for each gram of tyrosine, or about 18*8 c.c.

— potassium bromate solution. This method has not yet been much

applied to the investigations of the hydrolysis products of proteins.
Adrian Brown and E. T. Millar have, however, recently shown

1 On the assumption that cystine is the only sulphur-containing group.

GENERAL CHEMICAL CHARACTERS OF PROTEINS 39

that in the proteins they investigated (edestin and egg-albumin)
scission of tyrosine by tryptic digestion is a very rapid process
(whereas there is no scission by peptic digestion) ; the whole of the
tyrosine from edestin and egg-proteins was liberated in from one to
three hours. It is conceivable that the determination of tyrosine in
a protein by this method, after a short tryptic digestion, may afford
another valuable factor for the characterisation of proteins.

SECTION XVL— SALT FORMATION BY PROTEINS. COMBINA-
TION WITH ACIDS AND BASES.

One class of the proteins, vis., the protamines, react as strong
bases, giving alkaline solutions and characteristic salts with acids ;
others, of which caseinogen may be taken as an example, react as
acids, giving salts on treatment with bases. The majority of the
proteins, however, possess an amphoteric reaction, acting as bases
towards acids and as acids towards bases.

For reasons already mentioned in the introduction, the deter-
mination of the acidity or basicity of proteins offers several diffi-
culties. The chief of these is due to their high molecular weight ;
relatively large amounts of protein will require for neutralisation but
small amounts of acids and bases. Furthermore, as the proteins in
several instances act only as very weak acids or bases, the salts
readily undergo hydrolysis in aqueous solutions ; the ordinary methods
of titration, with the use of indicators, are not therefore available for
determining their combining weights with acids and bases. Con-
sequently it is necessary to employ indirect methods, of which the
following are the chief: —

A. Physical Methods.

I. The measurement of electrolytic conductivity when acids or
bases are added to protein solutions.

II. The measurement of the potential of concentration cells, and
the changes produced when proteins are added to either electrode.

III. The determination of the influence of the addition of pro-
teins on the depression of the freezing points of solutions of acids
and bases.

IV. The determination of the influence of the addition of proteins
on the rate of hydrolysis of cane-sugar and esters by acids and
bases.

B. Chemical Methods.

I. Direct titration in presence of indicators.

II. Determination of the solubility in acids and bases of proteins
which are insoluble in pure water.

III. Determination of the acidity of the filtrate from protein
precipitates produced by neutral salts or alkaloidal reagents in
solutions containing an excess of acids.

It is proposed to consider these general methods first, and after-
wards to refer to some of the more recent investigations on the salt
formation of individual proteins.

4

40 THE GENERAL CHARACTERS OF THE PROTEINS

•

A. Conductivity Method.

This method was first applied by Sjoqvist.

If a strong base be gradually added to a solution of a strong acid,
the volume of the solution being kept constant, and the electrolytic
conductivity be determined after each addition, it will be found that
it gradually diminishes, until a certain minimum is reached, after
which it gradually increases. This minimum represents the point
at which the acid is completely neutralised by the base. When
potassium hydroxide is added to hydrochloric acid the conductivity
curve, where the ordinates represent the molecular conductivity, and
the abscissae the amount of alkali, is represented by a descending
and an ascending straight line enclosing a sharp angle. The attain-
ment of the minimum point is due to the fact that both the acid and
the base are better conductors than the salt which is formed by their
neutralisation of one another. If, instead of a strong base, a weak
base like ammonia be employed, which is a bad conductor, the con-
ductivity rises but slightly after the neutralisation point is reached ;
at this point the curve becomes nearly parallel with the abscissa. A
somewhat similar curve is obtained when egg-white is gradually
added to a solution of acids, and by a study of such curves Sjoqvist
has endeavoured to throw some light on the salt formation of the
egg-proteins with acids.

Two additional factors have in this case to be taken into account,
viz. : (a) the conductivity due to the protein addition ; (b) the decrease
of conductivity due to the lessened velocity of the ions owing to the
increased internal resistance of the fluid produced by the addition of
the proteins.

The conductivity of the protein alone is readily determined.
This is probably due entirely to combined or adsorbed salts. Al-
though the protein solutions used in Sjoqvist's experiments were
carefully dialysed, they could not be obtained ash free. The ash
was found to consist of a mixture of calcium sulphate and phosphate.

The internal friction factor is not so readily determined. The
alteration of the conductivity of the acid due to this factor can only
be determined indirectly, viz., by determining the decrease of conduc-
tivity of a neutral salt solution produced by repeated additions of
known quantities of protein, and the ratio of this decrease to the
decrease produced in acids by other non-electrolytes which increase
internal resistance, and upon which acids do not chemically act. It was
found, for example, in the experiments of Sjoqvist, that each gram
of protein added to a 0*05 N sodium chloride solution produced (after
allowance for the conductivity of the protein alone) a diminution of
1*52 per cent, of the total molecular conductivity. Now the mean de-
crease produced by six other non-electrolytes, which increase internal
friction, was found by Arrhenius to be as follows: for NaCl, 2*21
per cent. ; for HC1, 1-84 per cent.; for HNO3, r88 percent. If,
now, the ratio of decrease produced by proteins is the same as for
other non-electrolytes, then the diminution of conductivity produced
by the addition of each gram of protein to 100 c.c. of a 0*05 N

hydrochloric acid solution is 1*52 x — 4 = 1*26 per cent. The actual

2'2I

diminution produced when egg-white solution is added to 0*05 N

GENERAL CHEMICAL CHARACTERS OF PROTEINS 4!

hydrochloric acid was much greater than this : the diminution was
due to actual neutralisation, and not to mere increase in internal
friction.

Sjoqvist determined the conductivity of the 0-05 N hydrochloric
acid solution after the addition of varying amounts of egg-proteins,
with the following results : —

Amount of protein in 100 c.c. o 0-72 1*08 2-16 3-03 4-09 4-70

£^pr .... 334'5 286-2 263-1 196-2 140-3 97-5 78-52

Amount of protein in 100 c.c. 5*22 6-26 6-71 7-83 9-40

^T .... 68-66 60-7 59-43 58-32 57-7

It is evident from the above numbers that in the last determina-
tions far more protein was added than was necessary to neutralise
the hydrochloric acid. After making corrections for the increased
conductivity, due to the protein added, and the diminished con-
ductivity due to friction, Sjoqvist calculated that the molecular con-
ductivity of the hydrochloride of the egg-proteins in 0*05 N solution
was 53 x io~7, and, by interpolation in the curve, that neutralisation
of 0*05 N hydrochloric acid solution was attained when 4*1 grams
of protein had been added to 100 c.c. From this the equivalent of
egg-protein was calculated to be 820. This number obtained with
sulphuric acid was found to be 840, and with nitric acid 720. The
average is not far removed from 800.

There is one other factor in these experiments which deserves
mention. It was noticed that in the case of the neutralisation of
hydrochloric acid by ammonia the curve first descends, and then
suddenly becomes almost parallel with the abscissa, forming thus
two straight lines enclosing a sharp angle. In the case of the protein
hydrochloric acid neutralisation curve there is no sharp break, but
the curve is rounded off. This is due to the fact that hydrolysis
of the protein hydrochloride can take place according to the equation

Alb. HCl + H2O = Alb. OH + HCl.

This hydrolysis is suppressed in the presence of a large excess of
protein (i.e., in the calculations of conductivity given above).

The amount of hydrolysis has been determined by Sjoqvist in
the following way. He calculated the molecular conductivity of a
solution produced by adding the inorganic constituents of 4 grams
of protein to 100 c.c. of 0*05 N hydrochloric acid (the quantity neces-
sary for " neutralisation," vide supra). Such a solution would con-
tain known quantities (which were determined by analysis) of free
HCl, CaCl2, H3PO4 and H2SO4 (produced by the action of a large
excess of HCl on the original salts CaSO4 and Ca3(PO4)2). For such
a solution //, was found to be 3 14 x 10 ~ 7. The conductivity was next
calculated for this solution when the free acids were neutralised by
egg-proteins without hydrolysis, and after making correction for
decreased conductivity due to viscosity. (The molecular conduc-
tivity of egg-protein hydrochloride was found as shown above to be
S3 x 10 ~ 7, and that of the sulphate and phosphate were determined
in a similar way in other experiments.) //, for this neutralised solu-
tion was calculated to be 55*66 x 10 " 7. The actual observed value
was 100*5. Now as 314 x io~7 represents the value for total

42 THE GENERAL CHARACTERS OF THE PROTEINS

hydrolysis of the hydrochloride (the protein having no conductivity),
and 5 5 '66 x 10 ~ 7 represents the value when no dissociation takes
place, then using the ordinary formula when x represents the pro-
portion hydrolysed —

314* + (i - *)55'66 = 100-5

whence x = 0*174, from which it was calculated that 17*4 per cent,
of protein hydrochloride was hydrolysed when a 0*05 N solution of
hydrochloric acid solution was neutralised with egg-proteins.

These experiments of Sjoqvist were carried out with the whole
of the proteins of egg-white, and therefore probably with a mixture.
Some of the calculations, furthermore, are approximations. They
have, however, been described in some detail, as they demonstrate in
a clear manner the salt formation when proteins are treated with
acids; they also indicate the methods for determination of the
hydrolysis of such salts.

A similar set of conductivity determinations with the proteins of
horse serum is due to Mellanby.

A. II. Electric Potential Method.

This method has been employed by Bugarsky and Liebermann.
According to the theory of Nernst, the electro-motive force of a
galvanic element constituted according to the following scheme —

Platinum laden with H2 | Acid | Base | Platinum laden with H2

depends on the concentration of the hydrogen ions in the acid and
base solutions.

The addition of a protein to either the acid or base would in
the event of combination and salt formation alter the concentration of
the various ions in solution.

The electro-motive force of such an element as the above can,
according to the Nernst theory of electrolytic solution pressure, be
expressed by the following equation : —

RT CH d\

ir = p- nat log —
fc0 *H

where R is the ordinary gas constant, T the absolute temperature,
E the quantity of electricity transported by I gram equivalent of
an electrolyte, CH the concentration of the hydrogen ions in the acid,
and ^H the concentration of hydrogen ions in the base. If the electro-
motive power be expressed in volts, the quantity of electricity in cou-
lombs, and the electrical energy I volt-coulomb = 10,000,000 ergs, and
the experiments be all carried out at a constant temperature (25° C.),

•p "T1

then -=- is a constant = 0*0256.

EO

If now the Brigg logarithms be substituted for the natural loga-
rithms,

CH
T = 0-0590 Iog10 — -.

H

The hydrogen ions in the acid depend on the dissociation of the
acid in the concentration employed in the experiments.

In the base solution (sodium hydroxide was used in the experi-

GENERAL CHEMICAL CHARACTERS OF PROTEINS 43

ments), if CH represent the concentration of the hydrogen ions, and
COH that of the hydroxyl ions, then according to the law of mass
action,

CH . COH = constant = K.

In the presence of sodium hydroxide the concentration in the

hydroxyl ions is decreased (owing to dissociation into Na and OH).
If now xH represent (as above) the concentration of the hydrogen
ions, and COH the concentration of the hydroxyl ions due to the
presence of the alkali, then (dissociation of water being negligible)

*H ' COM = K a°d #tr = n

COH

Substituting this value in equation (i), it is found that

c c

1T = 0-059 10g10 H K °H-

Suppose now that the ionic concentration of the acid be altered by
the addition of protein which forms a salt with the acid, a change in
the electro-motive force will follow. The concentration of the hydro-
gen ions in the sodium hydroxide (CNaon) will remain unchanged.
If CHCI represent the ionic concentration of the original acid solution
and CH that after addition of protein, and EQ the original electro-
motive force and E the changed force, then

EQ = O-OSg l°g

and E = 0-059 loglv, K

Whence,

^'HCK

E - EQ = O-O59 lOg

and similarly for alkalis, E - E0 = O'O$g log

The experiments were carried out with O'OS N acid and alkali
solutions with a 0*05 N sodium chloride solution interposed. The
electro-motive force was determined by a compensation method,
a Clark cell being used as standard. Values for E - E0 and

1 This equation is not strictly correct, for between the sodium hydroxide and acid
solutions there is a layer of sodium chloride, and there are consequently potential
differences between the acid | salt and alkali | salt solutions. If </> represent the former
of these differences, and ty the latter, then the equations become

E0= 0-059 log10

and E = 0-059 log,. „.

^ remaining the same in both equations and the ionic concentration in the alkali not
altering. Then

E - EQ = 0-059 Iog10

$ - ^o is sma11 compared with the other number and was neglected in calculating the

CHCI
value

44 THE GENERAI^ CHARACTERS OF THE PROTEINS

Iog10 cl were plotted as ordinates and abscissae, and a curve drawn.
CH

By means of this curve the value -~£l could be ascertained by each

experimentally determined value E - Eo.

The following were some of the results obtained :

Egg Protein + 0-05 HC1.

Egg-Protein -f- 0-05 NaOH.

Protein, Grams per
100 c.c. boluiion.

Per Cent. HC1

Combined.

Protein, Grams per
100 c.c. Solution.

Per Cent. NaOH
Combined.

O

O

0

0

0-4

9-0

0-8

14-4

0-8

18-9

1-6

27-4

1-6

33'3

3'2

60-2

3-2

60-2

6-4

97-0

6-4

99'56

12-8

99-88

12-8

99-67

From the above table it is clear that the egg-protein combines
both with acid and alkali, and it must be conceded that Bugarsky
and Liebermann have by this method clearly demonstrated the am-
photeric character. The addition of protein to sodium chloride in
another form of galvanic cell produced no change in electro-motive
force ; the combination took place only with bases and acids.

A. III. Depression of Freezing-Point Methods.

Bugarsky and Liebermann also employed this method. They
determined the depression of the freezing point caused by the addition
of definite quantities of protein to pure water (S). They then added
the same quantities to 0*05 N hydrochloric acid, sodium hydroxide
and sodium chloride solutions, and determined the depression in the
freezing points of these solutions. If D be the depression of the
freezing point of these solutions before the addition of protein, and
A the observed depression after addition of the protein, it was found
in the case of the acid and the base that A was less than D + S. In
the case of the salt A = approx. D + 8. This indicated the diminu-
tion of the number of molecules in solution, due to the combination
of the protein with acids or bases. No combination took place with
the salt. These experiments confirm the results obtained by the
electric potential method.

A. IV. Hydrolysis of Cane-Sugar and Esters by Acids and
Bases in Presence of Proteins.

This method was originally employed by Hoffmann for deter-
mining the free hydrochloric acid in gastric contents. It has also
been employed by O. Cohnheim for estimating the combining powers
of various products of protein digestion with hydrochloric acid, and
by Hardy for determining the amount of hydrolysis in the hydro-
chloric acid compounds of serum-globulin. More recently still it
has been applied by B. Moore for investigation of gastric contents

GENERAL CHEMICAL CHARACTERS OF PROTEINS 45

in pathological cases. Cane-sugar on hydrolysis is converted into a
mixture of glucose and fructose according to the equation

C12H13On + H20 = C6H1206 (glucose) + C6H12O6 (fructose),

whilst methyl acetate is converted by hydrolysis into a mixture of
methyl alcohol and acetic acid : —

CH3 . COOCH3 + H2O = CH3 . COOH + CH3 . OH.

The rate at which these reactions take place is proportional to
the concentration of the hydrogen ions present (i.e., in sufficient dilu-
tion of the acid), when the same amounts of ester or sugar are
employed in comparative estimations. If combination between acid
and protein take place the addition of the latter will diminish the
effective concentration of the former.

If now two parallel experiments be carried out with the same
concentrations of sugar or ester, both in the presence of known (but
not necessarily the same) concentrations of acids, but one in the
presence of known amount of protein and the other without, then
according to Wilhelmi's logarithmic law, the amounts of sugar or
ester hydrolysed can be represented by the following equations : —

where K is the reaction constant in the experiment when protein
is absent, and K' the constant when it is present, A is the original
amount of sugar or ester, and x and x the amounts hydrolysed in
the time /, in the presence and absence of protein respectively.
Then

K log A - log (A - x)
K' log A - log (A - *')'

If now d be the concentration of the acid when no protein is
present and Z be the effective concentration in the parallel experiment
with protein, then

K'.d

z = -ir.

By subtracting Z, found thus, from the original concentration, the
amount of acid that had entered into combination with the known
amount of added protein can be readily determined.

Cohnheim determined the combining powers of certain proteoses,
prepared by the methods of Kuhne and Chittenden, with the follow-
ing results (at 40° C.) : Prot-albumose in a concentration of 2^ per
cent, combines with 4*32 per cent, of its weight of hydrochloric acid
and in a concentration of 1*25 per cent, with only 3*5 per cent. ; in a
concentration of 5 per cent., however, it combines with as much as
4'9 per cent, of its weight of acid. Deutero-proteose combines with
5*48 per cent., hetero-albumose with S'i6 per cent, and antipeptone
with 15*87 per cent, of their weights of hydrochloric acid in 2\ per
cent, solutions at 40° C.

With the exception of antipeptone, which is a strong base, the
proteoses investigated appear to combine with less hydrochloric acid
in diminished concentrations. This is due to the fact that in dilute
solutions a certain amount of hydrolysis occurs : —
Alb. Cl + H2O = Alb, OH + HC1.

46 THE GENERAL CHARACTERS OF THE PROTEINS

The method has been applied by Hardy (see below, p. 53) for
investigating the amount of hydrolysis which takes place in dilute
solutions after a protein has been neutralised, the neutralisation
point having been determined by other methods.

The method is of somewhat limited application, owing to the fact
that most proteins are hydrolysed to some extent even by dilute acids.

B. I. Direct Titration in the Presence of Indicators.

This method is quite effective in the case of strong bases, such
as the protamines. It requires to be applied with some care in the
majority of cases, owing to a variety of circumstances, such as the
hydrolysis of the salts in dilute solutions, and the capacity for the
formation of acid and basic salts. Particular stress must also be
laid upon the choice of indicators. The method is discussed in
some detail below, in considering the researches on certain individual
proteins.

B. II. The Determination of the Solubility of Proteins which are
Insoluble in Water, in Acids and Bases.

This method is applicable to a limited number of proteins only,
e.g., to the globulins. This is also discussed in greater detail below.

B. III. The Determination of the Acidity of the Filtrates from
Protein Precipitates Produced by Neutral Salts or Alkaloidal
Reagents.

This method has been employed by Spiro and Pemsel, Cohn-
heim and Krieger, and von Rhorer.

Spiro and Pemsel added acids and alkalis in excess to protein
solutions. They then precipitated the proteins by ammonium sul-
phate, and estimated the acid or base remaining in the filtrate. They
assumed that the combination of acid or base with protein could be
precipitated by salts in the same way as the protein itself, and arrived
thus at conclusions as to the amount of acid or base which could
enter into combination with proteins. This method has been but
little employed, owing to the difficulty of maintaining solutions of
ammonium sulphate in a state of complete neutrality.

Cohnheim and Krieger employed a similar method, using, how-
ever, alkaloidal reagents, assuming that the acid salt of the protein
and the reagent enter into double decomposition, according to such
an equation as the following : —

Protein hydrochloride + Calcium phosphotungstate = Protein phosphotungstate + CaC!2.

If to protein containing excess of acid, calcium phosphotungstate
be added, the total acidity of the solution, as determined by titration,
would be diminished, owing to the combination of acid and protein,
and subsequent double decomposition of the salt thus formed with
the alkaloidal reagent. Other reagents used were sodium picrate,
calcium trichloracetate and potassium mercuric iodide.

The method of Cohnheim and Krieger was subjected to a critical
examination by von Rhorer, who has shown that, as originally
carried out, it is not free from certain errors.

When, for example, calcium phosphotungstate is used as precipi-

rENERAL CHEMICAL CHARACTERS OF PROTEINS 47

tant, the filtrate after precipitation of the protein should contain
calcium chloride, excess of calcium phosphotungstate and hydrochloric
acid. As hydrochloric acid is much stronger than phosphotungstic
the calcium salt of the latter is decomposed, and the filtrate contains
as a matter of fact chiefly calcium chloride and phosphotungstic acid.
The latter acid cannot be accurately titrated with the use of rosolic
acid, which was employed by Cohnheim and Krieger as an indicator.
Certain of the conclusions drawn by these observers, therefore, as to
the dissociation of the hydrochlorides of proteins are erroneous.

Another source of error has also been discovered by von Rhorer,
viz., that the bulky precipitates can adsorb certain quantities of acid
from solution. If a large excess of acid be present part of this
excess will be carried down ; it can, however, be recovered by suffici-
ently washing the precipitate. The statement of Cohnheim and
Krieger that the amount of acid entering into combination with
the protein depends on the amount originally present in solution is
erroneous ; the larger the quantity in solution the larger the quantity
carried down by the precipitate ; by thoroughly washing the latter,
and estimating the acid in the washings, the error due to adsorption
can be readily eliminated.

Von Rohrer found that the precipitation method yields reliable
results if calcium picrate or potassium mercuric iodide be employed
as precipitants (provided, of course, that the precipitates be thoroughly
washed). By means of the latter reagent, he estimated the equiva-
lent combining weight of crystallised egg-albumin as 981, whereas
Sjoqvist estimated it for dialysed egg-white at 800 (see p. 41).

Owing to its simplicity, the method seems worthy of further
application.

The Salt Formation of Individual Proteins.

The salt formation of some individual proteins has been investi-
gated in detail. Many of the general principles mentioned above
are well illustrated by these investigations. The following will receive
some more detailed consideration : —

(a) The salt formation of edestin, according to the investigations
of Osborne.

(&) The salt formation of caseinogen, according to the investiga-
tions of Lacqueur and Sackur.

(c) The salt formation of serum-globulin, according to the investi-
gations of Hardy.

The Salt Formation of Edestin (Osborne).

Owing to the ease with which edestin is obtained in quantity, and
to the fact that it can be recrystallised and obtained in an apparently
homogeneous state, it forms a very suitable substance for investigation.

Osborne noticed that all the crystalline edestin preparations ob-
tained by the deposition of the protein from warm salt solutions are
capable of neutralising small quantities of potassium hydroxide
solution when phenol phthalein is used as an indicator ; some even
are slightly acid to litmus. Out of twenty preparations examined,

ftr
eight required 0-2 to 0-5 c.c. - - potassium hydroxide to produce a

48 THE GENERAL CHARACTERS OF THE PROTEINS

neutral reaction to litmus ; the preparations required from 0*85 to 1-5
c.c. to produce a neutral reaction to phenol phthalein when i gram of
the protein was employed. The causes of this acidity were investigated
by treating suspensions of known quantities of edestin with potassium

/N\

hydroxide ( — j solution until neutral with phenol phthalein, filter-
ing off the undissolved protein and certain quantities of protein pre-
cipitated by the addition of alkali (as the edestin crystals are partially
soluble in water, as will be explained later), and estimating quantita-
tively the constituents of the filtrate. These consisted of a certain
amount of organic matter together with potassium and sodium salts
of hydrochloric and sulphuric acids. The following conclusions
were drawn by Osborne as to the nature of edestin crystals :—

(1) That edestin is a basic body which forms true salts with acids.

(2) That the preparations, as obtained by crystallisation from
salt solutions, react either weakly acid or neutral to litmus and are
salts of a basic protein.

(3) That by the addition of caustic alkalis to these preparations
until the mixture no longer acts acid towards phenol phthalein, these
salts can be decomposed ; by the analysis of the filtrate from the pro-
tein after this treatment the nature of the salts can be determined.

(4) The edestin preparations obtained by recrystallisation from
sodium chloride consist principally of chlorides (as the filtrate after
neutralisation with alkali contains chiefly chlorides) ; if prepared by
crystallisation from ammonium sulphate they consist chiefly of
sulphates.

(5) The edestin preparations, as ordinarily prepared by crystal-
lisation from sodium chloride, are appreciably soluble in water. The
part which is soluble in water requires twice as much alkali for
neutralisation, with phenol phthalein as indicator, as the part which
is insoluble.

(6) If the molecular weight of edestin be taken as 14,500, or
twice the simplest molecular weight as determined by direct analysis
(i.e., from C . H . N . S and O content), then the amount of hydrochloric
acid combined in the insoluble portion would correspond to a mono-
hydrochloride, whereas that in the soluble portion would correspond
to a dihydrochloride. The ordinary edestin preparations are, there-
fore, usually mixtures of chlorides, containing, however, a certain
quantity of sulphate. The relative amounts of these salts depend
on the relative amounts of chloride and sulphate present in the
liquids used either in the original extractions or in the recrystallisa-
tions.

It is of interest to note that all the edestin preparations, whether
free base, sulphate or chloride, have the same crystalline form. The
protein part of the molecule is so large in comparison with that of
the combined acid that the latter exerts no influence on the form of
the crystals. Analogous instances of isomorphism are known in the
case of certain minerals, and also in the case of haemoglobin and its
derivatives (e.g., O and CO haemoglobin). In the above-mentioned
experiments the method of direct neutralisation was employed (me-
thod B I.) ; another series was carried out with the object of de-
termining the amount of the free base edestin (which is insoluble
in water) which is soluble in acids of given strength (method B II.).

GENERAL CHEMICAL CHARACTERS OF PROTEINS 49

An edestin preparation, almost neutral to phenol phthalein, was
obtained by allowing the crystals to separate from a sodium chloride
solution containing sufficient sodium hydroxide solution to neutralise
the acid ; the amount necessary was determined by the titration of
a small aliquot part of the salt solution in the presence of phenol-
phthalein.1 Portions, each of a gram, of this preparation were sus-

N
pended in 20 c.c. of liquid containing quantities of — hydrochloric

acid varying from 2-14 c.c., in a series of stoppered bottles. After
shaking for two hours, and allowing suspended matter to settle, 10
cc. of clear liquid were decanted off, and the amount of edestin in the
solution was estimated ; the acid contents of the solution and residue
were also determined, and the distribution of the acid between the dis-
solved and undissolved edestin thereby ascertained. Similar experi-
ments were carried out with other acids and also with alkalis. The
results obtained by this method of experiment may be briefly sum-
marised as follows : —

(1) The free base edestin, if suspended in water, requires some-
what more than that amount of acid to dissolve it than is necessary
to form a dihydrochloride, on the assumption that edestin has the
molecular weight 14,500. If the acid be added little by little to the
suspension practically no solution takes place till more than half
this quantity has been added; the insoluble monohydrochloride is
first formed. Somewhat more than the theoretical amount of acid
is necessary to produce complete solution owing to the formation of
a more basic hydrolysis product being formed, which is insoluble in
water (edestan).

(2) The sulphates of edestin are less soluble than the chlorides,
and the existence of definite compounds has not been ascertained ;
ten times more sulphuric acid than hydrochloric acid is necessary to
dissolve a given quantity of edestin. Acetic acid, on the other hand,
dissolves nearly the theoretical amount (2 mol. acid : I mol. edestin),
as in this case little of the basic bye-product is formed. Phosphoric
acid acts as a monobasic acid, giving salts of the type B' . HJPO. and
B"(H2P04)2.

(3) In addition to acting as a base, edestin can also act as an
acid ; for solution of a given quantity of edestin one molecular equi-
valent of sodium or potassium hydroxide to one molecular equivalent
of edestin is necessary ; the solutions of the sodium and potassium
salts readily become turbid, owing apparently to hydrolysis. Rela-
tively larger quantities of the carbonates and of ammonium hydroxide
are necessary to bring about solution.

With reference to edestin, the chief points of interest are that the
existence of definite salts both with acids and bases can be ascer-
tained; with hydrochloric acid both a mono- and dihydrochloride
can be prepared, of which the latter only is soluble in water ; this
latter salt as well as the sodium and potassium edestin compounds
have a peculiarity, viz., they are insoluble in the presence of minute
quantities of a neutral salt, although they dissolve in more con-
centrated solutions ; in such solutions they show the ordinary pro-
perty of globulins.

1 Full details of the preparation of the free edestin are given in the original paper.

So THE GENERAL CHARACTERS OF THE PROTEINS

Many similar results with, however, some marked contrast, have
been obtained by Hardy in his investigation on serum-globulin ;
these analogies and contrast will be referred to later.

The Salt Formation of Caseinogen (Lacqueur and Sackur).

W. A. Osborne showed that the salts of caseinogen could be
divided into two classes, viz., those of the alkaline earths and those
of the alkalis, which can be distinguished from one another by the
facts that the former form opalescent solutions and cannot be
filtered through a Martin gelatin filter, whilst the latter form quite
clear solutions and can be filtered through gelatin. Osborne clearly
recognised the acid character of caseinogen and its salt -forming
capacity.

Various attempts have been made to determine quantitatively the
amount of different bases necessary to neutralise a given amount of
caseinogen,1 but the most accurate determinations are due to Lac-
queur and Sackur. They showed that the discrepancies of the
different observers are due to the fact that the calculations as to com-
bining weights are made with reference to the air-dried caseinogen,
instead of the substance dried at 110° C. This is important when
it is remembered that caseinogen preparations are somewhat hy-
groscopic.

Lacqueur and Sackur confirmed other investigators in their
statement of the fact that solutions of caseinogen in alkali, which are
alkaline to litmus, are acid to phenol phthalein. They showed further-
more that with phenol phthalein perfectly sharp neutralisation points
could be obtained with various alkalis, both by direct titration and
by the addition of alkali in excess, and subsequent titration of this
excess with acids. They found as a means of several concordant
experiments that i gram of caseinogen (calculated as free from

N
water) was capable of neutralising 8*8 1 c.c. of — sodium hydroxide

solution. The amount of water in each sample titrated was esti-
mated by drying at 110° C. ; the percentages of water were found to
vary between 8'2 and 15*5.

Lacqueur and Sackur also determined the electrolytic conduc-
tivity of solutions of caseinogen, which has been neutralised by
sodium hydroxide in the presence of phenol phthalein in various
dilutions, with a view of ascertaining the basicity of the acid casein-
ogen, as Ostwald and Walden have shown that the more polybasic
an acid is, the greater is the decrease in the amount of dissociation
when the solution of its sodium salt is concentrated. This dissocia-
tion influences, of course, the conductivity, and if Al represent the
equivalent conductivity at dilution vl and A2, the equivalent con-
ductivity at dilution z/2, then between dilutions 32 and 512 (i.e., I
gram equivalent in 32 litres and I gram equivalent in 512 litres)

— J— j — 2 has the following average values : for dibasic acids 0*15,

•^•i

for tribasic acids 0*22, for tetrabasic acids 0*29, for pentabasic acids
0*35, and for hexabasic acids 0*39.

1 The reference to these earlier investigations is given in Lacqueur and Sackur's
paper.

GENERAL CHEMICAL CHARACTERS OF PROTEINS 51

N
Lacqueur and Sackur added to — sodium hydroxide solutions

the calculated quantities of caseinogen necessary to neutralise it (a
solution in which v = 20 was thereby obtained) ; this was diluted to
varying concentrations, and the conductivities determined with the
following results : —

Per Cent. Caseinogen. v (Dilution). A (Equivalent Conductivity).
2-84 40 46-5

1-42 80 51-3

071 160 56-2

o-355 320 63-0

0-1775 640 69-5

Whence A™~A^ = 0-33.

•^640

From this it follows that caseinogen is at least a tetrabasic acid.

N
Now as i gram neutralises 8'8i c.c. — sodium hydroxide, its

equivalent combining weight is - - =1135. On the assumption

O ool

that caseinogen is a tetrabasic acid its molecular weight would be
4540, or, on the assumption that it is hexabasic, 6810.

The sodium salts of caseinogen can, like other protein salts,
undergo hydrolysis in solution. The solution of caseinogenate of
sodium, which is obtained by just neutralising caseinogen with
sodium hydroxide, is slightly opalescent and the opalescence dis-
appears on addition of excess of alkali. Furthermore, the point of
neutralisation, when litmus is employed as indicator, varies very
appreciably with the dilution of the solution. The condition of equi-
librium may be represented by the equation

Kas. nNa + nH2O ^ Kas. nH + nNaOH ;
or possibly acid salts may be formed —

Kas. nNa + mH2O f£ Kas (n - m)Na + mNaOH

(cf. Hardy1 on serum-globulin).

The statements with regard to hydrolysis have been confirmed
by Lacqueur and Sackur by the measurement of the internal friction
of sodium caseinogenate solutions. The magnitude of the internal
friction depends chiefly on the concentration of the caseinogen ions ;
very small quantities of free acid diminish the -friction (thus setting
free caseinogen from its salts), whilst correspondingly small quantities
of alkali increase it.

The Salt Formation of Serum-Globulin (Hardy).

To determine the combining powers of acids and bases with serum-
globulin Hardy employed two of the chemical methods (B I. and B
II.), viz.) direct titration and determination of the amount of globulin
(which is insoluble in pure water) soluble in given quantities of acids
and bases.

By the second of the two methods the more accurate results were
obtained. Various acids of the concentration 0*005 N were added
to suspensions of globulins containing from 0*28 to 4*18 grams in

1 See p. 53-

52 THE GENERAL CHARACTERS OF THE PROTEINS

100 c.c. The acid was in each case added gradually until the mix-
ture was only slightly opalescent and matched a given standard.
The amount of acid necessary to produce this given degree of opal-
escence was thereby determined. A large number of determinations
at different temperatures were carried out. If the mean amount of
hydrochloric acid necessary to produce this grade of solution of i
gram of globulin be taken as unity, the relative amounts of other
acids (given in equivalents) necessary to produce the same result is
given in the following table : —

Monobasic Acids. Dibasic Acids. Tribasic Acids.

HC1 ro H2SO4 rgi Citric acid 3-0

HNO3 0*995 Tartaric acid 1-994 H,PO4 2-9

CHC12.COOH ro Oxalic acid i-g H^BO, Very large excess

CC13.COOH ro

CH2C1 . COOH 1-05

H . COOH 1-25

CH3.COOH 5-2

CH2.CH3.COOH 7-56

It will be noticed here that the solvent power of strong acids is a
molecular function, and that HC1 = H2SO4 = H3PO4. It may be
recalled that Osborne, in the case of edestin, also observed that
HC1 = H3PO4)and that salts of the type B . H3PO4 are obtained.
The weaker acids require relatively far larger quantities to produce
the same grade of solution ; this is due to the readiness with which
salts undergo hydrolysis in dilute solutions.

Generally the results obtained indicate that serum-globulin forms
salts of the same type as the amino-acids, i.e., salts in which a mole-
cule of acid and base combine to form additive compounds without
replacement : —

/NH2 /NH2.HC1

CH2/ + HC1 = CH2/

\COOH \COOH

Similar experiments were carried out with bases. Taking NaOH as
unity, the relative solvent actions may be represented by the following
figures : —

KOH i

NaOH i

NH4.OH 0-98

Ba(OH)2 2-008

Urea and aniline also dissolve serum-globulin, but the quantitative
relationships were not determined. One point of considerable interest
appears from these numbers, viz., ammonia, which is a weak base,
has practically the same solvent power as sodium and potassium
hydroxides. Now the weak acids like acetic acid have considerably
less solvent power than the strong mineral acids, and these facts
indicate that serum-globulin has more decidedly marked acidic than
basic functions. In this respect it differs markedly from the plant
globulin edestin.

In respect to its action on bases, the behaviour of globulin is some-
what peculiar in that I molecule of sodium hydroxide is equivalent
to i molecule of barium hydroxide.

Measurements of the acidity and basicity of globulin by titration
in presence of indicators were also carried out. Freshly precipitated
and washed globulin reacts as acid to dialysed litmus, gives no colour

GENERAL CHEMICAL CHARACTERS OF PROTEINS 53

with phenol phthalein, and only slightly depresses the orange tint of
methyl orange. Such a reaction might be due to the carbonic acid
held in combination by the protein.

On titration with alkalis in the presence of phenol phthalein certain
marked phenomena were observed. It was noticed that just half as
much sodium hydroxide or potassium hydroxide was necessary to
produce solution as was necessary to produce a permanent red solu-
tion ; in the case of barium hydroxide, on the other hand, the point
of solution and the point of neutralisation to phenol phthalein coin-
cided. These facts suggest that globulin has two replaceable hydro-
gens, and can form acid salts, of which those of sodium and potassium
are soluble, and that of barium insoluble. The same relative amount
of ammonia as of sodium and potassium hydroxide is necessary to
produce solution ; as might be expected, however, the neutralisation
point with phenol phthalein is somewhat indeterminate. It was found,
as a general result of the neutralisation method, that 10 x 10 ~ 5
equivalents of alkali were necessary for the solution of I gram of
globulin, and 20 x 10 ~ 5 were necessary for neutralisation with
phenol phthalein as indicator ; 20 x 10 ~ 5 equivalents of strong
acids were also necessary to produce solution.

It is of interest to compare those numbers with the numbers
obtained by Hardy for dialysed acid and alkali albumin from egg-
white : —

Pink to Phenol-
Solution, phthalein.

Acid albumin from egg-white Na(OH) 27-3 57

'475 gram per 100 c.c. NH4(OH) 29 100

Ba(OH)2 57 65
Alkali albumin from egg-white Na(OH) 36

*363 gram per 100 c.c. NH^OH) 37 —

Ba(OH)2 77

Here again the molecular and not the equivalent quantities of
base exert the same solvent action, and there is evidence of the
formation of acid and basic salts.

Further investigations were made as to the acid and basic
functions of serum-globulin by the employment of physical methods.

By means of conductivity methods the basicity of globulin was

determined. The value — -1— - — ^ (see Lacqueur and Sackur's work

«i

above, p. 50) between ^ = 32 and v =32x32 was determined; as
a result, it was concluded that globulin is pentabasic.

The hydrolysis of both the acid and basic salts was determined
by ascertaining the rate of inversion of cane-sugar, and of hydrolysis
of methylacetate by acids and bases in the presence of globulin salts.
Owing to the change of the latter in the presence of acids, no very
conclusive results were obtained as to the hydrolysis of the acid salts.
The results, such as they were, indicated under the conditions of
experiments a distinct amount of hydrolysis : —

Glob. HC1 + HaO = Glob. OH + HC1.

With alkalis, under similar conditions, but little hydrolysis was
indicated. With a concentration of 2576 grams of globulin in a
litre neutralised by 17-12 x lo"5 equivalents of NaOH for each
gram of globulin, the hydrolysis, as indicated by methylacetate
catalysis, was only o'288 per cent, *>., only 0-288 per cent, of the

54 THE GENERAL CHARACTERS OF THE PROTEINS

sodium globulin compound was hydrolysed by water under the con-
ditions of the experiments.

Measurements were also made of the conductivities of the differ-
ent salts in gradually increasing concentrations. It was found that
in the case of the hydrochloride the molecular conductivity increases
much more for increase of volume than does the conductivity in the
case of the sodium hydroxide compound. This is due to the fact
that the acid functions of globulin are far more marked than its basic
functions, for from the equation

My = (i - *)/*y + */*HC1 or NaOH,

where My is the total conductivity, x the fraction of the salt
hydrolysed, yu,y the conductivity of the non-hydrolysed salt, and
/^HCI or NaOH the conductivity of the salt or acid, when yu,Hci or NaOH is
greater than //,y (as is the case with the globulin salts), the more x
increases with dilution the greater My must become. In the case
of the sodium hydroxide compound of globulin the increase of con-
ductivity with dilution is not markedly greater than it is in the case of
the inorganic salts mono- and disodium-hydrogen phosphates ; there is
little evidence of hydrolysis. In the case of the hydrochloride the
conductivity determinations indicate marked hydrolysis.

Determinations were also made of the ratios of the conductivities
of solutions of various acids and salts, both alone and after satura-
tion with globulin. In the case of very weak acids, such as boracic
acid, when a large amount of hydrolysis would be expected from the
equation

My would be nearly equal to //,acid. This is as a matter of fact
the case, for alt - = 0*9. In the case of hydrochloric acid,

M boracic acid

where the amount of hydrolysis is comparatively small, the value

is 0*24. In the case of ammonia, the free base conducts but

Af acid

little ; in fact, distinctly less than the salt formed by the combination
of globulin and ammonia, and the value alt -- is 2*3.

/^ammonia

All these facts lead to the conclusion that globulin forms with
alkalis salts which readily undergo ionisation and are good con-
ductors, but which, in contradistinction to the acid salts, are hydro-
lysed but little with water.

These salts of serum-globulin have certain characteristic properties
which are not shared with ordinary salts. It has already been men-
tioned that the solubility of globulin in acids was determined by
adding acids to a suspension until a definite grade of solution was
reached, which in most cases was that of minimal opalescence (a
process described by Hardy as that of matching). Conversely, if
such a solution be dialysed against distilled water, acid passes out,
and the globulin solution becomes more and more opaque, without
the separation at any time of a solid phase. Such an action Hardy
ascribes to hydrolysis and the formation of basic salts, which would
take place according to the equation
*GHAc

where G represents globulin and Ac acid.

GENERAL CHEMICAL CHARACTERS OF PROTEINS 55
With bases (B) a similar reaction would take place —

xGE + jHOH = (GH)y(GB)x_y + yBOH.

With increase in the value y\x> the size of the molecules of the
globulin salt molecules would increase until finally they are large
enough to diffract white light. Such molecules take part in electric
transport, and their velocity, as measured indirectly by conductivity
and by the " boundary " method, is exceptionally high, much higher,
in fact, than, according to Ostwald's law, could be accounted for by
ions containing such a large number of atoms. When large mole-
cules, such as the globulin salts, become ionised and take part in
electrical transport, they are said, according to Hardy, to form " pseudo-
ions," and when the magnitude of these latter become sufficiently
large, they attain the properties of matter in mass, being defined by
a surface, and moving under the influence of a surface contact differ-
ence of potential.

Somewhat similar to the phenomena just mentioned is the forma-
tion of opalescent solutions of caseinogen salts. In spite of these
peculiarities there is considerable evidence of a true chemical re-
action between serum-globulin and acids and bases. It is also
possible that globulin forms compounds with salts of the type GNaCl,
which readily hydrolyse with water. This would account for the fact
that such compounds are only stable in the presence of a large excess
of salt, i.e., they are soluble only in comparatively concentrated salt
solutions, from which they are reprecipitated on dilution. A globulin
acid salt is also but slightly soluble in dilute salt solutions, owing to
the double decomposition —

GHAc + NaCl = GNaCl + HAc.

The globulin alkali salts, however, are more readily soluble in salt
solutions than simple globulin. This Hardy considers to be due to
the compounds in question being of the following types : —
/NHa.HCl .,NH2.NaCl

R\ R\

\COOH \COOH

for the compounds of acids and salts. The one could only be formed
from the other by double decomposition. The compounds with
bases would be formed by the replacement of the hydrogen in the
carboxyl group.

The above chemical considerations go far to explain the peculiar
solubility conditions of the globulin type of proteins.

The Question of Pseudo-acid and Pseudo-base Formation.

Cohnheim has expressed the opinion that proteins may belong to
that class which Hantzsch has termed pseudo-acids and pseudo-bases.
Of such a class typical examples are the nitro-paraffins, e.g., nitro-
ethane. This body can exist in the two forms —

CH3 . CH2 . NO2 True nitro-form which is a neutral body.

CH3 . CH : NO . OH Pseudo nitro-form which is an acid body.

In the presence of alkali the true nitro-form is slowly converted into
the pseudo-form —

CHS . CH2 . N0a + NaOH = CH3 . CH : NO . ONa + H2O.

5

56 THE GENERAL CHARACTERS OF THE PROTEINS
|

If the sodium salt thus obtained be acidified it is reconverted into the
free acid and sodium chloride —

CH3 . CH : NO . ONa + HC1 = CH3 . CH : NO . OH + NaCl.
Such a solution has at first the electrolytic conductivity of the acid
pseudo-form + that of sodium chloride. After a time, however, the
pseudo-form gradually reverts to the true nitro-form, the conductivity
diminishes, and the solution attains finally the conductivity due to
sodium chloride alone. In a like manner the neutralisation of the
true nitro-body does not take place immediately, but there is a certain
latent period. The higher the temperature the shorter is this period.
Similarly certain pseudo-bases have been shown to exist, and
both are characterised by this latent period of neutralisation. The
conductivity changes, moreover, when neutralisation takes place, are
characterised by a high temperature coefficient, owing to the influence
of the increased temperature on the rate of change.

Neither the latent period of neutralisation nor the high tempera-
ture coefficient for the conductivity has been observed by Hardy in
his investigations on the serum-globulin, and he comes to the conclu-
sion that the hypothesis of Cohnheim that proteins act as pseudo-acids
and pseudo-bases is not justified by experimental facts. Similar
conclusions have also been arrived at by a different experimental
method by von Rhorer.

In concluding this section, emphasis must be laid upon the
anomalous behaviour in many respects of the proteins on treatment
with acids and bases. This behaviour is probably to be ascribed to
their capacity as colloids to adsorb inorganic substances, to which
reference has been already made in the introduction. This fact
renders it often impossible to discriminate between physical adsorption
and chemical combination, and for this reason some of the conclusions
recorded above must be accepted with reserve.

SECTION XVII.— THE PRECIPITATION OF PROTEINS BY SALTS
OF THE HEAVY METALS.

It has already been mentioned that the addition of salts of the
heavy metals, such as those of copper, mercury and lead, to protein
solutions produce precipitates. Numerous attempts have been made
to determine the composition of such precipitates and to isolate
definite protein salts of the heavy metals. There is a large literature
on the subject, and the various investigators have arrived at results
which are by no means concordant. The precipitates, which have
received the largest amount of investigation, are those produced by
the addition of copper salts. Certain of the products isolated have
a relatively large copper content, as the following analyses, due to
Ritthausen, of vegetable protein precipitates show : —

CuO. Ash.

" Gluten-caseinogen " from wheat 16-97

Legumin from peas I5'6i i'2i

„ from broad beans 14-10 3-05

,, from oats 13 '53

Conglutin from lupines i3'38-ii-6o o043-2'i6

The copper precipitate obtained from milk caseinogen also shows a
high copper content — 127 to 13*6 percent., according to the analyses
of Ritthausen and Pott.

Very widely differing results have been obtained by analyses of

GENERAL CHEMICAL CHARACTERS OF PROTEINS 57

the precipitate obtained by adding copper salts to a solution of egg-
white. Harnack, by precipitating in the presence of excess of pro-
tein, obtained a precipitate containing 1*35 per cent. Cu; when,
however, he precipitated in presence of excess of the copper salt, he
obtained a product containing 2*64 per cent. Cu. Other investiga-
tors have obtained products in which the percentage of copper varied
between 07 and 4' 15 per cent. Equally unsatisfactory and variable
are the results obtained by the precipitation with salts of other heavy
metals.1

The formation of the precipitates in question has more recently
formed the subject of investigations by Galeotti and by Pauli, who
have called attention to the complexity of the question.

Galeotti noticed that if a 3 per cent, albumin solution be allowed

N

to fall, drop by drop, into an — copper sulphate solution, a thick pre-
cipitate first forms, which redissolves on the addition of an excess
of protein, so that finally a homogeneous solution is obtained. If
more copper sulphate be added to this clear solution a precipitate is
formed again, which will redissolve on addition of excess of protein.
If an albumin solution be added, drop by drop, to a concentrated
copper solution, a precipitate will form as the drops first enter the
salt solution, but will disappear on shaking ; on addition of more
albumin a precipitate continues to be formed, which dissolves less
and less readily after each addition, until finally it becomes per-
manent. If, on the other hand, concentrated copper sulphate be
added to a protein solution a precipitate is formed, which gradually
dissolves on the addition of excess of the salt, giving a clear green
solution which contains protein in large quantities.

Silver salts behave in a somewhat different way. If an albumin
solution be allowed to drop into a I per cent, silver nitrate solution
a flocculent precipitate forms, which redissolves in an excess of pro-
tein. If, on the other hand, a solution of silver nitrate be allowed to
fall into an albumin solution, no precipitate forms until a certain
definite quantity of the salt has been added ; at this point a pre-
cipitate commences to form, which increases in quantity with each
additional drop of the salt. If albumin be added again, the pre-
cipitate redissolves.

Galeotti draws from his investigations the following conclusions :
I. Proteins form no definite compounds of* constant composition
with salts of the heavy metals ; the precipitates which form when
protein solutions and solutions of salts of the heavy metals are
mixed are simply loose compounds, the composition of which
depends on the conditions of precipitation. II. The precipitation
phenomena are often reversible,2 in the sense that the precipitates
are soluble in excess of either reagent. III. The composition of
the precipitate depends on the composition of the supernatant liquid
and is determined by the thermodynamical laws of chemical equili-
brium. IV. For the system albumin (egg or serum), water and salt

1 A table of these results is given in Galeotti's paper. References to the earlier
literature are given in Harnack's paper.

2 Galeotti uses the expression reversible, but only in the sense described in the
text, i.e., solubility on adding excess of either reagent. The reversibility is to be dis-
tinguished from that of the precipitation by, e.g., salts of alkalis, where the precipitates
redissolve on addition of water.

5*

58 THE GENERAL^ CHARACTERS OF THE PROTEINS

(copper sulphate or silver nitrate), it is possible to solve the problem
of chemical equilibrium by a graphical method, and, given the per-
centage composition of a complex, to determine into how many
phases it will separate, and the composition of each phase.

Pauli has investigated in some detail the conditions of precipita-
tion by zinc sulphate. He found that there were two maxima of pre-
cipitation when this salt was added to an egg-protein solution of a
certain strength. The precipitating power increased gradually be-
tween concentrations corresponding to 'OOI to 0*05 normal. From
this point the precipitating power gradually decreased, until the
concentration corresponding to that of a normal solution was reached.
From this point onwards till the concentration 2 N was attained,
no precipitation whatever took place. On increasing the concentra-
tion beyond this point, however, precipitation recommenced, and
increased with increasing strength of the salt solution until a second
maximum was attained. The precipitates obtained with the lower
dilution were irreversible, that is to say, they would not redissolve on
addition of water ; those obtained at the higher dilution were, on the
other hand, reversible and redissolved on dilution of the supernatant
fluid. The numbers just given refer only to solutions containing a
definite amount of egg-protein. In this particular case the limits
between which the maximum formation of the irreversible precipi-
tate and the incipient formation of the reversible precipitate took
place were 0*05 and 2 N. These limits are wider with a lower
protein concentration and narrower with a higher concentration.

These examples are sufficient to indicate the extremely complex
conditions governing the composition of the precipitates produced
when solutions of the salts of heavy metals are added to protein
solutions. In the majority of cases compounds of definite chemical
composition cannot be isolated. This statement does not, however,
preclude the possibility that certain proteins of more acidic character,
such as caseinogen, should form salts of definite chemical com-
position, and further investigation of precipitates with high copper
content, such as those derived from plant proteins and caseinogen,
seems desirable.

Various theories have been advanced to account for the nature
of the precipitates produced by salt precipitation. These have been
summarised in the paper of Pauli referred to. They may be more
fittingly discussed in dealing with the general chemical physics of
colloidal solutions and the theory of salt precipitation of colloids. In
the present state of knowledge the precipitates produced by the salts
of the heavy metals cannot be regarded as of sufficiently definite
chemical nature as to make them of value in fixing the characters of
the protein from which they are derived, and for this reason they
have been treated in a somewhat cursory manner.

SECTION XVIII. — THE OXIDATION OF THE PROTEINS.

In recent years the chemical examination of proteins, with a
view to elucidation of the constitution, has been confined chiefly to
the study of the hydrolysis products. This method, apart from the
value of the results yielded, is a logical one, in view of the fact that
all evidence with regard to the biological utilisation points to the
fact that the first stage of protein degradation is one of hydrolysis.

GENERAL CHEMICAL CHARACTERS OF PROTEINS 59

Nevertheless several investigations on the direct oxidation of
proteins without previous hydrolysis deserve mention.

Oxidation with Permanganate.

The oxidation with permanganate solution has been the subject
of repeated studies.1 The chief investigations are due to Maly and
von Furth. Other investigators were Be*champ, Sabbotin, Pott,
Briicke, Chandelon, Low, Siegfried, Bondzynski and Zoja, Bernert
and Ehrmann.

Maly, under the name of oxyprotosulphonic acid, described a pro-
duct which he obtained by oxidising egg-white with half its weight
of potassium permanganate in the cold. Under the name of peroxy-
proteic acid he described a product which was obtained by slowly
adding small portions of permanganate to a protein solution, at room
temperature, until only very slow oxidation took place. This acid,
isolated by precipitation with salts of heavy metals (mercury,
lead, etc.), gave a marked biuret reaction, but was not precipitated
by the ordinary alkaloidal reagents. On gentle treatment with
baryta water a large amount of ammonia was evolved, and separation
of oxalic acid in the form of its barium salt took place. By con-
tinued hydrolysis with baryta, lasting for several days, glutamic acid,
leucine, formic, acetic and ben zoic acids were obtained.

Further investigations on the oxyprotosulphonic acid from egg-
white were carried out by Bernert and Ehrmann.

The most complete of the recent investigations on the perman-
ganate oxidation products are due to von Furth. He oxidised
caseinogen with four times its weight of potassium permanganate at
room temperature, the reaction taking several weeks to complete.
A product was obtained which was resistant to further action of the
oxidising reagent. This was shown to consist of at least three pro-
ducts of high molecular weight, which gave the biuret reaction, but
not the Millon, xanthoproteic or Hopkins reactions, and which
could be separated by fractional precipitation with silver nitrate
(A), lead acetate (B), and mercuric acetate (C). From the three
peroxyproteic acids thus obtained, the esters could be readily pre-
pared by means of alcoholic hydrochloric acid. On treating these
acids with barium hydroxide a scission took place of oxalic acid, in
the form of its barium salt ; there was also a considerable loss of
nitrogen. The substances obtained in this way were designated by
von Furth as desamino-proteic acids, and they yielded on hydrolysis
glutamic acid, leucine, benzoic acid and ammonia. Unlike the
oxyproteic acids, from which they were derived, they were no longer
resistant to further oxidation with permanganate ; the scission of the
oxalic groups had left a new position of weakness in the molecule,
and the desamino-acids readily oxidised on further addition of the
oxidising agent, yielding a mixture of substances of a new class
called the kyroproteic acids, which gave a marked biuret reaction.
By means of lead acetate, these could be separated into acids which
contain a large quantity of oxygen, which could be readily oxidised
further. The kyroproteic acids, furthermore, on treatment with nitrous
acid, readily lose half their nitrogen, relatively five times as much as is
lost by caseinogen by similar treatment.

*A succinct account of the earlier literature is given in von Fiirth's paper, 1905.

6o THE GENERAL,CHARACTERS OF THE PROTEINS

There is not sufficient experimental material to draw any very
definite conclusions as to the course of oxidation of the proteins, but
some suggestions of Hofmeister seem reasonable. If the typical
polypeptide grouping of a protein be represented as follows—
— NH— CH . CO— NH . CH— CO-

I I

R R'

the first stage of the oxidation would be —

— NH— CH— CO— NH . CH— CO-
COOn COOH
from which, by CO2 scission, the grouping

— NH— CH2 . CO— NH . CH2 . CO—
would be formed, which on further oxidation would yield a group

— NH . CO . CO— NH . CO . CO—

From such a group, by hydrolysis, oxalic acid and ammonia would
be obtainable. It is of interest to note that Zickgraf, Seemann, and
Kutscher and Schenck have obtained by the oxidation of proteins
oxaluramide —

CO — NHX

I >o

CO.NH2NH2/

which contains the complex — NH— CO— CO— NH—

The hypothesis is, however, insufficient to explain all the facts
connected with the oxidation, as the amount of amide nitrogen,
which can be eliminated with nitrous acid, does not correspond with
the amount of oxalic acid.

Oxidation with Hydrogen Peroxide and Ozone.

The oxidation of proteins by hydrogen peroxide has been studied
by Wurster and by F. N. Schulz, and the action of ozone has been
studied by v. Gorup-Besanez and Harries and his pupils. Wurster
and Schulz noticed that egg-albumin (Wurster used egg-white solu-
tion and Schulz the crystallised product) on standing with excess of
hydrogen peroxide in neutral solution at 37° C. deposited after a
time the protein in an insoluble form. The product thus obtained
was subjected to a detailed examination by Schulz, who found that
it did not differ very greatly from the original protein in its per-
centage composition. The ratio of hydrogen, carbon and nitrogen
was practically unchanged, but it contained about 2'6 per cent, more
oxygen. Schulz designated this substance- oxyftrotem, and he showed
that, unlike the oxyproteic acids, it yielded the ordinary protein re-
actions. It represents, therefore, a simple oxidation product.

Blumenthal and Neuberg have shown that gelatin, in the pre-
sence of an iron or copper salt, yields, on oxidation with hydrogen
peroxide, acetone. The same product has been obtained by Orgler
from egg-albumin, using the same method.

The action of ozone on caseinogen has recently formed the sub-
ject of an extensive research by Harries and Langheld. By pro-
longed action of ozone on the sodium salt considerable chemical
change took place. A product was obtained which gave a marked
biuret reaction, but no reaction with the xanthoproteic, Millon or
Hopkins reagents (cf. oxyproteic acids). It gave a precipitate with

GENERAL CHEMICAL CHARACTERS OF PROTEINS 61

phenylhydrazine, indicating the presence of aldehydic and ketonic
groups, and could be separated into three fractions by means of
phosphotungstic acid and lead acetate. This operation was carried
out as follows : the original product was precipitated by phospho-
tungstic acid (in the presence of sulphuric acid) ; to the substance,
regenerated by barium hydroxide from this precipitate, lead acetate
was added, and the greater part precipitated (fraction A) ; the filtrate
from the lead precipitate formed the second fraction (B, i.e., the
fraction precipitable by phosphotungstic acid, but not by lead
acetate); from the filtrate from the phosphotungstic precipitate a
third fraction was obtained (C), which was not precipitable with lead
acetate. These three fractions were hydrolysed by Emil Fischer's
method, but no very essential difference between them could be
detected. Fraction C yielded, however, practically no leucine, where-
as fractions A and B did. The other amino-acids obtained by
hydrolysis were alanine, valine, aspartic and glutamic acids. Glycine
and proline could not be isolated. It is noteworthy, furthermore,
that tyrosine, phenylalanine and tryptophane appeared to be absent ;
it seems probable that ozone had attacked the aromatic groups of
the protein molecule.

Action of Nitric Acid.

The principal product obtained by the oxidation of nitric acid is
oxalic acid. On carefully dissolving proteins in nitric acid and
afterwards adding water, yellow substances, known as xantho-pro-
teins, can be obtained. These have been investigated by von Furth
(Habilitationsschrift, Strassburg, 1899).

SECTION XIX.— THE ACTION OF HALOGENS ON PROTEINS.

Investigations on the action of halogens on proteins date back to
1848, when Mulder described a "protein chlorous acid," which he
obtained as a precipitate when chlorine was passed into a solution of
egg-albumin.

Since that date the preparation of halogen derivatives of proteins
has been the subject of numerous investigations ; owing, however, to
the complexity of the possible reactions, and the variations in the
composition of the products obtained under different conditions of
experiment, the preparation of halogen derivatives has, up to the
present, been but little applied to the characterisation of individual
proteins ; the composition of the products depends too much on the
details of the preparation.

Chlorine Derivatives of Proteins.

In more recent times the chlorine derivatives have been studied
by Rideal and Stewart, Hopkins and Pinkus, Blum and Vaubel,
Habermann and Ehrenfeld, and by Panzer. Rideal and Stewart,
and Hopkins both confirmed the older observation of Mulder with
regard to the formation of a precipitate when chlorine is passed into
a protein solution, and the two former investigators proposed to
found upon this reaction a method for the quantitative determina-
tion of proteins.

Hopkins and Pinkus passed chlorine into a cold protein solution
till the latter was saturated, when a thick precipitate suddenly
formed; this was purified by solution in a I per cent, sodium hydroxide

62 THE GENERAL CHARACTERS OF THE PROTEINS

solution, and precipitation from the latter by dilute acetic acid.
From egg-albumin (not crystalline) a product containing 3*62 per
cent, of chlorine was obtained. By a modification of the method, pro-
ducts containing other quantities of chlorine could be obtained. Thus
by chlorinating in the presence of potassium chlorate, dissolving the
precipitate thus formed in alcohol, and adding ether to the alcoholic
solution, a derivative was obtained which contained 6-41 per cent, of
chlorine. As will be noticed later, a definite series of bromine
derivatives could be obtained from proteins, each with a constant
amount of bromine, by varying the method of preparation. It was
not found possible to obtain such a definite series either with chlorine
or iodine preparations.

Blum and Vaubel prepared chlorine preparations from proteins
in the presence of sodium hydroxide ; they state that it is impossible
to obtain products containing the maximum amount of combined
chlorine in the presence of free acid, which is formed by the action of
halogen on the protein ; consequently they carried out the reactions
in the presence of sodium hydroxide, which was added from time to
time to neutralise the acid as it was formed in the reaction. They
obtained from egg-albumin and caseinogen preparations containing
2 per cent, of chlorine.

Habermann and Ehrenfeld prepared a chlorine derivative from
caseinogen. They dissolved 100 grams of the protein in 700
c.c. of 5 per cent, potassium hydroxide solution, to which was added
50 grams of potassium chlorate; through this solution hydrogen
chloride gas was passed. After completion of the reaction the
liquid was filtered off from the potassium chloride formed during
the reaction, and diluted with water, whereupon the chloro derivative
separated, which contained between 13 and 14 per cent, of halogen.

Panzer also prepared a chlorine derivative of caseinogen. He made
a paste of i kilogram of the protein with 4 litres of 20 per cent,
hydrochloric acid ; to this 450 grams of potassium chlorate were
added in small quantities at a time, and considerable evolution of heat
took place. After cooling water was added, and the undissolved sub-
stance filtered off. This was a chlorine derivative containing about
8*3 per cent, of chlorine — considerably less than that in the prepara-
tion which Habermann and Ehrenfeld obtained by a similar method.

Bromine Derivatives of Proteins.

Some preliminary investigations on the action of bromine on
proteins were published by Loew, but most of our knowledge on this
subject is due to the investigations of Hopkins and Pinkus. As
already mentioned, they obtained different products with definite
bromine content by varying the methods of preparation. They
obtained three distinct bromine derivatives from egg-albumin.
Derivative I. was obtained by treating the protein solution in the
cold with bromine until the solution was distinctly coloured ; a
precipitate (crude bromination product) formed suddenly ; this was
dissolved in I per cent, sodium hydroxide, from which it was
precipitated by the addition of acetic acid. The precipitate was
dialysed against water, and then washed with alcohol, in which it
is only slightly soluble. It contained 3*92 per cent, of bromine.
Derivative II. was prepared by dissolving the crude bromination

GENERAL CHEMICAL CHARACTERS OF PROTEINS 63

product when still moist in hot alcohol, and allowing the solution to
drop into ether ; a product was thereby obtained which was easily
soluble in alcohol, and which could be purified by repeated solution
in this solvent and reprecipitation with ether. It contained 10*82
per cent, bromine. Derivative III. was prepared by dissolving the
moist crude bromination product in alcohol containing bromine, and
pouring the solution thus obtained into ether containing bromine.
A product resulted which, after washing with ether, contained 14*9
per cent, of bromine.

These derivatives can be converted one into the other. Thus,
for example, by dissolving derivative II. in sodium hydroxide solution,
and then adding acetic acid, a product containing only about 4 per
cent, bromine can be obtained. Conversely, if derivative I. be
added to alcohol containing bromine, and the solution thus obtained
be thrown into ether containing bromine, a product is obtained
which contains about 15 per cent, bromine (derivative III.).

Derivatives of the third class, containing the largest percentage
of bromine, were prepared from several other proteins. The
bromine content of the derivatives thus obtained may be regarded
as characteristic for each protein. From varying fractions of egg-
albumin substances containing from 1279-16*48 per cent, halogen
were obtained ; from serum of different fractions the bromine deriva-
tives contained from 12*1 5-12-94 per cent ; from serum-globulin they
contained from 13*53-14*03 per cent. The bromine derivative of
caseinogen contained 11*17 Per cent, that of proto-albumose 16*30-
17*12 per cent., and of deutero-albumose 17*63 per cent.

It was not found possible to obtain such definite series of
derivatives of chlorine or iodine derivatives ; nevertheless, evidence
was obtained that such series existed, although substances with definite
halogen content were not always obtainable. It was not found possible
either to obtain iodine derivatives, corresponding to derivative III.

Blum and Vaubel also obtained bromine derivatives of proteins,
using the method already mentioned, viz.) treating with halogen in
slightly alkaline solution ; from egg-albumin and caseinogen they
obtained products with between 4 and 5 per cent of bromine.

Iodine Derivatives of Proteins.

The earlier investigations on the iodine derivatives are due to
Bohm and Berg, and Jendrassik.1 The two former noticed the de-
colorisation of iodine by protein solutions, and, by coagulation, iso-
lated a product from which, by dialysis and washing, the iodine could
be removed. Jendrdssik determined the amount of iodine which
could be decolourised by a protein solution. Other investigations
were published later by Liebrecht and by Lepinois, who obtained
products containing respectively 17*8 and 21*6 per cent iodine, part
of which, at any rate, was in stable combination. The chief
systematic investigations on the iodine derivatives, apart from the
ones on the halogen derivatives generally of Hopkins and Pinkus,
and Blum and Vaubel, already mentioned, are those of Hofmeister,
Kurajeff and C. H. L. Schmidt.

Hofmeister worked with crystallised egg-albumin. He treated
20 grams dissolved in 400 c.c. water with 10 grams potassium

1 These earlier investigations are discussed in Hofmeister's paper.

64 THE GENERAL, CHARACTERS OF THE PROTEINS

iodide, 5 grams potassium iodate, and 4 c.c. concentrated sulphuric
acid for four hours on a water-bath. A brown precipitate was thereby
obtained, which was dissolved in ammonia and precipitated from this
solution by acetic acid ; this solution and precipitation was repeated
several times, and a product was obtained, which, after washing with
potassium iodide and then water, was found to contain 8*93 per cent,
of iodine.

Kurajeff carried out a series of experiments under varying condi-
tions. The treatment with iodine was carried out at 40°, both in acid
solutions and in the presence of magnesium carbonate ; he used as
source of iodine in some experiments, potassium iodide and iodate
in the presence of acids, in others iodine dissolved in potassium
iodide, with small quantities of iodate. When crystallised serum-
albumin was treated with these varying mixtures at 40-50° for from
three to seven days, preparations were obtained containing from
about 10-12 per cent, of iodine. By means of varying mixtures at
100° (one to six hours), preparations containing between 11*48 and
1 2*28 per cent, of iodine were obtained.

Hofmeister's method appears to give more constant results.
With egg-albumin (crystallised) at the lower temperature, with
iodine, potassium iodide and potassium iodate, preparations con-
taining from 8-29-8-42 per cent, iodine were obtained. When,
however, potassium iodide, with iodic and sulphuric acids, was used
(five days at 40° C), a preparation containing only 5-94 per cent of
iodine was obtained. It may be recalled that Hofmeister's preparation ,
obtained by treatment with potassium iodide and iodic and sulphuric
acid for four hours at 90-100°, contained 8-93 per cent, iodine.

Blum and Vaubel claim to be able to obtain products with con-
stant iodine content by their method already referred to ; they
propose to designate this the Blum- Vaubel iodine number, and to
make the preparation under the following conditions : The protein
is either dissolved or suspended in water, to which sodium bicar-
bonate is added, and the mixture is then warmed to 40-50° C.
Iodine dissolved in potassium iodide solution is then added in small
quantities at a time, till the iodine colour becomes permanent (i.e.,
does not disappear after half an hour). The mixture is then cooled
and filtered, sodium hydroxide is added in excess, and immediately
afterwards acetic acid. If necessary alcohol or acetone is added
to complete the precipitation. The precipitate is purified by reso-
lution in alkali and reprecipitation by acid, and then washed by
water and alcohol until the washings are free from iodine. The per-
centage of iodine contained in the dried preparation is the " iodine
number " of the protein. The iodine numbers found for various pro-
teins by Blum and Vaubel are as follows : Serum-globulin, 8'45 ;
serum-albumin (preparations by different methods), iix>2 and 9^93 ;
" -muscle-albumin," 1037 ; egg-albumin, 7*1.

The above examples show sufficiently the variations in the iodine
content of preparations prepared by different methods.

Properties of the Halogen Derivatives of the Proteins and Nature of
the Action of the Halogens.

The properties of the halogen derivatives have been to a great
extent indicated. They are readily soluble in alkaline solutions,

GENERAL CHEMICAL CHARACTERS OF PROTEINS 65

from which they can be precipitated by means of acids. Some of
them are, however, soluble in excess of acid. The higher bromine
derivatives are soluble in alcohol, but insoluble in ether, chloroform,
benzene and other organic solvents. They cannot be precipitated
from their solution in alkali by the majority of the alkaloidal reagents,
although they can be " salted out " by the addition of ammonium
sulphate. None of them have been obtained in a crystalline form.
They give the biuret reaction, but not the reactions of Millon or
Adamkiewicz (Hopkins).

The possible reactions are very complicated, for the amount of
hydrochloric, hydrobromic or hydriodic acid eliminated in the re-
action is far larger than that which can be accounted for by the mere
substitution of hydrogen by halogen. Concurrent oxidation must
also take place.

Some light has recently been thrown on the action of chlorine on
amino bodies by the researches of Raschig and of Cross, Bevan and
Briggs.

Raschig has shown that chlorine and alkaline hypochlorites act
upon ammonia with the formation of chloramines : —

NH3 + M'OCl = NH2C1 + M'OH.

Chloramine is capable of reacting with iodides with the liberation
of free iodine : —

NH2C1 + 2HI = NH3 + HC1 + I2.

Proteins seem to undergo a similar reaction, and it is suggested
by Cross, Bevan and Briggs that such a reaction may serve as a
measure of the reactive amino groups in the protein molecule. It is
possible that the treatment with hypochlorites in alkaline solution
will bring about a simple substitution, and that by means of the
iodide reaction the chloramine groups in the protein molecule can
be estimated. Evidence has already been brought to show that the
hypochlorite does not react to such an extent as free chlorine. It
is, furthermore, only the chlorine in the chloramine radical which
sets free iodine from combination with iodides.1

The action of halogens is, however, far more complex than mere
substitution in the sense of the chloramine reaction. Substitution
may take place in groups other than the amino groups. Attention
has been already called, furthermore, to the fact that far more of the
hydrogen halide is set free in the reaction than can be accounted
for by mere substitution. Oxidation must take place concurrently
with substitution. Some idea as to the reaction may be obtained by
the comparison of the empirical constitution of the halogen de-
rivative with that of the original body. Hofmeister, KurajefT and
the other investigators give several analyses which throw light on
this point One example will suffice, viz., the analysis of crystallised
egg-albumin, and its iodo derivative as prepared by Hofmeister : —

c. H. N. i. s. o.

Albumin 53'28 7-26 15*00 i«i8 23-28

lodalbumin 47^92 6'6o I4'i7 8-95 1*26 2foo

From these numbers it will be evident that in the formation of the

1 Further research on this " chloramine factor" is necessary. A method is sug-
gested in the paper of Cross, Bevan and Briggs, which is, however, only of a prelimin-
ary character.

66 THE GENERAL CHARACTERS OF THE PROTEINS

iodo derivative some carbon complex which is poor in nitrogen has
been eliminated from the molecule ; this Hofmeister assumes to be
a carbohydrate group.

Again, Blum and Vaubel, by their method of forming these
halogen derivatives in the presence of alkalis, bring some evidence
to show that a separation of some sulphur body takes place, although
this probably does not happen when other methods are employed.

Another significant fact with regard to the properties of the
halogen derivatives is their failure to give a positive result when
tested by the Millon and Adamkiewicz (Hopkins) reaction. This
fact indicates either that the aromatic groups are so substituted by
halogen that they fail to give these reactions, or that they are en-
tirely destroyed.

Investigations on the groups that have been eliminated during
halogenisation have been undertaken by Schmidt, who has isolated the
simpler bye-products of the reaction. He has shown that ammonium
iodide and ammonium iodate are always formed when the iodine deriva-
tives are prepared from egg-albumin by the Hofmeister reaction. This
indicates the scission of some amido group. Investigations with
simpler bodies, such as ammonium salts, urea, arginine, aspartic acid
and guanidine, showed that iodine is capable of causing the scission
of NH2 groups from those bodies where the group is combined
through a carboxyl or imido group to the molecule, such as in urea
and in guanidine. In other bodies, such as aspartic acid, where the
NH2 group is directly united to a carbon atom, no iodate could
be detected as a bye-product of the reaction. The reaction with
ammonium salts may be represented by the following equation : —

6(NH4)2SO4 + 61 + aH2O = 5NH4I + NH4IO3 + 6NH4HSO4.

It appears, therefore, from the somewhat limited number of Schmidt's
experiments, that the estimation of the free ammonium salts elimin-
ated may form a measure of the number of amido groups in the
protein molecule. The reaction is, however, somewhat complicated,
and the scission of amido groups cannot be directly measured by
the iodate eliminated, for, owing to concurrent oxidation, relatively
large quantities of hydriodic acid are formed simultaneously, and
this, when it reaches a certain concentration, reduces the iodate
according to the following equations: —

HI + NH4IO3 = NH4I + HIO3.
HIO3 + sHI = 61 + aH2O.

The amount of iodate formed can, therefore, only reach a certain
limit.

Other bye-products isolated by Schmidt are iodoform, carbonic,
formic and acetic acids, and possibly also para-iodopyrocatechin.
These have been estimated quantitatively, and their formation is
ascribed to the destruction of the tyrosine complex. Certain con-
clusions of Schmidt's later papers are not in full concordance with
those arrived at earlier, especially with reference to the formation
of iodates only by certain amino groups.

Sufficient has been said to show how complex are the possible
reactions which can take place when proteins are treated with
halogens. Nevertheless, it is conceivable that certain standard con-
ditions for halogenisation can be fixed, by means of which halogen-

GENERAL CHEMICAL CHARACTERS OF PROTEINS 67

protein derivatives can be prepared with fixed halogen content, which
could serve for the characterisation of individual proteins. The
bromination method of Hopkins and Pinkus, and the iodination
method of Blum and Vaubel, might, under certain stringent specified
conditions, be employed, and in this way a factor obtained for the
characterisation analogous somewhat to the " Hiibl number " for
fats. The suggestion of Cross, Bevan and Briggs for a "chlor-
amine " number is also worthy of further attention.

SECTION XX. — THE ACTION OF NITROUS ACID ON PROTEINS.

When proteins are treated with nitrous acid a considerable froth-
ing and evolution of nitrogen takes place. This action has formed
the subject of many investigations. Schiff, on treatment of egg-
proteins with nitrous acid, obtained a product which no longer gave
the biuret reaction, which was insoluble in water, and to which he
gave the name desamido-albumin, on the assumption that nitrous
acid had destroyed the amido groups (z>., groups containing the
complex — CO . NH2). The action of nitrous acid on proteoses and
peptones formed the subject of investigations by Paal and Schrotter,
who obtained products differing in many properties from the original
substances. In more recent times the action of nitrous acid has
formed the subject of investigations by Levites, and more especially
by Skraup and his pupils. Both these investigators failed to confirm
the observation of Schiff, that the product of the reaction yields no
biuret reaction. Levites found, moreover, that the products contained
as much amide-nitrogen as the original bodies. Skraup and his
pupils have prepared these desamido-proteins from several pro-
teins (caseinogen, gelatin, serum-globulin), and subjected them to
hydrolysis. They found that, with one exception, the hydrolysis pro-
ducts did not differ very essentially from the original proteins. They
failed, however, to isolate lysine from the desamido-proteins, although
this base was obtainable from the proteins themselves. The ele-
mentary compositions also did not differ very greatly from those
of the proteins ; the ratios of the carbon, hydrogen and nitrogen
were approximately the same; in the case of caseinogen thedesamido
body contained considerably less phosphorus than the protein from
which it was derived. The yield of desamido body varied consider-
ably with the different proteins, although approximately the same
methods of preparation were employed (mixture of sodium nitrite
and protein solution treated with acetic acid) ; from gelatin the
yield was about 100 per cent, of that of the protein, from caseinogen
70 per cent., and from crystallised egg-albumin about 50 per cent.
Obermayer, and more recently Treves and Salomone, have stated
that diazo compounds can be obtained from proteins by the action
of nitrous acid, and from these dye-stuffs can be prepared.

The Amino-Index.

As to the mechanism of the action of nitrous acid little is yet
known. E. Fischer and Koelker have shown that nitrous acid, when
acting on polypeptides, causes an evolution of nitrogen, the quantity
of which bears no very definite atomic ratio to the nitrogen of the
polypeptide. It is possible from the colour of the products that

68 THE GENERAL CHARACTERS OF THE PROTEINS

nitroso bodies are formed, which on hydrolysis might be expected to
yield the same products as the substances from which they are
formed, with evolution of ammonia (Paal).

On the other hand, the free amino groups in the molecule, as well
as the amide-nitrogen groups, might be attacked, as in the case of
asparagine —

2C4H8N203 + 2HN02 = 2C4H605 + 2N2 + 4H2O.

The amount of nitrogen evolved might serve as a measure of the free
amino and amide groups in a protein. It must be remembered,
however, that the polypeptides also act on nitrous acid in a some-
what indefinite manner (Fischer and Koelker) ; furthermore, in highly
complex substances, such as the proteins, it is also conceivable that
only a limited number of amino or amide groups may be the subject
of attack, owing to stereo-chemical reasons. Nevertheless, the amount
of nitrogen eliminated by nitrous acid treatment under certain speci-
fied conditions may be a fixed quantity and a characteristic of each
individual protein, and the determination of this so-called amino-
index has formed the subject of recent investigation by Horace
Brown and his co-workers. It has so far only been applied to a
limited number of proteins. When a pure amino-acid, such as
aspartic acid, is treated with nitrous acid, twice as much nitrogen is
evolved as the acid itself contains, one-half being derived from the
nitrous acid, e.g. : —

C4H7O4N + HNO2 = C4H6O5 + N2 + H-jO.1

In the case of proteins, therefore, one-half the nitrogen evolved,
expressed as a percentage of the total nitrogen, gives the apparent
proportion of the nitrogen present in the amino form, or, at any rate,
that proportion which will react with nitrous acid. This is termed by
Brown the amino-index, and is represented by the symbol Aol.
The following are the amino numbers for a series of proteoses and
peptones obtained by fractionating malt proteoses : —

Malt proteoses I. 4/0

II. 5-o

„ III. 20'0

Malt peptone I. 10*9

II. 19-3

Method.

The method of determining the amino-index adopted by Brown
and his co-workers is a modification of that previously employed by
Sachsse and Kormann for determining the amino-acids present in
certain technical samples. The principle consists in the treatment
of the substance under examination with nitrous acid in statu nascendi,
and the measurement of the nitrogen evolved after absorption of the
surplus nitric oxide carried over by the gas. Certain errors were
found by Brown to be inherent in the method as originally suggested
by Sachsse and Kormann. The chief sources are due to (i) the
residual air in the apparatus, or dissolved in the liquid ; (2) difficulties
attending the production of carbonic acid of high degree of purity
when this is employed for freeing the apparatus from air ; (3) diffi-
culties associated with the complete absorption of nitric oxide with

1 With asparagine containing an amido group only relatively half as much nitro-
gen is evolved. See equation already given above.

GENERAL CHEMICAL CHARACTERS OF PROTEINS 69

ferrous sulphate. In order to get the apparatus air-free the tube in
which the nitrous acid is allowed to act on the protein is connected
with the carbonic acid generating apparatus by means of a special
form of trap, from which the air is expelled by means of steam and
carbonic acid gas. The latter is evolved by the action of hydrochloric
acid on sodium carbonate solution in an apparatus specially designed
for the mixture of the liquids, in such a way as to ensure a steady
evolution of gas. The evolved nitrogen is collected over potash
solution in a modified form of Lunge nitrometer, in which it is
mixed with excess of oxygen, obtained by the electrolysis of water ;
in this way the nitrogen is freed from the nitric oxide. The excess
of oxygen is afterwards absorbed in alkaline pyrogallol solution, and
the volume of nitrogen evolved is finally measured over pure water.
Precautions must be taken to free all the liquids used in the ex-
periments from air. For full details reference must be made to the
original paper.

SECTION XXI. — ACTION OF FORMALDEHYDE ON PROTEINS.

From the presence of amino and imino groups in proteins it is
to be expected that they would enter into reaction with formalde-
hyde. The first observations on such a reaction are due to Trillat
and Hauser. The former noticed that on addition of concentrated
formaldehyde solution, egg-white was converted into an opaque gelat-
inous mass, whilst the latter noticed that gelatin, on treatment with
formaldehyde, was converted into a hard, insoluble substance. Blum,
somewhat later, noticed that the addition of a small quantity of
formaldehyde to egg-white solution caused the latter to lose its
capacity for coagulating on heating ; it remained, however, clear after
the addition of the aldehyde. The observations of Blum and of Trillat
do not appear to be concordant ; but the apparent discrepancy was
later explained by Schwarz, who showed that dilute solutions, especi-
ally in absence of salts, remain clear on addition of formalin, and
lose their coagulability, whereas more concentrated solutions become
turbid, but can be made to coagulate by the addition of salts. These
phenomena indicate that formaldehyde is capable of acting on proteins.

The reaction was studied in greater detail by Benedicenti. He
added dilute (2 per cent.) solutions of formaldehyde to protein solu-
tions, and estimated quantitatively at given intervals the amount of
formaldehyde which had not entered into reaction ; for this purpose
he employed hydroxylamine hydrochloride, which reacts with the
aldehyde according to the equation

NH2 . OH . HC1 + H . CHO = CH2 : N . OH + HC1.

By titrating the hydrochloric acid with standard potassium
hydroxide, using methyl orange as indicator, the amount of formalde-
hyde in a solution could be estimated. It was noticed that when
methyl orange was employed as indicator the proteins themselves
acted as bases ; a certain amount of acid was therefore necessary to
neutralise the solution to this indicator before the addition of the
hydroxylamine. This alkalinity diminished as the action of formal-
dehyde proceeded ; this fact indicated that the alkalinity was due to
the presence of amino groups, and gradually diminished, as the
formaldehyde condensed with them to form methylene derivatives.

70 THE GENERAL CHARACTERS OF THE PROTEINS

The reaction with dilute formaldehyde solutions (e.gty 4 c.c. of a
2 per cent, solution added to 10 c.c. of a protein solution) was
somewhat slow ; the maximum amount of aldehyde had not, as a
rule, entered into reaction until after two to three weeks ; after this
interval it was found that I gram of gelatin (a 10 per cent solution
of which had been boiled to prevent subsequent setting to a jelly)
combined with 0*0135 gram formalin; 10 c.c. fresh egg-white
combined with 0375 gram, 2 grams powdered egg-white with
0*0360 gram, 10 c.c. blood-serum with 0*315 gram, 3 grams fibrin
with 0*0345 gram, and 5 grams caseinogen with 0*0294 gram for-
maldehyde. The compounds thus formed were no longer digestible
when treated with pepsin, but could be decomposed when distilled
with steam, and a digestible protein could be thereby recovered ; the
formaldehyde could also be quantitatively recovered in the distillate.
Similar results to those of Benedicenti have been obtained recently
by Treves and Salomone.

SchifT has also investigated the action of formaldehyde on
proteins. He added a concentrated solution of formaldehyde (40
per cent.) to a solution of proteins, and then estimated the acidity of
the latter. The reaction which takes place is assumed to be similar
to that which takes place with the amino-acids. The amino group
entering into reaction with the aldehyde forms methylene derivatives ;
the alkalinity due to the presence of such groups is thereby elim-
inated, and the acid, which before treatment acts practically as a
neutral body to most indicators, now becomes strongly acidic in
character, and can be titrated directly with alkalis, with the use of
phenol-phthalein. By using this method Schiff found that I gram
molecular equivalent of potassium hydroxide neutralised 3,231 grams
of egg-albumin and 4,680 grams of gelatin, after solutions of the
latter had been treated with formaldehyde. The titrations were
carried out in some cases directly after mixture of the proteins with
the aldehyde, and in other cases after the mixtures had stood for
twenty-four or forty-eight hours. The same amount of alkali was
required for neutralisation in each case. The result is not quite in
accord with those of Benedicenti, who found the reaction was only
complete after two or three weeks ; he used, however, only very weak
aldehyde solutions, whereas Schiff used the undiluted commercial
preparation (40 per cent).

The results seem to indicate that the reaction may be of use in
estimating the amino and carboxyl groups in individual proteins, and
thereby obtaining other factors for their characterisation. It has
been already employed by Sorensen in studying the process of
digestion of proteins by enzymes. As hydrolysis proceeds and the
polypeptide groups are broken down, the number of free amino and
carboxyl groups in a given amount of the solution increases ; by treat-
ing the products of digestion with formaldehyde at different intervals,
and then titrating the mixture with barium or sodium hydroxide,
using phenol- or thymol-phthalein as indicator, Sorensen has succeeded
in obtaining a new factor for the study of proteolysis by enzymes.

It seems possible that the amino and carboxyl factors in any
protein may be determinable by a similar method, if the suitable
experimental conditions be ascertained. Such factors might be of
value for their characterisation.

PART III.

BIOLOGICAL METHODS FOR THE IDENTIFICATION
AND DIFFERENTIATION OF PROTEINS.1

SECTION XXII.— THE PRECIPITIN REACTION.

WHEN the necessity arises for differentiating between nearly allied
proteins of different origins the ordinary chemical and physical
methods entirely fail. No reliable chemical methods exist for dis-
tinguishing, for example, between human blood and the blood of
other species, or between the muscular tissue of one animal and the
muscular tissue of another. Yet, in actual practice, both in forensic
medicine and in the ordinary routine of food -inspection and analysis,
the necessity for determining differences of this description frequently
arises. For this purpose biological methods, and especially the so-
called precipitin reaction, have hitherto been almost exclusively
employed.

The first observations dealing with this subject date from 1 897 and
are due to Kraus, who showed that by the injection of typhus bacilli
into an animal a serum was produced, which not only caused ag-
glutination of the bacteria, but also produced a precipitate with the
filtrate of the culture medium. The reaction was found to be specific
for certain substances contained in this medium.

In 1899 Bordet and Tschtistowitsch obtained quite similar re-
sults with animal cells and cell products. By the injection of horse-
serum, eel-serum, cows' milk, etc., into rabbits, sera could be obtained,
containing the so-called precipitin, which gave thick precipitates
with the substances used to produce them, and with these substances
(the so-called precipitinogens) only. A relatively simple biological
method was, therefore, available for distinction between proteins
from various sources.

A large number of investigations followed the observations of
Bordet, which were undertaken with the main object of determining
how far the reactions were distinctly specific.

Bordet himself showed that the serum of a rabbit, immunised
against cows' milk, produced a precipitate with this milk and not with
that of a goat, and Wassermann and Schiitze, Uhlenhuth and others
found a similar specificity for precipitins produced by egg-white
and blood. Of special interest are the blood precipitins. If human
blood be injected into a rabbit a serum is produced which gives a

1 There is a very large literature on this subject, which can be only very briefly dealt
with in this place. Excellent summaries are given together with the principal literature
references in the papers of L. Blum and G. Blume. Reference should also be made to
the exhaustive monograph on the blood test by Nuttall.

71 6

72 THE GENERA!, CHARACTERS OF THE PROTEINS

strong precipitate with human blood, but not with that of a goat or
a dog. Such a serum will, however, precipitate the blood of a
species nearly allied to that from which the precipitinogen is ob-
tained, and the precipitin produced by the injection of human blood
will precipitate that of an anthropoid ape, although the amount of
precipitate formed will differ quantitatively in different cases. The
specificity of the precipitin reaction is, therefore, not absolute.

In order to produce precipitins the substance injected must be
foreign to the animal employed. With a protein derived from the same
species precipitins (the so-called " iso-precipitins ") are obtained only
in exceptional cases. To produce the most favourable results the
animals employed for producing the precipitins should not be too
closely allied to the animal from which the precipitinogen has been
obtained, although Uhlenhuth has recently succeeded in obtaining a
precipitin for hares' blood, but not for that of a rabbit, by in-
troducing into the latter the blood of the former animal.

In addition to the experiments with native proteins, experiments
for production of precipitins from chemically changed proteins have
also been carried out. Precipitins have been produced from crys-
tallised egg- and serum-albumins, although, according to Obermayer
and Pick, the precipitin-producing property of these substances is lost
after repeated recrystallisation. The precipitinogenic property is not
lost, however, by heating, and precipitins can be produced by the in-
jection of coagulated proteins. The property does not appear to
be lost even by boiling with -J per cent, hydrochloric acid or sodium
hydroxide solutions; neither does it appear as if the property is
readily lost by the tryptic digestion of the precipitinogens, although
it is readily lost by the peptic digestion. For this reason it is not,
as a rule, possible to produce precipitins by the administration of
precipitinogens per osy although the formation may, in certain cases,
take place when a particular protein is ingested in such large quanti-
ties that it escapes the action of the peptic juice.

Proteins, therefore, which have undergone considerable changes,
either by oxidation or hydrolysis, still possess the property of pro-
ducing precipitins, which are specific for the species of animal from
which they have been obtained. There is, however, according to
Obermayer and Pick, another class of changed proteins in which
this kind of specificity has been lost. lodo-, nitro- and diazo-pro-
teins, for example, will also yield precipitins, but these are specific,
not for a particular animal species, but for other substances of the
same class ; thus a precipitin which has been produced by the in-
jection of an iodo-protein from ox-serum is not specific for ox-serum,
or even the iodo derivatives from ox-serum, but will precipitate iodo-
proteins from other sera, and even an iodo-protein derived from a
plant. Furthermore, although an animal cannot, as a rule, produce
a precipitin for one of its own proteins, it can produce one for a
changed protein. In this way a xantho-protein precipitin has been
produced by the injection of xantho-protein derived from rabbits'
serum into a rabbit. Obermayer and Pick think that the animal
species specificity is due to the aromatic groups, and that this
particular kind of specificity is lost by the treatment of the protein
by reagents, such as halogens, nitric acid, etc., which have a
tendency to destroy these groups.

BIOLOGICAL METHODS 73

The precipitate formation takes place most readily in neutral
solution, and is impeded by the presence of mineral acids or alkalis ;
the presence of salts is necessary, those of calcium exerting a speci-
ally favourable influence.

On heating a serum containing a precipitin, its capacity for
forming precipitates is lost; the changed precipitin (precipitoid)
retains, in spite of this fact, its capacity for combining with a sub-
stance contained in the precipitinogen.

As to the actual chemical nature of the precipitin, precipitinogen
and precipitate but little is known ; the substance of the last-named
appears to be derived chiefly from the precipitin containing serum
(Welsh and Chapman), although opinions on this point are not
unanimous. The precipitates, furthermore, are soluble in excess of
the precipitable substances.

The conditions of chemical equilibrium have been recently
investigated by Hamburger and Arrhenius. The precipitin used
in their experiments was contained in the serum of a calf which
had been immunised against horse's serum. In one set of experi-
ments a constant quantity of the calf-serum was added to varying
quantities of diluted horse-serum (i in 50) ; the mixture was
allowed to remain at 37° C. for one hour, and then centrifuged in a
funnel-shaped tube ending in a graduated capillary tube, in which
the precipitate formed could be collected and measured. In a
second series of experiments the quantity of horse-serum was kept
constant and that of the calf- serum varied. If the solubility of the
precipitate in physiological saline (the diluent used) were the only
factor which caused the diminution of the quantity of the precipitate,
then, according to the Guldberg-Waage hypothesis —

Concentration of precipitinogen x concentration of precipitin =
K(reaction constant) x concentration of dissolved precipitate.

The quantity of dissolved precipitate was, however, larger than
could be accounted for by the above equation, and Hamburger and
Arrhenius came to the conclusion that a soluble substance is formed
by the combination of the precipitate with some substance in the
precipitinogenous body (i.e., horse-serum), and that the conditions
are analogous to those existing in the Ca(OH)2 : CO2 reaction.

A detailed account of the precipitin reaction and of the theories
that have been advanced to explain the precipitin formation need
not be discussed here ; it remains, however, to consider briefly the
practical applications of the reaction and the technique of the
methods employed.

The method for determination of the species origin of a protein
is due chiefly to Wassermann and Uhlenhuth.1 It has been studied
in great detail by Nuttall, and has been recently modified by A.
Schulz and extended by him to the quantitative estimation of
mixtures of proteins. Owing to the fact already mentioned that
the precipitin reaction is not strictly specific for a protein of any
given species, great care is required in the application of the re-
action.

1 A recent detailed account of the method for technical purposes has been published
by Uhlenhuth and his co-workers (Arbeiten aus dem Kaiserlichen Gesundheitsamt, 1908,
vol. xxviii., pt. 3).

74 THE GENERA^ CHARACTERS OF THE PROTEINS

The precipitin is generally prepared by several injections, gene-
rally intraperitoneal, but sometimes subcutaneous or intravenous,
following one another at intervals of from three to six days.1 Rabbits
are the animals commonly employed. The more precipitin a serum
contains the less specific is it, i.e., the more readily will it precipitate
proteins other than the precipitinogen. For practical purposes, there-
fore, it is not advisable to employ precipitins of very high grade ; if
sera be obtained which give precipitates with bodies other than the
precipitinogens, it is advisable to dilute them before use. In de-
termining the origin of a sample of blood the material to be in-
vestigated (clothes, etc.) is extracted with physiological saline, and
the extract is filtered through a Berkefeld filter and diluted, so that a
solution containing 0*1 per cent, protein is obtained. To 2 c.c. of
such a solution o* I c.c. of the precipitin-containing serum is added.
The more nearly the protein in the material under investigation is
allied to the precipitinogen employed for the preparation of the
antiserum, the greater the dilution in which a precipitate will
appear. To determine, therefore, the origin of a given sample of
blood (e.g.y human blood), samples of other bloods should be used as
controls ; the precipitin prepared by immunising a rabbit against
human blood, for example, will give with the material under investi-
gation a precipitate in much greater dilution, should it contain human
blood, than it would if it contain blood from any other species.
Furthermore, the more nearly allied the species to that 'from which
the precipitinogen has been derived, the more readily will its protein
give a precipitate with the precipitin. This reaction has been ex-
tensively employed by Nuttall for determining the genetic relation-
ships of different species.

Another method of applying the precipitin reaction has been
recently introduced by Neisser and Sachs. When haemolysis of red
blood corpuscles is brought about by a serum the latter contains
two different bodies, both of which are necessary for the process,
viz., the heat-labile complement and the heat-stable amboceptor.
Gengou has shown that, when a precipitate is formed by bringing
together precipitin and precipitinogen in the presence of a haemolytic
serum, the complement disappears, even when the amount of pre-
cipitate is so small as to be hardly visible. A haemolytic serum
can be tested as regards its haemolytic power towards a given
suspension of red blood corpuscles. To a similar quantity of the
same serum may be added a precipitin-containing serum which is
not haemolytic towards the same corpuscles. If to such a mixture a
protein be added containing a substance which will form a precip-
itate with the precipitin, it will lose its haemolytic properties. In
this way Neisser and Sachs have succeeded in detecting human
blood in dilutions of I in 10,000, or even I in 100,000. For sug-
gested explanations of these phenomena reference must be made to
the original papers.

To illustrate the method of employment of the precipitin reaction
a short description of Schulz's method for the quantitative estima-
tion of proteins in mixtures is appended.

1 Nuttall generally used 5-10 c.c. of serum for each injection, but in some cases
smaller quantities. For details see his Monograph, pp. 54, 55.

BIOLOGICAL METHODS 75

Method.

Schulds Method of Protein Estimation by Means of the Precipitin
Reaction. — The experimental basis of the method depends upon the
fact that if an antiserum be added to varying dilutions of an extract
of its precipitinogen in physiological saline solution the turbidity due
to the precipitate formation will appear earliest in the most concen-
trated solutions ; the more dilute a solution the longer will be the
time interval before turbidity appears. The strength (value) of an
antiserum can be determined, therefore, by ascertaining the greatest
dilution of the precipitinogen, which is just sufficient to give a tur-
bidity within a given time interval, which for the purposes of experi-
mental work has been chosen as sixty minutes. If the precipitinogenic
protein be mixed with other proteins a more concentrated extract
will have to be employed to yield a turbidity within sixty minutes,
than would be the case if the pure precipitinogenic protein, unmixed
with others, had been employed. By determining the ratio of this
dilution to the dilution of the extract of the pure precipitinogenic body
necessary to produce a turbidity within one hour, the amount of this
substance in a given mixture can be ascertained. The principle of
the method can be best illustrated by the example given by Schulz.
A given amount of a mixture of horse flesh and other muscular
tissue, weighing 50*57 grams, containing x grams of the former,
was extracted with 100 c.c. of normal saline (fe). This extract gave
a turbidity within one hour with a given quantity of a serum pro-
duced by immunising a rabbit against horse flesh, when it was diluted
1 60 times. An extract of pure horse flesh gave a turbidity under the
same conditions and with the same quantity of the same serum, when
the extract was of such dilution that 820 parts corresponded to I part
of the meat. The value of the serum used, W, was, according to

Schulz's method of expression, = — . From these data the amount of

horse flesh present in the mixture can be calculated from the equation

= - — , whence;tr = 19*5. That is to say, 50*57 grams of tissue

lOOA? o2O

contained 19*5 grams of horse flesh. The amount actually added
to the mixture was 19*0 grams.

For the purposes of experimental work a serum not too rich in
precipitin should be employed. It should be prepared by intra-
peritoneal injection of the protein, the quantity of which is to be
estimated in a given mixture. The antiserum should be steril-
ised by filtration through a clay filter and kept in hermetically
sealed glass tubes of O'6 c.c. capacity. Its value should be freshly
tested against a pure precipitinogen whenever a quantitative estima-
tion is carried out. The extracts of the precipitinogen and sample
under investigation should be carefully filtered and perfectly clear.
Physiological saline is used for the extraction of the material, which
should be finely disintegrated, and kieselguhr serves as a good filter-
ing medium. As there is a loss of protein during filtration, the
precipitinogen and the sample under investigation should be treated
in as nearly as possible the same way and the extracts filtered through
exactly similar filters the same number of times. Various dilutions
of the extracts can be readily prepared in series. To 0*9 c.c. of each

76 THE GENERAL CHARACTERS OF THE PROTEINS

dilution contained in test-tubes of 53 x 7 mm., O'l c.c. of the pre-
cipitin containing serum, measured from a i c.c. pipette graduated in

- — c.c., is added, the mixture is shaken and the series of test-tubes
100

allowed to stand for one hour. The dilutions of the pure precipi-
tinogen and of the sample under investigation, which produce tur-
bidity within this time, are readily observed, and the quantity of
protein to be estimated can be readily calculated by the method
already described. The dilutions are arranged by Schulz according
to certain definite scales. The method has its limitations, for whereas
it can be successfully applied to the estimation of a given description
of muscular tissue in mixtures, even when the latter are not quite
fresh, it fails when applied to egg-proteins.

The Complement Removing Action of Neisser and Sachs. — The
following method ,was employed for distinguishing between human
blood and blood originating from other species. 0*1 c.c. of anti-
serum (i.e., serum of an animal immunised against human blood)
+ 0*05 c.c. complement (fresh guinea-pig serum) + varying quan-
tities of normal sera of different origins, made up always to a volume
of i c.c. with physiological saline, were mixed and allowed to stand
for one hour at room temperature ; to each of the test mixtures was
added i c.c. of a 5 per cent, suspension of sheep's blood + 0*0015
c.c. of amboceptor-containing serum (serum of a rabbit immunised
against ox blood, such a serum acting haemolytically also towards
sheep's blood) ; the mixture was then allowed to stand at 37° C.
for two hours. Quantities of 'ooi c.c. of human serum caused total
inhibition of haemolysis ; in presence of similar quantities of serum
from monkeys a moderate amount of haemolysis took place ; whilst
much larger quantities ('Oi c.c.) of sera from the rat, pig, goat, rabbit,
ox and horse were incapable of preventing complete haemolysis.

A simplification of the above process is possible. Normal
rabbit's blood haemolyses sheep's blood, and this can be employed
instead of a prepared immune serum. It was found in Neisser and
Sachs' experiments that 0*25 c.c. of rabbit's serum could completely
haemolyse i c.c. of a 5 per cent, suspension of sheep's blood. 0^25
c.c. of this serum was, therefore, mixed with the liquid supposed to
contain human blood and the corresponding antiserum and allowed
to remain for one hour at 37°. I c.c. of the 5 per cent, sheep's-blood
suspension was then added, and the mixture incubated again at 37°
for two hours. The absence of haemolysis indicated the presence of
human serum.

The principle of the method has been applied to the identification
of many proteins other than those contained in blood and serum. An
account of the researches with literature references is given in the
of Blume.

BIOLOGICAL METHODS 77

CONCLUDING REMARKS.

In the foregoing pages the chief properties of the proteins have
been passed in review with the object of determining those which
might serve for the purposes of isolation and identification. As a
result it must be admitted that the methods available at the present
moment are extremely defective.

The separation of the proteins from one another depends almost
entirely on their differences of solubility in alcohol, water, salt solu-
tions, or dilute acids and alkalis. To the incompleteness of the
separation by differential extraction or by salt precipitation attention
has been already drawn. Furthermore, there are large classes of pro-
teins, to which even these methods are inapplicable, viz., those which
are quite insoluble in the solvents mentioned. For the separation of
mixtures of proteins of these classes no methods are available.

The methods for the identification of proteins are again extremely
defective. The unreliability of the physical constants has been
repeatedly emphasised. There remain the biological methods, which
in recent years have received considerable attention, and a few isolated
chemical factors, such as the sulphur content and the distribution of
nitrogen in the hydrolysis products. The biological methods are,
however, in many cases uncertain, and whilst they are generally
available for the physiologist or pathologist, they are entirely beyond
the scope of the worker whose only resource is a laboratory devoted
to pure chemistry ; the biological reactions, furthermore, require a
considerable interval of time for their accomplishment. For these
reasons their general application must be limited, and they are, for
the most part, quite unavailable for the purpose of the technical
examination of products, such as falls, for example, within the range
of work of the food analyst.

For these reasons reliance will have to be placed chiefly on the
purely chemical methods for the identification of proteins. Much
work remains to be done in the elaboration of such methods, and it
is not too much to hope that, with the rapidly increasing knowledge
of proteins, a reliable technique will be developed in the near future,
such as exists already for the identification and differentiation of fats.
It is a necessity for the physiologist, the pathologist and the technical
chemist.

BIBLIOGRAPHY.

"SALTING OUT" OF PROTEINS.

BOMER. Zinksulfat als Fdllungsmittel fur Albumosen. Zeit. f. anal. Chem., 1895,

34. 562.
BURCKHARDT. Beitrdge zur Chemie und Physiologic des Blutserums. Arch. exp.

Path. Pharm., 1882, 16, 322.

DE*NIS. Nouvelles etudes chimiques. Paris, 1856.
DENIS. Memoir e sur le Sang. Paris, 1859.

HALLIBURTON. The Proteids of Serum. Journ. Physiol., 1884, 5, 152.
HALLIBURTON. On Muscle Plasma. Journ. Physiol., 1887, 8, 133.
HALLIBURTON. The Proteids of Milk. Journ. Physiol., 1890, n, 448.
HAMMARSTEN. Ueber das Paraglobulin. Pfliiger's Archiv, 1878, 17, 413 and 18, 25.
HASLAM. The Separation of Proteids, I. Journ. Physiol., 1905, 32, 267.
HASLAM. The Separation of Proteins, II. Journ. Physiol., 1907, 36, 164.
HEYNSIUS. Ueber das Verhalten der Eiweissstoffe zu Salzen von Alkalien und von

alkalischen Erden. Pfliiger's Archiv, 1884, 34, 330.
HOFMEISTER. Zur Lehre von der Wirkung der Salze. Arch. exp. Path. Pharm.,

1887, 24, 247 and 1888, 25, i.
KAUDER. Zur Kenntniss der Eiweisskorper des Blutserums. Arch. exp. Path.

Pharm., 1886, 20, 411.

KUHNE. Erfahrungen uber Albumosen und Peptone. Zeit. f. Biologie, 1892, 29, i.
KUHNE AND CHITTENDEN. Ueber die Peptone. Zeit. f. Biologie, 1886, 22, 423.
LEWITH. Zur Lehre von der Wirkung der Salze. Arch. exp. Path. Pharm., 1887, 24, i.
M£HU. Methode ^extraction des pigments d'origine animate ; applications diverses du

sulphate d'ammoniaque. Journ. de Pharm. Chim., 1878, 28, 159.
OSBORNE AND HARRIS. Precipitation Limits of Globulins. Journ. Amer. Chem. Soc.,

1903, 25, 837.
PANUM. Neue Beobachtungen uber die eiweissartigen Korper. Virchow's Archiv,

1852, 4, 419.

PICK. Untersuchungen uber die Proteinstoffe. Zeit. physiol. Chem., 1897, 24' 24^-
PICK. Zur Kenntniss der peptischen Spaltungsprodukte des Fibrins. Zeit. physiol.

Chem., 1899, 28, 219, and Beitr. chem. Path. Physiol., 1902, 2, 481.
PINKUS. The Precipitation of the Proteids with Anhydrous Sodium Sulphate. Journ.

Physiol., 1901, 27, 57.
SCHAFER. Notes on the Temperature of Heat Coagulation of Certain Proteid Substances

of the Blood. Journ. Physiol., 1880, 3, 181.
STARKE. Beitrdge zur Kenntniss des Serum- und Eieralbumins . Maly's Jahresber.,

1881, ii, 17.
VIRCHOW. Ueber ein eigenthumliches Verhalten albuminoser Flussigkeiten bei Zusatz

von Salzen. Virchow's Archiv, 1854, 6, 572.
WENZ. Ueber das Verhalten der Eiweissstoffe bei der Darmverdatiung. Zeit. f.

Biologie, 1886, 22, i.
WEYL. Beitrdge zur Kenntniss thierischer und pflanzlicher Eiweisskorper. Zeit.

physiol. Chem., 1877, i, 72.
ZUNZ. Ueber den quantitativen Verlauf der peptischen Eiweissspaltung. Zeit. physiol.

Chem., 1899, 28, 132. Beitr. chem. Physiol. Path., 1902, 2, 435.

DEGREE OF SOLUBILITY OF PROTEINS IN SALT SOLUTIONS.

HARDY. Colloidal Solution. The Globulins. Journ. Physiol., 1905-6, 33, 251.

MELLANBY. Globulin. Journ. Physiol., 1905-6, 33, 338.

OSBORNE AND HARRIS. Solubility of Globulin in Salt Solutions. Amer. Journ.

Physiol., 1905, 14, 151.
PAULI. Die physikalischen Zustandsdnderungen der Eiweisskorper. Pfliiger's Archiv,

1899, 78, SIS-

78

BIBLIOGRAPHY 79

SOLUBILITY OF PROTEINS IN ORGANIC SOLVENTS.

MAYER AND TERROINE. Siir les proprietes des precipites d'albumine par V alcohol. Bull.
Soc. de biol., 1907, 62, 317.

OSBORNE AND VOORHEES, OsBORNE AND CHITTENDEN, OSBORNE AND HARRIS AND

OTHERS. Alcohol Soluble Proteins from Corn, Maize, Rye, Barley, etc.

Numerous papers. Amer. Chem. Journ., 1891, 1892, 1893 (13, 14, 15), and

other papers.

RAMSDEN. Some New Properties of Urea. Proc. Physiol. Soc., July, 1902.
RITTHAUSEN. Die Eiweisskorper der Getreidearten. Bonn, 1872.

SEPARATION OF PROTEINS BY PRECIPITANTS OTHER THAN SALTS.

LANDSTEINER AND UHLIRZ. Ueber die Adsorption von Eiweisskorpern. Zentr. Bak.

u. Par. I., 1906, 40, 265.
MICHAELIS AND RONA. Eine Methode zur Entfernung von Kolloiden aus ihren Losungen,

etc. Biochem. Zeitsch., 1906, 2, 219; 1907, 3, 109; 1907, 4, n ; 1907,5,365;

1907, 6, i.

CRYSTALLISATION OF PROTEINS.

ABDERHALDEN. Resorption des Eisens, etc. Zeit. f. Biol., 1900, 39, 143.

FREMY AND VALENCIENNES. Recherches sur la composition des oeufs dans la serie des

animaux. Compt. rend., 1854, 38, 469, 525 and 570. Ann. de chim phys., 1857

[3l» 50, 129.
GrJRBER. Krystallisation des Serumalbumins. Sitzungsber. physik. med. Gesellsch.

Wiirzburg, 1894, P- J43'

HARTIG. Uebtr das Klebermehl. Bot. Zeit., 1850, no. 5, 881.
HOFMEISTER. Ueber die Darstellung von krystalli sir tern Eieralbumin. Zeit. physiol.

Chem., 1889, 14, 165.
HOPKINS. On the Separation of Pure Albumin from Egg- White. Journ. Physiol.,

1900, 25, 306.
HOPKINS AND PINKUS. Remarks on the Crystallisation of Animal Proteins. Journ.

Physiol., 1898, 23, 130.
INAGAKI. Zur Kenntniss der Eiweisskrystallisation. Verh. der phys. med. Gesellsch.

Wtirzburg, 1906, 38, 17.

LEIPZIGER. Edestin, etc. Pfluger's Archiv, 1899, 78, 402.

MASCHKE. Krystallisierte Proteinverbindung. Journ. prakt. Chem., 1858, 74, 436.
OSBORNE. Crystalline Vegetable Proteins. Amer. Chem. Journ., 1892, 14, 208.
RADLKOFER. Ueber Krystalle proteinartiger Korper pflanzlichen und thierischen

Ursprungs. Leipzig, 1859.

RITTHAUSEN. Die Eiweisskorper der Getreidearten. Bonn, 1872.
SCHIMPER. Ueber die Krystalle eiweissartiger Substanzen. Zeit. f. Krystallographie,

1880.

SCHMIEDEBERG. Ueber die Darstellung der Paranusskrystalle. Zeit. physiol. Chem.,
1877, i, 205.

WALTHER. Zur Kenntniss des Ichthulins und seiner Spaltungsprodukte. Zeit. physiol.
Chem., 1891, 15, 477.

ZINOFFSKY. Ueber die Grosse des Hdmoglobinmolekdls. Zeit. physiol. Chem., 1885,
10, 16.

THE TEMPERATURE OF COAGULATION OF PROTEIN SOLUTIONS.

ARONSTEIN. Ueber die Darstellung salzfreier Albuminlosungen vermittelst der

Diffusion. Pfluger's Archiv, 1874, 8, 75.
HAAS. Ueber das optische und chemische Verhalten einiger Eiweisssubstanzen,inbeson-

dere der dialysirten Albumine. Pfluger's Archiv, 1876, 12, 378.
HALLIBURTON. The Proteins of Muscle. Journ. Physiol., 1887, 8, 133.
HEYNSIUS. Ueber Serumalbumin und Eieralbumin und ihre Verbindungen. Pfluger's

Archiv, 1876, 12, 549.

KUHNE. " Protoplasma und Contractilitdt" Leipzig, 1864.
WOLFGANG OSTWALD. Influence of Electrolytes on Coagulation Temperature. Abs.

Chem. Soc., 1908, i., p. 375.

PAULI. Ueber die physikalische Zustandsdnderungen der Eiweisskorper. Pfluger's
Archiv, 1899, 78, 315, and Part VI., Beitr. chem. Physiol. Path., 1907, 10, 53
(with Handovsky), and Beit. chem. Physiol. Path., 1908, u, 415.

80 THE GENERAL CHARACTERS OF THE PROTEINS

»

SCHMIDT (ALEX.). Weitere Untersuchungen des Blutserums, des Eiereiweisses und der

Milch durch Dialyse mittelst geleimten Papieres. Pfluger's Archiv, 1875, n, i.
STARKE. Beitrdge zur Kenntniss des Serum- und Eier albumins. Maly's Jahresber.,

1881, ii, 17.
WINOGRADOFF. Ueber Darstellung und Eigcnschaften salzfreier Eiweisslosungcn.

Pfluger's Archiv, 1875, ii, 605.
References to Table : —

Fredericq. Zenlralbl. f. Physiologic, 1890, 3, 601 (with references to papers in the
Bulletin de 1' Academic Royale de Belgique).

Freund and Joachim. Zeit. physiol. Chem., 1902, 36, 407.

v. Fiirth. Ergebnisse der Physiologic, 1902, I (i), no.

Halliburton. Journ. Physiol., 1887, 8, 133.

Hammarsten. Zeit. physiol. Chem., 1884, 8, 467.

Hewlett. Journ. Physiol., 1892, 13, 798.

Lacqueur and Sackur. Beitr. chem. Physiol. Path., 1902, 3, 193.

Magnus-Levy. Zeit. physiol. Chem., 1900, 30, 200.

Morner. Zeit. physiol. Chem., 1893, 18, 61.

Preyer. " Blutcrystalle," Jena, 1871.

Starke. Maly's Jahresber., 1881, n, 17.

Weyl. Zeit. physiol. Chem., 1877, i, 72.

OPTICAL ROTATION OF PROTEIN SOLUTIONS.

References to Table : —

Gamgee and Croft Hill. Ber., 1903, 36, 913.

Gamgee and Jones. Beitr. chem. Physiol. Path., 1903, 4, 10.

Hopkins. Journ. Physiol., 1900, 25, 306.

Morner. Zeit. physiol. Chem., 1893, 18, 61.

Osborne and Harris. Journ. Amer. Chem. Soc., 1903, 25, 842.

Willcock. Journ. Physiol., 1908, 37, 27.

MOLECULAR WEIGHT DETERMINATIONS BY CRYOSCOPIC METHODS.

A. DEPRESSION OF FREEZING POINTS.

References to Table: —

Bugarsky and Liebermann. Pfluger's Archiv, 1898, 72, 70.

Ciamician and Zanetti. Maly's Jahresber., 1892, 3.

Paal. Ber., 1892, 25, 1202 ; 1894, 27, 1827; and 1902, 35, 2195.

Sabanejew. Ber. Referatband, 1893, 385.

Sabanejew and Alexandrow. Ber. Referatband, 1891, 558.

B. DIRECT DETERMINATIONS OF OSMOTIC PRESSURE.

MOORE AND PARKER. Osmotic Pressure of Colloid Solutions. Amer. Journ. Physiol.,

1902, 7, 261.
MOORE AND ROAF. Osmotic Pressure of Colloidal Solutions. Biochem. Journ., 1907,

2,34-

REID. Osmotic Pressure of Proteids. Journ. Physiol., 1904, 31, 438.
REID. Osmotic Pressure of Hemoglobin Solutions. Journ. Physiol., 1905, 33, 12.
STARLING. Function of the Glomeruli. Journ. Physiol., 1896, 19, 312, and 1899, 24,

3i7-

THE "GOLD NUMBER".

SCHULZ AND ZSIGMONDY. Die Goldzahl und ihre Verwerthbarkeit zur Charakterisierung
von Eiweissstoffen. Beitr. chem. Physiol. Path., 1903, 3, 137.

ZSIGMONDY. Die hochrothe Goldlosung als Reagens auf Colloide. Zeit. anal. Chem.,
1901, 40, 697.

FRACTIONAL FILTRATION OF PROTEINS.
BECHHOLD. Kolloidstudien mit der Filtrationsmethode. Zeit. physikal Chem., 1907,

60, 257.

BECHHOLD. Ultrafiltration. Biochem. Zeitsch., 1907, 6, 379.
CRAW. On the Filtration of Crystalloids and Colloids through Gelatine ; with Special

Reference to the Behaviour of Hcemolysins. Proc. Roy. Soc., 1906, 778, 311.
MARTIN. Separation of Colloids and Crystalloids. Journ. Physiol., 1896, 20, 364.
OSBORNE (W. A.). Caseinogen and its Salts. Journ. Physiol., 1901-2, 27, 398.

BIBLIOGRAPHY

81

THE NITROGEN DISTRIBUTION IN THE PROTEINS.

GUMBEL. Ueber die Vcrtheilung des Stickstoffs im Eiweissmolekul. Beitr. chem.

Physiol. Path., 1904, 5, 297.
HAUSMANN. Ueber die Vertheilung des Stickstoffs im Eiweissmolekul. Zeit. physiol.

Chem., 1899, 27, 95 ; 1900, 29, 136.
HENDERSON. Zur Kenntniss des durch Sduren abspaltbaren Stickstoffs der Eiweisskorper.

Zeit. physiol. Chem., 1900, 29, 47.
KUTSCHER. Ueber die Verwendung der Phosphorwolframsdure bei quantitativen Bestim-

mungen der Spaltungsprodukte des Eiweisses. Zeit. physiol. Chem., igoo, 31,

215.
OSBORNE AND HARRIS. Nitrogen in Protein Bodies. Journ. Amer. Chem. Soc., 1903,

25, 323-
SCHULZE AND WINTERSTEIN. Ueber das Verhalten einiger Monaminosduren gegen

Phosphortvolframsaure. Zeit. physiol. Chem., 1901, 33, 574.
SCHULZE AND WINTERSTEIN. Ueber die Trennung von Phenylalanin von anderen

Aminosduren. Zeit. physiol. Chem., 1902, 35, 40.

SKRAUP. Ueber den sogenannten Amidstickstoff der Proteine. Monatsh., 1908, 29, 255.
See also Transactions of the Guinness Research Laboratory, Vol. I., Part 2.

THE SULPHUR, PHOSPHORUS AND HALOGEN CONTENT OF PROTEINS.

BAUMANN. Ueber das Normale Vorkommen von lod im Thierkorper. Zeit. physiol.

Chem., 1895, 21, 319.

DRECHSEL. Zur Chemie einiger Seethiere. Zeit. f. Biologie, 1896, 33, 90.
DURING. Ueber Schwefelbestimmungen in verschiedenen animalischen Substanzen, etc.

Zeit. physiol. Chem., 1896, 22, 281.
HARNACK. Ueber lodospongin, die iodhaltige eiiveissartige Substanz aus Badeschwamm.

Zeit. physiol. Chem., 1898, 24, 412.
HUNDESHAGEN. Ueber iodhaltige Spongien und lodospongin. Zeit. angew. Chem.,

1895, 473.
C. TH. MORNER. Zur Kenntniss der organischen Geriistsubstanz des Anthozoenskeletts.

Zeit. physiol. Chem., 1907, 51, 33; 1908, 55, 77.

OSWALD. Die Eiweisskorper der Schilddriise. Zeit. physiol. Chem., 1899, 27, 14.
OSWALD. Zur Kenntniss des Thyreo globulins. Zeit. physiol. Chem., 1901, 32, 121.
OSWALD. Weiteresuber Thyreo globulin. Beitr. chem. Physiol. Path., 1902, 2, 545 ; and

other papers.
PLIMMER AND BAYLISS. Separation of Phosphorus from Caseinogen by the Action of

Enzymes and Alkali. Journ. Physiol., 1906, 33, 439.
PLIMMER AND SCOTT. Trans. Chem. Soc., 1908, 93, 1699.
References to Table : —

(3) Bang. Zeit. physiol. Chem., 1899, 27, 463.

(13) Chittenden and Osborne. Amer. Chem. Journ., 1892, 14, 20.

(20) Chittenden and Solley. Journ. Physiol., 1891, 12, 23.

(2) Fleroff. Zeit. physiol. Chem., 1899, 28, 307.

(i) Goto. Zeit. physiol. Chem., 1902, 37, 84.

(8), (17) Hammarsten. Pfluger's Archiv, 1880, 22, 431.

(16) Hammarsten. Zeit. physiol. Chem., 1885, 9, 273.

(19) Levene. Zeit. physiol. Chem., 1901, 32, 281.

(26) Lubarch. Encyclopaedic der Mikroskopischen Technik. Berlin, 1903.

(5) Michel. Wiirzburger phys. med. Gesellsch. N.F., 1895, 29, 117.

(9) Osborne. Amer. Chem. Journ., 1893, *4> no- 8.

54), (12) Osborne and Campbell. Journ. Amer. Chem. Soc., 1899, 21, 477.
7), (10), (14), (15) Osborne and Voorhees. Amer. Chem. Journ., 1893, 15, 392.

(4) Schulz. Zeit. physiol. Chem., 1898, 24, 449.

(24) Schwarz. Zeit. physiol. Chem., 1893, 18, 487.

(25) Siegfried. Habilitationschrift. Leipzig, 1892.
(18) Walter. Zeit. physiol, Chem., 1891, 15, 477.

THE TYROSINE FACTOR FOR PROTEINS.

A. J. BROWN AND E. T. MILLAR. The Liberation of Tyrosine during Tryptic Proteolysis.

Trans. Chem. Soc., 1906, 89, 145.
H. J. MILLAR. A New Method for the Direct Estimation of Tyrosine in Mixtures of

Amides and Amino-Acids. Trans. Guinness Laby., 1903, I, 40.
REACH. Quantitative Untersuchungen iiber das Tyrosin als Spaltungsprodukt der

Eiweissstoffe. Virchow's Archiv, path. Anat., 1899, 158, 288,

82 THE GENERAL CHARACTERS OF THE PROTEINS

SALT FORMATION OF PROTEINS.

BUGARSKY AND LiEBERMANN. Ueber das Bindungsvermogen eiweissar tiger Korper fur
Salzsdure, N atrium hydroxid und Kochsalz. Pfliiger's Archiv, 1898, 72, 51.

COHNHEIM. Ueber das Salzsdurebindungsvermogen der Albumosen und Peptone.
Zeit. f. Biologic, 1897, 33, 489.

COHNHEIM AND KRIEGER. Das Verhalten der Eiweisskorper zu Alkaloidreagentien

zugleich elne Bestimmung der gebundenen Salzsdure. Zeit. f. Biologic, 1900,

40. 95-
ERB. Das Salzsdurebindungsvermogen einiger reiner Eiweisskorper. Zeit. f. Biologic,

1901, 41. 309.

HARDY. Colloidal Solutions. The Globulins. Journ. Physiol., 1905, 33, 251.
LACQUEUR AND SACKUR. Ueber die sduren Eigenschaften und das Molekulargewicht des

Caseins und seine Spaltung beim Trockenen. Beitr. chem. Physiol. Path., 1902,

3> J93.
OSBORNE. Der basische Charakter des Proteinmolekuls tind das Verhalten des Edestins

zu bestimmten Mengen von Sdure und Alkali. Zeit. physiol. Chem., 1901, 33,

240.

VON RHORER. Ueber die Bestimmung der Sdureverbindungsvermogen der Eiweissstoffe.

Pfliiger's Archiv, 1902, 90, 368.
SJOQVIST. Physiologisch.-chemische Betrachtungen uber Salzsdure. Skand. Archiv,

i895, S 277-
SPIRO AND PEMSEL. Ueber Basen und Sdurecapacitdt des Blutes und der Eiweisskorper.

Zeit. physiol. Chem., 1898, 26, 233.

THE PRECIPITATION OF PROTEINS BY SALTS OF THE HEAVY

METALS.

GALEOTTI. Ueber die sogenannten Metallverbindungen der Eiweisskorper nach der
Theorie der chemischen Gleichgewichte. Zeit. physiol. Chem., 1904, 40, 492.

HARNACK. Untersuchungen uber die Kupferverbindungen des Albumins. Zeit. physiol.
Chem., 1881, 5, 198.

PAULI. Untersuchungen uber physikalische Zustandsdnderungen der Kolloide IV.
Eiweissfdllung durch Schwermetalle. Beitr. chem. Physiol. Path., 1905, 6, 233.

OXIDATION OF PROTEINS.

BECHAMP. Recherches sur les produits d' oxidation des substances albuminoids par le

hy permanganate de potasse. Ann. de Chim., 1889, 57, 291.
BERNERT. Ueber die Oxidation von Eiweiss mit Kaliumpermanganat. Zeit. physiol.

Chem., 1898, 26, 272.

BLUMENTHAL AND NEUBERG. Deutsch. med. Wochenschrift, 1901, no. i.
BoNDzf NSKI AND ZOJA. Ueber die oxidation der Eiweisstoffe mit Kaliumpermanganat.

Zeit. physiol. Chem., 1894, 19, 225.
BRUCKE. Ueber elne durch Kaliumpermanganat erhaltene Satire. Sitzber. d. Wiener

Akademie, 1881, 83, III., 7.

CHANDELON. Beitrag zum Studium der Peptonisation. Ber., 1884, 17, 2143.
EHRMANN. Ueber die Peroxyprotsdure. Inaug. Diss. Strassburg, 1903.
VON FURTH. Beitrdge zur Kenntniss des oxydativen Abbaus der Eiweisskorper. Beitr.

chem. Physiol. Path., 1905, 6, 296.
GORUP-BESANEZ. Ueber die Einwirkung des Ozons auf organische Verbindungen.

Annalen, 1859, no, 96.
HARRIES. Ueber Versuche zur Spaltung des Caseins vermittelst Ozon. Ber., 1905, 38,

2990.
HARRIES AND LANGHELD. Ueber das Verhalten des Caseins gegen Ozon. Zeit. physiol.

Chem., 1907, 51, 342.
KUTSCHER AND ScHENCK. Die Oxydation von Eiweissstoffen mit Calciumpermanganat.

Ber., 1904, 37, 2928.

Low. Ueber Eiweiss und Oxydation desselben. Journ. prakt. Chem., 1885,31 (2), 129.
MALY. Ueber die Oxydation von Eiweiss mittelst Kaliumpermanganat. Sitzber. d.

Wiener Akademie, 1889, 97, II.
MALY. Ueber die • bei der Oxydation von Leim mit Kaliumpermanganat entstehenden

Korper und die Stellung von Leim zu Eiweiss. Sitzber. d. Wiener Akademie,

1889, 98, II., Monatsh., 1889, 10, 26.
ORGLER. Ueber die Entstehung von Aceton a^ls krystallisirtem Ovalbumin, Beitr,

chem. Physiol. Path., 1901, i, 583.

BIBLIOGRAPHY 83

POTT. Oxydationsversuche mit Kaliumpermanganat. Journ. prakt. Chem., 1872, 5 (2),

355-'
SCHULZ. Ueber die Oxydation von krystallisiertem Eieralbumin mit Wasserstoffsuper-

oxyd. Zeit. physiol. Chem., igoo, 29, 86.
SEEMAN. Ueber die Oxydation des Leims nnd des Eieralbnmins mit Calciumperman-

ganat. Zentrbl. f. Physiologic, 1904, 18, 285.
SIEGFRIED. Zur Kenntniss der Spaltungsprodukte der Eiweisskorper. Ber., 1891, 24,

427.
SUBBOTIN. Einiges uber die Wirksamkeit des ubermangansauren Kalis auf Albumin.

Chem. Zentralb., 1865, 594.
WURSTER. Ueber das Verhalten von WasserstoffsuperoxydgegenEiweiss. Ber., 1887,

20, 263.
ZICKGRAF. Die Oxydation des Leims mit Permanganaten. Zeit. physiol. Chem., 1904,

41, 259.

THE ACTION OF HALOGENS ON PROTEINS.

BLUM. Ueber die lodzahl der Eiweisskorper. Zeit. physiol. Chem., 1899, 28, 288.
BLUM AND VAUBEL. Ueber die Halogeneiweissderivate. Journal prakt. Chem., 1897,

56, 393 ; 1898, 57' 365-

BOHM AND BERG. Beitrdge zur Pharmakologie der lodes. Arch. exp. Path. Pharm.,

1876, 5, 329.
CROSS, BEVAN AND BRIGGS. The Chloramine Reactions of the Proteins and Technical

Applications. Journal Soc. Chem. Indust., 1908, 27, 260.
HABERMANN AND EHRENFELD. Uber Proteinstoffe. Einwirkung des nascirenden Chlors

auf Casein. Zeit. physiol. Chem., 1901, 32, 467.

HOFMEISTER. Ueber iodirtes Eieralbumin. Zeit. physiol. Chem., 1897, 24» I59-
HOPKINS AND PINKUS. Zur Kenntniss der Wirkung von Halogenen auf Proteine. Ber.,

1898, 31, 1312.

JENDRASSIK. Quoted by Hofmeister.
KURAJEFF. Ueber Einfuhrung von lod in das krystallisirte Serum- und Eieralbumin.

Zeit. physiol. Chem., 1899, 26, 462.
LEPINOIS. Action de Viode sur les matieres albuminoides. Journ. Pharm. Chim.,

1897, [6] 5. 56i.

LIEBRECHT. Ueber lodderivate von Eiweisskorpern (Casein). Ber., 1897, 30, 1824.
LOEW. Ueber Eiweiss und die Oxydation desselben. J. prakt. Chem., 1885, 31, 129.
MULDER. Ueber Protein. Journ. prakt. Chem., 1848, 44, 488.
PANZER. Ueber ein gechlortes Casein und dessen Spaltung durch rauchende Salzsdure.

Zeit. physiol. Chem., 1901, 33, 131.
RASCHIG. Monochloramin. Ber., 1907, 40, 4586.
RIDEAL AND STEWART. Note on the Determination of Proteids by Chlorine. Analyst,

1897, 22, 228.
C. H. L. SCHMIDT. Ueber die Bedeutung der lodsdurebildung bei der lodirung des

krystallisirten Eieralbumins. Zeit. physiol. Chem., 1901, 34, 55.
C. H. L. SCHMIDT. Quantitative Bestimmung der bei der lodirung von Albuminstoffen

entstehenden lodsdure und lodwasserstoffsdure. Zeit. physiol. Chem., 1901, 34,

194.

C. H. L. SCHMIDT. Zur Kenntniss der lodirungsprodukte der Albuminstoffe, I. Zeit.

physiol. Chem., 1902, 35, 386.
C. H. L. SCHMIDT. Zur Kenntniss der lodirungsprodukte der Albuminsiojfe, II.

Zeit. physiol. Chem., 1902, 36, 343.

ACTION OF NITROUS ACID ON PROTEINS.

BROWN. Description of an Improved Apparatus for the Sachsse-Kormann Process.

Trans. Guinness Research Laboratory, 1903, i, Part I., 30.

BROWN. The Nitrogen Question in Brewing. J. Inst. Brewing, 1907, 13, 394 (p. 415).
FISCHER AND KOELKER. Verhalten des Leucylisoserins und verwandter Korper gegen

salpetrige Sdure. Annalen, 1905, 340, 177.

LAMPEL. Ueber das Dcsamidoglobulin. Monatsh., 1907, 28, 625.
LEVITES. Ueber Desamidoalbumine. Zeit. physiol. Chem., 1904, 43, 202.
OBERMAYER. (Abstract of patent specification.) Ber., 1894, 27, Abstracts, 354.
PAAL. Ueber die Desamidirung des Glutinpeptons. Ber., 1896, 29, 1085.
SCHIFF. Ueber Desamidoalbumin. Ber., 1896, 29, 1354.

84 THE GENERAL CHARACTERS OF THE PROTEINS

SCHROTTER. Beitrdge zur Kenntniss der Albumosen. Monatsh., 1898, 19, 211.
SKRAUP. Ueber Desamidoglutin. Monatsh., 1906, 27, 653 ; and 1907, 28, 447.
SKRAUP AND HOERNES. Ueber Dcsamidocasein. Monatsh., 1906, 27, 631.
SKRAUP AND KAAS. Ueber die Einwlrkung von Salpetrigesdure auf Ovalbumin.
Annalen, 1907, 351, 379.

TREVES AND SALOMONE. Ueber die Wirkung der salpetrigen Sdure auf die Eiweisstoffe.
Biochem. Zeitsch., 1907, 7, n.

ACTION OF FORMALDEHYDE ON PROTEINS.

BENEDICENTI. Ueber die Einwirkung des Formaldehyde auf einigen Proteinstoffe.

Archiv f. (Anat.) u. Physiol., 1897, 219.
BLUM. Ueber eine neue Klassc von Verbindungen der Eiweisskorper. Zeit. physiol.

Chem., 1896, 22, 127.
HAUSER. Ueber Verwendunge des Formalins zur Conservirung von Bakterienkulturen.

Munch, med. Wochensch., 1903, pp. 567-655.
SCHIFF. Trennung von Amin- und Sdurefunction in Losungen von Eiweisskorpern.

Annalen, 1901, 319, 287.
SCHWARZ. Ueber Verbindungen der Eiweisskorper mit Aldehydcn. Zeit. physiol.

Chem., 1901, 31, 460.

SORENSEN. Enzymstudien. Biochem. Zeitsch., 1907, 7, 45.
TREVES AND SALOMONE. Ueber die Wirkung der salpetrigen Sdure auf die Eiweiss-

stoffe. Biochem. Zeitsch., 1907, 7, n.
TRILLAT. Sur les proprietes antiseptiques de la formaldehyde. Compt. rend., 1892,

114, 1278.

THE PRECIPITIN REACTION.

BLUM, L. Uber Prdzipitine. Zentralbl. f. allgemeine Path. u. path. Anat., 1906, 17, 81.
BLUME, G. Ueber die Methoden und die bisherigen Ergebnisse der Komplementbindung.

Zentralbl. f. Bak. u. Par. I., 1907, 40, 609.
BORDET. Sur V agglutination et dissolution des globules rouges. Ann. de 1'Inst.

Pasteur, 1899, 13, 73.
GENGOU. Sur les sensibilatrices des serums actif contre les substances albuminoides.

Ann. de 1'Inst. Pasteur, 1902, 16, 734.
HAMBURGER AND ARRHENIUS. Ueber das Wesen der Prdzipitinreaktion. Maly's

Jahresber., 1906, 36, 970.
KRAUS. Ueber spezijische Reaktionen in keimfreien Filtraten aus Cholera, Typhus,

etc., erzeugt durch homologes Serum. Wiener klin. Wochenschr., 1897, p. 736.
NEISSER AND SACHS. Bin Verfahren zum forensischen Nachweis der Herkunft des

Bluts. Berliner klin. Wochenschr., 1905, p. 1388, 1906, p. 67.
NUTTALL. Blood, Immunity and Relationship. Cambridge, 1904.
OBERMAYER AND PICK. Beitrdge zur Kenntniss der Prdzipitinbildung. Wiener klin.

Wochenschr., 1902, no. 22.
OBERMAYER AND PICK. Ueber die chemischen Grundlagen der Arteigenschaften der

Eiweisskorper. Wiener klin. Wochenschr., 1906, no. 12.

SCHULZ, A. Die Technik quantitativer Eiweissbestimmung, mit Hilfe der Prdzipi-
tinreaktion. Zeit. z. Unters. der Nahrungs- u. Genussmittel, 1906, 12, 257.
SCHUTZE. Ueber ein biologisches Verfahren zur Differ en zierung der Eiweissstoffe in

verschiedenen Milcharten. Zeit. f. Hygiene, 1901, 36, 55.
TSCHISTOWITSCH. Etude sur I 'immunisation contre le serum d'anguille. Ann. de 1'Inst.

Pasteur, 1899, 13, 406.
UHLENHUTH. Neuer Beitrag zum spezifischen Nachweiss von Eiwciss auf biologischem

Weg. Deutsche med. Wochenschr., 1900, p. 734.
UHLENHUTH. Eine neue Method zur Unterscheidung der verschiedenen Blutarten.

Deutsche med. Wochenschr., 1901, pp. 82, 260 and 780.
WASSERMANN AND SCHUTZE. Ueber eine neue forensische Methode zur Untersuchung

von Menschen- u. Thierblut. Deutsche med. Wochenschr., 1902, no. 27.
WASSERMANN AND SCHUTZE. Ueber die Spezificitdt der Eiweissprdzipitierenden Sera

und deren Werthbestimmung fur die Praxis. Deutsche med. Wochenschr., 1903,

no. 15.
WELSH AND CHAPMAN. The Main Source of " Precipitable " Substance and on the role

of the Homologous Proteid in Precipitin Reactions. Proc. Roy. Soc., 1906, 788,

297.

INDEX.

ACETONE, from oxidation of proteins, 60.

Acid and basic functions of proteins, 3-6, 39.

Acidity of filtrates, after precipitation of
proteins in presence of acids, 46, 47.

Acids and alkalis, influence on coagulation
of proteins, 22.

Aleurone grains, 18.

Amide-nitrogen, 34.

Amino-index of proteins, 6, 67.

Ammonium sulphate as albumin precipi-
tant, 9.

BIOLOGICAL methods for investigating pro-
teins, 71-77.
Biuret reaction, 30.
Blood precipitins, 71.
Bromine derivatives of proteins, 62, 63.

CARBOHYDRATE group in proteins, 31.
Caseinogen, salt formation of, 50, 51.
Chemical characterisation of proteins, 6.

— composition of proteins, 32.
Chlorine derivatives of proteins, 61, 62.
Chromo-proteins, 33.

Colloids, i.

Colour reactions of proteins, 31.

Complement removing action, 76.

Concentration cells, use of, in investigating
salt formation of proteins, 39.

Copper compounds of proteins, 56.

Cryoscopic methods for determining mole-
cular weight, 24.

Crystallisation of proteins, 16-21.

Cystine factor of proteins, 6, 36.

DENIS, protein investigations of, i.
Depression of freezing-point by proteins,
24.

EDESTIN, preparation of, 19.

— salt formation of, 47-50.
Egg-albumin, crystallisation of, 19.

salt formation of, 40-42.

Electrolytic conductivity of proteins, 26,
39, 40-42.

FORMALDEHYDE, action on proteins, 69, 70.
Fractional filtration of proteins, 28.

— precipitation of proteins, 8-15.
Freezing-point of protein solutions, 24, 39,

44.

GLOBULINS, 9.

Glyco-proteins, 31, 33.

" Gold number " of proteins, 27.

Gorgonin, 38.

Gum-mastic, precipitant of proteins, 17.

HAEMOGLOBIN, crystallisation of, 21.
Halogens, action on proteins, 61-67.

— content in protein, 37.

— protein derivatives, 6.
Heat-coagulation temperature of proteins,

21-24.
Heavy metals, precipitation of proteins by

salts of, 56-58.
Humin nitrogen, 34.
Hydrolysis of cane-sugar and methyl acetate

in presence of proteins, 39, 42-44.

INORGANIC solvents of proteins, 18.
Iodine derivatives of proteins, 63, 64.
lodoproteins (natural), 37.
Isolation of proteins, i.

KYROPROTEIC acid, 59.
LIEBIG'S views on proteins, 2.

MAGNESIUM sulphate as globulin pre-
cipitant, 9.
Milk, separation of protein constituents of,

9-

Mulder's views on proteins, 2.
Myosins, 9.

NITROGEN content and distribution in pro-
teins, 32-36.

Nitrous acid, action on proteins, 67.
Nucleo-proteins, 33.

OPTICAL rotation of proteins, 24.
Organic solvents of proteins, 16.
Osmometers, 26.

Osmotic pressure of proteins, direct deter-
mination of, 25.
Oxaluramide from proteins, 60.
Oxidation of proteins by hydrogen per-
oxide and ozone, 60.

by nitric acid, 61.

by permanganate, 59.

Oxyprotein, 60.
Oxyprotosulphuric acid, 59.

PEPTONES, precipitation of by salts, 9.
Peroxyproteic acid, 59.
Phosphorus content of proteins, 37.
Polypeptides, 3.
Precipitants of proteins, 30.
Precipitin reaction, 71-76.
Prosthetic groups, 3, 32.
Protamines, 32, 33, 39.
Proteoses from Witte's peptone, n.
Pseudo-acids and bases, proteins considered
as, 55, 56.

86

INDEX

QUALITATIVE distinctions between proteins,
32.

REACTIONS of proteins, 30-32.

SALT formation of proteins, 39-56.

— solutions, solubility of proteins in, 15.
"Salting out" of proteins, i, 8-15.

Salts, precipitating capacity of, 10.
Serum-globulin, salt formation of, 51-56.

solubility of in salt solutions, 16.

Serum, separation of protein constituents

of, 9.
Sodium chloride as globulin precipitant, 9.

— sulphate, use of, for salting out pro-

teins, 12.
Solubility of proteins in acids and bases,

39-

Sponges, iodine content of, 38.
Sulphur in proteins, 32, 36.

THYROID gland, iodine in, 37.
Titration of proteins, 39, 46.
Tryptophane reactions, 31.
Tyrosine factor of proteins, 6, 38.

UREA solutions, solubility of proteins in, 17.

VITELLINS, 9.

WITTE'S peptone, fractionation of, n, 15.

XANTHOPROTEIC reaction, 30.

YOLK platelets, 19.

ZINC sulphate as precipitant of proteoses,

12.

ABERDEEN: THE UNIVERSITY PRESS

OMOHOi

11