CELLULOSE

CELLU LOSE

AN OUTLINE OF THE CHEMISTRY OF

THE STRUCTURAL ELEMENTS OF PLANTS

WITH REFERENCE TO THEIR

NATURAL HISTORY AND INDUSTRIAL USES

BY

CROSS & BEVAN

(C. F. CROSS, E. J. BEVAN, AND C. BEADLE)

NEW EDITION

LONGMANS, GREEN, AND CO.

39 PATERNOSTER ROW, LONDON

FOURTH AVENUE & SOTH STREET, NEW YORK

BOMBAY, CALCUTTA, AND MADRAS

1916

All rights reserved

PREFACE

TO

THE NEW EDITION

THE call for a further edition of this work is an occasion
for a brief general review of progress. Of progress there are
certain outstanding landmarks, scientific, technical, and indus-
trial, which to the initiated are comprehensive measures of
development. The continuous growth of ' Cellulose ' as a
section of organic chemistry is 'evident in the comparison of
our original edition (1895), on the one hand, with articles on
the subject in the Dictionaries and Encyclopaedic Works of the
preceding period — e.g., Watts' ' Dictionary of Chemistry ' and
Supplements, 1878-1882 ; and, on the other hand, with
C. Schwalbe's ' Chemie der Cellulose' (Berlin, 1911), a
monograph of some six hundred pages, and E. C. Worden's
'Nitrocellulose Industry' (London, 1911), a technical mono
graph of sections of the subject of such industrial importance
as ' Celluloid,' Artificial Fibres and Explosives.

The plan and purpose of our more modest volume differs
from both of the above. It was designed rather as a study of

358831

vi Preface to the New Edition

illustrative types of necessarily c natural ' substances, classified
on the basis of ascertained chemical variations, and of deriva-
tives resulting from well-known processes of synthetic addition,
or of more or less destructive resolutions, chiefly by processes
of hydrolysis and oxidation. These relationships set out in
the logical order of the systematic science are of permanent
significance, and until we have an accepted theory of the state
of matter in these organic and organised products of the plant
world it would be superfluous to devise any other plan. Its
basis is that of an orderly empiricism, and in absence of any
criterion of actual molecular constitution especial prominence
is given to the weight statistics of the reactions described. Any
molecular units in this region must be more or less arbitrarily
fixed ; a unit of 100 is justified by convenience, and percentage
results are as exact as molecular proportions. In the case of
certain of the esters of the typical cellulose, these appear to
have a simple relationship to the parent substance, and the use
of molecular formulae therefore justified. But even in such
cases the reactions of esterification appear to be continuous,
and therefore molecular formulae have only a certain value as
applied to the limiting cases or end products. It is open to
question whether, even in the case of the nitrates, we gain
anything by a quasi-molecular terminology.

The limiting nitrate being the tri-nitrate (C6 formula),
representing 14*1 p.ct. N, a collodion nitrate of 11*2
p.ct. N is correctly described as an 80 p.ct. Nitrate

(80 = — J* x 100). The molecular proportion is n = 2*02,
14*1

and the body approximately a dinitrate. But the formula
C6HgO3.(NO3)2 is of no significance as identifying a chemical
individual. The specialist could produce a series of nitrates
of .this formula, differing widely in characteristics, e.g.t solu-

Preface to the New Edition vii

bi lilies, viscosities of solution, and mechanical properties of con-
tinuous solid forms prepared fr-om solutions.

In accordance with this general and, at the same time,
limited plan of the original, we have produced volumes of
' Researches ' intended as supplements, in which the growth
of the subject-matter is recorded in the form of abstracts of
original memoirs. To these we have added critical notes
intended as adumbrations of extensions of general theory,
towards which all workers may be presumed to be striving,
whether of set purpose or less consciously.

In the twenty years' period under consideration a con-
spicuous advance has followed the systematic investigation of
typical colloids, towards a general theory of the colloidal state.

Cellulose is the prototype of colloidal matter, and the
exhaustive study of celluloses and derivatives in their infinitely
varied forms, natufal and artificial, promises to be the most
important source of material for generalisations as to the
condition of matter in colloids. At present there are none
which command acceptance, and it is outside the scope of a
text-book to discuss hypotheses and tentative theory.

It is particularly in view of the somewhat nebulous aspect
of this region of potentially exact science that we abstain from
any present attempt to recast our original volume.

It will be implied from this confession that we have little
to note of really novel and fundamental importance — that is, as
contributing to a positive solution of problems of constitution
or of the more immediate problem of the reacting unit of
the typical cellulose. But there have been notable additions
of experimental method, of new groups of derivatives, and
evident progress in the critical appreciation of the subject.

1 Normal Cellulose.' — Many bleached cottons, even those
produced by carefully regulated 'standard' processes, are

viii Preface to the New Edition

below the standard of ' normal cellulose.' It is necessary to
note this in view of the common assumption that chemical
filter-paper or ' Surgical Cotton Wool ' is a cellulose hors
concours, and may be taken as a ' pure cellulose ' for important
investigations, fulfilling the comprehensive definition of the
'normal pure cellulose/ viz., a bleached and otherwise
purified cotton which can be 'mercerised' without loss of
weight, and thereafter converted into Xanthogenic ester, dis-
solved as Viscose, and regenerated to Cellulose (Hydrate), also
without loss of substance.

The nearest approach to the normal, as thus defined, in the
form of an easily accessible material is the ' Madder bleached '
cotton fabric of the calico printer, which may be used in
investigations, as such, or may be previously disintegrated by
mechanical means to the state of relatively ' loose fibre.'

Cellulose — Constitution. — That the normal cellulose
is resolved by hydrolysis quantitatively into dextrose, was a
statement of the text-books which remained for years on the
basis of quite insufficient experimental proof (see p. 49). The
re-investigation by Ost (Researches, iii., pp. 39-43) of the
hydrolysis by solution in sulphuric acid (H2SO4.2H2O
— H2SO4.3H.2O) and afterwards diluting and boiling — the
process originally employed by Flechsig — has carried the
resolution to satisfactorily extreme limits, and supplied the
identification of the product, by more modern methods, as,
in effect, a homogeneous hexose, viz., dextrose.

The process of resolution by acid in a medium of acetic
anhydride with attendant acetylation (/£/</., pp. 43-47), has
given us the cellobiose, and a special cleavage of the cellulose
complex which is destined to be of great value in investigating
the constitution of the typical cellulose, and of other celluloses
by comparison. The quantitative results are, however, not

Preface to the New Edition ix

yet sufficiently exactly determined to furnish a basis of positive
conclusions. This mode of resolution, however, establishes
another point of analogy with starch. So also reciprocally
does the production of the well-defined xanthogenic esters of
starch and their close general analogy — with the better known
cellulose derivative — Viscose.

Nothing, however, in these results affects the evidence that
the actual configuration of the unit groups in the cellulose
complex is rather ketonic than aldehydic, nor the probability
that these groups are of cyclic type rather than of the normal
hexose character.

In this connection the ascertained fluorescence of cellulose
[W. N. Hartley, Chem. Soc. J., 63, 243, (1893)], which is
intensified in some of its derivatives, and notably the acetates
(Cross & Bevan : J. Soc. Dyers, May, 1916), will be found,
when further investigated, to furnish some more direct
evidence on this point than that of processes of decomposition,
of which the effects as regards the cellulose complex are, in
important respects, problematical. Generally, it must be to
physical methods that we have to look for the full comple-
mentary interpretation of the related values furnished by
purely chemical methods.

Separation and Estimation of Cellulose.— As a
general result of research and criticism, the method of separa-
tion and purification of cellulose from cellulose mixtures,
described in our original volume, are confirmed and accepted.
A valuable review of work and literature on this section is that
of J. F. Briggs' * Recent Progress in the Analysis of Cellulose
and Cellulose Derivatives ' : Analyst, V. 40 (1915), pp. 107-120.
This author gives full value to M. Renker's critical study
of methods, ' Ueber Bestimmungsmethoden der Cellulose '
(Berlin, 1 909). It is only necessary to add that, in dealing with

x Preface to the New Edition

mixtures containing cuticular cellular products, special attention
is necessary in choice of methods of attack. On this section
the reader must be referred to a very valuable general study,
* Bestimmung der Zellulose in Holzarten und Gespinstfasern ' :
J. Konig and Fr. Hiihn (Berlin, 1912). Although we and
others have had occasion to differ from these authors in regard
to both methods and results, the general value of the experi-
mental matter is not prejudiced thereby, and the brochure
should be carefully studied.

New Derivatives. — We may mention a new class of
acetic esters of considerable industrial importance, which are
prepared according to the Dreyfus patents. In these acetates
the cjegree of acetylation is sensibly below that of the hitherto
better known tri-acetates, and they are distinguished by solu-
bility in acetone. We have also to note the production of
derivatives containing the methoxyl and ethoxyl groups — that
is, of true cellulose etheis.1 So far the investigation of these
bodies has not thrown much light on the major problems, but
they are useful additions to the series of derivatives of cellulose,
and are of potential interest as an auxiliary to research.

Destructive Distillation. — An important contribution
is that of H. Rollings and J. W. Cobb : ' Thermal Study of
the Carbonisation Process' (J. Chem. Soc. 107, 1106). These
authors incidentally confirm the statements in the original
volume as to the exothermic character of the decomposition of
celluloses and ligno-celluloses when heated, and they have
added methods of investigation of considerable refinement and
general application to organic products of the cellulose-lignite-
coal series. An interesting study of the products of destructive
distillation is that of E. Erdmann and C. Schaefer [Berl.
1 W. S. Denham and H. Woodhouse : J. Chem. Soc. 105, p. 2357

Preface to the New Edition xi

Ber. 43, 2398 (1910)]. Amongst the more characteristic
products is Maltol, 3 Hydroxy-2 Methyl — y. Pyrone, which
they separated and identified. We have also found that
maltol appears in quite early stages of the decomposition
(J. Soc. Dyers, 1916, loc. cit.). This is of some importance as
it shows the labile tendency of cellulose to heterocyclic
ketonic groupings.

Ligno-celluloses. — We have extended our investigations,
the results of which have been recorded in our volumes of
Researches, and later in a communication to the Society of
Dyers and Colorists, 1916 (loc. cit.). As there had been no
contributions by way of criticism of our conclusions from other
investigators, we need only notice the point. With regard to
the applications of portions of the researches in quantitative
methods of estimating ligno-celluloses, the method proposed
by us [Cross, Bevan & Briggs, Berl. Ber. 40, 3119 (1907)],
based on the quantitative relationships of the reaction with
phloroglucinol, has been generally adopted, for its obvious
advantages over tests based upon colour intensity. These
matters are dealt with at length in our communication above
mentioned. We may call attention to a paper by E. Heuser
and R. Sieber, ' Ueber die Einwirkung von Chlor. auf Fichten-
holz,' Ztschr. ang. Chem. 26, 801 (1913), which confirms and
extends our investigations (pp. 101, 134) of the process of
chlorination of the ligno-celluloses and the composition of
the products.

General Technology. — Industrial developments of both
old and new manufactures of cellulose and cellulose derivatives
are of continuously growing magnitude. These industries are
so well known that it is not necessary to refer to the subject in
an outline manual such as the present. Of the newer develop-
ments it may be noted that the ' artificial silk ' produced by

xii Preface to the New Edition

the viscose process has taken an increasingly important place
in industry, and, in fact, may now be regarded as a staple
textile. + The cellulose acetates are applied . in continually
increasing proportions for special purposes, and there is a very
definite position established for these derivatives in various
forms as substitutes for the better-known manufactures based
on nitro-celluloses, chiefly celluloid in its various forms of
solid and of sheet or film.

The capital investments and interests involved in cellulose
industries are of enormous dimensions, and are increasing at
a rapid rat**- For information as to these the specialist
publications must be consulted.

4 NEW COURT, LINCOLN'S INN, LONDON, W.C.
June 30, 1916.

PREFACE

TO

THE SECOND EDITION

SOME of the published criticisms of the book appearing to
indicate a certain misunderstanding of its plan and purpose,
we think it opportune to further and more specifically explain.

The chemistry of cellulose is necessarily a chemistry of
colloidal, uncrystallisable substance : hence the relationships
of any derivative to the parent substance is not established
by a comparison of composition and properties, but requires a
complete statistical account of the reaction. Even then it is
generally impossible to measure attendant changes of mole-
cular condition, or, to put it more precisely, changes of
dimensions of the reacting unit, accompanying the formation
of derivative compounds.

It is consequently impossible in the present state of our
knowledge to write a chemistry of cellulose in the terms of
the thoroughly systematised branches of the science. We have
therefore adopted the plan of classifying the empirical subject-

xiv Preface to the Second Edition

matter, as a necessary first step in progress from the purely
empirical. From the methodical treatment of the subject-matter
— under such heads as Hydrolysis, Oxidations, Ester formation,
and other special synthetic reactions — a certain order has
been introduced, from which there result at least prominent
suggestions of the underlying constitutional relationships.

So far has this proceeded that it is possible at this date to
propose constitutional formulae for even so complex a body
as a lignocellulose, such as would be a consistent summary of
the experimental facts as to composition and reactions. It
would, however, be premature to attempt this in any other
way than as a working hypothesis to guide investigations.
There remains, indeed, still unsolved the problem of the actual
condition of matter in these complex colloidal forms rela-
tively to the gaseous and liquid states, and pending a solution
of this major problem, we can only continue on the basis of
progressive empiricism which we originally adopted.

The present is therefore in the main a reprint of the
former edition with a small portion of the text rewritten and
the addition of an appendix giving an account of more impor-
tant recent contributions.

4 NEW COURT, LINCOLN'S INN, LONDON, W.C.
February 25, 1903.

PREFACE

THE purpose of this short work on a large subject is to con-
solidate the scattered contributions of investigators, with more
especial reference to the work of the past fifteen years. By
this later work the subject has been considerably widened,
not merely through the growth of the subject-matter, but by
notable additions of experimental methods on the one hand,
and theoretical generalisations on the other. In reviewing the
present position of cellulose chemistry the work has taken the
form of a monograph as distinguished from a textbook. In
adopting the freer form of writing we have reserved a certain
latitude of treatment, in view of the fact that ' Cellulose ' has
not yet been accorded a definite position in the specialised
sections of organic chemistry ; and also because we find it
necessary to address ourselves to original workers, and from
time to time to point out with particular emphasis the weaker
points in the evidence for the theoretical conclusions. At
such points suggestions are given of subject-matter for further
research, which it is our desire to stimulate. In presenting
this work, on the other hand, to the masters and ' past masters *
of the science it would have been out of place to have adopted

xvi Preface

the more positive method of a textbook, or to have entered
into the more minute detail of a handbook.

In the incidental treatment of the technology of the subject
we have endeavoured to maintain the scientific perspective
rather than to discuss the practical details of processes.

The photo-micrographs included in the work are from the
expert handiwork of our friend Mr. J. CHRISTIE, F.R.M.S., of
72 Mark Lane, E.G., to whom we record our best thanks for
this interesting addition to the subject-matter.

The book is printed upon a paper carefully selected as
composed of the ' normal ' celluloses, and to the exclusion of
the inferior ' celluloses ' ordinarily employed for the manufac-
ture of printing papers. Upon the reasons for this preference
we have something to say in the text (p. 305).

We have to thank our friend Mr. J. C. CHORLEY for contri-
butions of experimental results, and for kind assistance in con-
nection with the proofs.

4 NEW COURT, LINCOLN'S INN, LONDON, W.C

CONTENTS

PART I

FAGS

THE TYPICAL CELLULOSE AND THE CELLU-
LOSE GROUP . . . . 4 . . . I

PART II
COMPOUND CELLULOSES

LlGNOCELLULOSES . . . . . . . . . 89

PECTOCELLULOSES AND MUCOCELLULOSES ' . , . 214
ADIPOCELLULOSES AND CUTOCELLULOSES . ... 225

PART III
EXPERIMENTAL AND APPLIED . . 242

APPENDIX I . » . ., . . . . . . 311

PHOTO-MICROGRAPHS 312

APPENDIX II (1903) 313

INDEX OF AUTHORS , 321

INDEX OF SUBJECTS . . . . . . . . 323

CELLULOSE

PART I

THE TYPICAL CELLULOSE AND THE
CELLULOSE GROUP1

CELLULOSE is the predominating constituent of plant tissues,
and may be shortly described as the structural basis of the
vegetable world. The ordinary flowering plant is a complex
structure, and its several parts are also complex — that is, are
made up of cells. These cells exhibit an infinite variety of
form, the main lines of differentiation necessarily conforming
with variations in function. The growing cell is, of course,
nitrogenous, the living functions depending upon its proto-
plasmic contents. What we have to deal with, however, is
the cell, less the cell-contents of whatever kind, whether
'organic' — that is, concerned in the assimilating or othei
living functions of the cell — or of the nature of by-products
of metabolism excreted or thrown off from the main stream
of matter undergoing elaboration into the essential structures
of the plant. We have to deal, in fact, with the cell-wall
or envelope, to which the term cellulose has been applied as
to a chemical individual. There are, as might be expected, a
great many varieties of cellulose, and the term must be taken
as denoting a chemical group. The celluloses, taken as a

1 Sometimes abbreviated to the short title « Cellulose.'

8

group, present the following characteristics : — colourless sub-
stances insoluble in all simple solvents, generally but variably
resistant to processes of oxidation and hydrolysis, non-nitro-
genous, and having the empirical constitution characteristic
of the carbohydrates, i.e. CnH2mOm. Their reactions are those
of 'saturated' compounds. Their empirical formulae and
relationships to the carbohydrates of low molecular weight
further indicate ' single-bond ' linking of their C atoms as
exclusively prevailing. It must be noted here that the typical
celluloses are not separated from the plant in a ' pure ' state,
but in admixture or in intimate chemical union- with other
compounds or groups of compounds. The latter are distin-
guished by greater reactivity, e.g. they readily yield to alkaline
hydrolysis (* pectic ' bodies), to oxidation (colouring matters),
or to the action of the halogens. In the latter is included
the very important group of lignified celluloses or lignocellu-
loses (woods) distinguished by the presence of keto-hexene
groups in union with the cellulose, and therefore combining
directly with the halogens. These points are sufficient to
indicate the principles underlying the general method adopted
in the laboratory for the isolation of cellulose from vegetable
raw material — which consists in (i) Alkaline hydrolysis —
boiling the tissue or fibre in solution of alkaline hydrates (1-2
p.ct. NaOH), and after washing (2) exposure to the action of
a halogen — chlorine gas or bromine water at the ordinary
temperature, and (3) second alkaline hydrolysis — boiling in
alkaline solutions, e.g. sodium sulphite, carbonate, or hydrate,
to complete the resolution and to dissolve away the products
formed from the non-cellulose constituents by the preceding
treatment.

In the case of very refractory substances such as the hard
woods, it is sometimes necessary to repeat the treatment with
the halogen. After such treatment and thorough washing the

The Typical Cellulose and the Cellulose Group 3

material is treated exhaustively with alcohol and with ether to
remove fatty or resinous by-products of the oxidation.

Cellulose obtained in this way from raw fibrous materials,
e.g. cotton, flax, hemp, ramie, is a white substance distinguished
by more or less lustre and translucency, retaining the structural
characteristics of the raw material, of 1*5 sp.gr., and as a
chemical individual distinguished amongst C.H.O compounds
by its negative or non-reactive characteristics.

In the brief account which ensues, of the general chemistry
of cellulose, cotton-cellulose is taken as the type. The points
of differentiation of other members of the group from the type
will be noted subsequently.

The empirical composition of the pure cellulose is repre-
sented by the percentage numbers

C 4".,".*;, & V.'V-.V- 44'2

H V>- V . . * . . 6-3

O . , ...... . , . 49-5

corresponding with the statistical formula C6H10O6. These
numbers represent the composition of the 'ash-free' cellulose.
All vegetable tissues contain a greater or less proportion of
inorganic constituents of which a certain proportion are
retained by the cellulose isolated, as described, or by any of
the processes practised on the large scale in the arts (infra).
The celluloses burn with a quiet luminous flame, leaving these
inorganic constituents as an ash, retaining more or less the
form of the original. In cotton the average proportion of
ash is o'i-o'4 p.ct. The composition of the ash has, no
doubt, certain specific relationships to the several celluloses,
their constitution and origin ; but such correlations are at
present too obscure for useful discussion.

In the preparation of filter paper for quantitative work it is
important to eliminate the ash constituents as far as possible,
and this is effected by treatment with hydrofluoric and other

£2

4 Cellulose

acids — * Swedish ' filter paper of good quality contains from
o'O3~o'o5 p.ct. ash constituents, and constitutes the purest
form of cellulose with which we can deal.

Cellulose and Water. Cellulose Hydrates.— All
vegetable structures in the air-dry condition retain a certain
proportion of water — or hygroscopic moisture, as it is termed —
which is readily driven off at 100°, but reabsorbed on exposure
to the atmosphere under ordinary conditions. The mean per-
centage of this ' water of condition ' varies from 6 to 1 2 in the
several celluloses ; and in any particular cellulose will vary
on either side of this mean number to the extent of 1-2 p.ct.
with the extreme range of ordinary atmospheric conditions
of temperature and tension of aqueous vapour.

The authors have made experiments on the ' drying ' of cellu-
loses in a current of carbonic acid gas. The 'hygroscopic
moisture' (6-8 p.ct.) is rapidly driven off from the air-dry fibrous
celluloses at 90-100°, and there is a further small loss of water
(i p.ct.) on raising the temperature to 180°. The loss at ico-
120° is 0*5 p.ct. ; after that the loss is slow and probably due to de-
composition. Gelatinous celluloses in the form of films (see p. 28)
when dried at 90- 1 00° also show a further loss, but much greater at
higher temperatures. Thus in one experiment an air-dry film lost
8-6 p.ct. on drying at 100° ; an additional 3*9 p.ct. on raising
the temperature to 160°.

In an earlier age of the science the question might have
been discussed whether this absorption and retention of water
is a chemical or physical phenomenon ; but this is rather a
question of terms. The main points to be noted are (i) the
property of attracting water is a property of the cellulose
substance itself, and is not in any way dependent upon the
form in which it occurs. The amorphous modifications of the
celluloses obtained by solution and reprecipitation in various
ways (infra) are equally ' hygroscopic.'

(2) The phenomenon is definitely related to the presence of

The Typical Cellulose and the Cellulose Group 5

OH groups in the cellulose molecule, for in proportion as
these are suppressed by combination (with negative radicles to
form the cellulose esters) the products exhibit decreasing
attractions for atmospheric moisture. It is to be noted that
some of these synthetical derivatives are formed with only
slight modifications of the external or visible structure of the
cellulose, of which, therefore, the phenomenon in question is
again shown to be independent.

(3) The 'condition' of the fibre- substance in respect of
hygroscopic moisture is an important factor of such properties
of fibre as make up its spinning qualities ; it also seriously
affects the tensile strength of papers and cellulose textiles.

(4) A study of the hydration and dehydration phenomena
of the celluloses indicates an unbroken continuity in the series
of cellulose-water compounds — or cellulose hydrates ; of which
series the ' water of condition ' or hygroscopic moisture of a
cellulose represents the final terms.

The proportion of water held by the celluloses in an atmosphere
saturated with aqueous vapour is necessarily very much greater
than in the ordinary atmosphere, partially saturated at the same
temperature. (See H. Miiller, Pflanzenfaser, p. 3.)

The ' moisture of condition ' is a factor of some moment, first in
the buying and selling of fibrous products, and secondly in the
processes by which they are worked up (spinning, and ' finishing').

(1) In a delivery of ioo/. value of a fibrous material, e.g.
paper pulp or half stuff, the ordinary variations in the atmospheric
moisture may occasion a difference of I/, to 2/. in the value. It is
important, therefore, to have a normal standard of reference. In
the case of wood pulp or cellulose in which there is a large com-
merce it is customary to fix this at 10 p.ct., which means that ioo
of air-dry pulp give 90 'dry' at 100° C.

If, therefore, in any test the percentage of dry pulp is estimated
at any figure, the corresponding percentage of 'normal' air-dry pulp
(10 p.ct. Aq) is obtained by adding \ to the percentage of dry pulp.

(2) Cotton-spinning is carried on under special and carefully
regulated conditions of temperature and atmospheric moisture,

6 Cellulose

which have been arrived at as the result of accumulated observa-
tion and experience. Raw cotton, however, is not by any means
a pure cellulose, and the spinning properties of the fibre are to a
certain extent conferred by the substances associated, in admixture
or combination, with the cellulose. There is no doubt, however,
that the physical properties of the cellulose are largely modified
by its water of condition ; and the fine adjustments of these ulti-
mate fibres to the conditions of the spinning frame, more espe-
cially in regard to the drawing and twisting, largely depend upon
the maintenance of an * optimum ' of hydration of the cellulose.

(3) Finishing processes— textiles and paper. — The * finish ' of
textiles and papers for the market is of very various kinds. The
last operations are those of closing and * surfacing,' and consist of
the mechanical treatments of beetling, mangling (textiles), calen-
dering, and glazing (textiles and paper).

The finish is considerably affected by the condition of hydra-
tion of the fibre, and this is affected by the method of drying up
(air-drying or hot-drying) and the amount of moisture present in
the fibre when submitted to the mechanical treatments. The
operation of causes of this kind is necessarily somewhat obscure.
The student should address himself to the work of observation of
the phenomena of hydration of the celluloses, studying all the
conditions which affect, and the changes which accompany, the
loss and gain of water.

It is evident from a very superficial examination of the plant
world that the celluloses originate in the gelatinous form, i.e. in
a condition of extreme hydration. Hydrates of identical cha-
racteristics are obtained on precipitating cellulose from solutions
in the several special solvents to be subsequently described.
These hydrates differ in certain respects from the anhydrous or
dehydrated celluloses ; thus they dissolve in strong nitric acid,
and in solutions of the alkaline hydrates of moderate concen-
tration, they also are more readily attacked (hydrolysed) by
boiling dilute acids and alkalis. It is necessary to keep this
in view in regard to the determination of cellulose in fresh
tissues. A previous dehydration of the tissue by air-drying or

The Typical Cellulose and the Cellulose Group 7

by long immersion in alcohol confers upon the cellulose a much
greater resistance to hydrolytic actions, with the effect of in-
creasing the proportion of cellulose surviving the treatments
previously described as necessary for the elimination of the non-
cellulose constituents.

Some more important aspects of these phenomena are dealt
with in a paper upon

The Hydration of Cellulose (J. Soc. Chem. Ind. 4).— In an
investigation of the celluloses of green fodder plants the authors
showed that by a preliminary artificial dehydration — by long im-
mersion in alcohol— the quantity of cellulose isolated by the usual
process of alkaline hydrolysis and oxidation was considerably in-
creased.

The following numbers obtained with a ^:rop of oats may be
cited as typical.

Percentage Cellulose Isolated*

(a) (3)

Directly After alcoholic dehydration Difference

Leaves 28-2 35'4 7'2

Stems 29-5 34-5 5-0

It is a matter of ordinary observation that the maturing of
vegetable tissues is attended by loss of water, and it is clear from
these results that the growing plant contains hydrated modifications
of cellulose, which by mere dehydration are converted into the more
resistant forms. It must also be recognised that the line of cellu-
lose has to be drawn in an arbitrary manner. Products which are
the residues of treatments of a certain degree of intensity must be
so defined, and are not to be regarded as chemical individuals in
the strict sense of the term.

The hydrates of cellulose generally react with iodine in
aqueous solution, giving an indigo-blue colouration. They also
exhibit an increased ' affinity ' for those colouring matters which
dye cellulose directly.

In all the more essential properties, however, no distinction
can be drawn between the celluloses and their hydrated modifi-
cation.

8 Cellulose

Solutions of Cellulose. — Cellulose is insoluble in all
simple solvents, water included. In presence of certain metallic
compounds, however, it combines rapidly with water, forming
the gelatinous hydrates just described, which finally disappear
in solution in the water. Of these solvents of cellulose the
simplest is (i) ZINC CHLORIDE IN CONCENTRATED AQUEOUS
SOLUTION (40 p.ct. ZnCl2). The solution process requires
the aid of heat (60-100°), and may be carried out as follows :
4-6 pts. ZnCl.2 are dissolved in 6-10 pts. water and one pt.
cellulose (cotton) stirred in till evenly moistened. The
mixture is set aside to digest at a gentle heat. When the
cellulose is gelatinised the solution is completed by exposure
to water-bath heat, stirring from time to time and renewing
the water which evaporates. In this way, a homogeneous
syrup is obtained. This solution is employed in the arts for
making cellulose threads or filaments which are carbonised
for use in the incandescent electric lamp ; the ' carbon ' so
obtained having a sufficient resistance to mechanical strain with
the suitable degree of electric conductivity (resistance) for the
requirements of the lamp. In preparing the cellulose thread
the viscous solution is allowed to flow from a narrow orifice
into alcohol which precipitates a hydrate — a hydrated cellulose-
zinc-oxide — of sufficient tenacity for manipulation as a thread.
It is freed from zinc oxide by digestion in dilute hydrochloric
acid and copious washing. The cellulose zinc chloride solu-
tion is also precipitated by water, retaining a much larger pro-
portion of water (of hydration). After thorough washing and
drying a product is obtained retaining from 18-25 P-ct*
ZnO ; the variation in the proportion of ZnO to cellulose, no
doubt, corresponding with variations in molecular weight of
the latter, and these depending upon the molecular condition
of the original cellulose and the conditions of the solution
process.

The Typical Cellulose and the Cellulose Group 9

(2) ZINC CHLORIDE AND HYDROCHLORIC ACID. — If the
ZnCl2 be dissolved in twice its weight of aqueous hydrochloric
acid (40 p.ct. HC1) a solution is obtained which dissolves
Cellulose rapidly in the cold. This alternative process has
certain advantages over the preceding, and is useful in labora-
tory investigations. So far it has received no industrial
applications. It is to be noted that the cellulose dissolved in
this reagent undergoes a gradual lowering of molecular weight
(hydrolysis).

This process of dissolving cellulose is of value in the investi-
gation of fibrous products in the laboratory in cases where an acid
solvent is preferable, and where it is necessary to avoid heat.

If to the solution of pure cotton cellulose in this reagent bromine
be added in quantity sufficient to colour the solution, the colour
persists for a lengthened period, showing that there is no absorption
of the bromine, and that, therefore, there are no C = C groups
in the cellulose molecule. With the lignocelluloses (see p. 138),
which are also soluble in this reagent, and are known by other
reactions to contain C = C groups, there is considerable absorption
of bromine.

It is also noteworthy that if this solution of cellulose be
coloured with Cr03, it persists for some time in the unreduced
state. There cannot, therefore, be any free CO.H groups in the
cellulose molecule, and the observation rather throws doubt on the

existence of such groups in an 'acetal' form —

(3) AMMONIACAL CUPRIC OXIDE. — The solutions of the
cuprammonium compounds generally, in presence of excess of
ammonia, attack the celluloses rapidly in the cold, forming a
series of gelatinous hydrates which finally pass into solution.
The solutions of the pure cuprammonium hydroxide are more
active in producing these effects than the solutions resulting
from the decomposition of a copper salt with excess of
ammonia. Two methods are in common use for the preparation
of these solutions, which should contain :

IO Cellulose

10-15 p.ct. e . . . Ammonia (NH,)
2-2-5 „ v.V£ , .. ... Copper (as CuO)

(1) To a solution of a cupric salt, ammonium chloride is
added, and then sodium hydrate solution in sufficient excess ;
the blue precipitate is thoroughly washed upon a cloth filter,
squeezed, and re-dissolved in ammonia solution of 0*92
sp.gr.

(2) Thin sheet copper is crumpled up, placed in a glass
cylinder and covered with strong ammonia. Atmospheric air
is drawn by aspiration so as to bubble through the liquid
column at such a rate as to amount per hour to about forty
times the volume of liquid used. In about six hours a solution
is obtained of the composition given above (C. R. A. Wright).

Under the latter conditions the action of the solution upon
the cellulose may be made simultaneous with its production.
For this the cellulose and metal are mixed together as inti-
mately as possible and exposed as described to the action of
aqueous ammonia and oxygen.

There are various ways of accelerating the preparation of the
cuprammonium solution from the metal. Thus, as compressed
oxygen is now an ordinary commodity it is easy to substitute the
pure gas for the atmospheric mixture, with the result that the
volume of gas passing through the solution may be considerably
reduced, and therefore the loss of ammonia lessened.

The oxidation of the copper is facilitated by contact with a metal
which is * negative ' to the copper in presence of ammonia ; or this
differential disposition of the copper to be attacked may be more
directly attained by means of the electric current, the copper to be
attacked being brought into conducting connection with the
negative, and the second metal with the positive pole of a battery
— the latter being inserted in a porous pot within the alkaline
liquid (Hime and Noad, English Patent 7716/89).

The solutions of cellulose in cuprammonium are of little
stability, the cellulose being readily precipitated by the

The Typical Cellulose and the Cellulose Group n

addition of neutral dehydrating agents such as alcohol, sodium
chloride (and other salts of the alkalis), and even sugar.
From a study of these solutions, indeed, Erdmann concluded
(J. Pr. Chem. 76, 385) that they were not solutions of cellulose
in the strict sense of the term, the cellulose being rather
gelatinised and diffused through the solution as a highly
attenuated (hydrated) solid of this description. Cramer, on
the other hand, showed by osmotic experiments that this
inference was unfounded and that the solution of the cel-
lulose may be regarded as complete. According to modern
views on the subject of solution generally, and the solution
of colloids in particular, the lines drawn by the older inves-
tigators of these phenomena are of arbitrary value ; gelati-
nisation being expressed as a continuous series of hydrations
between the extreme conditions of solid on the one side and
aqueous solution on the other. This point will be further con-
sidered later on.

The evidence goes to show that the solution process, though not
the result of an oxidation of the cellulose — such as would be attended
by reduction of CuO — is attended by a disturbance of the ' balance
of oxidation ' of the cellulose molecule. By prolonged contact with
the cuprammonium the cellulose does in fact appear to be oxidised
(to oxycellulose) (Prudhomme, J. Soc. Dyers and Col., 1891, 148).

The ammonia also undergoes oxidation, and the cuprammonium
solutions, after keeping, will be found to contain a considerable
quantity of nitrite (ibid.). Cotton cellulose does not appear to be
hydrolysed by the process of solution, that recovered from the solu-
tion by precipitation by acids &c. having approximately the same
weight as that of the fibre originally dissolved.

There are celluloses, on the other hand, which are partially
hydrolysed, and when reprecipitated the cellulose recovered is
found to be in defect, and the solution to contain dissolved carbo-
hydrates.

Further investigation of these points is much needed, i.e.
quantitative determination of the oxidation and hydrolysis of the

12 Cellulose

several celluloses under treatment with the cuprammonium reagent
The evaporation of the cuprammonium solutions of cellulose upon
glass surfaces gives a film of the mixed cellulose-cupric hydrate,
but of little tenacity. It will appear as we proceed that high tensile
strength of a film obtained from solutions of cellulose compounds
indicates a relatively high molecular weight, and conversely, a
brittle product is evidence that in forming the compound the mole-
cular weight or aggregation of the cellulose has been lowered.

According to recent investigations of E. Gilson (Chem.
Centr. 1893, ii. 530) cellulose maybe crystallised from its solu-
tion in cuprammonium. If such solution is left to stand in a
loosely closed vessel the ammonia escapes, cellulose being pre-
cipitated together with hydrated copper oxide. On removing
the latter by treatment with hydrochloric acid, the cellulose is
stated to remain in the form of nodular crystals. It is also
stated that when sections of cellulosic tissues are allowed to
remain for some time in contact with the reagent, then gradu-
ally washed with ammonia and water, the interior of the cells
are found to contain the cellulose in crystalline form. This
requires confirmation.

These cuprammonium solutions are, of course, deprived of
their copper by digestion upon zinc, the latter metal replacing
the copper in solution and, under carefully regulated conditions,
without precipitating the cellulose, so that a colourless solu-
tion of the latter in zinc-ammonium-hydroxide results. Some
of these solutions have been observed to be laevogyrate.
Cotton cellulose in i p.ct. cuprammonium solution was found
by Levallois to show a rotation of —20° ; the rotation, however,
is not constant, but varies with the concentration and the ratio
of cupric oxide to cellulose in the solution. These observations
have been called in question by Bechamp, but reaffirmed by
the former observer, and apparently on sufficient evidence.

On adding a solution of lead acetate to these solutions of
cellulose a precipitate is obtained of a compound of cellulose

The Typical Cellulose and the Cellulose Group 13

with lead oxide, but of variable composition ; the compound
«(CGN10O5.PbO) appears to result from the treatment of the
ammoniocupric solution with finely divided lead oxide.

This property of gelatinising and dissolving cellulose has
been taken advantage of in important industrial applications
of the cuprammonium compounds. Vegetable textile fabrics
passed through a bath of the cuprammonium hydroxide are
' surfaced * by the film of gelatinised cellulose, which retains
the copper oxide (hydrate) in such a way that it dries of a bright
'malachite' green colour. By this treatment the fibres are
further compacted together, and the fabric acquires a water-
resistant character. The presence of the copper oxide is also
preservative against the attacks of mildew, insects, &c. If the
fabrics are rolled or pressed together when in the gelatinised
condition they become firmly welded together on drying, and
a variety of compound textures are produced in this way.

These fabrics are sold under the style or description of
1 Willesden ' goods ; the manufacture being in the hands of a
company whose works are situated at Willesden. The
company's processes are based on the patents of Drs. J.
Scoffern and C. R. A. Wright (q.v.\

AMMONIACAL CUPROUS OXIDE. — According to M. Rosen-
feld (Berl. Ber. 12, 954) a concentrated solution of cuprous
chloride in ammonia dissolves cellulose rapidly.

The reaction of cuprammonium with cellulose, although iden-
tified with the name of Schweitzer, appears to have been first
noticed by Mercer. He employed a solution of ammonia of 0-920
sp.gr. saturated at the ordinary temperature with the cupric oxide
(hydrate) and diluted with three volumes of water. Mercer investi-
gated the reaction in regard to the influence of the conditions of
treatment, showing that it was retarded by the presence of salts,
and hence that the solutions obtained by decomposing the copper
salts with excess of ammonia were much less active than equivalent
solutions of the pure hydrate. He also showed that the activity of

14 Cellulose

the solution was considerably retarded by raising its temperature,
becoming very slight at 100° F.

Mercer's favourite method of demonstrating the reaction con-
sisted in applying a solution of cupric nitrate to cotton cloth in
spots, decomposing the nitrate by plunging the cloth into a weak
solution of caustic soda, washing to remove the alkali, partially
drying — in the air at ordinary temperatures, and exposing the cloth
to the vapour of ammonia. In this way the cellulose was fully
acted upon in the portions containing the oxide. The demon-
stration is an interesting one, and should be repeated by the
student.

Theory of Action of Cellulose Solvents. — The causes
underlying the processes of dissolution of cellulose above de-
scribed will become more apparent as we proceed in the dis-
cussion of its special chemistry. For the present it is sufficient
to point out that they depend upon the presence in the cellu-
lose molecule of OH groups of opposite function, basic and
acid, and that the compounds formed with the solvents are
of the nature of double salts.

Qualitative Reactions and Identification of Cellu-
lose.— The properties of cellulose which we have already dis-
cussed afford the means of identifying it : that is (i) by reason
of its resistance to the action of oxidising agents, to the halo-
gens and to alkaline solutions it is obtained as a residue from
the treatment of vegetable tissues by these reagents in succes-
sion ; (2) it is soluble to gelatinous or viscous solutions in
the reagents above described — viz. ZnCl2.Aq, ZnCl2.HCl.Aq,

and Cu<^ TT4^> ^rom wmcn it is obtained by pre-

cis rl3 — JN r!4U

cipitation in the amorphous form or as a gelatinous hydrate.
These hydrates react in many cases with iodine, giving a blue
colouration ; the reaction is determined upon the original
cellulose by simultaneous treatment with iodine and a de-
hydrating solution.

The Typical Cellulose and the Cellulose Group 15

The most effective reagent is prepared as follows : zinc is
dissolved to saturation in hydrochloric acid and the solution
evaporated to 2-0 sp.gr. ; to 90 parts of this solution are
added 6 parts potassium iodide dissolved in 10 parts water,
and in this solution iodine is dissolved to saturation. By this
reagent cellulose is coloured instantly a deep blue or violet.

A superficial examination usually suffices to identify cellulose in
the mass, and an examination with the microscope establishes the
histological characteristics of the substance. There are cases,
however, in which distinctive tests require to be applied, and these
will be selected in order of convenience. Thus, by means of the
chemical tests, cellulose has been identified as a constituent of
many animal tissues (see p. 87) ; in these cases, of course, it could
be identified in no other way.

It will be seen as we proceed that a number of the properties
of cellulose are common to many of the * compound celluloses '
which are widely distributed in the plant world ; these are, how-
ever, differentiated by the special reactions depending upon the
compounds or groups with which the cellulose may be com-
bined.

Lastly, the cellulose group proper includes a number of sub-
stances which are differentiated from the typical cotton cellulose
in some specific property. These will be noted subsequently.

Compounds of Cellulose. — The chemical inertness of
cellulose is a matter of every-day experience in the laboratory,
where it fulfils the important function of a filtering medium in
the greater number of separations of solids from liquids. The
functions which it discharges in the plant world as well as the
numberless uses which it subserves in the world of humanity
are all referable to the predominance of these negative
characteristics. Cellulose, however, is a poly-hydroxy- com-
pound, and enters into a number of reactions characteristic of
the alcohols. These reactions, and the products of synthesis
resulting from them, we shall deal with in order, proceeding
from the less to the more definite.

1 6 Cellulose

In a general way the inertness of cellulose may be compared
with that of inorganic salts, more particularly those which result
from the combination of the weaker acids and bases. Cellu-
lose in reaction shows both acid and basic characteristics, and, as
we shall see, these properties may be explained by proximity of
its OH groups to CO and to CH2 groups respectively within the
molecule.

It appears, moreover, that these OH groups are in a condition
of reciprocal suppression, requiring the application of powerful
reagents or severe conditions to bring them into reaction.

This condition of its OH groups appears to be associated with
the endothermic constitution or configuration of the cellulose
molecule. There is a good deal of evidence physiological and
chemical that the formation of cellulose is associated with the
absorption of energy beyond what may be taken as normal to a
saturated compound of the empirical formula C6H10O5.

DILUTE ALKALIS AND ACIDS. — It has been shown that
pure bleached cotton enters into reaction with the acids and
basic oxides when plunged even into cold and highly dilute
solutions of these compounds (Mills). In illustration of this
point the following results of experiments may be cited : —

Reagent

Temperature

Time

Weight absorbed

H2S04
HC1
NaOH

4°

M

>»

3 mins.
»

M

0-00495
0-00733
0-02020

The molecular ratio of the absorption of the two latter —
viz. 3HC1, loNaOH — appears to hold good for a somewhat
wide range of conditions ; and it may be noted that the same
ratio was observed for silk, though the observation can only be
regarded as a coincidence.

These reactions of cellulose have been by no means exhaus-
tively investigated ; as our knowledge of the group of celluloses
and of their differentiations one from the other is extended, it
becomes necessary to institute a careful comparison in regard to
this property of ' absorbing * reagents.

The Typical Cellulose and the Cellulose Group 17

An examination of the structureless cellulose regenerated from
solutions of cotton as alkaline thiocarbonate (see p. 29) shows an
important differentiation from the original cellulose (bleached
fibre). This form of cellulose, after careful purification, was found
to combine with the caustic alkalis in dilute solution, in much
larger proportion ; thus from solutions of 3-1 p.ct. Na^O the cellu-
lose removed, i.e. combined with, the alkali to the extent of 5-6
p.ct. of its weight. With the dilute acids, on the other hand, no
increased combination was observed.

This phenomenon has been more recently studied from the
independent standpoint of thermal equilibrium. It has been
shown that when pure cotton is plunged into dilute solutions
of the acids and alkalis, liberation of heat takes place (Vignon).
The rise of temperature was found to be slow, and, under the
conditions chosen for the experiments, ceases after the lapse of
seven to eight minutes. The following are typical results in
calories per 100 grms. of cotton.

— .

KOH

NaOH

HC1

H.SO4

Raw cotton
Bleached . .

l'30
2-27

105
2-20

0-65
0-65

O'6o
0-58

It would appear from these results that cellulose has the
properties of a feeble acid, and of a yet feebler base. From
the comparative insignificance of the * affinities ' involved, it
might be inferred that they could have but a small determining
value in regard to the uses or applications of cellulose. Recent
researches, however, have shown that the combinations of
cellulose with colouring matters, i.e. the dyeing properties of
the fibre-subslance, are largely dependent upon a play of
1 affinities ' of this order and narrow range. Vignon concludes, in
fact, from a careful and exhaustive survey of dyeing phenomena,
including the action of mordants, that they depend chiefly upon
the interaction of groups of opposite chemical function, viz.

c

1 8 Cellulose

basic and ' acidic,' present in the colouring matter or mordant
and the substance with which it combines.

This explanation certainly covers a wide range of such re-
actions, but we shall find that the molecular constitution of the
fibre-substance is also an important factor. This point will
be discussed subsequently.

Capillary Phenomena.— The absorption and trans-
mission of solutions by cellulose is attended by a number of
special effects. Schonbein appears to have been the first to
observe that strips of unsized paper, of which one end was
placed in an aqueous solution, e.g. of a metallic salt, will
absorb and transmit the water more rapidly than the dissolved
salt, which is therefore ' filtered out ' ; further, that to the
various salts, cellulose manifests varying degrees of re-
sistance to transmission in solution. These phenomena have
been further studied by Lloyd (Chem. News, 51, 51) for
metallic salts,1 and by F. Goppelsroeder (Berl. Ber. 20, 604) for
various colouring matters ; the results of their observations
constituting the beginnings of a method of * capillary analysis
or separation.' The subject is comparatively new and not yet
systematised, but the method is undoubtedly capable of con-
siderable extension.

Contrasted with the relatively feeble attractions of cotton
cellulose for the acids and bases of low molecular weight there are a
number of cases ot special combinations which take place in much
higher proportions.

Thus the fibre removes a considerable quantity of barium
hydrate on digestion with the solution ; and from solutions of the
basic salts of lead, zinc, copper, tin, aluminium, iron, chromium,
&c. the fibre takes up considerable but variable proportions of the
respective basic oxides. The formation of these compounds
underlies the well-known processes of ' mordanting ' practised by
the dyer and printer of textiles. The theory of these processes will

1 More recently by E. Fischer and Schmidmer, Lieb. Ann. 272, 156.

The Typical Cellulose and the Cellulose Group 19

be found fully treated of in the text-books of these arts. We can
only call attention to those properties which are common to the
group of basic oxides capable of acting as mordants, viz. (i)
they are all oxides of di- or polyvalent elements ; (2) they form
colloid or gelatinous hydrates ; (3) their salts dissociate in solu-
tion into acid and basic salt ; (4) they are soluble in the alkaline
hydrates either directly or in presence of 'organic' hydroxy- com-
pounds. Certain of the acid oxides of the metals are also removed
by cellulose from solutions of these salts, but in relatively small
proportion ; of these the stannic compounds are most important
from the point of view of application as mordants.

Amongst ' organic ' acid bodies, tannic acid is conspicuous for its
'affinity' for cellulose. From aqueous solutions of tannic acid
cotton fibre takes up as much as 7-8 p.ct. of its weight ; and the
process of mordanting with this compound is one ot the most
generally useful.

The combinations of cellulose with colouring matter open up a
number of interesting problems. A colouring matter may be said
to dye a fibre or substance when it forms with it a * lake ' com-
pound, a lake being merely a pigment form of the colouring matter
in which its essential physical properties are preserved. By recent
investigation the properties determining lake formation have been
shown to be definitely correlated with the molecular constitution of
the colouring matter, i.e. more particularly with the nature and dis-
position of its chromogenic groups (NH2: SO3H, COOH, NO2,
N.OH, and OH groups).

An excellent treatment of this subject will be found in two papers
by C. O. Weber in the J. Soc. Chem. Ind. 10,896, 12,650, to which
the student is referred. A discussion of these problems is outside
the scope of this work. It may, however, be pointed out that, as of
course the phenomena of dyeing depend upon reciprocal attraction,
we may confidently expect that further investigation will lead to a
correlation of the specific or selective attractions of cellulose for
colouring matters, with its molecular constitution. It should be
remembered that there are three factors of the problem : (i) the
constitution of the colouring matter ; (2) that of the substance with
which it combines ; and (3) the condition of the colouring matter
in solution, and the causes which determine its transference to the
solid with which it is brought into contact. Of these we have pre-

C 2

2O Cellulose

cise knowledge of the first only ; the ' theory of solution ' is a
recent development, and the third factor is at present to be dealt
with only speculatively ; and of the constitution of the celluloses
we have at present only a general knowledge.

The further investigation of these problems is therefore probably
the most promising direction from which to approach the position
of the actual molecular constitution of the celluloses.

The actions of dilute alkaline and acid solutions at higher
temperatures are, of course, more pronounced. The mineral
acids of concentration, equal to semi-normal at the boiling
temperature, rapidly destroy in the sense of disintegrating
cellulose fibres, producing an important molecular change in
the cellulose itself. The modified cellulose is brittle and
pulverulent, and will be more fully described as the product
of the action of concentrated hydrochloric acid, viz. as hydro-
or hydracellulose. The time required for completing this
change varies of course with the temperature and the concen-
tration of the acid. The acid treatments of cellulose textiles,
which are necessary incidents of bleaching and dyeing opera-
tions, are carried out as a result of practical experience well
within the limits of safety ; such treatments being for the most
part in the cold (<7o° F.) and at strengths of 0-5-2-0 p.ct.
(HC1,H2SO4). In dyeing operations requiring an acid bath
and the boiling temperature, ' free ' mineral acids are as much
as possible avoided, acetic acid being substituted. This acid
is without sensible action on cotton.

The action of the acids in disintegrating cellulose structures
is undoubtedly hydrolytic^ and of the same order, for instance,
as their action upon cane sugar. The ' inversion ' of saccha-
rose by boiling with the dilute acids is not, it must be remem-
bered, a simple process of hydration ; according to the usual
equation,

CiiHMOn + H2O = C6H12O6 + C6H12O6,

Dextrose Levulose

The Typical Cellulose and the Cellulose Group 21

these products of the hydrolysis being susceptible of ' conden-
sations,' in which the reverse action is determined. On the
other hand, it has been recently shown (Wohl) that the con-
ditions may be so chosen as to exclude the latter, viz. by ope-
rating in the cold and in presence of a minimum of water, in
which case we get the surprising result that the hydrolysis of
the sugar is determined by 3T>V?T P-ct- °f its weight of HC1.

Applying these considerations to the case of the more
complex cellulose molecule it is easy to see that it may undergo
a series of hydration changes, with attendant resolutions, with-
out any change of its empirical formula. The disintegrating
action of the dilute acids appears to be of this kind.

The action of the aqueous acids upon cellulose has been investi-
gated by various observers, amongst others by Grace Calvert,
Girard (Compt. Rend. 81, 1105), C. Koechlin (Bull. Mulhouse,
1888). The latter observer gives the results of a study of the
limiting conditions of action of aqueous sulphuric acid at various
degrees of concentration. What may be called the critical con-
centration of the acid lies between the limits of 60-80° B. Thus
with the mixture of 3 vols. of concentrated acid and 8 vols. water
— i.e. an acid of 69° B. — at the ordinary temperature, its action
upon cotton does not become evident until after three hours'
exposure. With an aqueous acid containing 100 grms. H2SO4 per
litre and at 80° C. the first appearances of change in the cotton are
noticed at the expiration of five minutes ; after thirty minutes'
exposure there is sensible disintegration ; and the completion of
the action, i.e. conversion into a friable mass of hydrocellulose,
requires an exposure of 60 minutes' duration.

To alkaline solutions at high temperatures, cotton cellulose
is, on the other hand, very resistant. Solutions of caustic soda
of 1-2 p.ct. Na2O are without sensible action upon cotton
at temperatures considerably over 100°. The principal opera-
tions in the process of bleaching cotton and linen textiles con-
sist in drastic alkaline treatments of this kind, whereby the
' non-cellulose ' constituents of the fibre are for the most part

22 Cellulose

saponified and removed in solution in the alkaline lye. The
oxidation processes which follow — viz. treatment with the
hypochlorites, permanganates, &c. in dilute solution — although
they may be regarded as the bleaching processes proper, really
accomplish very little beyond removing residues or by-products
of the alkaline treatment. It is also evident that resistance to
alkaline treatment is a very important condition in the every-
day uses of cellulose textiles.

H. Tauss has recently investigated the action of alkaline
solutions upon various celluloses at high temperatures (J. Soc.
Chem. Ind. 1889, 913; 1890, 883). Purified cotton cellulose,
digested with solutions of sodium hydrate of 3 p.ct. Na2O three
times in succession, is attacked and converted into soluble
products in the following proportions, increasing with the temper-
ature at which digested :

I atm. pressure .... I2'l p.ct.
5 » » .... 15-4 »

10 „ „ .... 20-3 „

Strong aqueous solution of ammonia is without sensible action
on cellulose until a very high temperature is reached. At 200°
combination ensues, and the entrance of the NH.2 residue into the
cellulose molecule is evidenced by the increased attraction of the
product for colouring matters, approximating to that of the animal
fibres. (L. Vignon.)

We have mentioned that digestion with 3 p.ct. solutions of soda
(Na.2O) at high temperatures produces a certain conversion of
cellulose into soluble substances. Solutions of 8 p.ct. (Na2O)
strength have been found to give the following results (Tauss,
loc. tit.} :—

I atm. pressure . . . 22-0 p.ct. dissolved
5 » » ... 58-0 „ „

10 „„ ... 59'0 „ „

In connection with these observations it is to be noted that a
process of estimating the cellulose in compound celluloses (wood)
has recently been proposed (Lange, Zeitschr. f. Physiol. Chem. 14),
jmd adopted by other observers, based upon the action of strong'

The Typical Cellulose and the Cellulose Group 23

solutions of sodium hydrate at high temperatures upon the ligni-
fied tissue. It is assumed that the non-cellulose constituents of the
woods (see p. 172) are exclusively attacked by the treatment :
which, however, is by no means the case, as the results of Tauss
(loc. tit.} sufficiently show. Quantitative results obtained by this
method have, therefore, only a limited value ; and, as estimations
of ' cellulose,' are subject to large and variable errors.

CONCENTRATED SOLUTIONS OF THE ALKALIS. — Cold solu-
tions of the alkaline hydrates of a certain concentration exert
a remarkable effect upon the celluloses. Solution of sodium
hydrate, at strengths exceeding 10 p.ct. Na2O, when brought
into contact with the cotton fibre, at the ordinary temperature,
instantly changes its structural features, i.e. from a flattened
riband, with a large central canal, produces a thickened cylin-
der with the canal more or less obliterated. These effects in the
mass, e.g. in cotton cloth, are seen in a considerable shrinkage
of length and width, with corresponding thickening, the fabric
becoming translucent at the same time. The results are due
to a definite reaction between the cellulose and the alkaline
hydrates, in the molecular ratio Ci2 H20 O10 ' 2NaOH, accom-
panied by combination with water (hydration). The com-
pound of the cellulose and alkali which is formed is decom-
posed on washing with water, the alkali being recovered un-
changed, the cellulose reappearing in a modified form, viz.
as the hydrate C12H20O10.H2O. By treatment with alcohol,
on the other hand, one half of the alkali is removed
in solution, the reacting groups remaining associated in the
ratio— C12H20O10 : NaOH. The reaction is known as that
of Mercerisation, after the name of Mercer, by whom it was
discovered and exhaustively investigated. Although, however,
it aroused a good deal of attention at the time of its dis-
covery, it remained for thirty years as an isolated observation,
Le. practically undeveloped. Recently, however, the alkali

24 Cellulose

cellulose has been made the starting-point of two series of
synthetical derivatives of cellulose, which must be briefly
described.

An interesting account of Mercer's researches on this subject
is given in 'The Life and Labours of John Mercer' (E. A. Par-
nell, London, 1886), a work which maybe particularly commended
to the young student.

From the points established by Mercer in connection with this
reaction the following may be further noted : —

At ordinary temperatures a lye of 1-225-1-275 sp.gr. effects
' mercerisation ' in a few minutes ; weaker liquors produce the
result on longer exposure, the duration of exposure necessary being
inversely as the concentration. Reduction of temperature produces,
within certain limits, the same effect as increased concentration.
The addition of zinc oxide (hydrate) to the alkaline lye also increases
its activity. Caustic soda solution of noo sp.gr., which has only a
feeble ' mercerising ' action, is rendered active by the addition of the
oxide in the molecular proportion, Zn(OH)2 : 4NaOH.

The condition of the cotton also affects the result. The ordi-
nary bleaching process, with its treatment with boiling alkaline
lye under pressure, brings the cellulose into a condition relatively
unfavourable, the best results being obtained by a preparatory
treatment consisting of (i) boiling with water only, (2) bleaching
in a warm bath (60-70° C.) of hypochlorite (bleaching powder) pre-
pared with addition of lime.

In regard to the physical changes of the fibre-substance result-
ing from the treatment, the effects in the mass, i.e. in yarn or cloth,
are seen in shrinkage of linear dimensions, with a corresponding
increase in thickness. The percentage of shrinkage observed is 20-
25. The ' mercerised' fabric shows an increase of strength, i.e. re-
sistance to breaking strain, of from 40-50 p.ct. Another important
feature of the * mercerised ' fabrics is an increased dyeing capacity.
These changes of form and in properties were investigated by
W. Crum (Chem. Soc. Journ. 1863).

The changes in the minute structure of the cell he showed
to be similar to those which accompany the process of ripening
— i.e. from the flattened riband form of a collapsed tube to the
cylindrical form resulting from the uniform thickening of the

The Typical Cellulose and the Cellulose Group 2$

cavity of the cell wall. Owing to this thickening the cavity of
the cell is almost obliterated. Another effect of the alkali is to
produce a peculiar spiral twisting of the fibre, which further
explains the shrinkage of cloth in the process of mercerising ; the
shrinkage being in part due to the felting together of the twisted
fibres, after the manner of wool fibres in the process of 'fulling 'cloth.

CELLULOSE THIOCARBONATES. — When 'mercerised' cot-
ton, or more generally an alkali-cellulose (hydrate), is exposed
to the action of carbon disulphide at the ordinary temperature,
a simple synthesis takes place, which may be formulated by the
typical equation :

X.ONa + CS2 = CS. °*.

The best conditions for the reaction appear to be when the
reagents are brought together in the molecular proportions :

C6H1005 2NaOH CS2 f „ Q1 .
162 2x40 76 L3°-4oH2OJ,

the second ONa group being in direct union with the cellu-
lose molecule, which reacts, therefore, as an alkali cellulose.
The resulting compound may therefore be described as an
alkali-cellulose-xanthate. It is perfectly soluble in water, to a
solution of extraordinary viscosity. The course of the reaction
by which it is produced is marked by the further swelling of
the mercerised fibre and a gradual conversion into a gelatin-
ous transparent mass, which dissolves to a homogeneous solu-
tion on treatment with water.

To carry out the reaction in practice, bleached cotton is
treated with excess of a 15 p.ct. solution of NaOH, and
squeezed till it retains about three times its weight of the solu-
tion It is then placed in a stoppered bottle with carbon
disulphide, the quantity being about 40 p.ct. of the weight
of the cotton. After standing about three hours at ordinary
temperatures, water is added sufficient to cover the mass, and

26 Cellulose

the further hydration of the compound allowed to proceed
spontaneously some hours (e.g. over night). On stirring, a
homogeneous liquid is obtained, which may be diluted to any
required degree.

Thus prepared, the crude solution is of a yellow colour, due
to by-products of the reaction (trithiocarbonates). The pure
compound is obtained either by treatment of the solution with
saturated brine or with alcohol. It forms a greenish-white
flocculent mass or coagulum, which redissolves in water to a
colourless or faintly yellow coloured solution. Solutions of
the salts of the heavy metals added to this solution precipitate
the corresponding xanthates. Iodine acts according to the
typical equation :

CS + ° CS + '• = *NaI + CS' ° CS"

Na

The compound, which may be described as a cellulose
dioxythiocarbonate, is precipitated in the flocculent form ; it
is redissolved by alkaline solution, in presence of reducing
agents, to form the original compound.

The most characteristic property of the cellulose xanthates is
(a) their spontaneous decomposition into cellulose (hydrate), alkali,
and carbon disulphide — or products of interaction of the latter.
When this decomposition proceeds in aqueous solution, at any
degree of concentration exceeding i p.ct. cellulose, a jelly or
coagulum is produced, of the volume of the containing vessel.
These highly hydrated modifications of cellulose lose water
very gradually, the shrinkage of the * solid ' taking place sym-
metrically. The following observations upon a 5 p.ct. solution
(cellulose), kept at the ordinary atmospheric temperature, will
convey a general idea of the phenomena attending the regene-
ration of cellulose from the alkali xanthate. The observations
were made upon the solution kept in a stoppered cylinder ;
after coagulation the solution, expressed from the coagulum of

The Typical Cellulose and the Cellulose Group 27

cellulose by spontaneous shrinkage, was removed at intervals.
Original volume of solution, 100 c.c.

rp- • , „ Vol. of cellulose Diff. from loo c.c.

Time in days hydrate ^^ expressed

Coagulation . . 8th day
First appearance ) .,
of liquid . . i

i6th , .98 c.c. . » . 2 c.c.

20th
25th
30th
40th
47th

83'5 • . • '• '6-5

72-0 .';-.•---. 28 'o
58-0 * v >:* 42-0
42-8 . . . 57-2
38-5 • • .61-5

The shrinkage from a 5 p.ct. to a 10 p.ct. coagulum of cellu-
lose hydrate is therefore extremely slow and fairly regular ;
from 10-12 p.ct. there is considerable retardation; and at
12-15 p.ct. the coagulum may be considered as a hydrate,
stable in a moist atmosphere. It follows from these observa-
tions that if a 10-12 p.ct. solution be allowed to coagulate
spontaneously, the resulting cellulose hydrate will undergo very
small shrinkage if kept in a moist atmosphere. These obser-
vations indicate the uses which can be made of the solution in
preparing cellulose casts and moulds.

As regards the problem of hydration and dehydration of
the cellulose there are, of course, other methods of approxi-
mately determining the * force ' by \\hich the water molecules
are held. It is a problem of wide significance, by reason of the
important part played by such hydrates in the economy of
plant life. Further investigations of the problem, therefore,
by the various known methods are being prosecuted.

(£) Coagulation by heat. — The solution may be evaporated
at low temperatures to a dry solid, perfectly resoluble in water.
If heated at 70-80°, however, the solution thickens ; and at
80-90° the coagulation, i.e. decomposition, is rapidly com-
pleted. If the solution be dried down at this temperature in

28 Cellulose

thin films, it adheres with great tenacity to the surface upon
which it is dried. On treatment with water, however, the
cellulose film may be detached, and when freed from the by-
products of the reaction the cellulose is obtained as a homo-
geneous transparent colourless sheet or film, of great toughness,
which, on drying, hardens somewhat, increasing in toughness
and preserving a considerable degree of elasticity. From the
properties of the solution and of the cellulose regenerated from
it, it will be readily seen that both are capable of extensive
applications.

QUANTITATIVE REGENERATION OF CELLULOSE FROM
SOLUTION AS THIOCARBONATE. — Very careful experiments have
been made to determine the proportion of cellulose recovered
from solution as thiocarbonate. Weighed quantities of Swedish
filter paper were dissolved by the process, and the solutions
treated as follows : (a) Allowed to ' solidify ' spontaneously at
15-18°. (b) Coagulated more rapidly at 55-65°. (c) Sul-
phurous acid was added in quantity sufficient to combine with
one-third of the alkali present in the solution — the resulting
solution being colourless : this was then set aside to coagulate
spontaneously. The regenerated celluloses were exhaustively
purified, by boiling in sodium sulphite solution, digesting in
acid, digesting in water, &c., and, repeating the treatments
until pure, they were finally dried for some days at 60° and
finished at 100°.

The following results were obtained :

Weight of original cellulose Weight of regenerated cellulose
(«) 17335 I*748o

(b) 17415 17560

(c) x -8030 I -8350

The results show a net difference of i'i p.ct. (increase), a
quantity which, for practical purposes, may be neglected. As,
however, the empirical composition of the regenerated cellulose

The Typical Cellulose and the Cellulose Group 29

indicates hydration to 4C6H10O5.H2O (infra), and a corre-
sponding gain of 2*7 p.ct, it appears that there is a slight
hydrolysis of even this very pure form of cellulose. From
subsequent observations (p. 61) it will appear that the hydro-
lysis falls upon an oxycellulose, probably present in all bleached
celluloses.

The cellulose regenerated from the thiocarbonate differs
from the original cellulose, so far as has been ascertained, in
the following respects :

(1) Its hygroscopic moisture, or water of condition, is some
3-4 p.ct. higher, viz. from 9-10*5 p.ct.

(2) Empirical composition. — The mean results of analysis
show C=43*3, H=6'4 p.ct., which are expressed by the
empirical formula, 4C6H10O5.H2O.

(3) General properties, in the main, are identical with
those of the original, but the OH groups of this cellulose are
in a more reactive condition. Thus this form of cellulose is
acetylated by merely heating the acetic anhydride at its boiling
point, whereas normal cellulose requires a temperature of 180°.
( Vide Cellulose Acetates.)

As regards reaction in aqueous solution we may notice that
it has a superior dyeing capacity, and also combines with the
soluble bases to a greater extent : e.g. if left some time in
contact with a normal solution of sodium hydrate it absorbs
from 4*5-5*5 p.ct. of its weight in combination.

Towards the special solvents previously described it
behaves similarly to the normal or fibrous cellulose; the
solutions obtained are, however, more viscous and less gela-
tinous.

THEORETICAL VIEW OF THE THIOCARBONATE REACTION OF
CELLULOSE.— The occurrence of this reaction, under what may
be regarded as the normal conditions, proves the presence in
cellulose of OH groups of distinctly alcoholic function. The

3O Cellulose

product is especially interesting, as the first instance of the
Synthesis of a soluble cellulose derivative— i.e. soluble in
water — by a reaction characteristic of the alcohols generally.
The actual dissolution of the cellulose under this reaction we
cannot attempt to explain, so long as our views of the general
phenomena of solution are still only hypotheses. There is
this feature, however, common to all the processes hitherto
described, for producing an aqueous solution of cellulose (i.e.
a cellulose derivative), viz. that the solvent has a saline cha-
racter. It appears, in fact, that cellulose yields only under the
simultaneous strain of acid and basic groups, and therefore we
may assume that the OH groups in cellulose are of similarly
opposite function. In the case of the zinc chloride solvents
there cannot be any other determining cause, and the soluble
products may be regarded as analogous to the double salts.
The retention of the zinc oxide by the cellulose, when pre-
cipitated by water, is an additional evidence of the presence of
negative or acidic OH groups ; and, conversely, the much more
rapid action of the zinc chloride in presence of hydrochloric
acid indicates the basicity of the molecule, i.e. of certain of
its OH groups. On the other hand, in both the cupram-
monium and thiocarbonate processes there may be a disturbance
of the oxygen-equilibrium of the molecule ; and, although there
is no evidence that the cellulose regenerated from these
solutions respectively is oxidised in the one case, or deoxidised
in the other, it is quite possible that temporary migration of
oxygen or hydrogen might be determined, and contribute to
the hydration and ultimate solution of the cellulose. But,
apart from hypotheses, we may lay stress on the fact that these
processes have the common feature of attacking the cellulose
in the two directions corresponding with those of electrolytic
strain ; and it is on many grounds probable that the connection
will prove to be causal and not merely incidental

The Typical Cellulose and the Cellulose Group 31

The thiocarbonate reaction more especially throws light on
that somewhat vague quantity, the ' reacting unit ' of cellulose.
We use this term in preference to that of molecular weight ; for
the latter quantity can be determined only for bodies which
readily assume the simplest of states, and which can be ascer-
tained by physical measurements to be in that state ; whereas
in the case of cellulose the ordinary criteria of molecular
simplicity are quite inapplicable. •;•:-,*.•»

We have formulated the synthesis of the thiocarbonate as
taking place by the interaction of C6H10O5 : 2NaOH : CS2 ;
or in approximate percentage ratio :

Cellulose : Alkali : Carbon Bisulphide = 100 : 50 : 50 ;
or, again, in terms of the constituents estimated in the analysis
of the product :

Cellulose : Alkali (Na2O) : Sulphur = 100 : 40 : 40.
If now the crude product be precipitated from aqueous
solution by alcohol or brine, and again dissolved and re-
precipitated, the ratio changes to 100 : 20 : 20 ; and, through
a succession of similar treatments, the ratio of alkali and
sulphur to cellulose continually diminishes the product, how-
ever, preserving its solubility. In fact, no definite break has
been observed in the continuous passage from the compound
as originally synthesised to the regenerated cellulose (hydrate).
It is clear, therefore, that the reacting cellulose unit is a con-
tinually aggregating molecule ; and if in the original synthesis
it appears to react as C6H,0O5, so in a thiocarbonate
containing, e.g. only 4 p.ct. Na2O, the unit is ioC6H10O5.
There being, moreover, no ascertainable break in the series,
we have no data for assigning any limiting value to the reacting
unit under these conditions. All we can say is, that the
evidence we have points to its being of indefinite magnitude ;
and we can see no a priori reason why it should not be so.

In discussing this reaction we have left out of considera-

32 Cellulose

tion the part played by the water. It may be noted that a
i p.ct. solution of cellulose (as thiocarbonate) will ' set ' to
a firm jelly of hydrate, of the volume of the containing vessel;
and that even at 0*25 p.ct. cellulose, gelatinisation of the
liquid occurs in decomposition. We have also pointed out
that a hydrate containing only 10 p.ct. cellulose is a sub-
stantial solid which gives up water with extreme slowness.

Cellulose, therefore, affords conspicuous illustrations of the
property which the ' colloids ' have, as a class, of ' fixing ' water,
and of the modes in which this property takes effect. In regard
to the causes underlying this peculiar relationship to water, we
know as yet but little. It is to be noted that the group of
colloids comprises bodies of very various chemical function,
acids, bases, salts and compounds of mixed function, as in the
complex carbohydrates and proteids ; the only possible feature
common to so varied a group would be that of molecular
arrangement, favouring the aggregation of the molecules, to-
gether with those of water, to groups of indefinite magnitude.
On this subject, however, conjectures must, for the present, do
duty for a theory which can only be shaped by further in-
vestigation.

Cellulose Benzoates. — The alkali celluloses also react
with benzoyl chloride, according to Baumann's method, to form
the corresponding benzoates.

(a) Mercerised cellulose. — This form of alkali cellulose,
treated with benzoyl chloride in the cold and in presence of
excess of alkali, gives a mixture of products, the numbers
obtained indicating that reaction occurs in the ratios,

C6H1005 : C6H5.COOH and C6H10O5 : 2C6H5COOH.

Cellulose Benzoic acid

Within the limits of Concentration, producing the specific
'mercerising' action — the lower limit being at about 12*5
p.ct. NaOH — the degree of benzoylation is inversely as the

The Typical Cellulose and the Cellulose Group 33

concentration of the alkaline solution. The fibrous benzoate
produced under these conditions shows necessarily a much
increased volume ; examined microscopically the features of
minute structure of the fibre are seen to be much accentuated.
The hygroscopic moisture of the product is 2-3 p.ct. of its
weight, i.e. from J to J that of the original cellulose. This
weakened attraction for atmospheric moisture invariably
attends the substitution of the OH groups in the celluloses
by acid residues.

(b) Soluble alkali celluloses. — The hydrates precipitated
from solution in the zinc chloride and cuprammonium solu-
tions dissolve in solutions of the alkaline hydrates ; and the
benzoates obtained from these solutions, by treatment with
benzoyl chloride, are curdy precipitates, which may be purified
by solution in glacial acetic acid, filtering, and reprecipitating
by water. Obtained in this way, the benzoates approximate

in composition to C6H8O3<Q cjj Q' Tnev melt at a mSn
temperature to a clear liquid, which solidifies to a transparent
resinous mass. By friction the compound becomes highly
electric, a property common to the esters of cellulose.

The compound dissolves in acetic anhydride ; on heat-
ing for some time at the boiling-point of the liquid, a partial
replacement of the benzoyl by acetyl groups occurs, and at
the same time a further acetylation of the cellulose.

The compound gives, on analysis, numbers corresponding
with the empirical formula,

C6H6O.O.C7H5O.(O.C2H3O)3.

This group of benzoates and mixed esters requires further
and exhaustive investigation, as a study of their composition
and constitution cannot fail to throw much light upon that of
the parent molecule.

We now leave the alkali celluloses, and the synthetical

D

34 Cellulose

products which we have seen to be obtainable from them, to
deal with those esters which are formed by direct synthesis with
acid radicals. It will become evident, as we proceed with the
description of the derivatives of cellulose, that it only yields,
with any facility, such as result from reaction of its OH groups
with mixed radicals. It will appear from the composition of
the resulting esters about to be described that the empirical
unit C6H10O5 contains at least three OH groups ; of the
remaining O atoms one certainly is carbonyl oxygen, though,
as it exists in the cellulose complex, it manifests no ' outside '
activity. It is brought into play, however, in a number of
the decompositions of cellulose, and those determined by the
non-oxidising acids are chiefly to be noted in regard to this
point. It will become evident also that the molecular changes
by which this group is set free are, in effect, decompositions,
in the sense of a breaking down of the complex. The function
of the remaining O atom is obviously a problem of moment
in regard to the question of the constitution of cellulose, and
it will appear that the solution of the problem presents con-
siderable difficulties.

Cellulose Acetates. — The acetylation of hydroxy-com-
pounds generally affords the simplest evidence as to their OH
groups, their number and disposition or arrangement, in the
molecule of the compound.

In the cellulose group, however, the problem is complicated
(i) by the difficulties of preparation and purification of the
acetates. The solutions are highly colloidal, and the ordinary
criteria of purity of the products are wanting.

(2) By the difficulties of analysis ; the direct ' saponification '
numbers often varying considerably from those obtained by
distillation of the volatile acid (after saponification), and both
numbers often failing to agree with the results of ultimate
analysis.

The Typical Cellulose and the Cellulose Group 35

(3) Some of the methods of acetylation certainly involve a
change of molecular weight, and we have no criterion of the
relation, in this respect, of the acetate to the original
cellulose.

It must be understood, therefore, that the cellulose acetates
about to be described are of undetermined molecular weight,
and afford only an empirical expression of the number of OH
groups in the undetermined unit «.(C6H10O5), which itself
may vary under acetylation.

These considerations affect the value of deductions to be drawn
from the composition of the acetates, as to the number of reactive
OH groups in the unit molecule of cellulose.

On a priori grounds we should expect a maximum of 4OH in
the unit C6H10O5. For a long time, however, the ' triacetate' was
considered to represent the highest degree of acetylation. This
acetate, however, was obtained at an elevated temperature (180°) at
which a variety of molecular complications are possible.

Acetates have been obtained by the following methods : —
(a) Interactions of cellulose and acetic anhydride. — On
boiling cotton with acetic anhydride and sodium acetate, no
reaction occurs. Heated at 180° in a sealed tube, in the
proportion by weight of i of cellulose to 6 of the anhydride,
the cellulose is converted into the triacetate (Schutzenberger).
With the reagents in the proportion of i : 2, a mixture of
lower acetates is formed. The latter are insoluble in glacial
acetic acid ; the triacetate, on the other hand, is freely soluble.
The solution is highly viscous, and passes with extreme
slowness through filter paper. Filtration, however, is greatly
facilitated by addition of benzene to the solution. The acetate
also dissolves when heated with nitrobenzene, the solution
gelatinising on cooling, even when highly dilute.

The cellulose acetates are easily saponified by dilute
solutions of the alkaline hydrates, more rapidly in presence of
alcohol (50 p.ct. vol.).

D&

36 Cellulose

(ti) Cellulose and acetic anhydride in presence of zine,
chloride. — In presence of a relatively minute proportion of
zinc chloride, cellulose and acetic anhydride react at 110-120°.
The product, according to Franchimont, is the triacetate
described above. According to later investigations, however,
the acetylation proceeds further under these conditions, the
numbers obtained indicating the formation of a tetracetate, and
sometimes still higher numbers have been obtained. The
following numbers show the quantitative relationship of the
higher acetylated derivatives of a compound CGH10O5 :

Yields on saponi-

fication
C H Acetic acid Cellulose

Triacetate . . C12H16O8 288 50-0 5-5 62-1 56-2
Tetracetate . . C14H18O9 330 50-9 5-6 727 49-1
Pentacetate. . C16H20O10 372 51-6 5-3 80-6 43-2

In dealing with the evidence as to the composition of these
products, however, we must further remember that the formula
C6H10O5 has only a statistical value, i.e. that the molecule
of cellulose is a complex aggregate ; and if the molecule is
partially resolved during acetylation, this may occur by way of
hydrolysis or addition of OH groups. It is more than
probable that, in presence of ZnCl2, the acetylation is com-
plicated in this way. It is always to be observed that, on
pouring the product of the reaction into water, a fluorescent
solution is obtained ; and, further, that the cellulose regene-
rated from acetates prepared by this method is oxidised by
cupric oxide in alkaline solution (Fehling's solution). These
reactions indicate the liberation of the CO groups of the
original cellulose during acetylation, and the reaction is not of
such simplicity that we can draw any certain conclusions from
the products as to the problematical O atom or atoms in the
unitC6H10O5.

(c) Cellulose and acetic anhydride in presence of iodine. —

The Typical Cellulose and the Cellulose Group 37

The addition of iodine in relatively small proportion (^V p.ct.)
determines the solution of cellulose in acetic anhydride at
120-130°. By this method an acetate is obtained free from
coloured by-products, and the yields of the product are remark-
ably uniform. In a series of experiments in which the propor-
tions of cellulose and anhydride were considerably varied,
the following yields were obtained per 100 parts cellulose :

176, 175, 177,174;

the yield calculated for the triacetate being 177. The uniformity
in these numbers, however, is somewhat illusory ; as in certain
cases the product is entirely soluble in acetone, and gives on
analysis the numbers calculated for a triacetate ; in other cases
the product is resolved into a soluble fraction giving high
saponification numbers (75 p.ct. acetic acid), and an insoluble
fraction giving low numbers (48 p.ct. acetic acid; calc. for
diacetate, 49). The evidence from these several processes
is somewhat conflicting, as to the composition of acetates
higher than the triacetate, and their relationship to the parent
molecule.

(d) Cellulose regenerated from solution as thiocarbonate. —
This form of cellulose has been found to react directly with
acetic anhydride, under what may be considered the normal
conditions. At 110-120° the cellulose is gradually dis-
solved, to a solution of quite extraordinary viscosity, which is
so marked that the limit of concentration is not higher than
10 p.ct. of the acetate (5 p.ct. cellulose), beyond which
point, i.e. the reaction is practically arrested. It is necessary,
therefore, to use a very much higher proportion of anhydride
to cellulose (20 : i) than in the reactions previously described.
The acetate thus obtained appears, from all its properties, to
be a true derivative of cellulose. Thus it may be prepared
in films of great tenacity and remarkable lustre ; and the

38 Cellulose

cellulose regenerated by saponification retains the film form,
shows no tendency to be further hydrolysed by the alkaline
solutions used for saponifying, and is unaffected by boiling with
alkaline cupric oxide. The analyses of this acetate show satis-
factory concordance of numbers with those calculated fora
tetracetate : /*[C6H6O.(O.C2H3O)4].

The specific gravity of this acetate is 1*210. It is soluble
in acetone, methyl alcohol, glacial acetic acid, and nitrobenzene.
It dissolves in concentrated nitric acid (as do the cellulose
acetates generally), and is precipitated on dilution, apparently
without change.

This compound is of ' critical ' value in elucidating the con-
stitution of cellulose. If the above formula be established by
further and exhaustive investigation, the cellulose ' unit ' must
be C6H6O.(OH)4 ; this is consistent with a cyclic arrangement
of the carbon nuclei, and probably a symmetrical disposition
of the OH groups. This question will be referred to subse-
quently.

Cellulose and Nitric Acid.— Cellulose Nitrates or
Nitro-Celluloses. — The action of nitric acid on starch
was investigated to some extent by Braconnot in 1833, who
found that a very rapidly burning body was produced, and
which was called xyloidine. Pelouze further investigated
this substance in 1838, and also similar bodies from paper,
linen, &c. which he held to be identical with the one
from starch. Schonbein is generally credited with the
discovery of gun cotton in 1846. It appears to have been
almost simultaneously discovered by Bottger, and also by
Otto.

Whenever cellulose, in any form, is brought into contact with
concentrated nitric acid at a low temperature a nitro-product, or
a nitrate, is formed. The extent of the nitration depends upon
the concentration of the acid ; on the time of contact of the

The Typical Cellulose and the Cellulose Group 39

cellulose with it, and on the state of the physical division of
the cellulose itself.

Knop, and also Kamarsch and Heeren, found that a mixture
of sulphuric acid and nitric acid also formed nitrates of cellu-
lose ; and still later (1847), Millon and Gaudin employed a
mixture of sulphuric acid and nitrates of soda or potash,
which they found to have the same effect.

Although gun cottons, or pyroxylines, are generally spoken of
as nitro-celluloses, they are perhaps more correctly described
as cellulose nitrates, for unlike nitro-bodies of other series,
they do not yield, or have not as yet done so, amido-bodies on
reduction with nascent hydrogen. Eder gives the following as
general properties of the cellulose nitrates: (i) when warmed
with alkaline solutions, nitric acid is removed in varying quan-
tities dependent on the strength of the alkaline solutions
employed ; (2) treatment with cold concentrated sulphuric
acid expels almost the whole of the nitric acid ; (3) on boiling
with ferrous sulphate and hydrochloric acid, the nitrogen is
expelled as nitric oxide ; the reaction is used as a method of
nitrogen estimation in the cellulose nitrates ; (4) the alkaline
sulphydrates, ferrous acetate, and many other substances
convert the nitrates into ordinary cellulose.

Several well-characterised nitrates have been formed, but it
is a very difficult matter to prepare any one in a state of purity,
and without admixture of a higher or lower nitrated body.

The following are known : —

Hexa-nitrate, C12H14O4(NO3)6,1 gun cotton. In the
formation of this body, nitric acid of 1-5 sp.gr. and sulphuric
acid of 1*84 sp.gr. are mixed, in varying proportions, about 3
of nitric to i of sulphuric; sometimes this proportion is reversed,

1 To represent the series of cellulose nitrates so as to avoid fractional
proportions the ordinary empirical formula is doubled and the nomen-
clature has reference to this double molecule.

40 Cellulose

and cotton immersed in this at a temperature not exceeding
10° C. for 24 hours : 100 parts of cellulose yield about 175 of
cellulose nitrate. The hexa-nitrate so prepared is insoluble
in alcohol, ether, or mixtures of both, in glacial acetic acid
or methyl alcohol. Acetone dissolves it very slowly. This is
the most explosive gun cotton. It ignites at 160-170° C.
According to Eder the mixtures of nitre and sulphuric acid do
not give this nitrate. Ordinary gun cotton may contain as
much as 12 p.ct of nitrates soluble in ether-alcohol. The
hexa-nitrate seems to be the only one quite insoluble in ether-
alcohol.

Penta-nitrate, C^H^C^NC^. This composition has
been very commonly ascribed to gun cotton. It is difficult, if
not impossible, to prepare it in a state of purity by the direct
action of the acid on cellulose. The best method is the one
devised by Eder, making use of the property discovered by De
Vrij, that gun cotton (hexa-nitrate) dissolves in nitric acid at
about 80° or 90° C., and is precipitated, as the penta-nitrate, by
concentrated sulphuric acid after cooling to o° C. ; after mix-
ing with a larger volume of water, and washing the precipitate
with water and then with alcohol, it is then dissolved in ether-
alcohol, and again precipitated with water, when it is obtained
pure.

This nitrate is insoluble in alcohol, but dissolves readily in
ether-alcohol, and slightly in acetic acid. Strong potash solu-
tion converts this nitrate into the di-nitrate C12H18O8(NO3)2.

The tetra- and tri-nitrates (collodion pyroxyline) are
generally formed together when cellulose is treated with a more
dilute nitric acid, and at a higher temperature, and for a much
shorter time (13-20 minutes), than in the formation of the
hexa-nitrate. It is not possible to separate them, as they are
soluble to the same extent in ether-alcohol, acetic ether, acetic
acid, or wood spirit.

The Typical Cellulose and the Cellulose Group 41

On treatment with concentrated nitric and sulphuric acids,
both the tri- and tetra-nitrates are converted into penta-nitrate
and hexa-nitrate. Potash and ammonia convert them into di-
nitrate.

Cellulose di-nitrate, Ci2H18O8(NO3)2, is formed by the
action of alkalis on the other nitrates, and also by the action
of hot dilute nitric acid on cellulose. The di-nitrate is very
soluble in alcohol-ether, acetic ether, and in absolute alcohol.
Further action of alkalis on the di-nitrate results in a complete
decomposition of the molecule, some organic acids and tarry
matters being formed. (See infra.)

The above account of the cellulose nitrates may be regarded
as representing a fair digest of the extensive literature of the
subject, so far as regards the composition and properties of the
more important and definite products. A better grasp of the
relationship of these products to one another and to the parent
molecule will be obtained from the researches of Vieille
(Compt. Rend. 95, 132), a short account of which follows.
From the title of this author's communication, ' Sur les degres
de la nitrification limites de la cellulose,' it may be concluded
that it is a study of the nitrations of cellulose (cotton) under
the condition of progressive variations, with the view of
determining the maximum fixation of the nitric group cor-
responding to such variations. The most important factor
of the process is the concentration of the nitric acid, which
was the variant investigated. The temperature was kept con-
stant— 11° C— and the nitrating acid (nitric acid only) was
employed in very large excess (100-150 times the weight of
cellulose), so as to avoid disturbance of the results by rise
of temperature or by dilution of the acid. The products
were analysed by Schloesing's method, and the analyses are
expressed in cc. NO (gas) (at o° and 760 mm.) per i grin, of
substance.

Cellulose

Analysis

Sp. gr.
of acid

Composition
(approximate)

of product
cc. NO

Properties of products

per i grm.

f

Structural features of cotton pre-

I-502
1-497

I N03H.iH20 {

2O2-I 1
I97-9 1

served ; soluble in acetic ether ;
not in ether-alcohol

1

C24H20(N03H)10010

I-496
1-492
I-490

I NO3H.iH2O |

194*4
187-3 '
1837

Appearances unchanged ; soluble in
ether-alcohol ; collodion cotton

C24H22(N03H)9Oll
C24H24(N03H)8012

Fibre still unresolved ; soluble as

1-488
I-483

} N03H.iH20 {

1657
164-6

above, but solutions more gelatinous
and thready

C24HM(N03H)A3
Dissolve cotton to viscous solution ;

I-476

)(

products precipitated by water ;

1-472

N03H.fH20

gelatinised by acetic ether ; not

1-469

(

ether-alcohol

,

C2tH28(NO:jH)An

6

f?8'fi

Friable pulp ; blued strongly by

- - . £ _.

1 — o (_)

iodine in KI solution ; insoluble in

I 4OO

1-455

I* A CO

122-7
115-9

alcoholic solvents
C24H3U.(N03H)A15

4DU

9

C24H32(N03H)4016

In regard to the time factor, or duration of exposure to the
acid required to give the maximum number, this was in all cases
controlled by observation. Thus with the acid HNO34H2O
(1*488 sp.gr.), after 48 hours the product was still blued by
iodine, and gave 161 c.c. NO ; whereas after 62 hours' ex-
posure the iodine reaction was not obtainable, and the maximum
number (165*7 c.c. NO) was obtained. At the slightly lower
gravity 1*483, an exposure of 120 hours was necessary. At the
still lower gravity when the cotton (nitrate) passes into solution,
the maximum is very rapidly attained (5 minutes).

The highest nitrate obtained as above, with nitric acid only,
is somewhat lower than when sulphuric acid is present. Under
these latter conditions the author regards the highest nitrate
obtainable as C24H18(NO3H)11O9.

THERMAL CONSTANTS. — By calorimetric observation on the

The Typical Cellulose and the Cellulose Group 43

process, it has been ascertained that the heat liberated is 11-12
cal. per unit of HNO3 reacting. This 'heat of formation'
is approximately equal to that which is observed in converting
starch into the corresponding nitrates.

HEAT OF COMBUSTION. — The total combustion of gun
cotton by free oxygen evolves heat equal to 2,300 cal. (H2O of
combustion Hquid\ or 2,177 cal- with tne H2O as gas or vapour,
per i grm. of the compound. Collodion cotton gives the corre-
sponding numbers 2,627-2,474 cal. ; gun cotton exploded in
confined space gives 1,071 cal. (H2O of combustion liquid).

PRODUCTS OF COMBUSTION of gun cotton exploded in a
closed vessel vary in relative amount and in composition
with the 'density of the charge,' or the pressure developed
at the moment of explosion. Thus the CO2 and H increase
with the density of charge ; theCH4 also, but, being present in
very small ratio (o'o-r6 p.ct. maximum), it may be neglected.
The following equations may be taken as fairly representing
the combustion of 2C24H18O9(NO3H)11, under varying con-
ditions :

Density of charge

o-oio . . . 33CO + I5CO2 + 8H2 + 2iH2O+nN2

0-023. . . 3oCO+i8CO2+iiH2+i8H2O+iiN2

0-200 . . .

0-300. . .

Under the ordinary conditions of explosion in firearms with
maximum density of charge, the quantities of gas produced
approximate more and more closely to the limit :

24CO -f- 24CO2 + i7H2 + i2H2O + nN2.

Under ' explosion,' it will be seen that no nitric oxide or
other nitrous gases are formed ; but when a slower combustion
takes place, with the products of combustion escaping freely

44 Cellulose

under a pressure nearly equal to the atmospheric — as in a ' miss
fire * — the percentage composition (by vol.) of the gases is

NO 247

CO ...» t . 41-9

C02 18-4

H 7'9

N 5-8

-CH4 1*3

100 '0

(See Karolyi, Phil. Mag. 1863, 266; also Abel, Phil. Trans.
1866, 269; 1867, 181.)

INDUSTRIAL USES OF THE CELLULOSE NITRATES. — These
products find a number of highly important uses both for de-
structive and constructive purposes. As far as these uses in-
volve, or are based upon, essential properties of the products,
they may be briefly noticed here.

EXPLOSIVES. — The products of which gun cotton— or other
nitrated celluloses— is the essential constituent are of three
main classes : (i ) containing the nitrates only ; (2) the nitrates in
admixture with inorganic salts containing oxygen ' available '
for combustion, or aromatic nitro-derivatives, &c. ; (3) the
nitrates in admixture with, or solution in, nitro-glycerin (blast-
ing gelatine, ballistite, or cordite). An account of these
modern explosives, with determinations of their constants of
explosion, will be found in a paper by Macnab and Ristori,
Proc. R. S. 1894, 56.

CELLULOID, XYLONITE, &c. — The lower nitrates are worked
up with solvents of a special character (acetone, camphor, &c.),
with or without admixture of various substances, into plastic
masses, which are then cut and moulded into articles of most
varied form and use.

COLLODION, COLLODION VARNISHES, COLLODION FILMS. —
The lower nitrates, dissolved in ether-alcohol or other solvents
(amyl acetate and benzene, &c.), form transparent solutions,

The Typical Cellulose and the Cellulose Group 45

which on evaporation leave the nitrate as a glass-clear film of
considerable elasticity and tenacity. The products, both in
solution and in the form of films, are applied in numerous
directions, chiefly in connection with photography.

It is important to observe that these nitrates preserve in a
remarkable degree the essential physical properties of the
original cellulose, which will be most obvious by comparison
of the above products with those obtainable with the cellulose
regenerated from solution as thiocarbonate. But this is still
better illustrated by the processes of converting the nitrates
into a continuous thread, available as a textile material. This
product is known as artificial silk. Various inventors have
devised means for 'spinning' solutions of cellulose nitrates
into thread, one of which may be briefly described as having
reduced the operation to one of extreme mechanical simplicity.
It is essential to the production of a thread of sufficient tensile
strength — as directly obtained — to stand the strain of the
drawing process, that the solutions employed contain a certain
minimum proportion of the dissolved nitrate. Dr. Lehner, of
Zurich,1 after investigating the various problems involved, found
that, whereas ordinary collodion containing such a proportion
of the pyroxylin (10-12 p.ct.) in solution is unworkable under
the prescribed conditions, the adding of dilute sulphuric acid
causes a molecular change, and gives the solution the requisite
fluidity. With such a solution the conversion into thread is
effected as follows : The solution, carefully filtered and free
from all bubbles, is caused to flow by way of glass tubes to a
lower level, where it is delivered through a much narrowed
opening with a steady constant flow. The shorter limb ending
in this fine orifice is contained in a glass cell filled with water.

1 See original German patent D.P. 58508/1890. The earlier pro-
cesses of De Chardonnet (1885) and du Vivier (1889) must also be men-
tioned. See D.P. 38368/1885 and 46125/1888. Also Br.Pat. 2570/1889.

46 Cellulose

On emerging, therefore, the solution is at once coagulated to a
transparent jelly, and of considerable toughness. On applying
a slight pull to the jelly, grasped with the fingers or forceps, a
thread is produced ; and on fixing the end to a light wheel re-
volving at a definite rate, the thread is drawn off continuously
of uniform diameter. Several threads being twisted together
in the usual way of ' silk-throwing,' the artificial textile thread
is produced. After being deprived of water of hydration the
threads acquire the high white lustre of ' boiled-off' silk.

In this state, however, it is the explosive nitrate, containing
11-12 p.ct. N. To fit it for consumption, therefore, the
' silk ' is ' denitrated ' by treatment with ammonium sulphide
in the cold. This process in no way affects the lustre of the
thread, and when properly carried out gives a product not more
inflammable than ordinary cotton.

The * artificial silk ' has been found to have a tensile
strength equal to 70 p.ct. of that of the natural product, of
the same degree of fineness. Its elasticity is inferior in about
the same proportion ; but it has a higher lustre and is pro-
duced at much less cost. It appears, therefore, capable of
considerable industrial use.

OTHER DECOMPOSITIONS OF THE CELLULOSE NITRATES. —
In addition to the explosive resolution into gaseous products,
of these cellulose esters, they are susceptible of a more gradual
process of decomposition, into which they pass spontaneously
under certain conditions — yielding a complex of products,
some of low molecular weight, e.g. carbonic, formic, oxalic,
saccharic, and nitroxy -acids ; others of closer relationship to
the original cellulose, gummy acid bodies which have been
described as belonging to the pectic series. Observations of
these decompositions have been made by various chemists
(Maurey, Bechamp, Kuhlmann, Pelouze, De Luca, Compt.
Rend. 28, 343; 37, 134; 42, 676; 59, 363; 59, 487;

The Typical Cellulose and the Cellulose Group 47

Divers, Journ. Chem. Soc. [2], i, 91), but have thrown but
little light on the chemistry of cellulose.

A resolution of similar character is determined by a gradu-
ated treatment of the nitrates with the alkaline hydrates (solu-
tion). This has been investigated by W. Will, from the more
theoretical point of view suggested by the title of the com-
munication containing his results, viz. ' Ueber Oxybrenz-
traubensaure, ein neues Product des Abbaues der Cellulose,'
Berl. Ber. 24, 400.

The process yielding this characteristic product, hydroxy-
pyruvic acid, consisted in treating the ether-alcoholic solution
of pyroxylin (with 11*2 p.ct. N) with a 10 p.ct. solution of
sodium hydrate, shaking the two layers of solution together
from time to time, until decomposition was complete ; or
setting aside for 24 to 30 hours, when it completes itself
at ordinary temperatures. The alkaline solution is acidified
and warmed, to complete the removal of the lower oxides
of nitrogen, and treated with phenylhydrazine in presence
of acetic acid (excess). The osazone of the ketonic acid,
COOH— CO— CH2OH, is thus obtained. The acid itself
was also directly isolated from the original alkaline solution
after neutralising, by precipitation as lead salt, and decom-
posing in the usual way with hydrogen sulphide.

The author's purpose in studying the reaction was the
elucidation of the constitution of cellulose ; and, although the
results so far are too fragmentary for the drawing of definite
conclusions, they indicate a direction in which the problem
may be successfully attacked. It is obvious that progress in
this direction must lie by way of processes of regulated dissec-
tion, and of these there are very few under sufficient control to
be available. It is therefore to be hoped that this decom-
position will be more fully investigated, especially as, from a
private communication, we learn that the characteristic product

48 Cellulose

is obtained in relatively large proportion, indicating a principal
direction of cleavage of the cellulose molecule.

Attempts to arrive at the molecular weights of these compounds,
benzoates and acetates as well as nitrates, by the method of Raoult
have, so far, led to no result. The esters of cellulose appear to
produce an abnormally large and, moreover, variable depression of
the glacial point of acetic acid — which is a general solvent of
these derivatives — such that no conclusion can be drawn from the
observed depressions, as to the molecular magnitude of these
compounds, in the undissolved condition ; and if we interpret the
depressions of freezing-point observed in the acetic acid solutions
according to the usually accepted view, we must regard the
molecules, when dissolved, as undergoing disaggregation or
dissociation. There is no a priori objection to this view, and it
appears, in fact, to be in harmony with many of the characteristics
of cellulose in reaction, viz. those in which it resembles, to a certain
extent, the inorganic salts.

Cellulose and Sulphuric Acid.— Cotton cellulose is
rapidly attacked and dissolved by concentrated sulphuric acid.
The initial product may, perhaps, be regarded as a cellulose
sulphuric acid, but a rapid molecular disintegration ensues,
and there results a series of sulphates of the general formula
C6MH10wO5w_Ar(SO4);t. The resolution of the cellulose molecule
is a progressive phenomenon, and is accompanied by increase
of dextro-rotation and reducing power (CuO) in the product.

The free acids are amorphous bodies, very hygroscopic,
soluble in alcohol and water ; on boiling the aqueous solution
they are completely hydrolysed to glucose and sulphuric acid.
The Ca, Ba, and Pb salts of the acids are obtained by neu-
tralising with the respective oxides in aqueous solution and
precipitating by alcohol. On boiling with water these salts
lose one-half their sulphuric acid according to the equation

)f + fH2S04.
(Honig and Schubert, Monatsh. 6, 708 ; 7, 455.)

The Typical Cellulose and the Cellulose Group 49

The reaction may be described, therefore, as a progressive
hydrolysis of the cellulose through a series of dextrins, to a
carbohydrate of minimum molecular weight. This transform-
ation of cellulose to a sugar was established early in the cen-
tury (Braconnot, 1819). Recent investigation has established
the identity of this sugar with dextrose (rotation, [o] = 53'o).
The process of hydrolysis consists in the following stages : the
cellulose (50 grms.) is dissolved in strong sulphuric acid (250
gr. H2SO4 + 84 gr. H2O) in the cold, and the solution allowed
to stand ; diluted till the acidity equals 2 p.ct. H2SO4, and
boiled 3 hours. The isolation of the dextrose in the crystal-
line form is accomplished in the usual way. (Flechsig, Zeitschr.
f. Physiol. Chem. 7, 523.)

The reaction between cotton cellulose — as well as other cellu-
loses of the cotton group (p. 79) — and sulphuric acid is, in regard
to ultimate products, of the simplest character, resulting in their
conversion into dextrose, and in quantitative proportion (Flechsig).

The reaction in the case of other celluloses, e.g. wood cellulose,
is more complicated. The initial solution in the concentrated acid
is dark coloured, and on diluting and boiling there is a consider-
able formation of insoluble products. Although dextrose is
obtained as one of the ultimate products of the hydrolysis, it is
only in relatively small quantity (Lindsey and Tollens, Annalen,
267, 371), and appears to be accompanied by other carbohydrates.
It will be shown subsequently that the celluloses of this group are
oxy celluloses, containing reactive CO groups and very readily con-
densing to furfural. The hydrolysis in such cases would, no doubt,
be attended by condensations and other complications.

The subject has been recently and more exhaustively in-
vestigated by A. L. Stern (Thesis for D.Sc. Lond. Univ. 1894),
and we give a short extract of the results which he obtained.

( i ) Composition of body produced by dissolving cellulose in sul-
phuric acid. — In addition to the determination of the empirical
ratios of the constituents in the products isolated as Ba salts,
they were examined in solution for optical rotation and reduction

5O Cellulose

of cupric oxide. The former are expressed in terms of (<F), and
the latter in terms of dextrose reduction K=ioo (o-4535 grm.
dextrose equivalent to i grm. CuO). (a) Solution of cellu-
lose at 5°. (b) Solution at 15°. In both cases the compound
obtained was C6H8O3(SO4)2Ba ; the compounds were with-
out action on Fehling's solution ; the ' opticity ' varied directly
with the temperature of the solution-reaction — viz. for (a) it
was + 24° ; for (b) + 54°. The yield of soluble Ba salt was
48 p.ct. of the theoretical; the residue of the cellulose remains
associated with the BaSO4, obtained on neutralising the acid
liquid with BaCCV

Hydrolysis of the product. — The compounds in solution were
treated for 30 minutes at 100° in presence of free sulphuric
acid (2 p.ct. on the solution). The acid products were
isolated in each case as Ba salts. Compounds of identical
formula were obtained — viz. C18H28O13(SO4)2Ba. The yield
amounted to (a) 95 p.ct., (b) 80 p.ct. of the total calcu-
lated. The remainder was converted into dextrose. The
opticities of the products were different, viz. (a) for Ba salt
+ 25° ; (b) for Ba salt, + 75°. The CuO reductions were, for (a),
K=23-3; for(^), K=i8-i.

(2) Graduated hydrolysis of the disulphuric ester. — The
initial product of the empirical composition C6H10O3(SO4)2
was then subjected to hydrolysis in graduated stages, the con-
ditions being as before, and the stages being defined by the
duration of the hydrolysis, the products being exhaustively
investigated at periods of 7, 15, 20, and 30 minutes. The results
are summarised as follows :

C6H8O3(SO4H)2, when boiled with 2 p.ct. H2SO4, yields
successively,

C6H803(S04H)2 C6H904.S04H
C6H803(S04H)2 3C6H904.S04.H.
1 This residue should be investigated.

The Typical Cellulose and the Cellulose Group 51

No sugar (dextrose) is formed down to this stage, the result
indicating a loss of sulphuric acid, and the formation of the
monosulphuric ester, C6H9O4.SO4H, as the limit. This pro-
duct, as the original disulphuric ester, is without action on
Fehling's solution.

Subsequently the following were obtained as products of
the further hydrolysis.

5C6H904S04H C12H19O9SO4H
2C6H9O4SO4H C12HI9O9SO4H.

This stage indicates further resolution of the monosulphuric
ester ; this takes place rapidly and is difficult to control.

The hydrolysis was then proceeded with for longer periods,
30-120 minutes. The results are thus summarised : sugar is
formed, and acid products, with increasingly less barium and
sulphuric acid. The sugar formed is dextrose. The following
products were investigated :

6H2S04.C12H1809 ioH.SO4C6H9O4.2HSO4Cl2H19O9

H2SO4C12H18O9 HSO4C12H19O9

H2S04C12H18O9 4C12H1909S04H

H2S04C12H1809 4C12H1909S04H.2C6H1206.

The corresponding Ba salts of these products c'ontain more
Ba than is necessary to saturate the acid group SO4H. It is
highly probable that one of the OH groups of the resolved
cellulose molecule acquires acid functions.

The series of degradation products of the cellulose molecule
are termed cellulose-sulphuric acids by the author ; but this
designation is misleading.

The most important features of this careful study of the
molecular dissection of cellulose are, (i) the fact that the
molecule can be very considerably resolved without freeing
the aldehydic CO groups ; (2) the differentiation of two of
the OH groups of the C6 unit, as having a superior basic or

E2

5 2 Cellulose

alcoholic function ; and (3) that with the breaking down of the
molecule, OH groups of the cellulose units are brought into
play with acid functions.

It will be noted that up to this point we have been dealing
with compounds of cellulose products obtained by synthetical
reactions with acid and basic groups and with salts ; in all of
which the reacting molecule is maintained at or near its
maximum weight (magnitude). We have mentioned incident-
ally, on the other hand, that the cellulose molecule, in the sense
of the reacting unit, is a variable quantity ; and that, while
under certain conditions the tendencies are towards aggregation
(thiocarbonate reaction), under others the tendency is towards a
progressive disintegration. This is notably the case in the reac-
tion with sulphuric acid just described, in which there is a per-
fectly graduated transition from the complex colloid molecule
to the simple dextrose unit, a crystallisable solid of low molecular
weight. These considerations lead up to the study of the

Decompositions of cellulose, which we shall find group
themselves under the headings —

(a) Decompositions determined by the non-oxidising acids
— the changes resulting from addition or subtraction of H2O.

(b) Decomposition by oxidants, with attendant or second-
ary effects of hydrolysis and condensation.

(c) Decompositions by ferments ; (d) by heat.

None of these decompositions of cellulose are of a simple
character. Any aggregate change of composition can, of course,
always be determined ; as, however, we have no knowledge of
the molecular magnitude and configuration, either of the parent
molecule or of its derivatives — i.e. such as preserve the general
characteristics of the celluloses — we are limited to the statistical
study of these reactions, together with general inferences based
upon their particular character.

The Typical Cellulose and the Cellulose Group 53

(a) NON-OXIDISING ACIDS. — (i) Sulphuric arid, of i'5-i*6
sp.gr. (H2SO4.3H2O), produces the effects previously de-
scribed, but in such a way as to be controlled within the
earlier stages of molecular resolution. Unsized paper plunged
into the cold acid, diluted to the above formula, is rapidly
attacked, the paper becoming transparent owing to the swell-
ing and gelatinisation of the fibres. The reaction quickly
becomes one of solution ; but if the paper be transferred,
after short exposure, to water, the acid compound is at once
decomposed and the resulting gelatinous hydrate of cellulose
precipitated in situ. The product, after exhaustive washing
and drying, is obtained as parchment paper. This modification
of cellulose gives a tough translucent sheet, necessarily very
much less absorbent than the original.1

The compound itself, from its resemblance to starch,
has been termed amyloid. It is represented by the formula
«(C12H22O11), the semi -hydrate of «(C6H10O5). Like
starch, the compound is coloured blue by iodine, and the joint
action of iodine and sulphuric acid is frequently used in dia-
gnosing cellulose. As a further result of the reaction, the
product differs from cellulose, in containing active CO groups ;
it reacts with phenyl hydrazine salts, and is oxidised by CuO in
alkaline solution.

Effects of a similar character are produced by treating
cellulose with concentrated solutions of phosphoric acid and
zinc chloride.

(2) Nitric acid, of 1-4 sp.gr., also produces (without oxi-
dation) an effect of a similar character. A short immersion
of unsized paper — e.g. filter paper — in the acid, followed by
copious washing, has a considerable toughening action, due to
superficial conversion of the fibres into a gelatinous hydrate.

1 See Guignet on ' Soluble and Insoluble Colloidal Cellulose, and Com-
position of Parchment Paper,' Compt. Rend. 108, 1258.

54 Cellulose

These changes are marked by a shrinkage in linear dimen-
sions of about TVth : the tensile strength of the paper thus
treated is about ten times that of the original. (J. Chem. Soc.
47, 183.)

(3) Hydrochloric acid, both in the form of gas and con-
centrated aqueous solution, rapidly disintegrates cellulose
tissues. The product, obtained from cotton, after washing and
drying, is a white powder which under the microscope is seen
to consist of angular fragments of the original fibres. It has
been termed hydrocellulose by Girard, who first described this
product, and hydracellulose (Witz), the latter term being, per-
haps, preferable. According to Girard, the product may be
represented as ^C^H^On)— as having, i.e. the same em-
pirical composition as the above-described amyloid. From
cellulose it also differs similarly to the latter, in the presence of
free CO groups and the greater reactivity of its OH groups.
Thus it reacts with acetic anhydride at its boiling point ; the
acetylation, however, does not proceed very far.

The product is dissolved by nitric acid (1*5 sp.gr.) without
oxidation, and from the solution a series of nitrates are obtained :
(i) the lowest nitrates, by spontaneous evaporation of the solu-
tion in their fibres ; (2) higher nitrates, by precipitation with
water ; and (3) the highest nitrates, by precipitation with sul-
phuric acid.

From these properties it may be concluded that, in general
molecular configuration, these derivatives are similar to cellulose,
but are so modified that the typical reactions take place under
much less extreme conditions.

The action of this acid we should expect to be one of
dehydration ; and, although the final product has the composition
of a hydrate, there is every reason to regard the hydration as
a secondary result, following the molecular rearrangement
caused by the initial dehydration.

The Typical Cellulose and the Cellulose Group 55

Although, therefore, the products resulting from the action
of hydrochloric and sulphuric acids (1*55 sp.gr.) are identical
in empirical composition, they are the very opposite in physical
characteristics, and the actions of these acids certainly take
very different courses.

It should be noted that the action of sulphuric acid at
greater dilution (1-3 sp.gr.) approximates closely to that of
hydrochloric acid, the product being a disintegrated and
friable mass of the hydracellulose.

The non-oxidising acids generally produce similar results, the
degree of action being proportionate to their hydrolytic activity.

A curious practical application of these processes of disin-
tegrating cellulosic tissues may be noted in evidence of the
fundamental chemical distinction of the vegetable (cellulose) fibres
from those of animal origin (silk and wool). The latter are very
resistant to the action of acids. From a wool-cotton fabric, there-
fore, the cotton is easily separated by soaking the fabric in dilute
sulphuric acid, and, after removing the excess of acid, drying down
on a hot floor. The disintegrated cellulose is then completely
removed by dusting out, leaving the wool unaffected. A similar
result is obtained with hydrochloric acid ; or by treatment with
certain chlorides which are dissociated, on heating, into hydrochloric
acid and basic oxide — e.g. aluminium chloride or chlorhydrate.

On the other hand, the animal fibre-substances are extremely
sensitive to the action of alkalis, to which, as we have seen, the
celluloses are very resistant. The student should compare the
constitution of the substances in question, so far as they have been
elucidated, with that of cellulose, and for that purpose should read
Berl. Ber. 1886, 850 ; J. Soc. Chem. Ind. 12, 426.

This activity being conspicuously feeble in the case of acetic
acid, this acid has but little action upon cellulose, and therefore
finds extensive use in the printing of cotton and linen fabrics.

Solutions of the mineral acids are extensively used in the
'souring' operations of the bleacher and dyer. They are
usually employed cold, and the operation of souring is always

56 Cellulose

followed by copious washing. Failure to remove the acid,
even the last traces, results in disintegration or ' tendering ' of
the fabric on drying.

(ff) Decompositions of Cellulose by Oxidants.— It
has been already pointed out that cellulose is comparatively
resistant to the action of oxidants ; that most of the processes
for isolating or purifying (bleaching) cellulose depend, per
contra, upon the use of oxidising agents, which readily attack
the 'impurities' with which it is combined or mixed in raw
fibrous materials. The cellulose resists the action of these
oxidising agents, and, further, withstands in a high degree the
action of atmospheric oxygen. It is this general inertness of
the compound which marks it out for the unique part which it
plays in the vegetable world and in the arts.

It must be again noted that this high degree of resistance to
hydrolysis (alkaline) and oxidation belongs only to cotton cellulose
and to the group of which it is the type, and which includes the
celluloses of flax, rhea, and hemp. A large number of celluloses,
on the other hand, are distinguished by considerable reactivity,
due to the presence of ' free ' CO groups, and are therefore more
or less easily hydrolysed and oxidised. The ' celluloses ' of the
cereal straws and esparto grass are of this type, and hence the
relative inferiority of papers into the composition of which they
enter. (J. Chem. Soc. 1894,472.)

On the other hand, we have now to study those processes
of oxidation to which it yields more or less readily.

A. OXIDATION IN ACID SOLUTIONS. — (i) Nitric acid
(i -1-1*3 sP-§r-) attacks cellulose at 80-100°, at first slowly,
then more rapidly, but tending to a limit at which the action
again becomes very slow. This limit corresponds with the
formation of a characteristic product of oxidation — oxy cellulose.
This substance, which is white and flocculent, when thrown
upon a filter and washed with water, combines with the latter to
form a gelatinous hydrate. It requires, therefore, to be rapidly

TJie Typical Cellulose and the Cellulose Group 57

washed with dilute alcohol. It amounts to about 30 p.ct. of
the cellulose acted upon, the remainder being for the most
part completely oxidised to carbonic and oxalic acids. On
ultimate analysis it gives the following numbers:

C 43'4\r w o
H ,.••$/ 18 26 16'

It dissolves in a mixture of nitric and sulphuric acids, and
on pouring into water, the nitrate C18H23O13(NO3)3 separates
as a white flocculent precipitate. From the low number of
OH groups reacting with the nitric acid, it may be concluded
that the compound is both a condensed as well as an oxidised
derivative of cellulose. This oxycellulose dissolves in dilute
solutions of the alkalis, and on heating the solutions they
develop a strong yellow colour. Warmed with concentrated
sulphuric acid it develops a pink colouration similar to that of
mucic acid. The compound exhibits generally a close re-
semblance to the pectic group of colloid carbohydrates.

The by-products of this oxidation are carbonic and oxalic acids,
together with the lower nitrogen oxides. The solution, examined
at any stage, appears to contain traces only of intermediate pro-
ducts of oxidation of the cellulose. The reaction is divisible into the
two stages: (i) the conversion of the cellulose into hydracellulose,
evidenced by its breaking down to a fine flocculent powder ; and
(2) the oxidation of the hydracellulose.

The oxycelluloses resulting from this process differ from those
formed by the action of CrO3 (infra), in giving small yields only of
furfural (2-3 p.ct.) on boiling with HClAq (ro6 sp.gr.). It is also to
be noted that the carbon is higher than that of the oxycelluloses,
giving large yields of furfural (p. 84). These points suggest that,
side by side with oxidation, combination of the negative oxy-groups
with the more basic groups of unattacked molecules takes place,
giving derivatives of the nature of esters. And, indeed, the re-
action may be even more complicated. It is clear, from the com-
position of the nitrate, that the proportion of basic OH groups is
reduced to a minimum.

58 Cellulose

The reaction requires further systematic research in the light of
our increased knowledge of the constitution of the simpler carbo-
hydrates and the simple products of their oxidation.

(2) Chromic arid, in dilute solutions, attacks cellulose with
extreme slowness ; in presence of mineral acids oxidation
proceeds more rapidly, but at ordinary temperatures is still
very slow. The action is, therefore, easily controlled within any
desired limit, the oxidation being in this case of course directly
proportionate to the amount of CrO3 presented to the fibre.
The oxidation is accompanied by disintegration, and the
insoluble product is an oxidised cellulose, or oxycellulose, the
yield and composition of which bear a simple relation to the
amount of oxidation to which the cellulose is subjected. Its
properties are similar to those of the oxycellulose above
described. It dissolves in a diluted mixture of sulphuric and
hydrochloric acids (57 p.ct. H2SO4, 5-5 p.ct. HC1), and on
diluting and distilling with HC1 of 1*06 sp.gr., is decomposed
with formation of furfural, C4H3O.COH, the yield of this alde-
hyde being proportionate to the state of oxidation of the
product.

This is illustrated by the subjoined results of observations :

Weight of CrO3 employed Yield of Yield of furfural

cellulose oxycellulose p.ct. of oxycellulose

47 1-5 93-o 4'i

47 3'0 87-o 6-3

47 4'5 82-3 8-2

(Berl. Ber. 26, 2520.)

The first effect of treatment with CrO3 appears to be that of
simple combination ; reduction to the Cr2O4 then ensues, and the
further deoxidation requires the presence of a hydrolysing acid.

From the statistics of the reaction it appears there is little ' de-
struction ' of the cellulose ; and, as the oxidation is not attended
by evolution of gas (CO2), we may assume that the reaction
consists simply in oxidation with the fixation of water. A certain

The Typical Cellulose and the Cellulose Group 59

proportion of the products are dissolved by the acid solution, and
of the insoluble residue (oxycellulose) a large proportion is easily
attacked and dissolved by alkaline solutions. The product is no
doubt, therefore, a mixture ; and, indeed, it would be hardly conceiv-
able that an aggregate like cellulose should be equally and simul •
taneously attacked.

The reaction is so perfectly under control that it must be
regarded as giving a regulated dissection of the molecule of cellu-
lose, and therefore is an especially attractive subject for exhaustive
investigation.

The carbohydrates of low molecular weight are similarly
oxidised by chromic acid, and the product of oxidation
similarly resolved with formation of furfural.

It is to be noted with cellulose, as with the carbohydrates
of low molecular weight, that by oxidation its equilibrium is
disturbed in such a way that carbon condensation is easily
determined. This fact is of physiological significance and will
be referred to subsequently.

(3) Of other acid oxidations which have not been particu-
larly investigated we may mention the action of Cl gas in
presence of water, of hypochlorous acid, and of the lower
oxides of nitrogen in presence of water. Generally the result of
these treatments is similar : the formation of insoluble products
having the properties of the oxycelluloses above described, and
soluble products which are oxidised derivatives of carbohydrates
of low molecular weight. These, however, are usually obtained
in relatively small quantity.

Atmospheric oxidation of cellulose — if it could be proved to take
place — would fall in this category, as cellulose surfaces undei
ordinary conditions of exposure would be found to be normally acid.
From the evidence we have of the condition of paper and textiles of
the flax group after centuries of exposure to ordinary atmospheric
influences, we may conclude that the oxidation of the normal cellu-
loses under these conditions is excessively slight.

60 Cellulose

B. OXIDATIONS IN ALKALINE SOLUTION. — (i) Hypo-
chloriteS) in dilute solution ( < i p.ct.) and at ordinary tem-
peratures, have only a slight action upon cellulose ; a fact of
the highest technical importance, since hypochlorite of lime
(bleaching powder) is the cheapest of all soluble oxidising
compounds, and the most effective oxidant of the coloured
impurities which are present in the raw cellulose fibres or
formed as products of alkaline hydrolysis.

While the normal celluloses withstand these bleaching oxida-
tions, there are many celluloses widely differentiated from the
cotton type which are eminently oxidisable, and, at the same
time, susceptible of hydrolysis. The 'celluloses' of esparto and
straw are of this kind (see p. 84), and the economic bleaching of
paper pulps prepared from these raw materials can hardly be
expected to follow upon the same lines as that of 'rag' pulp
(cotton and linen). A study of the factors involved in the pro-
cess will be found in a paper entitled ' Some Considerations in the
Chemistry of Hypochlorite Bleaching' (J. Soc. Chem. Ind. 1890).
These factors are — in addition to temperature and concentration
(C12O) of the bleaching solution — the nature of the base in union
with the hypochlorous acid, and its proportion to the acid. A
knowledge of the operation of these factors will enable the bleacher
to control a process which is usually carried out on an entirely
empirical basis.

The resistance of cellulose to the action of these solutions
necessarily has its limits, and when these are exceeded, the
fibre-substance is oxidised and disintegrated, and an oxy-
cellulose results. These effects are rapidly produced by the
joint action of hypochlorite solutions and carbonic acid. The
oxycellulose formed in this way acquiring the property of selec-
tive attraction for certain colouring matters — notably the basic
coal tar dyes — its presence in bleached cloth is easily detected
by a simple dyeing treatment consisting in immersing the
oxidised fabric in a dilute solution (0*5-2 "o p.ct.) of one of
these dye stuffs, e.g. methylene blue. Local over-oxidation

The Typical Cellulose and the Cellulose Group 61

may be diagnosed in this way with certainty, and bleachers'
damages may be thus ascertained and often traced back to
the operating cause in the light of this ' oxycellulose ' test.
(J. Soc. Chem. Ind. 1884.)

The oxycellulose or disintegrated fibre resulting from this
process of oxidation differs but little in empirical composition
from cellulose itself, probably owing to the fact that the more
highly oxidised products are dissolved in the solution of the
oxidant, which is, of course, basic. Its reactions indicate
the presence of free CO groups, and it readily undergoes
further oxidation by atmospheric oxygen, the oxidation being
much accelerated by temperatures over 60°. The OH groups
of this oxycellulose are also more reactive than those of the
original cellulose, acetylated derivatives being obtained by
boiling the product with acetic anhydride.

The facts in relation to the conversion of cotton cellulose into
oxycellulose by the action of bleaching powder were first made
known by George Witz in 1883 (Bull. Soc. Ind. Rouen, 10, 416 ;
11, 169).

Since when a number of papers have been published dealing
with special aspects of the phenomena— theoretical and practical.
Of these we may cite : Schmidt, Dingl. J. 250, 271 ; Franchimont,
Rec. Trav. Chim. 1883, 241 ; Nolting and Rosenstiehl, Bull.
Rouen, 1883, 170, 239 ; Nastjukow, Bull. Mulhouse, 1892, 493.

It is probable on many grounds that the oxidised products
obtained from cellulose by the action of the hypochlorites in
the manner described are mixtures of one or more oxycelluloses
with residues of unoxidised cellulose. More recent investiga-
tion has led to the conclusion that the extreme product of
oxidation is an oxycellulose of the empirical formula C6H10O6,
which is freely soluble in dilute alkaline solutions in the cold ;
and that cellulose oxidised by hypochlorite solutions is a
variable mixture of this product, with hydracellulose, and un-
altered cellulose. (Nastjukow.)

62 Cellulose

By drastic oxidation of cellulose by the oxyhalogen com-
pounds — i.e. by treatment with chlorine or bromine in presence
of alkaline hydrates — the molecule is entirely broken down to
the simplest products. With bromine, i.e. hypobromite, some
quantity of bromoform is obtained ; carbon tetrabromide is
also easily obtained and identified. (Collie, J. Chem. Soc.
65, 262.)

(2) Permanganates. — The permanganates in neutral solu-
tion attack cellulose but slowly, and they may therefore be
usefully employed as bleaching agents. In presence of alkalis
a more drastic oxidation is determined. The degree of oxida-
tion is, of course, dependent upon the conditions of treatment.
The following general account of a particular experiment and
its results will illustrate its main features.

22 "6 grms. cellulose, with 400 c.c. caustic soda solution ;
50 grms. KMnO4 added in successive small portions ; tempera-
ture, 40-50°. Proportion of cellulose to oxidising oxygen,
2C6H1005 : 70.

The main products were :

(o) Oxycellulose . . . IO'5 grms., approximately 50 p.ct,
()8) Oxidised carbohydrates in \ A

solution . . ./ 3'5 " " l6 "

(7) Oxalic acid . . 4-3 „ „ 2O „

(5) Carbonic acid, water, and "I

traces of volatile acids . /

(a) The oxycellulose gelatinised on washing, and was
similar to the product obtained by the action of nitric acid.

(/3) The oxidised carbohydrate in solution resembled
' caramel ' in appearance. The compound or mixture was
precipitated by basic lead acetate, and isolated by decomposing
the precipitate with hydrogen sulphide, filtering and evaporating.
On distillation from hydrochloric acid, furfural was obtained
in large proportion.

(3) Extreme action of alkaline hydratei. — "When fused at

The Typical Cellulose and the Cellulose Group 63

200-300° with two to three times its weight of sodium or potas-
sium hydrates, cellulose is entirely resolved, the characteristic
products being hydrogen gas, and acetic (20-30 p.ct.) and
oxalic acids (30-50 p.ct). Generally the reaction takes
the same course as with the simpler carbohydrates, resolution
of the cellulose into molecules of similar constitution no doubt
preceding the final resolution, which appears to be an
exothermic or explosive reaction.

Distinguished from the two groups of decomposition which
we have now considered — viz. those determined (i) by hydro-
lytic agents, (2) by oxidising agents (under hydrolysing con-
ditions)— are those of a more intrinsic character, determined
rather by the addition or withdrawal of energy, than by reaction
with outside molecules.

C. RESOLUTION BY FERMENT ACTIONS. — This group of
decompositions of cellulose is necessarily a very wide one. In
the ' natural ' world of living organisms, of course, no structures
are permanent ; and although cellulose distinguishes itself by
relative permanence and resistance to the disintegrating actions
of water and oxygen, the differentiation in this respect is only a
question of degree, and all cellulosic structures are subject to
the law or necessity of redistribution.

The directions of redistribution are chiefly three . viz. (i) In
the assimilating processes of the plant a cellulosic structure is
broken down, reabsorbed into the supply of plastic nutrient
material, and re-elaborated,

(2) Structures which have ceased to play a part in the
general organisation of the plant are cast off and then exposed
as ' dead ' matter to the play of the redistributing agencies of
the natural world. The processes of * decay ' take various
forms according to the conditions to which they are exposed.
The humus of soils, peat, lignite, and all forms of coal present
various forms of the residual solid products of the decay of

64 » Cellulose

cellulosic structures, the remainder having been dissipated
and restored to the general fund of matter in circulation, in the
gaseous form — viz. as CO2 and CH4.

(3) In the processes of animal nutrition plants and vegetable
substances are, of course, most important factors. In the course
of animal digestion the vegetable substances are attacked by
the fluids of the alimentary tract and resolved into proximate
constituents fulfilling the requirements of the organs of assimi-
lation ; and in addition to these decompositions, which are
largely hydrolytic in character, more fundamental resolutions
are observed in which the carbohydrate molecules are com-
pletely broken down, i.e. with formation of gaseous products.

Processes of the first of these three groups are well known
to plant physiologists ; tissues of a cellulosic character are
specialised to serve as reserves of nutrient material, or reserve
material is stored up within a cellulosic cell wall which requires
to be broken down in order that the supply may be liberated.
In the seeds of the Gramineae, more especially the barleycorn,
this process of reabsorption of a cellulosic tissue has been care-
fully studied, and there is no doubt that the process of breaking
down the cellulose is due to the action of a special ferment — a
cellulose-dissolving enzyme. It must be noted here that the
celluloses susceptible of this simple form of hydrolysis are ol
very different constitution from the typical cotton cellulose,
and the features of this differentiation will be discussed subse-
quently. Taking cellulose, however, as a general expression
for the colloid carbohydrates, we may regard them as having
the property of yielding to the hydrolytic activity of special
enzymes.

As the student is now considering ferment actions he may take,
in illustration of these general views, the alcoholic fermentation of
dextrose. The main products of this decomposition — alcohol and
carbonic acid — are so related that the decomposition may be ex-
plained as resulting from migration of oxygen in the one, and of

The Typical Cellulose and the Cellulose Group 65

hydrogen in the other and opposite direction within the molecule :
a decomposition, therefore, of the electrolytic type. Nothing is
known of the intermediate stages of the resolution of the C6H12O6
into 2 [C2H6OH + CO2] ; the decomposition is rather of an ex-
plosive character, and we have so far no means of investigating its
mechanism.

The sugars are, of course, not ' organised ' as such into cellular
tissue, but are built up into aggregates of specialised constitution.
The reabsorption of such aggregates into the general circulation
of nutrient material of the plant, as indicated under (i), is the re-
sult of proximate resolution of these aggregates. It must be borne
in mind here that changes of this order have been brought to
light by physiological and histological methods ; and with very
little regard to the chemistry of the changes or, indeed, the actual
composition of the tissue substance. Later investigations are
differentiating these tissue-substances altogether from the celluloses
of the cotton type, and in reading this section the student may be
reminded that in the classification of the celluloses (which
follows later, p. 85), it will be shown that so-called ' celluloses '
susceptible of hydrolytic degradation are of an inferior order of
molecular aggregation — probably rather resembling that of the
starch-dextrin series.

We may note here more particularly an important paper by
Brown and Morris (J. Chem. Soc. 1890, 57, 458) upon the
' Germination of some of the Gramineae.' It is generally known
that the cell wall of endosperm cells containing nutrient substances, to
be supplied to the embryo in its earliest stages of growth, are broken
down, as a preliminary to the appropriation of the cell contents.
The general mechanism of the process has been elucidated by
the above observers, even to the localisation of the cellulose-
dissolving enzyme (cyto-hydrolyst). This enzyme does not exist
in the resting seed but is formed in the process of germination.
From a cold water extract of an air-dried malt the enzyme is pre-
cipitated by alcohol. An extended investigation of its activity
showed that it rapidly disintegrates the parenchymatous tissue of
the potato, carrot, turnip, apple, beet, &c. The elegant methods
of experiment pursued by the above authors are typical of chemico-
biological work, and should be thoroughly mastered by the
student.

66 Cellulose

In the second group of resolutions, constituting ' decay,1
various micro-organisms play an important part. The extreme
resolution takes place according to the equation

C6H1005 + H2O = sC02 + sCH4

(Hoppe-Seyler, Ztschr. Biol. 10, 401). This decomposition
is determined by the amylobacterium, and may be taken as
typical. Pure 'fermentations' of cellulose have, however,
been but little investigated.

In the decay of plant structures we have to deal with a
complex of compounds and with celluloses of very various
character. Again, therefore, we can only treat of these
processes in their broad and general features. These are, in
the main, (i) complete resolution, of the kind described and
formulated above ; (2) a tendency in the precisely opposite
direction, i.e. towards condensation of the carbon nuclei to
still more complicated forms, accompanied by the splitting off
of water. These processes are concurrent as they are in the
decompositions by heat, about to be described. As visible and
tangible results of this tendency to carbon accumulation we
have the vast aggregations of peat, lignite, and coal in all its
forms in the earth's crust, which are the residues of the flora
of a past geological age. In the coal measures, moreover,
there is abundance of gaseous carbon compounds also stored
up. These being chiefly marsh gas and carbonic acid, the
process of coal formation, in its earlier stages, appears to have
been similar in all respects to those which we can observe
around us as attending the decay of vegetable matter in the
mass.

These decompositions are necessarily of a complex character,
and are, no doubt, largely dependent upon the presence of nitro-
genous substances. We may cite in illustration the disintegration
of leaf parenchyma in the well-known process of « skeletonising.'
Leaves of the poplar, pear, &c. are covered with water and set aside

The Typical Cellulose and the Cellulose Group 67

in a warm place (35-45°). In the course of a week or so the paren-
chymatous tissue is so far broken down and gelatinised that it is
easily detached from the ' skeleton tissue ' constituting the vena-
tion of the leaf. This is a cellulose, or rather lignocellulose (see
p. 92), of the more resistant type ; and the process affords a simple
means of differentiating the cellulose group in regard to resist-
ance to hydrolytic agencies.

The 'rot-steep' or retting of flax is another important illus-
tration. The separation of the bast fibres of the plant from the
cuticular tissues on the one side, and the woody stem on the other,
is greatly facilitated by the breaking down of the parenchymatous
tissue with which the bast cells are in contact ; it is this tissue
which rapidly ' rots ' under the treatment, and the process is
another illustration of ' natural ' differentiation of cellulosic tissues.

The third group of decompositions involves the much
debated question of the 'fate ' of cellulosic tissues in their passage
through the alimentary canals of animals, or, to put it more
narrowly, the feeding or nutritive value of cellulose. We may
take it that the typical cotton cellulose would not be sensibly
affected by a passage through the most powerful processes of
animal digestion. There are, on the other hand, a number
of celluloses which would be, and undoubtedly are, readily
digested ; and the further consideration of this point may be
deferred until we have dealt with the specific differences
exhibited by the various members of the cellulose group, more
particularly in relation to acid and alkaline hydrolyses, under
which groups of decompositions the digestive processes maybe
generally included.

Special mention may be made of the results of an investiga-
tion, by Horace Brown (J. Chem. Soc.), of the question of the
presence of a cellulose-dissolving enzyme in the digestive tract of
herbivora. After exhaustive inquiry the author establishes the
conclusion that the enzyme is secreted by the plants themselves,
and comes into activity under the favourable conditions provided
by the digestive organs and processes of the animal.

F 2

68 Cellulose

D. DECOMPOSITION BY HEAT. — DESTRUCTIVE DISTILLA
TION. — Destructive distillations of cellulosic raw materials con-
stitute an extremely important group of industrial processes ; and
if we include the coals in such a classification, the hydrocarbons
of coal tar must be regarded as products of a series of trans-
formations of cellulose, of which the final stages are determined
by destructive distillation. By the direct action of heat,
however, upon the celluloses proper, 'aromatic' products —
hydrocarbons and phenols — are obtained in relatively small
quantity. The main products are (i) gases : carbonic
anhydride, carbonic oxide, and methane ; (2) liquids : water,
acetic acid and furfural, methyl alcohol, and small quantities of
hydrocarbons and phenols ; (3) solids : paraffins and aromatic
hydrocarbons in small quantity, and the residual charcoal.

The proportions of these products necessarily vary with all
the conditions of the distillation, chiefly (i) rapidity of heating
and (2) maximum temperature attained. Recent investigations,
in which these conditions were carefully regulated and the
products of distillation estimated, have led to more definite
results than those of previous date. It must be borne in mind
that the term Cellulose has been used in a somewhat loose way,
and by some writers or compilers of articles as synonymous
with the cellulosic raw materials generally. * The destructive
distillation of cellulose ' has in consequence been described as
including the woods. ' It is important now to differentiate
between the products obtained from the typical cotton cellulose
and compound celluloses, such as the woods. These differences
will be noted more particularly when treating of the latter
group. At this point we give the results obtained for (a) raw
cotton, (b) bleached cotton (Ramsay and Chorley, J. Soc.
Chem. Ind. n, 872), and (c) cellulose (cotton) regenerated
from solution as thiocarbonate.

The Typical Cellulose and the Cellulose Group 69*

(c)

d)

(2)

CtJ

(l)

(2)

**** — •

fi)

(2)

Weight (grms. ) . .

45

60

50

67

45

54

50

Charcoal, p.ct.

33'33

30-00

33-00

34-33

34-44

36-0

42-0

Distillate „ . .

53'33

5O-OO

46-00

43-32

51-11

43*0

44-0

Carbon dioxide, p.ct.

6-66

9-53

11-00

5-22

777

IO'O

7'4

Oiher gases (diff.)

6-68

10-47

10-00

I7-I3

6-68

ii-o

6-6

P.ct. of cellulose
Acetic acid . .
Methyl spirit .
Tar .

100-00 100-00 100-00 100-00 loo-oo loo-o loo-o
Composition of Distillate.

2-44 1-31 175
97°

i'3i
7-07

12-00

J'S

2-0

10-24 — —

Vol. per loo grms. )
excluding CO2 • )

Carbon monoxide
Oxygen
Residual gas .

C.c.

4,900

Gases.

C.c.

4,500

Cc.

7,000

Composition p.ct.

76-90 85-74 76*20

3-66 2-80 3-34

19-44 11-46 20-46

C.c.
2,240

54*14

8-50

37*36

C.c.

2,20O

52-46

4-73

43'ii

Cc.

8,000

80-0
4-0
16-0

The decomposition by heat is accompanied, in the case of
two of the above products — viz. the raw cotton (a), and the
cellulose regenerated from the thiocarbonate solution (c) — by a
well-marked exothermic reaction.

The distillation being carried out in a glass flask heated in
an air bath, and the temperatures within the flask and in the
surrounding air space being carefully noted, it is observed that
at about 325° the former rises suddenly several degrees, and the
rise is accompanied by a rush of gases. The reaction is not
observed, however, in the case of the bleached cotton.

This exothermic resolution of the molecule we are not yet
in a position to interpret, though we may conclude that it is
the expression of some special constitutional feature. It is
attended with the formation of gaseous products, of which
the greater proportion are the oxides of carbon. It will be

7O Cellulose

seen that, although the proportions of gaseous products vary
considerably, the ratio of CO2 to CO shows a general con-
cordance, and is approximately that of their molecular weights.
There is, therefore, ground for supposing that the disruption of
the molecules is preceded by the accumulation of oxygen in
the one direction, and of hydrogen in the other direction,
within the molecule, reaching a maximum with the formation

f O

of a group prC>O, which is then split off explosively, and at

the same time resolved. The complementary phenomenon is
the further condensation of the residues to form the * pseudo-
carbon,' or charcoal, in which the carbon is accumulated
relatively to the hydrogen and oxygen, and contains ap-
proximately two-thirds of the carbon of the original cellulose.

The constitution of the carbonaceous residues of the process —
or charcoals — is at present problematical. The subject has been
discussed by the authors, in a paper on the Pseudocarbons (Phil.
Mag. May 1882), a name suggested for the designation of this group
of compounds — which may be taken to include the coal series. This
paper contains a general discussion of the composition of these
substances — chiefly devoted to showing that they are not to be
regarded as containing * free ' carbon. They are, in fact, C.H.O
compounds, and yield derivatives with chlorine, nitric acid, and
sulphuric acid, similar to those obtained by Sestini from the bromic
or ulmic group of compounds.

Synthesis of Cellulose.— With a large number of
carbon compounds it is possible to dissect them molecularly
in such a way that the component groups or residues may be
put together and the original molecule or compound recon-
stituted. This is the ordinary history of the synthesis of these
compounds, of which the modern science furnishes innumerable
instances. In the case of cellulose only one process has been
described which may be considered as a constitutional dissec-
tion, and that is, the breaking down of the molecule by sulphuric

The Typical Cellulose and the Cellulose Group 71

acid. In the final result the process may be interpreted as a
simple hydrolysis into dextrose molecules — that is, the acid plays
an intermediate part only, combining with the molecule by
simple synthesis, and interchanging with water molecules in
presence of excess of the latter. The intermediate terms of
the dissection process are not sufficiently under control to be
followed with that degree of precision which is possible in the
case of other complex carbohydrates, notably starch, which
are hydrolysed by relatively minute quantities of enzymes or
' unorganised ferments.' Even if this were possible there appears
at present no prospect of building up the cellulose molecule by
reversal of the process, as our much more complete knowledge
of the starch molecule has brought with it no suggestion of a
constructive process following inversely the lines of its hydro-
lytic dissection.

It appears, therefore, on the experimental evidence that
cellulose is built up of molecules of simple carbohydrates, but
in what manner there are none but hypothetical indications.
On the other hand, certain processes have been brought to
light which are undoubtedly direct syntheses of cellulose from
particular carbohydrates of low molecular weight. Of these two
may be cited as typical, one of which (a) is due to the action
of an unorganised ferment resembling diastase, the other (b) is
produced by a micro-organism.

(a) As a result of a change which is observed to be set up
' spontaneously ' in beet juice, a white insoluble substance is
formed, and separated in lumps or clots ; this substance has all
the characteristics of cellulose. After separating this insoluble
cellulose the solution gives with alcohol a gelatinous precipitate
resembling the hydrates of cellulose previously described.
These results are independent of the so-called viscous or mucous
fermentations. That the process by which the cellulose is
formed has the essential features of a fermentation process, is

72 Cellulose

seen from the fact that when the lumps or clots are transferred
to a solution of pure cane sugar, or beet molasses, a further
formation of the cellulose ensues. When the process proceeds
in neutral solution no carbonic anhydride is evolved ; but in
presence of acids this gas is evolved, and at the same time
acetic acid is formed in the solutions.

E. Durin, by whom these phenomena have been investigated,
(Compt. Rend. 82, 1078; 83, 128), regards the ferment as
allied to diastase, and states that fresh solutions of diastase itself
act on solutions of sugar to form the soluble cellulose, precipit-
able by alcohol. There is also some evidence that cellulose
may be formed from cane sugar in the plant by processes ot
this kind. It may be noted here that the general view current
amongst plant physiologists has been that ' starch is the material
from which plants elaborate their tissue- substances or cellulose.'
The recent researches of Brown and Morris, however, have rather
discredited this view, their elaborate and ingenious experiments
going to show that cane sugar is probably the immediate
mother substance from which the plant cell builds up cellulose,
starch being rather a reserve form for what may be regarded as
the excessive energy of assimilation in sunlight, being in turn
hydrolysed as required to feed the more continuous process of
tissue formation.

(&) A. J. Brown has more recently made observations upon
* An Acetic Ferment which forms Cellulose ' (J. Chem. Soc.
49, 432). The ' vinegar plant ' takes a membranous form;
which in microscopic examination is seen to be clearly differen-
tiated from the zoogloea form of the Bacterium Aceti. It is, in
fact, composed of bacterial rods of 2// length contained in a
membranous envelope. This envelope has the properties and
composition of cellulose.

Pure cultures of the organism placed in solutions of levulose,
mannitol, and dextrose, reproduce the growth in question, com-

The Typical Cellulose and the Cellulose Group 73

posed, i.e. of the bacteria enveloped in a ' collecting medium '
of cellulose. The proportion of cellulose formed, to the soluble
carbohydrate disappearing, is highest in the case of levulose.

It is remarkable that the cellulose formed, when hydrolysed
by sulphuric acid, gives a dextro-rotary sugar. The organism
also has the power of determining the oxidation of ethyl alcohol
to acetic acid, and of dextrose to gluconic acid. But its
characteristic property is that of the building up of cellulose
from the carbohydrates of lowest molecular weight, whence its
descriptive name Bacterium Xylinum.

The synthesis of cellulose is a problem involving the whole ques-
tion of ' assimilation ' of ' organic ' substance by the plant. It has
been held generally by physiologists for a long time that starch
is the first visible product of assimilation in the plant cell. On this
subject the student should read Sachs's classic work on * Vegetable
Physiology,' the investigations of this observer having contributed
in a very important degree to the establishment of the above view.
A priori, perhaps, it appears somewhat singular that the plant
should invariably proceed by way of starch to the elaboration of its
permanent tissue. Recent researches of Horace Brown and G. H.
Morris (J. Chem. Soc. 1893, 604) throw doubt upon the con-
clusion from the experimental side. Again, we recommend to the
student a careful study of the work of these authors, not merely for
the results obtained and described, but for the excellent plan of the
investigations. We give a few of the main conclusions in which these
investigations issued. ' It is perfectly true, as pointed out by
Sachs, that starch is the first 'visible product of assimilation ; yet
there can be little doubt (as was, in fact, anticipated by Sachs him-
self) that between the inorganic substances entering into the first
chemical process of assimilation and the starch there is a whole
series of substances of the sugar class, and that it is from the last
members of this series that the chloroplasts, under normal con-
ditions, elaborate their starch. Both under the natural conditions
of assimilation and the artificial conditions of nutrition with sugar
solutions the chloroplasts form their included starch from ante-
cedent sugar.'

Observations on the secretion of diastase by the leaves of

74 Cellulose

flowering plants, the variations of diastatic activity with the con-
ditions of assimilation, and the relations of diastase to the starch
and sugars (including maltose) present in the leaves lead to the
important conclusions which we give in the words of the original :

* Looking at the results all round, they are, it seems to us,
decidedly opposed to the view that either dextrose or levulose is
the first sugar formed by assimilation, and point to the somewhat
unexpected conclusion that, at any rate in the leaves of Tropaeolum,
cane sugar is the first sugar to be assimilated by the assimilatory
processes. There seems every reason to believe that this cane
sugar . . . functions in the first place as a temporary reserve
material, and accumulates in the cell sap of the leaf-parenchyma
when the processes of assimilation are proceeding vigorously.
When the degree of concentration of the cane sugar in the cell
sap and protoplasm exceeds a certain amount, which probably
varies with the species of plant, starch commences to be elaborated
by the chloroplasts, this starch forming a somewhat more stable
and permanent reserve material than the cane sugar, a reserve to
be drawn upon when the more easily metabolised cane sugar
has been partially used up.'

From these authors' experiments it also appears that, in the
translocation of the sugar through the leaf stalk into the stem, it
takes the form of dextrose and levulose. The former, however,
being more quickly used up in the respiratory process, there is a
larger proportion of the latter passing over into the general
metabolic circulation.

The starch, on the other hand, migrates in the form of maltose,
and this appears to be, in a sense, a starvation phenomenon — that is,
it is only put under contribution to the general supply of nutrient
material when, and in proportion as, the carbohydrates of lower
molecular weight are used up.

These researches obviously constitute an important advance
towards the elucidation of the elaborating functions of the plant
cell. What the actual first step may be in the building up of
tissue-substance, is still a matter of conjecture. The prominent
facts presented to us are, (i) that carbonic anhydride is decomposed
in the plant cell, the whole of the carbon being retained, and part
of the oxygen restored to the atmosphere ; (2) that this decomposi-
tion takes place under the influence of the protoplasmic contents

The Typical Cellulose and the Cellulose Group 75

of the living cell ; but, although, therefore, nitrogen must be
regarded as essential to the process, the plant builds up non-
nitrogenous materials, both immediate and ultimate. (3) That the
source of energy which determines these constructive changes is
that of the sun's rays ; that portion of the solar radiation chiefly
concerned being included between the wave lengths T^lhro —
T7rJ#Tnr mm., with a maximum effect corresponding to the yellow-
green of the spectrum.

Generally, it may be fairly assumed that the CO, of the atmo-
sphere is 'loosely' synthesised with protoplasmic products or chlo-
rophyll,1 and so brought within the range of the specific molecular
activities, representing what we know in the aggregate as vitality.

Constitution of Cellulose. — From the preceding general
account of the properties and reactions of the typical cotton
cellulose we might be expected to be able to deduce its consti-
tutional formula. We have, however, already pointed out that
no purely chemical synthesis of any compound similar to
cellulose has been attempted ; we are, therefore, without the
essential criterion of the correctness of any general formula
which might be proposed, if only as a condensed expression of
the relationship and functions of its constituent groups.

But although no such formula can be proposed having any
but a speculative and a tentative value, it will be a useful guide
to future investigation to sum up those reactions which throw a
direct light upon the function of the molecule as a whole, and
of its constituent groups.

(1) The resolution by sulphuric acid, and subsequent hydro-
lysis of the esters formed in the reaction, into simple carbo-
hydrate—viz, dextrose molecules. Cellulose is, therefore, in this
sense an anhydro-aggregate of the aldose groups C6H12O6.

(2) Partial resolution under the action of hydrochloric acid,
attended by the setting free of CO groups.

In cellulose the carbonyl groups are c suppressed ' ; that is,

1 This view is specifically formulated by E. Fischer, Berl. Ber. 1894.,
3231. (Dec. 10).

76 Cellulose

they either exist in combination — as in the acetals — or are
susceptible of an alternative form, the carbonyl becoming
hydroxyl oxygen.

(3) Complete proximate resolution, by { fusion' with alkaline
hydrates, into hydrogen, carbonic, oxalic, and acetic acids.
The yield of the latter tending to a maximum of 30-35 p.ct.
indicates that the grouping CO — CH2 is an important element
in the constitution of the unit groups.

(4) Negative characteristics. — These are (a) those which
characterise generally the saturated compounds — in which
group cellulose must be classified, (b) Resistance to alkaline
hydrolysis, (c) Resistance to oxidising actions up to a certain
limit of intensity, (d) Resistance to acetylation : requiring
either very high temperature or the presence of an auxiliary
(ZnCl2) for the determination of reactions of its OH groups
with the acid oxide.

(5) Synthetical reactions. — Of these the more definite are
those which yield the esters, viz. nitrates, acetates, and ben-
zoates. The highest nitrate obtainable appears to be the tri-
nitrate (hexanitrate in the C12 formula) ; the highest acetate the
tetracetate (C6 formula). A higher degree of acetylation has
been obtained, but there is undoubted evidence that this results
from molecular resolution (hydrolysis). The conclusion to
be drawn from the relationship of these esters to the parent
molecule is that, of five O atoms in the formula C6H10O5, four
react as OH oxygen with retention of the original configura-
tion of the molecule.

The thiocarbonate reaction further elucidates the functions
of the OH groups, and the resistance of the molecule to hydro-
lysis. It constitutes a further distinction of the celluloses from
starch, as a type of molecular configuration ; starch failing to
give any definite indications of this reaction, and, in contrast to
cellulose, being eminently susceptible of hydrolytic resolution.

The Typical Cellulose and the Cellulose Group 77

To sum up these more prominent points in the evidence of
constitution, we are entitled to regard cellulose as conforming, in
regard to its ultimate constituent groups, to the general features
of the simpler carbohydrates of well-ascertained constitution,
but differentiated by a special molecular configuration resulting
in a suppression of activity of the constituent groups in certain
respects, but on the other hand conferring greater reactivity is
others. This molecular configuration involves primarily the
question of the mode of arrangement of the carbon with the
qualifying hydrogen atoms within the unit groups — which, for
the reasons given, may be assumed to be of the dimensions of
C6 ; and, secondly, the grouping of these into the aggregate
which may be regarded as constituting the true molecule of
cellulose. Next in importance are those modifications of con-
figuration which are bound up with the disposition of the C
atoms.

In regard to carbon configuration the evidences are rather
indirect than determinable by the actual properties of cellulose
itself. The choice obviously lies between a chain and cyclic
formula for the unit groups. The balance of evidence is in
favour of the latter and on the following grounds : (i) the
general differentiation of cellulose in regard to stability, which
points to a symmetrical formula, as distinguished from the normal
chain upon which the hexoses are represented ; (2) the for-
mation of a cellulose acetate of the composition C6H6O (OAc)4,
in which only 2n carbon valencies are taken up in ' outside '
combination ; (3) the simple and manifold transitions of cellu-
lose— in the plant world — into keto R. hexene and benzene
derivatives. The process of lignification in the plant cell is
characterised by the formation of groups of the general form

/CH=CHV
CO/ >CH2

XC C/

(OH)2 (OH),

78 Cellulose

which remain intimately associated with the cellulose, of the
cell or fibre in combination, as a compound cellulose, there-
fore (lignocellulose, see p. 137). These derived celluloses
exhibit a close general conformity with the parent type — that
is, apart from, or in addition to, the special properties and
reactions due to the presence of the hexene ring, all the typical
characteristics of the cellulose proper.

Although, however, the hexene ring is thus shown to be
represented in compounds identified with the ' organic ' func-
tions of the plant cell, this does not appear to be the case with
the fully ' condensed ' benzene ring. Aromatic compounds are
formed in profusion, it is true, in the general range of plant life,
but when they appear it is in the unorganised form, i.e. as
excreted products of metabolism. The same appears also to
hold for the terpene series.

It may also be noted here that the supplies of raw mate-
rials— hydrocarbons &c. — for the enormous modern industry in
' aromatic ' products are derived from the products of coal
distillation, and therefore may be traced back to a cellulosic
origin.

The Cellulose Group. — Thus far we have been dealing
mainly with one member of the very numerous class of plant
constituents comprehended in the term 'cellulose.' While the
properties and characteristics of cotton cellulose are in such-
wise representative that this substance may be regarded as
the typical cellulose, the differentiation of this, as of every
other group of tissue constituents, in conformity with func-
tional variation, necessarily covers a wide range of divergencies.

The celluloses of the plant world, so far as they have been
investigated from the point of view of chemical constitution,
group themselves as follows :

(a) Those of maximum resistance to hydrolytic action, and
containing no directly active CO groups.

The Typical Cellulose and the Cellulose Group 79

(b) Those of lesser resistance to hydrolytic action, arid con-
taining active CO groups.

(c) Those of low resistance to hydrolysis, i.e. more or less
soluble in alkaline solutions and easily resolved by acids, with
formation of carbohydrates of low molecular weight.

Group (a). — In addition to the typical cotton cellulose —
which, it is to be noted, is a seed-hair — there may be included
in this group the following fibrous celluloses which constitute
the bast of exogenous flowering annuals : viz. the celluloses of
Flax (Linum usit), Hemp (Cannabis sativa), China Grass
(Rhea and Boehmeria species) ; and of the lesser known
Marsdenia tenacissima, Calotropis (gigantea), Sunn Hemp
(Crotalaria juncea).

As in the case of cotton, the celluloses of the fibres are con-
sidered in the form of the white (or bleached) and purified
residues resulting from the treatment of the raw materials by
processes of alkaline hydrolysis and oxidation more or less
severe in character. For the purification of the celluloses in
the laboratory the methods usually practised consist in (i)
alkaline hydrolysis, i.e. treatment with boiling solutions of
sodium hydrate, carbonate, or sulphite ; (2) exposure to bro-
mine water or chlorine gas ; or when oxidation alone is suffi-
cient for the removal of the * impurities,' to solutions of the
hypochlorites or permanganates (in the latter case followed by
a treatment with sulphurous acid to remove the MnO2 de-
posited on the fibre-substance) ; (3) repetition of (i) for the
removal of products rendered soluble by (2).

Special accounts of these raw fibrous materials are contained
in Spon's 'Encyclopaedia Industrial Arts'; 'Die Pflanzenfaser,'
Hugo Miiller (A. W. Hofmann's ' Bericht.' Braunschweig, 1877);
1 Report on Indian Fibres and Fibrous Substances,' Cross, Bevan,
and King (Spon, London, 1887) ; and ' Chemische Technologic d.
Gespinnstfasern,' O. N. Witt (Braunschweig, 1888).

So Cellulose

It has been already pointed out that these celluloses occur in
admixture or combination with other substances, often grouped
together in the term non-cellulose ; cellulose and non-cellulose
being usually separated jointly from the plant, and constituting
the 'raw fibre.' The raw fibre is therefore usually a compound
cellulose^ though in some cases a compound of a very weak order.
These points will be best illustrated by a careful study of com-
mercial flax. Flax is made up of the pure fibre, which is a
compound cellulose, with a certain admixture of the tissues with
which it is in contact in the stem. These adventitious components
are largely got rid of, first in the processes of breaking and scutching,
and afterwards in the further refining processes of hackling and
preparing, by which the spinner brings the fibre into the proper
condition for the twisting or spinning process proper. But the
yarn still retains residues of the cuticular cells and wood (sprit),
which then require to be broken down, or converted into cellulose,
by the chemical processes of bleaching. It is the former which
occasion the major difficulties of the linen bleacher. As a result of
the intimate association of the fibre with the cuticle of the stem,
flax, as finished for the market, contains an unusually large propor-
tion of oil -wax constituents, i.e. from 3-5 p.ct. of such bodies, soluble
in the special solvents. These may be separated by fraction-
ation into (a) ceryl alcohol and derivatives (esters), and (b} a mixture
of oily bodies of ketonic character.

For more detailed investigation of this group of flax constituents
see Hodges, Proc. R. I. Acad. 3, 460 ; and Cross and Bevan,
J. Chem. Soc. 57, 196.

This oil-wax complex plays an important part in the ordinary
process of flax 'line' spinning, and the failure of many of the
artificial processes of 'retting' flax may be explained by the
deficiency of the resulting fibre in these constituents. In the
breaking down of the cuticular celluloses, whether in the retting
(rot-steep) or bleaching process, these waxes and oils are separated.
Their elimination from the cloth necessitates the very elaborate
treatment by which the ' Belfast Linen Bleach ' is obtained.

These constituents are adventitious impurities, the bast fibre
itself being a pectocellulose (see p. 214), easily resolved by alkaline
saponification into cellulose on the one hand, and soluble modifica-
tion of the pectic group on the other. Although, therefore, the

The Typical Cellulose and the Cellulose Group 8 1

considerable loss of weight of flax cloth in bleaching (20-30 p.ct.)
falls mainly in the early alkaline treatment, the chief difficulties
are in the breaking down of the more resistant bodies derived
from the cuticle, including chlorophyll.

The celluloses of this group thus purified may be taken
as chemically identical with cotton cellulose, investigation
having so far failed to differentiate them. It must be noted,
however, that the several members of the group present dis-
tinct morphological characteristics, and differ also in such
external properties as lustre and 'feel.' These are in part
correlated with the differences in minute structure, but they are
no doubt in part differences of substance. So far, however,
we have no knowledge of the proximate constitution of these
substances, and can therefore say nothing as to the causes of
difference in this respect.

On the other hand, the essential identity of these cellu-
loses is established in regard to ultimate composition and in
reference to the following properties and reactions :

(1) Resistance to hydrolysis and oxidation, and other nega-
tive characteristics, indicating a low reactivity of the CO and
OH groups.

(2) Their relationships to the special solvents previously de-
scribed, including the thiocarbonate reaction.

(3) Formation of esters, nitrates, acetates, benzoates.

Of the above, it is sufficient in general laboratory practice
to examine cellulose in regard to ultimate composition, resist-
ance to alkaline hydrolysis, behaviour with solvents, and re-
actions with sulphuric acid (solution without blackening) and
nitrating mixture (H2SO4 and HNO3) ; the ' nitration ' pro-
ceeds without oxidation, and gives a higher yield of product,
160-180 p.ct. according to the condition.

Group (b). — These celluloses are differentiated from the
former group (i) by ultimate composition, the proportion of

G

82 Cellulose

oxygen being higher ; (2) by the presence of active CO groups ;
(3) in certain cases by the presence of the O.CH3 group.

The general characteristics of the group are those of the
oxy celluloses. It has recently been shown that these oxidised
derivatives of the normal celluloses are further characterised
by yielding furfural as a product of acid (HC1) hydrolysis.
The yield of this aldehyde is, in certain cases, increased by
previous treatment of the oxycellulose with a reagent prepared
by saturating sulphuric acid of 1*55 sp.gr. with HC1 gas. In
this reagent the oxycelluloses dissolve ; and on then diluting
with HC1 of i -06 sp.gr. and distilling, maximum yields of
furfural are obtained, the yield being a measure of the increased
proportion of oxygen beyond that corresponding with the
formula CeH10O5.

Celluloses of this class are much more widely distributed in the
plant world than those of the cotton type ; they appear, from recent
observations, to constitute the main mass of the fundamental tissue
of flowering plants, in which they usually exist in intimate mixture
or combination with other groups more or less allied in general
characteristics. It appears, from a survey of the contributions of
investigators to the subject of cellulose, that research has been very
much confined to the fibrous celluloses, more particularly to such
as receive extended industrial use. The time has come, however,
when systematic research is much needed to establish at least
a preliminary classification of the ' cellular ' celluloses upon the
lines of chemical constitution. Constitution, taken in relation to
physiological function, is an attractive subject of research ; and it is
in the plant cell, where synthetical operations are predominant, that
we have to look for the foundations of the ' new chemistry,3 which
may be expressed broadly as the relation of matter to life.

It is to be noted that the differentiation of many of these celluloses
from the typical cotton is, in regard to empirical composition, only
slight. There appear, on the other hand, to be more important
differences of constitution. Thus pine-wood cellulose dissolved in
sulphuric acid, the solution diluted and boiled, and further treated
by the isolation of crystallisable carbohydrates, yields these (i.e.

The Typical Cellulose and the Cellulose Group 83

dextrose) in only small proportions. (Lindsey and Tollens, Lieb.
Ann. 267, 370.)

Investigation has stopped short at this negative result. It would
be of interest, therefore, to isolate the products formed in the re-
action with the concentrated sulphuric acid, so as to characterise
them, at least generally. Until this is done, or some other method
proximate resolution is worked out in detail, we can only say that
the constitution of these celluloses is in some important feature
radically different from that of the typical cellulose.

An account of recent investigations of these * celluloses ' will
be found in Berl. Ber. 1893, and a more special treatment of
the subject, ibid. 1894, and J. Chem. Soc. 1894 (C. Smith).

Of this group of the natural oxycelluloses the following
have been more particularly investigated :

(i) Celluloses from woods and lignified tissues generally. —
Lignified tissues are made up of compound celluloses, to be
subsequently described (see Lignocelluloses, p. 91), from which
the celluloses may be isolated by a number of treatments, all
depending upon the relative reactivity of the so-called ' non-
cellulose ' constituents, which in combination with the celluloses
make up the compound cellulose, lignocellulose or wood sub-
stance. These non-cellulose constituents are readily attacked
and converted into soluble derivatives ; and there are various
industrial processes for preparing celluloses (paper pulp) from
raw materials of this class, depending upon the direct conversion
of the former into such soluble compounds. The isolated
celluloses show the following general characteristics (Berl.
Ber. 27, 1061) :

Elementary composition { £ ^.g^.'* ^ Yidd °f fur~

fural, by solution and hydrolysis (HC1), 2-6 p.ct. Reactions
with phenylhydrazine salts and magenta-sulphurous acid, indi-
cating the presence of active CO groups. These celluloses
are necessarily less resistant to oxidation and hydrolysis, but

G2

84 Cellulose

show in all other respects a close general agreement with the
normal cotton cellulose.

(2) Celluloses from cereal straws, from esparto, <5rV. — These
celluloses are isolated from the matured stem, or haulm, by
digestion with alkaline lye at elevated temperatures. They are
also of considerable industrial importance, being largely used
in the manufacture of the cheaper kinds of writing and printing
papers.

Recent investigation has shown that these celluloses are
strongly differentiated from the normal, and are in fact pro-
nounced oxycelluloses. The following are the characteristics
of difference :

Ultimate composition, after treatment with hydrofluoric
acid to remove siliceous ash constituents :

Oat straw cellulose Esparto cellulose
d) (2) (i) (2)

C . . . 42-4 42-4 4178 41-02
H . . . 5-8 5-8 5-42 5-82

"Yield of furfural by solution and hydrolysis (HC1) :

Oat straw cellulose Esparto cellulose

I2'5 I2'2

Reactions. — In addition to those with Fehling's solution,
phenylhydrazine salts, and magenta-sulphurous acid indicating
the presence of active CO groups, the celluloses give a
characteristic rose-red colouration on boiling with solutions ot
aniline salts. This reaction serves to identify their presence
in papers, and from the depth of the colouration, the percentage
may be approximately estimated.

Investigation has also established the following points in
regard to the oxidation and deoxidation of these oxycelluloses.

They are gradually oxidised in dry air at the temperature
of the water-oven, undergoing discolouration ; the yield of fur-
fural, by hydrolysis, showing a progressive increase. They are
deoxidised, on the other hand, by neutral and alkaline reducing

The Typical, Cellulose and the Cellulose Group 85

agents. Thus after lengthened exposure to solutions of zinc-
sodium hyposulphite, prepared by the action of zinc dust upon
sodium bisulphite, the yield of furfural — which is a measure of
the degree of oxidation — was reduced, in the case of esparto
cellulose, from 12-6 to 8-9 p.ct.

A still further deoxidation results from solution of these oxy-
celluloses as thiocarbonate, and regeneration of the cellulose
by heating the solution at 80-100°. The regenerated cellulose
approximates to the normal, yielding only 2 p.ct. furfural
on hydrolysis. It is to be noted, however, that esparto cellu-
lose, in common with all the celluloses of this group, is partly
hydrolysed to soluble derivatives by this treatment ; the re-
generated cellulose amounting to 80 p.ct. of the original weight
dissolved. The soluble portions yield furfural on hydrolysis,
amounting (in a typical experiment) to 4-0 p.ct. of the original.

The celluloses of this group are dissolved by concentrated
sulphuric acid to dark coloured solutions. On diluting and
boiling they are resolved into carbohydrates of low molecular
weight j dextrose appears to be invariably formed, and in many
cases also mannose ; but only very small yields of either
carbohydrate have been so far obtained.

Group (c).— This includes the heterogeneous class of non-
fibrous celluloses which we have defined as of low resistance
to hydrolysis, being easily resolved by boiling with dilute acids,
and being also more or less soluble in dilute alkaline solutions.
This group has been but little studied, and therefore can only
be generally characterised. Physiological research has shown
that there are a large number of cellular, as distinguished from
fibrous ' celluloses,' which are readily broken down (hydrolysed)
by the action of enzymes within the plant itself, whether as a
normal or abnormal incident of growth. Thus in the germina-
tion of starchy seeds, the cell walls (cellulose) of the starch-con-
taining cells are broken down, as a preliminary to the attack

86 Cellulose

upon the starch granules themselves, to form the supply of nu-
trition to the embryo. In an exhaustive investigation of the
germination of the barley, Brown and Morris have thrown a
good deal of light upon this particular point, which they empha-
sise in the following words : ' that the dissolution of the cell
wall invariably precedes that of the cell contents during the
breaking down of the endosperm is a fact of the highest physio-
logical importance, and one which for the most part has been
strangely overlooked.'

A similar, but abnormal, dissolution of cell walls is that
which occurs in the attacks of parasitic organisms upon the
tissues which they invade.

These processes are well known to physiologists, who, how-
ever, generally regard * cell- wall ' and * cellulose ' as substan-
tially identical terms. The chemical differentiation of the sub-
stances comprising cell walls is, on the other hand, an
entirely new field of research ; but although investigation
has not gone very far, the results are sufficient to show that
the celluloses of this order are enormously diversified. The
variations already disclosed are (i) those of the carbohydrates
yielded by ultimate hydrolysis, and (2) those of molecular con-
figuration or condensation. We have already seen that the
celluloses of the cotton group (a) yield dextrose as the ultimate
product of hydrolysis ; those of group (£) yield, in addition to
dextrose, mannose and probably other bodies ; and the group
we are at present discussing yield, in addition, galactose, and
the pentoses xylose and arabinose. In illustration we may cite
a few examples. Thus GALACTOSE has been obtained as a pro-
duct of hydrolysis of the cell walls of the seeds of Lupinus
luteus, Soja hispida, Coffea arabica, Pisum sativum, Cocos
nucifera, Phoenix dactylifera, &c. MANNOSE is obtained in
relatively large quantity from the ' ivory nut,' and from a very
large number of other seeds ; and PENTQSES, from the seeds

The Typical Cellulose and the Cellulose Group 87

of the cereals and of leguminous plants. It appears, therefore,
generally that a large number of plant constituents which have
been denominated by the physiologists as 'Cellulose' have
little more title to be considered as such than has starch.
However, external resemblances count for something, at least in
the beginnings of classification, and substances of the type we
are considering may be conveniently grouped with the cellu-
loses ; but we should propose to apply to them the term
PSEUDO-CELLULOSES, or HEMICELLULOSES— as has been pro-
posed by E. Schulze. Our group (c) of pseudo-celluloses may
therefore be defined as — Substances closely resembling in
appearance the true celluloses, but easily resolved into simple
carbohydrates by the hydrolytic action of enzymes, or of the
dilute acids and alkalis.

Animal Celluloses. — Tunirin — a compound of the em-
pirical composition of cotton cellulose, and resembling it in a
number of its properties and reactions — is isolated from the
mantle of Ascidia and other invertebrate species by exhaustive
hydrolytic treatments. Such resistant residues have been in-
vestigated by Schmidt (Annalen, 54, 318), Berthelot (Compt.
Rend. 47, 227), Lowig and Kolliker (J. Pr. Chem. 39, 439),
Schafer (Annalen, 160, 312), and more recently by Franchimont
(Compt. Rend. 89, 755). From these later investigations it
appears that the sugar obtained as the product of ultimate hydro-
lysis is identical with the dextrose obtained from the vegetable
celluloses. From this and its reactions generally, which differ in
some respects from those of the normal cellulose, Franchimont
concludes that the compound is undoubtedly a cellulose, but
of different constitution from the normal. Cellulose has also
been identified as a constituent of the protozoa. Investigations
of one of these organisms— Ophrydium versatile — by Halli-
burton showed the investing matrix of a colony of these ciliated

88 Cellulose

protozoa to consist in the main of a cellulose similar to that of
the Tunicata (Q. J. Micr. Soc. 1885, 445).

These scattered observations indicate that the special
constitutional type or configuration of the celluloses is not con-
fined to those of vegetable origin. There is, of course, no reason
in the nature of things that the distribution of the type should
be limited to the plant world. It is quite possible, in fact, that
the animal fibres, and more generally the colloids of the animal
skeleton, may prove to be of similar carbon configuration to
that of the celluloses. A systematic investigation of such a
possibility has, so far as we know, not been attempted. Sugges-
tions have been made — in, it is true, rather a wild way — that the
silkworm is engaged in converting the cellulose of the mulberry
leaf into silk. It is impossible to say, a priori ', how far the
digestive processes in an organism of this order may be destruc-
tive in character, but an exhaustive physiological investigation
would throw light on the point. It is clear, of course, that the
animal organism is not constructive in the same sense as the
plant cell, and it is an interesting subject for speculation and
experimental inquiry how far the vegetable products constituting
che food of animals are broken down by the digestive process ;
or, in other words, how far they may preserve their constitutional
features in being synthesised to ' animal ' products.

PART II
COMPOUND CELLULOSES

IN dealing with the isolated celluloses it has been shown that
the processes by which they are isolated or purified are based
upon the relative reactivity of the compounds with which the
celluloses are combined or mixed, in the raw or natural
products of plant life. These natural forms of cellulose are, of
course, multitudinous. Remembering the infinite variety of
the vegetable world, the endless differentiation of form and
substance of the tissues of plants, it might be presumed that
the chemical classification of these products would present
unusual complications.

Investigation, however, has shown, and continues to show,
that this great diversity of substance, as revealed by proximate
analysis, exists upon a relatively simple chemical basis. The
compounds constituting the fundamental tissue of plants may,
in fact, be broadly classified in correspondence with the three
main types of differentiation of the cell wall long recognised by
the physiologists, viz. lignification^ suberisation, and conversion
into mucilage. That is to say, in addition to the celluloses
proper and hemi- or pseudo-celluloses — which may be defined
as polyanhydrides of the normal carbohydrates, ketoses and
aldoses — there are three main types of compound celluloses in
which the celluloses as thus defined exist in combination with
other groups, as follows :

Lignocelluloses. — The substance of lignified cells and

90 Cellulose

fibres, notably the woods — of which the characteristic non-
cellulose constituent is a R. hexene derivative.

Pectocelluloses and Mucocelluloses.— Comprising a
wide range of tissue constituents — of which the non-cellulose
constituents are colloidal forms of the carbohydrates, or closely
allied derivatives, easily converted by hydrolytic treatments into
soluble derivatives of lower molecular weight, and belonging
to the series of ' pectic ' compounds, or hexoses, &c.

Adipocelluloses and Cutocelluloses.— The substance
of cuticular and suberised tissues — in which the cellulose is
associated with fatty and waxy bodies of high molecular weight.

To deal with these groups in detail would involve a survey
of the entire vegetable kingdom ; of which, on the other hand,
but a very small section has been subjected to systematic
investigation. It is true, of course, that an immense number
of proximate analyses of vegetable products have been put
upon record ; but the analytical methods adopted have been of
the empirical order, and their results, stated under such terms
as * crude fibre,' ' non-nitrogenous extractive matters,' cannot
be regarded as 'systematic' in the sense of constitutional
diagnosis, We shall therefore confine ourselves to an account
of typical members of the above groups, and such as have been
investigated by molecular, as opposed to statistical methods.

Fre'my has devised (Compt. Rend. 83, 1136) a system ol
chemical differentiation and classification of vegetable tissue con-
stituents, which, although it has found but little favour, and is in
fact generally rejected by critical writers on this subject, may be
briefly noted here.

The classification embraces (a) celluloses, including 'para-
cellulose ' and ' metacellulose ' in addition to the normal cellulose ;
(£) vasculose ; (c) cutose ; (d) pectose and pectic compounds ; and
(e) nitrogenous bodies.

The celluloses (a) are differentiated by treatment with the
cuprammonium reagent : * cellulose' dissolves directly ; ' para-

Compound Celluloses 91

cellulose' (epidermis of leaves, &c.) dissolves after boiling with
hydrolysing acids ; ' metacellulose ' (found chiefly in lichens)
remains insoluble after the acid treatment.

Vasculose is insoluble in cuprammonium ; it is readily dissolved
on heating at high temperatures with solutions of the alkalis. It is
attacked by all oxidising agents. It may be selectively attacked
by dilute nitric acid. Vasculose is said to abound in hard woods,
the hard concretions of pears, &c. It appears to be identical with
the lignocellulose of this treatise ; the non-solubility in cupram-
monium being a statement of doubtful value.

Cutose is the substance of the transparent cuticular membrane
of leaves &c. It has been further studied by Fremy, and the
results of the later investigations are given in this treatise, p. 229.

Pectose and pectic constituents. — These have also been further
investigated by Fremy, and his results are noted in connection
with the pectocelluloses, p. 216.

In the main, therefore, the lines of this classification are
adopted in this treatise, but that is probably because, and in so
far as, they have a physiological basis. Chemically speaking, the
classification is of little value, since it rests chiefly upon the actions
of hydrolytic agents. Fremy's experimental work, on the other
hand, is of a certain empirical value apart from the conclusions
drawn from the results ; but as it does not contribute to the
solution of constitutional problems it will not be reproduced here.
An exhaustive account of the researches will be found in Ann.
Agron. 9, 529.

Of the three groups of compound celluloses, the lignocellu-
loses stand first in order of importance. Not only are they by
far the most widely distributed, but they have a physiological
and a special chemical significance which mark them out as the
arena of some of the most interesting processes presented by the
many-sided synthetical activity of the plant cell.

Of the lignocelluloses there are two well-defined types :
(i) the bast fibre of the Corchorus species, known in commerce
asjute ; (2) the woods, i.e. the lignified tissues of perennial
stems. The former, being a simple tissue and an annual
growth, is a more promising subject for the investigation of the

92 Cellulose

general chemistry of lignifcation than the woods, which are, of
course, complex structures and subject to continuous modifica-
tion with lapse of time and in adaptation to the varying neces-
sities of the plant. For this, amongst other reasons, the jute
fibre has been more thoroughly investigated than the woods.
The results of these investigations will therefore be reproduced
at some length. It will simplify the treatment of the subject
if we first give a brief account of the fibre-substance in theo-
retical terms ; afterwards the methods of investigation by which
the theoretical conclusions have been established will be given
in greater detail, and in strict sequence of the lines upon which
the celluloses proper have been described.

Lignocelluloses.— (i) The Jute Fibre.— The jute
fibre-substance differs strikingly in composition and re-

(~* TT

actions from the celluloses. With its higher — ^_ ratio, viz.

6-1-5-8 there are associated the characteristics

-

O 47-9-47-2

of an unsaturated compound — i.e. it contains C=C groupings,
and these are localised in C6 rings. These rings are, further,
of ketonic or quinonic character (containing a CO group), and
appear to be linked, by O, into complexes of the magnitude of
CI8. They combine readily with chlorine, in presence of water,
and the resulting quinone chlorides are bodies of definite
properties and reactions.

A second characteristic constituent of the fibre-substance
is a furfural-yielding complex, which appears to be an oxycellu-
lose derivative, a polyanhydride passing by hydration into an
oxycellulose of the ordinary type.

The third main constituent is the cellulose of the fibre,
which can only be isolated by chemical treatments selectively
attacking the ' non-cellulose ' in which the two previously de-
scribed constituents are comprised. The reagents available for

Compound Celluloses 9$

the purpose are chiefly the halogens, the halogenated deriva-
tives of the non-cellulose being dissolved away by treatment
of the product with alkaline solutions. The cellulose thus iso-
lated is not homogeneous, but is made up of a more resistant
a- and a less resistant /3-cellulose— more or less resistant,
i.e. to the action of oxidants and hydrolytic agents. By other
reactions, therefore, in which the oxidising or hydrolysing
conditions are more severe, the /3-cellulose is converted into
soluble derivatives. Such are, digestions with dilute nitric
acid ; with permanganates, in presence of alkali in excess ;
with solutions of the bisulphites at elevated temperatures. This
/3-cellulose is characterised by the presence of O.CH3 groups.
The a-' cellulose ' is an oxycellulose.

There are other minor characteristics of the non-cellulose
portion of the fibre-substance which remain to be noticed.
These are (i) the presence of OCH3 groups, in larger propor-
tion than in the /3-cellulose ; (2) the presence of a CH2.CO
residue, which is split off as acetic acid, under various hydrolytic
treatments of the fibre-substance, probably in union therefore,
as a side chain, with the R. hexene groups ; (3) the presence of
a body giving the characteristic reactions of the pentaglucoses.
The pentosans are, in fact, obtainable in small quantity as
products of alkaline hydrolysis of the fibre-substance ; and the
furfural-yielding constituent of the non-cellulose, already de-
scribed as a condensed oxycellulose derivative, might be assumed,
on this evidence, to possess the pentose configuration ; but the
evidence available so far is not such as to give a definite
solution of this point.

These are the main points of constitutional differentia-
tion of the lignocelluloses from the celluloses proper. It has
been largely the custom to describe the compound celluloses
of this class as mixtures of cellulose and non-cellulose, the
latter being described generally as 'encrusting matters,' or

94 Cellulose

under the more special term lignin^ in recognition of its, or
their, well-defined individuality.

This view will be found inconsistent with the results of the
systematic investigation of this particular fibre-substance, as
indeed of the Mignified celluloses' generally. They are
found to be very uniform in composition ; the cellulose
and non-cellulose are in intimate combination, resisting severe
hydrolytic treatment ; and in a large number of reactions the
typical characteristics of the celluloses are preserved. Therefore
the substantive term Cellulose is used in describing them, with
the addition of the adjective or qualifying prefix. Where we
have to speak of the non-cellulose complex we shall use the
term lignone, indicating thereby its ketonic characteristics.

We may sum up these outlines of the constitutional
features of the jute fibre-substance in a general diagram :

Lignocellulose.

Cellulose Lignone (non-cellulose)

Cellulose o Cellulose )8 Furfural-yielding Keto R. hexene

Containing Containing complex group

oxidised groups O. CH3 groups and secondary constituents

°'CH3 and CH-CO

groups residue

With this preliminary general survey in view, the experi-
mental treatment of the subject matter will be more readily
appreciated in its bearings upon the constitutional problem.

Methods of Quantitative Estimation of Constituent
Groups. — The groups above described may be regarded as
the proximate constituents of the fibre-substances, and they
may be quantitatively estimated by particular methods of
proximate resolution, which must be described in some detail.

(i) CELLULOSE. — Chlorination method. — For the elimina-
tion of the non-cellulose, by conversion into soluble derivatives,
various methods are available. One method only gives maxi-

Compound Celluloses 95

mum yields of cellulose, and for the reason that it is based
upon a well-defined reaction of the lignone group, admitting of
perfect control. This group, or rather its R. hexene constituent,
reacts with chlorine gas, in presence of water, to form a quinone
chloride, without, at the same time, affecting its union with the
furfural-yielding complex. On treating the chlorinated fibre
with sodium sulphite solution, the lignone chloride is dissolved,
and at the same time converted into a brilliant magenta
colouring matter. The undissolved residue (75-80 p.ct.) is
the cellulose of the fibre. The process is carried out as
follows :

About 5 grms. of the fibre— weighed after drying at 100° — are
(a) boiled for 30 minutes with a dilute solution of sodium hydrate
(i p.ct. NaOH), which is kept at constant volume by addition of
water. The fibre is well washed on a cloth or wire gauze filter,
squeezed to remove excess of water, opened out, placed in a
beaker, into which (b) a slow stream of washed chlorine gas
is passed. Rapid reaction ensues, and the fibre changes in
colour, from brown to a bright golden yellow. To ensure
complete conversion of the lignone, it is necessary to leave the
fibre for some time (from 30-60 minutes) in the atmosphere
of Cl gas. (c] The chlorinated fibre is removed, washed once
or twice with water to remove hydrochloric acid, and placed
in a 2 p.ct. solution of sodium sulphite ; the solution is gradu-
ally raised to the boiling point, a small quantity of caustic soda
solution is added (0-2 p.ct. NaOH calculated on the solution),
and the boiling continued for 5 minutes, (d) The cellulose
is now thrown upon a cloth filter and washed with hot water.
It will be found to be almost pure, i.e. white ; but to remove
the last residues of the non-cellulose, it may be bleached by
immersion in a dilute solution of hypochlorite (0*1 p.ct.
NaOCl) for a few minutes, or treated with dilute permanganate
solution (o'i p.ct. KMnO4). It is well washed from these

96 Cellulose

oxidising solutions, treated with sulphurous acid on the filter,
well washed with water, squeezed, dried and weighed.

The cellulose estimations by this method give what may be
considered the maximum yield ; other methods attack the
/3-cellulose more or less, giving products which are dissolved
and removed. These methods may be briefly noticed.

(2) Bromine water (Hugo Miiller).— This halogen, and in
the form of aqueous solution, fails to saturate the R.
hexene groups in one operation ; hence the alternate treat-
ment with bromine water in the cold, and boiling alkaline
solution, requires to be once, twice, or even three times
repeated. The yield of cellulose is 2-5 p.ct. lower, and
the process is by comparison tedious.

The difference in yield is due to the attendant oxidation
and hydrolysis of the /3-cellulose.

The following experimental determinations bear upon this
point :

A specimen of jute gave the following percentages of cellulose
under various methods of treatment :

(0 73-74 P-ct- Bromine water (cold) and boiling aqueous
ammonia alternately till pure.

(2) 74-76 p.ct. Chlorination at ordinary temperatures, fol-
lowed by alkaline hydrolysis.

(3) 80-9, 80-6, 79-9, 82-0, 81-3, 84-5, in individual experiments in
which the treatments were varied as follows : (a) chlorination at
0-5° ; (6) followed by digestion in dilute sulphurous acid at 0-5° ;
(c} hydrolysis with sodium sulphite solution, at first cold, afterwards
raised to boiling.

From these results it appears that, by excluding oxidising con-
ditions and graduating the hydrolysis of the chlorinated derivative,
a considerable proportion is separated in the form of fibrous cellu-
lose ; this portion, under other conditions, is hydrolysed to soluble
derivatives.

(3) Nitric acid and potassium chlorate (Schulze). — This
method consists in a prolonged digestion 10-14 days) of

Compound Celluloses 97

the fibre- substance, at ordinary temperatures, with nitric
acid of no sp.gr., containing potassium chlorate (o'5-o*8 p.ct
of the weight of the fibre) previously dissolved in the acid.
The lignone is attacked jointly by nitrogen and chlorine oxides,
and largely converted into derivatives, soluble in the acid
solution. The /3-cellulose is also considerably attacked
(oxidised), and the action extends to the more resistant (a)
cellulose. The residues of the lignone are dissolved away by
boiling the washed fibre with dilute ammonia.

The process has been largely used in investigations of the
lignocelluloses ; but the results, both as to yield and composi-
tion of the cellulose, are, for obvious reasons, of subordinate
value.

(4) Dilute nitric acid at 50-80°. — By digesting the fibre-
substance with nitric acid (5-10 p.ct. HNO3) at 60°, the lignone
is entirely converted into soluble derivatives ; the /3-cellulose
is also hydrolysed and dissolved ; the residue of the treatment
being the more resistant a-cellulose. The interaction of the
lignone and the acid is of use in elucidating the constitution of
the non-cellulose groups, and will be subsequently described
from this point of view. As a process of cellulose isolation, the
reaction is carried out as follows :

The weighed fibre is placed in a flask and covered with
three times its weight of 10 p.ct. HNO3. It is heated for
some hours at 60° until the fibre has changed to a pale lemon
yellow colour, the solution being of a bright yellow. After
washing away the acid by-products the residual cellulose is
boiled with a solution of sodium sulphite, which removes
the last traces of lignone derivatives. It may then be finally
washed on a cloth filter, squeezed, dried and weighed. The
yields of cellulose by this method are 63-65 p.ct. of the fibre.

(5) Sulphite and bisulphite processes. — By digestion with
solutions of the sulphites of the alkalis, or of the bisulphites

H

98 Cellulose

of the alkaline earths, at high temperatures, the lignone groups
are attacked and dissolved, as the result of a specific reaction,
which will be subsequently dealt with in its bearings upon
the constitution of the lignone molecule.

As a ' cellulose process,' the reaction may be carried out as
follows :

Neutral sulphite. — The fibre is sealed up with five times its
weight of a 6 p.ct. solution of sodium sulphite (Na2SO3.7H2O).

Bisulphite of lime or magnesia. — The fibre is sealed up
with five times its weight of the bisulphite solution containing
3 p.ct. total SO2.

The digestion is carried out at high temperatures — either in
glass tubes or autoclaves of metal, according to the circum-
stances of the laboratory. In the latter case, iron vessels may
be used with the neutral sulphite, but to prevent reaction with
the metal, the solution should contain sodium carbonate (£ the
weight of the sulphite). For the bisulphite treatment a lead-
lined digester is required. The maximum temperatures
necessary are 180° for the neutral sulphite, 160° for the bi-
sulphite process.

The temperature is raised gradually to the maximum, at
which it is maintained for 2-3 hours, the entire duration
of digestion necessary being from 6-8 hours. At the expira-
tion of this time, the vessels are cooled off, the contents
thrown on to a cloth filter, the residual cellulose washed
thoroughly with hot water, and finally purified by treatment
with dilute hypochlorite (0-5 p.ct. NaOCl) or permanganate
solution. After washing from the oxidising solution, the cellu-
lose is treated with sulphurous acid, from which it is thoroughly
washed, squeezed, and dried for weighing.

The yields of cellulose are from 60-65 P-ct- > tne /3-cellu-
lose being, under this treatment, also hydrolysed and dis-
solved.

Compound Celluloses 99

Furfural-yielding Complex.1— It has been found, so far,
impossible to isolate this constituent of the fibre-substance ; in
the mean time, we are limited to indirect methods of arriving at
its constitution and its quantitative relationship to the ligno-
cellulose molecule. It has been established, by the elaborate
researches of Tollens and his pupils, that the condensation to
furfural of those carbohydrates which, from their constitution or
configuration, readily yield this special product of dehydration,
admits of such control that the yield of the aldehyde may be
regarded as constant, and an exact measure, therefore, of the
parent molecule. The general method of conversion is that of
boiling the substance with hydrochloric acid of 1*06 sp.gr. (12
p.ct. HC1). For the estimation of the resulting furfural, various
methods have been proposed and practised ; the final selection
resting with that which consists in converting the aldehyde
into its hydrazone, which is then gravimetrically estimated.

A careful survey of the evidence upon which this selection
is grounded, together with an elaborate account in detail of the
methods both of the conversion of the carbohydrate into the
aldehyde in question, and its estimation as described, will be
found in a recent paper by Flint and Tollens (Landw. Vers.-
Stat. 42, 381-407), which should be closely studied.

The process may be outlined as follows : (a) The weighed
fibre (5 grms.) is placed in a flask, covered with 100 c.c. of
hydrochloric acid of ro6 sp.gr. The flask is fitted with a
double-bored indiarubber cork, carrying (i) the connection
to the condenser, the usual bent glass tube, (2) the tubulus
of a stoppered 'separating funnel.' The flask is heated,
preferably in a bath of oil or fusible metal, so that the rate of
distillation is about 2 c.c. per minute. The distillate is col-

The designation * furfural-yielding complex' or carbohydrate may be
conveniently shortened to furfurose or furfurosan in accordance with the
modern nomenclature of the group.

II 3

ioo Cellulose

lected in portions of about 30 c.c., and when each such quan.
tity is obtained, a corresponding quantity of the acid is admitted
to the flask through the separating funnel . The distillation is
continued until a drop of the distillate ceases to give the well-
known reaction of the aldehyde (rose-red colouration with
aniline acetate, in presence of acetic acid).

(b) Conversion into hydrazone. — To ensure constant results
it is important that constant conditions be adopted. The
hydrochloric acid is neutralised with sodium carbonate, a
slight excess being used ; the solution is then made acid with
acetic acid. The distillate is made up to a constant volume,
and, as it is necessary to keep the proportion of sodium chloride
approximately constant, any deficiency of volume is made up
with a salt solution of corresponding concentration.

The phenylhydrazine solution is made up with 12 grms. of
the base, 7*5 grms. glacial acetic acid, and water to 100° c.c.

The formation of the hydrazone takes place according to the
reaction

C5H4O2 + Ph.N2H3 = C5H4ON2HPh + H2O.

and in any series of determinations with the same
substance, the quantity of phenylhydrazine solution necessary
to be added is, therefore, approximately known. The quantity
is controlled by testing the solution with aniline acetate, drops
of the solution being placed upon filter paper moistened with
the reagent. The solution is set aside for the separation of the
hydrazone, which is much facilitated by continuous stirring.
The hydrazone is collected in a filtering tube, containing a
perforated plate of platinum or porcelain upon which a circle
of filtering paper is laid, the filtration and washing of the pre-
cipitate being expedited by the use of the pump. The filter
tube, with its contents, is dried preferably in vacua at 60-70°,
or in a slow current of dried air at this temperature. It is
then weighed. Having been weighed together with the filter

Compound;t€,elbi?.o'se$> , > \ '', i M i />\ IO1

paper before the experiment, the difference of the weight gives
the weight of hydrazone obtained. From this weight, that of
the furfural is calculated as under.

Furfural = [Hydrazone x 0-538]. The factor 0*538 is the
mean number obtained from an extended series of experi-
mental determinations. The variations of the results obtained
from the theoretical are due to the slight and varying solubilities
of the hydrazone in salt solutions of varying concentration.
For more exact approximation, the factor must be selected
according to the exactly ascertained conditions of its precipita-
tion, and the corresponding value determined by Flint and
Tollens (loc. at.) being used in calculating from the hydrazone
to the aldehyde.

In the case of the lignocelluloses we are not able to calcu-
late the results to their final expression, i.e. the weight of the
parent substance from which the furfural is obtained, for the
reason that the constitution of the latter has not been
definitely ascertained. The fibre-substance certainly contains
pentosanes, and the pentose xylose has been isolated in small
quantity as a product of alkaline and acid hydrolysis (Tollens,
Berl. Ber. 1889, 1046). As, however, the evidence goes to
show that other furfural-yielding groups— probably oxidised
hexose derivatives — are present, we are limited to an approxi-
mate estimate of the quantity of the entire furfural-yielding
complex. This approximation is furnished by multiplying the
weight of hydrazone by i'i ; or, in other words, the weight of
the furfural-yielding constituents may be taken together at
about twice the weight of the furfural obtained.

Apart from all hypothetical considerations, however, the
yield of furfural is an important constant of the fibre-substance,
which may be determined as described, within the limits of
satisfactorily small errors of experiment.

Keto R. Hexene Constituent— The characteristic

1O2

reaction of this group is its combination with chlorine, th£
quantitative features of which have been the subject of careful
investigation. The chlorinated lignone is a body of
definite and uniform composition, represented by the empirical
formula C19H,8C14O9. It is still a complex containing a
quinone chloride, allied to mairogallol (C^HyClnOio) and
leucogallol — products of chlorination of pyrogallol, under care-
fully regulated conditions — in combination with the furfural-
yielding complex. The combination with the chlorine is
attended by molecular hydration, in consequence of which
the chlorinated lignone is split off, more or less, from its con-
densed union with the cellulose.

As in the preceding case, therefore, we are dealing with a
reaction which, though perfectly definite and characteristic of
constituent groups of the parent molecule, cannot be inter-
preted in terms of these groups without introducing hypothe-
tical considerations. The reaction will be discussed subse-
quently from this point of view. In the mean time it is sufficient
to point out that the reaction is uniform in its empirical features,
that these may be quantitatively studied, giving what we may
term the constants of chlorination of the fibre-substances, viz.
(i) the chlorine combining with the hexene groups of the
fibre-substance as quinone chloride ; (2) the chlorine com-
bining with hydrogen, and set free as hydrochloric acid. It is
found, in the case of jute, that (i) and (2) are approximately
equal, and therefore that the reaction is unattended by second-
ary oxidations of the fibre-constituents to any notable extent.

The following are the details of the methods of estimating
these constants of chlorination.

(i) Volume of chlorine disappearing in chlorination. — The
method of observation is fully described in J. Chem. Soc.
1889, 169. The fibre-substance is prepared in the usual way, by
previously boiling for 10-15 minutes in dilute alkaline solution

Compound Celluloses 103

(i p.ct. NaOH). This treatment is carried out in duplicate, one
of the treated specimens being washed off with water, dilute
acid (acetic), and finally water, and dried, for the estimation of
the loss of weight in the alkaline treatment, (a) The second
portion, after thorough washing from the alkaline treatment
and finally with distilled water, is squeezed so as to retain a
minimum of water, introduced into a glass bulb of extremely
thin walls, and sealed off. (b) The bulb is carefully introduced
into a bottle previously filled with chlorine gas, collected over
warm water and inverted with a glass plate placed on the mouth,
and in such a way that a minimum quantity of water is left
in the bottle. In any case a suitable quantity of coarsely
pounded glass should be introduced with the bulb, in order
to prevent the fibre being unduly wetted, which would retard the
absorption. The bottle is closed with an indiarubber cork, well
coated with paraffin, through which passes a bent tube. Through
this tube the bottle is brought into connection with any suitable
gas-measuring apparatus permitting accurate measurement of
the vacuum formed in the bottle as the reaction proceeds.
All parts of the apparatus being brought to a constant tempera-
ture, the bulb containing the fibre is broken by a blow against
the sides of the bottle. The chlorine is absorbed with rapidity,
and observations of the absorption are made from time to
time. If an apparatus such as a Lunge's nitrometer is used,
the apparatus is adjusted at its extreme mark, i.e. full of air.
As the volume of gas in the reaction flask shrinks, the liquid
levels in the measuring apparatus are adjusted in the usual
way. The reaction may be considered at an end when no
further absorption is noted during an interval of 10 minutes.
It is advisable to insert a stopcock between the reaction bottle
and the measuring apparatus, so that the latter may be cut off
after each observation of volume.

In calculating from the observed numbers, it is only neces-

IO4 Cellulose

sary to note that the gas disappearing in combination is a
volume observed under the conditions of the experiment ; it is
corrected therefore for temperature, barometric pressure, and
the partial pressure of aqueous vapour, and the corrected
number reduced to the weight of the chlorine taking part in
the reaction. This weight is 16-17 P-ct- of tne weight of the
fibre-substance.

The quantity of the latter which may be conveniently taken
is 1-2 grms. ; the quantity of gas, measured under ordinary
conditions, required for 2 grms. of the lignocellulose is 100-
1 20 c.c. It is advisable to take a reaction bottle, of capacity
equal to twice this volume.

The errors of experiment in such a determination are not
very considerable ; they may be minimised by keeping the
reaction bottle submerged in water of constant temperature,
and shielding it from the light, to prevent interaction of
the chlorine and water ; also by observing the precautions
usual in the measurement of gas volumes.

(2) Determination of HCl formed in the reaction. — The
quinone chloride formed in the reaction is slightly soluble in
water, but almost insoluble in a solution of common salt (20
p.ct. NaCl). By washing the chlorinated fibre with a neutral-
ised salt solution, the hydrochloric acid may be removed. The
reaction bottle being disconnected, the salt solution is poured
down the sides, the fibre and bottle being further washed once or
twice with the salt solution. Residues of chlorine are removed
by passing a current of air for a minute or two through the solu-
tion, which is then treated with standard alkali in the usual way.

The chlorine converted into hydrochloric acid is, in the
case of jute, approximately one half the total chlorine entering
into reaction, i.e. from 8-8-5 P-ct- The reaction appears,
therefore, to be one of simple substitution of hydrogen. In
the case of other lignocelluloses, yet to be examined, it is

Compound Celluloses 105

found to be in excess of the half, as a result of oxidising actions.
This point should be borne in mind.

(3) Control observations.— (a) The chlorine in combina-
tion in the chlorinated fibre may be directly estimated by any
of the standard methods by which the * organic ' molecule is
broken down and the chlorine liberated as hydracid.

The chlorinated fibre itself is, for obvious reasons, some-
what difficult to deal with. The chlorinated product may be
dissolved by treatment with pure sodium hydrate, by which
treatment it is largely decomposed. To complete the isolation
of the chlorine as sodium chloride, the solution and washings
are boiled down to dryness, and heated for some time at 200-
300° C. An iron dish may be used for this treatment. The
soluble chloride is dissolved out and precipitated as silver
chloride, in presence of nitric acid.

(b) The cellulose may be isolated in the usual way, by boiling
the fibre-substance with sodium sulphite solution, and further
treating the cellulose as described, p. 95. The resulting solution
and washings of the cellulose may also be employed for the
estimation of the chlorine, the organic products being destroyed
by oxidation with nitric acid. Sufficient silver nitrate being
previously added, the oxidation may be carried out in an open
flask attached to an upright condenser.

The chlorination has also been studied in a different way, (i)
for the estimation of the total chlorine combining ; and (2) for
proving that no destructive oxidation takes place.

Weighed quantities of the fibre-substance were chlorinated (a)
after boiling in water, (V) after boiling in i p.ct. NaOH solution.
Duplicate specimens were weighed after these treatments and
without chlorination, and the statistics are worked out upon the
weights of the fibre- substance after treatment. After chlorination
the products were transferred to a bell jar containing an ample
supply of solid potassium hydrate ; the vessel was exhausted, and
the fibrous products left for some days. After a second similar
exposure in vacuo over solid KOH, and with addition of sulphuric

io6 Cellulose

acid (in separate vessels), the specimens were exposed for a short
time to a temperature of 100° in a water-oven and weighed. The
combined chlorine was then estimated. The following are the

results :

(<*) (3)

Weight of fibre-substance chlorinated . , 1-912 I '6 1 2

Combined chlorine determined . . .0-142 0-153

2-054 1765

Chlorinated fibre obtained . • • • 2-038 1763

Loss due to oxidation . . . .0*016 0-002

These results, in conjunction with independent observations of
the hydrochloric acid formed, further confirm the conclusion as to
the simplicity of the reaction.

The percentages of chlorine combining — viz. (a) 7-4, (&} 9-4 —
vary on either side of what may be taken as the mean number, viz.
8-0 p.ct. In both cases it was no doubt impossible, under the con-
ditions of the after treatment, to entirely remove the HC1. The
difference in favour of (a) shows the importance of the preliminary
treatment with the alkali ; without this the chlorination is incomplete.

Estimations of Secondary Constituents. — (a)

Methoxyl (O.CH3) is estimated by the now well-known and, in
fact, standard method of Zeisel. The fibre-substance is boiled
with hydriodic acid ; the resulting methyl iodide is washed in an
apparatus of special construction, to remove traces of hydriodic
acid, and passed into an alcoholic solution of silver nitrate,
with which it reacts to form silver iodide. A constant current
of carbonic anhydride is passed into the reaction flask and
through the entire apparatus, so that the methyl iodide may be
continuously carried forward and quantitatively decomposed.

The calculation from silver iodide to methoxyl is, of course,
simple (AgI=O.CH3).

This constituent of the lignocellulose molecule we have
every reason to regard as a characteristic constant, and its
determination is therefore of importance. Its constitutional
relationship will be discussed subsequently.

(b) The C0.C7/2 residue. — In a number of decompositions

Compound Celluloses 107

of the fibre-substance, acetic acid is formed. The maximum
yield is obtained in the process of decomposing by digestion
with dilute nitric acid. As it is also formed in considerable
quantity by dissolving the fibre in sulphuric acid, in the cold,
and obtained from the solution by diluting and distilling, it
must, in such case, be regarded as a product of hydrolysis, and
not of oxidation of the lignocellulose. The problem of its
constitutional relationship will be discussed in due course.
The estimation of the acid in the latter case need not be dealt
with, but it is necessary to describe the method by which it is
estimated after being liberated by the nitric acid treatment.
The weighed quantity of fibre is placed in a flask and treated
with 4 times its weight of 5 p.ct. nitric acid. It is digested for
5-6 hours at 90°, with the flask heated in a water-bath, and
attached to an upright condenser. The lignone being entirely
resolved, the acid solution is poured off, and the fibrous residue
washed with hot water. It is then digested with 5 p.ct. of its
weight (original fibre) of sodium carbonate dissolved in a small
quantity of water, which completes the removal of the lignone
derivatives from the residual cellulose. The solution and
washings are added to the original acid liquid. It is now
necessary to destroy the residues of nitric acid before distilling
for the volatile acid. A small quantity of sulphuric acid is
added to the liquid, in a flask, which is then digested for some
hours upon metallic iron. The solution is then boiled for
some time, with the flask attached to an upright condenser.
Urea is then added, and the solution distilled, taking care that
the sulphuric acid is present in slight excess. For the complete
removal of the acetic acid it is necessary to distil over as much
of the contents of the flask as possible, and to repeat the
distillation at least twice, adding a certain volume of water to
the flask, and taking over an equal volume of the distillate.
The distillate is then made up to a definite volume and

lo8 Cellulose

titrated. A portion is drawn off and tested for nitric and
nitrous acids. If free from these acids, the titration number
may be taken as representing the acetic acid. Otherwise a
fraction of the distillate must be further treated for the elimi-
nation of the nitrogen acids. For this purpose it is acidified
with sulphuric acid, and digested for some hours with a ' copper-
zinc couple.' The solution is then poured off, and distilled, as
before described, from a slight excess of sulphuric acid.

These determinations of acetic acid are destined to con-
tribute, in an important way, to the solution of constitutional
problems, and the student should master the details of th^
somewhat laborious process above described.

In certain cases, other volatile acids may be formed. It
is advisable to control the results by testing the distillate
qualitatively, and should there be indications of the presence
of other acids, e.g. formic acid, a fraction should be redistilled
from chromic acid, and the distillate again titrated.

The distillates also may be divided, a portion being titrated,
and the acid in a portion converted into silver salt in which
the silver is determined.

Having thus described in general terms the constituent
groups of the typical lignocelluloses, and more specially the
methods by which they may be quantitatively estimated,
directly or indirectly, it is necessary to point out that so far
nothing has been said as to the mode of union of these groups
in the fibre-substance. The evidence on this side of the subject
will be given in due course. It is sufficient, for the present, to
remember that we are dealing, on an empirical basis, with the
well-ascertained chemical constants of lignification — constant
for any given lignocellulose, but varying considerably from
member to member of this wide and varied group of plant
constituents. It may not be out of place also to insert at this
point a caution against the possible inference that the above

Compound Celluloses log

groups are sharply separated from one another. As the
student becomes more familiar with the subject, he will find
the constituent groups 'overlapping' in an unmistakable
manner, with suggestions of probable genetic connections.

Keeping, however, for the present, to the strictly empirical
view, we proceed to the systematic account of the jute fibre as
the typical lignocellulose.

The JUTE FIBRE is the isolated bast of plants of the species
Corchorus (Order Tiliaceae), an annual of rapid growth, usually
attaining a height of 10-12 ft. in the few months required,
in the Indian climate, for the maturing of the plant. This
great length of stem is attained without branching, and the
separation of the bast from the wood and cortex is a manual
operation of the simplest kind. The plants, after being cut
down, are steeped or retted for a short period in stagnant
water ; the stems are then handled individually ; the wood
being broken, the bast is easily stripped and freed by washing
from the softened cellular cortex. The fibre is supplied to
commerce in long lengths, or strands, representing nearly the
full length of the parent stem. As, however, the lower portion,
6-8 ins. from the root upwards, is more or less reticulated, this
is usually cut off, and these rejections constitute the jute 'butts'
or ' cuttings ' largely used as the raw material for special
classes of wrapping papers. The textile fibre is of a brown to
silver-grey colour in the finer sorts. The individual fibres, or
spinning elements (filaments), are complex structures ; in cross
section they are seen to be bundles of the ultimate fibres, the
number of which varies from 7 to 20. The ultimate fibre itself is
of short length, 2-3 mm. ; it is of circular or polygonal section,
with a central canal sometimes nearly obliterated, from the
thickening of the cell wall. These bast fibres taper off at their
extremities, and are built up by apposition to form the complex
filament or bundle.

no

Cellulose

The fibres or filaments are somewhat matted together in
the strands by reason of the great pressure under which
the bales are packed, and also in part owing to the presence, in
the tissue, of mucilaginous or pectic bodies (parenchymatous
residues £c). Jute requires, therefore, a softening treatment as
a preliminary to the preparing operations of the spinner. It is
opened out from the bales, dusted, and passed through a series
of heavy fluted * breaking' rollers, being simultaneously sprinkled
with water and whale-oil. By this treatment the subdivision
and drawing of the fibres in the hackling, or combing, and
spinning processes is greatly facilitated. For the purposes of
laboratory investigation the fibre may be freed from adventitious
impurities by boiling in weak solutions of sodium carbonate,
washing well to remove soluble matters, and rubbing well in a
stream of water, to remove residues of cortical parenchyma.

The bast fibre thus obtained is somewhat harsh to the touch,
coloured as described, more or less, and having a certain amount
of lustre.

Its specific gravity is 1*436 (Pfuhl), 1*587 after purification
by boiling in alkaline solutions (Cross and Bevan).

The following resulis of proximate analyses of various specimens
are given by Hugo Miiller, Pflanzenfaser, p. 59.

—

Long fibre

Brown
cuttings

Nearly colour-
less specimen

Fawn
coloured

Ash.
Water .
Aqueous extract
Fat and wax . .
Cellulose .
Incrusting substances and |
pectic constituents. Dif- I
ference from 100 . J

0-68

9 '93
1-03
0-39
64-24

24-41

9-64
I-63
0-32
63-05

25-36

12-58

3'94
0-45

6174
21-29

To compare these results — chiefly for cellulose — with the

Compound Celluloses III

authors' results given in the text, the numbers must be calculated
to dry substance (i.e. multiplied by ri).

The authors have not isolated pectic acid from the fibre-
substance proper ; but jute * cuttings ' often contain a considerable
quantity.

Composition. — The inorganic constituents amount to
from 0*8-2-0 p.ct., and are obtained, on burning the fibre,
as a brownish-coloured ash, of which the preponderating
constituents are silica (35 p.ct.), lime (CaO 15 p.ct.), and
phosphoric acid (P2O5 1 1 p.ct.). Manganese is usually present
in small quantity (Mn3O4 075 p.ct.).

The organic portion, or fibre substance proper, varies some-
what in composition, the subjoined numbers representing the
mean range of variations :

C 46 -0-47-0 p.ct

H . . . . 6-3-5-8 „

In dealing with the jute fibre substance in contradistinction to
the jute fibre, the results are referred to the substance taken as dry
(100°) and when the result would be seriously influenced, as ash-
free.

For * statistical ' purposes, therefore, the fibre-substance may
be represented by the empirical formula C12H18O9. There
is plenty of evidence for the view that lignification is an in-
trinsic process of chemical change of cellulose, and it might
therefore be inferred that the process is one of dehydration :
Ci2^2oOio — H2O = C^HjgOg.

As an illustration of the superficial meaning of such
numerical relationships, we may cite here the results obtained
by A. Pears in cultivating the jute plant under the more
artificial conditions of growth in a ' hot house ' in this country.
A normal growth of the plant was secured, in the sense
that the seed saved gave a satisfactorily high proportion of
germination in the second year of cultivation, and from both
cultivations good specimens of the bast fibre were separated in

112

Cellulose

the usual way. The composition of these specimens was de-
termined as follows :

Fibre grown in 189* Fibre grown in 1893

C 43'0 43*5

H . . . 6-1

In further illustration of the results obtained in these ' artificial '
cultivations of the fibre, we reproduce the various numerical

Constituents and
reactions

Method

Jute pro-
duced in
England
(1892)

Normal
fibre

—

Moisture .

Drying at 1 00°

II '4

10-3

— _

Inorganic con- 1
stituents j

Ash

1-6

1-2

—

,1 p.ct. solution)

Alkali hydrolysis

NaOH
(i) 10 mins. j
boiling . J

14-8

8-0

i Loss of
t weight

(2) 60 mins. )
* boiling . i

20 'O

18-0

ii

,2O p.ct. solu-v

' Mercerisation '

J tion NaOH in L

I2'2

8-0

M

I the cold . )

Nitric acid reso-
lution . .

5 p.ct. \
J HNOs.Aq »

* 8 hours at 70°

37-0

37 -o

r Residue
1 oxycellulose

Cellulose .

Chlorination &c.

75*2

75*0

Chlorine absorp-
tion

J. Chem. Soc. \
55, 199 • '

137

16-6

—

Iodine absorp-

/ Excess of gg \
] normal solu- j-

6-0

6-0

tion . .

I tion in KI . J

/ Equal volumes \

Nitration .

ofHNO,i-5;l

130-0

145

__

1 H2SO4 i -82 . )

. Increase of

Ferric ferricyan-
ide reaction .

J. Soc. Chem. i
Ind. 1893 • >

133-0

124

j weight un-
1 der equal

^ conditions

Thiocarbonate
reaction

F. Chem. Soc. i
1893,837 • i

45-c

45'0

/P.ct. of
j fibre un-
( dissolved

Carbon percent-
age .

/Com bus tion\
\ with CrO3 and [
1 H2S04 . 1

43-0

46-5

—

Compound Celluloses II?

determinations given in the original paper (J. Chem. Soc. 1893,

967).

The preceding table contains the results of a more extended
scheme of investigation than is required for special and more
practical purposes. The results, however, all have the value of
* constants,' depending as they do upon definite properties of the
fibre-substance. It is an amplification of the scheme adopted by
Webster (J. Chem. Soc. 43, 23), working in collaboration with the
authors ; and, again, of that given by the authors in the Reports
Col. and Indian Exhibition, 1886.

For the fibre grown in 1893, fr°m the seed saved from the above,
the following constants were determined (J. Chem. Soc. 1894,471):

C H

Elementary composition • • t 43-5 6'O

Furfural yield 8-55 p. ct.

Cl absorption , 15-0 „

The chief feature of these results is the preservation of the
essential constitutional features of the lignocellulose with such con-
siderable variation from the normal in elementary composition.

Notwithstanding this wide divergence in composition, the fibre-
substance showed all the essential characteristics of constitu-
tion of the ordinary product. The observed difference is,
therefore, in the main associated with hydration ; and lignifica-
tion is evidently a process which- is independent of dehydrating
conditions.

The lignocelluloses, however, under normal conditions of
growth are progressively dehydrated, and in nearly all cases,
therefore, are characterised by high carbon percentage (46-51).
These considerations lead up to the general question of the
relationships of the jute fibre to water.

Jute Fibre and Water.— Lignocellulose Hydrates.
The hygroscopic moisture of ordinary jute varies, under normal
atmospheric conditions, from 9-12 p.ct, the variation being,
of course, chiefly dependent upon temperature and ' dew point,'
or rather the percentage saturation of the air with aqueous vapour.

1 14 Cellulose

In an atmosphere saturated at ordinary temperatures jute takes
up 23 p.ct. of moisture.

The hydration of the fibre-substance, in the more permanent
sense of definite combination with H2O molecules, is determined
under conditions which will appear in the succeeding sections
of the subject.

Solutions of Lignocellulose. — The jute fibre is attacked
and dissolved by the solvents already described under * Cellulose '
(p. 8), viz. :

(1) Zinc chloride — concentrated aqueous solution.

(2) Zinc chloride — solution in HC1 ; and

(3) Cuprammoniun solutions.

From these solutions the lignocellulose is precipitated, on
dilution (i and 2) or acidification, as a gelatinous hydrate ; the
precipitation is, however, incomplete — the proportion remaining
in solution varying from 15-25 p.ct, according to the condi-
tions of solution. There is, however, no difference in reactions
between the soluble and insoluble fractions, and on ultimate
analysis both are found to have the empirical composition of
the original fibre-substance. Although, therefore, the ligno-
cellulose is a complex of various groupings, it behaves in this
respect as a homogeneous product, and the bond uniting the
groups together is not resolved by simple hydrolytic agencies
(see infra, p. 134).

In the case of the ZnCl2.HCl reagent the fibre- substance is
progressively hydrolysed on standing. This is illustrated by the
following determinations of the proportion of the lignocellulose
reprecipitated from such solution.

(a) Precipitated at once : ppt. 78-4 p.ct. of the original
(£) After standing 16 hours : ppt. 29-4 p.ct. of the original

Qualitative Reactions and Identification of the
Lignocelluloses. — Whereas the reactions of the celluloses
ure mostly negative, jute (and the lignocelluloses generally) is

Compound Celluloses 115

distinguished by a number of characteristic reactions. In
addition to those already described as admitting of quantitative
estimations, the following are the more important :

(1) Salts of aniline (and many of the aromatic bases), in
aqueous solution, colour the fibre a deep golden yellow.

(2) The Coal-tar dyes generally combine freely with the ligno-
celluloses. In ' staining ' sections of plants and parts of plants
for microscopic observation, the lignocelluloses are dyed by
the majority of soluble coal-tar dyes. Their 'affinities' for
colouring matters are, in fact, similar to those of the animal
fibres, silk and wool, although differing radically from them,
not only in constitution, but in containing no nitrogen (NH2
groups).

(3) Phloroglucinol) in hydrochloric acid, gives the deep
magenta colouration characteristic of the pentaglucoses. The
reagent is prepared by dissolving the phenol to saturation in
HClAq (1-06 sp.gr.).

(4) Iodine is absorbed from its solutions in potassium
iodide in large quantity, colouring the fibre a deep brown.

(5) Chlorine combines with the fibre with avidity, as already
described ; the chlorination is made evident by treatment
with sodium sulphite solution, which develops a deep magenta
colouration. This reaction is characteristic. Conversely, the
fibre-substance may be employed as a reagent for the identifi-
cation of chlorine, or may, in certain cases, be used for absorb-
ing the gas.

(6) Ferric chloride colours the fibre-substance to a dark
greenish tint — the reaction being due to traces of tannins.

(7) Ferric ferricyanide — the red solution obtained by mixing
together ferric chloride and potassium ferricyanide in equivalent
proportions — gives a highly characteristic reaction (subsequently
described in detail), the fibre-substance rapidly decomposing
the compound to ' Prussian blue,' the pigment being taken up

I 2

n6 Cellulose

by the fibre-substance in very large quantity (50 p.ct. of its
weight).

(8) Chromic add, in aqueous solution, combines with the
lignocellulose, and is then very slowly reduced to the inter-
mediate oxide (CrO3.Cr2O3).

(9) Potassium permanganate is rapidly reduced, the MnO2
produced colouring the fibre a deep brown. After treatment
with sulphurous acid, which removes the oxide, the lignocellu-
lose will be found to have been bleached by the treatment.
On repeating this treatment once or twice, with dilute solution
of KMnO4, the lignocellulose is obtained of a cream or greyish-
white colour, the loss of weight sustained in the bleaching
being small (2-4 p.ct.).

Compounds of Jute Lignocellulose.— The fibre-
substance itself being a complex or compound cellulose, and
susceptible of decomposition (a) by hydrolytic treatment — in
which, however, the union of the constituent groups is pre-
served—and (ft) by reagents which selectively attack the con-
stituent groups, it is obvious that we are limited in the prepa-
ration of compounds which may be regarded as compounds of
the lignocellulose molecule as a whole. We shall first describe
those which result from reactions of the OH groups of the
lignocellulose. These are more active than in the celluloses.
We have already pointed out that the fibre combines freely
with colouring matters. The phenomena of dyeing being
now well established, as the result of interaction of salt-forming
groups in fibre- substance and colouring matter, i.e. a species of
' double salt ' formation, we may deduce from the considerable
and very general ' affinity ' of the lignocelluloses for the coal-
tar colouring matters, that they contain OH groups of both
acid and basic function, and much more disposed to reaction
than those of the celluloses.

Absorption of acids and alkalis from dilute aqueous solutions. —

Compound Celluloses 117

This phenomenon, already described as a property of the cellu-
loses, is more pronounced with the jute fibre. The following
absorptions have been determined by the authors from normal
solutions of the respective reagents :

Normal hydrochloric acid. — (a) Fibre digested with 8
times its weight of solution, 10 minutes at 15° C. ; (b) with 20
times its weight.

(a) ' (b)

HC1 absorbed . 0-85 i-i p.ct. on fibre-substance

Normal sodium hydrate. — Fibre digested with 20 times
its weight of solution.

(«) (*)
Na,jO absorbed . .3-0 3-6 p.ct. on fibre-substance

It is to be noted that the molecular ratio of the absorptions
is approximately that observed in the case of cotton, viz.
3HC1 : 10 NaOH.

The hydrolysing action of the alkalis (a) and non-oxidising
acids (b) may be regarded as an extension of this phenomenon.
(a) The alkalis and alkaline compounds in aqueous solution
attack the fibre-substance in the ratio of their hydrolysing and
saponifying activity ; and, as in the action of the solvents pre-
viously described, the lignocellulose is attacked as a whole.
In the systematic comparison of the vegetable fibres (i.e. com-
pound celluloses) it is important to determine their relative
resistance to alkaline treatments under standard conditions.
It is usual, for this purpose, to determine the loss of weight
sustained by the fibre on boiling with a i p.ct. solution of
sodium hydrate (i) 5 minutes, (ii) 60 minutes. Under this
treatment jute loses on the average —

(i) (M)

8-0 15-0

No change in the composition of the fibre- substance is
occasioned by the treatment, the portion dissolved showing the
essential characteristics of the original fibre. It is precipitated

1 1 8 Cellulose

in part on acidifying the solution, and the gelatinous precipitate
gives the characteristic reactions of the original fibre. The
fibre is further but slightly affected in structure and physical
properties by the treatment.

At temperatures considerably above the boiling point the
action of dilute solutions of the alkaline hydroxide (1-3 p.ct.
Na2O) takes a different course ; the non-cellulose is attacked
and converted into soluble derivatives, and the fibre-bundles
are more or less disintegrated. Such processes are, in fact, used
on the large scale for the preparation of paper-making pulp
(cellulose) from the lignocelluloses.

(b) On digesting the fibre with dilute solutions of the mineral
acids at 60-80°, the lignocellulose is again progressively
dissolved, the loss of weight sustained by the fibre being
proportional to the hydrolysing activity of the acid, and to the
conditions of the digestion. In this case also the lignocellulose
is attacked as a whole, the insoluble fibrous residue preserv-
ing the essential characteristics of the original fibre. The dis-
solved portion may be isolated — when sulphuric acid (5-7 p.ct.
H2SO4) is used as the hydrolysing acid — by neutralising the solu-
tion with barium carbonate, filtering, and evaporating to dryness.
The soluble modification of the fibre-substance is obtained as
an amorphous, brown, gummy solid, having the same empirical
composition as the original fibre.

On prolonged digestion with the dilute acids (5-7 p.ct.
H2SO4) the loss of weight sustained by the fibre approximates to
a limit at about 30 p.ct. As a result of the treatment, the fibre
is disintegrated, the residue being obtained as a mass of brittle
fragments. It is to be noted that the disintegration is not a
progressive dissection of the ultimate fibres — such as results
from the alkaline digestions above described — but is the result
of a change in the physical properties of the fibre- substance

Compound Celluloses 119

itself; the disintegration which ensues is characterised by
fracture of the fibre-bundles or filaments.

The following results of particular experiments may be cited :

Fibre digested with 7 p.ct. H2SO4

(1) 18 hrs. at 60-80° : loss of weight, 12-0 p.ct

(2) 12 hrs. at 80-90° : „ „ 9-7 „

(3) 42 hrs. at 80-90° : „ „ 23-0 „ '

The investigation of the products from (3) gave the following
results :

(a) Soluble. — Isolated by neutralising with BaCOs. Evapora-
tion, solution of residue in alcohol, evaporating solvent and
drying at 105° gave, on combustion :

Calc. CiaHl§0.

C 46-29 46-08 47*05

H 575 5-95 5 "88

This substance gave with Cl the characteristic quinone chloride,
and on boiling with hydrochloric acid, furfural.

On adding phenylhydrazine acetate to the concentrated solu-
tion of the product and heating at 90-100°, an osazone is formed:
it separates as a coagulum of characteristic greenish-yellow colour.
After washing and drying, the product may be purified by solution
in toluene ; from which solution satisfactory crystallisations are
obtained. A series of these compounds has been obtained with
melting points ranging from 110-130°, and containing from 9-10
p.ct. nitrogen. Their relationship to the original fibre-substance
has not yet been determined.

(b) Insoluble. — The brittle fibrous residue gave with chlorine the
characteristic reaction, and the cellulose isolated in the usual way
amounted to 75 p.ct. of the weight of the product.

The action of the acids proceeds, as stated, to a limit
which is determined by the concurrent effects of condensation.
If the fibre be then washed and boiled for a short time in
alkaline solution, it is again rendered susceptible of attack by
the hydrolysing acids, with further conversion into soluble
derivatives.

If the acid solutions are boiled, the dissolved product is

120 Cellulose

decomposed with formation of furfural and acetic acid ; when
formed at temperatures below 70°, it may be regarded as a
soluble hydrate of the lignocellulose, or more correctly a
derivative of low molecular weight in which the characteristic
groupings of the parent molecule are preserved.

Concentrated solutions of the alkaline hydrates. — The struc-
tural changes produced in the fibre by treatment with solu-
tions of caustic soda of 'mercerising' strength (15-30 p.ct
NaOH) are remarkable. The fibre-bundles are resolved more
or less ; the cell wall of the individual fibres undergoes con-
siderable thickening, such that the central canal is almost
obliterated. The visible effects of these changes of minute
structure are (i) a shrinkage in length of the strands of fibre
(15-20 p.ct.) ; (2) a considerable refinement of the spinning
units or filaments ; (3) the filaments have a wavy or crinkled
outline, resembling that of wool.

The following experimental results may be cited : 80-35 grms.
fibre (air-dry) treated with 300 c.c. of 25 p.ct. solution NaOH,
six hours in the cold ; washed, acidified, washed and dried ;
weighed, air-dry, 75-5 grms. Loss of weight, 6 p.ct. Shrinkage in
length, from 4 ft. to 3 ft. 8 in., i.e. 17 p.ct.

An extended series of experiments upon normal specimens of
the raw fibre, with varied conditions — e.g. concentration of alkali,
J5-33 P'ct« NaOH ; duration of treatment, 5 minutes to 48 hours —
showed an average loss of weight of 7-5 p.ct., with slight vari-
ations only on either side of this mean number. The same speci-
men of jute lost 1 1 -9 p.ct. in weight on digesting 48 hours in more
dilute alkali (6-5 p.ct. NaOH), and n p.ct on boiling for 5
minutes in alkali of the same concentration.

The cellulose constants of the fibre are unaffected by the treat-
ment.

The chemical changes are more complex than with the
celluloses, for reasons which will appear when the constitutional
relationships of the constituent groups of the fibre-substance

Compound Celluloses 121

are discussed. Empirically the results of the treatment are as
follows :

A certain proportion of the lignocellulose is dissolved ; but
the dissolved portion, as well as the fibrous residue, gives the
characteristic reactions of the original. When the latter is
chlorinated, and the cellulose isolated in the usual way, the
percentage yield is found to be unaffected by the treatment.
The character of the cellulose is somewhat altered, however, as
it is obtained in continuous strands ; and when dried, the
filaments of jute cellulose have a certain amount of coherence.
So far there is a general resemblance to the changes produced
in cotton cellulose on mercerising, i.e. the effects are chiefly
hydration changes. The differences, on the other hand, are
brought into evidence when the alkali-lignocellulose is exposed
to the action of carbon disulphide.

The thiocarbonate reaction which ensues is of a remark-
able character. The lignocellulose is gelatinised more or less
in the reaction, but on treatment with water it is not dissolved
to a homogeneous solution, but swells up enormously, the
hydration proceeding to almost indefinite limits. The following
results of an experiment may be cited :

4*5 grms. fibre, treated with excess of 12 p.ct. solution of
NaOH, squeezed, and exposed to CS2 (2-0 grms.) in a stoppered
bottle 24 hours. On treatment with water, the gelatinised fibre
occupied a volume of 300 c.c. ; and for separation of the un-
dissolved fibre, dilution to 750 c.c. was necessary. The
following determinations were made :

(a) Undissolved fibre . . . . . .427

(b) Dissolved — reprecipitated by HC1 • . • 43*3
(b1) Soluble after acidification ... . . . 14*0

(a) Gave the reactions of the original fibre-substance.

(b} Gave only a slight reaction with Cl and Na2SO3.
(b") Consisted mainly of the furfural-yielding constituent.

The following results of particular experiments are of interest :

122 Cellulose

(1) Ten grms. raw jute (with 10 p.ct. moisture), purified by
boiling in dilute solution Na.2CO3 ; washed, squeezed, and placed
in bottle with 4 grms. CS2. When evenly diffused, treated with 25
c.c. of 15 p.ct. NaOH and left 48 hours.

Insoluble product (after purifying), 7*053 grms. (dry) ; yield on
9 grms. (dry), 78-4 p.ct. Gave 4*1 p.ct. furfural on distillation
(HC1).

(2) Conditions exactly as in (i), with which it was comparative
in regard to effect of varying the reaction ; viz. in this case the jute
was first treated with the NaOH, and afterwards sealed up with
4 grms. CS2.

Insoluble product, 7*004 grms., 77*8 p.ct. The variation in
question was therefore without effect. The filtrate from the in-
soluble residue was treated with zinc acetate in excess, which has
been found to precipitate the dissolved fibre-constituents. The
precipitate was then decomposed with HC1 in excess, and the now
insoluble fibre-products washed, purified, dried, and weighed.
Weight, 0*980 grm. On distillation with HC1 this gave 0*031 grm.
furfural ; the filtrate from this insoluble product gave none.

The undissolved fibre was chlorinated, and the cellulose sepa-
rated in the usual way : 5728 grms. obtained, i.e. 81*8 p.ct. on pro-
duct, or 63 p.ct. of original lignocellulose. The results are as
follows :

The furfural-yielding groups have been attacked ; the total yield
of the aldehyde is reduced by 50 p.ct., the reduction falling chiefly
upon the portion hydrolysed and dissolved. The a- cellulose is un-
affected, the keto R. hexene groups also. The portion dissolved
appears to be the /3-cellulose and the furfural-yielding constituent
of the lignone.

(3) The above conditions were maintained, varying the duration
of action of the alkali ; it was left 48 hours before adding the CS2.
The results were :

Insoluble fibre, 6-82 grms. =75*8 p.ct.

Giving cellulose (after Cl &c.), 57^ grms. =64 p.ct. of the original.

Under constant conditions as regards the reagents, the results
are therefore independent of the mode of carrying out the reaction.
The reaction requires further investigation, as it appears capable
of throwing light upon the actual mode of union of the constituent
groups in the lignocellulose.

Compound Celluloses 123

The results obtained are, however, by no means constant,
but vary considerably with variations in the conditions of
treatment. The causes of these variations lie in the complex
character of the lignocellulose. In the celluloses alcoholic
characteristics predominate ; in the lignocelluloses the presence
of phenolic OH groups, of CH2.CO and CH2.CO.O groups,
and of ketonic oxygen (see p. 137) must largely modify
the alcoholic functions of the cellulosic OH groups. They
are, in fact, in condensed union with these more negative
groups, and this union is only partially resolved in certain
directions, and further cemented in others, by the alkaline
treatment. It is probable that the alkali may have the
effect of further condensing or synthesising the aldehydic
groups.

It is evident, on the other hand, that the entire molecule
is opened up for the entrance of water molecules, and the
prominent result of the reaction is the consequent hydration of
the lignocellulose. It is this aspect which leads us to describe
the reaction under the heading of the ' Compounds of the
Lignocellulose.' The attendant hydrolysis is a secondary
result which will be referred to subsequently.

Compounds of the Lignocellulose with Metallic
Salts. — We are still dealing with those synthetical reactions of
the lignocellulose (OH groups) which take place in presence
of water. The resulting compounds are necessarily of a
variable and ill-defined character, owing to the complex nature
of the lignocellulose, the tendency to selective reaction
with its constituent groups, and to consequent partial hydro-
lysis.

The combinations already described as taking place with
the alkalis and non-oxidising acids are seen to be of a feeble
and transitory character. With many of the salts of the heavy
metals the reactions are more pronounced, but they are also

124 Cellulose

of too indefinite a character to require more than a passing
notice. Generally we may regard reaction as occurring only
with such as undergo pronounced dissociation on solution in
water, which dissociation is exaggerated by the fibre-substance.
The lignocelluloses present merely a particular case of the
general theory of the action of 'mordants,' its combinations
with the 'mordanting' oxides being similar to those of the
cellulose, differing only in the higher proportions of the oxides
taken up.

There is one reaction, however, of a specific character,
already alluded to, which merits description in detail — that is,
the interaction of the lignocellulose and ferric ferricyanide.

This reaction has been described in a paper by the authors
in the J. Soc. Chem. Ind. 1893, the experimental portion of
which is now reproduced, with certain alterations of minor
import.

A REACTION OF THE LIGNOCELLULOSES AND THE THEORY
OF DYEING. — The red solution obtained by mixing aqueous
solutions of ferric chloride and an alkaline ferricyanide, which
may be regarded as containing ferric ferricyanide, reacts as is
well known with the more easily oxidisable * organic ' com-
pounds, oxidising them and being itself reduced to the lower
mixed cyanides, i.e. Prussian blue and similar compounds
(Watts' Diet. ii. 248). The reactions of this solution with
the lignocelluloses, and notably jute, are remarkable and cha-
racteristic. Not only is the conversion into the blue pigment
very considerable in proportion to the weight of fibre- substance,
but the colouring matter is deposited within the fibre in such
a way as to give the effect of a homogeneous dye. Thus if
this particular fibre, suitably purified by previously boiling in
dilute alkaline solutions and washing, be plunged into a \ deci-
normal solution of the reagent (prepared by mixing decinormal
solutions of the reagents in equal volumes) the fibre-substance

Compound Celluloses

125

rapidly dyes to an intense blue- black with a gain of weight
of from 20-50 p.ct. ; and the dyed fibre examined under
the microscope is of an intense transparent blue with all the
characteristics, that is to say, of a ' solid solution ' of the
colouring matter.

The following is a brief account of the results of quantita-
tive observations. In our first series of observations we used
a solution of ferric chloride containing Fe2Cl6, equivalent to
0-0976 Fe2O3 in 10 c.c., and an equivalent solution of ferri-
cyanide. Five equal portions, each weighing 2762 grms., of
jute fibre were treated with the mixture of the above solutions
in equal volumes, in the proportions and with the results given
in the subjoined table.

Vols. of solution

Increase
of
weight

Increase
per cent,
of fibre

Fe2O,

added as
FeaCN.

Fe30.
fixed as
blue
cyanide

Fen ic
chloride

Ferri-
cyanide

Cc.

Cc.

10
20

IO
20

0-489

67

177

0-195
0-390

OT23
0-230

30

30

0768

27-8

0-58S

0-373

40

40

I'025

37'i

0-781

0-510

50

SO

I-l6l

42-0

0-976

0-601

It will be noticed that the Fe2O3 fixed by the fibre is pro-
portional to the quantity taken, viz. approximately two-thirds
in each experiment ; but the corresponding increase in weight
of the fibre due to the blue cyanide fixed is somewhat variable.
The cyanide, in fact, is shown by analysis to vary in composi-
tion slightly in the ratio Fe : CN, more considerably no doubt
in its condition of hydration. The ash left on ignition of the
dyed fibre we find to contain no soluble basic constituents,
therefore no K2O appears to be fixed.

Analysis of blue cyanide fixed by the fibre. — A specimen
of fibre, dyed blue under the above conditions and weighted

126 Cellulose

to approximately 20 p.ct., was analysed, the total Fe being
determined as Fe2O3, the CN as NH3.

Fe2O3 = 6-i p.ct. = 4-27 Fe; N = 3'i5 p.ct. = 5*85 CN.
Whence the ratio—

or—

Fe:CN = i :3.

A portion of the fibre was further treated until the increase
of weight amounted to 40 p.ct., and then analysed with the
following result :

Fe2O3 = 14-0 p.ct. = 9-8 Fe N = 9-3 p.ct. CN.
Fe:CN = g: 0 = 0-175: 0-357 = 1: 2.

By the continued interaction of the fibre- substance and the
ferric ferricyanide the Fe'" appears to be deoxidised, and in
exhausting the fibre-substance with dilute alkali in the cold,
ferrocyanide is dissolved, as of course is to be expected.

It is obvious that we are dealing with an aggregate and
the product of a mixed reaction ; the ratios Fe2(CN)6 and
Fe3(CN)6 determined as above for the product in successive
stages are therefore not to be taken as more than approximate
indications of the composition of the colouring matter deposited
in the fibre.

The mechanism of the reaction and the question of the
composition of the resulting blue cyanide are further elucidated
by the following experiments :

Equivalent solutions were prepared as above, and in two
experiments with equal weights of fibre the solutions were
mixed in the proportions :

A. 3 of ferric chloride to 2 of ferricyanide ;

B. 2 „ „ 3 ..

Compound Celluloses 127

The results observed, which were somewhat unexpected,
are given below.

Increase of
weight in
fibre

Analysis of dyed product

Fe,0,

N

Ratio Fe : CN

P.ct.
20
17

P.ct.
9'6
8-4

P.ct.
4-8
4-0

Fe2(CN)6
Fe4(CN)n

The N in these analyses was determined by the Kjeldahl
method, in the previous cases by the soda-lime combustion.
It appears, therefore, that the blue cyanide deposited in the
earlier stages of the reaction is fairly constant in composition
notwithstanding considerable variations in the proportions of
the reagents in the solution from which it is abstracted by the
fibre ; and the ratio of Fe to CN in the fibre cyanide complex
is approximately 1:3. By the last-cited experiment it is also
shown that the rearrangement of the Fe and (CN) which takes
place within the fibre-substance is to a certain extent inde-
pendent of the condition of the Fe in the solution, i.e.
whether added as basic Fe or as in the ferricyanogen complex.

The main points of the reaction have been sufficiently
elucidated by the results described, and it is unnecessary to
put on record a large number of quantitative results which
merely confirm those already given. It should be noted, how-
ever, that the limit of the reaction has not been investigated :
on occasions we obtained increases of weight of 50 and even
80 p.ct., but at this degree of loading the natural lustre of
the fibre had given way to the dull and dusty look of a fibre
weighted to excess.

We have now to consider the mechanism of the reaction
which has been in some measure elucidated by further experi-
ments.

To explain it as the result of reduction of the ferric iron

128

Cellulose

either of the chloride or the ferricyanide by the fibre-substance
will be found to be inadequate. The following experiments
show that either reagent taken singly is but slightly affected by
prolonged contact with the fibre-substance.

Equal weights (2-835 grms-) of the purified fibre were
placed —

(a) In a solution (30 c.c.) of FeCl3— r6 grm. per 100 c.c.

(b) In a solution (30 c.c.) of K3FeCy6 — 3-3 grms. per
100 c.c.

(c) In a solution prepared by mixing the above (30 c.c. of
each).

After standing some hours (a) and (b) were squeezed and
interchanged, and left some minutes. The fibre from each
was then washed off, dried, and weighed with the following
results :

—

Weight of dyed fibre

Increase of weight

Colour

«)
*)
4

2-931
2-846

3-550

Ret.
3'3
0-3
25-2

Indigo-blue
Medium-blue
Blue-black

It appears, therefore, that the lignocelluloses absorb but
little1 oxide from a neutral solution of ferric chloride, and
there is only partial reduction of the oxide so fixed : and also
that ferricyanide is slightly reduced by the fibre-substance in
neutral solution and without sensible combination with the
ferricyanogen or ferrocyanogen group.

The reaction in question is therefore specific as between
the ferric ferricyanide and the fibre- substance.

That the formation and fixation of the blue product is not
the result of reduction in the liquid is further shown by the fact

1 The maximum we have observed to be taken up from a normal solu-
tion of the chloride is 0-4 p.ct., and that after 48 hours' immersion.

Compound Celluloses 129

that it is not appreciably affected by the presence of oxidising
agents such as chromic acid.

The alternative conclusion is that it is due to a coagulation
or precipitation of the ferric ferricyanide by the fibre-substance,
in the first instance, followed by a rearrangement of its con-
stituents by specific combination with the fibre constituents.

Collateral evidence in support of this view is afforded by
the behaviour of the ferricyanide with another typical colloid,
viz. gelatin. Solutions of gelatin give with the ferric ferri-
cyanide a voluminous coagulum of a greenish colour, and the
reaction is approximately a quantitative one, but of course
depending to some extent upon the conditions of precipitation.
The following relations were determined :

One series of experiments —

A white gelatin (containing 16-5 p.ct. hygroscopic mois-
ture and 2-86 p.ct. ash constituents) was weighed out in
quantities of 2 grms. (=1-613 dry and ash-free gelatin) and
dissolved. The solutions were variously diluted and treated
with half decinormal solution of the ferric ferricyanide added
from a burette.

(1) To completely precipitate in the cold 23 c.c. were
required.

(2) To completely precipitate at 50° C. 24*5 were required.

(3) To a third quantity 32 c.c. of the ferricyanide solution
were added.

The precipitates were collected, dried, and weighed ; the
weights were: (i) 1*990; (2) 2-031; and (3) 2*077 respec-
tively. The mean weight 2-032 shows an increase of weight
of 2 1 '6 p.ct. An estimation of Fe2O3 in the product gave
6-0 p.ct., the proportion being rather lower than in the
case of the fibre-cyanide product, which, with the same gain
in weight, contains 7 p.ct. Fe^Og. The coagulum is in this
case, however, only slightly blue, darkening gradually or*

K

130 Cellulose

standing. On treatment with a reducing agent such as dilute
sulphurous acid, the coagulum swells to a deep blue transparent
jelly.

This interaction of gelatin and ferric ferricyanide is there-
fore rather of the character of a simple coagulation or combina-
tion of colloids by dehydration.

We have reason for assuming a similar relationship between
the fibre-colloid and the ferricyanide as the first cause of the
precipitation. The conversion of the colourless into the
coloured cyanide may be then accounted for by what we know
of the constitution of the fibre-substance.

We have in this complex all the conditions (i) for a
deoxidation of Fe'", (2) for union with ferric and ferrous
oxides, and (3) combination with HCN.

Such changes as would be determined by these relations,
when brought into play, are of the minor order and consistent
with the characteristics of the product, i.e. an intimate mole-
cular union of the complex fibre-substance, slightly oxidised at
the expense of ferric oxide, and the ferroso-ferric cyanide.

Further investigation has confirmed the interpretation
given in the communication which is reproduced above. It is
evidently a reaction in which the entire fibre -substance takes
part. The ferroso-ferric cyanide being a saline compound, and
the lignocellulose containing both acid and basic groupings in
combination, and being in that sense analogous to the inorganic
salts, the reaction may be regarded as in the main a species of
double-salt formation. In this respect it stands on the same
footing as the majority of dyeing reactions. But in the forma-
tion of the blue ferroso-ferric cyanide from the red ferricyanide
the special chemistry of the lignocellulose comes into play.
It may be fairly assumed that the deoxidation of the ferric
oxide is due to aldehydic groups, the fixation of hydro-
cyanic acid may be referred to aldehydic and ketonic oxygen,

Compound Celluloses 131

the fixation of ferroso-ferric oxide more particularly to the
quinone or keto R. hexene groups of the non-cellulose ; the
resulting combinations being, however, rather of a ' molecular '
character, they reunite to form the coloured lake or 'double salt '
represented by the dyed fibre. Without reference, however, to
explanations of the mechanism of the reactions, which are for
the present more or less hypothetical, the following are the
facts to be emphasised in conclusion :

(1) It is a reaction in which the lignocellulose manifests
itself as a homogeneous compound.

(2) It is unique in the range of dyeing phenomena both in
regard to the formation of the colouring matter by definite
chemical reaction, and the very large proportion in which it is
fixed by the fibre-substance.

This aspect of the reaction will be found discussed in the origi-
nal paper (loc. cit.\ and in a second communication on the sub-
ject (J. Soc. Chem. Ind., April 1894) in reply to criticisms by
C. O. Weber (ibid., March 1894).

Compounds of Lignocellulose with Negative
Radicals. — (a) Lignocellulose esters. — (i) Benzoates. — The
interaction of the lignocelluloses with the alkaline hydrates
and benzoyl chloride has been only superficially investigated.
Fixation of the benzoyl radical certainly takes place ; the fibre-
substance gains considerably in weight (36 p.ct), and analyses
of the products give results corresponding with the empirical
formula C19H22O10, which represents the fixation of i benzoyl
residue upon the empirical molecule Ci2H18O9 of lignocellu-
lose. This proportion is about one half that of cellulose
(C12H2oO10) under the same conditions. The result accords
with the observation of the partial yielding only of the fibre-
substance under the thiocarbonate reaction. From both we
may conclude that the ratio of alcoholic OH to the total oxygen
of the lignocellulose is low in comparison with cellulose.

K2

132 Cellulose

(2) Acetates.— The fibre-substance reacts directly with
acetic anhydride at its boiling point ; the product shows a con-
siderable gain in weight upon the original lignocellulose, and its
reactions are altogether different. It will be obvious that the
product in this case will not stand in simple relationship to the
parent molecule. In the first place the CO.CH2 groupings of
the lignocellulose are liable to further condensation and re-
arrangement ; under the condensing action of the anhydride
furfural groupings may be completed, which would then react
with the anhydride to form furfuracrylic acid (Berl. Ber. 1894,
286), which, again, would condense with the OH groups of the
cellulose ; and lastly, the keto R. hexene rings are open to
reaction with the anhydride in various ways. In view of these
various directions of probable reaction, and the further com-
plications in regard to the analysis of the product, resulting from
the presence of CO.CH2 and CH2.CO.O residues in the ligno-
cellulose molecule, the investigation of the reaction is deferred
until the constitution of the fibre-substance itself is elucidated.
It cannot then fail to afford confirmatory evidence of great
value, both as to the constitution of the constituent groups
and their mutual relationships within the molecule.

(3) Nitrates. — The ' nitration ' of the jute fibre has been
studied by O. Miihlhauser (Dingl. J. 283, 88) and the authors.
On plunging the fibre into the well-cooled 'nitrating' acid;
(H2SO4-fHNO3) it is instantly coloured to a dark red. After
remaining in the acid for about 5 minutes, evolution of
gaseous products is observed. If the fibre be then removed
and rapidly washed, the red colour of the nitrate gives place to
a golden yellow. The product when dry is found to be some-
what weakened (disintegrated) by the treatment, and harsher to
the touch than the original jute. It is, of course, explosive, and
takes fire at 160-1 70°. It is soluble in acetic acid and acetone,
and is gelatinised by nitrobenzene and acetic acid.

Compound Celluloses

133

The following results of experiments (Miihlhauser) may be
cited. The fibre-substance was purified previously to nitration
by boiling in alkaline solution (i p.ct. NaOH), thorough wash-
ing, and drying at 100°. The acids used in nitration were
of maximum strength.

The results may be given in the form of a table as under :

Ratio of acids in nitrating
mixture (by weight)

Proportion of
mixture to
fibre (by
weight)

Duration
of nitration
(hours)

Yield of
nitrate
p.ct.

Containing N
p.ct. (Eder's
method)

(i) (a)

iH2S04 iHNO3

IO I

I

130

12-10 II'SO

^ j > »>

15 I

*i

I32

12-26 12-04

3 5 » »l

15 I

3

136

12-03 II'SO

2 » n

15 *

3-4

145

12-03 11-96

The products were purified by exhaustively washing, diges-
tion in dilute solution of Na2CO3, and again washing.

An extended series of observations by the authors established
the following points :

(1) A gradual increase in yield with increase in duration of
exposure to the nitrating acid (at ordinary temperatures) up to
5-6 minutes, the maximum of 145 p.ct. being then attained.

(2) After that oxidation supervenes, soluble products are
formed, and the yield of insoluble nitrate diminishes.

(3) The increase in yield again observed on prolonged
exposure is a secondary result of the decomposition, alcoholic
OH groups being liberated from combination and then taking
part in the reaction.

A microscopic examination of the nitrated fibre (Miihl-
hauser) showed that nitration by prolonged exposure was
attended by resolution of the fibre-bundles into ultimate fibres,
and these showed a shrinkage in volume (diameter).

With the view of testing the homogeneity of the product,
specimens were exposed to the graduated action of alkaline

1 34 Cellulose

solutions. One of these, digested 52 hours with sodium
hydrate solution (i p.ct. NaOH), sustained a loss of 22 p.ct. in
weight. The insoluble residue, on analysis, was found to
contain 12*3 p.ct. N. A second specimen, heated with a
0-5 p.ct. solution of sodium carbonate 3 hours at 90-100°,
lost 25 p.ct. of its weight. The residue gave, on analysis,
12-25 p.ct. N.

From these observations, which the authors can fully
confirm, the important conclusion was drawn by Miihlhauser,
and may be given in his own words :

'Auch in diesem Falle hatte eine gradweise Abspaltung
nicht stattgefunden : die Zerstorung erstreckte sich, wie in
alien Fallen, auf das ganze Molecul.' The lignocellulose
behaves under nitration as a homogeneous body. It is important
to note at this point the convergence to this same conclusion,
of the evidence drawn from three independent lines of investi-
gation : (i) the general physiology of the elaboration of the
fibre ; (2) the resistance of the fibre-substance, so far as regards
the union of the constituent groups, to the action of hydrolytic
agents ; and (3) the homogeneous nature of the products of
synthesis, such as the nitrates just described. This evidence
compels the view that the fibre-substance is not merely a
mixture of cellulose with ' non -cellulose ' constituents, but that
these are compacted together into a homogeneous, though
complex molecule, by bonds of union of a strictly * atomic '
character.

(b) Compounds of Lignocellulose with the Halo-
gens, (i) Chlorine. — The reaction of the fibre-substance with
chlorine has been already described. We have now to deal
more particularly with the product.

The derivative in question is dissolved in large proportion
by treating the chlorinated fibre with alcohol after first washing
(to remove HC1) and squeezing. The alcoholic solution may

Compound Celluloses 135

be concentrated by evaporation, and on then pouring into
water, the product is precipitated in yellow flocks. On washing,
and drying at 100°, and analysing, it gives the following

results :

Calc. C19H1

C 42-82 .... 42-85
H 3'40 . . . . 3-38
Cl 26-83 .... 26-69

The product, obtained as above described from various
specimens of fibre, has been repeatedly analysed. It has been
analysed after fractional precipitation from solution in glacial
acetic acid ; also, after, then again exposing for a long period to
an atmosphere of chlorine gas, followed by suitable purification,
and always with results in close concordance with the above.
It is evident, therefore, that we are dealing with a complex of
a definite character. This complex is the lignin of earlier
observers, but which, in recognition of its ketonic characteristics,
is better termed lignone. From the fibre-substance exposed
to the action of dilute sulphuric acid (5 p.ct. H2SO4) for some
hours at 60-80° previously to chlorination, a derivative is
obtained, having identical characteristics and composition.
This further confirms the definite character of the lignone
complex, and the resistance of its constituent groups to hydro-
lytic actions.

The chlorination of the lignocellulose evidently resolves in
great measure the union of the lignone to the cellulose residue,
as the lignone chloride is largely dissolved away by exhaustive
treatment with simple solvents. The residue of chloride which
ultimately resists the solvent action gives the characteristic
reaction with sodium sulphite, and is probably therefore the
same product. The splitting off of the chloride is evidently a
secondary result, no doubt an effect of hydrolysis. It is to be
noted that the presence of water is essential tc the reaction ;

136 Cellulose

the fibre-substance in the dry state does not react with chlorine
even when heated with the gas (60-80°).

The residue, after removing the lignone chloride, is a
cellulose, containing, i.e. no ' unsaturated ' groups, but yield-
ing, on distillation with HC1, from 4 to 8 p.ct. furfural. It is to
be regarded, therefore, as a mixture of a normal cellulose (a)
and a cellulose (/?) which is readily condensed to furfural.
Since the total weight of furfural obtainable from the ligno-
cellulose is not affected by the chlorination, it may be con-
cluded that the ' furfuroids ' of the original lignocellulose are
in the main associated with the cellulose complex, and from
the yields of cellulose, that the principal constituent is this
J3 cellulose constituting 20 p.ct. of the complex. It appears
from later researches of the authors that a proportion of
actual furfural derivatives, notably hydrodyfurfurals, are pre-
sent in the lignocellulose, to which, in fact, certain of their
characteristic colour reactions are to be ascribed. • These,
however, are small in amount, and being easily removed,
without affecting the essential character of the lignocellulose,
may be regarded as products of secondary changes.

In the reaction of the lignocellulose with chlorine it is
found that HC1 is formed approximately equal in weight to
the Cl, combining as lignone chloride. It is to be concluded,
therefore, that the reaction is simple and unattended by
secondary oxidations of any moment.

The lignone complex when chlorinated, though readily
removed from the cellulose, has not been further resolved by
any treatments which can be accounted for by quantitative
statistics. The evidences as to its constitution are as follows :
(i) As regards the constituent group which combines with
chlorine. The lignone chloride when carefully heated gives a
sublimate containing chloroquinones ; treated with nascent
hydrogen it yields trichloropyrogallol ; the reaction with
sodium sulphite is identical with that of the chlorinated deriva-

Compound Celluloses 137

tives of pyrogallol, viz. mairogallol and Icucogallol. These
chlorides in turn have been shown to be derived from oxy-
quinone groups of the general type —
/CH=CHV

co<c__c>CH2

(OH)2 (OH)2

(Hantzsch and Schniter, Berl. Ber. 20, 2023.)
the presence of which in the lignone complex consistently
accounts for the most characteristic features of the ligno-
cellulose. (2) The residue of the lignine complex, while of
similar empirical composition, i.e. approximately, C2nH2nOn,
has very different constitutional relationships. Since it readily
breaks down under the action of dilute chromic acid in the
cold, to acetic acid as a main product, it might be formulated
by one of the many alternative CO-CH2 groupings ; and with
the keto-R.-hexene groups above the entire complex may be
expected to show the constitutional features of the pyrone group.

The further chlorination of the lignone chloride in solution in
glacial acetic acid has been studied by the authors ( J. Chem. Soc.
1883, 43, 18-21).

The products investigated were obtained from jute (a), and
from the fibre (f.v.b.) of the monocotyledonous Musa paradisiaca
(b). The analysis of the further chlorinated products showed a
higher percentage of chlorine (3375), the results also establishing
for both products the empirical formula C38H44C1UO16. The re-
action needs further investigation.

The similarity should be noted of the empirical formulas of the
halogenated derivatives of these unsaturated fibre- compounds with
those established by Sestini for the so-called sacchulmic com-
pounds (Gazzetta, 1882, 292 ; J. Chem. Soc. 1882, 1182). These
compounds are obtained from the carbohydrates by various
processes of dehydration, and, more particularly, spontaneous de-
compositions or decay of vegetable (cellulosic) matter (' humus ').

(2) Bromine. — Bromine attacks the lignocelluloses in
presence of water ; the brominated compound which results

138 Cellulose

resembles the quinone chloride above described, but the re*
action with this halogen is relatively incomplete. After remov-
ing the brominated product by hydrolysis with alkaline solutions,
and again exposing to bromine water, further reaction of the
same kind ensues. Proceeding in this way the lignone con-
stituent is completely removed as alkali soluble derivatives,
and cellulose is isolated.

If, on the other hand, the lignocellulose be dissolved in
the ZnCl2.HCl reagent (p. 9) and bromine added, the con-
ditions are more favourable for combination. On precipitating
by water, after standing some time, a brominated derivative is
obtained, containing 10*2 p.ct. Br (equivalent to 4-5 p.ct. Cl).
After standing 16 hours, during which period the cellulose
is largely hydrolysed to soluble derivatives, a brominated
derivative is obtained, containing 19-5 p.ct. Br (equivalent
to 8 p.ct. Cl). Even under these conditions, therefore, the
bromine is taken up in considerably less proportion than
the chlorine. When the lignone is completely isolated
from the cellulose, e.g. by digestion with alkalis at elevated
temperatures, it is then brominated in higher proportion.
Compounds C17H,4Br4OG, C16H12Br4O5, C27H28Br4O10 have
been obtained from the non-cellulose of esparto, isolated from
the alkaline by-products of the papermaker's boiling or pulping
process (p. 209 ; J. Chem. Soc. 41, 94).

(3) Iodine. — The lignocelluloses absorb iodine from its
aqueous solution and are coloured a deep brown. The re-
action has been quantitatively investigated, showing that jute
takes up 12-5 p.ct. from decinormal solution in potassium iodide,
but the proportion varies according to the concentration of
the solution and conditions of the digestion. When these are
kept uniform the proportion of the halogen absorbed is con-
stant. The resulting compound, however, is of a loose descrip-
tion, the iodine being easily removed by solvents.

Compound Celluloses

139

The following experiments with — iodine solution in potassium

kfcj >.. JO

iodide may be cited :

Weight of fibre

Vol. and composition of solution

Absorption p.ct.

2-II7
2-635
2726
2-463
2-500

60 C.C.A!
60 „
60 „
30 »»
30 „

12-2

n-3

I3-0

9-0
9-8

The absorption, therefore, depends upon the ratio of fibre-sub-
stance to iodine solution. This is more clearly shown by the
following parallel determinations :

Weight of fibre

Vol. and composition of solution

Absorption p.ct.

2-223

2'374
2-560

22-2 C.C.«S
237 ,»

25 '6 „

1 + 22 c.c. Aq

+ 47 '4 ,»
+ 76-8 „

6-01
4-8
3-2

It was finally established that, on digesting the fibre-substance at
1 8° C. with twenty times its 'weight of the — iodine solution as
ordinarily prepared, the absorption is constant at 12-9-13-3 p.ct.

It is to be noted that the celluloses also absorb a certain
proportion of iodine under similar conditions, viz. 3-4 p.ct.
when digested with 20 times their weight of the decinormal
solution.

Decompositions of Lignocelluloses, with Reso-
lution into Constituent Groups.— We have already
shown that the lignocelluloses are attacked by hydrolytic
agents and partially resolved into soluble products. These
products, though doubtless of lower molecular weight than
the original fibre-substance, preserve its essential characteristics,
and the results show that the lignocellulose reacts as a homo-
geneous compound. When exposed, on the other hand, to the

140 Cellulose

action of bodies which selectively attack particular groups, its
highly complex constitution is brought into evidence.

The reactions with the halogens just described, although
reactions of combination, also partake of the character of
decompositions, as the evidence has shown. We have now to
deal with the decompositions of the lignocellulose in their
order, and to emphasise the evidence which they afford as to
the relationships of the constituent groups of its complex
molecule.

(i) Non-oxidising adds. — (a) Hydrochloric add. — The fibre-
substance boiled with the acid of moderate concentration
(1-06 sp.gr.) is profoundly attacked. Furfural distils, and may
be quantitatively estimated as already described (p. 99).

The residue is a brownish-black mass of high carbon
percentage, presenting some features of resemblance with the
original, chiefly in its reactions with chlorine and nitric acid.
It is an ill-defined complex, however, and has been only super-
ficially investigated.

(b) Hydriodic add acts similarly in the earlier stages of
its action. The reaction with this hydracid is made use of in
the quantitative estimation of the O.CH3 groups of the ligno-
cellulose. The following determinations have been made in
normal specimens :

(i) (2)
OCHS . . 4'5 4 '6 p. ct. of lignocellulose

The acid acts, of course, as a deoxidising acid, and the
residue of the reaction is deserving of investigation with the
view to determine the limit of deoxidation.

(c) Sulphuric add. — The dilute acid at the boiling tempera-
ture acts similarly to hydrochloric acid ; the volatile products
of the decomposition are furfural and acetic acid. In the con-
centrated acid the lignocellulose dissolves, forming a purple
brown solution. On pouring the solution into water a

Compound Celluloses 141

1 condensed ' product is precipitated in dark brown flocks \
when dried it has the following composition :

C 64-4

H .... 4-4

O . 31-2

On diluting and distilling, acetic acid is obtained. The
amount formed in this way is 4-5 p.ct. of the lignocellulose.
Acetic acid is therefore a product of hydrolysis of the ligno-
cellulose, which contains a certain proportion of CH2.CO.O
groups.

Nitric acid (dilute), in presence of urea, acts as a
non-oxidising acid, and similarly to the above.

Other acids act in similar directions, and in greater or less
degree, according to the nature of the acid and the conditions
of its action.

Alkalis. — The hydrolysing action of the alkalis in boiling
aqueous solution has already been discussed.

At elevated temperatures (150-180°) solutions of the
alkaline hydrates (2-3 p.ct. Na2O) effect a complete resolution
of the cellulose and lignone, the latter being obtained in
solution in the form of acid derivatives. In addition to
acetic acid the solution contains acids of high carbon per-
centage, which are precipitated on adding mineral acids to the
alkaline solution. These bodies have been investigated by
Lange (Zeitschr. Physiol. Chem. 14, 217), but as the products
described by him were obtained from lignocelluloses of another
group — viz. the woods — and under more severe conditions of
treatment, they will be dealt with subsequently. Jute, however,
yields very similar products, viz. acid bodies of high carbon
percentage (60-6 1), giving Cl substitution derivatives. The
cellulose retains residues of these bodies, but they are easily
eliminated by treatment with hypochlorite solution. The
cellulose is then obtained as a white pulp, consisting of the

142 Cellulose

disintegrated fibre-elements or ultimate fibres. The yield from
normal specimens is about 60 p.ct. Only the more resistant
cellulose a survives the treatment, the cellulose ft together with
the lignone complex, being converted into soluble derivatives.

Extreme action of alkaline hydrates. — With the caustic
alkalis in concentrated solution and at temperatures exceed-
ing 120°, much more drastic decompositions take place,
the entire molecule being attacked. For complete resolution
into simple molecules (oxalic, acetic, and carbonic acids) the
proportion of alkaline hydrate to lignocellulose requires to be
2-3 to i, and the temperature raised to 250°, and maintained
at that point for some hours. Thus, on heating jute for 8
hours at 250° with 3 times its weight of KOH, the yields of
the main products were : acetic acid, 37-0 p.ct. ; oxalic acid,
53'3 P-ct- of tne lignocellulose (J. Soc. Chem. Ind. n, 966).
The action is an oxidising action, in the sense that hydrogen
is expelled ; gaseous carbon compounds (CO, CH4) are formed
in relatively small quantities.

DECOMPOSITIONS BY OXIDANTS.— (i) Acid.— Certain of
these are more important as contributing to the elucidation of
constitutional points.

(a) Chromic arid.— The direction of attack of this oxidant
depends upon the auxiliary conditions, chiefly upon the
presence of hydrolysing acids. With the CrO3 alone, the
interaction with the lignocellulose is at first one of simple
combination ; afterwards the CrO3 fixed is gradually deoxidised.
Under these circumstances the lignocellulose suffers a very
slight loss of weight. In presence of acids, however, the fibre-
substance loses in weight, and the insoluble residue is affected
more or less. The following results may be cited in illus-
tration :

CrOz alone. — (i) 4:5 grms. jute, containing 07 p.ct. ash
constituents, digested 18 hours in i p.ct. solution CrO3

Compound Celluloses

143

(250 c.c.) at 15-18°. Weight of product, 4-55 grms. ; with ash,
5*0 p.ct. j CrO3 fixed, equivalent to 4/2 p.ct. Cr2O3.

(2) 1-8 grm. fibre j 100 c.c. i p.ct. CrO3 ; 16 hours at 15*.
Product, 1-82 grm. ; ash, 4-3 ; CrO3 fixed, equivalent to 37 p.ct.
Cr2O3.

CrO3 and acetic acid. — 1*8 grm. fibre; 100 c.c. i p.ct.
CrO3 containing 4 grms. acetic acid : (a) digested 16 hours,
(b) digested 20 hours.

Product

P.ct. of original

Ash p.ct.

C p.ct. in product

(a) 170
(*) 1*65

94 '4
917

1-8
1-8

43*2 42-8
42 -o

The products were largely soluble in dilute alkaline solutions ;
the lignocellulose reactions were faint ; the characteristics of
the products were those of the oxycelluloses.

CrO3 and sulphuric acid (dilute). — (i) 0*9 grm. fibre ; 100 c.c.
solution containing 0-874 CrO3 and H2SO4 : (a) r6 grm., (£)
3-2 grms., (c) 4-8 grms.

Product

P.ct of original

CrO, consumed

(a) 0-840
(b) 0-801
(c) 0768

93'3

89-0

85'3

0'5IO
0-470
0-430

} Traces only of
gaseous pro-
ducts evolved

The solution of the fibre-substance increases with the increase
of hydrolysing acid ; the deoxidation of the CrO3 slightly de-
creasing.

(2) Jute, o'9 grm. ; 100 c.c. solution containing variable
quantities of CrO3 and H2SO4 as under (CrO3 solution, 0-842
CrO3 per 10 c.c. ; H2SO4 = 7-59 per 10 c.c.) :

—

d)

(2)

(3)

(4)

CrOj solution . . -,
H2SO4 solution . i :*.,
Yield of oxycellulose .

IO C.C.
IO C.C.

72-8 p.ct.

15
15
65-0

20
2O

53'3

25
25

45*0

144 Cellulose

The oxycelluloses obtained were soluble in alkaline solutions
and in nitric acid (1-43 sp.gr.). Under these more severe con-
ditions there is an increasing evolution of gas, and from (4)
75 c.c. were collected.

In the process of oxidising with chromic acid in presence of a
hydrolysing acid (H2SO4), acetic acid is formed. Oxidised by its
own weight of CrO3 in presence of excess of normal sulphuric acid,
the fibre- substance yields from 12-13 P-ct« C2H4O2.

It is evident that chromic acid oxidations of the fibre-sub-
stance can be controlled within any prescribed limits. From
the investigations, of which the above are typical series of
experiments, it was concluded —

(i) That the keto R. hexene groups yield most readily to
the action, and may in fact be selectively attacked and elimi-
nated ; (2) that with a net loss of weight of 10 p.ct. the ligno-
cellulose is converted into an oxycellulose containing 42-43
p.ct. carbon, and yielding the same percentage of furfural (HC1
distillation) as the original fibre. The furfural-yielding com-
plex is not, therefore, radically affected by the treatment. (3)
As the amount of oxygen expended (CrO3 deoxidised) is rela-
tively small — approximately i mol. per unit weight Ci2H18O9
of lignocellulose — and would appear to be chiefly consumed
in oxidising the portion passing into solution, the relatively
large reduction in carbon percentage of the insoluble residue is
due to simultaneous fixation of water. (4) It appears, in fact,
that the furfural-yielding complex is by such action converted
into an oxycellulose.

Chromic Acid and Sulphuric Acid (Cone.). — When the ligno-
celluloses are dissolved in concentrated sulphuric acid, the addition
of chromic acid determines complete combustion of the carbon to
gaseous products CO2 and CO. The proportion of CO formed is
usually very small. As both gases, however, have the same mole-
cular volume, a determination of the total gas evolved gives by
calculation the carbon contents of the substance. The method is

Compound Celluloses 145

available for analytical purposes, and will be found fully described
in J. Chem. Soc. 53, 889.

As I mgr. of lignocellulose gives approximately 0-9 c.c. COa
under ordinary conditions, it will be seen that trustworthy results
can be obtained with very small quantities of substance ; and as
the entire operation takes only a very few minutes, the method is ex-
tremely useful for rapid approximate analyses of products obtained
in the course of investigation.

(V) Nitric add. — In the interaction of nitric acid with the
fibre-substance, in presence of sulphuric acid, it has been
already shown that decomposition (oxidation) supervenes after
a few minutes' exposure. The acid (1*5 sp.gr.) alone attacks the
lignocellulose still more rapidly and energetically ; as the oxi-
dation is of a ' wholesale ' character, its investigation would not
throw much light upon the constitution of the fibre-substance.
The acid of i -43 sp.gr. acts more gradually ; there is direct
combination in the first instance attended by deoxidation. A
yellow product is obtained differing but little in weight from
the original, and containing 2*0-2-5 P-ct- N. With the pro-
gress of the oxidation there is considerable disintegration of
the fibre-substance and conversion into soluble derivatives,
but of an ill-defined character. With the dilute acid, on the
other hand, a very gradual resolution ensues, and the reaction
has been carefully investigated. The main results determined
are these : the lignocellulose is entirely resolved into insoluble
cellulose (a) and soluble derivatives of the remaining groups,
with a proportion of acid (HNO3) equal to 25 p.ct. of the fibre-
substance. The specific action of the acid takes place at any
dilution not exceeding 30 Aq : iHNO3 (by weight), and any
temperature within the range 40-100°. The most convenient
conditions are with the acid at 7-10 p.ct. HNO3 and temperature
60-80°. Under these conditions there is considerable evolution
of gas, and of very complex composition.

The course of the reaction may be thus described. The

L

146 Cellulose

lignocellulose is changed in colour to a bright yellow which
gradually changes to lemon yellow, and after some hours' diges-
tion, to white. If the digestion be interrupted at the yellow
stage, the fibre washed and digested with boiling alcohol, a
bright yellow solution is obtained ; and on driving off the
alcohol a gummy body is left, characterised by great instability,
reducing Fehling's solution in the cold, yielding furfural on
boiling with acids, and progressively decomposed on heating at
1 00° (in presence of water) with evolution of gaseous products.
The substance retains from 1-2 p.ct. N, but in a very un-
stable form, being entirely split off on heating with water.
This ill-defined product we may term, for obvious reasons, the
intermediate body. These results will be appreciated from
the following statement of the final products of the decom-
position.

Ligitocellulose and Dihite Nitric Acid.

— .*.

Solid products : Cellulose a Oxalic acid Intermediate

63-66 p.ct 4*0-5-5 p.ct. body, 5-3-5-8
Volatile acid: Acetic acid, 14-18 p.ct.

Gaseous products \ From HN03 From fibre- sub stance

N204. N202. N20. N2. HCN CO,. CO. HCN
(Representing about 50 p.ct.
oftheNoftheHNOJ

The most notable features of the decomposition are —
(i) As regards theHNO3 — (a) The reaction depends upon
the presence of nitrous acid ; the addition of urea entirely
arrests the specific action of the acid, and it then behaves
exactly as the non-oxidising mineral acids, (b) The direct
deoxidation of nitric acid never proceeds beyond the formation
of NO ; the presence of N2O indicates the formation of a
hydroxime, and its decomposition by further reaction with
nitrous acid. The formation of HCN also appears to result from
the dehydration of a product of this nature, and this conclu-
sion is confirmed by the observation that HCN appears in

Compound Celluloses 147

greatest quantity at the end of the reaction and as the tempera-
ture is raised to 100°. (c) The presence of N2 indicates a still
further deoxidation — or hydrogenisation of the N to ammonia.

(2) As regards the fibre-substance, the keto R. hexene
groups are rapidly oxidised, and entirely broken down. No
'aromatic' products are formed, and the result is in perfect
accord with the general view we have taken of their consti-
tution.

The furfural-yielding complex and the ^-cellulose group
are more gradually resolved, and both probably contribute to
the large yield of acetic acid. It is obvious that the constitu-
tion of both groups is of a special type, unlike the normal
grouping of the carbohydrates.

The reaction has been also studied in connection with
another group of the lignocelluloses — viz. the woods — and will
be again referred to (p. 212).

Joint action of oxides of nitrogen and chlorine. — F. Schulze's
method of eliminating the non-cellulose groups of the ligno-
celluloses, in the isolation and estimation of cellulose, has
been already described. It consists in a prolonged diges-
tion in the cold with nitric acid (n sp.gr.), with addition of
a small proportion of potassium chlorate. The reaction has
not been investigated in regard to the by-products. The
mechanism of the decomposition will be evident from what
has been stated in regard to the actions of chlorine and of
nitric acid upon the lignocellulose.

(2) ALKALINE OXIDANTS. — The actions of this group of
reagents are of considerable technical importance, as upon
them depend the various bleaching methods in common
practice ; but they have not been sufficiently investigated to
throw light on theoretical points.

Potassium permanganate acts, of course, as an oxidising
agent pure and simple. The limit of deoxidation in basic or

L2

148 Cellulose

neutral solution is the oxide MnO2, which is deposited upon,
and in intimate combination with, the lignocellulose. If re-
moved by treatment with sulphurous acid it does not further
attack the lignocellulose, the treatment merely revealing the
bleaching action accomplished in this stage of the deoxidation.
By treatment with sulphuric acid the lignocellulose under-
goes further oxidation, and with hydrochloric acid chlorine is
liberated and combines with the fibre-substance. These re-
actions are, however, of little importance. The permanganate
bleach is too costly for general adoption in the case of jute
fabrics. From its simplicity, it is a useful treatment in the
laboratory, whether for removing coloured impurities from the
raw fibre, or from cellulosic products separated by any of the
processes already described.

Hypochlorites. — Bleaching powder solution (calcium hypo-
chlorite) and the equivalent sodium compound act, in pre-
sence of excess of the base, as oxidising compounds ; but as
by oxidation and attendant hydrolysis, acid derivatives are
formed from the lignocellulose, the use of a * neutral ' solution
of the bleaching solution often leads to chlorination of the
fibre-substances, owing to liberation of hypochlorous acid.
Neglect of this probability has led to disastrous results in the
bleaching of jute piece goods ; and a full discussion of the
matter, in both practical and theoretical bearings, will be found
in the Bull. Soc. Ind. Mulhouse, 1880.

The danger is avoided by ensuring the presence of excess
of base ; this is more easily controlled in solutions of the soda
compound, which are therefore to be preferred. After
bleaching with the hypochlorites, the fibre or fabric should be
well washed, and plunged for a short time into sulphurous acid
solution which removes the last traces of oxidising compounds.
After again washing, the lignocellulose may be dried without
change of colour.

Compound Celluloses 149

Hypobromites. — The hypobromites of the alkalis attack
the lignocelluloses profoundly ; amongst the final products of
decomposition bromoform and carbon tetrabromide are ob-
tained in some quantity. (Compare N. Collie, J. Chem. Soc.
1894, 262, which contains the results of a general investiga-
tion of the reaction.)

In regard to the alkaline oxidants it may be said generally,
in conclusion, that they attack the non-cellulose constituents
of the lignocelluloses in greater degree, but their action extends
to the cellulose also. They are therefore of little present use
as * pioneer reagents,' and have moreover secured no syste-
matic investigation.

Other Decompositions of Lignocellulose.— There
are a number of decompositions remaining to be described
which do not fall within any group classification : these will
now be dealt with in the order of their importance.

Interaction of lignocellulose with sulphites and bisulphites. —
Jute, when heated at high temperatures with solutions of the
alkaline sulphites, or of the bisulphites of the alkaline earths,
is directly resolved into cellulose (insoluble) and soluble com-
pounds of the lignone complex with the sulphites. A similar
treatment of pine-wood, attended by the same results, is
the basis of the now highly developed ' sulphite wood pulp
industry. The process and the reactions upon which it is
based will be described in detail in a later section ; and as the
general principles apply to the lignocellulose with which we
are dealing, it is unnecessary to anticipate the fuller treatment
of the subject.

The theory of the reaction is deducible from the following
considerations : The lignocellulose when heated with water
only, at high temperatures (140-160°), is profoundly attacked ;
a considerable proportion of the fibre-substance passes into
solution (hydrolysis), and the residue of disintegrated fibre

1 5O Cellulose

resembles the product obtained by digestion with dehydrating
acids. It is obvious, b priori, that a reaction of this kind will
proceed to the limit representing equilibrium between the
hydrolysing and condensing influences. If, now, a substance
be present capable of uniting with the products of hydrolysis
in such a way as to prevent them entering further into reaction,
the resolution will proceed without secondary complications to
the limit determined by the constitution of the lignocellulose.
Sodium sulphite is a reagent fulfilling these conditions : acid
products combine with the base, and aldehydic products with
the bisulphite residue. In this case, however, the hydrolysis
being thrown chiefly upon the water, a high temperature (160°)
is required to effect complete decomposition. By substituting
bisulphites, the hydrolysis is aided from the first by sulphurous
acid, and the decomposition is completed at lower temperatures
(130-140°). That the hydrolysing action of sulphurous acid is
a powerful factor is evident from the fact that an aqueous
solution of this acid, containing 7-8 p.ct. SO2 (which, of course,
requires to be prepared under pressure), will itself resolve the
lignocellulose at the lower temperature of 95-105°. The reac-
tion with the bisulphites is, however, in many respects simpler,
and complete decomposition is effected with solutions con-
taining 3-4 p.ct. SO2.

The yield of cellulose is 63-66 p.ct. of the lignocellulose,
and is composed therefore of the more resistant a cellulose ;
the /3-cellulose is hydrolysed under these conditions also, and
passes into solution with the lignone. The soluble derivatives
preserve the features of the original lignone ; combining with
the halogens to form substitution products, and yielding furfural
on boiling with hydrochloric acid. All the reactions of the
product indicate that it is a sulphonated derivative of the
lignone complex of the original fibre. For the further discus-
sion of the reaction see p. 198.

Compound Celluloses 151

The following observations upon the behaviour of the fibre when
treated with water at high temperatures may be cited. The experi-
ments were conducted in glass tubes.

(1) Heated 12 hours at no°C. : Slightly attacked. Loss in
weight, ii-o p.ct.

(2) Heated 10 hours at 120-130°: Onry slightly attacked.
Heated further 10 hours : Fibre disintegrated. Loss of weight,
27-5 p.ct. Solution contained furfural.

(3) Heated 9 hours at 140° : Completely disintegrated. Loss
of weight, 22-6 p.ct.

Analysis of Disintegrated Fibre.

C 48-30 p.ct. 1 Yield of cellulose (Cl method), 76*8 p.ct ;
H 5 •! 6 p.ct. /calculated on original fibre, 59*5 p.ct.

(4) Heated with water and barium carbonate 9 hours at 140° :
Colour changed to brown. Fibre not disintegrated. Loss of
weight, 20-0 p.ct. Product yielded : cellulose, 79-3 p.ct. ; calcu-
lated on original, 63-5 p.ct.

(5) Heated with solution sodium sulphite (5 p.ct.) 10 hours at
120-130° : Loss of weight, 19-0 p.ct. Fibre disintegrated ; fibre
and solution colourless. Cellulose p.ct. on product, 84-6 ; p.ct. on
original fibre, 687.

(6) Heated with sodium bisulphite solution (2*6 p.ct. SO,,) 10
hours at 115°: Fibre disintegrated; fibre and solution colour-
less. Loss of weight, 19 p.ct. Cellulose p.ct. on product, 73-5 ;
p.ct. on original fibre, 65 '4.

Animal digestion of the lignocelluloses. — The urine of
the herbivora contains hippuric acid as a characteristic con-
stituent. The origin of this compound, and more particularly
its benzoyl group, has been the subject of considerable dis-
cussion and controversy, but the evidence points unmistakably
to the lignocelluloses of the various fodders as the source of
the product. It appears from what we now know of the con-
stitution of the lignone complex, that its R. hexene and
CO— CH2 groups may, without unduly straining the probabili-
ties, be regarded as undergoing transformation, in the processes

152 Cellulose

of animal metabolism, to the compound in question. The
problem has been specially investigated by Meissner and
Sheppard (1866), Stutzer (Berl. Ber. 8, 575), Weiske (Ztschr.
Biol. 12, 24).

Spontaneous decomposition of the lignocelluloses. — Jute is
sometimes baled in a damp state, or wetted by sea water in
course of shipment, and the fibre in the interior of such bales
is found to undergo considerable chemical change, attended
by structural disintegration. A specimen of the fibre thus dis-
integrated was found to present the following features :

Soluble in water lo-o p.ct.

Soluble in I p.ct. NaOH .... 23-0 ,,
Cellulose 60-4 58-8

The aqueous exhaust of the fibre was astringent to the
taste, was precipitated by gelatin solution, and gave coloured
reactions with iron salts. It was digested on barium car-
bonate, filtered, evaporated, and the residue resolved by
alcohol into (i) a soluble body of 'neutral' characteristics,
which on analysis gave numbers expressed by the empirical
formula C26H34O16. This substance, on fusion with potash,
gave some phloroglucol and a large yield of protocatechuic
acid. (2) An insoluble body, the Ba salt of an acid, which on
analysis gave numbers expressed by the formula BaC29H42O29.
The investigation of these products dates from 1880
(J. Chem. Soc. 41, 93), and they were not examined for
determination of the now well-established 'constants' (p. 157).
From the above results, only the main features of this sponta-
neous decomposition of the lignocellulose are evident, viz. a
resolution of the lignone complex into more and less oxidised
groups ; the latter representing transition to aromatic products
of definite and ascertained relationships, the former having
features in common with the group of pectic compounds. But
they have the additional interest of suggesting, in a very direct

Compound Celluloses 153

way, the origin of the astringent substances or tannins, widely
distributed throughout the plant world ; and not only of the
tannins, but more generally derivatives of the trihydric phenols.
This problem is of the greatest interest both from the chemical
and physiological standpoints. It involves varied transitions
from aliphatic to cyclic compounds, and a prominent feature
of the synthetic activity of the plant, the elucidation of which
is the immediate objective of organic chemistry.

Of the endless variety of excreted products in vegetable growth
— e.g. essential oils, waxes, alkaloids, and ' aromatic ' products —
the tannins have specially attracted the attention of physiologists.
An important monograph on the subject has recently appeared :
Grundlinien zu einer Physiologic des Gerbstoffs, by G. Kraus,
Leipzig, 1888. The work contains the results of extensive experi-
mental investigation of the origin, distribution, fate, and function
of tannins in normal growths. A resume of the evidence shows
generally :

(1) That the tannins are formed in leaves under the same con-
ditions as are necessary for general assimilation, but is an inde-
pendent process. The tannins thus formed are transmitted through
the leaf stalk, and distributed through the permanent structures.

(2) They are also formed in processes of growth in the dark, e.g.
growth of rhizomes, unfolding of buds, &c.

(3) Also in isolated cells and tissues, where they remain.

(4) The tannins take no further direct part in plant assimilation 5
they are end-products.

In dealing with the question of the sudden increase observed
in passing from the sap to the heart wood of many trees, the
author speaks as follows : ' The only satisfactory explanation would
be the assumption that the tannin in this case is formed locally, i.e.
in the wood-tissue itself, the parent substance being the tissue of
the medullary rays and wood-parenchyma. As to the possible
mechanism of such a process, however, we are in total darkness.'
That, we venture to think, is no longer the case.

Destructive Distillation.— The decomposition of jute
by destructive distillation has been specially investigated by

1 54 Cellulose

Chorley and Ramsay, who obtained the following results
(J. Soc. Chem. Ind. n).

Weight of fibre . • • .71 grms. 73 grms.

P.ct. P.ct.

Charcoal 2871 32-87

Total distillate . . . • S7'7Q 43 ^5

Carbonic anhydride , . . — 12*33

Other gases (and loss) ... — 1 1 '65

loo -oo

Volume of gas .... 3,000 c.c. 2,5000.0
„ „ per 100 grms. . . 4,220 „ 3,420 „

Composition of Products p. ct.

P.ct. vol.

(Carbon monoxide . . 78 '80 85 -29
Oxygen . . . .3-01 173

v Residual gas . . . I8'I9 12-98

P.ct. fibre

fTar 1478 6-85

Distillate j Acetic acid . , . . 0-40 1*40

I Methyl alcohol ... — 10-08

The chief features to be noted in the products are the low
yields of charcoal (compare p. 69) and acetic acid, and the
high yields of carbonic oxide and methyl spirit. The thermal
features of the decomposition are remarkable : heated gradually
to 320° the temperatures within the distilling flask and external
to it follow the ordinary course ; but at 320° (external) the
temperature within the flask rushes up to 375°, the change
being marked by a much increased evolution of gas.

The destruction of a complex substance such as the ligno-
cellulose, by heat, involves a highly complicated web of reac-
tions, which it would be impossible to disentangle in detail
and in such a way as to throw light on the fate of particular
groups. In the main there are, of course, the two opposing
factors at work — dissociation, giving products of lesser, to those
of the least molecular weight (gases) ; and condensation, giving

Compound Celluloses 155

products of greater complexity, up to those of indefinite
molecular weight (charcoal or pseudo-carbon). More definite
features of the condensation are the closing of the C4O ring
(furfural) and the further condensation of the hexene to
benzene rings. But to study these and other changes in
reference to the parent molecule, it would be necessary to carry
out an elaborate series of quantitative observations, varying
not only the physical conditions of the distillation (temperature,
time, &c.), but the chemical factors by the admixture with the
fibre-substance of reagents of known function. Until we have
such results the imagination is free to go to work upon such
slender materials as are available.

General Conclusions as to the Composition and
Constitution of Jute Lignocellulose.— Having thus set
forth the general chemistry of the typical lignocellulose, it is
important to select and bring together those facts which bear
more particularly upon the problem of its constitution. This
problem, it may be remarked, cannot be divorced from its
essentially physiological aspects : a plant is an assemblage not
merely of products, but of processes ; and in investigating a plant
tissue, we have not merely to ascertain the quantitative rela-
tionships of its constituents, but from the point of view of
physiology or organic function to distinguish between organic
and excreted products ; further, to endeavour to arrive at their
genetic relationships. The history of every tissue is one of
continuous modification, and the excreta of plants are, in many
cases, the last links of a long chain of transformations. Where
such compounds are formed as by-products of the assimilative
processes, we cannot as yet hope to have any definite clue to
their origin ; but where they originate independently, either by
intrinsic or extrinsic modification (e.g. oxidation) of a tissue-
substance, the clue may be expected to be found in the consti-
tution of the tissue-substance itself. Or, to put it in another

156 Cellulose

way, if we find associated with a tissue an excreted product of
general constitutional resemblance thereto, we cannot avoid
the suggestion of genetic relationship. The suggestion be-
comes an hypothesis upon which investigation can proceed.
The lignocelluloses, for instance, afford many indications of
such relationship to the tannins. In the jute fibre, tannins are
always present in small quantity ; the characteristic R. hexene
groups of the lignocellulose occupy a definite and close rela-
tionship to the trihydric phenol, pyrogallol, to which many of
the tannins stand in direct constitutional relationship ; and we
have described an instance of * spontaneous ' transformation of
the lignocellulose into a substance having the essential charac-
teristics of the tannin group. The general discussion of this
question belongs to a later section of our subject (see p. 179).
It is introduced here in order to show that we cannot attempt
to formulate a molecule of a lignocellulose on the lines of a
carbon compound of ascertainable molecular weight and such
relationships of its constituent groups as are sharply defined
and verifiable by synthesis. It is true that when attacked in
detail the lignocellulose is resolved into well-differentiated
groups, which may be regarded with reservation as consti-
tuents of the parent molecule ; the reservation being that
unless and until the lines of cleavage are proved to be in-
variable, we cannot consolidate the results of various direc-
tions of resolution into a homogeneous view of the parent
substance.

We will now point out how far the problem is solved by the
evidence available.

(i) THE LIGNOCELLULOSE A HOMOGENEOUS COMPOUND
RATHER THAN A MIXTURE. — The evidence for this conclusion
is as follows — (a) Physiological : general uniformity in composi-
tion and reactions ; does not vary with age of fibre (i.e. from root

Compound Celluloses 157

upwards) nor with thickening of cell wall (incrustation) ; pre-
serves essential features through wide range of differences in
empirical composition, resulting from differences in conditions
of growth, (b) General resistance to resolution (into proximate
constituents), by the action of solvents and hydrolytic agents
generally, (c) Behaviour in synthetic reactions, chiefly in ferric
ferricyanide reaction and formation of nitrates ; resistance of
molecule to resolution.

(2) GENERAL CHARACTER OF LIGNOCELLULOSE CONSIDERED
AS A WHOLE. — The alcoholic characteristics of the lignocellulose
are inferior to those of cellulose : the reactive OH groups are
fewer in proportion ; CO groups of aldehydic, ketonic, and acid
function are present in union, more or less, with the more basic
OH groups. Th: characteristic reactions of the compound
(lignone group) are those of unsaturated compounds, and it is,
by comparison with the celluloses, greedy of oxygen.

(3) CONSTANTS OF THE FIBRE IN REACTION. — In this con-
nection we refer only to such reactions as throw light upon
the relationships of constituent groups, and therefore reactions
of decomposition.

Cellulose. — Cl method : average yield, 7 5*0 p.ct. ; raised, by
minimising conditions tending to hydrolysis and oxidation, to
78-82 p.ct.

Br method : average, 72-0 p.ct. ; may be raised similarly
to 74-76 p.ct.

Nitric acid (dilute) method, 63-66 p.ct. Alkali method,
56-60 p.ct. Bisulphite method, 60-63 P-ct-

The cellulose is a variable, the variations being due to
greater or less hydrolysis. The lignocellulose contains a
cellulose of resistant characteristics and a cellulosic constituent
which is either isolated as cellulose or dissolved with the
lignone complex according to the treatment. This latter

158 Cellulose

cellulose, when isolated, contains O.CH3 groups. The whole
cellulose complex gives the following results on analysis :

Calc. C)8H32018
Ultimate analysis . | jj ^ 6

Proximate . . O.CH3 1-2 Furfural . 6-8 p. ct.

The cellulose is a hydrate, and a mixture of two celluloses :
the /3-cellulose contains the methoxyl groups, and gives
furfural with condensing acids.

Lignone— Chlorination. — Cl combining with lignone, 8*0
p.ct. ; Cl combining as HC1, 8'o p.ct, calculated on the ligno-
cellulose. Composition of lignone chloride, C19H18C1409
(containing 267 p.ct. Cl), from which we may assume
Ci9H22O9 as the approximate formula for the lignone
complex.

From these statistics, and on the assumption that there are
no hydration changes of any moment, we may calculate the
lignone complex to constitute a little over 20 p.ct. of the
lignocellulose.

Continuing this statistical and approximate method of
investigation and calculating to carbon percentages —

Cellulose (anhydride), 44-4; lignone, 57-8.
80 x 44-4 -T- 100 = 38-52
20 x 57-8 -i- loo = 11-56

47 *o8 p.ct. C in lignocellulose.

These results confirm the evidence of the * quantitative '
character of the chlorine reaction.

From a study of the ester reactions of the lignone chloride
and of the original lignone group, it is to be concluded that
the former contains not more than one alcoholic OH group,
even when isolated by treatment of the chlorinated fibre with

Compound Celluloses 159

sodium sulphite solution. Both in union and under resolu-
tion, therefore, the two main complexes preserve their general
character of complex anhydrides. The union resists the
action of all simple hydrolytic treatment, but yields at once
when the condition of oxidation or the specific attack of
negative groups (NO2, C12O) is superadded.

It has been shown by the quantitative statistical study of
these several reactions, and by the composition of the pro-
ducts, that each has its characteristic line of resolution or
cleavage of the original lignocellulose complex. While the
lines of separation of the a-cellulose and ^-cellulose residues
are well marked, it is doubtful whether the /3-cellulose is as
sharply separated from the lignone complex. Under the
chromic acid treatment it certainly appears that a portion of
the latter is converted into a cellulose (oxycellulose), as a
result of attendant hydration ; moreover the general features
of the lignone are largely those of products obtained by the
action of condensing agents upon the carbohydrates, and it is
not improbable that the general configuration might be so
retained that combination with water would restore the carbo-
hydrate, i.e. cellulose, character. Again, if the /3-cellulose is a
keto- cellulose, as it appears to be, it may exist in a condensed
form in the lignocellulose, and thus have features as much in
common with the lignone as with the a-cellulose.

These considerations justify the use of the group-term
lignocellulose, and at the same time show that the constituent
groups lignone-cellulose must not be too easily regarded as
fixed quantities. It leaves open the question as to whether
the lignone is not genetically connected with the cellulose.
This is an important physiological probability which will be
met with again in considering the chemistry of the woods,
i.e. the lignocellulose of perennials.

160 Cellulose

Furfural.

P.ct. of lignocellulose.

Yield from original fibre-substance . . 8-9
Yield after chlorination .... 8-9

P.ct. of products.

Yield after CrO3 treatment .... 8-9
Yield from isolated cellulose Cl method . y-fc

Th'e origin of this characteristic product of decomposition
is localised mainly in the cellulose complex, the group from
which it is derived being isolated as a cellulose (/3-cellulose)
by the chlorination method. The lignocelluloses in their
* natural' condition appear also to contain hydroxyfurfurals
in small proportion, and to which their characteristic colour
reactions with phenols, especially phloroglucinol, are probably
due.

Methoxyl. — The presence of O.CH3 groups is another
characteristic feature of lignification, i.e. of a lignocellulose.
In jute the total yield is 4-6 p.ct. ; the major proportion of the
methoxyl is localised in the lignone complex. A certain pro-
portion appears in the cellulose isolated by the chlorination
process, which is further suggestive of the relationships of the
lignone to the /3-cellulose previously discussed (p. 159).

Assuming that the whole of the O.CH3 is contained in the
lignone complex, the empirical formula assigned to this may
be calculated to contain two such groups, and would become
C17H16O7.2OCH3, a formula similar to that arrived at for a
product obtained from the lignone of coniferous woods (p. 201),
viz. C24H2408.(OCH3)2<

Acetic acid is an important product of resolution of the
lignocelluloses by the action of hydrolytic and oxidising
agents, and under conditions of very limited intensity. The
source of this product is the lignone complex, and when this
is broken down by treatment with chromic acid (in presence
of sulphuric acid) the yield amounts to 50-70 p.ct. of its

Compound Celluloses 161

weight. The conditions of its formation point to its being a
product of hydration rather than oxidation ; it is probable
that more complex ketonic acids are first produced, and
further resolved on distillation, especially in presence of excess
of the oxidant.

It appears from this that the lignone complex contains,
associated with the oxyquinone groups, a large proportion of
CO.CH2 groups, the configuration of which remains undeter-
mined. It is probable that groups allied to dehydracetic acid
are represented, and a pyrone grouping of a portion of the
complex would account for the production of acetone as a
first product of destructive distillation (pp. 154-206).

(2) Other Types of ' Annual ' Lignocelluloses.— The
chemistry of the jute fibre might be presumed to cover the
essential features of the lignification of bast fibres generally ; so
far as investigation has gone, this appears indeed to be the case.
This statement, of course, must not be taken as suggesting
identity of constitutional features. Comparative investigation
of the bast tissues of the dicotyledonous annuals generally has
not as yet been attempted. Such work is called for, and it is
impossible to predict the influence which the results might
have in extending our grasp of the physiology of the exogenous
stem. Of those which are lignocelluloses, jute is undoubtedly
typical, and the methods adopted for this fibre may be ex-
tended to the group.

The process of lignification, however, is by no means limited
to particular tissues, and we have now to deal with other
representative cases of the formation of lignocelluloses in
' annual ' structures.

* GLYCODRUPOSE.' — The hard concretions of the flesh of the
pear are composed of a lignocellulose giving the typical re-
actions of the jute fibre. This product was investigated some
years ago by Erdmann (Annalen, 138, 9). The concretions

M

1 62 Cellulose

are isolated from the parenchyma of the fruit in which they are
imbedded, by long boiling with water, rubbing down to a pulp,
washing away the cellular debris, and thus, by continued
mechanical action and washing, entirely freeing them from the
matrix of softer tissue. The substance of these concretions
gives constant results on elementary analysis, expressed by the
empirical formula C24H36Oi6 ; to this complex Erdmann gave
the name Glycodrupose^ and he regards it as resolved on boiling
with hydrochloric acid according to the equation :

C24H36O16 + 4H2O = 2C6H12O6 + Ci2H20O8.

Glycodrupose Glucose Drupose

Drupose, on fusion with potash (KOH), yields aromatic
products amongst which pyrocatechin was identified. Glyco-
drupose, on boiling with dilute nitric acid, gives a residue of
pure cellulose. These results were repeated by Bente (Berl.
Ber. 8, 476), and in general terms confirmed, though the ana-
lytical results varied somewhat from the above.

From our present point of view, the interpretation of these
results by these investigators is open to question in more than
one direction, but they certainly establish the following points :

The concretions represent a compound cellulose, of which
the non-cellulose is easily converted into aromatic derivatives.
This compound cellulose gives constant results when analysed,
whether for its elementary or proximate constituents, and is
therefore a chemical individual. The authors, on the other
hand, in investigating the product some years ago, noted a very
close resemblance in all the reactions of this complex with
those of jute. It may therefore be included amongst the Ligno-
celluloses. It may be also noted that the formula assigned to
the complex by Erdmann differs by only one O atom from
the empirical formula which we have used for the jute sub-

stance :

Ci2H18Og CiaHjgC^

Glycodrupose Jute lignocellulose

Compound Celluloses 163

The authors have made (1883) the following determinations of
the constituents of these concretions :

Inorganic constituents (ash) . . • . 0-91

Cellulose . . . . . ,26-0 34*2

1 Furfural 18*0 p.ct.

Loss on boiling in 12 p.ct. HC1 (30 rains.) . 53-6

In conclusion, we can only call attention to the desirability of
re-investigating the product, and, upon the evidence of close
similarity to the typical lignocellulose, of adopting, at the outset,
the general plan of investigation laid down for such compounds.

The formation of a lignocellulose under such totally dif-
ferent conditions from those which obtain in a flowering stem
is of especial significance in regard to the physiology of the
production of such compounds.

THE LIGNOCELLULOSES OF CEREALS. — Both in the straws of
cereals, and the seed envelopes of the grain, there is a typical
and characteristic process of lignification. With the formation
of quinone-like bodies, as in jute, there is associated the produc-
tion in the tissue of a large quantity of pentosan derivatives.

The composition of brewers' grains has been carefully inves-
tigated by Schulze and Tollens (Landw. Vers.-Stat. 40, 367),
and an abstract of their results is given in Section III. p. 259.
From the more recent results of Tollens this material has been
found to yield 16-03 P-ct- furfural, corresponding to 26-93 P-ct«
of pentosan. A considerable proportion of the pentosan con-
stituents may be directly hydrolysed to pentaglucose ; on the
other hand, a not inconsiderable proportion is so intimately
united to the cellulose as to resist hydrolytic treatments of
some severity. The lignone constituent was not specially

1 The furfural was estimated by the colorimetric method of comparison
with a standard solution of furfural (V. Meyer, Berl. Ber. n, 1870), the
only method available at the time. As a specimen of jute similarly investi-
gated gave 10 -6 p.ct. furfural, it is probable that the above determination
is 2-3 p.ct. in excess of the true number.

M2

164 Cellulose

investigated by Tollens in regard to its more characteristic
groups, the researches being chiefly directed to the furfural-
yielding groups. What we have to emphasise is the recognition
by Tollens that in this tissue-substance the various groups are so
united as to constitute a homogeneous complex. This tissue
has the closest resemblance to the grain-bearing straws^ which
have been recently investigated by C. Smith and the authors
(J. Chem. Soc. 1894, 472 ; Berl. Ber. 1894, 1061).

The starting-point of these researches was the observation,
already noted (p. 84), that the celluloses, isolated from their
stem tissues, themselves give a large yield of furfural when
boiled with hydrochloric acid ; at the same time none of the re-
actions of the pentaglucoses. It appears from these researches,
and from subsequent results, that from germination, continu-
ously with the growth of the stem, there is a steady increase in
the proportion of furfural-yielding constituents, and that these
are mainly utilised in building up the permanent tissue of the
stem. These results are noted paripassu with lignification, and
they further generalise the chemical features of the process
which were brought out in connection with the jute fibre — viz.

(1) the cellulose of a lignified tissue is, when isolated, found
to be invariably an oxidised and furfural-yielding cellulose ;

(2) in the non-cellulose, pentosan groups are present in associa-
tion with an easily hydrolysable oxycellulose, and with unsatu-
rated or keto R. hexene groups.

As lignocelluloses, the straws are generally differentiated from
the typical lignocellulose, (i) by their structural complexity ;
(2) by their lower carbon and proportionately greater oxygen
percentage ; (3) by the relative susceptibility of the non-cellu-
lose to hydrolysis ; (4) by the much lower percentage of cellu-
lose and the composition of this cellulose.

As a consequence of these differences, the straws are more
easily attacked by the thiocarbonate treatment. The following

Compound Celluloses 165

are the results of an experiment carried out under the usual
conditions.

Undissolved by treatment . • . 40*3 p.ct

Soluble and reprecipitated by acids . • . 32-4 ,,

,, and not reprecipitated by acids . « 27-3 ,,

The straws and products of this class have thus been
investigated in various directions, but by no means exhaus-
tively. A systematic investigation on the lines of research
herein indicated would be a valuable contribution to our
knowledge.

' Crude Fibre.' — ' Rohfaser.' — In connection with the ligno-
celluloses of cereals, the opportunity arises to discuss an artificial
product with which agricultural chemists are familiar under the
above description. In arriving at the nutritive value of food-stuffs
it is necessary to discriminate between digestible and indigestible
constituents. It has long been known that to the former belong
chiefly the proteids, the water-soluble carbohydrates and fats ; and
to the latter, in general terms, the cellular tissue of vegetable
food-stuffs. Between these two extreme groups lies the aggregate
of compounds known as * non-nitrogenous extractive matters.' It
will be evident from discussions in this treatise (p. 86) that this
complex admits of being resolved, by various processes of hydrolysis
and oxidation, into carbohydrates of known constitution, or deriva-
tive products which determine the constitution of the groups from
which they are formed. This aggregate is dissolved by treatment
with weak hydrolytic agents, acid and alkaline, and the residue is
the complex in question, known as crude fibre. A standard process
for estimating this complex, which has been largely, in fact gene-
rally, used by agricultural chemists, is that known as the * Weende
method.' This consists in boiling the material to be analysed with
dilute sulphuric acid (1-25 p.ct. H2SO4), and afterwards with
dilute alkaline solution (1*25 p.ct. KOH), washing, drying, and
weighing the residue. As the process of animal digestion may be
briefly defined as an exhaustive series of hydrolyses under alter-
nately acid and alkaline conditions, the method in question cer-
tainly gives a crude measure of the proportion of the material
resisting the natural process of digestion. On the other hand, as

1 66 Cellulose

an ' aggregate ' method it is open to a good deal of objection ; and,
with the general advance of chemical and physiological methods of
observation, the time has come for a revision of the subject, in
order that the line separating ' digestible ' from ' indigestible *
matters may be defined more in accordance with directly ascer-
tained facts.

In order to show in general terms the nature of the constituents
' digested,' i.e. dissolved by the artificial process, we give an abstract
of a report upon ' Determinations of Crude Fibre and their
Defects,' by C. Krauch and W. v. d. Becke, Landw. Vers.-Stat.
27, 5 (1882).

The residue from the treatments by the Weende method is
generally assumed to be * cellulose and woody fibre,' and, by in-
ference, that these constituents resist the attack of the boiling acid
and alkali. These authors determined the proportions dissolved
from typical food-stuffs by the two treatments, together with the
elementary composition of the aggregates, with the following
results :

(a) Dissolved by the boiling dilute acid ; (b] by the alkali ; and
(c) residue.

(«) (*) GO

Rye (gram) . • • • 52-12 26-48 21-40

Meadow hay . . . ,28-30 21-85 49*85

Clover hay .... 19-47 26-17 54'3&

Elementary Composition of Aggregates

(a) (b) Residue

Rye . .

Meadow hay .
Clover hay .

C

47-6

50-12
42-99

H
6-03
7-08
6-44

0

46-36
42-80
50-57

C

55-12
56-42
51-12

H
7-68
6-49

6-35

0

37-23
37-09

42-53

C

55-n

46-38
49-08

H

7-58

6-36
6-63

O

37-03
47-26
44-29

The above results are calculated with exclusion of the nitrogenous
constituents (albuminoids) and ash. From the high C percentage
of the constituents dissolved, it is evident that the lignocelluloses
are attacked.

In more direct criticism of the assumed digestibility of the
• N-free extractive matters,' the authors investigated cereal ' meals.'
The starch was estimated by the malt extract process, and the
4 N-free extractives ' by the Weende method, with the following
results :

(a)

(j)

to

I5'56

15-89

17-35

2-53

2-74

5-63

:tractives . . . 65-87

65-23

65-28

>re . 8-33

9-17

6-84

771

6-97

4-90

Compound Celluloses 167

(a) (3) to

Starch 62-48 42-02 26-61

N-free extractives . . . 70-38 64*8 66-50

These specimens were selected in accordance with gradations
in recognised feeding value from a to r, gradations corresponding
approximately with the ascertained proportions of starch, but alto-
gether at variance with the numbers for * N-free extract'

In further illustration of the same point the authors cite the
following more complete analysis of meals (Brunner, Landw.
Ztg. Westfal, 1877, p. 19).

Proteids

Fat

N-free ej

Crude fibre

Ash

Direct estimation of starch by

malt method . . . 27-93 3°'4 53^3

It is again evident that the * N-free extractives' are not a
measure of the nutritive value ; but, on the other hand, by a direct
estimation of the starch, the method becomes more complete.

The authors then completed their investigation by taking as the
basis of observation food-stuffs deprived of fats, by extraction with
ether-alcohol, and starch, by digestion with water and malt extract
at 50-60°. The residue, which they termed * Grundsubstanz,' was
then subjected to the Weende method of hydrolysis ; and by deter-
minations of elementary composition of the residues, the com-
position of the dissolved constituents was arrived at.

The specimens investigated were three grades of wheat-brans
(pollards) and two specimens of rice meal. The materials operated
on, viz. residues from the treatments above described, had the
following composition :

Brans Rice meal

<«)' (5 'to (<0~~ """to

C . • .51-82 50-38 48-32 51 '3 39-2

H . • . 7-00 6-34 6-38 7-09 5-12

N . . . 3-17 2-74 0-84 5-14 0-58

O . . .37-22 39-81 43-37 32-17 34-82

Ash . . 0-79 073 1-09 4-30 20-28

1 68 Cellulose

or, calculated to C,H,O compounds only —

(*) (*) to <<*) to

C . . .52-06 50-28 48-69 53-92 48-34

H . . .7-07 6-37 6-42 7-63 6-38
O . 40-87 43-35 44-89 38-47 45-28

(a} The following were the results of the first treatment, boiling
in 1-25 p.ct. H2SO4, in regard to the percentage and elementary
composition of the N-free constituents dissolved :

(«) (*) to (</) to

• «<* «* 5-15 20-72 8-84

ed

Elementary f C *8<96 47'O3 49'I7 50-26 4O'4

composition) « 6'34 5'46 6-48 7-46 8-78

<- O 44-70 47-51 44-35 42-28 50-81

(b) And on subsequently boiling with dilute potash —

(«) (*) to (rf) to

Proportion •)

dissolved }• 20<I4 I2'7S 23-44 20-55 20*57

Elementary fC S7'i7 57'63 5O'83 57'62 53'44

composition H ^5 7'O2 5*62 7-31 6-66

IO 35'i8 35-35 43-55 35-07 39-60

It may be noted that the albuminoids are almost entirely
removed by these treatments, as is evident from the determinations
of N in the residual crude fibre, viz. :

N 0*17 o-oo o-oo 0-42 o«37

The conclusions drawn from these results are, that the con-
stituents dissolved are of relatively high C percentage, that they
consist in large proportion of the lignocelluloses of the raw material,
and that therefore the residual crude fibre is a product of purely
empirical, and in fact arbitrary value.

The subject is also exhaustively treated by A. Miintz, in a
brochure entitled * Recherches sur 1'alimentation et sur la produc-
tion du travail' (Ann. Agron. 1877-8, No. 2). This contains the
results of a very elaborate inquiry into the muscular work of the
horse in relation to the food consumed. The inquiry involved the de-
termination of the nutritive value of typical fodders, as its necessary
basis, and the author's conclusions in regard to the * cellulose '
constituents of these food-stuffs are noteworthy. After showing

Compound Celluloses

169

that these all contain celluloses easily hydrolysed by dilute acids
to * glucoses ' — thus clearly anticipating the later work upon the
celluloses of Class C (p. 85) — the author makes the following
statement in regard to the cereal straws : —

' L'exemple le plus fmppant de cet effet nous est offert par la
paille, dans laquelle le microscope ne nous fait decouvrir aucune
trace d'amidon et dans laquelle pourtant on dose d'apres les
methodes ci-dessus indiquees 20 p.ct. d'amidon ' : the method
consisting in hydrolysing with dilute sulphuric acid (2 p.ct H2SO4)
at 1 08°, and titrating, in alkaline solution, with Fehling's solution
as in glucose estimations. Celluloses thus easily hydrolysed may
be assumed also to be digestible by the animal organism, and to
have a value equal to that of starch. In the case of starchy fodders,
however, the author adds to his scheme a method for the exclusive
estimation of starch, and selects the process of diastatic conversion.
In determining crude fibre (described in the original as ' cellulose
brute') the method of alternate digestion (at 108°) with dilute acid
(2 p.ct. H2SO4) and alkali (5 p.ct. KOH) was adopted. The residues
from these treatments were found to have the composition, and
to be formed in the proportions, given in the annexed table :

—

Oats

Beans

Bran

Hay

Straw

Yield p.ct. . . .

12-42

8-61

6-38

29-83

39*77

{C
jT
Q

45'55
6-28
48-17

45-00
6-48
48-52

49-93
6-99
43-08

48-05
6-36
45-59

47-62

6-35
46-03

the high C percentage being referred to the ' lignin ' group remain-
ing in combination with the cellulose. The proportion of nitrogen
in these residues surviving the treatment is from 0-2 to 0*3 p.ct. —
which is neglected in the calculations. In criticising this process
the author clearly states that it is of purely statistical and approxi-
mate value, even as a measure of non-digestible constituents ; and
as a determination of cellulose, valueless, if not altogether mis-
leading.

Having found in effect that from 20-25 P-ct- of the substance of
straws yields to acid hydrolysis in the same way as starch, and
that the constituent so attacked appeared to be of the nature of
cellulose, the following method is proposed for estimating the

Cellulose

cellulose proper (cellulose rtelle} : The substance (straw) is di-
gested with dilute hydrochloric acid (7 p.ct. HC1) for some hours,
washed, and treated with strong ammonia until all matters soluble
in this reagent are removed ; it is then treated with cuprammonium
solution. After exhaustive action of this solution, the cellulose is
precipitated from the solution by the addition of acetic acid in
slight excess, washed, dried, and weighed. The following results
were obtained :

Process No. I.
Cellulose estimated as described by solution in cuprammonium . 49*44

Process No. 2.

Cellulose dissolved in the process of estimating crude fibre . 21 -99
Cellulose present in ' crude fibre ' (3977 p.ct.) calculated from
its composition * . . . . . . . . .26-54

Total . . . 48-53

The concordance of these numbers is perhaps misleading, since
the lignocelluloses are attacked by cuprammonium. The errors of
the two processes compensating one another, the author arrives at
a determination of cellulose approximating to that of the now
standard methods.

In dealing with the group of substances inctttermine'eS) i.e. the
residue of constituents not determined by the standard methods
of analysis, the author arrives at conclusions similar to those con-
tained in the paper above noted. By the statistical method he
finds the composition of this complex in typical fodder plants to
be as under :

—

Oats

Maize

Beans

Bran

Hay

Straw

c . . .

46-8

45'5

44'I

45 '6

51*3

477

H . . .

6'2

6-5

6-2

6'3

6-2

6-1

0 ...

47-0

40-0

497

48-1

42-4

46-2

numbers which indicate large variations in composition, and there-
fore in nutritive value.

The author's elaborate method of proximate analysis certainly

effects a more complete resolution of this heterogeneous group ;

and the complete scheme of investigation, devised to minimise

the errors of the standard methods, is worthy of attention. It would

1 I.e. as a mixture of cellulose and lignin.

Compound Celluloses

171

take us too far from the purpose of this discussion to reproduce
the scheme in full detail ; it will be sufficiently grasped from the
subjoined statement of the complete results of analysis of straw.

Constitu-

Proportion of elementary
constituents

r

ents esti-

mated p.ct.

C

H

O

N

Hydrolysable cellulose (equi-
valent to starch)

J2I-99

977

1-36

10-86

—

Glucose

0'34

0-14

O'02

0-18

—

Fatty matters
Crude cellulose (crude fibre) .

1-26
3977

0-97
18-96

0-15

0-14
18-28

"

Pectic acid .

0-89

0-37

O-O4

0-48

—

Albuminoids .

3-95

2-12

0-28

0-92

0-63

Sum of elementary constitu-
ents .....

I -

32-33

4-38

30-86

0-63

Total, by elementary analysis
of original straw

48-27

6-35

4475

0-63

Differences = Elementary con-
stituents of substances un-

1 31-80

I5-94

1-97

13-89

o-oo

determined

J

Whence is deduced the per-

^

centage elementary compo-

1

50-I3

6-19

43-68

sition of undetermined con-

|

p.ct.

p.ct.

p.ct.

stituents ....

J

Total

100-00

With the aid of the methods of more recent introduction (see
p. 261), the group of ' undetermined constituents' of the older
analytical schemes may be much more completely resolved ; and
these methods, added to those above outlined, afford a scheme of
sufficient completeness for all the present requirements of agricul-
tural or physiological research.

We have devoted some space to the consideration of these
results, not only on account of the importance of the subject, but as
an illustration of the statistical method of inquiry to which chemico-
physiological investigations have been largely limited. By the
later work on the chemistry of the more complex carbohydrates,
the way is opened for direct investigation of the nutritive value of
the group of constituents formerly aggregated as N-free extrac-

172 Cellulose

lives. The subject is of wide interest, involving questions of import*
ance to the agriculturist and physiologist ; but the methods by which
it requires to be attacked are for the most part purely chemical,
and such as are described in more or less detail in this treatise.

Having now dealt with the more prominent types of ligni-
fication in plants and tissues of ' annual ' growth, it remains
to deal with the lignocelluloses of perennial stems.

(3) Woods and Woody Tissues.— The stem of an
exogenous perennial is a complex of structural elements of
varied form and function. Of these we may distinguish three
main groups : (i) vessels, (2) wood cells proper, and (3)
medullary tissue. When compacted together to form the per-
manent woody tissue, these groups appear to be indistinguish-
able chemically ; they all undergo ' lignification ' ; are charac-
terised by the same reactions ; and, although it has been
stated that they are variously resistant to the action of destruc-
tive reagents, the variation has not been satisfactorily referred
to any fundamental differences of composition. These points
are well discussed by Sachsse in his Chemie u. Physiologic
d. Farbstoffe, Kohlenhydrate u. Proteinsubstanzen (Leipzig,
1877), and we abstract a short resume of his treatment of the
subject.

To the action of the concentrated non-oxidising acids
(HC1.H2SO4) the vessels are generally more resistant. This is
especially the case, however, in the earlier stages of growth.
Thus sections of fleshy roots (Daucus Carota\ treated alter-
nately with dilute potash and concentrated hydrochloric acid
(2-3 days' digestion), are disintegrated by the treatment, the
cellular tissue being entirely broken down ; the vessels, how-
ever, survive, and are isolated free from the cellular matter in
contact with which they were built up. But with older tissues,
on the other hand, no such differentiation is observable : the
three groups of structural elements are equally attacked by

Compound Celluloses 173

purely hydrolytic treatments, however, they may be varied, both
as to reagents and conditions of action.

Similar results are noticed with oxidising agents. Thus
chromic acid solution (20 p.ct.) attacks the parenchymatous
tissue of young growths much more rapidly than the vessels,
which, with a careful regulation of the treatment, may by its
means be isolated more or less perfectly. In the woods the
reagent appears to attack the vessels more rapidly than the
wood cells and medullary tissue ; but any difference of action
is not such as to permit of an isolation of the one or other
group.

Schulze's reagent (HNO3 and KC1O3) also attacks the
several groups more or less uniformly, the differences noted
being rather as between woods of different ages and species ;
thus sap wood is more resistant than heart wood, and the ' soft '
than the ' hard ' woods.

In view of these results, the methods of classification
adopted by Fremy will be seen to rest upon very insecure
foundations. The basis of the classification is the resolution
of the substance by successive treatment with HC1 (dilute and
cone.), H2SO4.H2O, cuprammonium, alkaline hydrates, &c.
Thus the following individuals have been isolated and defined
as follows : cellulose, paracellulose (soluble in cuprammonium
after treatment with acids), metacellulose (insoluble in cupram-
monium) and vasculose (Fremy, Compt. Rend. 48, 862 ;
Urbain, Ann. Agron. 9, 529). This classification has been
severely criticised by Kabsch (Pringsheim, Jahrb. f. Wiss. Bot.
3, 357), and is entirely rejected by Sachsse (loc. «"/.), Hugo
Miiller (Pflanzenfaser, p. 7), and other authorities ; and it is
unnecessary to add anything to the criticisms of these writers.

There have been many attempts to resolve the woods into
proximate constituents, but the authors have for the most part
concluded, from their investigations, (i) that the fundamental

174 Cellulose

tissue of the woods — i.e. the woods freed from adventitious con
stituents such as tannins, colouring matters, resins, &c. — is
composed of substances which cannot be resolved by hydro-
lytic treatments into proximate components ; and (2) that there
is a striking uniformity in composition of the fundamental tissue
of the woods, notwithstanding their structural complexity ;
and this uniformity embraces not merely the individuals of a
species, but extends over the widest range of such products.

Mention should be made here of a general property of the woods,
and lignocelluloses discovered and investigated by C. Wurster,
viz. that of fixing atmospheric oxygen in the form of a peroxide,
giving the reactions of the typical hydrogen peroxide. One of
these is the oxidation of the methylated derivatives of paraphenyl-
endiamine to red colouring matters — and this reaction is equally
and generally characteristic of the woods. Wurster has, in fact,
reduced the reaction to an approximate quantitative estimation of
the proportion of ' mechanical wood pulp ' in papers. The reagent
in question is incorporated in definite proportion with a pure
cellulose paper, which is used as a 'test paper' ; the paper to be
tested being moistened with water, and the test paper pressed into
the moistened part. The depth of colour is compared with standard
coloured papers constituting a scale, the oxidising effects producing
the colour being also expressed in terms of normal iodine solution.
The percentage of wood in the paper corresponds to the depth of
colour produced on the test paper.

For an account of Wurster's researches, which have been
extended to a number of organic products, and are of considerable
physiological interest, see Berl. Ber. 1887, 20, 808 (Quantitative
Bestimmung des Holzschliffes im Papier), and also pp. 256, 263,
1030, 2631, 2934. From these it appears that the reaction, in the
case of the woods, is the expression of quinonic constitution of
characteristic (hexene) groups.

EMPIRICAL COMPOSITION (ELEMENTARY ANALYSIS) OF
WOODS. — The uniform composition of the woods has been for
many years regarded as established on the basis of the analyses
of Chevandier (Ann. Chim. Phys. [3], 10, 129), which are still

Compound Celluloses

175

retained in all text-books, and may therefore be reproduced
here.

—

Beech

Oak

Birch

Poplar

Willow

Carbon . . .
Hydrogen . .
Oxygen . . .
Nitrogen . . .

49-89
6-07

43'"
0-93

50-64
6-03
42-05
1-28

50-6I
6-23
42-04
I -12

50'3i
6-32

42-39
0-98

5175
6-19
41-08
0-98

A series of determinations has also been made by Gottlieb
(J. Pr. Chem. [2], 28, 385). These include the elementary
analyses of the woods, and the calorific equivalents in heat units
per i grm. burned.

Inorganic

Organic

Calorific equivalent

Ash

Wood

C

H

N

Heat units per i grm.

o-37

Oak

50-16

6-03

__

4,620

o-57

Ash

49-18

6-27

—

4»7H

0-50

Hornbeam

48-99

6-20

—

4,728

o-57

Beech

49-06

6-ii

0-09

4,770

0-29

Birch

48-88

6-06

O'lO

4,771

0-28

Fir

50-36

5-92

0-05

5,035

0'37

Pine

50-3I

6-20

0-04

5,085

Cellulose 44-4

6-2

4,146

These results have been extended by G. W. Hawes (Amer.
J. Sci. [3], 7, 585) to the woody tissues of Acrogens — e.g.
Lycopodium, Equisetum, Aspidium, &c. — and from his inves-
tigations he concludes that the same general relationship
obtains for these as in the forest trees above given.

PROXIMATE ANALYSIS. — A large number of the woods are
characterised by special constituents, more or less of the nature
of excreta. To deal with these would lie outside the scope of
the present treatment of the subject. We are strictly limited
to the fundamental tissue of the woods, considered as ligno-
celluloses. Hugo Miiller (Pflanzenfaser, 150) gives the re-
sults of analyses of a representative series, the most important

Cellulose

numbers being the percentages of cellulose isolated by the BrAq
method. His results are comprised in the following table :

Wood

Water

Cellulose

Aq extract

Resin

Non-cellulose

Birch .

12-48

55-52

2-65

I-I4

28-21

Beech .

12-57

45 '47

2-41

0-41

39-I4

Box

12-90

48-14

2-63

0-63

3570

Ebony .

9-40

29-99

9-99

2'54

48-08

Oak .

13-12

39-47

1 2 -2O

0-91

34-30

Alder .

10-70

54-62

2-48

0-87

31-33

Lignum Vitae

10-88

32-22

6-06

1 5 '63

35-21

Lime .

10-10

53-09

3-56

3-93

29-33

Chestnut

12-03

52-64

5-41

I -10

28-82

Fir

12-87

53-27

4^5

1-63

28-18

Mahogany

12-39

49-07

9-9I

1-02

27-61

Poplar .

12-10

6277

2-88

i'37

20-88

Pine .

I3-87

56-99

1-26

0-97

26-91

Teak .

II-O5

43-12

3-93

374

38-16

Willow.

11-66

5572

2-65

1-23

28-74

An account of the applications of Fremy's method of proximate
analysis to the woods will be found in J. Chem. Soc. 1884, 46,
860 (abstracted from Urbain, Ann. Agron. 9, 529-547).

Thus oak wood was purified by treatment with alcohol-ether
(losing 4 p.ct), and afterwards with water and weak alkaline
solutions (losing an additional 10 p.ct.). The residue, or ligno-
cellulose proper, is treated with cuprammonium to remove
cellulose ; the residue boiled with dil. HC1, and again digested with
cuprammonium to remove paracellulose ; the residue is vasculose.
Analysed in this way, the lignocelluloses of oak and poplar give
the following results :

Cellulose Paracellulose Vasculose

Oak . . t .27-05 42-90 30-05

Poplar .... 34'io 45*95 I9-95

When this method has been employed in investigations, the
results should be compared with the results of the methods
adopted in this treatise. The reagents employed will be found to
be not selective in their action according to the distinctive consti-
tutional features of the component groups of the woods.

Schulze investigated the resolution of the woods into
cellulose (insol.) and non-cellulose (sol. derivatives) by the

Compound Celluloses

177

process of digestion with nitric acid (dilute) and KC1O3 (p. 96).
The cellulose was estimated and analysed j the following are
representative results :

Wood

Elementary composition of
isolated cellulose .•

Proximate composition of wood

c

H

Cellulose (esti-
mated)

Non-cellulose
(diff.)

Beech
Oak .
Alder
Acacia

4771
44-51
43-96
44-29

44 '54

6-05

6-oo

5-95
6-09
6-oi

48-41
45-87
47-97
52-94
58-II

51-59
54-I3
52'03
47-06
41-89

A further consideration of these results by the statistical
method, taking the empirical composition of the woods as
the basis of comparison, led Schulze to adopt the formula
CigH24O10 (C = 55*3 p.ct.) as approximately representing
the composition of the non-cellulose or lignone complex — a
formula which is in very close agreement with that which we
have adopted for the lignone of the typical lignocellulose.

N. Schuppe has also investigated the composition of woody
tissue, upon similar lines and with similar results (Pharm-
Journ. [3], 14, 52). He arrives at the formula C19H18O8 for the
lignone complex, and at the mixed expression

as representing the average composition of woody tissue.

It is evident from these results that woody tissue is similarly
constituted to the typical lignocellulose, the main difference
being the higher percentage of carbon (higher proportion of
lignone) and lower proportion of cellulose.

We have not as yet endeavoured to connect the process of
lignification in any definite way with the products — viz. the
lignocelluloses — but the moment has now arrived for briefly
setting forth a general theory of the process. It is many years

N

1 78 Cellulose

since Sachsse definitely propounded the view that the ligno-
celluloses are products of metabolism of cellulose.

Comparing cellulose, Ci8H30O15, withlignone calculated to
a C18 formula, viz. C18H24O10, the latter could be formed from
the former by dehydration (— 3H2O)and deoxidation (— O2).
Sachsse preferred to formulate the process hypothetically as
under (' Farbstoffe,' &c. p. 146)—

C24H40O2o — C6H6O5 — 5H2O = Cj8H24O10|

Cellulose Lignone

the unknown complex C6H6O5 undergoing further change,
either oxidation to CO2, or condensation to aromatic products
(tannins). This view is based entirely upon physiological
evidence, the chemistry of the process being hypothesis, pure
and simple. But we are now in a position to supply a more
substantial and consistent chemical basis.

(1) The celluloses isolated from the lignocelluloses are all
oxycelluloses — i.e. contain furfural-yielding groups. Oxidation
of the celluloses upsets the molecular equilibrium ; the oxy-
products are relatively unstable, and easily condensed to closed
chain derivatives. The celluloses also contain, in certain cases,
methoxyl groups.

(2) The non-cellulose contains, in addition to the closed
hexene rings and methoxyl groups, a characteristic and con-
densed cellulose derivative. This group we can only diagnose
by indirect means. But a careful review of the evidence leaves
no doubt that the cellulose on the one hand, and keto R.
hexene derivatives on the other, regarded as extreme members,
are connected by an intermediate product or group of products,
which can be transformed (a) into cellulosic derivatives ; (&) into
acid products of low molecular weight, chiefly acetic ; (c) into
closed ring compounds, amongst which furfural is prominent.

It is impossible to draw any line of separation between
these main groups : the ' intermediate ' constituent in seme

Compound Celluloses 179

reactions (e.g. chlorination) remains united to the R. hexene
groups ; in others passes into cellulose, and is isolated as such
(regulated oxidation by CrO3) ; in others it is broken down,
and this destruction takes place at the expense of cellulose.
These considerations are dealt with at greater length in a paper
by the authors, 'Celluloses, Oxycelluloses, Lignocelluloses '
(Berl. Ber. 1893, 2520).

(3) Regarding lignification as a process of continuous
modification of cellulose, and the woods as representing the
extreme limits of such a process, these should show an
increase in lignone at the expense of cellulose ; which is in fact
the case. Lignocelluloses in the first year of growth contain
70-80 p.ct. cellulose ; the woods, on the other hand, 50-60 p.ct.

(4) The woods often contain aromatic products of definite
and well-ascertained constitution. We have given some of the
evidence for the view that the tannins in heart woods are direct
products of transformation of the lignocelluloses of the tissue.
We shall presently see that aromatic products are formed in
large quantity in the process of destructive distillation of the
woods, some of the most characteristic of these being pyrogallol
derivatives. We have also seen that the origin of the hippuric
acid of the herbivora has been traced to the lignocelluloses of
the fodder plants. It is, therefore, generally established that
the lignocelluloses are connected on the one hand with the
celluloses, and on the other with the derivatives of benzene,
through a series of intermediate products ; and in the present
state of knowledge it is not difficult to account for the relation-
ships which exist as genetic.

(5) The most convincing evidence, however, is that
furnished by a general review of the numerical constants of the
lignocelluloses. We have already alluded to the general
uniformity in composition of the woods, and therefore for
our immediate purpose we select a typical wood (beech) for

N 2

i8o

Cellulose

comparison, on the more essential points, with the typical
' annual ' lignocellulose — viz. jute.

—

Elementary
composition

Proximate
resolution

Quantitative reactions of
non-cellulose

Carbon

Hydrogen

Cellu-
lose

Non-
cellulose

Methoxyl

Furfural

Cl combining

Jute
Beech .

46-5
49'I

—

75
55

25

45

4*0
6-2

8-2

12*8

8-0

12 -O

In regard to the number thus obtained for beech, i.e. after
merely boiling in water, it is necessary to point out the cause of
the difference from the number given on p. 195, viz. 8-o obtained
after exhaustion with alkali, and calculated on the product so
exhausted. The alkaline treatment removes a complex made up
of pentosan, acetic residue, and keto-hexene constituents, the
removal of which gives a residue approximating in composition to
the jute lignocellulose ; whereas in the complex removed, the pro-
portion of the R. hexene groups, which react with chlorine)
is much higher ; and hence the increased proportion of Cl com-
bining.

It is unnecessary to enlarge upon the simple relationships of
these numbers. Sachsse's theory of lignification, it must be
remembered, was based upon purely physiological grounds,
and could only be supported by the meagre evidence of such
empirical determinations as were then available. Now that
we bring to bear the results of quantitative determinations
based upon specific constitutional features, and find these
perfectly consistent with this theory, the time has arrived to
press its consideration as a generalisation of wide import, con-
cerning the constructive processes of the organic world. The
theory briefly expressed, and in its more enlarged scope, is this :
that the process of lignification consists in a series of pro-
gressive and intrinsic modifications of a cellulose or oxycellulose

Compound Celluloses 181

tissue, the products of modification remaining associated with
the residues of the parent substance in a state of combination
or of intimate mixture, the final products of metabolism
(aromatic products, pentosans, &c.) being excreted and taking
no further part in the organic processes of the tissue.

We pass now from general deductions as to the con-
stitution of the woods to the consideration of the results
of special investigations (a) of particular constituents or
reactions of the woods as a class ; (b) of particular woods.

(a) With the advance of analytical methods, characteristic
constituents of the woods may now be determined with pre-
cision. Estimations of furfural and methoxyl have been carried
out in extended series, and the results are significant for their
uniformity on the one hand, and for the indications of pro-
gressive variation with the progress of lignification on the
other.

(i) ESTIMATIONS OF FURFURAL. — Furfural-yielding consti-
tuents.— In a recent publication of Professor Tollens (Zeitschr.
Ver. f. d. Riibenzucker Industrie, 44, No. 460) he gives a resume
of results obtained by himself and others who have collaborated
with him in the investigation of the furfural-yielding con-
stituents of plant tissues. The following numbers have been
selected as the mean results for the more important woods :

Wood

Yield of furfural p.ct.

/ Beech
Hard woods \ Oak
1 Birch
Pine

12-6

107

137
S'O

De Chalmot has made a more extended series of deter-
minations (Amer. Chem. Journ. 16, 224), the results of which
are given in the table on p. 182.

There is therefore a striking uniformity amongst the woods

182

Cellulose

of the Dicotyledoneae in regard to this constituent or group
of constituents, and an equally striking divergence in the case
of the woods of the Coniferae. This differentiation of the
Coniferae will be dealt with in the section devoted specially to
the group (p. 197).

Species

Family

Liriodendron tulipifera

Magnoliaceae

9'5

"Magnolia acuminata

Magnoliacese

8-8

Prunus pennsylvanica

Rosaceae

9-8

Cercis canadense

Leguminosse

10-5

Acer dasycarpum

Sapindaceae

II'O

Liquidambar styraciflua

Hamamelideae

10-5

Ilex opaca

Ilicineae

12-3

Ampelopsis quinquefolia
Cornus florida

Vitaceae
Cornacese

10-4
10-8

Nyssa sylvatica

Cornaceae

IO'4

Fraxinus americana

Oleacese

87

Fraxinus platycarpa

Oleaceae

87

Juglans cinerea
Carya alba

Juglandaceae
Juglandacese

9-6
10-6

Salix spec.

Salicinese

10-5

Betula spec.

Cupuliferae

117

Fagus ferruginea

Cupuliferae

10-5

Quercus Phellos

Cupuliferae

10-8

Quercus alba

Cupuliferae

10-2

Quercus rubra

Cupuliferae

10-8

Quercus nigra

Cupuliferae

107

Ulmus americana

Urticaceae

87

Platanus occidentalis

Platanae

10-8

Juniperus virginiana

Coniferae

5'2

Pinus strobus

Coniferae

37

Pinus mitis

Coniferae

4 '4

Tsuga canadensis

Coniferae

3'0

De Chalmot has further investigated the question of the
life-history of the woods in reference to this group of con-
stituents, and has established the important conclusion that
they preserve a more or less constant ratio to the entire wood-
substance. The determinations were made upon the wood of
particular and successive rings in horizontal sections, the posi-
tion of the rings in the section determining the age of the wood-

Compound Celluloses 183

substance. The following observations may be cited (loc. cit.
p. 225) :

Wood of Quercus nigra.

Furfural
2-12 years old ..... 10-6 p.ct.

69 ,, 10-5 „

109-110 „ 10-3 „

Liriodendron tulipifera.

19-21 years old 9-8 p.ct.

59-6o „ 9-4 „

Magnolia acuminata.

Year rings of this tree not distinctly discernible.
Younger than i o years . . . 8 '8 p.ct.
Old heart wood . . . . 8-4 „

Platanus occidentalism

4-10 years old . . . . .9-8 p.ct.
71-79 „ (Heart wood) . . 97 „

Prunus pennsylvanica.

f(l) 4-12 years old .... I o -8 p.ct.

*(2) „ „ 10-4 „

,(i) 60-70 „ .... 9'8 „

1(2) „ „ 9-6 „

In the above it will be observed there is a slight diminution
in the furfural with age, but the maximum difference may be
taken at i p.ct. on the wood, or approximately 10 p.ct. on the
furfural-yielding constituents themselves.

In other specimens, on the other hand, a slight increase was
observed, thus :

Juglans cinerea.

Furfural

j (I ) 4-18 years old .... 9-1 p.ct

„ .... 8-9 „

60-65 » .... 9-6 „

0 »t • • • • 9'5 »>

Fraxinus americana.

2-7 years old ..... 8*9 p.ct

51-52 „ 9'4 »

<8

1 84 Cellulose

It is therefore established by these results that the furfural
yielding constituents of the wood-substance undergo very little
change with age. It is necessary to point out that De Chalmot
uses the term ' pentosan ' as identical with * furfural-yielding
compound,' but this requires some qualification. The forma-
tion of furfural is an empirical and an aggregate result, and,
while specially characteristic of the pentoses, is also a property
of certain oxidised derivatives of the hexoses, notably glycuronic
acid. It is probable that the furfural may be formed imme-
diately, in this case also from a C5 derivative, a product of
resolution of the glycuronic acid— a view which is supported by
the observation that the acid when boiled with hydrochloric
acid yields carbonic anhydride in quantity corresponding with
the equation

C6H8O6 = C5H4O2 + CO2 + 2H.2O,

viz. 26-5 p.ct. CO2. The yield of furfural, on the other hand,
is only 15*3 p.ct. ; but this discrepancy may very well result
from secondary condensations of the C5 aldose (Mann and
Tollens, loc. tit.), in consequence of which only the small pro-
portions are decomposed in the second stage according to the
equation.

The import of these qualifying considerations is, perhaps,
rather physiological than chemical, showing that a number of
minor changes may be taking place in the furfural-yielding
groups without affecting their proportion to the lignocellulose
as measured in terms of this end-product of their decomposi-
tion.

There is evidence that such changes do take place with age,
resulting in the formation of pentosans as such. The dicotyle-
donous woods all contain the body known as wood gum (Holz-
gummi), which appears to consist, for the most part, of xylaa
This substance is extracted by treatment of the ground wood
(sawdust) with solutions of sodium hydrate (2-5 p.ct. Na2O) in

Compound Celluloses 185

the cold ; from the solution the ' gum ' is precipitated on the
addition of alcohol. The yield varies from 10-20 p.ct. of the
weight of the wood. The gum is easily hydrolysed by boiling
dilute acids with formation of xylose. Jute, on the other hand,
gives only very small yields of this product, Tollens obtaining
only 175 p.ct. by digesting the fibre with 5 p.ct. NaOH solution.
This product also yields xylose on hydrolysis.

It is this difference of yield of the proximate product which
requires to be emphasised, as it is altogether out of proportion
to the relative yields of furfural. With the progress of
lignification, in fact, there is probably a progressive formation
of pentosan resulting from molecular changes within the
particular group. These pentosans differ from the parent sub-
stance or complex in readily yielding to hydrolysis, and to
this extent may be regarded as dissociated or split off from
the fundamental tissue-substance — in other words, as excreta
or end-products of metabolism. This view necessarily is in-
volved in the wider question of the physiological significance
of the furfural-yielding constituents of plants. De Chalmot
has published a series of communications in elucidation of
this question, under the titles 'Soluble Pentoses in Plants/
c Pentosans in Plants,' &c. (Amer. Chem. Journ. 15, 16). One
important result of these investigations is the conclusion that
pentosans are not formed in any perceptible quantity by the
assimilation process. This is equivalent to the statement that
they must arise by secondary transformations of the hexoses
before or after their elaboration into the permanent tissue of
the plant.

Investigations of the germination process in relation to
pentosans have given variable results : in some cases there is
an increase in the total pentosan, in others a decrease ; and in
certain cases the pentosans of the seeds, e.g. of Tropaolum
majus, appear to behave as * reserve materials.' The authors

Cellulose

have also investigated this question by studying the germina-
tion of barley. In the germination of barley there is not only
an increase in the * total pentosan,' but the early permanent
tissue is found to contain a considerable proportion of these
furfural-yielding constituents. Of these constituents, moreover,
more than 80 p.ct. resist the process of alternate digestion in
cold dilute acids and alkalis ; they are not therefore pentosans
in the ordinary acceptation of the term. On the other hand,
the pentosans proper are found in relatively large proportion
in the later stages of growth of the cereal straws ; and, again,
the evidence leads us to regard the pentosans as secondary
products of metabolism, in contradistinction to primary products
of assimilation. It is evident from this brief outline that the
physiology of the pentosans — their origin, fate, and general
significance — is still, in many directions, problematical. In
regard, however, to the narrower problem of bonification, we
may sum up the evidence as follows : The formation of furfural-
yielding products invariably accompanies lignification. These
products exist in the earlier stages of lignification in the
cellulosic form, but with age (perennial stems) are gradually
transformed into pentosans of relatively low molecular weight,
and ceasing to occupy any organic relationship to the tissue.
The proportion of these constituents is uniform (18-24 p.ct.)
over a wide range of woods hitherto investigated, and varies,
moreover, but little with the age of the wood ; the proportion
is, however, much less in the woods of the Coniferge (6-9 p.ct),
which therefore represent lignification of another chemical
type (see p. 197). So far no relation has been traced between
the percentage of ' pentosan ' and the physical properties of the
woods.

Before passing from this section of the subject we must
describe somewhat more in detail the characteristic product,
wood gum, already briefly noticed.

Compound Celluloses 187

Wood gum was first isolated and investigated by
T. Thomsen (J. Pr. Chem. [2], 19, 146), and Poumarede and
Figuier (Annalen, 64, 388). It is obtained from the woods of
the ash, elm, oak, beech, willow, cherry, &c., by digestion with
solutions of the alkaline hydrates as already described. The
following particulars of later investigations by Wheeler and
Tollens (Landw. Vers.-Stat 39, 437) are noteworthy. After
extraction from beech wood, and precipitation by alcohol, the
product is purified by digestion with alcohol and hydrochloric
acid, and washing first with alcohol and then ether. It is ob-
tained thus as a white powder. In alkaline solution it exhibits
strong Isevo-rotation (d)D = —69-6°. The yield is approxi-
mately 15 p.ct. of the wood. Hydrolysed with boiling acids it
gives a large yield of crystallisable xylose. Cherry wood under
the same treatment yields 12-13 P-ct- °f tne product, also
yielding xylose as the chief product of acid hydrolysis. It is
noteworthy, on the other hand, that ' cherry gum,' the well-
known exudation from the tree, yields the isomeric pentaglucose
arabinose as the chief product of hydrolysis.

The cereal straws also yield, under similar treatment, 14-17
p.ct. of the product, but retaining a large proportion of the in-
organic constituents of the straw, chiefly silica. This product
shows a stronger rotation, viz. (a)D = — 84-1°. With the aid of
heat in the alkaline digestion, a much larger yield of the pro-
duct (26 p.ct.) is obtained.

Wood gum is insoluble in cold water, but slowly dissolves on
boiling with water ; on cooling, the solution is strongly opal-
escent ; but on the addition of alkali in small proportion, a per-
fectly clear solution is obtained. The compound is insoluble in
aqueous ammonia, and, in the process of isolating it, the raw
materials are therefore usually subjected to a preliminary diges-
tion with dilute ammonia, which removes colouring matters, &c.

Numerous (elementary) analyses of wood gum have given

1 88 Cellulose

numbers corresponding approximately with the empirical
formula C6H10Or,, which is confirmed by more recent results
(Tollens). The only value of these results, however, is to esta-
blish a normal ' carbohydrate ' formula. The more important
problem of its constitution has been elucidated, as already noted,
by its yielding the pentaglucose xylose as the main product of
proximate hydrolysis (acid), and furfural as the ultimate product
(HC1). The most recent analyses of Tollens gave the following
results :

Furfural Xylose

(Specimen i . . . 3878 74-26

„ ii . . . 46-90 89-82

,, m . . . 48-08 92-02

„ iv . . . 33'30 6373

These specimens were from beech wood, variously pre-
pared : No. i by the process already described ; Nos. ii and
iii by extraction with alkaline solution, after boiling the raw
material with dilute sulphuric acid ; No. iv by extraction with
boiling milk of lime. It is evident from these results that
wood gum is a pentosan — the amorphous anhydro- aggregate
of xylose or xylan — mixed or combined with variable propor-
tions of a carbohydrate of similar empirical composition, prob-
ably a cellulose derivative. It also generally contains meth-
oxyl (2-6 p.ct. O.CH3). These observations further confirm the
view that the pentosans are derived from hexose groups, and
represent the final terms of a series of transformations of which
' wood gum ' as directly obtained may be taken as representing
the intermediate terms.

METHOXYL DETERMINATIONS. — The O.CH3 group, as a
chemical constant of lignification, has been brought into promi-
nence by the investigations of Benedikt and Bamberger ; their
most important communication on the subject appearing
under the title, * Ueber eine quantitative Reaction des Lignins '
(Monatsh. n, 260-267). Employing the perfected method of

Compound Celluloses 189

Zeisel, these observers have made an elaborate series of estima-
tions, the results being expressed as percentages of methyl
(CH3) calculated on the dry substance. They are as under :

A. WOODS.

CHS p.ct.
Maple . Stem .... Acer Pseudo-plat anus > L. 3 -06

„ . „ extracted1 „ „ . 3'°5

„ . ,, shavings . . ,, ,, • 3'°6

Acacia . Branch .... Robinia Pseud-Acacia, L. 2 -37

,, . Extracted ... „ „ • 2-45

Birch . 3 years old ... Betulaalba . . .2-57
Pear . . Stem .... Pyrus communis, L. • 3*21

Oak . . „ . . . . Quercus pedunculatus . 2-86

,,..,,.... „ „ • 2-63

Alder ....... Alnus glutinosa . .2*89

Ash . . Stem .... Fraxinus excelsior ', L. . 271

„ . . Shavings from stem „ „ . 2-69

,, . . Stem shavings extracted . ,, ,, • 2'66

,, . . Shavings from branches . /,, ,, • 3'O2

/Shavings from branches!
" ' '1 extracted . . .) "
Fir • . Stem .... Abies excelsa . • ,2*15

„ . . , ..... „ „ . . . 2-39

,, . . „ (central zone) . „ ,, . . . 2-59
„ . „ (sap wood) .,„„... 2-32

,, . . ..... Abies pectinata^ DC. . 2-45

Pine . . ..... Pinus sylvestriS) L. . 2-25

„ . . Stem .... Pinus laricis . .2-05

„..„.... „ • • 2-12

Cherry . ,, . , . . Prunus Avium, L. . 2-38

Larch . ,, . . . . Larix europaa, DC. . I '99

.„....„ „ • • 2-68

Lime. . ,, . . . . Tilia parvifolia . . 2-56

Mahogany . ,, . . . . Swietenia Mahagoni> L. 2 -66

Walnut . „ . . . . Juglans regia, L. . .2-27

„ . Shavings from stem . ,, „ . .2-69

Poplar . Stem .... Populus alba . . 2 *59

1 ' Extracted ' signifies previously exhausted with water, alcohol,

and ether. Otherwise the specimens were analysed without previous
preparation.

190 Cellulose

CH,p.ct

Beech • Stem • • . . Fagus sylvatica . 3 -02

. 2-62

. 270

. 2-92

. 275

. 2-31

shavings

, y •UATUIga • • , ,

Elm ,„.... Ulmus campestris

,, . * „ shavings extracted . ,, „

Willow .... Salixalba

B. FIBROUS PRODUCTS.— Natural and prepared.

Jute (Lignocellulose} I '87

Swedish filter paper o-o

Cotton o-o

Flax, unbleached . . . • Linum usilatissimum . O'O

Hemp ,, . . . . . Cannabis sativa . .0-29

China grass,, . . . . • Bohmeria nivea • .0-07

Sulphite {Cellulose) .... Pinus sylvatica • • 0*34

C. MISCELLANEOUS.

Cork .••.... Quercus suber . .2-40
it •••••••• » >» • • 2*47

Nutshells Juglans regia . . 3 74

Lignite (Wolfsberg) 2-44

Brown coal • • 0*27

From these determinations it is evident that the formation
of methoxyl groups is an essential feature of lignification, and,
moreover, that the formation takes place with remarkable
uniformity over a wide range of woody tissues. This uni-
formity is, indeed, such that Benedikt and Bamberger proposed
to adopt the ' methoxyl number ' as the quantitative measure of
any wood lignocellulose present in an unknown fibrous mixture,
e.g. for determining the proportion of ' mechanical wood pulp '
in papers. From the above table it would be easy to calculate
the degree of approximation (probable error) to be attained,
and we may be satisfied to note that the approximation is
sufficiently close to make such determinations distinctly valu-
able for the purpose in question. These authors were also
enabled to draw from their results certain conclusions of
physiological significance, viz. : (i) there is in the woods a slight

Compound Celluloses 191

progressive increase of methoxyl with age ; (2) there is a higher
proportion of methoxyl in the wood of the branches as com-
pared with the main stem ; (3) the proportion of methoxyl is
unaffected by ' extracting ' the wood, i.e. it is a characteristic
constituent of the wood-substance (lignocellulose) itself.

THE ACETIC RESIDUE. — Acetic acid is produced in a
number of the decompositions of the lignocelluloses (ante,
p. 1 60). It is obtained more readily, and in larger proportion,
from the (dicotyledonous) woods than from jute (et similid).
The following reactions producing acetic acid may be cited :

(1) Alkaline hydrolysis. — The solutions obtained by treating
beech wood with dilute aqueous alkalis contain acetic acid
(acetate of soda), which is separated by distillation after acidi-
fication. The proportion is large, amounting to 7-8 p.ct. on
the wood.

(2) Add hydrolysis. — Acetic acid is formed on digesting the
woods with dilute sulphuric acid at 60-100°. Larger yields
are obtained by dissolving the wood-substance in concentrated
sulphuric acid in the cold, diluting and distilling.

(3) Oxidising processes. — (a) Acid. — The wood, in fine
shavings, is covered with normal sulphuric acid, and oxidised
at ordinary temperatures, with its own weight of chromic acid
(CrO3) added in successive quantities. The solution on distil-
lation yields acetic acid, equal to 5-6 p.ct. of the weight of
the wood (dicotyledonous). The following are the results of
actual determinations :

Beech Sycamore Birch

5*0 p.ct 5'2p.cL 6-op.ct.

Oxidised with dilute nitric acid (10 p ct. HNO3) at
60-100° (ante, p. 146), very much larger quantities of acetic acid
are obtained, viz. from 10-15 P-ct- of the weight of the wood.

(b) Alkaline. — The maximum yields are obtained in the
drastic decomposition, determined by heating with the alkaline

192 Cellulose

hydrates at 200-300°, The quantity obtained in this way is
from 30-40 p.ct. of the weight of the wood, together with a
considerable quantity of oxalic acid.

(4) Destructive distillation of the woods (see p. 204) also
determines the formation of acetic acid. The following estima-
tions of comparative yields are given by W. Rudnew (Dingl. J.
264, 88 and 128), the woods being 'distilled' in glass vessels
at 150-300°.

Linden • • . 10*24
Birch . . .9*5
Aspen • . . 8'o6

Oak . . .7*9
Pine . . .5-6
Fir . 5-2

Wood celluloses (birch and pine), isolated by the Schulze
process, gave under similar conditions the following yields :

Birch cellulose . . , .6-2
Pine „ .... 5'0

From these results it is evident that the CO.CH2 grouping is
a characteristic constitutional feature of the lignocelluloses. It
also occurs in derivative forms amongst the products of decom-
position of the lignocelluloses by 'natural processes. Thus, e.g.,
in hippuric acid, benzoyl-amido-acetic acid (p. 151), and in
phloroglucol, regarded as 3CO.CH2, which occurs in the plant
world in a number of derivative forms, and is obtained from
several of the natural tannins as a product of fusion with
alkaline hydrates. We are not yet in a position, however, to
localise the CO.CH2 groups in the complex lignocellulose
molecule, and we cannot go beyond a summing-up of the
evidence in general terms.

(i) Acetic acid is a product of simple hydrolysis, both acid
and alkaline, of the lignocelluloses, the proportion being from
3-6 p.ct. of the parent substance.

The formation of an acetic residue is thus a characteristic
feature of lignification. If derived from a hexose group (cellu-

Compound Celluloses 193

lose), it should be formed correlatively with the furfural-yielding
compounds ; and the quantitative relations of the two certainly
confirm this view. Thus the hypothetical decomposition may
be formulated as under :

2C6H12O6 = 2C5H1005 + C2H4O2,
2x150 60

and the pentosans of wood represent in effect a percentage
approximately five times that of the acetic acid obtainable
by simple hydrolysis.

In the jute fibre also, the smaller proportion of the fur-
fural-yielding constituents is associated with a similar smaller
proportion of the acetic residue. The formation of both
therefore increases, paripassu, with age, which is in accordance
with the view of a common origin.

(2) The celluloses, and the 'carbohydrates' generally, are
susceptible of the ' acetic condensation.' The normal cellu-
loses, however, require the application of drastic treatments,
e.g. fusion with alkaline hydrates or warming with concentrated
sulphuric acid, both of which treatments are of an oxidising
character. The oxy celluloses , on the other hand — notably the
straw celluloses — give a considerable yield of acetic acid
(together with furfural) on long boiling with 10 p.ct. sulphuric
acid. The maximum yield is obtained by dissolving the oxy-
cellulose in the concentrated acid in the cold, diluting and dis-
tilling. In this way the authors have obtained a yield of
9-10 p.ct. of the acid, calculated on the oxycellulose.

These observations confirm the view that lignification is a
process of transformation taking place in oxidised celluloses,
or oxycelluloses, and following as a secondary result of the dis-
turbance of equilibrium set up by the oxidation.

(3) In addition to acetic residues — converted by hydrolysis
into acetic acid — there appears to be a CO.CH2 nucleus, a
dehydracetic residue, which is the source of the increased

o

194 Cellulose

yields of acetic acid under the action of dilute nitric acid. Of
this constituent of the non-cellulose groups we have some
indirect knowledge. Thus, in the case of jute, we have given
to the entire lignone complex the statistical formula CjgH^Og.
A portion of this, reacting with chlorine to form mairogallol,
may be approximately formulated as C18H18O9 = 3[C6H6OS].
If we therefore resolve the complex into

C6H6O3 2OCH3 'CnHjoO*

Keto R. hexene Methoxyl
group

we are left with the highly condensed group CnH10O4, con-
taining the furfural-yielding constituents, and also yielding
acetic acid as described.

The constitution of this complex must be considerably
removed from that of the ordinary carbohydrates. Whether
hexoses or pentoses are represented, either must be in the form
of a polyanhydride ; and the acetic residues are also probably
of the dehydrated or CO.CH.2 form.

The further investigation of this problem is the work of the
immediate future, and it is with the view of setting forth some
of the probabilities involved that the discussion has been
pushed somewhat beyond the limits of ascertained fact.

THE CHLORINATION REACTION. — The reactions of the wood
lignocelluloses with chlorine have not been systematically in-
vestigated. It must be remembered that a wood tissue is a
complex structure, and although it will have become evident
that there is a remarkable uniformity in chemical composition,
still a mixture is always less attractive as a basis of investigation
than a homogeneous substance such as the jute fibre. It has,
however, been sufficiently established by research that the re-
action of the dicotyledonous woods with chlorine is identical
in general features with that of the typical lignocellulose — i.e. a
yellow-coloured quinone chloride is formed, giving the same
brilliant colour reaction with sodium sulphite ; and on treat-

Compound Celluloses

195

ment with alkali there is a complete resolution into cellulose
(insoluble) and soluble derivatives of the lignone complex.

The coniferous woods, on the other hand, react somewhat
differently, the chief distinctions being that the wood-substance
is changed in colour to an orange red, and the product does not
give any marked colour reaction with sodium sulphite. In
both cases the percentage of chlorine combining with the ligno-
cellulose is the same as with jute, viz. 8*0 p.ct.

Comparative experiments upon four typical woods gave the
following statistics of reaction with chlorine. The results are
given in terms of the lignocellulose proper, i.e. the residue from
exhaustion with the alkaline solution (i p.ct. NaOH).

—

Pine

Beech

S yea. mo re

Birch

i \ ( Residue from alk. treatment, or »
* ' 1 lignocellulose : p.ct. on wood >

89

82

84

87-5

/ Cl. combining : p.ct. on (a)
(3) \ Acidity after chlorination calc.

7'S

7'5

9-0

7-0

( as HC1

23-5

19*5

21 'O

iS'O

(f) Cellulose : p.ct. on (a)

72-0

65-0

70'O

72-5

These must be regarded as preliminary results, but they serve
to confirm the view we have taken of the general and close
similarity of the woods to the typical jute lignocellulose. It has
not been determined whether the whole of the * acidity ' developed
in the above chlorinations is due to HC1, or to acid products (e.g.
acetic acid) split off from the lignocellulose.

The chlorinated derivatives have not been closely investi-
gated. The authors have isolated one of these products ob-
tained from a Spanish mahogany, the chlorination being pre-
ceded by the usual treatment with boiling dilute alkali (i p.ct
NaOH). This product was found to contain 30-4 p.ct. Cl.

In regard to investigations involving the chlorination of
these lignocelluloses, two points must be borne in mind : (i) As
regards preparation of the material. To ensure a complete
reaction the wood must be reduced to the finest possible
shavings. (2) In regard to the preliminary treatment with

02

Cellulose

boiling alkali. The woods are not attacked as a whole as with
the jute fibre, the furfural-yielding constituents (pentosans)
yielding much more readily than the fundamental tissue or
lignocellulose proper. In systematic investigations following
the lines laid down in the case of the jute fibre, the latter
should be taken as the basis of observation, and not the entire
wood substance.

The reactions which we have so far discussed are, in the
main, reactions of decomposition. Synthetical reactions of
the wood lignocelluloses have been still less investigated.
Here, again, there is little to attract the chemist in the pre-
sent state of our knowledge, owing to the necessary complexity
of the reactions involved. From such reactions as have been
studied, if only in a general and superficial way, it appears that
the proportion of reactive OH groups is still less in these ligno-
celluloses than in those of which jute is the type. Thus, to
select the reaction of nitration : The maximum yield of
nitrate is considerably lower in the woods than in jute ; more-
over, the reaction is complicated by a destructive oxidation
which supervenes at a very early stage of exposure to the
action of the mixed acids. The following series of determina-
tions of yield in the case of mahogany wood illustrate this
point. In (a) the wood was used in its raw state ; in (b) it was
previously purified by boiling in dilute alkaline solution.

Nitrating acid : equal volumes of H2SO4 and HNO3 (i*43
sp.gr.) in excess.

Duration of exposure
to acid

Yield of nitrated wood

(«)

(*)

Mins.

I
2

3
4
5

P.ct.
106-6
Il8'4
126-5
1127
108-8

P.ct.
115-6

121 -O
127-2

I25-3
I23-I

Compound Celluloses 197

After three minutes' exposure, therefore, in both cases
oxidation supervened, accompanied by conversion into soluble
products ; this destructive oxidation being much more marked
in the case of the raw wood substance. Jute, under similar
conditions of treatment, would have given a maximum of 145
p.ct, and the nitrate is much more resistant to the continued
action of the acid mixture.

These results are, of course, of slight value only ; but they
serve to give emphasis to the general conclusion that lignifica-
tion is a process of condensation and etherification of OH
groups, accompanied, and in part conditioned, by condensa-
tion in regard to carbon configuration. Similarly, also, the
woods show considerably more resistance to the actions of
solvents of cellulose than jute lignocelluloses ; notably to the
thiocarbonate reaction, to which they yield only in very slight
degree and after prolonged exposure.

From this general view of the reactions of the woods con-
sidered as a class of the lignocelluloses, we proceed to consider
special investigations of particular woods.

Woods of the Coniferae. — These woods are of very
great industrial importance, not merely for their uses as such,
but as the raw material for the preparation of the ' sulphite
wood pulp,' now produced on an enormous scale in connection
with the paper industry. The ultimate fibres of these woods
are of greater length than those of the dicotyledonous woods ;
in addition there are well-marked features of distinction in
chemical composition from the latter, which have already been
noted.

The chemistry of these woods was investigated some years
ago by Erdmann (Annalen, Suppl. 5, 223). The wood of
Pinus abies— purified from adventitious constituents by boiling
in acetic acid, and subsequent exhaustion with water, alcohol
and ether — gave, on ultimate analysis, constant numbers, viz.

198 Cellulose

C, 48-4 ; H, 6*3 p.ct. From these results, together with general
observations on the chemical behaviour of the substance, it
was concluded that it is a homogeneous compound, having the
empirical formula C3oH46O2i. Erdmann further concluded
that this compound is resolved by hydrolysis with dilute
acids into glucose (soluble) and a residue C26H26On, which
he terms lignose, and the original compound therefore
gtycotignosc. Lignose was further found to yield, on fusion
with potassium hydrate, pyrocatechol and protocatechuic
acid.

These results, or rather the interpretation of them, is incon-
sistent in many respects with the results of subsequent investi-
gations. The experimental facts, however, remain ; and the
researches are worthy of notice, as one of the earliest attempts
to elucidate the constitution of the lignocelluloses as a definite
chemical problem.

The * sulphite pulp * process would appear to offer a much
more promising field of investigation, since it not only deter-
mines a satisfactorily sharp separation of cellulose (pulp) from
non-cellulose (soluble sulphonated derivatives), but with the
minimum of chemical modification of either group. Notwith-
standing these specific advantages of the process, considered
as a method of proximate analysis, and numerous investigations
of the soluble by-products ('sulphite liquor'), the constitution
of the latter, and therefore of the original lignocellulose, still has
to be expressed in very general terms. The most important
contribution to the subject is that of Lindsey and Tollens
(Annalen, 267, 341), of which the following is a brief account.
The solution used in these researches was that resulting from
the * Mitscherlich process,' which consists in a prolonged diges-
tion of the wood— after subjection to a preliminary mechanical
disintegration— with a solution of calcium bisulphite. The
solution is usually prepared to contain CaO 1-35 p.ct, SO?

Compound Celluloses 199

4*4 p.ct., and is used in the proportion of 5-7 parts to i part of
wood. In the digestion, the temperature is gradually raised to
160° C.

The particular specimen of 'waste liquor* (1*055 sp.gr.)
used in the above researches contained 9*5 p.ct. of ' total
solids in solution' (dried at 100°), of which 0-58 p.ct. wasCaO.
The solution has a pale brownish-yellow colour, and reduces
Fehling's solution strongly.

A systematic examination for the presence of carbohydrates
of low molecular weight and known constitution gave for the
most part negative results as follows :

(a) On boiling with HC1 (16 p.ct. on the solution) after
evaporation to a suitable volume, traces only of levulinic acid
were obtained ; showing the general absence of such carbo-
hydrates. (Annalen, 243, 333 ; Berl. Ber. 22, 370.)

(b) On oxidation with nitric acid no saccharic acid was
formed, showing the absence of dextrose or dextrose-yielding
compounds. (Annalen, 249, 222.)

(c) On oxidation with nitric acid, traces of mucic acid were
obtained, showing the presence of galactose (or galactan) in
small proportion. (Annalen, 232, 186, 205.)

(d) The solution was acidified with sulphuric acid, boiled
some time, neutralised (CaCO3), filtered, evaporated to a
syrup, and boiled with strong alcohol. The clear solution was
poured off, the alcohol evaporated, and the resulting syrup
mixed with phenylhydrazine acetate. An insoluble hydra-
zone was obtained, which proved to be mannose hydrazone.
An approximate estimate of the quantity showed o^-o'S
p.ct. on the solution, i.e. about 6-7 p.ct. of the 'organic
solids.1

(e) On ' distillation ' with hydrochloric acid, furfural was
formed in some quantity. After precipitation of the bulk of
the organic substances in solution with lead oxide, a solution

2OO Cellulose

was obtained which gave the brilliant colour reactions of the
pentoses, and xylose was identified in the solution by precipita-
tion with phenylhydrazine.

(f) Experiments were also made with the view of deter-
mining alcoholic fermentation (yeast) of the dissolved com-
pounds. Small quantities of alcohol were obtained, but the
maximum yield corresponded to i'2 p.ct. only of carbo-
hydrate.

The major proportion of the dissolved organic substances
was found to be a gummy body with the usual ill-defined
physical properties of the class of organic colloids. On the
other hand, this body behaved, in many respects, as a homo-
geneous complex ; and although it was found impossible to
resolve it into proximate constituents of definite character-
istics, it yielded a number of synthetical derivatives, from
the analysis of which, compared with that of the gum itself,
the conclusion resulted as to the homogeneity of the
complex. The complex was obtained in various forms as
follows :

(1) As a precipitate on adding hydrochloric acid to the
original liquor.

(2) As a lead compound by precipitation of the wood
liquor with lead acetate.

(3) The lead compounds were decomposed by treatment
with sulphuric acid, and the solution treated with alcohol. A
portion of the gum was precipitated as a flocculent mass, and
a second fraction was obtained on evaporating the filtrate.

(4) A brominated derivative was obtained by treating the
original wood liquor with bromine.

The following empirical formulae represent the results of
ultimate analyses of these products, together with methoxyl
determinations.

Compound Celluloses 2OI

(1) From analyses of HC1 precipitate :

C24H24(CH3)2SO12.

(2) Calculated from analyses of PbO compounds t

C24H24(CH3)2S012.

(3) From direct analyses of gums obtained from PbO pre-
cipitates :

(a) Precipitated by alcohol : C24H24(CH3)2SO,2.
(d) Soluble in aqueous alcohol : C24H24(CH3)2SO|S.

(4) Brominated derivative : -. ;

C24H22(CH3)2Br4.SOn.

The S is present in this complex as a sulphonic residue
(SO3H) ; the parent molecule, i.e. the non-cellulose or lignone
complex of pine wood, may be regarded as approximately
C24H24(CH3)2O10. Certain definite conclusions may be drawn
from this empirical study of its derivatives.

(1) It is evident that it represents a highly condensed
molecule. Taking the 'carbohydrate' formula to which it
most nearly approximates, viz. C24H24O12, this represents

4C6H1206-i2H2O.

In addition to condensation expressed by dehydration,
CH— CH groupings are also represented, as appears from
the bromination of the product. The authors not having
prepared the corresponding chlorinated derivative, we are not
able to compare the grouping of the hexene rings of this com-
plex with those of jute, especially as the reactions of the
chlorinated wood are distinct from those of jute. There is,
however, an unmistakably close general similarity.

(2) It is evident that the condensation is of a type which
resists hydrolytic treatment of a very energetic character ; and
that the constituent groups of the lignocellulose are united
together by stronger bonds of synthesis than CMinking.

2O2 Cellulose

(3) With regard tp the mechanism of the reaction in the
original resolution of the lignocellulose, it is of a complex
character ; and the synthetic equilibrium of the products in
the resulting solution is, no doubt, different from that repre-
sented by the parent complex. We may very well assume
that the reaction involves the following factors : (a) the hydro-
lytic action of the sulphurous acid ; (b) the formation of
aldehyde bisulphite compounds ; (c) the probable sulphona-
tion of side chains of the general form #.CH : CH.COH,
as in the well-known reaction of cinnamic aldehyde with
sodium bisulphite ; (ct) the saturation of acid OH groups
byCaO.

These researches are, it will be seen, an important prelimi-
nary elucidation of the problem of the composition of this
interesting industrial product, and afford general conclusions
as to the constitution of the non-cellulose constituents of the
lignocelluloses, which entirely confirm the deductions given in
preceding sections of this treatise.

In regard to the pulp or .insoluble product of the original
reaction, which is, as already stated, an industrial product of
the greatest importance, it represents the cellulose of the wood
together with residues of the non-cellulose in small proportion,
not removed by the treatment. The presence of the latter is
marked by the colour of the product, which is usually a greyish-
pink. A large quantity of the pulp is used in this crude or
unbleached condition; but for white papers a preliminary
treatment with bleaching powder is practised, the proportion
required being from 15-25 p.ct. of the weight of the pulp.
The process is attended by a loss of weight of from 8-12 p.ct.,
owing to conversion of the more oxidisable constituents of the
pulp into soluble derivatives.

ru It is to be noted that the yield of bleached cellulose by this
process is, as in many other cases, considerably inferior to that

Compound Celluloses 203

obtained by the process of chlorination, &c. By the latter
the authors obtain 60-65 P-ct- °£ cellulose from the coniferous
woods, whereas the * sulphite process ' yields about 50 p.ct
This is another instance of the variable character of the 'cellu-
lose constants ' of fibrous products ; the cellulose being a
product of resolution or decomposition* and varying both in
character and proportion with the conditions of the treatment
by which isolated. .-••:•.•• --• • .

With the exception of the woods of the Coniferae, none
of the woods have been submitted to exhaustive investigation
so far as regards the fundamental tissue or lignocellulose proper.
It appears, in fact, that such investigations have only been
rendered possible by the general advances of the science
during the last few years, more especially in the province of the
carbohydrates. This work upon the carbohydrates of lower
molecular weight, together with the preliminary work upon the
general features of lignification recorded in this treatise, opens
out a very wide field for future work in the direction of reducing
the phenomena of elaboration and metabolism in the plant to
exact molecular expression. In regard to such investigations
we may point out here that of those types of lignification which
have been so far studied, four maybe selected as showing well-
marked features of differentiation, viz. :

' Annual ' products ' Perennial ' products

Cereal straws ; Jute bast Dicotyledonous woods ; Coniferous woods
1 ... . I

each and all of which call for extended investigation, i.e.
individually as presenting a problem of chemical constitution,
and comparatively with the view to tx^nnect the variations in
composition with variations in the physiological factors of
their origin and growth.

There are, of course, a number of woods characterised by
the secretion or excretion of particular products, sueh as more

204 Cellulose

particularly the dye woods, logwood, brazil wood, sapan, &C.,
&c. These characteristic products are well-defined, mostly
crystallisable compounds, the constitution of which is deter-
mined entirely without reference to the physiological problem
of their origin or their relationship, genetic or otherwise, to the
tissues in which they are stored up.

The purpose of this treatise is, however, strictly limited to
the chemistry of fundamental tissue ; outside this lies the
indefinitely wide territory of plant secretions into which we
make no attempt to enter.

We have now, in concluding our account of the lignocellu-
loses, to deal briefly with certain industrial processes which
throw further light on the chemistry of the lignocelluloses.

(i) DESTRUCTIVE DISTILLATION. — The products of the
destructive distillation of the woods are extremely numerous
and of varied constitution, comprising, in fact, representatives
of all the more important groups of C,H, and C,H,O com-
pounds. The formation of these products depends upon
various factors : (a) the composition of the wood itself, and
(^) the conditions of distillation.

Ramsay and Chorley have made careful comparative
investigations of typical dicotyledonous woods — oak, beech,
and alder— and their results afford a general idea of the influ-
ence of these factors. The tables on the opposite page may
be cited in illustration (J. Soc. Chem. Ind. 1892).

These results, as regards the solid residue (charcoal) and
gaseous products and their relation to the conditions of distil-
lation, are very complete and require no further discussion.
The increase of gas at the higher temperature of distillation is
formed at the expense of the charcoal, and CO at the expense
of CO2.

In addition to these observations on the products, the
authors also found that the distillations were marked, as in the

Compound Celluloses

20$

—

Oak

Beech

Alder

Weight of wooc

P.ct. of charcoa
P.ct. of distillat
P.ct. ofCO2ab
Difference to m

Volume of gas a

P.ct.

composition of
this gas

P.ct. of pitch
the wood

taken in grms. .
1

167

180

134

24-55
58-69
9-58

7-18

26-66
59'33
9'23
478

25'37
5970
970

5'23

sorbedbyKOH .
ike up loop.ct. .

fter absorbing CO2
CO

o

7,000 c.c.
7077

I'll

14-90
13-32

7,200 c.c.

73-14

I'02

1-49
18-71
5'64

6,000 c.c.
73-47
1-52
i-59

20-11
4-31

defines • • .
CH4 . . . .
N by difference .

from distillate on

9-58
613
1-36

n-ii

6-54
6-08

I5-67
5'90
11-17

P.ct. of acetic a
P.ct. of methyl

cid

alcohol ....

Maximum temperature in each case about 500°.

—

Oak

Beech

Alder

Weight of wood taken in grms. .

181

187

ISO

33*7

3d. '22

34-66

P.ct. of distillate

c6-3i?

jq. *.*

Z 3M7

C4/OO

P.ct. of CO., absorbed by KOH .
Difference to make up 100 p.ct. .

y* 03

6-40

3-49

7-49
4-82

3*+ "^

8-00

3-34

Volume of gas after absorbing CO2
P.ct. (CO

4,000 c.c.
92-25

5,000 c.c.
87-36
i*ii

4,000 c.c.
84-61
i-6<;

2-96

4.-IC

4"32

UbV difference.

4-89

7-38

9-42

P.ct. of pitch from distillate . .

7-69

c-«a

7'49

6"O2

"'33

e-76

P.ct. of methyl alcohol ....

j y
1-32

5-31

3 /u
1075

Maximum temperature ....

344°

380°

343°

case of jute, by a strongly exothermic reaction occurring in all
cases at about 320°.

From the general literature of the subject, which is some-

206 Cellulose

what scattered, we find that the following compounds have
been identified amongst the liquid products :

Aqueous distillate Tar

Acids Alcohol Aldehydes ', Hydrocarbons^

Chiefly Acetic. Chiefly Methyl Ketones, <SrV. Phenols, &c.
Also Formic f Acetaldehyde Paraffins

Propionic \ Furfuraldehyde Toluene

Butyric / Acetone Xylene

Valeriahic, &C. J Methyl-propyl ketone Creosol

Crotonicacid I Methyl ethyl „ Guaiacol

j Methyl formate Methoxy-deri-

( Methyl acetate vatives of py-

rogallol
Methyl • pyro-
gallol

Propyl - pyro-
% gallol

A large number of these derivatives are obviously formed
by secondary reactions. What chiefly concerns our subject is
to distinguish, if possible, the primary products of the decom-
position.

A careful survey of the evidence leaves no doubt that the
main products, as to relative quantity, are the primary pro-
ducts, viz. :

Methyl alcohol Acetic acid Furfural and Pyrogallol derivatives

In regard to the two former, interesting conclusions are
drawn by Ramsay and Chorley from their results (loc. #'/.).
It will be noted that these products are constant for the indi-
vidual woods, over a wide range of temperatures of distillation,
and it is probable that they are formed as it were explosively,
i.e. in the exothermic reaction above described.

In regard to the relationship of furfural and acetic acid to
each other, the following results of observations upon a par-
ticular distillation of alder wood may be cited.

Compound Celluloses

207

No.

Temperature

Time

Quantity of
distillate

Furfural

Acetic acid

Hours

Less than

Grin.

Grnu

X

200°

2

I C.C.

0-05

o-o

C.C.

2

200-2300

2

4-5

O'O9

1-28 a

3

230-250°

;., i»..

6-0

0-096

1-95

Min.

4

250-2700

25

4'S

O'05

0-88

5

270-290°

25

4'5

0*064

0-83

6

290-3I00

15

3 '5

0-086

0-80

7

310-320°

10

5'0

0-142

0-79

8

320-330°

10

5'0

0-170

0-78

9

330-340°

15

4'5

0-237

0-87

Totals . . .

0-985

8-18

Both are formed continuously, and increase towards the end
of the distillation. Special observations were also made on
the yield of acetic acid from oak (183 grms. of wood), with the
following results :

No.

Yield

Containing acetic
acid

C.c.

Grms.

I

Up to 120°

20

—

2

120-180°

10

OT756

3

180-240°

IO

0-7320

4

240-260°

10

I -0248

5

260-300°

10

I -4640

6

300-310°

IO

I '5372

7

310-322°

10

i -0980

8

322-350°

15

2-3424

9

350-450°

About 10

1-6104

On comparing the woods with one another the only very
noteworthy feature of difference is the very large yield of
methyl spirit obtained from alder wood. The acetic acid re-
sults are approximately constant for this series of woods.

In regard to the aromatic products it is important to note
the predominance of pyrogallol derivatives, and that many of
these are characterised by the OCH3 group.

208 Cellulose

This strongly confirms the view we have taken of the con •
stitution of the hexene constituents of the lignocelluloses. The
completion of the benzene ring under the conditions of the
distillation indicates its occurrence in the ordinary life-history
of the wood. Unfortunately the mechanism of the condensa-
tion is too complex to follow, and so, in fact, of the entire
process. What is required is an extended investigation of this
destructive decomposition under the conditions of variations
determined by the addition of reagents, added to promote
reaction in one or other direction. All that we can deduce
from the results of investigations as they stand is a general con-
firmation of previous discussions of the constitutional relation-
ships of the constituent groups of the lignocelluloses.

(2) PROCESSES OF DISINTEGRATION BY REAGENTS. — A.
Proximate resolutions. — The various processes of preparing
a papermaker's pulp from the woods admit of a simple
theoretical classification on the basis of the foregoing treat-
ment of the subject. The table on p. 209, from a paper o£
the authors (Forestry Exhibition Reports, Edinburgh, 1886,
No. 20), gives such a comparative survey, together with the
names of the inventors more prominently associated with the
origination of the several methods.

The general principles of the classification are briefly these :
The lignocelluloses are readily attacked by hydrolysing agents,
even water. The attack of these agents is accompanied by the
inverse processes of condensation, which may be and are many-
sided, owing to the presence of OH groups of the most varied
function. A limit is therefore reached when the product is
sufficiently condensed to resist further attack. There are two
general ways of extending the limit : (i) strengthening the
hydrolysing action, either by concentration of the reagent, or
by increase of temperature \ (2) preventing the reverse action
by fixing reactive groups in combination. These more active

Compound Celluloses

20Q

M

H

K
<

4

j 2

a

d
o

S3*

ea«S

i

2

!-<

ZW

§
1

g^ M-

li

o <: «

ueao

BACHET S

f TILGHMA
I PICTBT, j

EH

'~ -d

•5J

<

y5

¥ '3

A

F DISINTEGR

AQUEOUS ACID

Oxidising and hydn
Nitric acid.
Nitro-hydrochloric ;

ii

W
*^

tt
&

Reducing' and hydrd
Sulphurous acid.

<j

S

3

S

W

H

a

fa"

tf, VM

BISULPHITES.

Hydrolysis aided primarily
by sulphurous acid, and
secondarily by combination
of products with bisulphites,
also by prevention of oxida*
tion.

H'.HB

2io Cellulose

groups are chiefly two, viz. aldehydic and acid ; and hence the
employment of sulphurous acid and bisulphites on the one
hand, and alkalis on the other, for the purposes as indicated in
the table.

The application of these principles is well illustrated by
taking any of the processes in series of variations. Thus the
sulphite processes :

(a) Sulphurous acid alone is capable of resolving wood into
cellulose (insoluble) and non-cellulose (hydrolysed), and soluble
derivatives. The acid, however, owing to its feeble hydrolysing
power, requires to be used in 7 p.ct. solution (SO2) prepared
under pressure ; and, again, to prevent reverse action, the limit
of temperature employed is 105°. (See 'The Pictet-Brelaz
Process of Preparing Wood Cellulose,' Cross and Bevan.
Spon : London, 1889.)

The hydrolysing action of the SO2.Aq, ordinarily very
feeble, is perhaps also more powerful in relation to aldehydic
condensations.

(b) The bisulphites. — The addition of the base lowers the
hydrolysing action of the acid ; a higher temperature (150-160°)
is therefore required. The base, however, serves to saturate
acid groups, and the process is further aided by sulphonation
in the CH=CH groups ; the presence of the excess of bi-
sulphite prevents reversal by condensation of aldehydic groups.

(c) Neutral sodium sulphite. — In this case the hydrolysing
action is still feebler, and a higher temperature is required
(160-180°).

In presence of the lignocellulose complex, undergoing
decomposition, the sulphite is dissociated, the base going to
acid groups, and the acid sulphite residue to aldehydic groups.

In all the above processes, moreover, the resolution is aided
by deoxidation of the lignocellulose constituents, a certain
proportion of sulphate being formed.

Compound Celluloses 21 1

The alkaline processes involve reactions of a totally
different character. In the sulphite processes the resolution
of cellulose from lignone is a comparatively simple process ;
the latter is obtained as a soluble derivative but little changed
in essential chemical features from the condition in which it
existed in the wood (see p. 178). The alkalis determine, on the
other hand, a highly complex decomposition ; the products
are extremely numerous, and for the most part ill-defined. An
analytical study of these products will be found in the Papier
Zeitung, 1878, 226, 242. A prominent feature of the de-
composition is the liberation or formation of acid groups, and
the consequent ' saturation ' of the alkali. The hydrolysing
power of the alkaline solution is continually diminished, and
the alkali has therefore to be used in excess ; and according to
the amount taken in excess, so, inversely, is the temperature
necessary for completing the decomposition. The additional
presence of reducing agents, such as sulphides, appears to have
a certain influence upon the result. But since the organic
products themselves, in presence of the alkali, are of a power-
fully deoxidising character, any influence would probably be
traceable to specific reactions between the sulphur and the
constituents of the wood.

The acid processes we consider with exclusion of the
sulphite processes. They divide themselves into the two
groups : (a) resolution by non-oxidising acids (dilute H2SO4
and HC1) ; (b) by oxidising acids (dilute HNO3).

In the former we have a process of some theoretical interest,
consisting of boiling with HClAq, neutralising, and bringing
the solution into alcoholic fermentation. It is stated in the
historical notices of the industrial development of the process
that it was worked for some time on a commercial scale (Payen,
Wagn. Jahresber. 1867). In such treatments the limit of
resolution is rapidly attained, and the residue is in the brittle

P2

212 Cellulose

and friable condition of the more ' condensed ' products. The
limit is obviously determined by the inability of such acids to
enter into any permanent synthetical combination with the con-
stituent groups of the wood-substance, which remain therefore
open to mutual interaction ; hence their further condensation,
and the building up of more resistant forms of lignocelluloses.
The reaction with nitric acid, on the other hand, is of a totally
different order : the acid hydrolyses and oxidises in the first
instance, but the specific and characteristic decomposition which
ensues (described in detail, p. 146) is the result of direct synthesis
of the lower oxides of nitrogen with the lignone groups. The
final products of resolution of the non-cellulose, or lignone,
are the simplest acids — carbonic, acetic, and oxalic.

The extreme oxidising action of nitric acid has been investi-
gated by Wheeler and Tollens (Annalen, 267, 367) in the case of
pine-wood, the wood being heated for 6 hours at 90-100°, with 10
times its weight of nitric acid (1-4 sp.gr.) diluted with \ water. The
lignone constituents are, of course, entirely broken down in the
reaction, the greater proportion of the cellulose also. The residue
was an oxycellulose, amounting to 17 p.ct. of the weight of the wood,
having the composition :

Calc. C36H60O,,
=6C.H100.+0

C 43-41 • • • • 4372
H 6-19 .... 6-07
O 50-40 .... 50-20

Its properties - were those of the oxycelluloses obtained by the
action of nitric acid upon cotton.

The peculiar feature of this oxidation is the survival of a
residue so little different in empirical composition from the wood
cellulose itself.

The results of these various treatments confirm entirely the
views advanced as to the constitution of the lignocelluloses,
and, conversely, a grasp of these views enables us to predict,
with a satisfactory approximation, the results of treatments not
specifically investigated.

Compound Celluloses 213

B. Ultimate resolutions. — (i) The extreme action of the
alkaline hydrates at 200-300° is an industrial process of some
importance for the preparation of oxalic acid from waste wood.
An exhaustive investigation of the process, more especially of
the conditions determining maximum yields of the acid, has
been made by W. Thorn (Dingl. J. 210, 24). The optimum
temperature for the decomposition is 240°, but at this point a
relatively large proportion of the alkali is required for complete
decomposition, viz. 4 parts hydroxide to i part wood. Potas-
sium gives higher yields than sodium hydroxide : the maxima
obtained under the condition of heating in closed vessels
were, 52 p.ct. (NaOH) + 66 p.ct. (KOH) ; heated in thin
layers, the yield in the latter case was increased to 80 p.ct.

The maximum yields from various woods observed by the
author, and calculated on the dry woods, were :

Pine, 947 ; Poplar, 93-14 ; Oak, 83-4 ; Box, 86-4.

By a more graduated but still severe treatment with the
alkaline hydrates, G. Lange obtains the following characteristic
products of resolution :

(1) Cellulose (insoluble), and in the alkaline liquid.

(2) Two complex acids, or groups of acids (lignic acids),
described below.

(3) Formic and acetic acids, and traces of the higher fatty
acids.

(4) Protocatechuic acid and catechol ; (5) ammonia and nitro-
genous bases in small quantity.

The lignic acids are of definite composition so far as is
established by uniform results of elementary analysis, which are as
follows :

Wood from which
obtained

Lignic acid soluble in
alcohol

Lignic acid insoluble in
alcohol

Beech . . .
Ash , . .
Pine

C H
6i-5 5'5
61-6 5'5
61-3 5-0

59-o 5-4
58-8 5-2
60-5 5-2

214 Cellulose

Nothing, however, has as yet been established in regard to the
constitution of these products.

They are no doubt in the main derived from the non-
cellulose or lignone complex, but the author's numbers do not
warrant his view that the attack of the alkali is confined to this
complex ; it is evident that the cellulose is considerably attacked
also, and the method cannot be recommended as a process for
cellulose estimation. For the original papers see Zeitschr. PhysioL
Chem. 14, pp. 15, 217.

(2) Chromic add in presence, of sulphuric actd (cone.) de-
termines complete conversion of the carbon of the woods
into the gaseous oxides CO2 and CO ; the proportion of
the latter is small. The reaction is available, therefore, as
a combustion method, under the conditions previously de-
scribed.

Pectocelluloses and Mucocelluloses. — The second
great division of the compound celluloses are those of which
the non-cellulose constituents are related to the * pectic ' group
of compounds. Hugo Muller (Pflanzenfaser), in stating the
results of proximate analyses of raw fibrous materials, completes
the list of constituents with an undetermined aggregate de-
scribed as ' Incrusting and intercellular substance and pectic
constituents, calculated from loss ' — i.e. having determined ash,
water, water extract, fat and wax, and cellulose, the residue is
estimated by calculation, and stated under the above aggregate
description. The question of ' incrustation ' is rather morpho-
logical than chemical. In the lignocelluloses we have ample
evidence that the process of lignification is an intrinsic trans-
formation of tissue-substance. It is, of course, consistent with
this view that there should be a concurrent process of deposi-
tion of substance external to the cells themselves, destined for

Pectocelluloses and Mucocelluloses 215

the binding or cementing of the cells together into a compact
tissue. But it is important not to confuse effects, even though
they may proceed from the same cause. In the lignocelluloses,
the view of incrustation of the cells is too often transferred to
the chemistry of the cell-substance, which comes to be regarded
as composed of cellulose merely overlaid with non-cellulose
constituents which mask its reactions. These views, we main-
tain, should be kept distinct.

The morphology of cell formation teaches that with growth
a differentiation of a portion of the cell wall takes place, which
in the fully developed condition of the tissue constitutes the
division of cell from cell and completes their individualisation.
A true * intercellular ' region is then formed, and chemical
differentiation no doubt often takes a different course in this
region ; but it has not yet been established that differentiation
of the cell-substance does not take place simultaneously in the
same direction though in lesser degree. In the lignocellu-
lose this certainly appears to be the case ; in the pectocelluloses,
which we are about to consider, the problem is more difficult to
investigate by chemical and microchemical observation, owing
to the absence of any well-marked reactions of the pectio
compounds.

We must now give a brief outline of the chemistry of this
group. Their characteristic property is that of yielding gela-
tinous hydrates, in which they closely resemble the mucilage-
yielding constituents of many seeds, fruits, and rhizomes — e.g.
linseed, the seed of Plantago Psyllium^ the roots of Orchis
Morio, &c., many of the Salvia species, the fruit of the
quince (Cydonia vulgaris\ &c. While, however, the empirical
composition of the latter is that of the carbohydrates, viz.
CnH2mOm, the pectic group are distinguished by empirical
formulae with considerably less hydrogen in proportion to the
carbon and oxygen, the general approximate formula being

2i6 Cellulose

CmH3mOm. A second feature of distinction is that, while the
^

former yield on hydrolysis hexoses and pentoses, the latter give
the series of pectic acids, this distinction corresponding with the
higher proportion of oxygen which characterises the group. It
is necessary to qualify these conclusions somewhat by pointing
out that the empirical formulae determined by the original in-
vestigators of these compounds — viz. Fremy (Ann. Chim. Phys.
[3] 24> 5)> Chodnew (Ann. Chem. Pharm. 51, 355)— have been
called in question by later observers. Thus Reichardt (Archiv
d. Pharm. [3] 10, 116) concludes that they are to be regarded
as gelatinisable carbohydrates (see Tollens, Kohlenhydrate,

P- 243)-

On the other hand it will be found that the pectocelluloses
differ from the celluloses by increased proportion of oxygen ;
and their acid character is further shown by their retaining
a relatively large proportion of basic mineral constituents
(ash).

The general relationships of the group as determined by
the earlier observers are these : Pectose, the insoluble mother
substance of the group, occurs in mixture or union with the
cellulose of the parenchyma of fleshy fruits and roots, e.g.
apples, pears, turnips, &c. This is hydrolysed by boiling
dilute acids or alkalis, or by a ferment enzyme (pectase)
secreted in the tissue, to pectin (C32H48O32, Fremy), the
solutions of which readily gelatinise. By continued hydro-
lysis (boiling water) this is further modified to parapectin,
and by alkalis to metapectin and parapectic acid and pectic
acid (C32H44O3o, Fremy ; C12H16On, Regnault ; C12H16O10,
Mulder; C14H22O14, Chodnew).

The final product of hydrolysis is metapectic acid. To
this acid Fremy assigned the formula C8H14O9. Later investi-
gations have established its general identity with arabic acid —

Pectocelluloses and Mucocelluloses 217

a complex acid which is the main constituent of gum-arabic.
Gum-arabic yields, on graduated hydrolysis, a complex of
glucoses (galactose, arabinose) and a series of arabinosic
acids, e.g. C23H38O22, and compounds differing from this by
+ C6H10O5. It appears, therefore, generally, that the pectic
group are compounds of carbohydrates of varied constitution
with acid groups of undetermined constitution, associated to-
gether to form molecular complexes, more or less homogeneous,
but entirely resolved by the continued action of simple hydro-
lytic agencies ; and ti\z pectocelluloses are substances of similar
character in which the carbohydrates are in part replaced by
non-hydroly sable celluloses. The general characteristics of the
pectocelluloses are therefore these : they are resolved by boiling
with dilute alkaline solutions into cellulose (insoluble) and
soluble derivatives of the non-cellulose (pectin, pectic acid,
metapectic acid) ; they are gelatinised under the alkaline
treatment ; they are ' saturated compounds,' not reacting with
the halogens, nor containing any groups immediately allied to
the aromatic series.

Compound celluloses of this kind are enormously diver-
sified in composition, structural character, and distribution,
and the group, having none of the sharp lines of differentiation
and demarcation presented by the lignocelluloses, cannot be
handled at all in the same way.

We must confine ourselves, therefore, to the one or two
more definite types which have been investigated.

Flax. — Commercial flax is a mixed product. The bast
fibre proper constitutes from 20-25 p.ct. of the entire stem,
and is more or less imperfectly separated from the wood on the
one side, and the cortical tissue-elements on the other, by the
ordinary processes of retting and scutching. These residues are
visible with the naked eye, but are brought into clearer evidence
by means of reagents, followed by microscopic examination.

218 Cellulose

Thus the wood is an ordinary lignocellulose, and gives the
characteristic reactions ; the cortical tissue is again distinguished
from the fibre proper by reacting strongly with magenta- sul-
phurous acid. The presence of the cortical tissue is also
marked by the large proportion of • oil and wax ' constituents
present in the fibre (3-4 p.ct). Excluding these adventitious
constituents the fibre proper is a pectocellulose. That the non-
cellulose constituents of flax are pectic compounds was first
established by Kolb (Bull. Soc. Ind. Mulhouse, June 1868).
According to his observations, the precipitate obtained on
acidifying the alkaline solutions from the * boiling' of flax
goods consists of pectic acid.

The proportion of these constituents varies from 14-33
p.ct. in the different kinds of flax, the variations being in part
due to the plant, i.e. to physiological habit and conditions of
growth ; in part to the different methods of retting the plant —
and extracting the fibre. After well boiling with the dilute
alkali (1-2 p.ct. NaOH) the fibre-substance consists of flax
cellulose, with residues of the wood (sprit), cuticular tissues,
and oils and waxes associated with the latter. By exposure
to chlorine (after well washing and squeezing) the wood is
attacked in the usual way, and is then easily resolved by
alkaline treatment. To purify the cellulose it requires to be
boiled out with alcohol, and finally treated with ether-alcohol
to remove the oil-wax residues. In this way flax cellulose is
isolated in the laboratory in an approximately pure condition.
It might appear from the outlines of this laboratory method
that the bleaching of flax goods, which consists substantially
in the isolation of the pure flax cellulose, is a comparatively
simple process. This is not so, however. The exigencies of
economical and safe treatment of textile fabrics prescribe
certain narrow limits of chemical treatment ; and the removal
of the more resistant wood (lignocallulose) and cuticle

Pectocelluloses and Mucocelluloses 219

(adipocellulose) under these conditions involves a reiterated
round of treatments consisting of —

Alkaline hydrolysis . . Boiling in solutions of NaOH, NaaCO3,&c.

Oxidation \ Hypochlorite solutions and atmospheric

* * oxidation (grassing).

Souring .... Treatment with dilute acids in the cold.

It must be remembered, however, that the problem is not the
removal of the non-cellulose constituents of the fibre itself — these
disappear almost entirely in the earliest alkaline treatments —
but of compound celluloses of the other two main groups.

The further investigation of the pectose of flax fibre has not
been prosecuted according to the methods of later years. Such
investigations will, no doubt, be undertaken in due course.

FLAX CELLULOSE has been mentioned incidentally to the
general treatment of the celluloses. So far no reactions have
been brought to light in which it is differentiated from cotton
cellulose, with perhaps one exception, viz. its lesser resistance to
hydrolysis. Thus H. Miiller mentions (Pflanzenfaser, p. 38)
that flax cellulose isolated by the bromine method lost, on
boiling five times with a dilute solution of sodium carbonate
(i p.ct. Na2CO3), 10 p.ct. of its weight. The statements of
R. Godeffroy (abstracted in J. Soc. Chem. Ind. 1889, 575),
that flax cellulose is distinguished from cotton cellulose by its
reducing action upon silver nitrate in boiling neutral solution,
are erroneous, the reaction resulting from residual impurities,
which, for the reasons given, are extremely difficult to isolate.
Flax cellulose may therefore, for the present, be regarded as
chemically indistinguishable from cotton cellulose.

The oil and wax constituents of the raw fibre will be
described under the group of adipocelluloses.

OTHER PECTOCELLULOSES. — As far as investigation has pro-
ceeded, it appears that pectose, or pectose-like substances, are
associated with all fibrous tissues of the unlignified order.

22O Cellulose

And indeed in the lignocelluloses themselves pectous sub-
stances make their appearance with increasing age. Thus the
lower portions of the isolated jute bast — jute cuttings or butts
— when boiled in alkaline solution yield products which cause
the solution to gelatinise on cooling ; and the gelatinous product
is insoluble in alcohol, distinguishing it, as pectic acid, from the
products of hydrolysis of the lignocellulose itself, which are dis-
solved, after precipitation, by alcohol. It must be remembered,
however, that in the 'jute cuttings ' the adhesion of the bark
and cortical parenchyma to the true bast fibre is such that we
are dealing with a complex tissue, and the source of the pectic
acid may be in the parenchyma of the tissue and not in the
bast fibre. On the other hand, we have shown (p. 152)
that in the spontaneous decomposition of jute, lying in the
damp state, gelatinous acid bodies are formed, indistinguish-
able from pectic acid. It would not be difficult, therefore, to
account for the pectic constituents of the bast tissue towards
the root end, as products of degradation of the lignocellulose
itself.

Reverting, however, to the non-lignified fibres such as China
grass, or Ramie (Bohmeria species), and the 'nettle fibres' gene-
rally, hemp, and even raw cotton — these all contain pectic bodies
associated with the cellulose, which are hydrolysed and dis-
solved by treatment with boiling alkalis. But these pecto-
celluloses have not been sufficiently investigated as compound
celluloses to admit of any useful classification on the basis of
particular constitutional variations of their non-cellulose con-
stituents.

The monocotyledonous fibre-aggregates, whether fibro-
vascular bundles (Phormium, Aloe fibres, Musa, &c.) or entire
plants (Esparto, Bamboo stems, Sugar Cane), are largely made
up of pectocelluloses, with a greater or less proportion of ligno-
celluloses. But the constitution of these non- cellulose con-

Pectocelluloses and Mucocelluloses 221

stituents is as yet quite unknown, and we have therefore none
but the general basis of classification.

In the same way also the parenchymatous tissue of fruits,
fleshy roots, &c. — the typical pectocelluloses— must be, for
the present, dismissed with the bare mention.

The investigation of these substances belongs rather to the
province of general carbohydrate chemistry than to the
narrower cellulose group ; and the problems involved are in
many respects rather morphological and physiological than
purely chemical.

These same considerations apply also in great measure to
the mucilaginous constituents of plant tissues, though certain
of these have been investigated by modern chemical methods.
The relationship of these substances to cellulose is indicated

(a) by the histology of the tissues, which shows them to be
associated with the cell wall, rather than with the cell contents ;

(b) by their empirical composition, which is approximately that
of cellulose ; (c) by their reactions with iodine, by which they
are coloured variously from blue to violet, as are the hydrated
modifications of cellulose (Sachsse, Farbstoffe, &c. p. 161).
Beyond superficial observations of reactions (iodine) and
gelatinisation with water, these compound celluloses — which
may conveniently be termed mucocelluloses — had been but
little investigated (Sachsse, lac. cit.) until the systematic work
of Kirchner and Tollens, and Gaus and Tollens (Annalen,
J75> 205 > 249> 245)> upon the mucilages and gums. Of these
typical researches we give a brief account.

(i) QUINCE MUCILAGE was prepared by digesting 50 grms.
with i litre of warm water, pouring off, and repeating the
digestion ; filtering the mucilage by squeezing through cotton
cloth, and precipitating the dissolved product by the addition
of hydrochloric acid and alcohol. After washing with alcohol
and ether the product was dried, forming brittle fibrous masses ;

222 Cellulose

the yield amounted to 8-10 p.ct Ultimate analysis of the
products, retaining 5-6 p.ct. inorganic constituents, gave the
following numbers, varying between the extremes

C 46-52 44-17

H 5-88 6-15

Corresponding to the formula Cj8H28O14 C,8H30O,5

The product was then investigated for the presence of typical
carbohydrate groups.

Oxidation with nitric acid gave no mucic acid and no
saccharic acid. Galactose and dextrose groups are therefore
absent. On the other hand, furfural was obtained in some
quantity (6-45 p.ct. furfuramide) on distillation with acids.
The substance therefore contains pentose groups.

Hydrolysis with dilute sulphuric acid. — On boiling with
the acid, the product is resolved into —

And a mixture of

Cellulose Gummy bodies and Glucoses

Insoluble and Precipitated by alcohol Soluble in

amounting to from neutralised alcohol

23 p.ct. solution

From the soluble products it was found impossible to isolate
any glucose in the crystalline form. The solution, on the
other hand, certainly contained compounds of this group, since
it was strongly dextro-rotary, reduced Fehling's solution to
an amount equal to 62 p.ct. that of dextrose, and gave, with
phenylhydrazine, anosazone melting at 162°, and giving results
on analysis corresponding with a mixture of osazones of a
pentose and hexose.

It is evident from these results that the mucilage is com-
paratively resistant to hydrolysis ; by its behaviour, in fact, it is
shown to be much more nearly related to the cellulosic than
the starch type of ' saccharo-colloids.' It is for this reason
that we direct special attention to this remarkable group of
compounds, since their further investigation cannot fail to

Pectocelluloses and Mucocelluloses

223

throw light upon the problems discussed in the earlier sections
of this treatise.

(2) SALEP MUCILAGE was prepared from the tubers pre-
viously pulped by grinding in a mortar, the details of pre-
paration being exactly as for the above. The mucilage was
precipitated by alcohol in white threads which hardened under
further treatment with alcohol (dehydration).

The dry substance (retaining i -5 p.ct. mineral constituents)
gave on analysis numbers approximately those of cellulose
(C6H1005), viz. :

C . 44*58

H . 6-63

The hydrolysis of the product, by boiling with dilute acid
(1*25 p.ct. H2SO4), was investigated in relation to the influence
of the time factor upon the three products, cellulose, gum,
and glucoses : the former being estimated by direct weighing ;
the latter, in terms of dextrose, by titration with Fehling's solu-
tion ; the result, subtracted from the total dissolved products,
giving the yield of gum.

Duration of hydrolysis

Cellulose

Gum

Glucose

\ hour

16-84

—

11-46

I 5,

I4'93

49 '33

41-93

2 hours

44-92

53-29

3 »

11-58

29-91

71-27

4 »

12-76

1870

81-37

5 »

12-41

16-02

75-97

7 »

9-04

6-87

74-76

It will be noted that the sum of the percentages in some
cases exceeds TOO, and in some is in defect.

These observations are explained by the attendant phe-
nomena of hydration and dehydration ; and the disappearance
of * glucose ' after the fourth hour, when it reaches a maximum,
is evidently due to condensation of aldehydic groups.

224 Cellulose

The further investigation of the product established in this
case the absence of galactose 2xi&pentose groups, but the presence
of dextrose groups in small proportion. The product of
hydrolysis yielded a mixture of osazones in which the deriva-
tives of dextrose and mannose appear to be represented.

But again the constitution of the products of hydrolysis
is left in a state of incomplete elucidation. That the authors'
methods failed to solve these problems further than has been
shown is a further illustration of the complexity of the sub-
ject. It is, in fact, the expression of the difficulty invariably
experienced with products of the cellulose class, viz. resist-
ance to hydrolysis ; and it is from the internal evidence of the
difficulties experienced by such practised investigators that
we are the more inclined to regard these products, although
soluble in water, as cellulose derivatives.

In regard to other mucilages, we may briefly mention the
more important, in order to give some idea of the distribution
of these compounds in the plant world.

AMYLOID is the name applied to a mucilaginous product
obtained from the seeds of a number of the Leguminosse,
e.g. Tamarindus indica, Hymencza Courbaril, and Schotia lati-
folia (Schleiden, Beitrage z. Botanik, i. 168). It is soluble
in boiling water, partly also in cold. It is precipitated by an
alcoholic solution of iodine as a blue flocculent mas3.

A similar substance was obtained by Frank (Pringsheim,
Jahrb. f. Wiss. Bot. 5, 15) from the membranes of the coty-
ledon cells of Troptzolum majus. This product was also
definitely proved to be formed at the expense of starch.

LICHENIN is the soluble constituent of the membranes of
Cetraria islandica (' Iceland moss ') and other similar lichens
(Knop and Schnedermann, Annalen, 55, 164). It is extracted
by treating the plant product with cold dilute hydrochloric acid
and adding alcohol to the solution ; or by boiling with water,

Pectocelluloses and Mucocelluloses 225

after previously purifying the raw material by digesting with
dilute alkaline solutions in the cold. According to Honig and
Schubert (Wien. Akad. Ber. 96, [2] 685), lichenin is accom-
panied in the plant by an amorphous form of starch.

On hydrolysis lichenin yields crystallisable dextrose, and
on oxidation with nitric acid, saccharic acid. With glacial acetic
acid it yields an amorphous triacetate, C6H7O2(C2H3O2)3.

CARRAGHEEN MUCILAGE is obtained from the seaweed
Fucus crispus (C. Schmidt, Annalen, 51, 56) on boiling with
water. This raw material is characterised by the presence of
galactose groups, yielding 22 p.ct. mucic acid on oxidation with
nitric acid, and also crystallisable galactose on hydrolysis with
boiling dilute acids.

This concludes our brief notice of the pectocellulose group.
It appears that there are two well-marked sub-groups of these
products : (i) the pectocelluloses proper, occurring in structures
of a more permanent character — fibrous and parenchymatous
tissues of the stems and roots of Phanerogams ; (2) pecto-
celluloses occurring chiefly in seeds and fruits of Phanerogams
and the tissues of Algae ; distinguished from the former by yield-
ing to the action of water, giving the peculiar solutions known as
mucilages. Hence the proposed name mucocelluloses for the
parent tissue-substance having these properties.

These groups are further distinguished by the characteristics
of their hydrolysable constituents, the former yielding com-
plexes in which acid features predominate ; the latter yield
neutral solutions, and in fact, on ultimate hydrolysis, various
hexoses and pentoses.

Adipocelluloses and Cutocelluloses. — Cork and
Cuticularised Tissues.— The plant represents, in the one
view, an assemblage of synthetical operations carried on
within a space enclosed and protected from the destructive

Q

226 Cellulose

influences of water and unlimited atmospheric oxygen. The
protecting external tissues are those which we are about to
describe as constituting the third important group of com-
pound celluloses. These tissues contain, in admixture with
the tissue-substance, a variety of oily and waxy products
(easily removed by mechanical solvents), the presence of which
adds very considerably to the water-resisting property of the
tissue. It will be seen as we proceed, however, that the tissue-
substance, after being entirely freed from these adventitious
constituents or oily excreta, yields a large additional quantity
of such products when decomposed by 'artificial' processes
of oxidation and saponification. By this and by its empirical
composition (infra) the tissue-substance will be seen to contain
* residues ' of high carbon percentage and molecular weight, and
closely allied in chemical structure to the oil and wax compounds
found in the ' free ' state in the tissue as it occurs in the plant.
These groups are associated in combination in the tissue with
cellulose residues, and hence the description of such complexes
as adipocelluloses.

CORK in its ordinary form is a complex mixture containing
not only oils and waxes, but tannins, lignocelluloses, and nitro-
genous residues. The following are the results of elementary
analysis : (a) of cork purified by exhaustive treatment with
ether, alcohol, and water ; (£) of cork (Quercus suber) without
purification ; (c) of the cork tissue of the cuticle of the potato
(tuber) purified by exhaustion with alcohol.

(«) 0) Cc)

C . . . .67-8 657 62-3

H .... 87 8-3 7-1

O . . . .21-2 24-5 27-6

N •»;''i- <f* ''*•'??£ 2-3 1-5 3-0

The analyses are calculated on the ash-free substance
(Dopping, Annalen, 45, 286 ; Mitscherlich, Annalen, 75, 305).

Adipocelluloses and Cutocelluloses 227

These investigators succeeded in isolating cellulose from
cork, but by complicated and drastic methods of treatment,
such as would break down the greater proportion of the cellu-
lose into soluble derivatives. These treatments were : (i) drastic
oxidation with nitric acid ; (2) alternate treatments with boiling
dilute hydrochloric acid and 10 p.ct. solution of potassium hy-
drate. The proportion thus isolated amounted to 2-3 p.ct. only.

The authors, on the other hand, have observed that the non-
cellulose of cork is entirely converted into soluble derivatives
by the process of digestion at high temperatures with solutions
of the alkaline sulphites as described, p. 150. In this way a
residue is obtained preserving the form, i.e. cellular structure, of
the original cork, and amounting to 9-12 p.ct.

The details of a particular experiment were as follows : io'995
gnus, cork, 20 grms. Na.,SO3.7H2O, 2 grms. Na8CO3, dissolved in
500 c.c. water. Digested 3 hours at 75 Ibs., and 4 hours at 125 Ibs.
pressure. Residue bleached with sodium hypochlorite solution.
Yield of cellulose, 1*34 grms. ; 12*1 p.ct.

In the proximate analysis of cork M. Siewert found 10 p.ct.
of constituents soluble in alcohol, which were further resolved
into —

Wax, in crystalline form .•••••! *75
Fat acid, non-crystallisable • . • • • 2 '50
Acid (2), non-crystallisable and of fatty character • 2*25
Tannic acid, soluble in water . . . • 2 '50

„ „ soluble with difficulty . . . . I*oo

The crystallised wax is termed by Siewert phellyl alcohol,
C17H28O. It melts at 100°, and dissolves in 500 parts boiling
alcohol. The acid bodies are described as (i) decacrylic acid,
C10H18O2 (m.p. 86°), soluble in 52 parts boiling alcohol; (2)
eulysin, C24H36O3 (m.p. 150°), soluble in cold alcohol.

According to Hohnel (Wien. Akad. Ber. 76) and Kugler
(Dissert, on Suberin, Halle, 1884), the cork-substance proper
is a mixture of cellulose with lignocellulose and two charac-

Q2

228 Cellulose

teristic compounds, cerin and suberin. Cerin has the empiii
cal formula C2oH32O. Suberin is of a fatty nature, and yields
stearic acid and phellonic acid (C22H42O3) on saponification.

In regard to the nomenclature of these compounds, we may,
consistently with the plan of this treatise, adopt the terms Suberose
and Cutose for the compound adipocelluloses as they occur in the
plant, and Suberin and Cutin for the non-cellulose groups united
to the cellulose to form the entire complex.

This complex of substances has been further investigated
by Fliickiger (Arch. Pharm. 228, 690), who obtains glycerol as
a product of saponification, showing the presence of the more
ordinary glycerides of the fat acids ; he otherwise confirms, in the
main, the results of Kugler (supra). Still more recently by C. v.
Wissenburgh (Chem. Centr. 1892, ii. 516) ; but the later results
throw no further light on the more important aspects of the
problem. These are obviously the questions of the constitu-
tion of the main tissue-substance and the physiology of its
formation, as well as that of the waxy excreted products which
accompany it.

In suberose, as in cutose, it is to be noted that Fre'my over-
looks the cellulose residue. V. Wissenburgh (loc. «'/.) makes the
more positive statement that cork contains no cellulose. Both
observers appear to be in error on this point.

There is some more conclusive evidence on these points in
the earlier work of Dopping and Mitscherlich. It is found
in effect that when rasped cork — yielding to solvents 7-10
p.ct. of fatty constituents (supra) — is oxidised with nitric acid,
it yields 40 p.ct. of fatty acids, and acids identical with those
obtained from the oxidation of fats and oils under similar
conditions, chiefly suberic acid. It is evident, therefore,
that the cork tissue is largely made up of constituents standing
in very close constitutional relationship to the natural fats and
oils, though possessing very different physical properties.

Adipocelluloses and Cutocelluloses 229

These relationships have been more definitely made out by
Fre'my, in his investigations of the closely allied compound
which constitutes the epidermal or cuticular tissue of the leaves,
stems, &c., of Phanerogams. As the characteristic constituent
of cork is termed suberin, so Fremy terms this cuticular tissue-
substance cutin or cutose. To prepare this substance, the cuti-
cular tissue (' peel ') of the apple, e.g., is treated with boiling
dilute acids, followed by digestion with the cuprammonium re-
agent (p. 10) ; then again with boiling acid, and dilute alkali
(KOH) ; finally the residue is treated with alcohol and ether.
In this way a nitrogen-free residue is obtained having the
empirical composition —

C . . • . 73-66 p.ct.
H . n-37 »

O . 14-97 „

Not only by these results, but by the study of the proximate
resolutions of this substance it is shown to have the closest
relationships to the carbon compounds of the 'wax' class
(Compt. Rend. 48, 667).

In a later investigation of the products of saponification of
this substance, Fre'my worked upon a raw material similarly
prepared, but having the composition C 68'3, H 8*9, O 22-8.
This compound is termed cutose^ in substitution for cutin.
Cutose is slowly attacked by boiling alkaline solutions ; a pro-
duct is dissolved, of the same empirical composition as cutose
(comp. Lignocelluloses, p. 157), but of a fatty nature. It is pre-
cipitated on acidifying the solution ; the precipitate is soluble
in ether-alcohol, and when isolated is found to melt below
1 00°. Under more drastic treatment with alkaline solutions
the dissolved products are found to be a mixture. Precipitated
by acids and treated with boiling alcohol the mixture dissolves ;
on cooling, the solution deposits an acid (m.p. 85°) in yellow-
coloured flocks, which after fusion form a brownish translucent

230 Cellulose

friable mass. This is a compound of stearocutic acid and oleo-
cutic acid (infra), into which it is resolved by further treatment
with alkali. The alcoholic filtrate from the solid acid when
evaporated gives a viscous residue, of an acid body, okocutic
acid.

By the further action of very concentrated potash solution
in the yellow acid, stearocutic acid is formed. The potash
salt of this acid is white and translucent, insoluble in water and
cold alcohol, soluble in boiling alcohol. The free acid (m.p.
76°) is also insoluble in cold alcohol, slightly only on boiling,
but dissolves in benzene and in acetic acid on warming. The
acid also dissolves freely in alcohol in presence of oleocutic
acid. A similar result is seen with the potassium salt, which,
though insoluble in water, dissolves in an aqueous solution of
potassium oleocutate.

The composition of these two acids is as under :
Stearocutic Acid.

C 75'QO

H 1071

O 14-28

which is expressed by the formula C28H48O4. This formula 13
confirmed by the analysis of the salts of the acid.

Oleocutic Acid.

C 66-66

H . . . . . 7-91

O . . . . . 25-42

expressed by the formula C14H20O4.

Cutose is regarded by Fremy as a complex of these two com-
pounds, in the proportion of i mol. stearocutic acid : 5 mols.
oleocutic acid. These acids undergo alteration on heating
at 1 00° in presence of water, passing into insoluble modifi-
cations of higher melting point. The original molecular con-
dition, however, is restored on heating with alkaline solutions.

Adipocelluloses and Cutocelluloses 231

Fre*my also made observations upon suberin (or suberose),
which yielded similar products of saponification. He there-
fore concluded that the two products are substantially identical.
The products of oxidation by nitric acid are also indistinguish-
able, viz. chiefly suberic and succinic acids.

These results suggest, from the purely chemical standpoint,
that the cellulose of the tissue and the waxy products of excre-
tion stand to one another in a genetic relationship, and the
cutose or suberose occupies an intermediate position. The
question of a direct conversion of cellulose into wax taking
place in these cuticular tissues was definitely raised and dis-
cussed by De Bary in his investigations of this group of plant
constituents (Bot Ztg. 187 1). It appears from these researches
that wax-alcohols are certainly not contained in the cell-sap or
protoplasm, and that their origin must be in the cuticular
tissues themselves ; but the parent substance may be either
cellulose, or some compound built up with it in the ordinary
course of elaboration. This question is left for the present
undetermined. It should be borne in mind, on the other hand,
that we have a great number of direct observations upon the
physiological equivalence of the carbohydrates and the fats,
both in the animal and vegetable worlds ; and although the
mechanism of the transformation of the one into the other
group of compounds remains unelucidated, it is after all not
more difficult to imagine than the condensation to furfural. It
is, however, not the purpose of this treatise to carry discussion
into purely speculative regions ; and it is sufficient to state the
conclusion that there is ample ground for adopting as a work-
ing hypothesis that carbohydrates, or possibly cellulose, are
transformed into cutose or suberose, and these, again, into free
waxy bodies of lower molecular weight, the whole process repre-
senting the change known as cuticularisation or suberisation.

In all the researches above described there is no attempt to

232 Cellulose

travel outside the empirical region of preparing products of
decomposition answering to the more obvious criteria of de-
finite composition. Further investigation will need to attack
the more important constitutional problems presented by these
products. Some preliminary observations in this direction have
indeed been made.

Thus on p. 190 we have cited Benedikt and Bamberger's
determinations of the methyl number of cork (Quercus suber),
the percentage of CH3 as O.CH3 being 2-45, which is approxi-
mately the mean number determined for the woods. The
authors have made determinations of furfural, with results as

under :

Cork 4-5 p.ct.

Cuticle of apple (purified) . . 3-5 „

Also of the yield of acetic and oxalic acids resulting from fusion
with alkaline hydrate (3 parts by weight NaOH), the fusion
being completed at 280°, with results as under :

Acetic acid . . . . 6*0 p.ct.
Oxalic acid . . . . 2*1 ,,

These, however, are aggregate results, and cannot be specifi-
cally referred to the particular constituents of the cork, viz.
cellulose, lignocellulose, the suberin constituents proper, or the
1 free ' waxy constituents. They indicate, however, the direc-
tions of research in which results will be obtained, complement-
ary to the work of Fremy, on the more characteristic products
of decomposition of the adipocelluloses.

In another direction also results have been obtained
throwing some light on the constitution of these complexes,
viz. in investigations by Hodges (R. Irish Acad. Proc. 3, 460)
and the authors (J. Chem. Soc. 57, 196) of the cuticular
constituents of flax.

The flax was treated, in the form of yarn, with boiling
alcohol ; a bright (chlorophyll) green solution was obtained,

Adipocelluloses and Cutocelluloses

233

from which, on cooling, a flocculent magma separated. This
was filtered, and washed with alcohol, and on treatment with hot
water melted to a greenish resinous mass (product A).

The alcoholic filtrate on distillation gave a brownish-green
pasty residue (product B).

These products, examined for nitrogen and mineral matter,
gave the following results :

—

Nitrogen

Mineral matter

Phosphoric acid P2O8
p.ct. of ash

A

B .

O'OQ
0-29

17
i'i

7-0

18-4

These numbers are explained by the presence of residues of
chlorophyll.

The complex A yielded, on saponification with alcoholic
potash, a large proportion of ceryl alcohol (C27H35OH), which
was proved to exist in the original substance, in great measure
in the uncombined state ; and at the same time a mixture of
oily ketones, the solutions of which were dichroic (orange
green). These ketonic bodies were found to be soluble in
sodium acetate solution, reacting in this solution with phenyl-
hydrazine (acetate), yielding crystalline precipitates.

The residue from saponification of A, after removing these
compounds, was an inert resinous substance.

By regulated fusion with sodium hydrate it was decomposed,
with formation of a mixture of acid bodies from which cerotic
acid (C27H54O2) was isolated. The complex B contained (a)
carbohydrate derivatives soluble in water, and giving with boil-
ing hydrochloric acid a copious yield of furfural ; (d) ceryl
alcohol (about 10 p.ct.) and (c) the dichroic ketones (about
1 8 p.ct.) above described ; (d) acid bodies (50-60 p.ct.) of a
fatty character, giving soluble alkaline salts and insoluble salts
with the alkaline earth metals, insoluble also in alcohol.

234 Cellulose

In regard to the localisation of these compounds in the
flax fibre, it is evident that they are associated with the cuticular
tissue, which inference is directly confirmed by analyses of a
waste product of the spinning mill (f preparing ' process) known
as * hackler's ' dust. Hackling is a process of combing the
fibre, and the dust which accumulates in the treatment, when
examined microscopically, is seen to consist, for the most part,
of the cortical parenchyma and cuticular tissue. A specimen
of this product was found to contain 7-3 p.ct. mineral con-
stituents (ash) and 14-5 p.ct. moisture. Alcohol and ether
extracted 11*7 p.ct. of its weight (15 p.ct. calculated on the
dry, ash-free substance), the extract having the same general
characters as that obtained from the flax itself. Extracted
with petroleum ether, the proportion dissolved amounted to
8-4 p.ct. (10 p.ct. on the dry, ash-free substance). Estima-
tions of nitrogen in the original substance gave (i) 1*8 p.ct,
(2) 2-1 p.ct, of which one-sixth existed in the form of ammonia
or amido-compounds.

Lastly, in reference to this complex of cuticular by-products,
it must be remembered that the flax plant is subjected to the
retting or rot-steeping process, as a preliminary to the
mechanical treatments for separating the fibre (bast) from the
stem. The characteristics of the resulting spontaneous fermen-
tation are those of butyric fermentations. In such decomposi-
tions, of course, the wax- alcohols and ethers can take no part,
but the oily products of lower molecular weight (acids and
ketones) are no doubt in part formed as products of decom-
position of a parent substance susceptible of this species of
hydrolysis. The entire subject requires extended investigation.
Systematic research would, of course, localise the substance —
whether cellular tissue or cell contents — undergoing this parti-
cular decomposition, and the result would throw direct light
upon the origin of fatty substances in the normal living pro-

Adipocelluloses and Cutocelluloses 235

cesses. The importance of these constituents as auxiliaries to
the spinning qualities of the fibre lends the additional technical
interest to the results of such researches ; and more generally
the special functions of the adipocelluloses in protecting the
tissues beneath them from penetration by water point to the
desirability of adapting the artificial processes of waterproof-
ing (cellulose) textiles to the lines laid down by the natural
process.

These observations bring us to the close of a brief notice of
this third group of compound celluloses. The brevity of the
treatment is the expression of the very small amount of research
which has been devoted to the subject. There are many
scattered references to particular products isolated from vege-
table tissues by treatment with * fat and wax ' solvents. But a
description of these falls outside the scope of the present
treatise, which is confined to the treatment of tissue-con-
stituents, with such incidental references to adventitious or
excreted products as appear to be genetically connected with
the tissue in which they are found.

From this point of view it appears to be established, by
similarity of constitution, that the oils and waxes found in the
free state in association with the adipocelluloses are closely
related to the non-cellulose constituents of the latter ; and that
they are either degradation products of the latter, or both have
a common origin in some anterior form. It is this important
physiological problem which future researches must solve.

General View of the Cellulose Group.— The elabora-
tion of the compounds which constitute the subject-matter of
this treatise is, in point of mass-effect, by far the main work of
the vegetable kingdom. The functions of these compounds in
plant life are obviously in part structural or mechanical, in
part chemical. The cellular tissue of the plant being regarded

236 Cellulose

as the seat of the essential vital operations of elaboration and
metabolism, the fibrous and vascular systems — in addition to
taking their specialised part in the general distribution of
nutritive and other matters — are the strengthening elements,
whereas the cuticular tissues are mainly concerned in closing
off and protecting the tissues beneath from the unregulated
action of water and air. In regard to the chemical relationship
of the tissue-substances to the vital processes by which they are
formed, and to which they in turn contribute, we have little
but indirect evidence and conjecture to go upon. So far as
this evidence is of a purely chemical nature, it has been dealt
with in the preceding sections. But it depends in great
measure upon the results of investigation by physiological
methods, and for such results the special treatises must be con-
sulted. We have to deal, in conclusion, with those changes of
the tissue-substances, celluloses and compound celluloses
which accompany or follow the cessation of vital activity.

The term ' death ' is perhaps more difficult to apply to the
vegetable than to the animal organism. In a perennial plant
the active life, in the sense of the elaboration of new material,
is bound up with portions only of the structure. In an
ordinary forest tree, for instance, the leaves are these active
agents ; the trunk or stem tissues, on the other hand, are largely
depleted of the organic nitrogenous matter (protoplasm) upon
which the vitality depends ; they have ceased to live in the full
sense of the term, but, on the other hand, they are known by
casual observation to live, in the sense opposed to decay.
There are, therefore, various phases of life in the plant recog-
nised by ordinary observation, and more exactly defined by
the physiologist ; but these phases graduate by insensible stages
into the region where decay and chemical disintegration are
predominant. If, therefore, it is difficult for the physiologist
to draw the line between the life and death of a cell, it is still

Adipocelluloses and Cutocelluloses 237

more so for the chemist, who deals with the cell- substance.
Another phase or aspect of the vital process, of which also the
science of to-day gives a very slender account, is that of the
organic connection of the vast assemblage of cells which con-
stitute a vegetable organism — more particularly the subordina-
tion of each unit to the general life- history of the plant, and the
extraordinary complex of adjustments which this involves.
These, again, are hardly problems for the chemist, save, perhaps,
in regard to the organic relationship of the cell-wall to the
protoplasm by which it is formed.1 The most striking general
feature of the cellulosic group is that they are non-nitrogenous \ *
it might be reasoned, therefore, that they are, from the first,
excreta, and never live in the full sense of the word. Without
pushing these considerations into an argument of doubtful
value, we may conclude that from the moment of origin the
history of the cellulose group is one of progressive withdrawal
from the region of the vital processes proper, and that to again
enter that region they must undergo a process of proximate
resolution as a result of action from without. This reabsorption

1 The views obtaining on this point are summed up by Goodale at
p. 218 of < Physiological Botany,' under the heading ' The Relations of
the Cell Wall to Protoplasm.' They are, in the main, two : (i) The
cell wall is formed, by the solidification upon the exterior of a protoplasmic
mass, of matters previously dissolved in it, the pellicle thus formed being
regarded as an excretion or secretion. This view is held by Hofmeister
(Pflanzenzelle) and Sachs. (2) The cell wall is directly produced by
conversion of the outer film of protoplasm into cellulose, with which other
groups may be mixed or combined (Schmitz, Sitz. d. Niederrh. Ges. Bonn,
1880). This conclusion is based upon the observation that the volume of
protoplasm in a cell decrease $ pari pqssu with increase of the cell wall, and
upon the phenomena which attend thickening of the cell wall.

The distinction between these theories, however, is rather morphological
than chemical ; and the view expressed in the text may be taken as con-
sistent with either theory of the actual mode of formation of the cell wall.

2 We are not aware of any specific proof of the assumption that the
celluloses are ab initio non-nitrogenous. It is possible that they contain
NH.j residues in the earlier phases of growth.

238 Cellulose

of cellulosic tissues is a frequent phenomenon ; the process
by which the tissue is broken down is of the character of an
ordinary hydrolysis, i.e. is determined by enzymes, and postu-
lates therefore a cellulose susceptible of molecular disaggrega-
tion. The ' hemi-celluloses ' (p. 87) are compounds of this
order, and occur chiefly in seed-tissues, where they serve as
reserve materials for the early growth of the embryo on germi-
nation. The more resistant celluloses and compound cellu-
loses are obviously destined for more permanent functions, and
they have been dealt with in this treatise under the term * per-
manent tissue.' But permanence is a relative term. We have
already endeavoured to trace the changes and modifications
which these substances undergo in the normal life of the
plant : lignification, for instance, has been dealt with as a
progressive chemical transformation of celluloses, differing
probably ab initio from those of the normal type ; and evidence
has been given in terms of the constitution of these compounds,
showing them to be highly reactive and susceptible of modifi-
cation in various directions under treatment in the labora-
tory.

We have now to follow the fate of these substances in the
ordinary processes of the natural world. Under these cir-
cumstances, * death ' is succeeded by a variety of processes of
decay and destruction. These are in part intrinsic, in part
determined by outside agencies ; they are of the two kinds :
processes of resolution, and processes of combination or con-
densation, which are usually concurrent. The former have
been dealt with on p. 66. These are fermentation processes
attended with evolution of gaseous products. The latter are
defined by an extended series of their products, viz. humus,
peat, lignite, and the coals. The chief characteristic of this
series of degradation products of plant tissues is the accumula-
tion of carbon at the expense of oxygen and hydrogen. This

Adipocelluloses and Cutocelluloses

239

is illustrated by the following results of elementary analyses,
calculated in all cases on the ash-free substance (W. A. Miller,
Org. Chem. ed. 1869, 139-146).

Peat

Oak wood

Decayed
oak wood

Humus
from decayed oak

Dartmoor

Vulcain

c .

50-20

53-50

54-0

56-0

5973

59-57

H .

6-08

5-16

5*1

4'9

5-96

0 .

4374

41-34

40-9

39-i

31-82

32-38

N . - *

2*54

2-09

A p p r o x. i

fo r m u 1 a [

C34H48O22

—

—

—

CjoH^Og

—

(C,H,0) J

Coals.

—

Lignite
Bovey

Scotch

Wigan

Cannel

Newcastle

Anthracite

c .

67-85

78-46

82-29

87-97

91-87

H . ',-.;'

576

8-n

5'68

5-3I

3-33

o .

N . ' V

23-39 1
0-58]

1373

(8-31)

\2-i8]

672

( 3-01

1 0-84

A p p r o x. \

form u la [

C27H2807

C25H3203

C^H^O,

c^o,

C40H160

(C,H,0) )

The coals are chemical aggregates of altogether unknown
constitution. By certain observations, however, they are
definitely connected with the earlier members of the above
series ; thus they yield substitution derivatives with chlorine
(Cross and Bevan, Chem. News, 44, 185), and yellow acid pro-
ducts on treatment with nitric acid, of similar composition to
those obtained from humic compounds, natural and artificial
(infra). The products of their destructive distillation (coal tar),
which it is needless to say have been exhaustively investigated,
throw, it is true, a certain light on their constitution ; but they
are too complex for the establishment of definite relationships
to the parent molecules.

HUMIC COMPOUNDS : HUMUS. — These series of compounds

240 Cellulose

are normal constituents of all soils in which they fulfil im-
portant functions. They are produced in the ordinary decay
of buried vegetable matter. They are generally of acid pro-
perties, and are dissolved by alkalis to brown solutions ; they
have the characteristics of unsaturated compounds, being
readily acted upon by the halogens. They closely resemble the
products of decay of forest woods under ordinary conditions.
As chief constituents of such mixtures, Mulder recognised
geic acid, C20H12O7; humic acid, C20H12O6; and ulmic acid,
C20H14O6. These names and formulae, however, have little more
than empirical value. Products closely similar to these are ob-
tained on a large number of decompositions of the carbohydrates,
both simple (sugars &c.) and complex (starch, cellulose, &c.).
Those obtained by long boiling of the sugars with dilute acids
have been carefully investigated by Sestini (Gazzetta, 10, 121,
240, 355). The products are of two classes : (a) soluble in
alkalis, sacchulmic acid, €4411400! 6 ; (b) insoluble, sacchulmin,
€4 ^agOi 5 ] both yielding,however, the same substitution products
with the halogens, of which the following may be mentioned :
dichloroxysacchulmide, CnH8Cl2O6 (or C44H32O8O24), and
sesquibromoxysacchulmide, C22H18Br3O11 (or C44H36Br6O22).
Products very similar to these have been obtained by the
authors from the alkaline by-products of the esparto pulping
process, viz. by acidification and purifying the precipitate,
a body of constant composition, C42H48O8, converted by treat-
ment with HC1 and KC1O3 into the derivative C44H46C18O20
(J. Chem. Soc. 38, 668 ; 41, 94). The lignic acids described
by Lange (see p. 213), and obtained by the action of alkaline
solutions upon the woods, are also of similar composition.

On further treatment (fusion) with the alkaline hydrates, these
products yield ' aromatic ' derivatives, of which protocatechuic
acid is obtained in largest quantity (Lange, loc. cit. ; Demel,
Monatsh. 3, 769). It must be admitted, however, that the

Adipocelluloses .and Cutocelluloses 241

entire group of products under consideration is extremely
ill-defined, and requires to be much more exhaustively in-
vestigated to establish definite relationships with the carbo-
hydrates from which they result.

Reverting to the aspect of their natural history, it is clear
that the functions of this diversified group are completed in a
very remarkable way by this property of carbon condensation.
In proportion as they are attacked by destructive agencies, the
residue tends to constitute itself into a complex of increasing
resistance ; and so the chemistry of the vegetable world, which
depends in its proximate relationships upon the properties of
polyhydroxy derivatives of the C6 unit, is ultimately a most
striking manifestation of the properties of the carbon atom
itself.

R

242 Cellulose

PART III

EXPERIMENTAL AND APPLIED

THE foregoing general and theoretical treatment of the subject
will carry with it a number of suggestions to readers of original
and independent habit of mind. Students accustomed to the
critical appreciation of essays in the ordering of such subject-
matter will have noted the imperfect state of development to
which the investigations hitherto prosecuted have brought our
knowledge of the celluloses and compound celluloses ; and an
appreciation of these imperfections implies the possession of
the key to the effectual remedy, i.e. to those directions of
investigation most certain to lead to valuable contributions to
the subject. Students, however, are by no means generally of
this class. The habit of initiation and enterprise in scientific
work is perhaps rare, not because of want of capacity, so much
as of opportunity and the abiding stimulus of necessity. A
main purpose of this treatise is to encourage original investi-
gation by opening out, in more definite terms than has been
done, the view of a region of the res publica natura extremely
rich in possibilities. This view should not be limited, moreover,
to the more ardent few who have imaginative and speculative
tendencies. The mastery of method is in itself a useful objec-
tive, and the subject of cellulose involves a number of experi-
mental methods which are not merely essential to the work of
pioneer investigation, but a necessary part of the general
training of the chemist.

Experimental and Applied 243

In the following pages attention is directed specially to
experimental methods, whether for purposes of demonstration
or of training in analytical processes of general usefulness.
These methods have been described in general terms in the
earlier sections ; and for the present purpose we shall continue
to avoid minute description of practical details, referring the
student to the easily accessible accounts of the actual processes
to be found in current literature. At the same time, suggestions
are included for a course of study, whether of general treatises
or special publications on particular subjects, designed to
familiarise the student with the work of investigators in this
field of natural science.

Laboratory and Research Notes.

Morphology of Cellulose. — The study of cellulose and of the
plant fibres generally involves on every hand questions of form and
structure, i.e. the form and dimensions of the individual cell or
fibre, and the structure or anatomy of the tissue of which it is a
unit member.

Microscopic examination and investigation are, therefore, the
necessary complement of the chemical study of vegetable sub-
stances.

The histological study of minute structure is dealt with in all
the standard text-books of the microscope. The structure and
elaboration of plant tissues, and the anatomy of the exogenous and
endogenous stems are fully described in the text-books of botany
which must be consulted. Goodale's Physiological Botany (Ivison :
New York) is especially to be recommended.

The most important works treating specially of the minute struc-
ture of plant fibres are Wiesner, Mikroscopische Untersuchungen
(1872), Die Rohstoffe des Pflanzenreichs (1873) 5 Vetillart, Etudes
sur les fibres vege'tales textiles (Paris, 1876). Useful information
will also be found in Spon's Encyclopedia of the Useful Arts, article
'Vegetable Fibres' ; and * Indian Fibres and Fibrous Substances'
(Cross, Bevan, and King : London).

As a useful application of microscopic method the student

244 Cellulose

should sow some flax (linseed) and examine the stem at intervals
throughout the growth of the crop, noting the development of the
bast fibre — which constitutes commercial flax — and its structural
relationship to wood and cortex.

Mount as permanent preparations the typical paper-making
celluloses : cotton, flax, hemp, wood, esparto, and straw. Note
dimensions and typical characteristics. Study the photomicro-
graphs of raw materials (sections appendix), And an account of
methods of photographing, in Indian Fibres, p. 15.

Bleaching, or Isolation of Cellulose from Raw Fibres. — Take
flax and jute as typical raw materials.

Flax. — Extract the fibre in continuous extraction apparatus
(Soxhlet) with fat solvents — e.g. alcohol — ether. Make quanti-
tative estimations, and for study of properties of oil-wax extract
see Cross and Bevan, J. Chem. Soc. 57, 196.

Boil extracted residue with 2 p.ct. solution NaOH one hour.
Examine solution for pectic bodies and note properties. Wash,
and bleach fibre in 0-5 p.ct. solution sodium hypochlorite ; wash
off, treat with sulphurous acid, wash, dry, and weigh. Note that the
fibre though colourless has a blackish look. A second portion boil
as before, wash, and digest in bromine water some hours. Wash
and boil out in carbonate of soda solution. Wash and return to
bromine water. Repeat till brilliant white cellulose obtained.
Note the gradual disintegration of the woody portions (' sprit ')
which adhere to the fibre.

Jute fibre. — Boil 15 minutes in I p.ct. NaOH solution ; wash off,
squeeze, and expose to chlorine gas one hour. Note formation of
yellow chlorinated derivative of non-cellulose. Wash off and
place in 2 p.ct. solution sodium sulphite. Raise to boil, adding a
little NaOH. Wash off and treat residual cellulose with sulphurous
acid. Wash off, dry, and weigh.

Pine wood. — Take deal shavings or ' wood-wool ' and proceed
as with jute, repeating the treatment with chlorine until pure
cellulose obtained.

Contrast these laboratory methods with those of the bleach
works (cotton, linen) and the paper-maker's ' pulping ' processes,
for which see the standard text-books on these subjects, more
especially Chemistry of Paper Making (Griffin & Little : New
York, 1894).

Experimental and Applied 24$

Note the correlations of lustre and external appearance with
minute structure.

RAPID-COMBUSTION METHOD.— Study method of com-
bustion with chromic acid in presence of sulphuric acid (Cross and
Bevan, J. Chem. Soc. 53, 889).

Inorganic or Ash Constituents. — Burn specimens of pure cellu-
lose and raw materials and estimate ash. Note that the ash con-
stitutes an inorganic skeleton of the original. Soak a cotton fabric
(muslin) in solutions of ammonium chromate, aluminium acetate,
or magnesium acetate ; dry and burn. Note that a coherent ash
is obtained preserving the structure of the original fabric. A process
of this kind is used in making the hoods for the * incandescent '
gas burners j oxides of the rare earth-metals being used for the
purpose.

Take an ordinary filter or blotting paper ; estimate the ash, and
a second portion digest in dilute hydrofluoric acid in a lead or
platinum vessel. Wash off with distilled water, dry, and determine
the ash. It will be found to be considerably reduced, owing to the
removal of silica and siliceous compounds. The ash in cellulosic
raw materials is much higher than in the bleached celluloses ;
notably in the straws. The proportion of silica is exceptionally
high. It was frequently affirmed in the older text-books that the
rigidity of the straws was due to this constituent, but special culti-
vations in absence of silica give a straw of normal appearance
and rigidity. The structural properties of these stems are there-
fore referable to their 'organic' components.

Hydration and Dehydration. — For a wider discussion of hydra-
tion and dehydration types as applied to the elucidation of
the special chemistry of the carbohydrates, the student should
read a paper by Baeyer, ' Ueber die Wasserentziehung und ihre
Bedeutung fur das Pflanzenleben u. die Gahrung' (Berl. Ber.
1870, 363); or the translation by Armstrong (J. Chem. Soc. [2],

9, 330-

The absorption, osmotic and capillary phenomena resulting
from the interaction of cellulose and aqueous solutions should be
carefully studied. Much work remains to be done on this subject.
The student should acquaint himself with Pfeffer's methods of
determining osmotic pressures, and the bearings of the phenomena
of osmosis on the theory of solution.

246 Cellulose

Cellulose Solvents. — Prepare solutions of cotton and other
celluloses by treatment with zinc chloride. To dissolve I part
cellulose take 4-5 parts Zn.Cl2, dissolve in 5-7 parts water, add
the cellulose, heating in a porcelain dish over a water-bath. Stir
from time to time, and add water to replace that which evapo-
rates.

Precipitate the solution (i) by pouring into water. Wash till
free from Zn salt, dry, and weigh. Burn off the cellulose, and
estimate the residue of Zn.O.

(2) Pour into dilute hydrochloric acid. The precipitate will be
cellulose hydrate free from ZnO. Wash thoroughly, dry, and weigh.

Calculate to original cellulose, and show that the molecule is
hydrated. Control by examining the solution in which the cellulose
was precipitated for dissolved products of hydrolysis.

(3) Precipitate by pouring the solution into alcohol in a fine
stream, or by spreading the viscous solution as a thin film on glass.
Submerge the whole in alcohol, and detach the coherent film of the
cellulose compound. Digest with HCl.Aq to remove Zn.O. Wash,
and dehydrate by lengthened exposure to alcohol.

The progress of the action of the zinc chloride solution should
be followed up microscopically. Cotton fibres are mounted in the
strong solution, covered with cover glasses, and the glass slips
ranged on a hot surface, so that they may acquire a temperature of
80-90°. Examine from time to time under low and high power,
noting the progress of the disintegration, swelling-up, and rupture
of the cell wall. The structural points may be further differentiated
by staining with iodine. To apply the iodine without precipitating
the dissolved cellulose it is necessary, after dissolving the iodine in
a little potassium iodide in the usual way, to dilute with 40 p.ct.
zinc chloride solution. A drop of this solution may be added to the
solution on the slide without otherwise affecting the cellulose than
staining the fibrous residues.

Dilute ammoniacal solutions of cuprous oxide are without action
upon cellulose. An interesting demonstration of the difference
between the cuprous and cupric oxides in this respect may be carried
out as follows : Cuprous chloride is prepared by the action of hydro-
chloric acid upon copper, with addition of potassium chlorate in suc-
cessive small quantities. The chloride is washed with water free
from oxygen, and, after settling, portions of the magma of crystals

Experimental and Applied 247

and water are placed upon a filter paper, which is then transferred
to a bottle provided with an indiarubber cork. Through the cork,
glass tubes are passed, and so arranged that a stream of coal gas
may be led into the vessel to completely expel the air. After ex-
pelling the air, a little strong ammonia solution is introduced by
means of a separating funnel previously inserted through the cork.
No change in the substance of the paper is observed. The stream
of coal gas is now stopped, and atmospheric air is introduced.
Reaction rapidly ensues, with development of the blue colour of
the ammoniacal cupric compound, accompanied by solution of the
cellulose. The details of the arrangement of the experiment are
immaterial, and may be varied in a number of ways.

On the theory of the action of cellulose solvents, read a paper
by the authors in Bull. Soc. Chim. 1893, 295.

Cellulose Xanthate. — In preparing the solution of the cellulose
thiocarbonate for use in the laboratory it is better to employ a ' rag '
cellulose (cotton and linen) disintegrated by the process of * beat-
ing ' in a paper mill, which can be easily procured. If obtained
in the moist condition the cellulose may be air-dried previously
to treating with the 15 p.ct. solution of NaOH. The cellulose and
the alkaline lye may be thoroughly incorporated by grinding
together in a mortar in the calculated quantities to finish the
alkali cellulose so as to contain

Cellulose, 100 ; caustic soda (NaOH), 45 ; and water, 250-300 ;

or the cellulose may be treated with the 15 p.ct. NaOH solution
in excess, and, after standing some time, the cellulose may be
drained off on a filter of perforated zinc and squeezed to retain
three times its weight of the solution.

As the process of 'mercerisation ' requires some time for com-
pletion, it will be found advantageous to set aside the moist alkali
cellulose in a closed bottle for two or three days before sub-
jecting it to the solvent reaction. This reaction proceeds spon-
taneously at the ordinary temperature ; the alkali-cellulose and
carbon disulphide (40-100 parts cellulose) are brought together
in a stoppered bottle, vigorously shaken together for a minute
or two, and set aside for two or three hours. The resulting
yellowish mass is covered with water and allowed to stand some
hours, then vigorously stirred with addition of water, in quantity

248 Cellulose

calculated to produce a solution of any desired percentage strength
(cellulose).

The progress of the reaction should be followed by microscopic
examination of the fibre at intervals. The rupture of the cell walls
is preceded by an exaggeration of the structural details of the fibre,
which will be found useful in differentiating cotton from celluloses
of other types.

From the solution of the crude thiocarbonate the solid product
is precipitated, in the form of a gelatinous hydrate, by various
'neutral' dehydrating liquids or solutions: of these the most
convenient are alcohol and a solution of common salt, and the
modus operandi may be instructively varied in the following
ways : —

(1) Alcohol. — The solution maybe poured into a photographer's
developing dish to a depth of, say, £ inch, strong alcohol may then
be poured upon the solution, the dish covered with a glass plate,
and set aside. Coagulation of the cellulose product proceeds gradu-
ally, and after some time a coherent slab is obtained of a greenish
colour, the yellow by-products of the reaction having been dis-
solved by the alcohol.

Or the solution may be caused to flow in a fine stream into the
alcohol, when the cellulose xanthate (hydrated) will be precipitated
as a continuous gelatinous thread. In the latter case it is advisable
to add a certain quantity of alcohol to the crude solution, which can
be done without causing precipitation, the xanthate being soluble
in dilute alcohol. Such a solution will obviously be more sensitive
to the further action of the alcohol, and precipitation in slender and
uniform threads is facilitated.

(2) Brine. — If prepared from the commercial salt, the solution
should be freed from magnesia by previous treatment with sodium
carbonate and filtering. The precipitation may be carried out as
described under (i), or any surface may be coated with the
crude solution, and the xanthate precipitated upon the surface by
immersion in the brine. The xanthate, purified by any of these
treatments, may be redissolved in water, and again precipitated by
a repetition of the treatment. With each precipitation the ratio of
both alkali and sulphur to cellulose in the product is considerably
diminished.

The precipitated xanthate may be treated with solutions of

Experimental and Applied 249

suitable salts of the heavy metals, and the cellulose xanthate of the
metals prepared.

Cellulose is regenerated from the solution in various ways which
may be practised by the student : (a) the solution may be set aside
in any suitable vessel for some days ; (b] it may be heated at 80-
90° in a water bath ; (c} it may be spread upon a glass plate or
other surface, dried at 60° C., and finally heated for some minutes
at 100° C.

The masses or films obtained are washed to remove the alka-
line by-products ; and the cellulose bleached, if necessary, by treat-
ment with sodium hypochlorite in dilute solution (0*5-1 'O p.ct.
NaOCl). Of the special uses of the solution in the laboratory, the
following may be particularised : —

(1) In Microscopic Work. — In preparing fibres for cutting in
cross section, they may be worked up with the viscous solution into
a strand or pencil, the cellulose being regenerated spontaneously or
otherwise. Such a strand may be cut with or without a microtome,
or may be further * embedded ' in cellulose by submerging it in the
viscous solution and coagulating. In some cases the crude solution
may be used ; in others the solution should be previously dis-
colourised (and neutralised) by treatment with sulphurous acid.

(2) In diffusion experiments (osmosis) the regenerated cellulose
is of use in the preparation of membranes or diaphragms, the
cellulose being either used alone, or a compound diaphragm may
be made by coating paper or cloth with the thick solution, and
' fixing ' the cellulose by any of the methods described.

Important results may be expected from a study of osmotic
transmission through this form of cellulose. The passage of
crystalloids through the cellulose, especially in its hydrated forms,
appears to be exceptionally rapid.

Theoretical Notes. — Determinations have been made in which
known weights of pure cotton cellulose have been dissolved as
thiocarbonate, the cellulose regenerated, spontaneously and in
other ways, then carefully purified and weighed. The cellulose
recovered shows a slight increase upon the original. It may be
taken, therefore, that the molecule undergoes no permanent dis-
aggregation under the treatment, and it is even probable that the
cellulose maintains a high molecular weight in solution. In con-
trast to its behaviour under the severe alkaline treatment, it has

250 Cellulose

been noted that when dissolved in zinc chloride, more especially
in presence of hydrochloric acid, there is a progressive hydrolysis
of the molecule, with conversion into soluble products.

It is evident, therefore, that the hydration of cellulose has
nothing in common with hydrolysis ; the cellulose molecule
appears to have an indefinite capacity for combining with water
and for undergoing an extensive series of hydration changes with-
out affecting its fundamental constitution.

Cellulose is usually regarded as a * condensed ' derivative, pre-
senting more or less analogy with starch. But in resistance to
hydrolysis there is a marked and very important distinction, the
due interpretation of which will no doubt result in establishing for
the celluloses a special constitutional type.

We write the cellulose unit as C6H10O5, remembering that this
merely represents the empirical facts of its ultimate elementary
composition. It is of course evident that at least one of the O
atoms is present as carbonyl oxygen. Further, we may regard
the constituent groups of the molecule as grouped around the CO
or negative, and a CH2 or positive centre. There are many evi-
dences of a ' polarity ' of this character manifested by cellulose, as
by other * carbohydrates,' in reaction. The thiocarbonate reaction
proper we may regard as localised at the more negative OH groups,
i.e. those in proximity to CO, which are held in combination by the
alkali used in mercerisation or the formation of alkali cellulose.

The attendant phenomena of mercerisation indicate a consider-
able degree of hydration of the cellulose (combination with water),
associated with the entrance of the alkaline groups. This gela-
tinous condition of the alkali cellulose hydrate conforms with the
characteristics of the earlier stages of the process of solution of
colloids. The completion of the process requires in this case the
additional strain of the carbon disulphide, evidently acting as an
acid group. It is certainly noteworthy, as already pointed out,
that all the solvents of cellulose (as cellulose) are of a saline cha-
racter, and may be regarded as forming with the cellulose, by
reciprocal combination with its acid and basic groups, compounds
analogous to the" double salts. The actual mechanism of solution
we cannot pretend to follow. The modern theory of solution in the
hands of its most advanced exponents confines itself to the inves-
tigation of the molecular condition of substances in solution ; and,

Experimental and Applied 251

in the case of electrolytes, it is sufficiently established that solution
is attended by dissociation into ions.

This aspect of solution may very well be taken up in dealing
with the solution phenomena of the carbohydrates. The lower
members of the series, in solution, show in most cases considerable
sensitiveness to the action of acids and bases, and form a number
of * molecular ' compounds with salts such as sodium and calckim
chlorides — which may very well be considered as double salts-
This indicates a saline character of the molecule in solution. The
phenomena of alcoholic fermentation also point in the same
direction. The resolution of a dextrose molecule into alcohol and
carbonic acid has entirely the characteristics of an electrolytic
split ; and it is fair to assume that the molecule in aqueous
solution already manifests a ' strain ' in the direction in which
cleavage ultimately takes place. In regard to cellulose in solution,
it is also reasonable to assume a similar * polarisation ' ; and the
reactivity of the cellulose regenerated from solution is explained
by assuming that this polarity is in a measure retained.

In regard to the hydrates of cellulose, and the extraordinary
power which cellulose has of solidifying water, it appears that
cellulose must have the capacity of absorbing into itself the * energy
of condition ' of water.

It would be important to determine the physical constants
— volume and vapour tension — of the water, in the form of a series
of definite hydrates of cellulose. This would afford a measure of
transference of energy from the water to the cellulose.

Cellulose Benzoates. — These reactions have been studied only
under ordinary conditions of temperature, and under conditions
where there is considerable loss of the anhydrochloride in waste
reaction with the alkali. This impedes the direct synthetical
reaction between the cellulose and benzoyl residues, and causes
large variations of the product. The reaction is necessarily diffi-
cult to control, especially in the case of the insoluble alkali cellu-
loses, and requires further systematic investigation from this point
of view, which will also result in the formation of definite and uni-
form products, i.e. so far as is attainable with a reaction in which
the original substance and derivative are both insoluble in the
reaction medium.

The conditions requiring systematic variation are (i) the

252 Cellulose

temperature : the reaction should proceed at the lowest limit
of temperature ; (2) the conditions of mercerisation : the reaction
would probably be favoured by employing a certain proportion of
zinc oxide in replacement of sodium hydrate ; and (3) the liquid
medium. There is, of course, a rapid hydrolysis of the benzoyl
chloride by the solution of sodium hydrate ; and there are various
ways which suggest themselves of retarding this waste reaction.

The precipitation, as benzoates, of modifications of cellulose
soluble in alkaline solutions should prove a valuable aid to investi-
gation ; but, before the method can be applied to the investigation
of substances and mixtures of unknown or lesser known composi-
tion, it will be necessary to characterise the products obtainable
from the normal celluloses. There are in plant tissues a widely
diffused group of substances having the external characteristics of
the resistant celluloses, or celluloses proper, but which are readily
soluble in alkaline solutions. In the course of an ordinary
proximate analysis these are usually ' dismissed ' as ' non-nitro-
genous extractive matters ' ; which, of course, is but a very loose
description.

We have seen that hydrated modifications of the normal cellu-
lose are themselves soluble in alkaline solution, and it would
appear that in this condition they are generally more reactive.

There are two practical problems suggested by the properties
of these hydrates : (i) Do they yield to the processes of animal
digestion ? and (2) how far can the processes adopted by the
paper-maker for isolating cellulose from complex fibrous materials
be limited, to prevent the solution of celluloses either existing or
hydrated in the raw material or tending to pass into such modifi-
cations ?

The investigation of these questions will be found to be greatly
facilitated by a method of separating the soluble celluloses from
solutions. The benzoates give promise of affording such a method
of separation, and invite investigation from this point of view as
much as from that of their intrinsic theoretical interest.

Cellulose Acetates.— The * suppression ' of the OH groups of
cellulose is evidenced by its not reacting with acetic anhydride
at its boiling point ; whereas the poly-hydroxy-compounds gene-
rally are acetylated under these conditions, and many even react
with acetic acid itself. We may find some explanation of this in

Experimental and Applied 253

the evidence that we have of a species of saline or ethereal con-
stitution which characterises cellulose in many of its reactions. It
is probable that the OH groups of opposite function exert a
mutually repressive influence ; and acetic anhydride, being essen-
tially a condensing agent, is unable to determine the liberation of
the OH groups into the condition in which they can react.

On the other hand, the presence of certain reagents in relatively
small proportion is sufficient to disturb the equilibrium, and
reaction results. An elevated temperature also determines reac-
tion, and obviously for a similar reason, viz. that the equilibrium
of the molecule holds only for a certain range of conditions.

For these reasons it may be considered doubtful whether a
cellulose acetate in the strictest sense can be obtained. Acetates
are, however, obtainable which satisfy the * general ' & priori defini-
tion— viz. which by saponification yield a carbohydrate having the
negative characteristics of cellulose, and in contradistinction to
acetates which, though obtained from cellulose, yield on saponifi-
cation a carbohydrate more or less soluble in the alkaline solution,
and reducing CuO.

Parchmentising Process. — The student should make careful
statistical observations upon this reaction. An unsized paper
should be used — ordinary ' waterleaf ' or filtering paper. The acid
should be placed in a 'photographic' dish of suitable size; the
sheet immersed in this and, after suitable exposure to the action
of the acid, transferred at once to a dish of water. After staying
in this first wash water for some time they must be thoroughly
washed in a flow of water until neutral. The sheets having been
weighed before treatment, and the moisture estimated in portion
of the same paper — the product must also be weighed after drying.
It is necessary also to determine the cellulose dissolved by the
process, for which purpose the acid and first wash water may be
treated. These may be mixed and a certain fraction boiled, the
dextrose formed being estimated by any of the well-known methods ;
or the solution may be neutralised with lime evaporated, and the
total carbon estimated, by the method of combustion, with CrO3
and H2SO4.

The product may also be further examined (i) for loss of
weight in boiling with dilute alkaline solutions ; (2) CuO reduction
in presence of alkalis (Fehling's solution) ; and (3) degree ol

254 Cellulose

acetylation, which is a measure of the OH groups rendered 'free'
or reactive by the process. A systematic series of observations
embracing these points would be a useful contribution to our
knowledge.

1 Toughening ' Action of Nitric Acid upon Papers. — The student
should read the original paper of Francis (J. Chem. Soc. 47, 183)
and repeat his observations on the toughening of filter paper by
this process.

Preparation of Hydrocellulose. — This reaction may also be
studied by the student statistically. A given weight of a pure
bleached cellulose fabric, preferably in the form of rag, is exposed,
in the air-dry condition, to hydrochloric acid gas in a closed vessel.
The reaction is complete when the fabric breaks down on slight
pressure to an impalpable powder. It may, however, be more
conveniently washed in the original form of cloth, which it will
retain sufficiently — notwithstanding the disintegration which has
taken place — to withstand a current of wash water, carefully ad-
mitted to the bottom of the vessel. The product when pure may
be dried and weighed, and its properties and reactions determined.

The student should convert the product into the corresponding
nitrates. One part of the product (dried at 105° C. and cooled in
a dry atmosphere) is gradually stirred into 5 parts of concentrated
nitric acid (HNO3 of i'5 sp.gr. and colourless). The gummy solu-
tion of the nitrate may be treated in various ways for the separa-
tion of various products.

(a) It may be diluted with a small proportion of water, gradually
stirred in so as to keep the product dissolved ; the solution then
poured on to a glass plate, and the acid evaporated at a temperature
of 40° C. in a free draught of air. The product is then washed, and
detached as a film from the glass.

(b) The concentrated or somewhat diluted solution may be
poured into water or spread on a glass plate, which is then plunged
beneath water contained in a suitable vessel.

(c) The precipitation may be effected by sulphuric acid— the
acid being added drop by drop to the concentrated solution, or
the solution poured into the sulphuric acid. In either case the
mixture must be carried out without sensible rise of temperature.

Preparation and Diagnosis of Oxycelluloses. — The student of
cellulose technology will find it necessary to study the original

Experimental and Applied 255

papers (loc. cit.\ and to repeat the experimental demonstrations
therein given, of the formation of these oxycelluloses, and the
methods of diagnosing their presence in textile fabrics.

It is to be noted that the ordinary bleached celluloses of the
cotton group all give a small proportion of furfural (0-3-1*0 p.ct.) on
boiling with hydrochloric acid(ro6 sp.gr.), which may betaken as
indicating the presence of an oxycellulose in proportionate quantity.
It is probable also that the somewhat increased reactivity of
bleached cotton may be due to formation of oxycellulose.

Study of Methods of Bleaching. — The student should compare
the bleaching actions of potassium permanganate and sodium
hypochlorite — from the point of view of bleaching effect, and
relative consumption of oxygen. It must be noted that the per-
manganate (i.e. Mn2O7) is deoxidised to the dioxide MnO2, which
is deposited as the brown-coloured hydrate upon the cellulose ;
the depth of colour is therefore a measure of the oxidation which
has taken place, generally or locally. The oxide is removed by
treating the substance, after washing from the alkali simultaneously
set free, with a solution of sulphurous acid, the interaction of the
reagents producing dithionic acid. With the disappearance of the
brown oxide the bleaching effect produced by the original oxidation
is at once apparent.

While the action of the permanganates upon the cellulose is
necessarily that of simple oxidation, the hypochlorites may act in
two ways : (i) a simple oxidation ; (2) chlorination. The latter
effect is usually small, the proportion reacting in this way depend-
ing upon the temperature of the solution and also the nature
of the base with which the hypochlorous acid is combined. See
* Some Considerations on the Chemistry of Hypochlorite Bleaching '
(J. Soc. Chem. Ind. 1890).

Formation of Acetic Acid from Cellulose. — The student should
read the account of a systematic investigation of these reactions by
J. F. V. Isaac and the authors (J. Soc. Chem. Ind. 1892).

The maximum yields of 30^0 p.ct., which are obtained under
the most favourable conditions, point to a CO — CH2 'residue3 in
the cellulose molecule itself, as the immediate source of the acetic
acid.

Ferment-hydrolyses of Cellulose. — It is obvious that chemical
compounds destined to transmit, accumulate, and otherwise respond

256 Cellulose

to, the form of energy which we are bound to express by the some,
what colourless term ' vital force ' must exhibit a specially plastic
constitution. The peculiar sensitiveness of the carbohydrates as
a group to fermentation and other decompositions is not sufficiently
accounted for by the constitutional formulas assigned to them.

Destructive Distillation. — Theoretical Notes. — We may by care-
ful consideration form some mental picture of the consequences
of the addition of heat to compounds of this class. As a preliminary
it is necessary to remember generally the distinction between the
purely chemical and the chemico-physical view of the constitution
of matter. According to the former, cellulose is a compound
72.C6Hj0O6, of which the ultimate constituent groups — CO — CH2 —
CHOH — are arranged in a certain way in the C6 units ; and these,
again, are grouped together in a special configuration which, when
elucidated, will represent the molecular constitution of cellulose.
The correlative of this view is that of the intrinsic or internal
energy of the compound. This is expressed as a crude aggregate
— on the thermo-chemical view — as what is known as the ' heat of
formation,' which, in the case of carbon compounds, is the difference
between the sum of the heats of combustion of its constituents,
and the heat of combustion of the compound itself, expressed in
arbitrary units of mass, usually the gramme, but in molecular
ratios. The methods of determining combustion-heats are neces-
sarily subject to errors of observation of some magnitude ; but,
even with much closer approximation to accuracy, the number of
calories per gram-molecule of a compound evolved on burning is
still of only empirical and statistical value, even when applied to
compounds related in differential series.

Whatever the meaning of the constant of energy, which we regard
as associated with every particular configuration of matter, it is at
least obvious that all compounds with which we are familiar repre-
sent an equilibrium of matter and energy under the conditions of
observation ; and as combination is generally associated with loss
of energy, so, vice versa, the introduction of energy implies decom-
position or tendencies to decomposition. The communication of
energy in the form of the electric current, as in electrolysis, is
usually attended by more or less simple decompositions, more or
less easy, therefore, to follow. The effects of heat, on the other
hand, are complicated by recombinations of the products amongst

Experimental and Applied 257

themselves, more especially in the regions of high temperatures,
rendering it somewhat difficult to distinguish the primary from
secondary products.

Generally we may take the saturated compounds of lower
molecular weight as direct products of the decomposition : these are
water, oxides of carbon, methane, methyl alcohol, formic and
acetic acids, £c. Volatile compounds, either hydrocarbons or their
derivatives, containing unsaturated C — H nuclei, result from re-
arrangements of the component groups of the original celluloses ;
these are defines and aromatic hydrocarbons, furfural, phenols,
aromatic methoxy-derivatives, &c. ; the residue of the process — or
charcoal — represents the extreme limits of condensation of the
carbon nuclei.

One important general feature of the decomposition is notice-
able as connecting the decomposition with those determined by
electrolysis and fermentation in the case of compounds of similar
constitution, that is, the accumulation of hydrogen in the one direc-
tion, and oxygen in the other. Thus we have, on the one hand,
methane and a large number of compounds containing the CH3
and O.CH3 group.; and, on the other, CO and CO2. Acetic acid
represents an intermediate equilibrium, viz. of CH4 and CO2. We
have already seen that the introduction of the alkaline hydrates so
alters the character of the decomposition that acetic acid becomes
a main product of decomposition ; and the remainder of the re-
action consists in the production of the more fully oxidised oxalic
acid (with CO2), * balanced,' so to speak, by the formation and
liberation of hydrogen. These considerations illustrate the views
which may be applied to the elucidation of the very complex pro-
blems in dissociation, generally presented by destructive distillation.

Constitution of Cellulose.— Theoretical Notes.— It is, perhaps,
premature to approach the problem of the constitution of cellulose,
as the molecule of cellulose is at present an altogether unknown
quantity. It will be instructive to the student to study, for their
general bearing on the subject, the investigations which have been
devoted to the elucidation of the structure of the starch molecule,
notably those of O'Sullivan (J. Chem. Soc. 1872), Brown and
Morris (ibid. 1889, 449-462), Lintner and Dull (Berl. Ber. 26,

2533).

These researches are based on the study of the hydrolytic dis-

258 Cellulose

section of the molecule, which breaks down ultimately into maltose
when invertase is employed, or into dextrose when the hydrolysis is
completed by acids, the intermediate terms being the uncrystal-
lisable dextrines. Owing to the obviously simple relation of the pro-
ducts of resolution to the original molecule, we are justified in con-
cluding that starch represents an aggregate of considerable mole-
cular magnitude ; the ultimate constituent C6 groups being linked
together by oxygen. The evidence, in fact, goes to show that starch
is of the same constitutional type as the bioses (e.g. cane sugar and
maltose) and trioses (e.g. melitriose). The combination of two
glucose residues, to form a biose, takes place in one of two ways—
i.e. either one or both of the typical CO groups takes part in the
condensation ; in the former case (e.g. maltose) the free carbonyl
determines aldehydic properties, which disappear, in the latter case,
(e.g. cane sugar) with the suppression of the second CO group.

In starch, the linking of the glucose residue is of the dicarbonyl
type, and it consequently fails to react with phenylhydrazine,
and does not reduce Fehling's solution. In the hydrolysis of
starch the carbonyl linkings are broken down successively, the
process being similar to that which takes place with the triose above
named, which is first resolved into melitriose and fructose, the
former then splitting into glucose and galactose. (See Scheibler
and Mittelmeier, Berl. Ber. 26, 2930.)

Cellulose is identical with starch in empirical composition
(«,C6H10O5), and similar, in being an aggregate of hexose groups ;
but for resolution into the latter, cellulose requires the severe
treatment of solution in concentrated sulphuric acid.

To give full value to these characteristic differences, we must
pronounce for a corresponding difference in molecular configu-
ration.

The problem must next be considered from the point of view
of function of the reactive groups of cellulose. The student should
be reminded that carbon chemistry is very much a chemistry of
function. The idea of chemical function grew but slowly with the
study of the * inorganic ' elements, and indeed of the carbon
compounds also, so long as it was limited to individuals and
groups more or less isolated. It was a chemistry of jumps and
gaps. In modern 'organic' chemistry, on the other hand, the
carbon compounds are presented to us in extended series of

Experimental and Applied 259

progressive differentiations — which are therefore, in the limited
sense of the term, * organically ' related. The differentiations are of
two kinds : (i) the reactive groups (CO, OH, &c.) may vary in
number and position ; (2) the substantive group (C«HW) may vary
in configuration (aliphatic, cyclic, condensed, &c., hydrocarbons).
To trace the corresponding variations in function or reactivity is a
very important part of every investigation ; and the student should
diligently study the papers of original workers, and familiarise
himself with such general expositions of the science, from this
point of view, as he will find in the opening chapters of Beilstein's
great work, Handbuch der organischen Chemie ; and more espe-
cially, in regard to the carbohydrates, in Tollens' monograph,
Die Kohlenhydrate (Breslau, 1888).

The student should endeavour to form a critical estimate of the
position of cellulose in relation to starch and others of the
* saccharo-colloids ' (Tollens), to which it is most nearly allied.
Such an estimate must be based upon the comparative study of
reactions.

Comparing cellulose with starch we find the former resistant
to hydrolysis and acetylation, but giving the highly characteristic
thiocarbonate reaction. These differences are differences of function
or reactivity of OH groups on the one hand, and of the linking
together of the unit groups on the other.

We have then to consider whether these differences are
sufficient to constitute the cellulose group a special type of
constitution. We think they are.

Assuming the unit group C6H10O6, and decomposing this,
on the evidence before us, into C6H6O.(OH)4, we have but few
alternatives. The conclusion we draw can only be expressed
within the limits of these alternatives, and therefore in general
terms, as follows : (i) The C atoms of the unit groups are com-
bined in a closed ring ; (2) the linking of the unit groups is not an
oxygen, but a carbon linking ; (3) fhe synthesis of the unit groups
together may be assumed to occur between the CO of one group
and the CH2 of another, giving the alternative form CH — C(OH).

LIGNOCELLULOSES OF CEREALS.— The student should
read the special papers on the subject, Berl. Ber. 1894, and J.
Chem. Soc. 1894.

The cereal straws— in which maybe included the envelopes of the

82

260 Cellulose

seeds of these and allied species — are of complex constitution. The
tissues, in addition to being lignified more or less (see Lignocelluloses*
p. 92), contain, as an essential constituent, the characteristic com-
pound known as wood gum (Holzgummi}. Wood gum is the
colloid anhydride of the pentaglucoses, yielding the crystallisable •
pentoses, xylose and arabinose, when hydrolysed by boiling dilute
acids. The pentosans are dissolved by dilute solutions of the alka-
line hydrates in the cold, from which solutions they are precipitated
on acidification, and more completely in presence of alcohol. The
pentosans, when boiled with hydrochloric acid (1*06 sp.gr.), of
course yield furfural, and it is sometimes assumed that the quan-
tity obtained from straw is a direct measure of the pentosans pre-
sent. This conclusion requires qualification, and it is necessary to
divide the furfural-yielding constituents into the two groups : (a)
the pentosans, easily hydrolysed, and obtained as above described ;
(£) the oxycelluloses, resisting hydrolytic actions of some intensity.
Together with the investigations above cited, the student should
read an account of the researches of Schulze and Tollens (Landw.
Vers.-Stat. 40, 367) upon the composition and constitution of the
substance of ' brewers' grains.' This material consists obviously
of the more resistant seed-envelopes of the barley, the greater por-
tion of which remains unaffected by the malting and mashing pro-
cesses. The following scheme represents the method of examination
pursued by the authors with indications of the results obtained: —

Acid Hydrolysis. — Boiling with 4 p.ct. aqueous H2SO4

Solution Insol. Residue

contains pentose, (45 p.ct. of original

chiefly xylose, some still giving reactions of

arabinose pentosans)

Exhaustive treatment with Digestion with 5 p.ct.

cuprammonium, filtering NaOH.Aq

and precipitating acid ; 53-2 |

p.ct dissolved and repre- ] |

dpitated Solution.— Pentosan Residue

giving xylose sol. in cu-

on hydrolysis prammonium

with very slight
residue

Experimental and Applied 261

An important conclusion from these experiments is, that the
cellulose and pentosan constituents of these tissues exist together,
in combination rather than in mere admixture, the complex also
containing the characteristic lignin groups. (See Ligno celluloses,
p. 156.)

General Methods for Identification of Carbohydrate Groups
(Hexoses, Pentoses, &c.}

The researches of Tollens and others (Landw. Vers.-Stat. 39,
401) have contributed considerably towards the completion of
methods of proximate resolution of such mixtures as it has been
customary to define by the term ' non-nitrogenous extractive matters.
In the investigation, more particularly of vegetable food-stuffs and
fodder plants, the so-called * Weende ' method of Henneberg has
been generally adopted. This consists in the direct determination
of fat, protein, { crude fibre,' and ash ; and the sum of these consti-
tuents subtracted from 100 is represented as 'non-nitrogenous
extractives.' These are the less resistant carbohydrates in question,
and concerning these it is important to determine whether they
contain (i) the true (hexa) carbohydrates or their anhydrides ; and
(2) the presence or absence of (a) dextrose, (£) galactose, (c) levulose,
(d) other carbohydrates, more especially mannose ; and (3) the
presence or absence of the pentaglucoses. It is also often necessary
to estimate these compounds quantitatively. It is also necessary
to keep in view the probable presence of derivatives of these com-
pounds or group, more particularly the oxidised derivatives. It will
be obvious that no generally applicable scheme of analysis can be
laid down ; but the following methods of diagnosis are typical, and
should be carefully worked out by the student.

(1) General identification of true (hexa} carbohydrates by levu-
linic acid reaction (Tollens, Ann. 243, 315). The substance is
heated for 20 hours at 95-98° with HCl.Aq (no). Levulinic acid
is extracted by exhaustion with ether, converted into the zinc salt
(cryst.), and this into the silver salt C5H7O3Ag, yielding 48-43 p.ct
Ag on ignition.

(2) Identification of dextrose groups by Jormation of saccharic
acid. — The substance is oxidised by digestion with HNO3 (1-15
sp.gr.) ; the saccharic acid converted into the acid potassium salt,
and this into the silver salt, giving 50-94 p.ct. Ag on ignition.

262 Cellulose

(3) Identification of galactose groups by conversion into mucic
acid.1 — The substance is oxidised with HNO3 (ri5 sp.gr.); the
mucic acid crystallising out directly, owing to its insolubility. Galac-
tose yields 75 p.ct. of its weight of the acid.

(4) Identification of levulose groups by reaction with resorcinoL —
Levulose may be sometimes identified by the relative ease with
which it is converted into levulinic acid. Oxidising methods are
not available owing to the ease with which it breaks down into
acids of lower molecular weight. The reaction with resorcinol in
presence of HC1 — a fiery red colouration — may be relied on. The
reagent is prepared by dissolving 0-5 grm. resorcinol in 60 c.c.
HC1 of 1*09 sp.gr., with which the solution to be tested is gently
warmed.

(5) Identification of mannose as phenylhydrazone. — Mannose is
easily distinguished by its property of reacting in neutral dilute
solution with phenylhydrazine (acetate) to form an insoluble hydra-
zone, which may be further identified by determining its melting
point (188°).

(6) Identification of pentaglucoses and oxycellulose. — These are
the furfural-yielding carbohydrates, and their identification and
estimation depend upon their conversion into furfural by boiling
with HC1 (ro6). In quantitative determinations the latter is esti-
mated as hydrazone.

Lignocelluloses. — Laboratory Notes. — It will have been evi-
dent to the student that the chemistry of the lignocelluloses is
that of a highly reactive molecule, and therefore is different in
a great many respects from the celluloses. The reactions of
the fibre-substance have also been dwelt upon more in detail
than in the case of cotton cellulose, and it will therefore be un-
necessary to do more than collect at this point those reartions
which are typical and characteristic, and which are useful in the
laboratory either for demonstration or in the investigation oi
unknown materials.

(i) Qualitative Reactions.— Colour reactions with phloroglucol
(solution in HC1), aniline sulphate (aqueous solution) with iodine
(absorption with development of brown colour), ferric ferri-
cyanide (deep blue dye) with magenta sulphurous acid, and with

1 The methyl hexoses are similarly oxidised to mucic acid. (E. Fischer,
BerL Ber. 1894, 385.)

Experimental and Applied 263

coal-tar dyes in great variety (simply dyeing or staining phe-
nomena).

(2) Solutions of the Lignocelluloses. — In cuprammonium, in zinc
chloride (saturated aqueous solution), and zinc chloride in HC1 ;
partial solution under the thiocarbonate reaction. This reaction
should be carefully followed up with the microscope, more
especially the extraordinary combination which takes place with
water on covering the fibre, after the reaction, with water.

(3) Hydrolytic Agents. — (a) Concentrated solutions of caustic
soda (10-25 P-ct- NaOH) in the cold. This reaction should
be studied quantitatively, and should be also followed up with a
microscopic observation of the fibre under action.

(b] Dilute alkali solutions. — These constants, quantitatively
determined in terms of loss of weight, are important. The usual
conditions are a i p.ct. solution of caustic soda in large excess ; the
specimen being boiled for 10 minutes, and a second for 60 minutes,
keeping the volume of the solution constant. The fibre is then
washed off and treated with a little dilute hydrochloric acid, again
washed, and dried.

(c} Dilute acids. — The observation of the action of boiling
dilute sulphuric acid (i p.ct. H2SO4) is of use in differentiating
one lignocellulose from another. The fibre may be boiled for 60
minutes with an excess of i p.ct. acid, keeping the volume of
the solution constant. The fibre is washed free from acid, dried,
and weighed.

(4) Cellulose Estimations. — On this subject see Cross and
Bevan, J. Chem. Soc. 1882. The following methods should be
worked :

(a) Bromine method. — The fibre is boiled in dilute alkali, washed,
and placed in saturated bromine water and left for some hours,
afterwards washed and boiled in dilute ammonia. The fibre is
then washed and returned to the bromine water, and again boiled,
after digestion, in ammonia. The treatment is repeated so long
as any residues of a yellow colour are seen in the boiling alkaline
solution. The cellulose is finally washed with dilute acid and then
with water, dried, and weighed.

(b) Chlorine method. — This involves not only a determination
of cellulose, but of the amount of chlorine disappearing in reaction,
and also the amount of hydrochloric acid (Read, Cross, and

264 Cellulose

Bevan, J. Chem. Soc. 1889, 199). Instead of the bulb there
described as' blown on the end of a tube, a simpler method
consists in blowing a thin bulb and inserting the prepared fibre,
and after insertion, drawing of the bulb. It is then placed in the
reaction flask, containing an atmosphere of chlorine. The flask or
bottle is closed with a cork well covered with paraffin, and bored
with one hole to admit the glass tube connecting with the
measuring apparatus. When the levels are all adjusted, the bulb
is broken by a blow against the side of the bottle. For further
particulars see the paper above cited, also for the estimation of
the hydrochloric acid formed. The cellulose estimation, by boiling
the chlorinated fibre with sulphite of soda solution, &c., and
washing off, has been already described (see p. 95).

(c) Nitric acid process. — The weighed fibre is placed in a flask,
and digested with 5 p.ct. nitric acid at 60°. The gaseous products
may be collected and examined. Digestion is continued until the
yellow colour which at first results gives place to white. If
arrested at an earlier stage the residues of non-cellulose may be
removed by treating with weak alkaline solution. The cellulose
is washed, treated with dilute acid, again washed, and dried for
weighing. The yield is from 60-66°, considerably less, therefore,
than by either of the above methods.

In reference to the theory of the action, show that by adding urea
the specific action of the acid is entirely arrested, and it becomes
similar to that of hydrochloric acid or sulphuric acid dilute. The
specific action also has a limit in reference to the concentration
of the acid, which must contain at least 3 p.ct. HNO3.

(d) Chromic acid process. — This process should be investigated
with chromic acid only, and with the addition of acetic acid and of
sulphuric acid. In the first case, the action is extremely slow, and
there is considerable combination of the oxide with the fibre-
substance. The presence of the hydrolysing acid causes the
specific oxidation of the non-cellulose constituents. In dilute
solution this takes place without evolution of gas. The product
may be tested from time to time by washing off a small portion,
exposing to chlorine, and afterwards plunging into sodium sulphite
solution. When this reaction ceases, arrest the experiment, wash
off the product, dry, and weigh. After weighing, compare this oxy-
cellulose mixture with the celluloses obtained by the above

Experimental and Applied 265

processes. In experiments where sulphuric acid is added as the
hydrolysing acid, the solution should be distilled and the volatile
acid estimated.

(e) Alkali oxidations.— Study the continued action of hypo-
chlorite of sodium and of hypobromite, continuing the action until
the specific reactions of the fibre disappear and a residue of
cellulose is obtained.

(/) Sulphite processes.— F 'or these the experiments must be
conducted in sealed glass tubes or in a lead-lined digester. The
fibre is sealed up with 7 or 10 times its weight of a solution of
bisulphite of lime, and heated for 8 hours, raising the temperature
gradually to 140° in the case of jute, or 160° in the case of woods.
The products are thrown out into a dish and the insoluble cellulose
filtered off, and the soluble products may be examined for the
reactions described on p. 200. If neutral sulphite is used, take a
5 p.ct. solution of crystalline salt, and in this case raise to 160°.

(5) Furfural Estimations.— The method used has been subject
to extensive investigations, as stated in the text, the details
finally adopted being those of Flint and Tollens (Landw. Vers.-
Stat. 1893, 42, 381-407). The fibre is boiled with hydro-
chloric acid, of i -i sp.gr., in a flask attached to a condenser.
The volume is kept constant by the addition of acid of this
strength to the flask. The furfural which distils is converted into
hydrazone in the neutralised solution, and estimated as such, with
careful attention to all the precautions given in the paper above
cited. The results are uniform, and the method is an important
one for the student to master.

(6) Methoxyl Determinations. — The fibre-substance is boiled
with concentrated hydriodic acid. Methyl iodide is formed, and is
carried forward by a stream of carbonic acid through a special
apparatus. The full details of the method are given in the original
paper of Zeisel (Monatshefte f. Chem. 1885, vi. 989), and is also
described in Vortmann's Anleitung z. Chem. Anal. Org. Stoffe.
(Leipzig, 1891). This method is of growing importance in the
investigation of vegetable substances, and should be thoroughly
mastered by the student.

(7) Nitration. — The ' nitrating ' acid is a mixture of equal volumes
of nitric acid (1-5 sp.gr.) and sulphuric acid (1*83), previously
mixed and cooled. The weighed quantity of the fibre-substance,

266 Cellulose

dried at 100°, is added to this mixture. The time and condition of
treatment may be varied, and the influence of the variations noted.
The product is removed from the acid and at once dropped into
a large volume of water. It is then exhaustively washed until
entirely free from soluble acid products, dried and weighed. The
products may be analysed for nitrogen by the standard methods.

(8) Ferric Ferricyanide Reaction. — The solutions of ferric
chloride and ferricyanide are prepared at normal strength, and
mixed in equal volumes previously to the experiment. The fibre-
substance is weighed and plunged into the red solution and allowed
to remain, with occasional stirring. The deposition of the blue
cyanide within the fibre-substance should be carefully observed
under the microscope, and the gain in weight should be determined
under varying conditions of digestion. The blue dyed product
may be analysed in various ways to determine Fe and N.

(9) Dyeing Operations. — The fibre may be prepared by a pre-
liminary boil in weak alkaline solutions. It should then be dyed
up in the ordinary way, with coal-tar colours typical of the various
groups. It will be found that the lignocellulose has a very varied
dyeing capability, corresponding with the great variety of re-
active groups which it contains. A systematic investigation of
this question is very much wanted.

(10) Ultimate Analysis. — The elementary analysis of fibre-sub-
stance may be carried out by any of the standard combustion
methods. In respect to the mineral constituents (ash), these must
be determined in the usual way, by completely burning the fibre
in a platinum dish, and an allowance made for the carbonic acid
retained by the ash. With regard to the moisture, the fibre may
be dried at 100-105°, and in transferring to the combustion appa-
ratus, care must be taken that no absorption of moisture takes place.

The Woods may be treated in the same way as the lignified
fibres, provided they are previously reduced to a state of the finest
possible division. For this purpose the well-seasoned wood
should be cut into shavings, with a fine plane. As stated in the
text, there is much need for a thoroughly systematic examination
of the woods comparatively with the typical lignocellulose, and there
is also room for investigating those woods which contain colouring
matters belonging to the aromatic group. In view of the more
complex constitution of the woods, care must be taken, in bringing

Experimental and Applied 267

out the results, to distinguish between the fibre-substance proper
and constituents easily removed by hydrolysis. Thus, to institute
an exact comparison of the woods with the jute fibre, both
should be taken after preliminary boiling out with dilute alkaline
solutions under the same conditions, the residue being washed
and dried after this treatment, and weighed in this condition for
the several determinations. What is required is the comparative
determination of the essential 'constants of lignification,' that
is to say, elementary composition, hydrolysis numbers, cellulose,
chlorine combining, furfural, methoxyl, nitration, and ferric ferri-
cyanide reactions.

The Pectocelluloses. — This group involves the general methods
of investigation of the carbohydrates of lower molecular weight,
for the reason that the non-cellulose constituents are easily hy-
drolysed to soluble bodies by alkalis, and are then further broken
down by acid hydrolysis to carbohydrates of definite and known
constitution. In the examination of these compound celluloses,
therefore, the methods of investigation to be found in standard
works on the carbohydrates must be followed. The general
scheme is that given in the text, p. 261 ; and as typical raw materials
the following should be studied : flax, esparto, and fleshy paren-
chyma, such as found in the turnip, apple, pear, and similar
fruits ; as bodies yielding the pectic acid series, and for the group
of mucocellulose which yield the neutral carbohydrates (hexoses
and pentoses), the investigations of Tollens (p. 223) should be
repeated.

The Adipocelluloses. — This ground is in a very undeveloped con-
dition, and much investigation will be required to establish the es-
sential chemical features of this particular compound cellulose. It
would be necessary to distinguish carefully between excreted by-
products and the essential cuticular tissue. The most promising
direction of investigation seems to be that resulting from previous
treatment of the tissue by one of the sulphite processes. The
drastic oxidations with nitric acid, and treatments with concentrated
alkalis, are too severe to enable any conclusions to be drawn with
certainty, from the products obtained, to the constitution of the
parent molecule; but a preliminary resolution into cellulose and
non-cellulose, by a method involving a minimum of change, affords
a much better basis for such investigations. As far as we know,

268 Cellulose

this has only been investigated in a preliminary way, as stated in
the text (p. 227), and an exhaustive study on these lines offers
a most promising field of research.

INVESTIGATION OF RAW FIBROUS MATERIALS.

These various processes admitting of quantitative observation
of results — the results being the constants of the individual fibres
— having been studied with raw materials of known composition,
and belonging to one or other of the groups, may be applied to
fibres of unknown composition, and to complex mixtures, in which
two or more of the compound celluloses are represented. The
investigation of fibres of unknown composition proceeds on the
following lines : (a) a general examination with reagents, to deter-
mine whether lignified or not ; (£) a general histological examina-
tion, determining its structural characteristics. A specimen is
boiled for some time in caustic soda solution (i p.ct. NaOH), then
teased out, and the ultimate fibres measured. The length of the
ultimate fibre is one of the most important criteria of value of
a fibre. Cross-sections are cut and examined, to determine the
general features of the fibre-bundle : the average number of fibres
in the bundle, the dimensions, and the divisibility of the bundle.
For a detailed account of these methods the student must read the
special treatises on the subject (p. 243).

Proceeding with the chemical examination. If the microscope
has revealed the presence of cuticular tissues, the raw material
should be extracted with ether-alcohol in a continuous extraction
apparatus, and the quantity of extract determined. The residue is
treated for the estimation of cellulose by the methods previously
described. The attendant reactions must be carefully observed.
The percentage of cellulose is the most rmportant of the chemical
constants. The quality of the cellulose should be noted. It
should be examined for resistance to further hydrolysis, and to
oxidising agents, e.g. Fehling's solution, permanganate, &c. On
these results, it is classified as of the cotton or normal group, the
jute or wood cellulose group, or the esparto and straw group. The
general character of the non-cellulose constituents will have
appeared from their reactions ; further evidence is obtained by
examining the solutions from the preliminary alkaline hydrolysis,
and from the hydrolysis following the chlorination process. The

Experimental and Applied 269

latter should be carried out with certain precautions (see p. 96),
and the chlorinated fibre may be washed, and the washings titrated,
to determine the HC1 formed.

The fibre may then be subjected to any of the special treatments,
according to the group to which it is assigned.

Fibrous raw materials of more complex constitution must be
examined with reference to external appearance, and the direc-
tions of application for which intended. If for the isolation of
a textile fibre, the material is subjected to a prolonged boiling
with sulphite of soda solution (?, p.ci. NaaSO3). The progress of
the disintegration is watched, and when the adventitious tissues
are sufficiently softened, the disintegration is aided by mechanical
means — crushing, rubbing, &c. — and the cellular debris washed
away in a stream of water. When perfectly cleaned, the proportion
of fibre obtained is estimated. In the case of bast fibres, the entire
bast may be stripped from the wood at an early stage, and the
further purification proceeded with as described.

Where the raw material is to be examined for paper-making
purposes— although it is possible to calculate with a fair amount of
accuracy, from the quantitative data obtained on the small scale,
the probable value of the material to the papermaker— it is ad-
visable to carry out an experiment on a larger scale, under conditions
similar to those which usually obtain in the papermaker's pulping
process. For this purpose special apparatus is necessary. Thus,
to make a complete examination, the material requires to be boiled
under pressure with a quantity of caustic soda calculated from the
data obtained on the small scale, and this must be carried out in a
digester capable of resisting steam pressure up to (say) about 100 Ib.
per square inch. After the digestion the pulp must be removed,
washed on a wire gauze filter, and then put through a quantitative
bleach operation. For this purpose the pulp should be well broken
up, placed in about 30 times its weight of water (estimated), and
bleaching powder solution is added, calculated to (say) 20 p.ct
on the weight of the pulp obtained, which will be approximately
known from the cellulose estimations by the laboratory methods.
When the material is bleached, an aliquot portion of the residual
liquor is drawn off, and in it the residual hypochlorite is estimated
by the usual methods. The difference gives the amount actually
consumed in bleaching the pulp. The pulp is then washed off and

270 Cellulose

treated with a little antichlor (sulphite of soda), and again washed.
It is then pressed up in suitable moulds, which may be easily made
by attaching perforated zinc to square frames of wood. Further
than this, if it is required to make an actual paper-making experi-
ment with pulp, this must be prepared by beating in a model
beater, and then converted into sheets by the hand process. The
details of such manipulation are, of course, highly technical, and for
organising such a plant and process, the assistance of an expert
would be required. Having determined the yield of cellulose by
the laboratory method, and ascertained its characteristics, the yield
by the alkali boiling and bleaching process, and the proportions
of caustic soda and of bleaching powder required for the isola-
tion of the pure cellulose, complete data are at hand for valuing
the raw material by comparison with staple materials of the same
class.

The application of these methods to the investigation of green
plants in physiological investigations, or to fodder plants, green or
otherwise, is a province in which it is difficult to lay down definite
schemes. The choice of method must depend largely upon the
subject to be investigated, and for the present — that is to say, until
our knowledge is more complete — the selection must remain more
or less arbitrary. The following general considerations will serve
as guides in selecting methods suitable for particular inquiries.
Thus in green plants it is important to distinguish between what
we may call ' permanent ' or fundamental tissue, and cellulose or
lignocellulose. The fundamental tissue might be defined as the
assemblage of cells which constitute the plant, or part of the plant,
less the cell contents, including all excreted products. Therefore,
to isolate such a complex we must proceed by way of selecting
reagents calculated to remove particular constituents, or groups ot
constituents, with the least action upon the cell wall, or cell substance
proper, of whatever kind. In investigation of the permanent tissue
of the Graminese which the authors are prosecuting, the following
process is used for its isolation :

(1) The material is exhausted with boiling alcohol.

(2) It is digested for 6 hours in cold dilute caustic soda (i p.ct.
NaOH). It is washed off from this solution, first cold and then
boiling hot.

(3) It is digested for some hours in cold dilute hydrochloric acid

Experimental and Applied 271

(i p.ct. HC1), and again washed off cold and hot. The residue
from this treatment is defined as ' permanent tissue.'

If, now, a plant or plant-substance were to be investigated con-
taining a large portion of starch, it would be necessary to precede
these hydrolytic treatments by a process acting selectively on the
starch, viz. the substances reduced to a fine state of division, boiled
for a short time with water, left to cool, and treated with malt ex-
tract, being digested for some hours at the most favourable tem-
perature for conversion. After this, which should follow the alcoholic
exhaustion, the remainder of the processes may be proceeded with
in order. (Compare V. Stein, Exper. Stat. Record, 5, 613, from
Ugeskr. f. Landmand, 39, 706.) Such a residue will contain a cer-
tain proportion of ash constituents and nitrogen, for which, in certain
cases, allowance must be made, by the usual methods of determining
and calculating. The difference between this product and that
known as * crude fibre ' (Weende method) will be noted. The im-
portant aspect of these methods and their differences is appreciated
in dealing with a complex such as the constituents which yield
furfural. Of these, the product known as wood gum (pentosan) is
soluble in dilute alkaline solutions in the cold, and would be elimi-
nated under these treatments ; but the furfural-yielding constants
are only partially eliminated by the treatment, even from non-ligni-
fied tissues. But, in the mean time, it is not safe to follow on the
older lines, which usually were held to sharply divide group from
group. We may affirm generally that no hydrolytic process can
effect any such separation, and this is particularly to be noted in re-
gard to the furfural-yielding constituents. It is advisable, therefore,
to bear in mind that any process selected is more or less arbitrary,
and gives results which, while they may be perfectly valid under
conditions of strict comparison, are not to be interpreted outside
these comparisons except with reservation. This will be specially
appreciated when the results of the proximate analyses are taken
as evidence of feeding value. This entire subject is very much in
need of revision, and we hope that both the theoretical matter and
the experimental methods described in this treatise will contain
suggestions of methods by which these problems can be more
effectively solved.

The Analysis of Textiles and Paper. — The various processes of
quantitative determination that have been described are available

272 Cellulose

for the examination of fibrous mixtures, with a view of determining
their composition. In textile fabrics this matter seldom arises,
unless in the broader distinction of the vegetable from the animal
fibres. The subject is exhaustively treated by H. Schlichter
(J. Soc. Chem. Ind. 1890, 9, 241). The vegetable fibres are
separated as a group by the process of boiling with the alkaline
hydrates (5-10 p.ct. solution), which dissolves the nitrogenous
animal fibres, leaving the vegetable fibres not, of course, un-
acted upon, but sufficiently so to enable the method to be re-
garded as a quantitative separation. The distinguishing of the
various vegetable fibres, thus separated, from one another is only
possible by microscopic observation with the additional aid of reac-
tions. No general directions can be given for investigations of this
character. They involve experience of histological methods, and
acquaintance with the characteristics of minute structure, and of the
special features which render their quantitative estimation possible
within a sufficient approximation for all practical purposes. The
examination of papers involves the identification of the vegetable
fibres only. The composition of a paper is first generally indicated
by its appearance. It is only in the mixed class of white papers
that investigation has to be carried out in minute detail. The fol-
lowing are briefly the methods adopted, so far as these are chemical :

(1) The paper is treated with aniline sulphate solution in the
cold. The presence of mechanical wood pulp is indicated by the
yellow stain produced, and the proportion approximately by the
depth of colour. Many of the ground woods, it may be remarked,
yield a pulp of sufficient whiteness to be used in what may be called
white paper, and is frequently present in ' white ' and ' toned '
printing papers.

(2) The paper is boiled with the solution of aniline sulphate.
The presence of esparto and straw ' celluloses ' is indicated by the
characteristic rose-red reaction. Papers giving no colour reaction
with aniline sulphate are probably composed of rag fibres (cotton
linen), with or without bleached wood cellulose. It is evident that
the approximate composition of a paper is thus very quickly deter-
mined by its chemical reactions ; but it is often necessary (3) to
make quantitative estimations within narrower limits, and for these
a microscopic investigation is at present the only method to be re-
commended. The examination is, of course, facilitated by taking

Experimental and Applied 273

advantage of chemical reactions to differentiate the fibres, the
method then consisting in the approximate estimation by actually
counting the fibres visible in the microscopic field according to
their identity, and averaging the results over a sufficient number of
separate examinations, and, where possible, by separate observers.
This, again, is a species of investigation which requires considerable
experience, which cannot be communicated in the form of working
directions.

(4) For actual quantitative work, of course, any of the reactions
of which full details have been given in the earlier sections of
the work are available. Thus, for instance, in white paper, found
as above to be composed of rag fibres and celluloses of the Gra-
minece only (esparto and straw), it will be evident that, as the
latter yield 12-14 P-ct. of furfural on boiling with hydrochloric
acid, and the former at the outside 0-5 p.ct., a furfural estimation
in the usual way (p. 99) would give a close approximation to
the proportions of the two groups of cellulose. In the case of
mechanical wood pulp, if the proportion is high, there are two or
three reactions available as a quantitative estimation. First, the
statistics of chlorination according to the methods described on
p. 104. Secondly, estimation of furfural ; but this is only available
in the absence of celluloses of the Graminese. Thirdly, methyl esti-
mations ; which, again, depends on the ascertained absence of other
fibres also containing this group. Fourthly, the colour reactions
with derivatives of /-phenylene-diamine, as described on p. 174.
And, lastly, the elementary analysis might even be made and taken
as a basis for calculating the proportion. These brief notes will be
sufficient to show the student how to set to work in the laboratory
to examine these particular mixtures of fibres with the help of the
reactions previously described in detail.

Principles of Cellulose Technology.

Following these notes of laboratory and general experimental
methods, we shall briefly discuss the applications of theoretical
principles and deductions to the practical processes of the arts.
The celluloses and compound celluloses are familiar to us in
multitudinous forms, both ' useful' and * ornamental' ; and the
processes by which they are manufactured, or treated for various

T

274 Cellulose

purposes after being manufactured, involve the special chemistry
of the raw materials at every turn. It must be confessed that
the arts of spinning, weaving, bleaching, and dyeing have been
highly developed upon a very slender chemical foundation so
far as regards the raw materials themselves. There is no doubt,
on the other hand, that an ample field for technological develop-
ments will be opened up by the systematic application of the
more definite chemical knowledge now available. It may in
fact be affirmed as a general principle, established also by long
and invariable experience, that there are no results of chemical
investigation, however recondite they may appear, which are not
in their due order absorbed into the province of technology.
As it is the province of the technologist to give a complete
account of his processes in terms of the factors which contri-
bute to the result, it will be very evident, from the ensuing
discussion of cellulose technology, that much remains to be
done before the industries of fibre-preparing, spinning and
paper-making, bleaching, printing and dyeing, can be said to
rest on such a basis.

Preparation of Fibres from Fibrous Raw Materials.
Processes with this object divide themselves into two groups :
(a) for the separation of spinning fibres, (<£) of paper-making
fibres. While the latter are almost exclusively chemical, the
former are as exclusively mechanical, and require therefore but
a brief general notice, (a) The spinning fibres are mostly
obtained from annual growths. With the exception of cotton,
which is a seed-hair, they form part of complex structures, and
are themselves either localised into a special tissue (bast fibres,
see Appendix, figs. 1-4), or scattered more or less irregularly
(fibre-vascular bundles of monocotyledons, see figs. 5, 6).
Structurally they are differentiated, as fibres or elongated cells,
from the cellular tissue by which they are surrounded, the com-
ponent cells of which are spherical or cubical, with more or less

Experimental and Applied 275

elongated deviations. Chemically the ' fibres ' are differentiated
by their superior resistance to the attack of hydrolytic agencies.
It is a matter of common observation that ' fleshy ' structures
are more perishable than fibrous. The chemical constitution
of the tissue-substance of this less resistant order has been only
superficially investigated ; generally the parenchyma of flower-
ing stems may be classed with the pectocelluloses. Where the
fibrous raw materials are subjected to a preliminary treatment,
with the object of facilitating the separation of the fibres from
non-fibrous tissue, it is always a process of hydrolysis, and
usually the * natural ' or spontaneous process of fermentation.

Thus flax and jute, to select the prominent types, are treated
by the process of retting or steeping. This consists in sub-
merging the stems in stagnant water ; a spontaneous fermenta-
tion is set up, with the result that the less resistant (cellular)
celluloses are disintegrated and broken down. In the case
of flax the retted ' straw ' is dried off, still containing the fibre.
This is separated by the mechanical process of breaking and
scutching. In the case of jute the bast layer is separated at
once from the retted stem, by the manual operation of stripping ;
and freed from cortex and adhering residues of parenchyma,
by beating the strips upon water. It may be stated generally
that these processes have not been systematically studied with
the view of localising the effects produced. An investigation of
the subject with the more precise methods of diagnosis now
available would be a most valuable contribution to theoretical
and industrial science.

As an illustration of the desirability of more precise in-
formation, the history of the attempts to substitute the natural
by artificial processes, in the case of flax, may be cited. Vari-
ous chemical treatments of the stem or straw have been pro-
posed, and indeed worked. The yield of fibre in this plant
being relatively high (18-23 P-ct), and the value of the fibre

T2

276 Cellulose

being also high (40!. to 6o/. per ton), such treatments are not
precluded on economic grounds. Moreover, as the natural
conditions most favourable for retting are not to be counted
on in the capricious climates of temperate regions where the
flax is chiefly grown, an artificial process admitting of exact
control is very much to be desired.

The difficulties to be overcome are not so much those of
separating the fibre as of separating it in a condition as favour-
able for spinning as the product of the natural process or pro-
cesses.

The analysis of the fibre shows that, in addition to the
pectocellulose or fibre proper, there is present an unusual
proportion of oil-wax constituents (3-4 p.ct). It appears
from later investigations of the spinning process (infra) that
the 'natural' balance of these constituents constitutes the
' optimum ' of spinning properties. All the artificial processes,
which are usually treatments with hot alkaline solutions, disturb
this balance, removing both pectic and oily constituents.
Moreover, the oils found in the * natural ' product are in part
produced in and by the retting process ; and the pectic con-
stituents of the fibre are present, not only in different propor-
tion, but in different condition chemically. As a matter of
history, these processes have failed technically — i.e. in producing
a fibre with the high spinning qualities of the ordinary product —
and with commercial results more or less disastrous. Had
investigators based their labours upon the natural model, as
defined by exact chemical investigation, such failures would
have been obviated.

But investigation is still needed to elucidate (i) the changes
produced in the oil-wax components during retting ; (2) the
effect of the retting process upon the pectic constituents of the
fibre proper. Upon the results of such investigation it might
be possible to devise an artificial process giving similar results,

Experimental and Applied 277

but the balance of conditions to be observed is necessarily
one of very fine adjustment. Any artificial treatment hitherto
attempted resembles the natural in being a process of hydro-
lysis ; the reagents to be used have been of the alkaline group,
and employed at relatively high temperature — conditions
which make it extremely difficult to limit and regulate their
action.

The authors have made investigations of the ' retting '
action of dilute solutions of sodium carbonate, silicate, and
sulphite comparatively with the soda soaps, and with the
natural process. Of these several reagents the action of the
soda soaps alone resembles that of the natural process, the
* straw ' thus treated behaving in the scutching process very
similarly to the ordinarily retted product. But the scutched
fibre is of inferior spinning quality owing to the partial removal
of the pectic and oily constituents.

Treatments of spinning fibres, after removal from the plant,
are sometimes resorted to in order to improve the working
qualities of the fibre in the mechanical processes of refining
and drawing preparatory to the actual spinning process. The
great desiderata in a yarn are uniformity and strength, and
yarns are valuable in proportion to fineness. The spinning
unit in all the vegetable fibres, with the exception of cotton
(and perhaps rhea), is a complex or bundle of the ultimate
fibres. In the processes of hackling and drawing, it is sought
to reduce or divide the bundles to the maximum of fineness.
In flax the subdivision of the bundles is carried very far and
without auxiliary treatment. In jute, on the other hand, the
bundles are much more firmly compacted ; and, as a lignocellu-
lose, it possesses none of the ' gummy ' properties of the pecto-
celluloses, and is also relatively deficient in oily constituents.
This fibre is subjected therefore to a preliminary treatment with
oily aqueous mixtures of varying composition, the incorpora-

278 Cellulose

tion of which greatly improves the spinning qualities. The
treatment and its effects are, however, rather mechanical than
chemical. Hemp has been chemically treated for the same
general purpose, by a process devised by the authors, con-
sisting in a digestion of the fibre with dilute solutions of basic
sodium sulphite at high temperatures. This enables the fibre
to be drawn and spun to finer numbers, and has proved
especially valuable in the manufacture of the finer counts of
shoe-threads.

Rhea, or China grass, is a fibrous material that also requires
chemical treatment preparatory to spinning, but for a different
reason. This fibre is separated from the mature stems by
stripping the entire bast and cortex from the wood. The
ribbons thus obtained are treated by various processes for the
removal of what is commonly termed the 'gum.' The pectic
constituents of the fibre and parenchyma readily yield to the
action of alkaline solutions, and the disintegrated cellular
residues are then easily removed by mechanical operations.

The process of purification is, in the case of this fibre,
carried to the full extent of isolating a pure cellulose. The
ultimate fibre, being of the unusual length of 40-200 mm., is a
spinning unit of sufficient dimensions, comparable with the flax
filament. Its spinning qualities are, on the other hand, inferior
to those of flax, and it is probable that much better results
would be obtained with this fibre by spinning it in a condition
more nearly that in which it occurs in the plant.

Generally it may be said that the chemical treatment of
these fibres preparatory to the mechanical operations of the
spinner has been investigated on purely empirical grounds.
There are a number of questions of both theoretical and
practical import which await systematic inquiry. The purpose
of this superficial and general discussion of the subject is to
indicate some of the directions in which the theoretical con-

Experimental and Applied 279

elusions arrived at in the earlier sections of this work may
be applied.

SPINNING PROCESSES. — The various spinning processes for
converting into yarn the fibres obtained as above described are
for the most part purely mechanical operations. They depend
in an important way upon the minute structure of the spinning
unit, whether that is an ultimate fibre (cotton, rhea cellulose)
or a complex filament (flax, hemp, jute, &c.) ; and therefore
indirectly upon the r.hemical properties of the fibre-substance.
These questions are exhaustively treated by Vetillart in his
work upon the Vegetable Textile Fibres. There is one process
only which directly involves the question of the chemical com-
position of the fibre-substance, and that is, the 'wet process '
of flax spinning The history of flax spinning shows three
periods of development : (i) At first the fibre was spun dry in
the same manner as jute is at this day. (2) It was found that
the drawing properties of the fibre were much improved by
maceration in cold water, and the wet spinning enabled the
fibre to be spun to much finer qualities of yarn. (3) A still
further advance was made by the introduction of hot water,
this treatment taking place on the spinning frame, the roving
running through a trough of water kept at 50-60°, and receiv-
ing its final drawing and twisting immediately as it emerges
from the trough. This may be considered the universal pro-
cess of spinning fine flax line yarns at the present day.

There have been numerous attempts to realise a still
further improvement by alkaline treatments of the most varied
kind, either on the spinning frame itself — i.e. by adding the
alkaline reagent to the ' spinning trough ' — or by a previous
treatment of the roving. Such processes, however, have only
come into limited use. A more successful attempt to still
further raise the spinning qualities of flax is that of C. C.
Connor, of Belfast, who patented in 1888 a process based upon

280 Cellulose

the results of the authors' investigations of the constituents of
the fibre. From these results it appeared that a considerable
proportion of the oily constituents were of a ketonic character,
and were readily emulsified by treatment with solutions of such
salts as sulphite and phosphate of soda.

The addition of such salts to the ' spinning trough ' might
be expected to bring about a much more perfect distribution
of the oil-wax components throughout the fibre-substance than
is possible by treatment with hot water. The pectic con-
stituents, also being further attacked than by hot water,
might be expected to be brought into a more favourable
condition for yielding to the drawing action of the frame.
Experience has verified these predictions, and the working of
the process on the large scale has shown conclusively that in
the coarser Russian flaxes there existed an undeveloped margin
of spinning quality which is fully realised under the new
process. As it has also been shown that the weight of yarn
spun from a given weight of roving is n ^t sensibly different
from that obtained by the ordinary hot-water process, it is
evident that these alkaline salts, under the conditions adopted,
do not exert any undue solvent action on the constituents of
the fibre, but are limited in their action to bringing about the
optimum condition for drawing and subdividing the fibre-
bundles. The salts are used in the process in the form of 1-2
p.ct. solution, the proportion being adjusted to the quality of
the flax.

(ft) The paper-making fibres are obtained from very various
sources, largely from the rejections of the spinning and weaving
industries (scutching tow and waste, jute butts, spinning
wastes, rags and cuttings of all kinds). In addition to these
there are a number of vegetable raw materials which are
treated directly for conversion into fibre or pulp— e.g. esparto,
straw, wood.

Experimental and Applied 281

The processes of treating wood have already been discussed,
as they admit of classification on strictly theoretical lines, and
afford a useful illustration of the general principles of the
relation of the cellulose to the non-cellulose constituents of the
compound celluloses. In extending this classification to the
wider range of raw materials above indicated, it is necessary to
remember the general features of the three groups of compound
celluloses, and more particularly the conditions under which
they are resolved into cellulose and non-cellulose, observing
also that any process of resolution to be available for the
purpose in question must be limited in its attack as much as
possible to the non-cellulose components. The following may
be laid down as a broad principle of economy in such treat-
ments : effects required to be produced should be separately
accomplished, and obtained by specific reagents. It must be
conceded at once that this is an ideal seldom realisable. The
treatments of the papermaker are nearly always 'overhead'
treatments, in which one process and one reagent is employed
to work a very complex mixture of chemical decompositions.
But because practice tends to stereotype itself on the lines of
apparent simplicity, it is not for the chemist to accept this
order of things as unassailable. Experience has shown, and is
continually showing, that * division of labour ' in reactions is as
economical as it is in other branches of work ; and it is a
particular purpose of this discussion to suggest a careful re-
vision of these * overhead ' treatments, with the view of improv-
ing methods wherever possible.

The pectocelluloses from our present point of view need no
discussion. They are easily resolved by alkaline hydrolysis of
the simplest kind, i.e. boiling at the ordinary boiling temperature
with solutions of the alkalis. The resulting cellulose would be
approximately pure and structurally disintegrated, i.e. in the
condition of ultimate fibres.

282 Cellulose

The lignocelluloses present problems of a totally different
character. The more resistant members of the group — viz. the
woods — are ' pulped ' by various processes which have been
already described. Some of these depend upon a specific
attack of the non-cellulose constituents, whether by way of
synthesis with the reagents employed (sulphite processes) or
radical decomposition (nitric acid process) ; others may be
rather described as ' overhead ' treatments, in which a highly
complex series of chemical changes are determined which are
by no means confined to the non-cellulose constituents, but
affect the cellulose also, and prejudicially in regard to yield.
Such are the alkali processes. The typical lignocellulose jute
stands on a different footing from the woods. The latter are
used either (i) as ' mechanical wood pulp,' obtained by merely
grinding the wood ; (2) as ' chemical pulp,' which is a more or
less pure wood-cellulose, obtained by the processes previously
described. Jute, on the other hand, is largely, in fact chiefly,
used as a disintegrated and purified lignocellulose, ' pulped ' by
a process which leaves the cellulose and non-cellulose still in
intimate combination. The process giving this intermediate
product is that of boiling with lime at relatively low tempera-
tures (105-115°). It is applied to the rejected root ends
(* cuttings '), which contain also ' pectic ' (incrusting) consti-
tuents and residues of cortical and past parenchyma. These
are resolved by the treatment, and the fibre-bundles of the
lignocellulose proper are largely disintegrated, a certain propor-
tion being also hydrolysed and dissolved. The product (pulp)
is therefore a purified lignocellulose, in a condition easily
yielding to the subsequent mechanical operation of beating.

Jute may be treated for the isolation of a jute cellulose by
any of the processes described for the woods. A process also
used to some extent is that of chlorination, the fibre being first
prepared by boiling in a weak alkali, and, after washing,

Experimental and Applied 283

exposing in closed chambers to an atmosphere of chlorine gas,
afterwards removing the chlorinated product by again boiling
in alkali. This process is the laboratory method of isolating
cellulose applied on the large scale. In the form of cellulose,
however, jute comes into unfavourable competition with the
woods, and the process is therefore not much used.

The adipocelluloses come into consideration in this connec-
tion, merely as adventitious tissue-constituents. They are a
source of considerable difficulty on account of their resistance
to the attack of reagents. They occur, of course, chiefly in such
raw materials as are entire stems or leaves, e.g. esparto and
straw, and are characterised by admixture with chlorophyll
(esparto) and oil-wax constituents. For the removal of the latter
the alkaline treatment is relied upon, the conditions of which
for the purpose require to be much more severe than for the
non-cellulose of the fibre-constituents proper. The cuticular
cells themselves are only slightly attacked by the process, and
are obtained in the pulp, in which they are easily recognised
under the microscope by their very characteristic form. The
neutral waxes are obtained at the end of the boiling process in
mechanical mixture with the mass of pulp and liquor, and they
collect on the surface of the lixiviating vats used in the continu-
ous process of washing esparto pulp. The treatment of such
raw materials, in which all the compound celluloses are repre-
sented, is perhaps the best illustration of what has been expressed
by an ' overhead ' treatment. The ordinary processes are, in
fact, a crude aggregate effect, and it is more than probable that
means may yet be devised for more specific treatments, in
harmony with the broad principle of economic chemical work.
In all these industrial processes for isolating cellulose, moreover,
the yield is considerably less than what may be considered the
theoretical, i.e. the proportion isolated by the method of chlori-
nation (p. 95). For these raw materials which have been more

284 Cellulose

especially mentioned, the following may be taken as the com-
parative yields :

Laboratory method.
(Yield of dry cellulose on dry raw

material.)

Esparto . 50-55 p.ct.
Straw . . 50-55 „
Wood . . 50-55 „

Papermakers methods.
(Yield of air-dry pulp on air-dry

raw material.)

43-47 P-ct. Alkali process
33-37 „ „ „

35~43 » 5> }»

42-48 ,, Bisulphite „

It is obvious, therefore, that the cellulosic constituents of
the fibres are considerably attacked, and that there is an ample
margin for improved results in regard to quantity as well as
quality of the fibre produced (pulp).

BLEACHING PROCESSES. — These processes appear to divide
themselves into the two groups : (a) the bleaching of textiles ;
(b) of paper pulp. It will be evident, however, from the present
treatment of the subject, that bleaching is a process of purify-
ing a cellulose or compound cellulose from adventitious con-
stituents, whether mechanically mixed with the tissue or fabric,
or chemically united to the ultimate fibre-cellulose ; and on this
view of the subject bleaching treatments divide themselves
into (i) processes for the purification of 'a compound 'cellulose, with
removal of colouring (or discolouring) matters (jute textiles ;
flax yarn, ' creaming ' and half bleaching process ; linen textiles,
part bleaching ; pulps for wrapping and coloured papers) ; (ii)
processes for the isolation of a pure cellulose (cotton textiles,
linen textiles, papermaker's cellulose).

The bleaching process proper is the whitening or decolourising
process which follows such alkaline treatment as those already
described. The bleaching is invariably a treatment with oxi-
dising agents, usually alkaline ; ' bleaching powder ' or calcium
hypochlorite is the 'staple' reagent. Other hypochlorites
(sodium and magnesium), obtained by double decomposition
from the former, are largely used, and oxidising solutions obtained

Experimental and Applied 285

by the electrolysis of solutions of the chlorides (chiefly MgCl2)
are also now extensively used (Hermite process).

Chemically, therefore, the bleaching processes of the arts
consist essentially of the two treatments : (i) alkaline hydrolysis
followed by (2) alkaline oxidations.

In the processes of the first group the alkaline treatments
are of the milder order, the purpose being to dissolve and
remove the minimum of non-cellulose constituents, consistently
with obtaining a uniform and sufficiently high colour (bleach)
in the finished product. As therefore a large proportion of the
more oxidisable (non-cellulose) constituents is retained in the
pulp or fabric, the consumption of the bleaching agent in the
after process is relatively high. It is in fact used up, not in
selectively oxidising those constituents which are the colouring
matters of the alkali-boiled fibre or fabric, but obviously in
a general oxidation of the non-cellulose constituents in the
order of oxidability.

Two processes may be considered as typical of this group :

(i) Jute fabrics and jute pulp. — Jute itself may be whitened
considerably by regulated oxidations. In the case of this fibre,
however, it is difficult to control the action of bleaching
powder. The avidity of the lignocellulose for chlorine is such
that should any free hypochlorous acid be formed in the solu-
tion, chlorination of the fibre immediately results. The pre-
sence of the lignone chlorides in the fibre is a source of con-
siderable danger. Being unstable they are gradually decom-
posed, with liberation of hydrochloric acid, which rapidly
disintegrates the fabric. The neglect of this property of the
lignocellulose has led to disastrous consequences in manufac-
ture. An industry established some years ago for the bleaching
and printing of jute cloth was ruined through the wholesale
' tendering' of the goods from this cause. The process
adopted consisted in (a) boiling in weak alkaline solutions

286 Cellulose

(carbonate and silicate of soda), (b) bleaching with calcium
hypochlorite solution in a closed vessel (Mason Kier) ; after
which the cloth was washed, ' soured ' in weak acid, washed up,
and dried. The printing processes were those ordinarily employed
for cotton goods, the colours being developed and fixed by the
usual process of steaming, in an atmosphere of dry stream at
4 Ib. (per square inch) pressure. It was in the latter case that the
discolouration and tendering effect chiefly showed themselves.
The cause being traced, the remedy was easily devised, the
process being modified as follows : (a) in the bleaching
process, sodium hypochlorite was substituted for the lime
compound — and in this way the chlorination of the fibre-sub-
stance was arrested ; (b) as a last treatment, after souring and
washing, the goods were run through a solution of sodium
bisulphite (i p.ct. SO2), and dried after squeezing. In this
way a residue of the normal sulphite (Na2SO3) was left in the
cloth, and this was found to prevent discolouration in the
steaming process.

In this method of bleaching, the loss of weight of the fabric
was from 8-12 p.ct, the colour obtained being the pale
cream shade of the highly purified lignocellulose. The results
obtained by bleaching with permanganates are superior to those
with the hypochlorites, but at much greater cost. The process
is therefore but little used industrially.

(2) Linen yarn and doth: partial bleach. — In the linen
industry, in addition to the full bleaching of shirtings, sheetings,
cambrics, &c., there is a large practice in partial bleaching of
various grades. These processes are familiarly designated as
* whitewashing,' in contradistinction to the * bottom bleaching' :
in the former the non-cellulose constituents are only partially
removed, and the residues whitened by bleaching agents ; in
the latter they are entirely eliminated, leaving the residue of
pure flax cellulose. The partial bleaches in question are

Experimental and Applied 287

obtained by a light alkaline boil, followed by a treatment with
bleaching liquor (hypochlorite), these treatments being once or
twice repeated for higher grades of bleaching. The consump-
tion of bleaching powder is relatively large (10-30 p.ct. of the
weight of the goods), a considerable proportion being used up
in oxidations which do not contribute to the bleaching effect
proper. The processes are therefore not economical in the
strict sense of the term, and are capable of considerable
improvement in the direction of a more specific attack of the
coloured constituents of the yarn. In various grades of paper
making, also, similar half-bleaches are practised.

Jute (cuttings and waste) is boiled in lime and bleached
with bleaching powder solution, the resulting pulp being of a
yellow to a yellowish-white colour, still retaining a large pro-
portion of the non-cellulose constituents of the original fibre,
and giving all its characteristic reactions. Flax wastes (scutch-
ing tow) are boiled with lime or soda to soften and disintegrate
the residues of wood (sprit), and the pulp is bleached with
hypochlorites.

The principle of these treatments is, however, one and the
same for all, and is sufficiently illustrated by the examples
discussed.

(b) The second group of bleaching processes, of which the
goal is a pure cellulose (or oxycellulose), differ from the above
in this general and important particular : the chemical work is
thrown chiefly on the alkaline boiling processes, the bleaching
treatment proper being limited to the oxidation of the coloured
residues from these treatments. Thus in cotton bleaching, while
the consumption of caustic soda may be taken at 80-100 Ib.
per ton of cotton goods, the bleaching powder required is
less than 30 Ib. per ton, a proportion of which is wasted in the
unavoidable losses attending the washing away of residual
liquors. In both cotton and linen bleaching of this order,

288 Cellulose

moreover, the bleaching solutions are used in a highly dilute
form (0-5-2-0 p.ct. bleaching powder). In papermakers'
cellulose bleaches, while it is true that by far the greater pro-
portion of the chemical work of purification is thrown upon the
pulping process, the consumption of bleaching powder in the
bleaching process proper is in some cases considerable. In
the bleaching of rag pulp (cotton and linen) the average con-
sumption is from 2-5 p.ct. ; in straw and esparto pulp, 10-15
p.ct ; and sulphite wood pulp, 15-25 p.ct. In these latter cases
we have a further illustration of ' overhead ' treatments — i.e. in
order to produce a certain result in a given time and a single
process, a large amount of waste energy is expended. These
celluloses are, as we have already seen, very different constitu-
tionally from the normal type : they are easily hydrolysed, and
in the alkaline bleach liquor a considerable further proportion
of the fibre-constituents are dissolved and undergo oxidation of
a perfectly useless character. To minimise these wastes of the
oxidising agent, the practice of intermediate washing is some-
times resorted to ; and by thus separating the effects of hydro-
lysis and oxidation, the latter is controlled into the directions of
useful, i.e. bleaching oxidations. The economy of bleaching
powder which results is very considerable, and it is not a little
remarkable that so rational a plan is not more generally
adopted.1

Of the textile bleaches of this group there are two which
may be selected to illustrate general principles, viz. the cotton
bleach and the linen full bleach.

In COTTON-CLOTH BLEACHING the most important process
is the alkali boil. The treatment is varied to suit the great
variety of goods which undergo the process, but for our present

1 A very thorough treatment of papermakers' bleaching processes will
be found in Griffin and Little's « Chemistry of Paper Making '(1894),
chap. v. pp. 275-300.

Experimental and Applied 289

purpose we need consider but the one in which caustic soda
is used. With this reagent, in the form of a 1-2 p.ct.
solution of NaOH, cotton goods are effectively cleared of their
non-cellulose impurities in a single treatment. The conditions
of the process are : (i) a saturation of the goods with the
alkaline lye, usually effected by passing the goods in continuous
length through the hot liquor, removing the excess by squeezing,
and piling up in the * kier,' or boiling-vessel ; (2) the boiling
process, in which the goods are subjected to the further action
of the alkaline lye at temperatures of 105-115°, and under
corresponding steam pressures. The liquor is kept in circula-
tion through the goods, and the ' boiling ' is continued from
six to ten hours.

After this treatment the goods are washed free from the
alkaline lye and the dark coloured soluble products of the
action, and are then of a greyish-brown colour. The residual
impurities are then removed in the bleaching process proper,
which consists in exposing the goods to the action of bleaching
powder solution. The goods are then washed and ' soured,'
to remove basic residues. This round of operations is some-
times repeated, though with weaker solutions, in the case of
heavy goods, or of goods made of the more refractory
Egyptian cottons, which contain a red brown colouring matter.
The process, however, need not be followed into its technical
details. It is one of great simplicity, and aptly illustrates the
resistance of the normal cellulose to alkaline hydrolysis and
oxidation under somewhat severe conditions. The fibre itself
loses from 7-10 p.ct. in weight under the treatment. The
products removed in solution have been investigated by
Dr. E. Schunck, who resolved the dissolved products into
(a) Cotton wax, a neutral wax, melting at 80-86°, having the
composition C 80-3, H 14-4 ; (b) Fat acid, which appeared to
be a mixture of palmitic and stearic acids, The analytical

u

290 Cellulose

numbers obtained were C 75-5, H 13-0. (c) Pectlc acid, &
gelatinous acid body, having the composition and properties of
the acid described by Fremy. (d) Two colouring matters —
(i) soluble in alcohol, (2) insoluble — having the following
composition :

C 58-48 577

H 5-80 6-05

N 5-30 874

(Mem. Lit. and Phil. Soc. Manchester, [3] 4.)

In addition to these substances, which are constituents of
the fibre proper — including residues of cell-contents — the
alkaline treatment breaks down the residues of the seed
envelopes (motes) which survive the mechanical operations of
preparing, and find their way into the yarn. The proportion
of these by weight is, however, relatively insignificant, though
they are a source of some difficulty to the bleacher.

There can be little doubt that the cotton cellulose under-
goes certain molecular changes during the process of a normal
bleach. From what we know of its constitution and reactions
we may affirm that it does not remain inert under treatments
of this severity ; but our methods are not sufficiently refined for
differentiating the product from the cellulose as contained in
the raw cotton. There is perhaps one exception to be noted,
which is, that the bleached cotton yields from o*2-o'6 p.ct. of
furfural on boiling with hydrochloric acid, which may be taken
as an indication of the presence of a small proportion of
oxycellulose.

As already pointed out, cotton is very easily oxidised to
oxycellulose under the joint action of calcium hypochlorite (in
dilute solution) and carbonic acid. The researches of Witz,
who established the general conditions of these oxidations,
were carried out at a date (1882-85) wnen tnere were none but

Experimental and Applied 291

qualitative reactions (dyeing phenomena &c.) available for
demonstrating the formation of oxidation products. As it is
probable that condensation to furfural is a property of these
oxycelluloses, and the estimation of this product is reduced to
a method of precision, it would be important to investigate
the cotton in three stages, viz. : (i) in the raw state ; (2) after
alkaline treatments of varying degrees ; and (3) after bleaching
processes of various kinds and degrees, for the presence of
furfural-yielding constituents and their quantity.

The classification of cotton-cloth bleaches into ' market
bleach,' * madder bleach/ &c., involves no important question
of principle ; and for description in detail of the variations of
treatment practised in the several grades, the technological text-
books must be consulted. We would specially mention, in pass-
ing, the article on * Bleaching,' in Watts' Dictionary (Applied
Chemistry, new edition), which gives an excellent survey of the
history of development of the art. It may very well be assumed
by those familar with this history that we have arrived at terminal
excellence in the art. From the economical point of view it
is, perhaps, difficult to see any unexplored margin. But, on the
other hand, there is evidence of important recent progress in a
direction of improvement, which will be evident from the
following considerations. A web or fabric of cotton must be
always considered by the technologist from the point of view
of minute structure, the structure being that of the ultimate
fibre, complicated by the spinning twist and the interlocking of
the yarns in the weaving. The penetration of cotton goods in
the mass by liquid reagents is obviously a highly complicated
process. In the first place, complete penetration is probably
possible only by previous exhaustion of the air contained in
the tubes ; and, secondly, penetration of the substance of the
cell wall must involve osmotic phenomena. Osmosis is compli-
cated in two directions : first, by the filtering-out of the active

u 2

292 Cellulose

reagent employed in the treatment ; and, secondly, by physical
changes in the cotton itself or its non-cellulose constituents.
In the alkaline treatments of cotton it is of importance that
the action of the alkali, water, and heat should be as nearly
as possible equal and simultaneous throughout the mass. The
advance of the caustic alkali process over the successive treat-
ments with lime and soda ash of the older methods consists
chiefly in this, that by the more rapid action of the more
powerful alkali, secondary changes of the more oxidisable non-
cellulose constituents are reduced to a minimum ; and these
are dissolved away by a single operation, with a minimum
residue of products to be removed in the bleaching process
proper. In the ordinary processes of bleaching, the result
attained is simply measured by the appearance of the cloth.
The printer, however, requires something more than a good
white. The operations of calico printing in many cases involve
a dyeing process, not of the whole cloth, but of the design or
pattern printed with suitable mordants, the cloth itself being
required to resist the colouring matter of the dye bath. Many
* market bleaches ' are therefore very inferior in point of purity
of the cellulose to the * madder bleach ' of the printer, and
will dye up with alizarin and similar colouring matters, which
the latter will resist under the same conditions. It is in regard
to this important distinction, and the further refinement of the
bleaching process for the * madder bleach,' that progress con-
tinues to be made.

Linen bleaching. — The full bleach of flax goods, which
consists in the isolation of the pure cellulose, is a much more
complicated process than the bleaching of cotton-cloth, though
based upon identical principles, and involving for the most part
precisely similar methods.

The proportion of non-cellulose constituents in flax is very
high, varying from 20-35 P-ct- °f the weight of the fibre,

Experimental and Applied 293

according to the conditions of growth and the methods of
separating and preparing the fibre. The greater proportion,
being pectose-like substances, are easily attacked by alkaline
hydrolysis ; but the removal of a large weight of such products
from a mass of cloth is not an easy operation. The alkaline
treatments are therefore graduated, and are three or even four
times repeated before the cloth is considered ready for the
bleaching treatment. There are then the additional complica-
tions of the wood residue (sprit) and cuticular constituents
which very much protract the after processes, or bleaching
proper. These processes may be divided into series, the first
of each series being the process of treating with dilute solu-
tions of the hypochlorites. These involve prolonged exposures
(6-12 hours), the cloth being entirely submerged in the solution.
After this follows usually the souring process, and to this
succeeds a light boil in progressively weaker alkaline solutions.
These treatments, with intermediate washings, constitute the
* round.' After each round, the cloth, or rather the residue of
non-cellulose constituents, is in the most favourable condition
for the further attack of the oxidising or bleaching agent.
These processes are repeated until the impurities are finally
eliminated. In addition to these treatments, which are those
practised by the cotton bleacher, linen undergoes the process
of 'grassing/ i.e. is spread out upon grassfields and exposed
for one or two days to the action of light and air and the other
influences of the ' weather.' This process follows an alkaline
treatment of the cloth, whether in the earlier or later stages,
when the cloth is in the most favourable condition for the
action of the atmospheric oxygen. The linen is also treated
by a special process of mechanical rubbing with a strong soap
solution.

The complications of the process are such that the full linen
bleach takes from three to six weeks to accomplish. They

294 Cellulose

are due to the highly resistant character of the cuticular
tissues and by-products which are associated with these
tissues, or formed during the process of breaking them
down ; and, in lesser degree, to the wood residues. Both of
these have to be entirely eliminated without injury to the
cellulose.

The process is therefore a complete illustration of the gene-
ral chemistry of the compound celluloses, and the order of
their resistance to hydrolysis and oxidation, i.e. to the chief
destructive influences of the natural world.

It cannot be said that the process has been subjected to
exhaustive chemical investigation, such as would reveal the
steps by which the various non-cellulose impurities are broken
down. From the more theoretical account of these constitu-
ents in the earlier sections of this book we may, however, form
a tolerably correct estimate of the progress of the breaking-
down process. But at the same time a full investigation by
chemical and microscopic methods is much more to be desired,
and could not fail to throw considerable light upon the
important industrial problems involved.

DYEING AND PRINTING PROCESSES. — It appears, a priori,
that these processes of colouring the textile fibres are the result
of interaction of colouring matter and fibre-substance as a
definitely molecular phenomenon ; and the progress of in-
vestigation is confirming this view more and more. At this
stage, however, the * theory of dyeing ' is still the subject of
active controversy, and a decisive statement must therefore be
avoided. The discussion ranges itself round the two opposed
views of dyeing : (i) as a mechanical, and (2) as a chemical
process. At the present time, however, these terms have lost
much of the significance attached to them in the early days of
the controversy. In those days * solution ' itself was regarded
as a 'mechanical* or 'physical,' in contradistinction to a

Experimental and Applied 29$

' chemical,' process. As, however, the c constants of solution,'
i.e. the properties of bodies in solution, are now definitely corre-
lated with molecular weight, the distinctions obviously vanish
in this case, and the corresponding terms are absorbed in that
of more comprehensive significance — viz. ' molecular.' So also
it may fairly be stated in connection with the phenomena of
dyeing. If solution is defined as the homogeneous distribution
of one substance through the mass of another regarded as the
solvent, the dyeing process is a special case of transference of a
body from one solvent to another, and a dyed fibre is a solid
solution of the colouring matter in the fibre-substance. The
conditions determining the transfer, in the process, from water
to fibre-substance are certainly complex : they depend (i) upon
the constitutional relationships of fibre-substance and colouring
matter ; (2) upon osmosis and all those conditions by which it
is influenced.

In regard to the first and chief factor, a very superficial
view of dyeing processes points to the important influence of
the chemical properties of the fibre-substance. But, in
extending this view to a detailed discussion, we are met at once
by the great disparity between these two groups of carbon
compounds, i.e. fibre-substances and colouring matters, in
their relationship to the science. The latter are, as a class,
bodies of the most definitely ascertained constitution, and are
synthesised, in many cases, by 'quantitative' reactions from their
constituent groups ; whereas the constitution of the former is
still highly problematical in every direction. A comprehensive
view of dyeing phenomena is necessarily, therefore, deferred
until the latter group shall have been more fully investigated.
At the same time, we have positive knowledge of the reactive
groups of the fibre-substances, sufficient to indicate the part
which they play in dyeing phenomena ; and these reactions
have already been discussed, in the case of the celluloses, as a

296 Cellulose

species of double-salt formation. On the more general view
of dyeing, this is in fact a well-grounded hypothesis, viz. that
as the colouring matters available for dyeing show invariably
a * saline' constitution, and the formation of * lakes' with
inorganic bodies is due to reaction with salt-forming groups —
as also the fibre-substances in reaction show a similar differen-
tiation into acid and basic groups — the interaction of the two
groups of compounds in the dyeing process is, on the more
general view, a special case of double-salt formation. But even
should this hypothesis be found to afford a consistent general-
isation of the whole range of dyeing phenomena, it carries us
only a certain length as a theory of dyeing. We have next
to deal with the selective relationships of the two groups of
carbon compounds, i.e. the particular ' colouring affinities ' of
the soluble colouring matters or dye-stuffs. Speaking gene-
rally, for instance, the celluloses are resistant to such solutions ;
the number of dye-stuffs giving a direct dye on cotton is ex-
tremely limited. In striking contrast to the celluloses, on the
other hand, the lignocelluloses are distinguished by ' cosmopo-
litan ' relationships, resembling the animal fibres wool and silk,
in being dyed directly with a wide and varied range of colouring
matters. This at once suggests that the essential factors
of the dyeing process are molecular and constitutional, i.e.
chemical, in the narrow sense of the term, rather than structural ;
and this conclusion is strongly emphasised by everything which
has preceded this discussion in regard to the constitution of
these typical groups of fibre-constituents. Further, by chemical
modification of the celluloses, their dyeing capabilities are con-
siderably modified ; thus the oxycelluloses were shown by Witz
to exhibit not merely an increased attraction for colouring
matters of the 'basic' class, but a diminished attraction for
those of the class more acid in character and generally requir-
ing to be dyed with mordants. Of these two groups the fol-

Experimental and Applied 297

lowing were cited by Witz as typical. The oxycelluloses show

An increased attraction for
Methylene blue
Hofmann violet
Malachite green
Safranine red
Fuchsine red
Bismarck brown

A diminished attraction for

Diphenylamine blue, sulphuric

acid

Induline blue, sulphuric acid
Indigo sulphonate
Tropseoline orange
Eosine red

in comparison with the cellulose. (See Bull. Soc. Ind. Rouen,
[10] 5, 416 ; [n] 2, 169 ; Dingl. J. 250, 271 ; 259, 97 ; J.
Soc. Chem. Ind. 1884.)

Here also structural factors are eliminated, and the vari-
ables are again constitutional.

Selective attractions of more narrowly specific character are
exhibited, on the other hand, by both the celluloses and ligno-
celluloses, of which typical instances may be discussed.

Thus, in the case of the celluloses, modern discovery has
added to the coal-tar dyes a number of compounds which dye
cotton directly to full shades, and are therefore known as
cotton colours. Although, however, these are synthetic pro-
ducts, and therefore bodies of known constitution, no general
constitutional relationship of these compounds has yet been
established such as to account for their ' specific affinities ' to
the celluloses. This, of course, complicates the phenomena, and
shows that other factors, in addition to those of constitution as
ordinarily understood, contribute to the result. Of such we
may instance as probably operative the molecular condition of
the colouring matter in aqueous solution.

Of all the colouring matters having this particular relation-
ship to the celluloses, the most noteworthy is the dye-stuff
known by the trivial name 'primuline,' a complicated colour-base
derived from thiotoluidine. The sulphonic acid of this highly
' condensed ' product combines freely with cellulose when the
latter is treated with its dilute aqueous solution as in ordinary

298 Cellulose

dyeing process. The combination is of so stable a nature that
the base may be diazotised upon the fibre without loss, and
then may be further synthesised with chromogenic phenols
and bases to form a range of dyes of varying shades. Such 'in-
grain' colours constitute an important theoretical and practical
advance, and their production by synthetical processes upon the
cellulose itself is a further proof that the bond of union of dye-
stuff to fibre-substance is ' chemical ' as ordinarily understood.

Another application of these peculiar relationships of dye-
stuff to fibre results from the observation that the diazoprimu-
line upon the cellulose is in a highly photo-sensitive condition,
a brief exposure to sunlight sufficing to decompose it with
evolution of (gaseous) nitrogen. From this observation has
resulted the diazotype process of 'positive' photographic
printing (Green, Cross and Bevan, Berl. Ber. 23, 3131).

The important feature of this process, from the point of
view of the present discussion, is the sensitiveness of the diazo
derivative when prepared upon the cellulose basis, compared
with its relative stability in the free state. The most reasonable
explanation of this increased sensitiveness appears to be that
the product exists in the cellulose in a condition of solution-
dissociation, a solid solution of the product in the colloid cellu-
lose having the essential characteristics of solutions in liquid
solvents. According to this view, the diazoprimuline, being
molecularly disaggregated, is in a more ' responsive ' condition
to the decomposing action of the light-energy ; and hence the
decomposition. It is no purpose of this discussion, however,
to advocate any particular views, but merely to introduce the
various aspects from which this in many respects unique dyeing
process of the celluloses may be regarded, and to point out
that judgment as to the underlying causes must for the present
continue to be suspended.

The lignocelluloses afford a still more characteristic dyeing

Experimental and Applied 299

reaction in their property of taking up the blue cyanides from
solutions of ferric ferricyanide. It is not a question here of a
merely superficial oxidation of the fibre-substance by the ferri-
cyanide, and a staining of the fibre with the resulting blue
cyanide. From the detailed description previously given (p. 1 24)
it is seen to be a specific reaction between the fibre-substance
and the ferricyanide, taking place in altogether unique quan-
titative proportions. It does not depend upon any anterior
reduction by the fibre-substance, as it is unaffected by the
presence of powerful oxidising agents ; nor upon the relation-
ships to the fibre-substance of either ferric oxide or hydroferri-
cyanic acid, since in any other form of combination they exert
but slight action. From the evidence, it appears probable that
the lignocellulose takes up the ferric ferricyanide as a whole, in
the first instance — such combination having rather the- features
of a ' physical ' reaction — and then redistributes its constituent
groups in such a way that the ferric oxide is deoxidised with
formation of the blue ferroso-ferric cyanide. In this second
effect the constitution of the characteristic groups of the ligno-
cellulose is the active cause.

These two reactions or groups of dyeing phenomena have
been instanced, not only because they are of critical and unique
value as test-problems for any theory of dyeing, but as further
illustrating the varied aspects of the subject of cellulose
chemistry. With progress in the theory of dyeing, it is highly
probable that the effects themselves may come to be available
as criteria of constitution of the fibre-substances ; in the mean
time it is equally probable that further elucidation of these
problems in other directions may contribute materially to the
establishment of a theory more generally acceptable than the
much controverted views at present held.

In the processes of printing the vegetable textile fabrics the
same general considerations obtain. The treatments are, how-

300 Cellulose

ever, much more diversified ; and their scientific basis, so far as
regards the chemical function of the fibre-substance as an active
cause, is even less elucidated than in the more simple operations
of dyeing. In the absence of any specific contributions of
investigators, no attempt can be made to deal with so wide a
range of effects. With a wider knowledge of the chemical
functions of the constituent groups of the fibre-substances, it
will be easy to devise critical experiments in solution of the
very various problems presented.

The industrial uses of the celluloses and compound cellu-
loses are of wide and varied range. They depend, of course,
largely upon the external and physical properties of the natural
products : but if less obviously, certainly in a not less important
degree upon the special chemistry of these substances. Their
industrial value again depends upon the conditions of supply,
the agricultural questions of yield, and the economic questions
of production and preparation in a fit state for the further manu-
facturing operations by which they are finally shaped for use.

In the province of textile fibres this threefold qualification
constitutes an effectual limitation of the number available, and
the numerous abortive attempts to exploit others of the end-
less variety of vegetable fibres have invariably followed from
neglect of one or other of the essential conditions of qualifica-
tion. These qualifications are in effect the constants of the
fibres, all expressible in numbers, the results of measurements
or observations of quantitative relationships. Thus, to select
in illustration the flax fibre, the following are the * constants *
which mainly determine its value :

Agricultural (constants of 1 Yield of * straw ' per acre,
raw material) . . . / Yield of fibre on 'straw.'
Morphological or structural 1 L of uhimate fibre.

(physical constants of fibres) J

r Proportion of cellulose and resist-

Chemical . . . .4 ance of cellulose to hydrolysis
I and oxidation.

Experimental and Applied 301

There are many considerations of subsidiary importance :
thus, on the agricultural side, the habit of the plant and cost
of cultivation ; on the mechanical or structural side, the sepa-
ration of the fibres from the stem, the uniformity, fineness, and
divisibility of the fibre-bundles ; and on the chemical side,
the relationship of the cellulose to the non-cellulose constitu-
ents both adventitious (wood and cuticle) and essential (the
pectic constituents of the fibre proper). A careful considera-
tion of these quantities or properties as factors of value will
almost tempt the reader, if of a mathematical turn of mind, to
propose a numerical expression of value somewhat as follows :

Taking V = value (in the sense of utility),
Y = yield of fibre per acre, ,<n >
L = length of ultimate fibre,
P = percentage of cellulose in fibre,

then V = c. YLP (c being a constant).
The factors Y, L, P would require to be qualified by the
introduction of the subsidiary factors ; and although these are
not expressible in so definite a form, they can be brought to a
sufficiently exact approximation. It is not the purpose of this
inquiry, however, to attempt a complicated special discussion
involving considerations outside our general plan of treatment.
With this general suggestion of the relationships of our sub-
ject, taken as a whole, to industry, we revert to the considera-
tion of the purely chemical problems presented by the celluloses
and allied compounds in use. These problems are in effect
those of destruction and disintegration.

Of the textile fibres cotton and flax are by far the most impor-
tant, and the position which they occupy is very largely deter-
mined by the properties of their cellulose basis. This cellu-
lose is amongst C.H.O compounds very much what silver and
gold are amongst the metals, manifesting, that is, a high degree
of resistance to the chief disintegrating agencies of the natural

3O2 Cellulose

world — oxygen and water. Both fibres have been used from
the remotest antiquity, though the manufacture of cotton tex-
tiles in Europe is of quite modern growth. At the time of its
introduction it was used for padding and filling purposes and
for manufacture into paper. The spinning of the short staple
fibre into yarn is an art borrowed from the East, where it has
been practised from the remotest antiquity.

Of both cotton and flax, however, we have sufficient record
— in the substantial form of manufactured products — to be able
to pronounce them for practical purposes indestructible save by
the mechanical agencies of wear and tear. In ordinary use,
however, they require periodical cleansing ; and the severe
treatments of the laundry, chemical and mechanical, lead to
more or less rapid disintegration. Very little attention is paid
to this industry from the chemical point of view, of which the
chief regulating principles are those of economic and rapid
handling. Occupying as it does a somewhat ' inferior ' posi-
tion in human affairs, it appears to be beneath the notice of
technologists. The result is unfortunate, as the very common
experience of the household will testify.

The cleansing of vegetable textiles by alkaline solutions,
wherever and however practised, is a chemical process ; and it
is high time that laundry work, conducted as it now is upon
the scale of an enormous special industry, should be more
consistently organised as a chemical industry. Great progress
in this direction would be made by modelling the procedure
of the laundry upon the general principles of treatment of these
textiles in the manufacturing industries ; i.e. in the case of
cotton and linen goods the lines of treatment should be gene-
rally similar to those of bleaching and finishing — though, of
course, differing considerably in degree. As a matter of experi-
ence, the chemical disintegration of these textiles in the course
of laundrying is considerable, chiefly through ignorance or

Experimental and Applied 3°3

neglect of the chemical properties of the celluloses on the one
hand, and the cleansing agents employed on the other. This,
again, is a subject opened up in definite directions of inquiry
by the matter of this treatise, and it is to be hoped that the
chemical history of a shirt or tablecloth may come to be
written at no distant period, and with special attention to
those conditions which make for longevity.

Of the uses of vegetable textiles in their unbleached or
partially bleached conditions there is little to be said from the
chemical side. It should be remembered that half-bleaching
treatments are fraught with some danger, owing to the chemical
changes (oxidation or chlorination) in the residual non-cellu-
lose constituents ; and to minimise these dangers a final treat-
ment with sulphite or bisulphite of soda is to be recommended.
The authors have in mind not only the facts in connection
with jute bleaching mentioned on p. 286, but have been called
in to adjudicate upon damages occurring in the bleaching
' out ' of flax goods woven with creamed or half- bleached yarns.
These have been frequently found to retain substantial quan-
tities of ' chlorine ' (bleaching powder), and it is quite re-
markable the length of time of persistence of these residues of
hypochlorites in contact with flax goods. Their presence
must involve a gradual oxidation of the entire fibre-substance,
which together with the acidity of the oxidised non-cellulose
effects a steady disintegration of the fabric. So long indeed as
goods are treated altogether without reference to the molecular
results of the treatments, sound practice is the result of tra-
dition and correct intuitions, and the chances are far too
numerous in favour of malpractice. If at any time there
should be an extensive exposure of the secrets of the ' damage
room,' it might occasion wonder that a stronger case should
not have been made out for the scientific regulation of these
industries.

304 Cellulose

The second great branch of the cellulose industry is that of
paper. Here also we meet with a large proportion of fabrics
composed of unbleached or partially bleached materials, in.
reference to which there is little to be said from the point of
view of their chemistry. They are used for 'inferior' purposes,
such as wrappings ; they serve their purpose, and there are no
problems of especial import presented by the chemical history
of the fibres in this particular form.

But it is otherwise with papers used for writing and print-
ing. In this category permanence is a first desideratum.
Books and records have more than a passing value, and it is
essential that they should be committed to pages suitably
resistant both to chemical and mechanical wear and tear. On
the other hand, we may safely affirm that there is no public
opinion in this country upon this important subject. Where
preferences for high-class papers exist they are based rather
upon aesthetic and other recondite considerations than upon
any judgment as to composition and the relation of their con-
stituents to the destructive agencies of the natural world. On
this basis white papers admit of a very simple classification
into three main groups : (A) those composed of the normal
and resistant celluloses only — e.g. cotton, linen ; (B) those
composed of celluloses containing oxidised groups or oxy-
celluloses — e.g. wood-cellulose, esparto and straw celluloses ;
(C) those containing, in admixture with the above, ground wood
or mechanical wood pulps (lignocellulose), many of which are
sufficiently ' white ' as not to prejudice a paper from the point
of view of colour.

Of the above, Class A stands beyond criticism. From the
discussion of the chemistry of the celluloses it is evident that
they fulfil all the requirements of inertness, and this may be
taken as a confirmation of the extensive experience which we
have of the lasting properties of the celluloses. Throughout

Experimental and Applied 305

the middle ages these fibres were the staple raw materials for
production of papers, and in books that have come down to us
from these times there is sufficient evidence of resistance to
the natural processes of disintegration.

Fibres of Class B have been introduced in response to the
enormously increased consumption of paper in this century,
and it becomes important to consider how far they fail, or may
on chemical evidence be predicted to fail, in regard to the
properties which distinguish the former class. It is evident
that chemically they are of totally different constitution,
esparto and straw diverging from the normal type much more
considerably than the wood-celluloses. It is a matter of ob-
servation that all papers containing these celluloses are liable
to discolouration under the ordinary conditions of wear and
tear. Chemists will have made the further observation that in
the atmosphere of the laboratory, reference books, or rather
the paper upon which they are printed, are liable to peculiar
discolourations. Thus, in laboratories where coal-tar products
are handled it is a frequent experience that our journals
change from white to bright pink, and even where there is
no direct contact with the atmosphere of the laboratory
it is common to see the pages change to various shades of
brown. This browning can be produced in a very short time
by exposure to the heat of the water-oven, and it has also
been shown that under these conditions the fibre undergoes
oxidation which is sufficiently marked to be measured by
an increase of yield of furfural on boiling with hydrochloric
acid. It is clear, therefore, that these reactive oxycelluloses
are inferior in an important chemical sense, and their use hi
books is open to the very obvious objection that the books are
more perishable. Of course, it is perfectly true that a large
amount of literature is of the ephemeral kind, and in this pro-
vince such questions as we have raised do not enter ; on the

x

306 Cellulose

contrary, paper being very much cheapened by the use of these
celluloses, a great -ad vantage is gained. It must be insisted
upon, however, that authors and publishers should have a defi-
nite judgment as to the papers to which they commit their
productions, and it would be of the greatest utility to exhaus-
tively investigate these particular celluloses from the point of
view of their resistance to the natural processes of decay.

CLASS C. — The presence of lignocellulose is a more extreme
departure from the sound basis of composition represented by
Class A. The lignocelluloses are not only more generally
reactive than the celluloses of Class B, but are easily attacked
by atmospheric oxygen (see p. 174). Added to these chemical
defects they are inferior in the mechanical properties which
contribute to the strength of the sheet of paper, and therefore
papers of this class are only permissible where lasting properties
are a question of no moment whatever.

In addition to these questions of the composition of the
fibres or pulps, the practice of loading papers with china-clay,
sulphate of calcium, and so forth, is also another of the causes
which lead to disintegration of modern papers as compared
with those of former days. There is, of course, the other side
to this question, the addition of these mineral diluents having
certain positive advantages not to be overlooked. The danger
of any practices of this kind only enters when they are
not measured at their proper utility. Paper is largely * taken
for granted' by consumers. In a great many, perhaps the
majority of cases this unenquiring consumption is not attended
with any serious consequences ; but, on the other hand, it is
quite obvious that it is attended with dangers of a very grave
character, when we are dealing with records of value for all
time. This, of course, is largely a question for posterity, to
whom we are handing down a literature produced upon
grounds for the most part of mere commercial expediency.

Experimental aud Applied 307

It is high time, as we have said before, that a public opinion
should be formed upon this subject, and it can only be formed
upon a recognised classification of papers, based upon their
chemical and mechanical ' constants,' which are determinable
by laboratory investigation. In Germany considerable pro-
gress has been made in the fixing of standards of quality and
securing their adoption by the trade. This classification by
fixed standards has' been systematically worked out in the
Government Testing Station at Charlottenburg, and the records
of the institution contain a number of important monographs
upon the various factors of quality of papers. As these, how-
ever, contain no very direct contributions to the chemistry of
cellulose, we have only to call attention to the general result of
the investigations. In our own country the character of the
paper trade differs in many respects from that of the Continent ;
and this would necessitate a special classification and series of
standards. So far, however, as this classification is based upon
differences of chemical composition the lines of demarcation
are simple and sharp, and the general recognition of these will
initiate a movement in the direction of specific uses of papers
according to their qualities and properties.

Outside the province of textiles and papers there are many
other uses of cellulose of great industrial importance, many
of which have been dealt with incidentally in the foregoing
pages. The nitrates of cellulose are the basis of manufactures
which have been developed within our own period of history.
They are used on the one hand as a plastic and constructive
material, on the other as an explosive and destructive agent ;
these uses affording remarkable illustrations of the chemical
and physical properties of cellulose. In regard to the former,
the use of the nitrated compounds of cellulose is open to the
very obvious objection of high inflammability. The combined
nitric acid is in fact a necessary evil ; and from what we now

3o8 Cellulose

know of cellulose in aqueous solution as thiocarbonate (p. 25),
its 'gratuitous ' character becomes still more prominent. The
nitric groups are merely a factor of a particular process of
solution of cellulose ; they do not modify in any essential
respect the properties of the parent molecule, but render these
available by bringing the cellulose into a condition of homo-
geneous solution. Lehner's ' artificial silk ' process illustrates
these considerations in a very direct way. For the spinning of
the thread the solution as nitrate is necessary ; but the sub-
sequent process 'of denitration changes the physical properties
of the product in so small a degree as to escape detection
otherwise than by the application of special tests. The pro-
ducts known as celluloid, xylonite, &c., are not subjected to
any denitration process ; but the cellulose products obtainable
by means of the cellulose xanthate are so similar to these that
the plastic properties of cellulose itself are more than ever
apparent as the essential basis of these manufactures. The
same facts are illustrated by the acetates of cellulose. When
these are prepared under carefully regulated conditions they
exhibit the same properties in solution as the nitrates, i.e. high
viscosity and coalescence, on evaporation of the solvent, to a
homogeneous elastic solid. It is evident, therefore, that the
nitrates of cellulose in such uses will be subjected to the ordeal
of a severe competition, and in certain directions must be dis-
placed by the parent substance itself or by derivative com-
pounds at present known or yet to be discovered.

The manufacture of explosives composed exclusively or partly
of the cellulose nitrates is now an industry of enormous pro-
portions. For many years after the introduction of gun-cotton
as an explosive its application was limited by its denomination
as a 'high explosive/ i.e. for blasting and similar purposes.
The researches of later years have shown that by changing the
physical condition of these * high explosives ' their explosive

Experimental and Applied 3C9

combustion may be brought under perfect control, and they
therefore become available as propulsive explosives, i.e. in
artillery and small arms. In these directions they are rapidly
displacing the charcoal or black powders which have done so
much service to the human race in the past centuries ! A
special advantage of these nitrocellulose powders from the
military point of view is that, owing to their perfect combustion
to gaseous products, their explosion is a ' smokeless ' one :
hence their general and popular designation. The basis of these
* powders ' is a mixture of nitroglycerin and c nitrocellulose.'
The nitrates of cellulose are gelatinised by nitroglycerin, and
by varying the proportions homogeneous plastic mixtures of
varying consistency are obtained. With small proportions of
the cellulose compounds, 7-8 p. ct., a gelatinous mass is ob-
tained, known industrially as ' Blasting Gelatine.' With lower
proportions, gradations of consistency are obtained in the
mixture which is the basis of explosives of the ' Gelignite '
class. With the cellulose nitrates increased to 40-50 p.ct. a
semi-solid product is obtained, which is worked up into threads
or ribands and constitutes the military smokeless powders
(' ballistite,' * cordite,' &c.). The product resulting from the
mixture of these two ' high explosives ' burns quietly when
ignited, and, burning from the surface^ the combustion is
perfectly under control, and can be easily regulated to avoid
detonation. A second class of * powders ' is made by mixing
the nitrocellulose and a certain proportion of barium nitrate
with a smaller proportion of camphor or nitrobenzene to allow
of their being worked up to a suitable form. Such are the * E.G.,'
'S.S.,' and other ' sporting' powders. In many of the latter the
nitrocelluloses employed are prepared by nitrating the celluloses
of Class B (supra) isolated by the processes of the papermaker;
in some cases also nitrated lignocelluloses are employed.

These industries are in a highly developed condition, the

3IO Cellulose

manufactures being carried on with the greatest precision, on
the basis of an extensive empirical knowledge of the properties
of the products. It must be admitted, however, that, in the
absence of any precise knowledge or even accepted theories of
the constitution of the cellulose nitrates, there remains a vista
of progress to be opened out by the solution or partial solution
of this important problem.

So, in fact, it may be said, generally and in conclusion, of
the industrial uses and treatments of the celluloses. All of
great and some of the greatest importance in human affairs, and
all highly developed upon an extremely slender foundation
of exact knowledge of the raw materials, it is probably true
that the cellulose industries have in many directions attained
a position of terminal excellence, measured from the point of
view of an empirical technology of the subject-matter. It may
be said with greater certainty that future progress will go hand
in hand with the progress of scientific investigation.

It is a province of applied chemistry where, as in many
others, the distinctions between * Science ' and ' Practice ' exist
only in the minds of those who grasp neither the one nor the
other. Manufacturers and technical men, if they will only take
the trouble to inform themselves, must see that an enormous field
of natural products and processes about to be explored has
a number of industrial prizes and surprises in store ; scientific
men who have to undertake the pioneering work in this field
will find sufficient stimulus to effort in the promise of progres-
sive discovery. It is to be hoped that some suggestions of
matter for research will be conveyed in the foregoing brief
account of the present position of the chemistry of cellulose.

APPENDIX I

THE illustrations which follow, reproduced from sections of typical
raw materials from amongst those dealt with in the preceding
pages, are designed to convey an outline view of their general
features of structure and arrangement in the plant.

The subjoined scheme of classification of fibrous raw materials
is based upon these structural or anatomical features considered
as the necessary basis of their varied applications in the arts.
The selection of types in illustration has been made in accordance
therewith, and as it is a sufficient key to their selection and
arrangement it is reproduced (from Indian Fibres, p. 18) without
further comment or explanation in detail.

.2

Fibres-

Fibre

aggre-4

gates

DICOTYLEDONOUS

Bast fibres only, in bundles
or filaments.

Chemical Composition.

(A) Pectocelluloses.

(B) Lignocelluloses.
Examples : Flax (A) ;

jute (B).

Entire bast tissues.
Entire stems.

Chemical Composition.
(D) Mixtures of ligno- and

pecto-celluloses.
Examples : Adansonia ;

"woods.

MONOCOTYLEDONOUS

Fibro-vascular bundles and
fibre bundles, sometimes

enclosed in cellular sheath.

Chemical Composition.
Usually mixtures of (c)

pecto-, ligno-, and cuto-

cellulose.
Examples : Sisal ; Phor-

mittm tenax.
Whole plants or parts of

plants.

Chemical Composition.
Mixtures of pecto-, ligno-,

and cuto-celluloses.
Examples : Esparto ; straw ;

bamboo.

It is important to note that the cotton fibre — the chemical
prototype of the celluloses — does not fall within the above classifi-
cation. As a 'seed hair' it stands apart. The cutocelluloses are
non-fibrous, and constitute a structural class (E) also outside the
above, though occurring in C and D in admixture with the fibrous
constituents proper.

PLATE I.

A.1 i. FLAX — Linum usitatissimum. x i5ct
Transverse section of stem.
Beginning at periphery : —
Layer of cuticular cells.
Intermediate cortical parenchyma.
Bast fibres in groups —flax fibres proper. Note secondary

thickening of cell walls.
Cambium region.
Wood.

1 These letters refer to the grouping of the table, page 311.

PLATE II.

A. 2. RAMIE, RHEA, OR CHINA GRASS — Bcehmeria nivea
x 150.

Transverse section of bast region only.

Bast fibres, distinguished by their large area from adjacent tissue.

PLATE III.

B. 3. JUTE — Cor chorus capsularis. x 50.
Transverse section of stem.

Wedge-shaped complexes of bast bundles extending from the
cambium to cortex.

PLATE IV.

B. 4. JUTE — Corchorus capsularis. x 300.

Transverse section of portion of bast. Showing anatomy of
fibrous tissue, form of bast fibres, and thickening of cell walls.

PLATE V.

C. 5. SISAL HEMP — Agave Sisalana. x 300.
Transverse section of single filament.

Kidney-shaped complex of lignified fibres almost enclosing
vessels, the whole surrounded by parenchymatous tissue.

PLATE VI.

C. 6. OIL PALM LEAF — Eltzis guineensis. x 300.

Transverse section of part of leaf, showing two classes of fila-
ments : —

1. Large fibro- vascular bundle.

2. Smaller bundle of thick-walled fibres without vessels.

PLATE VII.

D. 7. ESPARTO — Macrochloa (Stipd] tenacissima. x 50.
Transverse section of leaf.
Upper side composed of projecting ribs and deep bays fringed

with siliceous hairs. Areas of chlorophyll-bearing parenchyma

inteispersed with fibro- vascular bundles.
Lower side composed of prcsenchymatous fibres. In central

region of each rib, bands or bridges of thick-walled lignified

cells extending from lower to upper epidermis.

PLATE VIII.

D. 8. ESPARTO — Macrochloa (Stipa) tenarissima. x 150.

Transverse section of central ribs, &c.

General features of preceding section in greater detail, and show-
ing more clearly the band or bridge of lignified tissue passing
from lower epidermis between the chlorophyll areas and sur-
rounding the large fibro- vascular bundle.

PLATE IX.

D. 9. STRAW (WHEAT) — Triticum vuigare. x 150.
Transverse section of stalk.
Hypodermal layers composed of strongly lignified and thickened

fibres with smalt fibro-vascular bundles.
Larger f.v.b. disposed through thin-walled parenchyma.

PLATE X.

D. 10. WOOD — Pinus sylvestris. x 150.
Longitudinal section.

Tissue chiefly composed of the characteristic tracheides with
numerous 'bordered pits,' intersected by medullary rays.

PLATE XL

D. ii. WOOD — Tilia grandiflora. x 150.

Longitudinal section.

Wood fibres, woody parenchyma, and large pitted vessels with
oblique septa.

PLATE XII.

E. 12. RAFFIA — Raphia Ruffia. x 300.

Transverse section of epidermal tissues constituting the commer-
cial fibre.

Cortex of upper surface, with bundles of hypodermal fibres with
strongly thickened walls.

PLATE XIII.

E. 13. RAFFIA — Raphia Ruffia. x 300.
Surface view.
Cortical cells with serrated outline and stomata.

PLATE XIV.

E. 14. BOTTLE CORK — Quercus suber. x 300.
Transverse section.
Thin-walled cork cells.

APPENDIX II

IN the period 1895-1900 succeeding the publication of the first
edition of this work, there have appeared a number of contribu-
tions to the general chemistry of ' Cellulose,' the more important
of which have been recorded and discussed in a volume of
'Researches on Cellulose' by the present authors, published in
1901.

It will be of interest to our readers to follow the main lines of
growth of the subject ; and we therefore give a brief outline, and in
very general terms, of these later developments.

Cellulose. — Constitution.— An observation of fundamental im-
portance is the direct conversion of cellulose into a crystalline
furfural derivative under the action of the halogen hydracids.
Empirically the reaction in the case of hydrobromic acid may be
expressed by the equation :

C6H1006 + HBr - 3H20 = C6H5O2Br

the product being a brom-m ethyl furfural. The condensation
takes place readily at 100° C. in presence of anhydrous ether. A
particular point of interest arises in regard to the generalisation of
the reaction as one specially characteristic of the ketoses, i.e. the
keto-hexoses. The conversion of the typical levulose is represented
as follows :

H H | HH

C C C
;H2o;;HOH|iOHH;

OH.C.C. C C C CH2OH - OC.C : C.C : C.CH3Br.

(H. J. H. Fenton. Chem. Soc. J., 1901, 361.)

The yields of the w-brom -methyl furfural from various forms of
cellulose were found to be high (33 p.ct.), higher indeed than from

Y

314 Cellulose

levulose. The reaction is therefore a main reaction, and shows
that cellulose under these conditions breaks down, at least in large
parf, to ketohexose units. By these investigations therefore the
polyaldose view of the constitution of cellulose is directly called in
question. We have found on other grounds that a ketonic
formula is to be preferred (ist ed. p. 77), the fifth O atom having
ketonic rather than aldehydic function. This is consistent either
with an open chain or closed ring formula for the assumed C6
unit. There are general grounds of preference for the latter. But
this is a matter of speculation and hypothesis.

A point to be noted in connection with Fenton's researches is
that the normal celluloses (of the cotton group) give higher yields
of the furfural derivative than the cereal celluloses (group C),
which on the other hand are characterised by high yields of furfural
under the action of aqueous condensing acids. This decomposi-
tion is referred by many chemists to the presence in the cereal
celluloses of a pentose anhydride. In view of these later facts the
explanation, which is on other grounds doubtful, becomes unneces-
sary. It is clear that the transition from the normal chain to the
C4O ring is equally characteristic of hexose as of pentose units,
and the assumption that * furfural yielding ' is equivalent to ' pentose '
or ' pentosane ' carbohydrate, falls away.

A second point to be noted arises in connection with the ex-
haustive study of the action of ethereal hydrobromic acid on the
celluloses. In a succession of treatments with the acid, diminish-
ing yields of the brom-methyl furfural are obtained, and the final
residue has the composition and character of the humic or ulmic
series of complex derivatives described on p. 240. « M. Gostling.
Proc. Chem. Soc. 18, 250).

It is probable from later investigations of our own that pyrone
groups are formed as an alternative or complementary course of
condensation of the carbohydrates, and are represented in these
complex products.

On the broad and general question of the actual constitution of
cellulose there is as yet but little positive evidence. It is a ques-
tion of proximate arrangement and configuration of ultimate con-
stituent groups which we assume to be of C6 dimensions, and to be
represented by the ordinary molecular formulas. But we have no
conception of a molecule of cellulose, and no data as to its dimen-
sions. We have positive evidence as to a reacting unit, but of
variable dimensions, and the more definite synthetical reactions of

Appendix II 315

cellulose are expressed in terms of these units. But in these
reactions the factor of mass, as distinct from relative molecular
mass, has to be taken into consideration ; and, to cite a particular
case, the recent elaborate investigations of W. Will on the nitra-
tion of cellulose, and the decompositions of the nitrates by heat,
lead to the conclusion that in both directions there are no breaks
of continuity corresponding with definite reacting units of relatively
small dimensions, (fnfra, p. 317.)

This problem of the relation of molecule to mass necessarily
also arises in regard to the structural peculiarities of cellulose.
The conversion of cellulose into films, threads, and generally into
solids of continuous dimensions, has shown that the mechanical
properties of these solids are a direct function of the molecular
state of the parent substances, whether celluloses or cellulose
derivatives. Thus the hydrocelluloses (p. 54) are formed from the
fibrous celluloses at the expense of their tenacity : similarly, when
converted through soluble derivatives into continuous solids, these
are brittle and of low tenacity. The normal acetates give tough
films ; but if the acetylation is carried to the point that chemical
disintegration begins, as evidenced by the presence in the product
of reactive CO groups, the product gives brittle films.

These considerations may be borne in mind in regard to the
future investigation of the problem. But the problem is without
present promise of solution, and it must be admitted that we have
no criterion of the kind or degree of association of the molecular
units in the complex aggregates of the cellulose group.

Esters. — On the general subject of the nitric esters of the
carbohydrates Will and Lenze have made investigations leading
to the conclusions that whereas the aldoses are fully esterified, the
hexoses giving pentanitrates and the pentoses tetranitrates, the
ketoses with n.OH groups yield nitrates containing n— 2. O.NO2
groups as a maximum, the two remaining OH groups passing into
the anhydride form. These nitrates of the ketose-anhydrides are
distinguished by much greater stability. (Berl. Ber., 1898, 68.)

The authors have investigated the reaction of formation of
these nitric esters under the usual conditions of treatment of the
celluloses with a mixture of nitric and sulphuric acid, and conclude
from their experiments that the latter acid reacts also with the
cellulose hydroxyls. The fixation of SO4H residues in some
quantity is proved by analysis of the products formed under
certain conditions ; and the fact has to be taken into account

Y 2

316 Cellulose

under all conditions of treatment, especially in regard to the very
important question of ' stability,5 and the uses of these products
as explosives. (Cross, Bevan, and Jenks. Berl. Ber. 34, 2496.)

The highest derivative in this series of esters being the
trinitrate — on the C6 formula — the fact is shown to be consistent
with the presence of 4.OH groups in the cellulose unit, which
now must be taken as finally established by the general recog-
nition of the highest acetate as a tetracetate, and as a true
cellulose derivative. A higher degree of acetylation implies a
hydrolysis of the cellulose, which is confirmed by a study of the
properties of such derivatives. These conclusions have been
verified and extended by the later investigations of Z. H. Skraup
of the acetylation of starch and cellulose. (Berl. Ber. 1899, 2413.)
In regard to the lower limits of acetylation it is stated in this
volume (p. 35, ist ed.) that the normal celluloses do not react with
acetic anhydride at its boiling temperature. Investigations by the
authors have shown that this statement, current in the text-books,
is erroneous ; a mono-acetate (C6) is formed under these condi-
tions. This product is insoluble in all the solvents of the cellulose
esters, and moreover resists the action of cuprammonium solutions.

The authors have further investigated the benzoates of
cellulose, and the conditions of their formation by interaction
of cellulose and benzoyl chloride in presence of alkalis. From
these esters mixed esters have been obtained by the action of
nitrating acid. The benzoyl residues are converted into nitro-
benzoyl, and further reaction ensues with the residual OH groups
of the cellulose.

The following conclusions appear to be justified : the highest
benzoate is the dibenzoate, or on the C12 unit the tetrabenzoate.
Taking 8.OH groups as the maximum in this unit, five only react
in these mixed esters, as compared with six as a maximum in the
simple nitric esters. (See * Researches on Ce-.ulose,3 pp. 34-4°-)

From points of view other than the purely theoretical, various
and important investigations of cellulose esters have been published
in recent years.

Lunge and Bebie have carried out an elaborate enquiry into
the constants of nitration of the normal cellulose ; chiefly concern-
ing the yields and composition of the nitrates under definite varia-
tions of the -more important chemical and physical conditions of
the reaction. The results constitute the most extensive series of
numerical records hitherto published, for which the original papers
must be consulted. (Ztschr. Angew. Chem. 1901, 483. See also
O. Guttmann, Chem. Ztschr. I. No. 12.)

Appendix II 317

The authors with A. Luck have also investigated the actions
of diluted solvents upon the fibrous nitrates, under the action of
which they are converted into dense structureless forms with
elimination of the products causing instability. The process is the
basis of technical developments based upon the more perfect con-
trol of the process of gelatinisation or * colloidisation,' an
essential condition of the use of these products as restrained
or progressive explosion. (A. Luck and C. F. Cross. J. Soc.
Chem. Ind., 1900.)

The most important event in connection with the scientific and
technical development of this subject has been the foundation in
Germany of the Research Institution of Neu Babelsberg, Berlin.
(Central Stelle fur Wissenschaftlich-technische Untersuchungen.)
This institution, mainly devoted to the technology of nitrocellulose
explosives, has published two brochures on the question of the
stability of the cellulose nitrates. Full abstracts of these com-
munications will be found in the Journal Soc. Chem. Ind. 1901,
609, 617 ; 1902, 1470-1. We can only notice here the main result
of the elaborate investigations of Prof. Will in its bearing on
the scientific side of the subject. It has been established that the
normal stable nitrates when heated at high temperatures in an
atmosphere of dry carbonic anhydride are continuously decom-
posed with a regular disengagement of nitric oxide, the decomposi-
tion taking place according to the typical equation :

C12HI5(N02)5010- C10H3N08 + 4NO + 6H2O + 2CO

and reaching the limit represented by the formation of the end-
product in question. The points to be noted in the composition
of this product are the retention of one-fifth the original nitrogen
and the loss of i-C atom for each C6 unit. Until the constitution
of this empirical residue has been elucidated we cannot go beyond
the statistical relationships established. The prominent general
feature of the decomposition or dissociation is its regularity, i.e.
continuity, upon which the * stability' tests are based. It suggests
a similar continuity in the original ester reactions.

We may briefly note here the publication of a book under the
title * Smokeless Powder, Nitrocellulose, and Theory of the Cellu-
lose Molecule,' by J. B. Bernadou : New York, 1901 (London :
Chapman & Hall, Ltd.). This work contains, in addition to the
author's interesting speculations and records of experimental

3 1 8 Cellulose

work, a re'sume' of important recent investigations of Vieille and
Mendeljeff.

Cellulose Sulpho-carbonates (Viscose). — The authors have
published an account of later researches into the nature and con-
stitution of this series of compounds. The main point established
is that the affinity of the cellulose xanthogenic acid is consider-
ably higher than that of the fatty acids, and generally higher than
that of the monocarboxylic acids. Consequently the solutions of
the crude compound may be treated e.g. with acetic acid in
excess without decomposing the alkali salts of- the cellulose
sulphocarbonic acid. The acetic acid, on the other hand, entirely
decomposes the by-products of the original reaction and reactions
of spontaneous decomposition. By this means the isolation of pure
compounds of this series is much facilitated, the separation from
sodium acetate on addition of alcohol being satisfactorily sharp.

The following stages in the process of reverse decomposition

,

have been established : The general formula C S\ having

XSNa

been verified with satisfactory precision, and X being the cellulose
residue of various dimensions, it is found that when freshly pre-
pared X lies between C6 and C12, and the compound is not
precipitated by dehydrating agents : as X approaches C,2 the
xanthate is precipitated by alcohol, and readily redissolves in
water : the C24 xanthate is precipitated by smaller proportions of
dehydrating agents from alkaline solutions, and is entirely pre-
cipitated by acetic acid ; in other words, is insoluble in water. The
cellulose when reaggregated to these dimensions is not soluble as
a sodium xanthate, but requires the further combination of its
OH groups with the alkaline hydrate to produce a soluble com-
pound. These stages are well defined, and by their general
recurrence in the course of investigations to the apparent exclusion
of intermediate stages, it is suggested, though it cannot be finally
affirmed, that the decomposition as it actually occurs in the solu-
tion takes place in the later stages by units of C12 dimensions.

The analysis of viscose solutions is obviously much simplified
by these observations. By volumetric estimation, using succes-
sively normal acetic and hydrochloric acids, the alkali combined
with the cellulose is determined, and the number can be confirmed
by titration with a standard iodine solution. (Berl. Ber., 1901,
34, 1513-20.)

Appendix II 319

Ligno-celluloses. — The authors have shown that the colour
reactions of the lignocelluloses with phenols are not characteristic
of the lignone complex as such, but are due to break-down products
— in all probability to hydroxyfurfurals. These bodies have been
prepared by the interaction of furfural and hydrogen peroxide in
presence of iron salts : they give reactions with phloroglucinol
and resorcinol, identical with those of the lignocelluloses in their
natural state. (Cross, Bevan,and Briggs. Berl. Ber.33, 2132.) These
reactive constituents of the natural lignocelluloses are easily
removed by treatment with oxidants in regulated small propor-
tions, the lignocellulose undergoing only small losses of weight,
and retaining its essential chemical characteristics unchanged.

Further studies of the lignone complex in the case of the jute
fibre have somewhat modified the conclusions set forth in the first
edition, and the text has been accordingly rewritten in those
portions.

The furfural-yielding constituent is more probably a cellulose
or an anhydride, and appears with the cellulose complex when
isolated by the chlorination process. The lignone is thus to be
considered as distinct from this /3-cellulose and from the hydroxy-
furfurals. These latter may be formed from the /3-cellulose by
processes of hydrolysis and condensation, and oxidation occurring
* naturally.' It is certain that active oxygon is always present on
the surface of the lignocelluloses, indicating a slow and progressive
auto-oxidation of the fibre substance. The observations and
ingenious investigations of W. J. Russell (Nature, vol. 65, p. 200)
have emphasised these phenomena by showing that they are
associated with * emanations ' which act upon sensitive photo-
graphic surfaces, and produce an image of the objects. Russell
considers that the evidence so far accumulated points to these
emanations being hydrogen peroxide.

The problem of the constitution of the characteristic lignine
complex is so far simplified. Its most important constituent
groups are : (i) the benzenoid group, combining directly with
chlorine ; and (2) a group or groups of approximate formula
C2m H2m Om, which break down by gentle oxidation and hydrolysis
finally to acetic acid as a main product, with probable formation
of ketonic acids of low molecular weight as intermediate stages.
The complex contains a minimum proportion of hydroxyl groups
and of methoxyl groups.

Some further light has been thrown on the relationships of

320 Cellulose

cellulose to lignone groups in the lignocellulose complex, by later
investigations of certain lignocellulose esters.

The benzoate prepared by treating with benzoyl chloride in
presence of alkali is a monobenzoate, calculated to the simplest
empirical formula C]2H18O9. The benzoyl group enters the
cellulose residue ; the lignone is unaffected, and when removed by
the ordinary treatment a cellulose benzoate is left as the end
product.

On boiling the lignocellulose with acetic anhydride, an acetate
is formed, which analyses as a diacetate of the empirical unit
C12H18O9. The complete statistics, however, appear to show
that the ester reaction is attended by internal dehydration through
interaction of other groups of the complex. The lignone group,
however, retains its general characteristics, and may be removed
by similar treatment as the original, and the cellulose is separated
in the form of a diacetate (C12).

The benzoate (supra) also reacts with acetic anhydride, and
the proportion of acetyl groups entering is not affected by the
presence of the benzoyl group.

These ester reactions taking place in the cellulose group, it is
further established that the lignone complex contains no OH
groups reactive under these conditions, and also that there are no
free aldehydic groups.

These reactions are of use in the investigation of the ultimate
constitutional problems which continue to engage the attention of
the authors, and to which it is hoped other chemists will be
attracted by the publication of these evidences of more definite
progress,

INDEX OF AUTHORS

ABEL, 44

Cross and Witt, 148

Armstrong, 245

Crum, W., 24

BAEYER, 245

DEMEL, 240

Bary, de, 231

Dopping, 226

Bebie, 316

Durin, E., 72

Bechamp, 46

Du Vivier, 45

Beilstein, 259

Benedikt and Bamber-

ger, 1 88, 232

ERDMANN, u, 161,

Bernadou, 317

197

Berthelot, 87

Briggs, 319

FENTON, H. J. H.,

Brown, A. J., 72

313

Brown, Horace, 67

Fischer, E., 262

Brown and Morris, 65,

Fischer and Schmid-

257

mer, 18

Brunner, 167

Flechsig, 49

Flint and Tollens, 99,

265

CALVERT, Grace, 21
Chalmot, de, 181, 185

Fluckiger, 228
Franchimont, 6l, 87

Chardonnet, de, 45

Francis, 254

Chevandier, 174

Frank, 224

Chodnew, 216

Fremy, 90, 91, 173,

Collie, 62, 149

176, 216, 229

Cross and Bevan, 7,

61, 70. 79» 80, 83,

113, 124, 131, 137,

GANS and Tollens, 221

138, 142, 152, 164,

Gilson, 12

208, 232, 240, 244,

Girard, 21

247, 259, 263

Godeffroy, R., 219

Cross, Bevan, and

Goodale, 237, 243

Briggs, 319

Goppelsroeder, 18

Cross, Bevan, and

Gostling, M., 314

Jenks, 316

Gottlieb, 175

Cross, Bevan, and

Green, Cross, and Be-

King, 243

van, 298

Griffin and Little, 288
Guignet, 53
Guttmann, O., 316

HALLIBURTON, 87
Hantzschand Schniter,

137

Hawes, G. W., 175
Hime and Noad, IO
Hodges, 80, 232
Hoehnel, 227
Hofmann, A. W., 79
Hofmeister, 237
Honig and Schubert,

48, 225
Hoppe-Seyler, 66

KABSCH, 173
Karolyi, 44
Kirchner and Tollens,

221

Knecht, 55
Koechlin, C, 21
Kolb, 218
Krauch and V. d.

Becke, 166
Kraus, 153
Kugler, 227
Kuhlmann, 46

LANGE, 22, 141, 214,

240

Lehner, 45
Lenze, 315
Lindsey and Tollens,

49, 83, 198
Lintner and Dull, 257

322

Cellulose

Lloyd, 1 8

Lb'wig and Kolliker, 87
Luca, de, 46
Luck, A., 317
Lunge, G., 316

MAC NAB and Ristori,

44

Mann and Tollens, 1 84
Maurey, 46
Meissner and Shep-

pard, 152
Mendeljeff, 318
Mercer, 24
Meyer, V., 163
Miller, W. A., 239
Mitscherlich, 226
Muhlhauser, 132
Muller, Hugo, 5, 79,

iio, 175, 214, 219
Miintz, 1 68

NASTJUKOW, 61
Nolting and Rosen-
stiehl, 6 1

O'SULLIVAN, 257

PARNELL, 24
Payen, 211
Pears, A., in

Pelouze, 46
Poumarede and Fi-

guier, 187
Prudhomme, II

RAMSAY and Chorley,

68, 154, 204
Reichardt, 216
Rosenfeld, 13
Russell, W. J., 319

SACHS, 73, 237
Sachsse, 172, 221
Schaefer, 87
Scheibler and Mittel-

meier, 258
Schleiden, 224
Schlichter, 272
Schmidt, 61, 87, 225
Schmitz, 237
Schulze and Tollens,

163, 260
Schunk, 290
Schuppe, 177
Scoffern and Wright,

J3 .
Sestim, 137, 240

Skraup, Z. H., 316
Smith, 83, 164, 259
Spon, 79, 243
Stein, 271
Stern, 49

Stutzer, 152

TAUSS, 22
Thomsen, 187
Thorn, 2 [3
Tollens, 101, 181, 259,

261
Tollens and others, 261

URBAIN, 173

VfiTILLART, 243

Vieille, 41, 318
Vortmann, 265

WATTS, 291
Weber, 19, 131
Webster, 113
Weiske, 152
Wheeler and Tollens,

187, 212
Wiesner, 243
Will, 47, 315, 317
Wissenburgh, 228
Witt, O. N., 79
Witz, 61, 297
Wurster, 174

ZEISEL, 106, 189, 365

INDEX OF SUBJECTS

ACETATES of cellulose, 34, 252, 316 .
Acetic acid, formation from cellu- j
lose, 255

— anhydride, action upon cellulose,
35, 316 ; upon regenerated cellu-
lose, 37

— condensation, 193

— residue in woods, 191 ; product
of simple hydrolysis of ligno-
celluloses, 192 ; characteristic
feature of lignification, 192, 319

Adipocelluloses, 90, 225, 226 ;
proximate analysis, 227 ; general
methods of investigation, 267

Aloe fibres, 220

Amylobacterium, 66

Amyloid, 53, 224

Aniline dyes, action on jute fibre,

IJ5

— salts, action on jute fibre, 115
Arabic acid, 216

Arabinose, 86, 216
Arabinosic acid, 216
Ascidia, 87

BACTERIUM xylinum, 73

Ballistite, 44, 309

Bamboo stems, 220

« Belfast Linen Bleach,' 80

Benzoates of cellulose, 32, 251, 316

Blasting gelatin, 309

Bleaching, isolation of cellulose
frcm raw fibres, 244, 255 ; linen
yarn, 286 ; jute cuttings, 287 ;
cotton, 288 ; ' maiket bleach,'
• madder bleach,' 291 ; linen, 292

' Brewers' grains,' composition of,
163, 260 ; method of examina-
tion, 260

Brom-methyl furfural, 313

Butyric fermentations, 234

CARBOHYDRATES, 2 ; general me-
thods for identification, 261 ;
nitration, 315

Carragheen mucilage, 225

Celluloid, 44, 308

Cellulose, I ; empirical composi-
tion, 3 ; hydrates, 4 ; their re-
action with iodine, 7 ; of green
fodder plants, 7 ; solutions of,
8 ; in zinc chloride, 8 ; in
zinc chloride and HC1, 9 ; in
ammoniacal cupric oxide, 9 ;
in ammoniacal cuprous oxide, 13,
246 ; threads or filaments in
electric lamp, 8 ; crystallised, 12 ;
theory of action of solvents, 14 ;
qualitative reactions and identi-
fication, 14 ; compounds of, 15 ;
with dilute alkalis and acids, 16 ;
with colouring matter, 19 ; capil-
lary phenomena, 18 ; action of al-
kaline solutions at high tempera-
tures, 22 ; action of concentrated
alkaline solutions, 23 ; thiocar-
bonates, 25, 318; their spontane-
ous decomposition, 26 ; their co-
agulation by heat, 27 ; quantative
regeneration of cellulose from
solutions of thiocarbonate, 28 ;

324

Cellulose

purification by alcohol and by
brine, 248 ; uses in microscopic
work, 249 ; theoretical notes, 249 ;
regenerated cellulose from thio-
carbonate, 29 ; reaction with
acetic anhydride, 37 ; theoretical
view of thiocarbonate reaction, 29,
316. Reacting unit, 31 ; benz-
oates, 32, 316 ; soluble alkali,
33 ; acetates, 34 ; interactions
with acetic anhydride, 35 ; and
acetic anhydride in presence of
zinc chloride, 36 ; in presence of
iodine, 36 ; nitrates or nitrocel-
luloses, 38 ; their general pro-
perties, 39 ; approximate com-
position (table), 42 ; thermal
constants, 42 ; heat of combus-
tion, 43 ; products of combus-
tion, 43 ; industrial uses, 44,
307 ; gradual decompositions, 46.
Action of sulphuric acid, 48 ;
transformation to a sugar, 49 ;
composition of body produced by
dissolving in HSO4, 49. De-
compositions of, 52 ; by non-
oxidising acids, 53 ; practical
application, separation of cotton
from wool fabric, 55 ; by oxidants,
56 ; in acid solutions, 56 ; in
alkaline solutions, 60 j resolution
by ferments, 63 ; resolution con-
stituting 'decay,' 66; by con-
densation of carbon nuclei, 66 ;
feeding or nutritive value of, 67 ;
destructive distillation of fibrous,
of regenerated from thiocarbonate,
68 ; tables, 69 ; constitution of,
reactions throwing light upon it,
75 ; theoretical notes on, 257 ;
three subdivisions in group, 78 ;
purification in laboratory, (Tgh;
* cellular,' 82, 85 ; Jfcprn wooas
and lignified tissues, KVp elemen-
tary composition, 83; yield of
furfural, 83 ; from cereal straws,
esparto, 84 j their ultimate com-
position, 84 ; yield of furfural and
reactions, 84 ; results from solu-
tion as thiocarbonate, 85 ; re-
generated from straw and esparto

cellulose thiocarbonate, 85 ;
pseudo- or hemi-, 87 ; a con-
stituent of protozoa, 87 ; com-
pound, 89 ; adipo- and cuto-, 90 ;
pecto- and muco-, 90 ; Fremy's
classification, 90 ; para- and
meta-, 90; ligno-, 91, 92; (see
Jute) o and £, from jute fibre,
93 ; general view of the group,
235 ; processes of decay and
destruction (tables), 239 ; morpho-
logy, 243 ; technology, principles
of, 273 ; preparation of fibres
from raw material, 276 ; flax,
and jute, 275 ; spinning, 279 ;
bleaching, 284 ; of jute, 285 ;
linen yarn, 286 ; jute cuttings,
287; cotton, 288 ; constitution of,

313

Cell-wall, differentation of sub-
stances composing, 86

Cerin, 228

Ceryl alcohol, 80

China grass, 79, 220, 278

Chloroplasts, 73

Coal, 66, 238

Collodion varnishes, 44 ; films, 44

Colloidal cellulose, 53

Combustion, rapid method, 245

Condition, water of, 5

Cordite, 44, 309

Cork, 225, 226

« Crude fibre,' Weende method of
estimation, 165

Cutin, 228

Cutocelluloses, 90. See Adipocellu-
loses

Cutose, 90, 229, 230

DECACRYLIC acid, 227
Dehydration, 245
Dextrose, 64, 74, 86, 222, 261
Diastase, 71 ; secretion by flower-
ing plants, 74
Diazotype process, 298
Drupose, 162
Dye woods, 204
Dyeing processes, 294
Dynamite, 309

Index

325

ELECTRIC lamp, 8

Enzyme (cyto-h)drolyst), 65 ; in

digestive tract of herbivora, 67,

216

Esparto, 84, 220
Eulysin, 227
Explosives, 44, 308, 317

FERMENT, acetic, forming cellulose,
72 ; hydrolyses of cellulose, 255

Ferric ferricyanide, action on jute,
115 ; theory of dyeing, 124 ; be-
haviour with gelatin, 129

Fibres, raw, investigation of, 269 ;
fibre constants, 300 ; numerical
expression, 301

Films, 4, 44

Filter paper, 3

Finishing processes, 6

Flax, 79, So ; retting and scutching,
217; cortical tissue, 217; flax
fibre proper a pectocellulose,
methods of isolation in laboratory
and in practice, 218 ; flax cellu-
lose, 219

Food in relation to work, Miintz's
researches, 1 68; tables, 169,170,
171

Fremy's classification, 173

Furfural, product of acid hydrolysis
of oxycelluloses, 82 ; reagent for
obtaining, 82 ; furfural-yielding
complex, 98 ; estimation of, 99 ;
oxidation of, 319. See also Jute

GALACTOSE, 86, 199, 216, 222, 262

Glycerol, 228

Glycodrupose, 161

Glycolignose, 198

Glycuronic acid, 184

Green fodder-plants, 7 ; investiga-
tion of, 270

Grundsubstanz, 167

Gum-arabic, 216

Gun-cotton, heat and products of,
combustion, 43. See Cellulose
nitrates

HACKLER'S dust, 234
Hackling, 80, 234
Hemi-celluloses, 87

Hemp, 79

Hexoses, identification, 261
Hippuric acid, 192
Humus, 238, 239, 314
Hydracellulose, 54
Hydration, 245
Hydrocellulose, 54
Hydroxyfurfural, 317
Hydroxypyruvic acid, 47

JUTE, 91 ; composition of fibre, 92;
furfural-yielding complex, 92 ;
cellulose isolated not homo-
geneous, 93 ; quantitative esti-
mation of cellulose constituents,
94 ; by chlorination, 94 ; by bro-
mination, 95 ; by treatment with
nitric acid and potassium chlorate,
96 ; with dilute nitric acid, 97 ;
by sulphite and bisulphite pro-
cess, 97; estimation of furfural-
yielding complex, 99 ; of keto R.
hexene constituent, 101 ; estima-
tion of constants of chlorination,
1 02 ; empirical formula, 102 ;
determination of HC1 in reaction,
104 ; control observations, 104 ;
estimation of secondary constitu-
ents (methoxyl) by standard me-
thod of Zeisel, 106 ; CO. CH2 resi-
due, 107 ; systematic account of
fibre ; 'butts ' or ' cuttings,' 109,
287 ; sp.gr., no ; analysis of va-
rious specimens (table), no; com-
position, in; empirical formula,
in; artificial cultivation by A.
Pears, 1 1 1 ; analysis of cultivated
fibre (table), 112 ; lignocellulose
hydrates, 113; solutions of ligno-
cellulose, 114, 263; qualitative
reactions and identification, 115,
262 ; action of aniline salts and
coal-tar dyes, 115, 262 ; action of
phloroglucinol, iodine, chlorine,
ferric chloride, ferric ferricyanide,
115, 262, 266; chromic acid,
potassium permanganate, 116;
compounds of jute cellulose, with
acids and alkalis, from dilute
solutions, 1 1 6, 263 ; from concen-

326

Cellulose

trated solutions of alkaline hy-
drates, mercerisation, 120 ; thio-
carbonate reaction, 121 ; com-
pounds with metallic salts, reac-
tion with ferric ferricyanide, and
theory of dyeing, 124; com-
pounds with negative radicals,
benzoates, 131 ; acetates, nitrates,
132 ; lignocellulose under nitra-
tion behaves as a homogeneous
body, 134; compounds with the
halogens, chlorine, 134; bromine,
137 ; iodine, 138, 263 ; resolution
into constituent groups, 139; by
hydrochloric, hydriodic, sulphuric
acid, 140 ; nitric, dilute, in pre-
sence of urea, 141, 264; by alkalis,
141 ; by acid oxidants, chromic
acid, 142, 264 ; chromic and sul-
phuric, 144 ; strong nitric, 145 ;
joint action of oxides of nitrogen
and chlorine, 147 ; by alkaline
oxidants, potassium, perman-
ganates, 147 ; hypochlorites, 148;
hypobromites, 149 ; interaction
with sulphites and bisulphites,
149, 265 ; animal digestion, 151 ;
spontaneous decomposition, 152;
destructive distillation, 153 ;
general conclusions as to com-
position and constitution of ligno-
cellulose, 155 ; lignocellulose
considered as a whole, 157, 319 ;
ultimate analysis, 266

KETO R. hexene constituent of

lignocelluloses, 101. See Jute
Kieselguhr, 309

LEUCOGALLOL, 102
Levulinic acid, 199
Levulose, 74, 262 ; condensation

T-°lf'313

Lichemn, 2 4
Lignification, 92
Lignin, 94
Lignite, 66, 238

Lignocellulose, 91, 92 ; esters, 131 ;
of cereals, composition of brewers1

grains, 163; straws, how differ'
entiated from typical lignocellu-
lose, 164. See Jute
Lignone, 94, 319, 320

MAIROGALLOL, 102

Maltose, 74

Mannose, 86; hydrazone, 199;

identification, 262
Meals, analysis of, 167 ; rice meal,

167
Mechanical wood-pulp, estimation

in paper, 174
Mercerisation, 23, I2OV
Metapectic acid, identical with

arabic acid, 216
Metapectin, 216
Methoxyl determination in woods,

1 06, 1 88, 265
Mitscherlich process, 198
Mucic acid, 199
Mucocelluloses, 90. See Pectocellu-

loses
Musa, 220

NITRIC acid, toughening action on

papers, 254
Nitrocelluloses, or cellulose nitrates,

38, 309, 315. 3i6
Nitroglycerin, 309

OIL-WAX complex in flax, 80
Oleocutic acid, 230
Ophrydium versatile, 87
Oxycellulose, 56, 82 ; preparation

and diagnosis, 254 ; identification,

262
Osazone, 222

PAPER, analysis of, 271 ; perma-
nence a first desideratum in writ-
ing and printing paper, 304 ; dis-
integration of modern, 306

Paper-making fibres, 280

Parapectic acid, 216

Parapectin, 216

Parchmenting process, 253

Index

327

Peat, 66, 238

Pectase, ferment enzyme, 2l6

Pectic acid, 216, 290

Pectin, 216

Pectocelluloses, 90 ; how distin-
guished from mucocelluloses, 215;
general characteristics, 217 ; flax,
217 (which see] ; China grass or
Ramie, nettle fibres, monocoty-
ledonous fibre aggregates, 220 ;
parenchymatous tissue of fruits,
221 ; mucilaginous constituents
of plant tissues, quince mucilage,
221 ; salep mucilage, 223 ; amy-
loid, lichenin, 224 ; carragheen
mucilage, 225 ; general methods
of investigation, 267

Pectose, 90, 216

Pentaglucoses, 93, 262

Pentosans, 93, 185, 186

Pentoses, 86

Phloroglucinol, action on jute, 115,
192

Phormium, 220

Powders, smokeless, 309 ; sporting,

309

Printing processes, 294
Pseudocarbons, 70
Pseudocelluloses, 87
Pyrocatechol, from woods of Coni-

ferse, 198

Pyrocatechuic acid, 198
Pyroxylins, 39

QUINCE mucilage, 221

RAMIE, 220

Retting, 67, 80, 234, 277
Rhea. See China grass
Rot-steep, 67

SACCHARIC acid, 261
Salep mucilage, 223
Schultze's reagent, 173
Scutching, 80

Silk, Dr. Lehner's artificial, 45, 308

Skeletonising, process of, 66 ; a

simple means of differentiation, 67

Spinning processes, 279

Stability of nitrocelluloses, 317

Stearocutic acid, 230

Straws, cereal, behaviour in thio-
carbonate reaction, 164 ; wood
gum in, 187

Suberin, 228, 231

Suberose, 228, 231

Sugar cane, 220

Sugars, 65 ; cane sugar first assimi-
lated, and probable immediate
mother substance of cellulose, 72

TEXTILES, analysis of, 271
Thiocarbonate of cellulose, 25, 318.

See also Cellulose
Tissue-substance, first step in

building-up of, 74
Tunicin, 87

VARNISHES, collodion, 44
Vasculose, 90
Viscose, 25, 247, 318

WEENDE method, 165

' Willesden ' goods, 13

Wood-gum, 187

Woods, 91 ; structural elements,
172 ; Fremy's classification, 173 ;
general property to form hydrogen
peroxide, and estimation of
mechanical wood-pulp in papers,
174 ; empirical composition, 174 ;
tables, 75 ; proximate analysis
table, 175 ; resolution into cellu-
lose and non-cellulose, 177 ; dis-
cussion of Sachsse's view that
lignocelluloses are the products of
metabolism of cellulose, 178 ;
estimation of furfural (table), 182;
de Chalmot on life-history of
woods, 182 ; wood gum, 184,
187 ; in cereal straws, 187 ;
analyses, 188 ; methoxyl deter-
minations, 1 88, 189 ; acetic
residue, 191 ; destructive dis-
tillation, 192 ; chlorination of
wood lignocelluloses, dicotyledo-

328

Cellulose

nous, 194 ; coniferous (table),
195 ; synthetical reactions, nitra-
tion, 196 ; chemistry of woods of
Coniferge, 197 ; investigation of
sulphite pulp process, 198 ;
empirical formulae with methoxyl
determinations, 201 ; yields of
pulp, 202 ; destructive distilla-
tion, 204 ; tables, 205, 256 ;
Ramsey and Chorley's tables,
207 ; disintegration by reagents,
proximate resolutions, 208; table,

209 ; sulphurous acid, bisul-
phites, neutral sodium sulphite,

210 ; alkaline processes, acid pro-
cesses, 211 ; ultimate resolutions,

extreme action of alkaline hy.
drates, 213 ; chromic, in pre-
sence of sulphuric, acid, 214, 266

XANTHATES of cellulose, 26, 247,
318. See Cellulose thiocarbon-
ates

Xylan, in wood gum, 184

Xylonite, 44, 308

Xylose, 86

'ZEISEL,' standard method of esti-
mating methoxyl, 106, 189, 265

PRINTED BY

SPOTTISWOODE, BALLANTYNE AND CO. LTD.
LONDON, COLCHESTER AND ETON ENGLAND

UNIVERSITY OF CALIFORNIA LIBRARY