• A NEW SERIES OF PLANT SCIENCE BOOKS

edited by Frans Verdoorn
Volume VI

THE CYTOPLASM

of
THE PLANT CELL

Alexandre Guilliermond was born in 1876.
He obtained his Ph.D. degree at the Sorbonne
in 1902 and worked for many years at the
University of Lyo7is. He came to Paris about
1921; at that time he was already famous for
his work on the development, sexuality, tax-
onomy and phylogeny of the Endomycetes and
Saccharomycetes ; the higher Ascomycetes; the
cytology of the Cyanophyceae and Bacteria;
starch formation; and especially on account of
his research and theories on the chondriome
and vacuome. He was appointed Chairman
of the Department of Botany at the Sorbonne
in 1935, and elected a Membre de I'lnstitut in
the same year. Most of his later papers and
those of his students (among whom Eichhorn,
Plantefol, Mangenot and Gautheret should
be mentioned) have been published in his own
journal, the Revue de Cytologie et de Cyto-
physiologie Vegetales. He married Helene
Popovici, at one time his assistant, the
daughter of Professor Popovici of lasi in Rou-
mania. His principal publications are being
listed on pages 229/232 of this volume; cf. also
"Titres et Travaux Scientifiques de M. A.
Guilliermond" (Lyon / Rey, 1921) and Chron.
Bot. 11:128* (1936).

THE

CYTOPLA5M

OF THE PLANT CELL

BY

Alexandre Guilliermond

Professor of Botany at the Soroonne, Pans - ALemhre ae I'lnstitut
AUTHORIZED TRANSLATION FROM. THE UNPUBLISHED FRENCH MANUSCRIPT

hy LENETTE ROGERiS ATKIN50N, Pk D.
Foreword Ly A\^ILLIAM 5EIFRIZ

Professor of Botan;^, University of Pennsylvania

1941
WALTHAM. MASS.. U.S.A.

X uolisJieo by tne UJironica Jjotanica C^ompany

First published MCMXLI

By the Chronica Botanica Company

of Waltham, Mass., U. S. A.

All rights reserved

New York, N. Y.: G. E. Stechert and Co.,
31 East 10th Street.

San Francisco, Cal.: J. W. Stacey, Inc.,
236-238 Flood Building.

Toronto 2.: Wm. Dawson, Ltd.,
70 King St., East.

Buenos Aires: Acme Agency,
Diagonal Norte 567.

Rio de Janeiro: Livraria Kosmos,
Caixa Postal 3481.

London W. 1. : Wm. Dawson & Sons, Ltd.,
83a Marylebone High Street.

Moscow: Mezhdunarodnaja Kniga,
Kusnetzki Most 18.

Tokyo: Maruzen Company, Ltd.,
6 Nihonbashi; P. 0. Box 605.

Calcutta: Macmillan and Co., Ltd.,
294, Bow Bazar Street.

Johannesburg, S. A.: Juta and Company, Ltd.,
43, Gritchard Street.

Sydney: Angus and Robertson, Ltd.,
89 Castlereagh Street.

Made and printed in the U. S. A,

FOREWORD

One cannot read this book by Professor Guilliermond with-
out being profoundly impressed with the thoroughness with which
it has been written. No one has yet presented, nor is any one for
a long time likely to present, so complete and authoritative an
account of the mitochondria story. Most of us have been rather
bewildered by the prolific, detailed, and contradictory literature on
mitochondria, chondriosomes, chondrioconts, chloroplasts, amylo-
plasts, plastids, etc. It is, therefore, a satisfaction to have it all
assembled in a relatively condensed form.

I remember the day, it was a meeting of the Royal Microscopical
Society in London in 1921, when a number of cytologists admitted
that they were at last convinced that mitochondria are not arti-
facts. Guilliermond handles the subject of artifacts very con-
vincingly. He is forced to because much of his work has been on
fixed material and much of it has been subjected to the usual cry
of "artifact".

Guilliermond uses one bit of evidence in support of the reality
of certain cell structures which I should like to apply to another
part of the cell, the existence of which has also been much ques-
tioned, namely, the spindle fibers. Proof of the reality of vesicles
is to be had, says Guilliermond, in the fact that they always
appear at the same stage of development of the cell whatever the
fixative employed. This is likewise true of spindle fibers.

Shimamura has recently offered evidence of the existence of
spindle fibers by showing that if a cell in midmitosis is centrifuged
and then fixed, the spindle with fibers is found to be centrifugally
distorted. Centrifuging cannot distort an artifact before it has
been formed by fixation.

Though supporting the reality of mitochondria. Professor
Guilliermond in no way enthuses unduly over their possible sig-
nificance in the life of the cell, as have some other workers. He
regards as untenable the theory that chondriosomes are symbiotic
organisms. He also discards the idea that chondriosomes are the
means by which every synthesis in the cell takes place. In reject-
ing the theory that chondriosomes are distinct organisms, a concept
based on their superficial resemblance to bacteria and the fact that
both are stained by the same dyes, Guilliermond shows a breadth
of mind characteristic only of the true scholar. He says, that
though the theory is untrue, it gave rise to much good research.
He concludes the section on the function of mitochondria by the

courageous and laconic statement that "Nothing is positively
known about the role of the chondriosomes".

Controversial matters Guilliermond handles with fairness and
dignity. Each problem is dealt with unemotionally. He refers to
SCHLEIDEN "as the promulgator of the cell theory" ; to Robert
HOOKE as the first to recognize the cellular organization of living
things, but the significance of this structure was not understood
for a long time thereafter.

I anticipated a rather orthodox cytological handling of the sub-
ject by Guilliermond, and so was pleased to find him as awake
to the contributions of the newer cytology as to those of the old.
He takes de Jong's concept of cytoplasm as a coacervate and applies
it to chondriosomes; vesiculation indicates that they too are
coacervates.

Guilliermond deals with the question of the physical nature of
the tonoplast. I should like to restate it somewhat differently, and
what I say of it is also true of the outer surface layer of proto-
plasm. All protoplasmic surfaces are probably coated with fats
or other substances which are immiscible with water, but this
does not mean that cell membranes are made of any substance
other than protoplasm. Cell membranes are immiscible with water
because protoplasm is immiscible with water. The immiscibility
of protoplasm in water is not primarily due to an oily surface.
Structural continuity is responsible. Protoplasm takes up water
just as does a sponge, silica gel, or gelatine. Living matter is not
a solution of salts, sugars, and proteins. Protoplasm holds to-
gether. It could not be a living system if it did not do so. Such
misunderstandings have arisen because protoplasm shows certain
properties of liquids, such as rounding up and flowing. To certain
students this can only mean that protoplasm is a liquid, and if it
is a liquid it must be a solution. The colloidal viewpoint clarifies
all this. Protoplasm does flow and therefore it is a liquid, but it
is elastic, possesses tensile strength and contractility, and imbibes,
that is to say, soaks up water. These are the properties of solids,
their presence in protoplasm indicates structural continuity.

Protoplasmic membranes, whether the inner tonoplast or the
outer cell membrane, are of living matter, capable of the same
physical and chemical changes as is the protoplasm which they
bound. The cell membrane is not an inert layer of oil; it is a
dynamic living system.

When amoebae and slime molds move forward, the advancing
surface is in a constant state of change. As the surface increases
in area, material is added from the inner protoplasm, and as it
decreases in area part of its substance is returned to the inner
protoplasm. In short, the membrane is protoplasm. Convincing
evidence of this is an interesting observation which led one of my
students to flatly deny the existence of the tonoplast when it should
have caused him to recognize that protoplasmic membranes are
alive ; that they are dynamic not static systems. He had observed
a particle within the vacuole of a plant cell moving at the same
rate and in the same direction as the streaming protoplasm. Close
observation revealed that not only the mass of protoplasm but its

surface, that in direct contact with the vacuolar sap, was also in
an active state of flow. It was this streaming interfacial proto-
plasm which caused movement of the vacuolar sap and the particle
within it. There was therefore no quiet layer between protoplasm
and sap, which meant to my student that there was no tonoplast.
He failed to appreciate that the tonoplast and cell membranes in
general are not inert skins but a surface layer of living proto-
plasm. The tonoplast is protoplasm ; in that sense is it formed of
a substance immiscible with water. It is, as Hugo de Vries said,
"a membrane differentiated and living".

Much of the misunderstanding in regard to protoplasm arises
from a failure to realize that protoplasm possesses both liquid and
solid properties. Its liquid character is real but superficial. Its
solid qualities are basic. That protoplasm is liquid is evident
from the fact that it flows, but it is in no way comparable to a
solution of salt in water. Viewing protoplasm as a liquid devoid
of solid properties was an excusable fault in the classical cytolo-
gists, but the modern physiologist is aware, as is GuiLLiERMOND,
of the true physical nature of living matter and the need of struc-
tural continuity. GuiLLlERMOND refers to this indirectly by attrib-
uting to protoplasm such properties as torsion, elasticity, and
immiscibility in water. Warren Lewis once expressed the need of
structural continuity in protoplasm when he stated that were it
not for the glutinous qualities, the tackiness of protoplasm, we
should all fall to pieces. In short, protoplasm holds together, and
this is as true of fluid protoplasm as of firm protoplasm.

GuiLLlERMOND concludes this book with the statement that the
future of cytology lies in the union of morphology and physiology.
In return for the admission by a morphologist that anatomy with-
out physiology is sterile, let me say to Professor GuiLLlERMOND
that physiology is meaningless unless supported by structure and
function.

The present volume is the first addition, printed in the New
World, to the list of books which Dr. Frans Verdoorn is editing
and publishing under the title of "A New Series of Plant Science
Books". The book was especially written for the "New Series" by
Professor A. GuiLLlERMOND ; it is the translation of an unpublished
French manuscript and not merely an English version of a previous
book by Guilliermond.

The translator is Mrs. Atkinson of Amherst, Massachusetts
(Ph.D. University of Wisconsin, sometime C.R.B. Fellow in Bot-
any, University of Louvain). Mrs. Atkinson has given us far
more than a translation of the original manuscript, for there was
much interpretation and rearrangement to be done. The translator
is to be congratulated on accomplishing a difficult piece of work
so well. Valuable help in editing the Ms. was also rendered by
Dr. J. DUFRENOY of Louisiana State University.

September 194-1

William Seifriz

CONTENTS

Chapter 1.
INTRODUCTION:- HISTORICAL SKETCH — Difficulties in the

STUDY OF CYTOPLASM. RECOMMENDED METHOD .... 1

Chapter 2.

GENERAL FACTS ON THE STRUCTURE OF THE PLANT
CELL, ITS CYTOPLASM AND MORPHOLOGICAL CON-
STITUENTS:- The CYTOPLASM AND ITS PERMANENT INCLU-
SIONS — The PARAPLASM 5

Chapter 3.

THE PHYSICAL PROPERTIES AND GENERAL CHARAC-
TERISTICS OF THE CYTOPLASM:- ITS APPEARANCE IN
living form — viscosity — rigidity — density — ecto-
plasmic layer — physiological properties of cytoplasm —
The cytoplasm with regard to vital dyes and fixed prepa-
rations 8

Chapter U-
THE CHEMICAL CONSTITUENTS OF CYTOPLASM :-

Proximate analysis of the cytoplasm — The chemical con-
stituents OF cytoplasm. Protides — LiPiDES — Various

products and mineral SUBSTANCES — WATER .... 23

Chapter 5.

PHYSICO-CHEMICAL CONSTITUTION OF THE CYTO-
PLASM:- Electrical characteristics of proteins — Phys-
ical CONSTITUTION OF CYTOPLASM — CELLULAR CONSTANTS AND

EQUILIBRIA — Ionic reaction of cytoplasm — Cellular
rH 29

Chapter 6.
THE PLASTIDS:- The plastids — The chloroplasts in algae

AND BRYOPHYTES — ThE CHLOROPLASTS IN VASCULAR PLANTS:
THEORY OF SCHIMPER — CHEMICAL NATURE AND STRUCTURE OF

PLASTIDS — Movement of chloroplasts 40

Chapter 7.
THE CHONDRIOME:- GENERAL CONCEPTIONS. What is meant

BY CHONDRIOME IN ANIMAL CELLS — THE CHONDRIOME IN PLANT

CELLS — The chondriome in fungi — Development of the

CHONDRIOME — PflYSICAL AND HISTOCHEMICAL CHARACTERISTICS
OF CHONDRIOSOMES 56

Chapter 8.
THE GHONDRIOME (continued):- THE CHONDRIOME AND ITS
DEVELOPMENT IN THE PHANEROGAMS. RELATIONSHIPS BETWEEN
CHONDRIOSOMES AND PLASTIDS. THE FACTS 70

Chapter 9.
THE RELATIONSHIP BETWEEN CHONDRIOSOMES AND
PLASTIDS:- INTERPRETATIONS 85

Chapter 10.
DUALITY OF THE GHONDRIOME:- Theory of the author

HiSTOCHEMICAL AND HISTOPHYSICAL CHARACTERISTICS OF

CHONDRIOSOMES AND PLASTIDS — DEVELOPMENT OF CHONDRIO-
SOMES AND PLASTIDS AMONG THE PLANT GROUPS — PHYLOGEN-
ESIS OF CHONDRIOSOMES AND PLASTIDS 94

Chapter 11.
HYPOTHESES RELATIVE TO THE ROLE OF CHONDRIO-
SOMES AND PLASTIDS 116

Chapter 12.
THE VACUOLES:- Early data. Insufficiency of methods.
Theory of Hugo de Vries 125

Chapter 13.
VITAL STAINING OF THE VACUOLES:- Colloidal sub-
stances in the vacuolar sap — Action of vital dyes on the
cells. Advantages of vital staining 129

Chapter lU.
DEVELOPMENT OF THE VACUOLAR SYSTEM:- First
stages in development — Chondriosome-shaped vacuoles

and CHONDRIOSOMES. CHARACTERISTIC DIFFERENCES — PHYSICAL
CHARACTERISTICS OF THE VACUOLES — CHEMICAL NATURE OF
THE COLLOIDAL SUBSTANCE OF VACUOLES — VACUOLAR pYL AND

rH 146

Chapter 15.
ORIGIN AND SIGNIFICANCE OF THE VACUOLES :-

Aleurone grains : their formation — Reversibility of form

IN THE VACUOLAR SYSTEM — THE PHENOMENON OF VACUOLAR
CONTRACTION — ORIGIN OF VACUOLES — THE PRESENCE IN
SOME CELLS OF SEVERAL DISTINCT CATEGORIES OF VACUOLES — DI-
GESTIVE VACUOLES 170

Chapter 16.
THE ROLE OF THE VACUOLAR SYSTEM AND HYPO-
THESES CONCERNING IT 188

Chapter 17.
GOLGI APPARATUS, CANALICULI OF HOLMGREN AND
OTHER CYTOPLASMIC FORMATIONS:- GOLGI APPARA-
tus and the canaliculi of holmgren in animal cells —
Possible relationships of the vacuolar system with the
apparatus of golgi and of holmgren — relationships be-
tween the golgi apparatus and the chondriosomes and

PLASTIDS — The so-called GOLGI apparatus IN PLANT CELLS

— Other cytoplasmic formations 191

Chapter 18.

LIPIDE GRANULES, MICROSOMES AND OTHER META-
BOLIC PRODUCTS:- Fatty degeneration — Other meta-
bolic PRODUCTS 202

Chapter 19.

CYTOPLASMIC ALTERATIONS:- ALTERATIONS PRODUCED IN

dying cells — Alterations produced by various physical
AGENTS — Alterations produced by parasites .... 208

Chapter 20.
SUMMARY AND CONCLUSIONS:- Cytoplasm — The chon-

DRIOME — The PLASTIDS — VACUOLAR SYSTEM OR VACUOME 214

BIBLIOGRAPHY 224

AUTHOR INDEX 244

INDEX OF PLANT AND ANIMAL NAMES 246

/ •

Chapter I

INTRODUCTION

Historical sketch:- It is well known that the cellular organiza-
tion of living things was first recognized by the English engineer,
Robert Hooke (1665), who, toward the end of the seventeenth
century, was looking at sections of cork with a view to finding out
what applications could be made of the recent discovery of the
microscope. He described the tissue as formed of alveoli resem-
bling a honey-comb. These alveoli he called cells. For a long
time, however, the significance of this structure was not under-
stood. The French botanist, Brisseau de Mirbel (1833) thought
that cellular tissue was composed of vacuoles hollowed out of a
homogeneous substance, which corresponded to living matter. It
was MOLDENHAWER (1812) who, for the first time, demonstrated
the individuality of cells. Having suxiceeded in separating the
elements of tissues by maceration, he proved that cells have a wall
of their own and cannot, therefore, arise as cavities in a homo-
geneous substance. Later DUTROCHET (1824), TURPIN (1827) and
Meyen (1830) considered cells as moi'phological entities but their
attention had been centered rather on the walls than on the con-
tents of the cells. In 1830 Meyen had discovered chlorophyll
grains, starch grains and crystals within the cavity of the cell.
In 1831 the English botanist, Robert Brown, who gave his name
to Brownian movements, discovered the nucleus in the epidermal
cells of orchids. Shortly after (1838), Schleiden, the promul-
gator of the cell theory, attributed predominating importance to
the nucleus to which he gave the name Cytoblast and which he
considered as the generator of the cell. According to Schleiden,
a cell is formed as follows: in a matrix, the cytoblastema, there
appears the cytoblast on whose surface a membrane then becomes
diflferentiated which lifts itself up like a watch crystal, grows, and
bursts away from the cytoblast, leaving an empty space into which
the matrix penetrates by filtration.

DuJARDiN (1835), in the cells of Infusoria, first accurately
described living matter, to which he gave the name sarcode.
Nageli (1866) perceived in addition that plant cells are occupied
by nitrogenous matter and voN Mohl at the same time described
it under the name of Protoplasma and attributed to it a primary
importance. According to this observer, the plant cell, made up
of the protoplasm, contains a nitrogenous primordial utricle, lining
its wall on the inside and enclosing the nucleus. This primordial
utricle is the seat of special movements already seen by B. CORTI
(1772) and Treviranus (1807). The rest of the cell is occupied
by cell sap.

Guilliermond - Atkinson — 2 — Cytoplasm

A little later, Cohn (1850), Thuret (1850) and Pringsheim
(1854) perceived that the zoospores of the algae lack a wall and
are made up exclusively of protoplasm. Max Schultze and DE
Bary (1859) finally established the fact, once for all, that the
protoplasm of plants is the essential substance of cells and corre-
sponds to the sarcode of Dujardin. Leydig (1857) defined the cell
as "a mass of protoplasm furnished with a nucleus". Max
Schultze (1861) defined it as "a mass or lump of protoplasm
endowed with vital properties".

While these conceptions were being established, the works of
VON MoHL, Meyen and Nageli were proving the inexactitude of
Schleiden's theory of cell origin and showing that cells multiply
by division. Thus the word cell (French cellule, from the Latin
cellula, little room) came to mean the contents of the cellular cav-
ity, a significance quite different from that given it by HOOKE, who
observed only the walls.

From then on, the conception of the cell was enlarged upon
but the knowledge of its structure, making only slow progress, was
to remain obscure for a long time. The early cytologists observed
only living cells. This presents serious diflflculties, for, with the
exception of unicellular organisms, observation of living material
can be carried out only after tearing or sectioning tissue, operations
which risk injuring the cells. Cells examined in a medium not their
own — water, for instance — may, during observation, undergo
serious alterations. Lastly, observation of living material never
allows the study of cell structure to be pushed sufficiently far, be-
cause, except for cases which are unusually favorable, the
different elements which constitute the cell show too small differ-
ences of refractivity for it to be possible to distinguish one from
the other with clearness, and this becomes, moreover, well nigh
impossible in embryonic tissue in which the cells are very small.

The introduction of the paraffin method on objects pre-
viously "fixed" has greatly facilitated the work of cytologists.
This method consists in fixing, i.e., coagulating, the cells by means
of various chemical reagents, then embedding the tissues in paraflSn,
cutting them in thin sections with the aid of a microtome and
finally staining them. Several stains may then be applied to the
sections, which fix this constituent or that, according to its affinity
for the stain, and superb preparations may be obtained and made
permanent in Canada balsam. But this method, convenient as it
is, has, nevertheless, the serious difficulty of reducing cytologists
to the study of dead cells only. Fixation, i.e., coagulation, of proto-
plasm, modifies the structure of the cell and exposes cytologists to
serious errors in interpretation. Finally, this method does not
allow the physical character or biological properties of the cyto-
plasm to be studied, although they are very important, and cytology
is reduced to pure cellular morphology. The use of the paraffin
method has led to the striking discovery of karyokinesis and has
rendered very great service in the study of the nucleus and, in par-

Chapter I — 3 — Introduction

ticular, of the chromosomes, whose form in the fixed material is
preserved with a minimum of distortion.

On the other hand, as far as the cytoplasm is concerned, the
paraffin method has, over an extremely long period, given only
mediocre results and cytologists have made the mistake of neglect-
ing too greatly the study of living material. As a matter of fact,
the most convincing data on the cytoplasm were for a long time
obtained exclusively from living cells: on the one hand, the
excellent discoveries of SCHMITZ concerning the chloroplasts
of the algae and those of WiLHELM ScHiMPER and Arthur Mayer
in the study of plastids of higher plants, and, on the other hand,
the classical experiments of Hugo de Vries on the vacuoles and
their role in the osmotic phenomena of the cell.

It is only since 1910 that knowledge of the cytoplasm has made
rapid progress. At that date, the timely appearance of the ultra-
microscope enabled A. Mayer and Schaeffer to make known some
essential data on the colloidal nature of the cytoplasm of animal
cells — data which can be applied to plant cells. At about the same
time, the introduction of mitochondrial viethods, new fixation tech-
niques for certain cytoplasmic lipides, led to the discovery of the
chondriosomes and made it possible to preserve the plastids, to
stain them clearly and to follow them through their whole life
history. A little later, the methodical use of vital dyes, which
accumulate in the vacuoles of living cells, made it possible to follow
the evolution of the vacuoles through all the stages of cellular
development. Finally, the invention of the micromanipulator and
of motion-picture photography, and the perfecting of methods of
observation in vivo, have contributed in large part to making
known to us the physical properties of the cytoplasm and of its
various morphological elements.

Difficulties in the study of cytoplasm. Recommended method :-

As a result of work carried on during the last twenty years by
means of the mitochondrial technique, or with the aid of the ultra-
microscope, or by the use of vital dyes, it has been possible to solve
definitely the problem of cytoplasmic structure. The discoveries in
this domain are still too recent to be accepted by all cytologists.
But if these are still being discussed, it is simply because they have
been arrived at by methods very different from those usually em-
ployed and because some cytologists are unable to verify them by
their own methods. The study of the cytoplasm, a substance which
is very easily injured, is infinitely more difficult than the study of
the nucleus. It is necessary to use a long and delicate method
whose general outline will be indicated here.

This method, which will be called the analytical method, i.e.,
analysis of the cell, consists of the following serial operations:
First, this method requires the use of fixed and stained sections
which are indispensable in bringing out the elements which are
considered as entering into the constitution of the cytoplasm. But
this method is always insufficient, for once these elements have been

Guilliermond - Atk inson —4— Cytoplasm

brought out by fixation and staining, their actual presence must
be checked by observation of living material in such cells as lend
themselves to it. This observation of living material permits a
proper appreciation of the value of the above methods. As the
very handling of living material runs the risk of producing altera-
tions, every precaution to avoid them must always be taken.
Fungi may be used which can be observed in their own environ-
ment, or organs studied which are sufficiently transparent to permit
their cells to be examined without any manipulation. Petri dishes
of a special type (used at the International Bureau for the Culture
of Fungi at Baarn) can be used if necessary. In the bottom of these
is an opening of 3 cm. which can be covered with a slide sealed
with asphalt cement (Fig. 88). In this way, seeds can be grown
aseptically and their roots during development can be observed with
an oil-immersion lens by turning the dish under the microscope.
Vital dyes must be used to enable us to follow such elements as the
vacuoles which are not well preserved by any other means. More-
over, each element whose presence has been recognized by fixation
and staining must be described by means of a systematic study of its
behavior with most fixatives and stains ; this study must be based
upon histochemical reactions of the element in question. This
method, called histochemical analysis, permits us to characterize
the element, to distinguish it from others, and to inform ourselves
as to its chemical nature. It permits us, besides, to distinguish
some elements which are revealed only by certain methods of fixa-
tion and staining. Histochemical analysis must be followed by an
histophysical analysis, i.e., analysis of the physical state of the
element, its viscosity, colloidal state, and so forth, an analysis in
which will be employed the micromanipulator, the ultramicroscope,
the polarizing microscope, the centrifuge, the plasmolytic method
and, if need be, motion-picture photography. This histophysical
analysis will supplement the description obtained by histochemical
analysis. Then, too, the development of the element throughout
the entire life of the cell must be followed in order to know whether
it constitutes a permanent element or is only transitory, and in
order to obtain an idea as to its significance by observing its be-
havior. Finally, it will be useful to supplement this method called
the analysis of the cell, which comprises the various operations just
enumerated and which has, itself, all the value of an experimental
method, by a further series of experiments designed to clarify the
role of the element studied. Among these experiments are vivisec-
tion, depriving a cell of an individual element by means of the
micromanipulator, and a study of nutritional influences on the be-
havior of the element under consideration. This is the cyto-
physiological method which we will turn to only occasionally here,
our aim being chiefly the morphological study of the cytoplasm.
The procedure specified is the only one by which precise facts may
be obtained on the morphological constituents of the cj^oplasm : the
plastids, chondriosomes and vacuoles.

Chapter II

GENERAL FACTS ON THE STRUCTURE OF

THE PLANT CELL, ITS CYTOPLASM AND

MORPHOLOGICAL CONSTITUENTS

The cytoplasm and its permanent inclusions:- In agreement with
Strasburger and Henneguy, the term cytoplasm^ is here under-
stood to mean all the living matter in the cell with the exception
of the nucleus. By the term protoplasm is meant all living matter
in the cell, i.e., both cytoplasm and nucleus. The cytoplasm occurs
in living cells as a colloidal substance, hyalin and homogeneous,
elastic and of a viscosity which is always superior to that of water.
This substance holds permanently in suspension a certain number
of small elements which resemble bacteria in form and dimensions
and are distinguished in living cells by a ref ractivity slightly higher
than that of the cytoplasm. These elements, which are called
chondriosomes, appear in the form of granules, rods and threads.

In addition to these elements there are, in green plants, the
plastids, whose form is very variable in the algae and which, in
higher plants, appear in green tissue as large globules, filled with
chlorophyll, which are derived from small elements very similar to
chondriosomes in form and histochemical constitution. The cyto-
plasm also contains small fluid cavities called vacuoles, composed of
water, containing crystalloid and colloidal substances. These vacu-
oles, which are very small and very numerous in young cells, swell
and usually run together little by little, to form, in mature cells,

1 Von Mohl designated under the name of protoplasm, the living substance of the cell,
i.e., the nucleus and that which in the present volume is called the cytoplasm. But Strasburger,
Henneguy. and most of the modern cytologists have reserved the term protoplasm for the
cell contents (with the exception of the wall), including the nucleus. Within the protoplasm
they distinguish the nucleoplasm or karyoplasm which is the nuclear substance, and the
cytoplasm, which is everything else. The term protoplasm, however, is often used as a
synonym for that which, in this volume, is termed cytoplasm. Hardy means by cytoplasm
all that is not nuclear and he considers the protoplasm to be all living matter within the
cytoplasm to the exclusion of the nucleus: the cytoplasm therefore includes the protoplasm
and the products of its activity, various inelusions not pertaining to the living substance.
BOTTAZZI, on the contrary, incorporated the nucleus into the protoplasm or bioplasm which, for
him, includes all the living substance of the cell, i.e.. the nucleus, the living ground substance,
which is not nuclear, and the chondriome. He reserves the term cytoplasm for all the cell
contents, i.e.. the total protoplasm and all the products elaborated by it. Among these last.
BOTTAZZI distinguishes: 1. the metaplasm including the products of cellular elaboration which
are permanent (cell walla) ; 2. the paraplasm, represented by the reserve substances or waste
products, which are only transitory in the cytoplasm (starch grains, fat globules, etc.), as
well as soluble substances formed by cellular metabolism (glycogen, inulin) which may be
detected by certain chemical reagents. Thus, for Bottazzi.' protoplasm means all that is
living in the cell including the nucleus and the chondriome. whereas the cytoplasm includes
the protoplasm and the products of its elaboration (metaplasm and paraplasm).

Guilliermond - Atkinson

Cytoplasm

an enormous single vacuole which occupies the greater part of the
cell, forcing the nucleus to the periphery. This vacuole is often
traversed by thin cytoplasmic trabeculae which radiate from the
nucleus to join the parietal cytoplasm.

Lipide granules are encountered in almost all, if not in all,
cells. They are scattered in the cytoplasm in more or less consid-
erable numbers according to the cells and their stage in develop-
ment. One also finds, moreover, various inclusions in the cyto-
plasm: reserve or by-products formed during cellular activity
(starch grains, crystalline proteins, various crystals, etc.) which
are localized either in the cytoplasm itself or in the plastids or in
the vacuoles. All these substances, however, are only transitory
products formed during cytoplasmic activity.

The paraplasm:- We now turn to a new consideration. It has

just been seen that in the cytoplasm
there are in suspension some elements
which are always present such as plas-
tids, chondriosomes, vacuoles and lipide
granules. Among these elements, a
distinction must be made between those
wiiich can be considered as belonging to
the living substance, to the architecture
of the cell, and those which simply re-
sult from its activity. Among the lat-
ter whose chemical composition is more
simple, there are some, like the starch
grains, which form in the plastids,
others, like many crystals, which are lo-
calized in the vacuoles, others still, like
the crystalline proteins, which are con-
tained in the cytoplasm itself. Now, the
cytoplasm is a substance which continues to exist permanently. It
presents a chemical composition still not well known but which
does not vary appreciably. It is a living substance. The plastids
are elements which are never formed de novo but, like the nucleus,
are transmitted by division from cell to cell. They, therefore, may
also be considered as living substance. The chondriosomes seem to
be of like nature. Such is not the case for the vacuoles. Although
always present in the cells, they appear to arise de novo and to dis-
appear only to be replaced by new ones. Furthermore, they enclose
substances which, all of them, are products of cytoplasmic activity :
reserve products, or waste products, or transitory products of
metabolism. They do not, therefore, seem to belong to the living
substances of the cell. This is true for the lipide granules
which are present in almost all cells, but which vary greatly in
quantity depending on the state of development of the cells : there
are cells which contain only a few, others in which the granules
accumulate in great quantity and fuse to form large globules filling
up the cytoplasm. These are also products of metabolism. Co-
existent with these visible products, there are others, like glycogen,

Fig. 1. — Diagram of a plant
cell. C, chondriosome. Gg, lipide
granules. N, nucleus. P, chloro-
plasts. V, vacuole.

Chapter II

7 —

General Facts

which impregnate the cytoplasm and can be detected only by using
certain microchemical reagents. All these products arise by cjrto-
plasmic activity and a distinction must therefore be made between
them and the cytoplasm by which they were pro-
duced. These products are grouped under the
term pm'aplasm or deutoplasm and are separated
from the cytoplasm, which, with the plastids and
chondriosomes, constitute living substance. Among
the products resulting from the activity of the
cytoplasm are some which are of permanent char-
acter, such as the cellulose wall. This is a cyto-
plasmic secretion which persists during the entire
life of the cells and can not, it would seem, be
considered as belonging to living substance.
These permanent formations of the paraplasm or
deutoplasm are specified as metaplasm. Lastly,
the name protoplast is used to designate all the
contents of the cell except the cell wall, i.e., the
protoplasm and paraplasm together.

In the cytoplasm, then, we shall have to con-
sider the cytoplasm itself, the plastids, the chon-
driosomes and the paraplasm, the most important
constituents of the last category being the vacu-
oles and the lipide granules.

These, and other paraplasmic formations not listed above, will
be studied in the succeeding chapters.

Fig. 2. — Semidia-
grammatic represen-
tation of a living
epidermal cell of Al-
lium. Cepa showing
cytoplasmic trabecu-
lae traversing the
vacuole and uniting
the parietal cytoplasm
with the nucleus.

Chapter III

THE PHYSICAL PROPERTIES AND GENERAL
CHARACTERISTICS OF THE CYTOPLASM

Its appearance in living form:- DUJARDIN, who first studied the
cytoplasm in living cells of ciliated Infusoria, has given an abso-
lutely exact description of this substance which he calls sarcode,
a description which modern observations merely confirm. "This
substance", he says, "appears perfectly homogeneous, elastic and
contractile, diaphanous and refracting light a little more than
water and much less than oil. One can distinguish in it absolutely
no trace of organization : neither fibre nor membrane nor an appear-
ance of cellular form".

There is nothing to add to this descrip-
tion. The cytoplasm appears to modern
observers just as it was described by Du-
JARDIN. In living cells, it appears to be a
homogeneous substance, as transparent as
glass, viscous, a little more refractive than
water and non-miscible with it. As has
been already stated, modern research has
shown that it does, however, always con-
tain in suspension numerous granules
(chondriosomes, lipide granules) and
vacuoles, of which more will be said later.
Hence the description of Dujardin can
be applied only to the cytoplasm itself,
omitting these elements which it contains.
For the study of the physical proper-
ties of the cytoplasm, the Plasmodium of
the Myxomycetes has been much used. It
is seen as a voluminous protoplasmic mass
of irregular appearance, lobed in the most fanciful manner and
enclosing numerous nuclei. This protoplasmic mass changes shape
constantly by virtue of the amoeboid movements which control its
displacement. It glides along the surface of its support and if this
latter be of decaying wood, it worms its way into the interior of
the wood, penetrates it only to come out again further on, then to
re-enter it, and so on. The huge dimensions of the Plasmodium
make it a valuable object for the study of the cytoplasm. The
classical experiments successfully perfoiTned by Pfeffer on the
Plasmodium of Chondrioderma difforme have shown that in order
to alter a small portion of cytoplasm, it is necessary to exert a
pressure of 8 mg. per sq. cm. The cytoplasmic strands of this
Plasmodium break when subjected to a tension of 120-130 mg. per
mm. This indicates a rather strong cohesion which, it may be
said, can be even stronger in other types of cytoplasm.

Fig. 3. — Fragment of a
Plasmodium of Chondrioderma
diffortne. A, ingested foreign
body. X about 50. (After
ZOPF) .

Chapter III — 9 — Physical Properties

Viscosity:- There has been a good deal of discussion concerning
the cohesive state of cytoplasm, i.e., its consistency. Most cytolo-
gists consider that the cytoplasm more nearly approaches a liquid
than a solid state. A few, however, believe it to be of a solid
consistency.

The Plasmodium of the Myxomycetes is very fluid. A proof of
this is in an experiment carried out on Badhamia utricularis by
the English mycologist, Lister. Lister noticed Plasmodia of this
fungus on the trunk of an old hornbeam growing in his garden.
The trunk was covered over with the fruiting bodies of Corticium
puteanum. The Plasmodia of Badhamia moved around on the sur-
face occupied by Corticium, actually consuming the fruiting bodies
and after their passage leaving the bark of the hornbeam as smooth
and clean as if no fungus had ever grown there. But, although
Badhamia assimilated the Corticium tissues, its spores, protected by
a resistant brown membrane, were not attacked and accumulated
within the Plasmodium which
took on a dark brown color. Lis-
ter collected one of these Plas-
modia on a glass plate, where it
moved about, leaving behind it as
evidence of its passage, a fine
brown network, formed of the
ingested spores. These had been
progressively dropped, being
poorly retained in the plasmodial
cytoplasm. This demonstrates

its weak viscosity. The plasmo-

dium, however, still enclosed fig. 4. - Two successive shapes, a. b,

many spores. Lister then put in ^^^^^ ^y the Plasmodium of a Myxomycete

•1 ,1 1 . « , ■ . as it moves in the direction of the arrow.

Its path a barrier of wet cotton.

The Plasmodium passed through rapidly, leaving in the cotton all the
remaining spores, and emerged showing the yellow tint characteris-
tic of it before it had taken up the fungus Corticium. Lister thus
brought about the filtration through cotton of plasmodial protoplasm
and in this way succeeded in demonstrating its very fluid consistency.
This somewhat crude evidence may be made more specific by
a detailed examination of the Plasmodium. It is composed of a
network of opaque, anastomosing veins which, in certain species,
may attain large dimensions: even a diameter of several milli-
meters in the case of the principal veins. Toward one border, the
network merges with a fan-shaped continuous layer of the same
material. This entire body moves at about the rate of 1 cm. per
hour and presents constantly changing contours ; nevertheless the
continuous layer is always in front and may be interpreted as
protoplasm flowing slowly over the substratum. The substance
composing the Plasmodium is, in fact, fluid and motile but shows at
its surface the ability to coagulate. A puncture in a vein allows
a drop of the internal fluid protoplasm to escape. This drop is
immediately coagulated on the surface. The internal protoplasm,

GuilHermond - Atkinson — 10 — Cytoplasm

then, under its immobile surface, flows with extreme rapidity. In
addition, the different veins and the flowing, continuous layers com-
prising the Plasmodium, show very curious, rhythmic but not syn-
chronous, pulsations. These are rendered quite visible by motion-
picture photography. Each vein shows an alternation of systole
and diastole ; each outer layer flows by jerks over the substratum,
becoming alternately thicker and thinner (CoMANDON and PiNOY,
Seifriz).

Under sufficiently high magnification the Plasmodium seems to
be formed of an homogeneous substance, holding in suspension in-
numerable granules, in particular, lipide droplets and ingested
debris, whose incessant displacement in different directions reveals
with great clearness the existence of cytoplasmic currents.

These currents are very irregular. They may be rapid or may
even rush along in a vein and, at a given instant,
V\ /# they will immediately slow up, then change direc-

^ " tion, accelerate, retard again, return to the orig-

inal direction and so on. In each vein the same
irregularities are observed but without any syn-
chronism whatever. Thus, the movement of the
entire mass of the Plasmodium in one direction
expresses the sum of all these movements in dif-
ferent directions and indicates that in this appar-
ent disorder, the protoplasm flows more in one
direction than in the other ; namely, in the direc-
tion of the advance of the organism.

Now these characteristics of protoplasm have
nothing unusual about them. Microscopical ex-
of the diff^enrdire™ aminatious of the contents of the most varied cells
tions taken by proto- reveal that there, too, protoplasm behaves as a

plasmic currents in a ^ . , , , mi it i • i -x i i i

portion of the vein- fluid substance. The bodies which it holds sus-
piTsmodium.'"" °^ ^ pended in it are in most cases more or less rapidly

carried along in its multiple currents. Here, as in
the extended body of the Plasmodium, these currents run side by
side, separated by calm borders. They ramify, anastomose and
change constantly. These are the phenomena of cyclosis for which
the cells of Elodea canadensis and the staminate hairs of Tradescan-
tia or of Celandine are classic subjects for observation. These
phenomena are found again in Spirogyra in which they appear
with great distinctness and of which more will be said later on.

All these facts indicate, therefore, that cytoplasm flows. This
presupposes a mobility of molecules found only in a liquid state,
which, in a word, is the essential property of a liquid state.
These movements are, however, much less accentuated in some
cases. They seem to be nearly absent in certain plant cells, among
others in yeasts and various fungi, in which the cytoplasm appears
to be much less fluid and shows no displacement of granules. This
seems also to be the case in most animal cells.

Other arguments resting especially on observations of plant
cells have been brought forward in favor of the fluid state of cyto-

Chapter III

— 11 —

Physical Properties

lf.^/

plasm. HOFMEISTER and Berthold showed for the first time that
if a section is cut out of a filament of Vaucheria, a part of the cyto-
plasm comes out, taking a spherical form by virtue of the law of
surface tension which characterizes liquids (Fig. 7). Since that
time, many similar experiments have been performed, notably by
Strugger on Chara cells. And finally, an experiment by KUHNE on
the cytoplasm of the Myxomycetes, can be explained only by a
liquid state of the cytoplasm. This worker succeeded in obtaining
an artificial muscle by enclosing in elastic tubes, fragments of the
Plasmodium (intestines of Hydrophilus piceus).

All these facts, added to those obtained by microdissection,
which will be spoken of further on, permit us to
conclude that the cytoplasm possesses, in general,
the properties fundamental to liquids: it flows
and has a surface tension which tends to make it
take the form of minimum surface, i.e., spherical
form.

Research carried out in these later years with
the aid of the microdissector, put into practice by
Chambers, has made great progress in the knowl-
edge of the viscosity of the cytoplasm. This
method consists in the use of a special instrument,
the micromanipulator, or microdissector, provided
with glass needles which can be moved mechan-
ically with great precision. This permits the
dissection of cells under high magnifications.

The work of Seifriz with the micromanipu-
lator on various plants (Mucoraceae, Fucus, pol-
len tubes, Plasmodia of Myxomycetes) also dem-
onstrates that the cytoplasm presents a consist-
ency which is very variable: now very fluid and
almost like water, now almost solid, even to the
state of a gel with a consistency of bread dough
or vaseline, in which the passage of the needle
leaves a gaping hole.

The cytoplasm is very fluid in the Plasmodium
of the Myxomycetes as long as the latter is in an
active state. If the point of a needle of the micro-
manipulator is broken off in the Plasmodium, the cytoplasm is
rapidly aspirated (c/. p. 37). Its fluidity is, however, always
greater than that of water, intermediate between that of water and
that of oil of paraffin, but from the moment that the Plasmodium
ceases growth, i.e., in the stages which precede sporulation, the
consistency of the cytoplasm increases: it equals that of oil of
paraffin, then that of glycerin and, finally, that of bread dough. In
Rhizopus nigricans, the cytoplasm is also liquid in the young por-
tions of the hypha. It is, nevertheless, less fluid than that of the
Myxomycetes and presents approximately the consistency of oil of
vaseline. On the contrary, in the older portions of the hypha of
Rhizopus, the cytoplasm grows thick. Its viscosity becomes that

//

in

Fig. 6. — Cyclosis
a staminate hair
of Chelidonium. The
arrows indicate the
direction of the cur-
rents in the various
cytoplasmic meshes.
n, nucleus. (After
VAN TiEGHEM and

COSTANTIN).

Guilliermond - Atkinson — 12 — Cytoplasm

of glycerin and then that of bread dough and may even attain and
exceed that of vaseline.

In the egg of Fucus the cytoplasm becomes more and more
viscous as the egg matures and its exchanges diminish. Immedi-
ately after fertilization it again becomes liquid and remains in
this state in the embryo. Areas seem to exist in the egg, there-
fore, where fluidity is more accentuated, — sort of centers of activity
where different chemical exchanges are produced.

It appears from all these observations then, that the viscosity
of cytoplasm is always superior to that of water and in numerous
cases as great as that of blood. Yet in certain cells, especially in
old organs, the viscosity can be much higher. Moreover in dehy-
drated organs, seeds, for instance, the cytoplasm can be more or
less solid.

Many attempts have been made to measure directly the viscosity
of the cytoplasm inside the cells. The easiest method to employ is

that which makes use of the law for
falling spheres established by the
physicist Stokes. The cytoplasm of
certain cells encloses starch grains
which are more dense than the cyto-
plasm and which have a tendency,
because of their weight, to fall
through the cytoplasm to the lowest
point in the cell. By using an hori-
FiG. 7. — Ruptured siphon of zontal microscope, the time required

Vaucheria showing tendency of lobed n ,■, • j^j?ii'ii • i

extruded contents to break up into for the grams to fall IS detcrmmcd

globular drops. At left, one droplet ^nd, by applying the physicist's form-
greatly magnified. (After VAN Tie- , ' ,, ^. "^ ., « j, . i

GHEM and cosTANTiN). ulac, thc viscosity oi the cytoplasm is

calculated.

By this process of determining the speed with which starch
grains and certain crystals fall through the cytoplasm, Weber was
able to state that the viscosity of the cytoplasm varies with the
cell under consideration. It seems to increase with the age of
the cells. Certain influences (increase in temperature, the action
of certain chemical substances, such as narcotics) can also cause
variations.

Heilbronn has employed the following method : an iron needle,
very fine but too heavy to be carried about in the cytoplasmic cur-
rents, is introduced into a Plasmodium. The needle being located
under the microscope, an electro-magnet in which flows a current
of increasing intensity, is brought near the preparation. When the
needle in the field of the microscope begins to orient itself in the
direction of the lines of force, the experiment is stopped. The
intensity of the current necessary to produce a visible deviation of
the needle is divided by the intensity producing the same effect on
the same needle plunged into pure water. With the viscosity of
water thus taken as a standard, the quotient measures the viscosity
of plasmodial protoplasm.

Chapter III — 13 — Physical Properties

The centrifuge has also been used on living cells. This brings
about a displacement of heavy inclusions, such as starch grains,
or of light inclusions, such as lipide granules, with a speed
and facility varying with the consistency of the cytoplasm. This
method is much less precise since the exact densities of the gran-
ules and cytoplasm are not exactly known.

These various methods have shown that the most fluid cyto-
plasm has a viscosity which is only 3-5 times that of water and
that the most dense cytoplasm (that of animal cells) reaches nearly
10,000 times the viscosity of water. Thus from the results obtained
by means of centrifuging, microdissection and other techniques, an
essential and very general conclusion may be drawn : the cytoplasm
of plants does not present at all times the same viscosity, for this
varies essentially with the physiological state of the organ under
consideration.

Rigidity:- The cytoplasm possesses, at the same time, a property
which, as will be seen later, is tied up with its physical state and
which is characteristic of cells, namely, a certain rigidity which
gives to it an elasticity of torsion. Rigidity expresses the physical
ties between the particles of the system in question, ties which are
lacking in true liquids. This rigidity can be brought out by micro-
manipulation. Thus, in displacing cellular inclusions within the
cell, it was observed that sometimes these return to their places
when pressure is released, the rigid surroundings acting as a spring,
and sometimes the inclusions are displaced as from a liquid without
rigidity. The stability or instability of form of these inclusions
after being deformed, for instance being drawn out between two
needles, also teaches us something of their rigidity and their varia-
tions. In this way, Scarth, for example, demonstrated the elastic-
ity of the cytoplasm of Spirogyra. The nucleus of one cell of this
alga, when pushed by a microneedle from one side of the cell to the
other and then left alone, was observed to return of itself to its
original position. Like viscosity, rigidity seems to be variable in
the cell. Microneedles sometimes penetrate very easily into a fluid
cytoplasm without reaction, and sometimes with difficulty into a
thick gell. There is no method for measuring the rigidity of the
cytoplasm.

Density:- By a micropycnometric method, Leontjew succeeded
in measuring the density of the Plasmodium of Fuligo septica and
has shown it to be, on the average, 1.040 for individuals collected in
dry weather while it does not surpass 1.016 for those collected in
wet weather, but eleven hours after sporulation it rises to 1.065.

Ectoplasmic layer:- The cytoplasm is not miscible with its ex-
ternal surroundings but remains always very sharply separated
from them. In cells with no skeletal walls, the cytoplasm is sur-
rounded by an external zone presenting a consistency greater than
that of its central part. This zone is called ectoplasm, or ectopias-

Gullliermond - Atkinson

14

Cytoplasm

mic layer, or plasmalemma, in opposition to the rest of the cyto-
plasm which is designated as endoplasm. This is the only mem-
brane which exists in the Plasmodium of the Myxomycetes, in the
Myxamoebae, as well as in various zoospores and spermatozoids
of the algae and fungi. In other material, especially in the lower
organisms, the equilibrium forms of the protoplasm when in the
presence of water consist of lobes or pseudopodia, irregular and
changing, blunt or spiny, perhaps in accordance with temporary
and local modifications of surface tension. There again everything
goes on as if there were an elastic layer around the cytoplasm.
Figure 8 represents the division into two tiny plasmodia of
Vampyrella. The individuals in the process of separation remain
for a long time united by a protoplasmic strand which becomes
increasingly thin. Then suddenly it breaks in the middle and the

Fig. 8. — Division in Vampyrella. ps, pseudopodia.

two points liberated by the rupture go back into the plasmodia to
which they belong as if pulled back in by the contraction of an
elastic membrane. In all other cells it is recognized that an anal-
ogous membrane lines the cellulose wall. When a plant cell is
plasmolyzed, for example an epidermal cell of Iris germanica, the
cytoplasm leaves the cellulose wall and contracts in the middle of
the cellular cavity to a perfectly delimited spherical mass as if it
possessed a delicate external membrane.

Following work by de Vries on the osmotic phenomena
of plant cells, there was often given to the region of demarcation
between the cytoplasm and the surrounding environment, the value
of a differentiated membrane to which the semi-permeability of the
cells has been attributed. In no case, however, can this membrane
be detected morphologically by any kind of staining. Therefore
its existence as a differentiated membrane is purely theoretical

Chapter III — 15 — Physical Properties

(Chodat and Boubier, Lepeschkin). Furthermore, certain ex-
periments are difficult to reconcile with the existence of such a
membrane. It has been known for a long time that when a filament
of Vaucheria is injured in such a way that the torn cellulose wall
allows portions of cytoplasm to run out, those coming in contact
with water immediately assume spherical form (Fig. 7). They
are seen at the same time to become enveloped in a delicate mem-
brane separating them from the surrounding substance.

The microdissection experiments of Chambers on animal cells
have demonstrated that the cytoplasm always shows at its periph-
ery a more refractive zone, resulting from a condensation of its
substance, constituting a sort of membrane. When trying to insert
a microneedle in a living cell, it is noticed that, if the pressure is
not too great, there is formed a depression in the cytoplasm at the
point of contact as if an elastic membrane separated it from the
instrument. But this membrane does not have an autonomous exist-
ence. If it is destroyed over a small area, a membrane immediately
separates by condensation from the cytoplasm at the wounded spot.
By this same method, Seifriz has demonstrated at the periphery
of the Plasmodium of the Myxomycetes a very elastic membrane,
capable of stretching and contracting with the greatest ease, and
resistant to a remarkable degree, in short, a limiting layer of more
dense cytoplasm. Whenever it is torn, a new one forms immedi-
ately, as long as the cytoplasm is alive, whether the surrounding
substance be air or water. The same phenomenon is observed
in pollen tubes. If the tube is torn, it is seen that the cytoplasmic
masses which are detached from it are surrounded forthwith by a
membrane. From these experiments it can be concluded that each
fragment of naked cytoplasm possesses the property of separating
itself immediately, by a delicate membrane, from the substance
surrounding it. It is comprehensible that the cytoplasm may be
finely divided, being at all times surrounded by a delicate mem-
brane, as, for example, in the Plasmodium of Badhamia which
passed through the cotton wad. It seems demonstrated, further-
more, that this membrane is not, as was first thought, a differenti-
ated membrane of special constitution, but a limiting surface, a
sort of physical membrane, brought about by a condensation of the
cytoplasm in its zone of contact with the surrounding environment.

The manner in which this membrane is formed can be com-
pared with the well known fact that in a complex liquid all mole-
cules capable of lowering the surface tension accumulate in the
peripheral layer. Now, proteins lowering surface tension and con-
centrating at the surface of their aqueous solution constitute vis-
cous films which reform immediately when torn. It even happens
that this concentration at the surface exceeds the maximum solu-
bility of the body in question. It then becomes flocculent and may
form a solid surface layer quite comparable to the ectoplasmic
membrane.

Certain theoretical considerations, necessary to explain the
rapid penetration into the cell of certain substances, brought Over-

Guilliermond - Atkinson — 16 — Cytoplasm

TON, however, to the supposition that the ectoplasmic membrane
differed in its chemical constitution from the rest of the cytoplasm
and that it was composed essentially of lipides. Other physiologists,
for similar reasons, have considered this membrane as a sort of
mosaic, formed of lipide and protein particles (Nathansohn,
Clowes). The objection raised was that there were no morpho-
logical facts to confirm this hypothesis, and that the work of
Mayer and Schaeffer, as well as that of Faure-Fremiet, demon-
strated that the cytoplasm is a lipo-protein complex and that the
lipides are therefore scattered throughout all the cytoplasmic sub-
stance. There are, however, strong reasons for thinking that the
ectoplasmic layer does not have the same chemical constitution as
the rest of the cytoplasm but that it differs in being very much
richer in lipides. Indeed, in a complex liquid the lipides (lecithin,
phytosterol, etc.) which act more strongly on surface tension than
proteins, must, according to GiBBS' law, accumulate more at the
surface. This constitutes an argument in favor of the existence of
an ectoplasmic layer richer in lipides than the rest of the cytoplasm
(NiRENSTEiN, Hanstein, Czapek). The work of Rayleigh, De-
VAUX, and Langmuir have shown that oil put into water comes to
the surface where it forms a thin layer, and that its molecules have
an orientation which brings about an electrical polarization, per-
haps comparable to that which is encountered in the ectoplasmic
layer.

Physiological properties of cytoplasm:- Protoplasm is irritable.
According to Sachs' definition, irritability is the capacity of living
organisms to react in certain ways under the most diverse influ-
ences of the external world. This very complex characteristic is
qualified as physiological or biological because until now its physico-
chemical mechanism has not been known. It is manifested by
specific reactions dependent upon the differences between cells.
This characteristic is common to all the morphological constituents
of the cell (nucleus, chondriosomes), but is especially easy to ob-
serve in the cytoplasm itself. One of the most wide-spread re-
actions of cytoplasm and the easiest to perceive is motility or
contractibility, i.e., the property by which the cytoplasm changes
from place to place.

This quality can be observed in the Myxamoebae and the Plasmo-
dia of the Myxomycetes where the cells lack walls. Here motility is
manifested by changes of form and by displacements. Other cells
possess permanent extensions of the cytoplasm in the shape of fla-
gella or vibratory cilia which execute movements of rotation and
oscillation and act as locomotor organs (zoospores and anthero-
zoids). Besides these external movements, the cytoplasm usually
shows internal movements, as discovered in 1774 by CORTI and later
observed by Treviranus. These movements are called rotation, cyc-
losis or circulation of the cytoplasm, and have been already men-
tioned in these pages. They are found in most plant cells having
walls, where they are the only manifestations of cytoplasmic motil-

Chapter III — 17 — Physical Properties

ity. They are very easily observed in the epidermal cells of the
membranaceous bracts of Iris germanica and in the bulb scales of
Allium Cepa. These movements are observable because of the rapid
displacements of the lipide granules contained in the cytoplasm
which are seen circulating in the trabeculae between nucleus and
cytoplasm. The chondriosomes also are carried in the current but
much more slowly. Very favorable objects for the study of these
movements are to be found in fungi of the genus Saprolegnia and
in the leaf cells of Elodea canadensis. In Saprolegnia, as well as
these movements of circulation revealed by the displacement of
lipide granules and chondriosomes, there are also sometimes
observed displacements of an entire portion of the cytoplasm which
carries along with it the nucleus and vacuoles, a movement which
seems to be due to a general contraction of the cytoplasm. These
movements can be easily observed in moist-chamber cultures. The
observations of Lapicque, and more recently of Mangenot on
Spirogyra, have made clear the general nature of these movements.
In direct light, the cells of this alga show a hyaline cytoplasm
which, in addition to the ribbon-like chromatophore which will be
discussed later, contains, in suspension, chondriosomes in the form
of granules and short rods. These are arranged in rows and move
toward one extremity of the cell or the other. The chondriosomes
of neighboring veins may move in the same or opposite directions,
then may mingle in great agitation and soon after disperse, either
by all going in one direction or by different chondriosomes taking
different directions. These chondriosomes are being carried by
currents which agitate the mass of cytoplasm, currents similar to
those of the Plasmodium and which, like them, appear to be rapid
and restricted. These currents form little adjacent veins compar-
able in certain ways to the convection currents which agitate a
liquid as it is being heated. In observing the deeper regions of
the cell, the nucleus is seen at the center attached to the walls by
several strands of cytoplasm. This cytoplasm lines the internal
face of the wall as a thin layer surrounding the vacuole. The
boundary between the cytoplasm and the vacuole changes con-
stantly: for instance, the moving cytoplasm accumulates at one
point forming a protuberance which immediately disappears while
another appears a certain distance away, then is absorbed, and
so on.

All these currents are accelerated in the presence of different
agents (electricity, various chemical substances, etc.) and are
halted by ansesthetics.

The cytoplasm with regard to vital dyes and fixed preparations :-

The cytoplasm is permeable to water but as long as it is in the
living state, it is little permeable to certain substances dissolved
in water. It is often said that cytoplasm is impermeable to most
stains. It has been known for some time, however, that certain
basic dyes, called vital dyes, such as neutral red, cresyl blue, Nile
blue and methylene blue, have the property of penetrating the

Guillierixiond - Atkinson — 18 — Cytoplasm

cytoplasm of living cells, but these dyes traverse the cytoplasm
without coloring it and accumulate exclusively in the vacuoles. They
stain the vacuoles only in the living state and disappear in them
with the death of the cells, reappearing in the cytoplasm and the
nucleus. They will be discussed later (p. 131) in connection with
the vacuoles. Some of these dyes, however, such as Nile blue and
methylene blue, can at the same time produce diifuse staining of
the cytoplasm, especially when used at a high pH.

It is shown today, by the work of KtJSTER, confirmed by ours in
collaboration with Gautheret, that, although some dyes never
pass through the ectoplasmic layer, without any reason for this
being evident, many others can more or less easily penetrate living
cells. Basic dyes are in general the only ones which penetrate the
cell. Like those already mentioned, they may accumulate exclu-
sively in the vacuoles under certain special conditions or else may
show a predilection for the chondriosomes which they stain first.
Most of these dyes end by staining the cytoplasm and the nucleus
a little before the death of the cell. Among these dyes, chrysoidine
stains the cytoplasm and nucleus superbly in cells which are unques-
tionably alive, clearly showing cytoplasmic currents. This is with-
out doubt explained by the nature of this dye which is readily
dissolved in lipides.

The acid dyes in general do not penetrate living cells. Eosine
and erythrosine, however, may color the living cytoplasm and
nucleus as Kuster has shown. But these dyes, incorrectly con-
sidered by this observer as the best suited to staining of the cyto-
plasm, penetrate living cells only with great difficulty and only a
short time before the death of the cells, without doubt because of a
modification of permeability produced at that moment. Aurantia,
another acid dye, penetrates living cells very slowly but immedi-
ately kills them.

It should be noted that all dyes capable of producing vital
staining of the cytoplasm do so only between slide and cover glass,
therefore in cells placed under defective conditions, and it may be
admitted that this is sublethal staining, i.e., staining which occurs
only in the period which precedes the death of the cells.

If plants are cultivated aseptically in media to which these dyes
have been added, it is noted that the plants grow but no coloration
of the cytoplasm is produced. The stains accumulate in the vacu-
oles. It is only in cells where growth has stopped that the cyto-
plasm, nucleus and chondriosomes may show staining.

Chrysoidine, for example, which so easily stains the cytoplasm
and nucleus in cells placed between slide and cover glass, accurnu-
lates only in the vacuoles in plants cultivated in a medium to which
this dye has been added, and is scarcely taken up by the cytoplasm
except in cells where growth has stopped. Recent research (GuiL-
LiERMOND and Gautheret) makes it possible to explain this differ-
ence in behavior. The staining of the cytoplasm between slide and
cover glass is, in reality, obtained under abnormal conditions and
usually with toxic quantities of the dye. This is, therefore, unques-

Chapter III — 19 — Physical Properties

tionably sublethal staining. Nevertheless in cultures to which vital
dyes have been added, staining of the cytoplasm is obtained but this
staining is purely transitory and is visible only in the earlier hours
of culture. The dye, first taken up by the cytoplasm is rapidly
excreted into the vacuole which appears to be the region of the
cell in which all toxic products accumulate. It is only when the
dye has accumulated in the vacuole that the cell begins to grow
and multiply. Consequently, only vital staining of the vacuole is
compatible with growth and any persistent staining of the cyto-
plasm is necessarily sublethal. The same is true for dyes which
stain the chondriosomes. The acid dyes, eosine and erythrosine in
particular, do not in general stain plants cultivated in media which
contain them.

It seems likely that, ordinarily, living cytoplasm does not have
an affinity for dyes. Cytoplasm in healthy cells stains only under
special conditions or in cells which have ceased to grow. More
often, staining takes place just before their death in cells which
are not healthy. In the yeasts, however, Nile blue first accumu-
lates in the cytoplasm and is then excreted into the vacuoles. In
any case, these dyes, interesting from the point of view of cellular
permeability, are of no service in a cytological study of the cyto-
plasm except as they stain the chondriosomes.

Once dead, the cytoplasm becomes, on the other hand, very
permeable to all substances dissolved in water. This can be ex-
plained by considering that when the cytoplasm is coagulated, the
micelles crowded at the periphery, forming the ectoplasmic mem-
brane, move away from each other and allow the substances dis-
solved in the surrounding medium, especially all the dyes, to pass
through freely.

Consequently, for the study of the cytoplasm, the method of
fixation and staining has been resorted to because observation of
living material does not suffice. Since the cytoplasm seems to
behave like a hydrogel which is very slightly alkaline, as will be
shown farther on, its fixation, i.e., its coagulation, can be brought
about only by acid dehydrating reagents. The acid reagents usually
used are composed of weak acids (acetic, picric), or even strong
acids (nitric, trichloracetic), suitably diluted and associated, in
most fixatives, with salts of heavy metals (mercury, platinum,
osmium). Such fixatives have a pU inferior to 3. Besides these,
neutral liquids like alcohol and formalin^ may act as fixatives, for
both have the property of precipitating the proteins, the former by
its dehydrating action, the latter by combining with the protoplasm
to form insoluble compounds still not well defined.

The most acid fixatives, those with an acetic acid base and
fixatives containing alcohol, which has a too strong dehydrating ac-
tion, usually cause sudden coagulation, producing in the cytoplasm
artificial structures - coarsely granular-reticulate - which, as will be
seen, led the early C5d:ologists to attribute to the cytoplasm a reticu-

^ Formalin is usually acid, almost always containing, furthermore, some fonnic acid.

Guilliermond - Atkinson — 20 — Cytoplasm

late or alveolar structure. These fixatives, furthermore, have the
disadvantage of dissolving the lipides, which are among the essen-
tial constituents of the cjrtoplasm. Reagents containing acid in a
very dilute state, i.e., the least acid fixatives, especially formalin,
act more gently. They transform the cytoplasm into a very finely
granular coagulate which preserves, in appearance at least, the
homogeneous aspect it has in the living state.

In general, fixed cytoplasm stains only with acid anilin dyes
(eosine, erythrosine, light green, etc.). This distinguishes it from
the nucleus which stains with basic anilin dyes. The cytoplasm
for this reason is said to be acidophilic while the nucleus is baso-
philic. This affinity of the cytoplasm for acids is generally attrib-
uted to the fact that in the nucleo-proteins which compose it the
nucleic acids may be entirely saturated with simple proteins. These
nucleo-proteins thus are presumed to act as bases in regard to the

acid dyes.

The use of the method of staining material after fixation has
led many cytologists, contrary to the now classic description of
DUJARDIN, to believe in the existence of a special organization in
the cytoplasm. So for a long time the cytologists endeavored to
investigate this structure, either by fixed and stained preparations
or by direct observation of living cells, but they encountered two
great obstacles. By fixing the cells, they completely upset the
constitution of the cytoplasm which, as will be seen, is in a colloidal
state and coagulation images were induced. In the second place,
by tearing off the epidermis or by sectioning or cutting the tissue
in artificial media in order to observe living cells, alterations of
the cytoplasm were caused leading to its death.

Cytologists who studied it did not, consequently, agree on the
structure of the cytoplasm and numerous theories were proposed of
which for historical interest, the principal ones will be summarized
here and in as brief a manner as possible :

1. Reticular theory, according to which the cytoplasm is com-
posed of a network of anastomosing filaments immersed in a sub-
stance which is more fluid and less refractive (Hanstein, Reinke,
Strasburger, Carnoy, etc.).

2. _ Filar theory, according to which the cytoplasm is composed
of a fluid mass in which are immersed filaments of more solid mate-
rial which are isolated one from the other (Flemming, Haber-
landt).

3. Alveolar theory, which seems to have had the most adher-
ents and which is still advocated by some cytologists. According
to this theory, the cytoplasm is formed by an assemblage of small
alveoli pressed one against the other with walls which are more
compact than the more fluid substance within (Butschli, Crato).

4. — Emulsion theory of KUNSTLER, according to which the
cytoplasm is composed of an infinite number of very small Protem
spheres with compact envelopes and semi-fluid contents. Accord-

Caiapter III —21— Physical Properties

ing to this theory, these spheres and not the cell, constitute the
essential morphological units of living matter.

5. — Granular theory of Altmann which assumes that the
cytoplasm is composed of a very great number of small bodies in
the shape of grains called bioplasts, isolated or united in small
chains, having the appearance of bacteria. These grains are con-
tained in a homogeneous ground substance. According to Alt-
mann, these bioplasts are morphological and physiological units
capable of division and of leading an independent life. They are
homologous to bacteria which themselves represent independent
bioplasts.

Let it be added that some cytologists used to think that all
these structures might be encountered in a single cell and repre-
sented, in consequence, only functional states of the cytoplasm. Yet
the research of Henneguy, Schwakz, A. Fischer, Chodat, etc.,
had shown much earlier that different fixatives produced as many
different structures and that most of the structures described in
the cytoplasm were only artifacts produced by the fixatives.

Since 1908, the use of new fixation techniques called mitochond-
rial, producing less violent coagulation and a better conservation of
the lipides of the cell, has led cytologists to return to the old idea
of Dujardin and to conclude that the cytoplasm is homogeneous
but encloses in suspension small elements called chondriosomes.
Observations of Faure-Fremiet on living cells of Protozoa (1910),
ours in plant cells (1913, 1919) and, finally, observations com-
pleted on Metazoan cells by using tissue cultures (M. R. and W. H.
Lewis, G. Levi, etc.) have confirmed these results and shown that
the cytoplasm, examined under favorable conditions, always appears
as an homogeneous and translucid substance, containing in suspen-
sion numerous, slightly more refractive chondriosomes. While this
work was being done, the use of the ultramicroscope had led physi-
ologists to conclude, even as early as 1904, that the cytoplasm is
in a colloidal state and in 1908 Mayer and Schaeffer were able
for the first time to show that this cytoplasm always appears
optically empty under the ultramicroscope and presents the char-
acter of a fluid hydrogel. From that time on, therefore, there
could no longer be any question of any structure in the cytoplasm
other than that inherent in its physical constitution.

All theories proposed for the structure of cytoplasm are there-
fore out of date today and are now of historical interest only. It
is easy to understand how fragile the cytoplasm is, now that its
colloidal state is known. Most fixatives bring about the disorgan-
ization of the chondriosomes by dissolving their lipides at the same
time that they cause a coagulation of the cytoplasm. This coagu-
lated state appearing as a network was responsible for the formula-
tion of the reticular theory. On the other hand, observation of
living material, necessitating most often cutting or tearing of the
tissue which is then examined in an artificial medium or under
imperfect conditions, produces a disturbance in the very delicate

Guilliermond - Atkinson — 22 — Cytoplasm

equilibrium of the colloidal system which we now know cytoplasm
to be, and brings about more or less important alterations. It has
been demonstrated that the spherical and alveolar structures are
in reality the result of alterations of the chondriosomes which swell
and are transformed into little spheres and large vesicles. As foi
the filar and granular theories, it will later be seen that they rest
on facts exactly observed but wrongly interpreted. Since the pro-
posal of these hypotheses, our knowledge of the cytoplasm has been
greatly enriched by observations of the following types of material :
fungi observed in their medium of culture, leaves of aquatic plants
examined also in their own medium, membranaceous bracts of Iris
preserved from all alteration by their impermeable cuticle, and
organs of plants growing in Petri dishes.

It is thus definitely demonstrated at the present time that the
cytoplasm is a homogeneous colloidal system. Its physical charac-
teristics will be studied further on. The method of fixation and
staining can not be used in the study of the cytoplasm itself, for
this gives only pictures of coagulation which furnish no idea what-
ever of its actual nature. However, modern cytological work
carried out either by special techniques called mitochondrial or by
direct observation of living cells, with or without vital staining,
has revealed the constant presence, in the cytoplasm of each cell,
of a chondriome and a system of vacuoles, constituents of which
we have already spoken and which will be the subject of the fol-
lowing chapters.

Chapter IV
THE CHEMICAL CONSTITUENTS OF CYTOPLASM

Proximate analysis of the cytoplasm:- It is impossible to analyze
the cytoplasm. Only the analysis of the protoplasm as a whole,
including the nucleus, can be obtained. When it is a question of
specifying the chemical constitution of the cytoplasm itself, we can
only have recourse to microchemical reactions which verify the
results obtained by proximate analysis but which can give only a
very inaccurate idea of the chemical constitution of the cytoplasm.

It is especially the proximate analysis of protoplasm which will
give us an idea of the chemical constitution of the cytoplasm. But
before discussing the principal results obtained in this field, it is
important to stress the exceptional difficulties attending work of
this type. At the present time, we possess only analytical processes
with which to study the chemical constitution of living matter.
Now, protoplasm is killed by the least analytical attempt and one
is immediately reduced to working on dead matter, i.e., on a sys-
tem quite obviously modified. As for the sjmthesis of protoplasm,
living or dead, it has never been realized and in the present state
of our knowledge it even seems impossible of realization. What is
more, it is not known even how to make the principal substances
entering into the composition of living matter — the proteins or
lipides, which will be discussed later — or even how to make the
very numerous organic products, such as starch or cellulose, which
are generally manufactured in the cytoplasm.

Now as Berthelot has said, "To know really the nature of
things, it does not suffice to destroy them. We must be able to
recreate them". Analysis teaches how a body may be torn down.
It does not indicate how it has been built up. Everything indicates,
on the contrary, that decomposition in most cases does not follow
in reverse direction the same steps as synthesis. The two processes
differ essentially in every respect from each other. These con-
siderations will make it possible to give the proper weight to the
data which we now possess on the chemical constitution of living

matter.

The Plasmodia of Myxomycetes are here again particularly fa-
vorable objects for analysis of living matter. They are made up of
naked protoplasm, i.e., there is no trace of those more or less thick,
always important, external envelopes which living matter secretes
at its surface in almost all organs. The Plasmodium of Myxomy-
cetes is, therefore, the most directly and easily accessible proto-
plasm. Several analyses have been published, that, for example,
of Lepeschkin in 1923 for the Plasmodium of Fuligo septica which
follows : —

Guilliermond - Atkinson — 24 — Cytoplasm

Analysis of original (fresh) material

Water 82.6%

Dry material 17.4%

100.0%
Analysis of dry material

Substances soluble in water

Protides' 26.5%

Sugars 14.2

40.7%
Substances insoluble in water

Protides

Nucleoproteins 32.3%

Nucleic acids 2.5

Albumins 2.2

Globulins 0.5

Lipoproteins 4.8

42.3%
Lipides

Glycerides 6.8%

Sterols 3.2

Compound lipides 1.3

11.3%

Mineral salts 2.2

Miscellaneous 3.5

59.3

100.0%

A study of this table admits of the following conclusion: the
protoplasm of the Plasmodium of Fuligo is composed essentially of
proteins in the proportion of 68.8% (including the lipoproteins) and
of lipides in the proportion of 15% (including the lipoproteins)
and in addition, sugars (14.2%) and mineral salts (2.2%).

The chemical constituents of cytoplasm. Protides^:- The pro-
tides comprise the proteins and their hydrolysis products, the
amino acids. The protides are the essential constituents of cyto-
plasm and are, as is known, the most complicated of organic com-
pounds. Their molecules are the largest. These proteins charac-
terize living matter which alone has the power to create them.

Proteins are classified in two groups according to their chem-
ical constitution and some of their properties : the holoproteins^ and
the heteroproteins^. The holoproteins are made up only of amino

' Cf. Footnote'.

* Translator's note. Protide is a more comprehensive term than any used in the classifica-
tion recommended by the Physiological and Biochemical Committees on Protein Nomenclature,
J. Biol. Chem. 4:XLVIII-LI (1908) and refers to proteins and their hydrolysis products,
amino acids, amines, amides, soluble proteins, etc.

» Translator's note. Simple proteins. The term protein is more generally used by
American chemists today than proteid and will be adopted here in translation of proteide in
the French text.

* Translator's note. Conjugated proteins.

Chapter IV —25— Chemical Constitue nts

acid groups and the hydrolysis of these holoproteins results only
in amino acids and their derivatives. The heteroproteins are much
more complicated and are divided into nucleoproteins, phosphopro-
teins and glycoproteins. The heteroproteins give not only amino
acids by hydrolysis but other substances belonging to different
organic groups.

The nucleoproteins may be considered as the combination of a
simple holoprotein and a nuclein, which is itself a combination of
holoprotein with a nucleic acid. Nucleic acids are esters of phos-
phoric acid containing, side by side with this acid, a sugar repre-
sented in plants either by a hexose or a pentose (ribose) as well
as organic bases of the purine series (guanine, hypoxanthine,
adenine) and of the pyrimidine series (cystosine).

The phosphoproteins are proteins which, like the nucleopro-
teins, give on hydrolysis, amino acids and phosphorus-containing
substances other than nucleic acids. As for the glycoproteins,
they decompose by hydrolysis into proteins and sugar. The phos-
phoproteins and the glycoproteins are, moreover, very imperfectly
known.

Holoproteins are always found in protoplasm but they seem to
play only a small part in the constitution of living matter. They
usually represent products of cellular metabolism and are chiefly
localized in the vacuoles where they constitute reserve products
(aleurone) . The greatest proportion of the constituents of living
matter seem to belong to the heteroproteins, among which the
nucleoproteins dominate. They do not seem to be exclusively local-
ized in the nucleus, contrary to an opinion often accepted, but they
are found also in the cytoplasm. In 1881, Reinke and Rodewald,
studying the chemical composition of the Plasmodium of Fuligo
septica, concluded that it is in large part made up of a phosphorus-
containing protein presumably formed by the union of a nuclein
and a protein. These results assigning to the cytoplasm a nucleo-
protein constitution have since been verified in material differing
greatly from the above. In 1892, Halliburton observed that the
nucleoprotein extracted by him from the kidney was in too great
quantities to have come from the nuclei alone and concluded cate-
gorically that this protein came above all from the cytoplasm.
Then in 1895, Halliburton isolated another nucleoprotein from
mammalian red blood corpuscles which do not have nuclei. These
analyses and many others published since, that of Lepeschkin
for example, actually indicate, it would seem, that the histochemical
proteins are nucleoproteins. It is admitted, however, that there
is a difference between the nucleoproteins of the cytoplasm and
those of the nucleus. In the former the nucleic acid is completely
saturated by the protein base while in the latter the saturation is
not complete and there is some uncombined nucleic acid. The
capacity of the nucleus to be stained with basic dyes is due, accord-
ing to this theory, to the uncombined nucleic acid which becomes
affixed to the basic dyes. So it would be explained that the nucleo-

Guilliermond - Atkinson — 26 — Cytoplasm

proteins of the nucleus are basophilic while those of the cytoplasm
are acidophilic.

Lipides:- As is known, lipides include substances of very varied
chemical constitution which fall into one group by reason of their
common properties: solubility in ether, chloroform and benzene
as well as diverse histochemical characteristics.

Lipides, but for rare exceptions, seem to be exclusively con-
tained in the cytoplasm. Analysis of plant tissues shows that they
constitute a considerable proportion of the cytoplasm, 15-25%, and
among these are the simple lipides (lipides ternaires) and the com-
pound lipides. The simple lipides^ include on one hand the glucer-
ides or true fats, esters of glycerol and a fatty acid, and on the
other, the sterols^ composed of alcohols of high molecular weight
which can be esterified by fatty acids. The compound lipides are
subdivided into phospholipides, esters resulting from the combina-
tion of an alcohol, inositol, with phosphoric acid and the phospho-
aminolipides, represented by lecithins, which like the glycerides, are
esters of glycerol but contain nitrogen and phosphorus and in which
two alcohol linkages of glycerol are combined, each to a molecule
of fatty acid, the third being united to a molecule of phosphoric
acid itself linked to choline and sometimes to betaine, a substance
possessing both the alcohol and the amino function.

A considerable proportion of protoplasmic lipides represent re-
serve products which accumulate at certain periods in the life of
the cells and are later consumed. Such is the case for most of the
glycerides and perhaps also for certain lecithins. Sterols and
lecithins are also found in the vacuoles where they are products of
metabolism whose role is not yet known. But the work of Mayer
and SCHAEFFER on animal cells, the results of which certainly apply
to plant cells, has shown that a notable part of these lipides repre-
sent an essential constituent of the cytoplasm. They are, there-
fore, an integral part of living matter.

These lipides are sometimes refractive globules in the cyto-
plasm (Fig. 9), or they are sometimes represented in the plastids
and chondriosomes but it seems as if a considerable proportion of
them were combined with the cytoplasmic proteins as compound
lipoproteins. It is difficult to detect them, for, being combined
with the proteins, they are invisible and constitute what is known
as masked lipides. Their existence may be demonstrated either by
histochemical analysis of certain tissues or by chemical analysis.

It is by this chemical analysis that notable quantities of lipides
have been revealed in tissues which do not show any in the state
represented in Figure 9 and in which no trace of them is found with
ordinary histochemical methods. Among these masked lipides there
are some, however, which are united to the proteins in an unstable

' The translator thinks these substances are probably complex as they contain both the
compound and derived lipides of Bodansky's classification. Bodansky. M. Introduction to
Physiological Chemistry. Wiley and Sons, New York. 1938.

^ Translator's note. The sterols actually occur partly as free alcohols and partly as esters
of fatty acids.

Chapter IV — 27 — Chemical Constituents

way, in weak chemical combination or by a simple physical adher-
ence. These may be disclosed by special histochemical methods
isolating them from their combination, i.e., they may be unmasked.
Others of these masked lipides are united as stable chemical com-
pounds in a very intimate manner with the proteins and can be
demonstrated only by destroying the protein substances to which
they are united, either by chemical hydrolysis or by digestion.
Chemical analysis, therefore, is the only process by 'which their
presence can be demonstrated.

It would seem as if this would hold true for plant cells. It
has been observed that tissues rich in proteins are also rich in leci-
thins. It is also probable that lecithins are among the most im-
portant chemical constituents of the plastids and chondriosomes.
Some authors, having ascertained a fixed relation between the
content of lecithin and chlorophyll in tissues, state that this pig-
ment exists in the chloroplasts, in a combined state, as chloroleci-
thin.

Lipide granules which are found in all cells are perhaps the
product of an unmasking of lipides which takes place normally
under certain conditions. This phenomenon seems to be more pro-
nounced during cellular degeneration resulting in fatty degenera-
tion (lipophanerosis) .

It must be added that Dujardin had observed that when the
cuticle surrounding the cellular body of certain Infusoria is injured,
the cytoplasm in contact with water forthwith forms droplets which
he calls "boules sarcodiques", balls of sarcode. Maggi, who later
took up the study of them, thought that they corresponded to the
figures observed when myelin (a compound lipide forming the
peripheral layer of axone in higher animals) comes in contact with
water. There are found then, at the surface of the myelin, fila-
mentous protuberances whose extremities finally swell out in the
shape of clubs. Many cytologists have found similar figures
(KoLSCH, Albrecht, Schneider, Prowazek, Faure-Fremiet,
KUSTER) both in animal and plant cells. These figures form in
large quantities all around the vacuoles during the phenomena of
plasmolysis. They form pedicelled buds which may break off, and
become introduced into the vacuole as vesicular buds and might
correspond to the myelin figures arising by separation of the lipides
from the cytoplasmic lipoprotein compound.

Various products and mineral substances:- It has been seen that
analysis reveals in the cytoplasm, but in smaller quantities, other
products of different sorts : sugars resulting from cellular metabol-
ism, localized mostly in the vacuoles, and, more important, min-
eral substances, among which potassium, magnesium, iron may be
especially mentioned. These minerals which seem to play a very
important role in the manifestations of living protoplasm, may
be in the form of ions or molecules or may be fixed by adsorption
to protein substances and enter into the cytoplasmic lipoprotein
complex.

Guilliermond - Atkinson — 28 — Cytoplasm

Water:- Water is an important constituent of protoplasm. In
fact, all protoplasm contains water interposed between its micelles.
This water is indispensable for the manifestations of vital phe-
nomena as these can be produced only if the protoplasm is in a
semi-fluid state. The water content of living substance varies
with age, i.e., according to the physiological state of the cells and
also, in certain plants, according to the environmental conditions.

In a general way, in the course of the life of a plant, the maximum
water content coincides with the maximum physiological activity.
The differences which exist in the viscosity of the cytoplasm, vary-
ing with the age of the cells, have already been mentioned. The
amount of water held in the cytoplasm varies from 80-90% by
weight. A part of this water may be removed but then the vital
activity is diminished. If it is almost all suppressed, the proto-
plasm in certain cases may not die but passes into a state of re-
tarded activity (anhydrobiosis) . This is the case normally in the
spores of bacteria and fungi, as well as in the seeds of phanero-
gams which, during maturation, become dehydrated until they con-
tain only 10-18% of water and sometimes, as in certain fatty
seeds, as little as 5-6%.

Chapter V

PHYSICO-CHEMICAL CONSTITUTION
OF THE CYTOPLASM

The cytoplasm, as has been seen earlier, appears to be con-
stituted essentially of proteins (68.8% of dry weight), water-
insoluble lipides (about 15% of dry weight), and a large propor-
tion of water (80-90% of fresh material) containing various
mineral substances in solution. It may be considered then as a
colloidal solution whose micelles are represented by protein mole-
cules united with lipide molecules, the intermicellar liquid being
water.

It seems that every cell contains the same chemical constituents
and that differences between the cellular types exist only in the
proportions of the chemical constituents present. One is obliged
to admit, however, that each type of cell has a protoplasm, and
therefore a cytoplasm, which is peculiar to it. The widely-varying
properties of the plastids are identified not only by the differences
of minimum and maximum temperatures necessary for their
growth, but by the elaboration products of their protoplasm, which
differ essentially from one cell to another. The diverse properties
of different plants are due to the individual constitution of their
protoplasm and, in particular, of the cytoplasm which is peculiar
to each. One is therefore driven to the idea of specificity of proto-
plasm but chemical analysis has not so far furnished any basis for
this notion. Nevertheless, the possible number of different nucleo-
proteins being theoretically indefinite, it is admissable that each
species of animal or plant be characterized by a special nucleo-
protein.

Electrical characteristics of proteins:- The work of MiCHAELIS,
Procter, Wilson, and especially of Jacques Loeb, confirming an
hypothesis formulated a long time ago, has demonstrated that the
proteins belong to the class called electrolytic colloids which behave
as ionisable electrolytes. Their micelles, seeming to correspond
to enormous molecules, ionize to form true micellar ions, or protein
ions, having an electric charge analogous to that resulting from
the dissociation of the valence bonds of an ordinary electrolyte.

The proteins correspond to electrolytes called ampholytes, that
is to say, possessing at the same time acid and basic functions.
They are formed, as is known, of amino acids combined one with
the next, a great number of times, to form enormous molecules.
Now an amino acid may be schematically represented by the
following formula:

Guilliermond - Atkinson — 30 — Cytoplasm

C H (acid function)

Radical C H:

N Hj (basic function)

There is, therefore, an acid function COOH and a basic function
NHz and this amino acid is an electrolyte in both senses, being
able by dissociation to give an H+ ion as well as an OH— ion. In
the first case there will be :

COO- H4-

R C H:

N H,

and with the addition of one molecule of water, in the second :

COOH

R CH:

N H3+ O H -

When the medium is acid, the dissociation of the acid function is
diminished or blocked by virtue of the law of mass action, whereas
all the OH— ions leave the radical of the ampholyte and the basic
function is manifested. The ampholyte, then, is composed only of
R'H+ and behaves like a cation. In the electric field it will migrate
to the negative pole. When, on the contrary, the medium is basic,
the dissociation of the basic function diminishes or is blocked,
whereas that of the acid function increases. The ampholyte is now
constituted only of the anion R*OH-. It behaves like an anion
and migrates to the positive pole.

Finally, and this is very important, for a certain intermediate
reaction between the two extremes, the dissociation of the acid
valence being equal to the dissociation of the basic valence, the
ampholyte is both an anion and a cation simultaneously and there-
fore behaves somewhat as though it did not have any electric
charge, and does not move in the electric field. This point of no
migration is called the isoelectric point. It is important to notice
that this isoelectric point does not generally correspond to chemical
neutrality but to a particular pB. (isoelectric pH indicated by the
symbol pHi).

At their isoelectric point, the ampholytes show, as has been
demonstrated by Jacques Loeb, a series of special properties: a

Chapter V — 31 — Physical Chemistry

minimum of solubility, a minimum of viscosity and of linkage with
peripheral electrolytes (i.e., minimum ability to combine), mini-
mum swelling in water, maximum rigidity and maximum insta-
bility in solution. At the isoelectric point the ampholyte is ex-
tremely apt to flocculate. It seems therefore that the fundamental
properties of cytoplasm must be dominated on the one hand by the
pB. of the intracellular liquids and on the other hand by the iso-
electric point of its constituents.

Physical constitution of cytoplasm:- The classical researches of
Mayer and Schaeffer (1908) made an important contribu-
tion to the study of the colloidal state of cytoplasm in so far as
animal cells are concerned. These results apply equally well to
plant cells and must be considered first.

Mayer and Schaeffer undertook first the study, with the ultra-
microscope, of colloidal solutions of different organic colloids.
They then in a comparative way were able to begin the study of
animal cells. They showed that these solutions may present
two very different optical appearances. Some, such as glycogen,
dialyzed albumin, and diastases, are characterized by a suspension
of a large number of micelles, brightly lighted and animated by
Brownian movement. These are the hydrosols. Others, such as
white of egg and nucleoproteins, appear optically empty and with-
out any visible micelles whatever. They are filtered with difficulty
and show very high viscosity. They are composed of micelles con-
taining a great deal of water, thus possessing a strong linkage with
the intermicellar liquid, i.e., water, so that the micelles do not
refract light and do not appear in the ultramicroscope. To these
latter liquids, Mayer and Schaeffer give the name fluid hydrogels.
Under certain influences, by hydration of their micelles, these hy-
drogels can be transformed into hydrosols. They appear as con-
fused luminous streaks — this is called the non-resolvable nebulous
stage — then as very fine granules which agglomerate progressively
and the field becomes completely starred with brilliant dots. This
is the resolvable nebulous state i.e., flocculation. This reversible
phenomenon may be succeeded by a more important and irreversible
modification, coagulation, accompanied by physical and chemical
phenomena which finally agglomerate the particles into a single
mass.

So Mayer and Schaeffer set off against each other two dif-
ferent physical states: the hydrosol, appearing in the ultramicro-
scope as a suspension of micelles vividly lighted and animated by
Brownian movement, a state similar to that presented by pseudo-
solutions of metals; and the hydrogel appearing as an optically
empty pseudosolution which is very viscous. These investigators
designate by the name jelly the semi-solid state of the hydrogel
which is much more viscous and endowed with a very definite
rigidity. This terminology is not accepted by all authors. Some
writers distinguish the hydrophobic colloids (mineral colloids)
whose micelles do not have any affinity whatever for water and

Guilliermond - Atkinson — 32 — Cytoplasm

form hydrosols, from the hydrophilic colloids whose micelles
have a strong affinity for water and produce solutions which are
optically void (fluid hydrogels of Mayer and Schaeffer). Other
authors have reserved for this latter group the name isocolloids and
differentiate between the semi-fluid state or isocolloidal hydrogel
and the fluid state or isocolloidal hydrosol.

The structure of the hydrogels is, moreover, variously inter-
preted. Certain authors consider the hydrogel as formed by an
assemblage of micelles lying more or less closely together and
heavily saturated with water (Bradford). Others consider it as
constituted of a network holding water in its alveoli (Hardy).
Still others including Procter, Wilson, Jacques Loeb, think that
there is no difference between a true solution and a protein hydro-
gel. They believe that the protein hydrogel is made up of a homo-
geneous dispersion of protein and water molecules, their propor-
tional amounts varying greatly and regulating the consistency of
the system. Lumiere, elaborating on this conception, separated
into two categories the substances which until then had been desig-
nated as colloidal : first, the micelloid colloids, in the state of hydro-
sols, formed of voluminous polymolecular particles and visible in
the ultramicroscope and second, molecular colloids, characterized
by the optically void state of their solution, formed of monomolec-
ular particles and taking on, as they solidify, the aspect of glue.
These last, according to Lumiere's view, are in reality true solu-
tions, distinguished from solutions of crytalloids only by the
enormous dimensions of their molecules to which they owe their
special properties.

Devaux, who shared this opinion, looks upon these colloids as
allied to strongly polymerized compounds, such as cellulose and
rubber recently studied by Staudinger, which Devaux considers
as formed of molecules characterized by being extraordinarily long
but still conserving transversely the dimensions of ordinary mole-
cules, i.e., molecules arranged in a line. This, according to De-
vaux, explains their colloidal properties, for example their ability
to swell up in certain liquid media and give viscous and ropy solu-
tions. Certain reasons would seem to support this opinion. It is
known, in fact, that two molecules of amino acids may combine by
peptide bonds and, by repetition of the process, form a long chain
constituted of molecules arranged in lines carrying laterally, like
oars, the radicals of various amino acids. Several polypeptides may
combine in this way to give a very long chain, straight or folded
over, such as keratin which, according to the work of Astbury,
shows linear molecules (X-ray patterns). The knowledge of pro-
tein structure is not far enough advanced so that it can be known
whether generalization can be made from this structure to include
cytoplasmic proteins. Nevertheless, some biologists recognize the
existence of the fibrillar structure and crystalline nature of organic
gels and in particular of the cytoplasm (Seifriz). This structure
would permit them to explain, in accordance with Seifriz's opin-
ion, both the pulsations observed in the Plasmodia of Myxomycetes

Chapter V — 33 — Physical Chemistry

(CoMANDON and Pinoy, Seifriz, Mangenot) and the elastic
properties of the cytoplasm.

Utilizing the results obtained from the study of protein solu-
tions, Mayer and Schaeffer with the ultramicroscope began the
study of animal cells living under favorable conditions. This study
permitted the authors to observe that the cytoplasm presents
in the ultramicroscope only a few lighted granules, which do
not show Brownian movement. They are not to be classed as mi-
celles for they are always visible in direct light. In addition to
these granules which belong to the paraplasm, the cytoplasm,
like the nucleoprotein solutions, does not show under the ultra-
microscope any visible micelle at all. Schaeffer, therefore, has

Fig. 9. — Ultramicroscopic view of epidermal cells of the
leaf of Iris germanica. 1, In the living cell the nucleus is
opalescent, the nucleolus and cytoplasm optically empty. The
granules correspond to lipide inclusions either (P) within
invisible filamentous plastids or (GO dispersed in the cyto-
plasm. 2, In the coagulated cell the nucleus and cytoplasm are
entirely luminous.

identified it with a fluid hydrogel. Now a fluid hydrogel behaves like
an electronegative hydrogel. Like all the alkaline or negative gels,
it becomes cloudy when it is put into acids : at first confused lumi-
nous streaks are seen, then ultramicroscopic granules which
soon become visible in direct lighting and assemble in a network.
Finally the cytoplasm becomes entirely luminous. It is then coagu-
lated. The salts of heavy metals and, in general, all substances em-
ployed as fixing agents, act in the same manner as acids, by making
the cjrtoplasm appear granular and vesiculate. The dehydrators
(alcohol, heat) act similarly. On the contrary, in the presence of
alkalis the cytoplasm remains optically empty.

Guilliermond - Atkinson — 34 — Cytoplasm

The ultramicroscopic observations of Fauhe-Fremiet on In-
fusoria and sexual cells of the Metazoa, followed by the work of
others (Aggazzotti, Marinesco, Mossa, etc.), have confirmed the
results of Mayer and Schaeffer on the optically void state of the
cytoplasm. Furthermore, observations on living cells of tissue cul-
tures by Levi and the micromanipulations of Chambers are equally
in accord with these results. It is, however, to be noted that some
cytologists, such as Chambers, have designated under the name
of hydrosol the state corresponding to the fluid hydrogel of Mayer
and Schaeffer.

The question of the colloidal state of cytoplasm has been much
more discussed in connection with plant cells. The first observers,
Gaidukov and Russo, using the ultramicroscope, came to the fol-
lowing conclusions: The cytoplasm of these cells appears as a
heterogeneous structure; its micelles are animated by Brownian
movement; the cytoplasm, therefore, offers the characteristics of
a hydrosol. This opinion is still held today by Lefeschkin. Price
has stated, on the contrary, that the cytoplasm of plant cells most
often appears optically empty, i.e., as a hydrogel. However, this
author concedes that it may pass from the state of a hydrogel to
that of a hydrosol. Pensa has described in plant cells a hetero-
geneous structure in a semi-fluid suspension medium, presenting
a luminosity always more or less marked. He hesitates to consider
it as a hydrogel or a hydrosol. In this substance he believes there
is a dispersed and solid phase represented by a few strongly lighted
microsomes and a dark liquid phase also containing some micro-
somes animated by Brownian movement.

The work of Lapicque and his students, that of Becquerel, and
our work have shown, on the contrary, that the cytoplasm of plant
cells takes on the same aspect as that of animal cells. It is always
optically empty in living cells and becomes entirely luminous when
it is coagulated, appearing like snow. Our research has proved
that the diverging results of the men cited above are explained by
the fact that these workers neglected to make a comparative exam-
ination of the living cells with lateral and direct illumination. That
which those writers who recognize a heterogeneous structure of
the cytoplasm describe as micelles, corresponds to the microsomes
or vacuolar precipitates, which will be discussed later. These are
not of micellar rank and are observed quite as well with direct light
as with lateral illumination. Furthermore, some authors, like
Pensa, seem to have carried out their observations under unfavor-
able conditions and to have described cells in which the cytoplasm
was already beginning to coagulate. Observations of living mate-
rial that we have carried out in the most varied plant cells (epi-
dermal cells of Monocotyledons, filaments of Saprolegnia and of
other molds, yeasts, etc.), whether with the ultramicroscope or the
ordinary microscope, have always shown us the cytoplasm as homo-
geneous, optically void, and translucent.

Today, therefore, it may be recognized as definitively estab-
lished that the cytoplasm in plant cells as well as in animal cells

Chapter V — 35 — Physical Chemistry

appears optically empty as Mayer and Schaeffer described it.
What remains still obscure in the question is the physical structure
which can be attributed to that optically empty state which charac-
terizes the cytoplasm, both in plant and animal cells. From every
point of view the cytoplasm appears as a colloidal solution or a
fluid gel. It is not, however, miscible with water. This is a fact
of capital importance which distinguishes it essentially from all
gels or solutions. To designate the peculiarity of this gel-like sys-
tem of not being able to mix with the excipient, water, although
water represents 90% of its weight, the Italian physiologist, BOT-
TAZZI, has created the term gliode. The gliode is a colloidal system
existing exclusively in living cells, particularly in the cytoplasm.
Keuyt and Bungenberg de Jong in recent work have sought
to explain this peculiarity by considering the cytoplasm not as a
hydrogel but as a coacervate. They have shown that a very slight
addition of the precipitating agent may bring about, at first, a par-
tial separation of the colloid from the solution, in the form of a
mass retaining a considerable quantity of the solvent and still
showing, in consequence, a more or less marked fluidity. This col-
loid which is fluid but more concentrated than the original system
is miscible with an excess of the solvent. This phenomenon is
called coacervation and the fluid mass, separated as described, is
called the coacervate. Now coacervate systems are very reminis-
cent of cytoplasm. Such systems can be prepared in the laboratory.
Bungenberg de Jong has been able to produce coacervates in which
water, proteins and lecithins, i.e., the normal constituents of proto-
plasm, are actually associated in artificial cells, presenting striking
analogies with real cells and showing a central body, like the nu-
cleus, and globules of oil, separated by invisible films. What is
more, these systems are stable when the pH value is in the neigh-
borhood of 7.4, i.e., when the pH values are comparable to that
which seems to exist in the cytoplasm. Of course these resemb-
lances are of a purely physical nature and do not in any way con-
cern the absolutely inimitable physiological activity of living
cytoplasm.

Cellular constants and equilibria:- Some substances exist in the
protoplasm in greatly varying proportions, depending upon internal
or external factors (nutrition of the organ and the condition of the
medium respectively). The glycerides are such substances and in
the cells constitute transitory deposits which, by their nature, are
capable of being mobilized. This is not true of the permanent
constituents of the cytoplasm, such as the proteins, lipides and
water. These latter show differing properties and keep their differ-
ing properties when in the cell. The stability of the cell is main-
tained only because its various properties are in equilibrium. The
permanent constituents are present in the cytoplasm in fixed pro-
portions, independent of external or internal conditions.

Mayer and Schaeffer have done some work on animals, the
results of which it would seem possible to extend to plants. They

Guilliermond - Atkinson — 36 — Cytoplasm

think that the properties of protoplasm are determined by the pro-
portions of its fundamental constituents. These are not distributed
in a hit-or-miss fashion, but are always found for the same type
of cell in invariable ratios which Mayer and Schaeffer call cellu-
lar constants.

It is thus that the fatty acid content is very variable from one
species to the next but for all tissues of a single species it always
fluctuates about a constant value.

In the matter of cholesterol, the content differs very greatly
from one organ to another in the same species but is rather con-
stant for a given organ no matter what the species. It is charac-
teristic of the organ under consideration. The ratio, cholesterol :
fatty acid, or the lipocytic coefficient, is characteristic of a given
species. It constitutes a cellular constant. There exists also a
mineral constant. Similarly the water content for each type of
tissue always fluctuates about the same value, therefore each type
of tissue in a species possesses a constant and specific content of
water of imbibition. Water is thus a cellular constant. The
research of Nicolle and Alilaire brings out similar results
for bacteria. Thus in a typhus bacillus there is found 85% of
water whereas in an anthrax bacillus there is only 75%.

The 2?H value, the isoelectric point of cytoplasmic colloids, and
the rH, the oxidation-reduction potential of cytoplasm, appear also
to be cellular constants.

Mayer and Schaeffer have shown, furthermore, that there
exists for each type of cell a constant ratio in the proportions of
these different constants. Thus a close relationship is observed
between the ratio of cholesterol : fatty acid, or lipocytic coefficient,
and the imbibition of water by cells, the higher the coefl^cient, the
more water imbibed by the cells. The lipocytic ratio is thus an
index of a cellular property.

This relation between lipocytic coefficient and the water con-
tent of a cell is explained in the following manner : cholesterol, as
has been said, has the property of absorbing a rather large quan-
tity of water by imbibition. The combined fatty acids, on the con-
trary, do not imbibe water. But for some time it has been shown
that cholesterol and fatty acid compounds are soluble one in the
other and this mixture becomes penetrable to water. Thus, the
more cholesterol there is contained in a protein gel or coacervate,
the more water there will be imbibed.

The outcome of this research is that the cells of organisms do
not show their known properties to the same degree. They have
general properties which are encountered in all cells (protoplasmic
properties) and properties which are related to the function which
they perform in the cell. The general properties depend upon con-
stituents whose value appears constant for each cellular type.

Ionic reaction of cytoplasm:- Unfortunately, in spite of their pre-
cision, the methods for measuring pB. are not easily applicable
to living cells. What is needed, if it were possible, is a method for

Chapter V —37— Physical Ch emistry

measuring cells which have not been injured, for the very delicate
protoplasm modifies its pK the moment that there is any alteration
in the cell. Vles has insisted, on the other hand, on the fact that
the measurement has no value unless during the entire experiment
the cells keep their normal CO2 pressure. Now the altered proto-
plasm in contact with air, unless special precautions are taken,
loses its CO2, which raises its pU, and it is difficult in the extreme
to realize ideal conditions.

Electrometric methods or Clark indicators are used to deter-
mine pB. values. The electrometric method has been used on
crushed cells, extracts of plant juices, and finally on living cells.

Vles, Reiss and Wellinger have advocated a technique which
consists of crushing the cells after instantaneous freezing and
effecting a measurement at the precise moment when the crushed
mass returns to the cryoscopic point. In this way all escape of
CO2 seems to be avoided. But more often one is limited to meas-
uring the pB. of plant juices obtained from crushed tissues by the
electrometric method, sometimes even in a pressure of 100 atmos-
pheres (Kappen). Another method consists of introducing micro-
electrodes into living cells (Crozier, Ellis, Peterfi, Small) but
this is applicable only in cases of large cells. This method likewise
causes injuries to the cells and undoubtedly brings about a mixture
of cytoplasm with vacuolar sap. It is moreover a difficult opera-
tion in plant cells because of the resistance of the cellulose walls.

As the electrometric method is extremely delicate and necessi-
tates costly apparatus, the method most used is the Clark indicator
method. This presents great disadvantages, for the point of color
change may be modified by the presence of protein matter and its
products of disintegration, such as the amino acids. Thus in col-
loidal gels the color change is aberrant in consequence of chemical
reactions. This is called the protein error. Other causes of error
result from the fact that the indicators dissolved in the lipides are
not dissociated and that then their color does not correspond to any
indicator. This causes a lipide error. The point of color change
may also be displaced by the presence of a strong concentration of
neutral salts : the salt error. It may also be changed by a modifica-
tion in the concentration of the indicator: concentration error.
Finally, the indicators do not penetrate living cells. Vles has used
the method of crushing cells which consists of producing a sudden
yet careful pressure on cells in an indicator solution between cover
slip and slide, causing a break limited to the cell wall. An immedi-
ate release in pressure leads then to the taking into the interior of
the cell of a small quantity of the indicator, the color of which can
thus be noted before any effusion of CO2.

Botanists, however, have usually been limited here also to the
use of indicators either on juices extracted from tissues or on
sections of tissues which, in being sectioned, were necessarily
injured.

A more precise method for measuring pH has been to introduce
dyes by micro-injection with the aid of a micromanipulator. The

Guiriiermond - Atkinson — 38 — Cytoplasm

results obtained by these different methods, that of micro-injection
excepted, are very divergent and have resulted in pB. values rang-
ing from 3.4-5 to 7.8, whereas, according to Vles and Reiss, as
well as according to Small, the pH is 5.1-6.2. There is reason to
wonder as to what is the value of results obtained by these methods.
First, no value can be placed on results obtained from plant juices
since they do not express the pR of living protoplasm, but only
inform us concerning that of the vacuolar sap, more or less strongly
modified by the treatment. One can not deny, furthermore, that
the other processes, that of Vli^s in particular, offer great disad-
vantages because the crushing practiced on living cells, no matter
how rapidly executed, does cause injuries and these lead rapidly to
the death of the protoplasm. It seems evident that protoplasm be-
comes acid immediately after death.

In addition, and this is the real objection, none of these pro-
cesses allows the total pH value of the protoplasm to be obtained
without regard to the parts which make it up. Now, among these
parts the vacuoles must be considered. As will be seen, they give a
clearly acid reaction. It follows that the results obtained have
no value, for they are only the pK value of the cytoplasm and vacu-
oles combined, and are obtained, usually, after the death of the
cells.

More precise results obtained by micro-injection, with the aid of
Chamber's micromanipulator and Clark indicators, have given a pH
value neighboring on neutrality: about 7 — from 6.8-7.2 (D. and J.
Needham, Rapkine and Wurmser). It is suspected, moreover,
that micro-injection itself can induce disturbances in the cell and
modify the reaction of protoplasm. The ideal would be to measure
the cellular pH by means of vital dyes. Unfortunately the Clark
fndicators do not ordinarily penetrate the cells and are all of them
more or less toxic.

Cellular rH:- It is possible to measure directly the cellular rH in
accordance with the oxidation state (colored) or reduction state
(uncolored) taken on by the dye in the cellular medium. This is
accomplished by micro-injection into the cells of oxidation-reduction
indicators. The rH value given by cells in the presence of air varies
from 12 to 14. The rH value of the same cells measured after an
anaerobic period is in the neighborhood of 7 (Brooks in Valonia,
Wurmser and Rapkine in Spirogyra) . The micro-injection method
has the disadvantage of injuring the cell and of introducing indi-
cators which are generally toxic. The use of vital dyes appears to
be the best method. It is unfortunate that these accumulate almost
entirely in the vacuoles and generally produce only sublethal stain-
ing of the cytoplasm. However, with Gautheret, we have used
Janus green. In yeasts, particularly in Saccharomyces cerevisiae,
this dye is taken up by the cytoplasm and is rapidly reduced there
to its rose derivative, which does not revert to its green form. If
the medium contains nutrients, the dye is excreted to the exterior
in the form of the rose derivative. If these nutrients are composed

Chapter V — 39 — Physical Chemistry

only of potassium nitrates (1% at pH 9), the rose derivative of
Janus green will stain the cytoplasm of yeasts and in the absence of
air will completely lose its color. It is, therefore, reduced to its leu-
coderivative. If the medium is aerated, the leucoderivative is oxi-
dized and returns to its rose form. This reduction from the rose to
the leucoderivative takes place at rH 5.2. This limit seems to be
reached at the end of a long time and under experimental condi-
tions. It seems, therefore, to be the lowest obtainable rH value.
The method is a safe one to follow as yeasts treated in this man-
ner can bud in a very active way after having excreted the dye.

Chapter VI

THE PLASTIDS

The plastids (Fr. plastes, plastides, leucites) :• It has been known
for a long time that chlorophyll is found localized in small bodies
which were first observed by Meyen (1828) and described by him
under the name of chlorophyll grains. These bodies were further
studied by numerous workers on the cells of the phanerogams,
where they appear scattered in the cytoplasm as numerous globular
or lenticular bodies measuring from 6-lOft in diameter. Nageli
(1846) first noticed that they multiplied by division. Later, VON

MOHL (1838-1856), and then Sachs (1852), es-
tablished the fact that these bodies are composed
of a substratum which is colorless, insoluble in
alcohol, protein in nature and which, like the cyto-
plasm, stains yellow with an aqueous solution of
iodine. On this substratum is fixed the chloro-
phyll which can be dissolved in alcohol, leaving be-
hind the colorless substratum. voN Mohl and
Sachs finally proved that these bodies are the
seat of starch formation. In experiments, now
classical, Sachs succeeded in showing that the
starch grains which are formed in the chlorophyll
grains are the direct products of assimilation in
the presence of chlorophyll. These bodies were
later the object of splendid research in the
phanerogams by Schimper (1883) and Meyer
(1883), who will be spoken of later. These bodies
were called chloroplastids (Schimper), autoplasts
(Meyer), chloroleucites (van Tieghem) or chlo-
roplasts (Errera). It is not necessary to stress
the importance of these bodies which are the cen-
ter of the formation of chlorophyll, starch grains
and many other products and which play an essen-
tial role in photosynthesis.

The evolution of these chlorophyll bodies in
the higher plants, the pteridophytes and phanerogams, remained
obscure for a long time. It is known, indeed, that in these plants
chlorophyll is not generally found in the egg or in embryonic cells
and appears only in the course of cellular differentiation and then
only in tissues exposed to the light. In embryonic tissue and in
all root tissue, chloroplasts are generally not encountered. The
question as to the origin of these bodies was raised but it was not
possible to answer it until very much later.

In the algae, chlorophyll is always contained in all parts of
the thallus and, in consequence, chloroplasts are found in all the

Fig. 10. — Euglena
viridis. St, stigma
(eyespot). Fl, flagel-
lum. Ch, chloro-

plasts. P, paramy-
lum.

Chapter VI

— 41 —

The Plastids

cells. These chloroplasts sometimes have the same form and di-
mensions as in the higher plants, but often these organelles are
greatly differentiated, of a complex structure and are present in
the cell only in very small numbers. Sometimes there is only one
to a cell. In that case it is voluminous and generally contains small
refractive bodies which are colorless, rounded, or angular, known
as pyrenoids (Fig. 12). At the surface of these grow the starch
grains which form a sort of crown about them. Because of their
complex structures the chloroplasts are often designated by a spe-
cial name, chromatophores, but we shall see that they correspond
to the small chloroplasts of phanerogam cells (Fig. 10). The life
history of these chloroplasts is not difficult to follow and SCHMITZ
(1882) showed that they are always transmitted from cell to cell by
division, quite like the nucleus. Stras-
BURGER was even able to follow the divi-
sion of the single chloroplast in living
Spirogyra during cell division.

The chloroplasts in algae and bryo-
phytes:- The study of the plastids will be-
gin with the algae. In the Cyanophyceae,
which are the most primitive of the known
algae and whose cellular organization dif-
fers from that of the others, there are no
chloroplasts. Their cells contain a chro-
matic network, the central body, which
occupies the greater part of the cell and
which must be considered as a primitive
nucleus. Surrounding this is a thin pari-
etal zone of cytoplasm which contains
chlorophyll in a diffused state to which is
added a blue protein pigment, phycocya-
nin, and sometimes a red protein pigment,
phycoerythrin. Some authors have called
this parietal layer a chromatophore (Fischer, Meyer), but this
opinion seems difficult to justify, since the parietal zone does not in
the least resemble a chloroplast and encloses, all around the cen-
tral body, small vacuoles which can be stained in living tissue with
neutral red. As well as the Cyanophyceae, the algal Flagellates
(Phytoflagellaceae) are considered to be most primitive algae and
are thought to be the ancestors of all the other algae. In these,
the chloroplasts in each cell are sometimes numerous, as in many
Euglenas, and look like those found in the phanerogams. In other
Euglenas there is only one star-shaped chloroplast. In the Peri-
diniales, which are also classed among the flagellated algae, the
chloroplasts may be small, numerous, rod-shaped bodies or may be
reduced to a single body appearing as a fine network and occupying
the entire cell (Chatton). In many flagellated algae (Chrysomo-
nadales, Polyblepharidales), however, there is only one bell-shaped
chloroplast which occupies one of the poles of the cell, a position

Fig. 11. — Motile cell of
Chlamydomonas surrounded by a
cell wall which forms a beak
between the flagella at the an-
terior pole. Stippled region
shows chloroplast. st, stigma.
p, pyrenoid surrounded by o,
starch, n, nucleus, v.p., pul-
sating vacuole whose role is
imperfectly known. (X 1650).

Guilliermond - Atkinson

42

Cytoplasm

encountered in many zoospores of the Chlorophyceae. In the
Diatomaceae, a group often considered as closely related to the
flagellated algae, there are a small number of large, plate-like
chloroplasts, often only two, occupying a marginal region of the
cell.

In the Chlorophyceae, the chloroplasts appear as large organ-
elles of rather complex structure. In the Ulothricales, for ex-
ample, there is only one chloroplast per cell (Ulothrix, Draparnal-
dia) in the shape of a deeply dentate ring girdling the elongated
cask-shaped cell near its middle. In the Siphonales and the Siphono-
cladiales, the chloroplasts are variously disposed and appear in
many different shapes. There may be one single chloroplast per
chamber or cell, distributed throughout the cytoplasm in the shape
of a network whose numerous swellings are occupied by pyren-
oids (Cladophora) , whereas in algae, such as Acetabularia,

.chr.

L — )

1

A. rhr^

Fig. 12. — Chloroplasts of the Chlorophyceae. 1, Draparnaldia
(After ScHMiTZ). 2, Zygnema. 3, Mougeotia scalaris. Two views.
(After Palla). 4, Oedogonium. (After Schmitz). chr, chloroplast.
py, pyrenoid. n, k, nucleus, a, s, starch, ky, tannin bodies of Palla.

the sections enclose numerous chloroplasts of irregular con-
tours, each chloroplast carrying several grains of carotin but not
having any pyrenoids and starch grains (Mangenot and Nardi).
In Vaucheria, the chloroplasts, distributed in great numbers in the
tubular thallus and never enclosing starch, are bodies of the dimen-
sion and form of those found in the phanerogams. They are also
small in Caulerpa and Derbesia, which, although lacking in pyren-
oids, seem often to elaborate starch by a rather strange and still
imperfectly known process (Ernst, Czurda).

In the Conjugatae, the chloroplasts are of a very complex form
to which particular attention must be called. There are only a
few or even only a single one present in each cell. Thus in Cos-
marium and Zygnema there are found in each cell two large star-
shaped chloroplasts whose long delicate arms radiate from a py-
renoid situated in the center. In others, Mougeotia for example,
the single chloroplast is a large smooth plate. A chloroplast of
M. laetevirens may attain a length of SOOix and a width of more

Chapter VI —43— The Plastlds

than 40fx. Each cell of Spirogyra contains one or more chloro-
plasts. Each is ribbon-shaped with slashed edges, rolled into a
spiral by more or less tight turns and contains numerous pyrenoids
lined up along its median region. Closterium contains several
large chloroplasts analogous to those of the Diatomaceae. They are
plate-shaped and occupy the marginal region of the cell.

In the brown algae or Phaeophyceae the chlorophyll is always
associated in the chloroplasts with a brown pigment of the carotin-
oid group, fucoxanthin. Because of this pigment which masks the
chlorophyll, the chloroplasts are often called phaeoplasts. Among
these algae, the Fucales (Fucus, Pelvetia, Cystosira) enclose
chlorophyll in all parts of the plant except in the antherozoids. The
phaeoplasts appear in all the cells as numerous bodies somewhat
analogous to those in the phanerogams. Chlorophyll is, however,
formed only in very small quantities in the apical cell and in the
oosphere where the spindle-shaped chloroplasts are very small.
In the antheridium, the phaeoplasts lose their chlorophyll which
is replaced by a crystal of carotin^ and distribute themselves among
the antherozoids in such a way that each of these encloses a single
plastid, lacking chlorophyll, but containing instead a carotinoid
pigment.

In the Laminariales, the phaeoplasts are encountered in all
cells equally and exist in the zoosporangia. Each zoospore con-
tains a typical small phaeoplast possessing a carotin crystal.

The red algae or Rhodophyceae, form in their chloroplasts, in
addition to chlorophyll, a red protein pigment, phycoerythrin. This
is present in such abundance in most species that it masks the
chlorophyll. The chloroplasts then are red in color and are called
rhodoplasts. In the Rhodophyceae, there are grouped together
organisms of very different structure: the lower Rhodophyceae,
the Bangiales, possess an extremely simple thallus composed of one
type of cell. The more evolved Rhodophyceae, the Rhodymeniaceae,
the Delesseriaceae, on the contrary, possess a complex vegetative
structure with differentiated tissues — meristem, assimilating, and
conducting tissue.

In the Bangiales and some Nemalionaceae, such as Chantran-
sia and Rhodochorton, which are very primitive red algae, the cells
are all similar and contain each a rhodoplast of characteristic form
(regular disc or star-shaped, ribbon-shaped, etc.), varying with the
species and provided with a pyrenoid. The chlorophyll apparatus,
therefore, in these algae shows the characteristics of that in the
inferior algae (Chlorophyceae). In other Rhodophyceae the shape
of the rhodoplasts is closely connected with the various parts of

^The term stigma or eyespot is used for an orange-red body found in the zoospores of
most flagellate algae and in the antherozoids of many green or brown algae. The stigma arises
sometimes, as is the case for the antherozoids of Fucus, from a plastid which loses its
chlorophyll and elaborates carotin; but it also often seems to be a differentiated portion of the
chloroplast in the ease where the cell contains only one large chloroplast or a specialized
chloroplast, as in the Euglenas. This stigma plays a photostatic role in cells. The carotin
seems to render the cell sensitive to light and the cell is oriented in the direction of
greatest or least light intensity (positive or negative phototactic response).

GuilHermond - Atkinson . — 44 — Cytoplasm

the vegetative structure, varying according to the cells in which
they are found. They are irregular, or angular, disc-shaped bodies
or twisted ribbons. Moreover, chlorophyll is lacking in the rhizoids
and sexual organs.

In the Charophytes there is little chlorophyll in the apical cell
and small ovoid chloroplasts are often found there containing a
large number of starch grains. In this group also, chlorophyll is
lacking in the egg and in the antherozoids.

In the bryophytes, the chloroplasts are numerous and similar
to those of the phanerogams. Yet in Anthoceros there is only one
crescent-shaped chloroplast per cell, situated near the nucleus. It
is transmitted by division from cell to cell.

It is therefore clearly proved by the study of algae such as the
Conjugatae where the chloroplasts are present in all cells, are
voluminous, and number only one or two per cell, that these organ-
elles are transmitted from cell to cell beginning with the egg.
However, their behavior during fertilization is not yet very clear,

for it is very difficult to observe them in the
zygote. In the Desmidiales (Closterium and
Cosmarium) , in which the cells contain two
chloroplasts, Klebahn has stated that they fuse
two by two in the zygote. Nevertheless, it is
difficult to accept this supposed fusion, for it
has never been observed in other cases. In
another Desmid, Hyalotheca dissiliens, in
which the cells have only one chloroplast, it
seems well established by the recent work of
PoTTHOFF that one of the two chloroplasts, de-
— A Dividin ^ived from each of the gametes brought to-
chioropiasts in 'leaT cells gether in the zygote, begins to degenerate and
of Eiodea canadensis. B. (jigappears vcry rapidly. This is likewise re-

Final stages in division *^^ ./!;'»' x rz J?

of one chloroplast. ported for other Conjugatae. In Zygnema, tor

example, where there are two chloroplasts in
each cell, four are found in the zygote (P. A. Dangeard and KuRS-
SANOW) and it seems well established by the work of Kurssanow
on Zygnema stellinum that the two chloroplasts carried by the
male gamete soon degenerate. This is also reported for Spiro-
gyra crassa (Chmielevsky) , S. longata and S. neglecta (Trondle) .
In the Conjugatae the chloroplasts of the new individual would
then have an exclusively maternal origin. In some algae, how-
ever, the sequence of events may be otherwise. In a diatom Rho-
palodia, for instance, each cell encloses a single chloroplast with
two pyrenoids. The cells destined to form the gametes divide in
their interior to form two gametes each. Now it appears that
the single chloroplast divides to furnish each gamete with half
a chloroplast possessing a single pyrenoid, because in the zygote
the two half-chloroplasts resulting from the union of the
gametes fuse to form a single chloroplast with two pyrenoids
(Klebahn).

Chapter VI

— 45 —

The Plastids

The chloroplasts in vascular plants: theory of Schimper:- In
vascular plants, pteridophytes and phanerogams, as has been said,
chlorophyll is not present in embryonic cells and is formed only in
organs exposed to light as the cells of the organs differentiate.
Chloroplasts are therefore found only in green tissue. The question
as to the origin of chloroplasts remained a long time unsolved. The
older botanists concluded that these bodies were formed by differ-
entiation from the cytoplasm or that they were the result of the
transformation of starch grains. Schimper (1881-1885) showed
for the first time that the chloroplasts are derived in reality from
small colorless bodies, protein in nature, which are found in all
the colorless tissues of plants. To these bodies he gave the name
leucoplastids. These were later called trophoplasts (Meyer) and
leucoleucites (VAN TiEGHEM). We shall call them leucoplasts^.
These are very small, rounded,
or rod-shaped elements which
are found scattered in great
numbers in the cytoplasm and
which are transmitted from cell
to cell by division. All these
leucoplasts have the ability to
form starch grains, i.e., to con-
dense in the form of starch, the
hexoses elaborated in green tis-
sue during photosynthesis and
later transported to colorless
tissues (roots, tubercles, etc.).
Whenever the hexoses accumu-
late in green tissues in too high
concentrations, the leucoplasts
become amyloplasts.

One of the classic examples
of this formation of starch
through the agency of the leuco-
plasts is to be found in the root
of Phajus grandifolius in which

Schimper succeeded in following the entire process. In this root,
the leucoplasts are indeed rather large, rod- or spindle-shaped,
bodies enclosing each along its axis a needle-shaped crystalloid cf
protein, the product of its elaboration. The starch grain arises in
a peripheral region of the thicker part of the leucoplast. Minute
at first, this grain grows and very quickly protrudes beyond the
plastid which it no longer covers except on one side.

Leucoplasts develop in various ways depending on the organs
in which they are found. In leaves, they have only to grow larger

Fig. 14. — Chloroplasts containing
protein crystalloids from fruit cells. 1,
Maxillaria triangularis. 2, 3, Cerinthe
minor. 4, Phajus grandifolius. (After
Schimper).

^Arthur Meyer distinguishes between the autoplasts (chloroplasts) and the trophoplasts,
the latter comprising the leucoplasts and the chromoplasts. Van Tieghem replaces the term
of plastid or plast by that of leucite and distinguishes the leucoleucites (leucoplasts), the chloro-
leucites (chloroplasts), and the chromoleucites (chromoplasts). The term plastid having been
used by some biologists to mean cell, we replace it therefore by that of plast (which signifies
elaborative body) which seems to us much more justified than that of leucite (white body)
and we shall distinguish the leucoplasts, amyloplasts, chloroplasts and chromoplasts.

Guilliermond - Atkinson

— 46

Cytoplasm

and become green to be transformed into chloroplasts (called
chloroplastids by Schimper). In flowers and fruits, Schimper
has shown that the carotinoids, xanthophyll and carotin, always
form in plastids which he calls chromoplastids but which we shall
call chromoplasts. These may either form directly by transforma-
tion of small leucoplasts, which may or may not have previously
formed starch, or they may, in other cases, arise by a metamorpho-
sis of chloroplasts whose chlorophyll disappears and is then re-
placed by a carotinoid. Pigment may appear in the chromoplast
either in the amorphous or in the crystalline state. In the former,
which is always the case when the pigment is xanthophyll, the
chromoplasts affect the same globular or lenticular form as the
chloroplasts. In the crystalline state, which is frequently the case
when the pigment is carotin, the crystals are contained in variable
number in the chromoplasts, and usually take the form of needles

Fig. 15. — Inclusions of the plastids. 1, chloro-
plasts in epidermal cells of Hedera leaves. 2, chlo-
roplasts in palisade parenchyma of leaves of Aehy-
ranthes Verschaffelti. 3, leucoplasts in young buds
of Carina Warszetvickzii. A, starch; C, protein crys-
talloids. 4, chromoplasts in the flower of Neottia
Nidus-avis. C, carotinoid crystal; P, protein crys-
talloid. (After Schimper).

delicately curved in hooks, spirals or slender rods. Sometimes they
appear as triangular or rectangular tables, hollow tubes, or spiral
ribbons. They are all rhomboidal prisms. The crystals are ar-
ranged in parallel or divergent bundles in the chromoplast which,
following the contours of the crystal, takes on most varied aspects :
rods, spindles, triangles. In some cases, as in the carrot root, the
crystals, when once formed, have worn out the greater part of the
substance of the chromoplast which made them, and appear either
free in the cytoplasm or simply surrounded by a very thin and
barely perceptible plastidial envelope. In other cases, there may
be formed in chromoplasts not enclosing crystalline pigments, sev-
eral needle-shaped crystalloids of protein which bring about an
elongation of the plastid and give to it the shape of a spindle or
rod. It even happens that chloroplasts contain both pigment crys-
tals and protein crystalloids as, for example, in the fruit of Loni-
cera xylosteum, in which the chromoplasts are pear-shaped. The
pigment is distributed m the thicker portion of the chromoplast
as small needle-shaped crystals, and a protein crystalloid occupies
the long axis of the chromoplast and spins out at one of its extremi-
ties in the shape of a tenuous appendage. (Figs. 14-17).

Chapter VI

47 —

The Plastids

ocfDoo
o

o

(9

o

o o

^ >1

All these organelles - leucoplasts, chloroplasts, chromoplasts -
belong therefore to the same category of elements to which SCHIM-
PER gave the general term plastids which we shall replace here by
that of plast (Fr. plastesy. They have multiple potentialities,
manifestations of which vary with the organ in which the plastids
are found. They may remain in the state of leucoplasts and have
as their only function the condensation of hexoses into starch,
playing the role of amyloplasts; or else they may be transformed
into chloroplasts or chromoplasts; or from chloroplasts they may
be changed into chromoplasts. They all have
the ability to elaborate starch and to produce
protein crystalloids within themselves in the
manner already described for leucoplasts of
the root of Phajus grandifolius and for the
chromoplasts of the fruit of Lonicera Xylos-
teum. Thus chloroplasts of Cerinthe minor
are traversed by a needle-shaped crystalloid
of protein which is prolonged freely at its
two extremities.

SCHIMPER considers that starch can not
arise directly in the cytoplasm but is always
the product of plastidial activity. He de-
scribes in all details the process of starch
formation through the agency of the plas-
tids whether they be leucoplasts or chloro-
plasts. Let us recall that starch grains grow
by apposition and are made up of a dark
hilum surrounded by alternately light and
dark layers. The hilum is the portion first
formed and the alternate layers correspond
to zones of different water content which are
developed about the hilum during the growth
of the starch grains. The hilum may occupy
the center of the grain (central hilum) and
is then surrounded by regular concentric lay-
ers, or it may be situated at one of the poles
of the grain whose concentric layers widen
and become more and more numerous at the
opposite pole (eccentric hilum). There are simple starch grains,
compound starch grains, i.e., several grains stuck together, and
half-compound starch grains which are separate at first but be-
come united by common concentric layers. The difference in type
depends upon the method of formation within the plastid. The
simple grain with central hilum arises in the middle of the plastid
as a small granule which grows, forming regular concentric layers
about the hilum, and remains surrounded on all sides by a plastidial
wall which grows thinner as the grain enlarges. The simple grain

® i ^ D a

Fig. 16. — Transforma-
tions of chloroplasts in meso-
phyll cells of petals. A,
Lilium tigrinum. 1, caro-
tin granules at the periphery
of green plastids; 2, starch
grains in plastids whose
chlorophyll has disappeared;
3, completely formed chromo-
plasts whose starch has been
absorbed. B, Gladiolus var.
1-3, chloroplasts with large
starch grains; 4-7, xantho-
phyll replaces chlorophyll,
starch is absorbed, carotin
granules appear; 8-13, caro-
tin crystals appear.

^Translator's note. The French words plastides and plastes are rendered in English by the
one word plastid, so that this distinction loses its force in translation. "Plast" is used only
in compound words such as "chloroplast", "chromoplast", "leucoplast" etc.

Guilliermond - Atkinson

— 48

Cytoplasm

with eccentric hilum, on the contrary, arises at a point near the
periphery of the plastid and, in this case, by enlarging, soon pro-
jects beyond the plastid, which no longer covers it except at one of
its poles where it takes the form of a cap. The hilum surrounded
by the earliest formed layers is then found at the extremity of the
grain opposite the plastidial cap. The grain, no longer coming
under the influence of the plastid in that region, ceases to grow
and enlargement no longer takes place except at the point of con-
tact with the plastid, i.e., at the pole opposite the hilum. There
the concentric layers become increasingly numerous and thick.
The compound grain arises by a plastid forming several starch
grains inside itself, instead of only one. These are in contact with
one another and remain small. They are semi-compound grains
if they become surrounded by common concentric layers.

On the basis of all his investi-
gations, as well as those of
SCHMITZ on the algae, Schimper
was led to consider the plastids
as component parts of the cell,
incapable of arising de novo, be-
ing transmitted by division from
cell to cell beginning with the
egg, so that plastids of higher
plants, according to him, are com-
parable to the chloroplasts which
are encountered permanently in
many algae, but with this differ-
ence : in the algae the plastids al-
ways keep their chlorophyll and
remain as green plastids, while in
higher plants they appear first as
leucoplasts and do not become
chloroplasts except in stem and

leaf tissue.

Arthur Meyer (1883) working at the same time, confirmed
the theory of Schimper which was verified also as far as the
chromoplasts are concerned by Courchet (1888).

It is fitting to call attention to the observation that the plas-
tids, at the same time that they are elaborating starch and pig-
ments, are capable of producing inside themselves small re-
fracting granules which reduce osmic acid (Fig. 18). To these,
which are sometimes very numerous, a lipide constitution has been
attributed. These granules were described long ago by Nageli,
Godlewski, Schimper, Meyer. They have been the object of
more recent research by Meyer who has opposed the idea of their
lipide nature and considers them to be composed of a-(3 hexylene-
aldehyde, a waste product of photosynthetic assimilation found in
the products of the distillation of the leaves. Meyer calls these
formations elaborated by the plastids Autoplastensekret. We shall
see, nevertheless, that this interpretation has not been confirmed

,^£5^==^

F^G. 17. — Various aspects of chromo-
plasts derived from chloroplasts in the meso-
carp of the fruit of Rosa canina (in vivo) .

Chapter VI — 49 — The Plastids

and that these granules present the histochemical characteris-
tics of lipides and not those of aldehydes (Cf. p. 209).

Just what these granules signify is still very obscure. In
cells in which the plastids never form starch, the granules are
often considered as assimilation products replacing starch. It has
been noticed, sometimes, that these granules appear in large
numbers during the period preceding the formation of starch and
of chlorophyll and carotinoid pigments, only to disappear as soon
as these products have been formed. It was therefore thought
that they might be an intermediate product contributing to the
formation of starch or pigment. It has also been shown that very
frequently similar granules appear in great numbers in plas-
tids as they degenerate in the cells of flowers which are beginning
to take form. In this case, the presence of the granules can
only be explained as a breaking down of the plas-
tidial lipoprotein complex, i.e., as a process called
lipophanerosis (demasking of lipides. Cf. p. 203,
205).

Finally, large globules have recently been de-
scribed in the plastids of various Cactaceae
(Cephalocereus, Echinocereus, etc.). They pre-
sent the histochemical characteristics of phyto-
sterol and form in the plastid exactly as do
starch grains. These globules always precede
starch formation and disappear the moment that
starch appears. It was therefore supposed by ^J^^;i[^n7 ^Sfof
those who did the work that these globules of ™^ Veig'^a^ith lip-
phytosterol constituted a material which served '^^ granules in the

. "^ , „, 1 , r^ ■»«•• chloroplasts.

m the buildmg up of starch (Savelli, Miss
Manuel) (Fig. 19).

For a very long time it was impossible to obtain paraffin sec-
tions of preserved and stained plastids, at least of leucoplasts,
for they were destroyed by all fixatives then used, and only the
much more resistant chloroplasts were obtained. Schimper's and
Meyer's observations were therefore made exclusively from very
well executed studies of living material. These observations were
necessarily incomplete for it is absolutely impossible to distinguish
plastids in living embryonic cells and until the present writing they
have been seen only with the very greatest difficulty in colorless
differentiated tissue, such as the root, and then only in favorable
cases. The classical observations of Schimper on the method of
starch formation were based for the most part on examples par-
ticularly favorable for study, such as the root of Phajus grandi-
folius, in which the leucoplasts are very massive. Therefore the
theory of SCHIMPER and Meyer was very largely hypothetical and
was based chiefly on what seemed to them likely to be true, and
on what can be observed in many algae in which the chloroplasts
behave like component parts of the cell.

Numerous workers contested this theory of starch formation,
among others, Belzung, who did not succeed in distinguishing

Guilliermond - Atkinson — 50 — Cytoplasm

leucoplasts in living cells and asserted that starch arises most
often within the cytoplasm itself and that chloroplasts form by
cytoplasmic differentiation. The development of the plastids was
very imperfectly known and even the most characteristic forms
which the plastids take on were not known. Progress in this
matter was not marked until much later, from 1910 on, when re-
search was carried out with so-called mitochondrial technique
which makes possible the preservation of unaltered plastids, clearly
stained on paraffin sections. This will be taken up later. These
methods have made it possible to follow with the greatest precision
the entire life history of the plastids during cellular development ;
to bring out different aspects which up to then had not been per-
ceived; and to make important discoveries which have confirmed
the ideas of Schimper and Meyer by completing them and giving
them precision.

Chemical nature and structure of plastids :-

Plastids for a long time were thought to be ex-
clusively protein in nature. It was however be-
lieved, without anyone being able to furnish proof
for it, that plastids must enclose a lipide sub-
stance, probably lecithin, in which the chlorophyll
was supposed to be dissolved. Gautier, Hoppe-
Seyler and Stoklasa even formulated the hy-
pothesis that chlorophyll is a chlorolecithin. The
presence of lipides in chloroplasts is not ques-
tioned today and Menke, chiefly, seems to have
piS ilf 'aTaSIcS- furnished a proof of it. Granick, by triturating
matous cell of Echi- g^^d then ccutrifuging tobacco and tomato leaves,

noeereita procumbens iij-ij j^- i.-j_ j?ii„

showing large inciu- was able to isolate a certam quantity oi chloro-
^}^^^ °KJ?^^^'if^''°^' plasts and to obtain their microchemical analysis.

/ Affpjt Miss Manu-

EL). Granick thus managed to separate proteins from

lipides. Recent work, which will be discussed
later, has shown by histochemical reactions that in reality all plas-
tids, like the cytoplasm, are composed of lipoproteins in which,
however, the lipides (compound lipides, probably lecithins) are
much richer than in the cytoplasm and give to the plastids their
color characteristics.

The importance of the plastids in photosynthetic assimilation
has led authors for a long time to attribute a structure to these
organelles. The leucoplasts appear homogeneous and the chloro-
plasts, and chromoplasts especially, have been studied from this
point of view.

Let us recall that these organelles are formed of a lipoprotein
substratum on which the chlorophyll is fixed (chlorophyll a and
b), accompanied by a small quantity of carotinoid pigments (caro-
tin and xanthophyll).

Pringsheim, Tschirch, and Chodat held that the chloroplasts
consist of a spongy material impregnated with a fluid substance of
an oily consistency (lipochlorine of Pringsheim) serving as a sub-

Chapter VI —51— The Plastids

stratum for the pigments. Schmitz attributed a fibrillar structure
to algal chloroplasts and Schwarz considered the chloroplasts of
higher plants to be composed of two substances, the one, chloro-
plastin, formed of basic filaments, lying closely together, containing
chlorophyll grains, and the other a colorless substance, metaxin,
interposed among the filaments.

SCHIMPER and Meyer held that chlorophyll in chloroplasts was
found as inclusions, the grana, so small as to border on the limits
of visibility. These grana they found to be very numerous and
diflficult to distinguish, giving the impression that the piginent is
found in a diffuse state in the plastidial substratum. According to
these two investigators, xanthophyll is distributed in the same way
in chromoplasts and only carotin, when it is not in a crystalline
state, exists in the form of clearly differentiated grana.

Various arguments of a theoretical nature next led physiologists
to think that the chloroplasts must have a homogeneous structure.
A comparative study of the spectra of a chlorophyll solution and
of a living green leaf show, indeed, that the absorption bands do
not occupy the same positions in the two cases, and that those for
the living leaf coincide with those shown for a colloidal solution of
chlorophyll. It is known that solutions of chlorophyll break down
rapidly in light in the presence of oxygen. There is then an
oxidation of chlorophyll. In the chloroplast, on the contrary, the
chlorophyll manifests a great stability which could only be shown
by a colloidal solution of chlorophyll united to other colorless col-
loids. Certain facts, furthermore, lead to the supposition that
chlorophyll may be combined in the plastidial substratum either
with proteins or with lipides. Hence it has been thought that
chlorophyll must be uniformly distributed throughout the entire
mass of the chloroplast and that the latter is constituted of a
hydrogel formed of various colloids, some colorless, others colored
(Siebold), or formed of colloids containing chlorophyll chemically
combined with colorless colloids. According to Ponomarew, Le-
PESCHKIN, and Kuster, the gel of the chloroplast is in a semi-fluid
state and the variations of form undergone by this organelle may
be explained by a too weak tension between the cytoplasm and the
chloroplast. Investigations of Price, Scarth, Lapicque, Guillier-
MOND and Mangenot have shown, moreover, that chloroplasts, al-
though sometimes giving the impression in direct light of having
structure, are always optically empty under the ultramicroscope
and are not seen except for their color and their faintly luminous
contours. This confirms the opinion stated above. Finally, mito-
chondrial technique, which best conserves the chloroplasts and
makes their lipides insoluble, always reveals these organelles as
absolutely homogeneous, while those techniques which dissolve
lipides bring out a heterogeneous structure which may be supposed
to correspond to an alteration.

More recent cytological work, however, would tend to prove, on
the contrary, that the chloroplasts do have a structure. According
to ZiRKLE, they are formed of a chlorophyll-containing hydrogel

Guilliermond - Atkinson — 52 — Cytoplasm

pierced by colorless pores, encircling a central vacuole which, when
visible, contains starch. The plastid is coated by an adsorbed layer
of cytoplasm. ZiRKLE thinks starch is formed in the vacuole of
these chloroplasts. Living chloroplasts, however, are very delicate
organelles and one runs the risk of some alteration taking place
during observation, so that it is wise to be very cautious in the
acceptance of this structure which has never received any con-
firmation.

More interesting is the later work of numerous authors whose
results are in agreement and will be briefly analyzed here. In the
first place the works of Menke, Kuster, Heitz, and Friedl Weber
have shown that chloroplasts are clearly birefringent. This is
particularly easy to demonstrate in the chloroplasts of the Con-
jugatae {Mougeotia, Zygnema, Closterium, Spirogyra) . In polar-
ized light the chloroplasts are luminous and show a green polariza-
tion color. Menke found that in the chloroplasts of the Diatoma-
ceae, the polarization colors vary from reddish brown to red. Ac-
cording to Weber, reddish brown to red polarization colors are not
peculiar to the chloroplasts of the algae (diatoms, Spirogyra, Zyg-
nema, Vaucheria) but are also found, together with green polariza-
tion colors, in the chloroplasts of higher plants (Polygonatum
officinale, Bellis perennis, Elodea canadensis). Weber finds that
the perception of these colors depends upon the intensity of the
light in which the plants are observed. Furthermore, the investi-
gations of Menke, Doutreligne, Wakker, Wieler, Heitz, Weber,
Hubert, Geitler, Deschendorfer, Beauverie, Weier and Strug-
GER seem to confirm in chloroplasts the existence of a structure,
described by Schimper and Meyer fifty years ago, according to
which chlorophyll is fixed on grana suspended in a colorless stromal
These grana according to Wieler are enclosed within a peripheral
layer of the stroma. Heitz thinks the grana are small bodies,
measuring from 0.3-1.7/i. in diameter, and appearing flat and
discoid in shape, which explains the striated structure often visible
in chloroplasts when observed in certain positions. Doutreligne
describes them as rods or granules which, according to the condi-
tions present, may assemble like strings of beads, may separate,
or may become confluent. BEAUVERIE confirms these data without
admitting, however, that the structure is general, for he thinks
that in certain plants the chloroplasts may be homogeneous.

The chemical nature of the grana is still unknown. Wieler
concludes that they are formed of an essential oil in which the
chlorophyll is dissolved. Heitz considers them simply as grains

'Some even older observations would seem to support this view. Chodat, for instance, in
the pseudobulb of Calanthe Sieboldi, described chloroplasts which are round at first but capable
of changing to a dumb-bell shape by elongation and stretching at their middle region. Now
the two swellings remain filled with chlorophyll while the slender part connecting them becomes
colorless. In the mesophyll of leaves and bracts of Iris germanica as well, we have seen large
chloroplasts showing at one extremity, or both, a sort of slender, colorless appendage.
Embebger observed that in the bulb scales of Lilium eandidum the chloroplasts are surrounded
by a colorless layer which seems to indicate that chlorophyll, like the starch grain, has been
laid down within the plastid.

Chapter VI —53— The Plastlds

of chlorophyll. Still others think they are lipides holding the pig-
ment in solution (Menke, Weber).

With the study of this structure the question of the double re-
fraction of chloroplasts has been brought up again. While KtJSTER
concludes that double refraction is due to the ground substance of
the chloroplast, Menke, Heitz, and Weber attribute it to the
grana which, accordingly, they believe correspond to doubly refrac-
tive lipide inclusions. Menke and Weber observed that after a
prolonged stay in water, the chloroplasts are the seat of the pro-
duction of green myelin filaments which correspond to the fibrils
of SCHWARZ and which arise from the grana^ Now WEBER
showed that these myelin figures when seen in polarized light pre-
sent the same double refraction and same red color as do the grana,
which he thinks tends to prove that in the chloroplasts only the
grana are doubly refractive and that these elements correspond
to liquid cristals^.

It is known, however, that GiROUD had recorded that the chon-
driosomes of animal cells, which we will see later on are akin to
plastids, appear doubly refractive. Weber to confirm his opinion
examined chondriosomes and leucoplasts of various plants in polar-
ized light but was not able with certainty to prove that they are
doubly refractive. It is known, of course, that carotin in chromo-
plasts, when it is not crystallized, appears as clearly separated
granules in the colorless substance of the plastid. Now in examin-
ing such chromoplasts in polarized light, Weber was able to demon-
strate that the double refraction is localized exclusively at the
level of the pigment granules. This seems to demonstrate that the
substratum of the plastids is not doubly refractive and that this
characteristic is due to the grana in chromoplasts and chloroplasts
alike. The problem of the structure of the chloroplasts is certainly
very complicated, since it is difficult to avoid alterations in living
material during observation, but the tendencies are manifestly in
favor of a heterogeneous structure. Most authors (Heitz, Doutre-
LiGNE, WiELER, Deschendorfer, Weber, Pekarek, Geitler) agree
that the chloroplasts are composed of small lipide discs containing
chlorophyll and embedded in a hydrophilic stroma. Weber thinks
these discs are responsible for the fiuorescence of chlorophyll and
the double refraction of the chloroplasts.

Recently Frey-Wyssling has attributed a lamellate submicro-
scopic structure to these discs, consisting of a series of parallel
layers of protein, lecithin, chlorophyll and carotinoid pigment.
The scheme of this author implies that the phytol groups of the
chlorophyll molecules are interposed between aliphatic chains of the
lecithins, the porphin residues being arranged in a monomolecular

'By placing Spirogyra in a 1-2% solution of sodium oleate, Weber saw the spiral bands
of the chloroplast swell and form protuberances on their surfaces reminiscent of myelin
figures and sphaerocrystals. These formations are green like the chloroplasts and correspond
to the lipide phase of the chloroplast.

^Webeb thinks these grana may correspond to the granules considered by Meyer to be
formed of crP hexylene-aldehyde, a product of photosynthesis, which we will speak of again
farther on.

Guilliermond - Atkinson

54 —

Cytoplasm

layer. This supposed submicroscopic structure makes it possible
to form new hypotheses for the mechanism of chlorophyll assimila-
tion (Baas Becking and Hanson, 1937).

As for the pyrenoids which are found in the chloroplasts of
many algae, they are still variously considered and opinions as to
their significance are not yet well fixed. SCHMITZ, Chmielevsky,
and LUTMAN consider them permanent organelles, multiplying by
division. Schimper, Klebs and Strasburger saw them disappear
and maintain that they are bodies which form de novo, an opinion
which today seems demonstrated. It was thought that the pyren-
oids constituted a reserve protein elaborated by the plastids. Other
authors, on the contrary, think they play an important role in the
foiTnation of starch which always arises in their interior (Lutman,
McAllister) or in contact with them (Chadefaud).

In terminating this discussion as to the structure and chemical
nature of chloroplasts, attention is called to the ability of chloro-
plasts to reduce silver salts, especialFy in a 1% solution of silver

A B C

F'iG. 20. — Positions taken by chloroplasts in cells of seedlings of Saccorhiza bulhosa in
intense (A) and weak (C) illumination. B, intermediate position. (After Sauvageau).

nitrate. This property, discovered by MOLISCH and called the Mo-
LISCH reaction, is connected neither with chlorophyll nor the caro-
tinoid pigments, is produced only in living tissues and seemed to
MOLISCH to be independent of light. MOLISCH attributed it to the
presence of formaldehyde in the chloroplasts. Investigations of
Gautheret have led to important information about the conditions
under which the MOLISCH reaction is carried out. This research
showed that light plays a role in the reduction of silver nitrate by
chloroplasts: the reduction begun by the action of light may con-
tinue subsequently in the dark. Gautheret's work proved finally
that this reaction is not due to the presence of formaldehyde in the
chloroplast but to reducing substances still unknown.

More recently GiROUD and his collaborators claimed that the
Molisch reaction is to be attributed to the presence of ascorbic acid
within the chloroplasts which, therefore, these workers believe
to be the source of this substance.

Movement of chloroplasts:- It has been known for some time
that chloroplasts are capable of moving from place to place and,
depending upon the intensity of the light, are capable of placing
themselves on one side or the other of a cell. But for a long time
it has been said that they play only a passive role in these move-

Chapter VI —55— The Plastids

ments, which those who hold this opinion think are caused ex-
clusively by cytoplasmic currents. From observations of WEISS,
SCHIMPER, and Kuster it appears that chloroplasts are capable of
becoming distorted and of showing amoeboid movements. This is
also recorded for leucoplasts, as will be seen later. Now Sau-
VAGEAU showed by observations of seedlings of Saccorhiza bulbosa
that when exposed to intense light the chloroplasts, during their
movements, present contractions and dilations which can be ex-
plained only by movements which they make themselves. Senn,
who recognizes the capacity of chloroplasts to move themselves,
attributes this to the presence of a cytoplasmic sheath surrounding
them which he calls a peristromium. This, according to Senn,
gives rise to pseudopodia which permit the chloroplasts to change
shape and place. The existence of this peristromium, however, has
never been confirmed and remains very problematical.

Chapter VII
THE CHONDRIOME

General conceptions. What is meant by chondriome in animal
cells:- As the chondriome was observed first in animal cells, it
seems necessary before beginning its study in plant cells, to recall
as briefly as possible what is understood under that heading in
animal cytology.

The observations of Altmann cited in the previous chapter,
although exact, did not at first hold the interest of cytologists and
the bioplasts described by him were for a very long time confused
with the granules of Arnold brought out in most animal cells in
the time that followed by means of vital stains. It is known now
that these granules correspond to vacuoles and, in consequence,
have a quite different significance. It was believed that these
granules were merely artifacts and it was not until very much later
that the work of Benda, Meves, Regaud and Faure-Fremiet dem-
onstrated, by the use of special methods similar to those of Alt-
mann, the constant presence of small organelles in the cytoplasm
of animal cells. (Figs, 21, 22). They were somewhat similar in
shape and dimension to bacteria and were afterwards identified with
the bioplasts of Altmann and the fila of Flemming. These ele-
ments, whose width does not exceed 0.5-1^, were named chondrio-
somes by Benda (1906) and plastosomes by Meves. The general
term mitochondria is often applied to them. They appear now as
granules called mitochondria (Benda), now as rods or long, undu-
lating, sometimes branched filaments called chondrioconts. One of
these shapes may change into the other. The granule is capable of
elongating into a rod, then into a filament and this latter may, in
turn, fragment into granules. Mitochondria assembled in small
chains were called chondriomites by Meves. It was later recognized
that this very rare shape sometimes represents a transition form re-
sulting from a fragmentation of chondrioconts but more often still,
corresponds merely to an alteration of the chondrioconts caused by
the fixatives (Levi). Meves proposed the term c/ionc^nome (1908)
or plastome (1910) for the entire chondriosomal content of a single
cell.

The chondriosomes are made up of lipoproteins, very rich in
lipides (phosphoaminolipides), and can not be brought into evi-
dence except by the use of special fixation techniques, called mito-
chondrial, which do not affect the phosphoaminolipides. Fixatives
containing alcohol or acetic acid, i.e., those which were most cur-
rently used before the discovery of chondriosomes, do not reveal
them. This explains why, although visible in living material, they
were able to pass unperceived for so long. Mitochondrial tech-
niques consist in fixing cells with a mixture of chromic and osmic

Chapter VII

— 57 —

The Chondriome

acids (method of Benda and Meves) or with a mixture of formol
and potassium bichromate (method of Regaud), and then in fol-
lowing the fixation with a more or less prolonged treatment of a
3% solution of potassium bichromate, an operation called post-
chromatization which renders the chondriosomal lipides insoluble.
Once fixed, the chondriosomes stain clearly with iron haematoxylin,
acid fuchsin and crystal violet. They appear in the homogeneous,
barely-stained cytoplasm as intensely stained elements, with a very
clear outline and they very much resemble bacteria. It has been
shown that chondriosomes are made up of a lipoprotein complex
and that their affinity for stains is due to the lipides which they
enclose. Fixatives containing alcohol or acetic acid destroy these
lipides and the chondriosomes lose their chromaticity.

In young cells, mitochondria generally predominate among the
chondriosomes. At a later stage they elongate into chondrioconts
which is the most usual form found in mature cells. The chondrio-
somes are permanent formations and
many cytologists consider that they are
incapable of forming de novo and increase
in number only by division of pre-existing
chondriosomes. At first the chondrio-
somes were regarded as organelles in
whose interior were formed most of the
products elaborated in the cell (fat, zymo-
gen, pigments) and whose role was the
same as that of the plastids in chlorophyll-
bearing plants. But observation of living

material using tissue-culture technique ..

does not confirm this idea. The very accurate observations of Noel
on the liver of rats are the only ones made so far which seem favor-
able to this belief. Noel has shown that when an exclusively nitro-
genous diet is given the rats, the chondriosomes of their liver cells,
which are normally in the state of chondrioconts, become large,
round bodies filled with protein. It seems, therefore, that the
chondriosomes accumulate protein and act as proteoplasts. The
phenomenon is reversible and, if the nitrogenous food is sup-
pressed, the proteoplasts lose their protein and resume the form
of chondrioconts. The role of the chondriosomes is still very ob-
scure in spite of this single observation.

The chondriome In plant cells:- Chondriosomes in plant cells
were discovered even as early as 1904 by Meves who found them
in the nurse cells of pollen of the Nymphaeaceae. His results were
subsequently confirmed in various organs of the phanerogams by
the research of a certain number of investigators (Smirnow,
NicoLOSi-RoNCATi, Bonaventura) , among whom Duesberg and
Hoven will be given a special place, for in 1910 they produced ex-
cellent figures of the chondriome in embryonic cells of the pea and
bean.

Fig. 21. — The chondriome.
A, frog's liver. B, salamander's
liver. C, frog's kidney. Re-
gaud's method.

Guilliermond - Atkinson

— 58

Cytoplasm

Beginning with the year 1910, the aspect of the question
changed and the almost simultaneous work of Pensa (1910),
Lewitsky (1911), and our own work (1911) showed a relationship
between chondriosomes and chloroplasts. But, as will be seen,
from that moment on, investigators found themselves face to face
with a problem which it took several years of patient and laborious
research to solve.

A study of chondriosomes in the different plant groups has
demonstrated that these elements exist in every cell except, how-
ever, among the bacteria, where it has not yet been possible to
reveal them, and in the Cyanophyceae in whose cells it is at present
demonstrated that they are not found.

Fig. 22. — Connective cells from tissue cul-
tures of guinea pig prepared by the mitochon-
drial method. (After Maximov).

The chondriome in fungi:- The study of the chondriosomes is
relatively simple in plants lacking in chlorophyll, i.e., the fungi,
where we described them for the first time in the ascus of Pustularia
vesiculosa (1911). Chondriosomes were, after that, cited in the

most varied fungal groups:
Myxomycetes (VONWILLER,

CowDRY, Lewitsky, Mange-
not), Plasmodiophoraceae
(MiLOViDOV), Chytridiaceae
(PoissoN and Mangenot),
Blastocladiaceae (Winslow
Hatch), Saprolegniaceae (Ru-
dolph, Guilliermond), Pero-
nosporaceae (Lewitsky, H.
Edson, Dufrenoy, Saksena,
Miss Syngalowsky), Der-
matophytes (Grigoraki, Ne-
groni), Mucoraceae (Guil-
liermond, MoREAu), Hemiascomycetes (Guilliermond, Varit-
CHAK), lower Ascomycetes: Endomyces Magnusii, Endomyces fibu-
liger (Guilliermond), yeasts (Janssens and van de Putte,
Guilliermond, Henneberg, Negroni, Tredici, Verona), higher
Ascomycetes: Pezizales, Penicillium glaucum (Guilliermond,
Jannsens and Helsmortel), Ustilaginaceae and Uredinaceae
(M. and Mme. Moreau, Beauverie), Autobasidiomycetes (Guil-
liermond, Beauverie, Sarazin, Miss Duchaussoy).

In almost all fungi, the chondriosomes predominate in the form
of chondrioconts, generally very elongated, sometimes branched,
and orientated in a direction parallel to the longitudinal axis of the
hyphae. In the Myxomycetes and Plasmodiophoraceae, however,
there are present only mitochondria or short rods.

Development of the chondriome:- The chondriosomes are found
at all times in the cytoplasm of fungi. They are distributed among
the spores in the sporangia and asci and are likewise found in the
conidia {Penicillium glaucum) and in the buds of yeasts (Fig. 23).

Chapter VII

59

The Chondriome

In the Myxomycetes and Plasmodiophoraceae in which CoWDRY,
Lewitsky, and Milovidov have studied the chondriosomes in all
stages of development (spores, zoospores, myxamoebae, plasmodia),
these elements remain constantly in the state of mitochondria or
short rods and never become chondrioconts. During sporogenesis
they are distributed among the spores (Fig. 24).

The development of the chondriome in one of the Blastocladia-
ceae, Allomyces arbusculus, is known from a recent study of WlNS-
Low Hatch. In the mycelium the chondriome is represented ex-
clusively by long slender chondrioconts which appear to thicken
at the extremities of the hyphae. At the time of gametogenesis,
walls at the extremities of the hyphae cut off two gametangia, the
female being terminal, the male subterminal, both enclosing nu-
merous chondrioconts which are
more abundant in the female
than in the male gametangium.
The chondriosomes are distrib-
uted about the various nuclei of
the two gametangia, forming
around each nucleus an en-
tangled network of chondrio-
conts. Then, at the close of
gametogenesis, the chondrio-
conts in each gamete undergo a
fragmentation by which they are
reduced to numerous, very small,
mitochondria. These subse-
quently grow, then fuse, forming
about each nucleus a sort of
reticulate mantle which seems to
be transformed later into a chro-
matic, homogeneous cap ap-
pressed to the nucleus on one
side (nuclear cap). Each gamete
when mature encloses, there-
fore, a nuclear cap seemingly of
mitochondrial origin, occupying
the regions of the cell opposite to the insertion pomt of the flagel-
lum. Hatch compares this cap to the limosphere of moss anthero-
zoids and to the "Nebenkern" of some animal spermatozoids (Dip-
tera). Nevertheless the mitochondrial origin of this nuclear cap
seems still to demand some verification (Fig. 25).

The development of the chondriosomes is known particularly m
the Saprolegniaceae in which we have been able to follow different
species {Saprolegnia, Achlya, Leptomitus) with the greatest accu-
racy during their entire development, not includmg the sexual pro-
cess (Figs. 26, 27) . The chondriosomes of these fungi appear m the
extremities of growing hyphae as relatively large mitochondria.
Immediately behind the tip, these elements begin to elongate, first
becoming rods, then thin, undulating and often branched, chondrio-

r

■^'

Fig. 23. — The chondriome in fungi. A,
developing basidium of Coprinus; B, young
basidia of Psalliota cainpeatris; C, young
sporangium of Rhizopus nigricans: D. conidio-
phore of Penicillium glaucum; E, tissue from
the foot of Psalliota campestris; F. yeast,
Sporoboloviyces roseus.

Guilliermond - Atkinson

— 60 —

Cytoplasm

conts. The chondrioconts become thinner and thinner as they
elongate which seems to indicate that they are formed by a pro-
gressive stretching of the mitochondria. In the young zoosporan-
gia, the chondriosomes are all in the state of mitochondria which
subsequently, together with the cytoplasm, assemble about each of
the numerous nuclei, thus outlining the circumferences of the fu-
ture zoospores which are seen as separated by hyaline areas. All
the zoospores when mature enclose a chondriome made up exclu-
sively of mitochondria which, at the time of germination, are trans-
formed in the germinating tube into chondrioconts.

The development of the chondriosomes is also very well known

•J?*,

\ •

"vs^i

«iP

^w,.;

■'■■\Zy D

-A-ii

Fig. 24 (left). — The chondriome in Myxomycetes and the Plasmodiophoraceae. A, frag-
ment of the Plasmodium of Physarum (original) ; B, spores of Fuligo septica (after Lewitsky) ;
C, spores of Hemitrichia vesparum (after Cowdry) ; D, spores and zoospores of Fuligo septica
(after Vonwiller) ; E, young Plasmodia of Plasmodiophora Brassicae (after Milovidov).

Fig. 25 (right). — Allomyces arbiisculus. 1, elongated chondrioconts in the vegetative fila-
ment. 2, 3, the chondrioconts grouped about the nuclei in the gametangium. 4, mitochondria
surrounding the spores as they are cut out. C, chondrioconts; L.G., lipide granules colored
brown with osmic acid; N, nucleus. Champy-KuU method. (After Hatch).

in the ascus of Pustularia vesiculosa. It was here that they were
studied for the first time in the fungi by us and then by JANSSENS
and Helsmortel. They present a series of interesting phenomena.
All of the ascogenous hyphae, those from which the asci will be
formed, show a chondriome made up of numerous chondrioconts
densely clustered about each nucleus. It is known that the ascus
forms at the terminal portion of these hyphae. This terminal,
recurved, crosier-like portion is occupied by a binucleate cell whose
two nuclei divide simultaneously. Then two transverse walls form,
marking off three cells, of which the middle one, that occupying the
arch of the crosier, encloses two nuclei whereas the others, that at
the extreme tip and that at the base, contain only a single nucleus.
The middle binucleate cell destined to form the ascus, contains at
first, around each of its two nuclei, a small mass made up of the

Chapter VII

61 —

The Chondriome

^r\

^ o

^ „ o o

o O - o

_ o

Qa

numerous, densely clustered, chondrioconts. Then the cell under-
goes a nuclear fusion and when that is completed, the two chondrio-
somal masses fuse around the single nucleus. After nuclear fusion,
the ascus enlarges and grows longer progressively to form an
elongated voluminous cell. During this process, the nucleus main-
tains a somewhat central position while the chondrioconts, which
have been grouped about the nucleus, spread out through the entire
cytoplasm which in this phase is filled with small vacuoles. At the
same time, at one or several points on their long axis, the chondrio-
conts usually form small swellings, each occupied by a vesicle (Fig.
28).

When the growth of the ascus is complete, but a little before
the first mitosis occurs, these vesicles disappear and the chon-
drioconts, all oriented in the direction of the longitudinal axis of
the cell, manifest a tendency to elongate. At this stage a cytoplasm
(sporoplasm), dense and rich in chondrioconts, differentiates in
the central region of the ascus while
the basal and apical portions des-
tined to make up the epiplasm, re-
main filled with small vacuoles. The
nucleus which occupies the center of
the sporoplasm then undergoes three
successive mitoses during which, it
is seen, the chondrioconts remain
scattered in the sporoplasm except
in the region occupied by the asters
where they are completely absent.
The divisions completed, each of the
resulting eight nuclei remains con-
nected with its aster by a small pro-
tuberance at the end of which the
centrosome still persists. It is not
long before the astral fibres them-
selves recurve so that, in section, each nucleus then appears as if
surmounted by a parasol on whose surface there will first form the
limiting membrane of the future ascospore. Now the whole region
of the future spore occupied by the centrosome and aster shows no
chondriosome at all. All these elements are found exclusively at
the opposite pole, where they form a compact mass of entangled
chondrioconts. The ascospores then enlarge and become surrounded
by a cellulose wall. Not until then do the centrosome and aster
disappear. At the same time the nucleus becomes centrally placed
in the ascospore and the chondriosomes are distributed throughout
the cytoplasm. Only a few chondriosomes remain in the epiplasm.

In the ascus of one of the Hemiascomycetes, Ascoidea rubescens,
Varitchak has described in more recent work, a vesiculation of
chondriosomes analogous to that recorded for the higher Asco-
mycetes.

It has also been possible to follow the life history of the chon-
driosomes during the development of some of the Agaricaceae, in

a

Fig. 26. — Fragments of protoplasm
of Achlya. A, at Ihe extremity of a
growing filament; B, an older filament.
c, chondriosomes. Gl, lipide granules
blackened by osmic acid. n, nucleus.
Meves' method, stained with acid fuchsin.

Guilliermond - Atkinson

62 —

Cytoplasm

particular that of Agaricus campestris (Beauverie, Guilliermond,
Sarazin) and in Coprinus fimetarius (Miss DUCHAUSSOY). All
the hyphae which compose the plectenchyma of the foot and of the
cap of the sporophore have a chondriome almost exclusively made
up of long chondrioconts. In young basidia, before and after nu-
clear fusion, there is also found a large number of chondrioconts
lying parallel to the long axis of the basidium. As in the case of
the asci, the chondrioconts frequently show small vesiculate swell-
ings on their long axes. When the two nuclear divisions of the
basidium are completed, the chondrioconts fragment into short
rods and migrate with the cytoplasm and nuclei to the basidiospores,
each of which encloses a chondriome formed of numerous rods. In
the course of germination, these rods move into the germinating
tube where they multiply and elongate and become long chondrio-
conts in the primary mycelium.

Fig. 27 (left). — Development of the chondriome in Leptomitus. 1, chondrioconts in the
vegetative filament: 2, 3. fragmentation of the chondrioconts and grouping about the nuclei
during the formation of the zoospores; 4, granular chondriosomes in the zoospores; 5-7, ger-
mination of the zoospores, elongation of chondriosomes into chondrioconts. C. chondriocont.
N, nucleus. Meves' method, stained with acid fuchsin.

Fig. 28 (right). — Development of the chondriome in the ascus of Pustularia vesiculosa.
1. very young ascus after nuclear division; 2-4, vesiculation of the chondriosomes during
growth of the ascus. 5. first mitosis; 6, fragment showing clear zone about each nucleus
corresponding to the aster; 7, formation of the ascospores; 8, young ascospores; 9, older
ascospores. C, centrosome; N, nucleus. Meves' method.

We have been able to observe the chondriome in some living
fungi: Endomyces Magnusii, Saccharomy codes Ludwigii but the
Saprolegniaceae are particularly favorable for this type of study,
as we have shown in our research. Meyer as early as 1904 ob-
served living chondriosomes in Achlya and described them as leuco-
plasts. We have since been able to follow with very great clearness
the entire development of the chondriome from the germination of
the zoospores to the formation of the zoosporangia in several living
fungi of this group: Saprolegnia, Achlya, Leptomitus. The sexual
cycle has not been studied as sexual reproduction is not easy to
obtain in culture (Guilliermond).

This study has confirmed results obtained with mitochondrial
technique and has thus very emphatically shown that the chondrio-

Chapter VII

63 —

The Ghondriome

somes are permanent elements which are found in all parts of the
fungi and which are never seen to arise de novo, even in prolonged
examination of living material. In addition, the presence of nu-
merous division figures of chondriosomes and the regular distribu-
tion of these elements among the zoospores seem to indicate clearly
that the chondriosomes are transmitted by division from cell to
cell and never arise de novo. Proof of this, however, is difficult to
furnish, but we shall see later that certain indirect arguments
drawn from the study of plastids in chlorophyll-bearing plants seem
favorable to the opinion that chondriosomes maintain their indi-
viduality in the course of development.

Research on the development of chondriosomes, carried out by
direct observation as well as by mitochondrial techniques, has not
furnished any information as to the role of these elements and has
not brought any concrete facts to bear on their participation in the
secretory phenomena so of-
ten attributed to them in
animal cytology.

It is true, as has been
seen, that in the asci of
Pustularia vesiculosa and of
Ascoidea rubescens and in
the basidia of the Agarica-
ceae the chondrioconts form
on their long axes vesicular
swellings which have the
same appearance as those
brought about by the forma-
tion of starch grains in the
leucoplasts of the phanero-
gams. This has also been
observed by Lewitsky dur-
ing the formation of the
oogonium in the Peronosporaceae and more recently by PoissoN
and Mangenot in Vampyrella Closterii during the period of diges-
tion. It was first supposed, by analogy with the ideas accepted in
animal cytology, that these vesicles, appearing when the reserve
products are elaborated in the basidium and ascus and then dis-
appearing during mitosis, might bear some relation to the forma-
tion of these reserve products. But it has been shown by observa-
tion of living material that none of these reserve products of the
ascus, metachromatin, fats, glycogen, arises in these vesicles. Meta-
chromatin is formed in the vacuoles (P. A. Dangeard, Guillier-
MOND). Lipide granules always arise in the cytoplasm apart from
the chondriosomes, as our observations have allowed us to show in
fungi, such as the Saprolegniaceae and Endoynyces Magnusii (Figs.
30, 31), where the chondriosomes are very clearly seen in living
form. Glycogen also appears in the cytoplasm, as we have been able
to establish by direct observation of various fungi treated with
iodine-potassium iodide reagent. This reserve product is found

! 1^-

Fig. 29. — Portions of filaments of Saprolcgnia
observed with the ultramicroscope showing (1) only
the lipide granules (Gl) illuminated, (2) the chon-
driosomes (C) also and (3) the coagulated proto-
plasm.

Guilliermond - Atkinson

64

Cytoplasm

especially at the borders of the vacuoles and about the nuclei, where
it forms as small islands which soon run together in large masses.
This elaboration of glycogen takes place without any direct partici-
pation of the chondriome, a fact confirmed by DuCHAUSSOY and Sa-
RAZIN. The significance of these vesicles, therefore, is completely
unknown. It will be seen that similar vesicles form where there is
the slightest alteration in the cell. It might be asked, therefore,
whether they are not attributable to fixatives. Yet these vesicles
always appear at the same stage of development, whatever the fixa-
tive employed, and subsequently disappear. It is therefore difficult
to accept this opinion, unless it be supposed that the chondriome
offers a much greater fragility in that stage of development of the
fungi at which they appear. Besides, we shall see that in living
Saprolegniaceae similar vesicles may appear and disappear during
the course of observation. There is the possibility that these ves-
icles indicate an elaboration of a product within the chondriosome
which reagents do not reveal. The question remains unanswered

for the time being.

C.'i'^i

oo

Fig. 30.

Gl

/ o
oo

€

6 0. o

'cP®

^%%i

Gl

It has been demonstrated,
moreover, by our work that
red pigments (carotinoids),
found in the paraphyses of
some Ascomycetes and in cells
of some yeasts (Sporobolomy-
ces), are not formed in the
chondriosomes, but are al-
ways scattered in small lipide
granules having no genet-
ic relationship with the chon-
driosomes^. The pigments of
the Myxomycetes are not con-
nected with the chondriosomes either. They are phenol compounds
which exist in the cytoplasm as sphaerocrystals (Mangenot). One
sees, therefore, that research on fungi, both by observation of
living material and with mitochondrial methods, does not confirm
the hypothesis formulated for animal cells, namely, that the chon-
driosomes participate in the secretory phenomena of the cells. It
is seen that the role of the chondriosomes in plant cells still escapes
us.

Portions of a living filament of
Saprolegnia showing successive stages of vesicula-
tion of the chondriosomes. C, chondriosomes. Gl,
lipide granules. N, nucleus.

Physical and histochemical characteristics of chondriosomes :-

It is possible, while observing living chondriosomes in the Sapro-
legniaceae to specify at the same time their physical and histochem-
ical characteristics. In these fungi, as has been seen, the chondrio-
somes appear as mitochondria only at the terminal portions of the
hyphae which will form the zoosporangia and zoospores. Every-
where else they appear as long, undulating, sometimes branched

'Other pigments of the paraphyses of Ascomycetes are, on the contrary, localized in the
vacuoles, as is the case in Galactinia succosa. Let us add that some fungi also enclose in their
Vacuoles a yellow pigment which is a flavin, as is the case in Eremothecium Ashbyii.

Chapter VII —65— The GhondrJome

chondrioconts. In whatever form they take, the chondriosomes
appear as elements of very little refractivity, being only slightly
more refractive than the cytoplasm. They are, however, always
visible but are more or less clearly singled out depending upon the
viscosity and density of the cytoplasm. Their visibility is sufficient
for a satisfactory motion picture to have been made of them (GuiL-
LiERMOND, Obaton and Gautheret).

The chondriosomes are slowly moved about by the c5d:oplasmic
currents. Their very irregular and ordinarily extremely slow dis-
placement may accelerate or stop brusquely and then begin again.
In the course of their movements, the mitochondria generally keep
their shapes but the chondrioconts modify theirs constantly. We
have been able to follow the same chondriocont during a half hour
(Fig. 70) and to draw all the variations in form which it under-
goes. When the chondriosomes are not moving, they are usually
rectilinear. During their movements, however, they may assume
the most diversely sinuous forms, appearing as S, Z, propeller-
shaped, etc. Their movements are reminiscent of Spirochaetes.
Often the chondrioconts meet, become entangled, then separate, but
we have never observed anastomosis, although they frequently
branch. Their ramifications which are, moreover, transitory, are
brought about by changes in shape necessitated by obstacles en-
countered by the chondrioconts along their path. Thus when a
chondriocont suddenly encounters a lipide granule, the chondrio-
cont branches and goes around it. The same effect can be produced
by currents moving in a direction opposite to that of the chondrio-
cont. This transitory branching, comparable to a sort of pseudo-
podium, shows that chondrioconts are made up of a semi-fluid and
very plastic substance. They are capable of becoming shorter by
thickening, or of increasing in length by growing thinner. They
sometimes appear spindle-shaped or even show on their long axes
small, sometimes vesiculate, swellings which afterwards disappear.
These variations in form have been compared to amoeboid move-
ments and there has been an attempt to explain them as due to
modifications of surface tension (Milovidov).

Under the ultramicroscope, the chondriosomes are usually in-
visible or just barely visible but in a few cases they may be dis-
tinguished very clearly by their slightly luminous contours, espe-
cially in very large hyphae where the cytoplasm forms only a thin
layer about the vacuole. The chondriosomes are optically empty
like the cytoplasm and seem to behave as a hydrogel. It may be
supposed that they constitute a coacervate system in the cytoplasm,
since they are not miscible with it. They can also be made to
appear with a color different from that of the cytoplasm by the
Zeiss micropolychromar. (Fig. 29).

The chondriosomes of the Saprolegniaceae behave as extremely
fragile elements and, as Faure-Fremiet observed for animal cells,
the least disturbance in the osmotic equilibrium of the cell, such
as a too hypotonic medium or the slightest pressure on the cover
glass of the preparation, suffices to alter them. Alteration, i.e.,

GuilHermond - Atkinson — 66 — Cytoplasm

cavulation, identical to that observed in animal cells, consists first
in a swelling and increase in refractivity of the chondriosomes and
then in a transformation into vesicles formed by a watery liquid
within a thin, rather refractive layer which is sometimes thicker on
one side. Each mitochondrium and each short rod becomes trans-
formed into a vesicle, whereas chondrioconts of a certain length
form several vesicles on their long axes^ which are ultimately

separated. These vesicles swell
greatly and press against one an-
other appearing like an aveolar
structure of the cytoplasm. Some-
times they burst because of the
pressure of the liquid which they
enclose. Hypertonic solutions do
not modify the form of the chon-
driosomes as long as the plas-
molyzed cell is alive, but they
■■^^ immediately become vesiculate
when death occurs.

For a long time chondriosomes
were considered to be sensitive to
high temperatures. Investigations

of POLICARD, COWDRY, POLICARD

and Mangenot, carried out on
V^ animal cells as well as on plant
o^U cells and notably on the Sapro-

° o o legniaceae, led to the statement
4 that a temperature of 45-50°C.

sufficed to destroy the chon-
driosomes almost instantaneously.
More recent work by Famin has
shown that this opinion is erron-
eous. In the Saprolegniaceae the
chondriosomes merely become less
visible at a temperature of 45-50°

Fig. 31. — Endomyces Magnusii stained C, beCaUSe of a modification of

Tn^' ^^^'TJ^^:^^'^:^ Viscosity of the cytoplasm, and at
chondriosomes (VM) : 4, stages in vesicu- the Same time they undergo alter-

lt°e: GJ:'^ipMT'Snu.:s: I ''ZS^. ^tiou I fragmentation into balls,

then transformation into vesicles.
A study of them after fixation shows that their chromaticity has
been much diminished but that they persist until the temperature
attains the point which produces coagulation of the cytoplasm,
i.e., about 58°C.

The chondriosomes of fungi do not stain with neutral red,
cresyl blue, toluidine blue, Nile blue and methylene blue. These
dyes accumulate exclusively in the vacuoles. On the contrary.

'This vesiculation (cavulation) seems clearly to indicate that the chondriosomes are co-
acervates.

Chapter VII

67 —

The Chondriome

they stain selectively with Janus green, methyl violet 5B and
Dahlia violet. Recent research (Guilliermond and Gautheret)
has made known other vital dyes for staining chondriosomes :
gentian violet, crystal violet, Hoffman violet, methyl green, iodine
green, malachite green and Victoria blue. Among these, Janus
green is one of the least toxic. Used with Saprolegnia in weak
concentrations (0.0005-0.005% solutions) it stains only the chon-
driosomes, giving them a bluish green color in filaments which
continue to show strong cytoplasmic currents. Staining is there-
fore clearly vital. This vital staining of the chondriosomes by
Janus green is only transitory. It is observed that at the end
of a few moments the chondriosomes lose their green color,
whereas a rose tint appears in the vacuolar system. These ob-
servations are explained by the fact that the chondriosomes re-
duce Janus green to its rose derivative and this latter, having
more affinity than its oxidized form ^
for the vacuolar system, diffuses i \^\/. •
into it. If higher concentrations of luj^/"^^) \ ^
Janus green are used, it stains not j('\\(^,^^W^H-•^
only the chondriosomes but also the b'^XlljKv^
vacuolar system. The chondriosomes i \\f^ /3^^
then remain colored but the filaments i[ 4\^> "pf '^i
die rapidly and do so generally with-
out reducing the dye. At concen-
trations above 0.005% the dye stains
the chondriosomes at first but rapid-
ly causes them to become vesiculate,
then accumulates in the vacuolar ^ „ „ , , ^,

,, 1 1 j_ 1 • Fig. 32. — Portions of a filament of

system as well, and later brmgS SavroUgnia, showing the similarity of

about the death of the fungus. Most J^^) uying and (B) fixed tissue, in

, the latter the hpide granules are

of the other dyes, among them not visible, c, chondriosomes; cg, hp-

methyl violet and Dahlia violet, are i^^^hoT""'^'' ^' ""*''^"'' ^^^""""^'^
more toxic and stain living chondrio-
somes only in concentrations not exceeding 0.005%. At greater
strengths they stain the cytoplasm and nucleus as well as the
chondriosomes, which become vesiculate, and then very rapidly
bring about the death of the filaments. The attempts to grow
cultures of Saprolegniaceae in media to which vital dyes have
been added, has demonstrated the great toxicity of the dyes. It
has been possible, for example, to make one culture of Saprolegnia
grow in peptone bouillon containing up to 0.003% Janus green.
Under these conditions, it first reduced Janus green to its rose
derivative, which seems less toxic, and then developed without
showing any coloration whatever in its chondriosomes. With a
concentration of 0.004% Janus green, the fungus ceased to grow.
Dahlia violet and methyl violet proved even more toxic for this
same fungus which did not develop at all in a 0.001% solution of
this stain.

The reagent iodine-potassium iodide preserves the chondrio-
somes perfectly. It makes them more yellow than the cytoplasm

lu. V

Guilliermond - Atkinson — 68 — Cytoplasm

and renders them very apparent. A 1-2% solution of osmic acid
also preserves them and does not make them brown. The chon-
driosomes of fungi, like those of animals do not, therefore, reduce
osmic acid unless followed by a treatment of pyrogallol. On the
contrary, the methods of osmic impregnation recommended for
bringing out the Golgi apparatus, blackens the chondriosomes
very strongly and usually makes them vesiculate (Guilliermond).

The comparative study of living and fixed hyphae of the
Saprolegniaceae has also enabled us to demonstrate that all the
ordinary fixatives containing acetic acid or alcohol profoundly
alter the chondriosomes. Careful observation, however, shows
that they continue to exist in the cytoplasm, sometimes in a very
contracted state, sometimes vesiculate, in which case they are
more stainable than the cytoplasm. The mitochondrial fixatives,
i.e., those of Benda, Meves, Regaud, and foiTnaldehyde as well,
preserve the chondriosomes, on the contrary, as faithfully as
possible in the forms they show when alive. After the action of
these last fixatives, the chondriosomes stain clearly with iron
haematoxylin, acid fuchsin, and crystal violet. They behave,
therefore, exactly as do the chondriosomes of animal cells.

Investigations of Regaud, then of Faure-Fremiet, and of
Mayer and Schaeffer have proved that the chondriosomes of
animal cells are made up of a lipoprotein complex in which
lipides (phosphoaminolipides) predominate. Of these workers,
the last three named based their conclusions on the belief that
mitochondrial fixatives are all oxidizing agents which transform
the unsaturated fatty acids into hydroxyl acids. These are only
slightly soluble in alcohol and xylol, and are capable of being
strongly stained. The chromaticity of the chondriosomes, there-
fore, is due to the lipide substance which they contain. The work
of GiROUD has shown, on the other hand, the presence of proteins
in the chondriosomes. This author was able to obtain within
these elements all the reactions of proteins. These facts apply
as well to the chondriosomes of fungi as to those of animals. As
a matter of fact, the presence of lipides in the chondriosomes is
established in several ways: not only by their characteristics of
fixation and staining but also by their reduction of osmic acid
after treatment with pyrogallol or after prolonged infiltration
in the oven; by the fact that the chondriosomes stain by the
method of Dietrich-Smith, considered as characteristic of phos-
phoaminolipides ; by their reaction with indophenol blue, a lipide
indicator. Milovidov has also demonstrated that the chondrio-
somes of the Saprolegniaceae give all the reactions characteristic
of proteins.

It may, therefore, be concluded from their behavior, which is
quite analogous to that shown by the chondriosomes of animal
cells, that the chondriosomes of fungi are, like those in animal
cells, composed of lipoproteins much more rich in lipides than
is the cytoplasm. More recently, Milovidov has established the
fact, by a study of the Plasmodiophoraceae and Myxomycetes,

Chapter VII — 69 — The Chondriome

that these chondriosomes do not give the nuclear reaction of
Feulgen and furthermore, that their protein substance has nothing
in common with chromatin, contrary to the opinion expressed by
P. A. Dangeard who has called them chromatinosomes.

It is seen, therefore, that the study of the histochemical and
histophysical characteristics of chondriosomes in the fungi have
completed and made more accurate those carried out on animal
cells and have shown that, in both cases, the chondriosomes behave
in identical fashion.

Chapter VIII
THE CHONDRIOME (Continued)

The chondriome and its development in the phanerogams.
Relationships between chondriosomes and plastids. The facts :-

The first investigations on this subject were those of Pensa
(1910). Applying the Golgi method to various tissues of phanero-
gams, i.e., impregnating sections of living phanerogam tissue

'^i*vf?.

Fig. 33. — Development of chloroplasts in a young leaf of
the plumule of barley. 1, chondriome in the basal meristem; 2,
cells beginning to differentiate with some elements of the chondri-
ome showing thickening, and 3-5 their transformation into
chloroplasts; 7, cells in older regions showing chloroplasts. c,
young chloroplasts; gc, mature chloroplast. Regaud's method.

with silver nitrate followed by treatment with a reducing solution
of hydroquinone, he noticed that the chloroplast has the property
of reducing silver nitrate and appears strongly blackened by a
deposit of metallic silver on its substratum. Now, while studying
the chloroplasts in differentiating tissues, Pensa stated that these
elements appear first as very small bodies which present the form
characteristic of chondriosomes in animal cells. Yet this author
never reached the point of specifying the origin of these chon-
driosome-shaped elements. He did not find them in tissues lack-
ing in chlorophyll and stated that their property of reducing
silver nitrate is correlated with the presence of chlorophyll in
the substratum. Pensa, however, put forth the hypothesis that
the chloroplasts are derived from chondriosomes on which
chlorophyll accumulates, giving them the property of reducing
silver nitrate.

Chapter VIII

— 71

The Chondriome (cont'd)

Without knowledge of Pensa's work, Lewitsky, a student of
Strasburger, was working at the same time with mitochondrial
technique (method of Meves). Lewitsky (1911) showed in the
bud of Asparagus officinalis that the chloroplasts are built up from
minute elements looking like the chondriosomes of animal cells.
This investigator concluded therefore that the plastids, contrary
to the opinion of SCHIMPER, do not keep their individuality but
arise from chondriosomes which LEWITSKY considers originate, in
turn, from a differentiation of the cytoplasm.

At the same period (1911) and a little later, in cells of
plants belonging to very diverse groups (phanerogam seedlings,
nucellus, embryo sac, pollen, asci of Pustularia vesiculosa), we
were demonstrating by Regaud's method, the existence of chon-

o>®^^

m

5

a)

Fig. 34. — Various types of starch formation. 1, ■within mito-
chondria in young potato tuber; 2-4, compound grains within
chondrioconts in the meristem of a young root of castor bean;
5-7, compound grains within chondrioconts in bean root: 8, within
fusiform leucoplasts surrounding the nucleus; 9, leucoplasts show-
ing successive stages in starch formation. 8, 9, from the root
of Phajus grandifolius. Regaud's method.

driosomes quite similar in form, as well as in histochemical be-
havior, to those of animal cells. Our investigations led us to
consider, contrary to the opinion of Lewitsky, that the chondrio-
somes are permanent organelles, being transmitted by division
from cell to cell and incapable of forming de novo. We were
demonstrating besides, by a study of the plumule of barley, that
chloroplasts arise by the differentiation of some of the chondrio-
somes in cells of the meristem. Finally by a study of the potato
tuber and of roots of various seedlings, notably those of castor
bean, we were able to prove that starch never forms in the cyto-
plasm but is always the product of the activity of chondriosomes.
Our later investigations (1912-1923) as well as those of Pensa
(1912) and Lewitsky (1912), followed by many others, confirmed
these facts and if interpretations still differ, it nevertheless seems

Guilliermond - Atkinson

— 72 —

Cytoplasm

that the great majority of cytologists are at present in agreement
in recognizing that the plastids of Schimper are derived from
elements presenting the same forms as those of the chondriosomes.
The life history of the chondriosome in phanerogams will now
be studied in detail by first following the formation of chloroplasts.
As the phenomena are the same in all buds, it is sufficient to
choose a single example. The most favorable, because of the dis-
position of foliar primordia, is the bud of Elodea canadensis, first
investigated by Lewitsky. Afterward, it was the object of inten-
sive study for us and our reports were confirmed by Friedrichs.
If a longitudinal section of a bud of Elodea canadensis be ex-
amined after being fixed by Regaud's method
\ . , -2is« ^ ,^^ (fixation by a mixture of potassium bichromate

'l{"ly^<^^^'' ^^^ formaldehyde, and staining with iron
Cv^V<-==Crr haematoxylin), there may be observed in the

meristem of the stem and in the youngest foliar
primordia, a chondriome exactly like that of
many animal cells, composed of a mixture of
chondriosomes and granular mitochondria.
These elements have a diameter of about
0.5-1/x. (Fig. 36).

By following successively developed foliar
primordia, there may be seen with the great-
est accuracy, all the developmental stages of
the chondriome and it may be observed that
the chondrioconts differentiate into chloro-

plasts. The differentiation, manifested by a

N^H." / -i^ thickening of the chondrioconts, begins in
t> ^i!S^S>'=k Vx- those foliar primordia which are about 160/i

long. In those measuring about 200/i in length,
the chondrioconts form little swellings on their
long axes in which a small starch grain is
sometimes elaborated. As this grain is not
stained by iron haematoxylin, it looks like a
vesicle. Starch grains thus formed are only
transitory and soon disappear. The swellings
then gradually separate by rupture of the slen-
der portions between them. They increase in volume and, in ma-
ture cells, take on the appearance of large, rounded or ovoid, chloro-
plasts about 4-8)11 in diameter. These are distinguished from the
chondrioconts, from which they arose, by the modification which
they have undergone in their chemical qualities which gives them
a special resistance. They are preserved by all the fixatives which
destroy the chondriosomes. Henceforth these chloroplasts often
elaborate large starch grains.

During the differentiation just described, the granular mito-
chondria elongate first into rods then, in mature cells, generally
become typical chondrioconts.

In the axil of each leaf primordium there is found, as is known,
a small scale made up of a group of cells in which there is no pro-

Fio. 35. — A compari-
son (A) of the chondri-
ome of the vegetative
point of Elodea canaden-
eis with (B) that of the
liver of a mouse. Re-
eaud's method. X 3,000.

Chapter VIII

73 _ The Chondriome (cont'd)

duction of chlorophyll. In these cells, the chondriome always keeps
the characteristics which it shows in the meristem. It, therefore,
remains undifferentiated, made up of a mixture of mitochondria
and thin chondrioconts.

These phenomena may be verified in fresh material by studymg

Fig. 36. — Development of the two categories of chondriosomes in
the bud of Elodea canadensis. 1, diagram of longitudinal section of
bud; 2, mitochondria and chondrioconts in a foliar primordium at
level (A); 3, 4, chondrioconts transforming into chloroplasts in a
slightly older foliar primordium; 5, mature chloroplasts in cells of a
nearly mature leaf taken at (B). Some of the mitochondria have be-
come rods or chondrioconts; 6-11, details showing the same sequence of
events. C, level at which chloroplasts develop in the stem, also chon-
driosomes in Fig. 11. Regaud's method.

the living bud. The tip of the bud may be seen, without any al-
teration taking place within its cells, by stripping it of its oldest
leaves and putting it in water under a cover slip which is pressed
gently so as not to injure the vegetative point. It is seen that the
meristem of the stem and of the youngest foliar primordia do not
contain chlorophyll. In addition to a large nucleus, these cells
show only a confusedly granular cytoplasm in which it is possible

GuilHermond - Atkinson

— 74

Cytoplasm

to distinguish the chondriome. It is only in foliar primordia in
which the chlorophyll is beginning to appear that the chondrioconts
are visible. They are here impregnated with chlorophyll and all
the forms can be followed in sequence from these elements to the
large chloroplasts of mature cells. The other chondriosomes, how-
ever, are difficult to distinguish. The cells of the mature leaves
are, on the other hand, very transparent and very favorable for

Fig. 37. — Development of the chondriome in the castor

bean root. 1-6, meristem; 7-11, differentiating cells, plastids

forming starch; 12, differentiated parenchyma cell of the central
cylinder. Regaud's method.

the study of living cells. In them can be seen all that Regaud's
method brings out and it is possible to distinguish with great
clearness within the hyalin, homogeneous cytoplasm, large chloro-
plasts, often in the process of dividing, interspersed with chondrio-
conts whose slightly higher refractivity distinguishes them from
the cytoplasm.

It is easily possible to follow the formation of chloroplasts by
a study of the development of the chondriome in other buds, among
them barley which was the object of our first research (1911).
Here also, the chloroplasts are derived from chondrioconts which
thicken, and on their long axes form small swellings which after-

Chapter VIII

— 75 —

The Ghondriome (cont'd)

®

wards separate and then elaborate a grain of transitory starch.
It is not until this is absorbed that the swellings increase in vol-
ume and take on their characteristic appearance of large chloro-
plasts.

The living root of Elodea does not lend itself to study. On the
other hand, Regaud's method brings out in the meristem a chondri-
ome entirely similar to that of the vegetative tip. During the
differentiation of tissues all that can be obsei'ved is that a certain
number of elements of the chondriome, especially the chondrioconts,
without modifying their form or chemical quality, elaborate little
starch grains along their long axes, but this elaboration is not very
active. When the root is exposed to light, on the contrary, there
are formed in the course of cell-
ular maturation and by differen-
tiation of a part of the chondrio-
somes, chloroplasts similar to
those in the stem and leaves.

A study of the root of the
castor bean (Figs. 37, 38) is
more profitable and will serve as
an example here. The chondriome
of cells of the meristem is, here
also, composed of a mixture of
granules, rods and chondrio-
conts. In the central cylinder, a
part of these elements, especial-
ly the chondrioconts, elaborate
small starch grains directly. On
the long axis of the chondrio-
conts, there are seen to form small
vesiculate swellings occupied by
a sort of vacuole which corre-
sponds actually to a starch grain

left colorless by Regaud's tech- starch by the chondrlocont-shaped amyloplasts

nique. Soon, around this small

starch grain, others are seen to appear which give a spongy ap-
pearance to the swellings and thus a compound starch grain is
produced. This increases in size little by little while remaining
surrounded by a thin mitochondrial layer prolonged to a sort of
tail, the remaining portion of the chondriocont. In the cells of
the cortex, on the contrary, some of the elements of the chondri-
ome differentiate by thickening slightly and it is not until after
this is accomplished that they elaborate starch as first described.
Thus, some of the elements of the chondriome elaborate starch
and play the role of amyloplasts either immediately, or after thick-
ening slightly. It is easy to obtain the characteristic reaction for
starch by treating the preparation obtained by Regaud's method
with the reagent iodine-potassium iodide. The chondriocont is
stained by the haematoxylin while the starch grain becomes yellow-
ish brown, due to the action of the xylol which turns yellow the

Fig. 38. — 1,2,3A, the chondriome in the
castor bean root. 3B, in the bean root. 1.
portion of a parenchymatous cell of the
central cylinder showing the nucleus sur-
rounded by chondriosomes (C) and amylo-
plasts (A); 2, amyloplasts containing com-
pound starch (o) in a similar cell; 3, less
iiighly magnified stages in the formation of

Guilliermond - Atkinson

— 76 —

Cytoplasm

starch grains stained by the iodine. Moreover, simultaneous stain-
ing of the starch and the chondriocont may be obtained by various
more complicated processes. Milovidov, especially, has shown how-
to make such permanent preparations. These methods are much
more delicate and do not give constant results.

Starch formation takes place in the same way in the greatest
variety of tissues which are without chlorophyll: roots, tubers,
epidermis. Yet there are cases, as in the tuber of the potato, in
which the chondriome is represented only by mitochondria which
elaborate starch after having undergone a slight increase in vol-
ume. In such cases they take on the appearance of vesicles, due
to the production in their interior of a starch grain which mito-
chondrial methods do not stain. Sometimes the chondrioconts
which will later elaborate starch may acquire, before its production,
a much more marked increase in volume which makes it possible
to distinguish them very clearly from the other chondriosomes in

the mature tissue. This is
seen, for example, in the root
of Phajus grandifolius in
which by following the meri-
stem to the region of differ-
entiation, it is seen that some
of the chondrioconts take the
form of rods or spindles.
These chondrioconts are very
clearly bigger than the chon-
driosomes which continue to
exist side by side with them
but without increasing in size.
These enlarged elements cor-
respond to the amyloplasts
described by Schimper and
Meyer, through the agency of which the grains of starch arise. It
seems that the increase in volume is due to the formation in the
chondriocont of a needle-shaped protein crystal lying along the
long axis of the element, whose contours follow that of the crys-
tal. In other cases the amyloplasts assume the appearance in
mature cells of rather long rounded bodies (hairs of Tradescantia
virginiana) . It may be added that the simple or compound starch
grains instead of arising in the center of the swelling of the chon-
driosome, chondriocont, or mitochondrium, may form on its periph-
ery. The chondriosome then bears a vesicle whose wall is much
thicker on one side than on the other. The starch grain which
occupies the vesicle increases in size and ends by bursting out of
the chondriosome which is thus reduced, little by little, to a thin
cap, covering the starch grain in the region most distant from its
hilum (potato tuber, root of Phajus grandifolius). The starch
grain thus formed no longer remains surrounded by a continuous
mitochondrial layer as in the preceding case (Figs. 40, 41) .

• C C

Fig. 39. — Stages in starch formation in
potato tubers.

Chapter VIII

77

The Chondriome {cont'd)

It is difficult to check these phenomena by observation of living
material. Roots in general do not lend themselves to this type of
investigation. On the other hand, in the course of our research,
we have found exceptionally favorable examples in which the entire
process of elaboration of starch can be followed with remarkable
accuracy in living cells. In a fragment of the epidermis of the
anther of a young flower of Iris germanica examined in Ringer's
solution, a chondriome is observed with great clearness, composed
of thin, elongated, and undulating chondrioconts which sometimes
branch, interspersed with granular mitochondria and short rods. In
some cells there is no elaboration of starch; in others there may
be seen several stages in the formation of small, compound, very
refractive starch grains on the long axis of the chondrioconts.
Similar phenomena may be observed in epidermal cells of the
leaves, of the bracts, and of all very young floral parts. At later

OOG

\

4i

v7 * i f

Fig. 40 (left). — Successive stages in the formation of leucoplasta in the
root of Phajtts grandif alius . Regaud's method.

Fig. 41 (right). — Leucoplasta from the root of Phajus grandif alius showing
starch. 1, central cylinder; 2, cortical parenchyma. Regaud's method.

stages of development it is seen that the starch grains are absorbed
within the chondrioconts which persist after the disappearance of
the starch (Fig. 44) . It can also be seen in these same cells that the
chondrioconts are, at certain stages, the seat of a production of
small osmium-reducing lipide globules, clearly visible on the long
axis of these elements because of their strongly refractive power.
These granules which often completely fill the chondriocont are very
frequent in the monocotyledons. They can not be considered as
formed of a-^ hexylene-aldehyde (Meyer), for they present char-
acteristics of lipides and not those of aldehydes. The fact that
they are stained by Dietrich's method suggests that they are ma(ae
up of phosphoaminolipides. These granules, very numerous in
the young stages of development of leaves, bracts and floral parts,
disappear from the plastids as soon as the starch grains and pig-

Guilliermond - Atkinson

78

Cytoplasm

merits begin to form, except, however, in certain regions where they
persist during the entire hfe of the cells. Do they represent an
intermediate product from which starch and pigments are built
up, or do they result from a breaking down of the lipoprotein com-
plex which makes up the plastids (lipophanerosis) ? It is difficult
to say. In any case, these granules reappear in large numbers
in the plastids at the moment when the flower begins to form.
They are in this case products of disintegration of the plastids and
mark the beginning of plastidial degeneration.

The epidermal cells of perianth parts of the tulip are also par-
ticularly favorable objects for observation of the living chondri-
ome and in them it is possible to follow the formation within the
chondrioconts of a yellow pigment, xanthophyll. In the white tulip,
for example, the chondriome can be observed very clearly in a frag-

A
P

oOOq
oqOO

Fig. 42 (left) . — Epidermal cells of living young anther of Iris germanica,
showing refracting mitochondria, chondrioconts and strongly refracting lipide
granules. The two lower cells contain chondrioconts with small compound
starch grains (A) on their long axes.

Fig. 43 (right). — Epidermal cells of leaves of Iris germanica. A comparison
of the chondriome in (6) a living cell with (a) one fixed by Regaud's method,
showing that the chondrioconts in (a) correspond to the plastids in (b) ; that
the rod-shaped and granular chondriosomes are similar in the two cells; that
the lipide granules appear only in (b) . c, d, e, similar portions of the cell
showing (c, e,) starch bearing plastids in living and fixed material (Regaud's
method) respectively: d, plastids not forming starch; /, g, successive stages in
the vesiculation of living plastids, C, chondriosomes; P, chondriocont, plastid;
Gl, lipide granules; A, starch.

ment of the epidermis of the perianth. It is made up of a con-
siderable number of very elongated chondrioconts and granular
mitochondria. The bases of the perianth parts are almost always
yellow and, on examining the epidermis in this region, it is seen
that it is the chondrioconts which serve as substratum for the
xanthophyll pigment and consequently represent the chromoplasts.
The mitochondria, on the contrary, remain colorless. In yellow
flowers, however, the chondrioconts in all parts of the epidermis
appear yellow because of the xanthophyll.

In living epidermal cells of the perianth and those of the exo-
carp of fruit of monocotyledons, the chondriosomes can be observed
with greatest ease and the formation of carotinoid pigments fol-

Chapter VIII

79

The Ghondriome (cont'd)

lowed. A study of the latter can hardly be made from preparations
where the mitochondrial technique has been used, for, although
the plastids are stained, there is no indication as to what pigments
they contain. So in the flower of Clivia nobilis, it can be seen that
the orange-red pigment, carotin, arises directly from chondrio-
conts. Small starch grains are first elaborated and, at the moment
when these are absorbed, the carotinoid pigment arises in the in-
terior of the chondrioconts as small grains or more especially as
long needle-shaped crystals. In the flower of Sternbergia the
chondrioconts form several acicular crj^stals of carotin which give
them the appearance of thick spindles. In other cases, the chondrio-
somes in which the pigment will form are always chondrioconts

Fig. 44 (left). — Development of leucoplasts in living
epidermal cells in leaves of Iris pallida. 1, lipide granules
(GG) within the leucoplasts (chondrioconts) in a young leaf;
2, detail of leucoplasts; 3, later stage, leucoplasts containing
starch (A); A, absorption of starch, diminution in lipide
granules; 5, leucoplasts without starch containing lipide gran-
ules in adult leaf.

Fig. 45 (right). — • Living epidermal cells of a petal of
white tulip. At left, from a young flower; at right, from a
mature flower. C, chondriosomes; Chr, chromoplasts; GL,
lipide granules; O, fatty body.

enclosing small starch grains. When these are absorbed, small
vesiculate swellings enclosing a watery liquid are formed on the
long axis of the chondriocont. Small grains of carotin appear on
the walls of these vesicles which later become isolated by rupture
of the slender regions of the chondriocont which separate them.
They then appear as small rounded vesiculate chromoplasts (peri-
carp of the fruit of Asparagus officinalis and of Arum italicum) . In
the epidermis of the perianth of Iris germanica the phenomena are
a little more complicated. The yellow pigment, xanthophyll, first
appears in a diffuse state in the chondrioconts which contain small

Guilliermond - Atkinson

80 —

Cytoplasm

starch grains. Then, when the starch is absorbed, the chondrio-
cont thickens at the same time that large vesicles appear along the
element. These may disjoin by a rupture of the more slender por-
tions of the chondriocont which connects them, so that vesicular
chromoplasts are formed with tails of varying lengths. In other
cases the pigment begins to appear in chondrioconts which increase
their dimensions proportionally as the pigment develops, until
they have been transformed into large chromoplasts of the same
form and dimensions as chloroplasts.

Fig. 46 (left). — Transformation of chondrioconts into chromoplasts in living epidermal
cells. 1, formation of starch in very young petals; 2, starch being absorbed and replaced by
small granules and needle-shaped carotinoid pigment in older petals; 4, the same in an open
but young flower; 5, chromoplasts in an older flower. 1, 2, Clivia nobilis. 4, 5, C. cyrtanthi-
flora.

Fig. 47 (right). — Development of chromoplasts in living cells of the fruit of Asparagus
officinalis. 1, chondrioconts forming starch; 2, starch being absorbed; 3, carotin granules
forming on the borders of vesiculate swelling in the plastids; 4, fragmentation of chondrioconts
to form round vesiculate plastids containing carotin granules which tend to fuse; 5, chromo-
plasts and chondriosomes in a cell of the pericarp of a nearly ripe fruit.

Xanthophyll always seems to be diffuse in the substratum of
the plastid or else in the state of indistinct granules. Its iso-
mer, rhodoxanthin, on the contrary, appears as isolated, clearly
distinguishable granules. This is true of carotin and lycopin
if they are not in crystalline form. When crystalline, the crystals
give widely-differing shapes to the chromoplasts. These facts show
that whenever the chromoplasts do not arise by metamorphosis of
the chloroplasts as in the parenchymatous tissue, studied especially
by ScHiMPER, Meyer, and Courchet, they arise from chondrio-
conts which have first elaborated starch.

Chapter VIII

81 — The Ghondriome {cont'd)

The very beot material for the study of living cytoplasm is to
be found in the epidermal cells of flowers and various organs of
the monocotyledons, those of Iris and tulip among others, as well
as in the bulb scales of Allium Cepa, which will be taken up later,
and in the Saprolegniaceae which have just been studied. On these
forms we have been able to make the most accurate observations of
the chondriosomes that it has been possible to make up to the
present time.

We have been able to show, by a comparison of these observa-
tions with those on fixed and stained cells, that the mitochondrial
methods preserve the cytoplasm and its morphological constituents,
the chondriosomes and plastids, in a manner as faithful to the
form they present in life as it is possible to have it done. These
observations permitted us,
also, to specify the histo-
chemical and histophysical
characters of the elements.
This will be taken up later.

We have studied the
chondriome very accurate-
ly during the formation of
the embryo sac and of pol-
len grains in the Liliaceae
and, in particular, in Lili-
um candidum. In the young
ovary, all the cells of the
nucellus present a chondri-
ome made up of a mixture
of chondrioconts and of
mitochondria. The embryo
sac, which arises from a
cell of the nucellus, first
shows a chondriome simi-
lar to that of other cells of
the tissue, then, in the course of its differentiation, at the mo-
ment when synizesis begins, it is observed that a part of the
chondrioconts thicken and form small swellings on their long axes.
These grow little by little, often detach themselves from the chon-
driocont in which they rise by rupture of the thin portions which
connect them, then enlarge greatly, and take on a crystalline ap-
pearance. This seems to be due to the production in their interior of
protein crystalloids. These plastids, which we have called proteo-
plasts, because of their ability to elaborate protein, then appear to
be digested in the cytoplasm and their protein is thus utilized as a
reserve product. In the synergids, the egg and the antipodal cells,
on the contrary, no elaboration of protein is noted and the chondri-
ome remains about as it is in the embryo sac at the beginning of
differentiation, except for the appearance of small granules slightly
larger than the other elements of the chondriome. These granules
are the leucoplasts. In the tulip no proteoplasts are observed and

Fig. 48. — Development of chromoplasta in cells of
the petals of Iris gervianica, 1, very young stage
showing colorless chondriocont-shaped plastids and
granular chondriosomes; 2, the starch-bearing plas-
tids begin to fill with xanthophyll; 3, the starch is
absorbed as the chromoplasts increase in size; 4,
mature cell with variously shaped chromoplasts, most
of them showing vesiculate swellings.

Guilliermond - Atkinson — 82

Cytoplasm

the chondriome, consisting of a mixture of chondrioconts and mito-
chondria, undergoes no modifications during the development of
the embryo sac^ (Figs. 49, 50, 51, 52).

In the sporogenous cells of the pollen grains of Lilium can-
didum, the chondriome is seen clearly as short rods and granules.
In the pollen mother cells, only mitochondria are to be found.
Beginning with the period of synizesis, some of these mitochondria
which are to become amyloplasts, undergo a slight increase in size,
then, at the time of the heterotypic mitosis, they elongate into
chondrioconts and afterwards, in the pollen grains, break up into
mitochondria. The remainder of the mitochondria are unchanged
from the beginning. When the pollen grain is mature, only gran-

iff «x i^ < -kJ^iw

\J'\ ^ -^ v>. »> ,0<-A\

V:^

n

^^m

Fig. 49 (left). — Embryo sac of LUium candiduvi at the
beginning of differentiation. X 1500. Regaud's method.

Fig. 50 (right). — Formation of proteoplasts in the embryo
sac of Lilium candidum at the end of the second mitosis. Regaud'a
method.

ular mitochondria are found, among which a few larger than the
others elaborate compound starch grains.

Other investigators of the development of the chondriome dur-
ing the formation of pollen in other plants have produced data
more or less analogous (WAGNER, Mascre, Krjatchenko-Douze,
Prosina, Mrs. Luxemburg, Krupko, Miss Py). Recently there
has appeared Lewis Anderson's very good work on the develop-
ment of pollen in Hyacinthus orientalis. Investigations of NlCO-
LOSi-RoNCATi, Wagner, and others have shown that the chondrio-
somes in the spore mother cells of some species may collect in a
compact mass which surrounds the spindle as a sort of mantle dur-
ing the heterotypic division and divides (chondriocinesis) at the
same time as the nucleus. The significance of this grouping of
chondriosomes is not clear and one wonders if it does not corre-
spond to an alteration.

Chon(iriosomes have been observed in all the cells of the em-
bryo before the maturation of the seed and in the seed in the dor-

"LEWI3 Anderson finds this is also the case for the embryo sac of the hyacinth.

Chapter VIII

— 83 —

The Ghondriome (cont'd)

mant state (Guilliermond, Wagner). Some of the chondrio-
somes later, at germination, form the amyloplasts of the root and
the chloroplasts of the chlorophyll-bearing organs (leaves, etc.).
It has been proved by our research that chondriosomes exist per-
manently in all phanerogam cells and that they are transmitted by
division from one cell to the next.

Fig. 51. — Portion of the embryo sac of Lilium eandi-
dum. 1, 2, stages in the development of the plastids
(P) and the mitochondria (M) ; 3, digestion in the
plastids.

The origin of plastids, which for so long remained obscure in
the phanerogams, is now well known, through the use of mito-
chondrial techniques, by means of which a chondriome has been
demonstrated in embryonic cells analogous to that in animal cells.
The entire life history of this chondriome has been followed and
it has been shown that events take place as if the plastids arose
by differentiation of some of the elements of the chondriome.

Fig. 52 (left). — Development of the chondriome during the formation of pollen grains
in Lilium canadense. 1, sporogenous cells; 2, spore mother cell in synizesis, leucoplasts appear
slightly larger than other chondriosomes; 3, metaphase, leucoplasts have become chondrioconts;
4, anaphase; 5, pollen grain, leucoplasts slightly larger than other chondriosomes, various stages
in development of compound starch grains.

Fig. 53 (right). — Chondriome in pollen of Helleborus foetidus. 1, synizesis; 2, accumula-
tion of chondriosomes about the nuclear figures of the first division; 3, pollen grain; 4,
generative cell. (After Miss Py).

There is still one gap in our knowledge. This is the behavior
of chondriosomes during fertilization. It is still not known whether
the chondriosomes of male origin participate in this phenomenon.
In a recent work, however, Lewis Anderson reports having ob-

Guilliermond - A tkinson — 84 — Cytoplasm

served that in Hyacinthus orientalis the facts are in favor of a
passage of the chondriosomes of male origin into the egg. KiYO-
HARA, almost at the same time, described in certain phanerogams
the passage of plastids from the pollen tube into the oosphere.
More recently sfill, Mangenot in gymnosperms (Pine) was able
to follow the course of the chondriosomes of male origin because
of their size which is greater than that of the chondriosomes of
the oosphere. All the chondriosomes of the pollen tube and
oosphere are in the form of mitochondria but those in the pollen
tube are larger, and can be followed during their penetration into
the oosphere during fertilization, and can be recognized after fer-
tilization has taken place. During the development of the em-
bryo, however, the chondriosomes of male origin remain in that
portion of the oosphere which does not contribute to the formation
of the embryo and which will later degenerate. Therefore there
does not seem to be any mixing of chondriosomes of male and of
female origin.

Chapter IX

THE RELATIONSHIP BETWEEN
CHONDRIOSOMES AND PLASTIDS

Interpretations:- It was logical to admit from the facts already
displayed, which have been verified by many observers, that the
plastids described by Schimper arise by differentiation of some of
the elements of the chondriome during cellular development. This
opinion, which had been maintained from the beginning by many
workers, notably Pensa (1910), Lewitsky (1912-1913), Foren-
BACHER (1912), Maximov (1913), and which we ourselves were
among the first to formulate in our early work, is still held by
Lewitsky and his school (1925-1927), Alvarado (1918-1925),
MoTTE (1928), Gatenby and his collaborators (1930), Jutta VON
Loui (1930), Chalaud (1923), Lewis Anderson and others. It
was adopted by Wilson in his book, The Cell in Development and
Heredity (1925). Still the opinion has been variously expressed.
Thus, for certain investigators such as Lewitsky and Randolph,
the chondriosomes are not permanent components of the cytoplasm
but form de novo from it. Our first interpretation was quite con-
trary to this. Not having recorded any fact which would permit
us to think that the chondriosomes can arise by differentiation
from the cytoplasm, we had from the beginning considered them
as permanent components of the cell, incapable of forming other-
wise than by division of pre-existing chondriosomes. Therefore
at that time, we considered the plastids of the phanerogams as a
variety of chondriosome, differentiated in the course of cellular
development and having a special function. The plastids, there-
fore, we believed belonged to a much more general category
present in every cell, whether plant or animal. Then too, the
theory that we had formulated was only an extension of that of
Schimper and Meyer and in no way contradicted it. It is this
same point of view which Meves and Alvarado adopted.

Although based on incontestable facts, this theory, however,
raises very serious theoretical difficulties, for it can only be applied
to higher plants. In fact, although in the phanerogams the origin
of plastids had been for a long time only imperfectly known, this
was not so for the algae. In many of these plants, as has already
been said, the chlorophyll is present in all stages of plant develop-
ment and in that case chloroplasts are observed in all cells. These
chloroplasts are transmitted by division from cell to cell, beginning
with the egg. This has been well known since the work of Schmitz.
We have seen, besides, that in many algae there is in each cell
only a single, voluminous chloroplast which divides at each cellular
division. This chloroplast, however, can not be considered as
different from the chloroplasts of the phanerogams, for it offers
the same histochemical characteristics. There are found, more-

GuilHermond - Atkinson

— 86

Cytoplasm

^M

over, in the algae, bryophytes, and pteridophytes, all the interme-
diate stages between this chloroplast. of special form and chloro-
plasts such as exist in the phanerogams. Now the research of
Randolph has demonstrated in Vaucheria, which contains chloro-
plasts similar to those in phanerogams, that these chloroplasts are
found in all parts of the thallus at the same time as the chondrio-
somes. Our work on Spirogyra showed the chondriosomes to be
constantly present and distinct from the single permanent chloro-
plast which is characteristic of this alga. The work of Sapehin,
of SCHERRER, and of MOTTIER showed that in the bryophytes, too,
chlorophyll seems to persist in all stages of development, and that
all cells, even the egg and apical cell of the vegetative shoot, con-
tain both chloroplasts and chondriosomes.
There even exists, in this group, the genus
Anthoceros in which each cell contains
only one single crescent-shaped chloro-
plast adhering to the nucleus. Coexistent
with this organelle there are, however,
numerous chondriosomes. There can not,
therefore, be found in these plants any ge-
netic relationship between the chloroplasts
and the chondriosomes. It is to be added
that in centrifuging the cells of various
algae and cells of Elodea canadensis in the
process of division, BoROViKOV was able
to obtain cells without chloroplasts but
containing chondriosomes. He was never
able to observe in these cells any trans-
formation of chondriosomes into chloro-
plasts. These contradictory facts there-
fore must be explained.

Rudolph (1911), Scherrer (1913),
Sapehin (1913), Arthur Meyer (1914-
1921) and NoACK (1921) protested, not
altogether disinterestedly, against the
new results obtained in the phanerogams
by mitochondrial technique which, at least in appearance, seem to
invalidate the classical theory. They did not hesitate to consider
the chondriosomes and plastids as inherently different formations.
To explain the phenomena in the phanerogams, Rudolph, Scherrer
and Sapehin report that in the meristematic cells of these plants,
the plastids and the chondriosomes stain in the same way with
mitochondrial technique. According to these investigators, the
plastids appear as small grains and the chondriosomes as rods or
filaments. Now, as the plastids divide actively, Rudolph, Scher-
rer, and Sapehin believe they become dumb-bell shaped and are
thereby confused with the chondriosomes (chondrioconts). But
from the moment that the cells differentiate, the plastids enlarge
and appear as large bodies which it is no longer possible to confuse
with the chondriosomes, since the latter keep their original form

Fig. 54.

Portion of a fil-

ament of Vaucheria; division
figures of the chloroplasts
(C) and of the chondrio-
somes (M). (After Mange-

NOT).

Chapter IX

87 — Ghondriosomes & Plastids

and dimensions. Scherrer and Sapehin think, furthermore, that
chondriosomes are not permanent elements of the cytoplasm but
that it is more probable that they are merely reserve products.
Meyer, who was the instigator of this opinion, attributed to the
chondriosomes a ferronuclein-like constitution and called them
Allinantes. NOACK has carried this idea further and maintains
that there is not the least morphological or histochemical resemb-
lance at any time between (Chondriosomes and plastids. He finds
chloroplasts even in the meristematic cells of buds of Elodea cana-
densis and shows them to be different histochemically from the
chondriosomes, for they are preserved by all the fixatives which
destroy chondriosomes.

Jaretzky, in his German edition of Sharp's book, Einfuhrung
in die Zytologie appears to be of the same opinion, an opinion ex-
pressed, moreover, with insufficient knowledge of the question.
Geitler takes the same stand (Grundriss der Zytologie) and also

Fig. 55.
Scherrer),

Anthoceros. C, chondriosomes; P. plastid. (After

KtJSTER. The latter, particularly in Die Pflanzenzelle, tends to con-
sider the chondriosomes as heterogeneous formations resulting
from cellular metabolism.

P. A. Dangeard, having made observations exclusively on liv-
ing material, first thought, as will be shown further on, that the
formations described as chondriosomes belonged to the initial forms
of the vacuolar system (vacuome, p. 149) and that the refractive
granules corresponded to the microsomes of early authors.
These are encountered in all cells and will be discussed later (p.
203). According to Dangeard, the plastids, therefore, bear no
relation to these dissimilar formations, the chondriosomes. How-
ever, after more profound studies with mitochondrial technique,
Dangeard was obliged little by little to renounce his former opm-
ion. He had to recognize that the chondriosomes, which he at
first had found impossible to distinguish from the leucoplasts, are
discrete elements corresponding neither to young vacuoles nor to
microsomes, and that they evidently very much resemble the leuco-
plasts^

>P. A. DANGEARD. who, however, is not convinced as to the individuality of the chondno-
aomes says in his last re-statement of the question with reference to the multiplication of the
^Tndriosomes. "If this multiplication were to take place by division or fragmentation, the
chondriosomes would be akin to the plastids but distinguishable from them .

Guilliermond - Atkinson

— 88

Cytoplasm

MOTTIER (1918) finds that plant cells containing chlorophyll
constantly enclose plastids and chondriosomes which stain in the
same way. In meristematic cells of phanerogams he finds these
two categories of elements have the same form and are very diffi-

>P

Fig. 56. — A. Development of chloroplasts in aerial root of Chloro-
phytum Sternbergianum (After Meves). 1, meristematic cell, the
chondriosomes represented by chondrioconts (P) only, the granules (Gm)
MEVE3 believes to be metabolic products; 2-5, successive stages in the
transformation of chondrioconts into chloroplasts, granules unchanged.
B. Development of amyloplasts in pea root (After Mottier). 1, meri-
stem, amyloplasts (P) shaped like chondrioconts, mixed with small,
filamentous or granular elements (C) the only elements which Mottier
believes to be chondriosomes; 2, mature root cell, amyloplasts (P) form-
ing starch, the chondriome (C) unchanged; 3, detail, amyloplasts (P)
forming starch (A), division figure of chondriosome (C) at Cd.

cult to differentiate, still the plastids would always be recognizable
because of their slightly greater size. The two classes of elements
correspond to permanent components of the cell, incapable of aris-
ing de novo and multiplying only by division. Nevertheless this
American investigator considers them as radically different forma-

Chapter IX

89 — Chondriosomes & Plastids

tions, but without, however, bringing forth the slightest histochem-
ical proof in favor of this idea. Later (1921) he seems to have
abandoned his opinion. He states that chondriosomes can not
always be distinguished from plastids in meristematic cells of
plants and seems to admit that the chondriosomes are plastids
whose functions are multiple, the plastids of chlorophyll-bearing
plants being only a special variety.

••

• •••

• • •«,

/.••%.•

• •• •

m

Fig. 57. — Chondriosomes and plastids in leaf cells of Elodea
canadensis, I, in embryonic cells, II, at the beginning of differentiation.
o, the chondriome; b, plastids and c, chondriosomes drawn separately.
Ill, later stage, o, plastids; b, chondriosomes.

. On the basis of research carried out exclusively in the phanero-
gams, Meves (1918) expressed a theory which is the exact opposite
of that held by most of the authors just discussed. This eminent
cytologist observed that, in the meristem of buds, mitochondrial
technique brings out both chondrioconts and granules, and that all
the chondrioconts are transformed into chloroplasts in mature cells.

Guilliermond - Atkinson

90

Cytoplasm

whereas the granules persist after the differentiation of the chloro-
plasts. He considers that only the chondrioconts correspond to the
chondriosomes and the granules are not mitochondria but simply
metaplasmic granules. Thus, according to Meves, all the chondrio-
somes are transformed into plastids in the mature cells of the

n

ffi

>^i7f:rft

\/

LV::).-:.'^.

• • I

Fig. 58. — Elodea canadensis (fig. 57 cont.). I, II, further
differentiation, a, plastids; b, chondriosomes changing shape.
Ill, in a differentiated cell, a, the chondriome; b, plastids and
c, chondriosomes.

phanerogams and it is those elements considered by most authors
to be plastids, which Meves believes are represented by the chon-
driosomes, whereas those described by others as chondriosomes
have nothing in common with them.

BowEN (1926-1929) at first adopted the theory of Meves and
considered the plastids as corresponding to the chondriosomes of
animal cells. In later research he was led to modify his opinion
and to admit the existence in the phanerogams of two categories of

Chapter IX

91 — Chondriosomes & Plastids

>r^

elements which are different, but whose form is similar: the plas-
tids, peculiar to chlorophyll-bearing plant cells ; and other elements,
which, hesitating for some unknown reason to liken to the chon-
driosomes of animal cells, he groups into a category which he calls
the pseudo-chondriome,

Weier (1930-1933), having obtained impregnation of the chlo-
roplasts of Polytrichum commune by Golgi technique, felt justified
in likening the plastids of plant cells to the Golgi apparatus de-
scribed in animal cells. This theory was adopted by DuBOSCQ and
Grasse who, in a recent treatise, maintain that in animal cells
there are two sorts of permanent, closely allied constituents of a
lipoprotein nature, namely, the chondriosomes and the Golgi mate-
rial, or dictyosomes, the latter being comparable to the plastids of
chlorophyll-bearing plants.

Finally, Kiyohara (1936), after mak-
ing observations of living material car-
ried out under improper conditions,
thought he noticed that all the plastids
normally appear as vesicles and that it
is the mitochondrial technique which al-
ters them and makes them appear as
chondrioconts. But this Japanese inves-
tigator obtained plastids of vesicular
form by osmic impregnation, the tech-
nique used fot- the detection of the Golgi
apparatus. He thinks that chondrio-
somes do not exist in plant cells and that
all forms described under that name cor-
respond to images brought about by
alterations in the plastids. He reports,
however, that in mature cells there al-
ways exist, as well as the large vesicular
plastids, other much smaller vesicles,
but these he believes to be plastids in the
act of degenerating.

All these theories, aside from being essentially contradictory
are, unfortunately, at variance with the facts. They are the result
of hasty generalizations, founded on observations limited to certain
types of cells and carried out, most often, with defective techniques.
They give evidence of an insufficient knowledge of that which in
animal cells has been designated as chondriosomes and which have
been recognized in all the fungi.

It is now demonstrated by our research that in mature cells
the plastids without chlorophyll generally keep the shape charac-
teristic of the chondriocont. For example, it is in this form that
they appear in the epidermal cells which we have already described
(Iris, tulip, Allium Cepa) . It is therefore impossible to attribute,
as do Meyer and Sapehin, the form of chondrioconts to division
figures of plastids which have for the moment stopped dividing.
It has also been proved in the most evident fashion, by our re-

Fio. 59. — The chondriome. 1, 2,
Root of Cueurbita Pepo. 1, meri-
stem; 2, from differentiated cell of
parenchyma, some thicker chon-
drioconts (P) form starch, short
rods and mitochondria not per-
ceptibly changed; a few elongate
to thin chondrioconts. S, young
ascus of PtisttUaria veaievlosa.
4, frog's liver.

Guilliermond - Atkinson

— 92 —

Cytoplasm

search (1912-1923), as well as by that of FRIEDRICHS (1923), that
the meristem of the bud of Elodea canadensis does not contain
chloroplasts. This is contrary to the findings of NOACK, whose
error can only be explained by supposing that in his preparations
he confused tissues already differentiated and containing chloro-
phyll, with the meristem. It has been proved, besides, that among
the elements which constitute the chondriome of meristematic cells
in the phanerogams, it is impossible to distinguish those which will
later become plastids, from those which will remain chondrio-
somes. Both have the same shapes and histochemical character-
istics. Cytological work on animals (Levi), as we have said, has
established the fact that the chondriosomes are indeed permanent
and clearly characterized elements of the cytoplasm. This is dem-
onstrated, furthermore, by the research that we have done on the
Saprolegniaceae, in which the absence of plastids makes this study

more easy than in chlorophyll-bearing
plants, and in which we have been
able to follow the chondriome in liv-
ing material during the entire de-
velopment of these fungi. It is there-
fore not possible, for the time being,
to consider the chondriosomes as dis-
similar elements, or as products of
cellular metabolism (Allmantes).
The theory of ARTHUR Meyer, who
labels the chondriosomes Allinantes,
has never been verified, and is today
definitely invalidated.

It has not been confirmed, either,
that the plastids and the chondrio-
somes of embryonic cells of phanero-
gams can be differentiated by their
dimensions, as was thought by MoT-
tier who, without doubt, observed
only cells already in the process of
differentiation. Furthermore, our
research has shown that the chondri-
osomes which are not transformed into plastids are in no wise
exclusively shaped as granules, as Meves believed. It sometimes
happens, on the contrary, that in embryonic cells, the chondriosomes
exist as chondrioconts, whereas the plastids are represented by
mitochondria. This is so much the case, that when the chondrio-
somes appear as granules in meristematic cells, they almost always
have the appearance of typical chondrioconts after the differentia-
tion of the plastids has taken place, as in the leaves of Elodea
canadensis. Therefore they are obviously chondriosomes. (Figs.
57, 58).

There is no basis, on the other hand, for the opinion of Weier,
for if it is true that the plastids are blackened by osmic impregna-
tion because of their lipide constitution, it is also true that the

Fig. CO. — SimilaritV of the chon-

diionio in A. tho basidium of Apuricus

camiHstris; B, the ascus of Pustularia

vesiculosa: C, frog's liver; D, frog's

kidney. Kcgnud's method (C, D, drawn

from preparations of Professor Pou-

CAKD).

Chapter IX — 93 — Chondrlosomes & Plastids

chondriosomes coexistent with them behave in the same manner.
In the fungi, especially in the Saprolegniaceae, where there are no
plastids, it is easy to demonstrate that the chondriosomes are in-
tensely blackened by osmic impregnations, just as are the plastids
of chlorophyll-bearing plants. Furthermore, the Golgi material in
animal cells may be distinguished from the plastids by the fact
that it does not stain with mitochondrial techniques.

The opinion of Kiyohara results from an initial error of ob-
serving living material under defective conditions. This author
leaves out of consideration that characteristic property of chondrio-
somes of becoming vesiculate during alteration (cavulation). He
observed cells in which the chondrioconts were already transformed
into vesicles and misinterpreted this phenomenon, mistaking the
vesicles for normal shapes of the plastids and the chondrioconts for
their altered shapes.

Chapter X

DUALITY OF THE CHONDRIOME

The first data obtained in our laboratory by Emberger and
Mangenot on the pteridophytes and the algae as well as the
very meticulous study of the development of the chondriome in
certain phanerogams, notably in the bud of Elodea canadensis and
in the root of Cucurbita Pepo, led us, as early as 1920, to formu-
late a new theory which removes all the difficulties and accords
with all the facts drawn from the development of plastids in the
plant kingdom.

It has been seen that among the elements which constitute the
chondriome in meristematic cells of the bud of Elodea canadensis,
it is possible to distinguish between the chondrioconts which be-
come chloroplasts during cellular differentiation, and the granular
mitochondria which do not participate in this phenomenon but
elongate into rods and later into chondrioconts. Now it has been
known for a long time that chloroplasts have the ability to divide.
Our work has shown that it is only by this process that they in-
crease in numbers and that in differentiated cells the chondrio-
somes which persist along with the plastids are incapable of becom-
ing chloroplasts. This shows therefore that the two categories
of elements which constitute the chondriome of cells of the meri-
stem develop separately and seem independent, one of the other.

The study of the development in the chondriome in the root of
Cucurbita (Figs. 59, 61) will furnish a similar example. Here again,
there is observed in the cells of the meristem a chondriome com-
posed of two categories of elements : chondrioconts and mitochon-
dria or short rods. If, in the course of cellular differentiation, the
development of these elements is followed, it is seen that in the cen-
tral cylinder, where elaboration of starch is not very active, the
appearance of the chondriome does not change and one witnesses the
production of small starch grains only within the chondrioconts. In
the cortex, on the contrary, the chondrioconts undergo a consider-
able thickening and may be subdivided into rods or granules. The
other chondriosomes keep their original size but sometimes take the
form of chondrioconts which differ from the other chondrioconts
by being thin. Our first impression is that there exist at this mo-
ment in the cells, two categories of chondriosomes, the one consist-
ing of large chondriosomes, the other of little ones. The large
chondriosomes represent the amyloplasts. At certain periods they
are the seat of active elaboration of compound starch. Once having
reached maturity, the starch grain formed in the interior of the
larger chondriosomes considerably modifies the appearance of the
latter. The starch is seen as a large grain surrounded by a thin
mitochondrial layer. This layer is often prolonged as a tail, the

Chapter X

95— Duality of the Chondriome

remnant of the elaborating chondriocont. When the starch is util-
ized, its absorption takes place within the amyloplast. The grain
diminishes in volume while the outer mitochondrial layer grows
and regenerates a chondriocont which will function again later.

j»r

• / . o:

ocrV

Fig. 61. — Chondriome. A-D, Root of Cucurbita Pepo.
A, meristem; a, amyloplasts; a', inactive mitochondria. B,
cortical parenchyma; a, starch; b. amyloplasts: b'. inactive
mitochondria, in some cases (b") appearing like typical
chondrioconts. C. parenchyma of central cylinder; o.
starch; c, amyloplasts; c', inactive mitochondria. D, corti-
cal parenchyma of hypocotyl; a, compound starch grain in
chloroplast, d; d', inactive mitochondria. E, liver of frog.
F, basidium of Psalliota campestris. In E and F, the mito-
chondria form vesicles of unknown significance. Regaud's
method.

The chondrioconts, therefore, are not destroyed during elaboration
and it is always the same elements which function in the formation
of starch. Here again there are found the two categories of ele-
ments observed in Elodea canadensis and their shape generally
makes it possible to follow them separately during their entire
development. (Figs. 57, 58).

Guilliermond - Atkinson

— 96 —

Cytoplasm

But these are rare and almost diagrammatic examples. In most
cases, it is absolutely impossible among the elements which consti-
tute the chondriome in cells of the meristem to distinguish those
which will become plastids from those which will remain inactive,

>5 n> •'

I f '^ v.. A

B
Fig. 62 (left). — Chondriome, showing morphological similarity of
the meristem of pea root (A) and of pancreas cells of gruinea pig
(B), each fixed by Regaud's method and stained with iron haema-
toxylin. A, plastids and chondriosomes indistinguishable; B, chon-
driosomes and Claude Bernard granules. (After Cowdry).

Fig. 63 (right). — Detail of chondriosomes (A) in pea root and
(B) in mouse pancreas. X 1687. (After Cowdry).

for they are of similar form. Although in most cases it is the
chondrioconts which become plastids, there are numerous excep-
tions, and cases are found in which the granules, as well as the
chondrioconts, become plastids. It may even happen in some cases
that the granules alone form the plastids, whether the other ele-
ments present have the form of chondrioconts or whether they
also have the form of mitochondria. In the tuber of potato, for
example, only granular mitochondria are found (Fig. 39) and it is
through the agency of some of these that starch is elaborated.
Furthermore, the distinction which we have made between the two
categories of elements, i.e., chondrioconts and granules, in the bud
of Elodea canadensis and in the root of Cucurbita Pepo, is far from
being general. Whether we can differentiate between the two cate-
gories depends, in Elodea, upon the state of activity of the bud and
the period at which it was collected. There are buds in which this
distinction is much less clear, others where it can no longer be
made. The difference in size which we have noticed between the
starch-forming plastids and the chondriosomes in parenchyma cells
of the pumpkin root may itself diminish or disappear. For ex-
ample, in cells in which the starch grains have just been digested,
the chondrioconts which elaborated them, grow thinner and are in-
distinguishable from the other elements of the chondriome. Cells
in the meristem of Elodea and of the pumpkin are derived from
other embryonic cells in which the two categories are not dis-
tinguishable.

Chapter X

— 97

Duality of the Ghondriome

Theory of the author:- From the series of facts which we have
just related, however, there arises the idea that the chondriosomes
of cells of the meristem do not all have the same significance, al-
though morphologically and histochemically similar. One is led
to think that the chondriome of embryonic cells in phanerogams,
although appearing homogeneous, is composed of two categories of
chondriosomes, maintaining their individuality throughout cellular
development. One of these categories corresponds to the plastids
and may take on much larger dimensions during the course of
development, by virtue of its active elaborative power. The other
of these categories, which kept its original size after the plastids
had become differentiated and to which we have provisionally given
the name inactive chondriosomes, seems to have functions which
are as yet not definitely determined.

m, ■ '-i

/C^^v,/^

'V

Fig. 64 (left). — Chondriome in (1-4) diflferentiated colorless root par-
enchyma of Athyrium Filix-femina and in (5, 6) frog's liver. Regaud's
method. (After Mangenot and Emberger).

Fig. 65 (right). — Detail of chondriome in (A) Saprolegnia and in (B)
epidermis of tulip perianth. X 3000. Regaud's method.

These two categories have the same shape in the phanerogams
and it is almost always impossible to tell them apart in the meri-
stems and usually, also, even in mature cells which do not have
chlorophyll. They are, however, always perfectly distinct in some
lower plants (bryophytes, algae) in which chlorophyll is present
in all stages of development.

Thus considered, the plastids are not differentiated chondrio-
somes; they are a special type of chondriosomes. In fact, when
the life history of the chondriosomes is followed in the phanero-
gams, one is struck by the chondriosomal characteristics which the
plastids always maintain. The amyloplasts are usually typical
chondriosomes and are only occasionally a little thicker than the
other elements of the chondriome. They are not actually distin-
guishable from the inactive chondriosomes coexistent with them,
until they become chloroplasts. In this case they appear as thick-
ened bodies which, in the last analysis, are only hypertrophied chon-
driosomes containing chlorophyll. As for the variations in plas-
tidial shape, we do not yet know whether they are caused by a
growth of these elements or merely by imbibition.

Guilliermond - Atkinson

— 98

Cytoplasm

Figure 60 gives a very exact idea of these facts. Here, in the
pumpkin seedling, are shown side by side all the forms assumed
by the two categories of elements in the course of cellular differ-
entiation. It is seen that the two types of chondriosomes have
the same shape through all the stages of cellular development —
granules, rods, filaments — but these shapes are not always identical
for both types at a given stage in development and there are stages
in which one type appears as granules and the other as filaments.
This makes it possible to distinguish them at every stage. Further-

,— >i

T-C

J.^^..^

Fig. 66. — A-D, Epidermal cells of petal of tulip, c, leucoplasts; vi, chondrio-
somes; gg, lipide granules. B, beginning of change in leucoplasts. C, vesicula-
tion. D, cells fixed by Regaud's method. E-H, filament of Saprolegnia. n, nu-
cleus; c, chondriocont. F, beginning of change in chondriosomes. G, vesiculation,
V, vesicle. H, filament fixed by Regaud's method.

more, the two categories of elements are capable of division and
frequent stages in division are observed.

If these two categories of elements are compared with the
chondrioconts in liver cells of the frog or with those in fungi, as
represented in Figures 60 and 61 {cf. also pp. 85 and 113), it is seen
in a general way that it is the plastids which most resemble the
animal chondriosomes and those of the fungi.

In a general way also, the inactive chondriosomes are a little
smaller than animal chondriosomes and are less frequently found
as chondrioconts. The plastids in general have the same dimen-
sions as the chondriosomes of animal cells but in certain phases
become much more voluminous.

Chapter X — 99 — Duality of the Ghondriome

One may object to the above theory on the ground that the
two categories of elements do not have the same origin, that
they develop separately and do not possess the same functions.
This seems to imply that they are of different nature and that they
correspond to radically different formations. This would bring us
back to the first opinion of Mottier (1918). We shall see, indeed,
that in cytology one must be suspicious of analogies in shape, since
elements as different as young vacuoles and chondrioconts may
show in certain phases of cellular development entirely analogous
forms. These forms, moreover, are the forms of chromosomes as
well. The histochemical point of view must therefore be the decid-
ing factor.

Fig. 67. — Epidermal cells of bulb scale of Allium
Cepa, as seen under the ultramicroscope. 1, lipide granules
(Gl) . 2, granules, and the faintly luminous contours of
chondriosomes (C) and plastids (P).

We shall see, however, that this objection is not valid here.
Even before it was known that the chondriome of embryonic cells
of the phanerogams was composed of two types of different ele-
ments, CowDRY (1918) made a meticulous comparison between the
morphological and histochemical characteristics of the chondrio-
somes of pancreatic cells of the guinea-pig and those of the pea
root. In the latter, the two categories are blended in the meristem
and were not differentiated by the author. CowDRY concluded that
the chondriosomes are identical in the two cases, including, of
course, those which are transformed into plastids. (Figs. 62, 63).

Similar comparisons (Fig, 64) were undertaken by Emberger
and Mangenot (1920) between the chondriosomes of fern roots
(including the amyloplasts) and those of various organs of the frog
(kidney and liver). These observations led to the same results.

Guilliermond - Atkinson — 100 —

Cytoplasm

Histochemical and histophysical characteristics of chondriosomes
and plastids:- We proceeded ourselves to make a comparative histo-
chemical study of the two chondriosomal categories in chlorophyll-
containing plants with respect to the chondriosomes of the Sapro-
legniaceae. In the latter, there exists only one category of
chondriosomes which can be unquestionably homologized with the
chondriosomes of animal cells as we have seen earlier in these pages.
This study consisted first of a comparative examination, made
as accurately as possible, of the chondriome of epidermal cells from
portions of a tulip perianth (white variety) and of the chondriome
of Saprolegnia, both of which are particularly favorable for ob-
servation of living material. (Figs. 65, 66).

The living chondriome in these two different tissues has a sim-
ilar morphological appearance and refrac-
tivity. It is represented in the tulip by long,
thin, undulating and sometimes branching,
chondrioconts which correspond to the plas-
tids, and by the inactive chondriosomes in
the form of granules or short rods. In
Saprolegnia, except for the region at the tips
of the hyphae, the chondriome, as we have
seen, is formed exclusively of long and some-
times branched chondrioconts entirely sim-
ilar to the tulip plastids. Meyer, who had
observed them before the discovery of chon-
driosomes in the fungi of the same group,
did not hesitate to liken them to plastids.
These elements arise from mitochondria by
growth and elongation, just as do the tulip
plastids. It is in the mitochondrial form
that they appear in the extremities of the
hyphae of Saprolegnia and in epidermal
cells in perianth parts of very young tulip
flowers.

This study was completed later by sim-
ilar observations which we made on epider-
mal cells from the leaves of Iris germanica
and especially on epidermal cells from the bulb scales of Allium
Cepa. In these bulb scales, the plastids and the chondriosomes
present the same forms. They are both composed of a mixture of
mitochondria, short rods and chondrioconts. As the plastids do
not elaborate starch, it is very difficult to tell the two categories of
elements apart. The plastids, however, are often recognizable
because they are slightly thicker than the chondriosomes and are
longer when they are in the chondriocont stage. We will therefore
review briefly here the principal results which we have obtained
from this comparative study and add those reported by other
authors.

The two categories of chondriosomes, plastids and genuine
chondriosomes (Fr. chondriosomes proprement dits), of epidermal

f f
{tf
( I

X B

Fig. 68. — Epidermal cell
of Iris germanica. At right,
modifications in form ob-
served when a chondriocont
moves as from (A) to (B)
in the diagram at the left
where arrows indicate direc-
tion of current.

Chapter X

— 101 — Duality of the Ghondriome

cells which we have studied and the chondriosomes of Saprolegnia
show exactly the same refractivity. Slightly superior to that of
cytoplasm, this refractivity, although very slight, still permits the
chondriosomes to be adequately seen. Under the ultramicroscope
the two categories of elements of epidermal cells and the chondrio-
somes of Saprolegnia are distinguishable only under very favorable
conditions. When visible, they always have the same appearance
and are seen only because of their very faintly luminous contours.
With the Zeiss micropolychromar they are made to appear very
clearly with a different color from that of the cytoplasm, green on
a red background, for instance, or yellow on violet. Chondriosomes
and plastids of epidermal cells, as well as the chondriosomes of
Saprolegnia, behave like extremely delicate elements which the
least change in osmotic equilibrium, or the
least pressure on the cover glass of a prep-
aration, suffices to change into vesicles
(cavulation). In a hypertonic medium they
keep changing shape as long as the cell is liv-
ing but as soon as it dies, they become
vesiculate (Fig. 67).

We have already seen that in Saprolegnia
the chondrioconts are moved about slowly
by the cytoplasmic currents and that dur-
ing these displacements they change shape,
passing through the most varied forms.
They are even able to branch by growing
a kind of pseudopodium which afterwards
is retracted. In epidermal cells of tulip in
which cytoplasmic movements are very slow
or do not exist, nothing of this sort is ob-
served. In cells of Iris germanica and of
Allium Cepa, however, which are very favor-
able objects for study, we have observed for
the plastids these same displacements and
the same instability of form. (Figs. 68, 70,
150). During the movement from place to
place the plastids are capable of taking the

most irregular shapes. They are capable of shortening by becom-
ing thicker or of elongating by stretching out. They may form
swellings along their long axes which are sometimes vesicular or
they may put out transitory ramifications which are later retracted.
Analogous observations have been made by Emberger for the leuco-
plasts in the epidermis of the bulb of Asphodelus cerasiferus. This
proves that the plastids and chondriosomes are composed of a semi-
fluid, very plastic substance, are in the same physical state and
possess the same viscosity. During these observations, moreover,
we could follow under the microscope, the process of division of
the chondriosomes in the leaf cells of Elodea canadensis, that of the
plastids in the epidermal cells of Allium Cepa and of the tulip, in
which the displacements of the plastids and the cavulation (vesicu-

FiG. 69. — Cells from a
tuber of Fiearia ranuncu-
loides (A) before and (B)
after centrifuging. A, chon-
driosomes with and without
atarch dispersed in the cyto-
plasm; B, at one side of
the cell are found the nucleus
and the starch-bearing chon-
driosomes; those without
starch remain dispersed.
(After MiLOViDOV).

Guilliermond - Atkinson — 102 — Cytoplasm

lation) which they undergo as they are altered can be observed
(Guilliermond, Obaton, Gautheret).

It may be added that MiLOViDOV and Ortiz Picon have demon-
strated that the chondriosomes and plastids have a specific weight
rather like that of cytoplasm. After centrif uging, the plastids and
chondriosomes remain scattered throughout the cytoplasm and it is
only when the plastids contain starch grains that they are carried
with the nucleus toward one extremity of the cell (Fig. 69).

Recent work of Famin has proved that, contrary to previous
opinions (PoLiCARD and Mangenot), the plastids of epidermal
cells of tulip petals and the chondriosomes of Saprolegnia are both
resistant to high temperatures. Although their visibility in living
form is diminished and their chromaticity with mitochondrial tech-
niques is lost, they are not destroyed by these high temperatures.
The leucoplasts and chondriosomes of epidermal cells of bulb

scales of Allium Cepa and tulip
;^ X petals behave as do the chondrio-

^ r, _>,=-. ~>^ somes of Saprolegnia in regard to

^ ^ ^i C^!""^ vital dyes. They stain selectively

c:^ "y^ f:^. w^ with Janus green, methyl violet 5B,

;^ ^ '=:z^ • -=^ Dahlia violet and a certain number

^ "^ ^^ V^^ of dyes recently mentioned (GuiL-

X P !^^ o'"^^ LiERMOND and Gautheret). Em-

$ ( ;^ ^ ployed in 0.0005-0.005% solutions,

•^ 1 -< 2^^ Janus green stains only the chon-

g j ^ ^ driosomes and leucoplasts, giving

^ \ "^ -"^ them a pale bluish green color in

^ r co^ -> ^gjjg which are living and showing

X^ \ 1 3 cytoplasmic currents. In 0.01-0.02%

:^ -^ solution of the dye, the chondrio-

^^ -^ somes and leucoplasts are stained

1 ' B but in a manner clearly more ac-

^ centuated in the former than in

taken by a chondriocont observed for i-ii^- "^^i^v^ ^^, ^ ■^

half an hour. B, Epidermis of Allium the dye accumulates in all the va-
SiTVn'hTr! '"''" "" ' '"'"'''"* " cuoles which contain phenol com-
pounds (oxyflavanol and tannin
compounds). At higher concentrations it produces only sublethal
staining with vesiculation of the chondriosomes and leucoplasts,
resulting after a short while in the death of the cells.

The other dyes, which for the most part are very toxic, behave
somewhat differently. In low concentrations (0.0002-0.0008%)
they stain only the leucoplasts and the chondriosomes which are
colored in exactly the same way. The cells remain living for a long
time and show very active cytoplasmic currents. In solutions of
0.001% and above, the dyes accumulate at the same time in the
vacuoles containing phenol compounds and finally stain the nucleus
and cytoplasm to which they give a diffuse color in cells showing
active cytoplasmic streaming, but in which they rapidly cause
death. Staining is therefore sublethal.

Chapter X

103 — Duality of the Chondriome

As is seen, a single difference is shown between the chondrio-
somes and leucoplasts. Janus green stains the former more in-
tensely than the latter with a 0.01-0.02% solution of the dye. In
recent work, Miss SOROKIN has maintained that Janus green did
not stain the leucoplasts. This is inexact, but it is certain that
under some conditions Janus green stains the chondriosomes more
intensely than the leucoplasts. The chondriosomes and plastids of
epidermal cells as well as the chondriosomes of Saprolegnia are
preserved with the reagent iodine-potassium iodide which makes
them brown and renders them much more distinct than in living
material. Both these elements of the chondriome are preserved
with a 2% solution of osmic acid which does not turn them brown
but, if the preparation is treated with pyrogallol after being in
contact with osmic acid for half an hour or an hour, the chondri-
osomes and plastids appear gray. Both become even intensely
black after being for a long time
in a 40% solution of osmic acid
(method of osmic impregnation
used to reveal the Golgi mate-
rial). Lastly, the chondriosomes
and plastids of epidermal cells
and the chondriosomes of Sa-
prolegnia behave in exactly the
same way in regard to fixatives.
They are strongly modified and
lose their chromaticity when
treated with fixatives containing
alcohol and acetic acid, and are
preserved in their shapes and
are stained clearly with all
mitochondrial techniques (meth-
ods of Regaud, Benda, Meves,
Helly, Tupa, Volkonsky, Al-
VARADO's modification of Rio-
Hortega, etc.). They are stained
by Dietrich - Smith's method
(used for the detection of leci-
thins) and after sufficient time by indophenol blue. This behavior,
together with that described above, proves that the chondriosomes
and plastids have a similar lipide constitution.

MiLOViDOV was able, moreover, to show in the chondriosomes
and plastids all the protein reactions, just as Giroud had demon-
strated them for the chondriosomes of animal cells. The two
categories of elements, therefore, have the same lipoprotein con-
stitution. More recently, MiLOViDOV demonstrated that the chon-
driosomes and plastids in the roots of pea and of Allium Cepa do
not give the Feulgen reaction and in consequence do not contain
thjnuonucleic acid. Chloroplasts, nevertheless, oifer much greater
resistance to fixatives containing acetic acid and alcohol than the
other plastids and the chondriosomes. Furthermore, in regard to

Fig. 71. — Chondriosomes and plastids in
tlie prothallus of Adiantum capillus-Veneris.
1, vegetative cell; P, plastid. 3, very young
egg. 4, mature egg. 2, young embryo after
division of fertilized egg. Regaud's method.
(After Emberger).

Guilliermond - Atkinson

— 104 —

Cytoplasm

the vital stains for chondriosomes (Janus green, Dahlia violet,
methyl violet, etc.) the chloroplasts do not behave like the leuco-
plasts, in that the chloroplasts are not stained as long as the cells
are alive. On the contrary, as Strugger has shown, living chloro-
plasts are stained by rhodamine B which, at a sufficient concen-
tration, gives them, with time, a very characteristic yellowish color
in cells which show cytoplasmic currents and which remain alive
for a very long time. (Rhodamine B also stains the chondriosomes
and leucoplasts but very faintly.) The chloroplasts are distin-
guished from other plastids by their property of reducing silver
nitrate. This property was noted first by Molisch and, as we have
seen earlier, is manifested only in living cells. Ruhland and
Wetzel found it possible, with this reaction, to demonstrate the
presence of chloroplasts in the chondriosomal state in generative
cells of Lupinus luteus and of some other plants. Gavaudan in his

? 3 ' ^-^ 4

Fig. 72 (left). — Chondriome of antherozoids of Adiantum capillus-Veneriii.
1, 2, antheridial initial. 3, 4, sperm mother cell. 5, 6, stages in the formatioQ
of the antherozoid. Regaud's method. (After Emberger).

Fig. 73 (right). — Behavior of the chondriome during the life cycle of a fern,
I. leaf; starch-forming chloroplasts and chondriosomes. 2-3, formation of spores:
decreasing activity of plastids. 4, spore mother cells; inactive plastids indistinguish-
able from chondriosomes. 5, mature spore; plastids become active. 6, prothallus:
starch-bearing chloroplasts. 7, 8, sexual cells; second cessation of activity of
plastids. 9, egg; homogeneous chondriome. 10, developing embryo; certain chon-
driosomes secrete starch. 11, 12, adult plant; 11, amyloplasts of the root. 12.
chloroplasts of the leaves. (After Emberger).

study of the hepatics claimed that this property is common to all
plastids, even those lacking in chlorophyll, and considers it a means
of distinguishing plastids from chondriosomes. But our later re-
search, as well as that of Gautheret and of Mirimanoff, did not
confirm this assertion and proved that plastids without chlorophyll
do not reduce silver nitrate in living cells any more than do chon-
driosomes. This property which chloroplasts have of reducing
silver nitrate in living cells has nothing in common with the black
coloration of the plastids and chondriosomes in cells treated by
Alvarado's modification of Rio-Hortega's method, or with that
sometimes taken on when they are impregnated with silver (Golgi
method). It does, however, explain why Pensa found that only
the chloroplasts were stained when he treated living tissue with
Golgi's technique (silver impregnation).

Chapter X

105

Duality o£ the Ghondriome

All these facts lead us to the conclusion that the two categories
of cytoplasmic organelles in chlorophyll-containing plants both
show the characteristics of chondriosomes and it is evident that
there is no criterion, unless it is the ability of the plastids to form
starch and chlorophyll, for including the inactive chondriosomes
rather than the plastids in the formations known in animal cells as
chondriosomes. On the contrary, the plastids by their elongated
chondriocontal forms sometimes resemble the chondriosomes of
animals even more than do the inactive chondriosomes of plants.

Fig. 74. — Fern sporangia. Successive stages in the return to a homo-
seneous chondriome by resorption of starch and loss of pigment in the chloro-
plasts. 1-3, Asplenium Ruta-muraria. 1, sporogenous cells; 2, spore mother
cells; 3, tapetum and spore mother cells. 4, Pteridium; tetrad and tapetum.
Regaud's method. (After Emberger).

The plastids are sometimes, however, slightly larger. Further-
more, the inactive chondriosomes unquestionably have the charac-
teristics of chondriosomes from which it is impossible to separate
them, as Meves does, for they, too, show in a great number of cases
the form of typical chondrioconts.

These two categories of elements, therefore, fit the definition
of chondriosomes. They correspond to organelles seeming to be
incapable of forming other than by division, they have the shape

Guilliermond - Atkinson — 106

Cytoplasm

of granules, rods or chondrioconts, are able to change from one of
these forms to the next, and are characterized by a number of well
determined physical and chemical properties. The theory of the
duality of the chondriosomes has been remarkably confirmed by a
study of the life history of the chondriome throughout the plant
kingdom and is particularly well supported by the investigations
of Mangenot and Emberger.

Development of chondriosomes and plastids among the plant
groups:- Emberger (1920-23), in investigating the pteridophytes
(Figs. 71-74), found that the egg cell in the ferns contains a chon-
driome exactly like that in the animal cells, in which chondriome it
is not possible to distinguish the plastids from other chondriosomes.
However, in those prothallial cells from which the egg cell is de-
rived there are both large chloroplasts and small chondriosomes. As
these cells differentiate in the course of the formation of the egg,
Emberger has shown that their chloroplasts lose chlorophyll and

miH'

©

[0

d

Fig. 75 (left). — Selaginella Kraussiana. 1, vegetative tip of the
stem, each cell containing chondriosomes (C) and a slightly larger
plastid (P) appressed to the nucleus. 2, Diagram of developing
plastid during cellular differentiation. (After Emberger) .

Fig. 76 (right). — Selaginella Kraussiana. Plastid (P) in living
parenchyma cell of the stem. (After Emberger).

at the same time diminish progressively in volume, with the result
that they take on, little by little, the appearance of small elements
which it becomes impossible to distinguish from the chondriosomes.
The same phenomena occur in the formation of the antherozoids.
The fertilized egg resulting from the fusion of these cells shows,
therefore, a homogeneous chondriome. In the embryo, some of the
elements of this chondriome differentiate anew. In the leaves they
become chloroplasts, in the stem and root they become large chon-
drioconts which represent the starch-forming plastids.

In the epidermal cells of leaves which will produce sporangia,
both chloroplasts and inactive chondriosomes are encountered. In
the young sporangia, however, the chloroplasts again lose their
chlorophyll and appear as typical chondriosomes, indistinguishable
from the inactive chondriosomes. From the time that the spore
begins germination, these typical chondriosomes grow larger, be-
come impregnated with chlorophyll and take on again the character
of chloroplasts.

Chapter X — 107 — Duality of the Ghondriome

Analogous phenomena were found later (Cholodny, 1923) in
the submerged leaves of Salvinia natans which, as is known, look
like roots, as they have no chlorophyll and are reduced to veins.
The chloroplasts which are present at first in these leaves lose their
chlorophyll and take on the appearance of chondriosomes abso-
lutely indistinguishable from the genuine chondriosomes.

Investigating the Selaginellas (Figs. 75, 76), Emberger ob-
served that in cells of the meristem and in the spores where the
chondriome contains above all long chondrioconts, there is only a
single colorless plastid, which also appears as a chondriocont but is
a little larger than the others and is pressed against the nucleus.
This organelle, already pointed out by Haberlandt, Sapehin and
P. A. Dangeard, which is at first scarcely distinct from the other
chondrioconts, grows little by little during cellular differentiation
until in each cell of the leaf and stem there is a single chloroplast.
It is composed of a series of large swellings united by thin filament-
ous portions which are brought about by uncompleted divisions of
the initial plastid. This fact is particularly interesting from two
points of view. First, in Selaginella and Anthoceros (in which
there is also in each cell only one chloroplast, crescent-shaped in
this case, more or less appressed to the nucleus but always larger
than the small colorless plastids of embryonic cells of Selaginella) ,
there are found the intermediate steps between the plastids of the
phanerogams and the large chloroplasts of some algae. This shows
that there is no reason to consider the chloroplasts of the algae as
different from ordinary plastids. Secondly, the presence of this
solitary plastid, which can be followed through all cells of Selagi-
nella and which divides when the cell does, furnishes undeniable
proof that the plastids maintain their individuality during the
course of cellular development and arise always from the division
of pre-existing plastids. The behavior of the plastids in the phaner-
ogams makes this seem likely but does not sufficiently demonstrate
it.

The investigations of Mangenot (1922) brought out that the
algae behave differently, from the- point of view of plastidial de-
velopment, depending on whether the chlorophyll persists in all
stages of development or disappears in the sexual organs. When
it persists, the plastids are distinguished from the chondriosomes
in all stages of development, including the egg, by their size, their
shapes and their colors. Consequently chloroplasts and chondrio-
somes are coexistent at all times.

Mangenot demonstrated the presence of large chloroplasts and
small chondriosomes at all stages in the Siphonales. These
had already been encountered in Vaucheria by Rudolph. These
two categories of organelles, in spite of the difference in their
dimensions, have analogous shapes and divide at the same time
in some phases of development. Such is also the case in the
Fucaceae, in which, however, the chlorophyll and fucoxanthin lose
their intensity in the oogonium and in the apical cells. There the
phaeoplasts take on the form of small rod, or spindle-shaped

GuilHermond - Atkinson

108

Cytoplasm

organelles, differing only slightly from the chondriosomes. Fur-
thermore, the chlorophyll and fucoxanthin disappear in the mother
cell of the antheridium and the phaeoplasts in this cell come to
look like chondrioconts and are distributed among the antherozoids
in such a way that each encloses a single phaeoplast. This plastid
later is filled with a carotinoid pigment and becomes the stigma.
The stigma, then, is simply a structure derived from a phaeoplast.
Plants in which the chlorophyll does not persist but rather
disappears in the sex organs are found in the Rhodophyceae (Flo-
rideae) and the Characeae. In the Rhodophyceae, for example in
Lemanea, the cells of the thallus (Fig. 78, I and the upper portion
of A) enclose large, ribbon-shaped rhodoplasts, sometimes anasto-

o

s^^n

80,

^

p.p.

i.t. 1, 1, 1,1 . I.I. 1, 1. 1

o to a o ^

Fig. 77. — Fucus vesiculosus. Fusiform and rod-shaped plastids (pi) with mito-
chondria im) and fucosan granules (p.p.) in Fucus vesiculosus. 1, apical cell; 2, two
celled embryo. N, nucleus. Regaud's method. (After Mangenot) .

mosing to form a network, together with small chondriosomes. In
those portions of the thallus containing little chlorophyll (Fig.
78, Ii and lower portion of A), these elements grow thinner and
appear somewhat like chondrioconts. In the rhizoids (Fig. 78, I2),
in which neither chlorophyll nor phycoerythrin exists, the plastids
become very small and look so like the inactive chondriosomes that
it becomes impossible to tell them apart. The trichogyne and other
cells of the carpogonial branch (Fig. 78, A) develop from an
ordinary cell of the thallus containing large rhodoplasts. A regres-
sion of chlorophyll and of phycoerythrin may be observed in these
cells. The plastids lose their color and are transformed into small
rods becoming like the chondriosomes which are present with them
in the cell. Then the carpogonium shows a chondriome in which
all distinction between plastids and chondriosomes is impossible.
This chondriome persists in the first cells of the gominoblast fila-

Chapter X'

109 — Duality of the Ghondriome

SJ:S:^ ^r=~d>^'r^a.:^ ggj r^7^

Fig. 78. — Rhodophyceae. Life history of the plastids in Lemanea. A, fragment
of wall of cystocarp with two carpogonial filaments. B, Id. with gominoblast fila-
ments. C, detail of filament. D-G, carpospores. H, germinating spore. I, assimilat-
ing filament. Ii, filament near substratum. I2, rhizoid. J, cystocarp. (After
Mangenot).

N

Guilliermond - Atkinson

110

Cytoplasm

..-»..

ill

P'V-v ill

merit (Fig. 78, B, C) and then there can be followed in these cells
(Fig. 78, D-F) a differentiation of large rhodoplasts from some
of the elements of the chondriome. In the mature carpospores
(Fig. 78, G, H) there are found fairly large, well differentiated,
disc-shaped rhodoplasts.

In the Characeae, Mangenot found small chloroplasts and
chondriosomes in the apical cells but in the oosphere there is no
chlorophyll. In the cells which give rise to the oosphere, Mange-
not observed a regression of the small chloroplasts. They lose
their chlorophyll and are transformed into mitochondria or short
rods which can not be distinguished in the young oospheres from
the inactive chondriosomes. In the course of development of the
oosphere, some of the mitochondria and rods representing the
former chloroplasts, elongate and take on the shape of typical

chondrioconts, whereas the inactive
chondriosomes persist in the form
of mitochondria. The chondrioconts
then elaborate numerous starch
grains in the usual way. They cor-
respond therefore to amyloplasts.

Information on the development
of the plastids is still scarce in the
bryophytes. It has already been
seen that Rudolph, Scherrer, Sape-
HiN, and MOTTIER believed that
chlorophyll persists in these plants
in all stages of development. They
state that bryophytic cells always
contain chloroplasts and chondrio-
somes at the same time. The fact
is well demonstrated for Anthoceros
but is questioned for the other bryo-
phytes. Whereas P. A. Dangeard,
P. Dangeard, Gavaudan and Weier
tend to confirm it, Alvarado, Sen-
JANINOVA, MOTTE and Chalaud op-
pose it and believe that the chloroplasts are derived from the
chondriosomes. Alvarado has stated that in young paraphyses
of Mnium cuspidatum there are no chloroplasts but only chondrio-
somes of which some afterwards become transformed into chloro-
plasts. According to Motte, most mosses contain in the apical
cell of the stem both small lenticular chloroplasts and chondrio-
somes, while in some species {Grimmia crinita) only chon-
driosomes are found, a part of which develop into chloroplasts in
cells arising by division from this apical cell. Motte described a
regression of chloroplasts in the sperm mother cells. In the forma-
tion of these cells, the chloroplasts lose their chlorophyll and be-
come transformed into long chondrioconts. These fragment to
form granules that it is impossible to distinguish from the chon-
driosomes, which are coexistent with them. According to Motte,

'I

B

Fig. 79. — Chara fragilis. Develop-
ment of the chondriome in the egg. A,
young egg. B, beginning of differentia-
tion. C, later stage, appearance of
chondrioconts. D, mature egg; chon-
drioconts forming starch. E, detail of
(D).

Chapter X — 111 — Duality of the Ghondriome

the archegonium is formed from an initial cell containing only
chondriosomes and he finds that the egg also lacks chloroplasts.
These facts make it seem extremely probable that, at some stages
in development in the bryophytes, the chloroplasts are capable of
regression and of taking on mitochondrial form just as in the
pteridophytes and in certain algae.

However this may be, it follows from data presented for the
first time in the splendid work of Emberger and Mangenot, that
the chloroplasts may, under some conditions, lose their chlorophyll,
become considerably smaller, and take on again the size and shape
of typical chondriosomes. The form typical of chondriosomes and
the form typical of chloroplasts are therefore reversible and the
chloroplasts may be considered as chondriosomes containing chlo-
rophyll. The chloroplast is derived from a chondriosome and may
under certain conditions lose its chlorophyll and revert to the
state of the chondriosome, the state which is characteristic of the
functionally inactive phase of plastids, exactly as the amyloplast
resumes its initial form after the absorption of its starch. If this
reversibility of chloroplasts is not ordinarily observed in phanero-
gams, it is doubtless because, according to the research just dis-
cussed, the chlorophyll-containing tissues in these plants achieve
a state of differentiation too advanced for a regression to take
place such as occurs in plants less evolved.

Therefore the fact that chlorophyll is elaborated in a continuous
or discontinuous manner influences very notably the appearance
taken by the chondriome in chlorophyll-containing plants. In the
first case, the cells contain constantly and at the same time, both
large chloroplasts and small chondriosomes ; in the second case, on
the contrary, there are found, during the periods when chlorophyll
is lacking, chondriosomes which all together constitute a chondri-
ome analogous to that encountered in cells of animals and fungi,
and in which it is not possible to distinguish the plastids from the
genuine chondriosomes, and it is only during phases of elabora-
tion of chlorophyll that some of the chondriomal elements grow
and become large chloroplasts.

These facts will be illustrated by a study of saprophytic or
parasitic phanerogams in which chlorophyll is formed only in very
small quantities or, in some species, not at all. Among these,
Limodorum, a saprophyte which is poor in chlorophyll, contains
only very small chloroplasts. In the genus, Orobanche, in which
chlorophyll has disappeared, plastids elaborating starch are still
observable which in some portions of the plant also contain carotin-
oid pigments. In the stem of Monotropa, in which chlorophyll is
also lacking, there is no longer production of starch, except in
the endodermis, and in that region only, can the plastids be dis-
tinguished from the chondriosomes. This distinction is impossible
in the tissues of Cytinus Hypocistis, a plant more completely
adapted to parasitic life. This plant is also lacking in chlorophyll
and has lost its power of forming starch (Emberger and Mange-
not, unpublished observations).

GuilHermond - Atkinson — 112 — Cytoplasm

Considering this from a different angle, one of our students,
Gautheret, has shown that the production of chlorophyll may be
experimentally obtained in most roots when they are grown under
certain conditions (in the presence of light and in media contain-
ing a certain quantity of sugar). Thus, for example, in the root
of the barley, whose cells normally contain a chondriome composed
of a mixture of chondrioconts, rods and mitochondria, in which
plastids and genuine chondriosomes are indistinguishable, Gau-
theret has succeeded in obtaining large chloroplasts, entirely com-
parable to those encountered in the leaves of the same plant, by a
differentiation of some of these elements.

The series of investigations undertaken with the aid of mito-
chondrial methods, either by us in the phanerogams or in our
laboratory by Emberger and Mangenot on the pteridophjrtes and
algae, have permitted us definitively to solve this problem which
has remained obscure for so long, namely, the origin and life his-
tory of chlorophyll-containing plastids. The progress made since
the work of Schimper and Meyer may be judged by comparing the
state of this question when there were no methods for preserving
plastids in stained preparation and when the investigator had to
be content with incomplete observations of living material, to the
present status of the question with its very accurate data obtained
by mitochondrial technique, completed by observation of living
material.

It is now evident that the chlorophyll-containing plants possess
two categories of organelles which are permanently found in every
cell, both of which show all the characteristics of chondriosomes
in animal cells.

The first category, whose role it has not been possible to define
completely, corresponds to the chondriosomes found in cells of
animals and fungi. Its elements may be called inactive chondrio-
somes or even genuine chondriosomes.

The second category, peculiar to chlorophyll-containing plants,
corresponds to the plastids of Schimper. These are to be dis-
tinguished under the name plastids.

These two categories of organelles seem each to keep their
individuality in the course of cellular development and seem to
form only by division of pre-existing elements. This behavior,
difficult to demonstrate for the plastids in the cells of phanero-
gams, is absolutely proved by what is known regarding the chloro-
plasts of green algae, by the study of Anthoceros, and especially,
by the observations of Emberger on Selaginella. It is extremely
probable, also, that these characteristics are shared by the genuine
chondriosomes, since these elements are present in all cells, since
they have never been observed to form de novo or to disappear
and since they are capable of division. Indirect arguments in
favor of this opinion may be drawn, moreover, from the behavior
of the plastids which are very similar to them.

The two categories of elements have (except in many algae
in which chlorophyll persists in all stages), the same characteris-

Chapter X — 113 — Duality of the Ghondriome

tic forms : granules, rods and filaments, capable of changing from
one shape to the other. They offer, moreover, and this is much
more important, the same histophysical characteristics (same re-
fractivity, same viscosity, same process of alteration) and the
same histochemical characteristics, even to their behavior with a
great number of chemical reagents and dyes. For these reasons,
it is generally impossible to tell them apart in embryonic cells of
plants in which chlorophyll is not continuously elaborated. The
plastids, therefore, are distinguished from other chondriosomes
only by the fact that they are the centers of very active elabora-
tions which, considerably modify their shape. This is true in cells
lacking in chlorophyll in which plastids elaborate starch, and espe-
cially true in green cells in which the plastids become voluminous
because of the chlorophyll which they accumulate. Moreover, these
modifications of form may be only transitory. The amyloplasts in
cells without chlorophyll, as soon as the starch has been absorbed,
go back to their forms of typical chondriosomes. The chloroplasts
themselves may in some cases lose their chlorophyll and return
to their original state as chondriosomes. In a word, the chondrio-
somal form is the form taken by these organelles during the func-
tionally inactive period. The only distinction, therefore, that exists
between the animal and fungal cell on the one hand, and the cell
of the green plant on the other, is the presence of plastids, the
second category of chondriosomal elements. This distinction is re-
lated to the existence of the chlorophyll function which charac-
terizes green plants.

These facts, now exactly demonstrated by our work and that
of our students, carried on over a period of thirty years, have led
us to fomiulate the theory of the duality of the chondriome in
chlorophyll-containing plants. This consists in stating that the
chondriosomes of chlorophyll-containing plants are composed of
two categories having between them the same relationships which
the heterochromosomes bear to the autochromosomes, the first cate-
gory (the genuine chondriosomes) being very similar to the chon-
driosomes of animals and fungi, the second category (the plastids)
composed of a supplementary line of chondriosomes related to
photosynthesis, which characterizes these plants. It is obvious
that the two categories must, after all, possess differences in chem-
ical constitution. Otherwise it could not be explained why one
has functions which the other does not have. Nevertheless these
differences, probably very slight, do not appear during histo-
chemical analysis. In any case, both categories of elements have
very closely allied lipoprotein constitutions and form in the cyto-
plasm a disperse lipoprotein phase. They are differentiated only
by one of them manifesting a function which is lacking in the
other.

The question is therefore definitely solved as far as the facts
are concerned and it is demonstrated that the chondriosomes and
plastids are two individual cellular components with the same
lipoprotein constitution, capable of presenting identical shapes but

Guilliermond - Atkinson — 114 — Cytoplasm

developing side by side without any genetic bond between them.
This is undeniable and cytologists are more and more inclined
today to admit it. There is only one point which still remains hy-
pothetical and this is the identification of plastids with chondrio-
somes which many cytologists still refuse to accept. In order to
settle the question definitely we should have to know more pre-
cisely the chemical constitution of these two categories of elements
and this is impossible in the present state of science. We should
also have to obtain information on the phylogeny of plastids and
chondriosomes which now escapes us. That which is certain is
that the chondriosomes and plastids can only be regarded as very
closely allied formations. It is illogical to deny these incontestable
relationships as so many cytologists still do, for it is a much greater
assumption to consider them as essentially different formations
than it is to put them both together under the heading of chondrio-
somes. Also, our theory, which nothing has contradicted since we
formulated it nearly twenty-five years ago, seems to be the only
interpretation possible in the present state of knowledge. It has,
furthermore, the advantage of suggesting a series of working hy-
potheses, not only with respect to the role of the genuine chondrio-
somes which is at present almost completely unknown, but also
with respect to the physiological f unctionings of the plastids in the
elaboration of chlorophyll, their functioning in photosynthesis and
in the condensation of hexoses into starch. It is possible to imagine
that a day will come when the physico-chemical study of the cyto-
plasm will show us that the chondriosomes have a very general
function of which that manifested by the plastids is only one
specific part.

Phylogenesis of chondriosomes and plastids:- Nothing is known
about the phylogenesis of these two categories, chondriosomes and
plastids, whose chemical constitutions are so closely allied. The
fungi, which many botanists consider to be derived from the algae,
show, however, no trace whatever of plastids. There is only one
line of chondriosomes in them. In the Cyanophyceae (p. 41),
which by reason of the primitive structure of their nucleus, may be
considered as the most inferior algae known, it is impossible, as
we have seen above, to detect the presence of chondriosomes and
of plastids. The chlorophyll is diffuse in the cytoplasm of these
algae. It has sometimes been thought that the lipoprotein sub-
stance of plastids and chondriosomes was also diffused in the cyto-
plasm. This, however, is only an hypothesis based on the absence
of plastids and on the fact that chlorophyll can hardlj^ have as sub-
stratum any other than a lipoprotein substance. In all flagellate
algae, another inferior group thought to be the common ancestors
of algae and protozoans, there exist very varied forms, some with
chlorophyll, some without. All contain chondriosomes (Chade-
FAUD). We have seen that in chlorophyll-containing forms the
plastids sometimes look like the chloroplasts of phanerogams
(Euglenas, Peridiniaceae) , sometimes appear as a single voluminous

Chapter X —115— Duality of the Chondriome

chloroplast (certain Chrysomonadales). According to Chadefaud,
the forms possessing only a single chloroplast represent the most
primitive types.

Among the forms without chlorophyll, VoLKONSKY has reported
in Polytoma uvella the existence of a single leucoplast per cell,
which appears as a fine network spread throughout the cytoplasm.
This leucoplast elaborates starch. More recently Miss Rabinovitch
found the same organelle in Polytomella coeca. It has been possible
to suppress the chlorophyll by various cultural processes in the
Euglenas but it has not been possible previously to understand
what became of the chloroplasts in the forms deprived of chloro-
phyll. In the phylogenetic series, above the flagellated algae are
placed the green algae, such as the Chlorophyceae, w^hich often
possess only a single voluminous chloroplast. We then progress
through the Rhodophyceae, the Characeae and bryophytes to the
pteridophytes and phanerogams, in which the plastids are always
fragmented into numerous small elements similar to chondriosomes.

Chapter XI

HYPOTHESES RELATIVE TO THE ROLE OF
CHONDRIOSOMES AND PLASTIDS

It has just been seen that cytological investigations, carried
out during recent years, have shown that the cytoplasm always
contains in suspension various inclusions, among which the most
important are the chondriosomes to which, in chlorophyll-
containing plants, are added the plastids. The latter present a close
analogy to the chondriosomes but may be considered as a line dis-
tinct from these elements.

Immediately, this raised the question as to the role of the
chondriosomes and plastids. We find ourselves here on uncertain
ground, for it must be recognized that if, at the present moment,
our morphological knowledge is very advanced, we are still ex-
tremely ill-informed as to the role which must be attributed to
these various elements in the functioning of the cell. As Devaux
says, "The fact is that we do not sufficiently know the inner organ-
ization of the cell, for all that a microscopic study makes possible
is a first approximation, manifestly incomplete. We should, for
complete knowledge of the cellular mechanism, be able to reach
the molecules themselves, that is, the elementary particles of the
cell, and study them from the triple point of view of structure, of
molecular attractions and movements, as well as from the point
of view of reciprocal relations." Perhaps some day the progress
of physical chemistry will give us precise information on this
point, but for the moment we must be content with hypotheses
which are still very vague and which we will take up here as
briefly as possible. Investigations at the beginning of the study
of chondriosomes led various authors, Regaud in particular, to
attribute to the chondriosomes of animal cells, an important role
in the phenomena of secretion. According to this opinion the
chondriosomes are organelles through whose agency are elaborated
very diverse products of cellular activity: zymogen granules, fats,
pigments, i.e., the chondriosomes have a role exactly like that of
the plastids in chlorophyll-containing plants. Thus the chondriome
appears as the secretion apparatus of the cell. Although the in-
vestigations in plant cytology have demonstrated the very curious
fact that it is precisely the plastids of chlorophyll-containing plants
which show exactly the same forms and the same histochemical re-
actions as the chondriosomes, we have seen, however, that more
accurate research on animal cells and on the thallus of fungi,
employing the control methods of direct observation of living ma-
terial and of vital staining, have not been able to confirm this
secretory role except in exceptional cases (Noel). The direct par-
ticipation of chondriosomes in phenomena of secretion seems to us
therefore to be hypothetical and in reality we know nothing as to

Chapter XI — 117 — Role of Ghondriosomes

the role of these organelles. Nevertheless, the fact that the chon-
driosomes usually show no morphological evidences which can be
related to their participation in secretory phenomena does not
exclude the possibility that they may have a role in these phenom-
ena, and it is possible that later studies may succeed in demon-
strating this role which is suggested by the close parentage of
chondriosomes and plastids. So all cytologists are at present agreed
in attributing to the chondriosomes some important role in cellular
metabolism.

The role of the plastids in chlorophyll-bearing plants is, on the
contrary, very clear, for it is manifested morphologically by the
production within these organelles of chlorophyll, carotinoid pig-
ments, starch grains and so forth. However, we do not know at
all by means of what physico-chemical processes these phenomena
are brought about. At first an essential role in the elaboration of
these different products, as well as participation in the phenomenon
of photosynthesis, was attributed to the plastids. The plastids
were considered as small laboratories which were the seat of the
most important synthesis in the plant cell.

At the present time there is a tendency, rightly or wrongly, to
react against this way of thinking and to consider the plastids as
perhaps only the accumulation centers for certain substances manu-
factured by the cytoplasm itself. In any case, the work of modern
physiologists caused it to be thought that the plastids contribute
only in part to photosynthetic phenomena. Actually, we still do
not know how much to attribute to the chlorophyll, how much to
the substratum of the plastids and how much to the cytoplasm.
Nevertheless, it seems that the plastids do indeed play an important
part in photosynthesis. In any case, there is a role which cannot
be denied the plastids. It is the ability to condense the hexoses
into starch. In this case it is a question not of accumulation but
of actual synthesis. The influence of plastids on the form and
growth of starch grains is manifested by the fact that these prop-
erties are determined by the place where the grain appears in the
plastid. If the grain arises in the middle of the plastid, the hilum,
i.e., its oldest part, is the center and concentric layers, which are
the outcome of growth by apposition, develop regularly about it.
The starch grain in this case remains entirely enveloped by a thin
mitochondrial layer. When, on the contrary, the growing grain
forms on the periphery of the plastid, it very soon bursts out of the
plastid which then covers it as a sort of cap at only one extremity.
In this case the hilum is situated at the pole opposite to that occu-
pied by the cap, and the layers of growth do not form any longer,
except in those regions which are still in immediate contact with
the plastid. The grain then is eccentric in structure.

The fact that the starch grain is formed and is hydrolyzed in
the interior of the plastid led certain investigators to believe in the
existence in the substratum of the plastid of a diastase of rever-
sible action, capable of bringing about both synthesis and hydro-
lysis of starch (Meyer, Salter). Maige, on the contrary, thinks

Guilliermond - Atkinson — 118 — Cytoplasm

that there is in the plastid only a synthesizing diastase and that
the hydrolyzing diastase has its seat in the cytoplasm. The ex-
istence of diastase is, however, not necessary, if the hypotheses
are accepted as formulated by Nageotte and Devaux who con-
sider the plastids and chondriosomes as catalysts.

It is known, moreover, that plastids in the embryo sac of Lilium
candidum may enclose protein crystalloids which are used as re-
serve products. Some experiments seem to indicate that the plas-
tids have the ability to accumulate proteins and it is even possible
that they may be very important synthesizing ceriters (Ullrich,
Granick, etc.). Recently Volkonsky seems definitely to have
furnished proof that the reticulate leucoplast of Polytoma uvella
undergoes considerable variations in volume depending on the na-
ture of the nutriment furnished it. It expands greatly in media
rich in assimilable nitrogen. The leucoplast seems, therefore, to
be the region of the cell to which nitrogenous nutrients most readily
go, especially the amino acids, which are there transformed into
more complex products. This phenomenon is to be considered in
connection with the observations of Noel on chondriosomes in
livers of mammals, and seems to confirm the hypothesis of Robert-
son-Marston, of which more will be said later. Yet Volkonsky
says that this synthesis does not go beyond polypeptides and that
the formation of proteins is completed in the vacuoles.

It is seen that, in reality, we are still very insuflniciently in-
formed on the role of plastids. The close relationship of the plas-
tids and chondriosomes leads us to suppose that the two categories
of elements must have a single function which is very general
and that the function manifested morphologically by the plastids
is only a special example of it. Thus, while admitting with the
majority of cytologists that the chondriosomes have an important
role in metabolism, we can, at the same time, examine the various
hypotheses which have been proposed to explain the role of plas-
tids and that of chondriosomes and which may apply to both
categories of elements.

Purely for historical interest, the theory formulated in France
by PoRTiER (1919), then in America by Wallin (1922) may first
be mentioned. This theory held that the chondriosomes and plas-
tids represent symbiotic bacteria which are found present in all
cells and by means of which all syntheses take place in the
cell. It was based solely on the morphological resemblance of the
chondriosomes to bacteria and on the fact that with mitochondrial
technique the symbiotic bacteria which are encountered in certain
cells, notably those in the bacteria-containing root nodules of the
legumes, stain like the chondriosomes. The theory is untenable,
for even if it is true that the symbiotic bacteria stain as the chon-
driosomes do by these techniques, it signifies nothing since these
stains are not specific and since symbiotic bacteria show histo-
chemical behavior which makes it impossible to confuse them with
chondriosomes (resistance to alcohol, acetic acid, etc.). This the-
ory, successfully opposed by Laguesse, Regaud, Cowdry and Olit-

Chapter XI

119 —

Role of Ghondriosomes

SKY, DuESBERG, LEVI, and MiLOViDOV, is today abandoned by PoR-
TIER himself. Nevertheless it had the merit of initiating investi-
gations which have produced methods by which chondriosomes can
be distinguished in cells from symbiotic and parasitic bacteria,
CowDRY and Olitsky, Duesberg, and Milovidov have described
methods by which, in the cells of nodules of legumes and in the
adipose cells of cockroaches, symbiotic bacteria can be distin-
guished from the chondriosomes by means of differential staining.
By these methods MiLOVlDOV found that the symbiotic bacteria and
the chondriosomes, including the plastids, are both distributed to
the daughter cells during mitosis but not in the same manner. He
has demonstrated, besides, that centrifuging brings about a dis-
placement of the symbiotic bacteria in the direction of the centri-
fugal force but has no influence on
the chondriosomes and plastids. The
symbiotic bacteria, therefore, are
heavier than the cytoplasm and are
heavier than the chondriosomes and
plastids. (Figs. 80, 81).

This work on animal cells led
Regaud to consider the chondrio-
somes as "organelles having an eclec-
tic and pharmaceutical function in
the cell" i.e., as "electosomes". Ac-
cording to this theory, the chondrio-
somes by means of a physico-chemical
mechanism still unknown, draw from
the surrounding medium the mate-
rials necessary to the life of the cell,
transform them and finally release
the product of elaboration, so that it
may be excreted or kept in reserve.
An analogous theory was applied by
P. A. Dangeard to his "vacuome"
which he likened to the chondriome.

Mayer and Schaeffer, basing their idea on reports according to
which the fatty acids contained in the lecithins are made to play
the role of self -oxidizing bodies have suggested that the chondrio-
somes, by virtue of their lipoprotein constitution, might be the
center for the very general function of reduction and oxidation
and, in this way, have a role to play in the respiratory phenomena
and indirectly in cellular synthesis.

Along these same lines may be mentioned reports of a certain
number of authors who have considered that the chondriosomes
and plastids are the source of various diastases or other substances
playing an important role in the oxidation-reduction process in
cells. Various workers have revealed in the chondriosomes the
presence of oxidases or peroxidases (Marinesco, Prenant, Man-
genot) . Chodat and Rouge found that in plant cells the oxidases
are localized in the plastids. In these last years Joyet-Lavergne

Fig. 80. — Cells of lupin. Bac-
teria and chondriosomes differential-
ly stained. The bacteria (grey)
cluster at the poles during mitosis,
the chondriosomes (black) surround
the chromatic spindle. (After MiLO-
VlDOV) .

Guilliermond - Atkinson

120 —

Cytoplasm

and GiROUD have maintained that glutathione is found in the chon-
driosomes. Joyet-Lavergne reports having locaHzed vitamin A
in the chondriosomes and in the plastids, but he does not seem to
have brought forward sufficient proofs for this localization. He
even reports having proved by means of certain reagents that the
oxidation-reduction capacity of the chondriosomes in the Sapro-
legniaceae depends on the presence of this substance in the sub-
stratum. As a matter of fact, the reactions used to detect the
presence of glutathione in cells are not such as to permit it to be
localized in the cytoplasm with any accuracy. At the present time
there do not seem to be any microchemical reagents which can
localize glutathione in the chondriosomes and the hypothesis of
Joyet-Lavergne has not been verified. In fact, we shall see that

the chondriosomes do not seem to have of
themselves any reducing power. In any case,
they are incapable of reducing Janus green
to its leucoderivative, contrary to what has
been thought up to now. As for the oxidizing
role of these elements, it has not been con-
firmed either.

GiROUD and his collaborator report having
localized ascorbic acid, also, within the chon-
driosomes of animal cells by using an acid
solution of silver nitrate. These investigators
moreover, attributed the Molisch reaction
(pp. 54, 104) which characterizes the chloro-
plasts and which they obtained by the same
reagent, to the presence of ascorbic acid in
the chloroplasts which are thus reported to be
the locus of this substance. To verify this
hypothesis, GiROUD and his collaborators
measured the quantities of ascorbic acid
present in a great number of plants, and
found an evident relation between the pres-
ence of chlorophyll and that of ascorbic acid.
This conception, accepted by various investigators, Mirimanofp
was not able to confirm in his recent research. By measuring in
the organs of a rather large number of plants, very accurate quan-
tities of ascorbic acid, he was not able to obtain any direct relation
in these plants between the content of ascorbic acid and that of
chlorophyll. There are organs, totally deprived of chlorophyll,
which are nevertheless very rich in ascorbic acid and, conversely,
there are those poor in the acid which enclose a great deal of
chlorophyll. If sometimes a relation is found between the presence
of chlorophyll in the plastids and their richness in ascorbic acid,
this is, consequently, only indirect and can be explained only by
saying that ascorbic acid, like many other substances manufac-
tured in plants, is an indirect product of photosynthesis. More-
over, MiRiMANOFF has shown that the reaction used by GiROUD is
not specific for ascorbic acid in plants in which very frequently

Fig. 81. — Fixed and
stained cells of lupin
nodules. 1, bacteria re-
sembling chondriosomes.
2, 3, cell centrifuged
before fixing and stain-
ing; bacteria, heavier
than cytoplasm, left at
one side. (After MiLO-

VIDOV).

Chapter XI — 121 — Role of Chondriosomes

there exist phenolic compounds (tannins, oxyflavanol and antho-
cyanin pigments) capable of reducing silver nitrate. Moreover
this reagent is never reduced by the leucoplasts, the chromoplasts,
or the chondriosomes in tissue which is, nevertheless, rich in
ascorbic acid but lacking in chloroplasts and there is reason to be-
lieve that ascorbic acid is localized in the vacuoles. Mirimanoff
thinks that the Molisch reaction has a totally different significance
from that attributed to it by GiROUD. He thinks that it may be
compared to a photolysis, the activator being chlorophyll, the
hydrogen donator, glucose.

In addition, it has been supposed that the chondriosomes and
plastids are chemical catalysts. From this point of view a first
hypothesis was formulated by Nageotte to explain both the role
of the genuine chondriosomes and that of the plastids of green
plants. It is based on research in plant cytology which has dem-
onstrated that the plastids, regarded as a special category of
chondriosomes, are not destroyed during their operations, and it
attributes to chondriosomes and plastids the role of heterogeneous
catalysts. The homogeneous catalysts according to this theory are
represented in the cytoplasm by the diastases and the heterogeneous
catalysts by the chondriosomes and plastids.

Devaux later formulated a different hypothesis, according to
which the role of catalyst is played by the interfaces between the
chondriome and the cytoplasm. Devaux formulated an interesting
suggestion which gives significance to the heterogeneous structure
of the cell and brings out its real importance. He demonstrated
that all solid parts of the cell are orientated molecularly. Each
molecule or elementary particle not only occupies there a definite,
but an oriented, position. All the poles of like affinities occupy
one face of the membrane while all the opposite poles occupy the
other face. These facts made it possible for him to conclude that
the plasmic membranes in particular must constitute the principal
tools of the protoplasm. Applying these data to facts brought out
by cytologists, Devaux, in order to explain the prodigious work
which goes on in the cell, alleges surface actions which may be
reduced to, or classified as, the operations of surfaces in the proto-
plasm. Now all protoplasmic surface, external or internal, is
characterised by the formation of a coagulation membrane. The
polarized (catalytic) membranes are the scene of cellular activities
(activation by surface agency). There is a catalytic localization
of protoplasmic activity on the surfaces presented by the pro-
toplasm: between the cytoplasm on the one hand, and the nucleus,
chondriosomes, plastids, vacuoles and cellular membrane on the
other.

This hypothesis is based on the fact recognized by Otto War-
burg, that, in the eggs of the sea urchin, respiration takes place
essentially all along the protoplasmic membrane. Warburg ap-
plied this notion to the chloroplasts, whose activity he considers
to be purely a surface activity. Devaux says, "It must be con-
cluded that the cell is a system of numerous catalysts in the form

Guilliermond - Atkinson — 122 — Cytoplasm

of small closed sacs, automatically forming and maintaining them-
selves at the same time that they produce all the physico-chemical
transformations taking place in the cell. This establishes the
bond, heretofore mysterious, between cellular structure and vital
activity."

COWDRY and Lecomte du Nouy have shown, moreover, by
meticulous measurements, that the surface of the chondriosomes is
greater than that of the nucleus, although the total nuclear volume
is about five times greater than the total chondriosomal volume of
the same cell. The mitochondrial substance seems therefore to
realize a maximum surface with the minimum of material and
these investigators think that on these interfaces of considerable
area, certain substances may accumulate and the concentrations
attained may allow pluri-molecular reactions of the very highest
importance to take place between the interfaces.

Robertson has formulated a theoiy similar to that of Devaux
based on the data brought out by Marston who showed that dyes
of the azine series, such as Janus green, have a precipitating
action specific for proteases and that, moreover, the action of these
enzymes is doubled in the presence of an emulsion of lecithin, the
surface of the lecithin serving as the catalytic surface. Robertson
thinks, therefore, that the staining of the chondriosomes by Janus
green is an index of the presence of proteases in their substratum
and that the chondriosomes enclose proteases of reversible action,
capable of bringing about both the synthesis and the hydrolysis of
proteins. He believes that the chondriosomes may be the site of
proteosynthesis, a synthesis which takes place by virtue of the
lipide surface of the chondriosome which acts as the catalytic
surface. In the course of his research on the vacuolar system of
animal cells, Parat noticed that the cytoplasm and the chondrio-
somes seem to have a reducing power, while the vacuoles seem to
have an oxidizing power. This investigator is thus led to formu-
late the hypothesis that the chondriome brings about oxidation-
reductions by means of which cellular synthesis is carried out, and
that the second phase, or respiratory oxidation, which follows
these phenomena occurs in the vacuoles. Having observed, more-
over, that the chondriosomes and the vacuoles are often in intimate
contact, Parat supposes that the combination, chondriome + vac-
uome, is responsible for protein synthesis which, according to
Robertson, entails a lipide phase and an aqueous phase. This
author thinks, also, that the vacuolar system may be the region in
which operations begun in the chondriome are completed. Re-
cently Miss Le Breton, basing her ideas on those of Robertson
and Marston, and on investigations of Joyet-Lavergne, Giroud,
and Parat, sought to show that chondriosomes are found in the
conditions necessary and sufficient for them to be the center of
protein synthesis.

Recent work (Guilliermond and Gautheret) does not con-
firm Parat's hypothesis. The fact that chondriosomes stained
vitally with Janus green lose their color rather quickly, a fact

Chapter XI — 123 — Role of Chondriosomes

observed by many authors (Guilliermond, Parat, Sorokin) had
led to the acceptance of the idea that chondriosomes have the power
to reduce Janus green to its leucoderivative. It will be seen that
this interpretation is inexact and that the chondriosomes are in-
capable of carrying out this reduction. All that they do is to
reduce Janus green to its rose derivative, which, having less affinity
for the chondriosomes than for the cytoplasm or the vacuole, dif-
fuses into these latter. The chondriosomes, moreover, share this
property with the vacuole when the latter contains substances
capable of holding the dye and often the reduction even begins in
the vacuoles and is completed in the chondriosomes. On the whole,
Janus green is reduced wherever it is localized and the chondrio-
somes do not seem to have a more active role in this reduction than
the other elements of the cell. The chondriosomes can not then
be considered as having a reducing capacity and the vacuoles as
having an oxidizing capacity.

Nothing is positively known about the role of the chondrio-
somes. One fact, however, stands out from recent research. This
is that the chondriosomes, like the leucoplasts, have the property
of being stained selectively and in a transitory manner by a rather
large number of vital dyes, which dyes later accumulate in the
vacuole. It could therefore be supposed that they behave in the
same way in the absorption of various substances which the cell
takes from the external medium. According to this theory, these
substances are taken up by the chondriosomes, then thrown off
into the vacuole, either directly, or after having undergone some
transformation (synthesis), this transformation being brought
about during contact with the chondriosomes, by the mechanism
suggested by Devaux. This hypothesis, which seems to agree both
with that of Regaud and that of Devaux, explains the accumula-
tion of protein by the chondriosomes of the liver when an animal
is subjected to an exclusively nitrogenous diet (R. Noel) and
explains the capacity of leucoplasts to take up amino acids

(VOLKONSKY).

Apart from these hypotheses, it has been thought, also, that
the chondriosomes and the plastids might have a role in the phe-
nomena of heredity. This was the opinion of Meves. From work
carried out on plants with variegated leaves and branches, that is
to say plants presenting a mosaic of green leaves and colorless
parts lacking chlorophyll, various authors suggested a role that the
plastids seem to have in heredity. Thus, to cite only one example,
a variegated hybrid of Oenothera (O. rubridivaricata) , can form
flowers at the extremity of green branches and at the ends of
colorless branches. Now, after pollinating the flowers on the un-
colored branches by the pollen of flowers on uncolored branches,
Renner obtained entirely colorless seedlings. The pollination of
flowers of uncolored branches by pollen from flowers on green
branches has given a mixture of colorless seedlings, variegated
seedlings and normal green seedlings. In a second series of ex-
periments, Renner pollinated flowers from green branches by

Guilliermond - Atkinson — 124 — Cytoplasm

pollen of flowers produced on colorless branches. Most of the
descendants of such a cross are green, some are variegated but
none are colorless. To explain these data, the hypothesis has been
put forth that the oospheres and pollen grains within flowers grow-
ing on uncolored branches contain only plastids incapable of be-
coming green. In the first series of experiments, the oospheres
with modified plastids reached by pollen tubes with plastids equally
modified, give rise to colorless plants; when reached by pollen
tubes containing normal plastids, they produce variegated or green
plants because a small number of plastids of male origin have
penetrated the oosphere. During the succeeding divisions the
normal plastids are distributed by chance. When some cells re-
ceive only these normal plastids, while others receive modified
plastids of maternal origin, variegated plants are obtained. When
the normal plastids, although brought in in very small numbers, are
distributed among all cells of the embryo, an entirely green plant
is the result. Finally, when normal plastids, having been brought
in in very small numbers remain in a negligible quantity, colorless
plants are obtained. In the second series of experiments, colorless
plants were never procured because in this case the cytoplasm of
the oosphere, which is always in excess of that delivered by the
pollen tube, contains normal plastids. The few variegated seed-
lings give evidence of modified plastids having been brought in by
the pollen tube. Finally, to account fully for the differences between
the results obtained in the two series of experiments, it must be
admitted that the normal plastids, capable of becoming green, in-
crease in numbers more rapidly than do the modified plastids.
That is why the latter, although brought in in large quantity by the
pollen tube, are never represented exclusively in one plant.

The hypothesis is evidently plausible but it has no cytological
basis — at least up to the present — for we have seen that the be-
havior of the plastids and chondriosomes in fertilization is not
known.

Chapter XII

THE VACUOLES

Early data. Insuflficiency of methods. Theory of Hugo de Vries:-
Plant cells always contain, in their cytoplasm, watery inclusions
to which have been given the name vacuoles and which are the re-
gions of accumulation of numerous metabolic products. The
vacuoles contain a liquid called the vacuolar sap which holds in
solution very diverse crystalloid substances, such as mineral or
organic salts, mineral acids, sugars, and so forth.

In cells in the process of differentiation, several vacuoles are
generally found but in most mature cells there is only one enormous
vacuole which occupies the greater part of the cell, the cytoplasm
forming about it only a thin parietal layer containing the nucleus
appressed to the cell wall.

The splendid investigations of the Dutch botanist, Hugo DE
Vries, brought out the essential role of vacuoles in osmotic phe-
nomena of the cell and showed that a mature plant cell may be
likened to a small osmometer, composed of a semi-permeable ecto-
plasmic layer which lines the permeable cell wall on the inside, and
of a vacuole containing a solution of crystalloid substances. No
great modification is observed to take place if mature plant cells
are placed in water, for example, a staminate hair of Tradescantia
which has the advantage of being easily detached and of being
made up of large cells which lend themselves easily to observation.
However, it can often be observed that the vacuole dilates and
by its pressure causes a curvature of the cell wall. The vacuolar
sap, by virtue of the ciystalloid substances which it holds in so-
lution, has an osmotic pressure superior to that of pure water.
It is, therefore, hypertonic to water and so endosmosis takes place.
If the same cells are plunged into a solution hypertonic to the
vacuolar sap, these cells then show a remarkable phenomenon dis-
covered by DE Vries and called by him plasmolysis. There is pro-
duced an exosmosis which causes a contraction of the protoplasm.
The protoplasm becomes more and more detached from the cell
wall, and soon forms in the center of the cellular cavity a globular
mass separated from the cell wall, but remaining connected with
it by very fine cytoplasmic trabeculae which give evidence of ad-
herence of the cytoplasm to this wall. In elongated cells the pro-
toplasm as it contracts divides into several globular masses con-
nected like beads on a string by a thin cytoplasmic filament. This
phenomenon is caused primarily by the exit of water from the
vacuole, accompanied, of course, by a lack of imbibition on the
part of the protoplasm, but it is the contraction of the vacuole
which brings about the contraction of the protoplasm. Whenever
the cells on which the experiment is being carried out contain in

Guilliermond - Atkinson — 126 — Cytoplasm

their vacuoles an anthocyanin pigment which gives them a natural
color, it is observed that as the cells become plasmolyzed the color
of the vacuole is considerably accentuated. The pigment therefore
is becoming concentrated in the vacuole.

The phenomenon of plasmolysis takes place only in living cells.
A dead cell has become permeable and can no longer be plasmolyzed.
Therefore the capacity of the cell for being plasmolyzed serves as a
criterion as to whether the cell is living.

When carried out with certain precautions, plasmolysis does
not cause the death of the cell and may be followed by the converse
phenomenon, if the cell be placed in a solution of pure water.
Endosmosis then occurs which again brings about the dilation of
the vacuole and the cytoplasm once more becomes pressed to the
cell wall. The cell now recovers its normal aspect. It is said to
have been deplasmolyzed.

As the ectoplasmic layer which lines the interior of the cell
wall is generally only relatively semi-permeable, plasmolysis is fol-
lowed, at the end of a period varying according to the case, by a
spontaneous deplasmolyzing action which is brought about by the
gradual penetration into the vacuole of the substance dissolved in
the surrounding medium. Thus there is re-established an osmotic
equilibrium between the vacuolar sap and the external medium.

If plasmolysis is brought about too brusquely, however, it leads
to the death of the cell. As soon as this happens, the ectoplasmic
layer disorganizes and water enters the cellular cavity. The proto-
plasm is coagulated and the cell cavity now contains nothing but
protoplasmic coagulations floating in the water which has accumu-
lated within the cell cavity. Now, during this phenomenon the
vacuole, which is more resistant, remains stretched out in its
habitual shape and, if it contains anthocyanin, this pigment remains
localized within the vacuolar sap. The vacuole appears to be sep-
arated from the medium by a limiting semi-permeable layer. This
layer is resistant much longer than the ectoplasmic layer. The
vacuole may thus be preserved for a very long time. Then its
limiting layer is destroyed in turn, and the contents of the vacuole
blend with that of the cell cavity. In this way the vacuole may
subsist within the cell cavity after the death of the cell. This
phenomenon DE Vries called the isolation of the vacuole, a classical
phenomenon which has been seen since by numerous investigators :
TswETT, Guilliermond, Kuster, Weber, Hofler, Eichberger, etc.

This phenomenon demonstrates that the vacuole is limited ex-
ternally by a peripheral, semi-permeable layer. It is of the same
nature as that which surrounds the outside of the cytoplasm and
is called the endoplasmic layer, or perivacuolar layer. It is how-
ever much more resistant than the outer cytoplasmic layer.

As a result of his work, DE Vries was led to consider this semi-
permeable layer as a differentiated membrane and the vacuoles as
permanent components of the cell. He considered that the vacuoles
are made up of vacuolar sap, containing in solution various crys-
talloid substances, are surrounded by a semi-permeable, differen-

Chapter XII

— 127 —

The Vacuoles

tiated membrane, and are endowed with properties of secretion.
It is this membrane to which DE Vries gave the name tonoplast,
which he beheved secreted all the substances dissolved in the
vacuolar sap. De Vries, moreover, thought that the vacuoles can
not in any case form de novo, but are always transmitted by divi-
sion from cell to cell, like the nucleus and plastids. In short,
DE Vries compared the vacuoles to a sort of liquid plastid to which
VAN TiEGHEM gave the name hydroleucites.

At the time that DE Vries did his work, however, the origin of
vacuoles was unknown. All that was known, and had been known
for some time, was that several vacuoles appear scattered about in
the cytoplasm of cells which are beginning to differentiate; that
they enlarge and then coalesce until in mature cells there is only
one enormous vacuole occupying
the entire cell, pushing the cyto-
plasm and the nucleus back to the
periphery. Aside from this, noth-
ing was known. As a matter of
fact, it is generally impossible to
distinguish the vacuoles in living
embryonic cells. Hence, in the
opinion of the earlier botanists,
the vacuoles were lacking in these
cells and formed de novo in the
course of cellular differentiation,
as seems to be indicated in figure
82, reproduced here from Sachs.

Went, a student of de Vries,
succeeded, however, in revealing
in embryonic cells of certain
plants, the existence of small
vacuoles vv^hich increase in num-
bers by fission. This seemed to in-
dicate that the vacuoles do not

arise de novo, but keep their individuality during the course of de-
velopment, and consequently seemed to support the thesis of DE
Vries.

This theory, nevertheless, was not based on sufficiently solid
facts. It has not, as a matter of fact, been possible to bring out
the vacuolar membrane by means of stains and its existence is
manifested only by its property of semi-permeability. It is true
that there is the phenomenon of isolation of the vacuoles, which
obliges us to admit that there exists, at least in mature cells, a
semi-permeable membrane about the vacuoles, more resistant than
the rest of the cell. Chambers and Hofler who studied this mem-
brane during micromanipulation, describe it as a membrane of
inappreciable thickness, very cohesive and extensible, formed of a
substance non-miscible with water. The significance of the peri-
vacuolar layer is still very much disputed. Some observers regard
it as a differentiated membrane (Hofler), others as formed by a

Fig. 82. — Fritillaria imperialis. Devel-
opment of vacuoles in the cortical paren-
chyma. (After Sachs).

fe^'

Guilliermond - Atkinson — 128 — Cytoplasm

coagulation of the cytoplasm which is in contact with the vacuoles.
Pekaeek thinks it is protein in nature. Weber thinks it does not
belong to the cytoplasm but to the vacuole and is the outcome of a
condensation about the vacuole of the lipides found in solution
within it. According to Eichberger, this perivacuolar layer is of
the same nature as the ectoplasmic layer, composed, like it, of pro-
teins and lipides but much more resistant. He examined the layer
over a period of several days in abiotic solutions, for instance in a
5% solution of copper sulphate and over a period of from two to
three hours in a 20% solution of calcium citrate, which proves that
it acts like an inert membrane and not like a membrane differen-
tiated and living, in the sense of DE VRIES.

On the other hand, it is very difficult, if not impossible, to ob-
serve the vacuoles in embryonic cells and consequently to demon-
strate that they do not form de novo. Therefore the theory of
DE Vries and Went was contested by numerous botanists, among
others, Pfeffer and Nemec. Pfeffer, in particular, showed that
it is possible to produce artificial vacuoles in the Plasmodium of
Chondrioderma difforme by placing it in a solution saturated with
asparagin. The Plasmodium encircles the crystals of asparagin,
which, dissolving in the cytoplasm, form there vacuoles which are
indistinguishable in all ways from the pre-existing vacuoles. For
this reason, Pfeffer thinks that every particle found in the cyto-
plasm which will take up water more readily than it, is capable of
producing a vacuole. It is true that Pfeffer's experiment does not
demonstrate very much, for here it is a question of the production
of digestive vacuoles which are perhaps not the same as the other
vacuoles.

The question of the origin of the vacuoles remained uncertain
for a very long time, because neither observation of living material
nor fixed and stained preparations made it possible, in general, to
follow the development of these elements. Vacuoles in living em-
bryonic cells can not ordinarily be distinguished and in fixed prep-
arations the vacuoles are always altered by swelling.

Rapid progress in this field w^as not made until very recently
when investigators began to use vital stains.

Chapter XIII
VITAL STAINING OF THE VACUOLES

Colloidal substances in the vacuolar sap:- The existence of dyes,
called vital dyes (methylene blue, cresyl blue, Nile blue, neutral
red, etc.) has been known for some time. They have the property
of penetrating living cells and of coloring some of the cytoplasmic
inclusions.

Pfeffer first showed that methylene blue used in a 1% solu-
tion penetrates the cells of various plants (Azolla, Lemna, Spiro-
gyra) but accumulates exclusively in the vacuole and does not color
the protoplasm. Methylene blue brings about dark blue precipi-
tates in the vacuoles as a result of the flocculation of the tannins
contained in the vacuolar sap, with which the stain forms a com-
plex. Pfeffer stressed the remarkable property which the vacuoles
possess, of accumulating methylene blue and of taking on very
rapidly a coloration more intense than that of the solution of the
dye itself. The dye becomes, therefore, more concentrated in the
vacuole than it was in the solution. In staining tadpoles, Pfeffer
showed that methylene blue accumulates in certain inclusions of
the cytoplasm. Since that time it has been demonstrated that other
dyes, such as cresyl blue, Nile blue and neutral red, also have the
property of penetrating living cells, and numerous investigations
on plant cells as well as on animal cells have determined that these
dyes accumulate exclusively in the inclusions which are not a part
of the living matter (secretion granules of animal cells, vacuoles
of plant cells) .

In research on yeasts, we ourselves have demonstrated that the
vacuoles enclose granules, showing Brownian movement, which
have the property of being stained instantaneously by cresyl blue
and neutral red. These granules can be easily fixed with alco-
hol and formalin, after which they are stained deeply by basic
stains (anilin blue, anilin violet and haematein) which give them
a reddish tint. This has permitted us to identify them as meta-
chromatic corpuscles. These were cited first in the bacteria by
Babes and found later in the Cyanophyceae by Butschli, who gave
them the name of "red granules". In later research also, we
showed that vacuoles formed by hydration of aleurone grains (Fig.
84), when seeds germinate, accumulate neutral red and this dye
brings about the flocculation of the proteins which the vacuoles con-
tain in solution. But these are isolated observations. It remained
for P. A. Dangeard to establish the fact that the ability to accumu-
late vital dyes is a general property of vacuoles.

Taking up our work on the metachromatic corpuscles in the
fungi and algae, Dangeard demonstrated that these bodies are
visible only rarely in living cells and when they are visible, they

Guilliermond - Atkinson

130 —

Cytoplasm

always appear in much smaller numbers than when obtained by
vital staining or after fixation. These bodies, therefore, are gen-
erally the result of flocculation of a substance found normally in a
colloidal solution in the vacuolar sap. This flocculation occurs in
the presence of vital dyes or fixatives. P. A. Dangeard kept for
this substance the name metachromatin which we had pro-
posed to designate the substance constituting the metachromatic
corpuscles (volutin of ARTHim Meyer).

P. A. Dangeard had the idea of trying the effect of vital stains,
among others cresyl blue, on a very large number of cells of the
most varied plant groups. In every one he found that there existed
a colloidal substance dispersed in the vacuolar sap possessing a
strong capacity for taking up vital stains which precipitate it in
the form of corpuscles showing Brownian movement. These he
identified with the metachromatin of fungi. Thus he arrived at the
conclusion that all vacuoles enclose metachromatin, a specific sub-
stance of vacuoles, and he states that vital staining thus constitutes

a property of the vacuoles
which is both general and
characteristic.

Studying the origin of
vacuoles in most varied
plant cells by staining them
with the vital dye, cresyl
blue, Dangeard demon-
strated that vacuoles exist
in all embryonic cells but
are very different in ap-
pearance from ordinary
vacuoles. They are seen
here as numerous minute
elements composed of a very concentrated solution of metachro-
matin in a semi-fluid state. By their forms, as well as by their
dimensions, they are decidedly reminiscent of the chondriosomes.
It is these elements, swelling by imbibition and coalescing, which
finally become the large vacuoles characteristic of mature cells.
We shall not dwell on this matter here, but will study it later in

more detail.

Our research immediately afterward, confirmed, in part, the
facts observed by P. A. Dangeard. We recognized that the meta-
chromatin of fungi is normally found in a colloidal solution in
vacuoles. Vital dyes, for example neutral red, precipitate this
substance as corpuscles, intensely stained and showing Brownian
movement in the vacuolar sap which either remains colorless or
takes a diffuse stain. We also found, by using vital dyes in the
most diverse plant cells, that colloids are present as precipitable
pseudosolutions. Contrary to Dangeard's opinion, our work
showed, as we shall see further on, that the colloidal substances
contained in the vacuoles are of very different chemical natures and
do not show in the higher plants, nor even in many of the lower

Fig. 83. — Saccharomyces eerevisiae. Precipitation
of metachromatin corpuscles (CM) in vacuole: 1,
by neutral red; small vacuole at right contains one
corpuscle. 2, haematein, after fixation with formol;
GL, very refractive lipide granules. N, nucleus.

Chapter XIII

— 131

Vital Staining

plants, the characteristic properties of metachromatin, with which
they have in common only the ability of forming absorption com-
plexes with vital dyes.

It is well established, therefore, that the affinity of the vacuoles
for vital dyes, with some exceptions which will be taken up
later, is a general property of vacuoles and that this is due,
not to the presence in the vacuoles of a specific substance cor-
responding to metachromatin, but to colloidal substances whose
nature may vary radically, depending on the species in question.

The presence of colloids in the vacuoles has, moreover, been
confirmed by another method. This has been employed by Weber,
Frey, and Reilhes, who tried to determine the degree of viscosity
of the vacuolar sap by observing the rapidity with which certain
solid bodies (calcium oxalate or calcium sulphate crystals, lipide
concretions) contained in the vacuoles fall
in their liquid when the microscope is in-
clined. Frey was able to show by this meth-
od that the viscosity of vacuolar sap of Clos-
terium cells at 18° C. is about twice as great
as that of water. This viscosity increases
considerably in dead cells. After cells have
been fixed, it becomes impossible to obtain a
fall of the calcium oxalate crystals contained
within their vacuoles. This can be explained
only by the fact that they are surrounded
within the vacuoles by coagulated colloids
which prevent them from moving.

Finally, the more recent research of
Weber on "vacuolar contraction," which
will be taken up later, brought forward
new data in favor of the presence in the
vacuoles of colloidal substances which may
be true gels.

Fig. 84. — Zea Mays.
Living cell stained with
neutral red, from the
aleurone layer at the be-
ginning of germination of
the seed; aleurone grains
have become diffusely stain-
ing vacuoles containing one
or more deeply stained
protein bodies (P) and one
or more colorless globoids
(G).

Action of vital dyes on the cells. Advantages of vital staining :-

When colloidal substances endowed with an affinity for vital dyes
had been demonstrated to be generally present in vacuoles, there
remained the task of determining more specifically the action of
these dyes on the cells. Do they accumulate exclusively in the
vacuoles or may they stain at the same time other cytoplasmic in-
clusions? Have they an injurious action on the cells and are they
not capable of altering the shape of the vacuoles? Or may they
not even bring about the appearance of artificial vacuoles in the
cytoplasm? These are questions which were raised. It was all
the more important to elucidate them since the vacuoles of em-
bryonic cells appear as minute elements with very concentrated
colloidal contents which usually can be revealed only by the use
of vital stains. Very accurate work carried out recently on vital
staining — our own work carried out over a period of twenty years
and especially our recent work in collaboration with Gautheret —

Guilliermond - Atkinson — 132 — Cytoplasm

has shed considerable light on this question. The essential results
will be summarized here as briefly as possible.

It appears, as Kuster had already noticed, that living plant
cells are permeable to a great number of vital dyes. These may
accumulate in all the vacuoles, whatever their content. Others,
such as methylene blue, Bismarck brown and chrysoidine, may
stain the cytoplasm and the nucleus and may accumulate in the
vacuoles, but only if the vacuoles contain phenol compounds (tan-
nin, oxyflavanol and anthocyanin pigments). The cytoplasm and
nucleus are particularly easy to stain with chrysoidine.

In general it is only the basic dyes which penetrate living cells.
Under some conditions, however, some acid dyes will do this, but
the greater number of the dyes are not of importance in the ques-
tion which occupies us.

There are only a small number of vital dyes, all of them basic,
which are capable of being used for the study of morphological
constituents of the cytoplasm and these may be divided into two
groups from the point of view of their action on plant cells:

1. Dyes, which like Janus green. Dahlia violet, methyl violet
and a certain number of others, at first stain the chondriosomes
and plastids, for which they have an affinity, but can also under
some conditions accumulate in the vacuoles. (Among these dyes, Bis-
marck brown and methylene blue are very slightly toxic. We have
been able to germinate grains of wheat and have made Saprolegnia
grow in media to which these dyes have been added in proportions
from 0.0005-0.02%. In young wheat roots, as well as in Sapro-
legnia, the vacuoles which stain only between slide and cover glass
and only under certain conditions (especially when they enclose
phenolic compounds), accumulate the dyes during growth. Chry-
soidine, beginning with solutions of 0.005% is, on the contrary
very toxic).

2. Dyes, which, like neutral red, neutral violet, cresyl blue,
Nile blue, naphthylene blue and naphthylamine blue, do not stain
the chondriosomes and plastids but, under normal conditions, ac-
cumulate exclusively in the vacuoles.

The dyes of the first group are very toxic and do not produce
vital staining except when employed in solutions of low concen-
tration. At higher concentrations they are taken up by the chon-
driosomes in cells which remain alive for some time as shown by
their cytoplasmic currents, but in these cells the dyes rapidly cause
death. The staining is, therefore, sublethal. Recent work (Guil-
liermond and Gautheret) has shed a good deal of light on the
question of vital staining of leucoplasts and chondriosomes, which
until now had been rather obscure. This work has shown, as has
been already seen, that a certain number of dyes may be taken up
at first by the cytoplasm and nucleus, or by the chondriosomes and
leucoplasts, but this staining is purely transitory and the dye goes
into the vacuole very quickly. It is only when the cells have ex-
creted the stain into the vacuoles that the cells are capable of grow-
ing and multiplying.

Chapter XIII —133— Vital Staining

This work has demonstrated that the toxicity of the vital dyes
which stain the chondriosomes is approximately the same for
phanerogams as for fungi. The least toxic is Janus green. Ger-
mination of wheat seeds may be obtained in media to which 0.0005-
0.001% of Janus green has been added. The roots grow well in
solutions as high as 0.005% but at higher concentrations their
growth is inhibited and they soon die. The other dyes are much
more toxic. Wheat seeds will germinate in them, with the excep-
tion of methyl green and Victoria blue, only in solutions between
0.0002-0.0008%. The dye is never taken up by the chondriosomes
and leucoplasts but accumulates only in the vacuole (Janus green
generally in its rose form) . Elodea canadensis can be kept alive in a
0.0005% solution of Janus green or of methyl violet. At the be-
ginning, the dye is taken up only by the chondriosomes but after
a short while these elements lose their color and the dye accumu-
lates only in the vacuole (in its rose form in the case of Janus
green) .

It has also been possible in this work to follow the reduction of
Janus green to its rose derivative as it occurs in the course of vital
staining. If a strip of epidermis from a bulb scale of Allium Cepa
is colored between slide and cover glass in a 0.0005-0.005% solu-
tion of Janus green, the dye is too dilute to produce a macroscopic-
ally visible staining of the epidermis, but a microscopic examination
of the preparation shows that Janus green has been taken up by
the chondriosomes and leucoplasts. After several minutes, these
elements lose their color and there is no longer any trace of the
dye to be seen in the cell. If the same experiment is repeated in
solutions of 0.01-0.02% of Janus green, the strip of epidermis is
stained green macroscopically but at the end of one or two hours,
it changes to rose. Janus green has therefore been reduced to its
rose derivative. This reduction is irreversible and the rose deriva-
tive can no longer by re-oxidation resume its green form. If air
is excluded by sealing the slide with paraffin, reduction is more
rapid and can be followed under the microscope in a single cell.
At the beginning, the dye stains only the chondriosomes and leuco-
plasts, the chondriosomes more intensely than the leucoplasts, and
accumulates at the same time in those vacuoles which contain oxy-
flavanol compounds. At the end of about half an hour in cells m
which the vacuole is not stained, the chondriosomes and leucoplasts
lose their green color at the same time that the cytoplasm and
nucleus take on a rose tint. In cells in which Janus green has
accumulated in the vacuole, reduction often begins there. The
vacuole changes from violet to rose, more accentuated than the
green stain that it first showed, while the cytoplasm and nucleus
generally remain colorless. In reality the chondriosomes and leuco-
plasts only partially destain, for these two categories of elements
show a pale rose tint. The cells which have reduced Janus green
to its rose derivative can remain alive without air and show cyto-
plasmic currents for 48 hours. With concentrations above 0.02%
of Janus green, reduction of the dye to its rose derivative is some-

Guilliermond - Atkinson

— 134 —

Cytoplasm

times obtained but usually the cells die without having brought
this about.

The study of the behavior of the rose derivative of Janus green
in these cells explains these phenomena since the derivative does
not behave like Janus green. Whereas Janus green stains only the
leucoplasts and chondriosomes, or the vacuole if it contains oxy-
flavanol compounds, the rose derivative shows a greater affinity
for the nucleus and cytoplasm, as well as for the vacuole containing
oxyflavanol compounds, than it shows for the chondriosomes and
leucoplasts which it stains only faintly. Therefore, if one uses a
dilute solution of Janus green (0.0005-0.005%) the dye stains only

Fig. 85. — Vital staining with neutral red, except 03,
observed under the microscope. A, PeniciUium glaucum. 1,
before staining; 2, small deeply stained precipitates in the
vacuole showing Brownian movement; 3, fusion of small
precipitates to larger bodies; 4, precipitates appressed to
peripheral wall of vacuole, diffuse staining of sap. B,
Zygosaccharomyces Chevalieri. 1, small precipitates in
vacuole; 2, 3, fusion, bodies now appressed to wall of the
vacuole, sap diffusely stained. C, Saccharomyces ellipsoideus;
1, 2, as in A, 1-2; 3, cells fixed by formol stained with cresyl
blue which causes flocculation of the metachromatin from the
colloidal substances in the vacuole as numerous, deeply
stained bodies.

the chondriosome and leucoplasts but it stains them faintly. It is
almost immediately reduced to its rose derivative, however, and
this diffuses into the cytoplasm, the nucleus, or the vacuole, and its
concentration is too weak to produce in them any appreciable
coloration. On the contrary, a 0.01% solution of the dye taken up
by the chondriosomes and leucoplasts, is reduced to its rose deriva-
tive which, being less strongly retained by these elements, diffuses
into the cytoplasm and nucleus which it colors a pale rose in cells
whose vacuoles are lacking in oxyflavanol compounds. In the case
of the vacuole which, on the contrary, contains oxyflavanol com-
pounds and has accumulated Janus green, reduction goes on at the
same time in the chondriosomes and in the vacuole, and a part of
the rose derivative formed in the chondriosomes diffuses into the
vacuole. Often indeed, the reduction begins in the vacuole and
takes place later in the leucoplasts.

Chapter XIII

— 135 —

Vital Staining

This reduction is observed between slide and cover glass and
in tissue incapable of development. It is probable that under nor-
mal conditions, i.e., in culture or in an organ capable of growth,
the rose derivative, once formed, accumulates in the vacuoles be-
fore the growth of the tissue.

It has been seen that in Saprolegnia similar phenomena take
place. In many fungi, however, especially in the yeasts, Janus
green may be taken up by the cytoplasm at the same time as by
the chondriosomes, then be reduced there to its rose derivative (or
even to its leucoderivative in anaerobic conditions) and this deriva-
tive is later excreted into the vacuole.

Among the dyes of the second group are some which are toxic,
for example naphthylene blue and naphthylamine blue. There are
others which are less so, such as cresyl
blue and especially Nile blue. Finally,
there are still others which are a very
little toxic — neutral red and neutral
violet, by far the least toxic of all these
vital dyes. All these vital dyes pro-
duce coloration of the vacuoles which,
as will be seen further on, is essen-
tially a vital phenomenon, for it is pos-
sible only during the life of the plant.

Neutral red and neutral violet are
closely allied dyes. They behave in the
same manner and are particularly in-
teresting. From our research on the
behavior of neutral red, it is found
that except in rare cases, the dye is
not retained by the cytoplasm and that
the coloration of the living cell which
results, is almost always strictly lim-
ited to the vacuoles (Fig. 85) . We have
found only a few rare yeasts in which
neutral red may color certain lipide in-
clusions of a very special nature at the same time as the vacuoles,
although we have examined a considerable number of cells belong-
ing to the phanerogams and to the most diverse fungi observed
with the aid of this stain. In the Myxomycetes, however, Man-
GENOT showed that neutral red colors certain sphaerocrystals of a
phenolic nature. In the algae, some mucilaginous inclusions are
known with which tannins are often associated, which are stained
by the vital stains at the same time as the vacuoles. But it appears
that these inclusions, to be taken up later, may be considered as
vacuoles of a special nature.

The different phases of staining the living cell with neutral red
may be followed under the microscope. The vacuoles, as we have
said, appear in various forms. They may be small and contain
colloidal substances in very concentrated solutions, as in embryonic
cells, or they may be very large and contain a very dilute colloidal

M

C 1

Fig. 86. — Penicillium glau-
cum. Vital staining with neutral
red. A, B, C. 1, precipitates
in vacuole; 2, migration of pre-
cipitates into cytoplasm. D,
1-5, stages observed in a single
vacuole.

Guilllerniond - Atkinson

136

Cytoplasm

solution as in differentiated cells. Neutral red behaves differently
with the two types of vacuoles.

In the small vacuoles of very concentrated colloidal contents,
neutral red does not cause any precipitation and stains deeply and
homogeneously. These small vacuoles are not usually visible with-
out the assistance of vital dyes. There are cases, however, in which
they appear very distinctly because of the anthocyanin (red or
violet) which they always contain and which gives them a natural
color. Such is the case in the vacuoles of teeth of young rose leaflets
(Fig. 92), which will be mentioned later. The study of these
vacuoles in the living state makes it possible to see that they have
exactly the same shape as those which are obtained by vital stains
in cases where the vacuoles are not otherwise visible. There are
cases in which the small vacuoles of meristematic cells, because of

the high refractivity of their contents,
are perfectly visible without vital dyes
(barley root, wheat root, first leaves of
the bud in Iris germanica) . The stain-
ing of these small vacuoles with neu-
tral red may be followed under the
microscope and it may be observed
that this staining is not accom.panied
by any alteration — on condition, of
course, that the observation is not too
greatly prolonged, for at the end of a
certain time a swelling of the vacuoles
always occurs.

The large vacuoles with very disperse
colloidal contents are, on the contrary,
always visible without staining and it
is very easy with the microscope to
follow the different phases of their
staining with vital dyes. The phenom-
ena are very clear cut, particularly in the fungi (molds and yeasts) .
For example, by placing cells of Saccharomy codes Ludwigii grown
in a van Tieghem and Le Monnier moist chamber (Fig. 88) in a
drop of neutral red solution, it is observed that there are immediate-
ly produced in the vacuoles, a great number of granules strongly
stained and showing Brownian movement. These are the result
of precipitation of the vacuolar colloid through the action of neu-
tral red. It sometimes happens that these precipitates, carried
against the wall of the vacuole, pass through it and are deposited
in the peri vacuolar cytoplasm (Fig. 86), a phenomenon which is
also caused by fixatives and which leads to errors of interpretation.
This precipitation occurs even if an extremely dilute solution of
neutral red is used. The reaction is, therefore, very delicate and
the vacuolar colloid is very readily stained by the dye, but if the
solution is very dilute, the phenomenon stops with the production
in the vacuoles of small colored granules, showing Brownian
movement. If, on the contrary, the solution is more concentrated,

Fig. 87. — ■ Saccharomycodes Lud-
wigii. Vital staining. Neutral red
produces small precipitates in the
vacuole, showing Brownian move-
ment (o) which then fuse to form
larger bodies (o') sticking to the
periphery of the vacuole (a") and
then dissolve, leaving the vacuole
dififusely stained {a'"). B, similar
series.

Chapter XIII — 137 — Vital Staining

the granules quickly coalesce into a small number of large
globules (sometimes into a single globule) which come to
be closely appressed to the wall of the vacuole. Then they
diminish little by little in volume and disappear, while the
entire vacuole takes on a diffuse stain which later becomes more
pronounced.

To summarize these phenomena briefly : there is a precipitation
of the vacuolar colloid, followed, when the stain is sufficiently con-
centrated, by a dissolution of the precipitate and by the homo-
geneous staining of the vacuolar sap.

These phenomena may be compared with those described by
VON MoLLENDORFF in animal cells. He found, in staining with
neutral red the cells of the pronephros of amphibian tadpoles, that
basic dyes stain the fluid and acid inclusions of the cytoplasm (cor-
responding apparently to small vacuoles) and cause the precipita-
tion of the colloid of which they are composed, in the form of small
precipitates. Then, if an excess of the electropositive dye occurs
within the fluid electronegative inclusions, the precipitates formed
later in the fluid inclusions under the action of the dye are finally
dissolved and give to these inclusions a homogeneous color. To
explain this, voN Mollendorff supposes that the basic dyes pene-
trate the cytoplasm because of the lipides which it contains and in
which the dyes are soluble. Then they accumulate in the previ-
ously formed acid inclusions. There, the mixture of two colloids
of opposite signs, i.e., the electropositive dye and the electronega-
tive vacuolar colloid, produce precipitates. Then, when an excess
of dye is found in the vacuolar colloid, it communicates its charge
to the colloid, whereupon the precipitates are dissolved.

It seems justifiable to apply this reasoning to the staining of
vacuoles of plant cells with vital dyes and we shall see, further on,
that the work we have carried out on yeasts, in collaboration with
Gautheret, seems indeed to confirm voN Mollendorff's hypo-
thesis. There is yet to be explained why the protoplasm remains un-
colored. Neutral red has an oxidation-reduction potential which
makes it unthinkable that it could be reduced in the cytoplasm,
and it must be supposed that the dye traverses the cytoplasm in a
degree of concentration which is too weak for its color to be per-
ceptible, and then later accumulates in the vacuole.

The summary of vital staining which we have just given for
the vacuoles of yeasts with a dilute colloidal content applies to a
great number of cells, notably to the cells of the majority of fungi,
but it is not altogether general. In other cells, vital staining op-
erates in a different way. There are cases in which there is no
production of precipitates in those vacuoles which from the begin-
ning take a diffuse stain. There are other more frequent cases, in
which both precipitates and a diffuse staining of the vacuolar sap
occur at the same time. This diversity of behavior of the vacuoles
seems to depend on the nature of the colloids which they contain
and which are known to vary greatly from one type of cell to
another.

Guilliermond - Atkinson — 138 — Cytoplasm

It would seem today, in the light of investigations of BUNGEN-
BERG DE Jong, that the precipitation observed in vacuoles under the
influence of vital dyes may perhaps be explained, in many cases,
not by flocculation, but by the production of coacervates in the
vacuolar sap. According to this idea, there is a formation of a
complex between the positive dye and the negative vacuolar col-
loid. The colloid may then flocculate as in the yeasts, or else, if the
micelles are strongly bound to the dispersion medium, there is a
separation of a coacervate phase in the vacuolar sap which, how-
ever, still contains dispersed colloidal micelles, as the diffuse color
taken by the vacuolar sap indicates. If the dye continues to pene-
trate the vacuole, at first the coacervate phase becomes greater and
greater. Then the micelles of the coacervate take the charge of
the dye and, their repulsion becoming greater than their force of
coherence, the coacervate disappears and the vacuole becomes uni-
formly stained. An observation which would seem to support this
opinion further, is that the precipitates formed under the action
of the dye, seem often to be in liquid droplets, capable of changing

shape and even of becoming vacuolized.

^r«^^^^^^^^g5>^ Moreover, it seems that all intermediary

r^^lP^'^^^^^fcih^^ stages in the action of vital dyes between

m^^^^^^m/^^^'- -m coacervation and flocculation can be

^^^^^^^^^^^^^^f Another point very clear from our

^^^ 1^!' ^'iiy'illlf llM l*^ investigations is that staining of the

^^^r—s^J**^ vacuoles by neutral red is essentially a

Fig. 88. — Petri dish designed vital phenomenon. As a matter of fact,

ing material! '" °^^^'''"'*'°" °^ '"" although ncutral red is only slightly

toxic, if used in too strong concentra-
tions, it may cause the death of the cell. This is preceded by an
increase in the refractivity of the c>i:oplasm. Then suddenly the
vacuole becomes colorless, whereas the entire protoplasm takes on a
dark red color.

There is no formation of vacuoles or artificial granules pre-
ceding death in any of the cells which we have observed, and death is
always accompanied by destaining of the vacuole. Staining of the
vacuole is possible, therefore, only when the cell is living and this is
a general fact which applies to the staining action of all vital dyes.

This fact may be even more satisfactorily observed in the lower
algae, for instance in the Euglenas, which are endowed with move-
ment. The cells of these algae accumulate neutral red in their
vacuoles as long as they are moving but once their movements
cease and the cells die, the vacuoles destain and the dye colors the
cytoplasm and nucleus. The conclusion to be drawn is that the
staining of the vacuoles in a cell constitutes one of the best means
of determining if it is alive. Vital staining of the vacuoles and
that of the chondriosomes and leucoplasts are therefore clearly
differentiated. Vital staining of the last two is possible but when
it occurs it is transitory. Usually it is sublethal and does not dis-
appear when the cells die.

Chapter XIII

139 —

Vital Staining

We have tried, by cultures of various plants in nutrient media
to which neutral red has been added, to follow under the micro-
scope the development of the vacuoles during cellular growth. This
has, in addition, made it possible for us to judge the degree of
toxicity of this dye. We have been able to obtain cultures of Sapro-
legnia, for example, on soy bean bouillon to which 0.001% -0.002%
of neutral red or neutral violet have been added. The cultures were
grown in Petri dishes (Fig. 88) whose bottoms had a 3 cm. opening
covered by a cover slip sealed by asphalt cement (model used at the
International Bureau for the Culture of Fungi at Baarn). It suf-
ficed to turn the box over and place it under
the microscope to be able to follow the devel-
opment of the fungus under the oil immer-
sion lens and to follow the life history of the
vacuoles. Various species of Saprolegnia cul-
tivated undere these conditions developed as
well as those in the control cultures and fol-
lowed out their entire life history from ger-
mination of the zoospores to the formation of
zoosporangia. Now, during all their develop-
ment the dye accumulated in their vacuoles
and stained them superbly. In solutions of
neutral red beginning with a concentration of
0.005%, growth is somewhat retarded. The
fungus can stand relatively large doses of
neutral red (0.05%-0.06%), although when
the concentration of the dye is above a cer-
tain paint, it grows only very little.

With DuFRENOY and Labrousse, we were
able to germinate seeds of tobacco in pure
culture on Knop's solution to which had been
added doses of neutral red from 0.005%-
0.02%. We succeeded in doing the same
with grains of wheat, barley and lupin seeds,
in collaboration either with Obaton or with
Gautheret. The seeds germinated normally
under these conditions and by examining

their roots in the Petri dish under a microscope it was possible to
see that during the entire growth of the root, neutral red was
accumulated in the vacuoles of the meristematic cells, in the dif-
ferentiating cells of the cortex, in the root hairs and in the cells
of the root cap. Eichhorn found also that roots of Allium Cepa
grow perfectly in a solution of neutral red and show entirely
normal mitoses. Gautheret for a period of several months kept
alive the cells of the root cap of lupin in media to which neutral
red had been added. The vacuoles in these cells were stained. In
addition it may be recalled that Skupienski succeeded also in
obtaining the complete development of one of the Myxomycetes,
Didymium nigripes, in a medium to which neutral red was added.
The vacuoles were stained during the entire growth of this plant.

Fig. 89. — Saprolegnia
dioica. Zoospores grown
on gelatinized soy bean
bouillon containing 0.001%
neutral red. 1, before ger-
mination; vacuolar system
consisting of numerous glo-
bular bodies stained by the
dye. 2, id.; nucleus visible.
3, 4, fusion of small vacu-
oles into one large one con-
taining granules precipi-
tated by neutral red. 3, in
the germination tube small,
spherical, uniformly stained
vacuoles are formed which
in (5) elongate into fila-
ments, form a network and
then fuse to form large
vacuoles.

Guillierrnond - Atkinson — 140 — Cytoplasm

This is, then, a valuable method of following under the micro-
scope the entire life history of the vacuoles during cellular growth
and we shall see further on the uses to which it may be put.

A fact becomes evident from these investigations, namely, that
vital staining with neutral red is impossible if the medium is too
acid. The starting point at which staining occurs lies between the
pR values of 5.5 and 7, according to the type of cell in question. For
roots of phanerogams it is, for example, 5.5, for the Saprolegniaceae
6.5, for yeasts and Oidium lactis, 7. This is an essential fact pre-
viously brought out by Irwin and Pischinger whose work was
later confirmed by Bailey and Zirkle, Genaud, Chadefaud and
others.

It is curious to find that, with the exception of the Saproleg-
niaceae, most of the fungi are distinguished from all the other
plants by their behavior in the presence of neutral red. Cultivated
in media to which this dye has been added, they develop very readily
but are never stained. We wondered why fungi which accumulate
the dye so easily in their vacuoles when placed between slide and
cover slip in a solution of neutral red, do not ever accumulate it
while they are growing.

The species of Saprolegnia, in the conditions under which we
cultivated them, never appreciably modify the pB. value of the-
medium although other fungi make it more acid. The failure of the
latter to be stained could therefore be attributed to a rapid acidi-
fication of the medium. This hypothesis, which we accepted at first
and in which Becker and Skupienski later concurred, is, how-
ever, not to be retained, for acidification of the medium may be
considerably retarded and it is found that the fungi, even at a pH
favorable to staining, do not take up neutral red. Recent research
which we have carried out with Gautheret on Oidium lactis and
various molds, has shown that these fungi accumulate neutral red
only when they have ceased to grow. These same investigations
also proved that when the spores of these fungi are sown in a
medium to which neutral red has been added, they take up the
dye, at first, and then become destained as soon as they begin to
grow. If the cells of Saccharomyces ellipsoideus are sown, for ex-
ample, in a moist chamber on an aqueous medium containing 1%
peptone and 1 % glucose at a pB. of 7.5-8 to which has been added
0.005% neutral red and enough agar to hold them in place so
that they can be found under the microscope and if the cells are
examined microscopically, it is found that all the cells accumulate
the dye in their vacuoles even between slide and cover glass. The
accumulation is at its maximum at the end of half an hour, then,
after about three hours (two hours for some yeasts), the cells lose
their color and it is only then that they begin to bud. The loss of
color is brought about by a process which is the converse of that
by which staining was accomplished. The homogeneously stained
vacuolar sap loses its color and there are seen to appear in the
vacuoles, intensely stained granules which little by little lose their
color and disappear (Fig. 90).

Chapter XIII

— 141

Vital Staining

The oxidation-reduction potential of neutral red scarcely per-
mits this destaining to be attributed to a reduction of the dye and
various experiments seem to indicate that yeasts do not reduce
neutral red. The destaining of the vacuoles can only be explained,
therefore, by assuming a destruction of neutral red, or an ex-
cretion of it by the yeasts. The following experiment throws light
on this problem. One half gram of Saccharomyces cerevisiae at a
pH of 8 is sown in a big flask containing the medium described
above to which 0.005% of neutral red has been added. A sample
of the liquid is taken at regular intervals and centrifuged. The
sediment is examined under the microscope each time and the con-
centration of neutral red of the liquid is measured by Meunier's
photo-electric colorimeter. The experiment proved that, at the
end of half an hour, all the cells accumu-
lated neutral red in their vacuoles and that
the concentration of the dye in the liquid
was reduced from 0.005% -0.0015%. There
is therefore a very great absorption of the
dye by the yeast. At the end of half an
hour, the vacuoles begin to lose their stain
and the concentration of neutral red in the
liquid again increases and finally at the end
of an hour all the cells are destained and
the concentration of the dye in the liquid
has returned very nearly to its original
amount. The vacuoles are then complete-
ly destained. Now, various experiments
having shown that under these conditions
there is no absorption by the membranes
of the cells, one is obliged to conclude that
the yeasts, after having accumulated the
neutral red in their vacuoles, excrete it
into the medium. So it is only when they
are freed of the dye that they begin to bud.

Saprolegnia behaves differently from the yeasts, since it ac-
cumulates neutral red while growing. Nevertheless it seems also
to be able to excrete the dye under some conditions. In fact, al-
though it always accumulates neutral red during growth, it does
not keep it in the vacuoles unless the medium is poor in nutrients
(soy bean bouillon with agar). If the culture is grown in a richer
medium (peptone) it is found that, after having accumulated the
neutral red at the beginning of growth, the cells suddenly lose their
color at the end of 48 hours, i.e., at the moment of maximum
growth. From this point of view Saprolegnia occupies an inter-
mediate position between the fungi on the one hand, which absorb
neutral red only when their growth is arrested, and which reject
the dye as soon as they begin to develop, and the phanerogams on
the other hand, whose cells accumulate neutral red during their
growth and retain it in their vacuoles during their entire de-
velopment.

Fig. 90. — Saccharomycea
ellipsoideus grown on 1% pep-
tone and 1% glucose containing
0.005% neutral red. A. 1,
First strongly colored precipi-
tates are formed in the vacuoles;
2, The precipitates dissolve and
the vacuole is diffusely stained;
4, The color disappears; 5,
Budding. B. Similar pheno-
mena. 1-3, staining; 4, 5, loss
of color.

Guilliermond - Atkinson — 142 — Cytoplasm

The other dyes which stain vacuoles, i.e., cresyl blue, Nile blue,
naphthylene blue and naphthylamine blue, behave a little differ-
ently. Between slide and cover glass, if the solution is too strong,
they may, at the same time that they accumulate in the vacuoles,
stain the cytoplasm and the nucleus with a diffuse color. This
diffuse color occurs, furthermore, only in the phases which precede
the death of the cells. At all events, with these dyes as with
neutral red, the vacuoles destain and the staining of the cytoplasm
and nucleus becomes accentuated as soon as death occurs. The
coloration of the vacuoles by these dyes, as is the case for neutral
red, is therefore possible only as long as the cell is living. It may
be added that if the dyes are used in weak concentrations, in the
fungi especially, they may be reduced in the cells and one sees the
coloration disappear under the cover slip and reappear if the latter
is raised for purposes of aeration.

We have also been able to germinate wheat grains in a medium
to which these dyes had been added. Nile blue proved slightly
more toxic than neutral red and neutral violet ; cresyl blue is very
appreciably more toxic than Nile blue ; naphthylene blue and naph-
thyalmine blue are extremely toxic. These dyes are accumulated
exclusively in the vacuoles of root cells, just as is neutral red, and
brilliantly stain the vacuoles of meristematic cells, of root hairs
and of cells of the root cap. The staining persists as long as the
cells are in a living condition.

These dyes are more toxic for fungi. They do not generally
accumulate in the vacuoles but are reduced by the Saprolegniaceae
and by other fungi when cultivated in their presence.

In collaboration with Gautheret, some of the experiments on
Saccharomyces cerevisiae that we carried out with neutral red
were repeated, this time using Nile blue and cresyl blue. The
experiments showed that these dyes behave like neutral red but
have a much more complex action on the cells. These two dyes,
Nile blue especially, may stain the cytoplasm. At a high pU, for
example, Nile blue is at first retained exclusively by the cyto-
plasm which it stains diffusely. Then, at the end of a certain
time, the dye is taken up by the vacuoles and from there excreted
into the medium. When there is insufficient aeration, as is usually
the case, the dye, held at first by the cytoplasm, is reduced by the
cytoplasm to its leuco-derivative, in which form it goes into the
vacuole and finally, still in its colorless state, is excreted into the
external medium where it is re-oxidized in the presence of air.

Cresyl blue is only weakly retained by the cytoplasm and accu-
mulates especially in the vacuole, but it may be reduced, as is Nile
blue, before being excreted into the medium where it is re-oxidized.

Investigation of these dyes was carried further and the accumu-
lation of neutral red, Nile blue, and cresyl blue by yeast cells was
studied at different pB. values. The method consists of sowing a
known quantity of yeast (S. cerevisiae) in Schoen's medium (min-
eral salts, asparagine, glucose) to which 0.005% of the dye has
been added at pU values scaling from 3-10. At the end of twenty

Chapter XIII —143— Vital Stainin g

minutes, the liquid was centrifuged, the sediment examined under
the microscope and the concentration of the dye in the liquid
measured by the colorimeter. It was thus possible to plot in
terms of pU the quantity of dye retained by the yeast. A study
of the curves showed that the lower limit of pU at which accumu-
lation of neutral red begins is much lower than 7. It is at pB.
5.6. But at that point the staining of the vacuole is too weak to
be visible under the microscope and only the measurements give
evidence of the absorption of the dye. This is appreciable only
beginning with a pU of 6.0. From a value of 5.6, the absorption
gradually increases and the curve rises to a maximum at pR 7.6.
From this point the curve descends. The quantity of neutral red
absorbed at the maximum pB. value is found to be 0.0042 g. for

Amount absorbed in grs. per cent of fresh yeast

0.5 g. of yeast (suspended in a 100 c.c. solution) , which gives a con-
centration of 0.84 g. of neutral red for 100. g. of fresh yeast.

In the experiment with cresyl blue the lower limit at which
accumulation of the dye begins is also at pH 5.6 but the curve does
not redescend until about pH. 10. The maximum quantity of
cresyl blue absorbed is 0.0041 g. for 0.4 g. of yeast which makes a
concentration of 1 g. of cresyl blue for 100 g. of fresh yeast. The
curve for Nile blue is very different, for the lower limit is at pK
2.4 (absorption begins in a very acid medium) and steadily in-
creases to pK 10, at which point the curve flattens out. The maxi-
mum quantity of Nile blue absorbed by 0.5 g. of yeast in the
presence of 0.005 g. of Nile blue is 0.015 g. which makes a con-
centration of dye of 3 g. for 100 g. of fresh yeast.

These rather disconcerting results demand explanations. Two
facts are indeed surprising: first the descending portion of the
curve for neutral red after the maximum pK value is reached ; sec-

GuillJermond - Atkinson — 144 — Cytoplasm

ondly, the lower limit for accumulation of Nile blue placed at a pB.
value below 3 (2.4).

These apparent irregularities are easily interpreted. In study-
ing the accumulation of neutral red by yeast, it is noticed that the
quantity of dye retained by the cells varies incessantly, because of
the alternate excretion and accumulation of the dye. There is not
at any time, therefore, a true equilibrium and hence no real sig-
nificance can be attributed to the absorption curve of neutral red in
terms of pH values.

To obtain satisfactory results, a 1% solution of KH2PO4 must
be used. In it the yeast does not grow and does not excrete the
dye. Under these conditions, the amount of dye absorbed by the
yeast varies very little and a state of equilibrium is reached which
makes it possible to study accurately the accumulation of neutral
red in terms of pH. The curve obtained in this medium is similar
to that for cresyl blue. It begins at pB. 5.6, rises steadily to pB. 9
and then flattens out.

With regard to the second point, namely, the behavior for Nile
blue, it must be taken into account that this dye is absorbed by the
protoplasm of yeast and if the other dyes (neutral red and cresyl
blue) have their lower limit of accumulation set at pB 5.6, one
may think that the isoelectric point of their cytoplasmic colloids
must be situated a little below this pB and that there occurs in its
neighborhood a tightening of the micelles which prevents the
penetration of these dyes, for they have not, like Nile blue, an
affinity for the cytoplasm.

Our experiments showed, in addition, that the phenomenon of
the accumulation of vital dyes in the vacuoles seems, up to a certain
point, to follow the law of Freundlich and to correspond to an
adsorption^

The net result of these facts is that vital dyes, although not
having any injurious effects on cells when used in weak doses, do
behave as toxic substances and are always excreted into the vacu-
oles, where they remain (phanerogams), or from which they may
be excreted into the exterior medium. The facts show us at the
same time that the vacuoles are the centers of accumulation for
toxic substances collected in the cells. These facts show that cer-
tain dyes may be taken up by the cytoplasm but that this is always
transitory and the dye is very shortly excreted into the vacuole. It
is only when the dye has been localized in the vacuole that the cell
can grow and divide. There is an important difference, however,
between the behavior in the phanerogams and in the fungi (Sapro-
legnia excepted). Whereas the phanerogam cells can grow when
the dye has accumulated in the vacuole, those of the fungi can not
and excrete the dye accumulated in the vacuole into the external
medium. Once rid of the dye, the fungal cells begin to grow.
Nevertheless our recent research (Guilliermond and Gautheret)

1

^Translator's vote. A fuller discussion of the entire subject is to be found in an article by
Guilliermond, received after this volume went to press: La Coloration vitale d'apres des
Travaux recents (Montpellier Medical 1940, No. 1:1-24).

Chapter XIII -145- Vital Staining

minimizes this difference which formerly seemed inexplicable. In-
deed, this work has shown that Saprolegnia which grows with the
dye accumulated in its vacuoles can itself, in the course of develop-
ment, excrete the dye from its vacuoles into the external medium
under some conditions. This fungus, therefore, has a behavior
intermediate between that of the phanerogams and that of the
other fungi. It is not certain that the phanerogams themselves
may not be capable of excreting the dye in their vacuoles to the
exterior, but this is done with difficulty and under special
conditions.

The purpose for exposing here the essential data of these still
little known investigations before beginning the study of the vacu-
oles is to emphasize the following points. Among the dyes which
stain the vacuoles, neutral red, Nile blue and cresyl blue are only
very slightly toxic and, in general, bring about only a staining of
the vacuoles. In addition, the staining of the vacuoles can be pro-
duced only during the life of the cell and is essentially a vital phe-
nomenon. Furthermore, it is proved experimentally that neutral
red is much the least toxic of these dyes and that it is never retained
by the cytoplasm and, except in rare cases, accumulates only in
the vacuoles. It is a stain which is almost specific for vacuoles
and is the most useful dye for the study of these elements. It has,
besides, the advantage of not being reduced by the cells, as are
cresyl blue and Nile blue (in the case of the fungi). Neutral red,
therefore, is indisputably the best vital stain for vacuoles.

This being the case, the importance of the method of vital stain-
ing in the study of the vacuoles is readily appreciated. This vital
staining makes it possible not only to observe the vacuoles in
preparations between slide and cover glass but also to follow, at
least in most plants, the entire life history of these elements in
cultures which are growing on media to which vital dyes have been
added. This method of staining living material has led to im-
portant discoveries which we shall now set forth.

Chapter XIV
DEVELOPMENT OF THE VACUOLAR SYSTEM

First stages in development:- The starting point of the recent
discoveries on the question of the development of the vacuolar sys-
tem was an observation made by us on the mode of formation of
anthocyanin pigment in teeth of leaflets from the rose bud.
We noticed that the pigment begins to form at the extremity of the
tooth. In examining a tooth from tip to base, all the phases in

A ■^•M^sV

4f'^

3

ii^ ^>i

^Mi!^#

D

^ ^ '•. -l^f ^ ..V-

B"iG. 92. — Teeth of young, living rose leaflets containing anthocyanin pigment which
makes visible the vacuolar system developed from filamentous vacuoles which swell, anastomose
and fuse to form one large vacuole per cell. A, D, cells at tip; B, C. older cells. D, after Pensa.

the formation of anthocyanin may be followed. Now, we ob-
served that in the youngest cells, namely, those at the tip, the
pigment appears as minute, numerous, filamentous elements very-
like the chondrioconts. These elements, taken all together, are
exactly like a chondriome. In the region nearer the base, these
elements seem to swell and be transformed gradually into small
vacuoles, which, by their fusion, finally form a single enormous
vacuole, occupying the major part of the cell and enclosing the

Chapter XIV

147 —

The Vacuolar System

pigment in solution. By reason of the great resemblance between
the initial shapes in which anthoeyanin first appears and the shapes
of the chondrioconts, we were led to think, originally, that this
pigment arose in the elements of the chondriome which then be-
came transformed into vacuoles. This interpretation was founded
also on the fact that these shapes were preserved by mitochondrial
techniques. With the method of Regaud, for example, we ob-
tained at that time, both typical chondriosomes and elements of
the same form as the chondriosomes, but a little larger, stained
in the same way, but for which the staining was less stable and
which, if destaining was prolonged, lost all the dye and took a
yellow color. This color we attributed to the action of the potas-
sium bichromate on the anthoey-
anin. Now, between the typical
chondriosomes and these elements
which resemble them, there
seemed to exist all intermediate
stages. We thought at that
time, therefore, that the larger
elements, to which the potassium
bichromate gives a yellow color,
corresponded to chondriosomes
impregnated with anthoeyanin.
Our interpretation was a natural
one at the moment, for the chon-
driome was not yet well known,
the origin of the vacuoles was not
understood, and it was thought
that most of the pigments of ani-
mal cells were of mitochondrial
origin.

This observation was immedi-
ately investigated by a certain
number of workers among whom
some agreed with our interpreta-
tions (MoREAU, MiRANDE) and others contested them. Among the
latter Arthur Meyer thought, but without having proved it, that
the chondriosome-like elements which mark the beginning of the
formation of anthoeyanin, represent filamentous vacuoles. Low-
SCHIN, on the other hand, expressed the opinion that these fig-
ures correspond merely to anthoeyanin itself. According to him
the anthoeyanin is deposited in the cytoplasm in this form and
consequently there is merely a fortuitous similarity in shape be-
tween these figures and that of the chondriosomes.

Pensa, stressing the fact that the chondriosome-shaped ele-
ments of anthoeyanin always appear colored yellow by mitochon-
drial methods — probably in preparations too much destained —
was led to think that they have no relation to the chondriosomes.
Taking up Lowschin's theory and drawing upon his own observa-
tions, Pensa concluded that these figures correspond merely to an

Fig. 93. — Oil glands of a walnut leaf.
Regaud's method. 1, 4, preparations prop-
erly stained; chondriosomes (Af) stained
black, filamentous vacuoles (VA) containing
an anthocyanin-tannin complex stained yel-
low brown. 2, 3, preparations insufficiently
destained; 2, both elements black; 3, only
vacuoles visible. Preparations such as 2, 3,
might lead observers to think the chondrio-
somes are transformed into tannin-contain-
ing vacuoles.

Guilliermond - Atkinson — 148 — Cytoplasm

aggregated state of anthocyanin and thiat this pigment might,
according to the conditions in which the cell is at the time, assume
two different colloidal states in the cytoplasm: an aggregated
state characterized by the chondriosome-shaped elements scattered
throughout the cytoplasm and a state of dispersion, i.e., a state
of pseudosolution in the cytoplasm. By treating with alkaloids
those cells containing anthocyanin in the state of a pseudosolution
in the cytoplasm, Pensa claims to have obtained a return of the
pigments to the aggregated state, characterized by the produc-
tion of anthocyanin granules assembled in little chains or in a
network. Thus, according to Pensa, the chondriosome-shaped ele-
ments of anthocyanin do not always coincide with the stage of the
formation of pigment. But Pensa's interpretation is erroneous,
for the author did not understand that the pigment is dissolved
in the vacuoles and not in the cytoplasm, and that the phases which
he attributed to the state of pseudosolution of the pigment corre-
spond to the dispersion state of the anthocyanin within a single
enormous vacuole, occupying the major part of the cell and sur-
rounded only by a thin layer of parietal cytoplasm. That which
Pensa considers to be the cytoplasm is, therefore, nothing more
than vacuolar sap. Alkaloids do indeed bring about flocculation
(an aggregated state) of anthocyanin in the form of granules
showing Brownian movement. The granules are precipitates of
tannin absorbing the pigment as they form. These precipitates
may assemble in little chains or in a network, and were erroneous-
ly likened by Pensa to the figures observed in meristematic cells.
In the meristematic cells it is a question of small elements shaped
like granules or filaments, resembling the chondriosomes, made up
of a concentrated colloidal solution quite different from the vacu-
oles known at that time. Dangeard later gave these small ele-
ments their true title of young vacuoles. There is nothing in
common between these small vacuoles and the precipitations of
anthocyanin obtained by Pensa in mature cells.

Credit must go to P. A. Dangeard for having oriented this study
in a new direction. We have already said that in studying the
origin of the vacuoles in very diverse plants by means of vital
stains, among others cresyl blue, this investigator found in the
meristematic cells of plants and in the growing tips of fungal
hyphae that vacuoles always appear as numerous and minute ele-
ments in the form of granules, isolated or united in little chains,
or of filaments which often anastomose in networks, staining very
deeply and homogeneously with vital dyes. These very closely
resemble the chondriosomes and are composed of a very con-
centrated colloidal solution. They are elements which by hydra-
tion become transformed little by little in the course of cellular
differentiation into fluid vacuoles.

Later, taking up our observations on the origin of anthocy-
anin, P. A. Dangeard (C/. p. 130) demonstrated that, just as
Arthur Meyer had predicted, the elements which we had described
as chondriosomes in reality represented young vacuoles with very

Chapter XIV — 149 — The Vacuolar System

concentrated colloidal contents, the enclosed pigment giving them
a natural color. Consequently, the formation of anthocyanin, from
its beginning, is associated with the manner in which the vacuoles
are generally formed. Struck by the resemblance between young
vacuoles and the chondriosomes, as we ourselves had been, Dange-
ARD was led at the beginning of his research to liken them to chon-
driosomes and to think that the forms described under this name
in animal cells corresponded to certain aspects of vacuolar develop-
ment analogous to those forms encountered in plant cells.

He believed, furthermore, that he could demonstrate that the
colloidal substance of which the vacuoles are formed, corresponded
to a special substance found in all vacuoles, regardless of the type
of cell, which he identified with metachromatin (volutin of Meyer) .
Consequently, in spite of current opinion, there could be no rela-
tion between the chondriosomes and the plastids. Hence Dange-
ARD gave the name vacuome, or vacuolar system, to all the vacuoles
contained in a cell in the various phases of its development. This
expression was destined, in his own mind, to replace that of chon-
driome and the term mitochondrium that of mitochondrial sub-
stance. "The greatest error of cytologists," he said, "is to have
confused the chondriome and the metachromatin with the plastids.
This, at any rate, is what I am going to try to demonstrate. The
chondriome, which has been the object of so much investigation,
must, in my opinion, be considered otherwise than it has been to
the present time. It may be defined as the whole vacuolar system
in its various and successive aspects."

Starting from the fact that vacuoles are stained by dyes and
are capable of taking up almost the entire amount of dye in a
solution, Dangeard made his vacuome play an essential role in
the phenomena of nutrition of the cell. According to him, meta-
chromatin, the substance specific for the vacuome, plays at one and
the same time an osmotic and a selective role. It fixes the nutri-
ents as it, in turn, is fixed by the vital stains. In this way DANGE-
ARD explains the formation of anthocyanin pigments. These, ac-
cording to him, arise in the cytoplasm and are fixed by the
metachromatin of the vacuome by virtue of its selective power.
Thus Dangeard transfers to the vacuome, the hypothesis which
Regaud had proposed to explain the role of the chondriome.

Lastly, P. A. Dangeard and his son, P. Dangeard, think that
the vacuoles are permanent elements of the cell and multiply only
by division, thus agreeing with the conception of DE Vries, but
with this difference that, for the Dangeards, the vacuoles by de-
hydration of the metachromatin may become solid in certain
phases. This is true for aleurone grains and vacuoles of dormant
spores of fungi.

This notion of taking the chondriome for the vacuome, which
rests exclusively on observations made with vital stains, was inad-
missible, it being true that at that time it had already been
demonstrated that the chondriosomes are not stained by the dyes
used by P. A. Dangeard, but only, and then with difficulty, by

Guilliermond - Atkinson

150

Cytoplasm

other dyes which do not have a predilection for the vacuoles. More-
over, the chondriosome-shaped vacuoles are of temporary nature
and, even beginning with the first stages of cellular development,
these vacuoles take in water and become transformed into large
vacuoles, whereas it is known that the chondriosomes persist dur-
ing the entire life of the cell. The only question that could be
asked was whether, as we had supposed, the vacuoles were not
derived in some cases from the chondriosomes. It was not logical
to say that the chondriome and the vacuome were one and the
same thing. The interpretation of Dangeard has therefore been
abandoned for a long time now, even apparently by its author,
although without explicit statement to that effect.

Fig. 94. — Barley root. Stages in development of the
vacuolar system. 1-5, meristem; 6-8, adjacent region of differ-
entiation; 9, mature cells of cortical parenchyma. Vital staining
with neutral red.

Our investigations, beginning with that period, gave more
details and indeed confirmed the observations of P. A. Dangeard
in so far as the development of the vacuoles is concerned, but
they showed that the chondriosome-shaped vacuoles and the chon-
driosomes have in common only their shape, and that the vacuome
is a system completely independent of the chondriome. They
showed, furthermore, that the vacuoles are not characterized by a
specific substance corresponding to metachromatin and that the
colloids which the vacuoles contain are substances of very diverse
nature, having as a common property only their ability to absorb
vital stains. These facts were afterwards verified by a large num-
ber of cytologists and are today definitely accepted. The vacuoles
do not, in general, stain by mitochondrial techniques and it is the
very complex case of the leaflets of the rose which has caused all
the difficulty. It has been seen that in the cells in the young teeth of

Chapter XIV

151 —

The Vacuolar System

these leaflets (Fig. 92) there are chondriosome-shaped vacuoles con-
taining a substance which, when preserved with mitochondrial
methods, becomes yellow, but which may be stained by iron haema-
toxylin and which, if the destaining is not carried far enough, ap-
pears black. This would make one think that these bodies corre-
spond to chondriosomes impregnated with anthocyanin. In reality,
as we were able to demonstrate by later work, they are vacuoles
containing tannin with which anthocyanin is associated. Now,
tannin, like the lipide substance of the chondriosomes, is rendered
insoluble with fixatives containing potassium bichromate, and
may be stained by iron haematoxylin. On the other hand,
the study of plants in which
anthocyanin is not associ-
ated with tannin, makes it
possible to observe that this
pigment is not preserved by
mitochondrial techniques and
that the young chondrio-
some-shaped vacuoles in
which pigment forms do not
stain with mitochondrial
methods. Therefore, Dan-
GEARD, instead of correcting
a partial error committed by
us, made the mistake of gen-
eralizing it.

Let us examine in more de-
tail the development of the
vacuoles in the phanerogams.
In the barley root (Fig. 94),
for example, the phenomena
studied by Dangeard and
then by us, are particularly clear. If a very young root of a
seedling is studied in a solution of neutral red by crushing the root
gently, in such a way as to dissociate the cells without injuring
them, numerous minute elements are seen in all the cells of the
meristem. They are scattered about in the cytoplasm and are
stained deeply and homogeneously by the dye. In the very young-
est cells these elements are all small granules but they soon
elongate to undulous filaments which often afterwards anastomose
into a network. By their form and their dimensions these ele-
ments show a striking resemblance to the chondriosomes and an
observer not forewarned would easily take them for such. Never-
theless, they are distinguishable at first sight from the chondrio-
somes by a much greater diversity of appearance and by the fact
that they are reticulate. They seem to be composed of a very con-
centrated colloidal solution of semi-fluid consistency whose refrac-
tivity is such that they can easily be seen without vital staining.
In the course of cellular differentiation, it is observed that these
elements swell by absorbing water and then coalesce. They are of

Fig. 95. — Young flower of Iris germanica.
Early stages in the development of the vacuolar
system in the trichomas of sepals. Vitally stained
with neutral red.

Guilliermond - Atkinson

152 —

Cytoplasm

most diverse appearance : dumb-bell-shaped, granules arranged like
a string of beads, club-shaped, spindle-shaped, networks of monili-
form filaments. They then become transformed into small, spher-
ical vacuoles which always stain uniformly, but as they continue to
take in water, they soon appear only faintly colored. They some-
times contain a few deeply stained precipitates, which show Brown-
ian movement and are caused by the precipitation of some of the
colloidal contents of the vacuole under the influence of the dye.
These vacuoles fuse and finally, in mature cells, form a single large
vacuole which occupies the major part of the cell, pushing the
nucleus and cytoplasm to the periphery. The cytoplasm is now-
reduced to a thin layer around the vacuole which appears faintly
colored and shows only a few precipitates.

Fig. 96. — Anagallis arvensis. Stages in the formation of a glandular
hair on the corolla; anthocyanin pigments pi'esent from the first (in vivo).

The development of the vacuolar system in the root of the
wheat may be quite as easily observed and the phenomena take
place in the same way. In most of the phanerogams and the
pteridophytes\ moreover, an analogous development of the vacu-
olar system is recognized. Excellent examples are furnished by
the epidermal cells of very young leaves of Iris germanica, by the
cells of very young hairs from the sepals of the same plant (Fig.
95), by the glandular hairs on the leaflets of the walnut (Fig. 93),
by the leaves of Anagallis arvensis and others. In the glandular
hairs, the vacuoles, like those in the teeth of the rose leaflets, con-
tain from the very beginning an anthocyanin pigment which makes

' In the apical cell of the pteridophytes, however, there are found only large liquid vacuoles
and in the cells of the meristem which are derived from the apical cell there are no small
filamentous vacuoles (Emberger). It seems as if the apical cell had already passed through
a stage in which the vacuoles were filamentous and semi-fluid, after which these filamentous
vacuoles had taken in water and coalesced.

CHiapter XIV

— 153

The Vacuolar System

it possible to follow their entire development without using vital
dyes. The large vacuoles of mature cells, instead of staining dif-
fusely without precipitation or forming only a very few precipi-
tates, may behave differently. Very often, as for example, in the
epidermal cells of Iris germanica, the vital dyes do not stain the
vacuolar sap and only produce in the vacuole deeply stained bodies
showing Brownian movement or else bring about, at the same
time, both a diffuse coloration of the vacuolar sap and the pro-
duction in the sap of colored precipitates.

- v*-

• Vf f

i^

•••

F^G. 97. — Pea root vitally stained with neutral red.
1-4, meristem; numerous small filamentous vacuoles, uniformly
stained. 5-7, differentiating cells; swelling and fusion of small
vacuoles, whose colloidal substance is precipitated by the dye
as deeply stained bodies showing Brownian movement.

So the vacuoles seem to be at first small elements composed of
a very concentrated colloidal solution which, by taking in water,
gradually swell and coalesce during the period of differentiation
of the cells, and become large vacuoles containing an extremely
dilute colloidal solution. This progressive dilution of the vacuolar
colloidal solution is made evident by the fact that neutral red,
which at first gives the vacuoles a deep homogeneous color, stains
them only weakly when the cells have attained a certain degree of
dilution but generally brings about a precipitation of the colloid
as deeply stained bodies which show Brownian movement. The
precipitation is more or less copious depending upon the nature
and concentration of the colloid.

When the vacuoles have come to the end of their development,
they may, in the presence of vital stains, stain weakly and homo-
geneously without showing any, or at most very few, precipitates,
as in the roots of barley and wheat (except for the cells of the
root cap). In other, more frequent cases, vital dyes bring about
the formation of numerous deeply stained precipitates, which show
Brownian movement, at the same time that they stain the vacu-
olar sap diffusely (vacuoles in the epidermal cells of Iris germa-
nica) . In still other cases, the vital dyes at first cause only colored
precipitates. These later fuse and may dissolve in the vacuolar
sap which then becomes diffusely colored.

Guilliermond - Atkinson

154 —

Cytoplasm

Rather often, there are normally found in the vacuoles more
or less large, spherical bodies, or small granules united in mul-
berry-shaped masses, which are the result of a partial precipita^
tion of the colloidal solution within the vacuoles. These bodies
stain with vital dyes which, at the same time, bring about other
precipitates of the same or of different natures. The normal pres-
ence of these bodies in the vacuole could be explained by the fact
that the colloidal micelles contained in the vacuoles in some cases
do not possess a power of unlimited imbibition, so a time seems
to come when they cease to take in water. A disturbance in equi-
librium occurs and this leads to the production in the vacuolar
sap of a coacervate.

It may be added that rather frequently the large vacuoles of
mature cells, especially when they contain tan-
nins, continue to enclose a very concentrated so-
lution and in living material are exceedingly
refractive. These vacuoles which seem to be in
the state of a jelly do not form precipitates with
vital dyes or else form them with great difficulty.
In the latter case they show an intense homo-
geneous color. There are even cases in which the
jelly is almost solid and in such cases it becomes
very difficult to plasmolyze the cells, as is seen
in the vacuoles of the pericarp of Ilex Aquifolium
(Guilliermond, Chaze).

Vacuoles behave differently according to the
nature of their contents.

The development of the vacuolar system which
we have just described is of very general occur-
rence and has been observed in very widely sep-
arated plants (phanerogams: P. A. Dangeard,
Guilliermond, P. Dangeard, Bailey, Zirkle and
others; pteridophytes : Emberger).

All the vacuoles of the higher plants however,
do not follow exactly this development. Thus, in
studying the formation of the vacuoles in the bud of Elodea canor-
densis, it is observed that in all the cells of the meristem of the
stem and of the youngest foliar primordia, there are numerous, very
small vacuoles which are always globose and never filamentous.
These generally stain uniformly and deeply with neutral red, but
in certain cases there is seen in their interior a single, deeply
stained, corpuscle showing Brownian movement. This corpuscle
has been produced by a precipitation of the colloidal contents of
the vacuole. These vacuoles, which are not visible in living mate-
rial without being stained, seem to be constituted of a less concen-
trated colloidal solution than are those in ordinary cases. They
later swell up by taking in water and then gradually coalesce to
form, in mature cells, a single vacuole which neutral red stains
diffusely while also causing the production of numerous precipi-
tates. There are, therefore, no filamentous or reticulate formations

Fig. 98. — Bud of
Elodea canadensis vi-
tally stained with
neutral red. Vacu-
oles globular and uni-
formly stained; or
small and colorless
with one or several
colored precipitates.

Chapter XIV

155 —

The Vacuolar System

but only spherical vacuoles to be found at the beginning of develop-
ment of this type of vacuolar system.

Among the fungi, the Saprolegniaceae constitute particularly
favorable objects for the study of the vacuolar system, to which we
will have to return later. When the mycelium of a species of Sapro-
legnia is examined in a solution of neutral red, there are observed
at the growing extremities of the plant, vacuoles which at first
are generally small globular elements. These elongate and appear
as thin, tenuous canaliculi, more or less oriented in the direction
of the longitudinal axis of the hypha. These canaliculi anastomose
and form a delicate and complicated network. In this region they
appear more fluid and with too little difference in refractivity be-

FiG. 99. — Saprolegnia vitally stained with neutral red.
Development of the vacuolar system. 1-3, tips of growing
filaments; anastomosing canaliculi. 4, older filaments; fusion
and swelling of canaliculi. 5, later stage; precipitated bodies
in sap. 6, still later stage; single canal containing precipi-
tates.

tween them and the cytoplasm for them to be observed without
staining. A little further from the tip of the hyphae, these canali-
culi are observed to swell, little by little, then to converge, so that
in differentiated regions of the hyphae, there is only a single canal
running from one end of the siphon to the other. This canal
occupies the major part of the hyphae and in it the vital dyes pro-
duce both numerous, semi-fluid, precipitates and diffuse staining
of the vacuolar sap. The cytoplasm, accordingly, now constitutes
only a thin layer about this canal and the nuclei are appressed to
the wall of the siphon.

This vacuole, therefore, is of a special type; namely, a single
canal running from one end of the filament to the other and obvi-
ously adapted to the siphonate structure of the Saprolegniaceae.
This form can be found in Vaucheria but it is a rather rare type.

Guilliermond - Atkinson

— 156 —

Cytoplasm

In most fungi, except for the Phycomycetes, the vacuoles appear
in the tips of the growing hyphae as very numerous, small, globular
elements which sometimes stain uniformly and deeply with the
vital dyes and sometimes remain uncolored but contain a deeply
stained corpuscle showing Brownian movement (Fig. 100). These
vacuoles swell in regions farther away from the tip, then coalesce,
until, in the regions still farther away, they form large colorless
vacuoles filled with deeply stained corpuscles. These bodies swell
after a time, then may dissolve and give the vacuole a diffuse and
homogeneous color. This is also true in the yeasts in which there
exist in the bud several small, spherical vacuoles, sometimes uni-
formly stained with neutral red, sometimes not
stained at all, but containing colored corpuscles.
These small vacuoles fuse during the growth of
the bud until there is present only one large vacu-
ole or a few large vacuoles filled with stained cor-
puscles showing Brownian movement. (Fig. 101).
In some of the lower plants, the vacuoles de-
velop quite differently. In many algae, the Con-
jugatae for example, they are always in the form
of large liquid vacuoles. In other algae they
may, on the contrary, appear during the entire
cellular development as very small, semi-fluid,
usually globular, vacuoles, scattered about in
the cytoplasm and never undergoing hydra-
,„SXnJ;;-t.'^vftai tion. These vacuoles, formed of a very concen-
staining with neutral tratcd colloldal solution, stain homogeneously and
vacuoLf ''n/T*^co^tafn deeply with neutral red, usually without precipita-
intenseiy colored gran- ^jqj^_ -pj^jg jg ^j^^ ^ q£ vacuolc generally f ound

ules {CM) which . ., ^, , ^ ,, , ,t-, i t-. • t • j

show Brownian move- m thc Phytoflagcllates (Euglcuas, Peridmieae and
^spond"to''b^L''°of Volvocales observed by the Dangeards), in cer-
metachromatin ob- tain land forms of the Chlorophyceae (Pleuro-

coccus, Pleurastrum, Prasiola reported by PUY-
maly). The Cyanophyceae also contain similar
vacuoles which are found localized in the parietal
cytoplasmic layer surrounding the central body,
i.e., the region which corresponds to the nucleus (GuiLLlERMOND).
The bacteria seem to belong to this category. In them, by vital
staining or after fixation, there are observed, especially at the
poles of the cell, metachromatic corpuscles which seem to corre-
spond to small vacuoles with very concentrated metachromatin
(Guilliermond, Mile. Delaporte). Vacuoles of this type are en-
countered in the conidia, in the spores and in the zoospores, of
fungi and algae (zoospores of Saprolegniaceae and Ulothrix (Fig.
103), for example), and in pollen grains (P. Dangeard, Mile. Py).i
The data which we have just reviewed, confirm, therefore, the
observations of P. A. Dangeard. These data show that vacuoles

tained with Bouin's
fixative stained with
haematein. Gl, lipide
granules around vacu-
oles.

iMlle. Py has shown that in dehydrating pollen in a vacuum, a solidification of these
vacuoles is obtained. The vacuoles become comparable to aleurons grains withm the pollen
grains, although the pollen grains do not lose their viability.

Chapter XIV

157 —

The Vacuolar System

seem to be encountered in all cells in all phases of development and
that, in general, they appear in embryonic cells as minute elements
composed of a very concentrated colloidal solution and that they
sometimes show a great resemblance in form and dimensions to
the chondriosomes. These chondriosome-shaped elements swell and
are transformed into large liquid vacuoles containing very dilute
colloidal solutions. These facts do not, however, confirm Dange-
ard's interpretation in which he identifies the vacuolar system
with the chondriome. Although in some cells the young vacuoles
present forms almost identical with those of the chondriosome^,
there are other very numerous cases in which the young vacuoles,
on the contrary, have an appearance which does not permit of any
confusion with the chondriosomes. For example, in some algae,
the vacuoles remain constantly in the
state of large liquid inclusions. Nev-
ertheless, since it is evident that
these two categories of elements may
sometimes be easily mistaken for one
another, it is advisable to examine
here the characteristics which make
it possible to distinguish between
them.

hhpf^

®?^^^

Fig. 101. — Penicillium glaucum.
Vital staining with neutral red. 1-5,
filaments; one or more large, deeply
stained bodies precipitated in the
vacuole by the dye; small vacuoles in
branches seen to form de novo. 6-
19, germinating conidia; small vacu-
oles in germination tube seem to
form de novo.

Chondriosome-shaped vacuoles and
chondriosomes. Characteristic dif-
ferences:- The vacuoles, even in
their chondriosome-shaped state, are
essentially distinct from the chon-
driosomes in their histochemical
characteristics. An inherent differ-
ence rests in the fact that although
the vacuoles stain deeply with vital
stains (neutral red, cresyl blue, Nile
blue, etc.), the chondriosomes, on the
contrary, have no affinity for these dyes and are not colored in the
living state except with special stains (Janus green. Dahlia violet,
methyl violet), which usually show no marked affinity for the
vacuoles. Besides, as has been seen, staining of the vacuoles is
essentially a vital phenomenon and ceases when the cells are killed.
On the contrary, the chondriosomes stain only temporarily; their
coloration is stable only in dying cells and is then always accom-
panied by vesiculation, a state soon followed by the death of the
cells. The coloration persists, even after the death of the cell.
Meristematic cells may be found in which chondriosomes are vis-
ible in the living state and the independence of these from the
vacuoles may be made certain by vital staining, since the chondrio-
somes remain unstained. The same observation may be made for
the Saprolegniaceae in which the chondriosomes are always clearly
visible in the living state in all stages of development of the plant.
We have obtained, furthermore, in these and other fungi (Endo-

Guilliermond - Atkinson

158

Cytoplasm

tv ■;.•

myces Magnusii) a double vital staining of the chondriome and
the vacuolar system by means of using a mixture of a solution of
neutral red and Janus green or Dahlia violet. In this way, we were
able to follow the simultaneous development of the two systems
during the entire growth of the plant. The chondriosomes by this
method are colored green or blue and the vacuoles red. (Fig. 104) .
These initial forms of the vacuoles which look like chondrio-
somes, are very fragile, just as the chondriosomes are, and they
swell and fuse into larger spherical vacuoles during prolonged
observations with vital staining, but this alteration of shape has
nothing in common with the cavulation of the chondriosomes (c/.
p. 101). Solutions of osmic acid preserve and heavily blacken

the chondriosome-shaped vacuoles whenever they
contain phenolic compounds; otherwise it may
preserve them in a greatly swollen condition
without blackening them. It is known that the
chondriosomes are, on the contrary, very well
preserved with osmic acid but are not darkened
by it.

In preparations treated by mitochondrial tech-
niques, the chondriosome-shaped vacuoles usually
appear as uncolored canaliculi, sometimes anasto-
mosing in a network in the midst of the faintly
colored cytoplasm. Sometimes the contents of
the vacuole are colored, but this is rare. When
it does occur, however, they are always found
condensed by the action of the fixatives in the
middle of the colorless canaliculi so that it is not
possible to confuse the elements of the vacuolar
system with the chondriosomes. Sometimes also
the colloidal contents of large liquid vacuoles, de-
rived from the chondriosome-shaped vacuoles by
hydration, show bodies precipitated by the action
of the fixatives which stain with mitochondrial
technique. (Figs. 105, 106).
In all cases in which the vacuoles contain tannins, the chon-
driosome-shaped vacuoles appear as filaments, or as a network,
somewhat dilated by fixatives and colored yellow by potassium
bichromate, or blackened by osmic acid, according to the method
employed. The large liquid vacuoles resulting from their fusion
show, with the same methods, either granular precipitations or
large corpuscles stained yellow by potassium bichromate or black-
ened by osmic acid. But there is no method, in most cases, which
makes it possible to stain the colloidal contents of the vacuoles
after fixation, when they do not contain tannins, unless, perhaps,
the Golgi methods which will be discussed later (Fig. 107 and
108).

In most fungi and some algae, however, we have seen that the
colloidal substance of the vacuoles, or the metachromatin, made
insoluble and precipitated in the form of corpuscles by formalin

Fig. 102. — Cera-
tium strictum. N,
nucleus; V, vesicle;
Vac, vacuoles. (After
Dangeard).

Chapter XIV

— 159

The Vacuolar System

or alcohol, is stained deeply red by aniline blue or violet basic
dyes, as well as by haematein. It consequently shows a whole
series of histochemical reactions which are very well known and
are very different from those of the chondriosomes.

It is, therefore, well established that there does not exist the
slightest relation between the chondriome and the vacuolar sys-
tem; they are two independent systems which are coexistent in
the cell. There is between the young stages of the vacuoles and
the chondriosomes merely a coincidence of form. This close re-
semblance exists, however, only in a very limited phase in the
development of the vacuoles and seems to be explained by the fact
that the vacuoles are at that moment in a semi-fluid physical state
which is closely allied to the physical state of the chondriosomes.

Physical characteristics of the
vacuoles:- The chondriosome-
shaped vacuoles seem to be of semi-
fluid consistency and in a physical
state which approaches that of
the chondriosomes. Ultramicro-
scopic examination shows that in
general the chondriosome-shaped
vacuoles are no more visible than
the chondriosomes (Fig. 109). In
certain cases, however (teeth of
young rose leaflets, barley and
wheat roots), it is possible to dis-
tinguish them. They appear optic-
ally empty and are visible only be-
cause of their faintly luminous
contours, i.e., when they can be
seen, they show the same charac-
teristics in this respect as do the
chondriosomes and the plastids.

Fig. 103. — a, Cladophora; vital staining
of large central and small peripheral
vacuoles. 6, c, Ulothrix pseudo-ftocca;
arrangement of vacuoles in b vegetative
cells and in c zoosporangium. d, e.
Cladophora; zoospores which have ceased
swarming. /, Bryopsis plumosa; zoospore.
g, Ulva Lactuca; zoospore which has just
come to rest, at, stigma. (After Dan-

GEARD).

With the chondriosomes and the

plastids, they might be classified as a coacervate system.

The liquid vacuoles, derived by swelling from these chondrio-
some-shaped vacuoles, also appear optically empty even in cases
in which they enclose an abundance of colloidal substances (tan-
nin, metachromatin). Rather infrequently they are visible by
reason of their faintly luminous contours. Certain indirect meth-
ods, however, often make it possible to locate their position. Thus
in the yeasts, there exist in the cytoplasm bordering on the vacu-
oles, numerous lipide droplets which appear very luminous and
because of them the position of the vacuoles may be easily detected
(Fig. 110). There sometimes appear within the vacuoles strongly
lighted granules which show Brownian movement, but these are al-
ways visible in direct lighting and are not of the order of micelles.
The vacuole, then, in its liquid state seems to be composed either
of a colloidal solution whose micelles are very small and not

Guilliermond - Atkinson

160

Cytoplasm

visible, or of a very fluid hydrogel. Nevertheless this solution is
very unstable and easily precipitable, as we have seen in studying
the action of vital dyes on the vacuoles. Sometimes, however, the
large vacuoles of mature cells remain, as has been said, in the state
of a very concentrated colloidal solution, a sort of jelly. In this

case, they are generally visible in
the ultramicroscope because of
their luminous contours.

The filamentous vacuoles seem,
like the chondriosomes, to have a
specific weight rather like that of
the cytoplasm, although often it
is higher. For example, aker-
MAN found by use of the centri-
fuge that the filamentous vacuoles
existing under certain conditions
in the tentacles of Drosera are
heavier than the cytoplasm. H.
Clement, by the same process,
was not able to displace the chon-
driosome-shaped vacuoles contain-
ing anthocyanin in the teeth of
young rose leaflets. In recent work,
MiLOViDOV has shown that, in
the youngest cells of the teeth, the
chondriosome-shaped vacuoles con-
taining concentrated solutions of
tannin and anthocyanin become
oriented in the direction of the cen-
trifugal force and are therefore
heavier than cytoplasm. The large
vacuoles derived from them, and
containing a more dilute solution
of tannin, become oriented in the
opposite direction, i.e., centripetal-
ly, and are lighter than the cyto-
plasm. MiLOViDOV obtained simi-
lar results with barley roots. In
cells of the youngest part of the
meristem, the vacuoles are heavier
than the cytoplasm and are easily
displaced in the direction of cen-
trifugal force. In regions situated just a little above, the vacu-
oles which are beginning to take in water have about the same
density as the cytoplasm and are no longer displaced. In the region
in which the cells are already differentiated, the vacuoles are much
lighter than the cytoplasm and are displaced in a centripetal
direction. (Fig. 111).

Thus the small vacuoles which look like the chondriosomes are
semi-fluid or sometimes almost solid elements. They seem to be

fci?

"# #■ 6

Fig. 104.

Filaments and oidia of
Endomyces Magnusii. Double vital staining
with Dahlia violet, which stains the chon-
driosomes (c), and with neutral red, which
colors the young vacuoles (1, V), or in
older vacuoles {V) causes precipitation of
their metachromatic substances as deeply
colored bodies. Gl, lipide granules.

Chapter XIV — 161 — The Vacuolar System

composed of a jelly or of a coacervate. One is therefore led to
believe that the development of the vacuoles described in the pre-
ceding pages may possibly be nothing more than an unlimited
imbibition of small elements, shaped like granules or filaments,
which are in a jellied or coacervate state. This imbibition would
involve the transformation of the original gel into a very dilute
solution represented by the liquid vacuoles.

Chemical nature of the colloidal substance of vacuoles:- The
histochemical characteristics of young vacuoles, which we have
enumerated to establish a distinction between the chondriosome-
shaped vacuoles and the chondriosomes themselves, prove that the

M

Fig. 105. — Ricinus root. Meristem fixed by
Regaud's method. Black, filamentous or granular
chondriosomes (Ch) and vacuoles (V) , the latter
distinguished by an external hyaline region due
to a contraction of their colloidal contents during
fixation.

colloidal substances of the vacuolar sap have an essentially variable
constitution. Vacuoles are present in all plants but there does
not exist any substance characteristic of them, as is the case for
the chondriosomes. They contain diverse substances having noth-
ing in common except their property of fixing vital stains.

Vital staining alone reveals differences in coloration between
vacuoles. Cresyl blue, for example, changes color in vacuoles
which contain metachromatin (majority of fungi, certain algae).
It stains them a diffuse red and the enclosed corpuscles are colored
dark red. In the cells of phanerogams, the staining of the vacu-
oles is extremely variable. Whenever the vacuoles contain phe-
nolic compounds, they take a pure blue color with cresyl blue be-
cause of their acid pH. This same blue color is sometimes observed
when the vacuoles contain lipide substances (phytosterol or phos-

Guilliermond - Atkinson

— 162

Cytoplasm

pholipides reported by Reilhes) . In other cases the vacuoles take
on colors toward the violet.

Dangeard used the characteristic shown by blue vital dyes, of
staining red or violet those vacuoles which contain metachromatin,
as the only basis for his theory of the universal presence of meta-
chromatin in vacuoles. But this staining reaction (Fr. metachro-
masie) is not a characteristic belonging alone to the substance called
metachromatin. Metachromatin has been so named because it

Fig. 106. — Root of pea. A, meristem; chondrio-
somes (in) stained, vacuoles (P.V.) colorless. B, C,
meristem of central cylinder; contents of filamentous
vacuoles (P.V.) clearly distinguished from the chon-
driosomes by a peripheral hyaline region caused by
contraction of the contents of the vacuole by the
fixative. D, differentiated cortical parenchyma;
single large vacuole with densely stained bodies
(P.V.). m, chondriosomes. V, vacuole. Regaud's
method.

changes all blue and violet aniline dyes and haematein to a red color
after fixation. Now this still unexplained phenomenon has no rela-
tion to the color change obtained with vital staining. One can not,
as do the Dangeards, relate the colloidal contents of all vacuoles to
a single substance and call it metachromatin, since metachromatin
is a substance present in fungi, so named by reason of a color
change which it shows after fixation, not when vitally stained. Be-
sides, it has been seen that metachromatin is preserved by alcohol
and formol, whereas the colloidal substance of the vacuoles is usu-
ally destroyed by all fixatives, even those of mitochondrial tech-
niques.

If the blue coloration which the vacuoles sometimes take with
cresyl blue is often an index of an acid reaction, as is the case for

Chapter XIV

— 163 —

The Vacuolar System

the vacuoles containing phenolic compounds, the red or violet
coloration which they take in other cases is difficult to explain, for
it is known that cresyl blue in solution in pure water shows only
a single change in color : it takes on an orange tint for a pR value
of 11.2. Nevertheless, according to Mangenot and Mile Laurent,
cresyl blue at a pB. which is not accurately measured, shows a
change to violet, on condition that the medium contain diverse
colloidal substances (sodium silicate, dextrine, protein, etc.) or
even sugar (saccharose, according to unpublished work). This
fact has been confirmed by Chadefaud.

VCM 9''-^

immk

I

! 4 M

■■■ ¥1

.Ch

1

mm

2

HmfL

Fig. 107 (left). — Dematium. Bouin's method, stained with hemalum. 1, filament. 2,
germinating conidium. VCM, vacuole containing metachromatin precipitates, n, nucleus.

Fig. 108 (right). — Saprolegnia. Chondriome and vacuolar system. 1-3, Vital staining
with neutral red. 1, tip of filament; reticular vacuole (RV) ; other elements omitted. 2, older
filament; tendency of network to become a diffusely stained canal (V) containing deeply
stained bodies (CM), other elements visible but unstained. 3, still older filament; vacuolar
canal (V) loses its stain, other bodies as in (2). 4, mature filaments. Regaud's method
with iron haematoxylin; vacuolar canal unstained, lipide granules dissolved, chondriosomes
(Ch) strongly stained. 5, 6, mature filament. Meves' method with acid fuchsin; chondriosomes
(Ch) red, lipide granules (Gg) brown. Ch, chondriosomes. N, nucleus.

It must be added that the change to red in the vacuoles may
also depend on the chemical constitution of colloidal substances
which the vacuoles hold in solution. Mucilaginous substances, agar,
for instance, change cresyl blue to a color toward the red, no mat-
ter what the pH. According to LisoN, this color change in vital
staining is a histochemical reaction characteristic of all sulfuric
esters of high molecular weight. Now, mucilages or polyholoside
esters are the most important among them.

From these considerations it follows, therefore, that the change
toward red of the vacuoles, which is due to the most variable
causes, and which may be obtained in vitro by the addition of the
most diverse substances, can not possibly serve to characterize a
chemical substance.

In general, the vacuolar colloids appear in the higher plants
to be protein substances, perhaps proteins soluble in alcohol, and

Guillierrnond - Atkinson

164

Cytoplasm

are often associated or combined with tannins and perhaps with
mucilages. Lloyd and various other authors have shown that
tannin is often combined with mucilages in the state of a com-
plex, and it is thought that the vacuoles containing raphides en-
close mucilages. Furthermore, recent work has shown that in
certain cells the vacuoles contain a colloidal solution of phytosterol
or of phosphoaminolipides. These substances may become par-
tially solid in the epidermal cells of the Liliaceae (Fig. 112) . These
cells ordinarily contain a large inclusion, formed of a complex of
phytosterol and of phosphoaminolipides, which Mirande described
for the first time under the name of sterinoplast. This, he considered

Fig, 109. — Rose. Cells from a tooth of a leaflet under the
ultramicroscope, showing faintly luminous contours of the vacu-
oles (V) and of the nuclei (N) . 1, tip of tooth. 2-4,
differentiated cells.

to be a cytoplasmic inclusion, a sort of plastid, elaborating phy-
tosterol. The work of Miraton and of Emberger has demon-
strated that the sterinoplasts are not simple vacuolar concretions.
The recent work of Reilhes has established the fact that the vacu-
olar sap of these cells contains a solution of phosphoaminolipides
and of phytosterol which, in mature cells, becomes partially solidi-
fied in the vacuoles in the form of large bodies composed of a phos-
phoaminolipide-phytosterol complex. Inclusions, apparently of the
same nature, have been cited in the vacuoles by other authors : in
the epidermis of Iris, especially, in which they absorb anthocyanin,
when the vacuoles contain the pigment, and have been described
under the misnomer of cyanoplasts (Politis). Similar inclusions
have also been described in the epidermis of the flowers of Del-
phinium cultorum (SCHARINGER). We have shown as well that
in the cells of the root cap of barley and of wheat, there are
phosphoaminolipide solidifications and, by cultivating barley roots
in media very rich in sugar, Gautheret succeeded in making a great
number of lipide concretions appear in the vacuoles of most of the
cells. The vacuoles of Monotropa according to Weber contain large

Chapter XIV

— 165

The Vacuolar System

quantities of lipides. Furthermore, Buvat reported in the vacu-
oles of many roots, the existence of phosphoaminolipide concretions
sometimes combined with proteins.

In the algae, the colloidal substances of the vacuoles are also
very diverse. Proteins, tannins and mucilages seem to exist in
them but metachromatin and volutin are also frequently found.
This last substance, which is also found in the bacteria, especially
characterizes the fungi, in which, with the exception of some of

4 5

Fig. 110. — Saccharomycodes. (A), under the
ultramicroscope; (B), with direct light. A, 1-6 S.
Ludivigii. Vacuoles (V) usually invisible; position
marked by strongly lighted lipide granules (Gl;
A3, B5) surrounding them and corpuscles of
metachromatin (C; A2, B2) which they contain.
4, contour of vacuole visible. 6, cytoplasm coagu-
lated. 7-9, S. pastorianus. B, S. Ludwigii.

the Phycomycetes (Saprolegniaceae and Peronosporaceae) , its
presence is of general occurrence and in which it appears to play
the role of reserve product. In fact this substance accumulates in
the vacuoles of the epiplasm of the asci of yeasts and of the higher
Ascomycetes and is absorbed by the ascospores at the same time as
the glycogen and the lipides which are coexistent with it in the
epiplasm. Volutin offers histochemical characteristics which as
we have seen make recognition of it easy. Its chemical constitu-
tion is still not well determined but there are good reasons for
supposing that it is formed by a combination with nucleic acid
(Arthur Meyer).

From very recent work of Mile. Delaporte and Mile. Roukel-
MAN on yeasts, it seems that metachromatin corresponds to a
zymonucleic acid compound which has been extracted in great
quantities in yeasts. According to these investigators, the nuclei
of these fungi, as well as those of other plants probably, are com-

Guilliermond - Atkinson — 166 —

Cytoplasm

posed of thymonucleic acid, like those of animal cells, which ex-
plains why the Feulgen reaction is obtained just as well in plant,
as in animal, nuclei. In the Saprolegniaceae, finally, we have
found sphaerocrystals which appear to have some relation to the
phosphoaminolipides.

The vacuoles contain, as well as the colloidal material just
discussed, numerous crystalloid substances. The most wide-spread
of these are organic acids, halogen salts, nitrates and phosphates,
sugars (saccharose, glucose, fructose, etc.), heterosides and pig-
ments. Among the halogen salts we must mention the presence of

iodide in a dissolved state in the vacuo-
lar sap of numerous marine algae (Rho-
dophyceae and Phaeophyceae) . Its local-
ization in the vacuole may be demon-
strated in vivo by cresyl blue. This pig-
ment in the presence of iodized solu-
tions forms red crystals (Fig. 113) ar-
ranged in the shape of a bouquet (Sau-
VAGEAU and Mangenot). Among the
pigments may be mentioned the oxy-
flavanol pigments which are very pale
yellow in color and the anthocyanin pig-
ments which are red, violet, or blue.
Both types are extremely widespread
in green plants. Anthocyanin pigments,
when found in high concentration in
the vacuole, may crystallize there in
the form of needle-shaped crystals or
sphaerocrystals (Fig. 114). These two
types of pigment show histochemical
characteristics closely allied to tannins
(darkening with ferric salts, blackening
with osmic acid). The flavins may also
be cited. These are yellow pigments
playing at the same time the role of
hydrogen carriers and the role of Vitamin Ba. We have recently
found them in great abundance in the vacuoles of a fungus Eremo-
thecium Ashbyii where they crystallize in the form of needles or
sphaerocrystals. The alkaloids are very wide-spread in the vacuoles
of the phanerogams and may be recognized by testing with iodine-
potassium iodide (Bouchardat's reaction), which precipitates the
alkaloids in the vacuoles as brown granules.

In bringing this inventory to a close, there must be mentioned
asparagine crystals and leucine crystals (lozenge-shaped or sphae-
rocrystals) and especially calcium oxalate crystals. These latter
are found in the vacuoles of a great number of phanerogams, some-
times as long needles (calcium oxalate monohydrate or acid cal-
cium oxalate) belonging to the monoclinic series; sometimes in the
state of octahedral crystals, isolated or twinned — quadratoctahedra,
in the nature of crossed twins (calcium oxalate trihydrate) — be-

FiG. 111. — Diagrams of barley
root vitally stained with neutral
red and centrifuged. The chondrio-
some-shaped vacuoles of the meri-
stem, heavier than cytoplasm, are
displaced centrifugally; those tak-
ing in water in the region of
differentiation are not affected:
those still higher in the root, lighter
than cytoplasm, are displaced cen-
tripetally. (After Milovidov) .

Chapter XIV

167

The Vacuolar System

longing to the tetragonal system; sometimes as crystal dust. The
crystals of calcium sulphate are also very frequently found (Clos-
terium, Spirogyra) .

Fig. 112. — Lilium candidum. Lipide concre-
tions within the vacuoles. A, in epidermal cells
of the bulb; small granules (1) agglomerate into
mulberry-shaped masses (2), these, B. in older
scales become large bodies which are globular with
a denser center (4) or of radiating crystalline
structure (3). C, in drying outer scales, these
break up and myelin figures (5) appear which
finally, in dead cells, form masses of wound up
threads (6).

Vacuolar pH and rH:- It is very difficult to obtain an exact idea
of the vacuolar pH. It has been seen indeed that the change of
color of the vital dyes has only a very questionable value. Then,
too, the indicators of pH do not penetrate into the vacuoles and are
besides very toxic. However, evaluations have been attempted by
Crozier, Haas and Irwin, who sought to extract the vacuolar sap
of certain algae, such as Valonia, in which the articulations are oc-
cupied by an enormous vacuole. These evaluations have shown
that the vacuolar ^^H in Valonia has an acid reaction (5.0-6.7).
Crozier, Haas and Schmidt have used an interesting method by
making the anthocyanin pigments serve as indicators. These pig-
ments can be extracted from the petals and in mixing them with
buffer solutions of known 2?H, all their different possible colors may
be obtained and consequently the coloring which the petals show
naturally may be related to a known pH. The results obtained by
this method inform us as to the vacuolar pH. It scales from 3.0-8.0.

We have been able with Gautheret to find a way to evaluate
the vacuolar pH but only qualitatively. Clark and Perkins have

Guilliermond - Atkinson

— 168

Cytoplasm

shown that neutral red reduced in an alkaline medium (pB. 8.2 for
example) gives a leucoderivative which, by acidification of the
medium (pR 5.2 for example), is transformed into a second deriva-
tive, yellow in color and fluorescent, which is distinct from the

Fig. 113. — Cells of stipe of
Laminaria flexicaulis. Bouquet
crystals formed by the pre-
cipitation of iodide by cresyl
blue in the vacuoles. P,
phaeoplast. lo, fat globule.
(After Mangenot).

Fig. 114. — Epidermal cells of sepals.
Pigment in the vacuoles, partly in
solution, partly crystallized. 1-2 Del-
phinium. 1, D. Ajacis; pigment blue,
clusters of needle-shaped crystals. 2.
hort. var. with dark blue flowers; long,
entangled needle-shaped crystals. 3,
Verbena hybrid; large sphaerocrystals.

Fig. 115. — Mycelium of Eremothecium
Ashbyii. a, vacuole with flavin in solution;
b, vacuole with flavin crystals; gl, lipide
granules.

first. This latter is also obtained when neutral red is reduced in
an acid medium. Finally, although oxidation of the leucoderiva-
tive is rapid in contact with air, that of the fluorescent form is
very slow. Now, in treating wheat roots, or the epidermis of Iris
rich in phenolic compounds, with the leucoderivative of neutral
red, we have found that the vacuoles take a yellow coloration
which, in contact with the air, reddens only slowly. On the con-
trary, the cells of the bean root which are lacking in phenolic com-

Chapter XIV

— 169 —

The Vacuolar System

pounds remain uncolored with the leucoderivative, and the leuco-
derivative which they have absorbed becomes oxidized instantane-
ously m contact with air. It may therefore be believed that the
vacuoles of wheat and Iris are acid whereas those of the bean are
not.

Vacuolar rH has been evaluated by means of vital stains (Ma-
TIUJA Brooks) or by micro-injection of indicators in algae such as
Spirogyra. These measurements have given an rH of 16-18.

Fig. 116. — A, Petiole of Begonia; octahedral crystals,
isolated (lower cell), or twinned (upper left), in the
shape of a sea urchin (upper right). B, Leaf of Aloe
succotrina; raphides.

Chapter XV

ORIGIN AND SIGNIFICANCE OF THE VACUOLES

Aleurone grains : their formation. — Aleurone grains are found
in every seed. Their significance has been discussed for a long
time. These grains are to be found in large numbers in all cells of
the embryo and are especially voluminous in the cells of the coty-
ledons or of the endosperm. In the Gramineae they are localized
in the protein layer of the endosperm.

Aleurone grains vary in structure. In many cases (RiciniLS,
Cucurbita Pepo) they are composed of an amorphous protein mass
enclosing a crystalloid of the same nature and one or several
spherical bodies called globoids, formed of phytin. Sometimes
twinned crystals of calcium oxalate are found in the protein mass
or in the globoids. In other cases the aleurone grains do not con-
tain crystalloids and enclose in their protein mass only numerous
globoids (Gramineae). Lastly, there are cases in which the aleurone
grains are composed entirely of amorphous protein (Legumes).

Two opinions have been
formulated as to the
origin of aleurone grains.
Some authors considered
them to be plastids in
which were formed pro-
tein and the globoids.
Others contended that
they resulted from sol-
idification of the vacu-
oles during dehydration
of the seed. This latter opinion is now demonstrated to be true.
If, during the maturation of the seed, the vacuoles from any
part of the embryo or endosperm are examined using vital stains,
liquid vacuoles are found which are more or less large, according
to the type of cell and its state of development. These vacuoles
contain protein substances in solution and neutral red causes a
precipitation of these proteins within the vacuoles as deeply stained
bodies. In the course of development of the seed, the vacuoles be-
come much richer in protein. In the stages immediately preceding
maturation, i.e., at the time when dehydration takes place in the
seed, it is observed that the vacuoles sometimes have a tendency
to break up as they lose water, becoming smaller and smaller, and
less and less fluid. They stain deeply and homogeneously and are
now semi-fluid. They often at this moment become filamentous or
reticulate, analogous in form to those observed in meristems.
Later when the seed has reached maturity and passes into the
dormant stage, the semi-fluid vacuoles by more marked dehydra-

FiG. 117. — Aleurone grains in seeds, a, h, Ricinus:
one crystalloid and 1-2 globoides. c, Oenanthe Phel-
landriuTn; sphaerocrystal of calcium oxalate in protein
ground substance.

Chapter XV

-171

Origin of Vacuoles

tion take on the appearance of solid, spherical, very refractive
bodies. These bodies can, by crushing, be expelled from the cell
and then will not take up neutral red unless they have been im-
mersed in water for a long time.

The vacuoles are thus transformed into bodies of protein nature
whose dimensions are variable. They are very small in some cells
and very large in others. They are aleurone grains and are formed
by the solidification of the protein contained in solution in the
vacuole. As a result of losing water, a grain of protein has there-
fore been substituted for a vacuole. It is not astonishing, then, to

Fig. 118. — Ricinus. Endosperm. 1, young cell, starch (A)
forms in chondriconts. 2, just before maturation; starch is
absorbed, the large vacuole is fragmented into small vacuoles con-
taining protein crystals (C) and protein precipitates (P). 3, oil
globules (H) blackened by osmic acid, protein crystals (P). 4,
detail of (3). 5, dormant seed; A, aleurone grain, G, globoid.
6, during germination of the seed; aleurone grains (P) beginning
to dissolve. M, chondriosomes. 1, 2, 5, 6, Regaud's method. 3, 4,
Meves* method.

find along with the protein in these vacuoles, inclusions of phytin
and crystals of calcium oxalate, products encountered in many
vacuoles. Some aleurone grains may even contain oxyflavanol pig-
ments or anthocyanin. Investigations of Speiss and Chaze have
shown that aleurone grains of some varieties of maize contain an
anthocyanin pigment which gives them their characteristic black
coloration. At first red, this pigment appears in the vacuoles which
will later be transformed into aleurone grains, then remains ab-
sorbed by the protein of which the aleurone grain is composed. At
germination, when the aleurone grains are again transformed into
vacuoles, the anthocyanin pigment changes to red. Furthermore
Chaze has found oxyflavanol pigments in the yellow or white ker-

Guilliermond - Atkinson

172

Cytoplasm

1

nels of other varieties of maize as well as in the seeds of other
grains (barley, wheat, rye, oats). Aleurone grains stain deeply
with mitochondrial techniques which do not at all stain the liquid
vacuoles from which they are derived, except in periods preceding
solidification when the stains bring about flocculation of the col-
loidal solution, and the formation in the vacuoles of precipitates
stainable with iron haematoxylin.

At the time of germination, when the seed again takes up water,
the aleurone grains absorb it and become semi-fluid. At this period
they often show again a tendency to elongate into filaments capable
of anastomosing in a network.

At the beginning of hydration, the aleurone grains stain deeply

and homogeneously with vital dyes. Then
the filamentous and semi-fluid, reticulate
vacuoles derived from them swell as water
continues to be taken in and appear as
spherical liquid vacuoles which by coales-
cence gradually become transformed into
large vacuoles in which the vital dyes cause
strongly colored precipitates.

In sections prepared with mitochondrial
techniques, the forms which were reticulate
and filamentous, at the beginning of ger-
mination, show a heavily stained, compact
contour which is filamentous, or granular,
and which is surrounded by a clear zone.
^^ W I '^^^ spherical liquid vacuoles, which succeed

•15^ I * I them, still enclose stained corpuscles but

these corpuscles become less and less abun-
dant, as more and more water is taken into
the vacuoles and, finally, in the large vacu-
oles, no stained contents are found.

':-\.

•A

Fig. 119. — Tulip. Epi-
dermal cells of petals of a
dark red variety. Frag-
mentation of large vacuole,
containing anthocyanin, into
small, sometimes reticulate,
or filamentous, vacuoles
caused by plasmolysis in a
5% solution of NaCl.

Reversibility of form in the vacuolar sys-
tem. — Pierre Dangeard's investigations,
and our own, describing this development
show, therefore, that there exists a certain reversibility between
the two vacuolar forms. It has previously been shown that vacuoles
ordinarily appear under very diflferent aspects according to the age
of the cells :

1. As numerous, minute, semi-fluid vacuoles which have more
or less the shape of the chondriosomes.

2. As a small number of large, spherical, liquid vacuoles,
always containing colloidal substances, but in very dilute solutions,
i.e., vacuoles corresponding to the classical definition of vacuoles
and capable of fusing into a single enormous element. The first of
these is found in embryonic cells, the second in mature cells. If
the cells are greatly dehydrated, the spherical liquid vacuoles may
lose their water, become concentrated, and may again take on a
semi-fluid consistency and look like chondriosomes. A more com-

Chapter XV

173 —

Origin of Vacuoles

plete dehydration leads finally to the transformation of these semi-
fluid vacuoles into solid bodies (aleurone grains) by a solidification
of their colloids. The aleurone grains, by taking up water anew,
are capable of again assuming the semi-fluid consistency and ap-
pearance of chondriosomes and, by a continuance of this process,
may finally become liquid vacuoles.

The filamentous appearance of the vacuoles seems to be the
result of their semi-fluid state, for in the semi-fluid state, the
vacuoles are generally filamentous or reticulate and stain uniformly
and deeply with vital dyes. In the liquid state, the vacuoles are
generally spherical and are stained only weakly with the vital dyes,
which bring about a precipitation of the enclosed colloids as deeply
stained granulations showing Brownian movement. The vacuoles
are then composed of drops of a very dilute colloidal solution. In

Fig. 120. — Saprolegnia. Modifications in form of vacu-
olar system in a single branch, cultivated on 1% peptone
bouillon with 0.001% neutral red, in a van Tieghem and
Le Monnier cell. 1-6, spherical vacuoles fuse to form a
single canal which then, 7, 8, is transformed into a net-
work (After Mile. Cassaigne).

the solid state, they appear as globular bodies, which do not stain
with vital dyes unless imbibition has previously taken place, but,
on the contrary, always stain after fixation. The vacuoles, there-
fore, may pass from one form to the other depending upon the
water content of the cell.

This reversibility has been obtained experimentally, further-
more, in various cells, among others, in the epidermal cells of peri-
anth parts of red tulips (Fig. 119). In the open flower, these cells
contain a large vacuole, occupying almost the entire volume of the
cell, and containing a concentrated solution of red anthocyanin pig-
ment. Now, by plasmolyzing mature epidermal cells with a strongly
hypertonic solution, we have observed that these vacuoles, as they
lose water, may break up into small vacuoles which become semi-
fluid and appear granular, filamentous and reticulate. Similar re-
sults have been obtained in the Saprolegniaceae (Guilliermond)
and in the epidermal scales of Allium Cepa (Kuster) .

Guilliermond - Atkinson — 174 — Cytoplasm

If one observes more closely the changes in appearance of the
vacuoles brought about by hydration and fusion of small vacuoles
into one large one, and, inversely, if one observes the fragmentation
into very numerous, small, semi-fluid, filamentous or reticulate vacu-
oles — a sort of splitting up of the large vacuoles — one is led to admit
that in these phenomena the vacuoles play only a passive role.
Their contraction and division are brought about by the degree of
water taken into the cytoplasm which causes movements of the
latter, the consequences being felt by the vacuoles. The cytoplasm,
under some influences, may extract a part of the water contained
in the vacuole and may swell. This swelling is therefore produced
by movements of the cytoplasm, particularly by emission into the
vacuoles of lamellate prolongations which finally partition off the
vacuole into multiple vacuoles. These, in losing their water, be-
come very viscous and look like chondriosomes. In the reverse
process, the cytoplasm is capable of restoring to the vacuoles a
part of their water of imbibition, bringing about a hydration and
increase in volume of the latter. These then fuse into a single
large vacuole. It is phenomena of this type which must take place
at the beginning of the maturation of the seed and which must
take place during the process of plasmolysis; some of the water
from the vacuole passes into the cytoplasm and must cause a swell-
ing, to which may be attributed the fragmentation of the vacuoles,
and it is not until later that the cytoplasm in turn gives up its
water to the exterior. In germination, the contrary phenomena
must take place. The vacuole absorbs the water at first accumu-
lated in the cytoplasm, and, as the latter continues to dehydrate,
the vacuoles progressively swell and again fuse to form a very
large vacuole. This view of the matter seems, moreover, to be
confirmed by the fact that the viscosity of the cytoplasm increases
as the plant grows old, correlative with the development of the
vacuole which ends by occupying almost the entire cell.

This reversibility of vacuolar form is to be compared with a
remarkable phenomenon in the tentacles of the leaves of Drosera
rotundifolia, described long ago by Charles Darwin, and desig-
nated by him as aggregation. While studying the modifications
which occur in the pedicel of the tentacles as a result of the stimulus
produced by an insect, Darwin saw in each cell that the cytoplasm,
which was colored red by pigment before stimulation, broke up
before long into an aggregate of deeply stained corpuscles appear-
ing as granules, clubs, rods or filaments showing amoeboid move-
ments. The study of this phenomenon, taken up by various authors
(Gardiner, de Vries, Goebel, and akerman) has shown that in
reality the phenomenon observed by Darwin consists of a multiple
fragmentation of the vacuole and not of the cytoplasm. The cells
of the tentacle contain a single, very large vacuole filled with antho-
cyanin. At the moment of stimulation this vacuole undergoes a
great fragmentation. It splits into a large number of small chon-
driosome-shaped vacuoles. Immediately after stimulation, these
minute vacuoles fuse to constitute again a single large vacuole —

Oiapter XV

175

Origin of Vacuoles

the cell returns to its initial state. This is therefore a phenomenon
entirely comparable to that observed during the formation of aleu-
rone grains and during the process of plasmolysis.

akerman's work has shown that this phenomenon consists of
a modification in volume of the vacuoles as a consequence of a
great deal of water being taken in by the cytoplasm. The result
is a fragmentation of the vacuole, caused by swelling of the cyto-
plasm, and at the same time an increase in osmotic pressure. This
pressure increases by 5 atmospheres. Centrifuging revealed, as
has been seen, that, in the stimulated cell, the vacuoles are more
dense than the cytoplasm and, conversely, in the unstimulated cell,
the cytoplasm is more dense than the vacuole.

The study of these phenomena has been taken up recently by
DuFRENOY, Homes, and Kedrowsky working on Drosera, by QuiN-

FlG. 121. — Drosera rotundifolia. Glandular cells in the tentacles.
Vacuolar system made visible by red pigment in cells, may change from
large to small, spherical, filamentous or reticulate vacuoles. A, after
DE Vries. B, after HoMes.

TANILHA and Mangenot in Drosophyllum lusitanicum. These in-
vestigations show that it is necessary to differentiate two sorts of
physiologically distinct cells in the tentacles of Drosophyllum and
of Drosera. First, there are those which, covering the upper sur-
face of the tentacles, are secretory. They excrete the complex and
viscous liquid which forms a drop at the tip of each tentacle. These
cells, when they are not going through a period of digestion, possess
a group of small filamentous or granular vacuoles, an arrangement
probably having some relation to the loss of water which takes
place in the cell (Quintanilha, Dufrenoy, Homes, Mangenot,
Kedrowsky). Secondly, there are the cells which compose the
lower part of the head and the pedicel. Each of these cells con-
tains, except during the periods of digestion, a large fluid vacuole,
colored red by anthocyanin. During digestion, this enormous vacuole
becomes very finely divided into a large number of small, globular
or filamentous elements, colored violet-grey (Dufrenoy, Man-
genot). These small filamentous vacuoles are oriented parallel to

Guilliermond - Atkinson

176 —

Cytoplasm

the longitudinal axis, while the small vacuoles are accumulated
against the proximal wall of the cell (Mangenot).

This polarity of arrangement, which becomes more marked as
digestion becomes more active, clearly indicates that these cells
are traversed by a continuous flow of material caused by protein
digestion (proteolesis) at the level of the extremities of the

tentacles. Thus, the "aggregated" state
of the vacuoles, i.e., their irregular dis-
position, seems to correspond to differ-
ing physiological conditions — to a secre-
tion in the cells which cover the ex-
tremity of the tentacle, to an absorp-
tion in the cells of the stalk — but both
these processes indicate the passage of
a current across the cells. This same
arrangement of vacuoles is found in
young sieve tubes in the angiosperms in
which the fragmented and polarized
vacuoles are very polymorphic. In these
cells, large vacuoles and small filamen-
tous or reticulate vacuoles are found.
They are also found in the conducting
elements of the Laminariales and the
Rhodophyceae (Mangenot) (Fig. 123).
Recent work seems to suggest that
the vacuoles undergo a similar frag-
mentation each time that the cells are
in the process of active secretion (Man-
genot, Mile. Py, Thomas, Guillier-
mond). The reason for this is not yet
known.

Another phenomenon to consider
here is that of the frequent changes in
form of vacuoles in many cells. The
observation of a species of Saprolegnia
in van Tieghem and Le Monnier cells,
grown in a nutrient solution to which
neutral red has been added, made it
possible to record this phenomenon
under excellent conditions (Fig. 120).
In the extremities of growing filaments
the vacuoles generally appear as very
small elements shaped as granules, rods or filaments. These ele-
ments are carried along by cytoplasmic currants which cause them
to change shape constantly. They are capable of swelling and of
contracting, of passing from the shape of granules to that of fila-
ments and conversely. In the space of a few seconds, they are
frequently seen to fuse and form rather large spherical vacuoles
which, themselves, may give rise, by budding, to small vacuoles, or
may be completely split up into a multitude of small elements which

Fig. 122. — Drosophyllum
lusitanicum. Epidermal cells
of the pedicel of a tentacle.
Vacuoles in grey, arrows in-
dicate the direction of the
head of the gland. A, during
digestion of protein. B, dur-
ing inactive period. C, tip
of leaf with tentacles; one
gland showing pedicel, head
and drop of secretion. (Af-
ter Mangenot).

Chapter XV

177 —

Origin of Vacuoles

then elongate and anastomose to form a network. So the trans-
formation from large vacuoles to a network and the converse
transformation are here again observed. It is only a little farther
away from the tip of the filament that all the vacuoles fuse to form
a vacuolar canal (Mile. Cassaigne).

A prolonged observation of growing yeast without the aid of
vital stains makes it possible to see that
the large vacuoles, which appear rather
stable, in these fungi are themselves
subject to changes in shape. After a
period of stability their contours may
suddenly become irregular, angular,
may vary continually, contracting and
dilating and finally come back to a tem-
porarily stable form. Often they are
seen suddenly to put out long and thin
prolongations, which later contract as
the vacuole returns to its spherical
shape.

Ultramicroscopic observation of the
vacuoles of fungi (in cases where it is
possible because of the faintly luminous
outlines of these elements) has shown
also that their contours often manifest
slowly undulating movements. These
phenomena seem to be caused here not
only, (1) by differences of imbibition
between the cytoplasm and the col-
loidal contents of the vacuoles and (2)
by movements of the cytoplasm, but
also by modifications of surface tension.

6i&

Fig. 123. — a, Alaria escu-
lenta (Laminariales) , b, Bon-
nernaisonia asparagoides ( Rho-
dophyceae). Globular vacuoles
at one end of cell; plasmodes-
mic threads connect the cells In
the conducting tissue. (After
Mangenot) .

The phenomenon of vacuolar con- I I

traction:- In addition to these phenom- Jj

ena of vacuolar fragmentation, there
must be cited here another phenomenon
of quite a different nature, namely, a
particular state of the vacuoles which
was first observed by Weber. In the
flowers of the Boraginaceae (Symphy-
tum tuberosum, Anchusa italica and
Mertensia sibirica), this investigator observed that the vacuolar
sap of cells is a jelly, capable, under certain conditions, of contract-
ing "spontaneously" i.e., quite aside from all plasmolysis. This phe-
nomenon has been observed in flowers infiltrated with solutions
containing from 1%-0.1% of neutral red. There is then produced
a contraction of the colloidal contents of the vacuole whose con-
tours are curiously parallel to those of the cell itself. Weber pro-
poses to call this phenomenon vacuolar contraction and interprets
it as a syneresis, a name given by Graham to the spontaneous con-

Guilliermond - Atkinson — 178 —

Cytoplasm

traction of a gel when liquid is expelled. The phenomenon pre-
supposes the existence in the vacuole, around the contracted jelly,
of a liquid phase (serum) produced at the moment of contraction.
This phenomenon of vacuolar contraction seems to be very gen-
eral indeed. There is frequently encountered, in the epidermal cells
of flowers, a colorless space enclosing a contracted intravacuolar
mass and surrounded at the periphery of the cell by a thin cyto-
plasmic layer. Neutral red may be superimposed upon the natural
color of the vacuolar sap, when it is rich in anthocyanin compounds,
and will stain the contracted vacuolar mass intensely. Careful ob-
servation of the colorless space reveals that it is not empty. It is
occupied by a fluid. Plasmolysis of this modified vacuole, moreover,

is quite possible and water is lost from the
peripheral colorless fluid, which appears at
first very poor in dissolved material. Little
by little, over a period of from 24-48 hours,
however, the fluid reddens. The reddening
never reaches the intensity of that of the
contracted mass. It is evidence, however,
of a slow penetration of colloidal material
into the peripheral liquid. Then a second
contraction is frequently produced at this
time. The fluid separates into a new color-
less peripheral serum, and into a deeper red
region, contracted about the mass originally
isolated.

Weber thinks that the same phenom-
enon may explain the presence in many
phanerogam cells of two categories of
vacuoles which are adjacent in the cell, the
one liquid and lacking in tannin, the other
formed of a tannin jelly. These will be
taken up later. Weber underlines also the
potential importance of this phenomenon in
the realization of rapid changes in the
turgidity of cells, perhaps in the mechanism of certain organ move-
ments. He points out in this regard that the motor swellings of
leaves of the Leguminosae (Mimosa), as Mangenot has shown,
have vacuoles containing a tannin jelly, side by side with small
vacuoles which do not contain colloidal substance. Weber com-
pared this vacuolar contraction to the natural production of large
colloidal corpuscles observed in the vacuoles of the mature cells.
This contraction can not be attributed to a phenomenon of syne-
resis, for it consists only in the partial precipitation of the vacuolar
colloid. It seems, on the contrary, to correspond, as we have said,
to the formation of a coacervate within the vacuolar sap.

Fig. 124. — A, B, Sac-
eharomyces ellipsoideus vi-
tally stained with neutral
red, vacuoles in buds seem
to form de novo. C, D,
Oidium lactis, vacuoles in
branches seem to form de
novo.

Origin of vacuoles:- P. A. Dangeard and then P. Dangeard,
as a result of his own work on aleurone grains^ were led to adhere
to the theory of de Vries and Went and to admit that the vacuoles

Chapter XV — 179 — Origin of Vacuoles

are never formed de novo, but always arise by the division of pre-
existing vacuoles, and are transmitted by division from cell to cell.
But the Dangeards believe that it is not the vacuole itself, that is
transmitted, but the metachromatin, which they consider to be the
universal substance of their vacuome. According to them, this
metachromatin persists in a solid state after the disappearance of
the vacuole in the seed, as well as in the spores of fungi, and re-
forms vacuoles anew at germination by taking in water again.
This theory is difficult to admit in view of the fact that we know
that there is no single chemical substance which is characteristic
of vacuoles.

Actually, it is extremely difficult to study the origin of vacuoles
in the phanerogams because of the great number, and the small
size, of these elements in embryonic cells. It is known, moreover,
from what has just been said that vacuoles exist in all cells and
that they are capable of dividing and of fragmenting. It has been
observed besides that during mitosis the vacuoles are distributed
between two daughter cells. It is for this reason that Bailey and
ZiRKLE, without committing themselves on this subject, say that
they have never seen vacuoles form de novo and that nothing
proves that this phenomenon is possible. But neither the fact of
the distribution of the vacuoles between the daughter cells during
mitosis, nor that of their persistence in the aleurone grains of the
seed and their transmission to the embryo, proves that the vacuoles
can not rise de novo. Moreover, we can scarcely permit ourselves
to consider them as individualities of the cell, incapable of forming
de novo, when through their extreme instability of form, they may
in the space of a few minutes be split up into very minute elements
capable soon of fusing again.

Some fungi are more favorable for the study of the origin of
the vacuoles than are the phanerogams. In the mycelium of Peni-
cillium glaucum (Fig. 101) or of Oidium lactis, for example, lateral
branches may be observed to form from filaments already con-
taining large vacuoles and in these branches, which at first do not
have them, small globular vacuoles are seen to appear which can
hardly have any relation to the large vacuole of the filament from
which the branch arises. This is also true of the buds of the
yeasts in which there are small vacuoles which do not appear to
be derived from the large vacuole of the mother cell. We concluded
from these very clear facts observed by means of vital dyes that
vacuoles may be formed de novo (Figs. 124-126).

P. Dangeard has objected, and with reason, that vital stains
can cause alterations of the vacuoles, for example, their immedi-
ate fragmentation when in the process of dividing. It is certain
that vital dyes stop the multiplication of cells in certain cases, par-
ticularly in the fungi. Dangeard again took up the study of the
formation of vacuoles in yeasts and in following the budding of
these fungi in a moist chamber without vital staining, showed that
the large vacuole of the mother cell always puts out a delicate pro-
longation into the bud. The extremity of this prolongation is cut

Guilllermond - Atkinson

180

Cytoplasm

off and forms the small vacuole of the daughter cell. Dangeard
has sought more recently to demonstrate, but this time with vital
dyes, that the vacuoles of algal zoospores are always transmitted
by means of the filament put out at germination.

In Saprolegnia, growing on media to which neutral red has
been added and observed in van Tieghem and Le Monnier cells,
we have shown, however, that the vacuoles which in the zoospores
appear as small granules, fuse at the moment of germination to
constitute a single large vacuole and then, in the germination tube,
small globular vacuoles appear which do not seem to be derived
from the large vacuole of the zoospore. Now, if the objection of
Pierre Dangeard is sound in regard to vital staining of the
ordinary fungi which, when carried out between slide and cover

* > a • ■ • 7 •

A

* a J • 1 1

Fig. 125 (left). — Saprolegnia. Germination of zoospores in a van Tieghem
and Le Monnier cell on 1% peptone bouillon with 0.001% neutral red. 1, 2,
zoospore. 3-20, germination tube. The large vacuole extends into the germina-
tion tube and may fuse (15) with small vacuoles which form at the tip. (After
Mile. Cassaigne) .

Fig. 126 (right). -
vacuoles during budding
cell without vital dyes,
mentation and fusion.
vacuole of mother cell,
formed by a kind of
Cassaigne) .

- Saccharomyces pastor ianus. Types of formation of
of the yeast grown in a van Tieghem and Le Monnier

A, 1-8, Observation during 1 hr. 1-5, successive frag-
6, 7, prolongation sent into bud. 8, separation from

B, vacuole of bud formed de novo. C, vacuole of bud
budding from vacuole of mother cell. (After Mile.

glass do not grow as long as they keep the neutral red accumulated
in their vacuoles, this objection is evidently not sound in the case
of Saprolegnia, which can be grown on media to which neutral
red has been added. Mile. Cassaigne repeated this study and ob-
served the development of vacuoles, both in the germination tube
and in the growing filaments. She was able to see small vacu-
oles form de novo which later refused to constitute a single large
vacuole, or elongated into filaments and anastomosed into a net-
work. Nevertheless, since the conditions of observation were ad-
mittedly abnormal because of the presence of neutral red. Mile.
Cassaigne repeated the work of P. Dangeard on the yeasts in
which she followed budding without vital staining. Now, she ob-
served that actually the vacuole of the bud may arise from the
budding of the vacuole of the mother cell, as Dangeard indicated,
but that often this vacuole also arises de novo in the bud.

These observations seem therefore to furnish proof that these
vacuoles form de novo. In consequence of our research, we have

Chapter XV — 181 — Origin of Vacuoles

proposed an hypothesis to explain the formation of vacuoles. This
hypothesis is based, on the one hand, on the fact that colloidal sub-
stances contained in the vacuoles are of very diverse constitutions,
and, on the other hand, on the fact that the cytoplasm is constantly
the locus of secretory phenomena (production of reserve or of
waste products) . The hypothesis assumes that among these prod-
ucts, those which are in a colloidal state separate by an unknown
physical-chemical mechanism from the cytoplasm in the form of
colloids non-miscible with the cytoplasmic colloids and composing
a distinct phase of the latter. They appear in the form of small
elements. These, by virtue of their semi-fluid consistency and of
their physical state, which is rather like that of the chondriosomes,
are subject to the same laws which determine the shape of the
chondriosomes. This explains the resemblance in form of these
two elements. According to our hypothesis, these colloids possess
a capacity for taking up water which is stronger than that of
the cytoplasm, and when the cytoplasm has reached its maximum
point of imbibition, the excess water is absorbed within these ele-
ments which are gradually transformed into a true solution and
constitute the vacuoles. In these vacuoles during the different
stages of their development, there may accumulate by absorption,
according to this theory, all the products secreted by the cytoplasm
which are capable of forming solutions or pseudosolutions within
the vacuoles. This hypothesis, which resembles that of Pfeffer,
would apply at least to a great number of cases but probably not
to all.

The presence In some cells of several distinct categories of
vacuoles:- Recent research by Mangenot has drawn attention to
the existence of two distinct categories of vacuoles which are ob-
served in the mature cells of numerous plants. They were glimpsed
and very briefly cited some time ago by Went, Klercker and
Lloyd. Vacuoles which are rich in tannin, and very refractive, and
which reduce osmic acid instantly, are often observed in the same
cell side by side with, but in reality distinct from, other vacuoles
which do not contain tannin, are very slightly refractive and show
no reaction with osmic acid. The respective dimensions of each
are sometimes the same, or again the tannin-containing vacuoles
may be much more voluminous than the others, or conversely, may
be smaller, in which case they may appear as filaments or small
granules scattered in the cytoplasm around the vacuoles. Vital
dyes apparently stain these two categories differently. Cresyl blue,
for instance, stains the vacuoles containing tannin, blue or green,
and the other vacuoles, violet or rose. Cells with tannin-containing
vacuoles are very widespread in plants (Legumes, Mimosa, Ber-
beris, Eucalyptus, Oxalis, Monotropa, Hypopitys).

These two categories of vacuoles, the one acid and rich in tan-
nins, the other without tannins and seeming to have a high pH,
Bailey has found in the cambial cells of gymnosperms and arbores-

Guilliermond - Atkinson

182

Cytoplasm

cent angiosperms and MiLOViDOV pointed out their existence in epi-
dermal cells of rose leaflets.

More recent research has made it possible to show the rather
frequent presence of specialized vacuoles in epidermal cells of

Fig. 127. — 1-7, Fruit of Rubus fruticosus. Two types of
vacuoles. 1-4, exocarp. 1, young green fruit; large and
small colorless vacuoles (V.v). 2, ripening fruit; large
vacuole with raspberry-red anthocyanin pigment and small
colorless vacuoles. 3, ripe fruit; three vacuoles with red
pigment, numerous small vacuoles either with colloidal bodies
(g) or needle-shaped crystals (c), both colored violet by
anthocyanin. 4, ripe fruit; one large vacuole with red pig-
ment, smaller vacuoles with crystals of dark violet pigment,
isolated or in bundles. 5, mesocarp, ripening fruit; large
vacuole (V), with dilute solution of raspberry-red pigment,
numerous small vacuoles (v), with brick-red pigment and
raspberry-red colloidal granules. 6, 7, mesocarp, ripe
fruit; small vacuoles with one or more violet colloidal bodies,
sometimes also with needle-shaped crystals of pigment, isolated
or in groups, sometimes with the crystals only. 8, Wisteria
sinensis; epidermis of petal; large central vacuole {V) , with
red-violet pigment and small peripheral vacuoles (v), with
a concentrated solution of blue-violet pigment and crystalline
needles of dark blue pigment. 9, Hibiscus syriacus; epidermis
of petal; large central vacuole (V) , with red pigment and
dark red tannin bodies (P) ; small peripheral vacuoles (v),
with mauve pigment. 10, Canna indica; epidermis of leaf;
large vacuole (V) , with red pigment and a large spherical
crystal (S) ; one or more small colorless peripheral vacuoles.
(in vivo).

leaves, fruits, and flowers which contain anthocyanin pigments.
One of the most curious cases is to be found in the epidermis of the
petals of Wisteria sinensis, already reported by Went, in which
all the cells show two very distinct categories of vacuoles: one
large central vacuole and several small peripheral vacuoles. The
large central vacuole contains tannin and a reddish violet antho-

Chapter XV —183— Origin of Vacuoles

cyanin pigment. The small peripheral vacuoles lack tannin and
contain a very concentrated solution of bluish violet anthocyanin
pigment which is capable of partially or totally crystalizing into
long needle-shaped, dark blue crystals. Another no less interesting
example is seen in the exocarp and mesocarp of the fruit of Rubus
fructicosus, in which all cells likewise possess two sorts of vacu-
oles, the one large, solitary and centrally placed containing, at the
same time, tannin and a cherry-red pigment; the other, small,
spherical, extremely numerous, and scattered in the parietal layer
of the cytoplasm. These latter are without tannin and form at
first a brick-red pigment, but when the fruit is mature, there
appear in each of these vacuoles, large colloidal bodies, dark
violet in color, which show concentric zones. These are the result
of the precipitation of the colloidal content of the vacuoles which
has absorbed the pigment contained in the vacuoles. At maturity
the vacuolar sap changes from brick-red to pale violet, then to
white, whereas blackish, violet-blue crystals shaped like needles,
or sphaerocrystals, are deposited in the interior of the vacuole
between the colloidal bodies. In certain parts of the epidermis of
the petals of Hibiscus syriacus also, there are found in each cell
a large central vacuole, enclosing tannin as well as a raspberry-red
anthocyanin pigment, and small peripheral vacuoles enclosing a
mauve pigment.

In all the cases which we have just examined, the two categories
of vacuoles contain colloidal substances and have the property of
accumulating vital dyes, but this is not, however, universal. In a
very great number of cases (in the epidermis of leaves, stem and
petals of roses, in the petals of Lathyrus odoratus, Prunus japon-
ica, Camellia japonica, Tropaeolum majus, in leaves of Canna in-
dica, etc.), there are found together constantly, in each cell, two
categories of vacuoles: a large central vacuole containing tannins
or other colloidal substances as well as an anthocyanin pigment,
and small colorless vacuoles seemingly without any colloidal sub-
stances. Those of the second category sometimes contain very
minute crystals showing Brownian movement. In the elongated
cells of the inner portion of the fleshy pericarp of the fig there are
two categories of vacuoles of very curious appearance, each vary-
ing both in number and dimension in the cell. One type contains
violet-red anthocyanin pigment together with colloidal substances
and is very variable in shape, the larger among these having irreg-
ular contours which give them an angular appearance, the smaller
ones being chondriosome-shaped elements. The other type is color-
less, lacking in colloidal substance and all of them, no matter what
their dimensions, appear perfectly spherical. In this case, the vacu-
oles which do not contain tannins or other colloidal substances never
stain, which seems therefore to add further proof that vital stain-
ing of the vacuole is due exclusively to the presence in them of
colloidal substances. The case of the pericarp of the fig is par-
ticularly interesting because it shows us that the shape of the
vacuoles, whether irregular or like that of the chondriosomes, seems

Guilliermond - Atkinson

— 184

Cytoplasm

to be attributable to the viscosity of their contents, since the co-
existing vacuoles without colloidal substances are always spherical.
In all cases in which the cells contain two categories of vacu-
oles, these vacuoles become distinct very early and it is very diffi-
cult to determine their origin. It would seem, however, that the
vacuoles lacking in colloidal substances arise by exudation from
vacuoles rich in colloids, for, by plasmolysis, it is possible in some
cases to obtain experimentally the formation of similar small vacu-
oles in cells which do not contain any.

It would be natural, therefore, along
with Weber and Kuster, to relate this
phenomenon to vacuolar contraction and
to attribute it to a syneresis assuming that
the vacuoles not staining with vital dyes
are totally lacking in colloidal substances.
But we have seen that this is not always
so and in the fruit of the blackberry there
exist two categories of vacuoles both of
which contain colloidal substances. In
this case it might be supposed that these
two categories of vacuoles, which seem to
correspond to small accumulation and
transportation centers for various meta-
bolic products, are always distinct and
have no genetical connection, or else that
they arise by a differentiation from a
single category of vacuoles, but by the
phenomenon of coacervation and not of
syneresis (Cf. p. 177).

However this may be, this last ex-
planation does not apply to the lower
plants, in particular to the algae, in which
there are encountered still more fre-
quently, several categories of vacuoles in a
single cell. In the brown algae it has been
known for a long time that there exist
viscous inclusions which have been called
fucosan granules (Hansteen), or phy-
sodes (Crato), whose morphological sig-
nificance has been the subject of numerous
discussions. These inclusions stain vitally like the vacuoles and
yet are present at the same time with other large vacuoles whose
contours are more fluid and which also take the vital dyes but
stain differently. Cresyl blue, for example, gives the inclusions
a greenish blue tint, whereas the vacuoles take a violet-blue color.
These inclusions contain in fact catechin tannins, showing the
phloroglucinol-hydrochloric test, which explains the greenish blue
color which they give to cresyl blue. Although these phenolic in-
clusions are always separated from the other vacuoles even in the
beginning, it seems logical to consider them, as does Mangenot,

Fig. 128. — Fig. Living
cells from the inner part of
the fleshy receptacle. Two
types of vacuoles, one (V)
varying in size and shape
with colloidal contents and
anthocyanin pigment; the
other iv) , varying in size
but always spherical, with-
out colloidal substance.

Chapter XV —185— Origin of Vacuoles

as corresponding to specialized vacuoles, for the same reason that
we consider as specialized vacuoles, those encountered in the
phanerogams in which group of plants phenolic compounds are
always localized in vacuoles, often having forms corresponding to
those of the physodes. Chadefaud, who recently described simi-
lar physodes in the Phaeophyceae, thinks, on the contrary, that
these inclusions are chondriosomes which elaborate mucilages and
phenolic compounds. This opinion does not seem plausible to us.

The same peculiarities are found in Vaucheria in which the re-
cent work of Mangenot has shown, apart from the central vacu-
olar canal previously discussed, the existence of numerous small,
peripheral rod- or granule-shaped vacuoles formed of a very con-
centrated colloidal substance. These small vacuoles which P. A.
Dangeard confused with the chondriosomes, take a blue color with
cresyl blue whereas the vacuolar canal stains violet. Mangenot
thinks these small vacuoles are composed of mucilages with which,
in a great many cases, tannins are associated.

In Euglena viridis, as well as small vacuoles composed of a
concentrated solution of metachromatin, there are found in the
sub-cuticular cytoplasmic layer, spheres colored purple-violet by
cresyl blue and appearing as small vacuoles. According to Chade-
faud, these are specialized vacuoles containing mucus (Fr. corps
muciferes) . In this same region in other Euglenas, there are ob-
served elements shaped like bacteria, colored blue with cresyl blue,
rarely violet, which are capable of ejecting their contents as a long
filament. They seem to correspond to trichocysts such as are ob-
served in certain ciliates.

In Cladophora (Fig. 103), P. A. Dangeard cited two categories
of vacuoles, those centrally placed which are large and liquid, others
at the periphery which are small, semi-fluid elements shaped like
rods.

The existence of specialized vacuoles (with the exception of
those lacking in colloidal substances whose existence appears to be
connected with a phenomenon of syneresis), shows us that it is
not possible to consider the vacuome as a morphological entity in
the sense of Dangeard or to adhere to the theory of de Vries.
It is difficult, moreover, to keep for the term "vacuole" its classical
meaning and to limit it to liquid inclusions of the cell, since it is
now established that the well characterized vacuoles of the majority
of plants are themselves derived from semi-fluid inclusions whose
consistency is often greater than that of the cytoplasm itself, vacu-
oles which during the development of the cells again pass through
semi-fluid and even solid phases. Furthermore, we have seen that
in some lower plants the small inclusions, which, by the nature of
their contents and their predilection for vital stains, correspond un-
questionably to the vacuoles of more evolved plants, may remain
constantly in the semi-fluid state.

One fact, however, stands out very clearly from the investiga-
tions just reviewed. It is that the protoplasm itself, i.e., living mat-
ter, is incapable of being stained with vital dyes except in a transi-

Guilliermond - Atkinson — 186 — Cytoplasm

tory way. Either it excretes them into the vacuole or else it dies,
poisoned by them. It is only in products resulting from its metabo-
lism that the stains accumulate. We are thus brought back to
the idea expressed by many cytologists, VON MoLLENDORFF among
others, that vital dyes normally stain only that which we call the
deutoplasm, or paraplasm, m which are grouped all the products
arising from protoplasmic elaboration. The vacuoles belong in
this category and perhaps it would be suitable to include under
the general heading of vacuolar system (a term preferable to that
of vacuome, which involves the idea of morphological entity) all
the paraplasmic colloidal inclusions of the cytoplasm which are
not of a lipide nature or, at least, in which the lipides do not consti-
tute the essential element. These inclusions are composed of aque-
ous solutions of colloidal substances elaborated by the cytoplasm,
not miscible with it (doubtless forming a coacervate system sepa-
rate from the cytoplasm) and are characterized by a more or less
high concentration. They are however capable under certain physi-
cal conditions and in certain cells, by reason of their capacity for
taking in water which is stronger than that of the cytoplasm and
often unlimited, of becoming dilute and of taking on the aspect of
liquid inclusions or true vacuoles. In a word, the liquid vacuole
according to this interpretation may be formed each time that
there is deposited in the cell a product of secretion in a colloidal
state more capable of absorbing water than is the cytoplasm. So
the vacuolar system expresses a physical state, an aqueous phase,
separated from the cytoplasm, and containing various more or less
concentrated, colloidal and crystalloid substances of paraplasm
which may, according to the nature of these substances and the
conditions of the cell, have a rather high viscosity and which are
able to pass from the liquid to the semi-fluid or solid state.

If, in the great majority of plants, the vacuolar system appears
to us as a morphological entity, it is undoubtedly because plant
cells undergo a considerable hydration, and because, from the first
stages of their development, the paraplasmic inclusions whose en-
closed colloids take up water, are transformed into liquid vacuoles,
which run together very quickly and become a single and enormous
vacuole in differentiated cells. It follows logically that all the
products of metabolism capable of forming solutions or pseudo-
solutions with water would collect in this single vacuole. In some
lower plants and in animals, on the contrary, the cells would not
undergo this hydration and the paraplasmic inclusions would gen-
erally remain in the cytoplasm as concentrated colloidal solutions,
making a distinction among them more easy by reason of the
chemical contents characteristic of each.

According to this hypothesis it is possible to see, as does Mange-
NOT, a similarity between the vacuolar system and the lipide inclu-
sions — paraplasmic formations capable occasionally of being
stained in the living state because the vital stains are soluble in
lipides. In these inclusions, which are formed of neutral fats and
which are found in practically all cells, there accumulate all the

Chapter XV —187— Origin of Vacuoles

products of secretion of the cytoplasm capable of being dissolved
in them (phytosterol, lecithins, oils, carotinoids, etc.)- They may
remain scattered in the form of small inclusions in the cytoplasm,
or may fuse together, as in the spores of certain fungi and in the
adipose cells of animals, to constitute a single enormous fatty
globule occupying the entire cell.

A distinction may therefore be made in the paraplasm between
hydrophilic inclusions (vacuoles) and hydrophobic inclusions (lip-
ide inclusions).

Be that as it may, these investigations, taken all together, show-
that vacuoles are present in all cells, just as are the chondriosomes.
Although both are present, the vacuoles cannot, in any way what-
ever, be considered similar to the elements of the chondriome.
There is reason to think that they have no permanence, no indi-
viduality which is transmissible from generation to generation, or
as Parat says, "Only the group is significant, only the vacuome
is an entity, the expression of a cellular equilibrium, the bond in
metabolism, an 'aqueous phase' whose elements disappear and are
replaced by others."

Digestive vacuoles:- The theory which we have just formulated
in regard to the significance of the vacuoles permits us to incorpo-
rate the vacuoles of the Myxomycetes in the vacuolar system. It is
known that in this group as well as in the Amoebas, there do not
seem to be any digestive vacuoles which take up vital dyes but
there are vacuoles which are distinguished from ordinary vacuoles
by their exogenous origin. These vacuoles arise from food particles
surrounded by a little water in the cytoplasmic mass. If the hy-
pothesis which we have formulated on the origin of vacuoles be
admitted, it follows that in spite of their exogenous origin, these
vacuoles are in the same category as the others, contrary to the
opinion of Volkonsky who definitely separates them under the
name of gastriole.

There are other vacuoles, present in the flagellate algae which
are pulsating vacuoles. Their significance is still unknown.

Chapter XVI

THE ROLE OF THE VACUOLAR SYSTEM AND
HYPOTHESES CONCERNING IT

One of the most important functions of vacuoles is to regulate
the exchange of water which takes place in the cell by osmotic
phenomena. This was brought out by the classical research of DE
Vries. We have already mentioned this function (p. 125). Now
we must show the applications of it made by DE Vries.

From his experiments, this investigator thought out a method
by which the osmotic pressure of a cell might be determined. This
consists in placing fragments of living tissue (for example, stami-
nate hairs of Tradescantia, which have been mentioned before as
particularly favorable for these experiments) in solutions of a
known substance such as sugar, arranged according to concentra-
tion. There may then be found a limiting concentration at which
plasmolysis is just beginning, i.e., in which separation of the proto-
plasm from the angles of the cell wall is first detected. This limit-
ing phenomenon is considered as a criterion and it is recognized
that it corresponds to an equality in osmotic pressure : the solution
is therefore isotonic with respect to the vacuolar sap.

By this method DE VRIES made known one of the fundamental
laws of osmosis. By a series of progressive comparisons of dif-
ferent substances, it is demonstrable that they are isotonic when
each produces incipient plasmolysis of cells. By this biological
method, DE Vries was able to show that isotonic solutions are equi-
molecular, i.e., equal osmotic pressures are developed by an equal
number of molecules. The method is so sensitive for sugars that
DE Vries was able to determine the molecular weight for raffinose
about which chemists disagreed. Electrolytes, however, are ion-
ized in solution and each ion, acting as a molecule, increases osmotic
pressure. Consequently DE Vries had to introduce into this law
a coefficient of correction (isotonic coefficient).

Osmotic pressure of the cell sap varies according to the con-
ditions of the life of the plant : the osmotic value is 4-5 atmospheres
for aerial parts of aquatic plants, 12-14 atmospheres for cells in
the root of the bean, almost 100 atmospheres for the chlorophyll-
bearing parenchyma of various plants.

Normally, cells are always distended by their vacuolar sap.
This rigidity is called turgidity. The cells are entirely comparable
to a blown up balloon : the internal pressure manifests itself if the
membrane is pierced by a microdissecting needle and the proto-
plasmic and vacuolar contents escape with force just like the air
of a punctured balloon. Turgidity plays a considerable role in the
life of plants in maintaining their rigidity. When it is lacking,
the plants lose their rigidity and wilt. Cells have, moreover, the
means of regulating the concentration of their vacuolar sap in

Chapter XVI — 189 — Role of Vacuoles

such a way that it is always hypertonic with regard to the sur-
rounding medium. As soon as the concentration of the latter
increases, the cells hydrolize their reserve starch and the resulting
sugar goes into the vacuolar sap whose osmotic capacity increases.
This phenomenon has been given the name anatonosis. The pres-
ence of colloidal substances in the vacuole suggests that their role
is not reduced merely to osmotic actions but that they intervene
also in the processes of the passage of water in and out of the cell.
It is, in fact, this inward and outward passage of water, inter-
vening alternately between vacuolar colloids and cytoplasmic col-
loids which explains the reversibility of form of the vacuolar sys-
tem discussed earlier.

The vacuoles are accumulation regions, the reservoirs of a large
number of metabolic products or of reserves, and are particularly
regions of excretion of toxic substances, as the action of vital dyes
tends to indicate. In the vacuoles, there accumulate all the products
secreted by the cell which can be dissolved in water, forming true
or pseudosolutions (proteins, holosides, heterosides, tannins, fla-
vins, oxyflavanol and anthocyanin pigments, organic acids, alka-
loids, certain lipides, mucilages and so on). These various prod-
ucts may appear in the meristematic cells at the very beginning
of development of the vacuolar system, or at any stage whatever,
during the development of the system. It has been possible to
demonstrate, notably by microchemical reactions, that in the seed-
ling of tobacco, alkaloids appear in the cells of the meristem of
the root, in the chondriosome-shaped vacuoles formed by the hydra-
tion of aleurone grains (Chaze). There have been localized also
in the chondriosome-shaped vacuoles of the meristematic cells, cer-
tain heterosides, such as the saponarosides (POLITIS). This is
true for tannins (GuiLLlERMOND, P. Dangeard), the oxyflavanol
compounds and anthocyanin pigments (Guilliermond) . The vacu-
olar system is certainly more than a locus for the accumulation of
these various products. The presence of colloidal substances in
the vacuoles and their predilection for vital stains lead us to sup-
pose, as do the Dangeards, that the vacuoles can exercise a role
in absorption phenomena because of these very properties of ab-
sorption, imbibition and combination which bring about the pene-
tration of the dyes. Devaux believes the vacuolar system to be
the site of chemical affinities of the cell and explains that the vital
dyes penetrate the cell without staining the protoplasm and accu-
mulate exclusively in the vacuoles, i.e., in the non-living parts of
the cell, because the chemical affinities of the living substance are
masked by reciprocal saturation. So by his theory of polarized
(catalytic) membranes (C/. p. 121) only the non-living inclusions of
the cytoplasm, such as the vacuoles, are capable of fixing the dyes,
and there is localization of protoplasmic activity on the surface
presented by the protoplasm and the vacuole. It has been seen
that this opinion is not justified. If, actually, some dyes, like
neutral red, traverse the cytoplasm without ever staining it and
accumulate only in the vacuole, this is not true for other dyes which

Guilliermond - Atkinson — 190 — Cytoplasm

can stain the cytoplasm in living cells. Nevertheless this staining
is only transitory and the dye moves from the cytoplasm into the
vacuole. It is only after the dye has been localized in the vacu-
oles that the cells are capable of growing and no staining other
than vital staining of the vacuoles is compatible with growth.
The hypothesis of Devaux seems to be confirmed by the works
of Genevois and Genaud, who have shown that absorption of
salts by cells occurs exclusively along the cellular and vacuolar
membranes. It is necessary, however, to make reservations in
regard to the absence of chemical affinities from the cytoplasm,
since it has been seen that certain stains may, under some con-
ditions, be retained by the cytoplasm (pp. 18, 142). It has often
been supposed that the vacuolar system is not a simple center of
accumulation of metabolic products but that it is at the same time
the seat of phenomena of hydrolysis and of synthesis. According
to Kedrowsky and Volkonsky, the vacuoles are the secretion ap-
paratus of the cell and the seat of enzymes, particularly of pro-
teases, but this view seems to be exaggerated. There is reason to
believe that in the chemical phenomena which take place in the
vacuole, it is the cytoplasm which plays the active role, the vacu-
ole having only a passive role.

Let us add that Parat considers that in animal cells, methylene
blue is always reduced in the cytoplasm and in the chondriome
(rH <12), and that it is, on the contrary, re-oxidized by the vacu-
oles (rH <16), which does not seem to be true in plant cells.
Going back to the hypothesis of Robertson (p. 122), Parat thinks
that the pair : chondriome plus vacuole, presides over the synthesis
of proteins, which, according to Robertson, calls for a lipide phase
and an aqueous phase and thus gives a morphological basis for
this hypothesis. Parat considers further that the vacuome is the
crucible in which are completed the operations begun in the chon-
driome, but these points of view are very hypothetical and lack
a solid foundation.

It has been seen that this hypothesis of Parat is no longer
tenable, now that it is demonstrated that the chondriosomes do
not of themselves have a reducing role, contrary to what had
been supposed, and that the vacuoles may in certain cases be just
as capable of reducing actions as the chondriosomes.

Chapter XVII

GOLGI APPARATUS, CANALICULI OF HOLMGREN
AND OTHER CYTOPLASMIC FORMATIONS

Golgi apparatus and the canaliculi of Holmgren in animal cells :-
By using methods of silver nitrate impregnation, GOLGI (1898)
brought out in the cytoplasm of nerve cells (Purkinje cells and
invertebrate ganglia of Strix flammea) a network of very fine
filaments to which has been given the name internal reticular ap-
paratus of Golgi. This formation was made the object of impor-
tant studies by Cajal. Kopsch showed later that the G<)lgi
apparatus can also be brought out by osmic impregnation at 40° C.

This later method has the advantage over the preceding
that it is much easier to use, for, unlike impregnation with silver,
it does not result in so many failures. For this reason, it has been
the starting point for a great deal of research which has revealed
in most animal cells, formations which osmic acid blackens, just
as it does the Golgi apparatus described by GoLGi and Cajal. These
formations, in spite of their widely differing morphological aspects,
have been grouped with the Golgi apparatus on the single basis
that they stain like it. These formations are not generally repre-
sented by a network but by small elements scattered in the cyto-
plasm, appearing as spherical or ovoid bodies, composed of a
chromophobic substance surrounded by a chromophilic substance
which is thicker on one side than on the other. They are known
as dictyosomes or Golgi bodies.

Many cytologists today think that the Golgi apparatus is a
permanent feature of cytoplasm in the same way as is the chondri-
ome, and there has been described, during mitosis, a division of
the Golgi bodies between the daughter cells which has been called
dictyokinesis (Perroncito). Finally, scientists are coming to the
belief that this apparatus is the center of elaboration of metabolic
products. In brief, it is thought to play the role formerly attrib-
uted, first, to the ergastoplasm and, later, to the chondriome. So
the Golgi apparatus may be said to have supplanted the chondriome
for those who adher