Marine Biological Laboratory

RprP,vpH May 1, 1942

Accession No. 54964

Given By D^* Douglas rarsland
ivew York University

Place

THE STRUCTURE OF PROTOPLASM

^ Symposium ofi^

THE STRUCTURE OF
PROTOPLASM

A MONOGRAPH OF THE
AMERICAN SOCIETY OF
PLANT PHYSIOLOGISTS

Edited By
WILLIAM SEIFRIZ

<i6\c^-x

PUBLISHED, 1942

THE IOWA STATE

AT AMES, IOWA

COLLEGE PRESS

Copyright 1942, By The
Iowa State College Press

PRINTED AT THE COLLEGIATE PRESS, INC.
IOWA STATE COLLEGE, AMES, IOWA, U.S.A.

FOREWORD

This monograph is the printed record of a symposium on the
Structure of Protoplasm presented under the auspices of the Ameri-
can Society of Plant Physiologists at Philadelphia on December 30,
1940.

The meeting was held under the Presidency of Dr. F. P. Cullinan,
with Dr. William Seifriz as chairman. The symposium was the
partial fulfillment of a plan, long held by the chairman, for a series
of lectures on the structure of matter in which physicists and chem-
ists would build structural units to be used by biologists in an
attempt to construct a mechanism having some of the properties of
a living system. Though no physicists or chemists in the strict sense
actually took part in the symposium, yet several were indirectly
involved; thus Professor Herbert Freundlich, who was invited to
attend but could not come, offered a paper on Thixotropy which
proved to be his last manuscript. Another chemist, Dr. Linus Paul-
ing, has made a substantial contribution to the Monograph by aiding
in the writing of the Introduction. Throughout the chapters is
evidence of assistance from other physicists and chemists, notably
Dr. W. T. Astbury, Dr. H. Mark, and Dr. Kurt Meyer, two of whom
have made brief, last-minute contributions of their own which have
been put in a Supplement.

This volume is the first in what is planned to be a series of mono-
graphs to be published under the auspices of the American Society
of Plant Physiologists. It is appropriate that the first monograph
should deal with so basic a subject as the structure of living matter.
It is the wish of the authors that this monograph reawaken among
American botanists an interest in a study which occupied so much of
the time of European botanists during the past century. It was a
botanist, Karl von Naegeli, who gave biology and chemistry one of
the first fundamental theories of protoplasmic and gel structure,
and it was a botanist, Fr. Weber, who established the first journal
devoted solely to a study of protoplasm. By inaugurating its series
of monographs in plant physiology, with a volume on the structure
of protoplasm, the American Society of Plant Physiologists takes

vi Foreword

a forward step in advancing one of the most fundamental subjects in
science.

The Society is deeply indebted to all who contributed to the
success of the symposium and of this volume. Especial appreciation
is due Dr. William Seifriz for the knowledge, skill, and enthusiasm
which have made the project possible.

W. E. LooMis

Secretary-treasurer
Ames Iowa American Society of Plant Physiologists

June "l, 1941

TABLE OF CONTENTS
Foreword v

W. E. Loomis, Iowa State College

Introduction 1

William Seifriz, University of Pennsylvania

Microscopic Structure of the Cell Wall 11

Charles W. Hock, Textile Research Foundation, Bureau of Standards

Proteins and Protoplasmic Structure 23

Laurence S. Moyer, University oj Minnesota

Molecular Structure in Protoplasm 41

O. L. Sponsler and Jean D. Bath, University of California

Some Mechanical Properties of Sols and Gels and Their Re-
lation TO Protoplasmic Structure 85

Herbert Freundlich, University of Minnesota

Structural Differentiation of Cytoplasm 99

G. W. Scarth, McGUl University

Structural Differentiation of the Nucleus 109

C. L. Htiskins, McGill University

Protoplasmic Streaming in Relation to Gel Structure in the

Cytoplasm 12'

Douglas A. Marslayid, New York University

The Relation of the Viscosity Changes of Protoplasm to

Ameboid Locomotion and Cell Division 163

Warren H. Lewis, Wistar Institute

Physical Aspects of Protoplasmic Streaming 199

Nobnrd Kamiyc, University of Tokio

Some Physical Properties of Protoplasm and Their Bearing

on Structure 245

William Seifriz, University of Pennsylvania

Supplement:

Protein and Protoplasmic Structure, Communication

jroyn Kurt H. Meyer, University of Geneva 267

Letter From W. T. Astbitry, University of Leeds .... 270
Index 273

library) 30| ^7:1707-

[S^V ^^^^ J^i

INTRODUCTION

William Seifriz
The contributors to this symposium on protoplasm will deal with
present-day concepts on the structure of living matter. A brief
recounting of past theories will, therefore, serve as an introduction.
In addition, there is the matter of the future. But first I should like
to consider for a moment why it is of structure that we speak.

A symposium on protoplasm in which structure is the chief theme
suggests that structure is the most significant property of living
matter. This is true. One may even go so far as to say that it is a
specific structure of matter which distinguishes the living from the
nonliving.^

The chemical composition of protoplasm is important to life.
Water is the dispersion medium. Salts are required to maintain
physical qualities. Sugars are needed for respiratory oxidations.
Fats serve as reserve energy. And proteins are the material out of
which the structural framework of living matter is made. The com-
position is important, but it is secondary to arrangement, to organiza-
tion, to structure. The proteins, for example, hold a pre-eminent
position among the constituents of protoplasm because they are
structural material.

One may agree to the secondary position of composition and yet
regard emphasis on structure as arbitrary in view of the fact that
protoplasm is, above all, a dynamic system and not a static one. An
outstanding quality of living matter is activity, and activity is a
manifestation of energy. Energy, composition, and function are as
essential to a mechanism as is structure, and energy is particularly
significant in a living system. Indeed, one of the primary attributes
of life is the capacity of living matter to use energy in the synthesis
of protoplasm, whereas the synthesis of nonliving organic substances
involves the liberation of energy.

The primary source of energy in protoplasm appears to be the
oxidation of carbohydrates, but this is not in itself an extraordinary
reaction. It is not the kind of energy but the use to which this energy
is put which distinguishes protoplasm; and this uniqueness of living

'Philos. of Sci., 6:266, 1939.

[11

2 The Structure of Protoplasm

matter is very probably due to a specific arrangement of parts, that
is to say, to structure.

Often I am asked by physicists and chemists if much can be hoped
for in the way of an understanding of protoplasmic structure. Direct
observation, even with the highest powers of the microscope, is too
superficial to reveal anything fundamental. Chemical analysis is not
productive of significant results because once protoplasm is subjected
to severe treatment it is no longer living matter. The very structural
features which distinguish it from the nonliving are then destroyed.

The biologist interprets the basic structure of protoplasm some-
what in the same way that a chemist interprets the structure of a
compound, namely, by its behavior. The chemist sees nothing of the
links, rings, side chains, amino and carboxyl groups which he
attaches here and there, yet he builds up a structure with consider-
able confidence, which is the justification of stereochemistry. Just so
does the biologist work; he interprets the structure of living matter
in the light of its known physical properties. But this was not the
method of the older biologists, and herein lay their error. They told
only of what they saw.

Protoplasm viewed through the microscope appears to be a
suspension of fine granules. On this fact was based the granular
hypothesis of protoplasmic structure. Its weakness lay in the diver-
sity of the granules and the undue significance given them. Some
are admittedly important, as are the plastids and the mitochondria;
others, such as fat droplets, are but reserve food. These last are really
not granules at all but liquid droplets which make of cytoplasm a
fine emulsion. The importance attributed to the "granules" — they
were regarded by some as living units, morphologically and physiolo-
gically independent of the cell — and the necessary coarseness of any
optically visible structure, led to the discarding of the granular
hypothesis.

The belief that the basic structural unit of protoplasm is a spheri-
cal body has long persisted and has given rise to numerous hypotheses
expressed in terms of granules, globules, alveoli, and micellae. Such
suggestions are in part supported by fact. Thus, protoplasm is
unquestionably an emulsion, and when the emulsion globules are
under pressure and symmetrically arranged they assume the shape
of dodeca- or tetrakaidecahedra which are hexagonal in cross section.
These latter are alveoli. -

' Protoplasma 9:177, 1930.

Introduction 3

Emulsion hypotheses of protoplasmic structure were favorably
entertained by biologists, but an emulsion satisfies very few of the
requirements of a living system, and fails utterly as a mechanical
basis of inost physical properties of protoplasm and elastic gels.
Emulsions do not take up water by imbibition, whereas protoplasm
does. Emulsions do not coagulate. Emulsions are not elastic as is
protoplasm. A mathematical analysis led Hatschek^ to the conclusion
that an emulsion structure of elastic gels is untenable.

Biologists were misled by the presence in natural emulsions of a
stabilizer, a third substance, in addition to the water and oil of the
emulsion proper. It is this third substance, often a protein, which
exhibits the physical characteristics of protoplasm, and of jellies in
general, characteristics wrongly ascribed to a pure emulsion. Milk
illustrates the situation well. Superficially, milk is an emulsion of
butterfat dispersed in water, but when milk coagulates, the emulsion
plays no active part whatever. It is the stabilizing substance, a
protein, in milk which coagulates. And so it is with protoplasm.

The introduction of the cytological technique of fixing and stain-
ing gave rise to numerous theories of protoplasmic structure based
on what was seen in prepared tissues. Much controversy developed
over the reality of these structures in living protoplasm, because
fixation produces artifacts.

The fibrillar and reticular theories are the best known among
those based on classical cytological methods. Fibrillar and reticular!
structures in prepared sections of tissues may well be artifacts, but
an artifact is not without significance. The relatively coarse micro-l
scopic threads which form the visible fibrous entanglement or net,
may be artificially produced aggregates of threads of submicroscopic
dimensions. These latter are probably linear molecules. It seems
rather unlikely that a coarse fibrous coagulum should result from
the fixation of a fine granular suspension. Thus do the visible threads
of a reticulum in fixed dead material become evidence of submicro- 1
scopic threads in living protoplasm.

The first hypothesis of protoplasmic and gel structure which held
the serious attention of chemists as well as biologists was the contri-
bution of the botanist Karl von Naegeli. He postulated molecular
aggregates of colloidal size. These aggregates Naegeli called micellae.
They might have a random distribution, or as later thought more
probable, a symmetrical arrangeinent, like bricks in a wall (Fig. 1) .

'Trans. Faraday Soc. 12:17, 1917.

The Structure of Protoplasm

This was, and still is by some, assumed to be the structure of cellu-
lose. The concept had one great fault, there was nothing to account
for a strong bond between micellae (at c and d, Fig. 1) . A cementing

material was assumed,
but to attribute the ten-
sile strength of cellulose
to an intermicellar sub-
stance, rather than to the
micellae themselves, was
poor reasoning. Again
the fallacy lay in attempt-
ing to build an elastic
substance of high tensile,
strength out of granules. Protoplasm hangs together, and so does'
cloth; the latter we know to be made of threads. In order to satisfy
the elastic and tensile qualities of jellies, the "brush-heap" theory
of interlacing fibers was formulated.^

The brush-heap theory had several advantages; an entanglement
of fibers is elastic (Fig. 2) . A brush-heap was acceptable to both the
proponents of a molecular structure of elastic gels (A, Fig. 2) , and
to those who supported a colloidal or micellar brush-heap (B, Fig. 2) .
But the hypothesis failed as an explanation of certain optical prop-

FlG. 1

Fig. 2

B

erties, such as double refraction, and crystalline symmetry as indi-
cated by X-ray patterns.

Out of the micellar and brush-heap hypotheses there arose the
concept of symmetrically arranged overlapping fibers (Fig. 3).

* Protoplasm, New York, 1936.

Introduction 5

High tensile strength was particularly well explained by this
arrangement. But there was still reason to believe that some dis-
continuity is present in cellulose and like material.

An examination of natural cellulose with dark-field illumination

Fig. 3

reveals the presence of short rods, 1.5 to 2|.i in length, regularly
spaced in parallel rows (Fig. 4) .

Preston' compares the actual and theoretical tensile strengths of
cellulose and finds that the former is 20 kg/mm- for cotton and 100
kg/mm- for flax, whereas the theoretical tensile strength for cellu-
lose, based on the known chemical energy of the C-C link, is 900
kg/mm-. Discontinuity in the structural arrangement of parts would
account for the lowered tensile strength. The micellar hypothesis
appears, therefore, to still hold.

The demands made by elasticity indicate that jellies are built of
fibers. Low tensile strength, as in cellophane, suggests a scattered
distribution (Fig. 2) . High tensile strength, as in ramie fiber,
requires overlapping (Fig. 3) . And certain optical properties indi-
cate discontinuity (Fig. 4) .

The best interpretation, so it seems to me, of these somewhat
contradictory requirements is that indicated in Figure 5 where sym-
metrically arranged over-

trated at regular intervals ^~~ "^^ "" *""" """ '~~~ """" *"""

(See also Fig. 1, page 268) . Fig. 4

Out of the numerous facts
and suggestions so far mentioned, the most significant is the recogni-
tion that the structural unit of elastic jellies is a fiber. This is a great
gain over older concepts in which granules and globules were the
structural units of protoplasm.

The foregoing is the background of our subject. The present will
be dealt with by the contributors to this Monograph. There remains
the future.

The final word on the full meaning of protoplasmic structure will
rest with the biologist, for he alone knows the conditions to be met,

Biol. Revs., 14:281, 1939.

6 The Structure of Protoplasin

physical, cytological, and genetical. But the manner in which water,
salts, carbohydrates, fats, and proteins combine so as to fulfill the
biological requirements of a living system, only the chemist and
physicist can say. The direction which this cooperation between
physicist, chemist, and biologist is likely to take — in short, what the
future has in store for us in the matter of protoplasmic structure —
may be illustrated by a problem, a very fundamental one, namely,
the maintenance in protoplasm of two apparently incompatible prop-
erties, the capacity to flow, and the possession of structural qualities
necessary to satisfy elasticity and tensile strength. The biologist
presents the problem. The physical-chemist must answer, if answer

Fig. 5

there is to be. He suggests loosely attached cross-linkages, such as
those which are established by the hydrogen bond.

The problein of the simultaneous presence in organic systems of
fluidity and structural continuity is one which confronts the chemist
as much as the biologist. Highly branched, cross-linked, three-
dimensional polymers always exhibit elasticity, though fluid.

Side chains which make lateral connections between molecules
have long been recognized in stereochemistry. But the concept that
one end of the link is loosely attached and the other firmly so is
new, and of great biological importance. It has been suggested that
variations in the degree of hydration of the side chains constitute a
possible explanation of a weak attachment at one end. But it is
another interpretation of this feeble union which I have selected,
namely, the hydrogen bond.

In a chemical cycle, as in muscle, during which accessory mole-
cules go in and out of combination with the side chains of the protein
main chains, the latter take up a cycle of configurations. In all such
dynamic systems, of which protoplasm is one, there is a constant
shifting of ties between the structural units. The loose union which
permits this is very probably a hydrogen bond. Linus Pauling*'
believes this bond to be of greater significance for physiology than
any other single structural feature. It serves our present purpose
well because it is not a strong bond.

' The Nature of the Chemical Bond, Ithaca, N. Y., 1940.

Introduction 7

When an atom of hydrogen is attached to not one but two other
atoms and thus acts as a tie between them, the union involves the
hydrogen bond. In such a case, one of the points of attachment is a
firm covalent bond, and the other a weaker one, essentially ionic in
character. The latter is the hydrogen bond.

In general, the hydrogen bond is part of a situation in which
hydrogen appears to have two valences, one stronger with some
ionic character, the other weaker with more ionic character. The
strength of the latter weaker bond ranges from one-tenth to one-
hundredth that of the former, according to the atoms involved and
their distance apart. Because it is feebly attached, the hydrogen
bond shifts easily from one atom to another.

The use to which the hydrogen bond may be put in interpreting
protoplasmic behavior in respect to structural continuity is illustrated
in a simple system like water. For some time it has been known that
the water molecule is polar, i. e., it has an electric moment. It
possesses, therefore, the characteristics necessary for molecular
orientation. The two hydrogen atoms are positively charged, and
the oxygen atom is negatively charged. At a distance, such a mole-
cule is electrically neutral, but viewed nearby, it is positive or nega-
tive depending on one's position. The negative oxygen atom of a
water molecule will attract and hold the two positive hydrogen
atoms of another water molecule, and with a force which may equal
the hold on its own two hydrogen atoms. The valence bonds within
a water molecule may be regarded as chemical and those between
adjoining molecules as physical, but it amounts to essentially the
same thing in the end, for the hydrogen atoms and the oxygen atoms
exert forces on the surrounding molecules which are no less chemical
in nature than those holding the atoms within the molecule together.
Fluidity depends upon the shift that takes place between one mole-
cule and another. Water molecules are always in partnership, but
always changing partners. We thus see how polarity leads to orienta-
tion and it in turn to continuity in structure in liquid systems. That
this is true is indicated by the diffraction patterns which certain
solutions yield, indicative of some kind of arrangement.

In a somewhat similar manner, Pauling interprets the structural
continuity of water in terms of the hydrogen bond. Instead of the
classical H^O, it is presumed that each oxygen in water is surrounded
by four hydrogens, two of which, close to the oxygen atoms, are

8 The Structure of Protoplasm

joined to it by primary valences, and two, farther away, by the
hydrogen bond.

The situation existing in water is even more apphcable to pro-
teins where structural possibilities are infinitely greater; and proteins
are the building material of protoplasm. In amino acids the nitrogen
atom has one open coordinate position with which it may unite to a
hydrogen in the same or another molecule. This union is a hydrogen
bond; it may join the nitrogens of two amino acids.

The essential feature of the hydrogen bond, insofar as our present
problem is involved, is its ease in shifting. This permits fluidity
v/hile maintaining continuity in structure. But protoplasm does not
flow with equal ease at all times; it is often of high viscosity, and it
may be firm. The lateral bonds are therefore more securely held at
one time than at another, and at still other times they are tightly
locked, wholly preventing flow. If the hydrogen bond is to satisfy
these conditions, it must show considerable variability in firmness,
from a feeble contact permitting ready readjustment, to a firm grip
tightly locking the fibrous units of the living three-dimensional lat-
tice. Pauling points out that firmness of the hydrogen bond varies
with the electronegativity of the atom.

As one of the essential points of this discussion is the variability
in firmness of the weaker end of cross-chains, I want to dwell upon
it just a moment longer in order to emphasize an important differ-
ence between such a variability in nonliving and in living systems.
Usually, variability in the viscosity, elasticity, and tensile strength
of a solution of nonliving chain polymers is accomplished simply by
changing the concentration. At high concentration the chains will
join at more points. But in protoplasm the "concentration" remains
the same, yet the physical properties vary. This is accomplished by
change in the firmness of the cross-ties. How this change can come
about has been suggested. There is the further possibility that some
protons are added or subtracted with change in pH, forming or
breaking hydrogen bonds; change in salt concentration might also
alter their strengths.

The hydrogen bond thus provides a mechanism by means of
which continuity in structure, elasticity, and rapid changes in viscos-
ity in protoplasm may be explained.

Astbury' calls attention to another attribute of the hydrogen

' The hydrogen bond in protein structure, Trans. Faraday Soc. 34: 871-880,
1940.

Introduction 9

bond, namely, its probable function in the hydration of proteins.
Water of hydration is not our present interest, though an important
subject in biology. Reference is made to it in order to show further
application of the hydrogen bond.

The hydrogen bond will probably serve the biologist well in his
studies on the mechanism of protoplasmic behavior. But the physical
chemist promises still more for the future. Pauling'" calls attention
to resonance as a likely factor in the physiological activities of living
matter. Resonance describes the state of a molecule when one of
several alternate structures are possible, and the actual normal state
of the molecule is not represented by any one, but by a combination
of them. The molecule is then said to resonate among the several
possible electronic or valence-bond structures.

Little has as yet been done with resonance in the interpretation
of vital processes. The future of it in biology cannot, therefore, be
predicted with certainty, but it opens up a great field of possibilities.

The biologist is confronted with many qualities of living matter
which will remain pure physiological problems for some time to
come, but the list of properties which are now fully or partially
capable of pure physical or chemical analysis is a long one; it includes
elasticity, tensile strength, contractility, thixotropy, gelation, fluidity,
non-Newtonian behavior, streaming, amoeboid movement, structural
continuity, birefringence, symmetry, asymmetry, spirality, liquid
crystallinity, and selectivity.

It is truly an encouraging sign in the progress of science when
properties of protoplasm such as contractility and structural organi-
zation, which heretofore were so little understood, can now be inter-
preted in terms of folded polypeptide fibers, interlocking side chains, ^
hydrogen bonds, and asymmetry of the carbon atom. I

MICROSCOPIC STRUCTURE OF THE CELL WALL

Charles W. Hock
Research Associate oj the Textile Foundation, National Bureau oj Standards,

Washington, D. C.

Past discussions on the cell wall (1, 7, 12, 23, 26) have indicated
that a detailed knowledge of its structure will aid in the solution
of many botanical problems. Although there is a tendency to con-
centrate attention on the protoplast as the fundamental unit of the
plant, there can be no doubt but that the wall surrounding this unit
plays a role of considerable importance in its life. Since deposition
of the wall appears to be closely associated with protoplasmic
activity, a clear perception of its structure may be helpful in under-
standing the structure and behavior of protoplasm.

The walls of plant cells exhibit diverse structural patterns
depending upon the orientation of the cellulose and other constitu-
ents of which the wall is composed. Excellent photomicrographs
of cell walls which exhibit concentric, radial, ramifying, or even
more complex patterns, have already been published (4, 5) . In spite
of such differences, walls from various sources show certain simi-
larities in structure which appear to be fundamental. Accordingly,
a detailed description of the microscopic structure of a particular
cell wall may serve as a basis for comparison with other types.

Because of its economic importance the attention of many investi-
gators has been directed toward a study of the cotton fiber. As a
result, considerable information concerning the structure and
behavior of its cell wall is available. Furthermore, as material for
the study of cellulose deposition, cotton fibers possess several
advantages. On the day of flowering, or thereabouts, cotton fibers
originate as single-celled outgrowths of epidermal cells of the seed
coat, and their growth history can be traced daily thereafter. Where
fibers arise as part of a complex tissue, as is the case with wood
and bast fibers, the time of their origin is unknown, and their
development cannot be followed so readily.

The first evidence of the origin of cotton fibers is the appear-
ance on the day of flowering of a swelling on the outer wall of the
epidermal cells of the seed coat. The tubular outgrowths thus

[11]

12 The Structure of Protoplasm

formed elongate rapidly for a period of 15 to 20 days, when growth
in length ceases. In accordance with the description by Anderson
and Kerr (2) , the thin wall which encloses the protoplasm of the
cotton hair during this period of elongation is called the primary
wall. After growth in length ceases the thickness of the wall is
increased by a deposition of cellulose which comprises the secondary
wall. In discussing the structure of cell walls it is essential to
differentiate clearly between the primary and the secondary wall.
Since, in cotton fibers, deposition of the latter does not begin until
from 15 to 20 days after the fibers originate, young fibers are ideal
for studying the structure of the primary wall.

When young fibers with primary wall only are examined micro-
scopically with ordinary light, there is no evidence of structure in
the wall. The latter stains deeply and uniformly with ruthenium
red which, although not specific for pectic substances, is a satis-
factory presumptive test (2) . On examination with crossed nicols,
unstained primary walls show a low order of birefringence with
colors indicative of a predominantly transverse orientation (11, 17).
Although these fibers show only slight birefringence when un-
stained, they are clearly anisotropic after treatment with a cellulose
dye such as Congo red. This increase in birefringence upon staining
is due, apparently, to the double refraction of the dye molecules
which are preferentially adsorbed by the small amount of cellulose
which is present (12, 21) . Although a number of investigators have
suggested that the skeletal material in the primary wall is not
cellulose but a closely related substance, X-ray investigations
(11, 28), staining reactions (2, 17), and behavior in cuprammonium
hydroxide solutions (2, 17) indicate that it is probably cellulosic
in nature.

The cellulose in the primary wall of the cotton fiber has been
shown (2, 9, 10, 17) to be present as criss-crossing strands which
have a netlike arrangement (Fig. 1, A and B) . The reticulate
structure of the cellulose appears to become coarser and to show
greater birefringence as the wall increases in age (compare Fig.
1, A and B) . Likewise, the angle which the criss-crossing cellulose
strands make with the axis of the fiber appears to be greater at the
tip than at the base. These facts suggest that growth and expansion
of the primary wall involve the separation of fibrils and the deposi-
tion of new cellulose strands between them.

Cellulose is not, however, the only constituent of the primary
wall. The latter also contains wax and pectic materials which may

Microscopic Structure of the Cell Wall

13

be detected by applying various stains to the fiber (2, 17) , as well
as by direct chemical analysis (29) . When young fibers are purified
by the removal of wax and pectin, the delicate framework of cellu-
lose always remains. It appears, therefore, that the primary wall
is made up of a tenuous network of cellulose embedded in rela-
tively large amounts of wax and pectic substances. There is evi-
dence which indicates that the structure of the primary wall of

Fig. 1. Fibers stained with Congo red and photographed between crossed
nicols to show the strands of cellulose in the primary wall. A. Part of a 15-
day fiber placed at 45° with respect to the planes of the nicols. B. Tip of a
3-day fiber placed at 45° with respect to the planes of the nicols. Magnifi-
cation X 890.

cotton is similar to that found in other plant cells (5) . The orienta-
tion of the fibrils varies in different cells, diverse types showing,
however, a fundamentally similar structure.

Although the coarser details of the structure of the secondary
wall are often visible in unswollen material, the pattern is studied
to best advantage in expanded sections. By the use of suitable
swelling agents details of structure which could not be detected
otherwise, are rendered microscopically visible. Use of the swelling
technique in conjunction with X-ray data and polarization studies
offers a satisfactory method for observing the structure of the cell
wall.

14 The Structure of Protoplasm

When mature cotton fibers are placed in cuprammonium
hydroxide solution, or Schweizer's reagent, they immediately begin
to swell and twist, the rapid expansion of the cellulose of the
secondary wall usually resulting in the rupture of the primary
wall. As the cellulose pushes through the tears in the latter,
expanded regions resembling beads appear along the axis of the
fiber (Fig. 2) . After a short time in the reagent, however, the

Fig. 2. Mature cotton fiber in cuprammonium hydroxide solution, showing
irregular swelling due to the constricting influence of the primary wall.
Magnification X 120.

cellulose dissolves completely, leaving a residue which consists
principally of fragments of the primary wall, and to a lesser extent of
material from the lumen (16). This residue is isotropic, and chemi-
cal analysis indicates that it consists largely of wax and pectic
substance (29) . Full-strength cuprammonium reagent dissolves the
cellulose of the fibers in a relatively short time, leaving no micro-
scopically resolvable particles of cellulose (16) . Dilute solutions
of the reagent, on the other hand, swell the fibers greatly, but

Fig. 3. Cross sections of cotton fibers swollen in cuprammonium hydroxide
solution to show the layered structure of the secondary wall. A. Section of a
25-day fiber. Magnification X 950. B. Section of a mature fiber from an open
boll. Magnification X 420. C. Section of 26-day fiber showing lamellae, photo-
graphed between crossed nicols. Magnification X 700.

16

The Structure of Protoplasm

\

dissolution of the cellulose is retarded. Fibers swollen in this way
can be used for observing details of structure.

Swollen cross sections of cotton fibers clearly reveal a lamellate
pattern. On treating the swollen sections with a dye such as Congo
red, alternating layers appear to be deeply and lightly stained (Fig.

3, A and B) , and between

'■"Vl^"**^ ^'****^^ crossed nicols the alternat-
ing layers are strong and
weakly birefringent (Fig.
3,C). This layered struc-
ture of the secondary wall
can also be observed in
longitudinal view (Fig. 4) ,
where stripes running par-
allel to the fiber axis, and
extending from the lumen
to the primary wall, can be
seen. The first layer is laid
down on the day when sec-
ondary thickening is initi-
ated, the number increasing
thereafter as the fiber ap-
proaches maturity. This in-
crease in the number of
layers with increase in age
of fiber can be seen by com-
paring Figure 3, A with
Figure 3,B.

According to Kerr (20),
two adjacent layers, one
compact and one porous, are deposited every 24 hours during the
period of secondary wall deposition. Two adjacent layers together,
therefore, constitute a daily "growth ring." The denser of the two
layers which comprises each ring was found to be laid down during
the day, the other at night. Counts of rings in fibers of different ages
indicate that after the start of secondary growth, one ring is laid
down per day until the fiber reaches maturity (14, 20) . The width
of individual growth rings in unswollen fibers varies from 0.1 to 0.5
micron. The width of the growth rings appears to fluctuate with

t

'^"^

1

J^

Fig. 4. Longitudinal view of the layers
in the secondary wall of a swollen 51-
day fiber. Magnification X 500.

Microscopic Structure of the Cell Wall

17

variations in environmental conditions. For certain localities at
least, variations in temperature and the width of the growth rings
can be correlated (19). Thus, cotton fibers from different plants,
but grown under the same conditions, show similar rings, so that

Fig. 5. A. Mature fiber mounted in water and photographed between crossed
nicols to show the bands of extinction where the fibrillar orientation reverses.
Magnification X 180. B. A swollen fiber, photographed between crossed nicols,
showing a reversal in the direction of orientation of the first layer of fibrils
in the secondary wall. Magnification X 420.

it is possible to cross-date daily growth rings of cotton just as
annual rings are cross-dated in the trunks of certain trees.

Upon examination of single cotton fibers with crossed nicols,
much can be learned about the orientation of the cellulose. When
the fibers are placed approximately parallel to the plane of light
passing through one of the nicol prisms, the high birefringence of
the fibers is interrupted by dark extinction bands at irregular
intervals along the axis (Fig. 5, A) . These optical variations can be
correlated with differences in structure observed in swollen fibers.
Upon swelling, the latter clearly reveal a fibrillar structure (Fig.

18

The Structure of Protoplasm

5,B, and 6,A and B) . The bulk of the fiber is made up of innumer-
able fine fibrils oriented at an acute angle with respect to the axis
of the fiber. The fibrils make either an "S" or "Z" twist\ reversal of
direction (Fig. 5,B) taking place many times in a single fiber. At
least in certain types of cotton the direction of orientation of the
fibrils in the first-formed layer of the secondary wall is opposite

Fig. 6. Fibers^ free of wax and pectic substance, swollen in dilute cupram-
monium hydroxide solution. A. Showing differences in orientation between
the first layer of fibrils and those subsequently deposited in the secondary wall.
B. Irregular swelling caused by the constricting influence of the outermost
layer of fibrils which has become clumped at various points along the axis of
the fiber. Magnification X 420.

from that in layers which are laid down thereafter (Fig. 6,A) . In
other words, if the first layer of fibrils makes an "S" twist, all the
fibrils formed in subsequent layers make a "Z" twist. Under cer-
tain conditions of swelling, the outer layer of fibrils exerts a
constricting influence on the expanding inner layers of cellulose
(Fig. 6,A and B) . The fibrillar orientation in cotton seems to be
less complicated than in Valonia where the wall consists of some

' The fibrils are said to show an "S" twist if, when the fiber is held in a vertical
position, the spiral of the fibrils conforms in slope to the central portion of the
letter "S" and a "Z" twist if the spiral conforms in slope to the central portion of
the letter "Z."

Microscopic Structure of the Cell Wall

19

thirty layers, with the fibrils in alternate layers at a definite angle
to each other (22). In cotton, when the outermost layer of fibrils
changes its direction of winding, all the inner layers do likewise.
It is readily apparent that these reversals of direction of the cellu-
lose fibrils are responsible for the optical differences observed with

Fig. 7. A single swollen fiber being dissected with microneedles to show
its fibrillar structure. Magnification X 500.

crossed nicols and that the band of extinction is the place at which
the reversals occur. It can be shown, moreover, that the number of
extinction bands in a single fiber invariably corresponds to the
number of fibril reversals.

Many investigators (3, 8, 14, 15, 16, 17, 18, 22, 24) have called
attention to fibrils in the walls of widely different plant cells. There
is evidence, also, that the striations detectable in the intact wall are
related to the fibrils observed in the swollen material. A compari-
son of the work of Anderson and Kerr (2) with that of Berkley (11) ,
and the work of Astbury and Preston (3) shows, moreover, that
the cellulose molecules are oriented approximately parallel to the
long axis of the fibrils. Although a variety of physical and chemical
treatments reveal the presence of fibrils in the cell wall, probably
no method gives a more convincing demonstration of their presence
than microdissection. Upon handling cotton fibers with micro-

20 The Structure of Protoplasvi

needles, the fine fibrils which comprise the secondary wall can
easily be separated from one another (Fig. 7) .

In the cotton fiber the fibrils are grouped to give a layered pat-
tern. In an expanded transverse section they may be observed as
more or less round bodies (Fig. 3, A and B) . In the denser layers
of the wall there are, presumably, more fibrils per unit area than in
the more porous layers which are deposited during the day.
Although the secondary wall appears to be resolved into a system
of fine fibrils, it is unlikely that the fibrils are uniform and discrete
morphological units of constant dimensions. Their presence is,
however, indicative of the fibrous structure of the wall and the
manner in which it was deposited.

In the case of a heavily lignified secondary wall, it is possible
to remove the cellulose and leave a firmly coherent residue of lignin,
and conversely, it is possible to remove the lignin and leave the
cellulosic matrix. Following Frey-Wyssling's (13) conception of
the submicroscopic structure of cellulose (recently supported by
additional evidence from studies involving use of the electron
microscope (25) ) , Bailey (6) has suggested a similar configuration
for the finer visible structure of swollen secondary walls. Accord-
ing to this viewpoint the secondary wall consists of a matrix of
cellulose fibrils interpenetrated by a system of interconnecting
capillaries which commonly contain pectic substances, hemicellu-
loses, lignin, cutin, suberin, and other organic compounds. The
visible pattern of a wall may, therefore, be caused by variations in
the orientation of the cellulose fibrils, and also by variations in the
amount and arrangement of organic substances which are deposited
between the fibrils.

REFERENCES

1. Anderson, D. B. 1935. Bot. Rev. 1, 52.

2. Anderson, D. B., and Kerr, T. 1938. Ind. Eng. Chem. 30, 48.

3. AsTBURY, W. T., AND Preston, R. D. 1940. Proc. Roy. See. B 854, 54.

4. Bailey, I. W., and Kerr, T. 1935. Jour. Arnold Arboretum 16, 223.

5. Bailey, I. W. 1938. Ind. Eng. Chem. 30, 40.

6. Bailey, I. W. 1939. Bui. Torrey Botanical Club 66, 20.

7. Bailey, I. W. 1940. The Walls of Plant Cells in "The Cell and Protoplasm,"

Lancaster, The Science Press.

8. Balls, W. L. 1922. Proc. Roy. Soc. B 93, 426.

9. Balls, W. L. 1923. Proc. Roy. Soc. (London) B 72, 72.

10. Balls, W. L. 1928. Studies of Quality in Cotton, Macmillan & Co., Ltd.,

London.

11. Berkley, E. E. 1939. Textile Research 9, 355.

Microscopic StriLCture of the Cell Wall 21

12. Frey-Wyssling, A. 1935. Die Stoffauscheidung der Hoheren Pflanzen,

Berlin, Julius Springer.

13. Frey-Wyssling, A. 1937. Protoplasma, 27, 372, and 566.

14. Herzog. R. O., and Jancke, W. 1928. Zeitschr. phys. Chem. A 139, 235.

15. Hock, C. W., and Seifriz, W. 1939. Paper Trade Jour. 110, T53.

16. Hock, C. W., and Harris, M. 1940. Jour. Res. Nat. Bur. Standards 24, 743.

17. Hock, C. W., Ramsay, R. C, and Harris, M. 1941. Jour. Res. Nat. Bur. Stand-

ards 26. 93.

18. Iterson, G. van. 1933. Chem. Weekbl. 30, 6.

19. Kerr, T., and Bailey, I. W. 1934. Jour. Arnold Arboretum 15, 327.

20. Kerr, T. 1932. Protoplasma 27, 229.

21. Morey, R. D. 1933. Textile Research 3, 325.

22. Preston, R. D., and Astbury, W. T. 1937. Proc. Roy. Soc. B 826, 76.

23. Preston, R. D. 1939. Biol. Revs. 14, 281.

24. RiTTER, G. 1928. Ind. Eng. Chem. 20, 941.

25. RusKA, N., and Kretschmer, M. 1940. KoUoid. Zeitschr. 93, 163.

26. Seifriz, W. 1934. Protoplasma 21, 129.

27. Seifriz, W., and Hock, C. W. 1936. Paper Trade Jour. 102, T250.

28. Sisson, W. a. 1937. Contrib. Boyce Thompson Inst. 8, 389.

29. Whistler, R. L., Martin, A. R., and Harris, M. 1940. Jour. Res. Nat. Bur.

Standards 24, 555.

y

PROTEINS AND PROTOPLASMIC STRUCTURE

Laxjrence S. Mover
Department of Botany, University of Minnesota

It is obvious that some kinds of protoplasm, such as the chromo-
somes, have an intricate structure, but in the case of morphologically
undifferentiated cytoplasm it is not self-evident that structure is
present. It will be the purpose of this chapter to correlate recent
results in this field with what is known about proteins. No attempt
will be made here to discuss specialized structures such as the cell
membrane.

The physiological behavior of protoplasm indicates the presence
of some sort of structural continuity. Led by a desire to interpret
its properties in terms of structure, numerous investigators have
turned to the microscope for aid, yet have achieved little success.
For, under the microscope, protoplasm appears, to use the words of
von Mohl (27) , as "niemals einen klaren wasserigen Zellsaft . . .
sondern . . . eine zahfliissige, mit feinen Kornchen gemengte,
ungefarbte Masse." His description, it will be noticed, stresses its
gelatinous consistency. Later observers of living protoplasm have
emphasized the presence of emulsion droplets and granules but
have really left the subject of protoplasmic structure as obscure as
before; for it is hard to imagine how a disperse phase can furnish
the continuity needed to transmit stimuli. In any case, it has been
found that optically clear protoplasm can remain alive. Thus, an
amoeba may extend a pseudopodium free from granules. This can
be cut off with a microdissection needle yet retains its irritability
for a time (7) . Harvey (15) centrifuged sea urchin eggs into two or
more parts, one with the nucleus and the fat droplets, the other with
the pigment granules, yet each could be activated by chemical treat-
ment resulting in parthenogenetic cleavage. Thus it appears that
neither the nucleus nor the granules is needed to bring about cell
division. Hence it is generally realized that the hyaline, continuous
phase hides the answer to the problem.

The mere fact that the hyaloplasm is optically empty does not
signify that it is structureless. For instance, the nuclei in the pan-
creas of the white mouse cannot be seen in the living cells although

[23]

24 The Structure of Protoplasm

fixation reveals their presence (16). The nuclei of many other
animal cells are optically empty (7) , yet it would be foolhardy to
claim to a geneticist that no genie arrangement or differentiation is
present. Many other such cases in which the refractive indices of
cell constituents are too much alike to form visible edges can be
cited.

DOUBLE REFRACTION OF PROTOPLASM

Fortunately, there are other approaches to the problem than by
the use of the microscope alone. Of these, polarized light has been
employed with considerable success.^ It has long been known that
muscle, nerve, sperm, and certain other specialized cells are doubly
refractive (41) . Likewise, chromosomes, chloroplasts, and other cell
constituents may exhibit this property (41) . The question arises:
Is this birefringence due to the presence of anisotropic particles
(intrinsic double refraction) or is it produced by the regular arrange-
ment of isotropic rods or plates in an isotropic medium of different
index of refraction (form double refraction) ? It follows from the
theory of Wiener that the presence of form birefringence in an object
can be detected by varying the refractive index of the medium
bathing the particles, whereupon the magnitude of the double refrac-
tion changes, until, at a certain refractive index, the object appears
isotropic, i. e., dark in all positions between crossed nicols. On
raising the refractive index of the medium beyond this point, the
form birefringence reappears. If, however, the object is intrinsically
anisotropic, this disappearance of birefringence does not take place,
and the extent of the double refraction does not depend on the
refractive index of the medium. Upon the application of this test to
intracellular structures, it has been found that, although form
birefringence (generally rod birefringence) may be present, some
intrinsic double refraction may be superimposed on it. Under such
conditions, the curve of double refraction of the object vs. refractive
index of the medium reaches a minimum but does not drop to zero.
Since, in general, it is not possible to carry out this procedure on
living material and since one must assume that the imbibing
medium does not interact with the micells, its use with protoplasm
is limited. At all events, when an object appears bright between
crossed nicols it is clear that oriented anisodiametric particles are
present.

Although chromosomes, chloroplasts, and other specialized struc-
tures exhibit definite birefringence, attempts to detect double refrac-

Proteins and Protoplasmic Structure 25

tion in areas of morphologically undifferentiated cytoplasm were at
first without success. To some extent, this was due to the thinness
of the object and the consequent small magnitude of the phase
difference. Introduction of more sensitive compensators and the
use of higher light intensities has largely removed this difficulty and
has made it possible to detect double refraction even in extremely
thin strands such as spindle fibers (41). With these refinements in
technique, numerous investigators have succeeded in showing the
presence of doubly refractive elements in cyptoplasm itself, as well
as in the spindle fibers, cilia, and plasma membrane (41, 42) . Inas-
much as orientation of anisodiametric particles enhances double
refraction, it might be expected that streaming cytoplasm would
exhibit more brilliant birefringence than in the quiescent state, just
as solutions of myosin or tobacco mosaic virus become bright under
crossed nicols when streaming through a tube (9) . A consideration
of the magnitude of the shearing stress needed to produce noticeable
streaming double refraction in sols under experimental conditions
suggests that the low velocities encountered in cyclosis would hardly
be enough to cause sufficient orientation of submicroscopic proto-
plasmic particles. This seems to be the case, for such effects do not
appear to have been reported.

On the other hand, it is well known that many substances com-
posed of rod-shaped particles, e. g., a gelatin gel, can be statistically
isotropic and yield the same mean value of the refractive index in
all directions, due to random arrangement of the micells. Any force
causing orientation of the micells produces birefringence under such
conditions. Engelmann (10) pointed out long ago that apparent iso-
tropy of cytoplasm is not proof for the absence of doubly refractive
micells. Although cytoplasm may appear isotropic, pseudopodia
extended from it are often brightly birefringent (41). The birefrin-
gence persists for a short distance into the mass of the cytoplasm and
then disappears. If the pseudopodium is retracted, the anisotropy
fades away. When amoebae are rounding up before encysting, the
double refraction is much more definite (41). The spindle arising
during mitosis has been reported to be anisotropic in the living con-
dition by several observers (41). In these cases it is probable that
orientation of micells has been produced by activity.

As predicted by statistical mechanics, a system tends to assume
the most thermodynamically probable state when left to itself. In
substances such as rubber, which possess long-chain, flexible mole-
cules, the completely extended state is only one among the many

26 The Structure of Protoplasm

possible configurations of the molecule and, hence, will occur very
infrequently under these conditions (18, 21). Stretching the system
can be made to produce extension and alignment of the flexible
chains with a lessening of the probability of the system and a con-
comitant decrease in the entropy. Double refraction first appears on
deformation. When the stress is released, thermal agitation of the
different parts of the chain tends to produce more random configura-
tions and a more probable state. Contraction results as the chains
coil up. Likewise, in a system of inflexible rods, a random arrange-
ment is most probable. In either case, cooling would tend to produce
a more oriented, and warming, a more disoriented state. It is there-
fore especially interesting that Ullrich (47) has succeeded in pro-
ducing double refraction in the cytoplasm of the epidermal cells of
the bulb scale leaves of onion by cooling the tissue to 2°C. On
warming the tissue to 3° or 4°C., the effect disappeared. He reports
that death of the cells produced a sudden isotropy, so that it is
doubtful if the double refraction can be ascribed to injury.

With the exception of the cell membranes, the continuous phase
of protoplasm is aqueous. Several investigators (17, 38) have found
that water injected into cytoplasm is taken up by it without vacuole
formation. In many cases, it is probable that the high concentration
of water in protoplasm and the resultant hydration of the micells
interferes with double refraction phenomena. Rinne (39) has
emphasized that the high dielectric constant of water tends to pro-
duce a weakening of the electrochemical fields of force between
micells and reduce birefringence. In support of this, he has pointed
out that the double refraction of the myelin sheath of nerve decreases
as the water content increases, even though it is evident that the
orientation of its micells persists, as indicated by the retention of its
form. Ullrich (47) has also found that a reduction in the water con-
tent of protoplasm, brought about by plasmolysis with hypertonic
salt or sugar solutions, produced birefringence in onion cell cyto-
plasm.

ADDITIONAL EVIDENCE FOR ROD-SHAPED BODIES

The observations made with the polarizing microscope indicate
clearly that submicroscopic rods of some sort are present in cyto-
plasm, but they do not tell what the rods are or if any actual
structure is conferred on the protoplasm by their presence. Fortu-
nately, other experimental methods have been employed.

For instance, Moore (28) has found that the coenocytic plas-

Proteins and Protoplasmic Structure 27

modia of Myxomycetes could grow through pores in fine filter paper
averaging 1 u. If. however, a plasmodium was forcibly pressed
through bolting silk, the resulting pieces had to be larger than 200 u
to remain alive. Retardation of growth was noticed when the
diameter of the pores was less than 750 u. He suggests that pre-
formed rods, longer than 200 ^i and composed of submicroscopic
elements, are present in the protoplasm. Moore concludes that if
these are broken up or removed, life ceases. Centrifuging at 75,000
gravities for 5 minutes likewise retarded proliferation. This he
interprets as due to separation of these larger elements. On resting
for a sufficient time, it appeared that structure was re-established,
for proliferation began again. Here is an indication that not only
are long rod-shaped particles present, but also that they play a role
in the economy of the cell.

At once the question arises, what is the chemical nature of these
particles? Of all substances present in protoplasm capable of aggre-
gating to form such rods, the most prominent place is held by the
proteins. From the time of Reinke, all of the gross analyses of
protoplasm have emphasized the dominant position of the proteins.
With proteins making up so large a part of its substance, it would
be anticipated that they would exert a great influence upon its
characteristics. Let us compare what is known of their properties
with the properties of protoplasm.

GENERAL STRUCTURAL FEATURES OF PROTEINS

Recent work permits us to classify the proteins in two chief
groups: (1) the fibrous or extended proteins, such as wool, and (2)
the corpuscular or highly folded proteins, such as egg albumin. Mem-
bers of both groups are constructed from the same kinds of amino
acids. These appear to be joined into chains by peptide linkages to
form what Astbury (3) has designated the "backbone" structure
(Fig. 1) . While the backbone of silk is apparently fully extended, in
other protein fibers, such as hair and muscle, it is folded somewhat,
as shown by their X-ray spacings. The individual properties of
proteins are dependent on their side chains. In a fiber, adjacent
backbones appear to be held together by bonds of various strengths.
These bonds are due to interactions of the side chains (R, R', R", . . . ,
in Fig. 1) branching off from the backbone. These are of diversified
chemical nature (Fig. 2). Some are lipophilic (leucine) while others
are hydrophilic (aspartic acid, tyrosine, and lysine) . Of the latter
group, aspartic acid is an example of an acidic amino acid while

28 The Structure of Protoplasm

lysine is basic. These, and similar, side chains bestow upon proteins
their amphoteric properties. The presence of free lipophilic (hydro-
phobic) side chains in most fibrous proteins probably prevents them
from being dispersed by water.

In the corpuscular proteins, however, it appears that the folding
of the backbone is more complete, with most of the nonpolar side

R'

C. HgN COOH

/H\
HoN COOH

£

R*

H 0 R' H 0 r'* H
N C C N C C, N

/\h/\M/\h/\/h\/^

C N C C N C
R H 0 R H 0

Fig. 1. Back-bone structure produced by the union of amino acids through
the peptide linkage.

chains held inside and the polar groups sticking out on the surface.
There is no unanimity of opinion as to the mechanism of the folding
process, but it is probable that weak bonds, such as hydrogen bonds
(35) , are formed and that these hold the structure together (26) .
This configuration of these proteins appears to confer upon them
their water solubility. Recent measurements of sedimentation veloc-
ity, diffusion, viscosity, and electrophoretic mobility (1, 20, 32)
have emphasized that few of the corpuscular proteins approach
sphericity although their axial ratios are not usually very high.

It had been considered that the two groups were nearly inde-
pendent, but now it appears from investigations of denaturation and
X-ray structure that a corpuscular protein may be transformed into
a member of the fibrous group (3) . Denaturation produced by heat,
spreading on a surface, or by chemical treatment, apparently unlocks
the folded molecule so that it becomes extended (6, 25) . The slight
rise in the isoelectric point of egg albumin on denaturation (30)
indicates a change in the strength or number of the amino and
carboxyl groups, among other linkages (6, 26) . After extension has

Proteins and Protoplasmic Structure 29

occurred, X-ray photographs show crystal patterns similar to those
yielded by protein fibers.

Between the corpuscular and truly fibrous proteins are such
proteins as the muscle protein, myosin, and tobacco mosaic virus.
These, although so elongated that they exhibit streaming double
refraction, are not completely extended and disperse readily in

asporfic add lysine

COOH NHp

I J \

0 l"*^ H 0 T^'^H
N C C N C C N

/ W \ A Ah/ \ /H\ / \

C N C C N C
I HO I H 0
CH2

leucine tyrosine

Fig. 2. Portion of a hypothetical protein molecule showing typical side
chains.

aqueous solutions (9). The lability of proteins of this type has
probably retarded the discovery of many similar examples.

PHYSICAL CHARACTERISTICS COMMON TO PROTEINS AND

PROTOPLASM

Gel Formation. One of the properties of protoplasm is its capacity
to exist as a gel. It is a striking characteristic of many of the more
extended proteins that these too can form gels at low concentrations.
Staudinger (46) has emphasized that it is not enough for a particle
to be rod-shaped; to form a gel it must be able to form cross-
linkages and a net structure. Chains of polystyrene are dispersed by
organic solvents. Staudinger found that, if as little as 0.002 per cent

30 The Structure of Protoplasm

of divinyl-benzene was added to the styrene before polymerization,
cross-linkages were produced. The resultant polymer, although
indistinguishable by chemical tests from pure polystyrene, when
placed in the same solvents swelled enormously but did not dissolve.
In the case of a protein, like gelatin, gel formation depends on cross-
linkages between the main chains. These bonds may vary in
strength, from primary valence bonds, such as salt and disulfide
linkages, to van der Waals' forces (13, 35, 44) .

When collagen, the fibrous protein that constitutes connective
tissue, is heated in water, it disperses to form gelatin. Gelatin forms
statistically isotropic gels, yet, upon stretching these, the micells
become extended and parallel. A gelatin gel that has been dried in
the stretched state gives the X-ray pattern of a collagen fiber (14) .
Under these conditions, it is optically anisotropic and has a maximal
breaking strength in the direction parallel to the axis along which
it had been originally stretched. In other words, a fiber has been
produced. The modern conception of the structure of an unstretched
gelatin gel (37) would picture it as a tangle of chainlike molecules
held together at certain points where chains cross (Fig. 3a) . If por-
tions of these molecules become oriented parallel to each other,
a local anisotropy is set up (Fig. 3b) . Stretching tends to produce
a more orderly arrangement of longer portions of the chains until
the statistically isotropic areas of the gel nearly disappear (Fig. 3c) .

Seifriz (44) and, later, Frey-Wyssling (13) have proposed that
a similar structure characterizes cytoplasm, except that the cross-
linkages between the micells are constantly changing with changes
in its local metabolic states. Perhaps it is well to emphasize here
the importance of local regions in protoplasm. These become espe-
cially important when the question of shifts in bonds is considered.
It is possible, for instance, to arrive at values for the mean oxidation-
reduction potential or the pH of cytoplasm by the injection of suitable
indicators. These overall values may be of considerable use when
one cell is to be compared with another. But in the submicroscopic
realm in protoplasm, such experiments lose significance. In the
measurement of pH or oxidation-reduction potentials in nonliving
systems, a state of equilibrium throughout the system is essential.
It is obvious that living protoplasm is not in a state of equilibrium.
It undoubtedly has some regions of quite different pH from other
sections. For example, we have no right to predict from its over-
all pH value that the proteins in cytoplasm are above or below their
isoelectric points, for it is possible that local changes in hydrogen-

Proteins and Protoplasmic Structure 31

ion activity can produce a local situation completely different from
that occurring in a near-by region of the same cytoplasm. Yet it is
the local conditions prevailing in this unstirred system that can
bring about the formation or breakage of a bond between two adja-
cent protein molecules.

As local regions change under the influence of enzymatic
action, radicals break apart and join to other groups or remain in the
sol state for a time. It appears, however, that the framework can be

(a) (b) (0

Fig. 3. Gel structure, (a) Statistical isotropy in unstretched gel. (b) Effects
of moderate stretching, (c) Crystallite formation produced by extreme
stretching.

readily reformed by gelation. Just as with nonliving gels, conditions
inducing micellar orientation result in visible birefringence of the
cytoplasm.

Reactions to Shearing Stresses. Gels, in general, exhibit struc-
tural viscosity at low rates of shear. That is, their rate of flow is
not proportional to the shearing stress. When the shearing stress
becomes great enough, the structure of the gel may be broken down
so that it becomes a truly viscous sol. This has been observed with
dilute gelatin gels by Freundlich and Abramson (11) . On standing,
these sols reverted isothermally into gels; in other words, they were
also thixotropic. Not all gels exhibit this latter phenomenon, only
gels capable of forming very loose interconnections.

Pfeiffer (36) has reported that protoplasm shows structural
viscosity which depends on the rate of shear and that it, too, can
be thixotropic. His experiments were performed by sucking proto-
plasm, into a micropipette at various pressures. It has been observed
by Chambers (7) , among others, that agitation of the cytoplasm of a

32 The Structure of Protoplasm

marine ovum with a microdissection needle induced liquefaction. If
the egg was allowed to rest, it appeared to revert to its original
consistency.

Marsland, and Brown and Marsland (see, for instance, 19) have
found that the application of high hydrostatic pressure inhibited cell
motion, cell division, and protoplasmic streaming. As these properties
diminished in extent with increasing pressure, there occurred a
steady diminution in the structural viscosity of the plasmagel por-
tion of the protoplasm, until, at high pressures, complete inhibition
of these activities was attained and the viscosity curve reached a
constant, low value. When the pressure was released, the cells
reverted to normal. Marsland concludes that these physiological
functions are linked to a reversible solation-gelation mechanism in
the cell. Apparently the plasmagel of Artioeha, Arhacia, and Elodea
cells shows structural viscosity as well as thixotropy.- From experi-
ments with microdissection, Seifriz (44) has concluded that the
consistency of protoplasm is not uniform but varies from species to
species and with physiological states. Both he and Northen (33, 34)
have presented evidence to show that protoplasm can have struc-
tural viscosity. Under these conditions, anomalous viscosity and
thixotropy are strong evidence for a labile, micellar framework.

Elasticity. It is characteristic of many gels produced from rod-
shaped particles that they have rubber-like high reversible exten-
sibility, i. e., elastic properties. It was noticed by Freundlich and
Seifriz (12), for instance, that only those soap solutions which
contained rod-shaped aggregates were elastic. Gelatin (12, 14),
elastoidin (37) , elastin (21, 37) , and collagen (14, 37) , to name a
few proteins, can all be brought into an elastic state. These are
members of the fibrous or extended group.

In addition, both Scarth (40) and Seifriz (43, 44) have found that
cytoplasm is elastic. For instance, Seifriz found that a nickel par-
ticle, which had been pulled through cytoplasm by a magnetic field,
tended to revert to its initial position when the field was removed.
Scarth has pointed out that cytoplasm can be elastic even while
streaming. It was also observed by Seifriz that red blood cells on
being stretched are highly elastic. Mudd and Mudd (31) have like-
wise observed a high reversible extensibility in normal leucocytes
engaged in phagocytosis. Part of these latter phenomena may have
been due to the membrane.

As mentioned above, when rubber-like substances are stretched,
an extension of the molecular chains may take place without a change

Proteins and Protoplasmic StriLcture 33

in internal energy. If, however, the chains become oriented enough
to produce crystalHzation, the internal energy decreases, due to the
liberation of heat of crystallization (21, 37). In a normal elastic
solid, such as a steel spring, stretching results in an increase in the
internal energy by pulling the atoms out of their troughs of minimal
potential energy. Thus, while rubber becomes warmer on being
stretched, metals become cooler. This fundamental difference
enables one to detect rod-shaped molecules. It has been established
with the polarizing microscope and the X-ray that muscle owes its
physiological properties to the elongated protein molecules of myosin
(3, 37) . It is especially interesting, therefore, that muscle fibers on
contracting have been found to exhibit a thermal behavior com-
pletely similar to rubber (23). The contraction of muscle is thus
due to a coiling up of its myosin molecules.

Spinning Capacity. It is well known that many proteins possess
the capacity to be spun out into thin strands after denaturation. For
instance, Meyer and Jeannerat (22) have found that long threads
could be spun out from the viscous, liquid contents of the silk gland
of the silkworm. Although the contents of the gland were soluble at
pH 10, the threads were not. This indicates that they were denatured
in the spinning. Such threads were reversibly extensible on stretch-
ing, but if held on a stretch for 10-30 seconds, crystallization of the
micells took place and the elasticity disappeared. The stretched
threads yielded a crystalline X-ray pattern, but the liquid silk was
amorphous until it was dried. Seed globulins can be denatured in
urea solutions and can then be spun out into elastic fibers (3) . On
stretching, the fibers crystallize in the fully extended form. It is
interesting that these long chains were produced from corpuscular
proteins.

Many workers have observed that plasmolysis may produce very
long strands of protoplasm extending between the protoplast and the
cell wall (cf. 13) . After touching a plasmolyzed protoplast of onion
with a microdissection needle, Seifriz and Plowe (45) drew out
long, thin strands which were, in some cases, so fine as to be invisible,
except for globules along their lengths. When the tension was
released, the strands slowly contracted. It appears, therefore, as if
the surface of a protoplast, at least, can be spun out like some pro-
teins. Scarth (40) has likewise pointed out that strands of streaming
protoplasm, when stretched by osmotic swelling, snapped and
crumpled up like a solid thread, and then reverted to a fluid con-
dition.

34 The Structure of Protoplasm

The evidence presented so far shows a close correlation between
the behavior of protoplasm and the properties of proteins. It has
been shown that protoplasm can behave like a thixotropic, plastic
gel composed of elongated protein molecules arranged like a disor-
ganized brush-heap, except when in certain active states. Obviously,
not all protoplasm is the same; the lability of its framework permits
it to exist in extremely liquid forms.

BEHAVIOR OF SEA URCHIN EGG PROTEINS AND MYOSIN

Properties of the Proteins. With the exception of myosin, the
proteins we have discussed above have not been shown to take part
in protoplasmic structure. It is, therefore, especially interesting
that undenatured proteins involved in protoplasmic structure have
been isolated by Mirsky (24) from the eggs of Arhacia and Strongylo-
centrotus. The eggs were frozen at — 77°C., dried in vacuo, ground,
and then extracted with cold 1 M KCl at pH 7.3. Mirsky found that
about 83 per cent of the total protein in the egg could be dissolved
by this method. If, however, the eggs had been fertilized, the
soluble fraction would have been reduced to approximately 70 per
cent. In other words, about 15 per cent of the soluble protein is
rendered insoluble by the fertilization process.

Mirsky could show that the change in this protein fraction
(which may not be a single protein) was not associated with cell
division or the elevation of the fertilization membrane. The change
in solubility occurred in an interval of 3 to 10 minutes after fertiliza-
tion, with no further change detectable in the next 2 hours. It is,
however, in this interval that the first cleavage occurs. Mirsky
found that this decrease in solubility was accompanied by an
increased strength and elasticity of the egg. The processes of freez-
ing and thawing broke unfertilized eggs but did not break fertilized
eggs frozen in this time interval. The fertilization membrane was
not responsible for this phenomenon. He concludes that a skeleton
framework, capable of supporting the cytoplasm during develop-
ment, is produced by the alteration of this protein.-'

To test these results, Moore and Miller (29) investigated the
appearance of the Strongylocentrotus egg between crossed nicols.
Its optical behavior was somewhat obscured by the cytoplasmic
granules. After these had been moved to one side of the unfertilized
egg by centrifugation, it was found that the hyaline cytoplasm was
isotropic. If, however, the eggs were fertilized, the hyaloplasm
became clearly birefringent within 3 minutes. This likewise sug-
gests an orientation of structural elements.

Proteins and Protoplasmic Structure 35

Mirsky found it possible to isolate the labile protein fraction by
salting out the soluble protein from unfertilized eggs with ammonium
sulfate, whereupon the labile fraction came down first. It was then
redissolved. The physical properties of this protein in its soluble
state are precisely what might have been expected from the fore-
going discussion of protoplasmic properties. At pH 7, it appears to be
highly elongated. The viscosity of a 1.4 per cent solution was 9.6
times that of water. This value was influenced by the rate of shear,
so that solutions of the protein exhibit structural viscosity. Unlike
most of the corpuscular proteins, it showed streaming double
refraction.^

Myosin is very like this protein in many respects. Its solutions
show streaming double refraction and structural viscosity. They
can readily be brought into a thixotropic state and exhibit marked
elasticity (9, 25) . When muscle is brought into rigor, its myosin
becomes insoluble (25) . It thus appears that the physical properties
of two proteins associated with protoplasmic structure are strikinglj^
similar.

Physical State of These Proteins. Changes induced in the cor-
puscular proteins by heat, acid, etc., usually involve two steps:
(1) denaturation and (2) coagulation. In denaturing, the molecule
unfolds and the viscosity of the solution is increased. The mechanism
of this change is not clearly understood, but it is usually agreed
(26, 35) that weak bonds, such as hydrogen bonds, holding the mole-
cule together, are broken and the molecule becomes an extended
structure, approximately 10 A thick. The denatured product is not
rendered insoluble until the pH of the solution is brought into the
neighborhood of its isoelectric point, whereupon the protein coagu-
lates and loses its solubility in water (6) . This change is presumably
the result of a side-by-side association after the net charge is de-
creased. In most cases, denaturation of these proteins is accompanied
by a change in the number of sulfhydryl groups (25, 26) .

The fibrous proteins, on the other hand, are considered to be
already unfolded, i. e., denatured and coagulated. The X-ray inves-
tigations of Astbury (3) indicate that this unfolding varies in extent.
Astbury believes that keratin, as it exists in unstretched mammalian
hair, is in what he calls the a state. On stretching the hair, the
backbones become completely extended (|3 keratin) ; if, however,
relaxed hair is treated with steam, they become "supercontracted."
These states may not be produced by such simple foldings of the back-

36

The Structure of Protoplasm

bone as Astbury had originally postulated, but the true state of
affairs is probably somewhat like that shown in Figure 4.

Myosin and the protein isolated by Mirsky appear to have some
characteristics of both groups. They are much more elongated than
the corpuscular proteins, yet are soluble in water. Both show double

OC

f

collagen

supercontracted

Fig. 4. Possible folded states of the back-bone in fibrous proteins. (After
Astbury.)

refraction of flow. Both are exceedingly sensitive to the effects
of denaturing agents. Even salts, such as CaCL or LiCl, at low con-
centrations rapidly destroy the streaming double refraction of
myosin (9) .

Astbury (3) has presented evidence to show that myosin in
relaxed muscle is in the a state; on muscular contraction, he claims
it becomes "supercontracted." It is certain, in any case, that the
insolubility accompanying muscular contraction is not identical
with the denaturation of a corpuscular protein, for no alteration in
the number of sulfhydryl groups takes place (25) . On treating myo-

Proteins and Protoplasmic Structure 37

sin with acid or heat, the SH groups aUer in number (25), and
Astbury believes the extended j3 form is then produced. If so, it is
strange that the effect of these agents is to remove the streaming
double refraction (9) .

Likewise, when the sea urchin egg was fertilized, Mirsky found
no change in SH groups accompanying the decrease in protein solu-
bility (24) . These groups appeared, however, if the egg was treated
with acid. Of all the ways of coagulating myosin in vitro, Mirsky
finds that only dehydration produces no change in SH groups. Dehy-
dration readily renders the isolated egg protein insoluble. He sug-
gests, therefore, that both muscular contraction and fertilization act
as dehydrating agents which permit the protein molecules to come
close enough together to combine with each other and become insol-
uble without much intramolecular disturbance.

Banga and Szent-Gyorgyi (4) have recently proposed that,

"Whenever nature needs a mobile protein (serum albumin and globu-
lin, secretions like milk-proteins, hormones like insulin, different
enzymes, etc.) it applies the globular shape, and whenever it wants to
build a solid structure it applies the rod shape. Proteins have been found
to be mostly globular because unconsciously the rnobile and more
easily accessible proteins have been selected for study."

These authors have succeeded in dispersing some of the struc-
tural proteins from tissues like kidney and from chloroplasts and
find the extracts to be highly viscous and thixotropic and to exhibit
streaming double refraction. They believe this last property indi-
cates that these structural proteins belong to the fibrous group.
Banga and Szent-Gyorgyi used salt solutions containing 30 per cent
urea to disperse these proteins, a substance capable of inducing
marked changes in the structure of other proteins. Whether or not
these dispersed tissue proteins have the same constitution as the
native forms is questionable. They seem to be different from the
corpuscular proteins, for these do not exhibit streaming double re-
fraction after treatment with the salt-urea solution. This work is
an interesting beginning in a difficult, but important, field.

It is obvious that myosin and Mirsky's egg proteins are very
different in their properties from the corpuscular proteins. The
evidence indicates, furthermore, that they exhibit important differ-
ences from the truly fibrous proteins, in spite of the findngs of Banga
and Szent-Gyorgyi. Unlike the fibers, these intraprotoplasmic pro-
teins are characterized by an extreme sensitivity to heat or to elec-
trolytes. It is possible that they are extended to a fibrous condition

38 The Structure of Protoplasm

by denaturation. Indeed this may be the case with the proteins in-
vestigated by Banga and Szent-Gyorgyi. Perhaps these intrapro-
toplasmic structural proteins should be classified as a separate group.
At all events, it appears evident that many of the physical prop-
erties of protoplasm can be ascribed to its proteins and that these
provide a labile, structural framework within which metabolism
can take place. It has been suggested that the cell is a "protein
crystal." Although varying portions of the cell may exhibit transi-
tory birefringence, the notion of a whole cell as a crystal seems too
static. To use the words of Frey-Wyssling (13), "Das Haftpunkt-
system der lebenden Substanz ist daher nicht etwas Oegebenes, wie
z. B. bei Gelatine- oder gar bei Zellulosegelen, sondern bestandig ist
bei ihm nur der Wechsel!"

FOOTNOTES

'For a discussion of theoretical aspects of polarization microscopy, see
Ambronn and Frey (2), Schmidt (41), and Schmitt (42).

'It is suggestive that Coult (8) has found that germination of seeds is accel-
erated by shaking them. Shaken seedlings showed a significantly greater increase
in weight (wet and dry) over controls.

■The results of Beams and King (5) are interesting in this connection. They
found that fertilized Ascaris eggs retained their capacity to cleave, even after
long centrifugation at 100,000 to 400,000 times gravity. Some eggs divided in the
centrifuge. These centrifugal forces are more than enough to sediment srnall
corpuscular proteins. Beams and King conclude that Svedberg's stratification
does not take place and that either there is no need for definite spatial relation-
ships or else the protoplasm does not behave as a free dispersion of protein
particles. The latter hypothesis seems more likely in the light of Mirsky's evi-
dence for a structural framework after fertilization.

' Since the preparation of this chapter, Dr. Mirksy has brought the interesting
work of Bensley to my attention (Bensley, R. R., and Hoerr, N. L., Anat. Rec.
60:251-266. 1934; Bensley, R. R., Anat. Rec. 72:351-369. 1938). Bensley has
isolated structural proteins from the cytoplasm of liver cells in what appears
to be an undenatured state. The original articles should be consulted for
further details.

BIBLIOGRAPHY

1. Abramson, H. a., Gorin, M. H., and Moyer, L. S. The polar groups of protein

and amino acid surfaces in liquids. Chem. Rev. 24: 345-366. 1939.

2. Ambronn, H., and Frey, A. Das Polarisationsmikroskop. Leipzig. 1926.

3 AsTBURY, W. T. X-ray studies of the structures of compounds of biological
interest. Ann. Rev. Biochem. 8:113-132. 1939.

4. Banga, I., and Szent-Gyorgyi, A. Structure-Proteins. Science 92:514-515.

1940.

5. Beams, A. W., and King, R. L. Survival of Ascaris eggs after centrifuging.

Science 84:138. 1936.

6. Bull, H. B. Protein denaturation. Cold Spring Harbor Symposia on Quanti-

tative Biology 6:141-149. 1938.

7. Chambers, R. The physical structure of protoplasm as determined by

microdissection and injection. In Cowdry, E. V. General Cytology, pp.
237-309. 1924.

Proteins and Protoplasmic Structure 39

8. CouLT, D. F. Some observations on the effect of shaking on plants with

particular reference to Sinapis alha L. Protaplasma 32: 92-115. 1939.

9. Edsall, J. T., AND Mehl. J. W. The effect of denaturing agents on myosin.

II. Viscosity and double refraction of flow. Joiir. Biol. Chem. 133:409-

430. 1940.
10. Engelmann, Th. W. Contractilitat und Doppelbrechung. Pflilgers Arch. ges.

Plnjsiol. iJ: 432-464. 1875.
11 Freundlich, H., and Abramson, H. A. Uber die Thixotropie von Gelatine-
losungen. Zschr. phijsik. Chem. 231:278-284. 1927.

12. AND Seifriz, W. uber die Elastizitat von Solen und Gelen. Zschr.

physik. Chem. 104:233-261. 1933.

13. Frey-Wyssling, A. Submikroskopische Morphologie des Protoplasmas and

seiner Derivate. Berlin. 1938.

14. Gerngross, O., Herrmann, K., and Abitz. W. tjber den Feinbau des Gela-

tinemicells. Biochem. Zschr. 228:409-425. 1930.

15. Harvey, E. B. Parthenogenetic merogony or cleavage without nuclei in

Arhacia punctulata. Biol. Bui. 71:101-121. 1936.

16. Hirsch, G. C. Form und StofTwechsel der Golgi-Korper. Berlin. 1939.

17. Kerr, T. The injection of certain salts into the protoplasm and vacuoles of

the root hairs of Limnohiuvi spongia. Protoplasma 18:420-440. 1933.

18. Mark, H. The elasticity of long chain compounds as a statistical effect.

Chem. i?eu.23: 121-135. 1939.

19. Marsland. D. a. The mechanism of protoplasmic streaming. The effects

of high hydrostatic pressure upon cyclosis in Elodea canadensis. Jour.
Cell. Compar. Physiol. 13:23-30. 1939.

20. Mehl, J. W., Oncley, J. L., and Simha, R. Viscosity and the shape of protein

molecules. Scie?ice 92: 132-133. 1940.

21. Meyer, K. H. The state of aggregation of rubber and of substances with

rubber-hke extensibility. Chem. Rev. 25:137-149. 1939.

22. AND Jeannerat, J. Les proprietes des polymeres en solution. XI.

Sur la formation du lil de sole a partir du continu liquide de la glande.
Helv. Chim. Acta 22:22-30. 1939.

23. AND PiCKEN, L. E. R. The thermoelastic properties of mu.scle and

their molecular interpretation. Proc. Roy. Soc. London B 124: 29-56. 1937.

24. MiRSKY, A. E. Protein coagulation as a result of fertilization. Science 84:

333-334. 1936.

25. . Protein denaturation. Cold Spring Harbor Symjjosia in Quantita-
tive Biology 6:150-163. 1938.

26. — — • AND Pauling, L. On the structure of native, denatured, and coagu-

lated proteins. Proc. Nat. Acad. Sci. U. S. 22: 439-447. 1936.

27. MOHL, H. VON. Uber die Saftbewegung im Inneren der Zellen. Bot. Ztg. 4:

73-78, 89-94. 1846.

28. Moore, A. R. On the significance of cytoplasmic structure in Plasmodium.

Jour. Cell. Compar. Physiol. 7: 113-129. 1935.

29. AND Miller, W. A. Occurrence of birefringence in the fertilized egg

of the sea urchin. Proc. Soc. Exp. Biol. Med. 36:835-836. 1937.

30. Moyer, L. S. Electrokinetic aspects of protein chemistry. Cold Spring Har-

bor Symposia on Quantitative Biology 6:228-243. 1938.

31. MuDD, E. B. H., AND Mudd, S. The process of phagocytosis. The agreement

between direct observation and deductions from theory. Jour. Gen.
Phijsiol. 16:625-636. 1933.

32. Neltrath, H. The diffusion of proteins. Cold Spring Harbor Symposia on

Quayititative Biology 6:196-207. 1938.

33. NoRTHEN, H. T. Studies of protoplasmic structure in Spirogyra. I. Elasticity.

Protoplasvia 31:1-8. 1938.

40 The Structure of Protoplasm

34. AND NoRTHEN, R. T. Studies of protoplasmic structure in Spirogyra.

II. Alterations of protoplasmic elasticity. Protoplasma 3i:9-19. 1938.

35. Pauling, L. The nature of the chemical bond. Second ed. Ithaca and Lon-

don. 1940.

36. Pfeiffer, H. Further tests of the elasticity of protoplasm. Physics 7: 302-305.

1936.

37. PiCKEN, L. E. R. The fine structure of biological systems. Biol. Rev. 15:133-

167. 1940.

38. Plowe, J. A. Membranes in the plant cell. II. Localization of differential

permeability in the plant protoplast. Protoplasma 12:221-240. 1931.

39. RiNNE, F. Schwachung des feinbaulichen Zusammenhanges durch Wasser

und wasserige Lbsungen. Koll. Zschr. 61:304-308. 1932.

40. ScARTH, G. W. The structural organization of plant protoplasm in the light

of micrurgy. Protoplasina 2: 189-205. 1927.

41. Schmidt, W. J. Die Doppelbrechung von Karyoplasma, Zytoplasma und

Metaplasma. Berlin. 1937.

42. ScHMiTT, F. O. The ultrastructure of protoplasmic constituents. Physiol.

Rev. 19:270-302. 1939.

43. Seifriz, W. An elastic value of protoplasm, with further observations on

the viscosity of protoplasm. Brit. Jour. Exper. Biol. 2: 1-11. 1924.

44. Protoplasm. New York. 1936.

45. AND Plowe, J. Q. Effects of salts on the extensibility of protoplasm.

Jour. Rheology 2: 263-270. 1931.

46. Staudinger, H. The formation of high polymers of unsaturated substances.

Trans. Faraday Soc. 32:97-115. 1936.

47. Ullrich, H. Einige Beobachtungen iiber Doppelbrechung am lebenden

Protoplasten, an verschiedenen Zellorganellen sowie der Zellwand.
Plmita 26:311-318. 1937.

:S^

MOLECULAR STRUCTURE IN PROTOPLASM

O. L. Sponsler and Jean D. Bath

Laboratories of Physical-Biological Sciences, University of California,

Los Angeles

The principal aim in the sciences is comprehension; often a
secondary aim is control; and the more fundamental the nature of
the comprehension, the more certain is prediction and control.
With ideas somewhat like these in mind about ten years ago, we
began to turn from the molecular structure of cellulose (1) to the
possibilities of molecular structure in the components of protoplasm,
with the intention of carrying the molecular viewpoint as far as we
were able up into the submicroscopic or colloidal range and further,
if possible, into the microscopic.

With the realization that many vital activites, although not all
of them, may occur at the molecular level, we felt that attempts
which would elucidate molecular structure and spatial arrangements
would be worthwhile even though the results obtained were far
from either conclusive or complete. The structural goal towards
which we are striving is still very indistinct although the stepping
stones are surprisingly clear and well defined in many cases. It is
like climbing toward the top of a high mountain; we carefully pick
our steps that we may not stumble; we make false starts, often
finding the way blocked by a wall or a chasm, but consistently the
way leads upward, always intriguing, and keeping us wondering
what remarkable vista awaits at the top.

The comprehension we desire is of protoplasm on a molecular
level, in the range of molecular dimensions. To attain this, famil-
iarity is necessary with atoms considered as small particles; with
groups containing a few atoms, small molecules; with large molecules
containing the principal features of many small molecules; with small
submicroscopic particles, and with large particles close to the range
of visibility. In all cases we think of them as endowed with definite
properties of weight and of spatial dimensions, such as shape and
size. In addition to these they have other properties which may be
described in a general way as attraction and repulsion. These
categories are somewhat clearer when we say that the molecules

[41]

42 The Structure of Protoplasm

of the organic substances, sugar, fats, and allied compounds, are
"small molecules"; the proteins are to be thought of as "large
molecules"; while aggregates of an indefinite, although a small
number, of protein and other molecules, are considered as "small
submicroscopic" particles. The "large particle" may be composed
of many hundreds or thousands of these three. It is obvious that the
four size classes are merely arbitrary conveniences, although they
may have vague boundary lines.

The small molecules are, in size, about 5 to lOA; the large mole-
cules, about 20 to 50A; while the small particles range from 50 to
lOOA or larger; and the large particles, from say 500A to 5,000A.
Particles of 5,000A, or 0.5 micron, in diameter are at the limit of
microscopic visibility. Each of these size classes may be thought of
as furnishing structural units, or members, for particles of the next
larger class. A particle at the limit of visibility could conceivably
consist of approximately a million small 50A particles. This concep-
tion of multiple aggregates leads to the necessity for thinking in
terms of dimensional levels, since it is practically impossible to
visualize the details of one of the component molecules and at the
same time to visualize the whole structure containing a million of
these and to recognize the internal organization and many complex
details of structure which undoubtedly exist.

The particulate nature of cytoplasm is well known from micro-
scopic and ultramicroscopic observations (2, 3, 4, 5, 6) , which have
recognized the existence of particles ranging in size from the visible
far down into submicroscopic regions. Properties of these particles
are, however, very inadequately known. It seems reasonable to
think that properties, such as weight, shape, size, electrical features,
and other characteristics, must derive from their molecular compo-
sition, and that their capacity to take part in metabolic activities of
any sort is a consequence of these properties. Considerable evidence
from chemical and physical investigation (7, 8, 9) has accumulated
concerning the molecular components, especially the protein mole-
cule, which should be useful in attaining a comprehension of the
nature of these particles and their properties. A particle composed
of a thousand or more 50 A protein particles may be readily con-
ceived as having considerable internal organization and many
complex details of structure, which give it catalytic properties and
make of it a functional aggregate.

By the term, functional aggregate, we mean that the relative
arrangement of the constituents is such that certain reactions can

Molecular Structure in Protoplasm 43

occur which could not take place if these positions in space were
altered. In other words, the usefulness to the protoplasm of one
large molecule may be dependent upon the arrangement of a dozen
or more molecules immediately adjacent to it in the particle.
Organization comparable to this on a visible level may be observed
in the orderly behavior of chromosomes at mitosis; but organization
on a molecular level is not so obvious, although rather specific
arrangements of molecular structures seem to be essential in the
synthesis of glucose, and definite arrangements of enzymes are
involved in respiration.

To comprehend organization of molecules within particles of
visible size, and even of particles below visibility, seems at first
glance to be quite impossible because the number of component
molecules is so great. Fortunately, however, these molecules possess
many features in common, and many individuals of one type may
occur; as a result of this the particle as a whole, consisting of a
thousand such molecules, may appear in some aspects as a struc-
tural repetition pattern. Comprehension of one of these structural
units is a long step toward the comprehension of the whole.

We have just mentioned the possibility of a definite structural
interrelation between a dozen or more molecules within a single
particle as necessary for a particular activity. A molecular group
of this sort may be considered as a mechanism when by virtue of its
spatial configuration it is enabled to carry on a particular set of
activities, to serve a specific purpose in the cell. It is conceivable
that small mechanisms may become aggregated into still larger
composite particles. Thus several individual, molecular mechanisms
may become component parts of a single larger, more complex
machine. It is further conceivable that these micellar mechanisms
may in turn become aggregated into much larger submicroscopic
particles and that each of these may become a mechanism of great
complexity adapted to a still different sort of activtiy. In each case
the component building blocks, whether molecules or small aggre-
gates, lose their obvious individual identity when they become
attached into a larger particle.

Why then stop at a particular size? Why, indeed? A microscopic-
ally visible plastid is most certainly a mechanism since it may
function in building starch grains. It is a mechanism on a visible
dimensional level and may be thought of as built of millions of
smaller particles, but the individuality of these particles has become
lost in the formation of the larger structure. Or on a still higher

44 The Structure of Protoplasyn

level, the whole cell at mitosis becomes a single mechanism, ephem-
eral, yes, but a mechanism nevertheless, for a specific purpose, and
for a short time the particles of the cell may be considered also as
having lost their identity.

We must interpolate here, before taking up any of the specific
molecular characteristics, two points of view which are constantly
creeping into our ways of thinking of protoplasm; one is analytic in
nature; the other, synthetic. These terms are rather loosely descrip-
tive, but differences will appear as we proceed.

To understand a mechanism in protoplasm, we must first recog-
nize that it is a mechanism; and when we have broken it down to
its components, we must also, by some experimental means, recog-
nize the components as such. Thus we may step from the level of a
living structural mechanism, which we have recognized by physio-
logical experimentation, down to the level of the components, which
we study by physical-chemical methods. We assume that the
physiological processes by which we recognized the mechanism
have their origin in the physical or chemical characteristics of some
portions of the component pieces into which the mechanism has
been broken. In this analytic method we seek these active portions
and are likely to disregard the original framework. Opposed to this
analytic procedure the synthetic method starts with atoms and
molecules as building stones for the construction of larger particles
and mechanisms. Each step up is possible only because the charac-
teristic properties of the small structural building stones make
specific spatial arrangements possible. This building-up process
presupposes a knowledge of certain properties of the components
which will allow them to take up specific positions with respect to
one another, and as a result of their relative positions, willy nilly,
to form mechanisms essential to the activity of the cell. The analytic
method starts with a rather vague mechanism, known principally
by its performances, and attempts to comprehend the machine by
breaking it down and observing the behavior of the pieces which in
turn are not too well comprehended. In contrast to this the syn-
thetic method dwells more on the structure of the fairly well known
components but understands the behavior less when the components
are assembled. Obviously both methods are essential to a reason-
ably clear understanding of vital activities, but it is the latter method
which is of especial interest to us here. Of course, in any study,
these two procedures are often used interchangeably, and along

Molecular Structure in Protoplasm 45

with them argument by analogy creeps in perhaps a Httle too fre-
quently, and often in a rather far-fetched manner.

In this discussion, then, we are attempting to form a conception
of a type mechanism as viewed on a molecular level. For this pur-
pose we have chosen the only one for which there seems to be enough
information available to give us at least a tentative conception. This
is associated with respiration, and while the conception of the
mechanisms involved is still in a very hypothetical stage, an attempt
at the comprehension of them may nevertheless serve as an avenue
of approach to an understanding of the mechanisms on still higher
levels.

Investigations into the reactions and the materials necessary for
the processes or activities which we think of collectively as respira-
tion have given a fair idea of the various steps involved and of the
molecules required (10, 11) . Each step, it is thought, requires a
protein molecule with an attached small molecule; the two together
are considered as an enzyme. Many such enzymes are required for
the complete breakdown of a glucose molecule (10, 11) and there is
evidence (12, 13) that several of these may occur combined within
a single complex. That these enzyme systems are arranged in an
organized manner seems evident from the observations (14) that
certain activities are lost when the complex is disintegrated, and
that these enzyme systems are not rebuilt upon standing although
apparently all of the constituents are present.

The frequency of occurrence of these respiration complexes
locally in the protoplasm is not known, but that many thousand such
mechanisms are distributed throughout the cell seems not too unrea-
sonable. In the active cell they occur in a water medium in which
molecules and ions of various kinds are also present, and since these
may be expected to influence the complex, it may be well to first
obtain a general picture of the cell as a whole on a molecular level.

MOLECULAR CONSTITUENTS OF PROTOPLASM

From analyses of the protoplasms of many organisms the com-
ponents of active cells may be listed in a general manner as in
Table 1.

The proportions shown in Table 1 may be thought of as minima
of a sort but essential, nevertheless, for active protoplasm. By using
average molecular weights for the various constituents, a rough esti-
mate of the relative number of each may be obtained, as shown in
Table 2.

46

The Structure of Protoplasm

TABLE 1
Generalized Protoplasmic Analysis

Substance

Water

Proteins

Fatty substances ,

Other organic molecules
Inorganic ions

Percentage of
fresh weight

85-90
7-10
1- 2
1- 1.5
1- 1.5

TABLE 2
Relative Number of Molecules of the Various Types of Protoplasmic

Materials

Substance

Percentage
Used

Average Molecular
Weight Used

Approximate

Relative Number

of Molecules

Protein

10

85

2

1.5
1.5

36,000

18

700

250
55

1

Water

18,000

Fatty substances

Other organic

substances

10
20

Inorganic material . . .

100

These numerical proportions help us to grasp the cytoplasmic pic-
ture somewhat better since we may now think of one large protein
molecule surrounded by about 20,000 minute water molecules in
which the remaining relatively small molecules occur.

To make this still more vivid, dimensions of these molecules and
their shapes are needed. These may be obtained by recourse to two
fields: to the chemical, which furnishes the structural formula and
the interrelationships of the atomic components; and to the physical,
which supplies the size and valence angles of the respective atoms.
From this information it is possible to determine the dimensions of
the molecules and also to obtain a great deal of information concern-
ing their spatial relations and other properties of their constituent
atomic groups (21) . Space limits us at this time to brief descriptive
notes of only the most prominent spatial characteristics of these
molecular structures.

STRUCTURAL FEATURES OF COMPONENT MOLECULES

Of the component molecules in protoplasm, water is the smallest
and apparently the simplest in structure. It has, however, charac-

I

Molecidar Structure in Protoplasm 47

tertistics peculiar to itself which detract somewhat from its sim-
plicity, for, while each water molecule consists of one oxygen and
two hydrogen atoms, the atomic nuclei and constituent electrons are
so arranged that two residual negative electrical fields and two
positive fields occur at tetrahedral points in the molecule (15) . These
are effective in making the water molecule a permanent dipole (16) .
The more prominent features are indicated in the diagrams of Figure
1 and are mentioned because these characteristics are constantly
involved in the structure and behavior of protoplasm as well as in
the aggregation of the water molecules themselves (17) . The radius
of the spherical structure shown in Figure l.A, with two great circles
at right angles to each other, is about 1.38 A. At B the effect of the
intermolecular plus-minus attractions is shown as in ice where four
surrounding water molecules are coordinated tetrahedrally with a
central molecule (15. 18). Such groups vary in amount with the
temperature; and in themselves vary in number from one and two
at the boiling point to a group of several or many near freezing (17) .

When a portion of the water molecule occurs as a hydroxyl
group, — OH, on an organic molecule, one of the plus residual
points is missing due to the formation of the primary valence bond;
this leaves only three points for coordination with water molecules in
the surrounding solution. It is due to the nature of the oxygen atom
(19) that these negative residuals occur. Only one other atom, which
is of interest to us, is similar in this respect; this is nitrogen (19),
which has, however, only one negative residual charge. Thus, what
is said of oxygen and OH groups will apply to a considerable extent
also to NH and NH. groups. Residuals of this nature are completely
lacking in the carbon atom and in CH- groups.

There seems to be scarcely any need of mentioning here that the
four atoms, C, H, O, and N, are the principal components of proto-
plasmic molecules and that they make up nearly 99 per cent of
active protoplasm. One feature of these atoms which is less com-
monly known, although characteristic of the three larger atoms,
C, O, and N, was shown in Figure 1. This is the angle at which bond
formation occurs. This angle made by adjacent primary valence
bonds is practically 109° and is spoken of as a tetrahedral angle
(19) . Its occurrence will be noted in figures of the models presented
here wherever single primary bonds exist. The angle is relatively
rigid and therefore determines the shapes of molecules to a great
extent, as may be seen later in the models mentioned where the four
bonds of carbon reach out to form a tetrahedral structure; simi-

48

The Structure of Protoplasm

larly for the three of nitrogen and its single negative residual; and
also the two bonds of oxygen with its two negative residuals. When
a double bond occurs between atoms, this bond (19) makes an angle
of approximately 125° with the other single bonds present. It must
not be overlooked, however, that rotation on primary valence bonds

Fig. 1. A. Diagram to show tetrahedral distribution of the two hydrogen
nuclei and two negative residuals in a water molecule. The distance between
the oxygen and hydrogen nucleus is approximately 0.96 A, and the two hydro-
gen nuclei make an angle of about 105° with the oxygen nucleus (15). B. Dia-
gram showing coordination of several water molecules around a central one
resulting from the attraction of the negative residual of one molecule for the
hydrogen nucleus, or proton, of another (15).

may occur (19) so the shapes indicated may vary somewhat from the
model photographs shown.

The radii of the various atoms are included in Table 3. An
additive law is effective for these; thus when two atoms, such as

TABLE 3
Atomic Radii (19)

Atom

Radius for
Single Bond

Radius for
Double Bond

0.77 A

0.69 A

Nitrogen

0.70

0.63

Oxygen

0.66

0.59

Hydrogen

0.29

Molecular Structure i7i Protoplasm 49

carbon and oxygen, are joined by a single bond as in Figure 2, the
distance between their atomic centers is the sum of their single
bond radii, i. e., 0.77 + 0.66 = 1.43 A.

It thus appears that protoplasmic molecular structures have a
certain basic similarity regardless of the kind of atoms involved.

N O H

XX a

RADIUS 0.77A 0.7 1 A 0.66 A 0.29 A

BONDS 4 3 2 1

C-C C-N C-0 C-H

DISTANCE 1.54A 1.48A 1.43A I.06A

Fig. 2. Diagrams showing spatial properties of carbon, oxygen, nitrogen,
and hydrogen atoms. Lower portion shows interatomic distance between two
atoms joined by a single bond.

since the values of the angles and of the radii are practically the
same in all. Structural and cohesional differences arise, however,
due to the residual negative charge on the oxygen and nitrogen
atoms and to the absence of these residuals on the carbon atom.
These residual forces, it will be seen later, are responsible for the
formation of the hydrogen bond (19, 20) , a mechanism of consider-
able importance in the cohesion of molecules. In this mechanism
the proton, or hydrogen nucleus, of an — OH or — NHo group is
pulled toward the negative residual of another oxygen or nitrogen
atom. This attraction results in a closer approach (2.8 A) of the
two groups than would otherwise occur (3.5-4.0 A) . The hydrogen
proton, as shown in Figure 3, is always involved, and for this
reason the bond is spoken of as the hydrogen bridge.

The component of protoplasm which is universally recognized
(21, 22, 23, 24, 25) as being of most importance structurally is
protein. Various investigators using microchemical methods have
demonstrated its existence in cytoplasmic particles (26, 27, 28, 29,
30) ; and many other workers (31 and 64) have demonstrated the
importance of the proteins in respiratory mechanisms. At this point

50

The Structure of Protoplasm

the dimensional properties are of chief interest to us. It has become
well known that although many different kinds of proteins are
recognized they are all built on practically the same structural
scheme. Since the amino acids are all of the alpha form, they link
together to form a long peptide chain in which the adjacent amino
acid residues alternate from one side of the molecule to another.
The model in Figure 4 represents only a small portion of such a

chain but nevertheless shows
its characteristic features;
that is, the central zig-zag
backbone portion and the
various side chains, or amino
acid residues, which extend
out at right angles to it. The
backbone feature is the fea-
ture which is common to all
proteins, while the kind and
arrangement of the amino
acid residues give the protein
its specific qualities.

The dimensions of this
basic, structural chain are
nearly the same for all of the proteins (32, 33) . Along the backbone
each amino acid residue is allotted about 3.5 A or slightly less. As a
result of this, the length of the chain depends directly upon the total
number of residues. Thus, a chain of 300 residues is likely to be about
1,000 A long. The width varies somewhat depending upon the size
and the proportional content of the various residues. The shortest
residue, glycine, extends scarcely an Angstrom from the backbone;
while the longest, arginine, or perhaps tryptophane, may extend 6
to 8 A. The width of the chain as a whole, then, depends upon the
weighted average for the 300 or so constituent residues of various
lengths, and while it is slightly different from protein to protein,
depending upon the percentages of the various short and long
residues (32) , the width is, despite this, fairly consistently close to
10 A, rarely more than 12 A, and seldom less than 9 A. The thick-
ness of the chain through the backbone and normal to the 10 A
width is consistently 4.5 A for all of the many proteins which have
been measured (32) . A few are 4.4 A, and an occasional one is
4.6 A. From these values a generalized conception is obtained of
the chain dimensions, 4.5 A x 10 A x (3.5) 7i, where n is the number

H-1.1A

Fig. 3. Diagram of hydrogen bridge be-
tween two water molecules. Note attrac-
tion of residual negative charge on one,
for hydrogen nucleus, or proton, on the
other, and the separation distance of 2.8A.

Molecular Structure m Protoplasvi

51

of amino acid residues. Thus a protein of molecular weight 36,000
with about 300 amino acid residues of average molecular weight,
120, will have a chain length of about 1,000 A, a width of about
10 A, and a thickness of 4.5 A.

Although these generalized chain dimensions are well enough
established, the configurations which the chains may take and the
resulting sizes, shapes, and structural details in the formation of

^-tf^i

%

Fig. 4. Model photograph of a portion of a protein chain, made to scale
using known atomic radii and bond angles. Black balls indicate carbon atoms;
dark grey balls marked X indicate nitrogen atoms; light grey balls marked
with a dot indicate oxygens; small white balls, hydrogen atoms.

submicroscopic particles is less satisfactorily known. Practically all
of our knowledge concerning these comes from investigations of
proteins in bulk which have been treated in various ways for pur-
poses of purification. Material of this sort has been subjected to
many experimental procedures in attempts to determine the exist-
ence of elementary protein particles, or molecules, and their sizes,
shapes, and structural details.

Investigations as to size or sizes of the elementary particles have
been made by means of osmotic pressure and diffusion measure-
ments (34, 35), analytical methods (37), the ultracentrifuge (9,
38), and determination of the unit cell dimensions in crystalline
proteins (36 and 39). From this work it appears that the protein
particles may belong to groups of molecular weight 17,500, or some
multiple of this value, as 35,000 or greater; or submultiple, as 9,000.

52 The Structure of ProtoplasTn

Concerning the shape, a number of different experimental
methods show that many elementary protein molecules are globular
in outline rather than elongated as a completely extended chain,
although globular in this instance may mean anything from spherical
to considerable elongation. These methods include studies of vis-
cosity— concentration curves (40) , comparison of ultramicroscopic
observations (41) , determination of asymmetry by means of the
ultracentrifuge (9) , unit cell determinations of crystalline proteins
by means of X-ray analysis (36 and 39) , comparisons of double
refraction of flow (42) , and rotation in an electric field (43) . Certain
proteins, such as silk and gelatin apparently have their chains fully
extended for considerable lengths; others form particles having
various length-thickness ratios (44) ; and a few are practically iso-
diametric. A single chain of 1,000 A in length is about 100 times
as long as it is thick. Zein forms particles which are about 20 times
as long as their width; erythrocruorin, about 10 times; and insulin,
about 2 or 3 times; while for egg albumen the particles are nearly
isodiametric (44) . One of the most significant points, in addition to
the globular nature of many protein molecules, is that proteins
belonging to the same weight class may differ enormously in shape
(44).

Regardless of shape or size the particles seem to consist of poly-
peptide chains folded in some manner to give the respective shape.
Very little is known concerning the exact configuration which the
chain takes, but it seems probable that it is a function of the nature
and distribution of its side chains (45) and of its environment at the
time of synthesis (46) . Whatever model is accepted, the general
configuration of the protein chain in the globular particle must be
consistent with the observations that reflections from parallel
packets of protein chains may be obtained fairly easily from the
protoplasmic material upon the removal of water (33, 47, 48) . A
working model has been proposed (46, 49, 50, 51) for the configura-
tion of the chain in the globular particle. The essential feature of
this hjrpothetical configuration is that the peptide chain occurs in
parallel folds forming a monolayer in which the residues extend out
approximately perpendicular to the surface, thus making the layer
about 10 A in thickness. Packets containing about 1,400 of such
monolayers, piled one upon another, have been prepared, and the
individual monolayer thickness of lOA has been found by direct
micrometer measurement (50 and 52) . These monolayers may
either attach to one another to form a layered globular particle, or

Molecular Structure in Protoplasm 53

an individual monolayer may become folded to form a three-dimen-
sional particle. Such particles are probably held together by cystine
bridges, ionic linkages, or hydrogen bridges. The specific configura-
tion within the globular particle might be quite irregular and
depend upon the distribution of cysteine and ionizable residues and
on the placement of proline and short residues which make folding
possible, although the general plan might be similar to this hypo-
thetical model.

The features of the protein particle, size, shape, and configuration
of chain, just given, were taken from the results of work on proteins
in vitro. So far as the elementary particles in protoplasm are con-
cerned, reasoning by analogy, we may expect to find protein particles
of several shapes on this level where 1,000 A lengths or smaller are
under consideration; that is, as extended chains, as cylindrical or
globular structures, and possibly even as flattened discs.

A clean-cut parallelism of the chains as indicated for the mono-
layer has not been demonstrated for the particles in protoplasm,
although a certain degree of parallel arrangement is indicated by
X-ray (33) and polarized hght investigations (53). For convenience
in discussion we are assuming a model which has a greater degree
of parallel placement of the chains than is justified by experiment,
feeling that the convenience outweighs the chances of serious
misinterpretation. Also for convenience we are making use of a
cubical shape for the elementary protein particle as one which may
be more easily treated.

A chain of approximately 1,000 A in length, folded into a globu-
lar particle in the manner indicated above, may consist, for example,
of 20 folds, each about 50 A in length, arranged into three layers
of about seven in a row, as in Figure 5, making a packet 30 x 30 x 50
A of solid protein. The same chain length would make a block of
10 folds 100 A long but would be only about 25 A thick; that is,
about one-fourth its length. If, however, the particle consisted of
only four pieces of 250 A, its thickness would be about 15 A, or its
length would be nearly 20 times as great as its thickness. These are
comparable to the particle shapes mentioned as obtained experi-
mentally with proteins in vitro.

Particles such as these in protoplasm would no doubt take on
water molecules through hydrogen bridge formation wherever
oxygen- and nitrogen-containing groups were spatially available, as
at the ends of polar residues and along the backbone (54) . In a
hydrated particle where 40 per cent water or more is included, there

54

The Structure of Protoplasvi

I I B

->icQlQ5Ak-

■^**^%M''^*'^if^-

■^^^^-"^•^^M-^-

i_.

ca4.5A

'v^^lltiM -***^1iti»^- •'^^"^liti^

A

c

Fig. 5. Model photographs showing three faces of a packet of parallel pro-
tein chains. A. Three rows of seven chains each showing ends of the chains.
Note flat shape of individual units when viewed in this direction; width ap-
proximately 10.5A, thickness about 4.5A. B. Top view of same packet. C. Side
view.

is a strong probability that water molecules, by means of hydrogen
bridges, may form a seam between adjacent backbones (54). The
spacing may then become about 7 A instead of 4.5 A as when dry.
The hydrophilic end groups may furnish loci for hydrogen bridges
between the layers within a particle and also between particles.
Here the distribution of the oxygen- and nitrogen-bearing residues
is effective in furnishing few or many bridges according to the

Molecular Structure in Protoplasm 55

patterns formed by these residues on the surface. The presence of
the water molecules within the particle may spread the chains lat-
erally from the 10.5 A distance when dry to perhaps as much as
15 to 20 A when water is plentiful as in an active cell. This would
enlarge the protein packet just mentioned to a cube of about 50 A
on an edge.

Although the particles may appear regular in diagram, they will
no doubt be definitely irregular in outline due to the variable lengths
of the amino acid residues and perhaps to the uneven length of the
chain folds. Details of the structural features will depend to a
considerable extent upon the proportion and distribution of the
residues on the chain, as well as upon the number of chain folds
in the particle.

We turn now to a more intimate consideration of the residues
and their influence upon the formation of these larger structures.
There are about twenty different amino acids obtainable from prac-
tically every protein, and one protein differs from another in the
proportion of the amino acid residues it contains. Thus while gelatin
contains 25.5 per cent glycine (55) , casein has only 0.5 per cent (56)
and so for the variations in quantity of all twenty. While the amino
acids themselves have a generic similarity, in that each has a group
consisting of a carboxyl and an amino group attached to the alpha
carbon, they differ specifically in the remainder of the molecule (57) .
In general, in the case of cytoplasmic proteins, at least one-third
of the total residues are polar; the remaining two-thirds have
residues which contain neither oxygen nor nitrogen atoms and thus
are comparable to fatty substances in their behavior toward water.
In contrast to this about one-third of the amino acid residues have
either OH or COOH as oxygen-containing groups, and NHj or some
other nitrogen-containing group such as the guanidine group or the
imidazole ring. Those having an affinity for water coordinate with
various, but rather specific, numbers of water molecules; for
example, OH and NH- may each form hydrogen bridges with three
water molecules (54) .

In the formation of larger protein structures, ionization plays
an important part. From the amino acid analyses (8) of various
proteins whose residues are approximately one-third polar, it is
possible to ascertain that approximately 85 per cent of these polar
residues are ionizable at pH 7, the pH of cytoplasm; that is, about
one-fourth of the total residues on the chain or at least 60 out of

56 The Structure of Protoplasm

300 residues are capable of existing ionized at this concentration
of hydrogen ions.

The manner in which the end groups of the various amino acid
residues become ionized is indicated for the two general types.

A. The acidic groups lose a proton and thus bear a negative
charge. They include glutamic and aspartic acid residues.

— COOH ^ H H COO

B. The basic groups gain a proton and thus carry a positive
charge. Three types are of importance.

1. The amino group of lysine.

— NH, + H' ^ — NH.r

2. The guanidine group of arginine (19).

H H H +

N N

H II H + H II H

— N— C-N +H — N-C-N
H H

3. The imidazole ring of histidine (59) .

-C = CH -C = CH

I I + I I +

HN N + H ^HN NH

\ // \ //

C C

H H

Studies (58) correlating" the range of hydrogen-ion concentra-
tion in which these groups ionize (7) with the resultant hydrogen-
ion concentration of cytoplasm (23, 60, 61, 62) , indicate that glutamic
and aspartic acid, as well as arginine and lysine, are probably com-
pletely ionized at the pH of cytoplasm. Histidine may be expected
to be only partially ionized, and tyrosine, serine, and the amide
groups of asparagine and glutamine probably not at all (7) . The
relative number of positively and negatively ionized residues, as
well as their spatial distribution (63) , may be expected to have a
great influence upon the electrical field surrounding the elementary
protein particles. Some of these ionized groups will be distributed
throughout the interior of the particle, perhaps forming ionic link-
ages which may help to maintain the specific configuration of the
chain (46 and 51); other ionized groups may exist on the surface
of the particle. A particle, such as in Figure 5, consisting of three
layers, will have six faces over which charged groups may be dis-
tributed, and these should have at least six to eight ionized groups
on each face; some plus, some minus. In the model at least four of

I

Molecular Structure in Protoplasm 57

the six surfaces would contain ionized residues available for the
attachment of small organic molecules.

The interaction of one charged particle with another in its imme-
diate environment depends largely upon the pattern of the electrical
field surrounding it. The magnitude of this field diminishes with
distance. It is possible (58) , by arbitrarily placing a single positive
charge at various short distances out from the charged groups on
the surfaces, and by making use of Coulomb's law (65) in which
the force of attraction and of repulsion varies inversely as the
square of the distance, to gain some notion of the variation of the
resultant field around a particle. In general the resultant effect
is such as to give strong local areas of positive and of negative
fields close to the surface of the particle, while weaker resultant
fields occur farther away (58). The net charge may be effective
only at distances of 50-100 A and may be either positive or negative.
It may be thought of as the algebraic sum of the positively and
negatively ionized residues.

An additional feature of the protein molecule which may prove
to be significant when considering attachment of other molecules
is that concerned with residue arrangement. The frequency of
distribution along the chain of polar residues, such as lysine with
an amino group and glutamic acid with its carboxyl group, has a
strong bearing on the location of these groups on the surface of the
particle. It has long been realized (66) that, given twenty different
kinds of amino acid residues with no restrictions as to the relative
number of each, it is possible to obtain millions of different proteins
on the basis of composition alone without considering the number
of additional types possible when their arrangement is also varied.
Recent experimental studies (67, 68) of quite homogeneous pro-
teins, using carefully tested methods of amino acid analysis, seem
to indicate that the residues do not occur in random amounts but
instead are present to the extent of 1/2, 1/3, 1/4, 1/6, 1/8, 1/9, 1/12,
1/16, 1/18, 1/36 . . . etc., of the total residue number. Although
this generalization may be based on insufficient experimental evi-
dence, it furnishes an indication that some numerical rule may be
involved in the synthesis of the protein molecule in the living cell.
This observation has led to the hypothesis (67, 68) that the various
amino acid residues may occur at constant intervals along the
protein chain. Thus if there are 24 glutamic acid residues in a chain
of 288 residues a uniform spacing along the chain will place one at
every twelfth residue position; if there are 32 lysine residues,

58 The Structure of Protoplasm

every ninth will be of lysine; and so on for the various amino acid
residues.

While this notion of residue distribution still needs more sup-
porting evidence, there seems to be at least a probability, from the
biological point of view, that a certain amount of orderliness must
prevail in the formation of proteins. We are accepting this as plaus-
ible and are making use of it in a general way since it serves as a
basis for the arrangement of the residue end groups into patterns
or mosaics both on the surface and inside of the particle. Thus, if
it is necessary that the glutamic acid and lysine residue endings
occur at a specific distance from each other in order that a fatty
molecule or a respiratory prosthetic group may be attached, their
distribution may allow this to occur only once on a surface or at
most only a few times.

Of the molecules occurring in cytoplasm which are thought to
play an important part in the structural features of protein aggre-
gates the fatty materials stand out as the most prominent. They
occur in active protoplasm in quantities of at least ten to fifteen
molecules for every protein molecule of 36,000 molecular weight.
They form a miscellaneous class in which the common feature is
the excessively large amount of carbon-hydrogen groups and the
relatively infrequent occurrence of oxygen and nitrogen atoms.
The significance of these proportions of component atoms comes
out when it is recalled that the oxygen and nitrogen atoms have a
strong affinity for water due to their negative residuals, while the
carbons are lacking in this attraction for water. The resulting
insolubility in water has led us to think of these materials as fatty
substances.

The long-chain fatty acids, the neutral fats which are glycerol
esters of the long-chain fatty acids, the phospholipids, and the sterols
are included in this group. Of these the two most constant and
ubiquitous in occurrence in cytoplasmic material are the phospho-
lipids and sterols (69, 70, 71) . At least three methods indicate that
these materials may be found associated with protein material
within the heavier cytoplasmic granules. The methods include
solubility or dispersal of the large particles in fat solvents (72, 73,
74, 75, 76), blackening with osmic acid indicating the presence of
unsaturated fatty acids (77) and detection of the aldehyde grouping
of the acetal-phospholipids (78) .

The phospholipids include two general groups, the ester-phospho-
lipids represented by lecithin and cephalin and the recently discov-

Molecular Structure in Protoplasm

59

ered acetal-phospholipids (78) . Lecithin may be thought of as a
modified fat in which one of the three fatty acid radicals is replaced
by phosphoric acid plus a nitrogenous group, choline. Figure 6

shows a model of lecithin with its
characteristic polar and nonpolar
regions. Approximately half of the
phospholipids found in typical animal
cells are lecithins. The other half is
composed largely of cephalin, which
differs from lecithin mainly in the fact
that the choline portion is replaced by
colamine, or amino-ethyl alcohol.

Concerning the ionization of the
phospholipids it has been pointed out
(79) that lecithin and cephalin ionize
as shown in the diagram in Figure 7.
The isoelectric point of the former is
about pH 6.7 (80) . It seems probable
that in the case of lecithin both groups,
phosphoric acid and choline, are ion-
ized at the pH of cytoplasm (79, 81, 82,
83) . Cephalin, on the other hand, has a
weaker basic group (84) and may be
only partially ionized in cytoplasm. This makes it possible for certain
of the cephalin molecules to exist with a negative charge.

For many years organic chemists (71) struggled with a large
group of fatty compounds which apparently bore some structural

FiG. 6. Model photograph of
a lecithin molecule. Length
parallel to hydrocarbon chains
is equal to approximately
25A.

1

1

1
o=c

1

1

c=o

1

1

0

1

1

0 H

1 1

0 H H
II 1 1

— P — 0 — c— c
1 1 1

1
H-C-
1

1 1

-c — c —

1 1

0

1
H

H H

1 1 1
oe H H

I

R,

B

o=c c=o

I I

O O H O H H

III 'I II 0

(chX H— C — C-C — 0-P-O-C — c-nh,

III I II

H H H

,0

H

Fig. 7. Diagrams to show ionization of lecithin and cephalin. A. Zwitterion
form of lecithin. B. Zwitterion form of Cephalin.

relation to one another even though they were widely different in
their physiological effects (sex hormones, etc.). It is due to the
efforts of the X-ray analyst, however, that this work crystallized into

60

The Structure of Protoplasvi

our present understanding of the large group of compounds known
as the sterols (85) . The common feature of the sterols is now known
(71) to be a complex ring system plus miscellaneous small side
chains forming, in general, a rather flattened molecule about 5 A
in thickness, 7-7.5 A in width, and approximately 20 A in length
(85) . Minor structural variations give rise to the various sterols
of which perhaps the best known are cholesterol (86) , shown in

Figure 8, and ergosterol. They are found in both
plants and animals; those in the higher plants are
generally termed phytosterols. At the present
time very little is known of the specific role, either
structural or physiological, of the sterols in the
living cell.
^ »s In addition to the proteins and fatty substances

J#-r* which occur in relatively large amounts in cyto-

V^ "A^^— plasm, a small quantity of organic molecules of
V^IH^H great diversity play an important part in the
activities of the cell, although in the structural
framework they are insignificant. A considerable
number of them are known to be involved in
respiration, and when attached to proteins as
prosthetic groups form the active part of cellular
mechanisms, or enzymes (87) . Most of these have
been shown (64, 88, and 89) to occur in conjunc-
tion with specific proteins which appear to be
albumins and globulins.

The existence of cytoplasmic granules which
show intense respiration and hence contain
respiratory prosthetic groups has been amply demonstrated in
preparations from materials such as Arbacia eggs (126, 127) , liver
tissue (128) , pig's heart muscle (12) , and the breast muscle of the
pigeon (13). It has been reported (111) that similar material, con-
taining lipoids in addition to protein and respiratory prosthetic
groups, has been isolated from normal chick embryo. The associa-
tion of these three materials in cytoplasmic aggregates seems to be
more generally recognized as experimental work progresses (111) .
In order to show the diversity in structure of these small mole-
cules, a list of the better known is given, with here and there
photographic reproductions of molecular models and brief discus-
sions concerning the individual dimensions and properties. The
model photographs in Figures 9-12 inclusive are made to the same

Fig. 8. Photo-
graph of model of
Cholesterol. Black
balls indicate car-
bon atoms; grey
ball at bottom,
oxygen atom; and
small white ones,
hydrogen.

I

Molecular Structure in Protoplasm

61

scale as the portion of the protein chain in Figure 9D in order that
dimensional comparisons may be made.

The most natural classification of these small molecules is that
based on structure. The list may be a bit misleading for it is not
known whether all of these occur in all living cells or in what
amounts the various forms are associated in a single cell. The list
is given on the assumption
that many of them occur
in every living cell and
that the list is not ex-
haustive (88, 89).

They may be sorted out
into about six general
classes. The first three in-
clude the alcohols and car-
bohydrates, the aldehydes,
and the organic acids, such
as ascorbic (90) , succinic,
malic, pyruvic, lactic, and
amino acids (Fig. 9) .

The fourth group con-
sists of larger molecules
called nucleotides, which
are composed of phos-
phoric acid, a sugar, and a
nitrogenous base arranged
in the order given (8) .
They may occur as single

forms, such as riboflavinphosphate (91) and adenosine pyrophos-
phate; as dinucleotides, such as flavine-adenine dinucleotide (92)
and diphosphopyridine nucleotide (93) ; and as tetranucleotides

(94) , such as nucleic acid. These vary mainly in the nitrogen-
containing groups (Fig. 10) .

The fifth group includes the metal-containing prosthetic groups

(95, 11) , such as cytochrome, containing iron (96) ; chlorophyll, con-
taining magnesium (Fig. 11) ; and certain copper-containing groups

(95).

The sixth group consists of molecules containing sulfur, such as
glutathione (88) and thiamine pyrophosphate (97, 98) (Fig. 12) .

This partial listing and brief discussion of the fatty materials
and smaller molecules which act as prosthetic groups of enzymes

Fig. 9. Photographs of models of various
types of small organic molecules. A. Ethyl
alcohol. B. Ascorbic acid. C. Succinic acid.
D. A portion of a protein chain made to
same scale as the small organic molecules.

62

The Structure of Protoplasm

give us a somewhat broader view of the molecular constituents
of the particles in cytoplasm. Of the dozen or more prosthetic
groups, each has its own specific protein molecule, at least in vitro,

,^♦

Fig. 10. A. Adenosine triphosphate, a nucleotide plus two phosphoric acid
groups. B. Flavine-adenine dinucleotide. C. Nucleic acid, a tetranucleotide.

to which it may be attached. If we think of the attachment as being
associated with the patterns or mosaics formed by end groups of
particular amino acid residues on the surface of the protein particle,
then a specific protein may mean that in order to fit a prosthetic
group of one sort the configuration of the chain is different than
when it fits another prosthetic group. In other words it may not
mean that the amino acid composition of the chain is necessarily
different in order to have specific properties but that merely a
difference in configuration or arrangement of the chain lengths in
the particle would suffice, since this would alter the patterns formed
by the residue end groups.

With these conceptions of the spatial and electrical properties of
the various constituents of cytoplasmic particles, we see the need
for restrictions or limitations on the infinite number of ways in
which these materials could aggregate to form cytoplasmic particles;
without these a future comprehension of the larger structures seems
practically impossible. This aggregation, it has been pointed out
(99) , may be brought about by means of three sorts of bonding
forces: primary valence bonds, such as cystine bridges; bonds due
to electrostatic forces, such as charged groups, or ions; electrostatic
effects as observed in hydrogen bridges; and van der Waals'

4

Molecular Structure in Protoplasvi

63

cohesion forces. The most important of these are the electrical
effects arising from the existence of ionized groups, since they may
act at considerable distances as well as at close range, while the

9

r A>

Fig. 11. A. Portion of the prosthetic group of Cytochrome c. B. Chlorophyll
A. The large central ball in the porphyrin ring represents iron in Cytochrome
and magnesium in Chlorophyll.

other types are effective mainly at short distances, a matter of 1 to
about 5 Angstroms.

Here, again, we come to the borderline between the physical-
chemical and the biological viewpoints. Aggregates formed by all
possible combinations of these various molecules would produce
merely a heterogeneous mixture in the cell, while opposed to this,
organization is characteristic of the living protoplasm. Nevertheless,
the various means of attachment, no matter how they are brought
about, are undoubtedly the same in living as in inert matter; and

Fig. 12. A. Thiamine pyrophosphate (97, 98). B. Glutathione, glutamyl-
cysteinylglycine (88). The very lightly-colored ball near the middle in each
model represents a sulfur atom.

while it seems probable that the discrete molecules and particles,
such as we have selected from the proteins, fats and other organic
molecules, may not take exactly the same relative arrangement in
space in vivo as they do in vitro, comprehension in the latter seems
likely to aid in an understanding in the former.

To this end a few points concerning the attachment of various
combinations of molecules is introduced. We need not dwell on the

64 The Structure of Protoplas^n

cross-linkage of a residue on one chain to a residue on an adjacent
chain by means of primary valence bonds. These are merely
chemical bonds such as in any molecule. We are more interested in
what may be termed cohesion bondings through electrostatic and
van der Waals' forces. In all cases of combinations of molecules we
are thinking of the protein as being especially prominent, and there-
fore we are concerned with the surfaces or faces of the protem
particles.

It will be recalled that certain faces of the 50 A protein molecule
may have as many as fifty to sixty residue end groups exposed,
forming some sort of mosaic pattern. On this only about ten to
twenty are likely to be polar, and perhaps even fewer are ionized.
Thus the face will have patterns also based on ionized groups. Some
of these will be positively, some negatively, charged. The effect of
these charges is strongly localized at a short distance above the
surface, and resultants of these begin to appear as the distance from
the face increases, first in accordance with the arrangement of the
charges with respect to one another on the surface; and farther out,
perhaps at 50 A or more, the resultant effect is a blend of all charges,
internal as well as those on the surface, or is, in effect, the net
charge, or the algebraic sum, of all of the charges on the particle.
Interaction of two such particles may occur, providing their net
charge is opposite in sign. The resultant attraction of unlike and
the repulsion of like charges should produce a mutual rotation of
one particle with respect to another until their most attractive sur-
faces become adjacent. The surfaces should then turn until they
reach a position of best fit (99) . There is some doubt as to whether
the particles would be able to make direct contact because of the
possible presence of a 10 A layer of "electrically saturated" water
(16) around the ionized groups on their surface; that is, water
which is completely polarized and oriented. As the two oppositely
charged particles approached each other, they would reach a point
at a separation distance of about 20 A where their shells of similarly
oriented water molecules would touch; if their attraction for each
other is not large, it is possible that this might act as a barrier to
further approach (100) . However, in protoplasm other aggregating"
mechanisms may aid in squeezing out the water molecules and in
bringing the elementary protein particles into more intimate contact.
For example, van der Waals' attractive forces (19) become effective
at short distances of particle separation, and as the particles approach
still closer, hydrogen bridges between polar residues may become

Molecular Structure in Protoplasm 65

possible in addition to direct ionic linkages and cystine bridges.
Thus all three types may be influential in the production of aggre-
gates of protein particles. The more complementary the two faces
are to one another, the greater their interaction should be (99) .

Van der Waals' forces (65) are effective between all types of
atoms, while hydrogen bridges and ionic linkages, on the other hand,
are only possible between groups containing oxygen and nitrogen
atoms (19). Ionic linkages are possible only between oppositely
charged residues, such as ionized lysine and aspartic acid. In both
hydrogen bridges and ionic linkages, the atoms involved will
approach to a separation distance of approximately 2.8 A (19, 65)
in contrast to 3.5-4 A for van der Waals' distance (101) . A primary
valence bond, as in a cystine bridge, brings the bonded atoms to a
distance of approximately 1.5 A (19) .

The relative stability of these linkages is of interest. Ionic link-
ages and gross electrical effects are markedly affected by changes
in hydrogen-ion concentration (102). In addition to this, cystine
bridges are influenced greatly by the concentration of electron-
donating systems within the cell. Van der Waals' forces, on the
other hand, are practically independent of these two factors. Tem-
perature, of course, affects all forms of cohesion.

Recently evidence has been accumulating which indicates that
phospholipids and sterols may occur bound to the protein material
(103) . That the zwitterion portion of lecithin and cephalin, previ-
ously shown in Figure 7, may be effective as the means of attach-
ment is indicated by the similarity in the distance of 7 A between
the positive and negative charges on the lecithin, and the distance
of 7 A also between adjacent amino acid residues of the protein.
The lecithin might then become attached to two adjacent, oppositely
ionized residues of the protein through ionic attractions, or the
phosphoric acid group might bridge to a hydrogen-donating residue,
such as serine or tyrosine. If attached to the protein as shown in
Figure 13, it would project out about 25 A from the ends of the
amino residues (104, 105) . In certain instances the phospholipids
may act in blocking specific residue patterns where respiratory
prosthetic groups might otherwise become attached.

Concerning the attachment of the smaller molecules, the respira-
tary prosthetic groups, to protein particles, it seems too early to make
much more than suggestive statements, but one thing at least may
be sensed from the work which has been done; that is, degrees of
specificity seem to exist, or rather, some prosthetic groups may be

66

The Structure of Protoplasm

much more exacting than others in the type of residue or residue
combination to which they can become attached. For example,
small, uncharged organic molecules, such as acetaldehyde, ethyl
alcohol, etc., may require only a single polar residue; whereas
small ionized monocarboxylic acids, such as formic and acetic, may
require either a hydrogen-donating residue or a basically ionized
group; while others, such as hexose-diphosphate and the dicar-

•^♦Vi

Fig. 13. Model photograph showing dimensional relationship between ad-
jacent amino acid residues on the protein and the possible points of attach-
ment on the phospholipid.

boxylic acid, succinic, may require two adjacent residues for attach-
ment and proper orientation for reaction (Fig. 14) . In some of the
nucleotides only the ionized phosphoric acid group may attach to
the protein (106) ; while nucleic acid may require four adjacent,
positively ionized residues, such as those of arginine, for attach-
ment to its four negatively charged phosphoric acid groups. Evidence
(107) is accumulating which indicates that diphosphothiamine may
require two points of attachment, i. e., its phosphoric acid group and
the amino group of its pyrimidine ring. A somewhat similar attach-
ment was suggested for lactoflavinphosphate (108, 109) .

Degrees of specificity seem still more evident when it is noted
that in certain instances where two adjacent residues are required,
any two of seven different polar endings may be satisfactory; while
in other cases a more restricted combination is necessary. The
relative lengths of the adjacent residues may further restrict or

Molecidar Structure in Protoplasm

67

limit residue specificity; for example, if a hydroxyl group is needed
adjacent to an arginine residue ending, serine may be too short and
only a tyrosine residue would suffice to furnish the hydroxyl.

Again, in some instances, a large specific area consisting of as
many as four to six specific residues properly placed may be required
on the surface in order to provide for not only the active prosthetic
group but also for the substrate molecule with which it reacts; for

Fig. 14. Photographs of models showing dimensional relationship between:
(A) Ascorbic acid and adjacent amino acid residues; (B) Hexose diphosphate
and adjacent residues.

example, in the case of the pyridine nucleotides (110) the protein
determines the kind of substrate molecule with which it may react.
A further instance, one in which the conception of a prosthetic
group attached to the face of a particle gives way to a more complex
structure, is seen in the enzyme, cytochrome. It has been said (111)
"the heme residue is linked in two different ways to the protein, viz.,
by stable main valency bonds between the protein and one or two
vinyl side chains of the porphyrin, as well as by the usual coordinate
bonds between the central iron atom and some hemochromogen-
forming groupings in the protein component." To accommodate this
complex attachment the flat prosthetic group, shown previously in
Figure 11, would have to be partially enclosed by the protein chains
(129) . From such examples as these, where experimental evidence

68

The Structure of Protoplasm

points obviously to the necessity for attachment of certain small
molecules to specific larger ones in order to bring about particular
activities, and where the only means of attachment is through a
few sorts of residues restricted to limited localities on the larger

molecule to specific larger
ones in order to bring about
particular activities, and
where the only means of at-
tachment is through a few
sorts of residues restricted
to limited localities on the
larger molecule, the concep-
tion of specificity is given a
spatial as well as a structur-
al connotation.

There is still one more
class of components which
is involved in the construc-
tion of particle aggregates in
cytoplasm; this comprises the
inorganic ions. In general
the interactions of these
small ions and the protein

»

"^

r%

Fig. 15. Nucleic acid, a tetranucleotide,
plus a protein showing dimensional rela-
tionship between four phosphoric acid
groups and four adjacent amino acid resi-
dues.

molecules are of two sorts;
the ions may attach to the protein, and the protein may have a dis-
placing effect on the neighboring ions in the surrounding medium.
The kind and number of ions adsorbed will depend upon the nature
and distribution, both quantitative and spatial, of the charged groups
on the protein particle. The monovalent ions, Na+ and K+, will have
some tendency to associate with ionized carboxyl groups, but the
divalent ions, Ca"l + and Mg++, will have a more pronounced tend-
ency to become absorbed (112, 7) . The negative ions, NOs", SO4 ,

and PO4 , have been shown to associate with basic proteins (113)

probably with the residues of lysine, histidine, and arginine. Studies
of the displacement of the isoelectric point of various proteins with
changes of ion concentration of the solution (7, 114) give evidence
of ion association with the protein.

These examples of aggregation in which only one protein molecule
is involved with another protein, or with a fatty molecule, or with
another small organic molecule, or an inorganic ion, are instances
of simple interactions involving only two molecules at a time. It

Molecular Structure in Protoplasm 69

seems probable, however, that in the Uving organism several
organic molecules and ions become attached to a single protein
molecule to form a complex aggregate (115). The complexity to
which this is likely to lead may be brought out by referring to the
composition of active protoplasm as given in Table 2 where an
average situation is shown based on one molecule of protein of
molecular weight 36,000. A content of 2 per cent for the fatty
materials (69) corresponds to about 10 molecules, such as those of
the phospholipids, and the percentages of remaining materials indi-
cate about 20 small organic molecules and perhaps 50 to 100 inorganic
ions, also present on the average for each protein molecule of this
size. The small organic molecules include principally such substrate
molecules as those of hexose, triose, etc., since analyses indicate
very small amounts of the pyridine nucleotides (116) of cytochrome
(117) , thiamine-pyrophosphate (97) and riboflavinphosphate (118) .
In fact, computations show the number of these to be only 1 pyridine
nucleotide to about 20 molecules of protein of the size mentioned;
1 cytochrome to about 500 to 600 protein; 1 thiamine pyrophosphate
to about 1,700, and 1 riboflavinphosphate to about 1,300 protein
molecules of 50 A size.

If we now think of the protein as a cube of 50 A with six faces and
about ten polar amino acid residues on at least four faces, we will
have perhaps fifty potential places of attachment for the various
smaller molecules and ions, and in all there must be at least 150
molecules and ions available in the cytoplasm for each protein par-
ticle of 50 A in size, thus making it seem probable that some of the
small organic molecules are not attached to the protein but instead
are contained in the channels and submicroscopic vacuoles.

We now have some comprehension of an elementary protein par-
ticle to which several phospholipids as well as several inorganic ions
and perhaps a respiratory prosthetic group are attached. In the
living protoplasm, however, it seems reasonably certain that several
or many complexes of this sort are attached to form larger and
structurally more complex aggregates which serve particular pur-
poses. Down in the lower submicroscopic region, however, relatively
little is known concerning the aggregation of these conjugated pro-
teins with one another; further, there is very little evidence concern-
ing the quantitative distribution and localization of these materials
throughout the cell.

Aggregation of a number of small 50 A protein particles into a
larger complex involves the nature of the outer faces of the particles,

70 The Structure of Protoplasm

the number and distribution of the polar residue end groups. If the
particles could come together, face to face, a rather solid protein
complex would be formed; but with the large number of phospho-
lipids and smaller organic molecules present in protoplasm it seems
likely that some of them would intervene and prevent close face-to-
face attachment. We would expect, then, a rather wide separation of
perhaps 25-50 A due to the presence of the long hydrocarbon chains
of the phospholipids on some faces and less separation due to the
smaller organic molecules on other faces. It should be pointed out
also that inorganic ions may prove to be of considerable importance
in the formation of loose and transitory aggregates (122).

If we allow ourselves to speculate somewhat concerning the in-
ternal construction of the larger particles consisting of an aggre-
gation of perhaps thousands of these complex, conjugated proteins,
it seems reasonable to think that rather compact regions may occur
in addition to regions in which, either due to the overlapping of
particles or to loose aggregation (122), submicroscopic vacuoles oc-
cur. It is possible that it is in these regions that many of the "vital"
activities take place. In other words, the complex aggregates of many
simple 50 A particles are likely to be relatively porous with some
small and some large submicroscopic vacuoles. These loose, spongy,
complex aggregates may consist of a thousand component protein
particles, interpenetrated with water channels which widen here
and there to form the internal vacuoles, or "reaction chambers" as
they have been called (123). The walls of these "chambers" are
the faces of the protein particles to which fatty materials and res-
piratory prosthetic groups may be attached, probably in some spe-
cific pattern. The "chambers" may be relatively isolated from one
another within the same complex and are likely to vary somewhat
in their activity towards a substrate molecule. It seems possible
that the substrate molecule may be transformed in one manner in.
one "reaction chamber" and in quite a different manner in another.
By this means a channeling of reactions may occur which would not
take place if these materials were mixed and distributed in any
other fashion. Such a definite arrangement may be considered as one
phase of orgayiization in the particle as a whole.

The existence of complexes, whose activity depends upon the
integrity of the whole and whose organization is relatively easily
disturbed, has been demonstrated recently (14, 124, 125). The need
for such particular specific internal arrangements between the res-
piratory components has been sensed for some time (111) , and it

Molecular Structure in Protoplasm 71

seems probable that many vital activities take place in such or-
ganized structures.

With this conception of aggregates in which protein molecules
with attached smaller organic molecules are in turn held together
by cohesion forces of the same sort, hydrogen bridges, ionic bondings,
and van der Waals' forces, the enormous number of submicroscopic
granules in the fluid cytoplasm begin to take on the suggestion of
individual characteristics. Some may be too small or too simple
to perform any particular activity; others, although complex enough,
may be nonfunctional because they do not contain the proper con-
stituents or because of unsuitable arrangement of the constituents;
while some are definitely functional in character. A functional ag-
gregate is comparable to a large "floating laboratory" in which a
variety of separate, although coordinated projects are being carried
out.

Throughout this discussion, in which we have been concerned
with the molecular components of protoplasm and with the manner
in which their constituent atomic and molecular groups would allow
them to fit together, we have made use of the "particle" as a means
of keeping the molecular characteristics constantly in sight; and al-
though this has been a convenient means, it is based on in vitro
experimentation almost exclusively. The obvious question of the
biologist is: To what extent is it applicable to the living state? The
principal difference seems to lie in the random nature of the in vitro
world and the organization of the materials in the in vivo world;
further, in the living cell the functional aggregates are formed in
the presence of organized structures, of pre-existing mechanisms or
patterns.

We have come to a point where we wonder whether the elemen-
tary molecules interact simply by chance as recognized by statistical
methods or, instead of this freedom, have some orderliness brought
into their relationships by restrictions imposed upon them by their
living environment. The nature of this orderliness in construction
must be consistent with the behavior of the living cell, and our way
of thinking must now be altered to include the biological, as well as
the chemical-physical, statistical point of view. At this stage it now
becomes a matter of not what possible things can happen to protein
particles and to the other component molecules to transform them
into protoplasm, but instead, what has happened to bring about the
construction of active protoplasm from which these organic mole-
cules were obtained.

72 The Structure of Protoplasm

Perhaps it will clarify matters somewhat if we start with the raw
materials available to the plant and attempt to trace, in a general
way, their formation into larger aggregates. Inorganic ions and water
molecules are brought into the plant where carbon atoms from car-
bon dioxide have already been bound together into short carbon
chains with hydrogens and oxygens attached, as in glucose. These
we may think of as the basic substances. They are not combined at
random to form various substances, but instead are brought into the
presence of an already existing protoplasm, where only particular
constructions are permissible. The protoplasm present is a "going
concern" consisting of mechanisms which are capable of converting
these raw materials into particular molecular structures. We now
become confronted with the matter of construction, or shall we call
it synthesis. But synthesis of what, of more protoplasm, in place,
from basic materials, a sort of accretion; or of more mechanisms
similar to those already existing as component structures of the pro-
toplasm; or synthesis of large molecules, as those of protein, which
we may think of as building blocks used for construction of the
mechanisms? In other words, where does chance, as exemplified by
random interaction, enter into the formation of new protoplasm, and
at what stage does the pattern of pre-existing materials cease to be
of significance?

We are rather rapidly heading towards a dynamic picture, yet it
is the static, or only a single frame from a movie film, with which
we would momentarily be satisfied. If we should discover that it is
the larger molecules which are synthesized, then we have the task
of finding how they are put together to form a mechanism; and when
we know that, we still have to discover the organized arrangement
of the mechanisms which will produce active protoplasm. We have
omitted to mention synthesis of smaller molecules of which the larger
seem to be composed. The queries come up, which we repeat
although difficult to formulate clearly, as to whether the new proto-
plasm is built up by accretion in situ from the basic small-molecule
materials, or by addition of large already-elaborated molecules to
the existing protoplasmic structures, or by bringing already-built
mechanisms into proper relations for protoplasmic activity. The
difficulty of merely stating the questions is probably evident.

Many additional queries arise concerning "mechanisms," the
simplest of which may be the respiration complex which we have
just considered in a general way. Here the nature of the complex is
not yet well established, although small molecular mechanisms, en-

Molecular Structure in Protoplasm 73

zymes in this case, are fairly well known (64) . The larger complex
is thought to consist of several different enzymes which function in
a series, one reaction following another in more or less gradual steps
leading eventually to the disintegration of glucose. Each of the en-
zymes consists of a protein molecule and a prosthetic group. The
prosthetic groups, at least several of them, are known; and of the pro-
tein, the molecular weights of some are known, but little more in-
formation is available concerning them except that they are specific
for their particular prosthetic groups. The steps in the disintegration
process are known to some extent although not completely; we know
the materials which go into the machine and the details of a consider-
able amount of the disintegration as well as the final products; thus
here we have the parts of the machine and know how they perform
by themselves, but we do not know how to put them together to form
the complete structure; that is, we are ignorant of the arrangement
of the small mechanisms in the complex as a whole.

Recent evidence makes it seem probable that these small respira-
tory mechanisms may be associated with still other small mechanisms
to carry on even more complex activities. For example, the presence
of copper and iron along with chlorophyll, proteins, and lipoids has
been shown in isolated, intact chloroplasts (119, 120), thereby in-
dicating that respiratory mechanisms are probably associated with
photosynthetic mechanisms in the minute chlorophyll-containing
portions. Concerning this synthetic process we know with a con-
siderable degree of certainty that the raw materials which go into the
machine are carbon dioxide and water and that the products which
come out are glucose and oxygen; and we know here also the mole-
cular nature of part of the machine, the energy-absorbing chloro-
phyll molecule and in general the protein chain. This is a step in
advance of the leucoplast where the material going in, glucose, and
that coming out, starch grain, are known, but concerning the mole-
cular structure of the machine, practically nothing is known except
that it probably is mainly protein. In contrast to the chloroplast,
however, here the end product, the starch grain, is microscopically
visible.

In the microscope we see a leucoplast as a complex mechanism,
but as though at a great distance, for the details are invisible, al-
though the output of starch grain is seen. Here we see a mechanism
as a whole but know nothing of the smaller component mechanisms
or of their molecular details. By taking the starch grain apart we
have learned that the mechanism is capable of condensing glucose

74 The Structure of Protoplasyn

molecules and placing them in an orderly manner to form the starch
grain. It seems probable that, in detail, there are many small mechan-
isms within the plastid which are active in placing the glucose mole-
cules on the surface of the grain and also in producing an ether link-
age between them. Actually we know only that the starch grain
becomes larger; to our mechanistic minds there must be machinery of
some sort, but of its construction we know very little. In the plastid
we think of an orderliness of some sort, in the placement of the ma-
chinery which produces starch grains, since the shapes of the starch
grains are fairly specific with plant species, and as a matter of fact,
the behavior of the starch itself is in some respects almost as specific
as the structures of the grains. Despite this, it seems fairly reasonable
to think that the individual mechanisms which bring about the con-
densation of glucose to starch are more or less universal in plants.

From the biological point of view we think of the plastid as a very
large particle. Shall we think of its growth as consisting of a build-
ing-up process, in place, from small molecules, or large molecules as
of protein, or from ready-made small mechanisms? We come to
similar questions concerning a large protein molecule of some thou-
sand atoms. Is there a machine to manufacture these with the proper
amount of amino acids for each species and for each mechanism
where several specific proteins are apparently needed as in respira-
tion? Or is it more likely that the protein molecules are merely the
wreckage of the larger mechanism complexes which were built by
accretion of the small basic molecules? Questions of this sort and
the discussion of them leads primarily to a demand for experimental
answers, and each small point is likely to mean a large experimental
task.

Concerning these questions which have been posed here, there is
undoubtedly much more information in the literature than is known
to any one of us. Some of it is based on cytology or allied subjects;
some on morphological, some on physiological and some on other bio-
logical procedures. Here and there are items in chemical and phys-
ical journals. Our questions involve the interpretation of biological
experiments from a physical and chemical point of view; for example,
the microscopic measurement of the rate of deposition of a growing
cell wall, the transverse septum of an algal cell (121) . This is a bio-
logical experiment, yet the interpretation of these measurements
into the rate of deposition of glucose molecules is chemical in that
it gives a rate of ether linkage formation, and physical in that the
process of crystal formation may be studied.

Molecular Structure in Protoplasm 75

Experimental work is greatly desired which takes the molecular
viewpoint into consideration even though the direct purpose of the
experiment is based on a higher dimensional level. On the other
hand it may be possible that the molecular viewpoint and the sub-
microscopic viewpoint, where molecular structure of the particle is
taken into consideration, will help to direct the experimental at-
tempts to answer the many questions which are constantly arising.
Suggestive bits of information come out of the molecular dimen-
sional models and also from the conception of cohesion forces in their
manner of holding molecules and particles together. This informa-
tion seems directly applicable to the materials of protoplasm, for
regardless of whether we are dealing with large molecules, with
mechanisms, or with the summation of these into a protoplasmic mass,
we still are dealing with atoms and atomic groups and with the forces
which are involved in their interaction and in the formation of mole-
cular structures. Further, the atoms and the molecular groups occupy
specific amounts of space which must be reckoned with in their
movement within a concentrated mixture, especially when organ-
ized to the extent of cytoplasm.

SUMMARY

A very short summary may help to point out what we consider
to be important in forming conceptions of molecular structure, as we
now think of it in protoplasm. Specifically, the atomic radii and
their directive valence angles should be the same in vivo as in vitro.
They give shape and size to structures on the molecular level and
should actively influence structure on higher levels. The primary
valence bond may be considered as furnishing the binding strength
and closeness of atomic approach necessary to make the molecule
an entity.

The manner of attachment of molecules and particles to one an-
other must be the same in the cell as in vitro; that is, by means of
H-bridges, ionic and van der Waals' attraction. The H-bridge is effect-
ive directly between two atoms of oxygen and nitrogen only. Ionic
forces of attraction may be effective in a similar individual manner
between oppositely charged groups; and also as a resultant of several
or many charges. Thus particles may be held together by several
negatively and positively charged atoms pairing up directly and also
by the resultant field produced by all of the charges on the particle,
within as well as on the surface. Van der Waals' forces act much
less locally on a particle; instead, they act more like a diffuse field

76 The Structure of Protoplasm

around the particle as a whole. The strength of attachment will
depend to such a great extent upon the specific nature of the
approaching atoms and faces that little more can be said than that a
considerable although a relatively limited number of variations
occur. This must be evident when one considers the possibilities
existent in the approach, face to face, of two protein particles. An
H-bridge may be effective at one point, several ions may pair up at
other points, and van der Waals' forces will always be effective
when the surfaces are close together; that is, within 3 to 5 A.

Attachment of prosthetic groups to the protein is thought to
occur specifically. In some instances the small molecule may be
bound directly to the protein through hydrogen bridges, in other
cases through ionic pairing. Primary valence bonds and even van der
Waals' forces are not to be excluded. The spacing between the
atomic constituents which are thought to be active in the attach-
ment of the prosthetic group to the protein was shown to be com-
patible in certain instances with the distance of about 7 A between
adjacent residues on the protein chain.

These fundamental concepts of molecular models and inter-
molecular forces of attraction make it possible to determine the
nature of larger structures. The larger structures we may think of
as particles at one moment, principally because we are familiar with
microscopic particles in protoplasm; while at another moment we
may think of them as mechanisms, because we know from experi-
mental procedures that particles are involved in activities of various
sorts. From the composition of the particles we are enabled to
carry over the structural molecular concepts to the mechanisms,
but only insofar as we know the composition of the mechanisms.

Organization of these mechanisms, in some as yet undetermined

manner, in the cell, we assume at the present stage of investigations,

constitutes the physical basis for vital activities. In conclusion, we

feel that this organization within the cell, which is without doubt

responsible for the phenomena presented by a living organism, is

of such a nature that further elucidation of it may come from the

application of the fundamental concepts mentioned, to experimental

investigations involving both biological and physical-chemical pro-

2edures.

LITERATURE CITED

1. Sponsler, O. L. The Quarterly Review of Biology VIII, 1, 1933.

2. Gaidukov. N. Ber. d. Deutschen Bat. Gesell 24, 107, 155, 580. 1906; Dunkel-

jeldheleuchtujig und Ultraynicroscopie in der Biologie und in der Medi-
zin, Jena. 1910.

Molecidar Structure in Protoplasvn 77

3. Chambers, R. Jour. Exp. Zool. 23, 483. 1917.

4. Bayliss, W. M. Proc. Roy. Soc, Series B. 91, 196. 1920.

5. Price, S. R. A7i?i. Bot. 28, 601. 1914.

6. Taylor, C. V. Proc. Soc. Exp. Biol, and Med. 22, 533. 1925.

7. Lloyd, D. J., and A. Shore. The Chemistry of the Proteins. Philadelphia.

1938.

8. Schmidt, C. L. A. Chemistry of the Amino Acids and Proteins. Baltimore.

1938.

9. SvEDBERG, T. Proc. Roy. Soc. Lond. 127B, 1. 1939.

10. Kalckar, H. M. Chem. Rev. 28, 71. 1941.

11. Barron, E. S. G. Physiol. Rev. 19, 184. 1939.

12. Keilin, D., and E. F. Hartree. Proc. Roy. Soc. Lond. 125B, 171. 1938.

13. Stern, K. G. Cold Spring Harbor Symjjosia on Quant. Biol. VII, 312. 1939.

14. Needham, J. Order and Life. New Haven. 1936.

15. Bernal, J. D., AND R. Fowler. Jour. Chem. Phys. 1, 515. 1933.

16. Debye, p. Polar Molecules. New York. 1929.

17. Cross, P. C, J. Burnham, and P. A. Leighton. Jonr. Av\er. Chem. Soc. 59,

1134. 1937.

18. Barnes, W. H. Proc. Roy. Soc. Lond. USA, 670. 1929.

19. Pauling, L. The Nature of the Chemical Bond. Ithaca. 1939.

20. HuGGiNS, M. Jour. Org. Chem. 1, 407. 1936.

21. Sponsler, O. L. The Cell and Protoplasm. Science Press. 1940.

22. Pearsall, W. H., and J. Ewing. Brit. Jour, of Exp. Biol. 2, 347. 1924.

23. Seifriz, Wm. The Botanical Review 1, 18. 1935; Protoplasm. New York.

1936.

24. Frey-Wyssling, A. Jour, of Roy. Micr. Soc. LX, 128. 1940.

25. MoYER, L. This Monograph.

26. Parat, M., and J. Painleve. Compt. Rend. Soc. de Biol. 94, 745. 1926.

27. Giroud, a. Protoplasma 7, 72. 1929.

28. Milovidov, M. P. Compt. Rend. Acad, des Sciences 187, 140. 1928.

29. Shinke, N., and M. Shigenaga. Cijtologia 4. 189. 1933.

30. GuiLLiERMOND, A. Lcs Constituauts Morphologiques du Cytoplasme: Le

Chondrionie. Paris. 1934.

31. Volumes VI and VII of the Cold Spring Harbor Symposia on Quantitativie

Biology.

32. Astbury, W. T., and R. Lomax. Jour. Chem. Soc. Lond., p. 849, June, 1935.

33. Sponsler, O. L., and J. D. Bath. In press. Proc. Soc. Exp. Biol, and Med.

34. KuNiTZ, M., M. L. Anson, and J. H. Northrop. Join-. Ge7i. Physiol. 17, 365.

1933.

35. KuNiTZ, M., and J. H. Northrop. Jour. Gen. Physiol. 18, 433. 1935.

36. Crowfoot, D. Nature 135, 591. 1935; Proc. Roy. Soc. Lond. 164A, 580. 1938.

37. Bergmann, M., and C. Niemann. Joiir. Biol. Chem. 115, 77. 1936; Jour. Biol.

Chem. 118, 301. 1937.

38. Svedberg, T. Nature 139, 1051. 1937.

39. Bernal, J. D., D. Crowfoot, et al. Nature 141, 521. 1938.

40. Loeb, J. Jour. Gen. Physiol. 4, 73. 1921-22.

41. Chick, H., and Martin. Jour. Physiol. 45, 261. 1912.

42. VON Muralt, a. L., and J. T. Edsall. Jour. Biol. Chem. 89, 351. 1930.

43. Cohn, E. J. Cold Spring Harbor Symposia on Quant. Biol. VI, 8. 1938.

44. Neurath, H. Cold Spring Harbor Symposia on Quant. Biol. VI, 196. 1938.

78 The Structure of Protoplasm

45. AsTBURY, W. T. Cold Spring Harbor Symposia on Quant. Biol. 11, 18. 1934.

46. Pauling, L., and C. Niemann. Jour. Amer. Cheni. Sac. 61, 1860. 1939.

47. Wyckoff, R. W. G., and R. Corey. Science 81, 365. 1935.

48. AsTBURY, W. T., S. Dickinson, and K. Bailey. Biochem. Jour. 29. 2351. 1935.

49. Wu, H. Chinese Jour. Phijsiol. 5, 321. 1931.

50 AsTBURY, W. T., AND F. O. Bell. Cold Spring Harbor Symposia on Quant.
Biol. VI, 109. 1938.

51. MiRSKY, A. E., AND L. Pauling. Proc. Nat'l. Acad. Sci. 22, 439. 1936.

52. AsTBURY, W. T., F. O. Bell, E. Gorter, and J. van Armondt. Nature 142, 33.

1938; 143, 280. 1939.

53. ScHMiTT, F. O. Physiological Reviews 19, 270. 1939.

54. Sponsler, O. L., J. D. Bath, and J. W. Ellis. Jour, of Phys. Chem. 44, 996.

1940.

55. Mitchell, H. H., and T. S. Hamilton. Biochemistry of Amino Acids. New

York. 1929.

56. Karrer, p., and W. Kaase. Helv. Chiyn. Acta 2, 436. 1919; 3, 244. 1920.

57. Sponsler, O. L. Second Annual Symposium of the Society for the Study

of Development and Growth. Growth, Supplement, pp. 1-26. 1940.

58. Sponsler, O. L., and Jean D. Bath. To be published in the Biological

Reviews of the Cambridge Philosophical Society.

59. Hill, T. L., and G. E. K. Branch. Science 91, 145. 1940.

60. Chambers, R. B\d. Nat'l Res. Council 69, 37. 1929; The Harvey Lectures,

p. 49. 1926-27.

61. Small, J. Hydrogen Ion Concentration in Plant Cells and Tissues. Berlin.

1929.

62. Seifriz, W., and M. Zetzmann. Protoplasma 23. 175. 1935.

63. Cohn, E. J. The Harvey Lectures, p. 124. 1938-39.

64. Green, D. E. Mechanisms of Biological Oxidations. Cambridge. 1940.

65. Slater, J. C. Introduction to Chemical Physics. New York. 1939.

66. Fisher, E. Zeit. Physiol. Chem. 99, 54. 1917.

67. Bergmann, M. Chem. Rev. 22, 423. 1938.

68. Niemann, C. Cold Sjoring Harbor Symposia on Quantitative Biology VI,

58. 1938.

69. Leathes, J. B., AND H. S. Rarer. The Fats. London. 1925.

70. Maclean, H., and I. Maclean. Lecithin and Allied Substances. London. 1927.

71. Gilman, H. Organic Chemistry, Vol. II, New York. 1938.

72. Regaud, C. Compt. Reiid. Soc. Biol. 65, 660. 1908.

73. Faure-Fremiet, M. E. Arch, d' Anat. Micr. 11, 457. 1910.

74. LowscHiN, A. M. Ber. Deutsch. Bot. Gesell. 31. 203. 1913.

75. Lewitsky, G. Zeit. filr Botanik 16, 65. 1924.

76. Chibnall, A. C, AND H. J. Channon. Biochem. Jour. 21,242. 1927.

77. Bensley, R. B. Anat. Rec. 69, 341. 1937.

78. Feulgen, R., and Th. Bersin. Zeit jiir Physiol. Chem. 260, 217. 1939.

79 Fischgold, H., and E. Chain. Proc. Roy. Soc. Lond. lUB, 239. 1939; Biochem.
Jmcr. 28, 2044. 1934.

80. Chain, E., and I. Kemp. Biochem. Jour. 28, 2052. 1934.

81. Bull, H. B., and V. L. Frampton. Joiir. Amer. Chem. Soc. 58, 594. 1936.

82. Grun, a., and R. Limpacher. Ber. Chem. Gesell. 60, 151. 1927.

83. Rudy, H., and I. H. Page. Zeit. Physiol. Chem. 193, 251. 1930.

Molecular Structure in Protoplasm 79

84. Jukes, T. H. Jour. Biol Chem. 107, 783. 1934.

85. Bernal. J. D. Nature 129, 277. 1932; Jour. Soc. Chevi. hid. 51, 466. 1932;

An7i. Repts. Chem. Soc. Lond. 30, 423. 1933.

86. Bills, C. E. Physiol. Rev. 15, 1. 1935.

87. Lloyd, D. J. Biol. Rev. 3, 164. 1928.

88. Elvejhem, C. a. Respiratory Enzymes. Minneapolis. 1939.

89. Oppenheimer, C, and K. G. Stern. Biological Oxidation. New York. 1939.

90. GiROUD, A. L'Acide Ascorbique dans la cellule et Les Tissus. Berlin. 1938.

91. Theorell, H. Ergebnisse der Enzymforschung. VI, 111. 1937.

92. Warburg, O., and W. Christian. Biochem. Zeit. 298, 150, 368. 1938.

93. Warburg, O. Ergehnisse der Enzymforschung 7, 210. 1938.

94. Levene, p. a., and L. W. Bass. Nucleic acids. New York. 1931.

95. Stern. K. G. Cold Spring Harbor Symposia on Quant. Biol. VI, 286. 1938.

96. Theorell, H. Biochem. Zeit. 298, 242. 1938.

97. LOHMAN, H., AND ScHUSTER. Biochem. Zeit. 294, 188. 1937.

98. Stern, K. G., and J. W. Hofer. Enzymologia, 3, 82. 1937.

99. Pauling, L., and M. Delbruch. Science 92, 78. 1940.

100. Langmuir, I. Joiir. Chem. Phys. 6, 873. 1938.

101. Robertson, J. M. Chem. Rev. 16, 417. 1935.

102. Svedberg, T. Trails. Farad. Soc. 26, 740. 1930.

103. Heilbrunn, L. V. Biol. Bui. 71, 299. 1936.

104. ScHMiTT, F. O., R. S. Bear, and G. L. Clark. Radiology 25, 131. 1935.

105. ScHMirr, F. O. Jour. Applied Physics 9, No. 2. 1938.

106. Ball, E. G. Cold Spring Harbor Syviposium ori Qvxintitative Biology VII,

100. 1939.

107. Heegaard. E., and E. R. Buchmann. Unpublished Experiments. 1940.

108. Theorell, H. Biochem. Zeit. 290, 293. 1937.

109. KuHN, R., AND P. Boulanger. Ber. Chem. Ges. 69, 1557. 1936.

110. Warburg, O. Ergeb. d. Enzymforschung 7, 210. 1938.

111. Stern, R. G., Ann. Rev. Biochem IX, 1. 1940.

112. Greenberg, D. M., and C. E. Larson. Join. Phys. Chem. 43, 1139. 1939.

113. Robertson, J. B. The Physical Chemistry of The Proteins. London. 1924.

114. TiSLius, A., AND H. SVENSSON. Trans. Farad. Soc. 36, 16. 1940.

115. VON Przylecki, St. J., W. Giedroyc, and H. Rafalowska. Biochem. Zeit.

280, 286. 1935.

116. Bernheim, F., and A. von Felsovanyi. Science 91, 76. 1940.

117. Theorell, H. Biochem. Zeit. 298, 242. 1938.

118. GouREvncH, A. Bui Soc. Chim. Biol 19, 527. 1937.

119. Granick, S. Amer. Jour. Bot. 25, 558, 561. 1938.

120. Neisch, a. C. Biochem. Jour. 33, 293, 300. 1939.

121. Sponsler, O. L. Jour. Amer. Chem. Soc. 56, 1599. 1934.

122. Bernal, J. D. The Cell and Protoplasm. Science Press. 1940.

123. Warburg, O. Ergebnisse der Physiol. 14, 316. 1914.

124. Peters, R. A., and Thompson. Biochem. Jour. 28, 916. 1934.
325. Case, E. M. Biochem. Jour. 25, 568. 1931.

126. Navev, a. E., and E. B. Harvey. Biol. Bui 69, 342. 1935.

127. Sh.'Vpiro, H. Jour, of Cell and Comp. Physiol. 6, 101. 1935.

128. Warburg, O. Pflilgers Arch. Physiol. 154, 599. 1913.

129. Theorell, H. The Cell and Protoplasm. Science Press. 1940.

1

The American Society of Plant Physiologists wishes to show its
respect for the late Professor Herbert Freundlich by publishing
his portrait as a frontispiece to his chapter in this Monograph. The
article on Thixotropy which follows was the last manuscript writ-
ten by Professor Freundlich. Rare was the contribution from
Professor Freundlich's laboratory which did not enrich physiology
as much as it did chemistry, and none had a greater influence on
biological thought than did thixotropy.

Those who knew Professor Freundlich will remember him for
his intense interest in biology and his broad understanding of it.
Those few among us who had the good fortune to work in his
laboratory will remember him as one of the most inspiring teachers,
fair and tolerant colleagues, and gracious companions that a man
could desire. His was a truly great character. Biology has lost
one of its ablest and noblest friends.

[83]

SOME MECHANICAL PROPERTIES OF SOLS AND GELS AND
THEIR RELATION TO PROTOPLASMIC STRUCTURE

Herbert Freundlich

Editor's Note: This article was found among the papers of the late Professor
H. Freundlich in an unfinished state. The first part, dealing with the mechanical
properties of sols and gels, seems to be nearly complete. No notes concerning the
latter part were found. Except for minor styhstic alterations and correction of
typographical errors, the manuscript is printed as found. The references and
figures were added by L. Moyer and K. Sollner.

In older textbooks of physiology, let us say of about 1900, the
state of aggregation of protoplasm is left very unclear.^ It is fre-
quently considered to be fluid. On the other hand, it is emphasized
that life processes seem to require a well-defined "organization"
and that such an organization can hardly be imagined without some
kind of a more or less solid structure. There is practically no indi-
cation of the fact that the mechanical properties of protoplasm differ
essentially from those of normal liquids or solids and that they may
be correlated with its colloidal nature. References to colloids are
found, if at all, when the impermeability of membranes to proteins,
or swelling, or the scarcity of distinct crystals are discussed.

This outlook has completely changed during the last twenty
years. It has been found that there may exist a number of inter-
mediate stages between the normal liquid state and the state of a
crystalline solid. These are frequently observed in colloidal sys-
tems, particularly if they contain a sufficiently high amount of dis-
perse phase. Consequently, concentrated colloidal solutions may
differ essentially in their mechanical properties (viscosity, elas-
ticity, etc.) from normal, so-called Newtonian liquids. These differ-
ences may be manifold. Two limiting cases are of outstanding im-
portance; the first is that of thixotropy. Whereas the viscosity of a
Newtonian liquid is not changed by mechanical means, such as flow-
ing or stirring, a thixotropic system becomes less viscous while
flowing or when being stirred. This phenomenon is particularly ob-
vious if we have a thixotropic gel: It is liquefied by shaking and sets
again to a gel when at rest. We thus have an isothermal, reversible,
sol-gel transformation. -

[85]

86 Tks Structure of Protoplasm

We are dealing with a thixotropic change, too, if the viscosity
of a concentrated sol can be reduced by shaking or stirring and if it
increases again when at rest. The viscosity of such a sol differs in
other respects from that of a Newtonian liquid. It is anomalous, i.e.,
it does not obey Poiseuille's law for laminar flow; the amount of
liquid passing through a capillary in a given time depends on the
applied pressure in a different way. Speaking more specifically, the
viscous resistance is not directly proportional to the velocity gradient
of the laminar flow, as in Newtonian liquids. The characteristic
behavior of anomalously viscous, thixotropic sols and likewise of
thixotropic gels is correlated with the existence of a yield value.
This can be demonstrated in a graphic way by using a viscometer,
in which a ball is pulled through the viscous system in question by
a weight which can be increased at will; for different weights the
speed of the moving ball is compared.^' In a Newtonian liquid the
speed is proportional to the weight applied (curve 1 in Fig. 1) . In
a thixotropic sol or gel we have a yield value (A) : The weight must
exceed a certain minimum to cause the ball to move; only with
higher weights does its speed change in a way similar to that ob-
served in a normal liquid (cf. curve 2 in Fig. 1). Curve 1 refers to
glycerol, a typically Newtonian liquid, curve 2 to a thixotropic,
aqueous iron oxide gel. In other cases we may actually have a less
straightforward behavior: a curve which, at higher weights, is simi-
lar to curve 2 but which does not intersect the abscissa; instead, it
is curved and passes through the zero point.

Systems showing anomalous viscosity also exhibit a characteris-
tic behavior with respect to their electrical conductivity. The elec-
trical conductivity of an aqueous solution is markedly influenced
by its Newtonian viscosity; as a rule, it is, in a first approximation,
inversely proportional to the viscosity. For instance, the electrical
conductivity of an aqueous salt solution having a fairly high vis-
cosity, owing to the presence of a suitable concentration of glycerol,
is lower than that of a solution of the same salt concentration in
pure water. The anomalous viscosity of a thixotropic sol or gel, on
the other hand, leaves the electrical conductivity practically un-
changed. The electrical conductivity of a salt solution remains the
same, although it may contain so much gelatin that its apparent vis-
cosity is about equal to that of the solution containing glycerol.^ It
is well known that even gels of gelatin (containing electrolytes) or
of soaps do not differ in their electrical conductivity from the sols
from which they were produced. ' Presumably the movement of the

Sols and Gels — Relation to Protoplasmic Structure

87

ions is not influenced by the structure of thixotropic sols and gels,
which is very coarse compared to the small size of the ions.

Finally, not only gels but also sols of this type (gelatin, soaps,
etc.) show elastic properties. Small solid particles suspended in
such a sol may be moved by an external force but return to their

Fig. 1.

original position when the force stops acting, a phenomenon not ob-
served in Newtonian liquids.'' This elasticity in sols is not found so
regularly as the other characteristic properties just mentioned. The
presence of distinctly rod-shaped colloidal particles is perhaps more
decisive in causing elastic effects than it is with the other properties
characteristic of this kind of anomalous viscosity.

Thixotropic behavior is not exceptional; it is very common, pro-
vided that suitable concentrations of the colloid are chosen. Ex-
amples are: thixotropic, aqueous gels of many oxides' (aluminum,
iron, scandium, vanadium, titanium, thorium, etc.), of colloidal

88 The Structure of Protoplasm

bentonite,*^ of myosin,^ and tobacco mosaic virus/" of dibenzoyl-
cystine^^ and of barium malonate^- in a medium of water and alcohol.
Aqueous solutions of gelatin have been found to show thixotropy,
both as concentrated solutions and as gels.^^

Coarse suspensions containing particles with a diameter of 1^
and more may also be thixotropic: They behave as a liquid while
being shaken and settle to a solid paste when shaking stops. This
behavior is again found very frequently with aqueous suspensions
of clays, slates (Solnhofen slate) , and many powdered minerals
(mica, iron oxide, jet, etc.)^^; with finely powdered mercaptoben-
zothiazol in many organic liquids (benzine, carbon tetrachloride,
chloroform, benzene, toluene, etc.)^''; and in suspensions of many
pigments in oils.^'' Bentonite is a particalrly good example of a sub-
stance forming thixotropic systems both in colloidal and in coarse
suspensions.'^ The mechanical properties of coarse suspensions are
important because they enable us to understand better the mecha-
nism of these phenomena. Hence I shall have to refer to them fre-
quently, although protoplasm is certainly a truly colloidal system
containing very much smaller particles.

The other limiting case is that of dilatancy. So far, it has been
investigated only in coarse suspensions. Although observed and
named by Osborne Reynolds^ ^ in 1885, it has only recently been
recognized as a remarkable counterpart to thixotropy. Osborne
Reynolds used the term when describing the behavior of moist
quartz sand: It whitens and appears to be dry when the foot falls
on it and becomes wet again when the foot is raised. An aqueous
suspension of finely ground quartz powder (particle size 1 to 5 |.i)
at a concentration of about 44 per cent of the solid is strongly
dilatant^: A glass rod can easily be moved through the mass at
low speed, but an enormous resistance is set up if the speed is
increased above a certain limit. Using the viscometer mentioned
above, curve 3 (cf. Fig. 1) is observed: The suspension behaves like
a Newtonian liquid at low speeds but, from a certain higher speed,
the horizontal part of the curve implies a solid behavior. In this
part of the curve there are some intricacies which are better dis-
cussed later. Suspensions of intact starch grains in water are
strongly dilatant too."

There is every reason to believe that colloidal solutions may also
be dilatant. A colloidal solution of silicic acid of a suitable con-
centration and pH is found to be extremely viscous but Newtonian
in its behavior; it becomes hard and brittle, breaking up to a white,

Sols and Gels — Helation to Protoplasynic Structure 89

dry powder when crushed with a stout glass rod. Left to itself, the
powdered mass returns again to its original viscous liquid state. ^"^
A technical product, Nuodex Calcium-S (a colloidal solution of
about 10 per cent calcium naphthenate in a petroleum distillate),
shows a similar behavior.^ "^ A complete curve of one of these
colloidal systems corresponding to curve 3 in Figure 1 has not yet
been measured.

The behavior of coarse particles of suspensions under the micro-
scope gives us a clue concerning that property of the particles which
makes a suspension dilatant or thixotropic.^''- ^'' The particles of a
dilatant suspension are quite independent of each other; there is
not the least indication of coagulation. If they are settled on the
slide, they are all separated from each other; if dislodged and driven
into the liquid by a slight knock, they remain separated in Brownian
movement until they have settled again. On the other hand, the
particles of a thixotropic suspension are always found to be coagu-
lated to a certain degree, sticking together and forming clusters.
If brought into suspension in the liquid, they may be temporarily
separated from each other, but they always unite again to clusters.

This distinctive behavior allows us to understand many results
obtained with thixotropic or dilatant systems. For instance, in
order to make a concentrated iron oxide sol thixotropic a small
amount of a coagulating electrolyte like NaCl must be added, an
amount much smaller than that needed for actual coagulation.-"
In agreement with the positive C-potential of the iron oxide par-
ticles and with the Schulze-Hardy rule, the anions are especially
effective. Smaller concentrations of polyvalent anions are necessary
to obtain the same state of thixotropy, which is characterized by the
same time of re-solidification to a gel after the original gel has been
liquefied to a sol by shaking. The thixotropic state, therefore, is
often considered to be a primary stage of coagulation.

The following experiments done by W. Heller prove this concept
more quantitatively-^: If an iron oxide sol, which is just too dilute
in iron oxide to give a thixotropic gel on adding a certain amount
of electrolyte, is centrifuged after the electrolyte has been added,
a more concentrated, gelatinous sediment is accumulated at the
bottom of the vessel. This sediment is not formed on using the same
centrifugal force, if no electrolyte has been added. The sediment
can be reversibly redispersed in the liquid by shaking. If the sedi-
ment is separated from the less concentrated liquid on top, it is
found to be a normal, i. e., reversible, thixotropic gel. From the

90 The Structure of Protoplasm

way the rate of sedimentation is correlated with the concentration
of the added electrolyte, it can be concluded that reversible coagu-
lates are formed. These are very rich in water and contain a number
of colloidal iron oxide particles. This number increases greatly
with increasing concentration of the added electrolyte. Such revers-
ible coagulates, "geloids," are also formed in the original dilute sol
on adding electrolyte, but their concentration is not high enough to
let them coalesce to a coherent structure and turn the whole sol
to a gel. If, however, the geloids are concentrated by centrifuging,
they coalesce to a thixotropic gel. By investigating the change of
so-called conservative light absorption during the process of thixo-
tropic gelation, i. e., the light absorption caused exclusively by the
scattering of light (due to the presence of the colloidal particles) ,
it can be shown that the reversible coagulation causing thixotropy
is always accompanied by a certain degree of irreversible coagu-
lation, which increases strongly with increasing electrolyte concen-
tration.-- This irreversible coagulation is the chief factor in the
regular coagulation occurring at higher concentrations of electrolyte.

How many colloidal particles are contained in a geloid and
how their number depends on the nature of the colloid is not known.
These differences may be marked, as can be concluded from the
fact that the minimum concentrations of colloid, when a thixotropic
gel is formed, vary distinctly: For an iron oxide sol (of the Graham
type) , this concentration is about 5 gm. per litre, -^ for a VoO-, sol, it
is only about 0.1 gm. per litre. -^

Thixotropy may be sensitive to very small changes in the con-
centration of substances contained in the sol. Thus iron oxide sols
are particularly sensitive to H- and OH' ions"; the time of solidifica-
tion is strongly increased by an increase in H* ions, decreased
by an increase in OH' ions, i. e., an increase in H- ions has a liquefy-
ing effect and vice versa.-"' A pH change from 3.9 to 3.1 caused
the time of solidification to rise from 82 seconds to 150 minutes. By
dipping a silver plate into an iron oxide gel for 18 hours, the pH
changed from 3.4 to 3.8, producing a decrease in the time of solidifi-
cation from 33 minutes to 72 seconds. It was further found that
amino acids had a liquefying action upon these gels, independent
of the change they caused in pH, i. e., they increased the time of
solidification, although they increased the pH.-*^

It is important, also, from a biological point of view, that not
onlv electrolytes, but also suitable organic nonelectrolytes, are able

Sols and Gels — Relation to Protoplasmic Structure 91

to produce thixotropy. Alcohol added to a suitable iron oxide sol
makes it thixotropic.--^ We are probably dealing with a coagulation
due to "dehydration." These iron oxide sols may be considered to
be sufficiently hydrophilic to allow one to expect such effects of
dehydration as are discussed by Kruyt and Bungenberg de Jong
in their theory of the stability of hydrophilic sols and of coacerva-
tion.-^ If the alcohol is removed from the thixotropic gel by exposing
it to sulfuric acid in a desiccator, the gel is liquefied to a sol.

All these results show that substances causing a certain degree I
of coagulation lead to the formation of thixotropic systems. Inversely,
a strong peptizing agent, making the particles independent of each
other, can transform a thixotropic system (having the proper par-
ticle size) into a dilatant one. A technical dispersing agent, "Horn-
kem," a sulfonated product of vegetable origin, applied in a suitably
high concentration in aqueous solution, acts upon ZnO particles in
such a way that it produces a strongly dilatant suspension parallel
with a maximum degree of dispersion.-' In pure water, the same
ZnO produces a very stiff paste without any indication of dilatancy.
In this case it can even be observed that thixotropy requires a
medium degree of coagulation: In pure water the degree of cluster-
ing is too great to allow a thixotropic behavior of the suspension;
if, however, the right amount of a dilute solution of Hornkem is
added, the paste becomes thixotropic, whereas at higher concen-
trations of Hornkem a state of high dispersion and dilatancy is
reached.-'

It fits in well with these results that a dilatant paste can be
transformed into a thixotropic one by causing a certain degree of
coagulation of the particles. Hydrophilic colloids at low concentra-
tions favor the coagulation of hydrophobic particles, the phenome-
non of sensitization; whereas at high concentration they protect the
hydrophobic system. In this way, lecithin at low concentration may
coagulate aqueous quartz suspensions to a certain degree-'^; this is
proved by the marked increase in the rate of sedimentation of the
clustered particles after lecithin has been added. Only in this range
of sensitization by the lecithin are these quartz suspensions found
to be thixotropic, whereas they are dilatant in pure water.

The following facts cannot, presumably, be applied to biological
phenomena, but they deserve to be mentioned briefly, nevertheless.
The thixotropic or dilatant state depends strongly on the dispersion
medium. Whereas suspensions of quartz or of intact starch grains

92 The Structure of Protoplasm

are dilatant in water, they are thixotropic in organic liquids such
as benzene, carbon tetrachloride, etc.^ It is possible to pass gradually
from one state to the other by using two miscible liquids.

In these organic liquids, too, a small amount of a second sub-
stance may cause a marked change in behavior: A paste made from
a very fine iron powder and carbon tetrachloride is pronouncedly
plastic, a behavior which is practically always correlated with a
marked thixotropy.-'' On adding a small amount of oleic acid, the
paste is liquefied to a strongly dilatant suspension. The adsorption
of the oleic acid on the surface of the particles produces a state of
independence, and hence dilatancy.

Suspensions of finely powdered solids in organic liquids, such
as mixtures of oils, have been used for a very long time as paints.
It is, therefore, no wonder that painters have been acquainted with
the manifold, anomalous mechanical properties of suspensions,
though without having defined such limiting cases as thixotropy and
dilatancy. But it can hardly be doubted that many pastes of paints,
described as stiff and plastic and as having a marked yield value, are
thixotropic (in our terminology) if the concentration of the solid
is suitably chosen, whereas if they were dispersed to a mobile
liquid by the right dispersing agent, they would show dilatancy,
again at a suitable concentration of the solid. Green,^" at a time
when the phenomenon of thixotropy was hardly known, pubhshed
striking photomicrographs of plastic pastes of ZnO formed by sus-
pension in kerosene, which showed the ZnO particles to be coagu-
lated. When poppy seed oil was added as a dispersing agent, the
same particles were independent and dispersed. They are shown in
Figures 2a and 2b and are good examples of the state of the particles
in a thixotropic and in a dilatant system.

Our knowledge of the forces acting between the particles and
causing these phenomena is not yet sufficiently advanced to allow
me to state concisely, in the compass of a lecture, the processes
involved. Only a few points may be mentioned. The geloids in
thixotropic sols and gels are closely related to the factoids,^-' i. e.,
double refracting groups of oriented particles formed spontaneously
in concentrated sols containing nonspherical particles,-""' and to the
coacervates as defined by Bungenberg de Jong,-'^ i. e., hquid coagu-
lates of hydrophilic sols containing one or more kinds of colloidal
particles. In geloids, tactoids, and coacervates, the colloidal particles
are very far apart,-^^ up to many yi. It is, therefore, improbable that
the attraction between the particles is due to van der Waals' forces,

Sols and Gels — Relatio7i to Protoplasmic Structure 93

which act over smaller distances; an attraction due to electric forces
is more probable. The possibility has been considered that both
attractive and repulsive forces between the particles may be of
electric origin.''- But this concept is perhaps not easily reconciled
with the fact that the forces between the particles may be markedly
specific. Vanadium pentoxide, as well as benzopurpurin sols, form
tactoids which carry a negative electric charge. In a mixed sol of the
two, under proper conditions, both kinds of tactoids may be formed

■^- i>.'**»i*..

*'■■..■ -J* "^ ' J

B

Fig. 2.

side by side, containing only VoO-, and benzopurpurin particles
respectively.'^^ These two kinds of tactoids can be distinguished,
because those of V2O,-, are positively, those of the dyestuff negatively,
double refracting. The very specific behavior of some ions when
producing autocomplex coacervates must also be mentioned in this
connection.

The nonspherical shape of the colloidal particles strongly favors
thixotropy. But the opinion, which has been occasionally expressed,
that nonspherical shape is an indispensable factor for thixotropy
is not correct. Thixotropic systems are known whose particles
hardly deviate from the spherical shape, e. g., pastes of intact starch
grains in organic liquids.''

It is becoming more and more probable''^ that the thixotropic, i. e.,
isothermal change, sol ^^ gel, is the general phenomenon. When
dealing with the so-called, nonisothermal sol ^ gel transformation,
as we have it in aqueous sols of gelatin, agar, or methyl cellulose,^''
we must distinguish between (1) the actual sol ^ gel transforma-

94 The Structure of Protoplasjn

tion, which presumably has the same mechanism as the thixotropic
transformation and which is not changed essentially by a change of
temperature, and (2) a change in the solvation of the particles,
which depends pronouncedly on temperature.'^"' The latter factor
may secondarily influence the sol ^ gel change. Gelatin and agar
have the normal behavior — the gel is formed at low temperatures,
the sol at higher ones — whereas sols of methyl cellulose show the
so-called inverse, nonisothermal sol ^ gel transformation: The gel
is produced at higher temperatures and turns to a sol again on
being cooled.

The problems of thixotropy are always those of a special kind of
coagulation. Dilatancy is observed when the particles have a very
small, or perhaps even the least possible, tendency to be coagulated,
i. e., they do not stick spontaneously to each other at all. As long as
the suspension is exposed to weak mechanical forces, the particles
glide easily past each other, thanks to the continuous layers of liquid
between them. But if the forces applied exceed a certain limit, the
particles are brought much closer together. This may lead to a
displacement of the layer causing the independence of the particles —
the layer of oleic acid in the case of the iron particles, or the electric
double layer in aqueous suspensions — and the particles may be
made to touch each other. This will cause a high resistance toward
the acting force. As soon as the force stops acting, the particles
return to their independent behavior, which corresponds to a state
of equilibrium, and the suspension again assumes its liquid state.^'

We do not yet know whether thixotropy and dilatancy are
always produced by the same kind of mechanism or whether one will
have to distinguish different types of these phenomena in the future.
These do not represent the only characteristic features of anomalous
viscosity in colloidal solutions and in suspensions. One other phe-
nomenon may be mentioned, because it has been confused with
dilatancy.

Gels of V-Or, and of bentonite, suspensions of clays, etc., are
liquefied by strong shaking and stirring, but a weak movement — a
tapping or rolling of the vessel containing the liquefied system —
may markedly increase the rate of setting to a solid gel or paste.
This phenomenon has been called rheopexy.^*^ Striking experiments
were done by Hauser and Reed ^^ with aqueous bentonite sols; these
sols were fairly homodisperse, the particle diameter of the finest
dispersion being about 15 mji (a) , that of the coarsest 35 mf^i (b) . The
times of normal thixotropic setting were 42 seconds and 70 minutes

1

Sols and Gels — Relation to Protoplasmic Structure 95

for a and b respectively. The rheopectic effect was produced by
grasping the tube containing the sol between two fingers and oscil-
lating it like a pendulum about the point where it was grasped. The
times of rheopectic setting were about 10 seconds and 40 seconds
for a and b respectively. In rheopexy, the nonspherical shape seems
to be a decisive factor in causing the phenomenon. The gel (or
paste) produced by rheopectic action does not appear to have the
same structure as the one that has set spontaneously. The difference
between dilatancy and rheopexy is obvious: In dilatancy the final
state is liquid; the system behaves as a sohd only as long as the
external force is acting. In rheopexy the final state is solid; the
external force, causing the slight movement, only increases the rate
of solidification.

The mechanical forces causing liquefaction in thixotropy and the
increased rate of solidification in rheopexy may be replaced by the
action of ultrasonic waves.-^^' ^^ These phenomena deserve to be
discussed here, because the action of ultrasonics upon organisms has
already produced some interesting results and may lead to a better
discrimination of the manifold influence of mechanical forces upon
living systems.^^'*

Thixotropic gels like those of iron oxide, etc., are liquefied by
ultrasonics of sufficiently high energy.^^ Experiments of this kind
are done by simply dipping the test tube containing the gel into the
oil fountain that is formed above the vibrating quartz plate gener-
ating the ultrasonic waves; the plate is lying in an oil bath. This
effect is one of so-called cavitation.^" The waves produce periodic
dilations and compressions in the systems through which they pass.
The dilations may be intense enough to cause the liquid to tear, i. e.,
a cavity is formed, filled with the vapor of the liquid. This cavity
may collapse, if it again gets into a region of higher pressure. This
collapse of a cavity — to which we are referring if we speak of an
effect of cavitation — may lead to very high local concentrations of
energy and hence to effects like those of an explosion. It is appar-
ently mainly this phenomenon that produces the strongly destruc-
tive effects of ultrasonics.^" Cavitation is, for instance, the cause
of the strong dispersing action of ultrasonics upon mixtures of
organic liquids, such as benzene and water, where it has been inves-
tigated thoroughly. A collapse of cavities does not occur in vacuo,
the cavities formed simply increase in size; under a sufficiently high
outside pressure, on the other hand, cavities are not formed at all.^"
Hence, an action of ultrasonics due to cavitation is observed only

96 The Structure of Protoplasm

in a certain range of pressures and it may, therefore, be distinguished
from other actions of the waves not caused by cavitation. In vacuo
and at a suitably high pressure, thixotropic gels are not liquefied
by ultrasonics, which proves that the phenomenon is produced by
a collapse of cavities. ^^

The thixotropic decrease of structural viscosity in concentrated
colloidal solutions can also be produced by ultrasonics as an effect
of cavitation. This was shown for sols of gelatin where the effect
was partially reversible."'- In many cases effects are observed which
are due to cavitation, but are irreversible and which, therefore, must
be explained somewhat differently. In the sols of methyl cellulose
mentioned above, the structural viscosity left over, after cooling a
gel which had been formed at higher temperatures, can be made to
disappear by ultrasonics.^" In colloidal solutions of hemocyanin
{Helix pomatia) , ultrasonics split the molecules in halves; on pro-
longed action, a certain percentage of eighths appears. ^^ Perhaps such
a splitting effect occurs to a smaller or larger extent in many solu-
tions of highly polymerized organic substances.^-

Besides the destructive action due to cavitation, ultrasonics also
exert a coagulating effect which becomes conspicuous, both at high
energies and at such low ones that no more effects of cavitation are
observed.^"' Coagulation is very striking when ultrasonics act upon
fogs or smokes.^*^ It is also shown by the fact that, on emulsifying two
liquids like water and benzene in each other, a limiting value of
particle concentration is reached, where the dispersing effect is
balanced by the coagulation.^"^ In emulsions and suspensions of
suitable concentration, stationary ultrasonic waves cause an accumu-
lation of the particles in the nodes or antinodes, depending on whether
the particles are lighter or heavier than the surrounding medium.^^
Coagulation may be observed where this accumulation takes place.
Coagulation by ultrasonics may be due to several reasons which are
not easily distinguished from each other. Two may be mentioned:
(1) Attractive forces of a hydrodynamic nature are produced
between particles, if the particles are suspended in a medium exposed
to vibration.'' (2) The so-called orthokinetic coagulation may cause
collisions and hence a clustering of particles. ^"^ This latter kind of
coagulation occurs if the particles do not have equal sizes and hence
move with different rates, when exposed to external forces.

The marked effect of ultrasonics on some rheopectic suspensions
is due to coagulation. If a suspension of a kaolin ("Stockalite") in
an aqueous solution of NaCl was liquefied by shaking and exposed

Sols and Gels — Relation to Protoplasmic Structure 97

to ultrasonics, which were too weak to cause cavitation but strong
enough to cause coagulation, it solidified in 15 seconds, whereas it
remained liquid for 17 minutes if left to itself; it took 30 seconds to
cause the rheopectic solidification by tapping.'''' A dilatant suspension
is kept liquid by ultrasonics; it does not show the characteristic
increase in mechanical resistance, when treated with a glass rod.-^^
This effect is not due to cavitation.

Finally, there is still a third effect, which appears with very weak
ultrasonics: Rod- or plate-like particles of a suspension are oriented
with their long axes vertical to the direction in which the energy of
the waves flows. •''^

FOOTNOTES

'Verworn, M., Allgemeine Physiologie, Jena, (1895); Wilson, E. B., The Cell
in Development and Inheritance, New York, (1897).

' For a general survey of thixotropy, see H. Freundlich, Thixotropy, Actualites
scientifiques et industrielles, No. 267, Paris, Hermann et Cie., (1935).

'Freundlich, H., and Roder, H. L., Trans. Farad. Soc, 1938, 34. 308; R6der,
H. L., Rheology of Suspensions, Ph. D. Thesis, Amsterdam, 1939.

^ Sameshima, I., Bl. chem. Soc, Japan, 1926, 1, 255.

■ Laing, M. E., Jour. Phys. Chem., 1924, 28, 673.

'■ Freundlich, H., and Seifriz, W., Z. phijs. Chem., 1923, 104, 233.

• Szegvari, A., and Schalek, E., Kolloid-Z., 1925, 32, 318; 33, 326; and many later
articles, mostly by Freundlich and co-workers; for references, see Freundlich,
H., Kolloid-Z., 1928, 46, 289, and FreundUch, H., Thixotropy, 1935 (see footnote
2).

'Freundlich, H., Schmidt, O., and Lindau, G., Z. phys. Chem., Bodenstein-
iestband 1931, 333; Kolloid Beihejte, 1932, 36, 43; Hauser, E. A., and Reed, C. E.,
Jour. Phys. Chem., 1937, 41, 911.

' Edsall, J. T., and Mehl, J. W., Jour. Biol. Chem., 1940, 133, 409; Mirsky, A. E.,
Cold Spring Harbor Symposia on Quantitative Biology, 1938, 6, 150.

'"LaufTer, A. M., and Stanley, W. M., Kolloid-Z., 1940, 91, 62 (no mention is
made in this paper of thixotropy, but there is anomalous viscosity).

" Zocher, H., and Albu, H. W., Kolloid-Z., 1928, 46, 27.

'•' Zocher, H., and Albu, H. W., Kolloid-Z., 1928, 46, 33.

'Freundlich, H., and Abramson, H. A., Z. phys. Chem., 1927, 131. 278; Heller.
W., Compt. rend., 1936, 202, 1507.

" Freundlich, H., and JuUusburger, F., Trans. Farad. Soc, 1934, 30, 333;
Freundlich, H., and Jones, A. D., Jour. phys. Chem., 1936, 40, 1217.

'' Reckhnghausen, H. V., Kolloid-Z., 1932, 60, 34.
'"Green, H., hid. and Eng. Chem., 1923, 15. 122.
''Reynolds, O., Phil. Mag. (5), 1885, 20, 469; Nature, 1886, 33, 429.
'' Olze, A., and Daniel, F. K. quoted by Freundlich, H., and Roder, H. L.,
Trans. Farad. Soc, 1938, 34, 308.

" Freundlich, H., and Jones, A. D., Jour. Phys. Chem., 1936, 40, 1217.

'"Freundlich, H., and S511ner, K., Kolloid-Z., 1928, 45, 348.

"Heller, W., Compt. rend., 1936, 202, 61; Jonr. Phys. Chem., in press.

'^ Heller, W., and Quimfe, G., Compt. rend.. 1937, 205, 857; Heller, W., and
Vassy, E., Co-nipt, rend., 1939, 208, 812, Jour. phys. Chem., in press.

98 The Structure of Protoplasm

^ Kandelaky, B. S., Kolloid-Z., 1936, 74, 200.

" Goodeve, C. F., Trans. Farad. Soc, 1939, 35, 342.

='Freundlich, H., and SoUner, K., Kolloid-Z., 1928, 44, 309.

'" See, e.g., Bungenberg de Jong, H. G., La coacervation, les coacervats, et
leur importance en biologic. Volumes 1 and 2, Paris, Hermann et Cie., 1936.

"Daniel, F. K., hidia Rubber World, 101, Nos. 3 and 4, 1939.

'' Freundlich, H., Jour. Soc. Chem. Ind., 1934, 53, 223 T., and especially South-
ern, W. A., Ph. D. Thesis, London, 1939.

^ Verwey, E. J., and de Boer, J. H., Rec. trav. chivi., 1938, 57, 383.

'" As to the nature of tactoids, see: Zocher, H., Z. anorgan. u. allg. Chem., 1925,
147, 91; Coper, K., and Freundhch, H., Trans. Farad. Soc, 1937, 33, 348; Heller, W.,
ComiJt. rend., 1935, 201, 831.

■■'^ Heller, W., unpublished.

''Levine, S., Proc. Roy. Soc. London, (Series A), 1939, 170, 145, 165; Levine, S.,
and Dube, S. P., Trans. Farad. Soc, 1939, 35, 1125, 1141.

'= Freundlich, H., Enslin, O., and Sollner, K., Protoplasma, 1933, 17, 489.

'^ Roder, H. L., Rheology of suspensions, A study of dilatancy and thixotropy,
Ph. D. Thesis, Amsterdam 1939; Heller, W., Compt. rend., 1938, 207, 157.
''Heymann, E., Trans. Farad. Soc, 1935, 31, 846; 1936, 32, 1.
'' Freundlich, H., and Juliusburger, F., Trans. Farad. Soc, 1935, 31, 920.
" Hauser, E. A., and Reed, C. E., Jour. Phys. Chem., 1937, 41, 911,

^^ Freundlich, H., Kapillarchemie, Leipzig, 4th ed., vol. 2, 1932, 616; Freund-
hch, H., Rogowski, F., and Sollner, K., Z. phxjs. Chem., A., 1932, 160, 469; Kolloid
Beihefte, 1933, 37, 223; Freundlich, H., and Sollner, K., Trans. Farad. Soc, 1936,
32, 966.

'■° Burger, F. J., and Sollner, K., Trans. Farad. Soc, 1936, 32, 1598.

^^^ Some newer reviews of the action of ultrasonics on biological objects are:
Dognon, A. and Biancani, E. H., Ultra-Sons et Biologie, Paris, Gauthier-Villards,
1937; Hiedemann, E., Grundlagen und Ergebnisse der Ultraschallforschung, Ber-
lin, Walter de Gruyter and Co., 1939, pp. 179-185.

*" See, e. g., Bondy, C, and Sollner, K., Tratis. Farad. Soc, 1935, 31, 835; Sollner,
K., Jour. Phys. Chem., 1938, 42, 1071.

" Freundlich, H., and Sollner, K., Trans. Farad. Soc, 1936, 32, 966.

"Freundlich, H., and Gillings, D. W., Trans. Farad. Soc, 1938, 34, 649. See
there many additional references.

''Heymann, E., Trans. Farad. Soc, 1935, 31, 846.

"Brohult, S., Nature, 1937, 140, 805.

'' Solhier, K., and Bondy, C, Trans. Farad. Soc, 1936, 32, 616.

^^ Patterson, H. S., and Cawood, W., Nature, 1931, 127, 667; Brandt, O., and
Hiedemann, E., Trans. Farad. Soc, 1936, 32, 1101; Kolloid-Z., 1936, 75, 129; Brandt,
O., Freund, H., and Hiedemann, E., Kolloid-Z. 1936, 77, 103; da C. Andrade, E. N.,
Trans. Farad. Soc, 1936, 32, 1111; etc. The best summary in: Hiedemann, E.,
Grundlagen und Ergebnisse der Ultraschallforschung, Berlin, 1939.

" Bjerknes, V., Vorlesungen liber hydrodynamische Fernkrafte nach C. A.
Bjerknes' Theorie, Leipzig, 1900-01; Kbnig, W., Ann. Phys. (Wiedemann), 1891,
42, 353, 549. Brandt, O., Freund, H., and Hiedemann, E., Kolloid-Z., 1936, 77, 103;
Zeitschr. j. Physik, 1937, 104, 511; etc.

^^ Sollner, K., and Bondy, C, 1. c. 45; SoUner, K., Trans. Farad. Soc, 1936, 32,
1119; Brandt, O., and Hiedemann, E., and Brandt, O., Freund H., and Hiedemann,
E., 1. c. 46; for a review see pp. 199-201 in Hiedemann, E., Grundlagen und Ergeb-
nisse der Ultraschallforschung, Berlin, 1939.

v^\C/1^

STRUCTURAL DIFFERENTIATION OF CYTOPLASM

G. W. SCARTH

McGill University

Theories as to the basic structure of protoplasm usually assume
that there is a single master key to the problem. But if differ-
entiation exists in the physical properties of different regions
of the cell, it is necessary to consider the question of correspond-
ing differences in submicroscopic structure. The present paper is
concerned with the cytoplasm and, because of considerations of
time and relative importance, will deal only with the continuous
ground substance, omitting the various inclusions.

The type of cytoplasmic differentiation which seems to be most
general is that of a series of concentric zones, not usually defined
by any sharp vi.sible boundary but each showing distinctive phys-
ical properties when studied experimentally.

INTERIOR VS. CORTEX

As regards animal cells, viscosity tests of all kinds agree in
proving that the bulk of the internal cytoplasm of echinoderm eggs
and sometimes of amoeba and other protozoa is relatively fluid.
The same seems to be the case with most kinds of animal cells
though there appear to be exceptions such as epidermal and cer-
tain other epithelial cells. Along with the usual fluidity of the
interior, its water miscibility, solvent properties, and electrical con-
ductivity point to its being a sol with aqueous dispersion medium.

On the other, hand, a cortical zone of varying thickness always
has much higher consistency. This region combines elastic with
plastic or fluid qualities to a degree which is rarely if ever paralleled
in physical systems; the combination becomes more intelligible in the
light of its marked capacity for undergoing rapid reversible gel-
sol transformation. In amoeboid and other movements involving
vortical streaming, not only do reciprocal sol-gel changes occur
but the cytoplasm circulates between the internal and cortical
regions. However, the whole of the cortex does not liquify at
once; some coherent structure persists as a basis of organization.

[99]

100 The Structure of Protoplasm

As regards plant cells it was through the investigations of my
co-workers, Siminovitch and Levitt, on cytoplasmic behavior in
relation to frost resistance that the widespread occurrence in such
cells of the same differentiation into gelatinous cortex and liquid
endoplasm or mesoplasm was impressed upon me. In a non-
vacuolate type, the pollen mother cells of Trillium, an outer plas-
magel zone of variable thickness and an inner fluid one are easily
demonstrated with the aid of a little micromanipulation (Stern).
In vacuolate cells having a thick layer of cytoplasm, such as the
cortical cells of many trees especially in winter time, the same dis-
tinction is apparent when rupture occurs during deplasmolysis. A
thin cortical layer tears the rest spills out and mixes with water.
Even in the thinnest cytoplasmic layers, this differentiation is
revealed by a technique based on Chambers' method of applying
oil drops to the cell surface. An oil drop snaps into the surface of
a plant protoplast, taking the form of a biconvex lens. When the
protoplast is stretched (by deplasmolysis or pressure) the oil drop
is pulled out into a flatter and ultimately concavo-convex shape.
When again the protoplast is allowed to contract, the oil goes
through reverse change in shape, and if the contraction goes far
enough it becomes spherical. Evidently the tension on the edge
of the oil lens is that of an elastic, not a liquid, film. But while
the cortex is thus shown to shrink and stretch elastically, the meso-
plasm flows into and out of the angular space between the edge
of the drop and the surface of the vacuole, and must therefore be
quite fluid (Siminovitch and Levitt) .

PROTOPLASMIC SURFACE FILM OR ECTOPLAST

The cortical gel layer, when wide enough to be seen, often
shows little optical differentiation from the liquid endoplasm ex-
cept as regards restriction of Brownian movement. Both zones
are commonly granular. In contrast to this there is distinguish-
able on the surface of many cells a thin layer of more hyaline
appearance. This has been called ectoplasm or the ectoplast, but as
the term has also been applied to the cortical gel in amoeba. Cham-
bers prefers to call it the "protoplasmic surface film." Much of our
knowledge of this layer comes from Chambers' studies of echinoderm
eggs. He gives many proofs of its fluidity and shows that its con-
tinuity over the surface is essential to the integrity of the proto-
plast, whereas other membranes which normally exist external to

Structural Differentiation of Cytoplasm 101

this can be removed without harm to the cell. The hyaline film
on amoeba is also fluid and cohesive. According to Mast the fluid
film is here bounded by a more rigid pellicle, the plasmalemma.

In the case of plant cells a so-called hyaline layer on the surface
becomes visibly apparent only in very fine strands pulled out from
the surface. Incidentally, though it has been the custom to call the
film hyaline, it is by no means free from granules. But the very
fact that strands of fluid consistency can be pulled out from the
surface of a cell which has been dehydrated by plasmolysis until its
gel layer is rigid and brittle, is evidence for the existence of a differ-
entiated liquid layer overlying the gel. The fine strands which
connect a plasmolysed plant protoplast with the cell wall, being
composed entirely of the film substance, reveal some of its proper-
ties by their behavior. In strong plasmolytes they stiffen suffi-
ciently to crumple up on breaking, but tenuous as they are, this
stiffening seems to be confined to a still thinner surface film, since
under the same conditions of dehydration they become beaded on
standing due apparently to the running into drops of a more liquid
core. All this indicates that the protoplasmic surface layer may have
a denser film of its own — compare the plasmalemma of amoeba. In
spite of this capacity for hardening at its surface, the ectoplast
shows little sign of elastic extensibiUty. That quality of the proto-
plasm seems to reside mainly in the layer beneath.

Some properties of the surface film in plant cells are brought
out by the behavior of applied oil drops. On making contact with
the protoplasmic surface, a body of oil, even when large, relative
to the cell, immediately has its surface tension greatly reduced so
that it will often maintain shapes of nonminimal area. That the
surface-active material spreads as a film over the oil is shown by the
speed with which the change occurs and by the dragging of gran-
ules and chloroplasts with it, especially when the oil drop is large.
Further, the transport of these bodies through the cytoplasm to the
drop seems to prove bodily movement of a film over the proto-
plasm. The movement of the film is not prevented by plasmolysis
which is strong enough to make the gel layer quite stiff, which is
further proof that the surface film is distinct from the gel layer.

THE TONOPLAST

At the inner cytoplasmic surface in contact with the central
vacuole is another film, the tonoplast of de Vries, which is often

102 The Structure of Protoplasm

left intact when the rest of the cytoplasm has become disorganized
and separated from it. The isolated tonoplast persists as a fluid
semipermeable sac enclosing the vacuolar sap, and may preserve
its integrity for days or even weeks. By contrast, the external
surface film has never been completely separated from the proto-
plast, but probably this is due simply to its position, as local
elevation sometimes occurs naturally and may be produced artifi-
cially. Both surfaces of the tonoplast resemble the surface of the
ectoplast in showing immediate adhesion to oil but adhesion to
glass only after mechanical disturbance. An oil drop which has
snapped onto the outer surface of a freed tonoplast is deformed
somewhat by stretching of the latter but not to the same extent as
at the cell surface. It is possible that this slight display of elastic
property may be due to an adsorbed film of protein from the endo-
plasm. Chambers and Hofler emphasize the highly fluid, though
cohesive and extensible, nature of the tonoplast and the immisci-
bility of its substance with water.

On the whole the physical properties of the tonoplast and ecto-
plast seem to be very similar, and Plowe has described how the
two films may merge to form a single envelope.

IS THERE AN INNER GEL LAYER?

Since the cytoplasm of the plant cell is bounded on its vacuolar
side by a fluid layer similar to the surface film, the question arises
whether there is also an inner zone which corresponds to the cor-
tical plasmagel. The frequent elasticity of transvacuolar strands
and of strands pulled by micro-needles into the vacuole, as com-
pared with the relatively inelastic quality of strands pulled from
an isolated tonoplast, suggests that the elasticity of the former
resides in the core of endoplasm which such strands usually con-
tain. On the other hand, an oil drop applied to the surface of the
vacuole sinks into the cytoplasm much more easily than a similar
drop applied to the outside, which indicates that if an inner gel
layer exists it is much less substantial than the outer one.

OTHER CYTOPLASMIC FILMS OR MEMBRANES

It seems reasonable to regard the colorless film which covers
chloroplasts and can be elevated from them in pathological swell-
ing as being equally with the ectoplast and tonoplast a cytoplasmic
structure. Even the nuclear membrane, or part of it at least (since

1

Structural Differentiation of Cytoplasm 103

in elevation it sometimes splits into two), may belong to the cyto-
plasm. All these films show many points of resemblance, among
which is their relationship to the kinoplasm.

KINOPLASM
This is a structure of more variable occurrence than those al-
ready described, but it seems worthy of mention in a discussion of
the fundamental differentiations of cytoplasm because of its pos-
sible physiological importance and also because of its apparent
relation to the various surface films. Typically the differentiation
to which Strasburger gave the name kinoplasm appears as more or
less mobile filaments or tubules of very slightly higher refractive
index than the matrix in which they are imbedded. To be seen at
all, they must be in sharp focus, which may be the reason why they
have been observed mainly in plant cells in which the thinness of
the cytoplasm keeps them almost in one plane. The chains of
granules carried along by the kinoplasm are, however, easily
observed.

From the point of view of structure, the visibile connection be-
tween some of the kinoplasmic strands and the various surface
films is of interest. Strands can be seen to flow out of and into the
envelope of chloroplasts and of the nucleus. Sometimes they swell
to enclose small vacuoles. Streams of kinoplasm have been ob-
served to converge upon an applied oil drop at the moment of con-
tact. This points to a connection between the outer surface film
and the strands and also explains how granules and plastids im-
bedded in the cytoplasm but attached to the kinoplasm come to be
carried over the oil drop. It is not possible from a surface view of
the cytoplasm to observe at what level the strands lie. In a profile
view of the thicker portions of the cytoplasm of some cells, the
serial procession of granules which marks the position of a kino-
plasmic stream can be located. It ranges from the outer to the
inner boundary, and furthermore, streams may arch beyond the
general level of the cytoplasm into the central vacuole.

The shapes and movements of kinoplasm recall the typical myelin
processes which often adorn the inner surface of the tonoplast and
which have also been observed in a number of cases on the outer
surface of the ectoplast. It is not suggested that they are identical,
since myelin processes proper are random structures and often
(though not always) the result of an abnormal environment, whereas

104 The Structure of Protoplasm

the internal kinoplasm is a normal and presumably functional
organization. Yet the tendency of cytoplasmic surface films in
general to form these myelin-like outgrowths on both their surfaces,
has significance with respect to the composition and molecular
structure of both the films and the fibrils.

At this point attention may be drawn to Figure 1, which shows
diagrammatically the supposed position and relationship of the
differentiations described.

MOLECULAR STRUCTURE

On the basis of physical properties and probable structure, the
cytoplasmic differentiations fall into two groups. The plasmasol and
plasmagel, being reciprocally transformable, are merely different
physical states of one substance. The same relation seems to hold
between the films and kinoplasm. But the two pairs show no sign
of being interchangeable, and what is more significant, they ex-
hibit fundamentally different properties. Those of the sol-gel
region, the endoplasm, include elasticity, thixotropy, stream bi-
refringence, etc., which, as shown by the previous speaker, point to
elongated protein units as the building stones. On the other hand,
the properties of the film-kinoplasm complex are prominently
lipoidal, though no doubt proteins also enter into its composition.
Incidentally, while the film substance occupies much the lesser part
of the protoplast, yet (if such a distinction can be made) it is the
more essential part, being capable of preserving an independent
existence for a time, which the endoplasm is unable to do.

Let us see what these lipoidal properties are. Already men-
tioned is the tendency to myelin formation, a phenomenon confined
to a limited group of substances including phosphatides, cerebro-
cides and trioleates — all ingredients of protoplasm. It depends on
the assumption of paracrystalline structure, consisting of parallel
sheets of orientated molecules with the principal optical axis and
longer diameter of the molecules at right angles to the plane of
layering. The type of birefringence which results from this form
of structure tends to disappear with a high degree of hydration and
reappear on dehydration. Accordingly, it is significant that such
double refraction develops in the hyaline layer on the amoebocytes
of various invertebrates when partially dehydrated (Faure-Fremiet,
see Schmidt). Masses of myelin-forming material which, under
certain conditions, accumulate on the lining of the plant vacuole.

Structural Dijferentiation of Cytoplasm

105

afford confirmatory evidence of their composition by staining with
lipoid soluble dyes such as chrysoidin and by blackening with osmic
acid. These phenomena indicate that the film-forming substance is
rich in lipoids as compared with the protoplasm as a whole.

The lamellar type of structure, of which there is optical evi-
dence in myelin figures and in the dehydrated surface film of some
cells, is much more highly
developed in irreversibly
differentiated structures
such as chloroplasts and
the myelin sheath of nerve.
A nerve fiber may be re-
garded as a sort of cry-
stallized filamentous pseu-
dopodium in which the
axon or endoplasm is
largely a bundle of parallel
protein fibrils and the
sheath or ectoplasm con-
sists of concentric sheets
of orientated lipoid mole-
cules alternating with
sheets of protein.

Although it requires ex-
ceptional conditions to
produce visible double re-
fraction of ordinary proto-
plasmic films, molecular layering may be present at all times toward
their surface where the conditions for orientation operate more
strongly. (Contrary to what we might except, however, the usual
lipophilic adhesion of the cell surface would seem to indicate that the
hydrocarbon ends of the lipoid molecules point outward.) A closely
packed, superficial lipoid film is, of course, in harmony with the gen-
eral conception of the protoplasmic surface derived from the study
of permeability. The properties of the hypothetical plasma mem-
brane probably belong to an organized film of molecular dimensions
(perhaps including proteins as well as lipoids) rather than to the
ectoplast as a whole.

As to the colloidal state of the isotropic film substance, its appar-
ent immiscibility with water points to a coacervate rather than a

Fig. 1. Diagrammatic cross section of the
cytoplasm of a plant cell showing differentia-
tions, w — cell wall, e — ectoplast or proto-
plasmic surface film, pg — plasmagel layer or
cortical endoplasm. ps — plasmasol or liquid
endoplasm. k — kinoplasm. t — tonoplast. v —
vacuole, s — transvacuolar strand, m — mye-
lin processes, p — plastid.

106 The Structure of Protoplasm

sol. In a coacervate, according to Bungenberg de Jong, the col-
loidal units are attracted by electrostatic force but kept apart by
that of solvation. Thus the aggregrate, while coherent, may be
quite fluid. The whole cell has been compared by de Jong to a
compound mixed coacervate, but as already pointed out the liquid
endoplasm has normally the property of a sol. On the other hand,
so-called emulsion structure when it appears in cytoplasm may
well be a coacervation phenomenon, and reversible coacervation
(droplet precipitation) of colloids in the vacuole is a common oc-
currence. Compared with this the whole substance of a surface
film is suggested to be an irreversible coacervate mass having a
very low surface tension.

Of the types of coacervate which have so far been studied ex-
tensively by Bungenberg de Jong, those which most resemble the
protoplasmic films are the auto-coacervates of phosphatides. Among
the points of resemblance are:

relative irreversibility, due apparently to the formation of a
stable layer of orientated molecules at their surface;
lipophilic adhesion, due apparently to the hydrocarbon
chains of the orientated superficial molecules being directed
outward;

myelin forvnation under certain conditions;
vacuole formation as a result of reduced colloidal hydration
(shared by other coacervates) along with nonfusion of the
vacuoles when pressed together, due apparently to separa-
tion by a very stable bimolecular film; and
change from optical isotropy to anistropy with dehydration.
There is also fair agreement between the action of electrolytes
on the packing of phosphatide films and on cell permeability, re-
spectively. On the other hand, their effect on electrophoresis of
cells is unlike that of phosphatide drops, inasmuch as the sign of
the charge at the cell surface is not so easily reversed. Perhaps
further study of coacervates which contain proteins as well as
lipoids may furnish still better models of the protoplasmic surface
film.

In conclusion, let me point out the moral which my story is
meant to teach. It is the fallacy of taking a partial view to be a
true picture of the whole. That was the error of the blind men of
Indostan in their theorising on the structure of the elephant. I
cannot do better than parody this clever fable to suit the present

Structural Differentiation of Cytoplasm 107

In the original were six blind men. I shall list only four,
leaving the reader to complete the number as he sees fit:

case

It was four fundamentalists to learning much inclined,

Who went to see the Protoplast (though all of them were blind)

That each its structure might observe to satisfy his mind.

The first advancing hurriedly and happening to fall
Right through its soft interior at once began to bawl
"God bless me! But the Protoplast is very like a sol."

The second poked the animal and felt his staff repel
Its tough and springy cortex, so he began to yell
•• 'Tis evident the Protoplast is very like a gel."

The third approaching gingerly did only pinch and squeeze
Its slippery oleaginous hide when he began to wheeze
"It seems to me the Protoplast is just a lump of grease."

The fourth man, having punched and probed and proved its plastic state,

Watery yet indissoluble, did thus asseverate

"The Protoplast is a compound, complex co-a-cerv-ate."

And so these fundamentalists disputed loud and long

Each in his own opinion exceeding stiff and strong,

Though each was partly in the right and all of them were wrong.

SOME KEY REFERENCES

BuNGENBERG DE JoNG, H. G. (1936) "Actualites Scientifiques et Industriellesl"

Exposes de Biologie, La Coacervation 6 and 7.
AND J. Bonner. (1935) "Phosphatide auto-complex coacervates, etc."

Protoplasma 24, 198.
Chambers, R. (1938) "The physical state of protoplasm with special reference

to its surface." Amer. Nat. 72, 141.

AND K. HoFLER. (1931) "Micrurgical Studies on the tonoplast of

Allium cepa." Protoplasma 12, 338.
Frey-Wyssling, a. (1938) "Submikroskopishe morphologie der protoplasmas

und seiner derivate." Protoplasma monograph.

Plowe, J. K. (1931) "Membranes in the plant cell." Protoplasma 12, 196.

Scarth, G. W. (1927) "Structural organization of protoplasm in the light of
Micrurgy." Protoplasma 2, 189.

Scarth. G. W., J. Levitt, and D. Siminovitch. (1940) "Plasma membrane struc-
ture in the light of frost hardening changes." Cold Spring Harbor
symposium 8, 102.

Siminovitch, D., and J. Levitt. (1941) "Relation between frost injury and physi-
cal state of protoplasm, ii. The protoplasmic surface." Can. Jour, of
Research, 19, 9.

Schmidt, W. J. (1937) "Die doppelbrechung von karyoplasma, zytoplasma und
metaplasma." Protoplasma monograph.

STRUCTURAL DIFFERENTIATION OF THE NUCLEUS

C. L. HUSKINS

McGill University, Montreal

From studies on protoplasm in general, presented by previous
speakers in this Symposium, we may proceed to a consideration of
the elaborate structural features which are present throughout the
life of the nucleus, even in the "resting" stage when many of them
are not directly observable. These are today being studied from
many different points of view and by very diverse techniques. To
simplify, and perhaps to promote, discussion we may classify these
attacks upon the problems of the nucleus into four groups, though
recognizing, of course, that there are many studies which overlap
the boundaries that are, rather arbitrarily, here laid down. Apart
from simplifying discussion, the classification may also serve a
useful purpose in emphasizing that some studies of the nucleus pro-
ceed in rather extraordinary isolation from others and that general-
izations made by workers of one group often ignore the data of the
other groups or, in soine cases, draw unwarranted conclusions from
them. The diversity in points of view and of detailed opinion found
among students of the nucleus indicate in themselves the complexity
of the structures within the boundary of the nuclear membrane and
the scope of the problems that remain to be elucidated.

In the first of the four groups we may place all studies aiming at
the elucidation of the "submicroscopic" structure of the nucleus.
Often such investigations use wave lengths shorter than those of
visible light. The second level of analysis comprises all microscopic
studies with visible light of killed and fixed materials. The third
includes all the varied studies on the living cell made by the methods
of "experimental cytology." The fourth comprises the interpretation
of structure derived from observations of function made with the
techniques of genetics.

Workers in any of these fields should, of course, be familiar with
all four. Unfortunately, few of us are! Accepting a more limited
standard, those working on the first level viust, of course, be familiar
with the data of the second group. The second group must today
know the techniques and discoveries of the geneticists if they are

[109]

110 The Structure of Protoplasm

to avoid drifting again into the relative stagnation from which their
subject has fairly recently emerged. The geneticists interested in
nuclear structure must know the data of the first and second groups.
The third group needs to know all the data of all the groups.

Much of my own work falls jointly into fields two and four. At
present I am particularly impressed, perhaps depressed, by the
divergencies of opinion between workers on the second level — the
one with which I am most familiar. These I consider the greatest
source of danger for those workers of the other groups who seek to
bolster or check the hypotheses formed from their own observations
by the data of the second group. This difficulty in its relation to vari-
ous aspects of the general problems of nuclear structure will there-
fore be stressed.

Studies on the presence and distribution of nucleic acids made
jointly on the microscopic and submicroscopic levels, have been of
particular interest in recent years. The studies with visible light
have depended largely on the Feulgen reaction; those at a lower
level on the ultraviolet absorption technique developed by Caspers-
son. With the Feulgen method (the application of Schiff's reagent
to the aldehyde groups of the pentose constituents of nucleic acids) ,
desoxyribose (ribodesose-, "thymo-") nucleic acids give a positive
staining reaction, while ribose ("yeast-") nucleic acids do not stain.
This test is, however, of only limited value in making quantitative
determinations of thymonucleic acid in cytological preparations,
The method of Caspersson (1936, 1940) permits quantitative deter-
minations of nucleic acids, but the ultraviolet absorption spectra
with characteristic maxima at 2,600 A, "being determined by the
nitrogenous constituents," do not differentiate between the desoxy-
ribose and ribose types. For the study of the nucleic acid distribution
in the cell, a combination of both methods is thus indicated.

Studies by the Feulgen method have shown clearly that desoxy-
ribonucleic acid is concentrated in the chromosomes and is either
absent or present only in very much smaller amounts in the cyto-
plasm and nucleolus. Since thymonucleic acid appears to be the
"typical chromosome nucleic acid" Caspersson (1939) believes that
it is synthesized on the chromatid. The numerous objections to the
use of the Feulgen method in cytology have almost all been ade-
quately dispelled by Milovidov (1938) , Hillary (1940) , and others
who showed that when properly used, "chromatin" and only chroma-
tin is stained by it, as Feulgen claimed. The remaining difficulty,
that at certain stages in the maturation of the animal egg the chromo-

Structural Diferentiation of the Nucleus 111

somes cannot ordinarily be stained by Feulgen, has recently been
removed by Brachet (1940) , who finds, contrary to Koltzoff (1938)
and others, that the chromosomes are never entirely Feulgen-
negative during oogenesis. It is easy to see them throughout all
meiotic stages in Triton, and also in many insects, but they tend to
disappear at mid-prophase in the frog. Brachet concludes that it is
the changing size of the chromosomes and the degree of dispersion
of the thymonucleic acid that causes variation in staining capacity.
Here is the first of several examples where differences of opinion
between workers on the microscopic level can lead to extreme
divergence in conclusions about the submicroscopic. Koltzoff, find-
ing no reaction with Feulgen at certain stages, says that thymo-
nucleic acid cannot be an essential component of the gene, but only,
perhaps, its protector. Brachet, finding some Feulgen positive
particles at all stages, cannot agree.

With his ultraviolet absorption technique Caspersson (1936)
showed that: during mitosis nucleic acid is localized in and around
the chromosomes, and in collaboration with Hammersten and others,
he supplemented the absorption determinations with digestion
experiments in which trypsin preparations containing lanthanum
dissolve the protein and leave the nucleic acid as insoluble lanthanum
thymonucleate. With these two methods it was found that metaphase
chromosomes contain nucleic acids and proteins intimately mixed —
not as protein islands with surrounding zones rich in nucleic acids —
and that salivary gland chromosomes comprise segments alternately
rich in and free of nucleic acid. The former are the "bands" so
clearly seen after aceto-carmine fixation. Their highly organized
structure found after trypsin digestion, indicates that they are more
likely to be the seat of the genes than the interband regions which
were completely digested.

Mazia and Jaegar (1938) confirmed the trypsin results and per-
formed the complementary experiment of digestion with pepsin.
This, they found, "did not dissolve any constituent concerned in
maintaining the integrity of the chromosome." After treatment with
spleen nuclease, the chromosomes could not be stained with either
aceto-carmine or by the Feulgen method. They had not, however,
been disintegrated by the nuclease, as the ninhydrin reaction for
determining proteins stained them a deep blue while the cytoplasm
stained a lighter blue. These authors conclude that "the chromo-
some must have a continuous protein structure and the nucleic
acid must be attached in such a way that it may be split and the

112 The Structure of Protoplasm

pentose may be removed without affecting the continuity of struc-
ture." Since the protein of the chromosomes was not digested by
pepsin, they consider that it may be related to the protamines or
histones. The molecular structure of the chromosome which these
authors support is one of long polypeptide chains and parallel nucleic
acid molecules. Experiments with polarized light and X-ray diffrac-
tion studies reviewed by Frey-Wyssling (1938) and Schmitt (1940)
suggested the assumption of parallel structure and are, like Mazia
and Jaegar's data, against earlier concepts of a warp-and-woof fabric.
Caspersson (1936) observed that the nucleic acid content is
increased shortly before cell division, and after cell division it
decreases again. Later (1939), he showed that in grasshopper sper-
matocytes, nucleic acid "synthesis" is completed, i. e., the maximum
amount is reached, before "maximum condensation of the chromo-
somes." Since nucleic acids are present in large amounts before cell
division and evidence from other studies on the submicroscopic level
indicates that the self-reproducing viruses and phages also contain
appreciable amounts of nucleic acid in their molecules, an attempt
is made to correlate these data with observations from the second
level. Here trouble begins — or so at least it seems to me. The
generalization is first made that nucleic acid concentration is at a
maximum in the nucleus at about the same time as the chromosome
is believed to reproduce itself, and from this the conclusion is reached
that nucleic acid is fundamentally connected with gene reproduc-
tion. This is a plausible and, I should think, even highly probable
hypothesis. But, how is it supported by microscopic data? The
latter deal with chroviosome reproduction or "splitting," and there
is much evidence against and little in favor of the view that this is
the same thing as gene reproduction. As Caspersson points out,
chromosome "splitting" cannot be seen until there has been an
accumulation of nucleic acid, for it is this which stains. He says, for
this reason, that the observation of chromosome splitting in very
early prophase says nothing against the view that nucleic acid
formation and gene reduplication go together. But is it not also
reasonable to conclude that the association he has found says noth-
ing definitely for it? In itself the observation that nucleic acid con-
centration is at a maximum near the time that chromosome splitting
becomes obvious may mean nothing more than that nucleic acid con-
centration permits doubleness of the chromosome to be seen. How-
ever this may be, the attempted first and second level correlation runs
into further difficulties. The first of these is that descriptive

Structural Diferentiation of the Nucleus 113

cytologists, together with cytogeneticists who seek data from irradi-
ation experiments as well as direct observation, fall into two schools
regarding the time at which chromosome splitting takes place. And
the submicroscopic data that fit in with the views of one school
regarding meiosis seem to me to be against that school regarding
mitosis. Darlington (1937) and his school consider that meiotic
chromosomes reproduce themselves during late pachytene. This
fits Caspersson's data on the greatest concentration of nucleic acid
occurring just before the tetrad split becomes visible in grasshopper
spermatocytes. But Darlington is likewise convinced that repro-
duction of somatic chromosomes occurs during the resting stage
immediately preceding the mitosis in which separation of chromo-
some halves occurs — his whole "precocity theory" relating meiosis
to mitosis stands or falls on this diff'erence in the time of splitting
of meiotic and mitotic chromosomes. And during the resting stage,
when the chromosomes are believed by this school to split, the
nucleic acid concentration of the nucleus is at a minimum! One
should add that Darlington and La Cour (1940) state that "the
chromosome cycle is adjusted to have a maximum aggregate nucleic
acid attachment at metaphase" and that "during the resting stage
it is present in the nucleus in smaller quantity and largely unattached
to the chromosomes," but the basis for this opinion is not entirely
clear. However this may be, the other school, to which I belong,
is convinced that chromonemata (and genes) reduplicate themselves
at least one and probably two or more cell generations before the
mitosis in which the "split" becomes "effective." It is clear from the
work of Caspersson and Schultz that nucleic acid production is con-
trolled by genes, but to me the converse conclusion, that nucleic acid
is essential to gene reproduction, is not fully established by present
data, and it seems most unlikely that the obvious, wide separation of
chromosome halves that is seen in the earliest prophase of mitosis
is the chronologically immediate result of gene reproduction. Rather,
it is related to the onset of mitosis; and the occurrence of polytene
chromosomes, apart from all other evidence, shows per contra that
special physiological conditions not necessarily correlated with gene
or chromonema reproduction are involved in the initiation of nuclear
division.

Before proceeding to other aspects of Caspersson and Schultz's
work we must recall the differentiation of the chromosomes into
regions which Heitz (1935) termed "euchromatic" and "hetero-
chromatic." Here we must consider data from all four levels. Long

114 The Structure oj Protoplasm

ago, in the days when it was still commonly believed that the
chromosomes lose their identity at the end of mitosis and are formed
again in the prophase of the next division, it was noted that in some
organisms there are chromatic bodies in the "resting" or "energic"
nucleus. The chromosomes were observed to "form" in connection
with these, and they were therefore sometimes called prochromo-
somes. We cannot enter into any detailed discussion of these in
this short talk — for an outline of almost all that is known and of all
that can reasonably be guessed on the basis of present knowledge
reference may be made to recent publications of Geitler (1938) and
White (1940). In brief, the known facts from the second and third
levels are: (1) There are parts of the chromosomes that, in many
species of animals and plants, stain at all stages of the nuclear cycle.
(2) The sex chromosomes are particularly prone to bear such
"heterochromatic" regions. (3) Instead of staining at all stages,
they may in some species be relatively understained at times when
the "euchromatic" parts are deeply stained. (4) Heterochromatic
regions are often indiscriminate in their pairing capacity, and (5)
the nucleolus is very often formed in connection with one or more
of them. From the studies of Caspersson et al. we know (6) that
they are regions that differ from the rest of the chromosomes in
their nucleic acid cycle or total nucleic acid-producing capacity. From
the fourth level there may be added (7) that heterochromatic
regions seem to be devoid of genes that affect morphological char-
acteristics, and (8) that they affect the expression of genes adjacent
to them.

A. characteristic that may be related to heterochromatism is the
tei.dency for certain regions of the chromosomes to be understained
after exposure to low temperatures. This has been reported in a
number of plants, particularly by Shmargon (1938) and other Rus-
sian workers, but it is particularly striking in the genera Paris and
Trillium in which Darlington and La Cour (1940) have studied it.^
These understained "differential regions" of metaphase are, accord-
ing to Darlington and La Cour, the heterochromatic regions which
are deeply stained during the resting stage. They attribute the
attenuation at metaphase after cold treatment to relative "nucleic-
acid starvation" in these species which, they assume, normally have
a particularly high nucleic acid concentration.

Schultz, Caspersson, and Aquilonius (1940) conclude that the
heterochromatic regions of Drosophila melanogaster have the

' See also Wilson & Boothroyd, Cayi. J. Res. 19: 400-412, 1941.

Structural Differentiation of the Nucleus 115

capacity " (1) to form large amounts of thymonucleic acid (or. better
perhaps, thymonucleoprotein) in the chromosomes themselves; (2)
to form or affect the composition of the nucleoli; (3) to affect the
characteristics of neighboring regions translocated to them in such
a way as to change the developmental effects of these regions in
somatic cells; (4) to affect the content of the ribonucleic acids in the
egg cytoplasm of Drosophila." Conclusions (1), (2), and (4) are
based largely on ultraviolet data. Number (3) is a conclusion
drawn jointly from submicroscopic and genetic data; the problem
of the mechanism of the developmental effect, viz., variegation,
remains open and need not be entered upon here. Conclusion (4)
leads us to the problem of the transfer of materials between the
nucleus, nucleolus, and cytoplasm, which will be discussed after we
have considered ways of increasing the nucleic acid content of a
given nucleus by changing its chromosome constitution.

The old problems of the nucleo-plasma ratio and of changes in
nuclear volume began to receive renewed attention following the
observations of Jacobj (1935) on differences in nuclear size within
a tissue. He found that there were regular differences in size that
did not follow the probability curve but one with several maxima.
One maximum tended to be about double that of the preceding one,
i. e., nuclear volumes of 1:2:4:8 were more frequent than inter-
mediate values. In man almost all tissues were foujid to have nuclei
in nine classes, with volumes ranging from 1-256, which would
involve eight doublings. To account for this, Jacobj suggested
rhythmic growth, an alternation of longer rest periods and shorter
growth periods. Wermel also found a rhythm of nuclear growth in
tissue cultures, but showed that the volume doubling was not
reached in a single period of growth; the increase was first to one
and one-half times. Similar results were obtained in observations
on nuclear growth in the intestine of Anopheles and in the fat bodies
of silkworms (see Wermel and Portugalow, 1935). Fischer (1935)
found that growth of the follicular epithelium of some Orthoptera
occurs in two ways: (1) by mitotic increase in cell number, during
which the nuclei remain in the smallest size class, (2) by increase
in cell volume, with constant cell number, which occurs when the
secretory function of the cell starts. He states that the secretory
cells of the follicle do not divide mitotically at this time but that
after the last mitosis there is a doubling of nuclear volume and then
what is apparently an amitotic division. After this there is rhythmic

116 The Structure of Protoplasyn

growth of the nuclei of secreting cells, giving six size classes
1 : 2 : 4 : 8 : 16 : 32, all of which are correlated with distinct phases
of secretory activity.

It is, of course, well known that great variation in nuclear size
can occur, both normally and in experiments, without reproduction
of the constituent chromosomes or other permanent change in struc-
ture. Sometimes increase in nuclear volume is associated with
enlargement of the nucleolus. Metz and his students have shown
particularly clearly by in viva experiments that when hypertonic
Ringer solution is injected into the body cavity of Sciara larvae the
nuclear and chromosome volume of the salivary gland cells both
decrease by at least 25 per cent. With hypotonic solutions there is
an increase of at least 20 per cent. Complete recovery occurs. They
found also that reversible nuclear shrinkage occurs when cells are
under pressure. Upon asphyxiation of Sciara larvae with CO2 or N2
for 1-3 minutes, the nuclear volume of salivary glands remains
unchanged, but the chromosomes shrink and contract into a ball.
When shrinking they give out a clear fluid into the nucleus, and
when recovery occurs this is reabsorbed by the chromosomes.

The regular alterations in nuclear size described above for many
animal tissues occur also in plants (Hoppner, 1939) . They are often
correlated with multiplication of the chromosomes within the nucleus
or of the chromonemata within the chromosomes. Berger (1938)
has reinvestigated the polyploidy long known to occur in mosquitoes.
He found that in Culex pipiens, with n r= 3 chromosomes, the ileum
of the larva grows without mitosis until the nuclei are 8-16 times the
normal volume. When mitoses begin, up to 48 chromosomes are
commonly observed, and some nuclei have 96, which is 32-ploid.
The chromosomes must therefore have divided three or four times
during the long larval development without division of the nucleus
occurring. Geitler (1938) called this phenomenon "endomitosis"
and found it to occur in many tissues of insects such as the Mal-
pighian tubules, testis septa, follicular epithelium, fat bodies, salivary
glands, connective tissues, and epithelium of the mid-gut. A particu-
larly high degree of polyploidy was found in the salivary gland
nuclei of Gerris. The X chromosome of Gerris is heterochromatic,
and the number of chromosomes sets could therefore be determined
by counting the heterochromatic bodies; thus the number of chromo-
somes can be estimated though the nucleus is not dividing mitotic-
ally. As high as 2,048-ploid nuclei were found. It is not known
whether in insects the polyploid tissues which divide mitotically

Structural Differentiation of the Nucleus 117

owe their origin to endomitoses. Such somatic polyploidy is common
(see Oksala, 1939). Smith (1941) found the chromosome number
to be doubled in wing buds of sawflies (Hymenoptera) , i. e., to be
2n in males and 4n in females. In the ovariole wall it was octoploid
and in neural cells 20-30n. In leguminous plants, Wipf and Cooper
(1938) found that dividing cells in the root nodules, which are
involved in the process of nitrogen fixation, are characteristically
tetraploid.

The clearest case of endomitosis in plants occurs in spinach,
2n = 12, where somatic polyploidy has long been known, but its
origin due to endomitosis was shown only recently by Gentcheff
and Gustafsson (1939). They find that in the periblem cells of
roots from germinating seeds, the nuclei have as many as 96
chromosomes. This polyploidy is correlated with the presence of a
large amount of storage products in the periblem cells. The chromo-
somes are split at the beginning of prophase. At metaphase they lie
in pairs and do not separate. After a resting stage they are present
in the double number and unpaired. Levan (1939) obtained similar
results in auxin-treated cortical cells of spinach. In Allium, auxin
produced endomitosis differing only in that the kinetochores were
delayed in their division relative to the arms, and "diplochromo-
somes" were therefore present at metaphase. In these experiments
with auxin, cellular enlargement precedes the increase in nuclear
volume, and endomitosis appears to be initiated thereby. In most
of the other cases cited, chromosome or chromonema multiplication
precedes and apparently initiates the increase in nuclear volume.
This is also the case in polyploidy produced experimentally with col-
chicine. Colchicine apparently inhibits the centrosome or spindle
activity and sometimes the reproduction of the kinetochore, but not
that of the chromonema. Levan obtained cells with as many as 1,000
diplochromosomes, i. e., chromosomes divided except at the kineto-
chore, in Alliinn which normally has 16.

The giant chromosomes of the salivary gland nuclei of Diptera
are now generally, though not universally, conceded to be bundles
of chromonemata. They appear to result from successive endomitoses
which apparently differ from the foregoing cases chiefly in that the
chromonemata remain associated, instead of falling apart and
constituting separate chromosomes, and that homologues become
tightly paired. These differences may both well be concomitants
of the closer association which exists between homologous chromo-
somes in all cells of the Diptera— Metz (1916) showed that in poly-

118 The Structure of Protoplasyn

ploid as well as diploid cells of Drosophila, the homologous chromo-
somes all tend to lie together.

The fact that endomitosis provides a way in which the nucleo-
protein content of a nucleus may be increased, and the apparent
association of endomitotic chromosome or chromonema multiplication
with secretory or nutritional function have naturally led to a good
deal of speculation on the mechanism of the exchange which must
take place between the nucleus and the cytoplasm. Without going
seriously into this problem an indication of the present state of opin-
ion may be given by citing three papers which appeared in close
succession in the same Journal last year. Painter (1940) suggests
that rapidly segmenting eggs can draw the material for the synthesis
of their chromosomes from the chromatin or its derivatives that have
been produced by endomitosis and poured out into the cytoplasm by
the breakdown of the germinal vesicle, or by indirect absorption from
nurse cells. Calvm, Kodani, and Goldschmidt (1940) consider that
salivary gland chromosomes are essentially of the same structure
as mammalian egg "lampbrush" chromosomes from which they con-
sider chromatin is being sloughed off. Caspersson and Schultz
(1940) , however, point out that the high concentration of ribonucleic
acid in the cytoplasm occurs before the nuclear membrane breaks
down. They observe that the cytoplasm or nucleolus which does not
ordinarily stain with Feulgen does contain nucleic acids which must
be assumed to be of the ribose type. There is a high concentration of
these around the nuclear membrane in several rapidly growing
tissues and a gradient from it to the outside of the cell. The amount
of ribonucleic acid in the cytoplasm can be increased by adding
heterochromatin (in particular by adding extra Y chromosomes in
Drosophila) . It seems that more must be learned of the relationships
between different nucleic acids before we can proceed much further
with this problem. The same applies to the relationship between the
chromosomes and the nucleolus. Of the latter, Nebel (1939) says,
"The understanding of this problem has not progressed far during
the last fifty years" and "recent work has further emphasized the
complexity of the problem in that it is necessary to avoid all general-
izations." Caspersson and Schultz commit themselves to little in
this connection except to state that nucleoli are produced in hetero-
chromatic regions, that these regions "are especially concerned with
the ribonucleic acid metabolism of the cell," and that "activity of
nucleoU is closely associated with an intense synthesis of the cyto-
plasmic ribonucleic acids." Since the heterochromatic regions stain

Structural Differentiation of the Nucleus 119

deeply with Feulgen during the "resting" stage, they must, however,
be rich in desoxyrihonucleic acids at that time. The work of McClin-
tock (1934) on the nucleolar-forming body in Maize, and that of
Navashin, Heitz, and others on various plants, besides establishing
morphological and developmental relationships, indicates possible
lines of attack on problems of nucleolar function.

We may proceed now to a discussion of some features of chromo-
some structure for which submicroscopic interpretations are as yet
either very limited in scope or else highly speculative. Very diverse
opinions have been expressed by different workers on the function
or even the existence of the matrix of the chromosomes. Darlington
(1935a) dismisses it entirely but concedes that "the chromosome
thread probably has some sort of pelUcle." Heitz and Kuwada, respec-
tively, introduced the terms kalymma and hyalonema. These and
others are, as Nebel (1939) says, "synonymous terms for a morpho-
logical entity, at present insufficiently defined in terms of chemical
constitution." Some authors, including myself, consider that there
is evidence for both matrix and a sheath or pellicle. Nebel considers
each chromonema to have its own matrix and all the chromonemata
of a chromosome to have, in addition, a common matrix. However
this may be, it may be taken as established that the chromosomes
in mitosis and meiosis consist of chromonemata surrounded by a
substance for which we may as well use the now noncommittal term
matrix until more is known of its nature, structure, and function.

From this point we may logically consider next the problem of
the number of chromonemata and their arrangement within the
chromosomes during division. As Nebel (1939) has so recently
presented an extensive review of chromosome structure, only a brief
outline, particularly of those problems still at issue, will be given
here. The detailed nature of the coiling of the chromonemata within
mitotic or meiotic chromosomes and the mechanisms concerned in
coiling have engaged much attention since about 1925 when Kauf-
mann demonstrated it clearly in pollen mother-cell meiosis of plants
with large chromosomes. More recently spiral structure has been
established definitely in somatic chromosomes of plants, in grass-
hoppers (White, 1940) , sawflies (Smith, 1941) , a mammal (Roller,
1938) , and protozoa (Cleveland, 1938) . Though it has not yet been
clearly demonstrated in many animals nor in plants with very small
chromosomes, it seems a fairly safe assumption that the condensed
chromosomes characteristic of mitosis and meiosis in most organisms
all have a coiled structure and that spiralization is the chief mechan-

120 The Structure of Protoplasm

ism of "condensation" or better, of "packing" (Darlington) . Maxi-
mum spiralization occurs at metaphase or early anaphase in plants
with large chromosomes (and probably in almost all organisms) but
at the end of anaphase in some protozoa (Belar, 1926, Bauer,
1938) . For the purpose of analysis it is necessary to consider the
many different forms that coiling may take or stages that may occur
in the process. The chromonema within each chromatid of a mitotic
chromosome forms a helix which is best termed a standard coil
(Nebel, 1939) . The two chromatids are twisted about each other
like the two strands of electric flex, i. e., they are extended coils
turning in the same direction. This is termed relational coiling
(Darlington, 1935b) . In the prophase of mitosis, especially in the
first microspore division, the chromonemata are in loose spirals
which Darlington termed relic coils. They have now definitely
been shown, in pollen grain divisions, to be the residuum of the coils
of the previous division, as his term implies (Sparrow, Huskins, and
Wilson, 1941) .

At the metaphase and anaphase of meiosis, there is typically a
coil of much larger diameter. Fujii, in 1926, first described the
chromonema forming this coil as being itself a small-gyred spiral.
These two helices are now generally termed the major and minor
coils (Huskins and Smith, 1935) . In some organisms, e. g., Trillium,
the major coil persists relatively unchanged through both divisions
of meiosis and appears as a large- gyred relic coil in the prophase of
the first pollen grain division. In others, e. g., Tradescantia, the major
coil disappears during meiotic interkinesis and a new smaller-gyred
spiral appears in the second division.

In my opinion much confusion has resulted from the use of terms
with implications beyond those warranted by the existing data and
from failure to keep theory sharply differentiated from observation.
For instance, the major and minor coil were originally termed pri-
mary and secondary, and the terms were then interchanged as
evidence accumulated on their relationship. The small-gyred coil
of the second meiotic division in Tradescantia and that of somatic
cells have often been called minor coils, thus implying a relationship
with the minor coil of the first division of meiosis which is not yet
established. The frankly deductive hypothesis of coiling presented
by Darlington (1935b) relates all the different coils, and traces
them back to an assumed molecular spiral unitarily controlled
from the centromere or kinetochore. Certain of Darlington's fol-
lowers took the molecular spiral as a datum and when they found

Structural Differentiation of the Nucleus 121

changes in direction of the major coil or in the sahvary gland
chromosomes, explained them away, or in some cases ignored them,
because they could not, of course, be accommodated to a theory
having this basis. Similarly, many other workers presented figures
of drawn-out major spirals as evidence for the minor spiral concept
of Fujii, Kuwada, et al. Further confusion resulted from disagree-
ment on the number of chromonemata within meiotic chromosomes.
Kuwada and Nakamura (1933) , in one of the first presentations in
English of the work of the current Japanese school, showed the coil-
within-a-coil structure by means of a diagram containing four
strands. This diagram was adopted by the Darlington school since it
agrees with their concept of the number of strands. Kuwada and
Nakamura had, however, already pointed out in publications in
Japanese that there is a further split in each of the chromatids in
the second division, and they stressed this point in a publication in
English in the following year (1934) . They and others have since
brought forward much evidence for the existence of at least eight
strands in a meiotic bivalent. The Darlington school, however, still
sees only four. It is obvious that optical images can be given very
different interpretations as coils when such a disagreement exists
on the number of strands. It has also been pointed out (Huskins,
1937) that the minor coil of some workers is a tightly wound helix,
while that of others is a mere waviness. The former would most
easily be explicable as the result of its molecular structure; the
latter may be interpreted as the initiation of a spiral determined
by mechanical forces. There are also differences of opinion on the
relationship between the two major coils formed by the two chroma-
tids of a meiotic chromosome. Some authors have pictured them as
intertwined; to get them separated Matsuura has postulated breaking
and rejoining (genetic crossing-over) during anaphase — for which
there is no evidence. Most observers, however, find the chromatid
spirals twisted on their own axis but free of each other.

In the hope of resolving these differences of opinion on the
interpretation of microscopic images, efforts have been made, but
with very meager success, to obtain evidence from other levels.
Polarized light experiments were for some time held to prove the
existence of a minor spiral at right angles to the major spiral, but it
is now generally conceded that the existing data on birefringence
of chromosomes are equivocal (cf. Frey-Wyssling, 1938) . Innumer-
able experiments have been made to determine the number of
chromonemata through the breakage effects of X-rays. The results

122 The Structure of Protoplasm

have mostly been interpreted in accordance with the opinion the
investigator had previously formed from direct microscopic obser-
vation. The opinion is now gaining ground that X-rays are too crude
a tool for settling this problem. Fourth level data show that there
are four genetically effective strands in a meiotic bivalent so far as
crossing over is concerned, but this says nothing of the structure
within each of these.

In our McGill laboratory we have, during the last five years,
attempted to get more definite evidence on some of these problems
by starting with statistical analyses of extensive data from the largest
and most easily observed spiral structures. For this purpose Trillium
is exceptionally favorable material.

Our first significant observation was that the major coil changes
direction at numerous points in Trillium. By analysis of normal
material and comparison with pollen mother-cells that had different
degrees of lack of association between homologous chromosomes
and sister chromatids through temperature treatments, we have
been able to establish (Huskins and Wilson, 1938) that reversals
occur with random frequency at the kinetochore and at chiasmata,
and that elsewhere they may occur from unknown causes with a
frequency proportional to the length of chromosome involved. It
was next discovered (Wilson and Huskins, 1939) that the chromo-
nema more than doubles in length during formation of the major
coil — the nature of the elongation is not yet determined. These facts
seem decisively to rule out any hypothesis which involves an internal
torsion due to molecular pattern and directed as a unit. They are
fitted for the present by the simple assumption that a thread of a
resilient consistency will take up a spiral form if it expands in length
within a confined space— the sheath or pellicle. Almost surely this
concept will be found inadequate to explain some of the other spirals
now being studied.

The relational coil has generally been assumed to result from
torsion causing the chromonemata to twist around each other. Our
current data (Sparrow, Huskins, and Wilson, 1941) show that the
microspore chromatid relational coiling results directly from the
partial straightening out of the gyres of the major coil of meiosis
and is related to the plane assumed by the tertiary split in the major
coil. This cannot, of course, be the direct cause of the relational
coiling of homologous chromosomes in the prophase of meiosis —
which requires much more rigid analysis than it has yet received.
We are proceeding at present with analyses of standard somatic

Structural Differentiation of the Nucleus 123

coils. The torsion hypotheses of coiling and crossing over have also
been checked by analyses of the relationship of successive chiasmata
(Huskins and Newcombe, 1941). These reveal a more complex
relationship than previously envisaged, and experiments with Neuro-
spora (Lindegren and Lindegren, 1937 and unpublished) seem to
substantiate the results with fourth level, genetic data.

To sum up: The structure of the nucleus is highly involved and
changes in it are being correlated with function. Studies of the
nucleus are proceeding on four different levels which differ con-
siderably in the degree of accuracy or objectivity attainable. First-
level analyses of chromosomes have led to hypotheses on their mole-
cular pattern. Ultraviolet and Feulgen studies of the nucleus have
established the nucleic acid cycle during mitosis and interkinesis.
The conclusions from these regarding the time of reproduction of
the gene rest, however, partly on second-level or microscopic obser-
vations of the time of chromosome or chromonema reproduction,
regarding which there are sharp differences of opinion. They are,
therefore, much less definitely established than the first-level data
on which they may appear to be based. Microchemical studies on
genetically controlled materials involving heterochromatic regions
(combined first-, fourth-, and third-level studies) indicate: (a) that
heterochromatic regions are specialized for the production of nucleic
acid, (b) that they are concerned in the formation of the nucleolus,
and (c) that the position of genes relative to these regions affects
the expression of the characteristics they determine. The functions
of nucleoli of different types remain undetermined.

Second-level microscopic studies are now very frequently com-
bined with fourth-level, genetic analyses. Where the correlation is
experimental, very solid progress has been made. One school of
second-level cytology uses genetic data extensively for the deductive
formulation of cytological hypotheses (Darlington, 1935a and b) .
Very great advances have been made by this technique, but the
dangers inherent in the method are apparently not realized by all
those following it, nor is the admittedly tentative nature of some of
the conclusions drawn by this school realized by many workers on
other levels who attempt correlations of them with their own data.

Third-level studies by the methods of experimental cytology-
have not been considered to any extent in the above brief review,
since they are considered by another contributor to this symposium.
Only limited aspects of the problems of mitosis have been consid-
ered since this process obviously depends upon reactions of the cell

124 The Structure of Protoplasm

as a whole and only nuclear problems are here considered in any
detail. The coiled structure of meiotic and mitotic chromosomes is
being studied comprehensively by the deductive school and piece-
meal by others, including our own group, who are attempting to
remain more inductive in their approach. Coiling is being related
by both schools to mechanical function during nuclear division.
The structure of the resting or "energic" (Berrill and Huskins,
1936) nucleus is fairly evidently correlated with physiological func-
tion, and cytogenetic methods of attacking this problem are available
(cf. Stern, 1938) .

Progress is apparently being made most rapidly in nearly all of
the problems of the nucleus when attacks are made by associations
of workers with training in the different techniques of the four
"levels" into which, for the purpose of stimulating discussion,
analyses of nuclear structure are here somewhat arbitrarily grouped.

BIBLIOGRAPHY

Bauer, H. 1938. Chromosomenstruktur. Arch. Exp. Zellf. 22: 181-187.

Belar, K. 1926. Der Formwechsel der Protistenkerne. Erg. Fortschr. Zool. 6:

235-652.
Berger, C. a. 1938. Multiplication and reduction of somatic chromosome groups

as a regular developmental process in the mosquito, Culex pipiens.

Carnegie Inst. Wash. Publ. 496: 209-232.
Berrill, N. J., and Huskins, C. L. 1936. The "resting" nucleus. Am. Nat. 70:

257-261.
Bracket, J. 1940. La localisation de I'acide thymonucleique pendant I'oogenese

et la maturation chez les amphibiens. Arch. Biol. 51: 151-165.
Calvin, M., Kodani, M., and Goldschiwidt, R. 1940. Effects of certain chemical

treatments on the morphology of salivary gland chromosomes and their

interpretation. Proc. Nat. Acad. Sci. 26: 340-349.
Caspersson, T. 1936. Uber den chemischen Aufbau der Strukturen des Zell-

kernes. Skand. Arch. Phys. 73: Suppl. 1-151.
. 1939a. Uber die Rolle der Desoxyribosenukleinsaure bei der Zell-

teilung. Chromosoma 1: 147-156.
. 1939b. Studies on the nucleic acid metabolism during the cell cycle.

Arch. Exp. Zellf. 22: 655-657.
. 1940. Methods for the determination of the absorption spectra of

cell structure. Jour. Roy. Micr. Soc, Ser. 3, 60: 8-25.

and Schultz, J. 1940. Ribonucleic acids in both nucleus and cyto-

plasm, and the function of the nucleolus. Proc. Nat. Acad. Sci. 26:
507-515.

Cleveland, L. R. 1938. Longitudinal and transverse division in two closely
related flagellates. Bio. Bui. 74: 1-40.

Darlington, C. D. 1935a. The old terminology and the new analysis of chromo-
some behaviour. Ann. Bot. 49: 579-586.

. 1935b. The internal mechanics of chromosomes. Proc. Roy. Soc. B.

118: 33-96.

. 1937. Recent Advances in Cytology. 2nd ed., pp. 1-671.

Structural Differentiation of the Nucleus 125

AND La Cour, L. 1940. Nucleic acid starvation of chromosomes in

Tnllium. Jour. Genet. 40: 185-213.
Fischer, I. 1935. Uber den Wachstumsrhythmus des Follikelepithels der Lause

und Federlinge und seine Beziehungen zum Arbeitsrhythmus der Zelle

und zur Amitose. Zschr. Zellf. 23: 219-243.
Frey-Wyssling, a. 1938. Submikroskopische Morphologie des Protoplasmas und

seiner Derivate. Protoplasmamonographien 15: 1-317.
Geitler, L. 1938. Chromosomenbau. Protoplasmamonographien 14: 1-190.
Gentscheff, G., and Gustafsson, a. 1939. The double chromosome reproduction

in Spinacia and its causes. I-II. Hereditas 25: 349-386.
Heitz, E. 1935. Chromosomenstruktur und Gene. Zschr. Ind. Abst. Vererb. 70:

402-447.
Hillary, B. B. 1940. Uses of the Feulgen reaction in cytology. II. New techniques

and special applications. Bot. Gaz. 102: 225-235.
HusKiNS, C. L. 1937. The internal structure of chromosomes — a statement of

opinion. Cytologia, F. J. 2: 1015-1022.
and Newcombe, H. 1941. An analysis of chiasma pairs showing chro-
matid interference in Trillium erectum L. Genetics 26: 101-127.
AND Smith, S. G. 1935. Meiotic chromosome structure in Trillium

erectum L. Ann. Bot. 49: 119-150.

AND Wilson, G. B. 1938. Probable causes of the changes in direction

of the major spiral in Trillium erectum L. Ann. Bot. N. S. 2: 281-292.
Jacobj, W. 1925. Das rhythmische Wachstum der Zellen durch Verdoppelung

ihres Volumens. Arch. Entw. mechan. 106 : 124-192.
Roller, P. C. 1938. The genetical and mechanical properties of the sex chromo-
somes. IV. The Golden Hamster. Jour. Genet. 36: 177-195.
KoLTZOFF, N. K. 1938. The structure of the chromosomes and their participation

in cell metabolism. Biol. Zhumal 7: 2-46.
KuwADA, Y., AND Nakamura, T. 1933. Behaviour of Chromonemata in mitosis.

I. Observation of pollen mother cells in Tradescantia reflexa. Mem. Coll.

Sci., Kyoto Imp. Univ., Ser. B. 9: 129-139.
. 1934. Behaviour of chromonemata in mitosis. IV. Double refrac-
tion of chromosomes in Tradescantia reflexa. Cytologia 6: 78-86.
Levan, a. 1939. Cytological phenomena connected with the root swelling caused

by growth substances. Hereditas 25: 87-96.
Lindegren, C. C, AND LiNDEGREN, G. 1937. Non random ci'ossing over in Neuro-

spora. Jour. Hered. 28: 105-112.
Mazia, D., AND Jaeger, L. 1938. Nuclease action, protease actions, and histo-

chemical tests of salivary chromosomes of Drosophila. Proc. Seventh

Int. Genet. Con., p. 212.
McClintock, B. 1934. The relation of a particular chromosomal element to the

development of the nucleoli in Zea mays. Zschr. Zellf. 21: 294^328.
Metz, C. W. 1916. Chromosome studies in the Diptera. II. Jour. Exp. Zool. 21:

213-279.
MiLOViDOV, P. 1938. Bibliographie der Nucleal-und Plasmalreaktion. Protoplasma

31: 246-266.
Nebel, B. R. 1939. Chromosome structure. Bot. Rev. 5: 563-626.
Oksala, T. 1939. Uber Tetraploidie der Binde- und Fettgewebe bei den Odona-

ten. Ein Beitrag zur Kenntnis der sog. somatischen Polyploide der

Insekten. Hereditas 25: 132-144.

Painter, Th. S. 1940. On the synthesis of cleavage chromosomes. Proc. Nat.
Acad. Sci. 26: 95-100.

Schmitt, F. O. 1940. The molecular organization of protoplasmic constituents.
Coll. Net. 15: 1-7.

126 The Structure of Protoplasm

ScHULTZ, J., Caspersson. T., and Aquilonius, L. 1940. The genetic control of
nucleolar composition. Proc. Nat. Acad. Sci. 26: 515-523.

Shmargon E. N. 1938. Analysis of the chromomere structure of mitotic chromo-
somes in Rye. C. R. Acad. Sci. URSS. 21: 259-262.

Smith, S. G. 1941. A new form of spruce sawfly identified by means of its
cytology and parthenogenesis. Sci. Agr. 21: 245-305.

Sparrow, A. H., Huskins, C. L., and Wilson, G. B. 1941. Studies on the chromo-
some spiralization cycle in Trillium. Can. Jour. Res. 19:323-350.

Stern, C. 1938. During which stage in the nuclear cycle do the genes produce
their effects in the cytoplasm? Am. Nat. 72: 350-357.

Wermel, E. M., and Portugalow, W. W. 1935. Studien uber Zellengrosse und
Zellewachstum. XII. Mitteilung. Uber den Nachweis des rhythmischen
Zellenwachstums. Zschr. Zellf. 22: 185-194.

White, M. J. D. 1940. The heteropycnosis of sex chromosomes and its interpre-
tation in terms of spiral structure. Jour. Genet. 40: 67-82.

Wilson, G. B., and Huskins, C. L. 1939. Chromosome and chromonema length
during meiotic coiling in Trillium erectum L. Ann. Bot., N. S. 3: 257-270.

Wipe, L., and Cooper, D. C. 1938. Chromosome numbers in nodules and roots of
red clover, common vetch and garden pea. Proc. Nat. Acad. Sci. 24: 87-91.

PROTOPLASMIC STREAMING IN RELATION TO GEL
STRUCTURE IN THE CYTOPLASM

Douglas A. Marsland
Washington Square College of Arts and Science, New York University

I. INTRODUCTION

Reversible sol-gel transformations are commonly recognized in
protoplasmic systems, and for some time it has been thought that
these reactions may play an important role in physiological activity.
Thus Mast ('26 and '31), and Lewis ('39) have considered that the
movement of amoeboid cells depends upon a series of gelations
occurring at the anterior ends of the pseudopodia, and upon com-
pensating processes of solation in the posterior part of the cell. Aside
from this case, however, examples establishing the physiological
importance of sol-gel reactions have not been demonstrated very
clearly.

A. SOL-GEL EQUILIBRIA IN RELATION TO HYDROSTATIC PRESSURE

A relationship between hydrostatic pressure and the structural
characteristics of protoplasmic gels was first revealed by Dugald
E. S. Brown in 1934. Brown ('34c) found that the central mass of
the protoplasm of the Arbacia egg is relatively fluid compared to a
cortical layer, about 5 microns thick, which displays the properties
of a fairly rigid gel. When these eggs were centrifuged at atmos-
pheric pressure in a weak centrifugal field, the granular components
of the central fluid protoplasm were readily displaced, but the gran-
ules of the gelated cortex, mainly the pigment granules, remained
quite fixed. But when the centrifuging was done at increasingly
higher hydrostatic pressure, up to 10,000 pounds per square inch,
the cortical gel displayed a greater and greater degree of lique-
faction. In the higher range of pressure the pigment granules were
displaced with great rapidity, and at 10,000 lbs. /in.,- the cortical
gel offered less than 10 per cent of its atmospheric resistance.

Subsequent work has demonstrated that hydrostatic pressure
imposes solation upon the protoplasmic gels of animal and plant cells
generally. Among the animals, the effect has been shown in a num-
ber of different eggs;^ in two kinds of amoeba,- in the tentacles of

[127]

128 The Structure of Protoplasm

a suctorian^ (Ephelota) , in Paramecium,^ and in human erythro-
cytes.^ Among the plants a similar effect has appeared in the leaf
cells of Elodea,'' in the plasmodium of Physarum," and in the cells
of Spirogyra.^

A number of the cases have yielded quantitative measurements,
and these indicate that the relative magnitude of the liquefaction
induced by pressure is the same in all cases, regardless of whether
the initial gel at atmospheric pressure is quite firm or is relatively
fluid. Each increment of 1,000 Ibs./in.- reduces the rigidity by almost
25 per cent. Furthermore, within fairly broad limits, the effect is
reversible. Pressures up to 4,000 Ibs./in." may be maintained for
about an hour, and yet, when the cells are returned to atmospheric
pressure, the original structural characteristics of the gel are
i-estored within a minute, or perhaps within an even shorter time.
For higher pressures irreversible changes begin to appear much
sooner, depending upon the intensity.

B. STATEMENT OF THE PROBLEM

Since it is known that pressure induces solation, or conversely,
that pressure prevents gelation from occurring in protoplasmic
systems, a useful tool has been provided for analyzing the role of
such phenomena in various forms of physiological activity. At the
present time quite a number of studies are available, and the pur-
pose of this paper is to determine, so far as is possible, how the
manifold physiological effects of pressure may be related to changes
in the sol ^ gel equilibrium.

C. STREAMING ACTIVITIES IN CONTRAST TO OTHER PHYSIOLOGICAL

PROCESSES

Perhaps a general statement of the results, made in advance of
the detailed account, will provide a useful orientation. In general,
it seems valid to say that one group of physiological activities, the
group in which protoplasmic streaming is the common attribute,
appears to be particularly vulnerable to inhibition by very moderate
(below 5,000 Ibs./in.-) intensities of compression. This pressure-
susceptible group includes amoeboid movement, cyclosis, cell division
(in animal cells) , and the migration of pigment in the unicellular
type of chromatophores.

A second group of activities, which includes contraction in muscle,
conduction in nerve, and the motility of cilia and flagella, is not
inhibited by moderate pressures. In fact, the characteristic activities

Protoplasmic Streatiiing — Relation to Gel Structure 129

of the second group are considerably augmented or accelerated in
the lower range of pressure. Furthermore, it may be said that the
effect of pressure upon the activities of the first group is in propor-
tion to the effect upon the gel system, whereas this does not seem
to be the case for the second group.

II. METHOD

A. DIFFERENCES BETWEEN HYDROSTATIC AND OTHER TYPES OF

COMPRESSION

It is important to realize that the pressure used in the experi-
ments is of the hydrostatic type. Each cell or tissue is completely
surrounded by a liquid medium, and this medium serves to trans-
mit the pressure from the pump equally in all directions. This
condition eliminates all distortional injury, such as would result
if the tissues were compressed locally between impinging solid
surfaces. The difference is well illustrated by experiments on divid-
ing Arbacia eggs. When these cells are compressed between a slide
and coverslip, less than 5 Ibs./in.- suffices to cause considerable
distortion and to block the cleavage. Under hydrostatic compression,
however, the form of the egg remains unchanged, aside from a very
slight loss of volume, at pressures well above 10,000 Ibs./in.-, and the
capacity to furrow is not completely abolished until a pressure of
more than 5,000 Ibs./in.- is reached.

Another important experimental condition is the absence of any
gas phase in the system. Many of the early experiments in the field
were complicated by the fact that the pressure was applied through
the medium of a supernatant atmosphere. This made it difficult to
distinguish between the effects of the pressure per se, and the effects
of driving excessive quantities of the atmospheric gases into solution
in the protoplasm and the surrounding liquid medium. When the
gaseous phase is eliminated, the main effect of the pressure must be
mediated through small changes in the protoplasmic volume" and
through the consequent alterations in the fundamental molecular
pattern of the protoplasmic system.

B. RECENT TECHNICAL DEVELOPMENTS

All of the experimental methods cannot be given in detail, but a
brief consideration of two recently developed pieces of apparatus
will permit a clearer understanding of the ensuing work. The first
is the centrifuge-pressure bomb devised by Brown ('34c). This

130 The Structure of Protoplasm

apparatus makes it possible to centrifuge a tissue while the com-
pression is maintained at any desired level up to 14,000 Ibs./in.-
Thus one may measure, by the centrifuge method, the fluidity of the
protoplasmic system as a function of pressure. The bomb is divided
into two parts; (1) the experimental, or pressure chamber, and (2)
the control chamber. A needle valve seals the pressure into the
experimental chamber during the period of centrifugation, and since
the centrifugal radius is the same for both chambers, the control
(atmospheric), and the high pressure specimens are subjected to an
equal centrifugal force.

The second apparatus is the microscope-pressure chamber
described by Marsland and Brown in 1936. In this chamber, speci-
mens may be viewed during the period of compression at magnifica-
tions up to 600 diameters."" Upper and lower windows, 3 mm.
thick, permit hght to be transmitted through the chamber to a
special objective'^ which, despite an unusually great working distance
of 15 mm.,1" possesses a magnification of 15 diameters. Since the
specimens in the chamber tend to drop to the upper surface of the
lower window, the bomb is used with an mverted microscope, and
good images are obtained with oculars up to 20 X-

III. CELLULAR ACTIVITIES WHICH
INVOLVE PROTOPLASMIC STREAMING

A. AMOEBOID MOVEMENT

The dependence of this activity upon the structural character-
istics of the gel system of the amoeboid cell has been demonstrated
by several types of experiment (Brown and Marsland, '36, and
Marsland and Brown, '36) .
(1) Effects of high pressure, rapidly established

In this type of experiment, the amoebae were placed in the
pressure chamber and then, while under continuous observation,
were rapidly compressed at the rate of about 1,500 Ibs./in.- per

second.

The first effect of the compression is the cessation of the proto-
plasmic flow when the pressure reaches about 4,000 lbs. Then no
further change is noted until the pressure reaches approximately
6,500 lbs. At this point, within 0.5 second each elongate pseudo-
podium undergoes a sudden shrinkage in length and develops a
terminal sphere of the type shown in Figure 1. This abrupt and
rapid reorganization of the pseudopodium is followed by a more

B

4-^

#

■f^

^^H^

m 0

m

*^^.

m

#

%

*

-ti •

B , '

♦

Fig. 1. Solation of the plasmagel of the pseudopodia of Amoeba
duhia. In A, the specimens are at low pressure (1,500 Ibs./in.-). Note
that characteristically the pseudopodia are elongate cylinders.
B shows the same specimens 1 second after raising the pressure to
6,500 lbs. in.- Due to the greater fluidity of the plasmagel, each
pseudopodium undergoes a rapid reorganization, tending to become
spherical as a result of the tensional forces of the surface.

132

The Structure oj Protoplasm

100 5J

90

80

70

60

0 = Arbitrary point; all other values
are relative to this.

T =: Gel value, Amoeba.

J. = Gel value, unfert. Arbacia egg.

X — Gel value, cleaving. Arbacia egg.

+ — Rate of cleavage, Arbacia egg.

• — Gel value, Elodea cells.

0 — Rate of streaming, Elodea cells.

M — Gel value, myosin gel (rabbit)
pH = 65: temp. 23°-24°C.

50

:\

\

40

-

i\

\

30

-

\

^8\

\

M

20

-

0

\

10

I . .J —

1 —

1

10

PRESSURE -LBS./ 1 N^x 10^

Fig. 2. Proportionality between the effects of pressure on protoplasmic
streaming (and cell division) and the degree of solation imposed by pressure,
in various gel systems. The data for the myosin gel are entirely tentative,
since only points nearest the general curve were selected from several pre-
liminary experiments.

gradual change in the entire form of the amoeba such that in about
5 minutes all of the specimens become quite spherical.

This type of experiment clearly indicates that the tubular form
of a pseudopodium is maintained by rigid properties which are
inherent in the plasmagel layer. The data of Figure 2 reveal that
the rigidity of this layer at 6,500 lbs. is less than 20 per cent of its
atmospheric value. When the rigid properties of the gel are lost, the
pseudopodium behaves as a cylinder^ ^ of nonmiscible liquid sus-
pended in water and becomes unstable under the tensional forces of
the surface.

Protoplasmic Streaviing — Relation to Gel Structure 133

The foregoing reaction is exhibited most strikingly in the case of
Ainoeha duhia. With Amoeba jjroteus the formation of the terminal
spheres is not so well marked and requires slightly higher pressures.
In this case and in the case of the amoebocytes of the body fluid
of Arbacia the sudden collapse of many of the pseudopodia is repre-
sented chiefly by an abrupt shrinkage in length. In all cases, how-
ever, the cells gradually become spherical if the high pressure is
maintained.

A similar phenomenon has been observed by Kitching and Pease,
'39, who studied the behavior of the tentacles of the suctorian
protozoa. Ephelota coronata. In this case the disintegration, which
occurred mainly at pressures between 6,500 and 8,500 lbs. /in., -
was so complete that the suctorial processes broke up into tiny
isolated beads of protoplasm. This difference is probably due to the
greater length of the tentacular processes and to the absence of a
visibly differentiated surface membrane such as is present in the
Amoeba.

(2) Indications of a gradient in the tensile properties of the
plasmogel

It is significant to note in the Amoeba experiments that the
rigidity of the plasmagel varies somewhat in different parts of the
pseudopodium. Thus in a newly formed pseudopodium, which is
actively extending at the moment of compression, the terminal
sphere may involve all but a short remnant at the base, whereas in
older less active ones, only the distal tip may be involved. Assuming
that pressure produces the same percentage loss of rigidity in all
parts of the plasmagel system,^- these observations indicate that
normally the plasmagel is more fluid in the more recently formed
parts of the pseudopodium. They also support the view of Mast
('26) , namely, that a gradient of rigidity exists in the plasmagel
tube, decreasing as one proceeds from the base to the tip of the
pseudopodium.

(3) Measurement of the solating effect of the pressure

The experimental results (Brown and Marsland, '36) , which
clearly describe the solation of the plasmagel as a function of
pressure, are given in Figure 2. The gelation percentages^^ are
taken to be in proportion to the time of centrifugation" required for
the formation of a standard hyaline zone (see Fig. 3) , measured
at each of the indicated levels of pressure.

The formation of the hyaline zone (Fig. 3) requires the displace-
ment of all the formed bodies originally present in this region of the

134

The Structure of Protoplasm

protoplasm. The greatest resistance to such displacement occurs in
the peripheral protoplasm, that is to say, in the plasmagel rather
than in the plasmasol. At 7,000 X gravity the displacement of the
formed elements of the sol region is practically instantaneous (Heil-

brunn, '29) . Consequently,
it may be said that the so-
lation measurements ob-
tained deal entirely with
the gel portion of the
protoplasm of this cell.
(4) Equilibration to lower
pressures, steadily main-
tained

In view of the marked
effects of the sudden high
compression of the Amoe-
bae, it seemed important
to ascertain the character-
istic form and movement
of the specimens when
equilibrated to lower pres-
sures which do not induce
such a drastic liquefaction
of the plasmagel. In the
experiments the amoebae
were exposed to the de-
sired pressure until the
characteristic form was
assumed, a matter usually
of about 5 minutes. In
most instances the same group of animals was compressed at suc-
cessively higher pressures and then decompressed, step-wise, through
the same pressure range, sufficient time for equilibration being
allowed at each pressure.

The variation in form of the Amoebae as a function of pressure
involves a progressive diminution in the diameter of the pseudo-
podia which are formed. At lower pressures, i. e., up to 2,000 Ibs./in.,-
there is also a lengthening^'' of the pseudopodia, but thereafter both
the length and the diameter are reduced more and more. At 5,000
lbs. the pseudopodia, although numerous, appear as mere pin points
projecting from the surface of the otherwise spherical cell. Above

Fig. 3. The solation effect of pressure in
the Amoeba. Specimens E (experimental) and
C (control) were centrifuged simultaneously
for only 10 seconds. E was in the pressure
section of the centrifuge at 8,000 lbs./in.%
whereas C was at atmospheric pressure. The
much greater fluidity of the plasmagel of E
is indicated by the sharp centrifugal zoning.
An equivalent zoning of atmospheric speci-
mens requires more than 3 minutes of cen-
trifuging. Note also in E, that the contractile
vacuole was on the point of being thrown
out from the centripetal end of the cell at
the instant when the nucleus reached the
opposite end: o, the oil zone; h, the hyaline
zone; g, the granular zone.

Protoplasmic Streaming — Relation to Gel Structure 135

6,000 lbs. no pseudopodia can be formed or maintained, and the
Amoebae become completely spherical. With few exceptions the
equilibrium form is assumed by all of the specimens of an experi-
mental group, whether the particular pressure is approached from
a higher or a lower level. Apparently each characteristic form
represents a steady state with respect to the conditions imposed by
the pressure.

The extent to which the variations in the diameter of the pseudo-
podia may be accounted for on the basis of the increased fluidity of
the plasmagel constitutes an interesting question. Essentially, a
pseudopodium must be a tube of plasmagel through which the plas-
masol flows outward as extension is occurring, and the tubular form
of a pseudopodium must be maintained by rigid properties residual
in the plasmagel wall. Due to the solating effect of an increase of
pressure, a gel layer of equal thickness would possess a lesser
strength. On the basis that hydrodynamic factors are involved in
the flow of plasmasol, a reduction in the diameter of the pseudo-
podium can be regarded as a compensation for the diminished
strength of the wall. The situation would be analogous to a tube of
smaller diameter successfully conducting a fluid which is flowing
under a pressure that would be sufficient to rupture the wall if the
diameter of the lumen were greater.

Some further observations appear to justify the foregoing view-
point. At pressures below 3,000 lbs., a range in which an appreciable
protoplasmic flow continues, a few of the pseudopodia do not extend
steadily in one direction. Instead, the progress of the extending
pseudopodium is interrupted at intervals of about 15 seconds by
sudden ruptures in the lateral wall, just proximal to the advancing
tip. When a rupture occurs, the protoplasmic granules pour rapidly
out into the resulting lateral bulge, and then the flow stops altogether
for a moment. When the advance begins again, the direction is
deflected slightly toward the ruptured side. Such pseudopodia
assume a somewhat tortuous form. It would seem probable that this
phenomenon indicates an incomplete compensation in the particular
pseudopodia which are involved, and that the diameter of these
pseudopodia is reduced to a degree which is not quite sufficient to
provide stability during periods of active flow.

B. CYCLOSIS

A study of the effects of pressure upon protoplasmic streaming
in the leaf cells of Elodea (Marsland, '39b) , has also established a

136 The Structure of Protoplasm

dependence of this activity upon sol ^ gel reactions. In this case
it has been found that the rate of streaming diminishes^*^ in propor-
tion to the degree of solation which increasing pressure induces in
the nonflowing parts of the protoplasm (Fig. 2) .
(1) The velocity of streaming in relation to pressure

Qualitative observations on the rate of streaming as a function
pressure are relatively easy to make. At 2,400 Ibs./in.,^ the velocity
appears to be about half the atmospheric rate. At 5,500 lbs., the
chloroplasts continue to move, but the progress is almost imper-
ceptible. Between 6,500 and 7,500 lbs., depending upon the particular
cell which is being watched, streaming ceases altogether. However,
the flow begins again within about 1 minute after decompression,
provided the exposure in the higher range was not too prolonged.
In view of such a prompt reversal of the pressure effect, it is not
difficult to understand why Fontaine ('29) reported that pressures
up to 10,000 Ibs./in.- had no immediate effect upon the velocity of
streaming, since this worker was unable to see the cells until a few
minutes after decompression had occurred.

Quantitative measurements of the rate of streaming are some-
what difficult to obtain. It is necessary to select a cell which can be
brought into sharp focus in the pressure chamber, and with a stop
watch to time at least ten individual chloroplasts as they pass
through a complete circuit, or at least through a definitely fixed
major part of the circuit. Although considerable variation is found
from cell to cell, the same cell frequently maintains its individual
rate, calculated from the average of ten successive timings, for a
period of more than half an hour, provided that the hght and tem-
perature conditions are kept constant. In such cells it is possible
to measure the pressure rate relative to the initial atmospheric rate.
In each case the measurements were discarded if, after decompres-
sion, the particular cell failed to return, within 5 per cent, to its
original atmospheric rate of streaming.
(2) Measurement of the solation effect

The very marked solating effect of pressure upon the protoplasm
of Elodea cells is demonstrated in Figure 4. Leaf A (at 6,000
Ibs./in.-) and Leaf B (at atmospheric pressure) were centrifuged in
the same field for only 45 seconds. The greater resistance to the
displacement of the chloroplasts in the gelated protoplasm of the
atmospheric specimen is indicated by the absence of any clear sedi-
mentation zones. Such clear sedimentation zones may be obtained

Protoplasmic Streaming — Relation to Gel Structure 137

Fig. 4. The solation effect of pressure in the cells of Elodea canadensis. Lieaves
A and B were centrifuged simultaneously for 45 seconds, A at 6,000 Ibs./in." and
B at atmospheric pressure. The greater fluidity of A is indicated by the marked
sedimentation of the chloroplasts. Leaf C was centrifuged for 4 minutes at
atmospheric pressure.

in atmospheric specimens, but only by prolonged centrifugation
(about 5 minutes; see Fig. 4-C) .

When a leaf is placed in the control, and in the experimental
section of the centrifuge-pressure head, care must be taken that the
axis of the leaf is fixed in the proper position, i. e., parallel to the
radius of the head. Thus the displacement of the chloroplasts and
other heavier granules is toward one of the ends of the elongate cells,

138 The Structure of Protoplasm

rather than toward one of the sides. In each of the experiments the
centrifuging was started 1 minute after the pressure was estabUshed
in the experimental section of the bomb, and a uniform force, 810 X
gravity, was used. In view of the fact that protoplasmic streaming,
and consequently a redistribution of the sedimented chloroplasts,
would begin very shortly after the centrifuging was stopped, it was
necessary to fix the material in order to obtain an accurate measure-
ment of the degree of sedimentation. Heat fixation, obtained by
placing the bomb in boiling water for 45 seconds, gave excellent
results. This method, which avoids chemical contamination of the
chamber, was used in all the experiments.
(3) The plasmagel system in plant cells

Although it has not been usual to speak of a "plasmasol" and
"plasmagel" with reference to the protoplasm of plant cells, these
terms do not seem inapt, at least in plant cells which display proto-
plasmic streaming. A close observation of a cell of Elodea during
active cyclosis shows that, except under special circumstances,^'
only a relatively small portion of the protoplasm is involved in the
actual stream. Usually the flowing part is restricted to a fairly nar-
row channel which follows the side and end walls completely around
the cell. The protoplasm which lies above and below this channel,
i. e., along the upper and lower margins of the walls, and over the
"roof" and "floor" of the cell, does not flow. In these parts of the
protoplasm, the chloroplasts, even those which border directly upon
the flowing part, may maintain a fixed position for many minutes.
Thus the nonflowing portion of the protoplasm resembles the plas-
magel of an amoeboid cell, whereas the actively streaming part may
be regarded as plasmasol.

The gelation percentages, which in Figure 2 have been plotted
as a function of pressure, undoubtedly deal with the properties of the
plasmagel, rather than of the plasmasol portions of the protoplasm.
As in the case of the Amoeba, the present measurements are based
upon the assumption that the relative degree of gelation is in propor-
tion to the minimum time of centrifugation^^ required for the
formation of a standard zoning at each of the indicated pressures.

C. CELL DIVISION

The effects of pressure upon cell division as it occurs in the eggs
of various marine animals have now been reported by several
workers (Brown, '34c; Marsland, '36, '38, and '39a; Pease and Mars-
land, '39; and Pease, '40a), and there is general agreement that

Protoplasmic Streaming — Relation to Gel Structure 139

sol ^ gel reactions play an important role in the mechanism which
operates to cleave the cytosoma.^'' This view has also been supported
recently by other types of work (Schechtman, '37; Chambers, '38;
and Dan and co-workers, '37) .
(1) The plas7nagel systeyn of egg cells

On the basis of a number of recent observations, it is possible to
compare the egg cell with the Amoeba and to distinguish two well
differentiated parts of the cytoplasm: (1) the plasmagel,-" a corti-
cally situated layer,-^ about 5 microns thick, which displays marked
gelation changes during the different phases of division, and (2) the
plasmasoh the more fluid internal cytoplasm, which participates in
the protoplasmic streaming at the time of cleavage, and in which
the spindle and the asters are formed.

The existence of a differentiated cortical plasmagel layer in the
Arbacia egg was first demonstrated clearly by the centrifuge experi-
ments of Brown, '34c, and of Costello, '34. In these experiments it
was observed that the formed elements of the cortex undergo cen-
trifugal displacement much less easily than those in the medullary
region of the egg, and that displacement of the cortical pigment
requires higher centrifugal forces, and longer centrifugal periods.
And Costello further observed that the apparent viscosity of the
internal protoplasm increases markedly as the temperature falls,
whereas the cortical layer, as judged by the displacement of the
pigment granules, becomes slightly less "viscous" in the range
between 20° and 2° C.

Subsequently there have been a number of confirmatory observa-
tions. Motomura, '35, and Schechtman, '37, noted that the protoplas-
mic currents of the frog's egg, which appear at the time of the
cleavage furrow, involve only the deeper cytoplasm, and that visible
granules in the peripheral gelled layer are not disturbed by the
stream. Chambers, '38, determined that the thickness of the plas-
magel layer in the furrow region of the Arbacia egg is about 5
microns, by the method of watching the furrow impinge upon a mass
of oil which had been injected into the stalk between the blasto-
meres, replacing the plasmasol in this region. Furthermore, Cham-
bers has observed-'- that when one of the two potential blastomeres
of a dividing egg is punctured, practically all of the internal cyto-
plasm pours out from the unpunctured blastomere, through the
aperture surrounded by the impinging furrow, leaving a rind of
protoplasm which is especially thick bordering the furrow. This
cortical remnant displays the properties of a gel, not only in that it

140 The Structure of Protoplasm

does not participate in the flow of escaping plasmasol, but also in its
behavior when manipulated with micro-needles. Finally, the work
of Marsland, '38 and '39a, has confirmed and amplified the original
observations of Brown.

(2) The setthig of the cortical gel diLring active cleavage

An indication that the furrowing process might involve gelation
reactions occurring primarily in the cortical layer of the egg cell was
provided by Brown's observation ('34c) that when the furrows are
about to appear (and during the time when they are actually
cleaving the cell) , the plasmagel part of the protoplasm becomes set
much more firmly than in other phases of the division cycle (see Fig.
5) . This very marked shift in the equilibrium is indicated by the fact
that no displacement of the pigment granules of the cortex can be
obtained even in a relatively strong centrifugal field (18,000 X
gravity, Marsland, '39a) when the eggs are centrifuged during the
cleavage period; whereas, if the centrifuging is done shortly before
or shortly after this period, a clear displacement can be obtained
in relatively weak fields (7,000 X gravity) .

On the basis of the foregoing observations, experiments were
undertaken to determine the effects of pressure upon the visible
aspects of cleavage in the Arbacia egg-'' (Marsland, '38) , and upon
the sol ^gel equilibrium of the plasmagel system (Marsland, '39) .

(3) The cleavage block at high pressures

The first type of experiment involved subjecting the eggs to a
relatively high pressure (7,000 Ibs./in.-) applied suddenly at the
time when fairly deep furrows were visible in a majority of the cells.
As soon as the pressure is applied, the progress of each furrow
ceases. This is true whether the invagination has just begun or
whether it has already almost cleaved the egg. Not only does the
inward movement of the furrow stop, but soon a slow recession
begins (Fig. 6) and if the pressure is maintained, each of the
bilobed cells gradually-^ reverts to a sphere.

It seemed likely that this phenomenon might be comparable to
the slow rounding of amoeboid cells which occurs at a similar pres-
sure. Apparently the bilobed form of the cleaving egg is stable
provided that the cortical plasmagel possesses a sufficient degree of
rigidity. According to this view the cells become rounded under
the agency of surface forces as soon as, due to the solating effect of
the pressure, the resistance of the plasmagel drops below a certain
critical value.

The inhibition of furrowing by pressure is remarkedly reversible.

Protoplasmic Streaviing — Relation to Gel Structure 141

As soon as the pressure is released the furrow recedes no further,
and within less than a minute it begins to push inward once more
toward the axis. The delayed furrowing now continues rapidly to
completion, provided that the pressure inhibition has not been main-
tained for more than 14 or 15 minutes.-"^ No injury appears to result

A

mn

B

Fig. 5. The setting of the cortical gel of the Arbacia egg, which occui'S just
prior to the appearance of the cleavage furrow. A, two unfertilized eggs, and B,
a fertilized egg 5 minutes before cleavage time, all centrifuged together for 6
minutes at atmospheric pressure. The greater rigidity of the cortical gel of B
is indicated by the fact that there has been scarcely any displacement of the
pigment granules, which in this egg (Arbacia piLstulosa) are localized entirely
in the cortex. Note that the undisplaced pigment is readily observable in the
cortical part of the hyaline zone.

from the temporary suppression, for the eggs continue to develop
in an apparently normal fashion. They go through the second and
third cleavage at the same time as control eggs, and give rise to
blastulae which cannot be distinguished from untreated specimens.
(4) Ejfect of pressure on the rate of furrowing

In view of the marked effect of the relatively high pressure, it
was of interest to determine how cleavage might be modified by
lesser degrees of constraint. Soon it became apparent that lower

149

The Structure of Protoplasm

Fig. 6. Recession of the cleavage furrows as a result of hydrostatic compres-
sion. A, eggs (Arhacia puiictulata) 10 seconds after a pressure of 6,500 lbs. /in.'
was established in the chamber; B, two minutes later, pressure still maintained;
C, two minutes after decompression, which occurred immediately after photo-
graph B was taken. Note (bl and b2) that recession of the furrow occurred
even though the "blastomeres" were conne^ed by a mere strand of protoplasm
at the time when the pressure was applied. The fertilization membranes were
removed shortly after the spei'm was added.

pressures merely retard the rate at which the furrow intrudes upon
the division axis, and a quantitative study of this effect was under-
taken.

In these experiments the fertihzed eggs were allowed to develop
in the chamber at atmospheric pressure until 4 minutes before the

Proto-plasmic Streaming — Relation to Gel Structure 143

furrows were expected-'" to appear, at which time the pressure was
raised quickly to the desired level. In order to time the progress of
the intruding furrow accurately, it was found necessary to remove
the fertilization membranes by shaking the eggs before placing them
in the chamber. This procedure permits the dividing eggs to become
more elongate in the direction of the division axis and enables one
to see the furrows more plainly. In this way it is possible to deter-
mine precisely when the furrow begins to form and when it finally
reaches the axis of the dividing cell.

As the pressure is increased there is a marked increase in the
time required for the furrow to complete its passage from the equator
to the axis. At 2,000 lbs. /in.- the rate of progress of the furrow is
only half the atmospheric rate,-' whereas at 5,000 lbs. the progress
is slowed to about one-fifth of the original rate. At pressures between
5,000 and 6,000 lbs., abortive furrows are formed which fail to reach
the division axis before receding. Above 6,000 lbs. no furrows
appear although the eggs do become slightly elongate at the time
when cleavage is due.

When one plots the rate of furrowing as a function of pressure
(Fig. 2) , it becomes apparent that the retardation is in proportion to
the solation effect which the pressure exerts upon protoplasmic gels
generally. This would indicate an intimate relation between the
mechanics of furrowing and the capacity of the plasmagel system
to undergo a process of setting at the time when the furrowing is
active. However, direct measurements of the pressure effect upon
the exceptionally stiff gel which is formed at this critical time,
remain for consideration.
(5) Solation of the firm "cleavage gel"

The eggs of Arhacia pustulosa, which are available at the Naples
laboratory in adequate quantities throughout the year, were chosen
for these experiments. These eggs possess an unusually generous
number of pigment granules, and this pigment is confined, even in
the unfertilized egg, almost entirely to the cortical layer of the
cytoplasm. These two attributes are of particular advantage in the
experiments.

Heretofore, it had not been possible, using centrifugal forces
up to 7,200 X gravity, and pressures up to 7,000 Ibs./in.- (Brown,
'34c) , to cause any appreciable displacement of the pigment granules
if the centrifuging were done late in the division cycle, i. e., within
10 minutes of the time when the furrows were due to appear. In

144 The Structure of Protoplasm

tiie present experiments it was found necessary to use a force of
16,500 X gravity to secure an adequately rapid displacement of the
pigment granules through the firmly set plasmagel. But even this
force is not sufficient to dislodge the granules except under the
liquefying action of pressures greater than 1,500 Ibs./in.- At
atmospheric pressure, a force of 18,680 X gravity, the highest
immediately available at the Naples laboratory, gave practically no
pigment displacement even with prolonged centrifugation.

The centrifugal force used in the experiments was 16,500 X
gravity, and each sample of eggs, compressed to the desired degree,
was centrifuged for a period just sufficient to displace the cortical
pigment into a compact zone at the centrifugal pole of the cell. In
each case the eggs of a single female were fertilized and allowed to
develop in the usual fashion until the first furrows began to appear.
Then, without delay, the control and experimental samples were
placed in their respective sections of the centrifuge-pressure bomb,
and the desired pressure was established in the pressure section.
The operations were so fixed that, at the time when the pressure
was applied and the centrifuging begun, about 50 per cent of the
eggs possessed furrows of greater or lesser depth.

The very marked solating action of the higher pressures upon
the pigment-laden cortical protoplasm of the cleaving egg is clearly
shown in Figure 7. All of these eggs were centrifuged at the same
time for 2V2 minutes. It may be noted that each of the pressure-
treated specimens displays, in addition to the usual oil, hyaline, and
yolk zones, a sharp pigment zone, at the centrifugal pole. This zone is
absent in the atmospheric specimens.-"' Also it may be noted that the
one egg which had just completed the first cleavage by the time the
suppressing action of the pressure could take effect displays the
same physical properties as the others.

Assuming that the firmness (gelation percentage) of the cortical
gel is in proportion to the minimum centrifuge time required to pro-
duce a standard zoning of the pigment granules, in the range of
pressure wherein an effective displacement occurs, it is possible to
plot the gelation values as a function of pressure. These data (Fig. 2)
indicate, not only that the very firm plasmagel of the cleavage period
shows the same relative susceptibility to the solating effects of
pressure as other cellular gels, but also that the inhibition of furrow-
ing is quantitatively related to the concomitant shifts in the sol ^ gel
equilibrium.

Protoplasmic Streaming — Relation to Gel Structure 145

D. BEHAVIOR OF UNICELLULAR PIGMENTARY EFFECTORS

Recently, experiments have been started to determine the effects
of pressure upon the ebb and flov/ of pigment granules as it occurs
in the melanophores of Fundulus (Marsland, '40) . As yet the
results are far from complete, but there are clear indications that

1^*9

^

^ # i^

Fig. 7. Solation of the stiffer "cleavage gel" by pressure. All of these eggs
were centrifuged simultaneously for 2.5 minutes at the time when 50 percent
of them displayed furrows. The E (experimental) samples were at 7,000 Ibs./in.",
whereas the control (C) eggs were at atmospheric pressure. The solation of the
cortical gel of the pressure specimens is indicated by the pigment zones at the
centrifugal ends, and by the fact that all pigment has been thrown out of the
hyaline zones.

this type of streaming may also involve gelational phenomena. These
results, if borne out by further work, may provide a basis for decid-
ing M^hether the unicellular pigmentary effector is to be regarded
as a modification of visceral muscle (Spaeth, '16) or as an amoeboid
type of cell (Hooker, '14) .

(1) Suppression of the "contraction" phase of the pigmentary

response

In this type of experiment the isolated scale of Fundulus v/as

immersed in N/10 KCl solution and exposed to various degrees of

compression in the microscope-pressure chamber. At atmospheric

146 The Structure of Protoplasm,

pressure the scales in KCl assume the punctate form, that is to say,
all of the pigment granules are withdrawn from the numerous twig-
like branches which radiate out into the surrounding tissues, and
are aggregated in the central mass of the cell. As the pressure is
increased, however, by steps of 1,000 lbs., invariably there is a
greater and greater dispersal of the pigment (see Fig. 8) until at

1000 LSS / IN'

iOOO

4000 6000 7000

SINGLE MELANOPHORE ; SCALE OF FUNDULUS ISOLATED IN N/10 KCL

Fig. 8. Suppression of the "contraction" phase of the response of a unicellular
chromatophore.

about 7,000 lbs. the pigment reaches the extremities of the proto-
plasmic branches and all of the melanophores have becoine com-
pletely stellate.

The "expansion" of the melanophore appears to reach an equil-
ibrium value at each level of pressure, and if the pressure is kept at
a certain intensity the melanophores remain expanded to the char-
acteristic extent. This steady state is reached in 40-80 seconds after
each shift in pressure, and is approximately the same whether a
certain level is reached from above, during a step-wise decompres-
sion, or whether it is attained from below via step-wise increments
of pressure. Furthermore, at a given pressure each cell returns

Protoplasmic Streaming — Relation to Gel Structure 147

approximately to the same form when the experiment is repeated
a number of times, provided the exposure to pressures in excess of
5,000 lbs. does not endure beyond about 20 minutes.

Quantitative measurements of the chromatophoral expansion as
a function of pressure are difficult due to the extremely irregular
form of the pigment cell, and at the present time only qualitative
data are available. The same is true of centrifuging experiments
which have been undertaken to determine the effects of pressure
upon the gel properties of these cells. Without question solation does
occur, but several technical difficulties must be overcome before
accurate measurements can be made.

From a qualitative point of view, however, it would appear that
the contraction phase of the pigmentary response is limited by
pressure in a manner that parallels, at least roughly, the inhibition
of gelation which has been demonstrated in plasmagel systems
generally. It may be seen, for example, that the "half-expanded"
state of the chromatophore which occurs at about 2,500 Ibs./in.,-
corresponds to a gelation value of approximately 50 per cent, and it
seems probable that the other values, when they become available,
may likewise fall upon the general curve.
(2) Effects on pulsating chromatophore s

In this other type of experiment the chromatophores were induced
to pulsate by the method-"* of Spaeth, '16, before the isolated scales
were placed in the pressure chamber and exposed to pressures of
1,000, 2,000, . . . 8,000 Ibs./in.- Immediately it becomes apparent
that the pressure imposes a limitation upon the "contraction phase"
of the pulsation. At 1,000 lbs. the pulsations continue, but although
the outward flow of the pigment granules is complete and reaches the
distal extremities of the numerous protoplasmic branches, the inward
flow is curtailed and a reversal of the direction of flow occurs before
the end of the procession of granules returning from each branch
quite reaches the control mass of the cell. At 2,000, 3,000, . . . 6,000
Ibs./in.,- a greater and greater reduction in the amplitude of the
pulsations is witnessed, but the reduction is due entirely to a
further and further curtailment of the inward flow, for the outward
flow continues to reach the distal ends of the branches. Finally, at
7,000 lbs. all pulsations cease, and while the pressure is maintained,
all of the melanophores remain in a fully expanded condition. When
decompression occurs, however, provided it has not been delayed
beyond some 25 minutes, the chromatophores immediately undergo

148 The Structure of Protoplasm

a complete and vigorous "contraction," and within about 3 minuiei;,
full pulsations have been resumed.

(3) Coraparison between pseudopodia and the branches oj a pig-
ment cell

One further observation may be added to the evidence which
indicates that the effects of the pressure are being mediated through
solational changes similar to those which have been demonstrated
in other cases. At pressures of 7,000-8,000 Ibs./in.- occasionally it
may be seen that one of the several elongate protoplasmic processes
which radiate out from a particular melanophore becomes pinched
off, losing its connection with the central portion of the cell at a
greater or lesser distance from the origin. Such isolated portions of
the cell become rounded into a discrete mass while the higher
pressure is maintained, but show abortive attempts at "expansion"
and "contraction" when the pressure is reduced.

This observation indicates that the branches of the unicellular
chromatophore are in a sense comparable to the pseudopodia of the
amoeboid cell and that a reorganization of the elongate form may
occur when a profound solation of the protoplasm is induced. The
greater resistance in the case of the chromatophore may be due
largely to the fact that its branches are not free, but rather extend
out into the interstices between the other tissues (see Matthews, '31) .
Under such conditions, no doubt, the chromatophore branches receive
additional support, from contact with (or attachments to) the sur-
rounding cells, and this support is sufficient, in most cases, to prevent
collapse even in periods of complete protoplasmic liquefaction.

IV. CELLULAR ACTIVITIES WHICH DO NOT INVOLVE

A very brief and fragmentary consideration of the many experi-
ments which have dealt with the effects of pressure upon the physio-
logical activity of muscle, nerve, cilia, and flagella may suffice to show-
that this group of phenomena must be considered in a separate
category. The physiological activities which already have been
considered at length display certain common attributes. All are
progressively inhibited throughout the entire physiological range of
pressure, and in all cases, the degree of inhibition appears to be in
proportion to the suppressing action of pressure upon the formation
of protoplasmic gels. In contrast to this, the physiological activities
which remain for consideration are not inhibited, but rather are
augmented in the lower portion of the pressure range, and further-
more, no clear relationship has been demonstrated between the

Protoplasmic Streavung — Relation to Gel Structure 149

physiological effects and the effects of the pressure upon the physical
characteristics of protoplasmic gel systems.

A. MUSCULAR CONTRACTION

The marked increase in the tension developed during a single
isometric contraction^" at pressures between 100 and 6,000 Ibs./in.^
was reported, first for cardiac muscle (Edwards and Cattell, '27)
and then for skeletal muscle (Cattell and Edwards, '28) . For cardiac
muscle the maximum augmented tension is four to six times greater
than the atmospheric value and occurs at about 6,000 Ibs./in.-
(Edwards and Cattell, '30). For skeletal muscle, however, the
maximum additional tension is only 20-40 per cent greater than the
atmospheric value and is reached at pressures between 2,000 and
4,000 Ibs./in.- (Cattell and Edwards, '32) .

The pressure effect on tension in both cardiac and skeletal muscle
varies, however, markedly with temperature. In fact, a reversal of
sign occurs, at 5-8° C. for cardiac muscle (auricular muscle of the
turtle. Brown, '34a), and at 9-14° C. for striated muscle (sartorius
muscle of the frog, Cattell and Edwards, '32) . Below these critical
points a depression of tension occurs throughout all of the physio-
logical pressure range.

Brown ('36) has observed that the full measure of additional
tension is obtained only when the period of compression antecedes
the moment of stimulation, and that no extra tension is obtained
unless the compression intervenes before the first one-eighth of the
contraction has been completed (Fig. 9) . In fact, when the period
of compression is confined to the remaining seven-eighths of the
contraction phase, the tension amounts to less than the atmospheric
control value (Brown, '34b and '36) .

B. CONDUCTION OF THE NERVE IMPULSE

Grundfest and Cattell ('35) have demonstrated clearly that the
compression effects on the nerve impulse do not constitute a simple
case of progressive inhibition. These observations, on the form and
magnitude of the spike potential and on the rate of propagation,
were recorded by means of a cathode ray oscillograph, utilizing
grouped A fibres of the frog sciatic nerve, and in some cases, single
fibres of a sciatic-peroneal preparation.

The effects of moderate compression (below 5,000-6,000 Ibs./in.^)

c J, "O

O "TJ 0)

c c 8 g-

CO ■

o

O 01

•Sc

o

o
9

■on
c

c

CO

a

^-s

:;: G o

^8

+^ tn

^ C

O o

'-'0 0)

>> a

O m
3

,^ <" ^

l« m _

s <u o

CD S •-•
!-( -^ +^

o in CO

sec

CD
t^ o

a u
St

(« O

o o
a <u
X <"

T3 O

CO
CO

o
pq

O 0)
OJ

ao

S "m

Id
3
a"

OJ

0) CD

o

CD

a

C73

O

fa

0)

a'

a

o
o

CO ■

CO

C C O

o ? +^
3 O CO

.3 ^ 0) CO

0)

Protoplasmic Streaming — Relation to Gel Structure 151

are (1) an increase, up to about 10 per cent, in the height of the
spike potential, (2) an increase, up to 20 per cent, in the duration
of this potential, and (3) an increase, up to 12 per cent, in the rate
at which the action potential is propagated. At pressures above
6,000 lbs. the height of the spike and the rate of propagation fall
away, and at about 13,000 lbs. the nerve becomes practically non-
responsive. However, the duration of the potential continues to
increase throughout the higher range. At pressures above 5,000 lbs.
it is also observed that a single short shock evokes not one, but
several (up to four) , discharges, even in the single fibre.

C. THE MOTILITY OF CILIA AND FLAGELLA

The work of Pease and Kitching ('39), on the influence of pres-
sure on ciliary frequency in the gill filaments of the common mussel,
Mytilus edulis, clearly demonstrates the accelerating effect of sudden
increases in the pressure level, and the deceleration which occurs
with each abrupt drop of pressure (see Fig. 10) . In these experiments
a stroboscopic method, accurate to the nearest 10 beats per minute,
was used to measure the frequency of the ciliary beat.

It is interesting to note that, at least for moderate changes of
compression, the frequency tends to return to the initial basic value,
when the pressure is maintained at the higher or the lower level.
When pressures above 5,000 Ibs./in.- are maintained, however, after
the initial burst of activity, the basic rate falls away rapidly, and
within 10-15 minutes the cilia stop — an observation which is in
agreement with the earlier studies of Regnard ('84 and '91) and of
Certes ('84).

Pease and Kitching find no parallel between the changes of fre-
quency and the known effects of pressure upon protoplasmic gel
systems. They do, however, point out the striking similarity between
the ciliary frequency effects and the cardiac frequency results
obtained by Edwards and Cattell ('28).

Quantitative studies dealing with pressure effects on the motility
of flagella are entirely lacking, and the qualitative observations
which have been made are very fragmentary. However, it seems
safe to say that this type of movement is even less susceptible than
the ciliary type to inhibition by pressures within the range of
reversible action. For example, Marsland and Brown observed in a
cinematographic record of the action of pressure upon amoeboid
movement that the motility of a few small flagellates, which inad-

152

The Structure of Protoplasm

vertently had been included with the Amoeba in the pressure
chamber, continued without any sign of abatement while the pres-
sure was maintained at 7,000 Ibs./in.- for a period of about half an
hour. Also, Marsland and Rugh ('40) could observe no change in

Fig. 10. (From Pease, '39, p. 137.) A single complete record of changes in the
ciliary rate with rapid pressure changes of 5,000 Ibs./in.^

the motility of sperm (Rana pipiens) during a compression period
of 3 hours at 8,000 Ibs./in.^

V. GENERAL DISCUSSION
A. NATURE OF THE PRESSURE EFFECT ON GEL STRUCTURE

One important question which must be raised is whether the
influence of pressure on protoplasmic gel systems is a direct action
upon the sol ^ gel equilibrium itself, or whether the direct action
may be upon some other of the metabolic equilibria in the cell,
which in turn may mediate the effect upon the gel system.

At the present time it would seem that the weight of evidence
favors the direct action view. Freundlich ('37) , on the basis of work
by Heyman ('35 and '36), points out that gel equilibria fall into
three categories, according to volume changes'^^ which accompany
the setting of the gel. In the first type, which is exemplified by

Protoplasmic StreaTuing — Relation to Gel Structure 153

sodium oleate and other purely thixotropic gels, no change of volume
occurs when gelation or solation occurs at a fixed temperature, and
the shift of equilibrium which is induced by periods of shaking or
rest, is isothermal. The second type, exemplified by gelation or agar
gels, displays a small decrease of volume with setting, and heat is
evolved in the process. Finally, the third type, which is represented
by colloidal aqueous solutions of methylcellulose, shows a small
increase of volume and an absorption of heat during the setting
process. Thus it is found that gels of the second type tend to set
upon cooling, whereas those of the third type undergo solation as
the temperature falls.

From the principle of Le Chatelier, it appears certain that hydro-
static pressure would favor solation in the third type of gel, gelation
in the second type, and little or no change in the first. And the
experimental evidence in this regard, although scanty, clearly sub-
stantiates the predictions. Thus Marsland and Brown ('39) have
found that gelatin sols, which are known to belong in the second
category, undergo marked gelation at pressures in the physiological
range, whereas gel preparations of myosin extracted from rabbit
muscle are solated to a degree which corresponds, at least in quali-
tative fashion, with the solational effects of pressure on protoplasmic
gels.

In all probability, therefore, the various protoplasmic gels which
have proved so susceptible to solation under pressure, belong to
Freundlich's group III. This viewpoint is borne out by the truly
remarkable regularity of the effect in different systems — for in
every case which has been studied, each unit increment of pressure
produces the same proportionate loss in the tensile properties of the
gel which is being tested. If these changes were being effected
indirectly through the medium of other chemical reactions, such a
degree of regularity would scarcely be expected in so many different
kinds of cell. Furthermore, it has been observed (Costello, '34)
that the cortical gel of the Arbacia egg becomes less "viscous" as the
temperature falls, a behavior which would not be expected unless
this gel is of the third type.

The common component upon which the pressure acts in the
various cells probably will prove to be one or more groups of
protein substances which is capable of forming gels of the nature
of type III. And in this connection it is interesting to note that
Mirsky ('36) has extracted from the Arbacia egg (a cell which very
plainly shows the characteristic pressure-solation effect) a protein

154 The Structure of Protoplasm

substance which closely resembles myosin with reference to its
general chemical properties.

One cannot overlook the fact, however, that pressure also may
induce marked changes in other types of metabolic reactions in the
cell. Fontaine ('28) , for example, reports increased oxygen con-
sumption in a number of small marine animals exposed to pressure
in the range up to about 1,500 Ibs./in.,- and Deuticke and Ebbecke
('37) describe a greatly accelerated breakdown of phosphocreatin
and an increased lactic acid production in frog skeletal muscle during
a short period of tetanic contraction at 8,000 lbs. Furthermore, it is
well known that a great variety of organic reactions are markedly
affected as to rate, and as to the point of equilibrium, by pressures
within and beyond the physiological range (see, for example, Faw-
cett and Gibson, '34) .

B. RELATION BETWEEN GELATION AND THE PHYSIOLOGICAL RESPONSE

In seeking fundamental factors which may explain why in one
group of physiological activities (amoeboid movement, cyclosis, cell
division, and pigmentary responses) the inhibition imposed by
pressure is continuously in proportion to solation throughout all of
the pressure range, whereas this is not the case for a second group
(contraction of muscle, conduction of nerve, and the motility of cilia
and flagella) , two questions may be raised. (1) Is it true, perhaps,
that gelation reactions are directly concerned in the development
of the mechanical energy in the first group of responses but are not
concerned, at least directly, in the second; or (2) is it possible that
gelation, reactions are involved in both cases but that the metabolic
reactions which precede the active response and provide energy for
it are relatively immune to augmentation by pressure in the first
group, but not immune in the second?

An affirmative answer to the first question would provide the
basis for a simpler working hypothesis. However, the data of Brown
('34b and '36) , which indicate that augmented tension in muscle
depends upon a greater mobilization of energy prior to the actual
contraction and that the contractile process itself is actually
depressed even by relatively low pressure, make it doubtful that
this more restricted view is entirely tenable. In any event, however,
in view of the accumulated data, one can scarcely avoid the conclu-
sion that gelation reactions are directly concerned in the develop-
ment of mechanical energy, at least in one group of physiological

Protoplasmic Streaming — Relation to Gel Structure 155

activities, the group in which protoplasmic streaming constitutes
the common attribute.

C. POSSIBLE MECHANISMS BY WHICH GELATION MAY PRODUCE STREAMING

(1) VoliLme changes in relation to streaming

In an earher paper (Marsland and Brown, '36) it was suggested
that the motive force of streaming in the amoeboid cell might take
origin in the small changes of volume which occur during the gelation
and solation of the cytoplasm. This position now seems untenable in
the light of Heyman's ('35 and '36) work. In all probability the sign
of these changes is in reverse, that is to say, just the opposite kind
of volume changes would be necessary to give streaming in the
proper direction. The decrements of volume would need to occur
anteriorly, near the tip of the pseudopodium, where gelation is
known to be taking place, and the increments of volume would need
to be localized posteriorly where solation is occurring. To provide
volume changes of the proper sign, the plasmagel of the Amoeba
would of necessity belong to Freundlich's type II, but in this case
one would expect to obtain gelational effects from pressure, instead
of the solation which has clearly been demonstrated.

In the streaming of plant cells, however, the possibility still
remains that volume changes accompanying a series of sol ^ gel
reactions play an important role in generating the motive force
(Marsland, '39b) , but the question cannot be answered decisively
until the focal points where solation and gelation take place have
been localized more definitely. The high turgor which is character-
istic of the plant cell provides a favorable condition for the effective-
ness of such changes, since a greater fund of energy is mobilized
when volume changes take place in a medium under the constraint
of higher pressure.
(2) Contractile theories of streaming

According to Lewis ('39) streaming in amoeboid cells"- and the
movements which occur in dividing cells are motivated by elastic
and contractile properties which are inherent in gel structures gen-
erally, and in the plasmagel systems of certain cells in particular.
Given time, most gels do, of course, undergo a process of contraction
by which gradually the sol is squeezed forth from the colloidal
interstices into the surrounding medium, as, for example, serum is
expressed from a blood clot, or soap solution from a soap gel.

Such a theory presupposes, no doubt, that shrinkage in the

156 The Structure of Protoplasm

animate gel is accelerated by metabolic conditions established in
the cell, and also that these metabolic conditions create differences
in the contractile force in different parts of the plasmagel system.
In an Amoeba, for example, there would need to be a gradient such
that the contractile properties are greater in the older posterior
parts of the plasmagel than in the more recently formed anterior
parts of the system. In addition to the descriptive evidence for such
a gradient, which is provided by the work of Mast and of Lewis, the
pressure experiments of Marsland and Brown indicate a distinctly
lesser rigidity in plasmagel walls near the extremities of the pseudo-
podia, and a loss of this property would indicate, no doubt, a lesser
contractile potentiality.

An extension of the quantitation data yielded by the recent
experiments of Noburo Kamiya ('40) and of Norris ('40) may,
perhaps, provide a basis for deciding whether or not the elastic
contractile forces generated in a plasmagel system are adequate to
account for the vigor of the streaming. Fortunately both of these
workers are using the same organism, one of the myxomycetes,
PhysaruTTi polycephalum. The ingenious method of Kamiya appears
to give very accurate measurements of the flow pressures which are
developed as the protoplasm streams back and forth in the channel
between two parts of the plasmodium, and it is to be hoped that
Norris' technique of determining the elastic properties of the gel
system can be adapted to measure the changes of tension which
must occur concomitantly with the rhythmic alternations in the
direction of flow.

The viewpoint expressed by Seifriz ('37) represents somewhat
of a departure from the ideas of Lewis and Mast. Time-lapse motion
pictures enabled Seifriz to demonstrate very plainly that each
portion of the plasmodium of the slime mold displays alternating
periods of expansion and contraction and led him to believe that
these pulsations are analogous to the rhythmic beating of visceral
muscle as controlled by the autonomic nervous system. Funda-
mentally, perhaps, there may be little difference between the two
viewpoints, for in the last analysis it may be that the same forces
are involved in the setting and contraction of certain gels and the
contractile processes in the myofibrillae. If this be true, the marked
differences, in regard to the consequences of hydrostatic compression,
between muscle cells and cells which are characterized by proto-
plasmic streaming, are due to more superficial factors, as for

Protoplasmic Streaming — Relation to Gel Structure 157

example, differences in the metabolic reactions which are precursory
to the development of the mechanical energy.

VI. SUMMARY OF CONCLUSIONS

(1) The general effect of hydrostatic compression upon a num-
ber of different cells is to induce a uniform solation of gelated parts
(the plasmagel system) of the protoplasm. Each increment of 1,000
Ibs./in.- reduces the gel rigidity by almost 25 per cent.

(2) All of the protoplasmic gels which have been studied appear
to conform to the type (Freundlich, '37) in which the process of
gelation is accompanied by a small increase of volume.

(3) Certain types of cellular movement (amoeboid movement,
cyclosis, cell division, and the flow of pigment in unicellular pig-
mentary effectors) , in which protoplasmic streaming is the common
attribute, are especially susceptible to a pressure-induced inhibition.
In contrast, other physiological activities (muscular contraction,
nervous conduction, and the motility of cilia and flagella) are
augmented by low and moderate degrees of pressure and suffer
depression only in the higher range (above 5,000 Ibs./in.-) .

(4) The inhibiting effect of pressure upon the various kinds of
protoplasmic streaming is in proportion to the degree of solation
which the pressure induces in the plasmagel systems.

(5) To interpret these results, it is postulated that sol ^ gel
reactions constitute an intermediary mechanism whereby potential
energy is converted by the cell into mechanical energy — the mechani-
cal energy of the streaming protoplasm.

FOOTNOTES

'Brown ('34c), Marsland ('38 and '39), Pease and Marsland ('39).

"Brown and Marsland ("36).

'Kitching and Pease ('39).

'Ebbecke ('36).

° Marsland ('39b).

'Pease ('40b).

' Judging from the compressibility of muscle tissue (see Cattell, '36, p. 458) ,
the loss of protoplasmic volume in the range of pressure used in the present
experiments, would be about 2 per cent.

' Previous microscope-pressure chambers permitted observation of the com-
pressed specimens, but at much lower magnification (see, for example. Draper
and Edwards, '32) .

" Leitz, U. M.

'" This amount of working distance is necessary since the walls of the cham-
ber must be thick enough to support the high internal pressure.

158 The Structure of Protoplasm

"Reorganization of a fluid cylinder into drops will occur when the length
exceeds pi times the diameter, provided the viscous (or plastic) resistance of
the material is not too great.

'- This assumption seems fully justified from measurements on a variety of
cells. Each pressure increment induces the same percentage degree of solation
regardless of the absolute initial rigidity of the particular gel.

^'These values frequently have been referred to as the relative viscosity of
the protoplasm. This term, however, fails to distinguish between measurements
dealing with solated, and those made upon gelated parts of the system. Further-
more, in a truly viscous system some displacement of suspended granules occurs
even under very low centrifugal forces, which is not the case in the present
measurements.

''At constant centrifugal force, namely 7,000 X gravity.

''The elongation of the pseudopodia which occurs at pressures below 2,000
lbs. is probably related to the observation that in this range no plasmagel- sheet
(Mast, '26) is present at the pseudopodial tips. Apparently this part of the
plasmagel system, which tends to restrain the outflow of the plasmasol, and
which constitutes a gelation focus when retraction is to begin, is more labile
than the other parts.

'"Recently, Pease ('40) describes an acceleration of streaming in the Plasmo-
dium of the slime mold (Physarum polycephalum) at pressures between 1,000
and 3,000 Ibs./in.' This would constitute an exception to the general rule that
protoplasmic streaming is progressively inhibited by pressure in the physiologi-
cal range. Perhaps the discrepancy may be explained on the basis that in
Physarum the whole Plasmodium suffers reorganization in the range. The
acceleration may indicate that, as the contours of the protoplasmic mass are
changing, surface forces become active to augment the flow. This interpretation
is favored by the observation that the accelerated flow occurs chiefly in Plas-
modia which are not attached to the substratum but which are freely suspended
in the medium. A further factor in the acceleration may be found in the
reorganization which occurs in the channel walls. Thus the flow in a particular
channel might be augmented as a result of an anastomosis from a neighboring
channel. To obtain quantitative data in this connection, perhaps the technique
of Kamiya ('40) could be used.

"As for example, when streaming is about to stop, or when it is barely
beginning. At such times the channels and the direction of flow may be highly
irregular.

''At constant centrifugal force. In the case of Elodea this was 810 X gravity.
'"Recent work by Pease (personal communication) indicated that sol^gel
reactions may also be important in the migration of the chromosomes during
anaphase.

"' The use of the terms plasmagel and plasmasol with reference to the cyto-
plasm of egg cells was first proposed by Marsland ('39a) .

^' It is possible that the plasmagel system of the egg cell also includes some
of the medullary cytoplasm such as parts of the spindle and asters.

" This observation is recorded in one of Dr. Chambers' cinema films. See also
Chambers ('38).

-" The eggs of Arhacia punctulata were used in the study of the visible effects
of pressure upon cleavage, whereas those of Arbacia pustidosa served for the
measurements of the solation effects.

-* In from 1-10 minutes, depending upon the depth to which the furrow has
gone, and upon the intensity (above 7,000 lbs.) of the pressure.

''Apparently the internal conditions which result in cleavage persist for a
hmited time (about 15 minutes at 22° C). If the experimental inhibition is
prolonged beyond this, the tendency to cleave lies dormant until control eggs
have just begun the second cleavage. Now the eggs in which the "first" cleavage
has been suppressed, also begin to cleave, usually into four blastomeres directly,

Protoplasmic Streaming — Relation to Gel Structure 159

although a few three-cell and two-cell specimens may appear. Later, a number
of these embryos fail to gastrulate normally.

^The data of Fry ("36) were very useful in determining the schedule. The
eggs of most females start to cleave as follows: at 21X., in 60 min.; at 22% in
52 min.; and at 23°, in 47 min.

-' At 23° C. complete furrowing at atmospheric pressure requires an average
time of 3.1 min., in contrast to the 6.2 min. required at 2,000 lbs./ in.'

^ The lesser sharpness of the oil, hyaline, and yolk zones in the atmospheric
specimens represents an illusion due to the fact that they are seen through the
layer of undisplaced pigment. Sharp focussing shows that actually these zones
are quite equal in the atmospheric and the pressure specimens.

=^The isolated scales are immersed, first in N 10 NaCl for 15 min., and then
in N/IO BaCL for 7 min. When returned to the sodium chloride solution, pulsa-
tions begin in about 45 min. and continue for 2 hrs. or more.

'•"' Only the pressure effects upon tension in the single twitch will be consid-
ered in this paper. In other words, none of the work on tetanic contractions, or
upon contractures, will be included.

'•In these volume experiments, Heyman used a very sensitive dilatometer,
kept constant to 0.003 °C. The capillary which served to measure the volume
change was very fine, and a 1-cm. excursion of the meniscus corresponded to a
volume change of 0.0016 cc; and since the complete volume of the sol which was
turning to a gel was 80 cc, a change of volume amounting to 0.0002 per cent
could be measured very accurately.

=- See also papers by S. O. Mast (e. g., "26 and "31) and the paper by W. H.
Lewis in the present monograph.

LITERATURE CITED

Brown, D. E. S. 1934a. The pressure-tension-temperature relation in cardiac

muscle. Amer. Jour. Physiol., vol. 109, p. 16.
. 1934b. The effect of rapid changes in hydrostatic pressure upon the

contraction of skeletal muscle. Jour. Cell, and Comp. Physiol., vol. 4,

page 257.
. 1934c. The pressure coefficient of "viscosity" in the eggs of Arbacia

punctulata. Jour. Cell, and Comp. Physiol., vol. 5, p. 335.

1936. The effect of rapid compression upon the events in the isometric

contraction of skeletal muscle. Jour. Cell, and Comp. Physiol., vol. 8,

p. 141.
Brown, D. E. S., and D. A. Marsland. 1936. The viscosity of Amoeba at high

hydrostatic pressure. Jour. Cell, and Comp. Physiol., vol. 8, p. 159.
Cattell, M. 1936. The physiological effects of pressure. Biol. Rev., vol. 11, p. 441.
Cattell, McK., and D. J. Edwards. 1928. The energy changes of skeletal muscle

accompanying contraction under high pressure. Amer. Jour. Physiol.,

vol. 86, p. 371.
AND . 1932. Conditions modifying the influence of hydro-
static pressure on striated muscle with special reference to the role of

viscosity changes. Jour. Cell, and Comp. Physiol., vol. 1, p. 11.
Certes, a. 1884. Note relative a Taction des hautes pressions sur la vitalite des

micro-organismes d'eau douce et d'eau de mer. C. R. Soc. Biol., Paris,

T. 36, p. 22.
Chambers, R. 1938. Structural and kinetic aspects of cell division. Jour. Cell

and Comp. Physiol., vol. 12, p. 149.
CosTELLO, D. P. 1934. The effects of temperature on the viscosity of Arbacia egg

protoplasm. Jour. Cell, and Comp. Physiol., vol. 4, p. 421.
Dan, K., T. Yanagita. and M. Sugiyama. 1937. Behavior of the cell surface during

cleavage. Protoplasma, Bd. 28, S. 66.

160 The Structure of Protoplasm

Deuticke, H. J., AND U. Ebbecke. 1937. tJber die chemischen Vorgan^e bei der
Kompressionsverkiirzung des Muskels. Hoppe-Seyler's Zeitschr. phy-
siol. chem., vol. 247, p. 79.

Draper, J. W., and D. J. Edwards. 1932. Some effects of high pressure on develop-
ing marine forms. Biol. BuL, vol. 63, p. 99.

Ebbecke, U. 1936. tJber plasmatische Kontraktionen von roten Blutkbrperchen,
Paramacien, und Algenzellen unter der Einwirkung hoher Drucke.
Pflug. Arch. ges. Physiol., B 238, z. 452.

Edwards, D. J., and McKeen Cattell. 1927. Some results of the application of
high pressures to the heart. Proc. Soc. Exp. Biol, and Med., vol. 25, p. 234.

AND . 1928. The stimulating action of hydrostatic pressure

on cardiac function. Amer. Jour. Physiol., vol. 84, p. 472.

AND . 1930. The action of compression on the contraction

of heart muscle. Amer. Jour. Physiol., vol. 93, p. 90.
i'AwcETT, E. W., AND R. O. GiBSON. 1934. The influence of pressure on a number

of organic reactions in the liquid phase. Jour. Chem. Soc, Part I, p. 386.
Fontaine, M. 1928. Les fortes pressions et la consommation d'oxygene de quel-

ques animaux marins. C. R. Soc. Biol., T. 99, p. 1789.
. 1929. De Faction des fortes pressions sur les cellules vegetales. C. R.

Soc. Biol., T. 101, p. 452.
Freundlich, H. 1937. Some recent work on gels. Jour. Phys. Chem., vol. 41,

p. 901.
Fry, H. J. 1936. Studies of the mitotic figure. V. The time schedule of the mito-
tic changes in developing Arbacia eggs. Biol. Bui., vol. 70, p. 89.
Grundfast, H., and McK. Cattell. 1935. Some effects of hydrostatic pressure

on nerve action potentials. Amer. Jour. Physiol., vol. 113, p. 56.
Heilbrunn, L. V. 1929. The absolute viscosity of Amoeba protoplasm. Proto-

plasma, vol. 8, p. 65.
Heyman, E. 1935. Studies on sol ^ gel transformation. I. Inverse sol ?^ gel

transformation of methyl-cellulose in water. Trans. Farad. Soc, vol. 31,

p. 846.
. 1936. Studies on sol ^ gel transformation. II. Delatometric investi-
gations on iron hydroxide, gelatin, methylcellulose, silicic acid and

viscose. Trans. Farad. Soc, vol. 32, p. 462.
Hooker, D. 1914. Amoeboid movement in the corial melanophores of frogs. Anat.

Rec, vol. 8, p. 103.
Kamiya, N. 1940. The control of protoplasmic streaming. Science, vol. 92, p. 462.
KiTCHiNG, J. A., and D. C. Pease. 1939. The liquefaction of the tentacle of suc-

torian protozoa at high hydrostatic pressures. Jour. Cell, and Comp.

Physiol., vol. 14 (supplement of No. 3), p. 1.
Lewis, W. H. 1939. The role of a superficial plasmagel layer in changes of form,

locomotion and division of cells in tissue cultures. Arch. Exp'l. Zell-

forsch. Gewebziicht, vol. 23, p. 1.
Marsland, D. a. 1936. The cleavage of Arbacia eggs under hydrostatic com-
pression. Anat. Rec, vol. 67, p. 38 (supplement).
. 1938. The effects of high hydrostatic pressure upon cell division in

Arbacia eggs. Jour. Cell, and Comp. Physiol., vol. 12, p. 57.
. 1939a. The mechanism of cell division. Hydrostatic pressure effects

upon dividing egg cells. Jour. Cell, and Comp. Physiol., vol. 13, p. 15.
. 1939b. The mechanism of protoplasmic streaming. The effects of

high hydrostatic pressure upon cyclosis in Elodea canadensis. Jour.

Cell, and Comp. Physiol, vol. 13, p. 23.

Protoplasmic Streaming — Relation to Gel Structure 161

1940. The effects of high hydrostatic pressure on the melanophores

of the isolated scales of Fundulus heteroclitus. Anat. Rec, vol. 78, p
168 (supplement).

Marsland, D. a., and D. E. S. Brown. 1936. Amoeboid movement at high hydro-
static pressure. Jour. Cell, and Comp. Physiol., vol. 8, p. 167.

AND . 1939. The effects of hydrostatic pressure upon gela-
tion phenomena in animate and inanimate systems. Anat. Rec, vol. 75,
p. 141 (supplement) .

AND R. RuGH. 1940. Resistance of sperm of Rana pipiens to hydro-

static compressions; eflfect upon embryonic development. Proc. Soc.

Exp. Biol, and Med., vol. 43, p. 141.
Mast, S. O. 1926. Structure, movement, locomotion, and stimulation in Amoeba.

Jour. Morph. and Physiol., vol. 41, p. 347.
. 1931. Locomotion in Amoeba proteus (Leidy). Protoplasma, vol. 14,

p. 321.
Matthews, S. A. 1931. Observations on pigment migration within the fish mel-

anophore. Jour. Exp. Zool., vol. 58, p. 471.
MiRSKY, A. E. 1936. Protein coagulation as a result of fertilization. Science,

vol. 84, p. 333.
MoTOMURA, I. 1935. Determination of the embryonic axis in the eggs of Amphibia

and Echinoderms. Sci. Reports of the Tohoku Imp. Univ., 4th series,

vol. 10, p. 212.
NoRRis, C. H. 1940. Elasticity studies on the Myxomycete, Physarum polyceph-

alum. Jour. Cell, and Comp. Physiol., vol. 16, p. 313.
Pease. D. C. 1940a. The effects of hydrostatic pressure upon the polar lobe and

cleavage pattern in the Chaetopterus egg. Biol. Bui., vol. 78, p. 103.
. 1940b. Hydrostatic pressure effects upon protoplasmic streaming in

Plasmodium. Jour. Cell, and Comp. Physiol., vol. 16, p. 361.
AND J. A. KiTCHiNG. 1939. The influence of hydrostatic pressure upon

ciliary frequency. Jour. Cell, and Comp. Physiol., vol. 14, p. 135.

and D. a. Marsland. 1939. The cleavage of Ascaris eggs under excep-

tionally high pressure. Jour. Cell, and Comp. Physiol., vol. 14 (supple

ment of No. 3), p. 1.
Regnard, p. 1884. Note relative a Taction des hautes pressions sur quelques

phenomenes vitaux (mouvement ces cils vibratiles, fermentation). C. R.

Soc. Biol., Paris, T. 36, p. 187.
. 1891. Recherches experimentales sur les conditions physiques de la

vie dans les eaux. Masson, Paris.
ScHECHTMAN, A. M. 1937. Localized cortical growth as the immediate cause of

cell division. Science, vol. 85, p. 222.
Seifriz, W. 1937. A theory of protoplasmic streaming. Science, vol. 86, p. 397.
Spaeth, R. A. 1916. Evidence proving the melanophore to be a disguised type

of smooth muscle cell. Jour. Exp. Zool., vol. 20, p. 193.

THE RELATION OF THE VISCOSITY CHANGES OF PROTO-
PLASM TO AMEBOID LOCOMOTION AND CELL DIVISION^

Warren H. Lewis
The Wistar Institute of Anatomy and Biology
The viscosity of protoplasm is often readily changeable from sol
to gel, gel to sol, and to various intermediate states by unknown
internal factors and a few known external ones. The contractile
tension which protoplasm exerts varies more or less with its vis-
cosity. The contractile tensions exerted by protoplasm play import-
ant roles in the activities of cells and organisms, some of which are
to be considered in the following pages.

AMEBOID LOCOMOTION

The essential aspects of ameboid locomotion have been presented
by Mast in his classical account of the structure and locomotion of
the ameba.

Contraction of the posterior part of the superficial gel layer or
plasmogel drives the less viscous endoplasm, plasmosol, forward
through the body of the cell or organism into the softened or weak-
ened anterior end and there produces advancing pseudopods. As
the posterior part of the gel layer contracts and shortens, some of it
solates or becomes less viscous, mixes with and is carried forward in
the endoplasm. At the anterior end, the endoplasm which reaches
the lateral walls of the pseudopod gels and gradually extends or
builds up the body forward as rapidly as it is torn down at the
posterior end. A substratum is needed to move on. This generalized
statement applies to the ameba, to the slime mold, to leucocytes, to
different types of normal and malignant cells that migrate in tissue
cultures, and with certain special modifications probably to all
types of organisms and cells which have not developed specialized
organs of locomotion such as cilia, flagella, undulatory membranes,
and muscles.

The superficial gel layer is the motor organ in the ameboid type
of locomotion. The contractile tension which protoplasm exerts

' Aided by a grant from the International Cancer Research Foundation.

[163]

164 The Structure of Protoplasm

when it is in the gel state is the key to this, and probably also to the
action of cilia, flagella, undulatory membranes, and muscles.

Ameboid locomotion thus depends (a) upon a definite organiza-
tion of the cell or organism, namely, a superficial gel or viscous
layer and a central less viscous or fluid endoplasm, (b) upon the
contractile tension which protoplasm (a colloid) automatically exerts
when it gels, (c) upon local changes of viscosity: gel layer into
fluid or semifluid endoplasm, and endoplasm into gel layer, (d)
upon the regulation and the polarization of the viscosity changes,
and (e) upon a substratum.

Ameboid type of locomotion depends upon a definite organiza-
tion of the cell or organism, namely, a superficial gel layer and a
central less viscous or fluid endoplasm. The necessity for this par-
ticular type of organization is evident if one considers any other hypo-
thetical type of organization. If gelation were central or immediately
about the central area and the nucleus and the fluid or less viscous
protoplasm superficial, ameboid locomotion would be impossible.
A uniform viscosity of the protoplasm from surface to center
throughout the organism or cell, whether in the sol, the gel, or a
semigel state would also preclude ameboid locomotion.

Other important aspects of this organization are (a) the relative
and the absolute thickness of the gel layer and the endoplasm and
(b) the relative and absolute viscosity of the two layers. The thick-
ness and viscosity undoubtedly vary from moment to moment in
active cells and protozoa, less rapidly in inactive ones. They also
vary with the type and metabolic condition of the cell and the
organism.

The factors which bring about this fundamental type of organ-
ization are unknown. Probably all cells and protozoa have either
permanently or temporarily this type of organization, even though
they do not migrate (plant cells and eggs, for example) or migrate
with the aid of special organs such as cilia, etc. All cells and
protozoa are derived from cells or protozoa that multiply by cleav-
age, and cleavage, as will be shown later, is dependent upon the gel
layer. Some types of cells become much modified and lose their
power to divide. Among them there may be some that lose this
general type of organization. Any theory that attempts to explain
the factors involved which produce this fundamental type of organ-
ization will probably be applicable to all cells and to protozoa.
Heilbrunn (1940) has a theory that the high viscosity of the cortex
(gel layer) depends on the presence of the calcium ion, and that the

Viscosity Chaiiges of Protoplasm 165

low viscosity of the interior of the hving cell is due to the absence
or scarcity of free calcium. Various other theories might be sug-
gested, such as increase in the acidity or of the COo at the surface.
Jacobs (1922, p. 29) states, "A short exposure to carbon dioxide of
various cells causes a decrease, and a longer exposure an increase in
the viscosity of the protoplasm. It is suggested that carbon dioxide
may be an important factor in producing many of the natural changes
in protoplasmic consistency which have hitherto been unexplained."

We come now to our second important factor in ameboid loco-
motion, namely, the contractile tension which protoplasm (a colloid)
automatically exerts when it gels. This is a phenomenon common
to living and many nonliving colloids such as blood plasma, gelatin,
and silica gel. Some colloids exert enough tension when they gel
to flake off glass from the wall of the containing vessel. Mast (1926,
1931) believes that the contractile tension is due to an elastic recoil.
He states (1926. p. 404) that "there is an abundance of evidence
indicating that the plasmagel and the plasmalemma are elastic and
that they are usually somewhat stretched and consequently exert
inward pressure; that is, that amebae are usually turgid" (1926,
p. 405) . "The fact that amebae are usually turgid indicates that the
plasmagel, and possibly also the plasmalemma, is semipermeable
and that the plasmasol, and probably also the substance in the
vacuoles in the plasmagel, is hypertonic. If this obtains, it is plain
that there would be an excess inflow of water and a stretching of
the plasmagel and the plasmalemma until their combined strength
equals the diffusion pressure." (1931, p. 330) "An ameba is a
turgid system; owing to this the plasmagel is continuously under
tension. The plasmagel is elastic and consequently is pushed out at
the region where its elastic strength is lowest. This results in the
formation of pseudopods."

My idea of the origin of the contractile tension of the gel layer
differs from that of Mast. He assumes that the plasmogel is elastic
and that it is stretched by the inflow of water into the plasmosol.
In this stretched condition it keeps the ameba turgid and drives the
plasmosol into the weakened areas to form pseudopods. My idea
is that protoplasm, like most colloids, automatically exerts contrac-
tile tension when it gels (Lewis, 1939) .

I was led to adopt this view because Mast's theory did not seem
to offer an adequate explanation for the occurrence of constriction
rings. For example (Lewis, 1939, p. 415), "The contorted mitoses
observed in the division of the spindle cells of the spindle-cell

166 The Structure of Protoplasvi

sarcoma C37 are explained by the development of changing con-
traction bands of the plasmagel layer which produce constrictions
that result in marked distortions and lobulations of the dividing and
young daughter cells." During anaphase, telophase, and early
daughter cell stages (Lewis, 1939, p. 409) , "Constrictions appear
and disappear in different regions, squeezing the cells into lobulated
and elongated forms." Constriction rings appear regularly on
migrating lymphocytes and neutrophiles, occasionally on fibroblasts
and sarcoma cells, and always in cell cleavage (Lewis, 1939) . It is
impossible to see how an elastic recoil theory can explain such con-
strictions. It is difficult to explain the retractions of cell processes
and pseudopodia on the elastic recoil theory or the retraction of gel
sheets and gel strands of the slime mold that are far away from
the streaming endoplasm, or the excessive contraction of an ameba
that squeezes out all the endoplasm through a break in the plasmogel
and plasmolemma.

Presumably the contractile tension varies with the viscosity and
with the thickness of the gel layer. In the fluid state, it is practically
nil. In the gel state it is measurable. The "balance-pressure" of
Kamiya (1940) is probably a measure of the contractile tension of
the gel layer of the slime mold.

What are the limits of the contractility of the gel layer? The
following quotations from Mast and Edwards seem to indicate that
the gel layer can contract until all the endoplasm is squeezed out of
an ameba through an artificial hole in the plasmolemma and plasmo-
gel. Nothing is left but a small mass of plasmogel covered by wrinkled
plasmolemma. In other words, the plasmogel can contract until it
meets plasmogel. This is of especial interest. In the first place, if
the contraction of the plasmogel is due to the elastic recoil after
stretching, then we must imagine that stretching began at the
instant plasmosol appeared, if there was ever such a time. This
rather counts against the elastic recoil idea. We have other evi-
dence of the great contractility of the gel layer, as for example on
the tail buds of the lymphocyte and the cleavage furrow of dividing
cells. Mast (1926, p. 404) states, "There is an abundance of evidence
indicating that the plasmagel and the plasmalemma are elastic and
they are usually somewhat stretched and consequently exert inward
pressure; that is, that amebae are usually turgid. If pressure is
brought to bear on the cover glass with an ameba under it, the
plasmagel and the plasmalemma break and the plasmasol flows out,
leaving nothing but a small mass of plasmagel covered with the

Viscosity Changes of ProtoplasTn 167

plasmalemma, which is much wrinkled. This usually occurs even
if the pressure on the cover glass is relieved as soon as the plasmasol
begins to flow out. Under these conditions, the continued flow
must be due to inward pressure produced by contraction in the
plasmagel or the plasmalemma. The fact that the plasmalemma is
much wrinkled after the plasmasol has flowed out, clearly indicates
that the pressure is due primarily to contraction in the plasmagel."
Edwards (1933, p. 11) found "If N/1 NaOH or KOH is applied at
any point on the surface of ameba there is an immediate rupture
of the ectoplasm at that point through which the endoplasm flows
until nearly all of it has passed into the surrounding medium.
Nothing remains except a few crystals which adhere to the inner
surface of the ectoplasmic sheath."

The next important consideration concerns the local changes of
viscosity, gel layer into fluid or semifluid endoplasm, and endoplasm
into gel layer. This can be followed with ease in the ameba and
especially so in the slime mold under the oil immersion lens. These
changes are continually going on in all active cells and ameboid
organisms. In the ameba and the slime mold the numerous gran-
ules in the protoplasm act as indicators of such changes by the
presence or absence of Brownian movement or by their flow in
the streaming endoplasm. In both the ameba and the slime mold the
granules are enclosed in minute vacuoles in which they show a
limited Brownian movement even when the vacuoles are embedded
in gelated protoplasm. In the slime mold one can watch viscosity
changes in exceedingly minute areas with the oil immersion lens.
When the protoplasmic matrix gels, the vacuoles come to rest, and
when it solates, they show Brownian movement before they join the
endoplasmic streaming. There is no doubt about the changes in
viscosity, but the factors which bring them about are unknown.
Probably the factors which are responsible for the production of
the outer gel layer and inner less viscous endoplasm are internal.

Local changes of external pressure may evidently produce local
changes of viscosity as indicated by the following quotation from
Mast (1931, p. 328), which applies to an ameba in a capillary tube.
"The following results were obtained: If the pressure against the
advancing end of an ameba in the tube is increased, the direction
of flow in the plasmasol immediately reverses, and the opposite end
begins to extend, usually in several different regions, a protuberance
with a hyaline cap forming in each region. These protuberances,
however, soon unite, resulting in the transformation of the posterior

168 The Structure of Protoplasm

end into an anterior end with a typical hyaline cap. If the pressure
against this new anterior end is now increased, the direction of
flow again immediately reverses and a hyaline cap again forms at
the opposite end. This can be repeated many times on the same
individual, provided some little time is allowed for recovery between
successive reversals in the direction of pressure."

Internal pressures may also play a role. As endoplasm is forced
forward into the softened anterior end of the slime mold, the internal
pressure there increases. It is possible that when the pressure
reaches a certain point, gelation may occur and increase until the
resulting increase in contractile tension increases the internal
pressure to a point where solation results.

We come now to the regulation a^id polarization of the viscosity
changes. It is quite evident that there must be some sort of a
regulation of these changes, otherwise progressive locomotion could
not occur except accidentally. The continued or periodic softening
at the anterior end and the greater contractile tension at the
posterior end are necessary for progression. Pseudopodia may be
thrust out and withdrawn from various regions of an ameba or a
cell without locomotion. Some sort of a polarization is needed. The
thickness and the viscosity of the gel layer, which determines its
contractile tension, are undoubtedly regulated, as are also the
amount and viscosity of the endoplasm. There are probably both
internal and external environmental factors involved.

Various authors have questioned the validity of the theory that
the contractile tension of the gel layer plays an essential part in
locomotion. The older ideas that surface tension plays a leading
role can be dismissed without comment. More recently Marsland
and Brown (1936) have questioned the contractile tension theory
as will be seen by the following quotation: (p. 177) "Streaming of
the plasmasol in a forming pseudopodium is immediately stopped by
sudden compression to about 250 atmospheres." (p. 175) "On the
hypothesis that the flow is caused by a contraction of the proximal
portion of the plasmagel tube, forcing the plasmasol distally into
the forming pseudopodium the cessation of movement might be
attributed to a loss of tension throughout the entire plasmagel tube
due to the liquefying action of the pressure. On this basis the flow-
ing of the plasmasol would cease when, as a result of the increased
fluidity of the plasmagel, the force exerted by the plasmagel tube
becomes too small to overcome the resistance to flow. It is question-
able, however, that a constriction in the plasmagel tube is a neces-

Viscosity Changes of Protoplasm 169

sary factor in the streaming of the plasmasol. If the ameba is to
be considered as a contractile sac filled with fluid, one would
expect that the movement of the fluid would always be accompanied
by an inward displacement of the contractile wall. This, however,
is not always the case. Specifically, in an ameba capped with an
oil drop, a steady fountain streaming is maintained for hours despite
the fact that no change occurs in the oval contour of the cell. In
this instance it is evident that the constriction of the plasmagel
wall is not essential for active streaming . . . An alternative sug-
gestion as regards the cause of streaming would be that the volume
changes inherent in the sol-gel transformations may generate the
flow." Their objection to Mast's view is based on the observation
that "a steady fountain streaming is maintained for hours despite
the fact that no change occurs in the oval contour of the cell." I
wonder if they observed what went on inside of the ameba as
carefully as Mast (1926) did. He occasionally noted monopodal
amebae that were unattached except at the tip of the posterior end,
so that locomotion was prevented, (p. 400) , "In such specimens
the movements of the plasmasol and the plasmagel are precisely the
same as they are in moving specimens, but in reference to points
in space, the plasmagel actively moves backward instead of being
at rest. This gives the appearance of typical fountain streaming."
(p. 401), "This seems to indicate that in these specimens the plas-
masol flowed through the elongated canal in the plasmagel, that
when it reached the anterior end it spread out in all directions,
deflected backward, came in contact with the anterior border of the
plasmagel tube and gelated here, extending this tube which moved
backward as rapidly as it was built up, sliding under the plasma-
lemma which everywhere remained stationary." Mast does not
explain further, so I venture to extend the explanation a little.
Continuous contraction of the posterior end of the gel layer pulls
the plasmagel wall backward. This posterior end is also continually
solating and thus keeping a balance between sol and gel. It
would be rather difficult to explain this backward pull on the basis
that the contractile tension of the gel layer was due to an elastic
recoil from stretching. Such a sessile ameba might very well con-
tinue fountain streaming without much if any distortion of the
posterior end such as Marsland and Brown record. Their suggestion
that "the cause of the streaming would be that the volume changes
inherent in the sol-gel transformations may generate the flow"
seems to me unwarranted, because the same factors are involved in

170 The Structure of Protoplasyn

fountain streaming as in locomotion and as Mast (1926, p. 407)
points out "could in no way be involved in the forward extension
of the pseudopod, resulting in locomotion." The cessation of stream-
ing of the plasmasol by pressure is thus probably due to the decrease
in the viscosity of the plasmagel.

LOCOMOTION OF AMEBA

The locomotion of Ameha proteus can be considered as the type
representative of ameboid locomotion. The classical papers of Mast
on the structure and locomotion of Ameha proteus contain the most
important account of this. They contain a detailed study of the
structure which is essential for any understanding of locomotion
and of the various factors involved. Every visible detail, including
those seen with the highest powers of the microscope, is important
as are also all the changes which occur from moment to moment
during a period long enough to cover a complete cycle. Even with
the most complete record of the visible structure and all the visible
changes it undergoes in locomotion or any other process, one comes
immediately to the invisible structures and processes which are still
more fundamental, but unfortunately they are still obscure.

Mast's account of the structure and locomotion of the Ameha
proteus is too well known for me to consider it in any great detail.
The moving ameba has a plasmogel layer surrounding a fluid plasmo-
sol. The gel layer is thick over the sides of the tubelike body, some-
what less thick at the posterior end, and thin over the anterior
pseudopod. The outer plasmolemma and the necessity of a sub-
stratum need not be considered. Contraction of the plasmogel at
the posterior end drives the plasmasol forward and extends the
pseudopod at the weak anterior end. As the posterior end of the gel
layer contracts, it shortens and loses material because part of it
keeps going into solution. This sol mixes with and is carried forward
in the plasmosol. The posterior end is thus gradually shortened.
At the anterior end some of the plasmosol goes to the lateral wall of
the pseudopod and there gels as it comes in contact with the
anterior end of the thick plasmogel tube wall. The tube wall is
built up anteriorly as rapidly as it solates posteriorly. The follow-
ing significant observation by Mast (1923) is readily explained by
the above considerations, but not by any of the other explanations
of ameboid locomotion, (p. 259) "If a given granule in the plasmosol
is continuously observed it can be seen to move forward in the
neighborhood of the central axis of the ameba until it reaches the

Viscosity Changes of Protoplasm 171

enlarged portion near the tip of a pseudopod and then to deflect
outward where it sooner or later comes in contact with the edge of
the opening in the sac of granular plasmagel where it is caught
in the gelation of the fluid in which it is suspended and becomes
stationary in reference to points outside, regardless as to whether
it is to the right or the left or above or below the plasmasol. As
more and more substance from the plasmasol is deposited at the
anterior border of the granular plasmagel the ameba moves forward
and the observed granule approaches the posterior end where it
sooner or later comes to the inner surface of the plasmagel. Here
the substance in which it embedded goes into solution carrying the
granule into the plasmasol in which it is again transported to the
anterior end where it again enters the plasmagel, etc." (See also
Fig. 7, p. 409, Mast, 1926.)

Many questions remain unanswered, such as the regulatory
mechanism, the factors involved in solution and gelation, and the
origin of the contractility of the gel layer.

LOCOMOTION OF CELLS

The migrating mammalian lymphocyte exhibits typical ameboid
locomotion. It has a thin gel layer, a central less viscous endoplasm,
a persistent posterior end with a hyaline tail bud, and an anterior
pseudopodal end. It has, relative to its size, a very large nucleus.
The lymphocyte is one of the smallest cells of the body. The thin
cortex is not always recognizable, and one infers from the behavior
of the lymphocyte during locomotion that this layer must be a
contracting gel. Each time a pseudopod appears, a constriction ring
develops at its base where it joins the body of the cell. As the
lymphocyte advances, the constriction ring remains stationary in
relation to a point outside the cell exactly as does a particular
granule embedded in the anterior part of the plasmogel of the
ameba. As the lymphocyte moves forward, the body of the cell
posterior to it gradually decreases in size. The large nucleus as it is
forced through the constriction ring is constricted and distorted by
the ring, and when it is finally pushed through the ring it regains
its original spherical form. It is evidently elastic. When all but the
little tail bud has passed through the ring, the latter contracts
down and increases the length of the tail which undergoes shorten-
ing between each such addition. One cannot see very much of what
goes on in the cell. The flowing forward of the few mitochondria
can sometimes be seen. There is very little endoplasm and only a

172 The Structure of Protoplasm

few granules in it. I have not been able to see either the solation
at the posterior end or the gelation that extends the wall of the
body forward, yet from the manner in which it migrates, it seems
as though exactly the same fundamental events must occur as in
the ameba (Lewis, 1933, 1939) .

The neutrophilic leucocytes are somewhat larger than the
lymphocytes and have the same sort of ameboid locomotion. The
anterior pseudopods are more changeable and the tail bud shorter.
A constriction ring develops at the base of the advancing pseudo-
pod which, like the one on the lymphocyte, does not move forward.
The endoplasm contains numerous granules which stream forward
into the advancing pseudopod. It is quite evident that contraction
and solation of the gel layer posterior to the constriction ring con-
tinues until all but the tail bud is carried forward in the endoplasm
and built up in front of the constriction ring. The granule-free
gel layer is very thin. It has not been possible to see either the
solation at the posterior end or the gelation anterior to the constric-
tion ring. The presence of the constriction ring, however, is the
key to the idea that they as well as other species of white blood
cells have the ameboid type of locomotion.

Both the lymphocyte and the neutrophile move much more rap-
idly than a fibroblast and appear to have a less viscous endoplasm
than the latter. Fibroblasts and sarcoma cells move very slowly.
One cannot detect in the course of a few minutes any forward
movement. Such cells, however, occasionally develop a constriction
ring which indents and distorts the nucleus. These rings, like fixed
granules of the ameba and the constriction rings of the lymphocyte
do not move forward, yet the nucleus is slowly pushed through
them in the course of four or five hours, presumably by contraction
of the posterior part of the cell which pushes the nucleus and
rather viscous endoplasm very slowly forward. It seems quite
probable that fibroblasts, and sarcoma cells which are derived from
them, have a very slow ameboid locomotion because the endoplasm
is almost as viscous as the gel layer.

Ruffle pseudopodia and undulating membrane common on white
blood cells, macrophages, connective tissue cells, and sarcoma cells
consist of exceedingly thin sheets of hyaline protoplasm. They often
extend into the fluid medium of cultures and are always slowly
bending back and forth like the ruffles on a dress in a slight breeze.
They probably have an outer thin gel layer and a central somewhat
less viscous endoplasm. Slight local changes in the viscosity of the

Viscosity Changes of Protoplasm 173

gel layers would change the contractile tension and produce the
wavy motion. The motion of cilia, fibrillae, and imdulatory mem-
branes in general can be fitted into the idea that slight changes in the
viscosity of gel layers or gels alter the contractile tension which they
automatically exert because they are in the gel state.

LOCOMOTION OF SLIME MOLD {Phijsarum Polycephalum)

In studying the slime mold one should distinguish, as Howard
(1931) suggests, between growth, spreading, and locomotion. Growth
is actual increase in the mass of the protoplasm and is presumably
accompanied by nuclear division. Spreading may or may not be
accompanied by growth and locomotion. Locomotion is always
accompanied by spreading at the anterior end.

The minute pieces with which I propose to deal are examples
of locomotion and spreading. Most of the pieces for my prepara-
tions were made from the tips of the plasmodia that had spread out
over the surface of the water in the Petri dish (100 X 15 mm.)
cultures during the preceding night or on the same day from the
elevated filter paper at the center on which there was a generous
supply of oatmeal. There were considerable differences in the
age of the cultures, in the growth, and in the spreading. There were
likewise great differences in the behavior of the small pieces in the
different preparations. The pieces (6-12) in each preparation were
always from the same tip and as a rule behaved in much the same
manner. I have not as yet attempted to investigate carefully the
relation of the condition of the culture and the type of preparation
obtained from it. There are such very great differences in the
behavior of the pieces that an investigation of their relationship to
the condition of the culture and to the part of the plasmodium from
which they are taken must yield some very interesting results.
The preparations were made by taking a few millimeters from the
spreading tip of the plasmodium with a pair of forceps, transferring
to a small Petri dish with water and tearing up into smaller pieces
3 to 4 millimeters long. One of these was transferred to a thin
spread-out drop of water on a clean cover glass and then teased or
cut up with the needles into pieces a millimeter or less in diameter.
The cover glasses were sealed in the usual hanging drop manner over
a hollow slide or ring.

Small pieces show at first no or almost no internal movements
of endoplasm, due either to complete or almost complete gelation as
a result of the mechanical stimulation of tearing with the needles.

174 The Structure of Protoplasm

The endoplasm evidently gels as the pieces are divided because
it does not flow out at the cut end. In the course of a few minutes
or so, small pseudopodia begin to extend out from the periphery
and show typical elongation and retraction with each outward and
inward flow of tiny streams of endoplasm. The pseudopodia seem to
be confined to the peripheral border of the piece that is in contact
with or close to the cover glass. The entire periphery of good, active
pieces, which are the ones we are considering here, may show such
pseudopodia. Contractions and relaxations of the central piece were
sometimes noted. As the pseudopodia extend out radially farther
and farther, the tips become multiple, frequently come in contact and
fuse with neighboring ones to form a plasmodial network and sheet
back of the advancing tips or pseudopodia. In the course of an hour
or so, when most of the little chunk at the center has gone out into
the thin plasmodial net and sheet, some of the advancing ends become
retracting ends and the piece begins to migrate. Most of the active
plasmolets ultimately arrive at the periphery of the thin drop.
Whether this is due to some chemotactic response (Coman, 1940) ,
or to chance trapping is unknown. During the course of a day the
pieces undergo many changes of form, and the small ones some-
times fuse to form larger ones and sometimes the larger ones divide.
The smallest simple active pieces that have begun to migrate
show in the course of an hour or two an organization that is much
like that of an ameba in its fundamental aspects. They consist of
a fluid endoplasm, the plasmosol and a tubular body with a thick
gel wall, the plasmogel, which continues into a simple thick-walled
posterior end and a broad fan-shaped transition zone, which termi-
nates in a lobulated crescentic anterior end with numerous pseudo-
podia. The locomotion of one of these small pieces of plasmodium
is in general very much like that of the ameba, except for the
rhythmic reversals of the endoplasmic flow, the complications of the
transition zone, and many little minor solations and gelations which
frequently give rise to many very small temporary streams of endo-
plasm which connect with channels and tubes. Contraction and par-
tial solation of the posterior part of the gel layer forces the endoplasm
forward through the tube and channels to the anterior end, where it
forces out pseudopodia at weak places and builds forward the Plas-
modium at their bases. After about thirty seconds, gelation sets in
at the anterior end; the resulting contractile tension there, together
with partial solation reverses the endoplasmic flow and expands the
posterior weakened end. After about thirty seconds, gelation at the

Viscosity Changes of Protoplasm 175

posterior end increases the contractile tension and reverses the flow
and repeats the cycle. There is a net gain at the anterior end and
a net loss at the posterior end with each cycle.

Most of the active pieces obtained by tearing are somewhat
larger and more complex and have one or more branching and
anastomosing thick-walled tubes, one or two or more posterior
retracting ends, one or more thick, lobulated, anterior, advancing
ends which project from a broad flat plasmodial sheet with thick and
thin areas that grade into the tubular region. The thicker areas of
the sheet are traversed by branching and anastomosing channels
which merge into the tubes posteriorly and into minute, more or less
temporary, terminal channels leading to each lobule and pseudopod
anteriorly. The same type of organization found in the small plas-
molets prevails also in the larger ones and undoubtedly also in the
relatively enormous masses of the stock cultures under approxi-
mately the same conditions where growth and locomotion are fairly
rapid. When examining an especially selected small plasmolet a
millimeter or so long in a preparation 2-3 hours old, it is convenient
to consider separately the organization or structure and the behavior
of the different regions, namely, the posterior end, the tube, the fan-
shaped transition zone, and the anterior end.

Posterior ends differ considerably; they may be blunt or drawn
out into long tenuous tapering threads. The usual blunt ones have
a gel wall continuous with the tubular part, but it is not, as a rule,
quite as thick as the latter. Sometimes there is an outer hyaline zone
and sometimes not, depending upon the condition of the plasmodium.
In the course of a few minutes the hyaline zone may completely dis-
appear and then reappear again. Sharp hyaline spicules such as
Camp pictures at the posterior end may come and go in the course
of a few minutes. The limits of what one might designate as the
posterior end are rather indefinite as its gel layer merges into the
gel wall of the tube and they function more or less as a unit. One
might limit it to the region where during contraction and forward
flow of the endoplasm some of the gel layer solates and mixes with
the endoplasm. The two most interesting points about the posterior
end are its contraction and the partial solation of its gel layer during
the contraction. One can see this local contraction of the posterior
end and often in addition to this a more general shortening of the
posterior part of the tube which pulls the posterior end forward
rather rapidly at the time of the forward flow of the endoplasm.
Solation at the posterior end occurs during the contractile phase only

176 The Structure of Protoplasyn

and begins after the contraction and forward flow of the endoplasm
has started.

Solation at the posterior end during the contractile phase is
accompanied by the flow of great numbers of granular vacuoles into
the endoplasm from the inner wall of the gel layer where they pre-
viously had been at rest. Sometimes one can also see many tiny
streamlets of granular vacuoles flowing, often in single file, through
most of the thickness of the gel layer into the endoplasm. It is
evident that solation of the hyaline gel matrix in which the granular
vacuoles are embedded occurs during contraction. This is an
interesting relationship and is in marked contrast to what occurs
at the posterior end when the endoplasmic streaming is reversed. The
backward flow of the endoplasm expands and lengthens the posterior
end, but it does not usually lengthen to the original length that it
had at the beginning of the cycle if the plasmolets are moving for-
ward. There is a net loss at the posterior end during each cycle.
As the posterior end lengthens and expands, some of the endoplasm
comes to rest on the inner surface of the wall and seems to gel in
situ. The expansion at the posterior end is presumably due to the
internal pressure of the turgid system on the somewhat weakened
gel layer resulting from excessive solation during contraction.
Expansion and gelation appear to go on at the posterior end until
the gel layer becomes thick or viscous enough to automatically
exert a sufficiently increased tension to contract and overcome the
internal pressure of the endoplasm and reverse the flow. This is
undoubtedly correlated with tension changes at the anterior end
and in the whole system of tubes and channels. It is quite possible
that the weakened condition at the posterior end is in part due
to a general decrease in the viscosity of the gel layer throughout
most or all of its thickness.

The tubes consist of a thick gel wall through which the endo-
plasm flows back and forth. They are relatively permanent, but
still only temporary, structures. The length of time they persist
depends upon how rapidly the plasmolet is moving and changing
form. By the time a plasmolet has moved forward a distance equal
to its length, it has been completely torn down and rebuilt anew from
end to end. The posterior end tubes are the oldest part of the Plas-
modium. The condition of the slime mold at the time the preparations
are made seems to make considerable difference in the general
appearance of the gel layer of the tube. In some preparations it is
very much lobulated and very uneven; in others it is rather smooth.

Viscosity Changes of Protoplasm 111

The diameters of tubes differ as do also the thickness of the gel walls
and the diameter of the endoplasm. There are probably some definite
correlations between the size of a tube and the size and rate of
increase in size of the anterior region of the plasmolet to which the
tube delivers endoplasm and the amount of endoplasm which streams
through it. Tubes undergo changes in size. The thickness of the gel
walls differs both absolutely and relatively as regards the diameters
of the tubes and the diameters of the endoplasm. There is probably
a definite correlation between the thickness of the gel wall, the
viscosity of the gel, the size of the tube, and the internal pressure
of the endoplasm. There can be no doubt about the gel state of the
tube wall. Camp found it tough when stuck with a needle. In the
large culture dishes numerous ends often stick up into the air.

The tubes usually contract, that is, decrease in diameter and
length, during each forward flow of the endoplasm, and expand dur-
ing each backward flow, but not to their previous posterior extension.
The tubes are built up anteriorly by transformation of channels in
the transition zone.

Sometimes the endoplasm flows back and forth for a short time
without much of any change in the thickness of the gel wall, but
within a few minutes the same tube may show solation along its
inner surface with each forward flow and gelation with each back-
ward flow. This is not always confined to the inner surface, for
minute streams in and out of much of the thickness of the wall can
be detected here and there. Sometimes projections on the inner
surface of the tube deflect the stream of endoplasm for varying
lengths of time. Such projection seems to be due to quite local and
temporary increases in viscosity and contractions. The general
behavior of the tube is thus similar to the posterior end and acts
with it as a unit.

Tubes frequently have an outer hyaline zone, but this often
comes and goes as at the posterior end. The hyaline border is usually
uneven and more marked on the plasmolets of some preparations
than others. Sometimes a tube with a rather thick hyaline zone will
in the course of a few minutes lose it entirely, or part of it, and
the granular protoplasm will extend to the very periphery. My
explanation of the origin of the hyaline border is the same as that
of the hyaline cap of Mast. The granular gel layer at times permits
clear granule-free endoplasm to be squeezed through its pores to
the surface under the interface membrane. Here it gels or partly
gels. With subsequent changes in the viscosity of the granular and

178 The Structure of Protoplasm

hyaline zones they become mixed and the granules reach the
periphery.

Tubes not only contract and relax, but they shorten and lengthen,
bend and straighten, and undergo various contortions which indi-
cate that there is some sort of a contractile mechanism. This is
undoubtedly the gel layer. Tubes usually show a considerable
amount of shortening during the forward flow and the posterior end
is carried forward rather rapidly. During the backward flow they
lengthen but do not regain their former length. Sometimes tubes
contract locally and completely stop the streaming of the endoplasm
through the constricted region for a short time. The endoplasm
usually then streams in both directions from the constriction. Such
tubes usually relax and the flow of the endoplasm is resumed through
the tube. Sometimes there is a permanent obliteration of the flow
and the tube later disappears. Other tubes undergo a more or less
permanent increase in size. Such changes and alterations are corre-
lated with shifting increases and decreases of the advancing ends.
I have noted, occasionally, small pieces a few millimeters in size, on
the bottom of a Petri dish partly filled with water, send out tube-like
elongations at various angles which slowly squirm like a bunch
of worms.

We come now to the transition zone which spreads out fanlike
from the anterior end of the tube to the broad crescentic lobulated
anterior end. It is a difficult region to understand and to present.
There is nothing in the ameba which corresponds to it. One should
bear in mind that the slime mold moves forward by a tearing down
at the posterior end and building up at the anterior end with the
protoplasm which is transported in the endoplasm from the posterior
end and other retracting lateral ends and areas. It is obvious that
the plasmolet cannot continue to advance or build up in all regions
of the crescentic periphery if it is to move forward. One sector
builds up more rapidly than adjoining ones. The latter sooner or
later retract by contraction and solation and transport of the proto-
plasm to the more rapidly spreading or advancing sector. Corres-
ponding with this, there is a progressive enlargement and building
forward of the channels leading to the advancing sector and a diminu-
tion and disappearance of those that supplied the adjoining and
retracting ones. During this transformation the broad fan-shaped
plasmodial sheet undergoes transformation. The thick anterior part
becomes thin and a new thick part is built up in front in the active
sector. The middle part of the sheet develops thin areas between the

Viscosity Changes of Protoplasm 179

channels. These thin areas, as Camp noted, may become fenestrated;
the protoplasm partly retracts and partly solates and becomes incor-
porated into the tubes, contributing endoplasm and material to the
gel wall. If one followed the fate of tiny terminal channels leading
to the pseudopod which is destined to form the center of the advanc-
ing sector, it would be noted that this channel without moving for-
ward would gradually enlarge until it became the main tube with the
thick gel wall.

The channels in the anterior part of the transition zone through
which the endoplasm flows are surrounded by the broad gel sheet
and do not have definite walls. In the posterior part where the
channels are larger there is a gradual transition into tubes that
have more and more definite walls like those on the tubes into which
they are continued. Probably all the changes which the transition
zone undergoes can be correlated with the building up of the
advancing sector and "demand" for endoplasm. Camp has described
many of the changes which took place in the broad continuous sheet
which I have designated as the transition zone without indicating
their relation to the building up of the advancing sector. Camp
states (p. 327) that the "Protoplasmic movement or locomotion in a
plasmodiuni is so similar to the same phenomenon in the rhizopods
that when the correct explanation is found for the one, it will prob-
ably hold true in most of its aspects for the other." Although he
quotes Mast at some length, he does not apply the principles of
ameboid locomotion to the plasmodium. If one applies these prin-
ciples, all the visible changes fall into line.

The broad crescentic anterior end is thick, and it is often difficult
to see exactly what happens from moment to moment. The anterior
ends are usually more or less lobulated. In some preparations all
the migrating plasmolets show marked lobulation, while in others
it is almost absent. During the forward flow of the endoplasm, the
tip of each lobule expands and extends forward into one or more
pseudopodia or protuberances described by Camp. In some prepar-
ations the forward flow of the granular endoplasm usually completely
fills the rather rapidly expanding pseudopod. In other preparations,
the expanding tips often become filled with granule-free hyaloplasm
(Figs. 1-4) . Camp suggests that the granules cannot enter because
the hyaline pseudopod is too thin. This may apply to some cases,
but many of those that I have observed are thick and rounded and
have ample thickness for the granules. At the base of such a
hyaline cap there is a slightly rounded, bulging layer of granular

180

The Structure of Protoplasm,

CO

03

y
W)

c

">
o

S CO
• I— I

oo

■^3 C\J

S-i
HI

>>

CO

0)

CO

CO

u

<1)

"cO

C

3 P.

>T3

T3

C
o

3
o

O 75

O 0)

M D

c c

■r„ CO

c

CO

O CO

ft u

0)^
en

o 3

c

0)

CO

72 O

OJ CO

C« S-c

CO

TS

c

CO

cu
o
c

CO
S CO

o a,

en

CO -M
"I [ G

S 8

^ a;
"o G

> s

53 CO

> -

0) -a

OJ -f^

£ CO

00 0)

.S G

^ -35

« m

■a ^

O *

§• ^

^ l*

^ >>

3 CO

0) r-H

CO 3

3 CD

5 S-I

CO !-.

S-I •"

x; CO

o S
ft

u 73

s: 01

ft s

■- ft

a; 0.

-M o

rt "-2

o o

•^ CO

CO +3

«

"i a;

o.

t« .s

i3 o

■*-' in

CO

o

a ■*

CO

c

01

03 £

si

W C"

o o

O, ft

0) 1)

CO CO

M f/^ ■*

o;^ u g d u

tH 0) fe fo fe fe

CO

Viscosity Changes of Protoplasvi 181

protoplasm which is evidently in the gel state. This holds back the
granular endoplasm which piles up behind it, but seems to permit,
as Mast has suggested for the ameba, the hyaloplasm to filter through
its pores to form the hyaline cap. This rather thin, granular, plas-
mogel membrane often has a smooth outline bordering the base of
the hyaline cap, and its granular vacuoles are at rest. This membrane
may persist, or a small local break or two may occur and small
streams of granules suddenly pour through them into the cap. Then
the entire membrane may give way and the cap rapidly fill with
granules and all traces of the gel layer disappear. The anterior end
does not possess a thick, tough gel layer such as one can see and
manipulate at the posterior end and tube region. There is much
more protoplasm at the anterior end and in the anterior part of the
broad transition zone than in the tube and posterior end, but it seems
to be less viscous. With each forward flow of the endoplasm pro-
duced by the contraction of the gel layer at the posterior end and
on the tube, this mushy protoplasm is pushed forward within the
bulging boundaries of a rather thin gel layer of the lobes. At the
weakest areas, the pseudopods and hyaline caps form and bulge out
still more. Forward flow is probably much impeded and slowed by
the friction in the relatively large number of small and minute ter-
minal channels. The forward flow and expansion cease with the
increase in the viscosity and partial gelation of the lobes and of the
pseudopods. One can detect these changes by the absence of the flow
and the absence of Brownian movement of the granular vacuoles.
This is probably not a very viscous gel, but there are many such
ends. It seems probable that height of gelation is synchronized with
a weakening of contractile tension at the posterior end so that the
contractile tension which now develops at the anterior ends as a
result of the increased viscosity and gelation reverses the endo-
plasmic flow.

During the anterior contraction period each pseudopod contracts
and shortens. They do not always contract simultaneously, and an
occasional new one may form. The amount of contraction usually
falls short of the previous gain so that there results a net forward
gain with each cycle. The contraction usually involves more than
the pseudopodal tip and there may be a considerable amount of
retraction of the whole lobe. During the contraction there is a
backward flow of the endoplasm. Small, more-or-less temporary
streams from neighboring tips join to form larger and larger chan-
nels which unite and join to form the tubes. A large part of the

182 The Structure of Protoplasm

slightly gelated protoplasm of the pseudopods solates and joins the
endoplasm. The endoplasm is also enhanced by solation along the
borders of these small streams. The solation and backward flow
continues until the tips are weakened and the contractile tension
becomes less than that which develops at the posterior end. The
flow is then reversed, and the cycle is repeated.

There are a number of important differences between the anterior
end and the posterior end. The small plasmolets with a single pos-
terior end and one main tube about 100 microns in diameter may
have a dozen lobes at the anterior end with one to four pseudopodia
on each one. The main tube branches and rebranches in the broad
zone into about forty more or less temporary little terminal chan-
nels which supply approximately a corresponding number of pseudo-
podia. Some lobes enlarge, subdivide, and spread more than others.
The channels leading into such lobes increase in size as do also the
number of branches and the number of terminal channels. Some
lobes decrease in size and become retracting ends, and a reverse
series of events occurs until finally through contraction and solation
all traces of the lobe are lost. The whole response of the branching
system of channels and terminals to enlarging areas reminds one of
the growth of capillaries and the formation of veins and arteries in
growing regions of the vertebrate embryo. A fantastic comparison
can be made of the functional response of the tubes and channels of
the slime mold with the developing vascular system of man.

MICROSCOPIC STRUCTURE OF SLIME MOLD

The visible microscopic structure of the slime mold seems to be
somewhat similar to that of the ameba as described by Mast (1926) .
The thin areas of the transition zone can be examined with the oil
immersion lens, and Camp (1937) took advantage of them for his
observations on the minute structure of the protoplasm. His descrip-
tion which I now quote can be readily confirmed (p. 316) : ''The
protoplasm. Careful microscopic examination of very thin areas of
a plasmodial sheet reveals the presence of an optically homogeneous
substance which has an exceedingly faint bluish-gray color. This
is the substance which is generally recognized and referred to as
hyaloplasm. Scattered throughout the hyaloplasm there are numer-
ous granules, vacuoles, and particles of ingested material. The
granular bodies exhibit widely varying degrees of optical differen-
tiation, and they vary in size from approximately 0.2 microns to
slightly more than 1 micron. In general they are more or less

Viscosity Changes of Protoplasm 183

spherical, but some, especially the larger ones, may be irregular
in shape. The smallest granules are rather highly refractive; they
appear to be similar to those which Mast (1925 and 1926) observed
in Amoeba and designated as 'alpha' granules, and they are similar
to the microsomes described by Chambers (1924) and which he
believes to be always present in protoplasm. Many of the granules
are contained in very small vesicles which bear close resemblance
to vacuoles."

"The vesicles do not possess well-defined boundaries, and in
addition to containing one or more granules they seem to be filled
with a substance which is more completely hyaline and less viscous
than the surrounding hyaloplasm. That the material of these vesicles
is in a liquid state is indicated by the fact that these granules which
it surrounds are always in marked Brownian movement and can
very clearly be seen to move freely from place to place in the
vesicles." Granules not in vacuoles may at times exhibit very
slight Brownian movement.

To this account, I might add a few observations. Minute areas
of solation frequently occur in the thin areas, and one can see the
little vacuoles with their contained granules move out into the flow
of newly formed endoplasm. When endoplasm gels again in such
areas, the vacuoles can be noted as they come to rest. This indicates
that they either have or acquire a definite wall when the hyaloplasm
surrounding them solates. It thus seems quite probable that even
in the streaming endoplasm of the large channels the vacuoles persist.
One can sometimes detect them when the endoplasm comes to rest.
Some of the granules noted by Camp were not in vacuoles. These
I have also noted in great numbers. They are considerably smaller
than the vacuoles, are grayish in color, and are best seen in areas
that are too thin to accommodate the vacuoles. They stain with
Janus green and are mitochondria. They occur all through the gel
layer and endoplasm. The contractile vacuoles also noted by Camp
and others are often quite numerous in the neighborhood of the
thin areas and throughout the gel layer. They, like the granule-con-
taining vacuoles, are freed and carried into the endoplasmic stream
when the surrounding hyaloplasm solates. They also either have
or acquire a definite wall.

Camp does not mention the nuclei which are present in great
numbers. They are small, round, and have a single, rather large,
gray nucleolus. They are very difficult to see, because the nucleo-
plasm is optically similar to the gelated hyaloplasm, but once one

184 The Structure of Protoplasm

has learned to recognize them they can be detected in the neighbor-
hood of every thin area as grayish areas surrounded by the small
granule-containing vacuoles, with a darker gray nucleolar center.
They are much larger than the granule-containing vacuoles. A flake
of iodine placed under "the cover glass soon reveals great numbers
of nuclei and also a diffuse port-wine color in the protoplasm indi-
cating the presence of glycogen.

VISCOSITY OF SLIME MOLD

The following quotation from Camp expresses in general terms
rather obvious conclusions that one would reach from observations
and testings with a needle: (p. 372), "All of those considerations
would seem to lead to the conclusion that the viscosity of the
protoplasm of a plasmodium varies from place to place in the Plas-
modium, that it changes with changing external and internal condi-
tions, and that alterations in the physical state, i. e., changes from
sol to gel and gel to sol, take place more or less constantly through-
out all parts of a plasmodium. Normally, therefore, the protoplasm
of Plasmodia does not have an absolute and unchanging viscosity,
and variations in its viscosity seem to be due to alterations in the
physical state of the hyaloplasm."

Perhaps the presence of great numbers of small vacuoles, mito-
chondria, and nuclei and the relatively small amount of hyaloplasm
may have something to do with the apparent ease with which the
protoplasm or hyaloplasm changes from sol to gel and gel to sol.

CELL DIVISION— MITOSIS

Cell division consists of a series of interdependent reactions
which lead to a series of overlapping events, each of which takes
a considerable period of time. Most, if not all, of the visible reactions
are due to viscosity changes. The reactions lead to two essential
events, the formation of the daughter nuclei and the cleavage of
the cell into two parts.

Mitotic division of the nucleus is not always accompanied by cell
cleavage so that perhaps none of the events which result in cell
cleavage are necessary for mitotic division of the nucleus. The
events, however, which lead to cell cleavage or at least some of
them are probably dependent upon events which result in the mitotic
division of the nucleus.

A brief outline of the events as exhibited by an ordinary living
fibroblast in tissue culture is as follows:

Viscosity Changes of Protoplasm 185

The first-known event is the migration of part of the centrosome
material to the opposite pole. This has not been seen or reported
for Uving fibroblasts, but undoubtedly occurs some time before
prophase.
Events leading to the formation of the daughter nuclei.

The only visible structures in a resting nucleus are one to several
nucleoli, a homogeneous nucleoplasm, and a thin nuclear membrane.
In early prophase fine granules appear in the nucleoplasm. They
seem to become larger and larger and are recognizable as chromo-
somes toward mid-prophase. They ultimately almost fill the nucleus
except for a small amount of nuclear sap between them. During
the change from late prophase to metaphase the chromosomes
become more distinct and probably smaller and occupy less space
in what has now become the spindle than they did in the prophase
nucleus. The condition of the chromosomes in the resting nucleus
is obscure because they are invisible. It may be that they are
swollen and of low viscosity, and occupy the entire nucleus. Gelation
and then contraction of the chromosomes with a loss of fluid might
make them visible and also account for their small size as compared
with the relatively large amount of spindle material in metaphase
stage.

The chromosomes split and move to the poles to form the two
small compact masses and leave behind a large amount of exnuclear
sap which extends across the midline from one mass to the other.
These nuclear masses soon begin to increase in size as clear areas
appear and increase in size and number. As the nucleus enlarges,
the visible chromosome material gradually becomes more or less
dispersed as granules. Most of them disappear, and the few that
remain unite to form one to several nucleoli. It may be that there
is a reversal of the prophase process in that the chromosomes
take up fluid, swell and solate, and become invisible except for the
nucleolar part.

The nuclear membrane seems to disappear in late prophase just
before the chromosomes begin to move and does not become visible
until after the daughter nuclei begin to increase in size. It is very
thin and scarcely recognizable except as the sharp border of the
nucleus. It is probably in the gel state. Its fate is unknown. Does
it solate and disappear, or does it contribute to the spindle?

The chromosomes suddenly begin to move in late prophase. The
movements are scarcely perceptible to the eye, but in motion pictures
where events are speeded up they become quite pronounced. They

f

'^•^- ^

4

I

>

I

-*

I

* / ■

4*

in

i

9

i

}

m

"me'

\

00

»"?/

• *

•

^t s

' f

* ;

■ r ■ *».

«•

/ *<»

,•

v»

•x

''

'A •

'.*"."; ■

«.*v

'k'

<r^®

,^''

■5

■ •» •

• *

.J

■J .

5

^jMiiiijH

*-?.»*.

flH

■■w*,**

IF

%

V*

1

-•'1

•

m

1*

i

^

• •

r*

M
§

CO

0)

2

u

C
0)

^

(U

O

s

a

CS

m

CS

p^

^

CO

<1>

J3

OJ

<u

£

-<-'

+J

C

trt

X

x:

01

-<->

0)

T3

'%

r-"

1)

5

CO

CO
a

73
O

.2 S

1-7-! ■"

o
■a!

6
o

^ .i:

0)

O

o u

^

o

+j

^ «*-<

"„ o

>)

in

'^ tn

i3

dJ

o

CJ

o
a

t ^

CO
CO

a

5 °

CO

3 -^

0)

a

CO

■^ 0)

>>

Qj ji;

T)

a>

3 ^

S

g

.2 s

'3

c
o

CO 0)

1

. ,a

CM

^

O X!

CO t

(-1 m

.3 a;

o

Mt-i o

3

73^

(D O
O ^

.a a
3 £

M O

CO

2 6

■" CO
0, ^

.r: a

i3 o

CO "=

ri

>, "a

from a 1-d
■etraction of
ous dark fat

g telophase. (

ent inward b

later. Mid-

CO <u

Si M
0)

1 — 1 +J

3 2

u 3

brob

imple

CO c

OJ e

CO .S

x:

l5

a ^

O Sh
r-1 P

SH

t" "S

W) o

<ii

O 0)

OS _fo

*- V

CO

0) CO

o

j2

a

O

0) Ct3 -C

S

- s

W CO

«H

C/J

o

O S

*" 0)

2 s

■d

as

. >

a ^

fi
H

00^

&E

£H

1

w

o

lO ,»

o .^

■^ m

tH «3

t^ '

00

t/j

as
fa ^

CD C '"

- ci S 6 «1

ci

M

X

CO ^

<p

,r; -w

I

Viscosity Chaiiges of Protoplasm 187

are not Brownian movements. Since the movements continue until
after the chromosomes become arranged in the metaphase plate
as oscillations, they are probably produced by the same mechanism
that is responsible for the oscillation. Perhaps invisible fibers, or
spindle threads, form and gel sufficiently during late prophase to
exert tension on the chromosomes to which they are attached. Varia-
tions in the viscosity and contractile tension of these gel-like threads
which extend one from each chromosome to each pole would move
the chromosomes into the median plate, and there produce the
oscillations which move them individually in short paths in and out
of the median plate toward the poles. There are no visible bands
or fibers, but the fact that the chromosomes move individually a
little in and out of the metaphase plate would seem to indicate that
some sort of gel material exists and is also responsible for pulling the
chromosomes to the poles.

The nucleoli split up into chromosome-like material almost
immediately after the chromosomes begin to move. The splitting is
probably in some way connected with the movements of the chromo-
somes to which the compound nucleoli are presumably connected.
Eve.7its leading to cell cleavage. (Figs. 5-11.)

At the beginning of prophase the cell processes and pseudopodia
begin to retract, and the cell tends to assume a spherical form. The
withdrawal of the processes is often incomplete. Withdrawal of
pseudopodia is a common phenomenon and is probably due to an
increase in the viscosity of the gel layer, which results in an increase
of its contractile tension and a shortening or complete withdrawal
of the pseudopod. It is similar to the changes which go on at the
posterior end of the ameba or slime mold where contraction and
solation go hand-in-hand.

The tendency of the cell to assume a spherical form may be due
to an increase in the viscosity or thickness of the superficial gel
layer until it becomes fairly uniform over the body of the cell and
exerts a uniform contractile tension in all directions. There is con-
siderable evidence, as will be noted later, that the gel layer of eggs
increases in viscosity and thickness and becomes more rigid and
stable before and during cleavage.

The metaphase cell is approximately bilateral, with the chromo-
somes in the equatorial plane. The spindle and poles are surrounded
by a considerable zone of endoplasm with mitrochondria, fat
globules, and granules. The mitochondria are only approximately
bilaterally distributed. The fat globules vary in number. They may

i

s

*10

##

v

^^^

«.%M^^»

??.

11

Viscosity Changes of Protoplasm 189

all be at or near one pole, or irregularly distributed across the
equatorial plane. The endoplasm is rather dense in consistency, for
there is no Brownian movement or streaming of its contained gran-
ules. The superficial gel layer is relatively thin and hyaline in
appearance.

As the chromosomes move to the poles in anaphase, they converge
to form compact daughter groups and leave behind a homogeneous
granule-free interchromosomal material which is, or becomes, as
will be seen later, less viscous than the endoplasm. It evidently
consists of nuclear or spindle sap left behind as the chromosomes
move to the poles. It is sharply distinct from the surrounding endo-
plasm. As the cell flattens in the equatorial region during anaphase,
this interchromosomal material is squeezed from a blunt spindle
shape to a cylindrical shape by the contractile tension of the gel
layer. It probably plays an entirely passive role in this. It does not
mix with the endoplasm at this stage.

The flattening of the cell is probably due to the contraction of a
broad equatorial belt of the gel layer which has become more
viscous or thicker than the rest of the gel layer. A decrease in the
viscosity of the gel layer at the poles might also account for the
elongation. It is possible that both factors are at play. There is
scarcely any displacement of the endoplasmic granules and no
apparent diminution of its thickness in the equatorial region during
the flattening. This indicates that the endoplasm is rather viscous.

About the time, or shortly after, the chromosomes reach the
poles and after polar elongation of the cell has occurred, the constric-
tion groove, or furrow, appears in the equatorial plane. There are
some indications that the gel layer is somewhat thickened in the
region of the furrow. As the furrow deepens, it bends the endoplasm
inward and the deep surface of the latter projects into the soft inter-
chromosomal material. The endoplasm often has a row of fat globules
extending across the mid-plane; they are bent inward and finally
separated into two groups, one for each daughter cell when the
endoplasm is later divided. With the deepening of the furrow, the
endoplasm encroaches more and more on the interchromosomal
material and finally divides it into two parts.

Figs. 9-11. Cleavage of an adult rat fibroblast from a 2-day culture. X 1.100 dia.

Fig. 9. End of anaphase, cell elongated, homogeneous exnuclear sap elongated,
endoplasm with many small fat globules.

Fig. 10. Four minutes later. Mid-telophase, endoplasm with fat globules bent in-
ward, exnuclear sap about divided.

Fig. 11. Three minutes later. End of telophase. Connecting stalk.

190 The Structure of Protoplasm

Endoplasm then meets endoplasm in a broad equatorial area.
With the continued deepening of the furrow, by the contraction of
the gel layer at the bottom of the groove, the endoplasm is divided
and each half is incorporated into the endoplasmic layer of the
daughter cells. At the bottom of the groove, gel layer then meets gel
layer to form a stalk or band which connects the two daughter cells.
This band usually persists until the cells migrate away from one
another and stretch it to a thinner and thinner strand which finally
breaks. Endoplasm and interchromosomal material mix after the
former is divided and included in the daughter cells.

The observable steps in cleavage indicate that the interchromo-
somal material is less viscous than the endoplasm and the latter
less viscous than the gel layer or gel band which contracts. The
constriction band is presumably either a thickened or more viscous
band of the gel layer, which develops around the equator after the
cell has become bilateral. The cell becomes bilateral, or approxi-
mately so, when the chromosomes become arranged in the median
plane. It has already been noted that cells flatten in the equatorial
region as the chromosomes begin to move toward the poles and that
the constriction furrow begins after they have reached the pole. The
factors which produce the constriction band are unknown, but it
seems probable that the new bilateral set-up may have something
to do with it. Some metabolic products of chromosomes, centrosomes,
and interchromosomal material may be involved.

During cleavage of the fibroblast there are no indications of any
cytoplasmic currents or flow. As the endoplasm is bent inward by
the contraction of the gel band, fat globules in the neighborhood of
and at the bottom of the furrow are bent in; those at a distance are
scarcely disturbed at all.

My theory of cell cleavage is based on the idea that gels auto-
matically exert contractile tension and that the contractile tension
of the gel layer is increased in the median equatorial plane by an
increase of its thickness or of its viscosity, or of both in this region.
This increase automatically leads to the contraction which pinches
the cell in two. The existence of a superficial gel layer seems to be
a fact. The automatic contractility, which it exerts because it is in
the gel state, has already been considered. That the contractile
tension varies with the thickness and with the viscosity seems like
a probable assumption. Is there any direct evidence for the develop-
ment of an equatorial band? The gel layer of fibroblasts, on which
most of my observations are based, is rather thin, but one can some-

Viscosity Changes of Protoplasm 191

times see a slight thickening in the equatorial plane. When it con-
tracts and meets in the center to form the stalk which connects for
a time the two daughter cells, it exhibits marked thickness. Since the
theory is relatively new, no particular attention has been given to
this region with the idea that there might be such a thickening of
the gel layer which then automatically contracts and pinches the
cell in two.

Schechtman's (1937) observations on the cleavage of the eggs
of Triturus torosiis, according to my interpretation of them, not his,
lend support to the idea that cleavage is due to the contraction of a
thickened equatorial band of the gel layer. His observations and
interpretations are here quoted verbatim. They are followed by my
interpretation and comments (p. 222) : "Cleavage is indicated by
a contraction of the egg cortex at the site of the future furrow. This
is a contraction in the sense that the cortex becomes thickened and
bulges toward the interior. At the same time the surface of the
egg is displaced toward the site of the thickening." This is exactly
what one would expect when the thickened equatorial band of the
gel layer begins to contract. "The mid-portion of the contracted
cortex begins to expand within one to two minutes after the above
contraction (at temperatures ranging from 22° to 26° C). The
pigment of this expanding portion is rearranged in irregular streaky
lines, plainly indicating that the cytoplasm is being stretched. The
surface of the stretched material sinks below the general egg sur-
face much as does the surface of a fluid material stretched between
relatively firm supports." The continued contraction of the band
at the bottom of the furrow has stretched the adjoining gel layer,
especially its superficial pigmented surface, and thus the pigment is
pulled into streaky lines which extend into the furrow. The com-
parison with surface of fluid material is superfluous, since this
involves a gel layer only, not a fluid one. "Chambers gives other
evidence that this zone is liquid in his observation of Brownian
motion and his micro-dissection experiments." Chambers' evidence
that this zone is liquid does not hold for Triturus (Schechtman)
and Arbacia (Marsland) . "The stretched cortical material ('the
primary furrow') has a lower concentration of pigment per unit
surface and therefore, appears lighter than the rest of the upper
hemisphere." The cortical material stretched by the contraction of
the equatorial band naturally has a lower concentration of pigment
per unit area than it had before. "A secondary furrow appears at
about the center of the primary furrow. It gives evidence of addi-

192 The Structure of Protoplasm

tional stretching of its materials." The continued contraction of the
band produces the secondary furrow. "The pigmented cortex bound-
ing the hghtly pigmented 'primary furrow' becomes the site of
intense growth directed toward the egg interior." There is no
growth. The continued contraction of the band deepens the furrow
and pulls on the cortex adjoining the 'primary furrow', indicating
that the cortical layer is in the gel and not in the fluid state. "Vitally-
stained marks placed in this position are drawn out into long delicate
hair-lines as the furrow deepens." This is just what one would
expect if the cortex is stretched or pulled down into the furrow.

"The streaming of the peripheral cytoplasm from the sides of the
egg into the furrow, which has been described by a number of
persons, is noticeably absent in the cortex. The streaming observed
was in all probability a subcortical movement only, as has been
suggested by Motomura also. This is supported by the recent work
of Motomura, and of Brown, as well as of my own work, which shows
that the cortex (of the Strongylocentrotus, Arbacia, and Triturus
eggs, respectively) is a more or less rigid layer during cell division."
I suspect that the "streaming of the peripheral cytoplasm" was
invented to explain how the pigment granules moved toward the
equator and down into the walls of the furrow. This flow, or surface
current, as it has frequently been called, seemed to demand that the
cortex change from the gel to the liquid state. Thus error leads to
error. This flow, as already noted, is not a flow, but is a stretching
and pulling down of the gel layer into the furrow.

"A reasonable explanation of the observed cortical growth, in
view of its localized character, is that the increase may be due to an
inhibition process, which of course, does not imply that the cortex
becomes 'fluid'. Indeed, there are indications that the cortex of the
furrow, excepting a small part near its tip, does not differ much in
viscosity from the rest of the egg cortex. At least, the difference is
not great enough to give rise to the characteristic surface contours
of fluids in contact with relatively solid materials. It is well known
that the swelling pressure of bio-colloids may attain high values
under proper conditions, and the relatively fluid material which
Chambers ascertained in the equatorial region of the cleaving eggs
would offer little resistance. It is also possible that the growth of
the furrow cortex is by intersusception of clear cytoplasm of sub-
cortical origin; this might explain the part played by the cytoplasm
which some workers have seen streaming toward the furrow. Both
processes might be involved, since they are obviously not antago-

Viscosity Changes of Protoplasvi 193

nistic." The moment one assumes that the advance of the furrow
is due to cortical growth by an "inhibition process" or to a flow,
insurmountable difficulties are encountered in explaining the streaks
of the granules and of the dye materials. Schechtman recognized
that "the mechanism of cell division has certain features in common
with the sol-gel transformations generally regarded as important in
the formation of pseudopodia in certain amebae", but his interpre-
tation of the important facts which he contributed differ, as already
noted, from mine.

The material and the type of experiment employed by Schecht-
man would seem to be unusually favorable for an analysis of cell
cleavage. These experiments should be repeated again and again
with a continuous series of marks across the equatorial line. One
would expect that marks exactly on this line would be carried in the
constriction band to the very bottom of the furrow.

The following quotations from Marsland (1939) on "The Mechan-
ism of Cell Division" indicate that he recognized that an equatorial
band played an important role in cell cleavage (p. 21) : "At the time
of cleavage the superficial protoplasm of the egg, especially in the
region of the incipient furrow, becomes firmly set and displays the
properties of a rigid gel." The hydrostatic pressure effect upon
dividing egg cells (Arbacia) led him to the following conclusion
(p. 22) : "These results indicate that a gelation reaction by which
the plasmagel in and near the walls of the furrows undergoes aug-
mentation at the expense of the subjacent plasmasol may account
for the formation and progress of the furrow. According to this view,
the intrusion of the furrow is a process closely analogous to the
extrusion of a pseudopodium." The following quotations from his
discussion are significant (p. 20) : "These experiments give strong
support to the view that a gelation of the cortex of the egg, especially
in the equatorial region where the furrow will form, plays an essential
role in the mechanism of cell division." "Without question, just
before the appearance of the furrow, this part of the protoplasm
becomes firmly set and assumes a rigidity which is several times
greater than before." (p. 21) : "The view that a plasmagel girdle
in the equator of the cell undergoes augmentation from the sub-
jacent plasmasol on either side bordering the incipient furrow, and
thus, is pushed inward toward the equator is further supported by
the present experiments." (p. 21) : "Although it is evident that the
plasmagel over the entire surface of the egg undergoes a setting
process just prior to and during cleavage, some evidence has

194 The Structure of Protoplasm

appeared in the present experiments, which indicates that a maximum
rigidity is present in the equatorial region." "The role of the asters
in the cleavage phenomenon remains for consideration. If the
process which pushes the rigid walls of the furrow inward represents
the interpolation of plasmagel which derives from the subjacent
plasmasol, it is possible that the astral rays represent the lines of
flow whereby this translocation of materials is accomplished."

If Marsland had used the words "pulled" and "pulls" instead of
"pushed" and "pushes", his interpretations would agree perfectly
with the idea that the equatorial girdle exerts contractile tension.
The question as to how much the stretched plasmogel layer of the
furrow is augmented by gelation of the subjacent plasmosol is a
difficult one to answer. Schechtman's experiments show that the
plasmogel is stretched, that is, the superficial part containing pigment
granules and dye spots. It would seem probable, however, that there
is some augmentation by gelation of the subjacent plasmosol. There
is no necessity for assuming that the plasmosol adjacent to the
furrow is augmented by a translocation of materials along the astral
rays, unless the plasmogel consists of a different sort of protoplasm
than that of the adjacent plasmosol, and there is no particular
evidence for this.

DIVISION OF AMEBA

Chalkley in his paper on "The Mechanism of Cytoplasmic
Fission in Amoeba Proteus" states that "general cell locomotion and
cytoplasmic division may be intimately related as to mechanism."
With this I entirely agree. He also states that "with the approach
of the daughter nuclei to the surface, the elastic strength of the
plasmagel layer will be sharply lessened in their immediate vicinity.
The greater elastic strength of the plasmagel at the equator will
result in the gradual constriction in this region. This will press
the plasmasol into these weakened areas, and polar pseudopodia will
protrude. With their attachment to the substrate, locomotion will
be initiated, and the cell will elongate. The streaming of the cyto-
plasm will carry the nuclei outward, and their continued activity
will maintain opposing polar regions of low tension in the plasmagel.
The locomotion mechanism posited by Mast will thus be brought into
play and maintained in opposite directions, and will result in the
continual narrowing and stretching of the equator of the cell by
traction of the daughter cells, until the cell is parted, and fission
is complete."

Viscosity Changes of Protoplasm 195

If the elastic tension of the gel layer is due to a previous stretch-
ing by the osmotic pressure of the endoplasm, how can it contract
in the equatorial region until it meets there to form the connecting
stalk? Obviously, the gel layer could never have been stretched
that much in this region, so the elastic tension or recoil cannot account
for the constriction. The automatic contraction of a thickened or
of a more viscous equatorial band of the gel layer is probably respon-
sible for the cleavage of the ameba, as it is for the cleavage of cells.
The stretching and final rupture of the connecting gel stalk is
probably due to the ameboid movements of the daughter amebae in
opposite directions and in part, to the contraction and solation of
the stalk.

BLEB FORMATION

Most fibroblasts and sarcoma cells show bleb formation during
late anaphase and telophase (Fig. 7) . This has been noted by various
authors. No adequate explanation of this common, but rather
startling, phenomenon has been advanced. Perhaps they are analo-
gous to the hyaline caps on the pseudopods of the ameba (Mast)
and of the slime mold (Figs. 1-4) .

Blebs arise rather suddenly during anaphase and telophase.
They are usually free of granules. Occasionally, a stream of endo-
plasmic granules pours in. They evidently consist of clear fluid,
which gathers under the interface membrane or a new membrane
which enlarges as the blelDS enlarge. The fluid does not appear to
escape to the outside. It seems quite probable that increased pres-
sure, exerted in the equatorial region by the contraction of the
equatorial band of the gel layer during cleavage, forces endoplasmic
fluid through the pores or interstices of local weak places in the gel
layer and that these pores are too small to permit granules to pass.

Mast (1926, p. 420) states that "when a pseudopod is formed,
the inner layer of the plasmagel liquefies locally, and the remaining
portion stretches and bulges out, and at the same time becomes thin
and porous. Liquid from the plasmasol passes through this porous
layer of the plasmagel and collects under the plasmalemma forming
a hyaline cap." The pores are too small to permit the granules to
pass through. As already noted, almost precisely the same thing
frequently occurs at the advancing ends of the slime mold. The
hyaline caps are enormous as compared with the blebs of a fibro-
blast (Figs. 1 and 7), and consequently, offer better material for
observations. When the contractile pressure at the posterior end

196 The Structure of Protoplasm

drives the fluid endoplasm forward, it pours into the advancing buds
and stretches the thin gel layer. Suddenly clear, granule-free hyaline
fluid appears beyond the granular layer and expands. It does not
escape to the outside but is evidently enclosed by a thin invisible
interface protoplasmic membrane, and probably also by a new thin
gel layer.

BIBLIOGRAPHY

Brown, D. E. S. 1934. The pressure coefficient of "viscosity" in the eggs of
Arbacia punctulata. Jour. Cell, and Comp. Physiol. 5:335-346.

AND D. A. Marsland. 1936. The viscosity of Amoeba at high hydro-
static pressure. Jour. Cell, and Comp. Physiol. 8:159-165.

Camp, W. G. 1937. The structure and activities of myxomycete plasmodia. Bui.
Torrey Bot. Club 64:307-335.

Chalkley, H. W. 1935. The mechanism of Cytoplasmic fission in Amoeba proteus.
Protoplasma 24:607-621.

Chambers, R. 1919. Changes in protoplasmic consistency and their relation to
cell division. Jour. Gen. Phys. 2:49-68.

. 1924. General Cytology. Ed. E. V. Cowdry. pp. 294-295.

. 1938. Structural and kinetic aspects of cell division. Jour. Cell, and

Comp. Physiol. i2: 149-165.

1938. Structural aspects of cell division. Arch. exp. Zellforsch. 22:

252-256.

COMAN, D. R. 1940. Additional observations on positive and negative chemotaxis.
Arch. Path. 29:220-228.

Edwards, J. S. 1923. The effect of chemicals on locomotion in Ameba. I. Re-
action to localized stimulation. Jour. Exp. Zool. 38:1-44.

Heilbrunn, L. V. 1940. Protoplasm, and colloids. Publ. 14, Amer. Assoc. Adv.
Sci., pp. 188-198.

Howard, F. L. 1931. Life history of Physarum polycephalum. Amer. Jour. Bot.
18:116-133.

Hyman, D. H. 1917-18. Metabolic gradients in Amoeba and their relation to the
mechanism of amoeboid movement. Jour. Exp. Zool. 24:55-97.

Jacobs, M. H. 1922. The effects of carbon dioxide on the consistency of proto-
plasm. Biol. Bui. 42:14-30.

Kamiya, N. 1940. The control of protoplasmic streaming. Science 92: 462-463.

Lewis, W. H. 1931. Locomotion of lymphocytes. Johns Hop. Hosp. Bui. 49:29-36.

. 1931. On the locomotion of lymphocytes. Anat. Rec. 48: 52.

. 1933. Locomotion of rat lymphocytes in tissue cultures. Bui. Johns

Hop. Hosp. 53:147-157.

. 1934. On the locomoton of the polymorphonuclear neutrophiles of

the rat in autoplasma cultures. Johns Hop. Hosp. Bui. 55:273-279.

. 1938. On the role of a superficial plasmagel layer in division, loco-
motion and changes in form of cells. Arch. exp. Zellforsch. 22: 270.

AND M. R. Lewis, 1938. Studies on white blood cells. Pub. No. 51,

Cooperation in Research, Carnegie Institution of Washington, pp. 369-382.

. 1939. The role of a superficial plasmagel layer in changes of form,

locomotion, and division of cells in tissue cultures. Arch. exp. Zell-
forsch. 23:1-7.

. 1939. Contorted mitosis and the superficial plasmagel layer. Amer.

Jour. Cancer 35:408-415.

. 1939. Changes of viscosity and cell activity. Science 89:400.

Viscosity Changes of Protoplasm 197

. 1940. Some contributions of tissue cultvire to development and

growth. Growth Supp., pp. 1-14.
Marsland. D. a. 1938. The effect of high hydrostatic pressure upon cell division

in Arbacia eggs. Jour. Cell, and Comp. Physiol. 22:57-70.
. 1938. The effects of high hydrostatic pressure upon the mechanism

of cell division. Arch. Exp. Zellforsch. 22:268-269.
. 1939. The mechanism of cell division. Hydrostatic pressure effects

upon dividing egg cells. Jour. Cell, and Comp. Physiol. J3: 15-22.

1939. The mechanism of protoplasmic streaming. The effects of

high hydrostatic pressure upon cyclosis in Elodea canadensis. Jour. Cell,
and Comp. Physiol. 13:23-30.

Marsland, D. A., and D. E. S. Brown. 1936. Amoeboid movement at high hydro-
* static pressure. Jour. Cell, and Comp. Physiol. 8: 167-178.

Mast, S. O. 1923. Mechanics of locomotion in amoeba. Proc. Nat. Acad. Sciences
9:258-261.

. 1926. Structure, movement, locomotion, and stimulation in Amoeba.

Jour. Morph. and Physiol. 41:347-423.

. 1926. Structure of protoplasm in amoeba. Amer. Nat. 60:133-142.

. 1929. Mechanics of locomotion in amoeba proteus with si>ecial ref-
erence to the factors involved in attachment to the substratum. Proto-
plasma 8:344-377.

1931. Locomoton in amoeba proteus (Leidy) . Protoplasma 14:321-330.

ScHECHTMAN, A. M. 1937. Localized cortical growth as the immediate cause of

cell division. Science 85: 222.
Seifriz, W. 1938. Recent contributions to the theory of protoplasmic structure.

Science 88: 21.

PHYSICAL ASPECTS OF PROTOPLASMIC STREAMING

NoBURo Kamiya
I. INTRODUCTION

Protoplasmic streaming is maintained by a motive force which
overcomes the viscosity of protoplasm, and in so doing imparts a
specific velocity gradient to it. Until now the magnitude of this
motive force was not known, nor was it known how it changes.

The development of a suitable technique has made it possible to
measure the absolute value of the motive force generated in proto-
plasm.

The experiments were done on the plasmodia of the Myxomycete,
Physaruvi polycephalum. As is well known, the plasmodium of a
Myxomycete consists of stationary protoplasm of higher viscosity
and less viscous flowing protoplasm, both of which are readily inter-
convertible. Streaming is accompanied by changes in the contour of
the Plasmodium as a whole. These two processes, streaming of the
fluid, inner protoplasm, and change in contour of the whole body are
so closely related that one movement necessarily involves the other.
One remarkable feature of the protoplasmic streaming in Myxomy-
cetes is its comparatively great speed, and especially significant is
the characteristic reversal in direction of flow. The velocity of flow
changes according to the rhythmic pattern.

Before proceeding with the subject matter of this paper, it would
seem pertinent to state briefly the behavior of protoplasmic flow
under locally applied mechanical pressure, which behavior was so
characteristic as to lead me to undertake the experiments reported
here.

A cover-glass culture of a plasmodium is inverted on a glass slide
with enough water to keep the protoplasm from touching the slide.
Then, a certain point on the cover glass is gently pushed down by
means of a bent needle affixed to a micromanipulator. When some
part of a plasmodium thus mounted first touches, under applied
pressure, the surface of the slide, the movement of the interior, fluid
protoplasm is strikingly affected. As the mechanical pressure is
exerted unequally upon the plasmodium, because of the uneven
surface of the plasmodium, and the fact that pressure is applied only

[199]

200 The Structure of Protoplasvi

at one selected point of the cover glass, the fluid protoplasm is
pressed out of the parts to which the pressure is applied and goes
to those parts which are free from pressure. By this simple pro-
cedure the spatial relations of the fluid protoplasm can be observed
to change very readily in the network of strands or in the branched
courses throughout the fan-like expanses of gelatinous protoplasm.
The movement of the protoplasm induced by the mechanical pres-
sure is superimposed on the normal flow. Therefore, when applied
mechanical pressure affects the natural flow, a temporary accelera-
tion, retardation or reversal in direction takes place, after which
the protoplasm takes its normal course.

Such superimposed modified flow, controlled by mechanically
applied pressure, causes no injury to the protoplasm. When the
mechanical pressure is removed, the surface layer of the protoplasm
takes on its original form again, presumably because of its elasticity,
and the interior fluid protoplasm, which had been displaced, returns.
By careful control of the screw of the micromanipulator, a spasmodic
movement, as well as a rebounding of the interior protoplasm, can
be repeatedly produced, without giving rise to any observable dis-
turbances, provided that the applied pressure is not so strong as to
cause a structural disturbance of the protoplasm, such as was
observed by Balbach (1936) .

From the simple experiment above described one can say that
the streaming of the interior more fluid protoplasm is modified in a
manner directly dependent upon the pressure applied. These
observations verify the conclusion of Hilton (1908), who applied
mechanical-pressure by the tapping of a needle upon a small piece
of cover glass placed outside of the microscopic field and covering a
part of the body of a plasmodium. The application of localized pres-
sure can also be made with a blunt glass microneedle by pressing
it directly against the surface of the plasmodium, as reported by
Camp (1937).

The fact that a slight, unequally applied pressure produces prac-
tically no injury to protoplasm, yet induces artificial flow, suggests
the possibility that protoplasm is normally driven passively by a
pressure difference established in the system of a plasmodium. Cur-
iosity about such a behavior of protoplasmic flow led me to perform
further experiments in which air-pressure was used instead of
mechanical-pressure.

If each of two masses of protoplasm connected by an unbroken

Physical Aspects of Protoplasynic Streaming 201

strand are placed in separate air-tight compartments, thus permitting
the protoplasm to flow freely from one compartment to the other,
and if the air-pressure in one of the compartments is modified, then
the protoplasmic streaming in the connecting strand will be modified
accordingly. The tendency to flow from one compartment to the
other is thus opposed by a counter-pressure.

In order to realize such an experiment in practice, the following
technique has been arrived at after repeated modifications and
improvements.

II. METHOD

(a) Preparation of the Material. The stock culture of Physarum
polycephalum is grown on moist filter paper and fed powdered oats
(Camp, 1936). Then a tiny bit, say 1/20 gram, of protoplasm is
removed from the stock culture and placed on the surface of nutri-
tion-free (tap water) 2 per cent agar in a Petri-dish. After an hour
or two the protoplasmic mass spreads out into a thin round sheet,
and later into a network or branched system of vigorous strands
with a wavy advancing margin or several peninsula-like expanses.
When a comparatively straight part is found among these vigorous
strands, a small rectangular sheet of agar (ca. 15 X 25 mm., 1 mm.
in thickness) , including the selected protoplasmic strand, is cut out.
This rectangular sheet of agar, containing a strand of protoplasm
lying parallel to its longer sides, is placed on a glass slide of 24 X 60
mm. (Fig. lA) .

The next step in the preparation is to put new blobs of protoplasm
from the stock culture on to both terminals of the strand. The
strand and the two blobs of homospecific protoplasm soon fuse, and
the three separate parts now unite to make one dumbbell-shaped
Plasmodium (Fig. IB) .

Further procedure involves inverting this glass slide with the
agar sheet and the dumbbell-shaped plasmodium on to a specially con-
structed double chamber, shown in Figure 2. The glass slide covers
the top of this chamber. As shown in Figure 2, two glass partitions
(p) 1 mm. thickness, are fastened vertically across the chamber 6
mm. apart from each other. As the height of these partitions is about
1.5 mm. less than that of the side wall of the chamber, there remains
just this much distance between the upper edge of the glass partitions
and the under surface of the slide after the latter is inverted on the
top of the chamber. This space will partly be occupied by the con-
necting strand of the protoplasm and by the thin sheet of agar that

202

The Structure of Protoplasm

adheres to the sHde. The small space between the two partitions is
filled with a 2 per cent agar sol at about 40°C., the meniscus of which
is higher than the upper edges of the partitions. Before the agar is
transformed to a gel, which occurs at about 35°C., the glass slide
with the agar sheet and the dumbbell-shaped plasmodium is inverted
over the chamber. As the agar is still in a sol state, it fills up all the
spaces to be closed without injuring the delicate strand of protoplasm.

After this procedure the 2 per cent
agar soon gelatinizes, and the
chamber is divided into two air-
tight compartments. Figure 3A
shows a cross-section through the
double chamber at the agar wall
and Figure 3B, its longitudinal
section at the iniddle part.

Protoplasmic flow^ is stopped
temporarily because of mechani-
cal and thermal disturbances, but
it soon recovers. Meanwhile, the
two blobs of protoplasm at both
ends of the strand, which are in
the separate compartments, spread
out, having already fused with the
connecting strand (Fig. 4A) .
The observation chamber is now
put, with rubber gaskets, between two metal frames which are
tightly joined together by means of four screws. Thus is the pre-
paration of the material accomplished (Fig. 4B) .

It is a favorable characteristic of agar to have a large hysteresis
range, within which agar can exist either as a sol or as a gel. Two
hardly compatible technical requirements are thus fulfilled; namely,
the two separate compartments are kept air-tight against a consid-
erable pressure difference between them, and at the same time the
cross-wall does not block the flow of protoplasm in the delicate
connecting strand. Furthermore, agar is sufficiently transparent to
permit observation of the flowing protoplasm in the connecting
strand. It is by means of light transmitted through the agar wall that
the observation is made. Though it is not absolutely necessary for
simple observation of flow, one can get a clearer microscopic image
of the connecting strand if a rectangular prism of glass of suitable
size is mounted in the agar wall between the two partitions.

Fig. 1. A. An agar sheet, ag. with
a protoplasmic strand, st, placed on a
glass slide. B. Two protoplasmic
masses placed on the agar sheet at
both ends of the strand.

Physical Aspects of Protoplasmic Streaming

203

Fig. 2. Observation chamber with two partitions, p.

X\\\\\\\\\\\v ■ • ^4 : ■ , \K\\\\\\\\\>;-j

b B

B

Fig. 3. A. The cross section through the prepared observation chamber at the
agar wall. The broken line, e, shows the upper edges of the two glass partitions.
B. The longitudinal section through the middle part of the same chamber
divided into the two compartments, A and B. The protoplasm, a and b, in the
separate compartments and the connecting strand, st, are dotted. The shaded
parts in figures A and B represent agar. That part in which the hatching runs
from upper left to lower right represents an agar sheet adhering to the glass
slide, whereas that part in which the hatching runs from lower left to upper
right represents agar poured between the two partitions, p, just before inverting
the slide over the chamber, and which then serves to seal the chamber.

204

The Structure of Protoplasm

(b) Control of Pressure. In order to establish a pressure differ-
ence between the two compartments, the opening of one compart-
ment is connected with a passage leading to an aspirator tube and a
manometer, while the opening of the other compartment is kept open,
as shown in Figure 5.

Aspirator As, used for controlling the air-pressure, is made of a
thick, rubber tube, the middle part of which is placed between two
solid metal plates (shaded in Fig. 5) . The upper plate is equipped

so that it can be moved
gently up and down by
means of a screw S, with
the result that the inner
volume of the rubber as-
pirator tube can be con-
trolled most exactly.

After opening the stop-
cock SC, the aspirator is
pushed down gradually by
tightening the screw S.
The air pushed out by this
depression is let off at the
stop-cock SC. When the
aspirator tube is pushed
down to about half of the
stroke, the stop-cock SC
is closed. The whole sys-
tem consisting of compart-
ment B, the aspirator As
and the manometer M is
now air-tight. If the screw
S is tightened farther, the
inner pressure of the entire system is increased, i.e., the pressure
of compartment B becomes higher than that of compartment A. If,
on the contrary, the screw S is loosened, being turned in the opposite
direction, then the aspirator tube will expand because of the elas-
ticity of its rubber wall. The inner pressure of the system, including
compartment B, becomes, therefore, lower than atmospheric pres-
sure. The pressure of compartment B can thus be controlled to any
desirable degree within the necessary range, both in regard to
positive and negative compression, by means of a single screw. The
range desired is usually between + and — 30 cm. of water.

Fig. 4. A. The double chamber, after in-
verting the glass slide with the agar sheet
and the material. The thin strand of proto-
plasm runs through the agar wall. B. The com-
pleted preparation: finished observation cham-
ber fastened between metal frames with two
rubber gaskets.

Physical Aspects of Protoplasmic Streaming

205

The pressure difference between compartment A and compart-
ment B is ascertained by a water manometer M of 50 cm. in height.

Compartment A and compartment B will, hereafter, be referred
to as A and B for the sake of simplicity.

Ill EFFECT OF PRESSURE DIFFERENCE ON PROTOPLASMIC

STREAMING

' When there is no pressure difference between A and B, streammg
takes place normally along the connecting strand of the prepared

W^

Fig. 5. Diagram showing the arrangement of the whole system consisting of
the double chamber having two compartments, A and B, rubber aspirator, As,
controlled by screw, S, manometer, M, and stop cock, SC. Agar wall is shaded.
Protoplasm in A and B is designated a and h, respectively.

Plasmodium which penetrates the agar wall, showing the usual
rhythmic reversal in direction of flow. But when the two com-
partments are subjected to a pressure difference, the velocity of
the streaming, which otherwise would proceed normally, is modi-
fied in accordance with the extent and direction of the pressure
difference established (Kamiya, 1940) .

Let us suppose the protoplasm to be flowing from a to b (Fig. 5) .
If a slightly lower pressure causing suction is applied to B, then the
flow of the protoplasm along the connecting strand from a into b is
accelerated.

206 The Structure of Protoplasm

During the time in which the protoplasm flows in the opposite
direction, namely from h to a, the air-pressure of A must be lowered
below that of B, if the flow is to be accelerated. In practice, however,
the pressure of B is increased instead of decreasing the pressure of
A, in order to establish a pressure difference in this direction.

Protoplasmic streaming can be artificially speeded up in this way
to maximum velocities which are beyond those of natural conditions.
By establishing a pressure difference of, say 20 cm. of water, so as to
cooperate with the motive force developed in the protoplasm, one
can readily increase the velocity to more than 2 mm. per second
through a connecting capillary of 200!^i inner diameter. The normal
maximum is about 1 mm. a second.

Next must be considered the effect of a pressure difference,
which is established so as to oppose the protoplasmic force. In this
case, as expected, the velocity of the flow is retarded. When, for
instance, the protoplasm flows from a into b, the volume of the proto-
plasm b is, of course, increased. If the air-pressure of B is made
higher than that of A, then it is clear that a part of the motive force
generated in the protoplasm must be used in the work required to
bring about the expansion of the protoplasmic volume against the
higher pressure which has been artificially applied. Should the
pressure applied be stronger than the protoplasmic force developed
in the plasmodium, then the forward-moving protoplasm is forced
backwards. In other words, the motive force of the protoplasm is
overcome by the application of a counter-pressure.

If, now, the protoplasm is flowing as before, namely from b to a,
a lower pressure must be appHed to B, if the flow is to be retarded
or reversed.

From the foregoing description it is evident that by changing
the pressure of B both on the + and — sides of the manometer, the
direction and speed of the protoplasmic flow along the connecting
strand can be accurately controlled. Flowing protoplasm thus con-
trolled is driven by a resultant force which is the algebraic sum of
the shearing stress developed in the protoplasm and that caused
by the pressure difference artificially induced.

One would expect that artificial acceleration, retardation, or
reversal of flow would induce serious disturbances in the proto-
plasm, but a wholly normal flow is resumed as soon as the pressure
is released. It is rather surprising that such a remarkable modifica-
tion of the flow does not result in any pathological abnormality.

Physical Aspects of Protoplasmic Streayning 207

IV. MEASUREMENT OF THE MOTIVE FORCE

By controlling the air-pressure in B, it is possible to oppose the
motive force in such a way as to hold the protoplasm at a standstill.
The counter-pressure, which is just sufficient to prevent the proto-
plasm from flowing either forward or backward, is a measure of
the absolute value of the motive force responsible for the streaming
of the protoplasm. This counter-pressure has been termed "balance-
pressure" (Kamiya, 1940) .

The range of the absolute value of the balance-pressure is
usually within 20 cm. of water. But this value, as will be shown
later, varies from plasmodium to plasmodium of the same species,
and from rhythm to rhythm of the same plasmodium under the
same external conditions. The maximum value so far encountered
was 30 cm. of water.

Since the motive force developed in the protoplasm does not
remain constant, the balance-pressure must be adjusted accordingly,
if the protoplasm is to be kept immobile. In order to do this, one
must constantly watch the direction tendency of the moving proto-
plasm at some definite part of the connecting strand. As soon
as the equilibrium between the natural motive force and the balance-
pressure is broken and, consequently, the tendency of the flowing
direction is recognizable, the counter-pressure is increased or de-
creased, as the case may be, so that the protoplasm under observa-
tion always stays at the same place. With a little practice and skill
this procedure enables one to restrict the movement of the proto-
plasm within a range of 50ii. So sensitive is the movement of the
protoplasm that the slightest deviation from the balance point will
induce movement in an 8 mm. strand.^

When the protoplasm is kept motionless at a definite point in
the connecting strand, there is no longer a mutual transference of
the protoplasm between the two parts of the plasmodium. However,
it must be noted, that suspension of flow of the protoplasm under
observation does not mean the cessation of movement elsewhere
in the plasmodium. Local displacement of the interior protoplasm,

^This fact does not, however, necessarily mean that protoplasm has no
"yield value," since there is a possibility that displacement due to the slightest
shearing stress may be ascribable to an elastic deformation. The inherent
motive force which cannot be eliminated under normal conditions makes it
impossible to determine whether or not the slightest shearing stress is capable
of allowing protoplasm to undergo continuous deformation (flow), or whether
a certain amount of critical shearing stress is necessary to make protoplasm
flow.

208 The Structure of Protoplasm

though restricted, continues to take place within the protoplasmic
expanses at both ends of the connecting strand, which are spread
into more complicated forms than those schematically shown in
Figure 5, a and b.

Morphologically, the two parts of protoplasm in the different
compartments are still connected, but functionally, they behave just
as if they were two independent plasmodia, as long as movement in
the connecting strand is suspended.

As in the case of an artificial change in velocity, the cessation of
protoplasmic flow for a period of time by uninterrupted adjustment
of the balance-pressure also causes no visible abnormalities in the
protoplasm. Nor is there any indication of physiological disturbance
even after more than two hours, during which time the flow is
stopped. When the balance-pressure is released by opening the
stop-cock (SC in Fig. 5) , the protoplasm in the connecting strand
immediately starts to flow. The velocity depends upon the inherent
motive force in the protoplasm at the moment of release.

V. RELATED PROBLEMS

The experiment reported above gives a means of attacking prob-
lems covering important subjects of protoplasmic research.
The Capillary Method as a New Means of Measurement of Proto-
plasmic Viscosity.

It is impossible to make slime mold protoplasm flow through a
glass capillary without fatal resuhs, but through its own capillary,
such as a connecting strand, the protoplasm can flow quite normally.
A dumbbell-shaped plasmodium consisting of its own capillary and
two "reservoirs" of protoplasm is, as it were, in itself a capillary
viscometer. If one releases a balance-pressure of known value and
compares the velocities of protoplasmic flow at the time of release
in a treated and a control plasmodium, one can estimate the relative
viscosity of the interior flowing protoplasm under controlled con-
ditions. As the inner diameter of the connecting strand is not kept
exactly the same, some correction for this is needed.

The diameter and length of the connecting strand are both
determinable; the motive force responsible for the flow can now be
measured. By determining the velocity of protoplasmic flow through
the capillary of known diameter and length under known shearing
stress (motive force) , it seems not impossible to evaluate even the
absolute viscosity of protoplasm by applying Poiseuille's Law.
In passing, another point may be mentioned, namely, the

Physical Asjiects of Protoplasmic Streaming 209

possibility of causing protoplasm to flow under various rates of
shear. It may be that the capillary method will help throw some
light on the study of the anomalous, or non-Newtonian, flow of
protoplasm which is at the present time based upon a complex
mixture of fact and theory.
Motive Force versus Viscosity.

* As remarked by Ewart (1903, 61) , the speed of the flow is
dependent upon viscosity as well as upon the motive force which
drives the protoplasm. In fact, streaming must be considered in
terms of these two variables. There has been, however, no method
developed to measure the motive force, which is a fundamental factor
in the mechanics of protoplasmic flow. Heilbrunn (1937, 243) pointed
out the difficulty arising from an attempt to interpret the result
of experiments on protoplasmic streaming, since it was impossible
to decide to what extent a change in speed is due to a modification
of the motive force, and to what extent it is due to a change in the
viscosity of the protoplasm. It is hoped that the experiments reported
above may make some contribution to analytical studies on the effect
of physical and chemical agents on protoplasmic streaming. Some
of the observations made that might prove to be of value in this
respect, are as follows: ascertaining, when an external agent has been
applied that the protoplasmic flow is stopped because of an immense
increase in viscosity rather than a dying away of the motive force;
or ascertaining that the situation is different, namely, the flow is
retarded because of a reduction in motive force rather than an
increase of the viscosity.

VI. GRAPHICAL REPRESENTATION OF THE MOTIVE FORCE—
"DYNAMOPLASMOGRAM"

In section IV, it was stated that the balance-pressure must be
adjusted continuously in order to keep the protoplasm quiet. Since
the balance-pressure is a measure of the motive force, an increase
and decrease of it must mean a parallel change in the motive force.
Changes in the balance-pressure reveal autonomic variations of the
mechanical forces liberated in the vital system.

In order to determine in what manner and to what extent the
motive force changes in relation to time, instantaneous values of the
balance-pressure were read at 5-second intervals. The recording
was made by another observer, who watched the manometer.- The

' I am indebted to Miss A. Cantlin, Mr. J. Evans, Mr. R. Ferlauto, Mr. M. Ross,
Miss M. Uraguchi and others for their kind assistance in the experiment.

210 The Structure of Protoplasm

5-second sound signal was given by an apparatus operated by a
synchronous motor whose speed was governed by A.C. cycle con-
stancy. By plotting a series of these succeeding values as ordinates
against time as abscissas, undulating curves were obtained which
faithfully portray the distinguishing features of the changes which
the motive force undergoes in accordance with the autonomic
scheme of the protoplasm. The graphs thus obtained, which shall be
referred to as "dynamoplasmogram," give a complete pattern of the
rhythm in protoplasmic activity. All the characteristics of rhythm
such as wave form, frequency, polarity, and amplitude are portrayed
by graphical representation.

So striking and impressive is the rhythmic flow of the protoplasm,
that one is inclined to view the streaming as the significant thing,
when actually it is only the visible end-effect of the motive force.
The speed of flow is a function not only of the motive force, but also
of various factors which are not kept constant during the rhythmic
flow, such as thickness of the strand, viscosity,'^ etc.

Holding protoplasm quiet artificially is a method which has
decided advantages. By such means, the motive force can be meas-
ured directly under statical conditions, thus eliminating numerous
factors which would possibly exert a disturbing influence. It was
possible, through the development of special techniques, to measure
continuously rates of flow and volume of protoplasm transported.
When these values are plotted against time, they yield rhythmic
wave patterns comparable to the dynamoplasmograms. However,
the dynamoplasmograms will be considered exclusively as a standard
of rhythm in the present paper, because the motive force when thus
measured is purer in respect to the rhythmic functioning of proto-
plasm than any other observable phenomenon.

In order to understand a dynamoplasmogram, it is necessary to
remember that the protoplasm is being opposed and kept at a stand-
still. In other words, what one observes is quiet, but what one
measures is a motive force which still functions, i. e., the rhythm
continues. The value of the balance-pressure at any point in the
time scheme represents the motive force responsible for the flow
which would have taken place were the balance-pressure removed at
that moment. In virtue of the high sensibility of protoplasmic mobil-

' Viscosity may not remain constant under different rates of shear (velocity
gradient) during the normal rhythmic flow, since protoplasm is thought to be
non-Newtonian in nature (Pfeiffer, 1936, 1937).

Physical Aspects of Protoplasmic Streaming 211

ity to pressure variations, the experimental data obtained would
seem to have a high degr-ee of accuracy.

The measurements were made at room temperature (22°-25"C.) ,
which deviated not more than 0.5°C. during the time of experiment.

In all the following figures, the abscissas represent the balance-
pressure applied to B (Fig. 5) , but they are equal to the vital motive
force, if the direction from a to b is regarded as positive and the
opposite direction, viz., from b to a as negative.

Figure 6 represents the usual and most regular change in the
motive force. The undulating curve is beautifully smooth and the
points plotted are almost in line. The points at which the wave
passes through the base line, where the balance-pressure is zero,
correspond, if the protoplasm were free to flow, to the reversal
moments of the flowing direction. According to the concept generally
accepted, one rhythm involves the duration of time for a complete
progression and regression of the protoplasmic streaming (cf. Vouk
1910, 1913). Therefore, graphically, the distance on the base line
between alternating reversal points, including both + and — sides
of the wave, denotes, in a general sense, "one" complete rhythm.
Except for the wave groups at the beginning and at the end of the
wave train, the distances between any two adjacent maxima or adja-
cent minima have approximately constant intervals (periods) aver-
aging 93.7 seconds, and there is also mostly one point of inflexion
between the consecutive maximum and minimum of the curve. The
wave looks, on the whole, fairly regular and bears considerable
resemblance to the sinusoidal wave.

The wave in Figure 6 does, however, show slight fluctuations,
which indicate that there is quantitatively no constant factor in
respect to amplitude, wave-form, etc. At the beginning and end of
the wave train, the amplitude diminishes conspicuously, and the
wave no longer sustains a regular form; whereas, in the middle part
of the train, the waves swell. When one considers the "envelope",
or imaginary margin, of this wave train, it assumes the form of a
long spindle, the middle portion of which is six times as wide as
the ends.

It is extremely interesting to note that the change in amplitude
often takes place with considerable regularity, i. e., after a decreas-
ing amplitude period the wave swells out gradually to a maximum,
and after the lapse of a certain period of time, the magnitude
decreases again and is followed by the next increasing period. Fig-
ure 6 shows simply one long "waxing" period between two "waning"

212

The Structure of Protoplasm

Fig. 6.

Physical Aspects of Proto'plasmic Streaming

213

12

10

8

6

4

2

0

-2

-4

-6

-8

-10

- 12

Fig. 6 — continued.

214 The Structure of Protoplasm

periods. The alternation of waxing and waning periods is better
illustrated in Figure 7. During the periods of 9-13, 21-25, 32-36,
and 42-46 minutes on the abscissa, the waves have an increased
amplitude, and the shape approaches the harmonic wave. Between
any two of these adjacent wave groups, there are periods of decreased
amplitude where the wave form is irregular and clonic. The interval
of waxing and waning periods in Figure 7 remains approximately
constant for about 11 minutes; that is, the wave undergoes a similar
change in amplitude every 11 minutes. This is, however, not a
general rule, but applies only to the plasmodium used in this experi-
ment. The duration of the waxing and waning periods of a Plas-
modium, if any, differs from specimen to specimen; indeed, it
changes in one and the same plasmodium, as will be considered later.
Figure 8 shows a pattern similar to Figure 7, but the increase
and decrease of the amplitude takes place more frequently. Here
the irregular and clonic portions of the curve, which correspond to
the waning periods in the foregoing graphs, appear every 6-7 min-
utes, preceded and followed by only two or three smooth and reen-
forced waves. The patterns of the clonic parts of the waves are not
similar to each other.

From the graph discussed above (Figs. 7 and 8), it is evident
that the waves undergo pronounced variations in amplitude and that
these variations occur according to a rhythmic pattern. It must be
noted, however, that the regular alternation between waxing and
waning periods can not be observed in all specimens of plasmodia.
The waxing and waning of the wave train often take place at vary-
ing intervals and to a varying extent. Figure 9 is an example of this.
The wave magnitude changes from rhythm to rhythm, but tonic and
clonic changes occur at irregular intervals. During the periods 0-6,
26-29, 55-70 minutes on the abscissa, the waves show a decrease in
magnitude. A similar tendency is also recognized, though less promi-
nently, during 16-19 minutes on the abscissa. On the other hand,
the wave groups between two of these depressed portions of the
curve are comparatively tonic. The constancy of the waxing and
waning periods is not maintained here.

The foregoing graphs show that the motive force undergoes pro-
nounced variations not only in its magnitude, but also in its wave
form during the rhythmic succession of vital processes. Sometimes
a characteristic form appears repeatedly throughout several rhythms.
Even when an exactly similar wave form is not repeated, the change
is generally transitional in the course of the successive waves.

Physical Aspects of Protoplasmic Streaining

215

1x1

cr

D
(f)
<J)

q:

Q_

I

UJ

U
z
<

_l
<

CD

16
14

12

10

8

6

4

2

0

-2

-4

-6

-8

-10

-12

- 14

TIME IN MINUTES

Fig. 7.

216

The Structure of Protoplasm

Physical Aspects of Protoplasviic Streayning

217

Fig. 7 — continued.

In Figure 10 the form and amplitude of each successive wave is
almost the same. As the wave repeats nearly the same details of the
change, it is approxiinately periodic within this range of the time
scheme. It is noteworthy that the curve shows a special concavity
following each trough and also, a somewhat similar concavity fol-
lowing each crest. The wave as a whole is not symmetric, but it does
show a symmetric character. The rhythmic interval (period) is
only 84 seconds. The dash-dot line, which will be considered in the
following section, cuts the wave into two equal + and — areas.

In Figure 11 the wave form is very characteristic. The wave rises
suddenly, then falls rather suddenly at first, and later more gradually
until the trough is reached. Nevertheless, the form of each rhythmic
pattern is very similar, although the magnitude of the wave gradually

TIME IN MINUTES

Fig. 8 — continued.

220 The Structure of ProtojplasTn

decreases in this case. The distance between any two adjacent
maxima and minima of the waves is nearly constant for about 117
seconds.

Figure 12 shows another pattern of the rhythm. In this figure, the
crest is acute, while the trough is flat. The magnitude and the period
(ca. 132 seconds) are greater than those of Figure 11. Though the
amplitude diminishes gradually, as in Figure 11, the wave form re-
mains similar in each rhythm. The dash-dot line and the dotted
waves will be taken up later.

VII. GENERAL CHARACTERISTICS OF DYNAMOPLASMOGRAMS

The graphs presented in this paper are some of the representative
cases of more than 40 wave trains so far obtained in my experiments
with different specimens. By inspecting these graphs one can sum-
marize their general features as follows:

1. The pattern of the curve undergoes variations from rhythm to
rhythm in the same specimen under the same external conditions.

2. A marked change in amplitude sometimes recurs at regular
intervals (Figs. 7 and 8) , but this regularity is not always maintained,
even though the waves continue to change amplitude (Fig. 9) .

3. A similar wave form often reappears throughout several suc-
cessive rhythms (Fig. 10) . Changes in the wave form as well as in
amplitude are transitional (Figs. 11 and 12) ,

These general characteristics of the curve are, however, open
to possible question. One may wonder whether or not the peculiar
pattern of the wave has anything to do with a "grafted" plasinodium,
in which three separate portions (two protoplasmic blobs and a
connecting strand of protoplasm) were united in preparation. This
is answered by the fact that a plasmodium, the form of which, with-
out grafting, has been so changed that there are two protoplasmic
bodies connected by a single strand, also shows characteristics in
its dynamoplasmogram which are identical to those of the grafted
one.

In viewing the foregoing graphs, it is to be noted that the curves
show no sign of a special deflexion when they cross the base lines;
in other words, the points which correspond to the moviejit of rever-
sal are not characterized at all by the shape of the curve. Previous
investigators have almost implicitly adopted the moment of reversal
of flow as the only criterion for distinguishing one rhythm from
another; for other than this, there has been no gauge with which to

Physical Aspects of Protoplas^nic Streaviing

221

q:
CO

UJ 'i^
Q_ 5
uJ o

53

_J
<

m

12

10

8

6

4

2

0

-2

-4

-6

-8

-10

■12

TIME IN MINUTES

12

10

8

6

4

2

0

-2

-4

-6

-8

10

12

-

^

N

\

\

\

22/

- is\^^

/l6

17

18

.9 V

^—^20

2, ^

^

-

Fig. 9.

222

The Structure of Protoplasm

Physical Aspects of Protoplasmic Streaming

223

12

10

8

6

4

2

0

-2

-4

-6

-8

10

12

-

-^

^

v^ «,^

^^r

1 I I

^62

63\^

^^ 64

66

-

12

10

8

6

4

2

0

-2

-4

-6

-8

-10

- 12

Fig. 9 — continued.

measure the period of the rhythm. It is evident that this method of
separating one rhythm from another is most practical, but one must
keep aware of the fact that the instants of reversal, where the motive
force is zero, have neither a significant meaning in rhythmic activity,
nor do they imply the savie phase of rhythm as manifested by the
graphs (all except Fig. 10). A reversal moment simply means a
momentary equilibrium of the intraplasmic forces. Pease (1940) is
also of the opinion that cessation of flow indicates a temporary equili-
brium state by acting as a "trigger mechanism."

224

The Structure of Protoplasm

Those who have once observed protoplasmic streaming in a Plas-
modium will recall the "abnormal", but not unusual, phenomenon
that the protoplasm shows when it continues to flow only in one
direction for an excessively long period of time. From such "abnor-
mal" flow, one would necessarily be led to believe that protoplasmic
rhythm is often "disturbed." If one bases one's judgment of the

20

18

16

14

u

12

tr

in

D

B

(/)

,,^

(0

cc

b

UJ

UJ

<

4

q:

2

a.

?

1

0

LJ

Li.
O

-2

Z

5

-4

<

o

-b

_)

-8

<

-10

CD

-12
-14

TIME IN MINUTES
Fig. 10.

rhythm of protoplasm merely upon the reversal of flowing direction,
then one will overlook the following important situation.

Figure 13 shows a case in which the whole undulating curve is on
the -)- side of the zero-ordinate for about 12 ininutes. This means
that the motive force is exerted always in the same direction during
this tiine (from a to h, Fig. 5) . But nevertheless, the undulating
curve strikingly reveals a regular change in the force generated in
the protoplasm throughout that time.

Although Figure 13 represents an extreme case, it is generally
noticeable in all the foregoing graphs that the area included within
the curve and the zero-ordinate on the -|- side differs from that on the
— side. This difference between the + and — side indicates that
the motive force is not exerted equally in both directions, but
always more in one direction than in the other. Because of this
polar nature of the motive force, a plasmodium, were it free to
move, would advance by flowing forward a little more each time
and not retreating all the way back to its original position when it
reverses.

Let us suppose a smooth line cuts an undulating curve in such
a way that the area on the upper side of the line is equal to that on

Physical Aspects of Protoplasmic Streaming

225

the lower side. If the curve is periodic, or almost periodic, as in
Figure 10, this line can be found easily with the aid of a planimeter,
or by any other practical method. When the wave form changes
from rhythm to rhythm, as is the case in many of the wave trains
with which we are concerned, the line of which we speak is mathe-
matically indeterminate. But in practice we can find such a line with

TIME IN MINUTES

14 r

12

10

8

6

4 -

2

0
-2
-4
-6
-8
-10
-12
-14

Fig. 11.

a certain degree of approximation. This can be done graphically
by measuring the successive areas included within common tangents
of consecutive crests or troughs, and the experimental curve. These
areas divided by the distances between two contact points of the
common tangents show the mean distances from the tangents to
the line being sought.

The dash-dot lines XY in Figures 10, 12, and 13 are lines which
divide the waves approximately into equal -f- and — parts. These
lines show the "true" axes of the waves, the ordinate of which is a
measure of the polarity of the plasmodium in its absolute scale.
These true axes or "polarity lines," are generally (except Fig.
10) not parallel to the base line. Lack of parallelism shows that the

226

The Structure of Protoplasm

polarity, for which the displacement of the whole plasmodium is
responsible, changes its intensity spontaneously. It may tentatively
be assumed that polarity and rhythm are independent variables.
This is a permissible assumption until the concrete meanings of
polarity and rhythm are elucidated on a physical or chemical basis.

UJ

cr

D

(^ „
(/) cr
LJ ^

cr

Q.

I
UJ

U

<

CD

16

14

12

10

8

6

4

2

0

-2

-4

-6

-8

-10

-12

-14

TIME IN MINUTES

Fig. 12.

Although the true axes, or polarity-lines, are not shown in the above
figures, except in Figures 10, 12, and 13, nevertheless, it is possible to
find them in all cases with a limited degree of accuracy.

VIII. CHARACTER OF PROTOPLASMIC RHYTHM

We are now confronted with the fundamental problem of the
character of the rhythm operated in the protoplasmic system. If the
spontaneous increase and decrease in amplitude of the waves under
the same external conditions are regarded as due to a physiological
disturbance or to "fatigue" of the protoplasm, the question naturally
arises, just what is the physiological disturbance; what is fatigue
of protoplasm. A further question is the cause of the repeated

Physical Aspects of Protoplasviic Streaming

227

22

20

Id

16

14

u

12

DC

in

J)

8

c/l

(f) 5

b

UJ u

4

a: 5
Q- 1

2

1

()

UJ n

-?

V ^

-4

Z o

f;

< ^

_l

-a

<

- 10

m

- 12

TIME IN MINUTES

22

20

18

16

14

12

10

8

6

4

2

0

-2

-4

-6

-8

- 10

- 12

-

l^\

^

\

/ ^

I - -' V

^_V

/— \

v-/-

.\

A

/

\ /

\ ~ ~ — J

[ \

/

\J

\ /

'" \

\

-

1

t

V

8

9

10

1 1

12

13

14

•

22

20

18

16

14

12

10

8

6

4

2

0

-2

-4

-6

-8

- 10

- 12

Fig. 13.

228 The Structure of Protoplasm

appearance of a characteristic wave form in successive rhythms (cf.
Figs. 10, 11, and 12) of the vital process.

By studying many examples of wave trains, it seems most prob-
able to me that such extraordinary variability in the pattern is due
to nothing but the well-known physical phenomenon called inter-
ference.

Interference is one of the most remarkable phenomena asso-
ciated with wave motion. It is well known in acoustics, optics, and
other fields of science. The significance of interference in vital
functions has, however, been scarcely considered. In order to study
the rhythmic character of protoplasmic force, the rhythms being
regarded as a combination of a series of components, it is pertinent
to take up the general features of compounded harmonic waves.

A simple case is that of two superimposed harmonic waves
having equal wave lengths. The resultant wave is a harmonic curve
of the same frequency, no matter what the amplitudes and phases of
the component waves may be. The amplitude of the resultant depends
upon the amplitudes of the two component waves, and upon the
phase difference between them. If the components are always in the
same phase, the amplitude of the resultant wave is the sum of the
amplitudes of the components. If one of the two components is half
a period behind the other, the amplitude of the resultant is the
difference of the amplitude of the components. From this, it follows
that any number of harmonic curves of the same wave length may be
so formulated that the resultant is a single harmonic curve of the
original wave length. Conversely, if a harmonic curve is experi-
mentally obtained, it can be assumed to be the resultant of numerous
indeterminate harmonic components having the same wave length.
The resultant wave will exhibit a harmonic curve regardless of the
amplitudes and phases of the components. If, therefore, the pattern
of a dynamoplasmogram shows pronounced similarity to a simple
harmonic wave, this does not necessarily mean synchronous coopera-
tion of the intraplasmic forces; it means simply, that all the com-
ponent forces in one and the same plasmodium have the same
frequency of the motor mechanism.

The next case is very important for our subject; it is that in
which the frequencies of the two component waves differ slightly.
Because of a difference in wave length, the phase difference between
the two components does not remain constant. The amplitude of the
resultant wave is increased when the waves are in the same phase,
but when the waves are in opposite phases, the amplitude of the

Physical Aspects of Protoplasmic Streaming 229

resultant is reduced. Consequently, the resultant wave undergoes
pronounced variations in amplitude. In the former case the result-
ant is "constructive", whereas, in the latter case it is "destructive".

When two simple tones, the frequencies of which differ slightly,
are sounded together, a throbbing is heard. The two notes are said
to produce "beats." This is a fact well-known to acoustics, and is
represented in Figure 14 where the thin lines show two component
waves, the ratio of the frequencies of which is 8:9. The resultant
of these two component vibrations is shown by a thick line. The
amplitude increases, when crest meets crest and trough meets
trough, whereas, it decreases when crest meets trough. The number
of beats per unit time is equal to the difference in the frequencies
of the components during the same unit time. Therefore, the longer
the period of one beat, the less the difference between the frequencies
of the two components.

Although the component harmonic curves of different wave
lengths can no longer be compounded into a single harmonic curve,
as shown on the graph in Figure 14, the resultant is, nevertheless,
periodic, if the wave length of the components are commensurable.
If, on the other hand, the wave lengths are incommensurable, the
period of the resultant curve is infinite, i. e., the resultant is non-
periodic.

It is extremely interesting to compare the above-mentioned result-
ant wave with the experimental curve of the motive force generated
in protoplasm. The alternation of waxing and waning periods bears
a strong resemblance to the phenomenon of beats. In fact, increase
and decrease in the amplitude of the dynamoplasmogram (Figs. 7
and 8) are, in all probability, due to the interference of intraplasmic
forces, the rhythmic frequencies of which are slightly different from
each other. At the waxing periods, two main component rhythms
of the motive force reenforce each other, whereas, at the waning
periods these two main groups oppose each other. In Figure 7, there
are about eight waves in each "beat" of the dynamoplasmogram.
Since one of the waves must gain a wave length on the other for
each beat produced, one can say that the period of one of the two
main components must differ about Vs from that of the other. As
one "beat" continues here for about 11 minutes, it is possible to
evaluate the approximate difference in the period between the two
components for about 10 seconds. In the case of Figure 8, the period
ratio between the two main components will be ca. 3 : 4.

But the explanations above-mentioned are not quite enough.

230

The Structure of Protoplasm

Fig. 14

Physical Aspects of Protoplasmic Streaming

231

Fig. 14 — continued.

Indeed, the interference of two component waves having sHghtly
different frequencies is, in all probability, responsible for the major
change in amplitude, but it can not be the general principle govern-
ing the protoplasmic system. In order to explain the repeated
appearance of a peculiar wave form throughout several successive
rhythms, the example of which is shown in Figures 10, 11, and 12,
many subordinate components of different frequencies must be
attributed to the protoplasm.

On the basis of the mathematical principle known as Fourier's
Theorem, any one-valued periodic curve can be analyzed into har-
monic curves of suitable amplitudes and phases having wave lengths
of 1, 1/2, 1/3, 1/4, 1/5, and successive aliquot parts of the wave
length of the wave to be analyzed. A given wave can be resolved
into only one combination of simple harmonic curves of definite
amplitude and phases. Theoretically, an infinite number of com-
ponents is generally required to represent arbitrary periodic curves,
but, for practical purposes, the first few components are sufficient,
since the Fourier series converges rapidly for most wave forms.

The curve shown in Figure 10 is practically periodic within the

232 The Structure oj Protoplasm

limit of the time scheme under consideration. Therefore, this curve
can be analyzed into its Fourier components. The characteristic
wave form of this figure enables one to predict that both odd and
even harmonics are present. Figure 15 shows the analysis of the
experimental curve of Figure 10 into its first four components. This
analysis was done by means of a graphical method developed by
Wedmore (1895). The amplitudes of the 1st, 2nd, 3rd, and 4th
component, having the period of ca. 84, 42, 28, and 21 seconds are
respectively 11.75, 1.95, 1.45, and 0.70 cm. of water. When these four

Fig. 15.

components combine, they produce the resultant shown in the thick
line, which is almost the same as the dynamoplasmogram pattern of
Figure 10.

In Figures 11 and 12, very similar wave forms are repeated, but
the amplitudes do not remain constant during the experiments.
These curves are non-periodic within the range under consideration.
Although there is no general method of analysis for non-periodic
curves, "much information may be obtained from such curves by
making skillfully assumed analyses" (Miller, 1916, 141) . The gradual
decrease of amplitude is assumed to be due to an interference effect
between two components, the frequencies of which are nearly in
unison; it is not due to the gradual "exhaustion" of the protoplasm.
The amplitude will increase again after a certain period of time.

By varying the amplitudes and wave lengths of the components
and shifting them arbitrarily along the axis, very similar resultant
waves can be reproduced from three components. Figure 16 shows
the redrafting of Figure 11. The resemblance of both the experi-
mental curve (Fig. 11) and the resultant of the artificially combined
harmonic curves (Fig. 16) is striking. The serial ratio of the wave

Physical Aspects of Protoplasmic Streaming 233

lengths of the three components is 90: 87: 44; that is, two of the three
components have only slightly different wave lengths, whereas, the
other component has a wave length but half as long as the other two.
This component of short wave length is responsible for the peculiar
asymmetric pattern of the curve.

Figure 17, showing the resultant of three harmonic components,
resembles the experimental curve in Figure 12. The acute crests
and flat troughs result from a component whose wave length is half
of one of the other components, the wave lengths of which are

Fig. 16.

slightly different from each other. The three components of differ-
ent wave length are in the ratio 18:17:9. In this case the disturb-
ing effect of polarity must be considered. The strength and direction
of polarity change during an experiment, as shown in Figure 12,
where the true axis (dash-dot line) is not parallel to the base line
and crosses it, revealing a reversal of polarity. If one subtracts the
ordinates of the true axis, or polarity line, from the ordinates of the
curve, and plots these from the same base line, a new curve (dotted)
results, which represents the rhythm from which the polarity effect
is eliminated. The pattern of the "pure" rhythm (dotted) , after the
elimination of the polarity effect, bears a strong resemblance to that
in Figure 17.

It has been shown in the above figures that the series of com-
ponents of Figures 15, 16, and 17 give resultants which are very
similar to the protoplasmic rhythm experimentally deterinined.
These components postulated for the protoplasmic rhythm are pre-
sumed not to be purely abstract ones, but to have a real meaning,
each functioning separately as a physical entity in the vital system.
The motive force measured is regarded as an actiial resultant of
intraplasmic interference of various simple harmonic component
forces and, in part, of polarity. This concept necessarily implies the
co-existence of different frequencies and amplitudes in the motor
mechanism of one and the same plasmodium. Dynamoplasmogram

234

The Structure of Protoplasm

patterns may all be different, but the essential nature of the proto-
plasmic rhythm must be the same in all. The basic character of the
protoplasmic rhythm is also a simple harmonic function of time.
But the protoplasm is not driven by a single rhythm; it is driven
by a complexity of rhythms; in short, a plasmodium is polyrhythmic.

IX. DYNAMICAL ASPECTS OF INTRAPLASMIC CONDITIONS

It has been stated in the foregoing section, that the motive force
responsible for protoplasmic streaming is operated, not by a single

Fig. 17.

rhythm, but by a complexity of rhythms having unequal frequencies
and magnitudes. It must be noted, however, that the combination
of the components, which represent a definite portion of the dynamo-
plasmogram, may not represent other portions beyond the limit of
time scheme under consideration. Lack of accuracy of the analysis
is one reason for this. No general method of analysis is applicable
to the curves with which we are mostly concerned. But even if one
could analyze a limited portion of a dynamoplasmogram into an exact
series of components, the combination of the same components would
not represent the same dynamoplasmogram unlimitedly.

Physical Asyects of Protoplasmic Streaming 235

An important reason why the components postulated for one
portion of the curve can not represent other portions seems to be the
instability of the components. This means that the components them-
selves are not kept constant during an experiment; in other words, a
dynamoplasmogram contains variable components. Even when the
extrinsic factors are kept constant, the intrinsic factors of the proto-
plasm change spontaneously. For instance, as the protoplasm spreads
during an experiment, the surface-volume ratio of the plasmodium
may change; this will affect intraplasmic diffusion involved in many
physiological processes, especially in respiration. Change in intra-
plasmic diffusion is but one of the factors which possibly alter the
components of the rhythmic motive force and so restrict the analysis
to that portion of the curve which represents the period of an experi-
ment.

Spontaneous change in polarity, no matter what the real physico-
chemical meaning of polarity may be, also disturbs the dynamoplas-
mogram pattern. Polarity factors are also in part responsible for the
fact that components postulated for one portion of a dynamoplasmo-
gram can not represent other portions of the same dynamoplasmo-
gram beyond the limit of the time scheme. The meaning of polarity,
physically or chemically considered, is not clear, but it may be
considered as an entity apart from rhythm. Watanabe, Kodati, and
Kinoshita (1937) found that the anterior portion, in the direction
toward which the plasmodium advances as a whole, has always a
higher electric potential than the posterior part, although potential
difference between the anterior and posterior parts show a close
correlation with the rhythmic reversal of flow. It is probable that
the potential difference is, in some way, connected with the polar
nature of the motive force, if it is not the direct cause of the polarity
of the motive force.

To sum up, it is obvious that the rhythm of the motive force is
modified, to a greater or lesser extent, by internal factors under
constant external conditions.

The waves shown in Figure 9 change their amplitude and form,
but the periods of increased and decreased amplitude continue for
an irregular duration of time. There is no regularity, either in
regard to the wave form or to the change in amplitude. Nevertheless,
the basic rhythm must operate according to a regular rhythmic pat-
tern. Such a curve is regarded as containing many subordinate
components, the frequencies of which are not in simple ratios to one

236 The Structure of Protoplasm

another. Furthermore, all components do not necessarily remain
constant in frequency and amplitude, but can change spontaneously.

In Figure 7 the waxing and waning periods of the curve are
repeated without the details of the waves. The envelope, or imaginary
margin of the curve, is not periodic. Such is also the case in Figure 8.
If there are only two constant component waves, whose periods differ
slightly from each other, the envelope of the resultant wave is
periodic, no matter whether the frequencies of the two components
are commensurable or not. Furthermore, the resultant waves at
the waning periods do not exhibit such irregular and ragged pattern
as revealed by the curves shown in Figures 7 and 8.

It is probably true that the major change in amplitude is due
to the same principle as that of beats, as stated in section VIII. But
actually, there must exist more subordinate components which
would explain the irregular pattern of the waning periods insofar
as they diverge from the regular beat wave pattern. The wave pat-
terns of Figure 7 are, therefore, a compound rhythm consisting of
two main components having slightly different frequencies and many
subordinate components of small wave lengths. All these compon-
ents may be variable to a certain extent. The situation in Figure 8
is also similar to that of Figure 7.

The general situation of the intraplasmic force is really not
simple. It is a necessary conclusion of the analytical study of the
dynamoplasmogram that there must exist components having not
only slightly different frequencies, but also very divergent ones,
namely, frequencies twice, three times, and perhaps, even larger than
that of the lowest (fundamental) frequency. The course of reasoning
is rather abstract, yet without the concept of polyrhythm, it will be
impossible to explain the various types of dynamoplasmogram pat-
terns. As stated in this section, the components are unstable; they
may not remain constant during a period of time. But the com-
ponents, even though they are variable, must have an identity of
their own. One and the same plasmodium must produce many fre-
quencies of rhythm simultaneously.

The next important question with which we are confronted is
naturally the origin of the different components. The concept of
polyrhythm must be harmonized with the actual mechanism and
structure of protoplasm.

Mechanical engineers often study the vibration of machines by
separating it into a series of components by analytical methods.
Since each part of a machine has its own frequency of vibration, the

Physical Aspects of Protoplasmic Streaming 237

analytical method is a very helpful means of learning from what
part of the machine the vibrations come.

In the case of protoplasm, we do not know in what parts of the
Plasmodium the components having different frequencies originate.
Each component postulated for the motive force represents simply
the resultant of a group of waves having that frequency of different
amplitudes and phases generated in any locus of the Plasmodium.
Components having different frequencies may come from different
loci of one and the same plasmodium. It is, however, surprising that
components having not only slightly different frequencies, but very
divergent ones are operative in the same protoplasmic system.

A single string, when it vibrates, produces a series of "over-
tones" simultaneously, the frequencies of which are 2X, 3X, 4X, and
successive aliquot parts of the fundamental vibration. Whether
divergent frequencies of components of the protoplasmic force have
structurally or physiologically a common origin, just as in the
case of a series of overtones from the vibration of one and the same
string, or whether divergency of rhythm is attributable to struc-
tural heterogeneity of the protoplasm, is a problem for the future.

X. ANALOGY AS A MEANS OF ILLUSTRATION

An attempt to compare a complex mechanism, such as a Plas-
modium, which is composed of many unknown factors, with a simpli-
fied model may seem to be misleading. Nevertheless, an illustration
by means of a model will be helpful, insofar as it is concerned with
an explanation of the physical principle which governs the complex
living system, just as it governs the simple non-living model.

The first problem in construction for this analogy is to make a
model which, in gross form, resembles the plasmodium with its two
parts connected by a single strand of protoplasm. Figure 18 shows
a double chamber containing two water reservoirs connected by a
single tube. These two reservoirs are comparable to the two blobs of
protoplasm in the double chamber (cf . Fig. 5) . The bottom of each
of these two reservoirs can be moved up and down by means of a
pin-and-slot device. As there is a difference h between the two water
levels in reservoir a and b, the water has a tendency to flow from a
through the tube to b. The motive force of the movement of the
water through this connecting tube is expressed in terms of the
dijference in pressure (difference h in level) between the two
reservoirs, and not the absolute pressure H at the height of the

238

The Structure of Protoplasm

connecting tube. In order to prevent a flow from a to h, additional
air-pressure must be applied to compartment B. This additional
pressure is the same as the pressure difference corresponding to the
level difference h between the two reservoirs. Like this model, the

Fig. 18.

balance-pressure of protoplasm is not concerned with the absolute
inner pressure of a plasmodium, but with the difference in pressure
between the two parts.

When water in the connecting tube of the model is kept motion-
less, because it is opposed by a balance-pressure, the pressure along
the horizontal tube is the same as the hydrostatic pressure corres-
ponding to the height of the water level of reservoir a above the tube
(H in Fig. 18) . When, however, the balance-pressure is released and
water begins to flow, the inner pressure of the tube varies in accord-
ance with its horizontal position between a and b. The farther the
position of the tube is from reservoir a, the lower is the inner pres-
sure. This is true because the potential energy, represented by the
difference in level, must in part be expended in giving velocity to the
water (velocity head) and also, in overcoming the viscosity (resist-
ance head) . This is a well-known fact in hydraulics, and must also
in part be true in the protoplasmic system.

Physical Aspects of Protoplasmic Streaming 239

Using the same model, the case will now be considered in which
the two cranks are turned slowly with the same angular velocity
(change in pressure due to positive and negative acceleration of
water is neglected) , and the air-pressure in compartment B is con-
stantly adjusted so as to stop the movement of water in the connecting
tube. Since there is no exchange of water between the two reser-
voirs under such circumstances, their levels show motion which is
exactly the same as that of the pistons; this is simple harmonic
motion. The combination, either addition or subtraction, of two
simple harmonic motions having the same frequency gives rise to
another simple harmonic motion of the same frequency. This being
so, the balance-pressure which holds the water in the tube at a stand-
still and is equal to the pressure that is equivalent to the difference
between the two levels, can be expressed by a simple harmonic
function of time. This holds true, no matter what the phases and
amplitudes are, provided the angular velocities of the two cranks
remain the same.

The next case to be considered is that in which the two cranks
have slightly different angular velocities. The motion of the pistons
is of a simple harmonic nature, but each motion has different frequen-
cies. At some point in the time scheme, the pulsation of the pistons will
be in the same place, but the next moment the phase of the pulsation
in one reservoir will have advanced beyond that of the other. After
a certain period of time, periodic movement of the pistons in both
a and b will attain phases which directly oppose each other.

If, when the two pistons pulsate with slightly different fre-
quencies, one plots the change in balance-pressure of the model
against time, there will result a curve which reveals a pattern with
alternating periods of increase and decrease in amplitude exactly
as in the case of beats (cf. Fig. 14) . When the pulsation of the two
pistons are in the opposite phases, the amplitude of the curve will
reach its maximum (waxing period) . When the two pulsating
motions are in the same phase, the amplitude of the curve is at its
minimum (waning period) .

It is likely, though not necessarily so, that in one plasmodium,
when its dynamoplasmogram reveals a wave pattern resembling beats
(cf . Figs. 7 and 8) , the rhythmic frequency of the protoplasm in one
compartment is not the same as the frequency of that in the other,
just as in the case of the two reservoirs, when each has a different
pulsating frequency.

240

The Structure of Protoplasm

Early in this experimental work, the plasmodium was recognized
as polyrhythmic. Each reservoir in the model shown in Figure 18
has only a single frequency of rhythm, whereas, the motive force
generated in the protoplasm lying in one compartment will be of a
number of different frequencies existing simultaneously. At this
point, the analogy drawn between protoplasm and model ceases to
hold. In order to make the analogy hold for the polyrhythmic fea-

FiG. 19.

ture of a plasmodium, a second model must be constructed. Such a
model is shown in Figure 19. Here we have a double chamber con-
taining a two-reservoir system connected by a single tube like that
in Figure 18, but here each reservoir has an arbitrary number of
pulsating pistons of various sizes.

Let us suppose that a balance-pressure is applied to compartment
B, so as to keep the water in the connecting tube at a standstill.
When all the cranks of these pistons rotate slowly with the same
angular velocity, the levels of the two reservoirs undergo simple
harmonic motion, since any number of simple harmonic motions of
the same frequency is always combined, regardless of amplitudes and
phases, into one simple harmonic motion of the same frequency. If,
however, some of the cranks turn with different angular velocities,
the two levels no longer show simple harmonic motion. The differ-
ence, h, between the two levels to which the pulsation of each small

Physical Aspects of Protoplasmic Streaming 241

piston with its individual frequency makes its own contribution, is
not a simple harmonic function of time. The balance-pressure
under such conditions, when plotted against time, will produce wave
patterns similar to those revealed by dynamoplasmograms. The
wave patterns depend upon pulsating frequencies and phases of
these individual pistons, and also upon the capacity of a stroke of
each piston which represents the amplitude of that component.

It is self-evident that in a plasmodium the mechanism responsible
for protoplasmic flow has nothing to do with a pulsating mechanism
of pistons. The above models are presented as possible imaginary
machines which will produce, when suitably operated, a wave pat-
tern of the balance-pressure similar to that of the dynamoplasmo-
gram. For simplicity's sake, the factor of polarity, which is involved
to a greater or lesser extent in a plasmodium, is not taken account
of in these models. But the concept of intraplasviic interference of
component rhythms will be understood by such simple models.
Only as to the rhythmic character of the motive force are slime mold
protoplasm and the models illustrated in the above figures compar-
able.

XI. REACTION OF PROTOPLASM AS REVEALED BY THE
DYNAMOPLASMOGRAM

The experimental data obtained from the present investigation
are all concerned with normal protoplasm. The results of the experi-
ments show that the dynamoplasmograms are of divergent patterns
and change from time to time in the same plasmodium under the
same external conditions.

Despite variability in the pattern of the normal dynamoplasmo-
gram, it reveals much about the physiological conditions of patho-
logical protoplasm. Plasmodia treated with certain artificial agents,
such as anesthetics — this can be done by using a special double
chamber having four taps — show very peculiar dynamoplasmogram
patterns. Change in amplitude, wave length and wave form is often
characteristic of the treatment. Just as the heart specialist can learn
from the pattern of electrocardiograms more about the heart and
circulatory system than he can determine by any other method of
diagnosis, so the protoplasmologist will learn more about the proto-
plasmic mechanism from pathological patterns of dynamoplasmo-
grams than he can determine by microscopic observation. A better
understanding of the general characteristics of protoplasmic reac-
tions is to be had from a comparative study of dynamoplasmograms.

242 The Structure of Protoplasm

Especially helpful in an analysis of the effect of an external agent on
protoplasmic reactions is a study of two dynamoplasmograms, one
obtained from a plasmodium which had been treated in entirety by
an external agent, and the other obtained from a plasmodium of
which only one-half was subjected to the same agent. The latter con-
dition is easily accomplished by using the double chamber, the exter-
nal agent being applied only to one compartment and therefore, only
to one-half of the plasmodium. As a result, the plasmodium under-
goes a modification in polarity. The dynamoplasmograms plotted
from this, show that very often the whole wave is shifted to one side
of the base line without crossing it, but the rhythm continues.

Phenomena known as taxis, such as chemotaxis, phototaxis,
thermotaxis, rheotaxis, galvanotaxis, etc., have been studied in
slime molds by many research workers since the last century. All
taxic movements, no matter what kind they may be, result from the
polar nature of the motive force. A change in polarity due to
unequal distribution of external factors can be very exactly por-
trayed on the dynamoplasmogram on the quantitative basis. The
half-treatment-dynaTnoplasviogram offers interesting problems when
studied in comparison to the whole-treatment-dynamoplasmogram.

XII. RHYTHMICITY AS AN ATTRIBUTE OF PROTOPLASMIC ACTIVITY

It was stated that the rhythmic motive force does not disappear
even though the movement is arrested for a period of time. From
this fact one can assume that the displacement of protoplasm by
streaming can play no part in the causal chain of the mechanism of
reversal. Rhythmic mechanism has, therefore, nothing to do with
the relation of action and reaction brought about by the flowing
of protoplasm. Thus, the views which postulate that a possible
physical or chemical change accompanying the change in distribution
of protoplasm is the cause of the reversal may be discarded in the
light of the above findings. Rhythm is a more deep-seated attribute
of protoplasm.

The rhythmic activity of protoplasm discussed in the present
paper is concerned only with slime molds. It is, however, a question,
and an important one, whether or not polyrhythmic nature is the
specific attribute of slime-mold protoplasm or whether it is essentially
universal and, therefore, occurs in other protoplasmic systems. The
fact that a visible form of movement does not show periodicity can
not offer any cogency to the argument that the basic motor mechan-

Physical Aspects of Protoplasviic Streaming 243

ism is not rhythmic. If the rhythmic activity of each structural unit
be of a random nature both in regard to the phase and direction,
periodicity in the total mass will statistically tend to be canceled out.
Rhythm will be observable only when each unit cooperates or syn-
chronizes to a certain degree.

Some Euglenae, e. g. E. deses, exhibit a characteristic change in
shape and attendant streaming of the endoplasm, which is generally
known as "euglenoid movement" or "Metabolie." This is a sort of
amoeboid movement in the sense that the streaming of protoplasm
involves a change in shape. Under normal conditions the movement
is quite irregular and with no indication of rhythm. It is, however,
very interesting to point out that when some Euglenae are trans-
ferred into acid solutions or subjected to other unfavorable condi-
tions, the movement soon begins to show almost perfect periodicity
for a considerable time before death takes place (Kamiya, 1939) .
The rhythm is in this case pathological, yet nevertheless, it does show
that a periodic mechanism is inherent in an organism which does not
exhibit any periodic movement whatever under normal conditions.

In order to explain the streaming movements of protoplasm,
known as rotation and circulation, which observably are not
rhythmic, it may be necessary, regardless of the nature of motive
forces, to postulate a rhythmic cycle in the motor mechanism.
Indeed, rhythmic mechanism may exist in protoplasmic function, but
is concealed behind the non-rhythmic resultant. There is no reason
to believe that the mechanism responsible for other forms of rhythm,
such as ciliary movement, is essentially different from that respon-
sible for protoplasmic streaming. It is, however, a problem left for
the future, how far rhythmic mechanism is a universal attribute
of living matter.

It is true that the energy used for protoplasmic streaming is
f irnished ultimately by respiration. But the rhythmic output of
meclianical energy (motive force) does not necessarily imply a
rhythmic pattern in respiratory function. A periodic mechanism may
be ascribed to any definite part in the total chain of intraplasmic
reactions.

According to the view advanced by Professor Seifriz (1937,
1938) , streaming in slime molds is caused by a "rhythmic contrac-
tility" of the protoplasm, due probably to the shortening of protein
fibers by molecular folding or, possibly, helical contraction. Although
it is not within the scope of the present paper to indulge in specu-

244 The Structure of Protoplasm

lation on the concrete meaning of the polyrhythmic functioning of
protoplasm on a physical or chemical basis, it may be added that the
difference in frequency of component rhythms may mean a differ-
ence in frequency of the contracting and relaxing rhythm of proto-
plasm.

I wish to express to Professor William Seifriz my sincere appre-
ciation of his kindness in extending to me the courtesies of his
laboratory. I was encouraged by his constant aid and cooperation
during the course of these experiments. I am indebted to Professor
Jacob R. Schramm for permission to work in the Botanical Labora-
tory of the University of Pennsylvania. Finally, it is my most pleas-
ant duty to acknowledge with thanks the aid granted me by Mrs.
Curtin Winsor, who graciously made it possible for me to carry out
the present work by supporting my research.

LITERATURE CITED

Balbach. H. 1936. Uber Plasmadifferenzierungen in Plasmodien der Schleim-

pilze. Protoplasma 26:161-180.
Camp W G. 1936. A method of cultivating myxomycete plasmodia. Bui. Torrey

Bot. Club 63:205-210.
1937. The structure and activities of myxomycete plasmodia. Bui.

Torrey Bot. Club 64:307-335.
EwART, A. J. 1903. On the physics and physiology of protoplasmic streaming in

plants. Oxford.
Heilbrunn, L. V. 1937. An outline of general physiology. Philadelphia.
Hilton A E. 1908. On the cause of reversing currents in the plasmodia of

' Mycetozoa. Jour. Quekett Micr. Club 10: 263-270.
Kamiya N. 1939. Die Rhythmik des metabolischen Formwechsels der Euglenen.

Ber. d. Deutsch. Bot. Ges. 57:231-240.

. 1940. Control of protoplasmic streaming. Science 92:462-463.

Miller, D. C. 1916. The science of musical sounds. New York.

Pease, D. C. 1940. Hydrostatic pressure efJects upon protoplasmic streaming

in Plasmodium. Jour. Cell, and Comp. Physiol. 16:361-369.
Pfeiffer, H. 1936. Further tests of the elasticity of protoplasm. Physics 7:302-305.
. 1937. Experimental researches on the non-Newtonian nature of

protoplasm. Cytologia Fujii jub. vol. 701-710.
Seifriz, W. 1937. A theory of protoplasmic streaming. Science 86:397-398.
. 1938. Recent contributions to the theory of protoplasmic structure.

Science 88:21-25.
VouK. V. 1910. Untersuchungen iiber die Bewegungen der Plasmodien. I. Teil.

Die Rhythmik der Protoplasmastromung. Sitzungsber. d. kais. Akad. d.

Wissensch. Wien, math.-naturwiss. Kl. iI9: 853-876.
. 1913. Untersuchungen iiber die Bewegungen der Plasmodien. II.

Teil. Studien iiber die Protoplasmastromung. Denkschr. d. kais. Akad.

Wissensch. Wien, math.-naturwiss. Kl. 88: 653-692.
Watanabe, a., Kodati, M., itnd Kinoshita, S. 1937. Uber die Beziehung zwischen

der Protoplasmastromung und den elektrischen Potentialveranderungen

by Myxomyceten. Bot. Mag. (Tokyo) 5i: 337-349.
Wedmore, E. B. 1895. A graphical method of analyzing harmonic curves. Elec-
trician 35:512-513.

SOME PHYSICAL PROPERTIES OF PROTOPLASM
AND THEIR BEARING ON STRUCTURE

William Seifriz

The demonstrable physical properties of protoplasm indicate its
hidden structure. In the introduction to this Monograph, I dealt
with bonds between linear molecules. Neither the bonds nor the
linear molecules which they tie together is visible, but their pres-
ence is required to satisfy the needs of a fluid and elastic system.

There are many qualities of living matter which owe their
existence to specific structural features. One of these is thixotropic
behavior.

THIXOTROPY

Professor Freundlich, in his contribution to this Monograph, has
discussed thixotropy from a purely physical and chemical point of
view. He would have added some biological applications had he
lived to complete the chapter. I shall, therefore, do this for him,
and then tell in more detail of one application to which I have of
late devoted considerable attention.

Thixotropy, as originally described by Schalek and Szegvary^,
referred to the sudden collapse of a gel due to mechanical agitation.
A simple example of this in living matter is the commonly observed
liquefaction of protoplasm when disturbed by a needle. Mechanical
agitation is, however, just as likely to produce solidification of proto-
plasm as liquefaction, for the most common characteristic of death
is coagulation. I shall later suggest that both phenomena, when
they occur suddenly, may be included under the term thixotropic
behavior.

More closely resembling the original observation of Schalek and
Szegvary is the complete and sudden collapse of one-celled organ-
isms when severely punctured by a needle. Protozoa and the eggs
of algae often disintegrate instantaneously on being quickly punc-
tured with a needle; they literally blow up. Very striking also is
the complete collapse of a dividing echinoderm egg in midmitosis

'Kolloid Zeitschr., 32, 318: 33, 326. 1923.

[2451

246 The Structure of Protoplasm

when subjected to pressure. No vestige of the structural features
of the preceding karyokinetic figure remains.

Among his many previous references to physiological examples
of thixotropy, Freundlich- cites an experiment by Faure-Fremiet^,
who found that mechanical agitation would change an Amoeba
from the firm condition of its resting state to the fluid condition of
its active state. Peterfi^ described the liquefaction of the gelled
protoplasm of neuroblasts caused by the movement of a needle.

It is a pathological case of thixotropic behavior, involving not
collapse but the rapid setting of protoplasm, of which I wish to speak
in detail, and in so doing I should like to broaden the concept of
thixotropy so as to include the instantaneous and spontaneous set-
ting of a colloidal system. A truly thixotropic gel always under-
goes rapid gelation after collapse. As both phenomena are charac-
teristic of thixotropic behavior, there is no reason why emphasis
should be laid upon one any more than upon the other, especially,
since the instantaneous setting is as remarkable as the sudden
collapse by agitation.

The thixotropic collapse of a gel due to agitation is a form of
liquefaction which is characterized by the suddenness with which it
takes place. Just so is the instantaneous setting of a colloidal system
a form of gelation or gelatinization characterized by the rapidity
with which it is accomplished. The occurrence of either of these
phenomena in living matter may be spontaneous or the result of a
pathological condition. As I shall refer primarily to thixotropic
setting, it may be well to define it as the instantaneous and readily
reversible gelatinization of protoplasm or other colloidal system.

Much of the protoplasm of slime molds is in a state of continuous
flow. The movement is first in one direction and then in the other,
reversal occurring every forty-five to fifty seconds. Any pronounced
change in the rhythmical flow of the protoplasm is an indication of
a change in physiological condition.

If subcultures of slime molds, in particular, Physarum poly-
cephalum, are placed in a small gas chamber and subjected to certain
anesthetic agents, the streaming protoplasm comes to a sudden stop.
Recovery takes place within a minute or two, and there is no
observable injury. Three anesthetics accomplish this perfectly: car-
bon dioxide, cyclopropane, and chloroform (Fig. 1).

^ Chemisch Weekblaad (Amsterdam), 32, 739. 1935.
' Trans. Faraday Soc, 26, 779. 1930.
* Arch. exp. Zellforsch., 4, 143. 1927.

Some Physical Properties oj Protoplasm

247

The average time for the anesthesia of Myxomycetes by carbon
dioxide, administered with an equal proportion of oxygen, is one
minute, with recovery in the same time. A proportion of three parts
carbon dioxide and one part oxygen produces anesthesia in half a
minute, and permits recovery within two minutes. The time for
both anesthesia and recovery varies with the concentration and

Fig. 1. A Plasmodium anesthetized with carbon dioxide: left, the normal
protoplasm at the moment of reversal in flow; right, the anesthetized protoplasm.
In the latter the protoplasm is quiet and remains so for several minutes, fol-
lowed by a return to normal; note absence of any injury or observable pathologi-
cal change in the anesthetized protoplasm.

method of application of the gas. A substantial dose quickly given
will put the Plasmodium under within twenty seconds, and may
delay recovery for from eight to ten minutes. As low a concentra-
tion as 10 per cent of carbon dioxide is sufficient to produce anes-
thesia for brief periods.

Of general biological and medical interest are the unexpected
results that carbon dioxide proved to be the most successful anes-
thetic agent for Myxomycetes though that gas is one of the least
satisfactory for man; that ether, the most used among surgical
anesthetic agents, proved very unsatisfactory for slime molds; that
cyclopropane and chloroform are as successful in inducing the
anesthesia of very primitive forms of living matter as they are in
anesthetizing higher forms of life; and that ethylene, one of the

248 The Structure of Protoplasm

newer and best anesthetics for man, had no observable effect what-
ever on shme molds."" Ethylene not only failed to produce anesthesia
in slime molds, but it stimulated their growth, an unexpected result
which may be due to increased respiration.

That feature of the anesthesia of slime mold protoplasm by
carbon dioxide, cyclopropane, and chloroform which is of chief
interest to us at the moment is the instantaneous cessation of
streaming. There is little warning, occasionally a momentary tremor,
and then a sudden and complete stopping of flow. On removal of the
anesthetic agent there is resumption of normal activity, with no
observable damage.

It is only necessary to observe the sudden cessation of proto-
plasmic movement in a slime mold due to the application of a suit-
able anesthetic agent to be convinced that the protoplasm has under-
gone a pronounced change in its physical as well as its physiological
state, yet no change is evident other than a stopping of flow.

The instantaneous cessation of protoplasmic streaming caused
by carbon dioxide appeared to involve a setting or gelatinization of
the protoplasm. It also seemed likely that the failure of other
anesthetic agents to put a sudden stop to protoplasmic streaming
was an indication that no gelation had taken place, but there was
no direct proof of either deduction.

By means of an ingenious technique involving the application of
pressure, Kamiya is able to oppose and control protoplasmic flow
(pp. 205, 206) . If protoplasm is quiet, due to anesthesia, pressure
externally applied should cause it to move if it is in a liquid state,
and fail to move it if it is firm. Such proved to be the case. Proto-
plasm under ether, which produces only partial anesthesia of slime
molds, is readily moved by externally applied pressure, whereas
protoplasm under carbon dioxide cannot be made to flow by pres-
sure. Obviously then, under carbon dioxide the protoplasm is firm;
under ether it is fluid.

The foregoing facts form the basis of a theory of anesthesia,
applicable to those cases where suspended animation is brought
about quickly, as with carbon dioxide, cyclopropane, and chloro-
form when applied to slime mold protoplasm. The theory states that
anesthesia is due to the thixotropic setting of protoplasm. In the
case of higher vertebrates only nerves would be affected, whereas,

' Anesthesiology, 2, 300. 1941.

Some Physical Properties of Protoplasm 249

in slime molds all the protoplasm sets thixotropically when anes-
thetized.

A neurologist, on observing the instantaneous cessation of move-
ment of slime mold protoplasm when anesthetized by carbon
dioxide, characterized it as paralysis. I find the analogy interesting,
though prefer not using the term simply because less is known of
the mechanism of paralysis than of anesthesia. It may be true,
however, that paralysis, like anesthesia, is due to a thixotropic set-
ting of the protoplasm of nerve fibers.

Thixotropic gels possess large quantities of water, often with less
than 1 per cent of solid matter. A germanium or cadmium gel may
have as much as 99.8 per cent water. The jelly-fish contains but 2 or
3 per cent solid matter, yet maintains its form. Such systems must
be built of long and flexible fibers in order to set into firm and
elastic gels. Fibers may contact each other at greater distances than
do spheres, therefore, fewer units are needed to build a structure
of given volume.

Further evidence of a very open structure in certain jellies is
the transparency of some of them. Soap jellies are quite clear. This
is possible because of an open three-dimensional lattice built of long
fibrous units. The sudden locking of such intermeshing fibers
accomplishes thixotropic setting; and their release by mechanical
agitation is thixotropic collapse. In protoplasm, these linear units
are the familiar polypeptide chains.

ELASTICITY

That protoplasm is elastic is so obvious to many workers that one
wonders why it should ever have been questioned by others. The
earliest students of protoplasm knew it to be elastic. Additional
evidence of recent date comes from Northen" and Norris".

Proof of the elasticity of protoplasm is to be had in a number of
ways. The simplest is to stretch it with microneedles (Fig. 2) . On
release after stretching, the protoplasm usually shows pronounced
recoil, indicating high elastic qualities.

The elasticity of protoplasm may be demonstrated in another and
rather novel way, by inserting a metal particle and attracting it
electromagnetically. On elimination of the electromagnetic field the
particle retraces its path due to the elasticity of the protoplasm.
Evidence that the elastic qualities of protoplasm are not pathological

Protoplasma 31, 1-8. 1938.

J. Cell, and Comp. Physiol., 16, 313. 1940.

250

The Structure of Protoplasm

W^

is to be had in the migration of fibroblasts shown in moving pictures
by Dr. Warren Lewis. As the fibroblasts move, they form long
streamers attached to the substratum. These are later torn loose, or
broken, and then snap and recoil rapidly.

The elasticity of protoplasm varies, just as do its other physical

properties such as viscosity. Highly
viscous protoplasm may be poorly
elastic, resembling plastic paste,
whereas thin protoplasm is sometimes
capable of great extensibility, with
pronounced recoil. From the fore-
going it could be assumed that vis-
cosity and elasticity have a direct
bearing upon each other, but this is
not true; they are independent vari-
ables.

Very thin protoplasm is often
thought to be inelastic because the
elasticity cannot be readily deter-
mined.

That thin solutions may be highly
elastic is capable of nice demonstra-
tion in soap solutions.'' Sodium stea-
rate will, under favorable conditions,
exhibit pronounced elasticity though
the solution is but twice as viscous
as water. The elastic quality of so
thin a solution inay be demonstrated
by twirling the flask containing it and
observing the manner in which the
solution comes to rest. A wholly inelastic liquid, such as water or
glycerin, will come to rest slowly and show no return, but an elastic
solution will slow down, come to a sudden stop, and then return a
part of the circumference of the flask. Were it possible to apply some
such method to protoplasm, I feel confident that the thinnest proto-
plasm would show elastic qualities just as does the thicker protoplasm
in which elasticity is so readily demonsti'ated.

The effects of salts on the elasticity of protoplasm have been
determined; calcium increases the extensibility of protoplasm.

Fig. 2. The stretching of proto-
plasm with a microneedle. On
tearing, the protoplasm rebounds,
exhibiting high extensibility, ten-
sile strength, and recoil.

' Third Colloid Symposium Mono., New York. 1925.

Some Physical Properties of Protoplasm 251

sodium diminishes it, and magnesium has no effect at all." The
following series was obtained:

Ca > Sr > Mg > K > Li > Na .

This is in keeping with the known dispersing effect of sodium and
the aggregating effect of calcium on protoplasm.

Theories on the elasticity of gels have been numerous, one of the
oldest being based on surface tension. It was maintained that the
elastic qualities of protoplasm are due to surface forces, those same
forces which cause a free liquid droplet to round up into a sphere.
This is very unlikely in protoplasm, or in any gel, for the elasticity
of a glutinous mass like protoplasm or gelatin is not restricted to
the surface. This is obvious when a jelly is torn. Furthermore, the
elasticity of protoplasm, measured in terms of extensibility, far
exceeds that which could be accounted for on the basis of surface
tension. When, as is usually true, the surface layer of protoplasm
proves to be more elastic than the interior, this is due not to surface
tension, but to a denser knitting of the fibrous molecules at the
surface, i. e., to the same structural features which are responsible
for elasticity throughout the living substance.

Change in surface tension entered into a theory of the structure
of elastic gels. It was assumed that gelatin is a fine emulsion, and
that the physical forces responsible for elastic qualities of gelatin
resided in the interface between the dispersion medium and the
dispersed phase. By a neat mathematical analysis Hatschek^"
demonstrated that surface tension cannot possibly be responsible
for the elastic qualities of jellies. If the interfacial tension between
two liquids is responsible for the elasticity of the system, then the
magnitude of this tension is dependent upon the amount of surface.
Consequently, the product of the increase in surface, due to stretch-
ing, and a constant representing the degree of interfacial tension and
extent of surface, must equal the work done in producing elongation.
If stress is plotted against increase in length, the stress, in the
theoretical curve, increases until elongation reaches two and a half
times the original length, and then decreases, which is experimentally
untrue. If the foregoing theoretical curve, plotted on the assump-
tion that elastic qualities reside in surface forces, is compared with
experimental curves of elongations of gelatin and rubber, the two
are found to differ greatly in general character. Obviously, elasticity

'Rheology 2. 263-270. 1931.
'"Trans. Faraday Soc, 12, 17. 1917.

252 The Structure of Protoplasm

is not a surface or interfacial phenomenon, but a property of the gel
as a whole, the basic mechanism of which lies in structural con-
tinuity.

There then came into colloidal chemistry the stimulating concept
of a brush-heap of molecular or colloidal fibers. My earliest^^ inter-
pretation of the mechanism of elasticity in protoplasm was that of a
brush-heap, a haphazard distribution of molecular rods (Fig. 2, p.
4). The concept holds well for elasticity, but fails when other
qualities of protoplasm and jellies in general are considered. The
folded protein molecule (Fig. 4, p. 32) best meets all requirements,
and is the molecular form now commonly ascribed to elastic organic
substances and tissues such as albumen, wool, hair, and muscle.

A folded protein chain is the basic mechanism of both the
elastic and contractile qualities of protoplasm and certain other
organic substances. I shall reserve further discussion of it for the
section on contractility, to follow.

Though the folded chain is the most likely molecular form
responsible for the elasticity of gels, there are other possibilities of
which the helix is one. The molecular structure of rubber is thought
to be that of a modified helix.

That a hydrocarbon could be elastic seemed so unlikely that when
it was suggested at the inaugural meeting of the Society of Rheology
that those structural features responsible for the anomalous flow of
gelatin, soaps, and like substances are also responsible for their
non-Newtonian behavior, objection was raised on the grounds that
hydrocarbon oils were found to be anomalous by Hardy, and hydro-
carbons cannot be elastic. Since then, artificial rubber of high
elastic qualities has been made of polymerized hydrocarbons.

An interpretation of the molecular structure of hydrocarbons
such as will account for their elasticity, has been made by Mark^-,
Mack^"', and Meyer.^ ' They assumed the rubber molecule to be a
long chain, or molecular fiber fashioned as a helical spring, which
uncoils on stretching.

In order to account for an 800 per cent stretch in rubber, Mack
devised a mechanism wherein the axis of one molecule is perpen-

" Amer. Naturalist, 60, 124. 1926.

''Mark, H., Physical Chemistry of High Polymeric Systems, N. Y. 1940. (See
also Kautschuk, 6:2. 1930).

'Jour. Amer. Chem. Soc, 56:2757. 1934.

" Meyer, K., Die Hochpolymeren Verbindungen, Leipzig, 1940 (See also Chem.
Revs., 25:137. 1939).

Some Physical Properties of Protoplasvi 253

dicular to that of another to which it is attached. On stretching,
the axes of the molecules are pulled into line with each other.
Elongation is thus doubled. As one spiral coil will permit an exten-
sion of 300 per cent, a stretch of 600 per cent is thus obtained.
This surpasses the usual 400 per cent and approaches the possible
maximum of 800 per cent for rubber.

The source of energy was also given consideration.^'' The residual
valences of double bonds are insufficient because in stretched rubber
they are spaced too far apart. Mack, therefore, assumed that the
source of attraction lies in van der Waals' forces between hydrogen
atoms attached to carbon atoms.

Contraction results, says Meyer^^, because the extended state of
a long chain molecule is thermodynamically a less probable one
than the contracted state. The extended form has only a single
geometrical possibility, whereas the contracted form has many. The
extended chain will thus have a tendency to contract, and this
appears to be true not only for rubber, but for all elastic synthetic
polymers, for muscle, and for protoplasm. Thus we see that deduc-
tions made from considerations of rubber apply to living matter.

Proteins are the substances, and their folded chains the molecular
forms probably responsible for the elasticity of protoplasm. This is
the consensus of opinion, but it has had its opponents. Controversy
over the form of polypeptide molecules entered early into the stereo-
chemistry of proteins. Solid spheres, hollow spheres, chains, fibers,
coils, nets, and lattices have all had their advocates. By the time
most chemists had agreed on the polypeptide chain for proteins, with
side-chain linkages establishing two and three dimensional lattices,
there arose renewed support of the hollow sphere, based primarily
on mathematical analysis. ^^

The chief objection to a hollow sphere is the difficulty of doing
anything with it structurally. If the 32,000 molecular weight protein
is a complete and closed hollow sphere in itself, it would be difficult
to construct 64,000, 96,000, etc., molecules from it, as Svedberg sug-
gests is done in nature. Mechanical properties, such as elasticity,
tensile strength, and the extraordinarily high water holding capacity
of jellies would be equally difficult of interpretation. To attempt to
satisfy the physical properties of protoplasm with a spherical mole-
cule is rather like asking a weaver of cloth to make his fabric of
sand instead of threads.

^Science, 85. 76. 1937.

254 The Structure of Protoplasm

Globular molecules are said to occur in globulins, such as edestin,
and in normal albumen (egg white) , but this is true only in dilute
solutions and only when the molecules are not at surfaces. Further-
more, all yield X-ray photographs when denatured. According to
Astbury^"', the supposed globular protein molecule is a folded
polypeptide chain in which the linear fiber has been thrown into a
series of rings.

In some proteins, the natural molecular configuration is an
extended chain, which may be fully extended as in silk, (3 keratin,
and [3 myosin, or regularly folded in one dimension, as in the a- and
super-contracted forms of keratin and myosin (Fig. 4, p. 32) . When
the chains are fully folded, so-called "globular" proteins result;
among them are egg albumen, haemoglobin, pepsin, and insulin.
These "globular" proteins unfold on denaturation, and the liberated
polypeptide chains may then sometimes be spun into artificial protein
fibers. In the living cell there is probably an unlimited number of
reversible interchanges between various states of folding. From this,
it follows that a spherical protein may be transformed into a fibrous
one; possibly the reverse transformation may also occur. The
former change has been proven experimentally (see Moyer, p. 28,
this Monograph) .

Evidence is predominantly in favor of but one stereochemistry of
the proteins based on the convolutions of the polypeptide chain. It
may be, however, that the proteins fall into two groups, morpho-
logically considered, the soluble proteins of globular molecular
form, such as egg albumen, insulin, and edestin, and the insoluble
proteins with chain molecules, such as silk, hair, and wool. Proto-
plasm will contain both forms; the globular proteins will serve pri-
marily as nutrition and the linear ones for construction. It is to
the chain molecule that the elastic and general structural properties
of protoplasm are due.

CONTRACTILITY

Elasticity and contractility are very similar phenonena, for
elastic substances contract. Consequently, much that is said of the
one is applicable to the other. In the conclusion to this chapter, I
shall emphasize the wide-spread occurrence of rhythmic contrac-
tility in tissues, but it is a special case of which I wish first to speak,
that of the continuous rhythmic pulsation of the protoplasm of
slime molds.

' Nature, 237, 803. 1936.

Some Physical Properties of Protoplasyn 255

The mechanism of protoplasmic flow has long been a subject of
speculation. Change in surface tension, the sine qua non of so many
cellular activities, was once regarded as being responsible for the
streaming of protoplasm. There was also the suggestion that the one-
way shuttle type of flow in the filaments of coenocytes where move-
ment is first in one direction and then in the other, may be due to
hydration at one end and dehydration at the other end; but the
protoplasm flows equally well in both directions, even when fully
submerged.

With the introduction of colloidal chemistry into physiology it
became apparent that many cellular activities had their counterpart
in colloidal systems. There thus arose the suggestion that proto-
plasmic streaming is a cataphoretic migration of particles or the
electroendosmotic flow of an aqueous medium; if either, it is the
latter, for in streaming protoplasm the entire mass of material moves
and not just the suspended particles. In spite of some attempts to
prove that streaming protoplasm is associated with electric poten-
tials, there has been no convincing evidence that the potentials
measured are real in the sense of innate to the protoplasm, and if
real, are the cause rather than the result of the streaming.

One difficulty with most theories advanced in explanation of
protoplasmic movements is the lack of any explanation of a reversal
in the mechanism. The protoplasm of slime molds flows first in one
direction and then in the other. Reversal occurs normally every 45
to 50 seconds, the time of outward flow occupying about five seconds
longer than that of the inward flow, which accounts for the advance-
ment of the Plasmodium. Were a difference in potential responsible
there would have to be a change in polarity every 45 seconds, and
there is no evidence that such a change occurs, except the reversal
in flow.

If time-lapse moving pictures of the streaming protoplasm of the
myxomycete, Physarum polycephalum, are made, a remarkable
pulsation of the plasmodium is revealed. Pictures taken every five
seconds and shown at the usual rate of sixteen a second, which
means a speeding up of eighty times, show the plasmodium under-
going rhythmic contractility and relaxation in perfect synchronism
with the protoplasmic flow (Fig. 3) . The rhythmic contractions and
expansions when thus optically accelerated resemble the pulsations
of a heart. At each contraction and relaxation the direction of proto-
plasmic flow reverses, and with the outward flow the plasmodium
advances slightly.

256 The Structure of Protoplasm

As the rhythmic contraction of the protoplasm is synchronized
with the streaming, it seems Hkely that one is the cause of the other,
from which the only constructive deduction to be made is that the
flow is the result of the rhythmic movement.

Fig 3 The rhythmic contraction of two pseudopods of a shme mold Plas-
modium The top picture in each series is the fully expanded form and the bot-
tom pkture is the fully contracted form. The three left-hand pictures were
ph'to'^rlphed 20 and 2 J seconds apart; and the four right-hand P-^-es ^e-
ohotoeraphed 20 15, and 10 seconds apart. Note that m the left-hand pictures
the expaSed form at the top has a smooth contour, and the contracted form at
the bottom has a wrinkled contour.

Sovie Physical Properties of Protoplasm 257

The mechanism of protoplasmic movement in shme molds is,
then, one of rhythmic contraction and relaxation of the plasmodium,
with a total of 95 seconds for each pulsation, 45 seconds for systole
and 50 seconds for diastole, the additional 5 seconds in time of out-
ward flow account for advancement. Kamiya (p. 211) gives 93.7
seconds for one rhythmic period.

The precise nature of protoplasmic contractility is difficult of
interpretation, for in slime molds it is complicated by the constantly
varying shape of the plasmodium, and by a multiplicity of rhythms
(see Kamiya, p. 234). The latter are the expression of a too little
appreciated property of non-cellular protoplasm, namely, the pres-
ence of distinct regions, each with its own individuality, yet each
contributing to the function of the whole. Moving pictures revealed
three separate rhythms in slime (Figs. 14 to 17) , which number
Kamiya has now greatly increased. He postulates many rhythms
or frequencies of pulsations.

The discovery that a plasmodium is a polyrhythmic system would
at first thought seem to hopelessly complicate the situation, but
actually it is only through a multiplicity of secondary pulsations
that it is possible to account for a number of phenomena. Where the
situation is a simple one, it is not difficult to visualize the mechanism
responsible for protoplasmic flow. If there is a strand connecting
two masses of protoplasm, then the mass at one end contracts and
that at the other end expands, forcing the protoplasm through the
connecting strand. At the end of 45 seconds the second mass under-
goes contraction and the first mass now expands. The protoplasm
then flows in the opposite direction for an average of 50 seconds.
It is probable that in such a case contraction takes the form of a
peristaltic wave which passes along the strand from one end to the
other, and then returns, the process being repeated rhythmically
every three-quarters of a minute.

In a situation such as the foregoing, the outer cortical layer of
the protoplasm may be postulated as the region responsible for con-
tractility and protoplasmic flow; and, with reason, it may be assumed
that this surface layer is a gel, in contrast to the fluid protoplasm
which is in a state of motion within, the one being readily converted
into the other through a sol-gel transformation. Though in full agree-
ment with this viewpoint where there is one general inner region
of liquid and flowing endoplasm, surrounded by an outer firm and
contractile ectoplasm, as in Amoeba, yet I do not believe so simple

258

The Structure of Protoplasm

a mechanism is applicable to a myxomycete plasmodium, for the
following several reasons.

A Plasmodium is an aggregate of many regions, each of which is
a unit in itself. Each contains liquid, flowing protoplasm surrounded
by firm and quiet protoplasm, and each possesses its own rhythm.
Three rhythms were shown in my original moving pictures of Plas-
modia, and no two of the three were pulsating in unison. Were the
mechanism of flow situated in a single protoplasmic part, the peri-
pheral layer, it is hardly likely that there would be a duphcity of
rhythms. One may frequently observe in a plasmodium several

lines of flow which cannot

-e-

->-

/\

/V

possibly be the expression of
a single contractile mechan-
ism. Illustrations of two such
situations are shown in Fig-
ures 4 and 5. Figure 4 is a
diagram of two connecting
strands with arrows which
indicate six possible types of
flow all seen within a few
Such a situation can be adequately accounted for
of several frequencies not in perfect

Fig. 4. Six possible combinations in
direction of flow in two connecting strands.

minutes' time

only on the assumption

synchronism. There is a lag in the time of reversal in one strand

over that in the other, which means that there is more than one

center of contraction in the plasmodium.

Figure 5 illustrates another case where contraction is not cen-
tered in the surface layer, but at a point, or points, some distance
in from the surface.

The six possible types of flow shown in Figure 4 are reproduced
many times over by numerous small individual systems or centers
of activity throughout a plasmodium. In short, there is not one
surface layer of contractile protoplasm, but many secondary centers
of contractility.

A situation that one would have difficulty in interpreting on the
basis of a single contractile mechanism is that involving the flow of
protoplasm in a globule. Were a spherical plasmodium a shell of
firm contractile protoplasm enclosing a core of fluid non-contractile
protoplasm, then uniform contraction of the surface would compress
the inner protoplasm and not produce flow along defined channels;
but if this droplet of protoplasm consists of many individual centers

Some Physical Properties of Protoplasm

259

of activity, each with its own rhythm, then movement from one
part to another can be accounted for.

For these reasons, I regard protoplasmic movement in myxomy-
cetes as not identical to that in Amoeba. There is apparently in
Amoeba one cortical contractile layer and one major line of flow,
whereas, a myxomycete consists of many regions each exhibiting
rhythmic contractility which is not necessarily in perfect synchron-
ism with its neighbors.

The most convincing evidence against a contractile cortical layer
in slime mold is the wrinkled surface of a plasmodium at systole. Con-
traction should shorten and
therefore smooth out the sur-
face of a pseudopod as it is
emptied of protoplasm. The
lowest of the three photo-
graphs in Figure 3 is the
contracted state of the pseu-
dopod, yet its surface is much
wrinkled, whereas, the sur-
face of the expanded proto-
plasmic inass, at the top in
Figure 3, is smooth. Tension
at a surface should tend to

round up, not buckle, the contour of a contracting fluid drop. Wrink-
ling of the surface with a decrease in volume because of loss of proto-
plasm indicates that there has been no change in the total surface
area and therefore, no contractility of the surface layer. This is a
matter of great importance for the whole concept of rhythmic con-
tractility in protoplasm, and is to be explained, I think, by the con-
clusion expressed in the foregoing paragraph, namely, that in slime
molds there is not one single contractile mechanism which resides in
the surface layer, but many independent centers of rhythmic con-
tractility.

The conclusion to which I come is similar to that reached by
Pfeffer, and by Ewart, in relation to another physical force, that of
surface tension, then held to be responsible for protoplasmic flow.
Their deduction is as applicable to the present case as it was to their
own, namely, that the energy responsible for protoplasmic flow is
liberated throughout the living substance and not solely or primarily
at the boundary.

Another important matter which should be emphasized is that

Fig. 5. Simultaneous flow both toward
and away from the periphery of a Plasmo-
dium.

260 The Structure of Protoplasm

fluid protoplasm is contractile. Usually the contracting regions of
a Plasmodium are of high viscosity, but protoplasm need not be so to
exhibit contractility. The significance of this remark is twofold:
It emphasizes that the chief feature of protoplasmic movement is not
a sol-gel transformation but the contractility of protoplasm, and
that organic colloidal solutions have some of the properties of gels
no matter how thin they may be.

The physical differences between streaming protoplasm and quiet
protoplasm may be great, but they are purely physical and relative.
Quiet protoplasm is more viscous than flowing protoplasm; it may
be quite firm. It is usually more resilient, elastic, of higher tensile
strength, and more contractile than flowing protoplasm, but the
latter is by no means devoid of these properties. Fluid protoplasm
is elastic and contractile. The experiment cited on page 250, wherein
an elastic fluid was given a swish in a flask, is a demonstration of
contractility in a fluid system having a viscosity value but twice
that of water. The protoplasm which flows in the capillaries of a
slime mold is less contractile than the surrounding protoplasm
which is responsible for the flow, but it is, nevertheless, capable of
exhibiting contractility.

Before turning from a purely biological discussion of rhythmic
contractility in protoplasm to a physical interpretation of it, I should
like to restate and further support the important concept that a Plas-
modium is an aggregation of centers of activity each with an indi-
viduality all its own. The physiological state of any one center
differs from that of adjoining centers at any given moment very
much as adjoining cells in tissue differ, with this exception, that in
a Plasmodium the difference is less permanent.

Further indication that a slime mold is an assemblage of indi-
vidual units is to be had in the habit plasmodia have of breaking up
into microscopic bits at the time of sclerotium formation, and even
during the plasmodial stage. At sclerotium formation, the drying
Plasmodium fragments into many small sections, the whole resem-
bling an assemblage of cells. De Bary noticed this, and it has recently
been recorded for Hemitrichia vespariuvi by Ruth N. Nauss, who
describes the sclerotium as made of many isolated globules of resting
protoplasm. These subdivisions represent previous centers of activ-
ity which were physiologically distinct at the time the sclerotium
was formed. It is possible that they are uninucleate masses of proto-
plasm. The bits which are morphologically separate units in a
sclerotium cease to be so in a plasmodium where, however, they

Some Physical Properties of Protoplasm 261

retain their physiological individuality, the chief feature of which
is, for us at the moment, the characteristic rhythm each possesses, a
rhythm not necessarily in perfect synchronism with that of other
units.

The foregoing essentially biological discussion of rhythmic con-
tractility in protoplasm leads naturally to a physical interpretation
of the phenomenon. This would be a highly speculative venture were
it not for the great advance made by physical and X-ray chemists in
an understanding of the molecular structure of contractile organic
substances such as silk, wool, and hair. Just what has been accom-
plished in the case of the elasticity of rubber I have already stated
(p. 253) , and I added that as elasticity and contractility are part of the
same general phenomenon, consideration of the physical basis of
these properties of protoplasm would be postponed. We may, then,
now consider the molecular structure most likely responsible for
elastic and contractile properties in certain organic systems, particu-
larly protoplasm.

Protoplasmic contractility has been attributed to a variety of
forces. In the case of muscle, change in surface tension was a widely
accepted theory. But today it is generally conceded that proto-
plasmic contractility is due to the same molecular mechanism as are
the contractile properties of wool and hair, namely the folded poly-
peptide chain.

The elasticity of jellies and the high water-holding capacity of
thixotropic gels led to the general acceptance of a linear structural
unit. Contractile properties have now led to another feature of the
structural unit of elastic organic systems, namely, molecular folding
(Fig. 4, p. 32) . With it our picture is relatively complete.

All proteins are built of polypeptide chains with side chains
joining one molecule to another. Most proteins are crystalline.
Usually, two classes of proteins are recognized, the fibrous and the
non-fibrous or corpuscular ones, but there is so much in favor of a
close relationship between the two, and a small energy difference,
"that it is difficult to avoid the conclusion that fundamentally all
proteins are fibrous."^"

The general molecular pattern of proteins is now well known
(Fig. 4, p. 51, etc.).

Astbury'"^ divides proteins into four types according to their

'• Trans. Faraday Soc, 36, 233. 878. 1940.
"Tabulae Biologicae, J7, 90. 1939.

262 The Structure of Protoplasin

molecular patterns: extended (^-proteins, extended proteins of the
collagen type, folded proteins of the a-type, and folded proteins of
the supercontracted type.

To this scheme of molecular protein types, Astbury^^ and others
had previously added one other important structural feature, that
of the grid, formed by side-chains joining one protein main-chain
to another. If the paper on which the diagram of such a protein
grid is printed is thrown into a series of folds parallel to the side
chains, the a- or contracted protein molecular pattern is obtained.
A parallel alignment of grids results in the formation of a three-
dimensional lattice, or crystal.

It was a haphazardly assembled three-dimensional pattern, in a
general way similar to the symmetrical pattern postulated by Ast-
bury, which the earlier colloidal chemists had in mind when they
wrote of brush-heaps; these later became more orderly in design.

Astbury-"^ states that the contraction of muscle is due to the
supercontraction of the oriented myosin, the protein which is the
principal structural component of muscle. Similar suggestions had
been made by Mark^- and Meyer.^^ And now we may carry the
concept directly to protoplasm and say that the contractihty of
protoplasm is due to the supercontraction of its principal structural
proteins through the folding of molecular fibers symmetrically
aligned and joined one to the other by side chains so as to form a
three-dimensional lattice. As this is a general conclusion of far-
reaching significance in biology, it may be restated as follows: Con-
tractility, wherever it occurs in animate nature, whether in the most
elementary form of protoplasm or in highly specialized tissue, is due
to the shortening of protein fibers by molecular folding. The energy
for this work is supplied by the oxidative processes of the cell.

The physical basis of the mechanism of protoplasmic contrac-
tility has been set forth. The source of the energy is suggested.
There remains the interesting and rather speculative question,
whether the rhythm of protoplasmic movement lies in the chemical
reaction which supplies the energy, or in the mechanism where the
energy is used. If the energy supply is rhythmic, then there must
be a periodically reversible chemical reaction in protoplasm. A
number of autocyclic rhythmic reactions are known to occur in
protoplasm; the glycogen-lactic-acid cycle in muscle is one. A still

20

Fundamentals of Fiber Structure, London. 1933.
Proc. Roy. Soc. London, B 129, 307. 1940.

Sovie Physical Properties of Protoplasm 263

closer analogy is to be found in the rhythm of nerve conduction where
a cycle of oxidation-reduction reactions is assumed to pass along
the nerve.

The occurrence of autocyclic chemical reactions in protoplasm is
not at all unlikely, but it seems more probable that the rhythm of
protoplasmic flow is a visible expression of a rhythmicity which
exists in the vital mechanism and not in the supply of energy. The
living mechanism may show perfect rhythm even though the source
of energy is in no way rhythmic. Just what rhythmic control in a
vital mechanism means is as yet unknown.

In this connection the conclusion of Kamiya is important (p. 242) ,
that "views which postulate a possible physical or chemical change
accompanying the change in distribution of protoplasm as the cause
of the reversal may be discarded," because the rhythmic motive
force does not disappear when the streaming is artificially stopped
for a time.

In closing this chapter, I wish to give some indication of the
widespread occurrence of rhythmic contractility in living matter.
Heart muscles contract rhythmically. The muscles of the diaphragm
are induced to do so. Contractility occurs periodically as peristalsis
in the walls of the intestines, the ureter, the uterus, and the Fallopian
tube. Feeble rhythmic uterine contractility is by some thought
to be continuous. Voluntary muscles are likewise believed to
undergo rhythmic contractility during periods of apparent rest. The
bladder, not functionally contractile, can be experimentally made to
pulsate rhythmically.

Rhythm in the nervous control of organisms has long been known
and has received added confirmation of late in work involving the
electric recording of rhythmic waves from the brain. -^ That a
rhythmic train of events in the nervous system regulates certain
body functions is shown by the rhythmic movement of the respira-
tory muscles which is not intrinsic, as in the case of the heart. Such
induced rhythms are controlled by periodically recurring stimuli
from, the medulla. When the action of a single neuron is obtained,
a clear record of periodic neural impulses is obtained, which recur
in perfect rhythm.--

The references so far, other than that of slime molds, have had
to do with mukicellular organisms, with tissues; but individual cells.

;' Science, 81, 597. 1935.

"'Chemistry and Medicine. Minneapolis, p. 261. 1940.

264 The Structure of Protoplasm

whether of tissue or Protista, pulsate rhythmically. Dr. Warren H.
Lewis-'' has observed the pulsation of a single cell from chick heart
muscle in culture. The pulsation of a heart is the expression of the
many pulsations of its cells which may not be in perfect synchronism
with each other, the heart rhythm being merely the resultant.

Rhythmic movements are common in lower organisms. Kamiya-^
gives an interesting case of it in Euglena. When this Protozoan is
treated with certain reagents, notably acids, it undergoes rhythmic
peristaltic movements. The pulsation is induced, but the capacity
to undergo contractile motions and the rhymicity are innate qualities.

All nature is rhythmic. In animate nature the visible instances
of it are but outward signs of the basic rhythm in protoplasm. The
rhythmic contractility of slime mold protoplasm is then but one
instance of a very wide-spread phenomenon. In my belief, all proto-
plasm is capable of rhythmic contraction, but ordinarily exhibits it
only in those tissues where it is functional.

Some modern workers are inclined to strip protoplasm of the
properties of multicellular organisms, forgetting that the attributes
of higher forms of life exist because they are properties of proto-
plasm. The protoplasm of primitive organisms is not less intricate,
less responsive, nor less organized, than that of higher forms of life.
It is fallacious to hold that primitive protoplasm is devoid of nervous
response because it lacks a nervous system. Protoplasm is itself a
nervous system; the nerves of higher organisms are but protoplasm.
The properties of tissues are the properties of the protoplasm of
which they are made. The rhythmic pulsations of contractile tissues
in higher organisms are but grosser manifestations of the innate
rhythmic contractility of protoplasm wherever found.

ACKNOWLEDGMENT

I wish to express my gratitude to Mr. William Donner and his
daughter, Mrs. Curtin Winsor, and to the National Research Council,
for their kindness in supplying funds which have made much of
this work possible.

" Carnegie Inst. Pub. 363. 1926.

-' Ber. Deut. Bot. Ges., 57, 231. 1939.

SUPPLEMENT

Editor's Note:

When planning the Symposium of which this Monograph is a
record, it was my wish to have representatives from the three major
sciences which, combined, constitute physiology. Physicists, chem-
ists, and biologists were each to make their contribution to the
structure of matter. But unfortunately a number of those who were
the most interested and willing were too far away to attend. This was
particularly true of collaborators among physicists and chemists
abroad.

I attempted to reach our colleagues over-seas, but, out of the
chaos in present-day Europe, replies from only two of them were
received. These two messages are, each in its own way, exceedingly
fine. I am very happy to be able to add them here. The one is from
Professor Kurt H. Meyer of the Ecole de Chimie, Geneva, Switzer-
land, and the other from Professor A. T. Astbury of the Textile
Physics Laboratory, The University, Leeds, England.

The following is a translation of Professor Meyer's German letter,
in which I have made but one change, to convert the personal refer-
ence to an impersonal one.

[265]

PROTEIN AND PROTOPLASMIC STRUCTURE

Kurt H. Meyer

I should like to refer to a theory of protoplasmic structure which
Seifriz and I independently advanced at about the same time. On
the basis of the highly elastic extensibility of Amoebae, erythrocytes,
and other cells, Seifriz, in 1929' concluded that the ultimate struc-
tural units of the living substance are probably linear molecules or
micellae so arranged as to form a framework. We, on the other hand,
in 1928- and 1929'' showed that there are two types of pro-
teins, the fibrillar and the globular ones; the spherical form of the
latter was proved by Svedberg. We demonstrated the undoubted
presence of fibrous molecules, or elongated polypeptide chains, in
silk, sinew, and stretched muscle. From this it followed that "the
living substance is composed of a true network of primary valence
chains which at several points are coated by a hydrated layer, and
at other points are tied together by chemical bridges held by mole-
cular cohesion" (today one would say residual valences or hydrogen
bonds) . Proof that a macroscopic change in form, such as muscular
contraction, is accompanied by a change in form of protein chains,
led us to the assum.ption-' that visible contraction and elongation
in muscle or protoplasm is to be ascribed to the contraction and
elongation of fibrous protein molecules, which in turn are controlled
by chemical activities.

These were the older concepts. The question now arises, what
contributions have since been made in support of this theory. Under-
standing of the change in molecular form, especially reversible form
changes, is still very meager. Most noteworthy are the changes in
molecular configuration observed by Gorter, van Ormondt, and
Dam^ in egg-albumen films spread on water. At pH 5 the films
are quite flat but below this pH value they are shrunken to one-tenth
of their former area.

'Amer. Naturalist, 63, 410. 1929.

'Ber. dtsch. chem. Ges., 61. 1932, 1928.

'Bioch. Z. 214, 253. 1929.

' Proc. Acad. Wetensch. Amsterdam, 35, 838. 1932.

[267]

268

The Structure of Protoplasm

Fig. 1. This figure is by Dr.
H. Mark and illustrates Dr.
Kurt Meyer's concept of the
molecular and micellar struc-
ture of elastic organic sub-
stances. The chains are mole-
cules, the segments in which
represent glucose units, or
rather a block of glucose units,
say fifteen in each segment.
The micel, or-h, would contain
some 500 glucose units. Many
of the molecular chains ex-
tend through two or more mi-
cel. The loose ends of the
chains are "fringes," always
present, so Meyer and Mark
believe. A somewhat similar
scheme of micels with over-
lapping ends is given in Fig.
5, page 6.

The conditions which estabUsh mole-
cular form became evident from a com-
parison of acetylcellulose and acetyl-
amylose which are distinguished only
by the configuration of the glucosidal
bond, [3 in cellulose and a in amylose.
Molecules of acetylcellulose when in
solution are stretched and oriented by
streaming, as R. Singer has found in as
yet unpublished experiments; whereas,
the molecules of acetylamylose are not.
From this it follows that the molecules
of acetylcellulose are fibrous, and those
of acetylamylose are spherical.'^

Far better known is the relationship
between chain structure and rubber-
like elasticity. According to the theory
which was developed by us"*, and at
about the same time by W. Busse*^ and
E. Karrer', the tendency of rubber-like
matter to take its original form after
deformation can be ascribed to the fact
that the chain molecules themselves are
capable of deformation, and that parts
of adjacent chains can slide one over
the other just as do molecules in a
viscous fluid. Stretching and orienta-
tion force molecules, which when re-
laxed may assume any possible shape,
to take on less possible forms and ar-
rangements. Due to kinetic energy, the
thermodynamic probability that the
molecules shall assume a disordered
state becomes greater as soon as the ex-
ternal force causing deformation disap-
pears. It follows from this that only
systems containing chain molecules, the
inner mobility of which is not prevented
by crystallization or micel formation,
can possess rubber-like elasticity. In

Protein and Protoplasmic Structure 269

proteins such systems are always found as hydrated chains or chain
fragments. Chain parts when coagulated or arranged in micellar
fashion do not take part in elastic phenomena.

If a passively deformed object, for example, a stretched cell,
loses its tension, one can conclude, that whole chains slide over each
other; therefore, they must be free. If, on the contrary, no relaxa-
tion is recognizable, then the protein chains are connected to each
other to form a network.

With this, I come to the matter of contact regions or the region
of bonds between molecules in organic jellies. Gelatinized aqueous
gelatin and gelatinized starch paste are built in the following man-
ner: They consist of a loose network, the regions of contact being
formed by micellarly arranged parts of numerous chains held
together into a firm bundle. These parts or sections of numerous
chains are securely bound by "lattice forces" to form a bundle.
Segments of molecules protrude from the micellae ("Fransen-
micelle", after Gerngross) which are hydrated and which establish
ties between adjoining micelles. As a result, the micelles are united
into a network of elastic and solvated molecular threads^. In
solutions of high polymeres fewer and smaller regions of contact
are sufficient to make an elastic jelly out of a viscous solution.
Whether or not gelatinization takes place depends upon the medium,
especially its pH, and the presence of electrolytes. Very slight
changes in these properties are sufficient to bring about the forma-
tion of a jelly.

I think it not unlikely that such sol-gel changes in protoplasm
are to be ascribed to the same factors, acidity and electrolytes, just
as in all metabolic changes, such as the synthesis and decomposition
of organic acids and acetylcholin, the binding and dissolving of
phosphoric acid, etc.

Editor's Note: The concept which Professor Meyer gives us is
so important that I should like to repeat part of his conclusion, and
then add a figure. As I did not care to assume the responsibility of

'Kolloid Zeitschrift, 59, 208. 1932.
° J. Physic. Chem., 34, 2870. 1932.
' Physic. Rev., 39, 857. 1932.

' Further details in K. H. Meyer "Die hochpolymeren Verbindunsen," Leip-
sig. 1940.

" Helv. chim. acta 20, 1331. 1937.

270 The Structure oj Protoplasyn

the latter, I asked Dr. Mark, long collaborator with Professor Meyer,
to give me a sketch such as Professor Meyer would have made.

Professor Meyer tells us that the "points" of contact between
fibers are not points but regions, that there are no free chains, and
that the chains are hydrated over much of their surface. In short,
a chain, or part of it, is either linked to other chains or to water.
Free chains do not occur owing to the high van der Waals' forces of
the — OH or — CONH groups.

Editor's Second Note:

It is an especial pleasure to close this Monograph with a word
from Professor Astbury. His letter gives a promise of something out-
standing in the very near future, for which we shall all wait with
interest. Even greater than the information he promises, is the
courage he displays in carrying on with good humor under the pres-
ent trying circumstances. The second from last sentence in Professor
Astbury's letter expresses the spirit in which every contributor to
this Monograph took part in the original Symposium. The sympa-
thetic cooperation which Professor Astbury exemplifies is the basis
of all that is worth-while in science. I am sure, too, that I express
the sentiments of our Society when I say in reply to the last sentence
in Professor Astbury's letter, that we, his colleagues in America, in
particular those in biology who have kept so closely in touch with
his work, send our gratitude for that which a good physical chemist
has done for biology, and our hope that he and his fellow scientists
will soon be able to carry on in a peaceful England.

The following is Professor Astbury's communication:

Dear Seifriz:

I was delighted to hear from you again, for it is an even greater
pleasure than usual these days to receive letters from our American
friends. Quite apart from the fact that your continuous flow of
sympathy and help is tremendously heartening, you Americans have
the distinction of being just about the only correspondents we have
left now, outside our own Commonwealth!

Your letter by Clipper Air Mail was five weeks under way. Does
this mean that the molecular structure of protoplasm is so terribly
important that weeks of censorship are necessary to make absolutely
certain that the enemy learns nothing about it? . . . The reprint

Protein and Protoplasmic Structure 271

you mention, telling about the Symposium held at Christmas, and
your scheme of turning it into a Monograph, did not reach me. I am
sorry, for I should have liked very much to write something for you.
Had your request arrived in good time probably I should have
made a special effort, but there's so very little spare time these days,
believe me! You are right in inferring that we still carry on as much
as possible with work on the good proteins but, our hands and
thoughts are pretty full with other things too. . . .

I have become an officer in the Air Force and am functioning as
Navigation Instructor to the University Air Squadron. It is volun-
tary work, of course, and does not mean that I have abandoned
research, just that I have to try to fit in more things during my
waking hours (we don't get nearly so much sleep as we used to!)
and have rather less time to think about proteins, music, and all
the other things that make life worth living. I suppose that as the
war goes on there will be still less and less time or opportunity for
these nice things, but for the moment we are not doing too badly.
Our laboratory hasn't been knocked down yet and we are still alive.
Quite recently we have had a lot of excitement because of a
new advance we have made in the structure of a-keratin and (3-
myosin, the heroic problem of the nature of the intra-molecular
folds in proteins. I really think that we have got something this
time, as you Americans say; when I posted the manuscript I had
almost a 'nunc dimittis' feeling about it.

You are certainly right in ascribing to me the belief that the
folded protein chain is of more importance than any other structural
feature in biology, or rather, the capacity of the protein chain to
fold. Pauling thinks so too, I should say, to judge by his more
recent publications; in any case, the problem of the fold is the same
thing as that of the linkages that promote or maintain the fold, so it
may well be that resonance and the hydrogen bond are of more
importance to physiology than any other two facts in chemistry —
provided, of course, the hydrogen bond does turn out to be
pre-eminent among protein linkages, a thing we are not too sure
about yet, I'm afraid. It's a pity, in one sense, that there are still
so many things that we are not so sure about, though in another it
is very gratifying, otherwise we should all lose our jobs! I only hope
that God is not laughing too much behind His cloud at our efforts!

272 The Structure of Protoplasm

Speaking seriously, though, molecular biology has progressed an
amazing lot these last few years, and I am happy that I have been
privileged to take part in it.

"We all in this laboratory send our kindest regards and good
wishes to you and your associates.

Yours very sincerely,

(Signed) W. T. Astbury.

INDEX

Abitz, W., 38

Abramson, H. A., 31, 38, 40, 97

Accumulation

of particles by ultrasonics, 96
Acetaldehyde, 66
Acetic acid, 66
Acetylamylose, 268
Acetylcellulose, 268
Adenosine pyrophosphate, 61
Adsorption

and dilatancy, 92

Aggregates, 68
complex, 69, 70
functional, 42, 71
simple, 68

Aggregating mechanisms, 64
Aggregation, 62, 68

complex, 69, 70, 71

loose, 70

Albu, H. W., 97
Albumin, 60
Alcohol, 61
Aldehydes, 61
Alveoli, 2

Ambronn, H., 38, 40
Ameba; see Amoeba
Amino acids, 27, 28, 74

analyses, 57

differences, 55

similarities, 55
Amino acid residues, 50

distribution, 55, 57, 58

length, 50

number, 50, 55
Amino groups, 47
Amoeba, 23, 25, 32, 99, 100, 101,
131, 132, 133, 134, 155, 163. 165,
169, 170, 171, 172, 194, 195, 246,
267

division of, 194

locomotion of, 163 ff., 170-171
Amoeba dubia, 131, 133
Amoeba proteiis, 133, 170, 194
Amoebocytes, 104, 133
Amoeboid locomotion, 163 fiE., 171,
Amoeboid movement, 127, 130 fT.
Anderson, D. B., 12, 19
da C. Andrade, E. N., 98

130,
166,
257,

172

Anesthesia, 247 ff.
theory of, 248

Anesthetic agents, 246, 248

Anisotropy; see also birefringence,
double refraction
of dehydrated surface film, 104-105
of myelin forms, 104

Anomalous flow, 209, 252; see also non-
Newtonian

Anomalous viscosity, 32, 86, 87, 94;

see also structural viscosity
Arbacia, 32, 34, 60, 127. 132, 133, 139,

140, 141, 142, 143, 153, 191, 192

Arginine, 56, 67, 68

Arrangement in space, 63

Artifact, 3

Ascorbic acid, 61
model, 67

Aspartic acid, 56

Astbury, W. T., 8, 19, 20, 27, 35, 36,

37, 38, 254, 261, 262, 270
Atomic radii, 75
Atoms

radii, 48

size, 46

structure, 46

valence angles, 46, 47, 48

Autocomplex coacervates, 93

Balance-pressure, 166, 207 ff., 239, 240,
241

Bailey, I. W., 11, 20

Balbach, H., 200

Banga, I., 37, 38

Barium malonate, 88

Bauer, H., 120

Beams, A. W., 38, 40

Beats, 229, 239

Belar, K, 120

Bemsley, R. R., 40

Bentonite

thixotropy of, 88, 94
Benzopurpurin sols, 93
Berger, C. A., 116
Bergmann hypothesis, 57
Berkley, E. E., 19
Berrill, N. J., 124
Biancani, E. H., 98

[273]

274

Index

Birefringence, 24 ff., 104; see also ani-
sotropy, double refraction
of cotton fibers, 17
of cytoplasm, 24-26, 31
of young fibers, 12

Bjerknes, V., 98

Bleb formation, 195

deBoer, J. H., 98

Bondy, C, 98

Bonner, J., 107

Boothroyd, 114

Brachet, J., Ill

Brandt, O., 98

Brohult, S, 98

Brown, D. A., 32

Brown, D. E. S., 127, 129, 130, 133, 138,

139, 140, 143, 149, 151, 153, 154, 155,

156, 157, 168, 169, 192

Brownian movement, 89, 183, 189

and viscosity changes, 167, 181, 191
Brush heap, 4, 34, 252
Bull, H. B., 38
Bungenburg de Jong, H. G., 91, 92, 98,

106, 107
Burger, F. J., 98
Busse, W., 268

Cadmium gel, 249

Calvin, M., 118

Camp, W. G., 175, 177, 179, 182, 183,

184, 200, 201
Carbohydrate, 61
Carbon dioxide, 72, 73, 165, 246, 247,

248, 249
Casein, 55

Casperson, T., 110, 111, 112, 113
Cattell, McK., 149, 151, 157
Cavitation

caused by ultrasonics, 95

Cawood, W., 98

Cell

centrifugation of, 23, 27, 40, 127, 133,

136, 137, 138, 139, 140, 141, 143,

144
cooling and birefringence of, 26
locomotion of, 171 ff.
plasmolysis and birefringence of, 26

Cell cleavage, 34, 139 ff , 166, 187, 190,

191
Cell division, 138 ff., 184 ff., 193

Cell wall, 11, 13
growth, 74
microscopic structure, 11-21

Cell wall — continued
primary, 12, 13, 14
secondary, 12, 13, 14, 16, 20

Cellulose, 4, 12
tensile strength, 5

Cephalin, 58, 59, 65
Cartes, A., 151
Chalkley, H. W., 194
Chambers, R., 31, 38, 100, 102, 107, 139,
158, 183, 191, 192

Channeling of reaction, 70
Channels in protoplasm, 69, 70
Charged groups, 64
Chemical bonds, 64
Chloroform, 246, 247
Chlorophyll, 61, 73

model, 63
Chromosomes

birefringence of, 24, 121

differential regions in, 114

euchromatic regions of, 113, 114

heterochromatic regions of, 113, 114

matrix of, 119

migration of, 185 ff., 189, 190

pepsin, digestion of. 111

spiral structure of, 119 ff.

trypsin, digestion of. 111

Cilia, 128, 151

Clays, 88, 94

Cleavage, 34, 139 fE., 166, 187, 190, 191

Cleveland, L. R., 119, 121

Coacervates, 92

as colloidal state of surface films,

105, 106
autocomplex, 93

Coacervation, 91, 106
Coagulation

caused by ultrasonics, 96

orthokinetic, 96

proteins, 35

thixotropy and, 89, 94

Cohesion, 49, 63, 64, 75
forces, 71

Coils, 119 ff.

changes in direction of, 121

major, 120

minor, 120

molecular, 120

relational, 120

relic, 120

somatic, 122

standard, 120

torsion, hypothesis of, 123

Colloidal particles
rod-shaped, 32, 87

Index

275

Colloidal solutions

dilatancy of, 88

elasticity of, 85 ff .

mechanical properties of, 85 ff .

viscosity of, 85 fT.

Coman, D. R., 174

Complementary faces, 65

Conduction of nerve impulses, 149

Configuration, 62

Continuity

structural, 6, 7, 8, 252

Contractile properties, 155
Contractile tension, 163, 166, 168, 169,
173. 174, 175, 182, 187, 189, 190, 194

Contractility, 254 ff.

centers of, 258

molecular basis of, 262

rhythmic, 243, 254, 255, 256, 259, 260.
261, 263, 264
Coordination points, 47, 53, 55
Coper, K.. 98
Copper. 61

Cortical gel, 100, 127, 139, 140, 144, 153
Costello, D. P., 139, 153
Coulomb's law, 57
Coult, D. F., 38, 40
Cross-linkages, 64
Cyclopropane, 246, 247

Cyclosis, 135 ff.; see also protoplasmic
streaming

Cytochrome, 67, 69

model, 63
Cytoplasm, 55, 58, 59, 60. 68. 71. 75. etc.;
see also protoplasm

diagram. 105

gel structure in, 127-161

particulate nature, 42, 49

structural differentiation, 99-107

Cystine bridges, 53, 62, 65

Dam, 267

Dan, K., 139

Daniel, F. K., 97, 98

Darlington, C. D., 113, 114, 119, 120,

123
Dehydration

and thixotropy, 91

of proteins, 37

Deuticke, H. J., 154

Dibenzoylcystine, 88

Dilatancy, 88, 89, 91, 92, 94, 95
and adsorption, 92
and dispersion medium 91
and peptization, 91

Dilatancy — continued

and ultrasonics, 97
vs. rheopexy

Dinucleotides, 61

Diphosphopyridine nucleotide, 61

Dipole, 47

Diptera, 110

Disintegration

respiration, 73
Dispersion medium

and thixotropy, 91

Dognon, A., 98

Double bond angle, 48

Double refraction, 4; see also aniso-
tropy, birefringence

protoplasm, 24-26, 31, etc.

streaming, 25, 29, 35, 36, 37

factoids, 93
Draper, J. W., 157
Drosojjhila melanogaster, 114
Dube, S. P., 98
Dynamoplasmogram, 209 ff.

Ebbecke, U., 154, 157
Echinoderm eggs, 99, 100, 245
Ectoplasm, 167, 257
Ectoplast or surface film, 101-102
molecular structure of, 105-106

Edsall, J. T., 38, 97
Edwards, D. J., 149, 151, 157
Edwards, J. S., 166, 167
Egg-albumen, 28, 267
Egg albumin configuration, 52

Elasticity

colloidal solutions, 85 ff.

cortical cytoplasm, 99-100

inner gel layer, 102

muscle, 33, 253

protoplasm, 32. 249-254, etc.

rubber, 25-26, 32-33, 252. 253. 268

salts, effect of, 251

soaps, 33, 250

sols and gels, 87

structural continuity as the basis

of, 252
thin solutions, 250
theory, 251

Electric potential, 235
Electrical conductivity

of sols and gels, 86
Electrical field. 56, 57
Electron-donating system, 65
Electrostatic forces, 62, 63, 64

276

Index

Elementary particles
ionization, 56
of protein, 53

Elodea, 32, 128, 132, 135, 137
Elodea canadensis, 137
Emulsion, 2, 3, 106

and ultrasonic waves, 96

Endomitosis, 116 ff.
in Allium, 117
in Gerris, 116
in hymenoptera, 117
in mosquitoes, 116
in spinach, 117

Endoplasm, 100, 104, 105, 163, 164, 166-
168, 171-179, 181-183, 187, 189, 190,
195, 196, 243, 257

Engelmann, Th. W.. 25, 38

Enslin, O., 98

Enzymes, 67, 73

respiration, 45, 60, 61

Ephelota, 128, 133

Ephelota coronata, 133

Ergosterol, 60

Erythrocytes, 128, 267

Ether, 247, 248

Ethyl alcohol, 66

Ethylene, 247, 248

Euglena, 243

Euglena deses, 243

Euglenoid movement, 243

Ewart, E. J., 209, 259

Extensibility, 250, 251, 267; see also
spinning capacity
of ectoplast, 101

Fats, 63

Fatty acids, 58

Fatty material, 58, 60

Fatty substances, 46

Faure-Fremiet, 104, 246

Fawcett, E. W., 154

Feulgen reaction, 110, 111

Fibers

cotton, 11, 12, 14, 17

Fibrils, 13, 17, 18, 19, 20

Fischer, I., 115

Flagella, 128, 151

Flavine-adenine dinucleotide, 61

Fluidity, 6, 8

Fogs

coagulation by ultrasonics, 96

Fontaine, M., 136, 154
Forces between particles, 93

Formic acid, 66

Freund, H., 98

Freundlich H. M., 31, 32, 38, 82, 83, 84,
85, 97, 98, 152, 153, 245, 246

Frey, A., 38, 34

Frey-Wyssling, A., 20, 30, 38, 107, 112,
121

Fry, H. J., 159

Fujii, K., 120, 121

Fundulus, 145

Geitler, L., 114

Gels

collagen, 30

double refraction of, 25
elasticity of, 25-26, 32
mechanical properties of, 85 flE.
polystyrene, 29-30
stretched, 26, 30
structure, 29, 127 ff.
superficial layer, 163
thixotropic, 31, 85 ff ., 153
viscosity of, 31

Gel formation, 29 ff .
and solvation, 94

Gel layers, 195, 196
in cytoplasm, 99-100
inner, 102
superficial, 163, 190

Gel-sol transformation; see sol-gel

transformation
Gel values, 132
Gelatin configuration, 52, 55

Gelatin gels

birefringence, 25
electrical conductivity, 86
structural viscosity, 31
thixotropy, 88
X-ray pattern, 30

Gelation, 132, 168

relation to cleavage, 132

relation to physiological response,

154
relation to pressure, 132
relation to streaming, 132

Gelation percentages, 133, 138

Geloids, 90. 92

Gentcheff, G., 117

Germanium gel, 249

Germination of seeds

influence of shaking on, 40

Gerngross, O., 38
Gibson, R. O., 154
Gillings, D. W., 98
Globulins, 60
Glucose, 72, 73, 74

hidex

277

Glutamic acid, 56
Glutathione, 61

model, 63
Glycerol, 86
Glycine, 55
Goldschmidt, R., 118
Goodeve, C. F., 98
Gorin, M. H., 38
Gorter, 267

Granules

submicroscopic, 71
Green, H., 92, 97
Grundfest, H., 149
Guanidine group, 55, 56

Hammerstein, 111

Harvey, E. B., 23, 38

Hatschek, E., 3, 251

Hauser, E. A., 94, 97, 98

Heart muscle, 60

Heilbrunn, L. V., 134, 164. 209

Heitz, E., 113, 119

Helix povxatia, 96

Heller, W, 89, 97, 98

Heme, 67

Heniitrichia vesparium, 260

Hemocyanin

splitting by ultrasonics, 96

Herrmann, K., 38

Hexose, 69

Hexose diphosphate model, 67

Heyman, E., 98, 152, 155. 159

Hiedemann, E., 98

Hillary, B. B., 110

Hilton, A. E., 200

Hirsch, G. C., 38

Histidine, 56, 68

Hock, W., 11, 21

Hoerr, N. L., 40

Hdfler, K., 102, 107

Hooker, D., 145

Hoppner, 116

Howard, F. L., 173

Huskins, C. L., 109, 120, 121, 122, 123

Hyaline cap, 167, 168, 179, 180, 181, 195

Hyaline zone, 133

Hydrated protein particles, 53, 54

Hydration, 6

and dehydration. 104, 255

of proteins, 9, 37
Hydrocarbon, 252

Hydrogen bond, 6, 7, 8, 9, 28, 35, 49, 50,
267, 271

Hydrogen bridge, 53, 54, 62, 64, 65, 71,
75, 76

Hydi'ogen-donating residue, 66

Hydrogen ion, 65

Hydrogen ion concentration, 56

Hydrophilic groups, 54

Hydroxyl, 47, 55

Ice, 47

Imidazole ring, 55, 56

Insulin configuration, 52

Interference, 228

intraplasmic, 229, 233, 241

Ion, 46, 68
adsorbed, 68
concentration, 68
inorganic, 68, 70

Ionic bonds, 53, 71, 75, 76

Ionic linkages, 65

Ionization, 63

acidic groups, 56

amino and residues, 55, 56

basic groups, 56

protein, 55, 57

Ionized groups
distribution, 56

Iron, 61

Iron oxide, colloidal, 86, 88, 89, 90,
91, 95

Isoelectric point, 35, 68

Jacobj, W., 115
Jacobs, M. H., 165
Jaegar, L., Ill
Jeannerat, J., 33, 39
Jelly fish, 249
Jones, A. D.. 97
Juliusburger, F., 97, 98

Kamiya, N., 156, 158, 166, 199, 205, 207,

243, 248, 257, 263. 264
Kandelaky, B. C, 98
Kaolin, 96
Karrer, 268
Keratin, 35, 254
Kerr, T., 12, 16, 19, 21, 38
King, R. L., 38, 40
Kinoplasm, 103
Kinoshita, S., 235
Kitching, J. A., 133, 151, 157
Kodani, M., 118
Kokati, M., 235

278

Index

Roller, P. C, 119
Koltzoff, N. K., Ill
Konig, W., 98
Kruyt, 91
Kuwada, Y., 118, 121

Lactic acid, 61
Laing, M. E., 97
Lamellae, 15, 16
LauflEer, A. M., 97
LeChatelier, 153
Lecithin, 58, 59, 65, 91
Leucocyte, 32, 163, 172
Leucoplast, 73, 74
Levans, A., 117
Levels

dimensional, 42, 43, 44

molecular, 41, 45

Levine, S., 98

Levitt, J., 100, 107

Lewis, W. H,, 127, 155, 156, 159, 163,
165, 166, 167, 250, 264

Lindau, G., 97

Linkages, 6

Lipoidal properties of surface films,
104-105

Liquefaction of gels
by ultrasonics, 95

Liver tissue, 60

Locomotion

amoeboid, 163 ff.

of amoeba, 170-171

of cells, 171-173

of slime mold, 173-184

Lymphocyte, 166, 171, 172

Lysine, 56, 68

McClintock, B., 119

Mack, 252, 253, 262

Magnesium, 61

Malic acid, 61

Mark, H., 39, 252, 262, 270

Marsland, D. A., 32, 39, 127, 130, 133,

135, 138, 140, 145, 151-153, 155-158,

168, 169, 193, 194

Mast, S. O., 101, 127, 133, 156, 158, 159,
163, 165, 166, 167, 169, 170, 177, 179,
181, 182, 183, 194, 195

Matsuura, H., 121

Matthews, S. A., 148

Mazia, D., Ill

Mechanical pressure

effect on protoplasmic streaming,
199, 200

Mechanical properties
of sols and gels, 85-98
of paints, 92

Mechanism, 72, 73, 74, 75
cellular, 60
micellar, 43, 44
molecular, 43, 44, 45, 49

Mehl, J. W., 39, 97
Meiosis

precocity theory of, 113

Melanophores, 145 ff .

Membranes

and films, cytoplasmic, 100-103
coacervate resemblance, 106
lipoidal properties, 104, 105
molecular structure, 104, 105
plasmogel, 181
relation to kinoplasm, 103
undulating, 172

Metabolie, 243

Methylcellulose, 93, 94, 96, 153
Metz, C. W., 116, 117

Meyer, K. H., 33, 39, 252, 253, 262, 267,

270
Micellae, 3, 267

fringed, 269
Microdissection, 19, 32, 191; see also

micromanipulation needles (see

microneedles)

Micromanipulation, 33, 100-102; see
also microdissection
oil drop technique, 101

Microneedles, 23, 32, 33, 102, 140, 200,

249
Microscopic structure

of cell wall, 11-21

of slime mold, 182-184

Microscopic visibility, 41
Miller, D. C, 232
Miller, W. A., 34, 39
Milovidov, P., 110

Minerals

thixotropy, 88
Mirsky, A. E., 34-37, 39, 40, 97, 153
Mitosis, 165, 184 ff ., 245
Mitotic spindle, 25, 185, 187

birefringence of, 25

Mohl, H. von, 23, 39
Molecular structure, 75, 76; see also
structure

of cytoplasm, 104-106, etc.

of hydrocarbons, 252

of protoplasm, 41-79, etc.

Molecules; see also proteins
chain, 268
dimensions, 46

Index

279

Molecules — continued
fibrillar, 267
folded, 28. 252
globular. 254. 267
helix, 252, 261
linear, 267

number in protoplasm, 45, 46
organization of, 43
rod-shaped, 33
rubber. 252, 253
shapes, 46, 48
sizes, 42
sphere, 253, 254

Monolayers of protein, 52, 53
Moore. A. R., 26-27, 34, 39
Mosaics, 62, 64
Motive force, 155, 199 ff.

graphical representation of. 209 ff.

measurement of, 207-208

Motomura, I., 139, 192

Moyer. L. S., 23, 38, 39, 85, 254

Mudd, E. B. H.. 32, 39

Mudd, S., 29, 32

Muscle. 60, 128, 149
elasticity of, 33

Muscular contraction. 33. 35. 36. 37.
149, 267

Myelin forms, 26, 103, 104, 105, 106

Myosin, 29, 33, 34 ff., 153, 154, 254, 262

denaturation of, 36-37

physical properties of, 25, 29. 33. 35-
37

thixotropy of, 88
Myosin gels, 88. 132
Mytilns edulis, 151

Myxomycete Plasmodium, 27. 173-184,
199 ff.. 246-249. 254-260; see also
slime mold

anesthesia, 246 flf.

as an aggregate, 258, 260

centers of activity, 258. 259

locomotion, 173 ff .

microscopic structure, 182 ff .

nucleus, 183-184

polarity, 225, 226

protoplasmic streaming, 199 ff ., 255 ff .

rhythmic pulsation, 254 ff.

sclerotium, 260

viscosity, 184

Naegeli, K. von, 3
Nakamura, T., 121
Nauss, R. N., 260
Nebel, O. R., 118, 119. 120
Nerve, 128
Nerve impulse, 149

Neurath, H., 39

Newtonian liquids, 85. 86. 87

Nitrogen, 65, 75

coordination points, 47, 53. 54, 55

groups, 55, 58, 61

Non-Newtonian, 210
behavior. 252
flow, 209

Non-spherical particles
and rheopexy, 95
and thixotropy, 93
spontaneous orientation, 92

Norris, C. H., 156, 249

Northern. H. T., 32. 39, 249

Northern, R. T., 32, 39

Nucleic acid, 61
desoxyribose, 110
during cell division, 112
metabolism, 118
model, 68

relation to nucleolus, 118
ribose, 110. 115, 118
role in gene reproduction, 112, 113

Nucleic acid
starvation, 114
thymonucleic acid, 110

Nucleolus, 118. 119, 123

Nucleotides, 61, 66

Nucleus

change in volume, 115, 116
energic, 114, 124
in slime mold, 183, 184
nucleo-plasmic ratio, 115
structural differentiation, 109-126

Oksala, T., 117

Oleic acid, 92, 94

Olze, A., 97

Oncley, J. L.. 39

Organization, 63, 70, 76, 85, 164
of cell, 164
of molecules, 43
on visible level. 43

Orientation of particles

by ultrasonics, 97
Ormondt, van, 267
Orthokinetic coagulation, 96
Oxide gels, 87
Oxygen, 65, 75
Oxygen groups, 54, 55, 58

Painter, Th. S., 118
Paints, 92
Paralysis, 249
Paramecium, 128

280

Index

Paris, 114

Particles

cytoplasmic, 62

forces between, 93

in vitro, 71

orientation, by ultrasonics, 97

properties, 42, 56

protein nature, 42, 56, 58

size, 42, 51

size classes 41, 42, 51

Patterns, 65, 72

molecular, 64

structural, 43

Patterson, H. S., 98

Pauling, L., 6, 7, 8, 9, 39

Pease, D. C, 133, 138, 151, 157, 158, 223

Peptide chains, 50, 52

Peptizing agents, 91

Peterfi, 246

Pfeffer, W., 259

Pfeiflfer, H., 31, 39, 210

Phospholipids, 58, 65, 69

Phosphoric acid, 61, 66

Photosynthesis, 73

Physarum polycephalum, 128, 156, 173.
199, 201, 246, 255

Phytosterols, 60

Picken, L. E. R., 39

Pigmentary effectors, 145

Plasmagel, 32, 131, 133-135, 138, 156, etc.
of amoeba, 132, 155, 165, 166, 169,

170, 171, 195
of egg cells, 139, 140, 193, 194
of plant cells, 100, 138
of slime mold, 174
vs. plasmasol, 99-100, 104

Plasmalemma, 101, 165, 166, 167, 169.
170, 195

Plasmasol, 135, 138, 139, 165, 166, 167,
169, 170, 171, 174, 193, 194, 195, etc.

Plasmodium; see Myxomycete Plas-
modium

Plasmogel; see plasmagel

Plasmolemma; see plasmalemma

Plasmosol; see plasmasol

Plowe, J. A., 33, 39, 40, 102, 107

Poiseuille's law, 86, 208

Polar groups, 55
distribution, 70
number, 66

Polarity, 225, 235
reversal of, 233
spontaneous change in, 235

Polarized light, 24, 112, 121; see also
anisotropy, birefringence, double
refraction

Polypeptide chains, 249, 253, 254, 261,
267; see also peptide chains, pro-
tein chains

Polyrhythm, 234, 236, 240, 242, 257

Porphyrin, 67

Pressure, 129

balance, 166, 207 ff., 239, 240, 241
counter, 201, 206. 207
hydrostatic, 32, 127, 129
mechanical, 199, 200

Pressure bombs, 129-130
for centrifuging, 129-130
for microscopic examination, 130

Preston, 5, 19

Primary valence, 62, 63

Prosthetic group, 61, 62, 65, 67, 69,

73, 76
Proteins, 8, 46, 49, 53, 55, 60, 72, 253

aggregates, 58

and protoplasmic structure, 23-40

backbone, 50, 54

basic, 68

chains, 53, 55, 61, 73, 74

coagulation, 35

complex, 70

component, 67

configuration, 51, 52, 54, 56

corpuscular, 27, 28, 33, 35

cube, 53, 69

cytoplasmic, 33-38, 40

denaturation, 33, 35, 36

dimensions, 50, 53

discs, 53, 56

elasticity, 32-33, 34, 35

elementary particles, 69

fibrous, 27, 28, 35-37

folded, 28, 252

globular, 37, 254, 267

molecular weight, 51, 58

molecules proportional number, 69

monolayers of, 52, 53

packets, 52, 53, 54, 55

particle, 62, 63, 64, 65

particles and separation, 70

shape, 28

side chains, 50, 54

specific, 62

spinning capacity, 33

streaming double refraction, 25, 29,
35, 36, 37

structure, 27-29, 35-36, 50, 62

surface, 62, 64

synthesis, 72

thixotropy, 31-32

viscosity, 31, 35

Proton, 49

Index

281

Protoplasm

anesthesia, 247 flf.
atomic constituents, 47
centrifugation, 23, 27; see also cen-

trifugation of cells
cold, effect of, 26
components, 71
composition, 1, 69
contractility, 254-265
double refraction, 24-28. 31; see also

anisotropy
elasticity, 32-33, 249-254
emulsion droplets in, 23
gelatinization of, 248
"a going concern'", 72
liquefaction of, 245
local regions in, 30-31
mechanical properties, 85 ff .
microscopic appearance, 23
miscibility with water, 26
molecular constituents, 45, 46
physical properties, 245 ff., 263. etc
polyrhythm, 234, 236, 240, 242, 257
proteins of, 33-38, 40, etc.
rod-shaped bodies in, 26-27
statistical isotropy of, 25
structural viscosity of, 31, 32; see

also anomalous, non-Newtonian
thixotropic setting of, 246, 248, 249
thixotropy of, 31-32
water in, 26

Protoplasmic streaming (flow)
cause of, 169
effect of hydrostatic pressure on, 32,

135-138
effect of mechanical pressure on, 199,

200
effect of pressure difference on, 205-

206
fountain streaming, 169
in relation to gel structure, 127-161
mechanism of, 255 ff.
motive force, 155, 199, 207 ff., 209 ff.
physical aspects of, 199-244
theories of, 155, 255 ff .
velocity, 136, 206
volume change in relation to, 155

Protoplasmic structure, 1 ff. (introduc-
tion)
molecular aspects of, 41-79
physical properties in relation to,

245-265
proteins and, 23-40, 267-269

Protoplasmic surface film, 100-101

Protoplasmic volume, 129

Protoplast sti'ucture (a poem), 107

Pseudopod; see pseudopodium

Pseudopodium, 23, 130-135, 148. 165.
168, 170, 171, 172, 174, 175, 180, 181,
187, 193. 195, 256

Pseudopodium- — continued
birefringence of, 25
form of, 134-135
ruffle, 172
stability of, 135
withdrawal of, 187

Pyridine nucleotid, 67, 69

Pyruvic acid, 61

Quimfe, G., 97

Rana, 152

Raw materials, 72. 73
Reaction chambers, 70
Recklinghausen, H., 97
Reed, C. E., 97, 98
Regnard, P., 151
Residual charges, 49
Residue, length, 66
Resonance, 9, 271
Respiration, 60, 72, 73
Reynolds, O., 88, 97
Rheopexy, 94, 95

and ultrasonics, 96

vs. dilatancy, 95

Rhythm, 210 ff., 242 ff., 255 ff.

graphical representation of (Plasmo-
dium), 210 ff.

in Euglena, 243, 264

in Myxomycete Plasmodium, 210 ff.,
246, 255 ff.

interference, 228, 229, 233, 241

occurrence, 263, 264

polyrhythm, 234, 236, 240, 242, 257

Riboflavin phosphate, 61, 69

Rinne, F., 26, 39

R5der, H. L., 97, 98

Rogowski, F., 98

Rubber

structure, 252
birefringence, 25, 26

Rugh, R.. 152

Sameshima, I., 97

Scarth, G. W., 32, 33, 39, 99, 107

Schalek, E., 97, 245

Schechtman, A. M., 139, 191, 193, 194

Schmidt, O., 97

Schmidt, W. J., 39, 40, 104, 107

Schmitt, F. O., 39, 40, 112

Schultz, J., 113, 114

Schulze-Hardy rule, 89

Sciara, 116

282

Index

Sclerotium

fragmentation of, 260

Sea urchin egg
centrifugation, 23
double refraction, 34
fertilization, 34, 37
proteins, 34-37

Seifriz, W., 1, 21, 30, 32, 33, 38, 39, 40,
97, 156, 243, 245, 267

Sensitization, 91

Separation distances, 65

Serine, 56, 65, 67

Shmargon, E. N., 114

Side chains, 6, 8, 50, 54

Silicic acid, dilatancy, 88

Silk, 254

configuration, 52
elastic properties, 33

Simha, R., 39

Siminovitch, D., 100, 107

Singer, R., 268

Single bond angle, 47, 48

Slime mold, 163, 167, 168, 195, 246-249,
254-260; see also Myxomycete
Plasmodium

locomotion of, 173-182

pseudopods of, 180, 195

Smith, S. G., 117, 119

Smokes

coagulation by ultrasonics, 96

Soaps

anomalous flow, 252

elasticity, 32, 250

electrical conductivity, 86

jelly, 249

Sodium stearate, 250

Sols

mechanical properties, 85 ff .

Sol-gel equilibria, 128, 144
in relation to pressure, 132
in relation to temperature, 139, 153
types of, 152, 153

Sol-gel transformation, 85, 93, 94, 99,

127, 193, 257, 260

Solation of protoplasmic gels, 127, 175,
176
quantitative measurements, 128
relation to cell function, 128, 129
relation to protoplasmic streaming,

128, 129, 136
reversibility, 128, 140, 141
specific examples, 127, 128

Sbllner, K., 85, 97, 98

Solvation

and gel formation, 94

Southorn, W. A., 98

Sparrow, A. H., 120, 122

Specific area, 67

Specific protein, 62

Specificity
degrees of, 66
of molecules, 68

Spindle; see mitotic spindle
Spinning capacity, 33
Sponsler, O. L., 41
Stability, 65
Stanley, W. M., 97
Starch

dilatancy of, 88, 92

grain, 73, 74

thixotropy, 92
Staudinger, H., 29-30, 40
Stern, C, 124
Stern, E., 100
Sterols, 58, 59, 60, 65
Strasburger, E., 103

Streaming

double refraction, 25, 29, 35, 36, 37
of protoplasm; see protoplasmic
streaming

Strongylocentrotus, 34, 192

Structure

cellulose, 4

elasticity in relation to, 8, 257

emulsion, 2, 3

energy vs., 1

fibrillar, 17

gel, 4, 252

granular, 2

micellar, 3, 269

proteins, 27-29, 35-36, 41 ff., etc.

protoplasm, Iff., 23 ff., 41 ff., 85 ff.,
99 ff., 245 ff., 267 ff., etc.

synthesis, 72
Structural continuity, 6, 7, 8, 252
Structural differentiation

cytoplasm, 99-107

nucleus, 109-126

Structural viscosity, 31, 32, 35, 96; see
also anomalous viscosity, anoma-
lous flow, non-Newtonian liquids

Substrate, 67, 69, 70

Succinic, 66

Succinic acid, 61

Sulphur, 61

Surface films, 100-103
molecular structure, 104-106

Surface tension, 251, 255, 261

Suspensions
dilatancy, 89

Index

283

Suspensions — continued

thixotropy, 88, 89

ultrasonics, effect on, 96
Svedberg, T., 40, 253, 267
Szegvary, A., 97, 245
Szent-Gyorgyi, A, 37-38

Tactoids, 92, 93
in mixed sols, 93

Taxis, 242

Temperature, 65

Tensile strength, 5, 6

Tension

at a surface, 259

contractile, 163, 165, 166. 168. 169,
173, 174, 175. 182, 187. 189, 190.
194
Tetrahedral angle, 47
Tetranucleotides, 61
Thiamine model, 63
Thiamine pyrophosphate, 61, 69
Thixotropic gels, 31, 85 ff., 153
Thixotropic setting, 246, 249

anesthesia due to, 248

bentonite sols, 94

Thixotropic sols, 86

Thixotropy, 85-96, 245-249

and coagulation, 89, 94

and light absorption, 90

and nonspherical particles, 93

and peptization, 91

definition, 246

of gels, 31

of protoplasm, 31-32, 245 ff.

of oxide gels, 87, 89
Thymonucleic acid, 110
Tobacco mosaic virus, 29, 88
Tonoplast, 101-102
Tradescantia, 120
Trillium, 100, 114, 120, 122
Triose, 69

Tnturus torosus, 191, 192
Tyrosine, 56, 65

Ullrich, H., 26, 40

Ultrasonic waves, 95
coagulating action, 96
dispersing action, 95
liquefaction of gels, 95

Vacuoles

submicroscopic, 69, 70
Valence angles, 75

Van der Waals' forces, 62, 64, 65, 71,
75, 76, 92, 253, 270

Vanadium pentoxide, 90, 93, 94

Vassy, E., 97

Verwey, E. J., 98

Verworn, M., 97

Viscosity

anomalous, 33, 86, 87, 94; see also

anomalous flow, non-Newtonian

liquids, structural viscosity
calcium ion in relation to, 164
change due to CO2, 165
colloidal solutions, 85 ff .
gels, 31

local changes, 167
measurement, 208
motive force vs., 209
protoplasm, 31-32, 164 ff., 208, 209,

210
rapid change, 8

regulation and polarization, 168
slime mold, 176, 177, 184, 208. 209,

210

Volume change in relation to stream-
ing, 155
Vouk, v., 211
Vries, H. de, 101

Wall; see cell wall
Watanabe, A., 235

Water

aggregation of, 47, 48
electrical fields, 47
molecule, 7
radius, 47
shells, 64
structure, 47, 48

Wedmore, E. B., 232
Wermel, E. M., 115
White, M. J. D., 114, 119
Wilson, E. B., 97
Wilson, G. B., 114, 120, 122
Wipf, L., 117

X-rays

breakage effects, 121
diffraction, 112

Yield value, 86, 207

Zein configuration, 52
Zocher, H., 97
Zocher, H. Z., 98
Zw^itter ion, 65

1 ; ^ •>- '