INTERNATIONAL SERIES OF MONOGRAPHS ON
PURE AND APPLIED BIOLOGY

Division: ZOOLOGY

General Editor: G. A. Kerkut

Volume 5

THE BIOLOGY OF STENTOR

OTHER TITLES IN THE SERIES ON PURE AND
APPLIED BIOLOGY

ZOOLOGY DIVISION

Vol. 1. Raven — An Outline of Developmental Physiology

Vol. 2. Raven — Morphogenesis: The Analysis of Molluscan Development.

Vol. 3. Savory — Instinctive Living.

Vol. 4. Kerkut — Implications of Evolution.

Vol. 6. Jenkin — Animal Hormones.

Vol. 7. Corliss — The Ciliated Protozoa.

Vol. 8. George — The Brain as a Computer.

BIOCHEMISTRY DIVISION

Vol. 1. Pitt-Rivers and Tata — The Thyroid Hormones.
Vol. 2. Bush — Chromatography of Steroids.

BOTANY DIVISION

Vol. 1. BoR — Grasses of Burma, Ceylon, India and Pakistan.

Vol. 2. TuRRiLL (Ed.) -^ Vistas in Botany.

Vol. 3. ScHULTES — Native Orchids of Trinidad and Tobago.

Vol. 4. CooKE — Cork and the Cork Tree.

MODERN TRENDS
IN PHYSIOLOGICAL SCIENCES DIVISION

Vol. 1 Florkin — Unity and Diversity in Biochemistry.

Vol. 2. Bracket — The Biochemistry of Development.

Vol. 3. Gerebtzoff — Cholinesterases.

Vol. 4. Brouha — Physiology in Industry.

Vol. 5. B acq and Alexander — Fundamentals of Radiobiology.

Vol. 6. Florkin (Ed.) — Aspects of the Origin of Life.

Vol. 7. Hollaender (Ed.) — Radiation Protection and Recovery.

Vol. 8. Kayser — The Physiology of Natural Hibernation.

Vol. 9. Francon — Progress in Microscopy.
Vol. 10. Charlier — Coronary Vasodilators.
Vol. 1 1 . Gross — Oncogenic Viruses.
Vol. 12. Mercer — Keratin and Keratinization.
Vol. 13, Heath — Organophosphorus Poisons.

c«

'■r.

V

i^

■t-T

a

^

• -.

(/J

;S

ri

•^

o

2i

j:

Oh
72

i

bf,

"^

1>

~-

1)

c

r^

t?

3

■"^

^

X

-u

bf

"^

i

^

£

s~C

h c c u:

03

_-M ^

:i

C

> >-.

>C'

C

o

- -^

■^

c

C ir"

"^

1>

JZ ^

c«

r> c;

^^

;^-

1^

^

o

^

• 5f

■iJ

11

^

1 "

"^

o

^.2

■^

o

^^

•&

1

• c

C

u

_ !^

0

g

0

tj Jt

■^

U

^c

1

U

Si^

<i

u

^■§

<i"

w

•o Ji

>5

Oh

-S

C/2

"C '"

i^

S c

1

h

:S|

z

^«

O

C'

X
U.

& 2i

THE BIOLOGY OF

STENTOR

BY

VANCE TARTAR

Department of Zoology
University of Washington

PERGAMON PRESS

OXFORD • LONDON • NEW YORK ■ PARIS

1961

I'I'UCAIVION IM<i:SS l/l'I).

Heading ton Hill Hall, Oxford

I & 5 I'il'^-roy Siiimrr, London 11'./

iM':u(;aivi()N pki.ss inc.

122 luisl SS'li StrrrI, Nnu Vorii ;,'..', A^. V.

Stdlln ('enter O/o, ()<><> Wihhire Ihntleiuml

Los /hif^eles 17, (California

IM'IUCAMON I'KI'SS, S.A.R.I,.
..' / line (les L'eoles, Paris I ''

IMCUCAMON I'KI'SS (J.m.h.ll.
Kaiserstrassv 7^'j, Frati/ijurt am Main

(\>pyrij^lit (C) \i)()\
I'l'lUiJAMON I'hhss In<'.

Lihiiuy <>1 ( 'oii^Mcs.s ( 'iird No. f)0 i57'4

Sil in hnfuint 11 nn i.'pt. <in<l I'tinttd in (hinl IUil(nn tit

nil': MAY rRi;!-; pki:ss, s ri;vi;N aci: , iiiiicrs.

PREFACE

In this monograph I have attempted to summarize all that has
been learned about a certain group of ciliate protozoa pre-
eminently suitable for class-room study and research. To this end
1 have tried conscientiously to review all the literature of Stentor
so that the reader will need to turn to original sources only for
minor details. A few publications were not available to me though
I had reports of their contents, and these are so indicated in the
bibliography. Many points I have been able to substantiate
myself, and I have taken this opportunity to include previews of
work in progress and miscellaneous observations from my own
experience with these ciliates.

Naturally I am keenly aware of our indebtedness to all who
have labored in this field and of my responsibility for reporting
their studies accurately and commenting upon them fairly. When
I use the words ''seem" and ''apparently" with their goading
provisionality, this is not in derogation of a fine piece of work but
simply means that confirmation of results assures a firm basis for
further investigation. It is often surprising what differences may
appear in both procedure and interpretation when different
individuals undertake the same problem or even the same approach.
On the other hand, there has been unnecessary duplication of
effort for lack of a comprehensive review as here attempted, and
I have myself been at fault in this regard. Studies in which
Stentor appears as suitable material in a wider context — frequently
biochemical — may also be led astray because investigators are
not aware of relevant aspects of the biology of this animal which
are crucial for proper interpretations. I hope to have provided
the basis or background for extensive further researches.

The illustrations from my own studies do not represent general
conceptions but specific cases drawn from laboratory records.
Therefore they offer the basis for different interpretations, if these
need be made, as well as suggesting many directions for further
study.

VI PREFACE

I wish to thank Dr. Gerald Kerkut for proposing this book,
and the pubUshers for their care in its reaUzation. Emogean
Saunders Tartar, my wife, prepared the manuscript. My own
studies have been generously supported by the American Cancer
Society and, currently, by our National Institutes of Health.

Vance Tartar
Aquaterre

Nahcotta, Washington
U.S.A.

CONTENTS

Preface

I Introduction
II Form and Function in Stentor

III Behavior

1. Food selection

2. Swimming

3. Avoiding reaction and learning

4. Response to light

5. Response to heat and electric current

6. The question of sensory cilia . .

7. Cystment . .

IV Fine Structure

1. Feeding organelles

(a) Frontal field

(b) Oral pouch

(c) Membranellar band

(d) Gullet

2. Holdfast . .

3. Cytopyge

4. Contractile vacuole

5. Cortical structure

(a) The cell surface . .

(b) Granular stripes; nature of the pigment and

granules. .

(c) Clear stripes and their fiber systems

(d) Fiber systems of doubtful status

(e) The cilia . .

6. Fine structure of the nuclei

7. The endoplasm . .

V Grov^th and Division

1. Growth . .

2. The course of normal division . .

3. Nature and location of the fission line

4. Incitement to division . .

5. Persistence of division . .

PAGE
V

II
II
17
19
22
24

25

26

28
28
29
29

30

34
37
40
40
42
42

43
49
54
56
57
58

61
61
67

75
81

84

Vll

Vlll

CONTENTS

(a)
(b)
(c)
(d)
(e)

VI Reorganization . . . . . . . . . . . . 91

1 . The course of reorganization . . . . . . 91

2. Analysis of the reorganization process . . 94

3. Stimulus to reorganization and the significance of this
process . . . . . . . . . . . . 98

To replace defective mouthparts? . . 98

Response to change in the medium? . . . . 99

Need for nuclear reorganization? . . . . loi

For growth of the adoral band? . . . . . . loi

Need for adjustment of nuclear dimensions? . . 102

VII Regeneration . . . . . . . . . . . . 105

1 . The course of regeneration . . . . . . . . 105

(a) Oral regeneration and its requirements . . 105

(b) Regeneration of the holdfast . . . . no

(c) Reconstitution of the normal shape . . . . in

2. Nuclear behavior during regeneration . . .. .. 113

3. Effective stimulus to regeneration .. .. .. 115

4. Time for regeneration .. .. .. .. .. 117

5. Minimum size of regenerating fragments .. .. 120

6. Adjustments to proportionality of parts .. .. 123

7. Can mouthparts and membranelles be formed in situ? 127

8. Repeated oral regeneration . . . . . . . . 130

9. Blockage of regeneration .. .. .. .. 131

VIII Activation and Inhibition of the Oral Primordium 135

1. The course and spectrum of cell interactions . . . . 136

2. Timing the period of activation . . . . . . 142

3. Relation of the macronucleus to activation and

inhibition , . . . . . . . . . . . 143

4. Relation of intact feeding organelles to activation and

inhibition . . . . . . . . . . 144

5. Synchronization of developing primordia . . . . 149

6. Activation in reorganizers and dividers . . . . 152

7. Rerouting the oral primordium . . . . 152

IX Primordium Development . . . . . . 159

1. Normal location and development of the primordium 159

2. Primordium development under abnormal conditions 164

3. Determination, or the progressive specification of the

oral anlage . . . . . . . . . . . . 170

4. Induction of mouthparts formation .. .. .. 174

5. Repair, mending, and joining of primordia .. .. 176

X The Primordium in Relation to the Stripe Pattern 179

1. Nature of the normal primordium site . . .. 179

2. Production of supernumerary primordia 180

3. Abnormal primordia correlated with abnormal striping 184

4. Primordium formation in loci of minor stripe contrast 188

CONTENTS

IX

5. Competition among loci of stripe contrast;

regeneration and obliteration of primordium sites

6. Exceptions

XI Polarity

1 . Fixity of structural polarity

2. Rate of regeneration in relation to the polar axis

3. Gradients in head and tail formation

XII Fusion Masses of Whole Stentors
I. Simple masses and biotypes
2.
3.
4-
5.

Adjustments among formed ectoplasmic organelles

Larger masses and reduction of oral valency . .

Incomplete oral differentiation . .

Absence of fission

Tubes and ciliated vacuoles

XIII Reconstitution in Disarranged Stentors

1. Minced stentors . .

2. Other disarrangements of the normal cell pattern

XIV Analysis of Stentor Through Its Response to

External Agents

1 . Action of the membranellar band

2. Coordination of body cilia

3. Ciliary anaesthesia

4. Anaesthesia of myonemes

5. Comparison of osmotic effects to cooling

6. Acceleration of division . .

7. Changes in state of the protoplasm

8. Tests for an antero-posterior metabolic gradient

9. Acquired tolerance to external agents . .

10. Shedding of pigment and pellicle

1 1 . Shedding of the membranellar band . .

12. Morphogenetic effects . .

13. Inhibition of growth by X-ray, and other effects

14. Effect of temperature on size . .

XV Metabolism . .

1 . Effects of starvation

2. Storage and utilization of nutrient reserves

3. Respiration

4. Digestion . . . . . . " . .

5. Symbiosis with green algae

6. Parasites of stentor

7. Abnormal stentors

(a) Depigmented stentors

(b) Over-pigmented stentors

(c) Amorphous stentors

190
191

195
195
201

202

205
205
211
213
215
215
215

220
220
226

232
232
238
240
241

244
244

245
246
248
250
252

254
256

257

259
259
263
265
266
267
273
274
274
275
276

CONTENTS

XVI Behavior and Functions of the Nucleus

lO.

II.

12.

13-
14-
15-

Location of the macronucleus . .

Clumping of the nucleus

Nodulation

Equivalence of macronuclear nodes

Shape, size and number of nuclear nodes

Control of nuclear behavior

Necessity of the nucleus for oral redifferentiation

Reconstitution of shape in relation to the nucleus

Functioning and re-formation of vacuole and holdfast

in enucleates
Behavior of enucleates . .
Digestion in enucleates . .
Survival of enucleates . .
Consequences of excess nucleus
Consequences of reduced nucleus
Delayed renucleation

XVII

Tow^ARD A Genetics of Stentor

1. Interspecific chimeras and nuclear transplantations

2. Racial differences

3. Conjugation

XVIII Species of Stentor

XIX Techniques . .

1 . Collecting

2. Culturing . .

3. Survival on slides

4. Staining . .

5. Cutting methods. .

6. Grafting . .

7. Minceration

8. Enucleation and renucleation

XX Extensions . .

I. Stentor and other ciliates

Hypotheses concerning morphog
Stentors and cells
Stentor and metazoa
Theoretical considerations

Bibliography of Stentor
Other References Cited
Author Index
Subject Index

of ciliates

CHAPTER I

INTRODUCTION

What are stentors good for ?

One would like to say that these exquisite little organisms are
a sufficient wonder in themselves and that to study them as a part
of nature is an expression of natural curiosity and that happy
relationship between subject and object which carries its own self-
justification. Doubtless this delight sustains the investigator
throughout what would otherwise be the weary and protracted
pursuit of other ends. Moreover, the experience of science has
shown that pursuing a subject for its own sake is likely to turn up
clues to which a more ulterior approach would be blind.

Yet this hobby-like vitality of interest is not sufficient. Our
studies become truly exciting and fruitful to others only when
they lead to general principles on the theoretical level. Of necessity
we have to start with some specific organism, woefully unique in
itself, out of the immense variety of existing forms of life, yet we
want our study eventually to be relevant to general problems of
biology.

To rew^ord the question, we may ask what particular advantages
Stentor may have with respect to these larger ends, that an entire
book should be devoted to this one type of organism.

Most outstanding is that on stentors one can easily perform a
wider range of micrurgical operations than on any other uni-
cellular organism or tissue cell, remarkable though the experiments
with Amoeba and the single-celled alga Acetabularia have been.
These operations are made possible by what for lack of a more
subtle analysis we have to call the consistency of the endoplasm
which permits grafting whole animals or cell parts in any number,
combination, or arrangement desired. The relatively large size
of these cells is a help, though the largest are no bigger than the
period at the end of this sentence. Stentors, unlike amcEbas,
exhibit a high degree of visible cytoplasmic differentiation and in

2 THE BIOLOGY OF STENTOR

several species, including the commonest, the cortical pattern is
conveniently outlined by a series of pigmented stripes so that the
organization of individualities and the identification of local areas
and grafted patches is quite evident in the living material. This
offers many advantages. With cells and patches self-marked,
operations can be guided and specified, and the whole range of
classical grafting experiments and transplantations can be ex-
tended to the cell level of organization. Fixing and staining are not
required to follow the performance of grafts so that experiments
proceed rapidly and can be done in sufficient number for valid
conclusions. Complex, specific, asymmetric elaborations of form
increase the number of responses to alteration of the system which
we can observe and measure, and render Stentor highly relevant
to the great unsolved problem of organic form. A cytoarchitecture
which has repeatedly been postulated as necessary to explain the
orderly development of eggs is visibly displayed in stentors and
does in fact play a cardinal role in their morphogenesis.

Different species of Stentor can be grafted together almost as
readily as cells and cell parts of one species. The cytoplasms and
nuclei of two or even more species can be combined in any desired
proportions, and this is a new method of ''transduction" by
which not only different genetic material may be added to a cell
but also alien cytoplasm. These chimeras persist and do not fall
apart, cytoplasms mingle and nuclei are maintained at least for a
considerable time in foreign cytoplasm on which they often exert
a visible influence.

The macronucleus, which alone is significant in the vegetative
life of stentors, is clearly visible in the living animal. Enucleations
are not difficult. Stentors therefore provide additional examples
in which the contribution of the nucleus may be assessed by
determining the consequences of its absence. The more types of
cell in which this operation is possible the more likely we are to
come to general conclusions. Moreover, the extended form of the
nucleus in Stentor allows us to remove all or only a desired portion
of it. Such quantitative operations, when combined with additions
of enucleated or highly nucleated cytoplasmic masses vary the
ratio of nucleus to cytoplasm in extremes not heretofore possible.
Interesting consequences of this imbalance are evident in stentors.

The same properties which permit grafting also make possible

INTRODUCTION 3

the transfer from one cell to another of nuclei retained within a
thin envelope of endoplasm, allowing enucleated cells to be
renucleated at any time or the nucleus of one species to be sub-
stituted for that of another with practically no admixture of
cytoplasms.

When added to the simpler experiments on stentor fragments
in which parts become wholes, the possibilities afforded by these
operations and their permutations appear endless. The organic
integration by which new individualities become one can be
explored. Nucleo-cytoplasmic interactions and the nature of
species differences are opened to inquiry with fresh material.
Problems of polarity can be explored in heteropolar grafts of cells
and cell parts. Cell differentiation under a variety of conditions
occurs before our eyes. The intimate nature of aging, necrosis,
and damage by various external agents can be investigated by
testing the revival of " sick " animals after grafting them to healthy
cells or cell parts in stentors as in the important work which
Daniels (1958) is doing with giant amoebas. These are only a
sampling of what can be done.

Stentors share, now or potentially, experimental advantages
common to many protozoa. As free-living cells they are directly
affected by alterations in the fluid medium ; and it has been found
that certain substances added to the medium may produce profound
effects in the behaviour, reproduction, and morphogenesis of
stentors. Since stentors undergo sexual conjugation at times,
genetic experiments should eventually be possible. Irradiation or
other treatments at the time when the nuclear complement is
reduced to a simple anlage could produce mutations as genetic
markers and indicators of cell activities. If mating types appear
as in other ciliates we would have a differentiating characterization
in the expression of which the roles of nucleus and cytoplasm
could be investigated by direct operations. A fundamental need
in the cancer problem as well as of general biological under-
standing is to learn precisely what incites the cell to division,
whether it be an egg, a tissue cell, or a protozoan. The great
amenability of stentors to manipulation encourages us to search
for the answer in them.

Each of these experimental possibilities is important in itself
but their unique combination within one organism makes Stentor

4 THE BIOLOGY OF STENTOR

one of the classical types in biology. These leads can more intel-
ligently be pursued in all their modern implications, if one has a
thorough background in the biology of Stentor. For this we can
draw on all the many studies of stentors extending far back into
the previous century as well as our own good observations. In
what follows we shall therefore try to summarize all that is known
about the ciliate protozoan, Stentor. Many of these details may
seem tedious unless one keeps in mind the marvel that so much
can be learned through the patient efforts of a long procession of
able students about a single type of minute organism, and the
wonder that so tiny and seemingly insignificant an animal reveals
on close inspection so much integrated complexity of form and
function. a

CHAPTER II

FORM AND FUNCTION IN STENTOR

To BECOME acquainted with Stentor we begin with an account of
those features of the organism which are open to simple observa-
tion, emphasizing S. coeruleus which, because of its commonness
and reveahng pigmentation, has been the favored type for study;
but there is no reason to beheve that even httle-known species
differ fundamentally in basic Bauplan or manner of living.

Several of the early investigators were impressed by what
seemed to them to be the great variability in form of stentors, as
the species name polymorphus and multiformis imply ; but the form
of Stentor is no more indefinite than that of an earthworm merely
because it extends and contracts. The shape of Stentor is simply
that of a cone capable of extension or of contraction into a sphere.
Attachment is by the point of the cone, a small enough area to
permit voluntary release; while the feeding organelles are at the
broad end where they can most effectively produce a vortex
bringing particulate food to the animal from a large region of the
medium which lies beyond it like an imaginary extension of the
cone. Set free and swimming, the shape of the cell under what
seems to be minimum tension is, as Merton (1932) remarked,
that of a gently rounded cone. Shortening of contractile elements
pulls the cone into a sphere, and extension is the result of elasticity,
pulling by the feeding membranelles, and possibly transverse
contraction.

The surface is covered by alternating longitudinal stripes of
two kinds: bands of granules, often colored, forming stripes which
increase in width in an orderly way around the cell and, between
these, clear stripes of relatively constant width which bear the
rows of body cilia as well as a complex of fibers probably respon-
sible for longitudinal contraction and ciliary coordination (Fig. i).
To accommodate for the decreasing cross-section of the cone the
granular stripes taper near the apex and some of both kinds of

b THE BIOLOGY OF STENTOR

Striping stop short of the posterior pole. Cross-sections of con-
tracted animals show a corrugated surface with the clear stripes
lying in valleys between the raised granular stripes.

A thin pellicle forms the outermost surface of the cell. It could
conceivably be the product of secretory activity, special elaboration
of which may produce cyst walls and the cylindrical cases found
in some species. The pellicle is not completely elastic and on
contraction is thrown into transverse folds over the granular
stripes, causing the surface of the rounded animal to appear like
a scalloped theater curtain.

The graded variation in the width of the granular stripes
provides a fundamental asymmetry to the pattern of the cell. In
the oral meridian these bands are narrowest and they gradually
increase in size around the cell from left to right so that the widest
bands eventually come to lie next to the narrowest in a locus of
stripe- width contrast on the oral or ventral side. Stripe multiplica-
tion occurs in this region, the widest granular stripes being split
by the interpolation of new clear bands. Because this splitting
generally proceeds from the anterior end and does not follow all
the way through to the posterior, there results a triangle of shorter
stripes which was called the ramifying zone by Schuberg (1890).
This area is also the region in which the oral primordium appears.

All stentors attach by a temporary holdfast organelle at the
posterior end. In undisturbed cultures only a few animals will be
found freely swimming (Gelei, 1925), and this may be taken as the
usual condition in nature.

Stentors sink to the bottom in agitated cultures, their specific
gravity being greater than that of water. Attachment may serve
the purpose of keeping them with minimum expenditure of energy
in favorable locations toward the surface of the water where oxygen
is abundant. It may also be assumed that the effectiveness of the
feeding vortex created by the peristome is increased when the
animal is attached.

Adherence is firm. Animals just detached remain sticky at the
posterior end and Jennings (1902) saw them dragging trails of
mucoid material behind. If water is pipetted out of a culture,
the stentors remain fastened to the sides in a watery film. Rapid
evaporation may even leave animals stuck to the rim of the vessel
where they dry and die, seemingly unable to loose themselves.

FORM AND FUNCTION IN STENTOR 7

In pipetting animals from the sides of a vessel the holdfasts are
often torn off because of their firm adherence. Yet stentors can
release their hold at will under favorable conditions in order to
search for a better environment. Thus in unfed cultures many
animals will be found on the move as if searching for food.

Some stentors form cases and are still more sedentary. Jennings
(1902) described how roeseli seemed to explore the substratum with
its anterior end and when a likely spot was found the tail bent over
and attached. Then mucous was secreted over the posterior half
of the body while the animal moved backward and forward on its
side for about two minutes as it secreted an elongated cylinder.
The tube was later compacted somewhat by subsequent contrac-
tions of the stentor.

In describing the feeding organelles at the anterior end we shall
use the simplest and most unambiguous terms. Confusion and
synonymy have arisen in the past largely from unjustified attempts
to homologize these organelles with the upper parts of the human
alimentary canal, with the result that stentor should have both a
pharynx and an esophagus. There is said to be a buccal or cheek
cavity, yet what is called the mouth or cytostome does not in fact
open into this cavity but is homologous with the anterior pyloric
sphincter. The membranellar band is usually called the peristome,
but it does not encircle the cytostome, and the term "peristome"
was originally used to designate a special fold generally running
alongside this band (see Johnson, 1893). Regardless of more
precise designations, however, it will be convenient at times to
refer to the entire set of feeding organelles at the anterior end as
the '*head"; to the oral pouch, gullet, and cytostome as the
'* mouthparts " ; and to the holdfast as the " foot ".

The anterior end or frontal field is covered with alternating
clear and granular stripes the same as the lateral body wall from
which it is derived (Fig. i). Being newly formed, the granular
stripes there are narrow and the clear bands with their ciliary rows
or kinetics are close together. Bordering and almost completely
enclosing the frontal field is a band of membranelles which nor-
mally spirals always in one direction as shown in the figure.
Fully extended, the frontal field and bordering membranelles
take the form of a broad funnel. At the left side, in most stentors
the frontal field dips down sharply with its striping into an oral

8 THE BIOLOGY OF STENTOR

pouch which is often called the buccal cavity. The membranellar
band runs along the outer edge of this pouch and then coils
sharply inward into an invaginated tube which itself coils about
one turn into the cell. This tube we shall call by the rather non-
commital term of gullet. Food vacuoles are separated off the inner

IroniaL Held.

membrsLncUar band
clear harder stripe
oraiLpomJi

contractile vaciwle
cv pore

ruJlei

{ood vacude

macromiclesLr node.

micronus^us

hxMIast

Fig. I. Descriptive diagram of Stentor coeruleus.

end of the gullet, which therefore has a temporary film-like
closure to prevent endoplasm from escaping into the gullet but
capable of acting so as to allow passage of food into the cell.
Probably the ectoplasmic lining of the gullet simply ends here in a
thin membrane which can stretch and increase to form the wall

FORM AND FUNCTION IN STENTOR 9

of a new food vacuole and then close behind it as the vacuole is
pinched off into the interior. But vacuolar walls can arise de novo
as is seen when active rotifers are ingested and thrash around
inside the cell but are later re-encapsulated and digested.

In feeding, the membranellar band by coordinated beating
creates a powerful vortex which draws in particulate food organ-
isms, large or small. Impinging on the frontal field, particles are
moved by its cilia toward the oral pouch in which they are trapped
and concentrated, whirling around within the cavity. In this region
the food is apparently tested. If undesirable or in excess, particles
are then ejected over the outer rim of the oral pouch and carried
toward the base of the animal, away from the feeding vortex, by
the backward-beating lateral body cilia. If to be ingested, the food
is passed down the gullet by reason of its ciliated lining and is
further concentrated while peristalsis of the gullet forces the
material into the interior.

After digestion the residue in the food vacuole is cast out
through the left anterior wall of the cell below the pulsating con-
tractile vacuole. Especially when the stentor has been feeding on
tiny flagellates, many exhausted vacuoles accumulate and fuse in
this region, forming a very large bolas which requires about one
minute to be voided. Whether there is a permanent anal opening
or cytopyge may still be questioned.

On the other hand, the exit of the contractile vacuole is visibly
persistent. Moxon (1869) observed openings in the broad granular
stripes exterior to the contractile vacuole which is always located
in the anterior left side of the cell. These openings are evident in
pigmented coeruleus as clear spots. One, at least, of these openings
expands noticeably when the contractile vacuole is voided,
assuring their identification.

Larger species of Stentor have a moniliform macronucleus
composed of many nodes lying within a common nuclear mem-
brane. This nucleus lies underneath the ectoplasm and is deployed
in a characteristic way as shown in the illustration. Adherent to
the macronuclear nodes or nearby are many micronuclei. Smaller
species show a single compact macronucleus; and micronuclei,
which are very tiny in stentors, have not been seen in all species.

Endoplasm fills the interior of the cell and is in irregular
cyclosis, possibly because extension and contraction itself pro-

10 THE BIOLOGY OF STENTOR

duces sufficient mixing. The endoplasm has been described
variously as alveolar or reticulate and contains reserve materials
in the form of droplets and granules.

When maximally contracted, stentors become nearly perfect
spheres. Most stentors are also capable of remarkable extension.
When attached and feeding the body may stretch out to three to
six times the diameter of the contracted animal while the feeding
organelles expand vv^idely to produce the stentorian or trumpet
shape. All the complex structures of the cortical layer are therefore
capable of w^ide displacements though maintaining their precise
pattern and organization.

Next we shall consider what is known of the behavior of
stentors and then we can deal with the fine points of structure in
terms of which this behavior is to be explained and which
demonstrate the highly complex and precise achievements of
morphogenesis.

CHAPTER III

BEHAVIOR

In broad perspective, multicellular animals enjoy periods of
relaxation or inactivity, but their constituent cells are ever active
as long as life maintains. Unicellular animals share with tissue
cells this unresting activity, and stentors are no exception. In their
reproduction, continual search for food, and avoidance if possible
of unfavorable surroundings, the abiding impression is that
stentors are always busy. Cessation of swimming and attachment
by the holdfast is only the prelude to active feeding. If we define
behavior as altered response to changing conditions, unresting
stentors are continually behaving. Observing them even briefly,
one is struck by the appearance that their activity is not mechan-
istically simple, though they may be high-grade automatons. If
we place ourselves in the position of early investigators, the
wonder is renewed that even in these minute and lowly forms of
life we can undertake to analyze behaviour.

I. Food selection

Food selection in Stentor seems to have been clearly demon-
strated in a nice series of experiments by Schaeffer (1910). He
recorded the uptake by coeruleus placed for a time in a prescribed
suspension of particles as well as observing what happened when
single particles were introduced one at a time with a capillary
pipette into the feeding vortex. In one of the "hand feeding"
tests, for example, 12 Phacus were ingested and only three rejected,
while 13 indigestible sulphur particles were rejected and only
three taken in. In another test all 50 Phacus presented were
ingested, while 18 starch grains were rejected and only one
accepted. Size was not determinative because the starch grains
were four to one-eighth times the size of Phacus. Again, 21 Phacus
and I starch grain were eaten, while 7 Phacus ^ 12 grains, and 11
glass particles were rejected. Euglena was preferred to Chilomonas.

II

12 THE BIOLOGY OF STENTOR

Phacus and Eiiglena recently killed by heat or alcohol were eaten
as readily as live ones. Stentor even discriminated between two
species of Phacus, predominately accepting triqueter and rejecting
longicaudus but there were no observations on whether the latter
was actually indigestible.

In mass feedings from mixtures of equal parts, a stentor took in
1,500 Chlamydomonas, 85 Euglena, and 10 carmine particles. In
carmine alone 20 units were taken in. Hence generally less carmine
was eaten when food was present than when not, and carmine
was even rejected preferentially when much in excess over food
particles in the mixture. In India ink alone only 3 granules were
taken in, so the greater number of mistakes made with carmine as
compared with the smaller ink particles was the reverse of that
found with paramecia. These tests indicate that when fed, stentors
become more selective but they also rejected more of the favored
items as if no longer hungry.

Conversely, hungry stentors were found to be less selective.
This may explain apparently contradictory observations, because
an investigator testing the ingestion of a given type of material
would be likely to use starved, clear stentors in which confusing
food vacuoles were not already present. Thus on occasion stentors
will ingest considerable numbers of fine carmine and ink particles
(Schuberg, 1890; Jennings, 1902). Before Schaeffer's studies,
Jennings had therefore concluded that there is no selection after
the material reaches the oral pouch, ''dissatisfaction" with the
meal resulting only in cessation of feeding and turning in a new
direction. Prowazek (1904) observed that coeruleus ate free-living
Chlorella though it could only partially digest them; and I have
found plentiful "food" vacuoles in samples of this stentor left
unnutrified for nearly a month and apparently re-feeding on waste
materials. Johnson (1893) remarked that coeruleus eats the alga
Scendesmus in quantity but apparently does not digest it and
quickly passes this material through the cell. Hence even when
indigestible materials are eaten, the animal can short-circuit to
the cytopyge food vacuoles with useless contents.

On the evidence, food selection does occur in Stentor, though
by no means perfect and distinctly related in its acuity to the state
of the organism.

The next consideration is the basis for this selection, and

BEHAVIOR 13

Schaeffer's observations on this point are interesting though
negative. As noted, Hve and dead food organisms were not dis-
criminated nor were fragments versus whole organisms. When
mixed in sugar, beef extract, and other solutions, ink and carmine
particles were still predominantly rejected. Likewise, I have
observed that the empty but well-formed hulls of long-dead
rotifers were quickly rejected at the same time that their live
fellows were being eaten. Therefore neither size, shape, taste nor
activity of food particles seem to be the basis of selection, which
remains a considerable mystery. Schaeffer's results are perhaps
the more remarkable because his stock cultures were being fed on
something entirely different, viz. ''small paramecia", doubtless
with bacteria; and materials like sulphur grains, carmine, and
glass particles are probably never encountered by stentors in
nature. His demonstration of food selection is rendered more
credible by Lund's (1914) evidence of similar discrimination in
the related Biirsaria truncatella.

In a still more closely allied genus, Parafolliculina, Andrews
(1947) found that an increase of 10 °C doubles the rate of feeding
or the number of food vacuoles formed in a given period of time.
Very likely it is the same in stentors, increased feeding being the
consequence of thermal acceleration of the membranelles (Sleigh,

1956).

It remains, if possible, to locate the site of food selection as we
follow the course of digestible material during the feeding act.
Particles drawn in by the vortex produced by the membranellar
band impinge on the aboral side of the funnel-shaped frontal
field and are carried into the buccal pouch by the oralward
beating of the rows of frontal cilia. In clouds of carmine, according
to Schaeffer, the frontal field cilia may beat circularly, forming
balls of this material which then fall over the edge of the disc.
If so, this would constitute pre-oral selection, but Dierks (1926a)
could not confirm this behaviour; and I found that the feeding
vortex simply creates a locus of nearly zero water velocity at the
non-oral end of the membranellar band where granules collect
until a mass is built up and falls into the rejection current. Current
velocity over the frontal field itself is too great to permit such
accumulations as Schaeffer described.

14 THE BIOLOGY OF STENTOR

When items of considerable size, like small rotifers or hypo-
trichs, fall into the oral pouch they are trapped by its partial
closure as already indicated by Johnson (1893) and Andrews (1946).
As I often observed, the rim of this cavity is independently con-
tractile and closes a widely opened pouch until only a narrow
orifice is left while the rest of the frontal field remains completely
expanded. Cilia from extensions of the ciliary rows in the frontal
field into the pouch cause the particles to spin around inside.
Schaeffer stated that there was less of this looping if the animals
were either hungry or well-fed, as if in the first case they were in
a hurry to ingest the food while in the second they would not
bother to test it. He therefore thought that the oral pouch is the
organ of food selection; but Dierks maintained that selection
occurs principally at the opening into the gullet and I am inclined
to agree with him from observation that particles are not rejected
until after they have reached and spun around for a moment in
this region. As Schaeffer noted, particles may be rejected at the
same time that others are being ingested, so that selection is indeed
precise and implies a fine coordination. Even after particles enter
the gullet reversal of the cilia there may eject undesirable items,
but once they pass to the lower end the cilia invariably carry them
inward, according to Dierks.

Food is propelled in the gullet not only by specialized body
cilia but also by the spiral extension inward of the membranellar
band. The entire gullet seems to be formed by invagination and
extension of this band and of originally surface ectoplasm lying
adjacent, both spiralling inward. An orderly contraction of ecto-
plasmic myonemes thus carried into the gullet could therefore
produce the peristalsis observed by Dierks, which apparently
comes into play when large objects are swallowed. Dierks also
confirmed that Stentor is more selective of what it ingests as the
cell becomes replete.

The pouch and gullet, like the cell surface in general, are
capable of great extension and contraction. In cannibalization they
open wide enough to accommodate a fellow stentor nearly as large
as the predator. Gelei (1925) therefore thought that the fine mesh-
work surrounding the gullet which he observed in sectioned
animals is to prevent tearing of this organelle when greatly
stretched. He also found that the force of closure of the oral

BEHAVIOR 15

pouch was such that pieces of stentor prey could be bitten off.

As in many other cihates, the food vacuoles after they are
pinched off from the inner end of the gullet may be guided into
the interior by long fibrils dangling therefrom, first described in
Stentor by Schuberg (1890). But these could also serve a different
purpose. Andrews (1946) observed them in gullets everted by
pressure. When released the gullet can reinvert in only 10
minutes and this may be accomplished by traction of the fibers
in question.

Perhaps a nice point of morphology is that the rim of the oral
pouch over which rejected particles are dumped is definitely
below the level of the frontal field and membranelles so that
rejects probably do not return to the oral stream.

Gullet ciha and membranelles can work independently, for
Dierks noticed that ingestion may occur while the membranelles
circling the anterior end have for some reason stopped.

Cannibalism has been observed in the three species of Stentor
most commonly cultured and may also occur in others. Ingestion
of its fellows by coeruleus was first reported by Johnson (1893)
and was the subject of a special study by Gelei (1925) who also
noticed cannibalism in roeseli; and Ivanic (1927) claimed that
cannibalism occurs in polymorphus. I recorded indubitable evidence
of cannibalism in all of the 9 stocks of coeruleus which I have
growing in my laboratory. To paraphrase Gelei, at least three
problems come to mind in regard to this peculiar food choice:
Why stentors come to eat each other, how they are able to ingest
such large objects, and what the consequences are for the can-
nibalizer, particularly whether it is able to digest its own species
of protoplasm. These topics will be considered in that order.

Gelei noticed that hunger or the absence of other food organisms
is not the cause of cannibalism, as may be inferred from the fact
that stentors will ingest more of their fellows when they already
have one or more of these huge '* meals " in the process of digestion.
I found one coeruleus with five others inside. I also noticed can-
nibalism to be most frequent in cultures only a day or two fol-
lowing their nutrification. But neither is satiety the cause; for
cannibalism is found in starving samples, but not as frequently as
one might expect. These observations disprove Ivanic's contention
that cannibalizing stentors and other protozoa have a need and

l6 THE BIOLOGY OF STENTOR

therefore a hunger for their own type of protoplasm because of
some deficiency which has developed. Also it seems unreasonable
to suppose they could correct this lack by ingesting animals in the
same culture, therefore subject to the same deprivations. Likewise
we have to question Johnson's plausible and attractive suggestion
that cannibahsm should help the species to survive a period of
scarcity, although an ** ecological" study and sampling of un-
disturbed cultures might demonstrate this to be the case. Gelei
thought cannibalism a racial trait in coeruleus but its occurrence
without exception in a wide selection of stocks is against this
interpretation.

Stentors ingest only free-swimming stentors because the prey
is taken in by the attenuated tail end. By agitating the culture jar
to set animals loose, Gelei was able to increase the incidence of
cannibalism from about 1.5% to 4.6%, but never more. The
posterior point of the animal to be eaten is drawn deep into the
gullet where it is held in spite of the prey's rotating and attempting
to escape. In fact, the victims may usually manage to escape and
I believe no one has actually observed the swallowing, but this
may be due to disturbance by bright illumination under the micro-
scope. As mentioned, the rim of the oral pouch shows contractions
that may even bite off the posterior end of the prey. Ingested
animals are usually smaller than the predator, yet Gelei stated that
cannibals may swallow animals larger than themselves. Once inside,
the victim is not surrounded by a food vacuole at first; therefore
it remains alive and actively rotating for perhaps the better part
of a day, during which it can be released and always, in my experi-
ence, will recover. If not released the prey is eventually surrounded
by a membrane and enclosed as a food vacuole within which it
stops moving, dies, and becomes wrinkled within an hour. Stentors
will also eat individuals of a diflFerent species of the same genus.
I have frequently found polymorphus ingested by larger coeruleus.

There is no doubt that stentors, like other cannibal ciliates,
can digest their own kind. As Gelei described the process, coagula-
tion is the first sign of digestion and the corpse becomes friable.
Fat spherules then appear in "astonishing" numbers; this could
be due to the coalescence of pre-existing lipoid droplets which are
very tiny and not easily seen. The nucleus is digested, its nodes
falling apart and becoming progressively smaller and more weakly

BEHAVIOR 17

Staining. Pellicle and myonemes are digested slowly. Cilia are often
the last to be digested and are first attacked at their distal ends,
then gradually disintegrating and swelHng towards the base. Gelei
stated that pigment granules are not digested, and certainly the
last stage of the food vacuole is a bright spot of concentrated pig-
ment; yet the fading of stentors during regeneration and starvation
may imply, as Weisz (1949a) maintained, that the animals are able
to assimilate their own coloring matter. Cannibals doubtless
receive advantages from their prey as food. However massive these
meals may be and however easily most of the substance of their
like may be assimilated, giant forms do not result, as in the case of
Blepharisma (Giese, 1938) and Stylonychia (Giese and Alden,
1938). Regulation of size in Stentor is therefore such as to prevent
gigantism or the production of forms two or more times the
maximum normal size.

2. Swimming

That ciliates rotate when swimming and describe a spiral path
through the water was first formulated by Jennings (1899) who
pointed out that such movement serves the same purpose as in
projectiles: by rotating, an asymmetrical body can maintain an
over-all straightness in the direction of its course (Fig. 2A).
Rotation is always predominately in one direction in a given
species. Jennings (1899, 1902; Jennings and Jamieson, 1902) found
that S. roeseli and coendeus, like most ciliates, including Parame-
cium, rotate to the left, i.e., front end of the animal rotates clock-
wise. Slightly curved toward the oral side, stentors also tend to
swerve in this direction so their course is a spiral. Bullington
(1925), who has made the most extensive studies of swimming in
ciliates, confirmed that three unnamed species of Stentor rotate
and spiral to the left. On the basis of his surveys Bullington
remarks that Stentor is with Coleps the only genus of more than
one species in which all members spiral in the same direction.
{Paramecium calkinsi for example rotates to the right.) This
generalization may be valid, for I found that the new species
introversus rotates to the left, as does coeruleus, polymorphus and
roeseli. When backing up coeruleus continues to rotate to the left,
as Jennings earlier noted for polymorphus.

i8

THE BIOLOGY OF STENTOR

Fig. 2. Behavior of Stentor.

A. Avoiding response : {a) normal swimming with left rotation
and slow spiraling; {b) contraction on adverse stimulation; {c)
backward movement even if stimulated at the base ; {d) turning
to aboral side with resumed forward swimming in new direction.

B. Graded response of feeding animal to adverse stimuli: {a)
undisturbed posture ; {b) turning away ; {c) feeding vortex stopped
and beating of body cilia reversed; {d) sharp contraction; {e)
holdfast released and animal swims away, abandoning lorica {e)
if present. With repeated stimuli the response decreases in the
direction e-^a. (After Jennings, 1902).

BEHAVIOR 19

Like those of other ciHates, stentor fragments which are not of
bizarre shape continue normal swimming behavior. Jennings and
Jamieson (1902) observed that isolated heads, tails, and mid-body
fragments of coeruleus rotate and spiral to the left. This indicates
that the direction of beating of the membranellar band during
swimming is not contradictory to the movement produced by the
body cilia, as well as that it is not the asymmetric oral structures
which produce the rotation. Rather, it is to be inferred that the
body cilia do not beat directly backward but obliquely, as Parducz
(1953) has elegantly demonstrated for Paramecium. Not only is
the direction of beating of each cilium oblique, the whole ciliation
is so coordinated that metachronal waves move slant-wise over
the surface of the cell, giving the appearance of rows of grain
moved in succession by the same gusts of wind. This apparently
calls for a wave-like impulse passing over the cell surface or the
successive '' firing " of adjacent ciUa down each ciliary row, as well
as coordination between adjoining rows.

When a swimming stentor encounters a noxious stimulus it
swims backward a little, reversing the beat of the body cilia and
stopping the membranelles while pointing them forward, turns
toward the aboral side and swims forward again, quite as with
Paramecium (Fig. 2a). The membranelles then begin beating again
as they are pointed backwards presumably to aid in the forward
progression. Merton (1935) avers that the membranelles sometimes
help in swimming backwards but I find them always held quiet
then. He confirmed that posterior fragments can spiral forward
and also noted that they, too, are able to swim backward '' at will ".

A common behaviour of stentors is to scoot over the bottom of
the dish with the oral surface apphed thereto, apparently to explore
for and break loose food particles like a vacuum cleaner. This is
the preferred method of feeding in the related genus Condylostoma.

The effect of various chemicals in narcotizing and reversing the
ciliary beat, as well as cutting experiments concerning ciliary
coordination, will be considered in Chapter XIV.

3. Avoiding reaction and learning

After feeding, the next commonest response of stentors is to
manage to remove themselves from the reception of noxious
stimuli. Jennings (1902) made a careful and interesting study of

20 THE BIOLOGY OF STENTOR

this avoiding reaction in roeseli and coeruleus, prodding them with
a glass needle or, by gentle use of a pipette, injecting into the oral
vortex substances of weak chemical stimulation, like carmine
particles. As if to conserve their status quo, the animals performed
a series of distinctly different responses of increasing intensity
until the stimulus was avoided (Fig. 2b).

First, a light touch as from an impinging rotifer which could
serve as food provoked no response in coeruleus, which merely
went on feeding, and roeseli even made the positive response of
bending toward the source of stimulation.

When the stimulus was a little stronger, as from a large, hard
object or a cloud of carmine particles, the stentors '* turned away."
Remaining attached but twisting one or two turns on the axis, the
animals then bent toward the aboral side always and therefore not
necessarily away from the source of stimulation. This reaction is
quite like the avoiding response of Paramecium: a fixed response
without reference to the direction of stimulus, repeated if un-
successful.

Third, the membranelles might stop and body cilia reverse for
an instant, thereby propelling the carmine particles forward and
away from the anterior end. Feeding currents then continued and
the reaction was repeated several times if the particles were still
encountered. This response sometimes occurred instead of turning
away, but the variability may have been due to the difficulty of
providing precisely graded stimuli.

If the noxious chemical stimulus still persisted, or if poked with
the needle, the stentors instantly contracted, slowly extended again,
and re-contracted if conditions were still undesirable. This reaction
could continue for fifteen minutes if carmine particles were kept
available or prodding repeated.

Fifth and finally, the holdfast was set free and the stentors swam
away seeking a new environment. Sometimes co^rw/^w^ detached and
swam away after the first stimulus but usually the other avoiding
responses intervened. From my own observations it appears that
stentors from cultures which have recently been fed are more
likely to persist in the feeding response and to give the graded
response, as if bothered by the interruption of a good thing,
whereas unfed animals are more likely to detach and swim away
at once, as if the negative stimulus finally prodded them to

BEHAVIOR 21

''decide" to go in search for food. However, the case-bearing
roeseli would not abandon its home by mechanical shocks alone
even though Jennings struck it with a glass needle for an hour.

Normal reactions of avoidance were also shown by stentors after
their " heads " had been excised (Jennings and Jamieson, 1902).
But stentors never became accustomed to truly injurious stimuli
such as salt solutions (see also Merton, 1935) or sharp poking,
though they learned to put up with a lot of minor disturbance.
This may happen frequently in nature and stentors growing among
Tuhifex have been observed to continue feeding and not contract
though constantly struck by the worms. Jennings provided a simple
demonstration of this accommodation by attempted equal impacts
with a glass needle repeated each time after re-extension of the
stentor. After about a dozen strokes there was no contraction
response unless the animal was poked several times. The longer
this was continued the greater w^as the number of strikes which
were necessary to elicit contraction, although there was some
irregularity probably due to inequality of impacts. Eventually the
animals detached and swam away. Sometimes there was a ready
response only at first, repeated proddings then eliciting no response
until the animals swam away. But in these cases he noted that the
stentors did not remain oblivious to the blows but twisted continu-
ally and turned away as if to avoid them, finally detaching and
swimming away. Hence in accommodation there was a reversal
of the sequence of avoiding reactions, for example, contraction
later replaced by merely turning away. Similar abolishment of
major avoiding reactions occurred in other contractile ciliates
{Epistylis, Vorticella, and Carchesium) and was noted by Holmes
(1907) in Loxophyllum.

This orderly change in response was not due to reaction fatigue.
About a minute was required for re-extension and this should have
been sufficient for complete recovery. Also, stentors could be kept
continuously contracting for an hour at a time, but they very soon
ceased responding to weaker stimuli. Nor could the response have
been due to sensory fatigue because the animals showed continued
appreciation of the stimulus (by turning away) and because stentors
subjected to strong mechanical blows or injurious salt solutions
continued reacting indefinitely. Schaeffer (1910), for instance, said
that coeruleus would swim backward, without spiralling, continu-

22 THE BIOLOGY OF STENTOR

ously for three hours in a dish of dense carmine particles. Therefore
if learning be altered response due to previous experience, this did
in fact occur.

The weaker the stimulus the more rapid the accommodation.
To a jet of water from a capillary pipette coeruleus responded only
once by contraction and thereafter merely bent in a new position.

If stimulated while swimming, Jennings (1899) reported that
polymorphus contracts and backs up a short distance then turns to
the right side always and swims off in another direction. This
response was invariable, regardless of the point of stimulation.
Though the posterior end was less sensitive than other parts, a
sharp blow here elicited the same response, which therefore
carried the animal toward rather than away from the point of
stimulation. Unlocalized stimulation such as jarring the dish also
evoked the same avoiding response, as did diffuse chemical
stimuli.

Unlike Paramecium and Chilomonas, stentors were completely
indifferent to bubbles of carbon dioxide or solutions of acids, not
showing the spontaneous aggregations of the former (Jennings
and Moore, 1901-02).

Even to relatively strong solutions of cane sugar roeseli showed
no avoiding reaction and responded by sudden contraction only
after the cell became obviously affected by osmotic pressure
(Jennings, 1902).

4. Response to light

Jennings (1902) noted that roeseli does not respond to light of
ordinary intensities. According to Schulze (1951) polymorphus
shows contrasting reactions to hght depending on whether sym-
biotic Chlorella are present: green animals appropriately went to
the lighted side of the aquarium but white forms collected on the
dark side. Hammerling (1946) stated th^t polymorphus is sensitive
to strong light and that cultures had to be screened, but in nature
I have found these stentors fully exposed to brightest summer
sunshine. In coeruleus the reaction seems to vary with the strain.
Testing 8 races cultured in the same manner, fed on the same
day each week, and producing animals of about the same intensity
of green pigmentation, I found that two showed a strong negative
response to daylight and quickly accumulated on the side of the

BEHAVIOR 23

dish away from the window ; two showed no response ; the remain-
der showed, fair, weak, or doubtful response.

That Stentor coeriileus moves to the dark side of an elongated
aquarium was reported long ago by Holt and Lee (1901) and
interpreted tropistically as due to an orientation in the field of light
away from the source of illumination. An extensive study of this
subject was made by Mast (1906) with different conclusions.
Swimming animals placed in a dish lighted from one side simply
showed repeated random avoiding responses of backing up and
turning until they found themselves headed away from the light,
and then they continued swimming forward to the dark side. This
response kept animals confined to the darkened end of an aquarium,
as if an invisible wall were present. That the confinement was not
due to warming caused by the light was shown by the fact that
paramecia which are quite sensitive to heat but not to light swam
readily into the irradiated area. Attachment of course prevented
the avoiding response of stentors, which did not even lean
(tropistically) away from the light; but if the light was strong
enough the animals detached and then gave the characteristic
response. Mast concluded that the anterior end was most sensitive
to light because, when the water was a thin film so that the stentors
could not face the light source above and received stimulation
only on their sides, collecting at the darker end was slower. This
point is confirmed by my observation that decapitated stentors no
longer avoid the light (unpublished).

Contrasting wdth coer ulcus, the yellow S, niger shows a marked
attraction to Hght, according to the studies of Tuffrau (1957). He
states that all parts of the animal appear to be equally sensitive to
the stimulus because there was no orientation in a field of light, yet
the head end could be more sensitive if the response is not tropistic
but one of trial and error. Although there was some individual
variability, most of the animals accumulated rapidly at an illumin-
ated opening in the side of a covered tube, and the response was so
strong that a spot of light acted like a trap in preventing stentors
from leaving it after they entered. The shorter the wave length of
light, the stronger the attraction: red elicited almost no response
and the aggregations increased as the spectrum shifted to blue,
violet, and ultraviolet. Animals dark-adapted for fifteen hours
recovered photoresponsiveness in an hour or two. The rapidity of

24 THE BIOLOGY OF STENTOR

response varied directly with the temperature, and this may not be
solely due to increased rate of ciliary beating but also to the en-
hancement of photoreactions in the cell.

Dabrowska (1956) was unable to get coeruleus to associate an
electric current with its response to light, and hence learning by
conditioned response was not demonstrated.

5. Response to heat and electric current

Alverdes (1922) found th^t polymorphus reacted to heat with a
typical avoiding response. This response disappeared when the
anterior end was cut off, for the cell body then swam indiscrimin-
ately towards the heated end of the slide and was killed. To a i %
solution of table salt the headless cell bodies gave the avoiding
response just as whole animals. He concluded that warmth percep-
tion is limited to the anterior end while chemical sensation is over
the entire body surface. This interpretation is open to question,
first, because salt apparently compels ciliary reversal with contin-
uous backward swimming by direct action on the coordinating
mechanism, and second, because as Dierks (1926b) pointed out,
the response of the isolated heads was not indicated. Anterior and
posterior half fragments would better have been compared since
both are capable of typical avoiding reactions, with use of a less
noxious chemical stimulant like carmine particles.

With further regard to sensory localization, Roesle (1902)
claimed that the mouth is the part of stentor most sensitive to
electric currents but behavior of stentors with mouthparts excised
was apparently not studied for comparison. Roesle also reported
that an induction current stimulates Stentor to contraction only
when the direction of the current was parallel to the axis of the
animal, contrasting with muscle physiology in which stimulation
is independent of the orientation. Yet Hausmann (1927) found
that polymorphus contracted and backed up in an electric current
but without correlation with its direction. Stentor was more
sensitive than smaller ciliates of other genera so that at least the
intensity of the response may be correlated with body size if the
animals compared normally swim at the same speed. Dierks Roesle,
neuroid (1926b), repeating the electrical stimulation studies of
found that Stentor was more sensitive to the current than any of
the other cihates tested, which he credited to the presence of fibers.

BEHAVIOR 25

6. The question of sensory cilia

Certain of the body cilia behave differently from most. According
to Kahl (1935) the posterior cilia are strongly thigmotactic,
coming to a stop when they touch something substantial. Along
the ciliary rows it has been found that groups of cilia are stiff
and pointed outward while the remainder of the ciliation is actively
beating; and these have been called tactile spines, setae, or Tast-
borsten (Fig. 3). They may disappear and reappear. Hence Stein,
who seems to have first noticed them, thought they could be with-
drawn into the body. Johnson, with more probability, said that

Fig. 3. Sessile body cilia as seen in S. roeseli, possibly sensory.
(After Kahl, 1935.)

they were only temporarily rigid cilia which could start again
beating and only seemed longer than the others because they were
stopped. He suggested a sensory function for the cilia in the rigid
state because they are found mostly toward the anterior end where
stimuli would presumably be most frequent. Since Johnson saw
them both in coeruleus, which makes no case, and in roeseli which
does, they are not uniquely correlated with case building. He
found them most evident in the frontal view ; and Kahl states that
there is always one group of " bristles ", five to twenty in number
according to the species, in each kinety directly under the
membranellar band.

26

THE BIOLOGY OF ST EN TOR

7. Cystment

Hibernation in a cyst may be regarded as a response of Stentor
to adverse conditions. Stein's is the only published account and
characteristically he provided beautiful illustration of coeruleus
and polymorphus cysts (1867, Tafeln V and VI). Although the
final test of excystment was lacking, he can hardly have been
mistaken because the coeruleus capsules retained the blue-green
striping and the polymorphus were of similar form, with chlorellae
and colorless stripes. In both, the feeding organelles were
dedifferentiated and the animal rested within a flask-shaped cyst
with a gelatinous plug (Fig. 4). Only once have I seen coeruleus
apparently beginning encystment within a membrane inside which

A

B

Fig. 4. Cysts of (A) coeruleus and (B) polymorphus. (After Stein,

1867.)

it rotated for a day but then died. In nigery however, cystment
seems to occur readily, with small, spherical, brown cysts, though
again, stentors have not yet been seen emerging therefrom
(unpublished).

Altogether, stentors exhibit a considerable range of behavior
in their orderly swimming movements, avoiding responses, food
selection, and attachment or detachment of holdfast "at will".

BEHAVIOR 27

To these frequent reactions may be added the "decision" to
encyst or excyst, as well as the special response to each other
which leads to the joining of pairs in a specific orientation for
sexual conjugation. Of special interest is the accommodation to
repeated adverse stimuli which Jennings demonstrated and which
may be regarded as the best evidence of a primitive type of
learning in unicellular organisms.

The ease with which stentors lend themselves to micrurgical
operations should provide opportunities for analyzing the struc-
tural basis of coordination. After the immediate contraction
response to cutting, stentors show no evidence of '' pain " or
lingering effects of injury as the fragments or cut animals swim
away in a remarkably normal manner. Later they mend the cut or
regenerate missing parts, but between the time of cutting and
repair or reconstruction there is, it would seem, a sufficient period
for observing effects on behavior of specific lesions and ablations.
Are there circumferential cuts at certain levels of the cell axis
which abolish coordination between adjacent rows of body cilia so
that spiral swimming is prevented, or may such cuts prevent attach-
ment or detachment of the holdfast? Are stentors from which the
oral pouch has been removed or those — after a method which
will be described later (p. 172) — developing without a pouch still
capable of food selection? If not, could they be made to ingest
unusual materials relevant to analysis of their metabolism? If
decapitation abolishes the avoiding response to light in coeruleuSy
can selective ablations at the anterior pole demonstrate the sensitive
area ; and do similar operations also prevent the positive response to
light in niger? Do stentors temporarily bereft of ingestive organelles
still show the avidity of normal animals evidenced by quicker
accommodation to mechanical stimuli when in the presence of
food after a period of deprivation, or is the successful formation of
food vacuoles the necessary prelude to this behavior?

These questions are but a sample of possible ways in which
behavior on the cell level might be- investigated in a form so
amenable to operation as Stentor. Might the neurone itself, then
be capable of considerably more " behavior " than mere
''excitation", realizing as part of the nervous system something
of the potentialities which have been evolved in a different manner
but from the same cellular origins, in ciHates?

CHAPTER IV

FINE STRUCTURE

Bearing in mind our general survey of the morphology of Stentor,
we can now probe the intimate construction of its parts as the
structural basis of its living functions and behaviour. Every detail
we learn only increases our wonder that stentors are capable of
such remarkable feats of regeneration and reconstruction after the
cutting needle comes crashing through these complex and highly
organized cell differentiations.

I. Feeding organelles

The differentiations at the anterior end of stentor are formed
in a few hours from an oral primordium in the most dramatic act
of cytoplasmic elaboration shown by this animal. A band of
membranelles develops on the side of the cell and carves out an
area of the ventral striping to the right which it carries forward to
the anterior end, as the posterior terminus of the band invaginates
to form the gullet and the adjacent part of the isolated striping
indents to produce the oral pouch. An important consideration
is whether all growth and elaboration of oral parts occurs only
through primordium development. If so, the size and number of
membranelles should not increase thereafter, the gullet would not
increase in resting diameter or length, and the number of stripes
in the frontal field should remain the same after oral differentiation
regardless of increase in cell volume through growth. Indications
that this may be the case are found in the fact that adoral bands
and gullets abbreviated by cutting do not grow out again in situ
but initiate regeneration of a whole new set of feeding organelles.
But in regard to decrease in these structures through dedifferentia-
tion and resorption the situation is entirely different, for the
membranellar band and gullet can shorten in situ (see p. 125).

28

FINE STRUCTURE 2g

(a) Frontal field

The granular or pigmented stripes of this area are narrow and
the ciHary rows correspondingly close together, not only because
they are ventral fine stripes shifted forward, but also because there
is usually an extensive formation of new stripes within the arc of
the developing membranellar band. Some of the resulting pigment
stripes in coeruleus may be so narrow as to consist of only a single
row of granules. The disposition of the frontal striping follows
roughly the curve of the membranellar band (Fig. i). Some of the
stripes end at the border, while those nearest the oral side continue
as the lining of the oral pouch and proceed downward and spiraling
into the depths of the gullet.

Like their lateral progenitors, clear stripes of the frontal field
bear rows of small cilia, and contractile myonemes which are said
to be finer than those of the lateral body wall and without
branchings (Dierks, 1926a). Stevens (1903) thought that the frontal
stripes can multiply in situ and not only alongside the developing
primordium. This now seems unlikely because stripe splitting
within the field is not observed. Likewise, when for some reason
the stripes fail to increase adjacent to the anlage, the frontal field
is then deficient and remains so until corrected by a later re-
organization or re-regeneration.

The frontal stripe area is bordered by a wider clear stripe
followed by a bordering pigment stripe and finally the membranel-
lar band itself (see Fig. i). According to Maier (1903), who studied
niger but says that coeruleus is the same, there is a marginal ciliary
row which neither Schuberg nor Johnson noticed on the clear
border stripe, but no myoneme underlies it. Schroder (1907)
seems to have been the first to mention the pigmented border
stripe in coeruleus though Schuberg (1890) showed it in his
drawings ; I have found it along the right side of the developing
oral primordium when presumably the clear border stripe is also
laid down.

(b) Oral pouch

In some species (e.g., introversus and roeseli) the frontal field
simply dips downwards like a ramp and forms a curved trough
leading to the opening of the gullet. In coeruleus there is a definite
in-pocketing such that the pouch curves back under the frontal

30 THE BIOLOGY OF STENTOR

field, forming what Johnson called the buccal fold. Oralward the
wall of the pouch lies against the body ectoplasm, forming a thin
wall which is apparently what Johnson called the " velum ".
The membranellar margin of this wall bends inward, producing
the groove which Schuberg emphasized, and is capable, by
independent contraction, of increasing this over-hang or nearly
closing off the cavity below. It was Schuberg who first pointed out
that the oral or buccal pouch is not a part of the gullet but only a
modification of the frontal field, which later experiments confirmed
(see p. 172).

(c) Membranellar band

On first inspection membranelles appear to be merely large cilia,
and so they seemed to Stein (1867) and Simroth (1876); but Sterki
(1878) first noted that they are flat plates. Each lamella is formed
by many cilia clinging together in a sheet, presumably by the inter-
locking, as in the barbules of a feather, of lateral spurs recently
demonstrated through electron microscopy by Randall and
Jackson (1958). The same occurs in the formation of cirri in hypo-
trichs (Roth, 1956). All students had agreed that each membranelle
in stentors was made up of 2 rows of cilia and I found the same
in silver-stained polymorphus. It is therefore surprising that
Randall and Jackson describe from clear photographs triple rows
for the same species, except in the mouth region. This discrepancy
will have to be resolved. It is possible that strains could differ in
this respect and if so the number of rows could be an important
genetic character. In the ciliate Oxytricha membranelles are com-
posed of 3 rows of cilia (E. E. Lund, 1935).

The row of membranelles is held together by some sort of band
or fibers at the level of the ciliary bases, and when shed in salt
solutions the membranelles do not fall apart but come off as a
continuous ribbon (Tartar, 1957a). Schuberg thought this union
was accomplished by thickened pellicle; Schroder by a special
meshwork or membrane. Dierks saw a fiber (of coordinating
function he supposed) connecting the membranelles at this level,
and such connectives are clear in the electronmicrographs of
Randall and Jackson but afford no indication of whether the
fibers are supportive, contractile, or coordinating. If isolated by
crushing, Moxon (1869) saw in the membranellar band "tremulous

FINE STRUCTURE 3I

waves passing along it after patches of cilia were detached ", but
no one has confirmed this. That the oral lip is capable of indepen-
dent contraction Hke a sphincter suggests some means of developing
tension in the transverse direction. Whether actively or passively,
the membranelles of the entire band do come closer together
when the frontal field is contracted (Randall and Jackson).

Each membranelle has an extension into the endoplasm.
Difficult to observe, the structure and function of these processes
have occasioned conflicting interpretations, but recent studies with
the electron microscope clarify the issues. Schuberg (1890, 1905)
described a triangular sheet or lamella from the base of each
membranelle, apparently bordered by two converging fibers and
narrowing inward to an apex which continued as an end fiber,
while the end fibers of all the membranelles were joined together
by a deep-lying basal fiber. Most workers agreed with this picture
(excepting Schroder, Neresheimer, and Dierks) and EM studies
are confirmatory. Neresheimer (1903) interpreted the lamellae as
overlapping plates which were not joined by a continuous fiber and
served for anchoring the membranelles. This view was expanded
by Dierks (1926a) who described the lamellse as anchoring
rectangles with a twist which accounted for the other appearances
including the basal fiber. In the light of the EM studies this view
will be discarded, and also it may be added that in shed
membranellar bands nothing like these plates is found, as would
be expected if they had sufficient strength to serve for anchoring.

Faure-Fremiet and Rouiller (1955) made an EM study of
polymorphtis, niger and coeruleus. They find that each membranelle
is composed of two short, parallel rows of ciliary basal granules or
kinetosomes which are connected to each other laterally and give
off ciliary rootlets into the interior. Each bundle of rootlets com-
bines with that from neighbouring cilia to form the triangle
Schuberg described, which is now seen to be of a fibrous nature
throughout. Randall and Jackson confirmed this picture and added
new details (Fig. 5). The fibrils in the triangular bundle are fitted
together in an orderly stacking. These fibrils do not appear to be
striated, as is the case with some metazoan ciliary rootlets (Fawcett
and Porter, 1954). Continuing inward, the fibrils form a long
bundle, corresponding to Schuberg's end fiber, and these are in
turn joined together at their ends by a composite basal fiber. The

32 THE BIOLOGY OF STENTOR

connection is a smooth one, fibrils from the end fiber bending and
running into the basal fiber. Photographic evidence of these points
is very convincing. The whole basal structure of the membranelle
extends for about 20/x into the endoplasm.

cdi

hineiosome

Ciliary roois

membranelle

memhranelLar hand

inke r memhranelUr
connecliye

jfooz bundle

Jbasal liber

Fig. 5. Structure of membranelles as revealed by electron micro-
scopy. (After Randall and Jackson, 1958.)

The function of this intracellular structure of the membranelles
has been variously interpreted as contractile, nutritive, supportive,
or coordinating, beginning with Brauer (1885) who thought the
basal fiber a muscle which contracted the whole frontal field.
Maier (1903) modified this view with a speculation that the basal
lamellae retract or pull the membranelles inward while the basal
fiber draws them together; but the membranelles do not seem to
retract, and were the basal fiber contractile it would only pull the

FINE^STRUCTURE 33

tips of the end fibers together without necessarily compacting the
membranelles.

Neresheimer (1903), Schroder (1907), and Dierks (1926a) all
maintained that the basal lamellae were to give solidity and support
as anchors for the powerfully beating membranelles. It is difficult
to see how much support could be given because the membranellar
band is easily sloughed in salt solutions and when it comes off
there is no sign of supporting structure below the membranelles.

Schuberg denied a muscular function for the basal fiber since
it is thrown into convolutions when the animal is contracted and
he thought that the basal lamellae could achieve little anchoring
in a fluid endoplasm. He suggested a nutritive function for the
parts he described, though granting there was still no proof of
this. The membranelles start and stop together and they beat in
an orderly fashion, one firing after the other in regular series to
produce a metachronal rhythm. It therefore seemed to Johnson
that the basal fiber with its connections to the membranelles would
be suited to a coordinating function. But even this reasonable
interpretation is not without its difficulties. Neresheimer found
that the usual nerve anaesthetics had no effect on ciliary action in
Stentor ; and also, if the membranellar band is severed deep into
the interior, metachronal rhythm continues on both sides of the
cut although the basal fiber must certainly have been sundered.

The wide, clear marginal stripe of the frontal field running
along the inner margin of the membranellar band should not be
overlooked as a possible site of fibers coordinating the membranellar
beat. Clear stripes elsewhere carry fibers connecting the cilia, and
in Stylonychia the membranelles of the oral region are apparently
connected by a lateral fiber (Chen, 1944).

Although the electronmicrographs give no indication thereof,
the membranellar band shows an intrinsic polarity. Bands or
sections of bands similarly oriented will readily join and mend
together without a break but not otherwise, and reversed mid-
sections of a band are reincorporated only after they, invariably,
rotate 180° in reorientation. This polarization appears during
primordium development. The primordium can be cut through
transversely in many places without effect, the severed parts merely
heahng together ; but if a sector is cut out and replaced in reversed
position it develops separately (see Fig. 41K).

34 THE BIOLOGY OF STENTOR

From the complexity of its structure we appreciate that the
formation of the membranellar band is indeed an astonishing
achievement of differentiation, involving multiplication of ciliary
basal bodies to 15,000 (judging from the data of Randall and
Jackson), the precise alignment of these bodies in rows and the
grouping of these rows by twos or threes, outgrowth of long cilia
from these granules, ingrowth of ciliary rootlets and their precise
association into triangular plates ending in a fiber, together with
the elaboration of the basal fiber connecting the end fibers, not to
mention the coiling and shifting of the entire structure to the
anterior end. All this occurs within about 4 hours. The perform-
ance is the more remarkable in view of the fact that when the
membranellar band is forming it can be slashed through many
times with a glass needle without producing any apparent abnor-
mality of construction (Tartar, 1957c).

(d) Gullet

The membranellar band continues in a sharply spiraled course
down into the gullet, as does the ectoplasmic striping which lines
the oral pouch and is itself continuous with the stripes of the frontal
field. Opening into the right-hand side of the oral pouch, the gullet
shows a double spiraling: as a pendant tube it takes about one
complete turn as it penetrates into the endoplasm, while the wall
of the gullet is itself under sharp torsion. Dierks claimed that the
opening into the gullet is capable of closure, though no one else
has observed this.

In everted gullets of coeruleus Andrews (1946) saw the membran-
ellar band extending in a spiral to the lower end of the gullet,
while decreasing to half its usual width (Fig. 6a). In polymorphus,
Randall and Jackson describe the membranelles in this region
as bi- instead of tri-lamellar. In pigmented species the appearance
is often that only one side of the gullet is colored because the band
of membranelles is itself unpigmented. However, Dierks main-
tained, apparently erroneously, that the membranellar band does
not continue into the gullet. In any case, the gullet has its own
specialized ciliation. Gelei (1925) found that the ciUa here stained
differently, and he likened them to the pharyngeal cilia of turbel-
larian worms which serve in swallowing. Dierks even denied that
the kinetics of the frontal field continue into the gullet, being

FINE STRUCTURE 35

interrupted where new types of cilia begin. The latter investigator
made fine distinctions regarding the gullet tube. Its entrance he
called the cytostome which leads into a short passage called the
pharynx separated by a ridge from the extensive remaining portion

orsLpomUx

stomal Rim.

narrovj'm^ inembrandUa.r baiui

frontal Jield

myoneme.

C^laslome

Fig. 6, Structure of the gullet of 5. coernleus.

A. Exposure of gullet everted by pressure. (After Andrews,

1946.).

B. Schematic course of myonemes in the gullet lining,
affording possiblity of peristalsis. (After Dierks, 1926a.)

or esophagus. Each of these parts was.described as having its own
special ciliation: the ''cytostomial" cilia look like the pharyngeals
but have the special function of selecting the food particles, and
the esophageal ciUa were said to be different, the distance between
cilia increasing as they are followed down into the gullet. Andrews
agrees with Dierks that in the profundity of the gullet the

36 THE BIOLOGY OF STENTOR

membranelles take the form of isolated cilia not joined in sheets.

Dierks granted, however, that the myonemes of the frontal
field are continuous into the gullet, proceeding uninterrupted and
in a sharply spiraled course to its terminus and also, as in the
frontal field itself, showing no branchings (Fig. 6b). In the gullet
the myonemes become much thicker and presumably stronger,
according to Gelei, and their disposition more or less transverse to
the length of the gullet could provide for the peristaltic movements
in swallowing which have been observed. Gelei also described a
fibrous net surrounding the gullet, which he thought might serve
both to prevent the gullet from tearing when stretched and to
coordinate cilia and myonemes in a swallowing action. Such an
appearance may have been due to the system of vacuoles and inter-
spersed fibers found near the gullet by Randall and Jackson.
These numerous vacuoles have double or triple membranes and
it was therefore suggested that they might be formed by invagin-
ations from the gullet wall, as in the pinoc3rtosis of Amoeba. The
adjacent fibers run rather deeply into the interior and may
represent the pharyngeal fibers of Schuberg.

Following Schuberg we shall call the inner terminus of the
gullet the cytostome. Again in everted gullets, Andrews (1946)
saw the cytostome as a thin, clear membrane without visible struc-
ture which prevents the escape or regurgitation of endoplasm.
Dangling inward from the periphery of this cytostome he found the
long fibers described by Schuberg but not seen as such by others,
and he thought that they formed part of a permanent canal which
guides and might even propel ingested food into the endoplasm.
By Andrews' account, ingestion may occur in one of three ways :
small particles may collect at the bottom of the gullet and push out
the stomal film until it breaks off as a membrane surrounding
them; the film may be momentarily broken as objects like small
rotifers pass directly into the endoplasm; or the film may be
missing as the cytostome opens wide to admit larger organisms
or clots of food. In the last two cases the ingested animals thrash
around freely in the endoplasm but eventually are encased in a
food vacuole and die. I once found a stentor that had ingested a
cotton fiber with one end still protruding forward out of the gullet
and the other end passing through the cell and emerging through
the surface near the posterior end.

FINE STRUCTURE 37

2. Holdfast

A history of our knowledge of stentor's anchoring organ is given
in Andrews' (1945) most complete account of this organelle, which
confirmed and extended the early observations of Gruber (1878).
Stein (1876) thought that stentors fastened in part by means of a
tiny suction cup. This idea was revived by Dierks (1926a) who
described the myonemes as not continuing all the way to the
posterior pole but leaving the ectoplasm near the tail end to turn
inward toward the center of the cell where they took another bend
as they joined together to make a bundle pointing forward. The
result was a cone of contractile elements open toward the terminal
pole. Assuming that the recurved ends of the myonemes are
independently contractile while their anterior extensions remain
wholly relaxed, and that the posterior end of the animal could
somehow produce a tightly adhering cup, he conceived that this
arrangement produces a suction which is the principal means of
attachment. This scheme is highly dubitable. In the first place,
the study was made on killed and contracted animals, unattached.
Second, stentors can attach to the surface film where suction
should not be eflFective. Finally, among the other assumptions
mentioned, this conception was based on the questionable pre-
supposition that amoeboid processes with sticky substances could
not account for the firmness of adhesion which is observed.
Schroder (1907), however, also described myonemes as recurved
at the posterior end, but he did not advance the suction idea.
Instead, in the cone before-mentioned, he defined a special
cytoplasm from which the attaching organ was presumed to be
elaborated. I have sometimes observed the " hem " or sharp bend
in the cell contour toward the posterior end which Dierks des-
cribed as indicating where the myonemes turn inward but other-
wise found no confirming indications of his description in living
material. Whether Schroder's and Dierks' recurved myonemes
are artifacts of fixation can only be decided by successful preserva-
tion of animals in the fully extended^ state.

Johnson (1893) figured the myonemes as running without
deviation to the posterior pole and hence his conception of attach-
ment was quite diff"erent. Body striping was described, however,
as not continuing all the way to the pole itself but stopping short
to leave a small terminal area which, because of its absence of

38 THE BIOLOGY OF STENTOR

Structure, he designated questionably as endoplasm. This would
correspond to the polar plasma later described by Schroder. Weisz,
too, (1948a) emphasized that the posterior end is clear and
structureless, but Dierks denied that there is any such '* naked
protoplasm ". Probably the pellucid polar endoplasm is responsible
for this illusion. Rather it would seem that granular stripes, ciliary
rows and myonemes cannot continue all the way to a fine point
without an improbable anastamosis of unlike elements and that
therefore at the posterior end the construction would not be strictly
closed but allow an opening for extrusions.

The manner of attachment may vary with the nature of the
substratum. To clean glass, according to Johnson, terminal proto-
plasm adheres as a smooth disc ; against slime, pointed pseudopods
are given out from this disc ; and in attachment to the surface film
pseudopods become broad and branching as Andrews later des-
cribed more fully. Johnson also observed that stentors never attach
until stretched out and that the terminal cilia seem to feel about
for a place of attachment.

Enlarging much on these observations, Andrews (1945)
described that in the outstretched stentor seeking attachment the
posterior cilia on the stalk come to a stop projecting outward while
the terminal ciUa, somewhat like the scopula of a vorticeUid,
remain active, possibly seeking a favorable spot or effecting a
preliminary attachment. By focusing downward on the foot of
stentors attaching to glass wool or a cover slip he was able to give
the most complete account of the holdfast, one which also has
interesting implications concerning the relationship between cilia
and pseudopodia. An amoeboid disc of naked cytoplasm is first
extruded from the posterior end to adhere by its stickiness, and
some of the posterior cilia are apparently transformed into viscid,
rigid, acicular pseudopods which he called " radiants". (If this
does indeed occur it carries the surprising imphcation, contradic-
tory to the hypothesis of the French school (see Lwoff, 1950) that
formed cilia can transform into something else without new
growth from a kinetosome specifically determined to produce such
a structure.) Meanwhile some of the terminal cilia remain active
("undulants") but these gradually disappear with continued
attachment. Then the adjacent ectoplasm with its pigmented
stripes and ciliary rows is drawn out into extensive projections like

FINE STRUCTURE 39

pseudopodia which he called ''radicules" (Fig. 7A). This would
account for the observation that coeruleus even at low magnification
shows a green, stellate foot. In side view after long attachment
Andrews observed that the stentor is chiefly anchored by the

ackulsir psevudopod : j-a^a.nt

Fig. 7. Holdfast of S. coeruleus.

A. Underside of attaching holdfast, showing ectoplasmic
projections ("radicules") and active, undulating cilia said to
convert into thicker, acicular, attaching pseudopods.
B. Side view. (After Andrews, 1945.)

acicular and lobose pseudopodial processes, Hke a balloon anchored
by ropes (Fig. 7B), the openings between which would preclude
any suction. If forcibly detached, the holdfast remains somewhat
intact for a while and is so sticky that if touched with a needle
adherence is firm and immediate. But later, or when the animal
detaches itself at will, the holdfast is withdrawn. Then, according
to Andrews, its structural parts resume their former forms and
functions, which would imply that the striping of the pseudopods
again takes the form of the posterior cell wall and the aciculars
transform back into cilia. However this may be, it follows that
structureless cytoplasm is not the sine qua non of attachment.

40 THE BIOLOGY OF STENTOR

Johnson surmised that the pedal opening could be used for the
extrusion of pigment granules but in this he could have been mis-
led by appearances, as Weisz (1949a) noted, since the stickiness of
the detached holdfast is likely to pick up debris, including cast-off
pigment. However, Weisz stated that the discontinuity of the
ectoplasm at the foot can be used for the ejection of such waste
as the undigested pellicles of paramecia.

3. Cytopyge

Undigested material is usually collected and extruded at a single
site on the anterior left side, just below and to the left of the
opening of the contractile vacuole. As Johnson first described the
process, the spent food vacuoles, if small, accumulate by fusion
in this place. The pellicle then ruptures within one of the broad
granular bands and the waste is slowly defecated as the slit opens,
often so widely as to distort the adjacent striping; thereupon
closing without leaving a trace.

Whether the cytopyge has a persisting structure is still in ques-
tion. Moxon (1869) could find no pore but said the spot was
marked by an irregularity in one or two of the granular stripes and
Johnson found no fixed organelle, but Andrews (1946) made out
a long slit with definite lips. Nevertheless, it is certain that defeca-
tion can occur in other places, as Johnson first observed. I twice
observed coeruleus ejecting solid material in the normal manner
simultaneously at two points far distant from the normal site
(Fig. 8a). Defecation openings break or open through granular
stripes, clear stripes carrying too much structure to permit exit.

4. Contractile vacuole

Stentors have but one contractile vacuole always located at the
left anterior side of the cell. The structure of this excretory system
seems to be less well-developed than in Paramecium in spite of the
fact that Stentors are much larger, e.g., there are no star shaped
canals and Haye (1930) found few lipoid granules associated with
this system. Walls of the contractile vacuole were not blackened in
osmic acid (Park, 1929). Schwalbe (1866) is said to have been the
first to see the excretory pore, in polymorphus. With its pigmenta-
tion, coeruleus shows this part more clearly and Moxon described
the presence of two or three unpigmented spots in the colored

FINE STRUCTURE 4I

bands over the vacuole, one of which opens widely to void its
content. Independently, Maupas (1883) discovered these spots
which are evidence of persisting pores. Andrews (1948b) noted
that rarely two pores may open at the same time as well as that
the arrangement of the spots may vary in different individuals or
in the same specimen at different times. The presence of these
pores is easily confirmed (Fig. 8b).

CM

B

Fig. 8. Excretion in S. coeruleus.

A. Specimen showing location of contractile vacuole and
normal site of cytopyge (i) but also excreting from a second

opening (2).

B. Multiple pores in granular stripes, one of which is excret-
ing contents of the underlying contractile vacuole. (After

Andrews, 1948b.)

All students agree that the vacuole increases in size by the
confluence of smaller vacuoles or at least by the draining of their
fluid contents into the contractile vacuole. Although not as evident
as in paramecia, a system of collecting channels has been des-
cribed. In coeruleus, Maupas found variable canals formed by the
alignment of adventitious vacuoles which presumably ran together
and pushed their fluid toward the contractile vacuole. Johnson
confirmed this picture and maintained that in divisions one of a
pedally directed line of vacuoles becomes the contractile vacuole
of the posterior daughter cell, even before separation beginning to
contract regularly but not in synchrony with the old one and
acquiring excretory pores in the pigment stripes above. Johnson
also described a definite longitudinal canal in roeseli, recalling that

42 THE BIOLOGY OF STENTOR

in Spirostomuniy in which also the new vacuole is produced at
division simply by an enlargement of this canal. Anterior to the
new vacuole a segment of the longitudinal canal separates both
from this vacuole and from the posterior end of the anterior
daughter, producing the circumoral " ring canal " discovered by
Lachmann (Claparede and Lachmann, 1 858-1861) ; but this
severed portion of the canal soon atrophies and hence is seen only
in the young opisthe. Haye stated that polymorphus also has a
longitudinal canal but he did not describe or illustrate it. In
coenileus Andrews saw many channels emanating from the
contractile vacuole. One proceeded forward and led to a horizontal
ring canal underlying the frontal field ; others proceeded backward
toward the foot and from such might come the new contractile
vacuole during fission.

5. Cortical structure

The well-differentiated cortex or ectoplasm of Stentor is highly
extensible, considerably elastic, sharply contractile, capable in
part of being shed and regenerated, as well as bearing cilia with
the means of their coordination. These properties and functions
are to be related to the types of microscopic and submicroscopic
structure present. Even today our knowledge of this correlation
is still highly problematical; nor is it certain that all structural
details have been revealed, though electron microscopy has made
possible astonishing advances in this study.

From the standpoint of morphogenesis, the cell cortex with its
enduring pattern is of greatest importance. For it is from this layer
that other cytoplasmic organelles are elaborated, as when a mid-
section fragment regenerates a new head and foot. As will become
evident later, both holdfast and oral primordium formation are
intimately related to the polarity and pattern of the cortex, and
we may hope that the causal basis of this relationship will in time
be exposed.

(a) The cell surface

Proceeding from the exterior, there is a pellicle, long ago
demonstrated by Johnson who saw it lift off the cell on treatment
with weak osmic acid while remaining firmly attached to the ciliary
rows. Electronmicrographs of Randall and Jackson showed the

FINE STRUCTURE 43

pellicle to be a double or perhaps even a triple membrane, joined
to the body cilia because continuous with their outer walls. When
the pellicle is elevated the cytoplasm does not flow out into the
spaces provided and hence there is another film, the plasma
membrane, which was also shown to be double. One may regard
the pellicle as being a somewhat dispensable secretion of the cell
because salt treatments often produce the shedding of a layer which
is presumably the outermost, but the stentor remains intact,
appears not to be significantly affected, and probably re-secretes
the layer. These remarks are demonstrably true for Blepharisma
(Nadler, 1929).

In cross-sections of contracted stentors the surface is thrown
into a series of ridges parallel to the longitudinal axis of the cell.
Each ridge represents one of the granular or pigmented stripes
(" Rippenstreifen " of Butschli) while the alternating valleys are
the clear stripes (Zwischenstreifen). The bases of the rows of body
ciUa are implanted along the (animal's) left side of each valley or
clear stripe.

It was early noticed that on contraction the stripes of pigmented
or non-pigmented granules are thrown into folds, transverse
ridges, or crenulations while the clear stripes are not. Schroder
remarked that this pleating may cause the granules to be aligned
in rows, which indicates that these particles have some freedom
of displacement. Hence the original deception that these bands
were striated muscles. The appearance described impHes that the
pellicle has a limited elasticity, at least over the granular stripes,
and that it is more elastic or simply pinned down in the region of
the clear stripes.

Presumably there is a break in the pellicle permitting extrusion
at the holdfast and no pellicle over the cytostome.

(b) Granular stripes: nature of the pigment and granules
As already described in Chapter II, the ectoplasm is chiefly
characterized by alternating clear and -granular longitudinal stripes
or bands. In colored stentors the granules are pigmented, giving
the appearance of pigmented stripes. These stripes seem to be
without specialized structures other than the granules located in
them ; but the clear bands mark the site of complex differentiations,
including not only the ciliary rows but also a band which in living

44 THE BIOLOGY OF STENTOR

animals is often seen to be thrown into transverse waves or convo-
lutions. Therefore, the granular stripes appear to be merely fill-ins
where surface granules come to occupy spaces left between the
structured clear bands and membranelles. Accordingly, it is the
granular and not the clear stripes which should have been called
mere " between-stripes " (" Zwischenstreifen ") even though they
are more obvious to the eye. Moxon, for instance, found that in
crushed coeruleus the pigment stripes disperse while the clear
bands persist as refractile structures, and Schroder remarked the
same. The great change in width of the former speaks for the same
conclusion, contrasting with the uniformity of the clear stripes.
When the area of ectoplasm in coeruleus is greatly reduced, the
patch remaining stretches to cover the whole and this stretching
occurs mainly in the pigmented stripes which become very broad
(Fig. 9a). Furthermore, where stripe increase occurs it can be seen
that the pigmented bands adapt in width and contour to the
exigencies of the situation (Fig. 9B), which would not be the case
if they had to maintain a uniform and stable structure.

These stripes are characterized by uniform, spheroid inclusions
about ijLt in diameter. They are not fixed in place but capable of a
certain freedom of movement (Andrews, 1946) which Weisz
(1949a) called Brownian motion. In all colored species the
pigmentation is probably confined to these granules (though
S. Felici was differently described), for there is not a second set
of uncolored bodies. That cortical granules seem to be present in
uncolored stentors indicates that they serve some purpose besides
pigmentation. When pigmented, some granules can also be identi-
fied in the interior of the cell ; and Weisz indicated this to be the
main site of a putative metabolic function, the granules being
stored, as it were, in the ectoplasm and loosed into the interior to
be utilized during starvation and regeneration. Within the interior
of coeruleus, Andrews (1955) reported that the pigment granules
move between the endoplasmic vesicles as if gliding along films,
by a movement not yet explained.

Granules can also be cast off to the exterior by the action of
mild irritants (see p. 250). This effect resembles the discharge of
trichocysts, which are, however, always spindle-form. Hence
the granules have been called protrichocysts by Prowazek (1904),
Kahl (1935), and Faure-Fremiet et al. (1956).

FINE STRUCTURE 45

Extensive studies on pigment granules in coeruleus were made
by Weisz (1949a, 1950a). He found them to have a basophiUc core
of protein pigment, surrounded by a phosphoHpid shell, and to
give a negative test for RNA but positive for cytochrome oxidase.

A B

Fig. 9. Granular stripes adapting to space available.

A. Specimen from which most of ectoplasm was removed,
granular stripes of the patch stretching to cover the endoplasm.

The animal then regenerated.

B . Photograph showing granular striping of nonuniform width

and contour according to the space provided.

The bright red appearance of these green particles in reflected light
he attributed to phase interference by the outer shell. Andrews
(1946) regarded the pigment granules of coeruleus as not mito-
chondrial; but Weisz concluded that they, as well as those of
Blepharisma, are mitochondria, basing this on their enzymatic
content, lipo-protein composition, apparent involvement in
metabolism, and especially their selectivity for Janus green B stain
(with no other bodies so staining). Admittedly it was difficult to
make distinctions in staining a green body green.

A better test is the demonstration of villiform interior structure
typical of the protozoan mitochondrion. The electronmicrographs
of Faure-Fremiet et al. (1956) and Randall and Jackson (1958)
clearly reveal mitochondria in stentors, but these bodies appear too

46 THE BIOLOGY OF STENTOR

large and much too few in number to correspond to the ecto-
plasmic granules. It is possible that the latter, if they do in fact
contain an oxidase, may be a new type of oxidative center different
from mitochondria, located next to the ciliary rows to assist in
their energy metabolism. Such alternating rows of ''mitochondria"
and cilia have been described in other cihates by Horning (1927)
and Turner (1940). Clearly, the function of these granules calls for
further investigation. However these questions may be resolved,
Andrews was probably right in saying that the degree of pigmenta-
tion is a delicate indicator of the physiological state in stentors
(see p. 274). Healthy coeruleus capable of long survival on slides are
invariably well colored.

Nothing is known concerning the origin of the bodies in the
granular stripes. In the blue Folliculina ampulla, closely related
to stentors, Faure-Fremiet (1932) found many blue granules very
close to the macronucleus yet he did not suggest that they were of
nuclear origin. Stentor coeruleus also frequently shows pigmented
granules surrounding the macronuclear nodes. Perhaps it is
relevant to mention that when I grafted a nearly colorless coeruleus
to another which was deeply pigmented the fusion mass became
well pigmented throughout in the course of 4 hours, far more
rapidly than in the usual regeneration of pigment in faded stentors
left to themselves. A closer following of such cases as well as of the
regeneration of pigment in animals which have been artificially
depigmented through chemical treatments, or of similar studies on
colorless stentors in which the granules have been artificially
sloughed, would seem to offer considerable possibilities for
obtaining clues regarding their origin.

The chemical nature of the pigments themselves is of interest.
Of these there seem to be three, as Johnson remarked : the blue-
green which gives the name to coeruleus and is probably also found
in the similarly colored introversus, multiformis and amethystinus,
if not in the related blue Folliculinids ; the brown pigment in niger
and possibly also in Johnson's nigricans variety of S. igneus; and a
purplish-red in igneus which may be the same as the zoopurpurin
of Blepharisma. Only the first two have so far been studied.

The pigment of coeruleus was given the name "stentorin" by
Lankester (1873) in a pioneer work in which he remarked the
extraordinary stability of this substance, not dissolved by fat

FINE STRUCTURE 47

solvents, acids, or alkalis. But Prowazek (1904) found that
sulphuric acid turned it red, potassium hydroxide caused it to
become grass-green and osmic salts changed it to black. Weisz
(1950a) could bleach stentorin with chlorine gas or potassium
permanganate followed by oxalic acid. Stentors of this species
appear red by reflected light and blue-green by transmitted (see
frontispiece), which is also the appearance of blue Folliculinids
(Andrews, 1923). Correspondingly, Lankester demonstrated two
strong absorption bands, one in the red and one in the green.

Prowazek grew coeruleus at higher than normal temperatures
and one of the effects he reported was that the animals often
became more reddish and showed fluorescence, warming in
general producing deeper hues of color. He also found that most
animals which feed on coeruleus do not digest the pigment, though
the color may be altered somewhat in passage through the alimen-
tary canal, as specifically confirmed by Gelei (1925) for the worm
Stenostomum. Only certain species of a worm(?), Nuclearian,
which grew in some of Prowazek's stentor cultures could assimilate
their pigment and become colored throughout. Cannibalizing
stentors do not assimilate the pigment of their own species but
concentrate and eject it as a dark green excretion vacuole, according
to Gelei (1925) and Andrew^s (1955), and this has also been my
impression. So also in the resorption of oral parts in the trans-
formation of Folliculinids the blue pigment granules are not
metaboHzed (Faure-Fremiet, 1932; Andrews, 1949).

In niger, Maier (1903) noted that the yellow pigment was of the
granules and could be dissolved by chloroform to give a red
solution. This unique pigment was later studied by Barbier,
Faure-Fremiet, and Lederer (1956), who found it to be soluble in
alcohol and of two components. A minor component was brown in
color and eluted by ether. The major portion, eluted by ether with
2% ethanol, was a substance of red-violet color which they called
'' stentorol ". The latter could be dried to a dark powder, showing
in ultraviolet a red fluorescence which was changed to yellow or
blue after various treatments. Absorption spectra were obtained
using diflferent solvents, leading to the identification of the pigment
as a polycyclic hydroxyquinone. They were impressed by the
resemblance to hypericum, a photodynamic substance originally
discovered in plants of that name.

48 THE BIOLOGY OF STENTOR

The nature of the pigment in the cortical granules of coeruleus
has been reinvestigated and much enlarged upon by K. M. Moller
(i960). Instead of Lankester's (1873) spectrometric absorption
maxima of 662 m/x and 562 m/x he found with living or extracted
stentors, only, a strong band at 618, a weaker one at 568, and a
third and very weak band at 527 m/x. This difference can be
explained on the basis that Lankester used stentors concentrated
in the gut of an aquatic worm, which may have ingested photo-
synthetic organisms as well or even S. polymorphus with its algal
symbionts. Stentorin itself does not resemble chlorophyll; and
tests by Moller and C. Chapman-Andresen demonstrated that it
has no photosynthetic action : coeruleus grown in the light do not
incorporate C^" bicarbonate solution.

The predominant blue-green pigment is indeed resistant to
solubilization but is dissolved in acetone-water and completely
extracted by ethylenediamine. Solutions, hke the living stentors,
are dichromatic and appear green by transmitted and red by
reflected light, but this is no proof of fluorescence. However,
Moller discovered that some races of coeruleus have an additional,
ethanol-extractable pigment in the cortical granules which renders
them red fluorescent in ultraviolet light of wave lengths from 366
to 590 m/x. Because the fluorescence appears only after these
stentors are dead or dying — as when dried on filter paper or
killed with boiling water — he inferred that the pigment is
probably bound to some protein (or carbohydrate) carrier which
uniquely quenches the fluorescence in living animals. The produc-
tion of this fluorescence is a nuclear-dependent character (see
p. 322). Moller and A. H. Whiteley (unpublished) found the
alcohol-extracted pigment or aqueous homogenates only of
fluorescent stentors to be photolethal (killing action of pigment
plus strong light) to Paramecium caudatum, Colpidium, and to the
stentors themselves but not to non-fluorescent coeruleus. Yet
living fluorescent stentors seemed to affect non-fluorescent
animals, but not the reverse, in the same medium (separated by
a screen) causing the latter to become colorless, smaller in size,
and even fluorescent.

Both the fluorescent and the major pigment not extractable by
ethanol are in their spectrometric and chemical properties different
from yet quite similar to hypericin, a photolethal substance

FINE STRUCTURE 49

previously found only in certain plants, which is phototoxic to
herbivores. These tests by Moller showed that the pigments
which may comprise stentorin apparently belong to the mesonaph-
thodianthrone group of compounds also including the photo-
dynamic pigments hypericin and phagopyrin. The function of
pigments in coeruleus is still unknown. Like the apparently related
chromatic substances of niger they may render these stentors
sensitive to light. Alternatively or in addition, if the pigment in
igneus is the same as that in Blepharisma which Giese (1949) found
to be toxic to certain other protozoa, and if that of at least certain
if not all races of coeruleus and niger be phototoxic to some
predators as eaten, then stentor pigments might have some
protective value for their bearers.

(c) Clear stripes and their fiber systems

In the living animal the highly differentiated clear stripes show
a row of cilia on their left margin and in the center a wide, sub-
pellicular band best revealed by polarized light or phase micro-
scopy. This band doubtless represents the original " Muskelfaser "
described by Lieberkuhn in 1857. Later students of this minute
structure (notably Schuberg, 1890; Johnson, 1893; Nerescheimer,
1903; Schroder, 1907; Dierks, 1926a; and Gelei, 1926) published
varying accounts of its precise nature which are now rendered
obsolete by recent studies with electron microscopy. Earlier
accounts agree, however, that these bands run the length of the
animal, branch and rebranch in correspondence with the clear
stripes, are tapered toward the anterior end but much thickened
posteriorly, in cross-section appearing pendent from the pellicle
adjacent to the ciliary rows as shown in Fig. loc, being contractile
in function and hence deserving the name myoneme. The
possibility of fiber connectives between the basal granules of the
cilia, presumably required for their coordinated movement and
universally found in ciliates through subsequent study of silver-
line and infraciliature systems, was completely neglected.

After the strained efforts with light microscopy, the EM studies
come as a revelation, though partially anticipated by Gelei (1926).
To date, we have the reports of Faure-Fremiet and Rouiller (1955),
Randall (1956), Faure-Fremiet, Rouiller and Gauchery (1956) and
Randall and Jackson (1958). Fig. lOA attempts to combine in one

50 THE BIOLOGY OF STENTOR

diagram the accounts of these two groups insofar as they concern
fibers lying under the clear stripes.

It is now seen that the bands immediately underlying the clear
stripes are of lamellae, stacked edgewise and attached to each other
as well as to the pellicle by their outer margins, their inner edges
lying free. Each lamella is composed of a stack or layer of very
fine fibrils, regularly spaced, adjacent lamellae being connected by
even finer processes.

According to the French workers the main body of this
lamellation, which we shall refer to empirically as the ribbon
bundle, constitutes the "ectomyoneme", a contractile organ of
unique structure. Number of lamellae varies with the level of the
body. The ribbon nearest the ciliary row, somewhat different in
shape and often separated from the other lamellae, was identified
as the kinetodesma which they supposed to connect the kine-
tosomes or basal bodies of the cilia, and the whole was referred to
as the myocihary complex. This connection was demonstrated
by the British workers whose photographs indicate that all
lamellae in the pile achieve connection with cilia. They therefore
called the ribbon bundle the '' km band ", suggesting that all parts
are possibly involved both in ciliary coordination and in contrac-
tion. Fibers leave the bundle and bifurcate as they attach to
opposite sides of a kinetosome, corresponding to the peduncles
observed by Villeneuve-Brachon (1940) in light microscopy. It
was further suggested that a given fiber may terminate forward on
one kinetosome and posteriorly on another, presumably facilitating
coordination of ciliary beating.

Either the restricted kinetodesma of Faure-Fremiet et ah or
the entire km band of Randall and Jackson follows the rule of
desmodexy (Chatton and Lwoff, 1935b) in that the fibrous con-
nectives between the kinetosomes lie to their right ; but otherwise
the system is entirely different from the infraciliature of other
ciHates, with the exception of Spirostomum (Randall, 1956). In
other forms the interciliary fiber or kinetodesma appears simple
and single, and transverse connectives between the kinetodesmata
are often found. According to the pioneer work of Worley (1933,
1934) on coordination of body cilia in Stentor and two other
ciliates, stentors do have these connectives and a metachronal
wave down a line of cilia can escape around a small surface

FINE STRUCTURE

51

Older nuclear membrane
inner nuclear membrane,
granular nuclear malrix

Connectives helween

M jbaiids

macronuclear
node
/cinetodestna

Fig. 10. Fine structure in Stentor.

A. Stereo-diagram showing structure revealed by electron-
microscopy. (After Faure-Fremiet et al., 1956 and Randall,

Jackson, 195B.)

B. Drawing of stentor beginning to re-extend, showing
convolutions in the fibers of the clear stripes. (After Johnson,

1893.)

C. Section through ectoplasm showing parts distinguished by

early microscopists. (After Schroder, 1907.)

52 THE BIOLOGY OF STENTOR

incision because of them. Klein (1932) may have demonstrated
such transverse fibers in his silver staining of a cortical network
in S. tgneus; but the EM studies of other species have revealed
no specialized connections between adjacent rows. Another impor-
tant difference is that the system in Stentor is refractive to silver-
staining and neither the wet (Villeneuve-Brachon, 1940) nor the
dry (Weisz, 1949a) method gives the beautiful network demon-
strable by this means in most other ciliates.

If the ribbon bundles we are now discussing were indeed what
earlier workers described as the myonemes, as appears from
correspondence in location, certain of their remarks may still be
pertinent. Popoff (1909) stated that the bands were not only more
numerous but correspondingly wider in larger stentors, 57 of
which were studied in this connection. Anteriorly, Schroder (1907)
found fine extensions of their much-tapered ends leaving the
cortex and passing inward and forward to attach to the outer
margin of the membranellar band. Dierks (1926a) seems to have
seen something like this too. Johnson (1893) noted that the bands
are straight during contraction but much convoluted at the
moment of beginning extension before the lengthening of the cell
has again stretched them straight (Fig. iob). This observation
was confirmed by Faure-Fremiet et al. as the behavior of the
ribbon band. Gelei (1926) reported that the bands in the clear
stripes of the frontal field are not tapering but of uniform thick-
ness. In the expanded field he found the bands to be still slightly
sinuous and in the contracted field they were strongly coiled. In
this area the bands therefore did not seem to become straight
when contraction occurs, yet he still regarded their function else-
where as contractile.

A nice point was made by Gelei when he remarked that if the
fibers responsible for sharp contraction were fastened only at their
anterior and posterior ends they would, on developing a tension,
pull to the center of the cell and not form an arc following the
contour of the surface as is in fact observed. Contraction would
then draw the cell into the shape of a much-flattened sphere.
(Incidentally this very shape, with corresponding retraction of
the frontal disc, actually occurs in introversus, in which the
disposition of the contractile fibers may therefore be quite diflFerent
from that in all other known species of Stentor.) Therefore Gelei

FINE STRUCTURE 53

postulated that the contractile bands were attached at successive
points to the " epimuscular band ", by which I think he meant the
pelHcular clear stripe. The ectomyonemes or km bands fulfill this
requirement in their connection with the pelHcle. And this con-
nection not only allows the bands to draw the whole cell into a
compact sphere rather than merely pulling the head and foot
together, but also makes possible the continued contraction of
isolated fragments because the bands need no end anchors. It
might even make possible the independent contraction of different
sections of the band, thus accounting for Dierks observation that
in simple contraction the anterior part of the band shortens while
the posterior part remains unthickened and thrown into curves,
itself straightening and thickening if super-contraction follows.
If indeed contractile, the ribbon bundles as muscles should have
their antagonists (Ishikawa, 1912), which would be whatever is
responsible for drawing out the cell. Elasticity of the pellicle may
be one factor here and the accessory bands shortly to be mentioned
another, but this matter is quite uncertain.

Lying interior to the ribbon bundles, Faure-Fremiet et al. made
out a layer of trabecular cytoplasm which in the study of Randall
and Jackson seemed to be another set of bands under the clear
stripes, tapered forward, wider posteriorly, and having transverse
connectives (Fig. iga). In their composition, these bands and
their connectives showed only short fibrils more or less randomly
arranged but tending to align with the axis and not orderly
stacking of long fibers. This is the type of structure Faure-Fremiet
and his associates find in the '^endomyonemes" of stalked ciliates
like Vorticella. Therefore these bands may be contractile. Accor-
dingly, Randall and Jackson referred to them as "M bands".
The transverse connections may be what Prowazek (19 13) observed
in vivo: delicate transverse connections between the substance of
the clear stripes in the expanded ectoplasm of stentors ''exploding"
or deliquescing on the surface film. Randall and Jackson found
that these transverse connections between the M bands w^ere
prominent posteriorly but fewer at the forward end of the animal.
Their demonstration that the matrix of the bands is continuous
with that of the connectives would seem to imply that action of the
latter could not be independent, say, in causing extension of the
animal.

54 THE BIOLOGY OF STENTOR

Extension of stentors to over six times their length when
maximally contracted calls for an adequate explanation but is still
a mystery. Randall and Jackson's report states that the volume is
quadrupled on extension but I think they must have meant the
surface area.

Walls of the M bands as described by the British v^orkers are
surprisingly indefinite, becoming confluent v^ith endoplasmic
vesicles so that the bands are very intimately related to the endo-
plasm, and this could account for the trabecular appearance
observed by the French investigators. These bands seemed to
have no connection with the pellicle, which of course poses the
problem of how they could produce the movement of anything
but themselves. One also wonders why stentors should have two
parallel contractile bands when one would seem to be sufficient.
These and other problems of structure and function we hope will
be resolved by further studies in this actively developing field.
Apropos of this, Causin's (1931) surprising statement may be
repeated : that although a cut into the side of a stentor is followed
by prompt healing and does not initiate the formation of a regenera-
tion primordium, there results nevertheless a cryptic resorption
and replacement of the myonemes. This observation should
certainly be checked.

In addition to understanding the static structure and short-time
activities of the cortex, we need to learn how its elements grow and
dediflFerentiate, develop and increase in number, as well as how
they manage surprising performances in mending and realignment
after cutting and other gross disturbances. These capabilities seem
contradictory to the fineness and complexity of the structures
present and tax the imagination to conceive how they are possible.

(d) Fiber systems of doubtful status

Still other types of fibers have been reported in the clear stripes.
They were located adjacent to the myonemes and described as
unvarying in thickness and convoluted in the contracted animal,
therefore presumably nervous in function and not contractile.
We have to call them doubtful because these reports did not
present at the same time a clear description or, indeed, indicate
any awareness of possible kinetodesmata. Neresheimer (1903)
seems to have had a bias for completing the roster of '* tissues"

FINE STRUCTURE 55

in Stentor by identifying nerve structures. Following him closely
and yet insisting on the uniqueness of what he had found, Dierks
(1926a) also sought a nervous system because he thought that the
coordinated behavior of Stentor implied its existence.

Neresheimer called his fibers "neurophanes" to contrast them
with *' myophanes ", the term used by Haeckel for the myonemes.
In retrospect Neresheimer seems to have stained and been
examining the ribbon bundles, which may indeed have a nervous
function if they serve to coordinate the body cilia. But in whole
mounts he could follow these bands only from the posterior end
to the middle of the cell. Along this course the fibers branched and
some of them ended in minute bulbs or boutons which he regarded
as sensory but which Dierks and Gelei (1926) thought to be mere
optical artifacts. Most of the study was on pieces of ectoplasm
loosed from the animal by treatment with methylene blue, and
Schroder (1907) criticized the results as artifacts from injury and
distortion. In reply, Neresheimer (1907) admitted that he could
not find his neurophanes in all preparations but insisted that they
were evident in some. Apparently he saw something of the ribbon
bundles but could not divine their intimate structure and actual
extent. To demonstrate that his fibers were nervous in function
he treated stentors with drugs which act as nervous excitants and
depressants in metazoa and found they had a similar effect on
stentors but not on other ciliates in which "neurophanes" are
lacking.

The fibers described by Dierks were called "neuroids". He
pictured them as running close to but above the myonemes
(ribbon bundles) and present throughout their entire extent,
either ending in these bands or sending side branches to them.
The " neuroids " may very well have been kinetodesmata or strips
torn loose from the ribbon bundles, as Villeneuve-Brachon sug-
gested. In any event, nothing like them has so far been found with
the superior resolution of the electron microscope.

Although Dierks (1926a) himself questioned why stentors
should need ''nerves" when the myonemes are so intimately in
contact with other parts of the cell, he nevertheless considered the
"neuroids" nervous in function. This assumption was based on
the response of stentors to the potassium ion which causes them to
relax in the extended state as if the "neuroids" were anaesthetized;

56 THE BIOLOGY OF STENTOR

but on fixing the animals always contracted, as if the myonemes
were then being stimulated directly by the fixing agent.

Simultaneously, Gelei (1926) found the ribbon bundle exterior
to the *'endomyonemes" and described it quite accurately within
the limits of light microscopy. Identical in location to the so-called
neurophane or neuroid, he regarded this fibrous band as giving
support and attachment to the myonemes, therefore ''skeletal"
in function.

(e) The cilia

Cilia comprising the oral membranelles are evidently not only
longer but also of larger diameter than the body ciha, according
to Randall and Jackson. They found that the body cilia oi poly-
morphus measured 20 /x in length, while in light microscopy they
appear to be 10 /x (Andrews, 1945, found them to be 13 /x in coeruleus).
This discrepancy may be due in large part to the fact that the cilia
have very fine tips, not easily visible. Thus, in a pioneer work of
Schuberg (1905) on coeruleus cilia stained by the Golgi method it
was shown that the proximal two-thirds of the cilium stains darker
and is of uniform diameter, the distal third being much narrower
and pointed at the end. Because the freed ciha were curved, he
foretold the view now generally held, that the contraction of the
cilium is intrinsic. He also noted that fixation seems to preserve
the cilia in phases of their rhythmic beating, thus anticipating the
interesting work of Parducz (1953). This general picture was con-
firmed with the electron microscope by Randall and Jackson whose
figures also show that the fine tip is prolonged into the length of
the wider portion as its axis.

Randall and Jackson unmistakably show that at the posterior end
of the cell the body cilia are paired and no longer form a single
row, raising the question of how in division the proter, acquiring
a new posterior end out of the middle parts of the cell, could
develop a double row. Possibly there is new growth there, as
Johnson first suggested.

Electron microscopy reveals in stentors the universal fine struc-
ture of the cilium. The outer layer is continuous with the pellicle
and the axis shows the typical 9 + 2 fibers (Faure-Fremiet and
Rouiller, 1955; Randall and Jackson, 1958). A characteristic
septum or basal plaque was found at the level of the cell surface

FINE STRUCTURE 57

where the two central fibers end, and there also is to be seen a
" kinetosome " or ampule terminating the central fibers. According
to the latter report, the base of the cilium, with its prolongation
of the peripheral fibers, continues inward as a cylinder without
rootlets extending into the endoplasm but sometimes showing
minute granules in longitudinal rows along its cylindrical wall.*

6. Fine structure of the nuclei

Light microscopists describe the matrix of the macronucleus as
of homogeneous granules in a sort of meshwork with one clear
spherule, the nucleolus, usually found in each node of a chain
nucleus. The nuclear membrane swxUs loose in distilled water
as a highly birefringent and therefore well-organized layer;
significantly Hke the shell of the pigment granules, its composition
was indicated to be phospholipid (Weisz, 1949a). Resting on the
macronuclear membrane. Park (1929) described osmiophihc,
bleb-Hke bodies, i to 22 for each node. He suggested that they
might be secretory droplets, reminiscent of the parabasal body
associated with the nucleus in flagellates.

Electronmicrographs reveal further details (Faure-Fremiet and
Rouiller, 1955; Randall and Jackson, 1958). The granules within
the macronucleus turn out to be clusters of filaments, possibly
beaded, these masses being more or less equidistantly spaced
within a clear nuclear sap. The outer layer of the macronucleus
is porous, showing curious tubular processes extending and
branching into the endoplasm, w^hile the inner membrane appears
to be a system of tubular vesicles joined by sheets, resembling
spaghetti laminated in plastic.

In his cytochemical studies, Weisz (1949a, 1950b) made
Feulgen and Millon tests which indicate that protein and nucleo-
tide or potential nucleotides are homogeneous in concentration in
the macronucleus at all times, except of course during conjugation
when the old macronucleus is resorbed. His methyl green tests
suggested however that the nucleotide. — desoxyribonucleic acid —

*From observations on stentor membranelles, Sleigh (i960) ingeniously
integrated motor and recovery strokes of cilia as resulting from one wave
of localized contractions passing up and around the cilium, the propulsive
phase occurring when bending starts on one side at the base and the rest
of the cilium is straight.

t^8 THE BIOLOGY OF STENTOR

varies in degree of polymerization along the nuclear chain. Faure-
Fremiet and Rouiller speak of this DNA as in the form of micro-
somes. In dark-field illumination I have found that the exposed
macronuclear nodes are often a glow^ing light blue, which may
indicate something of their composition or state as aifecting the
scattering of light.

The micronuclei of stentors were first described by Maupas
(1879) and later by Johnson (1893). These very small nuclei
reside on or near the macronuclear chain. Multiplying mitotically,
they are typically chromosomal, as further substantiated by their
behavior during conjugation (see p. 329).

7. The endoplasm

The interior cytoplasm w^as examined by Weisz (1949a) who
found that it did not stain with basic dyes and only diffusely with
acidic. Neutral red was taken up by the living coeruleus and stained
various inclusions so that stentor may be said to have a 'Vacuome" ;
this dye was segregated by the contractile vacuole. Chromidial
nets and metachromatic volutin granules were not present.

By introducing minute electrodes into the cell, Gelfan (1927)
went to much trouble to prove that the electrolyte concentration
in Stentor is higher than that in the surrounding fresh water medium,
a conclusion which could have been inferred from the pulsation
of the contractile vacuole in voiding water imbibed through
osmosis. The specific conductance was lower in stentors than
in three other ciliates tested. Internal conductivity decreased with
injury, presumably due to the leakage of electrolytes from the cell.

The endoplasm of stentors presents a foamy appearance which
was first emphasized by Butschli in keeping with his theory of the
alveolar nature of protoplasm. Correspondingly, Randall and
Jackson found by electron microscopy that the endoplasm con-
sists of numerous vacuoles within a matrix which shows many
small particles and vesicles. An endoplasmic reticulum was not
revealed. The vacuoles have a definite membrane and seem to be
especially numerous in the sub-cortical regions. Within the endo-
plasm are also found typical protozoan mitochondria. Randall and
Jackson (1958) found them to be Janus green B positive and
having triple membranes of equal width, if one counts the material
between outer and inner layers as the third. The enclosed tubular

FINE STRUCTURE 59

vesicles terminate on the inner membrane. Both membrane and
vesicle w^alls showed small opaque particles and the interstices
gave the appearance of a finely particulate matrix. Empty mito-
chondria were found, as well as others without bounding mem-
branes amongst the normal forms of these bodies, reminding one
of Weisz's (1949a) suggestion that the mitochondria are utilized
in starvation and regeneration.

Food reserves are also present in the form of fat droplets and
glycogenoid granules. In addition, multiformis and introversus have
yellow, brightly refringent bodies or crystals in the endoplasm,
the nature of which has not yet been determined.

Altogether, this detail of fine structure represents about as
extensive and intensive a cytodifferentiation as we are likely to
find, comparing favorably with that of the most complex hypermas-
tigont flagellates in arthropods and ruminant commensal ciliates.
The number of definable parts which have been "compacted**
into the minute volume of a stentor is quite amazing and attests the
extremely fine-grain structure possible to organisms. For instance,
there are about 32,500 fibers in the complement of ribbon bundles
or ^m-bands alone, not to mention the countless cortical granules,
etc. Not only in number but in their greater diversity the minute
parts of an organism like Stentor stand in contrast to the cyto-
differentiation of most tissue cells. This difference in manifest
complexity is one of the reasons why some biologists have hesitated
if not refused to call protozoa cells or unicellular forms. Yet a
stentor represents no more or less a separate nucleocytoplasmic
system than a neurone. And a bridge between protozoan and tissue
cell may perhaps be found in the egg; for if one refuses to call a
fertilized egg a cell, all seem to agree that its cleavage products are
cells and from either of the first two cells, let us say, the whole
complex multicellular organism can be derived by embryogenesis.
It may therefore not be too much to infer that such a cell is
intrinsically as complex as Stentor, if not more so, but manifesting
this complexity in development through multicellularity instead of
more directly in itself.

What, then, shall we *'do" with all the complex cytodifferentia-
tion we find in Stentor ? One approach is to study, if possible,
certain types of parts in themselves. For example, the ^m-bands

6o THE BIOLOGY OF STENTOR

may represent a unique contractile structure the elucidation of
which might define a specific parameter of muscle physiology.
This is the orientation of Randall and his co-workers. Or by
emphasizing diflFerences in the parts of descendant individuals one
might explore new aspects of genetics such as the importance of
cytoplasmic inheritance, as in the work of Sonneborn ; and poten-
tially this approach could be most fruitful in Stentor in which
micrurgical exchange of cytoplasms and nuclei is not difficult.

One may also consider all the fine structure from the standpoint
of epigenetics or morphogenesis. Obviously the criterion of re-
generation and other types of epigenetic performance, is the fine
structure also what "does" the morphogenesis? Is the fine struc-
ture the cause or the result of morphogenesis? For instance, in the
simple healing of an incision in Stentor the ribbon bundles and
their many fibers in the clear stripes apparently rejoin; therefore in
addition to their function of conduction and contraction have these
fibers also the capacity for guiding their reintegration? In some
way of which we have yet no understanding all the fine structure is
integrated, an obvious inference from the integrity of the organism
and its normative tendencies which will receive specific documenta-
tion as we proceed into the experimental studies. It is as if, in
addition to its specific physiological functions, every meridional
unit of a stentor " knows " when one of the mouthparts is missing
or the oral structures are misplaced, for all seem to cooperate in
correcting the deviation from the norm, some parts, often quite
distant from each other yet somehow in effective communication,
taking the leading roles.

For example, if the head is rotated i8o° on the body, the
mouthparts disconnect from the membranellar band, migrate into
the frontal field and are resorbed — ^this specific disjunction and
taking down of structures being in its way a compounding of the
marvel of their original construction. Then, in a far distant part
of the cell a new oral primordium begins on some signal, and
experiment indicates that all parts of the cortex support anlage
formation and are involved in the timing of its development.

But this is to anticipate a part of our story, yet such phenomena
pose the further problem of the integration of all the parts of
Stentor in terms of the fine structure w^e have described if not
some further principle.

CHAPTER V

GROWTH AND DIVISION

I. Growth

Division in Stentor is of course usually preceded by growth to
the definitive size. Authentic structural growth and de-growth
occur, as well as stretching of parts. The first is seen, for example,
as increase in the number of lateral clear stripes with their fibrous
elaborations ; the second, in resorption of striping or macronuclear
nodes under various conditions ; and the potentialities for stretching
are shown when small patches of ectoplasm come to cover the
surface (see Fig. 25A) or when isolated nodes change from round
to spindle form (see Fig. 82A). We still have no comprehensive
understanding of growth in the individual stentor cell and no
investigator has yet addressed himself directly to this problem,
but the pigmented stentors, especially, offer many advantages for
such a study. In the first place, most stentors contract into a
sphere and a fair estimate of their volume can be obtained by
measuring one diameter. Grafted pairs can be identified by their
pigmented stripes, which are often seen to increase both in number
and in length (Fig. 11 a). Natural markers sometimes are found in
the cortex of the cell and indicate growth by their apparent dis-
placement (b,c). Similar markers for following the growth of the
ectoplasm could be made by small disturbances of the pigment
striping, for the smaller these disarrangements are the longer they
persist before correction (Schwartz, 1935).

Differential growth of major cell constituents is to be seen in
the recovery of endoplasm in stentor "skins" (see Fig. 25B), as
well as in the rapid recovery of the macronuclear mass when all
but a single node has been removed from the cell (see Fig. 86c).
In normal stentor cultures I have occasionally found specimens
which were much longer than usual, with half the normal number
of lateral stripes, as if growth in length had occurred at the expense
of growth in width. On isolation this disparity was later corrected.

61

62

THE BIOLOGY OF STENTOR

B

Fig. 1 1 . Observations suggesting ways of studying growth in
stentors.

GROWTH AND DIVISION 63

The lateral striping undoubtedly increases in length, though
we do not know when or how this occurs. Johnson said that during
division the striping at the posterior end of the future anterior
daughter cell lengthens as a new tail pole is formed for it. But we
do not know, for instance, how new body cilia in a growing longi-
tudinal row could be interpolated between those already present,
or whether cilia are added only at the end of a row, or increased in
number only during fission.

Increase in the number of lateral stripes is much more obvious.
Brauer (1885) first suggested that the shorter stripes which do not
run from pole to pole are new ones resulting from multiplication,
and that pigment stripes multiply by the interpolation of a new
clear stripe was proposed by Johnson (1893). The number of
granular and clear stripes increases with the size of the animal.
Largest specimens of coeruleiis have about 100 stripes of either kind.
Tiny individuals from starving samples which were about one-
sixteenth the maximum cell volume had approximately 66 stripes of
either type. When a cell divides transversely, the division products
have about 80 stripes of each kind at their circumference because
about 20 are carried into the frontal field of the opisthe or posterior
daughter cell. Presumably, interfissional increase is from 80 to 100
as the volume is doubled. The number of stripes thus increases
with the volume but less rapidly, some of the surface increase
probably being accommodated by widening of the granular stripes.
These remarks agree well with earHer conclusions of Popoff (1909).

A. A case of differential increase in length and breadth of a
grafted patch, a: Patch bearing primordium grafted into back
of a host from which mouthparts were then excised, b: Primor-
dium resorbed but length and number of stripes of the graft also

promptly increased.
B. Shift in relative location of a tube formation, a: Grafted
pair developed an adventitious, gullet-like tube, a: In process
of reorganization-regeneration next day, with tube now displaced
far posteriorly, possibly indicative^ of differential growth and
resorption in the ectoplasm.

C. Problems of growth indicated by another marker, a:

Grafted pair developed a tube which was resorbed {b) leaving

patch of dense pigment granules at the surface, c: Specimen

reorganizing two days later with pigment clot now far to left of

primordium site.

64 THE BIOLOGY OF STENTOR

Very small species of Stentor, like multiformis, are not miniatures
of the larger since they contain fewer stripes which are therefore
relatively much wider in proportion to the cell volume.

The lateral stripes tend toward a certain maximum number.
When stentors were cut in two longitudinally, slender fragments
were produced which had half the normal complement of stripes.
The aboral halves not bearing the mouthparts were followed
because stripe increase in them is more easily seen than in oral
halves with their finer stripes, as Stevens (1903) noted. Stripe
multiplication occurred at the line of heal (Tartar, 1956c), and in
5 days the specimens regained the normal width and number of

Fig. 12. Increase in lateral striping.

A. Non-oral longitudinal half has less than half the normal
number of stripes {a), h: New fine granular-stripes appear as
regeneration primordium forms at suture. Continued splitting
of granular bands with interpolated clear stripes {c) increases

stripes to the normal number {d).

B. Parabiotic graft of two oral halves has somewhat more
than the normal number of lateral stripes yet increases the com-
plement to near zN, in correlation with the double individuality.

Stripes (Fig. i2a). When two of the oral longitudinal halves were
grafted together in homopolar parabiosis producing artificially a
more than normal sum of stripes, stripe increase still occurred
and very wide doublets with about twice the normal number of

GROWTH AND DIVISION 65

Stripes were produced (Fig. i2b). From such experiments we may
eventually learn how the number of lateral stripes and fibrous
bands is controlled.

Stripe multiplication apparently can occur in any meridian of
the cell. When it occurs on the dorsal side where the pigment
stripes are wide, the pigmented stripes seem very quickly to attain
the width normal for that area after they have been split in two by
the interpolation of new clear stripes. Figure 13 thus shows

Fig. 13. Specimen suggesting that splitting of granular stripes

in wide-stripe area quickly leads to broadening of these stripes

in harmony with those adjacent.

branches of the pigment stripes as wide as the " stem ". This illus-
tration also demonstrates that the splitting of the granular stripes
can occur in either direction. But the greatest stripe increase occurs
at a specific region in the side of the cell where the mouthparts are
located. This area has become a key to stentor morphogenesis.

Brauer (1885) had early described in coeruleus that posterior to
the mouth there is a ''fiber which may give up to 10 secondary
members lying against each other". Essentially this triangular area
is a place where about 25 clear stripes and an equal number of
alternating pigmented bands do not run all the way to the posterior
pole but are bounded on each side by stripes which do (see Fig. i).
Since the stripes become shorter as they approach the left boundary
stripe and because that stripe takes something of a diagonal course,

66 THE BIOLOGY OF STENTOR

the appearance is that the left boundary stripe is branching.
Schuberg (1890) therefore called this area the "ramifying zone",
evidently the area of stripe multiplication and also the site of the
oral primordium. In the left anterior corner of this triangle the
widest pigment stripes begin splitting into narrow stripes. As
the split proceeds posteriorly the next wide stripe to the left begins
splitting, wdth the result that a series of stripes of ever increasing
length is formed to the right and the characteristic ramifying zone
is thus achieved. At least this is the general impression, though
other details doubtless need to be added. If so, growth takes a
spiral course, as it were, with the zone of increase gradually moving
to the animal's left as new short stripes are added and older stripes
to the right increase in length until they reach the posterior pole.
Correspondingly, the oral primordium which appears in this zone
would continually shift leftward, with the result that new mouth
parts appear always somewhat to the left of those preceding. Such
spiral growth recalls that of the fruiting body in certain fungi
(Delbriick and Reichardt, 1956).

We have already remarked that pigment stripes are mere fill-ins
and their splitting is doubtless due to the emergence within them
of new clear stripes with their ciliary rows and fibrous structures.
Each new clear stripe would then not be connected with others,
corresponding to the description of Villeneuve-Brachon (1940).
Later the clear stripes do join together and cut off the split
branches of a pigment band. Older figures thus show fibrous struc-
tures of the clear stripes as branching and re-branching in the
ramifying zone. From what we now know about kinetics and
myonemes it is evident that anastomosis would entail great struc-
tural difficulties and we have to leave this problem until appro-
priate EM studies are available. Because the number of stripes
tends toward a fixed upper limit, the need for stripe multiplication
may therefore vary, possibly being minimal in stentors that have
lived for a long time without dividing. This would account for the
observations of Johnson, and much later of Dierks, that the
ramifying zone is variable in its aspect, sometimes even un-
identifiable as such.

Multiplication of clear stripes could not occur by simple
splitting since this would leave one branch with cilia and one
without. One branch would have to migrate sub-cortically and

GROWTH AND DIVISION 67

then push up through the adjacent pigmented stripe somehow.
Apparently the new clear stripes with all their attendant complex
differentiations arise in situ, but of their origin we know nothing.
If we accept the genetic continuity of the kinetosomes and regard
them as fibrogenic granules which produce not only the cilia but
also the fibrils of the ribbon bundles in the clear stripes (see
Lwoff, 1950), the new kinetosomes will have to be traced to their
progenitors.

Growth of other parts of the stentor cell present their own
special problems. The young daughter cell has a membranellar
band proportionate in length to its size. When full-grown the
length has increased and is still proportionate. Morgan (1901a)
therefore thought that this organelle grows in length and implied
that the number of membranelles increases. But if this were so,
Stentor would need two ways of producing membranelles : through
primordium formation and in situ. If the length of the membranellar
band is abbreviated by cutting, compensating growth does not
occur, contrary to a dubious observation of Stevens (1903). There-
fore it seems more likely that increase in length is accomplished
by the spreading apart of membranelles, already present, as
obviously occurs during the development of the oral primordium.
This point could be settled by counting the membranelles. Both
the total mass and the surface area of the macronucleus increase,
together or separately. Growth of the nucleus will be discussed
in a chapter devoted to that organelle. For the present, we may
merely remark that the trophic macronucleus, perhaps like the
giant salivary gland chromosomes of insects, represents a form of
nuclear material which adapts to the size of the cell and not vice
versa (cf. Goldschmidt, 1940).

2. The course of normal division

From regeneration studies we may say that a stentor could
multiply by simply cutting itself in two, the resulting daughters
then regenerating those structures which they lack. This would
be cell fission in the strict sense of the term and does occur under
unusual conditions of experiment. A stentor might even cut itself
into several fragments each of which would be viable, for neither
the whole nor the half represents a minimum unit of potential
organization. But in nature division is accomplished by trans-

68

THE BIOLOGY OF STENTOR

Fig. 14. Stages in division as seen in living S. coenileus.

Stage o. First indication of fission: a splitting of granular

bands at the bulk-center of the cell on the oral side. (Fine striping

of frontal field and ciliary membranelles shown here but omitted

in remaining sketches.)

GROWTH AND DIVISION 69

forming the parent organism into two individualities the morpholo-
gies of which come to exclude each other and are finally separated
by fission. To paraphrase one of the ablest students of Stentor:
the situation in ciliates is the reverse of that in metazoa since all

Stage I . Initial appearance of oral primordium for the pos-
terior daughter cell, as a transverse rift in the lateral ectoplasm.

Stage 2. Primordium enlarges by extending anteriorly with
a new curvature to the right. Anlage has a faint glisten in
reflected light but no cilia are apparent yet.

Stage 3. Primordium increases in length and membranellar
cilia are visible but not yet grown to their final length. Continued
multiplication of striping within curvature of the anlage.

Stage 4. Primordium grown to nearly its full length and oral
cilia are organized into closely-packed membranelles which beat
in slow^ metachronal rhythm. Moniliform macronucleus still
shows no change.

Stage 5. An enlargement or etched space appears at posterior
end of the primordium, site of the future mouthparts. Anlage
now embraces many fine stripes. Macronuclear nodes begin

fusing.

Stage 6. Posterior end of membranellar band coils inward
sharply to form gullet and cytostome. During this stage severing
of stripes begins at each side of anterior end of the primordium
and progresses on both sides around the cell to form the fission
furrow. Macronucleus fused to a compact mass. Fading or
partial dedifferentiation of oral pouch and gullet, begun at stage
5, is now at its maximum.

Stage 7. Primordium migrates posteriorly, its anterior end
being cut out of the anterior daughter cell whose stripes heal
together at once in a herringbone pattern which will for a long
time distinguish proter from opisthe. Gullet and cytostome are
now nearly complete and fine striping enclosed by the anlage is
being carried forward as the new frontal field. Compact macro-
nucleus elongates to rod shape and begins renodulation from the

ends.

Stage 8. Oral pouch is now formed as an inpocketing of the
frontal field adjacent to the cytostome. Primordium migrates to
its definitive position at anterior end of opisthe and body striping
becomes parallel to the membranellar band. Rod-shaped macro-
nucleus is divided by constriction of the furrow which has nearly
separated the daughter cells and formed the new posterior pole
of the proter. Mouthparts of proter redefined, membranellar
band probably proportionately reduced from length in original
cell. Fission products then twist apart. (After Tartar, 1958c.)

yo THE BIOLOGY OF STENTOR

the "embryology" occurs before instead of after reproduction
(Johnson, 1893). (An exception is found in budding, such as chain
formation in lower worms in which new individualities are com-
pletely formed before separation, an analogy with protozoan
reproduction first suggested by Gruber (1885a).) A self-cutting or
fission line is indeed formed in StentoTy but it is neither the cause
nor the necessary result of the conversion of one individuality
into two.

MultipHcation by fission in Stentor was first observed by
Trembley (1744). With continued refinement of the microscope,
further details of division were given by Stein (1867), Moxon
(1869), and Cox (1876); yet it remained for Schuberg (1890) to
provide the first really comprehensive account of what takes place.
Stentor division as a developmental process was beautifully and
accurately drawn in the illustrations of Johnson (1893) and
Schwartz (1935). Visible changes during division have been
designated as a numbered series of stages (Tartar, 1958c) and are
shown in Fig. 14.

Restricting the story to the best-known species, coeruleus, the
first sign of the formation of a new individuality is a splitting of
the pigment stripes on a diagonal in the mid-ventral region
(Fig. 14-0), first noted by Stevens (1903) and very probably
representing the insertion of new clear bands. In this area of
stripe multiplication a rift soon appears as the very beginnings of
the new set of feeding organelles for the future posterior daughter
cell. This primordium lengthens at both ends according to
Johnson and broadens quickly to its definitive width while the
long oral cilia develop within it. At this time the anlage is usually
in the form of a crescent and this appearance is generally diagnostic
of an early divider. (Occasionally the anterior end of the primor-
dium may extend straight forward towards the old oral region, as
it does in reorganizing animals, so one cannot always be sure.)
There is increasing multiplication of stripes as pigment bands at
the anterior border of the primordium split into 2, 4, and 8 rows
(Fig. 15A) (Schwartz, 1935). It is these new, fine stripes embraced
by the presumptive membranellar band which will form the new
frontal field of the opisthe. Such additions may increase the
circumference of the cell according to Stevens and Schwartz,
though this is certainly not obvious.

GROWTH AND DIVISION

71

A

Fig. 15. Details of division in S. coerideus.

A. Eight, four and two-fold splitting of granular bands to
produce the fine bands of the new frontal field enclosed by a
stage-2 primordium. The number of interpolated kinetics

(clear stripes) is correspondingly increased.

B. Macronuclear division according to Johnson, showing
clumping of nodes, followed by rod formation, pinching in two
of the nucleus by the dividing cell and beginning renodulation.
A preliminary constriction of the massed nucleus (c) may go to
completion {d') with larger portion extending into and contri-
buting to the nucleus of the proter. (After Johnson, 1893.)

C. Persisting fission in late divider after posterior excision;
rod-form macronucleus distributes itself accordingly and is
unequally — but proportionately — divided. (After de Terra,

I959-)

72 THE BIOLOGY OF STENTOR

Continuing its development, the posterior end of the primordium
begins to coil inward to form the gullet (stage 6). At this time the
cell usually shows a central contraction tending slightly toward a
dumbbell shape, but this constriction is not coincident with the
future furrow (Johnson) and makes its appearance earlier, as our
figure shows.

A fission line then appears at both sides of the anterior end of
the primordium. To the right it cuts off the presumptive frontal-
field striping and runs approximately perpendicular to these stripes
which have been somewhat distorted by the movements of the
anlage. To the left, the furrow runs sharply posteriorly while
cutting obliquely across the wide granular stripes in this area, the
two ends of the fission line moving more and more transversely as
they proceed around the cell to meet on the lower dorsal side. By
being oblique, the furrow can cut the primordium, which runs far
anteriorly, into the posterior cell and yet divide the parent into
approximately equal daughters. The fission line is made evident
by a change in the pigmented stripes which leaves a colorless band
across each one. Possibly this may be caused by the formation of
new transverse contractile structures, pushing the granules aside
and later responsible for constriction at the furrow.

Only when the membranellar band is fully formed and the gullet
begins to develop (stage 5-6) does the macronucleus undergo a
relatively rapid series of changes. At this time the nodes of the
nucleus begin to coalesce within the common nuclear membrane.
According to Johnson, this fusion occurs at separate loci because
it sometimes may result in a premature breaking of the chain.
Eventually the nucleus is compacted into one more or less spherical
mass in the center of the cell, though unsuccessful enucleation
experiments performed at this time indicate that occasionally one
or more nodes may remain isolated. Johnson described the
clumped nucleus as then showing a preliminary constriction
which lasts for about half an hour, then disappearing as the nucleus
elongates into a rod (Fig. 15B). But sometimes this constriction
was completed, the nucleus then and there separating into two
parts, not always equal. When unequal, the larger part showed a
secondary division later, resulting in a more equal allocation of
the nuclear material. These observations of exceptional behavior,
as well as the fact that division usually occurs later in the rod

GROWTH AND DIVISION 73

Stage, indicate that clumping is not for the purpose of dividing the
nucleus equally. Hence Johnson suggested that both the clumping
and the preliminary constriction are a recapitulation of phylogeny,
harking back to the form in fission of the nucleus in less specialized
protozoa. Nor is there any evidence of macronuclear reorganization
occurring during clumping; for Johnson found that the character
of the macronuclear matrix remained unchanged and he stated
explicitly that there was no indication of linear arrangement of
threads or the formation of something like chromosomes.

At the earliest, a new contractile vacuole for the opisthe makes
its appearance in the proper location at stage 4. Its formation is
therefore probably not initiated by the division furrow, which is
not yet visible, though Weisz (1951b) found that cutting the stripes
of non-dividing stentors transversely would induce the temporary
formation of a posterior vacuole and I have confirmed this.

As the division line cuts around its upper end, the primordium
can bend more sharply and move backwards into the future
opisthe. Schwartz described how the anlage shifts with reference
to the striping on the left side so that these lines, at first parallel,
come to lie at right angles to the new membranellar band. The cut
ectoplasm of the future proter, or anterior daughter, closes together
immediately as the primordium migrates posteriorly, with the
result that a herringbone-pattern of stripes is formed which is
somewhat asymmetrical because the furrow ran more sharply
posteriorly on the left side. Anterior can be distinguished from
posterior daughters long after separation because this pattern may
persist for three days afterward in starved animals. Gradually the
abbreviated stripes grow posteriorly to reproduce the typical
ramifying zone.

In the meantime the anterior portion of the opisthe bearing the
primordium has been bulging outward while gullet and oral pouch
have been forming and shifting forward. The clumped macro-
nucleus then elongates parallel to the main axis to form a long
rod, which begins to nodulate simultaneously at both ends. As
constriction continues the stripes of the proter are drawn together
to form its tail and possibly extend in length as they narrow to a
point. The half-nodulated macronucleus now divides in two;
Johnson thought that its division is autonomous and the same is
implied in Causin's (1931) report that even in regeneration the

-74 THE BIOLOGY OF STENTOR

macronucleus can divide within the single cell, the separated parts
then rejoining. Yet it has been found (Popoff, 1909; de Terra,
1959) that if cell division is unequal the macronucleus is likewise,
quite as if this nucleus were passively pinched in two at the rod
stage by the constricting furrow (Fig. 15c). The proter is now
connected with the opisthe only by the tail pole, still attached at
the aboral end of the latter's membranellar band ; and up to this
time the daughters have continued to coordinate their backward
and forward swimming together (Gruber, 1886). Final separation
seems to be due to a twisting apart which sunders the fine con-
nection between the two cells.

According to Johnson's account the micronuclei swell and
undergo mitosis after macronuclear division is completed, i.e.,
within the essentially separate daughters. Each daughter thus
achieves about the same number of micronuclei as the parent cell.
After completing nodulation there are about the same number of
macronuclear beads in each product as there were in the original
animal. Therefore these nodes are half the size of the original ones
and nuclear growth consists largely of increase in the size of the
new nodes, though occasionally one segment may later divide in
two. This doubUng of the nodes of the macronucleus was first
noted by Balbiani (1882) and later confirmed by Johnson; Stolte;
and Tartar (1959c). It is therefore plausible that the macronucleus
clumps together to make possible its renodulation at once into
twice the original number of nodes.

De Terra's (1959) studies on coeruleus have shown that the
uptake and incorporation of radiophosphorus is very rapid before
division but drops to one-twentieth of this rate when the macro-
nucleus is compacted and fission is in process, indicating that
nuclear increase does indeed occur by growth of the nodes and
not when the nucleus is in the coalesced stage immediately
preceding fission.

The time required to complete the act of division is probably
quite variable but about 6 hours would be a reasonable average.
The first stages having to do with primordium formation proceed
more slowly. Fusion of the nodes of the macronucleus can occur
in one hour according to Johnson and the nuclei renodulate as
rapidly. Timing of the complex events in division presents special
problems in the integrated action of the cell, and one possibility

GROWTH AND DIVISION 75

is that any specific event triggers the next following (see p. 295).
From the general description of division wq see hov^ the
daughter cells are composed of parts both old and new. Endoplasm
and nuclear material are halved, but the macronuclear beads are
reconstituted and new micronuclei appear as division products of
the old. The original feeding organelles go to the proter. They
may undergo a slight dedifferentiation during fission, with the
oral pouch temporarily disappearing as such, but there is no
comprehensive regression and rediff"erentiation as in Bursaria
(Schmahl, 1926) or Condylostoma (Villeneuve-Brachon, 1940).
Nevertheless, the original oral structures which are at first too
large are gradually reduced to proportionate measure in some
manner which is not yet understood (Weisz, 1951b). The proter
also retains the original contractile vacuole but it has to form a
new tail-pole and holdfast. The original tail goes to the opisthe
and is also at first too large, but the posterior daughter has a new
set of oral structures formed entirely independently of the old.
It also develops a new contractile vacuole, though this may be
but an enlargement of contributory channels of the old. The striped
ectoplasm is divided largely unchanged between the two daughters,
although there may be a growing out of stripes and fibers in the
formation of the new tail as there is also a post-fissional stripe
multiplication in the opisthe to form a new ramifying zone
(Schwartz, 1935). There is, however, a marked decrease in the
number of lateral stripes because those in the ramifying zone are
shifted to the frontal field. Presumably the old body cilia are
passed on unchanged, for there is never a time when ciliation is
lacking. One should keep in mind, however, the amazing possibility
described by Schmahl for Bursaria truncatella^ in which a ciliary
molting seems to occur, resorption of old cilia and formation of
new ones occurring simultaneously and therefore easily overlooked.

3. Nature and location of the fission line

We still know practically nothing of what happens at the division
furrow. It seems highly improbable, as Johnson remarked, that
the line represents the edge of a plane passing through the interior
of the cell though Weisz (1956) conceived that there might be
some sort of separation or pre-division of the endoplasm which
later comes to expression on the surface. There is no obvious

76 THE BIOLOGY OF STENTOR

rupture in the surface of the cell as Schuberg (1890) first thought,
for even with the most drastic manipulation of dividers no gaping
or separation occurs along the fission lines. Yet it is reasonable to
suppose that there is a severing of the granular stripes and fibrous
structures in the clear stripes, because we know that the striping
also has a strong tendency to heal together when cut and this
procHvity would have to be overcome. The severance is, however,
not necessarily irreversible. Popoff (1909) described one case and
I have seen another in which division was aborted and the fission
line disappeared without a trace, showing the pigment stripes
again running continuously from pole to pole.

All that we can be certain of at present is that the pigment
granules are moved away at the levels where the colored stripes
cross the future furrow. Something of how this occurs may be
shown in the aboral longitudinal half of a stage-4 divider which
still continued on its course and attempted division. As shown in
Fig. I 6a, the granules at mid-level in each stripe were seen in one
place to have shifted from the center of the stripe posteriorly and
this may have been the prelude to the complete depigmentation of
the stripes in the adjacent region. This appearance resembles that
of stripe multipUcation, and it is possible that new, short, posterior
kinetics were being introduced which pushed the pigment granules
aside as they formed double rows of cilia demonstrated by Randall
and Jackson for the new tail pole. The half-cell did not complete
division, but it did form a secondary tail-pole, very likely because
the body striping remained severed. This case is also significant
in showing that although the furrow normally begins at the
anterior end of the oral primordium this is not essential to furrow
formation.

The fact that the fission line does not form all at once but
progresses in two directions around the cell suggested to Weisz
(1951b) that there are two waves of dissolution, each beginning at
one point on a given stripe and spreading radially until it touches
and sets off a new center of dissolution in the next adjacent intact
stripe, like the firing of a fuse. This would not explain, however,
why the fission line moves sharply posteriorly on one side of the
primordium; nor why, in Stevens' (1903) observation of longitu-
dinal halves of dividing stentors, the furrow stopped short by two
pigment stripes on each side of the line of heal; nor why the line

GROWTH AND DIVISION 77

Stops at ''indifferent" striping (Fig. i6b). The latter blockage is
probably not because the indifferent component is not in a "state
of division" since Weisz (1951b) had shown that, although removal
of patches of ectoplasm in the path of the presumptive furrow still

D

Fig. 16. Pertaining to the fission line (S. coenileus).

A. Continuation of furrow formation in the non-oral half of
divider cut longitudinally at stage 4. Note how pigment granules
withdraw from the middle of granular stripes and accumulate
posteriorly.

-78 THE BIOLOGY OF STENTOR

permits normal fission as the division line crosses the suture, if
other ectoplasm from the same dividing animal is shifted to fill
the gap the furrow then does not cross over. Evidently, as Weisz
remarked, the path of the fission line is strongly and uniquely
determined and cannot be initiated by local point-to-point processes
alone.

Fixity of the presumptive furrow is shown by the observation,
first reported by Johnson, that if the anterior or posterior end of a
divider is cut off just before the furrow is to appear, unequal
daughters are formed because the fission line appears in its normal
place. If at the same stage a cell is prematurely divided by cutting,
fission still occurs along the predetermined oblique path half of
which lies in each fragment with the result that small blebs are
separated (Fig. i6c). An artificial cutting of the body stripes is thus
not used as a substitute for the normal fission Une. Also, if pre-
furrow dividers are cut through transversely, first on one side and
then on the other so that the two halves remain fused together,
division still occurs but it is oblique and not in the line of heal.
Still more convincing is the experiment in which these cut halves
are rotated 180° upon each other so that the body striping does not
match or heal together, as evidenced by obvious discontinuities in
the granular stripes; for even in this case relatively normal and
equal division occurs with both parts of the severed primordium
going to the opisthe and division was obviously not in the line of
created discontinuities (Fig. i6d).

These cutting experiments not only attest the fixity of the fission
line, even before it is visible as such; they also strongly indicate
that division in stentor is not due to ingrowth of separating mem-
branes, surface tension changes, or other mechanisms which have

B. Stage-5 divider grafted to a non-divider. Fission continues
on the divider side but furrow stops when it meets indifferent
striping. Daughter cells do not separate because held together

by the partner of the graft.

C. Stage-4 divider cut in two continues fission along the pre-
determined fission line resulting in separation of small blebs.

D. Stage-4 divider with anterior rotated on posterior half,
separating the primordium into 2 sections. Both parts of the
anlage still go to the opisthe. Fission line does not follow the

discontinuity of striping at the suture.

GROWTH AND DIVISION 79

been invoked to explain cleavage in eggs. And the old notion of a
constriction band which must remain a complete ring in order to
exert a pull is of course precluded. There is no doubt that con-
striction occurs, which is probably from point to point on the
furrow. Localized contraction together with the cutting of the
lateral stripes and bands are probably the two agents directly
involved in fission.

We now consider observations and experiments relating to the
question of how the division line is determined so that it should
be at a certain level on the cell, normally such as to produce
daughter cells of equal volume. This line appears to be precisely
laid out as a perfectly smooth curve without indefinite zig-zagging.

Suggestively, there is a cell constituent which comes to follow
this contour during division. This is the complement of glycogenoid
granules which comprises the carbohydrate reserves of Stentor.
Weisz (1949a) had previously noted that these reserves are about
equally distributed between the two daughter cells though initially
lying at the posterior end, and later I supplied an exact account of
their distribution (Tartar, 1959a). In a well-fed pre-divisional
stentor the granules lie in a broad sub-cortical band at the posterior
end, exclusive of the pole itself and interrupted or missing in the
post-oral meridian (Fig. 17). About stage 5, considerably before
the first visible indication of a furrow, half of the granules migrate
forward. Those left behind become somewhat more diffuse than
they were before. The anterior border of the migrated complement
is also irregular, but its posterior boundary forms a sharp line
precisely defining the path of the fission fine which soon appears.
When furrowing occurs, therefore, it merely segregates the
reserves which were previously divided. These events do occur at
just the time when the fission fine is being determined with respect
to its location. But if immediately preceding this stage the carbo-
hydrate reserves are excised from the cell by cutting off the
posterior end where they still reside, division can still occur in their
absence. Therefore the peculiar behavior described is not the
cause but rather seems to be the sign of other factors which
locate the fission line.

Whatever it is that determines the path of the division furrow,
there is further evidence of the pervasive nature of this agent from
unpublished experiments in which excision of parts far distant

8o

HE BIOLOGY OF STENTOR

from the circumference of the cell produce grossly unequal
daughters, not accounted for by the relatively small loss of cyto-
plasm. Enlarging upon an experiment by Weisz (1951b), I found
that if the mouthparts or the membranellar band or both are
minimally excised or caused to be shed by salt treatments from
early dividers, proters are later produced which are only about
half the volume of the opisthes (Fig. i8a). That this difference is
too great to be due to the ablations alone is obvious, and confirmed

Fig. 17. Predivision of carbohydrate reserves (S. coeruleus).

A. Normal distribution of glycogenoid granules in a
subcortical band at the posterior end, open in the primordium

meridian.

B. Separation of granules into two groups at stage 5, the
posterior border of the anterior aggregation precisely coinciding
with the future fission line and the posterior granules somewhat

diffuse.

C. Division leaves about half the carbohydrate reserves in

each cell. (After Tartar, 1959a.)

by the fact that such operations on dividing animals at stages 5 and
6, when the division line is already determined, do yield products
which are approximately equal. It therefore appears that such
interferences have a marked effect on whatever determines the
level of the fission line, shifting it far forward from its usual
position. A rare case of division in a fusion complex of parts of
two stentors suggests that the fission line may also be laid down
far posterior of its normal location (Fig. i8b).

GROWTH AND DIVISION

8l

Fig. 1 8. Experimental conditions affecting location of the
division furrow.

A. Head only of stage-4 divider excised; furrow is shifted
forward with result that proter is only half the volume of the
opisthe. Proter begins regeneration only after fission is

completed.

B. Oral side of a coeruleus grafted transversely to animal from
which anterior end was excised. Unusual subsequent dividing
off of an anucleate product without an oral primordium {x)
indicates how extensively process of division may be upset by

m i sarrangements .

By centrifugation, Popoff (1909) was able to produce unequal
fissions in coeruleus. The macronucleus was also unequally divided,
yielding, for example, a small cell with 3 nodes and a larger one
with 1 6. Presumably the level of the fission line was located other
than normally, though no details w^ere given. Prowazek (1904)
likewise reported unusual cases of shift in the fission level leading
to unequal daughter cells, as did Packard (1937), without being
clearly aware of what he was observing. Altogether, these studies
show that the fission line becomes fixed beyond altering only late
in division, and can be shifted in its location by earlier influences.

4. Incitement to division

No one yet knows what causes a cell to divide and Stentor is no
exception. In all the experiments on stentors by myself and others
no operation has been established as promptly and invariably
leading to cell division. Yet the search for the inciting cause is so
important that it is appropriate to discuss the few eflForts that have
been made in this direction with stentors. Generally, stentors
attain a certain maximum size before dividing, i.e., dividers are

82 THE BIOLOGY OF STENTOR

found only among the largest animals. All we can assert, however,
is that whatever precipitates division is usually correlated with
size increase, for many circumstances demonstrate that size alone
is not the determining factor. Stentors of smaller than maximum
volume can divide, and yet when many stentors are grafted together
the combined ** cell ", of extraordinary mass, generally shows no
tendency whatever toward fission.

The idea that division is caused by a progressive deviation from
some normal ratio of macronuclear volume to cytoplasmic volume
(Popoff, 1909; Causin, 1931) does not seem to be confirmed in
Stentor. It is sufficient to say that in a variety of experiments in
which nucleus or cytoplasm is at once added or subtracted there is
no clear evidence of division occurring promptly as a consequence.
Nor does micronuclear mitosis trigger division (Weisz, 1951b)
because (a) mitosis occurs near the end of fission (Johnson),
(b) reorganizers and regenerators also show mitoses though not
dividing, (c) emicronucleate stentors divide (Schwartz), (d) and
many other amicronucleate ciliates reproduce normally.

In certain unpublished experiments I have found that when the
membrane liar band or the mouthparts alone are removed from
rather large specimens the stentors almost always promptly
divided and the resulting anterior daughters with the abbreviated
feeding organelles then regenerated a new set (see Fig. 39B). The
same operation performed on the smallest stentors in the culture,
however, yielded no divisions at all. Hence the combination of size
with this specific operation seems to have done the trick, but further
study is needed.

Weisz (1956) conceived that the problem of division in Stentor
could be approached by determining the effect on smallest, post-
fissional coeruleus of large pre-divisional and dividing cells grafted
to them. First he fused largest animals which were soon to divide
but which had not yet produced a division primordium with
smallest cells or products of recent division which therefore would
not be expected to divide until they doubled their size. Division
of the larger partner was then greatly delayed but this could be
attributed to injuries of operation since ungrafted controls also
postponed fission if they were sliced into. After this delay a
division primordium finally appeared in the larger partner,
followed by an induced primordium in the smaller, promptly if

GROWTH AND DIVISION 83

the connection between the two components was quite intimate;
and the complex then divided as a unit, often separating into two
proters and a doublet opisthe.

When the larger partner had already entered division and
carried an early division primordium, its division was still delayed
and the anlage was resorbed after grafting. Again, this response
could be attributed to the operation, because control animals also
resorbed the primordium with cutting injuries and did not re-
commence division until 6 hours later. In the graft combination,
primordia then appeared in both large and small components
which divided simultaneously about 6 to 7 hours later.

If the dividing partner was in mid-stage division with a well
formed membranellar band, grafting then resulted only in arrest
of the anlage and not in its resorption. An induced primordium
then appeared in the small component which divided along with
the larger. But if the divider had already advanced to the stage at
which mouthparts were beginning to form at the posterior end of
the primordium, then no secondary anlage was induced and fission
was largely confined to the side of the original divider. Yet there
still could be some effect on the smaller component. If the macro-
nuclear nodes of the divider had not yet clumped, then those of the
small partner coalesced also. Weisz further stated that a division
furrow sometimes extended around the smaller component, but I
suspect from what has already been said concerning blockage of
the division line that in these cases the furrow passed above or
below the smaller graft.

These results were interpreted by Weisz as demonstrating that
pre-divisional animals or stentors in early stages of fission can
induce fission in graft partners which otherwise would not have
divided, and that this induction is produced by some influence
emanating from the dividing cell and passing to its partner. I am
obliged to say, however, that these demonstrations are in need of
further clarification and control before such conclusions can be
asserted with certainty. If dividers. are grafted to regenerators
one might expect that division would be the more easily induced
because the non-dividing partner already carries a primordium;
instead, I found that fission occurred only on the side of the
divider and did not include the partner, which merely regenerated.
Furthermore, combinations like Weisz's did not always yield the

84 THE BIOLOGY OF STENTOR

same result and frequently when dividers were grafted to non-
dividing animals the complex then simply reorganized doubly,
the single division primordium first resorbing, to be followed by
two anlagen which served only to replace the original feeding
organelles and no furrow formation occurred (see Figs. 38D and c).
In only one case did I obtain what appeared like a prompt and
indubitable induction : a stage-2 divider was grafted to a small non-
divider, the division primordium was not resorbed and another
was induced in the smaller component, whereupon simultaneous
division occurred; and yet the same result was obtained in one
case when a stage- 1 regenerator was used in place of the divider,
though fission then proceeded more slowly. Further studies,
however, may firmly establish a phenomenon of induced division,
and if so, this would afford great potentialities for causal analysis
of division in Stentor.

5. Persistence of division

Clues to the nature of the fission process may be sought in its
persistence in spite of often drastic operations. Long ago, Balbiani
(1891c) reported a case in which a longitudinal half of a stage-6
divider completed fission without either a nucleus or the division
primordium. In other instances, whether of aboral or adoral
halves, the division products did not separate though it was clear
that the cortical striping had been divided into two systems
because double cell shapes resulted. This was also the experience
of Stevens (1903) who obtained division without separation in
aboral halves and even in one enucleated oral half, which correlated
with her studies of the year preceding showing that enucleated
halves of sea urchin eggs are still capable of division. Much later
Schwartz (1935) described one instance of complete division into
two daughter cells after removal of the macronucleus from what
was, to judge from his drawings, a stage-6 divider. This has been
confirmed by de Terra (1959) and myself. Yagiu (1951) found the
same in Condylostoma, and Suzuki (1957) in Blepharisma. These
cases show that there is some ''momentum" in the processes of
division, or that after the primordium is well formed the final
shifts in the disposition of the anlage as well as the cutting of the
cortical striping into two systems and even their total separation
can be effected.

GROWTH AND DIVISION 85

-In de Terra's study on coeruleus it was demonstrated that during
fission the uptake of radiophosphorus (P^") dropped to one-
twentieth its rapid predivisional rate and was in fact the same as
that of enucleated dividers from which the compacted macro-
nucleus had been removed. This indicates that the large nucleus
is not very active biochemically at the time of fission and thus
helps to explain why division can continue to completion in its
absence.

Extension of studies on persisting division called for operations
on still earlier stages of dividers. In recent tests yet unpublished I
found that division in Stentor coeruleus can go ahead after some
rather drastic operations and often when the division process is
by no means nearing completion. Dividers in stages 2 to 5 were
cut in two transversely but the two halves allowed to heal in place.
The primordium also rejoined its parts and division could be
consummated in a perfectly normal manner. Stage 4 dividers were
cut longitudinally and the oral half rotated 180° on the other in
heteropolar orientation; furrow formation still occurred and
division was nearly complete although the division products did
not separate (Fig. 19A). The mouthparts of dividers in stages 2, 3
and 4 were excised and the cells either split down the back and
opened out flat or cut and spread out in three parts like a clover
leaf, and still division often followed, yielding proters which
regenerated the missing mouthparts later. Isolated longitudinal
halves of dividers cut before there is any visible sign of a furrow
(stage 5) could cut the striping and form furrows. Usually the
fragment did not actually separate into two pieces, as others have
also found ; yet in two cases at the preceding stage 4, longitudinal
oral halves did complete division. These tests clearly indicate that
the division process is not so delicate and precisely adjusted that
disturbances cause its undoing.

The oral primordium was excised from dividers as early as
stage 3 and the animals continued division. Because the headless
opisthes later formed oral structures through a regeneration
primordium we can infer that the macronucleus was also divided.
Well-formed primordia at stage 4 were also circumscribed and
rotated 180° in situ and subsequent fission was still successful.
These experiments, as well as certain of the aforementioned,
show that although the fission line begins at the anterior end of the

86

THE BIOLOGY OF STENTOR

primordium this site is by no means the necessary initiator of
furrow formation. The same is also indicated by a case in which
the primordium of a stage-3 divider was shifted to the posterior
end; the cell still divided in two (Fig. 19B). Nevertheless if for
any reason the dividing stentor resorbs the oral primordium itself,
division is then not continued (Weisz, 1956; Tartar, 1958c).

A

B

Fig. 19. Persistence of division in spite of major disturbances;
separation of a heteropolar pair.

GROWTH AND DIVISION 87

Division, but without separation of the products, can even
occur though both nucleus and primordium are excised before
there is any visible beginnings of the fission line (Fig. 19c).

Total removal of the macronucleus as early as stage 4 does not
always preclude division even though complete development of
the primordium cannot occur without nuclear support. However
necessary for the original stimulus to divide, the macronucleus, as
Yagiu (1951) also found in Condylostoma, does not seem to be the
immediate trigger for the actual process of fission. (Micronuclei
are of course not concerned, because Schwartz (1935) showed that
growth and division occur in stentors from which these nuclei
have been removed.) At stage 5, when the division line is presum-
ably being determined, the cell can be cut through with a glass
needle following exactly the path which the furrow will take, or in
stage 6, when the furrow is visible, it can be slashed through with
the needle around its entire course, and still division is completed.
Together, these operations indicate that the division furrow is not
a structural elaboration ; for if it were, the nucleus would presum-
ably be indispensable for the synthesis of new parts, and because
any specialized ** organelle of division " would be destroyed by the

A. a : Left half of stage-6 divider rotated in polarity reverse
of that of right half, with no fission line yet begun, b, h' :
Furrows develop along predetermined course in both halves, as
seen in ventral and dorsal views, but not joining, c: Hence
daughter cells held together by cytoplasmic connections, only

later pulling apart.

B. S. coeruleus continues division and anlage develops com-
pletely though stage-3 primordium shifted to the posterior end.
a: The operation; b: resulting arrangement; c: fission

consummated.

C. Continued fission of stage-6 divider after both nucleus
and primordium removed before any sign of cutting of stripes
to form the fission line, a: The operation, excision of the
anlage patch and clumped macronucleus. b: Division nearly
completed but products held togethe-r by a cytoplasmic bridge
probably due to cortical pattern disturbances from cutting.
c: Substance of opisthe now largely absorbed into the proter.

D. Tail-to-tail heteropolar grafts easily pull apart; but even
head-to-head pairs like this, in which the heteropolar striping
does not join, also can separate or "divide" neatly by a course

which was not observed.

88 THE BIOLOGY OF STENTOR

Operations just described. Instead, it appears again that furrow
formation involves only the severing of longitudinal structural
elements of the ectoplasm, v^hich was merely hastened or abetted
by the needle. Constriction is probably due to the action of contrac-
tile elements largely already formed.

Cutting of stripes alone does not result in fission however. The
ectoplasm of non- dividing cells can be completely cut through
around the equator and division never follows, the cortical struc-
tures merely healing together, often without leaving any indication
of the operation. Conversely, in abnormal situations one some-
times finds stentors in which complete furrowing seems to have
occurred but division does not follow. These cases indicate the
importance of constriction in division, for it is almost certain that
had constriction occurred, this type of specimen would have been
divided in two.

Division is not the only means by which two separate stentor
individualities can become separated. Fusion masses of two or
more stentors show a strong tendency for the components to pull
apart. This is especially the case in the heteropolar pairs, whether
joined by the heads (Fig. 19D) or the tails; for when stripes of
opposite polarity meet it is quite evident that they do not join and
at this locus of discontinuity a separation may occur. Weisz (1951a)
remarked that separations of tail-to-tail telobiotics are ''strikingly
reminiscent of vegetative division"; yet they are different in that
pulUng apart requires a long time for completion, as if sharply
localized constriction could not occur at all.

By growing coeruleus in what were probably rather putrid
cultures of beef extract, Stolte (1922) produced animals with highly
vacuolated endoplasm which showed many anomalies of division.
Animals could divide into three parts instead of two, producing
posterior daughter cells without primordia or nuclei, or showed
very unequal divisions yielding abnormally small opisthes. A case
of partial vertical fission was even described which resulted in an
animal with two holdfasts. Since the conditions were obviously
abnormal, the interpretations offered seem dubious; yet these
observations suggest, as Weisz (1956) proposed, that the endo-
plasm is important in division, vacuolization greatly disturbing
whatever its function may be.

Postponed fission, in which division is much delayed but eventu-

GROWTH AND DIVISION 89

ally realized, may be regarded as another manifestation of the
persistence of division. I have observed (see Fig. 64A) repeatedly
that dividing stentors do eventually undergo fission even though
the original process may be cancelled by causing the primordium
to be resorbed, or by intervening reorganization or regeneration
even with loss of cytoplasm (Tartar, 1958b). The response is as if,
once stimulated to divide, a stentor is bound to do so eventually,
in spite of intervening catastrophies. This recalls the interesting
hypothesis of Swann (1954), originating from studies of egg
cleavage. He conceived that, as a separate mechanism, the cell
builds up a reservoir of something which is essential to or stimula-
tive of division alone, so that this store is depleted only by division.
Adapting this idea to Stentor, greatly postponed division could be
the consequence of presence and persistence of a reservoir of this
factor which is not exhausted by other intervening acts of
morphogenesis.

Reproduction by division in a form like Stentor normally in-
volves first the transformation of one individuality into two,
followed by the physical separation of the two individualities pro-
duced. The integrative tendency of the organism toward unitary
wholeness, which theoretical biologists have generally emphasized,
is therefore suspended or violated during reproduction in ciliates.
From this observation, together with numerous phenomena in the
regeneration of multicellular forms, we are led to suppose that the
organism is in an important aspect beyond individuality, though
tending to individuate as one or more than one, depending on
circumstances. Of this we shall have more to say in the con-
cluding chapter. For the present it is sufficient to say that
"wholeness" is no metaphysical principle which organisms are
compelled to maintain and is in fact transgressed every time a
stentor divides.

Stentor also bears on another issue which in the past at least
has been prominent in biology : namely, whether a fully differenti-
ated cell is capable of division. If not, then regeneration of
metazoa would imply either dedifferentiation of cells or the
presence of "embryonic cells" still capable of rapid fission and
pluripotential differentiation (see Bronsted, 1955). Basing his
argument largely upon the fact that apostomatous ciliates undergo

90 THE BIOLOGY OF ST EN TOR

detorsion of the lateral striping (kinetics) preceding fission, and
thus apparently return to a more primitive state of differentiation,
Lwoff (1950) maintained that division does require a more
embryonic state and that '* The ciliates have solved the problem of
perpetuating complex adult structure by cyclical dedifferen-
tiation ". Although this statement may apply to apostomes and to
forms like Euplotes which form new^ feeding organelles for both
proter and opisthe, it is not apropos of Stentor in which pre-
existing cytoplasmic differentiations are obviously passed on to
the daughter cells, and therefore cannot be generalized. I think
that the important point is that most ciliates do not dedifferentiate
before or during fission until they are quite formless and then
divide. For it is apparent enough in the example of Stentor that
maintenance of the complex structures of lateral striping which
continue their ciliary and contractile functions throughout this
process and are simply cut in two is not incompatible with
division, and that therefore cell division does not necessarily require
that a cell regress below a high state of differentiation.

Persistence of cortical differentiations in dividing stentors also
precludes surface tension changes as a means of cell division in
these forms and their allies, and the separation of asters in a
mitotic figure is also ruled out. Studies on cleaving eggs in which
these and other forces seem to be operating hence cannot be taken
as characteristic of all cell divisions. Perhaps we can learn from
Stentor of other factors equally important or effective, especially
with regard to multiplication of tissue cells. It has been amply
demonstrated above that stentors can be operated upon during
fission in many ways which afford a promising approach to prob-
lems of cell division, as it has also been shown that the questions
of differential growth even within the confines of a single cell are
in this organism amenable to experimental analysis.

CHAPTER VI

REORGANIZATION

At seemingly irregular intervals stentors form an oral primordium
which serves neither for division nor regeneration but merely
replaces oral structures already present. This has been called
reorganization or physiological regeneration. As these terms imply,
it has been supposed that this act is a necessary renewal of worn-out
organelles, but this is by no means certain and the real meaning of
the act is still to be ascertained. The problem of this peculiar
redifferentiation of the cell is not unique to Stentor. Reorganization
also occurs in the related Condylostoma (Tartar, 1957b) and
Bursaria (Lund, 1917), as well as in many other ciliates.

The course of reorganization in coeruleus was well described
by its discoverer, Balbiani (1891a). To this description Schwartz
(1935) added many significant details in the most complete and
best illustrated account in the literature, and additional points
were contributed by others, as will be noted.

1. The course of reorganization

The first indication of reorganization is the beginning of an oral
primordium. Approximately at mid-body level below the mouth-
parts a splitting of pigmented stripes occurs as in division,
but multiplication of clear and granular stripes is not so extensive
(Fig. 20). The good reason for this is that a complete new adoral
field is not to be formed but only an addition to the old one. As
the primordium lengthens and develops, its anterior end meets
the old membranellar band at the point where the latter begins
to form the margin of the oral pouch. At stage 5 the original
mouthparts begin to dediflferentiate. Macronuclear beads start to
fuse and form a compact mass by stage 6, beginning renodulation
at stage 7. According to Weisz (1949a), some of the terminal nodes
may not fuse and if this occurs they break free. He also reported
that endoplasmic streaming carries the clumped nucleus as well

91

92

as

THE BIOLOGY OF STENTOR

any free nodes forward to a position directly under the
developing primordium and he claimed that this is a precise and
invariant event in reorganization, though neither Schwartz nor I
confirm this. Essentially, the macronucleus clumps and renodulates
without dividing.

b ^ c ^ d

Fig. 20. Course of reorganization (S. coeruleus).

a: Morphologically complete animal shown with stage-3
primordium. b: Stage-4 anlage intirriately joining with original
membranellar band, c: During stage-6 of primordium develop-
ment the old band between joining point and mouthparts,
together with those parts, is resorbed. d: New mouthparts
and addition to membranellar band moving into place and
carrying new fine stripes into the frontal field which now shows
a double pattern, 2 swirls.

Since the granular carbohydrate reserves in the posterior end
of the cell undergo remarkable shifts in division, their behavior
during reorganization should be noted. Weisz reported that the
granules remained in place without change, but I observed that
generally they diffuse forward under the ectoplasm though not
separating into two groups (Tartar, 1959a). The chief resemblance
between reorganizers and dividers is that in both cases a primor-
dium is formed although feeding organelles are already present.
In the tendency of the carbohydrate reserves to migrate there is
also a slight similarity, but points of distinct difference are that
no second contractile vacuole is formed during reorganization
(Johnson, 1893) and of course no fission line.

The anterior end of the anlage now fuses with the original
membranellar band, often causing a slight deflection where it
joins ; and the section of the old band between this juncture and
the mouthparts then dedifferentiates and is resorbed along with
the last traces of the gullet. The original oral pouch apparently is

REORGANIZATION 93

not resorbed but simply rises and flattens out to the level of the
adoral field with whose stripes its own are continuous. As the
primordium migrates forward and carries with it some ecto-
plasmic striping to the right, new and old frontal-field stripes are
brought together, but though homopolar they do not join. Instead,
the anterior ends of the new stripes are pulled over to the point
where the old mouthparts dissolved and the resulting frontal field
therefore shows two swirls of striping (Fig. 2i).This doubleness
is an enduring character which identifies stentors that have under-
gone reorganization, a sign which is often useful in following the
performance of experimental animals. Sometimes one finds
stentors with three disjunctive systems of striping in the frontal
field, indicating that these animals have twice reorganized, though
they may be proters from an ensuing division.

Fig. 21 . Anterior end view of a coeruleus which had reorganized
twice, showing muhiple pattern of frontal field.

After the reorganization primordium is nearing completion,
a secondary stripe multiplication occurs just below the newly-
forming oral region quite as in the opjsthe of dividers, as pointed
out by Schwartz. This stripe increase will form a new ramifying
zone and completes the reorganization process. It now remains to
report what has been done toward analyzing the sequential events
in reorganization and above all to inquire into its possible signifi-
cance for the life of the organism.

94

THE BIOLOGY OF STENTOR

2. Analysis of the reorganization process

There have been few experiments on reorganizing stentors and
this area of study is prickly with paradoxes, but I have some
unpublished data which is suggestive. These relate to the central
question whether resorption of mouthparts is uniquely character-
istic of animals in process of reorgnization.

When the aboral half of the membranellar band is removed, an
oral primordium is formed and as it moves into place the original
mouthparts remaining are resorbed (Fig. 22A). Is this regeneration
or reorganization? Such cases are like regeneration in compensating

A

a ^ b ^ .' c ^ d

Fig. 22. Observations relating to reorganization.

A. When one half or more of the membranellar band is excised

delayed formation of a regeneration primordium follows and

original mouthparts are resorbed as in reorganization.

Regenerant then gains a peristome of normal length.

REORGANIZATION 95

for an excised part, but they resemble reorganization because the
mouthparts are replaced. Both Schwartz (1935) and Weisz (1951b,
1954) called this reorganization, and this may be permissible; for
if, in such cases, the well-formed primordium at stage 4 is removed
by a minimal excision or caused to be resorbed, the mouthparts are
still completely resorbed, the animals then regenerating later
(unpublished). The same occurs with true reorganizers, bearing
complete feeding organelles (b). It would appear that in reorgani-
zers and regenerators with mouthparts, these structures are pre-
determined to be resorbed. In contrast, an adaptive resorption of

B. Primordium of reorganizer excised at stage 4, before visible
resorption of mouthparts has begun. Oral structures neverthe-
less subsequently resorbed though there are no new ones to take

their place. Regeneration follows.

C. a: Stage-5 primordium grafted to a non-differentiating
host, b: Developing anlage breaks into the oral zone, joining
with the original membranellar band, a section of which is
resorbed to permit incorporation of the new mouthparts. This
imitates band resorption in reorganization, but the host mouth-
parts are not resorbed and a doublet stentor results (c).

D. Stage-5 reorganizer with head circumscribed and rotated
180°. Original mouthparts are resorbed, although now on side
opposite to primordium. Anlage breaks into the peristome, free

ends of same join on far side and a normal stentor results.

E. Specimen reorganizing after anterior was rotated on the
posterior half. The primordium formed far down on the side of
the cell yet the old mouthparts and adjacent section of mem-
branellar band were resorbed as if the anlage were normally
located. Later the new organelles moved forward, joined with
the original membranellar band and produced a normal stentor.

F. Stage-4 reorganizer grafted to non-differentiating stentor.
a: By stage 6, the reorganizer (right) had induced a transient
reorganization primordium in its partner. Both original sets of
mouthparts were resorbed, as also the out-of-phase accessory
anlage. c: Partner left without oral structures now regenerating,
with induced re-reorganization in the other animal. Oral resorp-
tion may therefore extend to other mouthparts present in the

graft systein.

G. a: Parabiotic graft of two stage-2 dividers, b: No
division occurred, and the anlagen moved forward instead but
the original mouthparts were not resorbed as in reorganization.
c: Specimen then produced two reorganization primordia and
all four pre-existing sets of mouthparts were resorbed as the

usual doublet was formed {d).

96 THE BIOLOGY OF STENTOR

sections of the membranellar band is shown by the observation
that late regeneration primordia grafted into the backs of non-
differentiating cells can and frequently do break into the oral ring
through a localized resorption of the membranellar band (c). This
interpolation also occurs in reorganizing stentors in which the
head has been rotated 180° in place, but now the mouthparts are
also resorbed though on the far side of the cell (d).

There are other evidences that oral resorption is a separate and
predetermined part of the reorganization process. In one interesting
case the stentor was transected and the anterior half rotated 180°
on the posterior. This specimen then reorganized. Because of the
disarrangement of the striping, the anlage remained for a long
time in a diagonal position across the center of the cell yet the
mouthparts and adjacent membranellar band were resorbed long
before the primordium moved forward (e). It is also pertinent
that mid-stage reorganizers can induce transient primordium
formation in non-differentiating cells to which they are grafted;
reorganization goes to completion on the reorganizer side and the
mouthparts of the other component are also resorbed although
there is not a new set to take their place (f).

On the contrary, it appears that in the state of division or
regeneration there is little predisposition for the feeding organelles
to be resorbed. When two stage-2 dividers were grafted together
in homopolar parabiosis division did not continue, the two
primordia moved forward but the original mouthparts remained
intact (g). True reorganization then followed, in which all four of
the existent oral parts were resorbed.

These experiments suggest that in reorganization the mouth-
parts are somehow invisibly dissociated, or cut off morphologically,
and therefore usually predestined to dedifferentiation. This
interpretation seems to be supported by the fact that if re-
generation is induced by excising the head and a new head is then
grafted back into place later, then, if the primordium continues
developing, reorganization ensues and the old mouthparts are
resorbed (see Fig. 37B). Also if the heads of non-differentiating
stentors are circumscribed and rotated 180° in place, reorganiza-
tion follows in the majority of cases. So in both types of experi-
ment it would seem that isolating the whole set of feeding organ-
elles has the same effect as the hypothetical disjunction of the

REORGANIZATION

97

oral region only. It should also be mentioned that although injury
to the cell usually causes resorption of early primordia in dividers,
this occurs very rarely in reorganizers. In every case w^hen early,
stage-2 reorganizers w^ere split into a clover-leaf shape the primor-
dium was not resorbed and the animals completed reorganization
after the parts of the cell fused together. Such persistence of the
anlage almost always did not occur in dividers unless the mouth-
parts were also excised at the time the cell was split. Therefore the
reorganizers behaved as if their mouthparts were not present, i.e.,
as if these parts were effectively, if cryptically, isolated somehow
from the rest of the cell. This hypothetical, morphological dis-
junction of parts would be of a subtle nature, however — possibly
at the level of fine fiber structures — for when I tried to duplicate
it by sectioning the membranellar band with a needle at the point
where it meets the oral pouch, the band merely mended together
and there was no reorganization. Obvious isolation of mouthparts

Fig. 23. "Autotomy" of mouthparts.
a: Head of stentor rotated 180° with mouth now opposite the
primordium site, h: Several days later the mouthparts — such
as are resorbed in reorganization :^— separate from the mem-
branellar band and move into the frontal field, with ends of the
band rejoining behind them. Then a "reorganization" primor-
dium appears, c: Old mouthparts cut into the frontal field are
then resorbed, together with part of the membranellar band
which permits integration of the anlage to produce a stentor of
normal orientation.

98 THE BIOLOGY OF STENTOR

by cutting and shifting does act as a stimulus to primordium
formation although no portion of the feeding organelles is excised.
Even when heads were rotated 180" in place and primordium
formation did not occur at once, a remarkable readjustment of the
cell pattern took place (Tartar, 1959b) as shown in Fig. 23. The
mouthparts, including the oral pouch and its membranellar margin
were autonomously severed and thrust in towards the center of
the frontal field while the membranellar band closed together. In
the primodium site there appeared a "reorganization" anlage
which eventually broke into the oral ring and provided a new set
of mouthparts, now in the correct location. In this performance
we see that the very structures which are resorbed in the normal
course of reorganization can in fact be " autotomized ".

3. Stimulus to reorganization and the significance of this
process

The seemingly adventitious occurrence of reorganization in
stentors, which appear to be the same as their non-reorganizing
fellows, gives the impression of a quite unnecessary act which
leaves the animal just as it was before. Hence the enigmatic
character of reorganization. Yet we naturally assume that ciliates
would not go through this complicated process without good
reason, and several hypotheses have been advanced in the case
of Stentor, though there is none w'hich has not left its residue of
paradoxes.

(a) To REPLACE DEFECTIVE MOUTHPARTS?

On discovering reorganization, Balbiani (1891a) suggested that
the process is for replacement of worn-out ingestive organelles.
The act would therefore be essentially the same as regeneration
which is evoked by removal of these parts. Having well observed
that the entire membranellar band is not replaced, Balbiani
(1891a) assumed that the mouthparts are the most "used" and
therefore the most subject to deterioration; but there was also a
hint in his initial report that aging alone might result in these parts
eventually becoming defective. Among modern students of Stentor,
Weisz (1954) accepted this interpretation of the raison d'etre of
reorganization and emphasized (1951b) that injury or defect
might be either structural or functional, justifying the term physio-

REORGANIZATION 99

logical regeneration. In either case there should be less successful
feeding and one would expect that reorganizers would appear
under-fed, transparent, and with few if any food vacuoles. But as
I recently pointed out (Tartar, 1958c) reorganizers are quite as
replete as their fellows. In fact, Weisz (1949a, 1954) almost
implied this himself in explaining that pigment granules and
carbohydrate reserves are not decreased and utilized in re-
organization as they are in regeneration because reorganizers can
continue feeding.

There are other strong objections to the defect hypothesis.
Johnson (1893) independently discovered reorganization in
Stentor and he seems to have followed Balbiani's interpretation,
yet he described a case of two successive reorganizations in
coeruleus which cannot be explained on the improbable assumption
that the mouthparts just formed by the first primordium had
become defective through use or aging. Then Morgan (1901a)
noted that in most instances the old feeding organelles of re-
organizers are still active and appear entirely normal, though this,
he said, w^as not always the case. In my own studies, I talUed 36
cases in which newly-formed feeding organelles, wholly normal
in appearance, were promptly subjected to reorganizational
replacement, quite apart from the fact that in graft stentor com-
plexes repeated reorganization is the rule (Tartar, 1954). These
cases cannot be explained on the defect hypothesis unless one
supposes, against all appearances to the contrary, that the pre-
ceding differentiation was inadequate.

If the mouthparts wear out, this should occur sooner in proters
which retain the old ones; yet Hetherington (1932b) did not find
reorganization in the continued isolation of proters for five
generations, i.e., of feeding organelles five generations old. Finally,
and most conclusively, one can specifically injure the mouthparts
by thrusting a needle down the gullet and cutting laterally, where-
upon the injury is simply repaired and no reorganization follows
(Tartar, 1957c) (see Fig. 33B).

(b) Response to change in the medium?

Hetherington (1932b) was strongly of the opinion that re-
organization does not occur in stentors under constant conditions
of culture, and that reorganization if it occurs at all, is brought

\Uj I I IBRARY ]:>!

-^'.V '^^ss-

lOO THE BIOLOGY OF STENTOR

about by changes in the medium, not necessarily unfavorable,
such as transfer from old to fresh culture fluid. It may be that
reorganization is a response to disproportionality of cell parts (see
below) and that under the most uniform conditions growth in all
parts proceeds so harmoniously that no disproportion arises. But
Hetherington's argument is vitiated by several contradictions.
First, he says that no physiological regeneration occurs in stentors.
Then he admits that he did find *'reorganizers" in unchanged
stock cultures of coeruleus. To explain this, he asserted that such
animals were regenerating from cryptic injuries; and he stated
that renewal of mouthparts is not the same as reorganization though
he did not offer a different definition. He said that his animals
were invaded by bacilli from which they were freed by repeated
transfers into new medium, during which reorganizations were
frequent; but then it might be held that the infection was really
the cause of reorganization. Hetherington's contribution, then,
was to direct attention to changes in the culture medium as a
possible cause of reorganization; and to raise, if not resolve, the
question whether replacement of worn out or injured mouthparts
should not properly be called regeneration, as reasonably as when
excisions are the inducement.

That '* depression " conditions in the culture may be the cause
of reorganization, though not the only or principal one, was also
suggested by Balbiani (1891a), and Weisz (1949a) assumed the
same; but Causin (1931) found that unfavorable conditions never
seemed to cause reorganization. Merely adding new water to the
cultures was said to bring about reorganization (Weisz, 1949a).
Yet it is difficult to see how such a mild stimulation as change in
the medium could elicit reorganization when the most severe
cutting injuries involved in many stentor experiments do not.
I therefore also question Causin's (1931) remark that if the tailpole
is cut off a stentor the cell then undergoes a partial reorganization
as if in response to a mild injury. He did not describe what
happened beyond saying that the nucleus did not clump together
completely.

Stentors in small drops under cover slips are incited to divide as
well as to reorganize, according to Balbiani (1891a), but this cer-
tainly does not occur with regularity in depression slides. I
reported (Tartar, 1958c) that a dilute solution of methyl cellulose

REORGANIZATION lOI

brought about extensive reorganizations in a stentor sample, but
this procedure was not easily reproducible.

(c) Need for nuclear reorganization ?

In one of his cytochemical studies, Weisz (1950b) reported that
in the chain macronucleus of coeruleus a gradient in affinity for
methyl green seems to develop in anticipating reorganizers and
pre-fissional animals, the posterior nodes staining less intensely.
In both cases, after clumping and renodulation the nuclear beads
stained uniformly. In this there is the implication that re-
organization might be to reinstate uniformity of composition of
the nucleus in animals vv^hich for some reason are not yet able to
divide. But Weisz did not say so explicitly, perhaps because he
found reorganizers in "all cycle stages", i.e., at any time during
the interfissional period.

(d) For growth of the adoral band ?

It will be recalled from the account of the reorganization process
that important new additions to the membranellar band and the
frontal field take place, while only the gullet and the border of the
oral pouch are obviously resorbed. Therefore a considerable
enlargement of the head should result. Schwartz (1935) carefully
counted the membranelles and found that approximately twice as
many are added as are resorbed. This suggested to him that re-
organization may be a periodic growth process serving in part for
the increase in the length of the membranellar band as well as
perhaps the enlargement of the mouthparts. Favoring this concep-
tion is the finding that if for any reason the primordium produces
too small a head, with a short membranellar band and limited
frontal field, reorganization soon occurs with resulting enlargement
of these parts (Tartar, 1958b).

In at least one case, however, I found that when an extra head
was grafted to a stentor and this fused with the original to form
a supernormal number of membranelles, reorganization never-
theless occurred. There are other arguments against the growth
hypothesis. When there are repeated reorganizations the membran-
ellar band does not become of exaggerated length. We are obliged
to assume that the immediate increase results in a compensatory
resorption of membranelles in some part of the band. Only in

102 THE BIOLOGY OF STENTOR

grafted doublet and triplet stentors does the frontal field and
membranellar band become much enlarged over the normal, as if
the excessive girth of these complexes could support a larger
structure. It seems clear that reorganization is not an essential
growth process, for otherwise it should occur with great regularity.
Considering only the opisthe, a daughter cell starts with a set of
feeding organelles which appears to be proportionate to the cell
volume and therefore about half the size of those of the parent
cell. If the membranellar band can increase only by adding new
membranelles through primordium formation, reorganization
should occur always before the next division and probably at a
certain time when disproportion sets up a tension. But re-
organization does not occur with regularity and, ever since
Balbiani, it has been observed that stentors of any size can be
found reorganizing. This includes even very tiny individuals,
which I can vouch for, as well as animals in which the oral struc-
tures do not appear in any way disproportionate to the cell size.

(e) Need for adjustment of nuclear dimensions ?

Although Balbiani did not find an increase in the number of
macronuclear nodes following reorganization, Johnson reported
that this was usually the case. Of i8 reorganizers, he found that 14
increased the number of macronuclear nodes, 2 remained
unchanged in this respect, and 2 even decreased the number of
nodes. The increase was sometimes to twice the original number
of nuclear beads, but the new ones seemed to be smaller. Therefore
he suggested that reorganization is for the purpose of increasing
the active surface but not the size of the macronucleus, or that
the surface-volume relation is adjusted as required, even in the
direction of decrease. I may mention here that I have also observed
cases of decrease in nodal number following reorganization of
regenerated stentor fragments that contained too much nuclear
material.

A striking demonstration of this correlation between nuclear
size and reorganization was given by Schwartz when he showed
that reorganization could be induced at will by cutting out most
of the nuclear beads. Weisz (1951a) and I have confirmed this.
When only one or two nodes were left, they at first enlarged the
surface by becoming spindle shaped, just as Prowazek (1904)

REORGANIZATION IO3

had previously observed. This earlier investigator also noted that
without primordium formation and nuclear clumping there are at
most only one or two nodes which may divide. Schwartz com-
pletely cinched the point by showing that regeneration of the
depleted macronucleus occurs only after primordium formation,
when there is also a mitotic division of the micronuclei leading to
their increase in number. Moreover, such an increase in the nuclear
complement seemed definitely to be called for, because he found
that the "entire metabolism" of stentors with reduced nucleus is
upset. After great reduction of the macronuclear volume there
may follow a series of reorganizations, according to Schwartz, with
the implication that in each only a limited increase in the macro-
nucleus is possible. This I have also observed.

That mitotic multiplication of the micronuclei occurs during
reorganization when there is the increase in the number of macro-
nuclear nodes, Schwartz deduced as follows. If one assumes that
in division there is but a single mitosis of each micronucleus so
that the total number is only doubled, then the demonstrated
presence of about the usual number of these nuclei in both
daughter cells of a stentor which had previously been induced to
reorganize by removing most of the macronucleus (and therefore
most of the adhering micronuclei) implies that the micronuclei as
well as the macronuclear nodes must have increased during the
reorganization. Furthermore, it has been demonstrated directly
in Blepharisma that micronuclear mitosis accompanies reorganiza-
tion (Suzuki, 1957).

It may well be that anything which may lead to a macronucleus
becoming too small for the cell volume results in reorganization.
The essence of reorganization would then not lie in the fact that
the mouthparts are replaced, for this also occurs after excision of
substantial lengths of the membranellar band. Nor would it be a
response to injured or worn-out mouthparts, since this is essentially
regeneration and the evidence for this condition's being the neces-
sary cause of reorganization is overwhelmingly in the negative.
Changing the medium does not in my experience act as a stimulus
to epidemics of reorganization, and reorganizers are found in
cultures that have not been altered. Hence it would seem proper to
regard reorganization as a wholly spontaneous and intrinsic
response to certain disproportionalities or disarrangements of parts

104 '^"^ BIOLOGY OF STENTOR

of the cell which is for the purpose of bringing them into a more
normal relationship.

Deserving special emphasis is the point that the macronucleus
seems to be as much dependent on a cytoplasmic primordium
formation for its growth as the primordium is dependent on it.
Hence it could be that when the macronucleus needs adjusting to
the cell volume, the mouthparts are cryptically ** autotomized ", as
suggested above, in order to incite anlage formation without which
the nucleus cannot undergo extensive alterations.

CHAPTER VII

REGENERATION

Stentors have long been the preferred subject for studies on
regeneration in the protozoa because of the large size of common
species, their amenabiHty to cutting operations, and the elaborate
system of cortical differentiations which calls for a substantial
performance in morphogenesis and provides a definite end-point
for experiments. It must have been a dramatic moment v^hen
Nussbaum (1884) extended to the ciliates the earlier experiments
of Greeff, 1867, and Brandt, 1877, on heliozoa in demonstrating
the general *' divisibility of living matter" at the cell level. Of
course cells divide, but now it was shown that man could do the
dividing himself with similar results. A year later Gruber (1885a,
1885b) published his finding that, in contrast to division, stentors
can be cut into three pieces, each of which could produce a new
individuality, and his drawing of the regeneration of a trisected
stentor was reproduced in dozens of textbooks. He proved that
regeneration was in fact complete, for the fragments not only
regained the normal form but could then subsequently grow and
divide. These studies were carried forward by Balbiani in a series
of notable early investigations. Following these pioneers, investiga-
tors have turned repeatedly to Stentor as a form in which regenera-
tion and reconstitution can be studied within the confines of a cell,
the " structural unit of life ".

I. The course of regeneration

(a) Oral regeneration and its requirements

Excision of the head or any appreciable portion of the feeding
organelles leads to oral regeneration. A primordium appears on
the side of the cell and then moves forward to the anterior end as it
develops a new set of ingestive structures. If any part of the
original membranellar band and frontal field remain, they persist
and are integrated into the new head (Stevens, 1903). But if only

105

io6

THE BIOLOGY OF STENTOR

Fig. 24. Stages in oral regeneration following excision of feeding
organelles (S. coeruleus.)

Stage I. Anterior end healed over and primordium appears
as a rift across fine striping near area of widest stripes. (Omitted
stages correspond to those in the development of the oral anlage
in division — see Fig. 14.)

Stage 5. Primordium with expansion at posterior end where
mouthparts will form. Multiplication of fine stripes within the
arc of the anlage which will form the new frontal field.
Macronu clear nodes coalescing.

Stage 6. Invagination of the end of the primordium to form

cytostome and gullet. Nucleus compacted, but usually not as

much as in division. Stripe multiplication below anlage will

form a new fine-line zone and recover approximately the normal

number of lateral stripes.

Stage 7. Gullet and cytostome now well formed and oral
pouch invaginating as primordium moves to anterior end.
Macronucleus renodulating. (After Tartar, 1957c.)

REGENERATION I07

the membranellar band suffers ablation, the old mouthparts are
resorbed as the new ones take their place. Regeneration has been
staged according to visible changes in the primordium (Tartar,
1957c) as in Fig. 24.

Just as regeneration can occur in starving metazoa, so in Stentor
the process imposes no nutritive demand. Weisz (1949a) remarked
that in regeneration of coeruleus there is an extensive loss of pigment
granules which he presumed to be utilized in supporting primor-
dium formation, since this occurred in posterior and middle
fragments but not in anterior pieces which do not have to produce
a new set of feeding organelles. I too have often noticed a fading
in the animals, which seems to be correlated in degree with the
number of times they undergo primordium formation, though
this is not always apparent. Carbohydrate reserve granules may
be utilized in regeneration, if this can be dissociated with their
employment in mere survival; and Weisz (1948b) claimed that
oral regeneration could not occur in the absence of these reserves
or their potential equivalent in the form of food vacuoles, but this
could not be confirmed (Tartar, 1959a). Regeneration or further
development of a regeneration primordium already begun can,
however, be greatly delayed by cold (Morgan, 1901a).

Apart from the necessity for the presence of a segment of the
macronucleus, the character of cutting injuries and ablations im-
poses few limitations on regeneration potentialities. Central-disc
fragments with widely exposed endoplasm folded upon themselves
to cover the wound surfaces and neatly regenerated (Fig. 25A).
Collapsed stentor "skins" from which almost all the endoplasm
has been squeezed out easily regenerated and recovered the
normal plump form (b), quite as in similar tests with Condylostoma
(Tartar, 1941b). When almost all the ectoplasm was sliced off, the
patch remaining greatly stretched to cover the exposed endoplasm
and regeneration was consummated (c). But endoplasmic spheres
completely bereft of ectoplasm never regenerated, though they
remained intact and alive (insofar as they resisted bacterial attack)
for two days (Tartar, 1956c).

These tests effectively dispose of the notion (Prowazek, 1913;
Sokoloff, 1924; Weisz, 1948a) that the ratio of ectoplasm to endo-
plasm (how measured ?) cannot be altered far from an optimum if
regeneration is to be possible, as well as the opinion that wound

io8

THE BIOLOGY OF STENTOR

healing is an important factor in regeneration (Causin, 1931;
Weisz, 1948a). Weisz's (1951b) statement that any portion of the
endoplasm is capable of supporting regeneration is subsumed in
the fact that no significant amount of endoplasm is needed at all.

Fig. 25. Regeneration of coeruleus under severe conditions.

A. Mid-ring fragments regenerate in spite of extensive wound
surface and exposure of endoplasm, because the piece folds to
cover surface with ectoplasm and anlage promptly appears in
short section of original primordium site. A normal stentor can

be formed within a day.

B. Specimens with collapsed ectoplasm after removal of
practically all the endoplasm by vigorous pipetting can regenerate

and fill out the cell shape within a day.

C. In nucleated endoplasmic spheres with almost all the
cortical layer excised the remaining ectoplasm stretches to cover,
with granular bands becoming excessively broad and pale.
Here the reconstitution was abnormal and the primordium,
appearing on the "wrong" side, produced a stentor of reversed
asymmetry. Usually such specimens, with greatly reduced
ectoplasm do not live or regenerate, possibly because even
maximum stretching cannot achieve a cortical continuum with

no "edges". (After Tartar, 1956c.)

REGENERATION

109

A

C

Fig. 26. Experiments concerning holdfast regeneration.

A. Removal of posterior portion of primordium-site sector
is followed by temporary tail formation from cut ends of the
lateral striping; but the anteriorly located extension is later
resorbed in favor of holdfast reconstitution at original posterior

pole. (After Weisz, 1951b.)

B. Temporary tail-pole formation at suture may occur when
anterior is rotated on posterior half. Misaligned stripes do not
rejoin and projection occurs in oral meridian of anterior half,

but is soon resorbed as stripe patterns interpenetrate.

C. a: Sector with stage-3 regeneration primordium grafted
heteropolar into a non-differentiating host, b : Anlage is resorbed
and temporary pedal pole formation occurs from posterior end
of graft, c: Extra tail resorbed, graft patch diminished, and
specimen reorganizing doubly, d: Short, anterior primordium
contributes only a sector {x) of membranellar band, forming no

mouthparts.

no

THE BIOLOGY OF STENTOR

(b) Regeneration of the holdfast

This occurs much more readily and quickly than oral regenera-
tion. It can even take place in the absence of the nucleus (Tartar,
1956c). Within 2 hours after excision a new attachment organelle
is formed (Morgan, 1901a; Weisz, 1951b). By removing holdfasts
and posterior portions of the left boundary stripe of the ramifying
zone, Weisz showed that a new tail was then produced at the
posterior terminus of the stripes remaining, even if this led to a
holdfast appearing forward and projecting laterally (Fig. 26a).
(Actually a substantial portion of the ramifying zone must have
been removed.) The regenerated holdfast then moved toward the
posterior pole, possibly through an accelerated growth of the
striping anterior to it. When the original holdfast was not removed
the new one was soon resorbed. I have found that when a stentor
is cut in two transversely and the anterior half rotated 180° on the
posterior so that the lateral striping is out of alignment and does
not rejoin, a new tail is sometimes formed and projects temporarily
from the oral meridian of the anterior part (b). Likewise, if the
ramifying zone is circumscribed and rotated in place its posterior
end regenerates a new holdfast projecting forward (c). Consonant
with these results, Weisz offered two important principles of
holdfast regeneration: first, the presence of one good organelle

Fig. 27. Unusual tail-pole and holdfast formation in folded
non-oral halves, a: Longitudinal cut through the axis to yield
aboral half lacking widest and narrowest pigment stripes, b:
Wound healed by folding which brings head and tail-poles
together, polarities indicated, c: Lateral striping is self-severed
across the sharp bend of the fold, giving same appearance as a
fission line. Oral primordium develops where widest granular
stripes lie adjacent to their attenuated extensions, d: Cut ends
of striping drawn together to form a new pole. Original half-tail
extends temporarily but is later resorbed as the new holdfast
becomes functional {e). (After Tartar, 1956b.)

REGENERATION III

tends to inhibit the formation or persistence of an extra holdfast,
and second, that free posterior ends of one or more stripes in the
ramifying zone are inductive of tail formation.

Surprisingly, a new posterior pole and holdfast can be formed
in a way which one would never expect to occur in the usual life
of stentors (Tartar, 1956b). As Balbiani had noticed, longitudinal
aboral halves tend to fold on themselves to close the wound, and if
this situation persists, a new pole is formed at the point of bending.
There one observes that the pigment stripes are severed just as in
the formation of a division furrow, the cut ends of these, and
doubtless of the fibrous alternating clear bands as well, are then
brought together at a point from which a holdfast emerges (Fig. 27).
Sometimes when the original half-holdfast persisted and moved
posteriorly to a more normal location, it was nevertheless later
resorbed and replaced by the new organelle produced in such an
odd manner.

A stentor with single head but two tail poles and holdfasts,
like a specimen found in nature by Faure-Fremiet (1906), was
produced when Balbiani (1891b) split the posterior end. This
dupHcation can also be produced in Condylostoma (Tartar, 1941b),
but in either genus it is much more usual for the two parts simply
to fuse together again.

(c) Reconstitution of the normal shape

Examples already given are enough to indicate the strong
tendency of stentors to reconstitute the normal shape and contour
of the cell. Later discussions will show that this capacity is indeed
phenomenal, though easily passed over because of the slow pace
with which it is pursued. For the present it is sufficient to say that
no shape distortion of a stentor has yet been produced from which
the animal could not recover in time. The gradual nature of the
processes involved was emphasized by Schwartz (1935), who
showed that minor discrepancies in the striping persisted for a
long time.

Apart from such minute disruptions, the shape of a stentor
seems to be strictly a function of the pattern of the striping
(Tartar, 1954). When from aborted cleavage or for some other
reason there is a break in the striping, the contour of the cell
shows a corresponding deviation from normal (Fig. 28A) and if the

112

THE BIOLOGY OF STENTOR

Striping breaks into many patches the whole cell becomes knobby
when expanded (see Fig. 71B). Likewise, longitudinal fragments
remain thin and elongate until they recover the normal comple-
ment of stripes (see Fig. i2a). If two stentors are grafted together
at random, there is no arrangement from which they cannot shift
and integrate into a normal shape (Fig. 28b) (Tartar, 1954).

Fig. 28. Pertaining to cell shape in S. coeruleus.

A. Any discontinuity in the lateral stripe pattern results in

corresponding modification of cell shape.

B. Even head-to-head telobiotics can reconstitute a single
normal shape by jack-knifing and fusing. (After Tartar, 1954.)

C. Doublets with mouthparts proximate tend to form double

"cleavage" shapes.

Doublet stentors are usually wide, but if they become single, they
or their progeny recover the normal number of lateral stripes.
When doublets persist and retain essentially two sets of body
striping there is a strong tendency to develop a Siamese twin shape
(c), showing again that cell shape depends on the disposition of
the ectoplasmic striping.

REGENERATION II3

2. Nuclear behavior during regeneration

Not until the primordium is already half developed (stage 5)
do the macronuclear nodes begin to coalesce; their fusion is not
so complete as in reorganization and especially division. But
Causin (1931) reported that regeneration is like abortive fission in
that the macronucleus divides, the parts later rejoining. No one
else has observed this. The point needs checking, in view of Yow's
(1958) recent work on Euplotes, showing that in regeneration two
ciliary anlagen are produced, just as in division, though one is
promptly resorbed and hence was overlooked by previous
investigators.

In any event, the compacted regeneration nucleus renodulates
and the number of new nodes may not be the same as originally.
Johnson, confirming Balbiani (1889), remarked that there is in-
variably a slight increase in the number of nodes after renodulation.
The average increase was from 12 -6 to 16 macronuclear beads. If
the regenerating fragment was cut so as to contain few nodes to
start with, there was a substantial increase during regeneration.
Hence primordium formation in regeneration can be used for
correcting a decreased nucleo-cytoplasmic ratio, just as Schwartz
found for reorganization. This adjustment of the nuclear size to
the size of the fragment was confirmed by Weisz (1949a) and is in
accord with my own observations.

Prowazek (1904), too, found that the number of macronuclear
nodes always increased during regeneration. He further stated
that this increase might occur even if the stentor was only diagon-
ally cut or injured, but he does not seem to have followed his
animals closely enough to exclude the possibility that an inter-
vening reorganization had not occurred. Evidently he believed
that any substantial cut or deletion of oral parts resulted in a
nodal increase which was also an increase in the absolute size of
the macronuclear material, for he stated that after the invariable
nuclear hypertrophy there then occurred a subsequent reduction
to the normal nucleo-cytoplasmic ratio. Thus in some cases he
found that one node of a series was absorbed, but it might have
fused with another. These observations should be checked
especially with regard to real changes in the macronuclear volume.
Increase in the number of macronuclear nodes following re-
generation was explained by Schwartz (1935) in the following

114 THE BIOLOGY OF STENTOR

manner. The macronucleus increases substantially as a rule only
during cell division ; therefore pre-division stentors, as they grow,
will come to have a decreased nucleo-cytoplasmic ratio or the
need for more nuclear material, which will be redressed only
during subsequent fission. If regeneration is then brought about,
a stentor can take this opportunity of primordium formation to
make up its lack and increase the number of macronuclear nodes.
Then he found that when this happened and the stentor was caused
to re-regenerate there was now not an increase in nodal number
because the normal nucleo-plasmic ratio had already been
achieved; and if substantial parts of the cytoplasm had been
removed there might even be a decrease in number or coalescence
of nodes. However, in all this Schwartz doubted that there was an
actual change in macronuclear volume and believed it more likely
that the adjustment was largely an increase or decrease in the
effective surface of the nucleus. Yet, in grafts of two stentors
sharing but one macronuclear node I found an indubitable increase
in nuclear mass at the end of regeneration (see Fig. 86b).

When a stentor is transected across the longitudinal axis the
macronucleus is distributed about proportionally; the posterior
fragment has to regenerate a new set of feeding organelles, while
the anterior does not and serves as a control. Comparing these two,
Weisz (1949a) found that in the posterior piece only does macro-
nuclear coalescence occur, as an accompaniment of primordium
formation. More recently, Guttes and Guttes (1959) have found
that mitotic division also occurs only in the posterior fragment,
or at least this was demonstrable in 17 out of 125 cases. No mitoses
could be found in either the anterior fragments or in uncut
controls not undergoing fission. If the exact time of mitosis is
somewhat variable, this could account for their not always finding
it. They noted the similarity between their results and those of
Schwartz, who deduced that micronuclear multiplication occurs
during reorganization along with macronuclear increase. The
results showed that mitotic multiplication of micronuclei, as well as
increase in the number of macronuclear nodes (see above) can take
place in regeneration ; for only the posterior fragments would have to
form an oral primordium to replace the missing feeding organelles.
This is in accord with demonstrations of mitosis in other ciliates
during regeneration (Lewin, 191 1 ; Suzuki, 1957; and Yow, 1958).

REGENERATION II5

The Guttes assumed that both fragments regenerated. This is
true only insofar as the anterior fragments had to regenerate the
holdfast ; but this makes no demands on the nucleus and can even
occur in its absence. It has long been known (e.g., Morgan, 1901a)
that anterior halves need not and do not form an oral primordium.
Therefore the most important difference between the two types of
fragment is that oral anlagen formation occurs only in the posterior
ones, and this is somehow related to corresponding changes in
both macronuclei and micronuclei.

To the Guttes, however, the only difference between the frag-
ments was that the posterior halves lacked the feeding organelles.
Their interpretation is accordingly highly questionable; for they
suggested that in the posterior halves the phosphoryolytic energy
utilized in membranellar beating could now be diverted toward
promoting mitosis. Historically, this explanation stems from the
Henneguy-Lenhossek hypothesis, 1898, of the homology between
mitotic centrioles and the fibrogenic basal bodies of flagella and
cilia. But the application cannot be valid if we accept Schwartz's
deduction (see p. 103) that mitosis also occurs during reorganization
of stentors, during which the original membranellar band is
retained and continues actively beating as the new one from the
reorganization anlage joins with it. Instead, it may be concluded
that regeneration, reorganization, and division are so similar that
each gives the cue for macronuclear and micronuclear increase;
and it may be the developing primordium which provides this
encitement, as in part suggested by Weisz (1951b).*

3. Effective stimulus to regeneration

That cutting injuries alone, without excision of parts, do not
result in regeneration or reorganization has been pointed out many
times. Morgan (1901a) tells how he cut the cell nearly in two
without effect, even if the cut passed through the membranellar
band. I have found, however, that if the feeding organelles are cut
in two and displaced, or if for any other reason a good set of mouth-
parts and a good membranellar band are present but not joined,

* According to Uhlig (i960) " regeneration " can occur without primor-
dium formation, evidenced only by fusion and renodulation of the macro-
nucleus and formation of a new contractile vacuole under the wide-stripe
areas. Division also occurred without anlagen formation (cf. Fig. 18B).

Il6 THE BIOLOGY OF STENTOR

regeneration will then ensue (Tartar, 1957c). Causin (1931) found
neither primordium formation nor nuclear changes in coeruleus
which were cut into repeatedly. Yet (if a repetition be allowed for
completeness of this account) he remarked, without giving further
details, that when so cut the '' sectioned myofibrils degenerate and
new ones appear in the pigment bands to replace them." Weisz
(1949a) also spoke of an extensive reorganization of "contractile
equipment " during regeneration and these hints deserve pursuing.
Causin likewise considered that alteration of the nucleo-plasmic
ratio would be a sufficient stimulus to regeneration, but whatever
anticipations he may have had in this direction are probably
covered by Schwartz's demonstration that reorganization follows
excision of major parts of the macronucleus.

Even substantial portions of the lateral body wall and endoplasm
can be removed without inciting regeneration, but removal of any
portion of the feeding organelles is a sufficient stimulus to re-
generation. Excision of all mouthparts of course produces prompt
primordium formation. If the gullet or the oral pouch only is
removed regeneration also occurs, or if in morphogenesis mouth-
parts are produced which lack either of these organelles, or are in
any other way incomplete, they will be replaced by a new set
through regeneration (Tartar, 1957c). The mere act of primordium
formation therefore does not satisfy the requirements of re-
generation, for there seems to be a feedback mechanism which
informs the cell whether the resulting differentiation has been
complete.

When only the aboral half of the membranellar band is removed
regeneration is much delayed as a rule but does occur eventually.
The only exception is that, if division intervenes, the abbreviated
feeding organelles, now on the proter, may be approximately of
right proportions for this smaller cell and then regeneration does
not always occur. These results are reminiscent of Taylor's (1928)
studies on Uronychia, in which he found that the removal of one
cirrus, or the sectioning of critical neuromotor fibrils which could
then not rejoin, constituted sufficient stimulus for regeneration.

Long ago Johnson observed that a double monster stentor
regenerated doubly, on both sides, though it needed to renew only
one of the mouths. Such observations were greatly extended with
the technique of grafting two stentors together. In doublet animals

REGENERATION II7

with two complete sets of feeding organelles I found that if one of
the mouthparts developed incompletely, or if one mouth was
excised, or if one complete set of feeding organelles was removed
without leaving remnants behind, then the remaining set, normal
and fully formed, still did not prove sufficient. Regeneration always
occurred on the defective side with simultaneous reorganization
on the other. The only time when this did not take place was when
the doublet was transforming into a single stentor and one of the
primordium sites was disappearing (Tartar, 1954). Regeneration
therefore may be said to occur whenever a primordium site is not
subtended by a complete set of feeding organelles normally joined
together in one unit.

4. Time for regeneration

Clocking the time for regeneration may afford some hint
regarding the nature or the order of magnitude of the processes
involved. At least we can designate the minimum period within
which any postulated reaction must be able to accompUsh a visible
result, and this should offer some guide to hypothesis. A point
which is obvious, yet perhaps deserving explicit statement, is that
regeneration of lost parts is enormously more rapid in ciliates than
in multicellular animals.

We have noted that an excised tail-pole and holdfast in Stentor
coenileus can be re-formed in one to two hours, and little or no
synthesis of new structures may be involved. Relating oral as well
as pedal regeneration to temperature, Weisz found that lowering
the temperature 10 degrees increased the time by a factor of about
1-6. He also claimed that the presence of intact feeding organelles
hastens foot formation, yet it is possible that such formations are
retarded when the head is excised merely because an added burden
is thrown upon the cell (Child, 1949).

Oral regeneration is by elaboration of a primordium and requires
more time. An important distinction was emphasized by Weisz
when he separated a preparatory period, as the interval between
excision of parts and the beginning of anlage formation, from the
time required for the development of the primordium itself. The
former he found to require about 4 hours as a rule, though the
figure can be pushed closer to three if one is careful to watch for
the inconspicuous stage- 1 anlage. Development then proceeds

Il8 THE BIOLOGY OF STENTOR

at the rate of about one stage per hour and the total time for
regeneration from the moment of cutting is around 8 to lo hours
(Weisz, 1955). What may occur during the preparatory period is
discussed later (p. 138).

In a study of several ciliates other than Stentor but including
the spirotrichous Spirostumumy Sokoloff (19 13) stated that the
larger the fragment the sooner it regenerates, but his data indicate
that differences appear only when there is a marked disparity in
size of the pieces. The differences were explained on the basis
that a hypothetical physiological harmony has to be established
before regeneration and that this, rather than regeneration itself,
takes more time to accomplish in tiny fragments. Weisz (1948a)
did not find such differences in Stentor coeruleus and stated categor-
ically that, other conditions being the same, the time for both oral
and holdfast regeneration is independent of the initial size, pro-
vided the piece is large enough to permit any regeneration. In a
recent series of tests I have found, however, that when the head
and tail-pole of coeruleus were excised and regeneration times
measured for the main cell body and its own polar fragment the
time for the initial appearance of the oral anlage was with two
exceptions always greater in the smaller pieces, and the difference
was often considerable (unpublished). Size therefore may have a
bearing on regeneration rates.

The same tests — in which the posterior fragment was
" favored " by the holdfast — render questionable Weisz's (1948a)
contention that the presence of a foot increases the speed of oral
regeneration. Therefore, Child's (1949) criticism of this point also
may be valid.

In aboral, longitudinal halves which lack the normal primordium
site Weisz (1951b) found that oral, pedal, and contractile vacuole
regeneration were much delayed — oral, as much as 30 to 40
hours. He attributed this delay to the time required for other
stripes to assume the morphogenetic role normally played by those
in the part removed. I too have found that the preparatory period
in such fragments is usually very protracted, but there appear to
be contradictions that need resolving because this was not always
the case and some of these fragments did regenerate promptly
(Tartar, 1956c). Likewise, when only the primordium site was
removed along with the mouthparts, the time for beginning

REGENERATION II9

primordium formation was exceedingly variable, ranging from
5 to 12 hours (Tartar, 1956a).

It is relevant here that in doublet stentors, with only one set of
feeding organelles removed, regeneration is usually prompt
(Tartar, 1958b), again indicating that the remaining set of intact
organelles offer no inhibition to a primordium site which is not
subtended by one of its own. In single animals, however, the time
for beginning anlage formation does vary inversely with the extent
of oral ablations, recalling a similar rule by Zeleny (1905) for
metazoa. Thus Morgan found that the more of the membranellar
band removed the sooner regeneration followed, and Weisz (1948a)
confirmed this. A similar relationship was demonstrated in the
hypotrichous Ur onychia by Taylor (1928). Even w^hen there are
no ablations, re-regeneration occurs if for any reason the differen-
tiation of the oral primordium is incomplete, and the more
incomplete the sooner (Tartar, 1957c). I also noted the time
relation in regard to the length of membranellar band removed
and found in addition (Tartar, i959d) that if the gullet, buccal and
oral cavity are neatly removed so as to leave almost the entire length
of membranelles intact regeneration is still retarded. These experi-
ments indicate that any portion of the feeding organelles is partially
inhibitory of primordium formation, but all are required to prevent
this formation entirely.

Sokoloff and others believed that the ratio between volume of
nucleus and volume of cytoplasm cannot vary too greatly if re-
generation is to be possible, but Weisz (1948a) found that re-
generation times are the same in comparable fragments regardless
of the number of nuclear nodes included, provided of course that
at least one was present. He therefore discounted the idea of
necessary nucleo-cytoplasmic ratios. With this I can agree in
regard to the range of differences in the ratio which one finds in
fragments from a single animal, yet it will be shown later (p. 306)
that the extreme decrease in the ratio of nucleus to cytoplasm
which is made possible by grafting exp^eriments does indeed result
in very tardy regeneration.*

*Uhlig (i960) emphasized the correlation between " age " and regenera-
tion time : this period was shortest in young, post-fissional animals which
were also more reactive in producing primordium formations at multiple
primordium sites from disturbances of cell patterns.

I

I20 THE BIOLOGY OF STENTOR

5. Minimum size of regenerating fragments

Given at least one macronuclear node, how small may a frag-
ment be and still regenerate? In the earliest cutting experiments
on stentors, Gruber (1885b) had already found that not only halves
and thirds but even smaller fragments of coeruleus regenerate and
form tiny Stentors. The embryologist Lillie (1896) raised the
question of the limits of divisibility of stentors as leading to
significant theoretical implications. Fragmenting the ciliates by
shaking, he found that no piece smaller than i/24th the volume of a
large polymorphiis regenerated completely, and the minimal size
for coeruleus w^as i/3oth. Lillie was impressed by the fact that such
fragments are still of considerable size, since they were about 80 ju,
in diameter, and therefore emphasized that the cytoplasm is as
important as the nucleus to regeneration, postulating that there
is a ** minimal organization mass" below which the complete,
potential form of Stentor could not find representation. This size
limit should be absolute rather than relative ; therefore he expected
that it would not be exceeded even if one started with smaller cells
for cutting. Morgan (1901a) found that pieces no larger than
I /64th of the whole coeruleus could regenerate and this was later
confirmed by Stolte (1922). Morgan's minimal fragments were in
fact only slightly smaller than Lillie's but they were cut from
larger cells. Recalling that there are also lower limits to the size of
regenerates in Hydra, Tubularia, and Planaria, Morgan offered a
first-order explanation for the failure in regeneration of very small
pieces in both metazoa and ciliates, namely, that there is simply
insufficient material to produce the typical form.

Sokoloff (191 3) pursued this problem further in the ciliates
Dileptus and Spirostomum. The first is suitable because the macro-
nucleus is finely subdivided and widely distributed, and the second
because the very elongate shape lends itself to cutting tiny frag-
ments. Pieces i/8oth the volume of the whole cell could regenerate.
Although fragments i/iooth of the normal size could be cut, these
did not regenerate or survive for long. Therefore Sokoloff (1934)
seems to have settled on the idea that there is really no theoretically
significant limit to the divisibility of ciliates, and that in practice a
limit is imposed only by the circumstance that in smallest frag-
ments the wound surface with its exposed endoplasm is relatively
^-SoJ-ajcge that the pieces become vacuolated and soon disintegrate.

REGENERATION 121

This conception was taken to the extreme by Weisz (1954)
when he stated that size is not a Hiniting condition of regeneration
in protozoa and that theoretically one molecule of deoxyribonucleic
acid surrounded by a shell of cytoplasm should be able to re-
constitute the organism. He therefore regarded the explanation of
Lillie and Morgan concerning minimal size as untenable. Weisz
(1948a) found successful regeneration in pieces of coeruleus as
small as 70 /x in diameter; yet he reported and later emphasized
(Weisz, 1953, 1954) that even much larger fragments could be
produced which are incapable of regeneration. The crucial point,
he thought, was w^hether or not a fragment contains a portion of
the normal primordium site and hence presumably specialized
kinetosomes which alone can produce an oral primordium. Yet
this explanation is contradicted not only by Causin's (1931)
demonstration of the dispensability of the primordium site but also
by Weisz's own experiments, mentioned above, showing that
regeneration can occur in aboral halves, though much delayed.
And I have found (Tartar, 1958b) that nucleated primordium
sectors, or just the part of the stentor cell which contains the
primordium site, can regenerate completely, with mouthparts,
only if of sufficient size.

When a sample of coeruleus is set aside for a week or two without
added nutrients the animals starve until individuals are produced
which are much smaller than normal daughter cells. Starting with
these starvation dw^arfs, I cut off substantial portions of the posterior
pole and found that pieces as small as 75 ft in diameter or only
I /123rd the volume of large, pre-starvation stentors, could re-
generate completely and survive for over 6 days (Fig. 29). Although
these tiny stentors had much fewer than the usual number of
membranelles, the width and length of these organelles when
measured proved to be very nearly the same as in large animals,
and these relatively oversized organelles caused the anterior end of
the tiny animals to shake and shudder with their beating.

Therefore it seems to me, as previously suggested (Tartar,
1941b), that a Hmit to reconstitution of the normal form is imposed
simply by the fact that the units of ectoplasmic structure are each
of a nearly constant size or incapable of '' miniaturization ", so that
with decreasing volume there will come a point beyond which
the formation of anything like a normal set of feeding organelles

°/ \^

|uj I LIBRARY 1>:|

122

THE BIOLOGY OF STENTOR

is impossible with such units. Failure of smallest pieces to re-
generate would then be due neither to pathological changes nor
to insufficiency of material but rather to structural incompatibility
between the size of the parts and what is to be made from them.
Tiny fragments can produce some oral cilia and membranelles but
it may well be that there is a jamming when these parts attempt to
coil tightly inward to produce a gullet.

Fig. 29. Regenerated *S. coeruleiis of near minimum size, a:
Tiny and large stentors drawn to same scale. Note that mem-
branelles are of same width and length in both. Pigment stripes
are also of similar widths, hence minute form had only about 20
as compared with 100 for the large animal, b: Enlarged view
of regenerant, which has but one macronuclear node and very
few stripes in the frontal field.

If units of ectoplasmic structure in ciliates, such as oral cilia,
body cilia, and trichocysts, are of a standard, nearly invariant size
for any species of ciliate (cf. Bonner, 1954; Ehret and Powers,
1959) this should simplify the problems of growth; for one would
then need only to explain their increase in number, and further
hypothesis regarding their adaptive size would not be necessary.
This seems to be one of the crucial theoretical points involved in
these small-fragment studies. The other resides in the amazing
fact that organic form is largely independent of size and, outside
the limitation just mentioned, it is possible for stentor shapes and

REGENERATION 123

feeding organelles to be produced in an enormously wide range of
sizes. It seems that nature herself has already explored these
possibilities, for the tiny, blue-green Stentor tmdtiformis appears
in almost every respect like the tiniest regenerate of coeruleus.

6. Adjustments to proportionality of parts

Tiny fragments form primordia which are very short though
apparently of normal width and therefore regenerate a set of feeding
organelles proportionate to their size except that the individual
membranelles are relatively large. In the other extreme, Balbiani
(1891b) noticed abnormally large mouthparts in some of his
double monsters and I, too, have occasionally seen the same, as
well as very large frontal fields and unusually long membranellar
bands in the products of stentor grafting. Hence the normal upper
limit in size of these organelles can also be exceeded.

When regeneration is induced by excising the mouthparts only,
the new membranellar band joins with the old one. Therefore one
might expect that when the entire head is removed the regenerated
membranellar band would be smaller ; but in this case the primor-

FiG. 30. Proportionality of parts in S. coeruleus.

Anterior half of transected stentor is at first too short and with
too-large head. Membranellar band and frontal field are then
reduced to half original size without primordium formation, as
the cell extends and a new tail-pole and holdfast are formed.
Posterior half is at first too long, then regenerates a smaller set of
feeding organelles, as the posterior pole is proportionately
reduced. (After, Morgan, 1901a.)

124 '^"^ BIOLOGY OF STENTOR

dium grows to a greater length, extending far forward, and so the
size of the regenerated feeding organelles is the same and
proportionate.

Of special interest is the finding of Morgan (1901b) that frag-
ments whose parts are rendered disproportionate by the cutting
do not wait, as conceivably they might, for gradual differential
growth to right the imbalance but adjust to proportionality
relatively soon. Morgan cut unfed coeruleus in two transversely and
observed in the anterior fragments that the stalk which was at first
too short then gradually lengthened, while the original membran-
ellar band, initially too large, became reduced to half its starting
size without formation of a new one, and proportionality of parts
was regained (Fig. 30). In posterior fragments the stalk was at
first too long, but it gradually came to assume normal proportions
and the regenerated feeding organelles were of course of smaller
and proper size. Reviewing his own studies, Morgan (1901b) then
added the statement that the regenerated organelles on the
posterior fragment are in fact too small and that they "later
become larger until the characteristic form is reached". This
would imply an improbable growth in situ, a question which will
be dealt with shortly.

Prowazek (1904) said that he confirmed Morgan's original
findings and noted that they imply, with reference to the anterior
fragment, that there should be an imperceptible resorption of
portions of the old membranellar band to make it proportionate
in size. Such adjustment he thought was exhibited in a dramatic
way in the case of a stentor which divided unequally, producing
a smaller than normal proter which carried the now much too
large original ingestive structure. The feeding organelles then
gradually regressed until they appeared to be completely resorbed
while a new primordium was forming to produce a head of proper
proportions. Yet this behavior may be regarded as anomalous
because it does not occur even in the most abbreviated anterior
fragments in which there is more occasion for it.

Even in normal division the original head, which is passed on to
the anterior daughter, is at first too large but on separation both the
proter and the opisthe seem to have feeding organelles which are
equal in size and proportionate. According to Weisz (1951b),
adjustment occurs in the presumptive proter during the last stages

REGENERATION I25

of division, whereby the original feeding organelles are reduced in
size. The partial regression of the mouthparts at this time, in
which disappearance of the oral pouch as such is particularly
conspicuous, may represent the initial steps toward a remodeling
of the mouthparts on a smaller scale, but further changes are not
easily followed.

We do not yet understand what determines the size or scale of
mouthparts formed anew. Experiments here are contradictory.
When a stage-3 regenerator was cut in two transversely through
the primordium and the anterior half rotated 180° on the posterior
the short anterior half anlage produced a tiny mouth while the
posterior section of equal length was completely employed in
forming a large one (Fig. 31 a). If the two fragments were entirely
separated, however, each portion of the primordium produced a
small and proportionate gullet and oral pouch in addition to the
membranellar band (b). An odd case, in which the regeneration
primordium was unusually short, produced a tiny set of mouth-
parts in a large stentor (c) ; but when a nucleated primordium sector
was isolated from a stage-4 regenerator the mouthparts were still
proportionate to the fragment although the anlage was of normal
length (d). When tail-poles were grafted into the frontal field and
reorganization followed, the mouthparts produced on the graft
were proportionate to its size, as were those on the host (e). Hence
in some cases the length of the primordium and in others the size
of the cell seemed to determine the scale of the parts produced.

The most exaggerated requirement for an adjustment of cortical
organelles is occasioned by producing fragments which consist of
the head only (Tartar, i959d). By circumscribing the membran-
ellar band and cutting carefully around the oral pouch and gullet
so as not to disturb them, fragments were cut which contained
only the feeding organelles intact, the frontal field, a little endo-
plasm, and usually one or two of the most anterior macronuclear
nodes. Much shorter than the anterior fragments cut by Morgan,
these pieces folded on themselves in healing to produce spheres in
which the membranellar band was thrown into coils like the
stitching on a baseball (Figs. 32 and 86c). In these specimens there
was no primordium formation, but the membranellar band soon
decreased in length as it became normally disposed and the
mouthparts were later gradually reduced in size, while ecto-

126

THE BIOLOGY OF STENTOR

Fig. 31. Observations relating to proportionality of mouthparts
in S. coeruleus.

A. Regenerator in stage 3 is transected and the halves rotated
on each other. Both sections of the as yet undetermined pri-
mordium produce mouthparts. Those from the anterior part
are very small; posterior anlage almost entirely used to form a

very large set of mouthparts. Yet —

B. If the halves of such a specimen are separated, equal,
proportionate and medium-sized oral differentiations are

produced.

C. From deletions to the primordium site a very short
regeneration anlage was produced, forming much too small a

set of mouthparts for the size of the animal.

REGENERATION 127

plasmic striping grew out and the normal form and proportions
of a stentor were reconstituted on a small scale. But nothing of
this happened if no nuclear beads were included and the fragment
then remained until death about four days later just as it was after
cutting and healing. It would therefore seem that the nucleus is
essential in both the formation and the dedifferentiation of oral
structures. These cases demonstrate how capable is Stentor in
adjusting its parts to normal proportions.

Fig. 32. Adjustment of size of parts in nucleated, isolated head
of S. coeruleus. Feeding organelles and frontal field are excised
without injury but with minimum lateral ectoplasm. In folding
to cover the wound the fragment becomes much contorted.
Membranellar band decreases in length and lateral striping
gradually grows out to form a tail-pole. Later the mouthparts
are also decreased in proportion. Adjustment occurs without
primordium formation but only if nucleus is present.

7. Can mouthparts and membranelles be formed in situ ?

In the normal course of life new feeding organelles in Stentor
are formed only through the development of an oral primordium ;
yet there are hints in the literature that this may not be the only
pathway to oral differentiation, although no really convincing
demonstrations have been offered. In respect to the mouthparts,

D. Primordium sector isolated from a stage-4 regenerator.
Development continues and size of mouthparts is proportionate
not to the original cell or the length of the anlage but to the size

of the fragment.

E. Tail pole was grafted to frontal field of a stage-2 regenera-
tor. First sketch shows an additional anlage now induced in the
graft. On developing, the primordia produced mouthparts

proportionate to the size of the part in which they arose.

128

THE BIOLOGY OF STENTOR

Morgan commented that if a portion is removed the remaining
parts seemed to reconstitute the normal ingestive structure, though
this was generally replaced later by a new one. In my experience
an isolated gullet can attain a neat opening on the surface and
attaches to the correct end of a remnant of the membranellar band,
while the severed oral pouch with its membranellar border also
does not remain as cut but coils sharply to form a pigmented
depression with the shape of the inside of an abalone (Fig. 33A).

Fig, 33. Relating to reconstitution and formation in situ of
mouthparts.

A. Gullet severed inside of stentor, isolated oral pouch widely
displaced, gullet opening destroyed by anterior incision. Gullet
finds neat opening to exterior and joins adoral end of adjacent
membranellar band, while oral pouch coils sharply as if attempt-
ing mouth formation. Regeneration follows.

B. Two types of gross oral injury which are followed by
mending without regeneration: sectioning mouthparts but
leaving them close together, and thrusting an eyelash into the

gullet and out through opposite side of the cell.

C. Before regeneration, adoral end of the membranellar
band may produce a small pit, or a tight coiling (D).

REGENERATION I29

Yet neither part reconstitutes a complete mouth, even if either one
is completely removed. If the gullet is severed from the oral pouch
and the structures are left adjacent, or if the mouthparts are severely
injured in place, in the vast majority of cases the parts will rejoin
and perfect mouthparts be reconstituted without the formation
of a regeneration primordium. That some remodeling can occur
in situ was indicated by the fact that in one case an unusually long
and wide gullet was produced. In another case an eyelash was
thrust down the gullet and out the side of the cell yet no regenera-
tion followed and the stentor was later capable of forming food
vacuoles (Fig. 33B).

Morgan also noted that some of his aboral, anterior fragments
formed a small oral pit at the proximal end of the membranellar
band remaining (Fig. 33c), and a similar effort toward oral re-
generation was also observed by Causin (193 1). I have observed
these formations too, as well as the tendency for the cut, proximal
end of the membranellar band to form at least a tight little coil (d).
I further reported (Tartar, 1956a) a case in which good mouthparts
were apparently reconstituted from the buccal pouch alone, as
well as the formation at least of an apparently complete gullet
instead of merely a pit at the end of an adoral band (Tartar, 1956b).
One may at least conclude that the mouthparts are quite capable
of repairing themselves.

This may also be said for the membranellar band. If the band
is cut in two or small sections of it removed, the parts simply heal
together and no regeneration ensues. Whether some compensatory
growth of membranelles occurs if some are excised has not been
precisely determined. Stevens (1903) found in oral longitudinal
halves "some evidence" that the abbreviated membranellar band
increased in length. But the formation of regeneration primordia
in stentors from which half the band has been excised speaks
against the formation of membranelles in situ. Were this possible,
such regeneration would then not be necessar}^

In Prowazek's important if miscellaneous paper of 1904, he first
reported that the membranellar band in coeruleus is shed when the
ciliates are subjected to a weak solution of table salt. Then he noted
that after 24 hours a new membranellar band was regenerated in the
same place (an derselben Stelle). This is all he says. The point is
not developed further, nor was this remark italicized, as was his

130 THE BIOLOGY OF STENTOR

habit in emphasizing major issues in the remainder of the paper.
Hence it seems to me that all later commentators have mis-
interpreted this passage as a statement that membranelle formation
can occur in place or without primordium formation. But this is
not to exclude that such development may occur after a fashion,
anomalous as this w^ould be. Schw^artz repeated Prowazek's salt-
shedding experiments and stated that in some cases there W2is
clearly a neo-formation of membranelles in situ. His explanation
was that only the cilia of the membranelles had been cast off,
leaving the basal bodies intact, from which new cilia may have
grown ; and he remarked that if this can occur, such replacement,
rather than primordium formation, should be the method of
renewing supposedly worn-out feeding organelles. I have myself
noticed a few similar cases. In one of these, a stage-3 divider was
treated with sucrose and it shed the membranellar band. The
division primordium remained but showed abortive development,
while around the anterior rim of the cell there appeared within
about 4 hours shorter than normal oral cilia which beat in meta-
chronal rhythm. Such cases indicate that if carefully graded treat-
ments were employed, a renewal of the large oral cilia if not the
entire membranelle might be firmly established. Yet it is certain
that in most experiments of this sort the entire band comes off and
the regeneration primordium is soon formed (see p. 252).

8. Repeated oral regeneration

Since the formation and development of an oral primordium
involves the production of thousands of new, large, oral cilia as
well as other parts, one wonders whether there is an inexhaustible
reserve for such synthesis. Gruber (1885b) cut and presumably
decapitated a coeruleus on 5 successive days and each time
complete regeneration followed until the animal finally became
necrotic and too small for further operation. With the same large
species, Prowazek (1904) also performed successive cuttings. In
one tabulated case an animal was cut nine times during which
macronuclear beads were not removed, and this animal always
regenerated. The material of the macronucleus seemed to have
been substantially drawn upon, because it was finally reduced
from II to only 2 nodes. He also reported 3 cases in which the
animal was repeatedly cut or wounded and compelled to re-

REGENERATION 131

generate and that these then became able to regenerate without
the nucleus. This surprising result was explained in terms of the
then-popular chromidial hypothesis, whereby a nucleus can be
stimulated to extrude chromidia, which can then substitute for it
(see p. 299).

Hartmann (1922) posed the question of whether division could
be indefinitely postponed by repeated cutting ablations on a
feeding cell. That this is the case, he demonstrated for Amoeba
and the fresh water worm Stenostomum, as well as for Stentor
coeruleus. Stentors were fed on Colpidium and allowed to grow but
were cut before they attained division size. Hartmann noted that
a cut could produce either oral or headless remainders and,
although his account is not clear in this regard, I assume from his
statement that regeneration occurred and that this was oral
regeneration and not merely holdfast renewal or recovery of normal
shape. In one tabulated case a stentor regenerated 25 successive
times during 52 days, without fission, while the controls divided
35 times. These results indicated, that if there is an accumulation
of some factor disposing the cell to fission, this factor is reduced by
excisions ; as well as that indefinitely repeated regeneration seems
to be possible within one individual if fed, and that frequent fission
is not essential to survival.

9. Blockage of regeneration

Although stentors regenerate with the greatest regularity and
can even re-regenerate repeatedly or exhibit a succession of re-
organizations in starved fusion complexes, I have encountered a
half-dozen cases among thousands in which, for some un-
accountable reason, otherwise healthy appearing coeruleus failed
to regenerate the feeding organelles though surviving for many
days. A similar number of instances were found among starving
animals, which is enough to give the impression that stentors
cannot form regeneration primordia without carbohydrate reserves
as Weisz (1948b) asserted. Yet a direct pursuit of this question
showed that even the most pellucid animals without food vacuoles
or demonstrable glycogenoid granules were still quite able to
regenerate (Tartar, 1959a). On the other hand, it is common
enough to find that necrotic stentors or animals which have an
apparently decreased vitality from being long isolated on slides

132 THE BIOLOGY OF STENTOR

are unable to consummate regeneration. Improper healing of cut
animals is supposed to offer a blockage to regeneration according
to Sokoloff (1924) and Weisz (1948a) ; but my experience is that the
healing capacity of stentors is sufficient for neat repair after any
cutting operation except an extreme reduction in the ectoplasm
which alone prevents apposition of cut surfaces.

Nevertheless regeneration can be blocked in Stentor by treat-
ment with certain chemical agents. Weisz (1955) tested the effects
on regeneration of over 20 compounds, including substituted
purines and pyrimidines and a variety of anti-metabolites. The
most effective, in the sense of producing reversible blockages
without toxicity, was acriflavin, a mixture of 2,8-diamino-io-
methyl-acridinium chloride and 2,8-diamino-acridine. These
compounds or their allies are bacteriostatic, and some of their
effects on ciliates had already been explored (Robertson, 1925).
Weisz reported that acriflavin has a graded sequence of effects on
coeruletis, depending on concentration and duration of exposure.
First there was some paralysis of ciliary beating and cell contrac-
tion, followed by more or less complete shedding of the peUicle.
Oral primordium formation might then be merely delayed, or pre-
vented entirely, the animals then dying. When primordium forma-
tion occurred there were graded effects in the completeness of the
development of the anlage. The primordium might appear briefly
and then be resorbed without any attempt at re-regeneration. Oral
formation might be arrested at stage 4, producing a membranellar
band which developed no further. Sometimes the band could
assume the normal curvature but failed to coil inward and develop
the gullet and associated mouthparts. These inhibitive effects
could be reversed or counteracted by other agents: adenine,
guanine, thymine, uracil, folic acid, RNA, and DNA, the two
latter, presumably the commercial product from yeast, being the
most effective. Interpreting these findings, Weisz postulated that
development of the oral primordium is a series of separate morpho-
genetic events interconnected by acriflavin-sensitive transition
reactions. Kinetosomes might be affected in several of their
functions, first in the promotion of ciliary beating, then in their
synthesis of new cilia, and finally in some morphogenetic activity
by which membranelles and other complex organelles are pro-
duced. Application of compounds which reversed the effect of

REGENERATION 133

acriflavin had the same effect whether administered before or
during the acriflavin treatment, and hence it appeared that inhi-
bition by acriflavin is non-competitive. He could not say whether
the effect is physical or chemical.

This inviting biochemical approach to cell differentiation as
expressed in oral primordium formation in Stentor is being pursued
further by A. H. Whiteley. He is finding (unpublished) that both
the purine analogue, 8-azaquanine — which gave no effect for
Weisz — and the pyrimidine analogue, 2-thiocytosine, completely
block anlagen formation in coeruleus. The inhibition is reversible,
and regeneration of animals returned to lake water indicates that
this result is probably not due to toxicity but to interference with
the formation of nucleic acids which incorporate purines and
pyrimidines. Moreover, in the case of 8-azaquanine the effect is
counteracted by the presence of normal components of nucleic
acids, i.e., hydrolyzed yeast RNA or by RNA directly. And the
impHcation of RNA in primordium formation is further indicated
by Whiteley's finding that a certain concentration of the RNA-
destroying enzyme, ribonuclease, can also block regeneration.
The abolition of this effect by added RNAimpKes that the RNAase
was in fact producing this blockage through destruction of
ribonucleic acids.

Similarly, but at a wider range of concentrations, 5-methyl-
tyrosine prevented regeneration without appreciable toxic side-
effects. Since this compound is an antimetabolic analogue of
adenosine found in most proteins, the result, in this case was
probably due to the blockage of protein synthesis. Therefore it
appears that primordium formation in which thousands of new
cilia are produced does involve extensive protein synthesis and
not merely the translocation of proteins already formed, as well
as that RNA is equally implicated, in accordance with the hypo-
thesis that RNA guides protein synthesis (Brachet, 1957).

A satisfactory elucidation of the intimate material basis of the
elaboration of cell differentiations is rendered promising in regard
to Stentor by the fact that several treatments inhibit oral anlagen
formation, presumably by affecting separate, essential aspects of a
complex process. Even simple salts in very dilute solution also
delay or prevent regeneration or inhibit primordium develop-
ment (see p. 254). Moreover, regeneration may be blocked by

134 THE BIOLOGY OF STENTOR

morphological disarrangements without chemical additives, as
when reversing the single primordium site often if not always
precludes the formation of an anlage (p. 197).

Oral regeneration is thus often the preferred phenomenon for
study because by oral ablations we can induce at will the bio-
chemical and epigenetic processes involved in primordium forma-
tion. But there is no reason to suppose that the fundamental
features of anlagen development in regeneration are different from
those in the more autonomous performances of division and re-
organization. Instead, it is perhaps reasonable to suppose that
from the means providing for the basic requirement for reproduc-
tion by division were developed the capacities for reorganization
and regeneration which seem far less significant for survival of
the species.

CHAPTER VIII

ACTIVATION AND INHIBITION OF
THE ORAL PRIMORDIUM

When a stentor which is in the process of developing an oral
primordium is intimately grafted to a normally feeding partner
not producing a new set of feeding organelles, both animals no
longer continue on their original ways but now act upon each
other with significant and visible consequences. These inter-
actions were first explored in fusions of regenerating to non-
diiferentiating stentors. Regenerators often caused the partner to
produce a primordium and undergo parallel reorganization. This
type of interaction may be called induced reorganization as
formulated by Weisz (1956). Alternatively, the influence may
proceed in the other direction, suspending the regeneration and
causing the regenerator to resorb its anlage. This reaction may be
referred to as induced resorption of the primordium (Tartar,
1958b). Another way of exhibiting these interactions is to graft
cell sectors bearing primordia on to various hosts. When grafted
to stentors which are themselves undergoing regeneration, the
extra primordia are accepted, supported, and continue developing;
but if implanted on to non-diflFerentiating stentors primordia are
resorbed though the patch itself is incorporated into the lateral
striping of the host.

The range and basis of these reciprocal influences have been
quite extensively explored (Tartar, 1958b, 1958c). The results can
be explained in terms of two contrasting cell states: activation, in
which something of the whole cytoplasm is involved in supporting
primordium formation and development, and inhibition, which is
equally pervasive and tends to block or counteract the processes of
cell redifferentiation. Either of these states is sufficiently potent
to spread from one cell to another with which it is intimately
joined, in the one case to force a precipitous primordium formation,
and in the other, to cause the complete resorption of an anlage
which is already well begun.

K

135

136 THE BIOLOGY OF STENTOR

I. The course and spectrum of cell interactions

Enlarging upon these statements, we shall at first and for the
most part confine our account to regenerators and their parts
interacting with non-differentiating stentors (Tartar, 1958b).
When a sector bearing the primordium and a few macronuclear
nodes is cut out of a regenerator, development of the anlage con-
tinues as the fragment regenerates a small stentor; or when the
sector is grafted into the back of another regenerator, both host
and donor primordia continue differentiating and produce a
doublet or bistomial stentor. These tests show that such sectors
contain all that is necessary for anlagen development and that the
grafting operation itself has no effect on this process. But if a

B

Fig. 34. A. Induced primordium resorption. Sector of a stage-3
regenerator grafted to a non-differentiating host (a), b: Anlage
is promptly resorbed, but not the multiplied fine striping
encompassed by it. c: Specimen undergoes regeneration-
reorganization because added primordium site does not subtend
mouthparts, and a doublet is formed (d).
B. Accelerated reformation of primordium after excision.
a: Primordium and site removed, b: Rift soon appears in
previously closed line of heal, c: Primordium appears in rift
within an hour after operation and there is no multiplication of
adjacent lateral stripes, d: Only relatively few and broad stripes
are hence carried into the frontal field.

ACTIVATION AND INHIBITION: ORAL PRIMORDIUM 137

sector bearing a mid-stage primordium is grafted onto a non-
differentiating stentor (without primordium and not in process of
regenerating, reorganizing, or dividing), the primordium is
promptly taken down and resorbed — not sloughed. Notice in
Fig. 34A that the patch itself is not resorbed, nor are the newly
multiplied fine striping to the right of the primordium site;
instead, the sector becomes part of the lateral striping of the host.

This experiment shows that something besides cytoplasm and
macronucleus is necessary for regeneration. The cell must also be
in a state of activation. The nuclei of the non-differentiating host
could have been replaced by others from an actively regenerating
stentor and primordium resorption would still have taken place;
likewise if the host's primordium site had been replaced with the
grafted sector. Therefore the state of inhibition (or its opposite,
activation) seems to characterize the cytoplasm; and not merely
the part adjacent to anlage formation, but every part of the cyto-
plasm. For it is clear that an inhibiting influence was spreading
from the host, across the grafted patch, to the primordium,
resulting in its dissolution. There are indications, though not yet
conclusive, that the endoplasm as well as the macronucleus is
indifferent, with cell states characterizing the cortex alone.

A reciprocal influence appears to occur in the later history of
this type of case; for now the grafted primordium site, lacking
subtending oral structures, is incited to produce a regeneration
primordium and brings the host along with it into activation, with
the result that combined regeneration and reorganization occur to
produce a doublet stentor.

Rapidity and success of induced resorption depends upon the
stage of development of the imposed primordium. Early anlagen
to stage 3 can be completely resorbed within about 2 to 4 hours.
Stage-4 primordia which already have a well developed membran-
ellar band can also be dissolved, but this requires many hours
during which the anlagen crumples and is gradually taken down,
though complete resorption may not occur. From stage 5 onward,
the primordia do not seem to be resorbable under any conditions,
yet they do not remain unaffected when grafted to non-
differentiating hosts. Late primordia shrink in length or become
compacted and convoluted as if the ectoplasm were not co-
operating in their deployment, and mouthparts are not developed

138 THE BIOLOGY OF STENTOR

or remain incomplete. Examples of this abortive development will
be noted later.

A state of inhibition can therefore adversely affect primordium
development at any stage until final oral structures are formed,
or conversely, a state of activation is essential during all this time.
The initial appearance and preparation of the anlage also requires
activation. For incipient regenerators at what may be called
stage O will not even begin primordium development if grafted
to inhibitive, non-differentiating partners. The inhibition is in
fact then so strong that the regenerator usually does not begin
regenerating until the following day.

Returning again to our typical experiment, consider now what
happens to the regenerating stentor after the primordium sector
has been removed. A new anlage can appear within one hour,
although an hour and a half is closer to the average interval.
This precipitous re-formation of the anlage is most simply ex-
plained on the basis that the cell was already activated.

An accelerated renewal of the anlage of a quite different order
of magnitude (6 vs. 9 hours) was noticed by Weisz (1956) in
comparing dividers, which had resorbed their primordia because
of injuries, with injured pre-fissional animals. This time difference
he attributed to the persistence of an " anarchic field " or multiplied
store of new kinetosomes which remain ready to supply materials
for the new primordium. Yet, when an anlage is resorbed there
is no rift left in the ectoplasm to indicate that kinetosomes remain,
and one would expect an "embryonic" anarchic field also to be
resorbed since the earlier and more nascent the primordium the
more easily it is resorbed. Moreover, in regenerators in which a
new primordium could appear within the surprisingly short time
of a single hour, a relatively large sector bearing the anlage was
excised so that any anarchic field adjacent to the primordium
would also surely have been removed. For in the related Fahrea
the new kinetosomes lie between the kineties immediately adjacent
to the anlage and in Stentor they seem to be coincident with the
primordium itself (Villeneuve-Brachon, 1940), so it should be
impossible to cut out the anlage without also removing its progen-
itors. I therefore cannot agree with Weisz's explanation, nor accept
his claim to have effected this separation of primordium and
precursors in other experiments. The long preparatory period of

ACTIVATION AND INHIBITION: ORAL PRIMORDIUM I39

about four hours between inducement of regeneration and first
appearance of the primordium is probably occupied, not by
developing an anarchic field or other assemblage of formed
materials for the anlage but in transforming the cell from a state
of inhibition to one of activation (Tartar, 1958b).

In the rapid re-formation of anlagen in regenerators minus
primordium sectors the primordia themselves are normal and lead
to successful regeneration, but their manner of appearance is un-
usual. As shown in Fig. 34B, the line of heal simply reopens and an
anlage appears in the rift. Apparently there is no time for con-
comitant stripe multiplication in the presumptive frontal field,
and the primordium simply cuts out and carries forward some of
the relatively wide striping on its right side. The frontal field is
correspondingly abbreviated and reorganization therefore often
follows.

A similar appearance is also found in induced reorganization.
If a stage-3 regenerator is grafted to a smaller non-diflFerentiation
cell the latter exerts an initial influence by causing the arrest or
even partial regression of the regenerator's anlage, though later
the regenerator is dominant and induces normal primordium
formation, with stripe multipHcation, in the partner which then
reorganizes simultaneously (Fig. 3 5 a). But when a stage-4
regenerator is used no transient regression of the original anlage
occurs, and the induced primordium may be forced to appear so
rapidly that there is neither stripe multiplication nor normal
growth in length of the anlage (b). As indicated in the first example,
the impression is unmistakable that in mis-matched grafts there is a
contest and conflict between primordium activation and inhibition,
the flnal outcome of which is only decided after some time.
Figure 35c illustrates a case in which an incipient regenerator was
grafted to a small non-diflferentiating partner: a regeneration
primordium soon appeared and an anlage was induced in the other
component, then regression of both primordia occurred, after
which both were revived and regeneration-reorganization went
to completion. If in balance, with the forces of inhibition apparently
equalling those of activation, neither resorption nor development
occurs ; the primordium is not merely arrested but seems abortive
as it takes on a crumpled appearance, and so the graft complex
remains for a half-day or more until an entirely new start is made

140

THE BIOLOGY OF STENTOR

(d). Considering this case as a mid-point, the range of interactions
was from one extreme of prompt and complete primordium
resorption to the other, or precipitous induction of anlagen
formation.

Where the final result will lie within this spectrum depends
upon the stage of the original primordium, the relative volume of
the two cells, and the intimacy of their union. As we shall see in

A

B

C

D

Fig. 35. Activation-inhibition reactions in parabiotic stentor

grafts.

ACTIVATION AND INHIBITION: ORAL PRIMORDIUM 141

the next section, the stage of differentiation of the anlage is
probably significant as marking the waxing and waning of a w^ave
of activation. That this activation or the reciprocal state of
inhibition characterizes some aspect of the whole cell is shown by
the importance of the relative size of the two graft components.
A large regenerator induces reorganization in a much smaller non-
differentiating partner, if the latter is larger it forces the regenerator
to back down and resorb its primordium. If the two cells are equal,
anlage resorption also occurs, and this seems to indicate that the
force of inhibition is stronger than that of activation. However
that may be, the two forces or cell states are seen to be quantitative
and potentially measurable.

On the other hand, the stimulus which starts the whole course
of regeneration is stronger than the forces of inhibition, as indeed
it must be if primordium formation is to be possible at all. Thus

A. Induced reorganization, a: Large stage-3 regenerator
(activated) grafted to small non-differentiating partner (inhibited
with respect to anlage formation), b: Initial partial regression
of the primordium under influence of partner, c: Revival of
regeneration primordium and induction of reorganization
primordium in small partner, d: Regeneration-reorganization,
with resorption and renewal of oral structures in the reorganizer,

producing a doublet stentor.

B. a: Stage-4 regenerator grafted to small non-differentiating
partner, b: More advanced regeneration primordium does not
suffer partial regression and a reorganization anlage is induced
so rapidly that no concomitant stripe multiplication occurs
{y, cf. x). c, d: Regeneration-reorganization produces a doublet.

C. a: Stage-o regenerator (stripes splitting in primordium
site) grafted to small non-differentiating cell — immediately
following operation to show how cells are split down the backs
opened out and pressed together, b: Regenerator continues to
stage 2, induced primordium in stage i (predominance of activa-
tion), c: Conspicuous regression of both anlagen (predominance
of inhibition), d: Revival of primordia leading to doublet
formation through regeneration on one side and reorganization

on the other.

D. Abortive primordium development, a; Stage-4 regenera-
tor grafted to non-differentiating animal of same or larger size.
b, c: No induction. Advanced primordium arrested, shortened,
crumpled — neither developing nor resorbing and showing no
normal membranelles. d: Simultaneous regeneration and

reorganization occurring much later.

142 THE BIOLOGY OF STENTOR

if a large and a very small non-differentiating stentor are grafted
together and the mouthparts then excised from the minor compo-
nent, simultaneous regeneration and reorganization then occur in
the graft complex (Tartar, 1954). For now the reorganization
primordium is not to be regarded as induced by the regenerator ;
instead the stimulus to regeneration somehow passes from the
small cell to the larger, causing it to produce its own state of
activation.

Moreover, in some cases, stage-i regenerators did induce
reorganization in non-differentiating partners which were much
larger than they. Here it is possible that something of the powerful
original stimulus to regeneration, whatever its nature may be,
lingers in the early regenerator to boost its inductive influence.

The relevance of the intimacy of union on the timing and final
result of the interaction between a differentiating and a non-
differentiating stentor will be important in analyzing the basis of
the mutual influences (Weisz, 1956). When the two partners are
firmly but not broadly joined, the reorganization primordium
induced by a regenerator is noticeably tardy in appearing (Fig. 36A) ;
when the joining is tenuous, there is no induction at all (b).

B

Fig. 36. Barriers to induced reorganization, shown in tail-to-tail
telobiotics with one head excised.

A. When union is broad, regeneration in one induces
reorganization in the other partner, but with considerable delay.

B. If connection is tenuous, no induced primordium forma-

tion occurs. (After Tartar, 1958b.)

2. Timing the period of activation

When a late stage-4 regenerator is grafted to a smaller non-
differentiating stentor, there is usually the transient induction
of a beginning reorganization primordium; but the regenerator

ACTIVATION AND INHIBITION: ORAL PRIMORDIUM 143

now overtakes the reorganizer, and as the original primordium
goes into its final development the induced anlagen is resorbed
(see Fig. 38E). Stage-5 regenerators are no longer able to induce
reorganization in a partner cell. Therefore we may say that as
anlage development goes to completion the state of activation
ceases and is replaced by a state of inhibition.

When does activation begin? This time can be determined by
several tests. If a sector bearing the primordium of a regenerator
is grafted into a regenerating stentor the transplanted anlage
continues its development along with that of the host; but if the
primordium is grafted into a regenerator in which the primordium
has not yet appeared, the transplanted anlage is resorbed. When
regeneration is induced by causing the membranellar band to be
shed in salt solutions and when some of the salt is carried over
with the specimen and regeneration is thereby much delayed, such
cells are also not able to support primordia grafted to them
although it may have been many hours since the stimulus to
regenerate was given. And if stage-2 or 3 primordia are implanted
on non-differentiating cells whose heads or mouthparts are then
excised, the stimulus to regeneration in the host is not itself
sufficient to support the primordium development and the anlage
remains for a long time in arrested development or may even
become partially resorbed, but is finally revived and continues
differentiation as the host primordium itself appears and develops.
Considering these results and allowing for an appreciable time-lag
in the effects upon each other of host and graft, we can conclude
that activation is not developed to an effective state until shortly
before the primordium appears.

3. Relation of the macronucleus to activation and inhibition

The cell states relating to primordium formation and develop-
ment seem to reside in the cytoplasm and are possibly restricted
to the cortical layer or ectoplasm. The nuclei probably respond
to changes in the cell state, as when macronuclear nodes condense
and micronuclei undergo mitosis simultaneous with the passing
of the cell from its state of activation to one of inhibition ; but they
do not seem to be the bearers or determiners of these cell states.
The evidence for this is, briefly, that enucleated non-differentiating
stentors cause as prompt and as complete a resorption of anlage in

144 '^"E BIOLOGY OF STENTOR

nucleated primordium sectors grafted to them as nucleated hosts,
and the macronuclear nodes of an early regenerator can be
replaced by those of a non-differentiating cell without stopping
the course of regeneration (unpublished). In the latter experiment
"non-regenerator" nucleus clumps and renodulates on cue just
as the original nucleus would have done. It therefore appears that
the nucleus simply responds to any demands made upon it by the
cytoplasm without taking the lead in cell redifferentiation, though
of course the macronucleus is essential to primordium formation.
Yet the presence of the macronucleus seems to be necessary for
achieving a state of activation in the cytoplasm, as suggested by
the following experiment. Both feeding organelles and macro-
nuclei were removed from coeruleus and after five hours re-
generation primordium sectors were grafted to them. Normally
the hosts would have been in active regeneration by this time but
now, lacking the nucleus, they behaved exactly like non-
differentiating hosts, causing resorption of the grafted anlage. It
follows that the nucleus is not only very probably essential to
protein synthesis in the elaboration of the oral primordium but is
also necessary for the achievement of the postulated state of
activation in the cytoplasm. Another finding which points to the
same conclusion is that if regenerators with early primordia are
enucleated the anlage are then soon resorbed. Not only is there no
further synthesis of ciliary proteins, or whatever is involved in the
further development of the primordium ; the developing organelles,
in contrast to remnants of those already formed, are actually
taken down and resorbed, so that it appears that the nucleus is
necessary for the maintenance as well as the achievement of the
state of activation.

4. Relation of intact feeding organelles to activation
and inhibition

Because removal of all or of a substantial portion of the feeding
organelles initiates their complete regeneration, it is natural to
suppose that the formed parts had exerted an inhibition on the
production of their like. Indeed, it is clear from the experiments
recounted above that non-differentiating stentors are continually
inhibiting primordium formation because they even cause
resorption of already well-formed anlagen grafted on them. This

ACTIVATION AND INHIBITION: ORAL PRIMORDIUM 145

relationship is common to regeneration in general, whether of
plants or multicellular animals; for it is a general rule in re-
generation and embryological studies that formed parts prevent
neo-formations of their like and so allow the organism to attain
stability and unity of form (see Child, 1941; and Rose, 1957).
That there is specific inhibition between formed and potential
structures can be demonstrated on the cell level in Stentor where
it presents special problems as well as unusual opportunities for
analysis.

The first exploratory experiment in this direction was performed
by Prowazek (1904) when he cut dividing coeruleus in two trans-
versely. If the animals were in an early stage of fission, the half of
the primordium remaining in the anterior fragment was resorbed,
but not in the posterior piece ; yet he was not aware of the full
significance of this simple test. Today we can say that the portion
of the anlage in the anterior fragment was resorbed because of
the presence of the intact feeding organelles, and conversely, that
their absence in the posterior piece permitted the maintenance and
continued development of its section of the primordium.

Weisz (1956) later found that it was sufficient merely to sHce
into an early divider to cause total resorption of the entire primor-
dium. I have also found that a single slice into the cell, merely
removing the tail tip (Tartar, 1958c), or even a too long exposure
to the quieting agent, methyl cellulose, may cause stage i and 2
dividers to resorb the primordium. Even at stage 4 the anlage
may be completely resorbed in the adoral half of dividers cut in
two longitudinally. Early primordium sectors cut from these
dividers, including the mouthparts but not much of the membran-
ellar band, also resorb the anlage when isolated but not if the
original mouthparts are also excised from the piece. The response
of regenerators to cutting Weisz found to be entirely different,
for the primordium is then never resorbed because of injuries. This
point has also been adequately confirmed; following a standard
maximal disturbance in which the regenerator was cut into three
sections and spread out widely, the anlagen were never resorbed
(Tartar, 1958c).

A simple explanation for this difference between dividers and
regenerators is at once apparent. It is not because the division
primordium is uniquely subject to reversal of its development

146 THE BIOLOGY OF STENTOR

(Weisz, 1956), for we have seen that regeneration primordia can
also be caused to be resorbed. The difference lies rather in the
simple fact that dividers have an intact set of feeding organelles
but regenerators do not. Thus if the injurious cut through a
divider is such as to remove the entire feeding organelles or the
mouthparts, then, as in Prowazek's original experiment, the
division primordium is not resorbed. And therefore the simplest
interpretation is that presence of intact organelles is the cause of
resorption. Then, as Weisz himself suggested, in division (as in
reorganization) the primordium site is somehow enabled to escape
the inhibitive action of the existing feeding organelles and to
produce an oral primordium in spite of their presence; and I
would add that cutting injuries in some manner nullify this
delicate escapement, thus enabling the formed parts to re-exert
their full inhibitive force.

Inhibition by the intact feeding organelles would also explain
why dividers do not produce a new primordium at once after
anlage excision, as do regenerators.

The situation in dividers can be duplicated in regenerators by
grafting a new head in place of the one that was removed (Tartar,
1958c). If the regenerator had not yet produced a primordium, it
was prevented from doing so ; or if it already had an early primor-
dium, this was then resorbed (Fig. 37). When tails were grafted
instead of heads, primordia were not resorbed. This is evidence
that the formed feeding organelles exert an inhibitive action on
primordium formation and development. With regenerators
which had progressed beyond stage 2 the effect was not as marked
and only partial resorption or merely arrested development
occurred. But when the primordium was completely resorbed in
recapped regenerators the majority of the specimens later re-
organized. This suggests that complete healing may not have
occurred, with complete union of lateral striping, and thus set the
stage for later escapement of the primordium site in re-
organization.

Similarly, it may be that in division, as in reorganization, there
is some temporary and invisible severance of connection between
the lateral body striping and the feeding organelles sufficient to
break the path of oral inhibition and permit the formation of a
primordium although intact feeding organelles are present.

ACTIVATION AND INHIBITION: ORAL PRIMORDIUM 147

A

Fig. 37. Oral inhibition of primordium development.

A. Stage-2 regenerator is recapped with head from another
stentor. By handling only pendent portion of donor, which is
subsequently excised (a), injury to feeding organelles is
prevented, b: Development is stopped and the anlage resorbed.
This occurs neither on injury, alone, of regenerator nor after
implanting tail poles; therefore a specific inhibition by formed
oral structures. Most specimens reorganized later {c, d), and
some divided instead.

B. When primordium was already at stage 4, it was not
resorbed (a) and served for reorganizational replacement of the

grafted feeding organelles. (After Tartar, 1958c.)

Reorganization and division would then be like regeneration in
that the oral structures may be "self-excised", and if so, the
regeneration response to cutting off the head or mouthparts would
be not so much an adaptive behavior as a gross imitation or
artificially induced performance of something that happens
cryptically in the recurring processes of fission and reorganization.
This in turn would at last answer Gruber's (1885a) question why
stentors should be so capable of regenerating from injuries such as
they are not likely to encounter in nature, as well as explain to a
considerable extent his original conception of the close similarity
between regeneration and division, a point repeatedly emphasized
by later students of ciliate morphogenesis (see Balamuth, 1940).
We need to learn how these formed feeding organelles exert the

148 THE BIOLOGY OF STENTOR

inferred inhibiting effect upon the primordium site. They probably
do not act directly, because the primordium site and anlage are
at some distance from these structures. Moreover, in tandem
grafts the head of the anterior cell effectively inhibits regeneration
in the posterior partner, the head of which has been excised,
though the distance between ingestive organelles and the posterior
primordium site is then abnormally great. Nor do these organelles
give off some '' inhibitory substance ", since regeneration will occur
if the mouthparts are merely cut and separated or the intact head
rotated in place. Not the materials of the organelles but their
proper pattern and relationship to the whole is essential to their
inhibitory effect. Moreover, non-differentiating stentors from
which the mouthparts have just been excised still can induce
resorption of early regeneration primordia grafted to them. The
tendency of the normal primordium site to form anlagen is appar-
ently stronger than that of other loci in the lateral ectoplasm, and
therefore requires a stronger inhibition. This is indicated by the
finding that fusions of six aboral halves promptly regenerate,
whereas anlagen formation in these grafts without normal primor-
dium sites is long delayed if one set of intact feeding organelles is
present (Tartar, 1956a). In contrast, when one set of feeding
organelles is removed from a doublet stentor, the remaining set is
insufficient to prevent, or often even to delay, regeneration in the
" unsaturated " primordium site left on the cut side.

As a working hypothesis it is suggested that formed oral struc-
tures act upon the lateral stripe pattern, with which they are
connected, in such a way as to render this pattern inhibitive of
primordium development. The entire cell-body ectoplasm would
be involved in this inhibition, as indicated by the fact that the
larger the volume of cytoplasm the greater the inhibition exerted.
This state of inhibition could then be transmitted across the
borders of a grafted sector, rendering the included striping in the
patch also inhibitory and producing resorption of the primordium ;
or the state of inhibition could be transmitted in a similar way over
the ectoplasm of an adjoining cell. Conversely, when the head is
excised or the mouthparts removed, oral inhibition is dis-
continued and the pattern of the body striping gradually trans-
forms, with the aid of the nucleus, from a state of inhibition to one
of activation which is to be characterized in the same way. The

ACTIVATION AND INHIBITION: ORAL PRIMORDIUM 149

important point is that the development of the oral primordium
is not a strictly local affair except in morphological terms, that the
entire ectoplasm appears to be a continuum, that every part of this
ectoplasm — even far from the primordium site — can affect the
primordium development by either hindering or supporting it,
depending on the intrinsic state of that cytoplasm.

5. Synchronization of developing primordia

In stentor grafts or complexes of more than one individuality
there is a strong tendency for both or all oral primordia to com-
plete their development together although they may have begun
at different times. This synchronization was first indicated by
Johnson in his observation of redifferentiation in an adventitious
double-tandem monster of coeruleiis. The anterior individuality
had a complete set of feeding organelles but the posterior lacked
the mouthparts. An anlage first appeared in the posterior compo-
nent, then somewhat later primordium formation also occurred
in the anterior component which had no need for regeneration;
but in spite of the difference in the time of their appearances the
two primordia soon fell into phase and developed simultaneously.
This case may therefore be regarded as the first observation of a
regenerator inducing reorganization in its partner ; and it suggested
that in such double systems both parts tend to do the same things
together and at the same time.

Even within a single primordium the parts tend to develop
together when they might do otherwise. Thus if half of an early
anlage is excised there is a compensating growth in length of the
primordium but a difference between younger and older developing
membranelles is not detectable (Tartar, 1957c). Evidently the
older part waits while the growth of the new part is accelerated.
This effect is still more striking in cases in which an original
primordium later extends into a new primordmm site which is
often produced by graftings. This and other examples described
in Fig. 38 show how an anlage extension or an induced re-
organization primordium may differentiate very rapidly in order
to catch up with the first anlage, often apparently cutting short its
growth in length in its haste to develop. Converselv, in the achieve-
ment of simultaneity of development, arrest and delay of one of
the anlagen is often noticeable.

50

THE BIOLOGY OF STENT OR

^3

a be

Fig. 38. Synchronization of primordia within a graft complex.

A. a: Anlage of a stage- 1 regenerator excised and patch with
stage-3 regeneration primordium implanted, b: Stage-3 anlage
arrested while an extension occurs adjacent to wide-stripe area
of host which forms a new primordium, both the latter in stage i .
c: All three anlagen synchronous by stage 4. d: Primordia join,
with metachronal waves of membranelles continuous in direction
of arrows. Implanted anlage, with its extension, forms a V-shape
which undergoes stomatogenesis at the point and a doublet
stentor is formed.

ACTIVATION AND INHIBITION: ORAL PRIMORDIUM 151

These are only a sample of many observations (unpublished) in
which synchronous development within the same system by arrest
of one primordium, or acceleration of the other, or both occurred.
But if the phase difference between the two anlagen is great,
simultaneity cannot be achieved and the older primordium
overtakes the much younger one, causing the system to pass into
a state of morphogenetic inhibition, resulting in resorption or
very incomplete development of the younger anlage (Fig. 38E).

Synchronization of primordia, often involving astonishing
accelerations and delays in development, should have important
implications which are only coming into view. For one thing, it is
clear that each primordium is not given a start and a source of
substrates and a suitable environment to proceed on its own.
Instead, the two primordia are as it were in continuous "com-
munication" with each other though they may be at opposite
sides of the cell. One suggestion that comes to mind is that there
is a competition for substrates which the younger primordia are
able to take up more avidly. But the supply does not seem to be
limited, since induced primordia are formed and present anlage

B. a: Telebiotic with narrow connection has stage-3 anlage
at end from which the feeding organelles were excised, b:
Original primordium at stage 5 ; induced anlage appeared at other
end and developed precipitously to stage 4. c: Both anlagen
synchronized at stage 6 and regeneration with induced

reorganization continues.

C. a: Smaller non-differentiating stentor grafted to stage-3
regenerator, b: Original anlage now in stage 4 and induced
reorganization primordium has developed so rapidly that it is
now in the same stage, c, d: Synchronous regeneration and

reorganization to produce a doublet.

D. a: Stage-3 divider, with mouthparts excised, grafted to
non-differentiating stentor of same size, b: Division primor-
dium develops to stage 4 but becomes crumpled as it is arrested
and waits for induced reorganization to attain the same stage.
c: Synchronous regeneration-reorganization proceeds from

stage 4 onward. The graft complex did not divide.

E. Stage-5 regenerator grafted to non-differentiating partner.
A reorganization primordium is induced {a) in the partner, but
the regenerator continues development and passes out of the
stage of activation. Thereby induced anlage is "overtaken" and
can neither develop to normal length nor produce mouthparts

to replace those resorbed.

152 THE BIOLOGY OF STENTOR

readily extend themselves into new primordium sites. Another
possibility is that somehow the anlage does receive morphogenetic
guidance from the surrounding ectoplasm which acts as a unit,
and that the ectoplasm as a whole gives "information" only one
step at a time, instead of a single command to make a primordium.
However this may be, we see again that the cell makes a strong
attempt to act together in all its parts as a single integrated unit.

6. Activation in reorganizers and dividers

It is natural to suppose that the state of activation which is not
of the nucleus but of the cytoplasm and can be transmitted from
one cell to another, or from a host cell to a grafted patch, is to be
found whenever an oral primordium develops. Therefore re-
organizers and dividers should also be in this state. This can be
tested by determining whether they continue to support oral
differentiation in regeneration primordia grafted to them, in the
same way that regenerating cells do. They do. Reorganizers
support regeneration primordia (Tartar, 1958b), likewise for
dividers. But in the case of dividers the intact feeding organelles
seem to exert a greater effect than in reorganizers and the mouth-
parts usually have to be excised if a grafted anlage is not to be
resorbed along with the host's, following the injury of cutting.
Conversely, both division and reorganization primordia are
resorbed when grafted to non-differentiating cells. We may con-
clude that oral primordia arising under any circumstance require
the same type of cytoplasmic as well as nuclear support.

7. Rerouting the oral primordium

This state of activation, or readiness to support primordium
development which is common to all re-differentiating stentors,
points to a basic similarity of dividers, reorganizers, and regener-
ators which has often been remarked. It was Gruber who first
noted that oral regeneration is accomplished through the formation
of a lateral primordium like that appearing in the normal course of
division. The unique characteristic of fission is not anlage
formation but the development and constriction of a division
furrow, and this aper9u of Johnson's is amply confirmed by the
fact that dividers as early as stage 3 can proceed to complete
separation after the primordium is excised. Otherwise, events in

ACTIVATION AND INHIBITION I ORAL PRIMORDIUM 153

regeneration and division are very similar. In both, the macro-
nuclear beads coalesce. Causin even described an instance of
temporary division of this compacted nucleus in a regenerating
stentor, though this is probably exceptional. In both there is
mitotic division of the micronuclei (Guttes and Guttes, 1959).

Reorganization is obviously similar to regeneration in that a new
set of feeding organelles is produced while the original individuality
of the organism is retained, and the accompanying nuclear changes
are similar. Schwartz (1935) commented on the resemblances
between reorganizers and dividers: in both there is oral primor-
dium formation in the presence of an already complete set of
feeding organelles; and in reorganizers as in dividers there can
occur the mitotic multiplication of micronuclei as well as an
increase in the number of macronuclear nodes. A basic similarity
in division, reorganization, and regeneration was recognized by
Weisz (1949a) who conceived of these processes as alternative
responses to a graded series of stimuli increasingly forceful in
their extrinsic character.

In all three programs of morphogenesis, oral primordium
formation occurs and a basic similarity is best demonstrated by
the fact that the anlage can be rerouted to serve other ends than
that for which it was originally "intended". In other words,
morphogenesis can be Preprogrammed"; for it can be shown
that a stentor is not irrevocably set upon one course from the start.
Johnson, for example, described a case in which a reorganizing
coerideiis seemed to have transformed itself into a divider. At first
the primordium ran all the way forward to contact the original
membranellar band, as is characteristic of reorganizers, but then
a secondary contractile vacuole developed and the anterior portion
of the anlage was resorbed, whereupon the cell divided. I read this
report with some scepticism because I have never seen resorption
restricted to one section of the primordium; although I have
observed three instances in which a coeruleiis which should have re-
organized divided instead. These were from regenerating animals,
the primordium and neighboring Ectoplasm and endoplasm of
which had been excised so that they had already suffered a con-
siderable reduction in volume, which is supposed to preclude
division (Weisz, 1956). A new anlage was then produced so
rapidly that no stripe multiplication occurred and the resulting

\^-

J R A R y

'V

154

THE BIOLOGY OF STENTOR

frontal field and head was much too small in relation to the size
of the cell. This disproportion is almost always the occasion for
reorganization, defined as the spontaneous replacement of major
portions of an intact set of feeding organelles by a new one ; but in

Fig. 39. Rerouting of the oral primordium.

A. Reorganization anlage presumably serving for division.
a: Primordium excised from stage-3 regenerator, h: New
anlage promptly formed in line of heal, without stripe multipli-
cation, c: Hence regenerated frontal field and head are too small.
d: This disproportion is usually the stimulus to reorganization
and an anlage altogether like that of reorganizer (not D-shaped
as in divider) is formed, e: Yet animal may divide instead of
reorganizing, even though its original volume was considerably
reduced by excision.

ACTIVATION AND INHIBITION: ORAL PRIMORDIUM 155

these instances the stentors divided instead (Fig. 39). One may
suppose that for some obscure reason the primordium which
developed for the purpose of reorganization was used instead for
the division, nevertheless, of animals considerably smaller than
the maximum size.

It is also possible, but again not indubitably demonstrated,
that regenerators can be converted into dividers. In experiments
already described, all of the membranellar band but none of the
mouthparts or all of the mouthparts but none of the membranellar
band were removed from larger animals, with the result that
division almost invariably occurred, thus representing cases in
which an operation which would ordinarily be expected to incite
regeneration led to fission instead. The primordia did appear at
first exactly like those of regenerators, but subsequently the
anterior ends of the anlagen bent to the right as division was
accomplished in the typical manner (b).

Regenerating stentors can easily be converted into reorganizers.
When a complete head is grafted to a regenerator to replace the
one which was excised and the regeneration anlage is not resorbed,
then this primordium finds attachment to the intact membranellar
band and the preexisting mouthparts are resorbed as they are
replaced (see Fig. 37).

Dividers are frequently converted into reorganizers by many
types of operation which permit the continued development of
the division primordium but somehow block furrow formation.
Causin reported that mere transfer of dividing stentors onto a
slide often resulted in their undergoing reorganization instead,

B. Regenerator becomes a divider, a: If either membranellar
band (as shown) or the mouthparts only are excised, a regenera-
tion primordium is called for and appears {b) but serves for
division instead {c) even though ablation decreased cell volume.

d: Proter regenerates proportionate oral structures later.

C. Divider becomes reorganizer. a: Stage-3 divider grafted
to oral half of a small, non-differentiating stentor. b: No
induced primordium in small partner ; anlage of divider develops
to stage 6 without commencing division or reorganization, c:
Anlage finally used to reorganize larger animal ; later the mouth-
parts of the partner were also resorbed though having no
reorganization primordium from which to replace them. Next
day the specimen performed double regeneration-reorganization.

156 THE BIOLOGY OF STENTOR

and this was confirmed by Hetherington (1932b). When mid-stage
dividers were greatly disturbed by cutting and spreading them out
in a clover-leaf pattern, the separated parts then healed together
and the primordium continued to develop, but almost all of the
specimens reorganized instead of dividing (Tartar, 1958c).
Division was usually only thereby postponed, and successful
fission with a new primordium generally occurred some time later.
Likewise, when heads of early dividers were circumscribed and
rotated 180° on the body, there occurred an initial partial regression
of the division primordia, probably due to the cutting injury as
such, after which the anlagen continued developing but moved
forward instead of posteriorly and the animals reorganized instead
of dividing. In three cases a stage-3 divider was grafted to a small
non-differentiating stentor or to the oral longitudinal half of such
an animal. The primordium served only to replace the feeding
organelles of the divider and, surprisingly enough, the mouthparts
of the partner were also gradually resorbed though there were
none to take its place (Fig. 39c). When early dividers were grafted
to regenerators, regeneration proceeded on one side while re-
organization instead of division occurred on the other as already
mentioned in connection with dividers failing to induce division.
Even when two stage-3 dividers were grafted together in homo-
polar parabiosis they reorganized doubly instead of dividing. It is
clear that furrow formation is not determined from the beginning
of the division process but is inaugurated much later, so that shifts
from division to reorganization are possible.

Likewise dividers can easily be changed into regenerators.
Causin had at least one case in which he cut off the anterior right
hand corner with membranelles of an early dividing coeruleus the
primordium of which then served for regeneration instead of
division. When primordium sectors were cut and isolated from
dividing animals, these pieces made no attempt to divide but
used the anlagen to regenerate the missing ingestive organelles
(Tartar, 1958c). It was also shown that if the head or feeding
organelles are excised from dividing cells they then regenerated
instead, postponed fission with the formation of a new division
primordium usually occurring sometime later. In conclusion it
may be said that in their beginning phases fission, reorganization,
and regeneration are more similar than different, so that a stentor

ACTIVATION AND INHIBITION: ORAL PRIMORDIUM 157

embarked upon any one of these courses is not irrevocably deter-
mined to pursue no other.

The experimental analysis reviewed in this chapter demonstrates
that stentors alternate through at least two cell states involving
some pervasive aspect of the cell. A prolonged state of inhibition
of oral primordium formation which maintains the status quo of
the formed organism alternates with another and more transitory
state promoting redifferentiation of feeding organelles which
prevails during regeneration, reorganization, and division.
Moreover, the stimulus to regeneration appears to be another
condition separable from the subsequent activation, transmissible
to any grafted partner regardless of size and resulting in its
parallel reorganization. Whether there is a " division state " or
predisposition to fission which is Hkewise transmissible in fusion
complexes is still obscured by contradictory evidence.

Besides clarifying the question of division, we next need to know
in what parts of stentor these cell states reside. Present evidence
suggests that the nucleus is not involved, since macronuclei can
be exchanged between regenerators and non- differentiating
stentors without effect. A nucleus or some nucleus is essential for
primordium formation and development but this organelle
apparently does not take the lead. Enucleated non-differentiating
stentors are as capable of inducing anlagen resorption as nucleate.
Preliminary tests in which stentors bereft of the endoplasm show
the same inhibitive influence suggest that cell states reside in the
cortical layer. If so, these states characterize the entire ectoplasm
because the effect is quantitative and depends on the relative sizes
of the joined stentors. Every part of this or some other pervasive
feature of the cell may be involved in the cell states of activation
and inhibition and somehow capable of affecting what occurs
locally at the primordium site, as indicated by the quantitative
relationships.

After those parts of stentor which "carry" or take the lead in
establishing cell states are identified, the next step according to
conventional procedure would be to obtain a biochemical
characterization of the changes in these parts. It is natural to
suppose that intercellular transmission in grafts would occur via
the semi-fluid endoplasm which flows and mixes between the two

158 THE BIOLOGY OF STENTOR

partners. For instance, this endoplasm, during primordium
formation and development, might be charged with an unusual
amount of RNA in support of the extensive synthesis which then
presumably occurs. But if it is the semi-solid ectoplasm which is
involved, the transmission would be more probably something
like an electrical excitation of a more novel character. Moreover,
the synchronization of developing anlagen without indication of
competition for substrates suggests that not one substance or
state of excitation is concerned but a series, paralleling the stages
in primordium development.

If nourished stentors are continually undergoing structural
growth and not merely stretching or extending the distance
between their formed parts as seems evident in the case of the
lateral striping (see Fig. iia), how is this possible when during
the same period synthesis and morphogenesis in oral redifferentia-
tion is being so effectively inhibited that a stentor can even cause
the regression of the primordium of another stentor and even after
that anlage has been well-started ? Yet extensive nodal increase in
the macronucleus does seem to require the state of activation or its
final phases, since this increase occurs only during the last stages
of primordium development and it appears that reorganizers may
instigate anlagen formation in order to accomplish this nuclear
increase. Evidently the different parts of the pattern of cortical
differentiations, however, constitute a very precise responding
system in respect to growth ; and this is also indicated by specific
resorption of extra mouthparts or in a disproportionately long
membranellar band, when all other parts remain apparently
unaffected.

In the resolution of such problems relating to cell states in
Stentor I think we may expect interesting discoveries which may
in turn prove relevant to cell differentiation in general.

CHAPTER IX

PRIMORDIUM DEVELOPMENT

An acutely felt omission in our data on Stentor is the lack of
silver-stain or electron micrographic studies of the developing
oral primordium. We have therefore no idea of what happens on
the level of fine structure during the most dramatic act of cyto-
differentiation. Yet much can be said in simple description of the
forming anlage and its relation to the pattern of lateral striping.
This relationship is two-fold: first, some of the ectoplasmic
stripes and bands adjacent to the primordium join with it to
complete the integrated parts of the ingestive apparatus, and
second, the anlage arises in definite correlation with the topo-
graphy of the cell surface.

I. Normal location and development of the primordium

At its inception the oral primordium seems to violate the
cortical pattern because it makes its appearance as a break in the
ectoplasm, cutting across the striping. The unpigmented rift sug-
gested to Johnson that the primordium originates in the endoplasm
and breaks through to the surface. He further argued that the
ectoplasm is too thin to supply the materials needed for this
extensive elaboration, besides being too highly differentiated to
participate in such ** embryonic" formations. In the related
Bursaria triincatella, Schmahl (1926) also found that the primor-
dium gives the appearance of breaking through the ectoplasm, yet
his cross-sections clearly showed him that the anlage lay entirely
in the surface. On the basis of other ciliate studies (see Lwoff,
1950) it is probable that the anlage is formed entirely in the ecto-
plasm and requires cortical derivatives such as kinetosomes for
its composition. Villeneuve-Brachon (1940) described accumula-
tion of kinetosomes in the early primordium, and these, in Stentor
as in the related Fabrea, seem to arise by multipHcation of granules
in the existing ciliary rows.

159

l6o THE BIOLOGY OF STENTOR

If the anlage has to cross the striping, it is apparent that the
structural components of the clear stripes would have to be
sundered and the pigment granules pushed aside to make room
for the primordium. Much simpler would be merely to have the
stripes spread apart and permit the anlage to form parallel to them ;
and this does occur in Folliculina ampulla, in which the primordium
follows the contour of the stripes (Faure-Fremiet, 1932). In this
and other forms (see Lwoif) one could speak of a '' stomatogenic
kinety ", if all kinetosomes of the primordium arise in connection
with a single kinety. But even in the related Semifolliculina,
Andrews (1923) described the oral primordium as cutting across
the lateral striping. Also like Stentor, there is in the Ophryoglenids
no single kinety which produces the primordium, according to
Mugard (1947). In the latter there seems to be good reason for this
type of development. Where the primordium site cuts across the
lateral stripes these are bent and a small section cut out of each
kinety, the sections then combining to form the anlage. This does
not occur in Stentor, and there are certainly more membranelles
produced than kinetics which are cut by the anlage. Although the
anlage of Stentor may come to lie largely parallel to the lateral
striping, even those of the '' French school " did not maintain
that it arises from a single ** stomatogenic kinety" (Chatton and
Seguela, 1940). All we can say at present regarding the elaboration
of the membranellar band is that kinetosomes appear from some-
where in the rift provided for them, sprout cilia, and align them-
selves in a series of parallel rows to make the membranelles. This
corresponds to Schmahl's descriptions of Bursaria in which he
observed first single cilia with separated basal bodies later coming
together as membranelles.

Normally, the primordium always appears on the ventral side
of the cell at about one-third the distance in contracted animals
between the mouthparts and the posterior pole. This precise
localization of the anlage was emphasized by Schuberg (1890)
who correlated it with local differences in the pattern of lateral
striping. Thus the primordium appears in what he called the
ramifying zone, a zone of abbreviated striping bounded right and
left by bands which do run from pole to pole. Schwartz (1935)
however has pointed out that as the primordium increases in
length its anterior end may overstep the left boundary stripe, so

PRIMORDIUM DEVELOPMENT l6l

that there is nothing magically restrictive about the ramifying
zone as far as primordium formation is concerned.

In fact, Morgan (1901a) soon found that oral regeneration
occurred readily enough even after the normal site of the anlage
was removed, and Stevens (1903) confirmed this by showing that
in longitudinal aboral halves lacking this site entirely the oral
primordium appeared in the line of heal. Faure-Fremiet then
posed explicitly the question of whether, if the primordium always
appears at the same place in stentors, there is some specialized
potential restricted to this area; but his student Causin (1931)
likewise found that the normal primordium site could be com-
pletely eliminated without preventing regeneration. Therefore
there are not localized potentialities for oral differentiation in one
region of the cell. This point has been amply confirmed by later
students of Stentor, including myself. Weisz regarded the oral
primordium as arising from a single stomatogenic kinety next to
the left boundary stripe of the ramifying zone. He stated (1953,
1954) that not only tiny fragments but also pieces larger than half
the cell can be cut which do not regenerate because they lack the
specialized kinetosomes of this meridian ; although reporting that
longitudinal aboral halves can regenerate and that in division the
primordium bends so that it eventually touches the left boundary
stripe, from which it follows that the anlage originates away from
this stripe. That the ectoplasm is virtually totipotent throughout
as regards oral differentiation will become even clearer as our dis-
cussion proceeds, and this is not contradictory to the fact that the
oral primordium usually appears in a certain place.

Stages of visible change in the anlage in regeneration have been
defined (Tartar, 1957c) and are altogether comparable to those
of division (Fig. 40). The first sign of primordium formation as
seen in coeruleus is a scooping of the pigment granules to each side
as a rift crosses about 10 granular stripes (Moxon, 1869). A groove
with slightly projecting flanges is evident at later stages in cross-
sectioned view, as shown. Stripe multiplication also occurs with
the splitting of granular bands both immediately above and below
the primordium. The primordium extends from both ends,
cutting across more stripes posteriorly, the anterior end reaching
forward. Now the anlage has a glistening appearance, presumably
due to cilia growing out from kinetosomes included within it.

1 62

THE BIOLOGY OF STENTOR

3^ <"

W '-' <u

P (U be

0 bo 03

— I 4-> C

^-, en a

.2 c -M

•• o ^

^f §

S 5 <u

"S "^ •

Jh O N

(U 4->
?5 ^

b£).2 c

ai »H C

U (D ^

1 S o

5 a «^

qj =! ^ ^^ ^ ^ ^

<u

B o

.2 -^ - ^

^ ^ 5 go

en ^ 4->

£ oj C ex 3

TS i3 £

w

C3

c«

;^ C« C
I ^ -

0) c«

^■Si* a»j-3

be

(At

^ O jH X!

«U C =^ 8

13 '::; -O rt^ S^ *^

(U C^

^ s ?^ s

3 O
u
o o

be

c
o

N U

(A

O ,C ;3 J

c <^ •£

C > C <u
be C ^ W)

> - ^ c

bC

<i> S

a 2 ^
t bc c

= .5 C

(U O D *^
> - 3 O

O ^

o "5

c/3 bc^^

O >.T3

Oh (U TJ

(U 03 CI

'ii (n "^

III

-^ a; :3 .^
. +i ^ vw ^
C ^ o D >

a> ^ a-ts Qj bp^

'^ cu ►L'l::; be B

QJ • -^ flj ^ — V ■— ' I

>..^ ^ ?^C^ O ^
<v C^^-'^ti >

^ ^ ^-.& o c;5 be c

<^ ^ - § =^ J3

„ ^ . ^ o -6 •£ I

o

^ °" 5 « o t^

g^

:3
be

u ^ C ^.
-^ Ti ^ ^

03 ^ — ' T3

!-- tJ <U I-l

=^ be o ^, ^ -.
^ 5 .^^ §._^-2 -

bc^ (u o :;:=! ^
5 t; .S t: f«

o "> _,

C .ir 03

(U u u

O ^ R

Jrl <u be

lU C O!

^ «=! t^

P ^

6 M

i^ ^-p-5 6

bco p-t^ i) a; .5 13

<u

t3 -C ^ a; c3 ^ .5 o

<L> Cu • "r 73 ^ be,iH

t^ ^^ S o a> ^

be • be oj j2 c "S

03 »;5 v7 be -^ o

^ .in rt •• •£ c ^ ^

drl ^>;ttH03OH^

M f^-S PQ u^ 2^ £ U

be
c

'c
a

<v
S

-u

O

PRIMORDIUM DEVELOPMENT 163

This 1 have called stage 2. At stage 3 the anlage has become some-
what longer and the ciha are clearly visible in it but they have not
attained their final length. Even so, as Johnson observed, the cilia
begin beating on their first appearance, at first slowly and without
coordination. Oral cilia then attain their definitive length, and
transverse stripes in the rift indicate the formation of membranelles
which now beat slowly but in metachronal rhythm. The membran-
elles are at first very close together and they will produce a
membranellar band longer than the primordium as the distance
between them later increases (Stevens, 1903; Schwartz, 1935).

As now deployed, there runs immediately to the right of the
membranellar row a pigment stripe and to the right of this a clear
band (the border stripe of Schuberg), and further to the right
another pigment stripe, which three will form the border stripes
of the frontal field. In the meantime, in an extensive area to the
right of these, considerable stripe multiplication has usually
occurred with formation of many kinetics separated by very fine
granular stripes, as Moxon first noted. It is these fine stripes which
will form the new frontal field, as well as the lining of the buccal
pouch and of the gullet in part. At stage 5 the posterior end of the
anlage enlarges a bit and begins to make a sharp bend to the right
in the initiation of mouthparts formation. A spiraling ingrowth of
the end of the primordium forms the gullet (stage 6). In this
invagination the terminal membranelles are carried down inside
the cell, as well as the posterior ends of the fine striping, and
shortly a gullet lined with bright refringent oral cilia and pigmented
ectoplasm is produced. It seems Hkely that there are some further
additions at this time to produce the complete lining of the gullet,
but this is not known for sure.

At stage 7 the ectoplasm adjacent to the membranellar band and
just forward of the spiral gullet begins, in coeriileiis, to depress
and form the oral pouch as the anlage starts shifting into its final
position. With further development the entire anlage moves into
its final position, carrying the new stripes with it as the enclosed
frontal field (stage 8). In this migration the primordium which at
mid-stage was roughly parallel to the lateral striping comes to
assume a position at right angles to it and this may involve cutting
and shifting of stripes, as Moxon remarked, as well as compen-
sating growth in length of the stripes below the new mouthparts

164 THE BIOLOGY OF STENTOR

(Schwartz, 1935). As the two ends of the new membranellar band
approach each other the enclosed fine stripes are bent into arcs in
the frontal field.

There are certain modifications of the primordium in different
types of morphogenesis. Speaking teleologically, the primordium
in reorganization needs only to replace the original mouthparts
which will be dissolved, and, as Schwartz emphasized, there is
accordingly less multiplication of fine stripes than in either
division or regeneration. In dividers the anlage, already at stage-2,
may appear semicircular as its anterior end also bends to the right
and cuts across lateral striping, and stripe multiplication may be
observed along both ends (see Fig. 15A). On the contrary, in
regenerators and reorganizers the anterior end of the primordium
usually runs straight forward to the anterior pole or to a pre-
existing adoral membranellar band. The D-shaped primordium
is usually diagnostic of dividing animals. All primordia begin in
stage I at the same site and level of the cell. This was remarked in
reference to regenerators and reorganizers by Johnson. My im-
pression is that it is also true of dividers, Weisz (1951b) to the
contrary.

Earlier students of Stentor in the heyday of the recapitulation
theory saw an evolutionary significance in the lateral origin of the
oral primordium in stentors. Both Schuberg (1890) and Johnson
(1893) regarded this as a return to a more primitive design in
heterotrichous ciliates. Spirostomum, presumably more primitive,
retains the lateral disposition of the membranellar band and
resembles regenerating stentors in stage 6. In Fabrea the mouth-
parts would be shifted half-way forward from their original
posterior location having the appearance of stage 7 in stentors.
In Stentor itself they would eventually achieve the wholly anterior
disposition. Finally, in Folliculinids the highest development
would be achieved, in which there is an enormous extension of
the membranellar band in two folds projecting outward from the
anterior end.

2. Primordium development under abnormal conditions

Such is the normal development of the oral primordium so far
as we now know. Its behavior under unusual circumstances may
give us some insight into the processes involved (Tartar, 1958b).

PRIMORDIUM DEVELOPMENT 165

First, multiplication of lateral striping is not essential for anlagen
formation, as shown when a large regenerator, for example,
induces the precipitous formation of a reorganization primordium
in a small, non-differentiating animal grafted to it. An anlage
appears which cuts across several stripes which show no splitting
whatever and an extensive presumptive frontal field is not pro-
duced (see Fig. 34B). Dissociability of stripe multiplication and
membranellar band formation is further indicated by the reciprocal
process of induced primordium resorption; for when this occurs
the new band disappears entirely but there is no regression of the
newly-multiplied fine stripes and ciliary rows which then contrast
sharply with the neighboring stripes remaining (Fig. 34A).
When a newly forming membranellar band is caused to be shed
by treatment with salt solutions at a relatively late stage in anlage
development, the multiplied fine striping can still be moved for-
ward to form a new frontal field, indicating that these stripes are
not merely shifted passively by the migrating adoral band (Fig. 41B).

The disposition of the original anlage rift at stage i is not
rigidly fixed. It may run almost perpendicular to the striping or
its course may sometimes be considerably canted at about 30° to
the striping. In one case it slanted downward instead of upward
though this did not interfere with normal oral formation (Fig. 41 a).
The primordium opening need not cut across the lateral striping
at all. Thus Stevens (1903) observed that in longitudinal half
fragments the anlage appears in the line of heal which runs from
pole to pole. But this type of development is. shown most clearly
when the primordium, excised from a regenerating animal, is
replaced at once. The line of heal, at first tightly closed, opens to
permit the formation of a new primordium entirely parallel to
the striping (see Fig. 34B). Since such specimens were already in
regeneration, the new anlage develops very rapidly and there is
apparently insufficient time for stripe multiplication. After stage 6,
posterior ends of such replacing primordia curl to the right,
cutting through some of the relatively large preexisting stripes
which are carried forward to form new frontal fields of abnormal
appearance.

The primordium need not develop as a unit or single entity
from the start. When through abortive fission a head-like struc-
ture remains athwart the primordium site, anlage development

1 66

THE BIOLOGY OF STENTOR

Fig. 41. Abnormal primordium developments {S. coeruleus).

A. Anlage slanting in direction opposite from the normal,
satisfactory reorganization nevertheless accomplished.

B. Stage-5 divider treated with dilute sea water to cause
shedding of membranellar band and all but posterior end of the
division primordium. Though the posterior daughter formed
only a gullet, this organelle and the multiplied striping of the
prospective frontal field moved anteriorly to their normal

positions. Both cells then regenerated.

PRIMORDIUM DEVELOPMENT 167

C. Stage-4 divider transected and posterior fragment dis-
carded. In continued ''division" the anterior half of the anlage
was shifted to the posterior end of the anterior fragment, later
moving forward and forming a crescent of membranelles. After
the mouthparts were excised, a regeneration primordium
appeared which bridged the interrupting crescent. A nearly
normal animal was regenerated (though with an extra tube in the
frontal field) either by resorption of the crescent or its incorpora-
tion into the new membranellar band.

D. Island primordia formed when post-oral sector was reversed
180°. Anlagen formation on both sides of the patch. Islands of
membranelles joined with each other and with the second
primordium but complete mouthparts were not formed. After
several transformations the specimen eventually became normal.

E. Capacity for erosion of ectoplasm shown in narrow-loop
primordium. Graft of 4 coeruleiis produced one normal, one
ring- and one loop-primordium. Enlarged view of latter in
second sketch shows "erosion", lifting and buckling of ecto-
plasmic striping enclosed within the membranellar band, leading
eventually to the separation of a bleb of ectoplasm. Only

membranelles and oral pouch were formed

F. Etching of clear band alongside an oral primordium. a:
Stage-4 reorganizer (anterior) grafted in tandem with a stage- 1
regenerator, b: Anlagen synchronized at stage-5, interrupted by
oral remnant of the regenerator and stomatogenesis confined to
the posterior primordium. c: Oral remnant either resorbed or
incorporated. Clear band to right of primordium, apparently
from dissolving of ectoplasmic structures, permits viewing
through the interior of the cell with its food vacuoles to the
ectoplasm on the far side. An elongated singlet produced which

soon reorganized with two anlagen.

G. Dissolving of ectoplasmic striping in presumptive oral
region {x) at stage 5 — observed in experimental animals but may be
an exaggeration of a normal process preceding oral invagination.

H. Divider in stage 2 transected through the anlage and

halves rotated. Essentially there v/as no growth in length of the

sections of the primordium, as if blocked by abutting stripes.

Oral differentiation w^as incomplete in both halves, but the

original mouthparts incurred reorganizational resorption.

I. Development of a V-shaped primordium. These are
formed in addition to the host primordium when a fine stripe
sector {x) is implanted into the back of a regenerator. The point
of the V invaginates to form a good cytostome and gullet though
two membranellar bands are involved. Resorption of subtended
portion of old peristome permits entrance of new fine stripes into
the frontal field.

M

1 68

THE BIOLOGY OF STENTOR

J. Development of a loop primordium. To a stage-3 divider
was grafted a stage-2 regeneration primordium. As shown in
first sketch, the host anlage is developed to stage 5 and the
implanted primordium extended to form two parallel memhra-
nellar bands separated by only a single stripe. The specimen
divided, with the loop primordium going to the proter, resorbing
one of the bands and forming no mouthparts but only a curl.

K. Polarity in early stage-4 primordium. Mid-section was
reversed in situ. Parts did not rejoin, as when merely transected,
but developed separately.

PRIMORDIUM DEVELOPMENT 169

may occur both above and below it with a single mouth produced
only in the lower segment (Fig. 41c). In one unusual instance, in
which the cell was cut in two longitudinally and the halves rotated
180° on each other, an adventitious primordium appeared first as a
series of islands of oral cilia which later connected to produce a
good membranellar band though mouthparts were not formed (d).
These cases suggest that primordium formation is the result of
local episodes of elaboration along its length rather than a
differentiation proceeding from a single center.

Other observations indicate that dissolving or etching of formed
parts may be involved in the later stages of primordium develop-
ment, as if in this way space is provided for the evolving parts.
Thus in Bursaria truncatella (Schmahl, 1926) and in Condylostoma
magnum (Tartar, 1941b) the oral groove seems to be scooped out of
the cytoplasm by an active process of dissolution. Something like
this was seen in the development of an unusual stentor primordium
in the form of a pinched ring. The enclosed striping was dissolved
and cut out as a pendent tongue of cytoplasm (e). In two cases the
breakdown or dissolving of ectoplasmic structures along the whole
length of the primordium was conspicuous, and one of these is
shown (f). This was in a tandem graft of two Stentor coeruleus in
which, at stage 6 in development of the joined primordium, a wide
break in the striping was seen to the right so that one could clearly
see through the transparent plasma membrane into the cell interior,
and the gap was later covered by scattered pigment granules not in
rows. This may have been an exaggerated picture of what happens
when a place is provided for the clear border stripe to the right of
the membranellar band, as well as a demonstration that pigment
granules tend to invade any open or unstructured area of ectoplasm.
The case is also reminiscent of normal events in primordium
formation in Folliculina^ already mentioned, in which a gape or
spreading of the longitudinal striping to a distance of 15/x occurs.
When primordium formation occurs without stripe multiplication
there is sometimes the appearance at stage 5 of a dissolving of the
ectoplasm near the posterior end of the anlage (g) and this might
be regarded as a localized etching of the structured cortex to permit
the inward invagination of the anlage in gullet formation.

Lengthening of the primordium can be blocked by incompatible
striping. Thus when cell and primordium were cut transversely

170 THE BIOLOGY OF STENTOR

and the anterior half of the stentor rotated 180° on the posterior,
there was no extension of the primordium from the cut end and
very short membranellar bands were produced (h). This suggests
that increase in length occurs in all parts of the primordium and
not merely at its ends.

When the stripe pattern is abnormal after certain grafting
operations V-shaped and even looped primordia may be formed
(Fig. 41, I and j). The former can produce mouthparts of normal
appearance, and the latter make an attempt to do so, though the
conditions for invaginations are certainly quite atypical.

These and other types of oral development show that cyto-
differentiation is not so delicately precise a process that inter-
ferences cannot be surmounted. They also demonstrate the
important point made by Driesch that the same result in organic
development can be achieved by several routes, even exceeding the
usual experience of the organism.

3. Determination, or the progressive specification of the
oral anlage

Only early-stage anlagen are resorbable (Weisz, 1956; Tartar,
1958c). Stage 3 appears to be the time of transition, after which
oral primordia become self developing systems resistant to resorp-
tive influences, though some early stage-4 anlagen, not actually
resorbed, developed astomatously or with incomplete mouthparts.
Weisz cut off the tails of late dividers including the posterior end
of the primordium and found that the anlage then produced no
mouthparts. Similarly, Lund (19 17), working with Bursaria
truncatella, stated that minor injury to the anlage of the gullet did
not prevent development of the oral primordium but led to
abnormality in the oral structures produced. He concluded that
" there appears to exist in the anlage of the gullet a definite part
which corresponds to a definite structure in the fully differentiated
gullet ".

These isolated indications led to a comprehensive demonstration
on the cell level of something very much like the determination
of parts in developing embryos (Tartar, 1957c). It is well known
that embryonic anlagen are at first modifiable but later not. This
developmental principle seems to be simply a statement that once
a complex is well underway it cannot be modified and there is no

PRIMORDIUM DEVELOPMENT 171

turning back, which is a rather universal generalization; yet in
the development of the relatively new science of biology it was
very important to find that the elaboration of the organism is not
magical and immutable but is a process which occurs in time and
is subject to some analysis by operative manipulations. In many
respects it can be shown that this principle also appHes to Stentor.

First we shall discuss deletion experiments on the oral primor-
dium. Minimal excisions of parts of the oral primordium, always
necessarily including some of the surrounding ectoplasm and
endoplasm, were performed on regenerating stentors in stages 4
and 5 when the membranellar band is already well formed but there
is still no visible indication whatever of developing mouthparts.
When the anterior halves of such anlagen were removed, develop-
ment continued and complete mouthparts were formed but the
membranellar band was only half its normal length (Fig. 42A).
Re-regeneration then occurred about a day later to produce a set
of feeding organelles in normal proportion to the size of the cell.
If the posterior half or third of the primordium was excised, only a
considerable length of membranellar band was produced and no
mouthparts at all, with re-regeneration now following sooner (b).
Extensive removal of all but the posterior end of the anlage could
lead to formation of a perfect set of mouthparts without any
membranellar band at all (c) ; and removal of only the posterior
tip, at stage 5 or early stage 6, resulted in complete absence of the
gullet although oral pouch and membranellar band were normal (d).
Finally, by removing a penultimate section of the anlage at stage 5
it w^as possible to produce heads in which a normal gullet termin-
ated the membranellar band but the oral pouch was missing
entirely (e). Therefore, it is clear that by stage 4 the oral anlage is
determined and any ablation of its parts results in corresponding
deletions in the organelles formed.

The same operations were then performed on early primordia.
Even when extensive sections of the anlage were removed, the lack
was then compensated by additions to the primordium, increasing
it to its normal length, and conlplete, proportionate feeding
organelles w^ere produced (f). At stage 3 some specimens showed
corresponding defects while others did not, so that it may be
concluded that fixity or determination of the primordium occurs
in late stage 3.

172

THE BIOLOGY OF STENTOR

Fig. 42. Developmental determination of the oral primordium.

Removal of portions of the stage-4 anlage results in corres-
ponding deletions : A. Posterior half produces mouthparts and
short membranellar band. B. Anterior two-thirds forms no
mouthparts, only peristome. C. Posterior fourth forms only
gullet with opening.

Removal of parts at stage 5, before stomatogenesis has visibly
begun. D. Excision of posterior tip of anlage results in absence
of gullet and cytostome, only membranellar band and oral pouch
being formed. E. Penultimate deletion results in absence of oral

PRIMORDIUM DEVELOPMENT I73

Now it is very suggestive that it is just at this time that the
primordium becomes susceptible to sloughing when stentors are
treated with salt and other solutions (see p. 253). This shedding
response indicates that the anlage has become disconnectable from
the surrounding ectoplasm and this in turn suggests not only that
the primordium has become an integrated unit but also that it may
now be isolated from morphogenetic influences emanating from
its immediate surroundings. As will be developed shortly, there
are strong indications that mouthparts are induced at the posterior
end of the anlage by reason of its relationship to the surrounding
parts at the posterior end of the cell.

Parenthetically it should be noted that in one salt treatment of
a stentor with a stage-4 primordium it happened that only a
section of the membranellar band in the region of the presumptive
oral pouch was sloughed, and again there was formed a perfect set
of feeding organelles except that the pouch was missing. Since the
ectoplasmic striping alongside the anlage is in these cases never
affected, it may be assumed that the lining of the presumptive
pouch had not been removed. From this it may be inferred that a
certain section or bend of the developing primordium induces
pouch formation in the adjacent ectoplasm on its right and that
in the absence of this influence this material remains unaltered.

Progressive determination of the oral primordium is also demon-
strated by shifting its location on the cell. When stage-2 primordia
were circumscribed and shifted to the anterior pole they developed
without forming mouthparts (Fig. 43A), but when this operation
is performed upon later anlagen, oral development is then complete.
Early primordia shifted to the posterior end developed adequate
mouthparts and a well-formed though somewhat shortened mem-
branellar band (unpublished). In one case a grafted anlage
developing at the posterior pole even showed an additional gullet
formation (b).

It therefore appears that the normal surroundings of the

pouch only, with adequate gullet attached to a long peristome.

F. Removal of large portions, whether anterior or posterior, of

stage-2 primordium has no eff"ect ; remaining section of the anlage

elongates and forms complete set of feeding organelles. (After

Tartar, i957c.)

174

THE BIOLOGY OF STENTOR

posterior end of the primordium act upon it to cause formation of
mouthparts and that when the anlage is shifted away from this
environment such induction is missing. Later primordia in stage 4
or 5 have already received this influence and are then semi-
autonomous systems capable of complete self-differentiation
regardless of their surroundings.

Fig. 43. Oral induction associated with the posterior pole.

A. Stage-2 primordium of regenerator shifted to the anterior
pole develops no mouthparts but only membranellar band.

B. Stage-3 regeneration primordium implanted heteropolar
on a stage-3 regenerator. Primordium patch slips to posterior
end where it forms good mouthparts and an extra oral pouch (y)

and gullet (x) in addition.

4. Induction of mouthparts formation

As just indicated, there is accumulating evidence that the
posterior end of the cell in Stentor has an inductive action on the
end of the developing membranellar band which causes it to
invaginate and form mouthparts. Perhaps the first indication of
this relationship was in experiments in which a sector bearing the
primordium site was reversed in situ and the original mouthparts
excised to initiate regeneration (Tartar, 1956b). An oral primor-
dium then appeared in the reversed patch but mouthparts now
were formed at both ends (Fig. 44A). Formation of the normal
mouthparts at the original posterior end of the anlage may be
regarded as due to the influence of its own surroundings of
"posterior" ectoplasm, and that of the additional formation at the
other end as being produced by an influence of the adjacent tail-
pole, passing across the graft and affecting the originally anterior
end of the anlage. This experiment has been confirmed by Uhlig

PRIMORDIUM DEVELOPMENT

175

Fig. 44. Stomatogenesis in relation to the posterior pole.

A. Primordium site reversed in situ and mouthparts excised.
Regeneration anlage forms best mouthparts at its posterior end
(above) but oral formations are also induced at the other end by
adjacent posterior pole. Polarity of primordium prevails as
metachronal beat of membranelles proceeds from posterior to

anterior (arrows). (After Tartar, 1956b.)

B. Wide stripe patch implanted transversely in mid-region
develops a primordium but forms no mouthparts as does the
other anlage in the normal primordium site extending to the

posterior end. (After Tartar, 1956b.)

C. Lateral graft of an extra tail-pole induces abortive mouth-

parts formation in the middle of the primordium.

(1959) who emphasized that the symmetry of the induced mouth is
always that of the inducing pole, i.e., the anterior end of the primor-
dium shows a double curvature which results in its coiling in the
normal direction. This was also shown in fusion masses in which
for some reason mouthparts formed at the "wrong" end of the
primordium yet coiled in the normal direction (Tartar, 1954,
Fig. 11). The pattern of the posterior ectoplasm therefore deter-
mines not only that coiling and invagination shall occur but also
the direction taken. Uhlig also noted that in these double-ended
formations the original polarity is functionally dominant, for the
metachronal beating of the membranelles originates at the normal
mouth and progresses without interruption to the other.

Several other observations gave evidence that mouth formation
depends on the geometric relationship of the anlage to the topo-
graphy of the cell. Ectopic primordia, developing in primordium
sectors grafted transversely across the lateral striping of the host
did not form mouthparts (Fig. 44B). Sometimes when primordium
sites or primordia were reversed in place there was not a bipolar

176 THE BIOLOGY OF STENTOR

differentiation, as described above, but instead no mouthparts
were formed at all. In these cases, as in the absence or incomplete-
ness of oral formation in stentors grafted in complex, random
orientation, the failure of oral differentiation may be attributed to
the mutual cancellation of polar gradients (Tartar, 1956b).

Starting from these impHcations, Schwartz's student, Uhlig
(1959, and unpublished thesis), has pursued this matter in a
demonstration of morphogenetic gradients in Stentor coeruleus,
suggestively similar to those which have been postulated for the
cleaving sea urchin egg. To mention only two of his experiments,
Uhlig found consistently that when primordia were grafted trans-
versely across the axis of the cell they never formed mouthparts,
suggesting that although the anlage arises in this manner the
assumption of its later anterio-posterior position is for more than
*' historical " reasons, namely, to align it with a morphogenetic
gradient which will insure its complete and proper development.
He also found that when an extra tail pole was grafted laterally
alongside a developing primordium this had the effect of inducing
an additional if somewhat incomplete formation in the middle of
the anlage (Fig. 44c). The interpretation that this induction may
be due to the operation of some gradient steepest at the posterior
pole will be discussed in a later section on polarity (p. 202).

Hence there is good evidence that the oral anlage is induced to
form mouthparts by its normal surroundings. Once this inter-
action has taken place and although there are yet no beginnings
of mouthparts, the primordium is then determined and can
develop completely regardless of where it is placed.

5. Repair, mending, and joining of primordia

The oral anlage can endure drastic cutting injuries without
total blockage of development (Tartar, 1957c). For example, a
stage-3 primordium was transected in many places yet produced
an apparently complete set of feeding organelles, though these
were later replaced by a new set (Fig. 45A). Usually, however, this
operation, or the comparable one in which the whole length of the
anlage is slashed through several times with the point of a glass
needle, generally did not prevent the formation of a good mem-
branellar band but the mouthparts were lacking (b), especially
when later-stage primordia were used. As in continued develop-

PRIMORDIUM DEVELOPMENT

177

Fig. 45. Rejoining in oral primordia.

A. After cross-cuts, a: Stage-3 primordium of a divider
transected 6 times, b: Anlage continues developing in normal
time, severed parts rejoining as the animal reorganizes instead
of dividing, c: Perfect but tiny mouthparts formed, possibly

178 THE BIOLOGY OF STENTOR

ment after deletion of parts, these tests show that the oral primor-
dium need not be complete or remain always continuous for the
full differentiation of its separate parts.

It is equally clear that in these interrupted primordia there is a
strong tendency for the parts to rejoin. This association is perhaps
best shown by tandem grafts in which the ends of two separate
primordia were made approximate (unpublished). Two stentors
could be grafted together in homopolar telobiosis without any
disturbance of the lateral striping or injury to the primordia, which
were brought into alignment. In almost every case in which the
anlagen were in mid-stage development, they fused together as a
continuous membranellar band (c). A complete set of mouthparts
was formed only at the posterior end, although almost always at
the point of joining there appeared an accessory oral pouch (see
also Fig. 4 if).

We have therefore learned much of how the oral primordium
develops under both normal and abnormal conditions, but how
the precise and elaborate feeding organelles are guided to their
perfection remains a mystery.

because one section (x) which may have been pushed into
heteropolar orientation was not incorporated.

B. After splitting, a: Stage-5 reorganizer with anlage sliced
through three times, b: Reorganization proceeds, and a well-
formed membranellar band is produced but no mouthparts. c:
Specimen regenerating because differentiation was incomplete.

C. Joining of tandem primordia. a: Tandem graft of two
stage-3 regenerators cut just beyond the ends of the primordia.
b: Anlage join though not originally touching, development
continues, and a space clear of lateral striping (x) develops
alongside as in Fig. 4 if. c: Complete mouthparts formed only
at posterior end, only an extra oral pouch (y) being produced at

mid-level. (After Tartar, 1957c, in part.)

CHAPTER X

THE PRIMORDIUM IN RELATION
TO THE STRIPE PATTERN

We return now to the site of the primordium to learn what
pecuHarity this region may have that it should serve as the place
where the new feeding organelles always normally originate.

I. Nature of the normal primordium site

Schuberg in 1890 had already described a special geometry for
this area which he called the ramifying zone. There he noted
especially that the lateral striping does not run all the way to the
posterior pole and his figures clearly and correctly show that in
this region the granular stripes are the narrowest of any on the cell
and the ciliary rows correspondingly close together. Causin (1931)
and others have clearly seen and depicted these differences in the
striping but it is not uncommon to find pictures of Stentor which
completely ignore this distinction. That the oral primordium
always appears at a definite position on the cell was Schuberg's
main point. It soon became apparent that the narrow granular
stripes are found in the region posterior to the mouth because the
wide stripes to the left have been split to permit the interpolation
of new clear stripes with their kinetics and attendant structures.
The primordium site is thus also the region of stripe multiplication,
and indeed both processes run simultaneously during anlage
formation

Morphologically, the ventral area may be characterized as the
place where the oldest and broadest granular stripes meet the
newest and narrowest in a locus of sharply contrasting stripe
widths. This asymmetry of the lateral striping is found in all
species of Stentor. That the kinetics are not equidistant in other
ciliates, and close together in the region of oral formation, may be
the case in some, though this is by no means obvious. In the related
Folliculina, Faure-Fremiet (1932) could find no ramifying zone

179

l8o THE BIOLOGY OF STENTOR

and described the pigmented stripes as not being graded in width.
Yet in Paramecium the oral anlage does appear near the junction
of two differently patterned areas (Ehret and Powers, 1959).

2. Production of supernumerary primordia

If the crucial feature of the primordium site lies in something
correlated with the visible appearance of contrast between wide
and narrow granular stripes, then it should be possible to elicit
primordium formations in atypical loci by creating such areas of
contrast by operative manipulation. This has proved to be the case,
for a wide spectrum of experiments has shown that primordia
appear wherever and however wide- and narrow-stripe areas of
the ectoplasm come together to create a locus of sharp contrast in
stripe widths (Tartar, 1956a, b, c).

First it was shown that the primordium site need not be in the
normal position in order to produce an anlage and that a single
animal or simple coeruleus graft complex can produce and develop
more than one oral primordium. Thus when an extra primordium
site cut from one stentor, and not necessarily carrying any macro-
nuclear nodes, was grafted into the back of another animal and
regeneration induced by excising the mouthparts, anlagen appeared
simultaneously in both normal and ectopic sites and a doublet or
bistomial stentor was always produced (Fig. 46A). It was then
found that when only the sector characterized by fine pigment
stripes in which the primordium first appears at stage i is
implanted, an extra primordium still appears and it forms on the
side of the graft where the narrowest stripes of the implant lie
adjacent to wide stripes of the host (b). If such a patch was reversed
and implanted heteropolar the extra primordium then appeared
on the other side where the contrast in stripe widths was now the
greatest (c). This anlage soon assumed the polarity of the host, but
on forming the gullet the posterior end curled to the left instead
of the right and an incompletely developed set of mouthparts of
reversed asymmetry was produced. In the controls in which sectors
bearing wide stripes were grafted into the back, or wide stripe
area, no extra primordium formation occurred.

Taken together, these experiments are enlightening. The suture
as such, produced by grafting, is not the cause of anlagen forma-
tion. Instead oral differentiation occurs onlv if and where an area

PRIMORDIUM IN RELATION TO STRIPE PATTERN l8l

characterized by fine striping lies adjacent to an area bearing wide
stripes. The fine stripes need not be to the right of the wide, as is
normally the case, but can lie to the left in an arrangement which
is just as effective. Yet in the reverse arrangement the direction of
coiling is reversed, as if the anlage always bends away from the
wide stripes and into the fine, even though this results in a primor-
dium of reversed asymmetry. And finally, it is evident that the
wide- and fine-stripe areas are able to interact even though
heteropolar.

One of the most convincing demonstrations of the correlation
between primordium formation and locus of contrasting stripe
widths (l.s.c.) was provided by splitting the fine line area of the
primordium site by introducing a narrow sector bearing wide
stripes. Now there were three loci of stripe contrast, the original
primordium site and on each side of the implanted patch, and on
regeneration three anlagen were produced, one in each Ls.c.(d).
In further development the two primordia on the left usually
joined to form a V-shaped anlage which might not form mouth-
parts. It should be added that control tests showed that neither a
mere splitting of the fine-line zone nor the implantation there of
an additional fine-line sector had the effect of producing super-
numerary primordium formations.

Many other types of graft were made to produce juxtapositions
of wide- and fine-stripe areas in all possible combinations, and
these are also illustrated in Fig. 46. They showed that primordium
formation is always correlated with this juxtaposition but is
independent of the orientation of the striping; for in addition to
the modifications already mentioned, the contrasting stripes may
lie at right angles, or abut end to end homopolar or heteropolar.
This shows not only that the two types of area can interact regard-
less of orientation but also implies that the developing oral cilia
and their kinetosomes are autonomous in their precise alignment
into membranelles, because normal membranellar bands were
always produced regardless of the^ disposition of the adjacent
ectoplasmic striping. For example, when anterior halves of
regenerators were rotated 180° on their posterior halves (g)
membranellar differentiation which was to all appearances normal
occurred at the new l.s.c. where the lateral- striping lay at right
angles to the primordium instead of roughly parallel to it.

1 82

THE BIOLOGY OF STENTOR

tiilll Ipii liiHlt Isfii lii ^ li

lii lill! ilil ^iiiiiiiii! liiiii tm m

Fig. 46. Experiments establishing empirical correlation between
primordium formation and loci of stripe-width contrast.

A. Sector bearing primordium site grafted into back of the
host. On excision of mouthparts, primordia are formed both in

the host's and in the implanted site.

B. Sector of fine-stripe grafted into wide-stripe area pro-
duces extra regeneration primordium on the side with the finest

stripes.

PRIMORDIUM IN RELATION TO STRIPE PATTERN 183

Such cases also illustrate another point. In the normal primor-
dium site the anlage appears in and across the fine-stripe area near
that part of the cell which carries the broadest pigment stripes.
Broad and fine stripes as well as the kinetics between them are of
course homopolar and presumably in intimate continuity since the
wide-stripe zone is continually transforming into the fine-line zone
by stripe sphtting. If there is interaction between the two areas,
this would seem to be an action at a distance and not something

C. Same, implanted heteropolar. The anlage now forms on
the other side where the stripe contrast is greatest. Initially, the
ectopic anlage bends at its posterior end, but later the polarity
of host predominates and mouthparts are induced at the other

end, incomplete and of reversed asymmetry.

D. Fine-line zone of host split by an implanted sector of wide
striping. Three anlagen are formed in correspondence to the
three loci of stripe contrast : one in the host primordium site {x)
and one on each side of the graft. Adjacent anlagen join to form

V-primordium which may or may not produce mouthparts.

E. Pair of stentors grafted by wound surfaces from removal
of the heads develop primordia in the normal sites but extending
around the suture where wide abut narrow stripes in heteropolar

orientation.

F. Head of one stentor replaced by fine-line zone of another,
grafted at 90°. Anlage extends from primordium site into the

newly created l.s.c.

G. Anterior half rotated 180' on posterior. Anlagen appear
in separated halves of the primordium site and are joined by
extension which may run halfway around the specimen where

wide stripes abut fine stripes end-to-end.
H. Left half rotated 180° on right. First sketch shows anlage
formed in normal primordium site and extended to new l.s.c.
where fine stripes meet wide-stripe area of the other half. Second
drawing shows third primordium induction in minor l.s.c.
produced on the opposite side.
I. Wide-stripe patch implanted transversely into fine-line
zone of host develops an extra primordium, which however forms
no mouthparts, probably because too distant from posterior pole.
J. Aboral half, folded upon itself, develops regeneration
primordium where the widest stripes lie next to their narrowing
prolongations.
K. Summary diagrams showing primordium formation
(indicated by bar) to be correlated with loci of contrasting
pigment stripe widths but independent of the orientation of
those stripes. (After Tartar, 1956a, b.)

f^y^- «o4nv.

184 THE BIOLOGY OF STENTOR

immediately taking place between the materials of the " last "
fine stripe and the ** first " wide stripe. In certain graft combina-
tions one obtains what may be called the locus of stripe contrast
in the strict sense of the term. For instance when the anterior half
is rotated on the posterior, wide stripes abut fine stripes but
careful examination shows that wide and narrow bands do not fuse
together even though they are homopolar, and this may also be
true of the associated ciliary rows. In heteropolar telobiotics (e) it
is even more obvious and probable that joining does not occur, as
a definite suture persists. When patches are grafted parabiotically
but in reverse orientation, the fact that they usually retain their
original polarities and slip on or creep away from each other, is
indicative that heteropolarity of adjacent striping prevents the
most intimate structural union, although the line of heal may
appear quite perfect. Now, in such cases the anlage seems to
appear within the suture and therefore at an l.s.c. in the strict
sense rather than in the sense of mere adjacency, as in the normal
primordium site. In either type of formation it is clear that there
is something about the juxtaposition of wide- with narrow-stripe
areas which locates primordium formation. Extra anlagen can be
produced at will in an almost mechanical way which is not teleo-
logical, for the Stentor may form primordia which it could better
do without. Moreover, the number and length of oral formations
is clearly not limited by some " critical metabolite " sufficient
only for one. Both repeated and multiple regeneration are possible
for this cell.*

3. Abnormal primordia correlated with abnormal striping

If primordium formation is reUably correlated with loci of
contrasting stripe widths, then this should also be manifested in
aberrant stentors with unpremeditated misalignments of the
patterns of cortical striping, leading to primordia of very abnormal
forms. A sampling of such cases is shown in Fig. 47. Perusal of
these sketches will show with what persistence all loci of stripe
contrast are filled out with primordia, often exceedingly bizarre.
Even though the l.s.c. may be very contorted or of extraordinary

*According to Uhlig (i960) the size of the anlage is also directly corre-
lated with the degree of stripe contrast at the primordium site.

PRIMORDIUM IN RELATION TO STRIPE PATTERN

85

A 1

Fig. 47. Abnormal primordia.

A. Odd specimen found in culture and showing how the
anlage follows the contortions of the locus of stripe-width

contrast.

B. Elongate specimen from aborted fission now regenerating
with extensive primordium formations in long l.s.c.

C. Dividing stentor showing how anlage follows contour of
the fine stripes and not the broad, which are considerably

misplaced.

D. Short V-primordium from small, anterior patch of fine
striping forms no mouthparts and merely replaces a portion of

the original membranellar band.

E. Stage-2 regeneration primordium grafted posteriorly into
a stage-5 reorganizer. Regeneration anlage extended and formed
a loop primordium which moved to the anterior end, broke into

the peristome and formed good oral pouch.

F. Reorganizing double stentor with odd anlage in one half
consisting only of an oval of long oral cilia, not organized into

membranelles. Specimen then died.

1 86

THE BIOLOGY OF STENTOR

length, anlagen formation reliably follows its contour. Yet when
the wide stripes were bent and the fine stripes not, the primordia
was normal and straight, confirming that the anlage comes from
the fine-line zone. Neighboring sectors of fine striping put in
wide-stripe areas usually result in V-shaped primordia. If short
and far anterior, such anlagen contribute only a section of mem-
branelles to the regenerant ; if long and extending posteriorly, well
formed mouthparts are generally produced at the angle, even
though this involves the co-operative coiling of two membranellar
bands in a way that does not naturally occur (see Fig. 411). When
a small patch of fine striping becomes surrounded by wide-stripe
areas the patch is encircled by a new membranellar band as a
continuous ring. Especially when in the form of a loop, such anlagen
still attempt to form mouthparts toward the posterior end but
these are never very complete (Fig. 41J). When in the form of
small rings, they contribute at most a section of membranelles
enclosing an extra oral pouch. Schwartz informed me (1958) that
the same occurs when small patches of wide striping find them-
selves surrounded by a fine-stripe area, and that this may occur
even though the patch is very small, suggesting to him that the
materials of the primordium probably do not come from wide-
stripe areas. Perhaps the case shown in Fig. 47F was of this origin ;
at least it carries the same implication.

Fig. 48. Cases of primordia (x) curving in the wrong direction:
into the wide stripe area. A. In main portion of a graft of 12
coeruleus reorganizing on the fifth day. Only one of the anlagen
is abnormal and it formed only a short tube and no adequate
mouthparts. B. In a doublet stentor, but in this case the
abnormal primordium apparently formed good mouthparts
notwithstanding.

PRIMORDIUM IN RELATION TO STRIPE PATTERN 187

In four instances it has been observed that the primordium
curved in a paradoxical direction, the posterior end bending
tow^ard the wide-stripe area instead of the narrow. Two cases are
shown in Fig. 48. They remain entirely unexplained.

One of the most remarkable relationships between the stripe
pattern and oral anlage development is found in cases of reversed
asymmetry. For these are always correlated with a reversed
position of wide- and narrow-stripe areas (Tartar, 1956b, c). My
best case is shown for the first time in Fig. 49. This was from a

Fig. 49. Reversed asymmetry in »S. coeruleus. After rotating
left half on right and anterior half on posterior, stripe pattern
reconstituted with stripes graded in opposite direction from
normal. The primordium appears in the l.s.c. but coils into the
fine-line zone which is now to the animal's left. Macronuclear
chain is also on reversed side from normal. Fair but incomplete
mouthparts formed and specimen reorganized four times
without achieving an adequate oral differentiation.

Specimen of coeruleus which had been '* drawn and quartered ",
i.e., the anterior half was first rotated 180° on the posterior, then
the animal was cut longitudinally and the left half rotated on the
right. Naturally, this operation resulted in great disturbance of the
cortical stripe pattern and when realignment was achieved the area
bearing widest stripes lay to the right instead of to the left of the
fine-line zone. Correspondingly, the anlage coiled in the reversed

l88 THE BIOLOGY OF STENTOR

direction but it was never able to form a complete set of mouth-
parts even though it undertook 5 successive regenerations. Every
time the primordium persisted in coiling in the " wrong "
direction and the oral differentiation could not be perfected. The
character of the stripe pattern therefore not only determines where
the primordium will be formed but also the direction of its
asymmetry. Still to be explained is the barrier, possibly on the
level of molecular asymmetries, which blocks the completion of
cytodifferentiation. Other ciliates are able to produce what appear
to be complete mouthparts of reversed asymmetry (Lund, 191 7;
Faure-Fremiet, 1945a; Tartar, 1941b; Yagiu, 1952), yet they are
incapable of feeding, possibly because the ciliary organelles beat
in the wrong direction, and hence reproducing lines cannot be
estabHshed.

4. Primordium formation in loci of minor stripe contrast

One naturally asks how great the difference between two ecto-
plasmic areas must be in terms of visibly different pigment stripe
widths in order to occasion primordium formation. Longitudinal
aboral fragments lack both the widest and the narrowest stripes,
but they give evidence that stripe widths are graded in an orderly
manner all the way around the cell because in the line of heal the
bands on the right are slightly narrower than the ones on the left,
and it is here that the primordium always appears. During and
after anlage formation there is multiplication of stripes in this
region and a normal primordium site is regenerated. When two
such fragments are grafted together in homopolar parabiosis there
should be two minor l.s.c, and if regeneration occurs rather
promptly two primordia are accordingly formed and a doublet is
produced. One such combination was made with very narrow
fragments in an attempt to eliminate all stripe differences, but one
primordium did finally appear at 19 hours in a minor l.s.c. which
contained no fine striping (Fig. 50A). It would appear that any
stripe difference, however minor, is sufficient to locate the place
where the primordium will break through.

As a rule, but not always, the regeneration of longitudinal aboral
halves is delayed as compared with whole animals in which the
mouthparts have been excised. Causin regarded this delay as due
to the necessity for first regenerating a typical primordium site.

PRIMORDIUM IN RELATION TO STRIPE PATTERN 189

Fig. 50. Consequences of minimum contrast in granular
stripe widths.

A. Graft of halves of two stentors with fine-stripe areas
removed. After delay of 19 hours two primordia finally formed

in the sutures with very minor stripe contrast.

B, No regeneration without stripe contrast. Aboral side with
medium-width striping is isolated by a cut but left attached to
oral portion so that the latter, in swimming will draw out the
aboral fragment and keep it from folding. The parts soon
separate. Last sketch shows fragment which failed to regenerate,
presumably because it lacked a l.s.c, though it lived and was

active four days.

but primordium formation occurs before. Weisz (1951b) ascribed
the delay to time required for the transformation of the ciliary
row next to the widest stripe into a stomatogenic kinety, but there
is no evidence for a key kinety in normal oral differentiation.
Instead, it may well be the case that minor l.s.c. are quantitatively
less potent in exciting primordium formations and therefore do
so more gradually, and further evidence for this surmise will be
given in the next section. If so, a quantitative time factor would
be available for investigation in connection with anlage induction
by l.s.c.

An attempt has been made to p^roduce nucleate longitudinal
fragments which are so narrow as to contain no significant contrast
in stripe widths (Tartar, 1956c). A longitudinal cut was made far
to the right of the central axis of a stentor but the two parts were
left joined at the tail pole. This was in order that the larger portion
by active swimming would draw out the smaller and prevent it

IQO THE BIOLOGY OF STENTOR

from folding upon itself with resulting stripe distortions (Fig. 50B).
The pieces soon pulled apart but not before straight-line healing
had occurred. In 4 cases, the narrow fragments did fail to form
primordia though they had no mouths, retained sufficient macro-
nuclear nodes, and survived and were active for about 5 days. It
is therefore possible that in an almost completely symmetrical
system there may not be sufficient difference or anisotropy of
pattern for oral differentiation. Such specimens presumably were
entirely capable of producing primordia, since much smaller
fragments do, but apparently failed to do so because of the
absence of any guidance in where to produce them. The attempt
to produce anisotropic systems by grafting together patches of
fine striping from the center of the cell and comparable patches
of wide striping, however, was not successful; always sufficient
l.s.c. remained and primordia were formed at these places.

5. Competition among loci of stripe contrast; regeneration
and obliteration of primordium sites

Although the number of primordia is usually equal to the
number of loci of stripe contrast, this correspondence is apparently
modified by competition among primordium sites. Thus it usually
happened that in grafts of two longitudinal aboral halves in which
membranellar band remnants remained and delayed the onset of
regeneration a primordium was formed in only one of the minor
l.s.c. and a single stentor resulted, as if one site became dominant
over the other (Tartar, 1956a). This effect could work either way.
Two aboral halves were grafted with a complete primordium site
and regeneration was delayed; two primordia were produced,
instead of one in the major primordium site. This was explained
by noting that in tardy regeneration one of the minor l.s.c. had
time to regenerate something like a normal primordium site with
major stripe contrast. When a number of aboral halves were
grafted together with a single normal primordium site, the latter
produced the sole initial anlage and only one set of mouthparts
was regenerated. In later reorganizations the minor l.s.c. had their
effect in multiple oral differentiation, and eventually the original
primordium site disappeared (Fig. 51 a). This decline of one
primordium site appears to have been by a shifting forward and
gradual absorption of the fine-line zone, but in most cases the same

PRIMORDIUM IN RELATION TO STRIPE PATTERN I91

is accomplished simply by a widening of the narrow striping so
that the contrast with neighboring areas disappears (see Fig. 58B).
Stentor graft complexes in general often show a waxing or waning
of supernumerary l.s.c, always tending eventually toward the
single form.

Fig. 51. Resorption and formation of loci of stripe-width
contrast, a : Graft of 6 aboral halves without fine striping plus
one primordium site, b: Initial oral regeneration only from the
site with its maximum l.s.c. c: Reorganized now from three
primordia as fine striping multiplies in two sutures with minor
l.s.c. Original fine-line zone being resorbed. d: Re-reorganized
from the two new l.s.c. only, original l.s.c. nearly obliterated.
(After Tartar, 1956a).

6. Exceptions

The correlation between loci of contrasting stripe widths and
primordium formation is not without its puzzling exceptions.
Primordia have on rare occasions been observed to be formed
w^here no l.s.c. was evident, and still less frequently the primor-
dium appeared at some place other than the good primordium site
w^hich w^as present. These exceptions w^ere so few that almost all
of them can be presented in Fig. 52. In two cases of doublish or
mixed up stentors, as explained in the caption, the single or
secondary primordium appeared far from the major locus of stripe
contrast (a and b). Another case showed a primordium in the
primordium site but also another on the opposite side of the cell
where there was no significant stripe contrast (c). Other cases
showed primordium formations or extensions in regions where
the stripe widths were apparently uniform (d, e, f, g). A special
case was a double primordium with a single kinety or clear stripe
separating the two halves which also appeared in an area without
significant stripe contrast (h).

192

THE BIOLOGY OF STENTOR

Fig. 52. Exceptions to formation of primordium in association
with a locus of stripe-width contrast.

A. Anterior half of a stentor, with intact feeding organelles,
was grafted to two minced stentors. Next day two division
anlagen were formed, one in the l.s.c. and another (jc), paradoxi-
cally, amidst medium-width striping.

B. Specimen resulting from rotating anterior half 90° on the
posterior developed primordium only in the minor l.s.c.

C. Fifth day reorganization in a regenerator to which an
additional regeneration primordium had been grafted.
Secondary anlage (x) appeared where a primordium site had

been but which no longer showed contrasting stripe areas.

D. Regeneration in posterior tip cut from a stage-2 divider,
the anlage appearing in a site of negligible contrast in stripe

widths.

E. Anterior aboral corner fragment of a coeruleus bearing
anlage without significant l.s.c. but showing some branching of
striping. Later normal primordium site was formed by further

stripe multiplication.

F. Tiny fragment with regeneration primordium developing
amidst a few uniform granular stripes. Stomatogenesis did not

occur and the specimen died 2 days later.

PRIMORDIUM IN RELATION TO STRIPE PATTERN 193

Although primordia form at sutures where wide-stripe areas
He heteropolar to fine striping (Fig. 46H), local reversing of the
wide-stripe half of an l.s.c. completely blocked anlagen formation
and no regeneration occurred though the specimens lived for a
week or longer (i).

These apparent exceptions may point up the fact that the
correlation between primordium formation and loci of contrasting
pigment stripe widths is a purely empirical one. It is probable
that the granular stripes as such have nothing whatever to do
directly with primordium formation. They contain no kinetosomes
as possible progenitors of the basal bodies of the primordium.
Instead of narrow-stripe areas we might just as well have spoken
of close-together areas, referring to the fact that the clear stripes
with their kinetics here are not so far apart and their opposite
would then be " wide-apart areas " (cf. UhUg, 1959). At stages i
and 2, in fact, the anlage arises wholly within the fine-line zone
which is not an area of stripe contrast, but the point is that this
place is near a wide-striped area and that experiments show this
to be significant. All we can say is that there is a condition which
is usually associated with the juxtaposition of areas bearing wide
stripes with areas bearing narrow which is conducive to primor-
dium formation. In the exceptions to the rule this crucial condition
may well be present though not in its normal association with
contrasting stripe widths. Yet the whole question of interaction at
the l.s.c. still remains to be explored.

It is attractive to regard the wide-stripe areas as inducing
primordium formation in fine-stripe areas adjacent or near by.

G. Two stentors grafted at right angles. Anlage develops in
the only intact primordium site but extends in suture toward
remnant of membranellar band along locus without contrasting

stripes.

H. Regeneration primordium in narrow sector grafted to a
divider. At stage shown the division anlage is developing, while
the grafted one extended to form two membranellar bands
separated by a granular stripe an^ surrounded by uniform
medium-width striping. (Further development shown in Fig.

41 J-)
I. Wide-stripes alongside primordium site reversed in place
and no primordium formation occurred though the specimen
survived for a week without mouthparts.

194 THE BIOLOGY OF STENTOR

Conditions of embryological induction seem to be fulfilled: a
fine-stripe area remains morphologically inert as ventral ectoplasm
until it is brought into association with a wide-stripe area. The
latter itself never produces further elaborations, but in its associa-
tion with a fine-stripe area there are produced the oral structures
of Stentor.

Whatever the difficulties to be resolved as we learn what actually
happens at the primordium site, the concept of the l.s.c. is a useful
guide. It explains why a graft complex produces more than one
primordium — because it has more than one primordium site or
l.s.c. — and this is not explainable in terms of either cytoplasmic
or nuclear volumes. It explains why any nucleated fragment cut
from locations far from the normal primordium site can neverthe-
less regenerate: because of the graded stripe widths around the
cell, it is almost impossible to produce a piece which on healing
will not bring stripes into juxtaposition with other stripes that are
not so wide and hence produce a sufficient anisotropy to occasion
primordium formation.

It is therefore remarkable that on the cell level in Stentor we
find something very much like induction as manifested in the
embryogenesis of amphibia. In both cases there is the evocation
of a major elaboration determining the principal axis of the
organism — neural tube in salamander and feeding organelles in
Stentor — around which a new individuality can be organized (see
Fig. 55d). This evocation in both can be brought about by the
juxtaposition of certain parts, and is followed by a regionalization
or secondary induction, which in amphibia determines which end
of the tube will form the brain and in stentors is represented by
mouthparts formation under the influence of the posterior end of
the cell. What the significance of this striking parallel may be we
cannot yet say, but the consequences of the fact that induction
need not be intercellular could be of considerable theoretical
importance.

CHAPTER XI

POLARITY

Polarization as a graded " difference " is probably a precondition
for the achievement of persisting organic form, for we cannot
imagine how potentiahties for development could begin in an
entirely anisotropic system. Although specific form is not explain-
able in terms of polarities, which are almost universal in the organic
world, the guidance of development is traceable thereto; and
polarity may well be intimately involved in the first stages of
differentiation itself (see Bonner, 1958). For radially symmetric
organisms like Acetabidaria, cellular slime molds, and higher
plants, antero-posterior polarization may be sufficient, but Stentor
is asymmetric in the position of its mouthparts and the pattern of
the lateral striping. In the ciliate we may therefore expect to find
transverse as well as axial gradients in some intimate property of
the cortical cytoplasm ; and in addition, the structural elements of
the ectoplasm, which persist in fragments and in whole cells in
reorganization and division, have a built-in polarity and asymmetry.
The importance of polar differences in explaining form has been
emphasized above all by Child (1941). Consciously or not, the
tradition he established has continually been drawn upon. Child
also included Stentor in the scope of his investigations, yet his
findings in this context are here considered in a different chapter
because they seem, for the most part, to be more indicative of
structural differences. In the autonomous differentiation of stentors
there is better evidence for gradients than in responses to external
agents.

I. Fixity of structural polarity

There is abundant evidence that structural polarization charac-
terizes the formed components of the ectoplasm. As already
described, ciliary row s or kinetics are intrinsically polarized because
they follow the general rule of desmodexy in ciliates : the fiber or

195

196 THE BIOLOGY OF STENTOR

fibers connecting the cilia in Stentor are always to the right of the
ciliary meridian so that one cannot turn a kinety upside down and
have the over-all pattern remain the same. Endomyonemes or
M-bands also show graded differences in tapering widths and in
density of lateral connections. Polarity may even characterize the
granular stripes, for in pigmented forms grafted in heteropolar
orientation there is always a white line or space where granules
are not continuous at the place of abutment. Since we do not
conceive of a polarity in these granules themselves, it follows that
the ectoplasm in which they reside may itself be polarized, though
little differentiated otherwise.

Balbiani (1893) recognized that original polarities were retained
in Stentor fragments as shown by the direction of their swimming;
and Prowazek (1904) made the same claim on the basis that folded
longitudinal halves eventually draw themselves out in correspon-
dence with the original polar axis. Causin (1931) found that
triangular fragments cut from the middle of the cell retained their
polarity although the " bulk axis " was at first at right angles to
this, but Weisz (1951b) provided a more convincing demonstration.
He cut stentors in such a way that the future site of the primordium
was bent around over the anterior end of the cell, or conversely,
part of the general striping was bent around so that its forward end
pointed backward toward a much-shortened primordium site.
When anlagen appeared they followed the striping, as did the
general reconstruction of shape, in complete disregard of the
" bulk axis ". It is in fact difficult to conceive how a fluid endo-
plasm could have an axis at all. The main point is that intrinsic
polarity persists in the striped ectoplasm, no matter how oriented.
Weisz's inference, that the polarity of the so-called stomatogenic
kinety determines the polarity of the oral primordium and that of
the entire cell cannot be the case, however, because the primordium
first appears at nearly right angles to the striping and at a consider-
able distance from the kinety in question, and because reversing
the whole sector containing the primordium site does not result
in reversing the polarity of the entire cell.

For intrinsic polarity within the ectoplasm is best demonstrated
by altering the orientation of parts of the cell or separating parts
and turning them around. The simplest response is that the
disarranged parts shift into homopolar alignment. If a patch is

POLARITY

A.

into

B.
C.

G

Fig. 53. Adjustments correcting heteropolarity.
When primordium site is implanted transversely it rotates
harmony with axis of the host, striping of the latter

accommodating.
Left half rotated 180° on right, but parts eventually come

into homopolar alignment.
Same operation performed on doublet stentor leads to
homopolar alignment of the two individualities.

D. Fine-line zone grafted heteropolar into host with mouth-
parts removed. In this case regeneration anlage formed only at
the host primordium site, possibly because the reversed patch

was gradually resorbed in situ.

E. Wide-stripe sector implanted heteropolar often shifts
forward or posteriorly, being resorbed or isolating itself from

the host pattern.

F. Early primordium may fail to develop even when grafted
to a regenerator in same stage of development if implanted

heteropolar.

G. Alternatively, both early primordia may be resorbed but
when re-formation occurs there is no anlage induced in the

reversed primordium site.

H. Reversed primordium-site sector commonly leads to

formation of three individualities, the lowest in the figure

being of reversed asymmetry. (In part after Tartar, 1956b,

1957c, 1958b.)

198 THE BIOLOGY OF STENTOR

grafted transversely across the striping of the host it will rotate in
correspondence with the host axis (see Fig. 44B), and even when a
stentor is cut in two longitudinally and the halves rotated 180° on
each other they often rotate back into their original alignment
(Tartar, 1957c), as shown in Fig. 53. Grafted pairs could shift into
homopolar orientation from any initial arrangement (Tartar,
1954). The original polarity of the components is obviously
retained and becomes the basis for their extensive movements
with reference to each other in the reorientation (see Fig. 28B).

Another response which sometimes occurs when patches or
sectors are completely reversed is that the misoriented piece
becomes resorbed in place (Fig. 53D) (Tartar, 1958b). Commonly
the ectopic patch creeps towards the anterior or posterior end of
the host where it is gradually resorbed (e). A less drastic expression
of this way of resolving the conflict in polarities is observed in the
suppression of such pieces, as when early primordium sectors are
grafted heteropolar on to regenerating hosts and the anlage then
fails to develop, or when the induction of a secondary primordium
fails to occur in an additional primordium site implanted in
reverse (Tartar, 1958b) as we show in (f) and (g). These cases
may be of great interest for their implication that the instigation
and support of primordium development involves geometric
relationships in the entire cortex and are not solely the result, say,
of a substance like RNA being released within the cell and affecting
formative loci regardless of how they lie.

Reversed sectors may not be resorbed but creep to one end of
the host and establish new individualities (Tartar, 1956a) clearly
demonstrating the autonomous polarity of the implant (h).
Alternatively — and most intriguing for the problem of polarity —
the patch may remain in place and produce an astonishing
disturbance of the form and morphogenesis of the cell (Tartar,
1956b). In Fig. 54A is shown one case of four in which regeneration
never occurred though the animals survived 6 days — striking
instances of inhibition of regeneration apparently due to polar
conflicts ; and (b) is an example of the transient chaos which may
develop before polar discrepancies are resolved. Nuclear nodes
also are often abnormally located, indicating that the overlying
ectoplasm guides the movements of the macronucleus. From these
extraordinary disturbances it may be inferred that, although the

POLARITY

199

polarity is intrinsic within each part, there is an interaction of some
sort by which heteropolarity may lead to extensive disharmonies
beyond the original misalignment.

Fig. 54. Disturbances in heteropolar systems.

A. Primordium sector of stage-4 reorganizer reversed in situ.
Absence of stomatogenesis associated with posterior end of anlage
lying now in the frontal field. The inverted patch {x) apparently
was gradually resorbed but no regeneration occurred during

7-day survival. Nuclear distribution abnormal.

B. Stage-2 regeneration primordium sector grafted hetero-
polar to regenerating stentor in same stage. Both anlagen were
resorbed; then two new ones produced the incomplete oral
diflferentiations shown in the second sketch as the shape became
grossly abnormal. Specimen is re-regenerating with single

primordium.

Polar conflicts may be resolved by the larger part becoming
dominant (Fig, 55A). A compromise may result in heteromorphosis,
in which a secondary polarity is responsible for the formation of
an extra set of feeding organelles but the lateral striping has,
throughout, the polarity of the major portion of the specimen so
that the secondary oral differentiation is of reversed asymmetry (b).
These forms are however less frequent and less well-developed in
Stentor than in other ciliates. When the major mid-section of the
cell is reversed, all parts retain their original polarities and
multiple formations occur (C).

Shifting the head to the posterior end does not result in reversal
of polarity. The most frequent result (unpublished) was that a new
set of feeding organelles eventually regenerated at the original
anterior end of the major cell body, and the displaced head became

200

THE BIOLOGY OF STENTOR

Fig. 55, Observations concerning polarity.

A. Heteromorphic specimen from abortive fission of a
divider grafted to a regenerator, showing tendency toward

resorption of conflicting part.

B. Heteromorph with continuous striping and therefore
reversed asymmetry in the smaller part, consequence of hetero-
polar implant (later separating) which set up a secondary polar

axis.

C. When major portion of stentor is reversed (head and tail
trade places) all parts retain their polarities and mid-portion

regenerates separate head (x) and tail (3;).

D. Head excised and grafted to replace tail at posterior end
does not reverse the cell polarity nor prevent regeneration of the

"host" but organizes a new individuality.

the center of organization and growth of a secondary cell shape
(Fig. 55D). From the standpoint of oral inhibition these cases were
also interesting as showing that a displaced set of intact feeding
organelles can much delay primordiiim formation but not prevent
it entirely.

All these experiments so attest the fixity of polarity in every
part of the cortex that one wonders whether reversal of polarity is
ever possible. It would seem that the best place to look for such a

POLARITY 201

reversal is in folded aboral halves in which the striping bends and
breaks to form a new holdfast (see Fig. 27), for in these cases both
posterior and anterior ends of the severed striping meet at the new
posterior pole. Yet such specimens will have to be followed very
closely to determine whether the stripes in reversed orientation
are not subsequently resorbed.

2. Rate of regeneration in relation to the polar axis

A further manifestation of polarity is of course to be found in
the fact that heads are always regenerated at anterior ends of
fragments and tails at the posterior. As in regeneration of metazoa,
the organism can produce either anterior or posterior structures
from almost any level of the body, the choice depending on the
original polarity. Although coelenterates and turbellarians may
produce heteropolar heads on very short pieces, this does not occur
in stentors. Isolated heads do not become heteromorphic (see
Fig. 32) nor do disc-shaped fragments whose longitudinal dimen-
sion is brief (Fig. 25A). Nevertheless, it is conceivable that the rate
of oral regeneration might vary with level of cut, as in flatworms
(Bronsted, 1955). That this is not the case within a single cell such
as Stentor was first shown by Gruber (1885b) who found that
animals, minus the head only, regenerated a new set of feeding
organelles as rapidly as posterior fragments. This was confirmed
by Weisz (1948a), who also found (1948c) that the relative growth
rates of fragments from any region were the same. Contrary to the
experience with stentors, Sokoloff (19 13) reported that middle
pieces of Spirostomum regenerate faster than the ends but I think
this work requires checking.

Although denying an axial gradient in speed of regeneration,
Weisz (1948a) stated that oral regeneration is hastened by the
presence of a holdfast, mid-pieces regenerating more slowly than
posterior fragments. Child (1949) regarded this difference as
probably incidental to the fact that middle fragments have to
accomplish two regenerations, of both head and tail. This could be
tested in other ways, as I have done (unpublished). If the presence
of a holdfast hastens oral regeneration, then stentors from which
the head only is excised should regenerate sooner than animals
from which head and tail are removed, but they do not. On Weisz's
assumption, the former should also regenerate in the same time as

202 THE BIOLOGY OF STENTOR

small tail pieces since both have holdfasts, but I found the
regeneration of the latter to be always slower. Slower regeneration
of oral parts in tail pieces may be the consequence of another
aspect of polarity which will now be discussed.

3. Gradients in head and tail formation

Popoff (1909) had found in abortive fissions of Stentor that even
though the daughters did not separate, a new tail pole with hold-
fast, projecting laterally, was produced for the anterior cell because
the lateral striping had been severed by the fission line. Weisz
(1951b) then showed that foot formation could be brought about
by excisions of post-oral striping, but the nearer the anterior end
the more incomplete and temporary was the pedal diff"erentiation
(see Fig. 26a). Notice that the holdfast forms not merely where
the stripes come together (at the posterior pole) but also differen-
tially along the whole side of the cell, wherever ablation creates a
new terminus of polarization.

Uhlig (1959) confirmed that there is a gradient in tail formation,
highest at the posterior end and diminishing anteriorly. This
gradient is strongest on the ventral side where the oral primordium
is also formed, as shown by the appearance of a secondary tail
projection in this region when anterior halves are rotated on the
posterior (see Fig. 26b). The polar pedal gradient is therefore
involved with the circumferential gradient in stripe widths, since
it is on the ventral side that the locus of stripe width contrast
determines both the location of the oral anlage and the side on
which the new tail-pole will appear.

Now, the polar gradient in foot formation is also coincident with
that responsible for the induction of mouthparts formation. When-
ever, but only when, an end of the oral primordium lies near a part
of the posterior end, or its entirety, are mouthparts produced.
This inductive relationship has already been discussed (p. 202)
but should now be considered further within the context of polar
gradients.

Following the implications of double oral differentiation in
reversed primordium sectors (see Fig. 44A), Uhlig (1959) has
explored this matter thoroughly and concluded that the inductive
action is strongest just beyond the posterior pole, diminishing
anteriorly. Because his detailed report is not yet available, I have

POLARITY 203

supplied sketches from my own experiments (including Fig. 44),
which may therefore be regarded as generally confirmatory.

Fig. 56. Observations regarding induction of stomatogenesis by
the posterior end.

A. In an oral half of a stage-3 divider the anlage was not
resorbed but extended all the way to the posterior pole and
produced no mouthparts, presumably because inducing region

is anterior to the pole itself.

B. Two extra tail poles engrafted led to multiple stomato-

genesis, with complete but ectopic gullet (x).

C. Specimen with two tail poles due to shift of primordium
site produced a stentor with unusually large mouthparts,

possibly due to the double tail.

Primordia far from the posterior end produce no mouthparts
(Figs. 47D, and 26c). That the oral-inductive gradient stops short
of the posterior pole is indicated in Fig. 5 6a, showing incomplete
oral differentiation in a primordium extending too far posteriorly.
It is possible, also, that inductive action may be compounded by
the presence of multiple posterior ends. When three tails were
grafted, double mouthparts were produced in the host (b), and in
another case three posterior poles may have been responsible for
unusually large mouthparts formed (c). Astomatous oral differen-
tiation in large fusion masses may be due to the mutual canceling
of oral induction gradients in these random grafts. A similar initial
astomatous development in isolated sectors bearing division pri-
mordia (Tartar, 1958c) may likewise have been due to the frag-
ments at first containing insufficient polar regions, a situation later
corrected by regeneration of the posterior pole.

204 THE BIOLOGY OF STENTOR

Uhlig also regards the appearance of a primordium or new
membranellar band in connection with the locus of stripe width
contrast as expressing a circumferential gradient in propensity for
anlage formations, and this may be a fruitful way of regarding
these events. Certainly the granular stripe widths, or as he perhaps
more pertinently states, the distance between the fibrous clear
stripes, form an orderly gradient around the cell. Primordium
formation might therefore be regarded as always occurring at the
*' head end " of this gradient, or where the finest pigment stripes
are found. Whatever explanatory virtue the polarity or gradient
concept may have would then be applicable to happenings in this
region. Yet there are some difficulties which still need to be
resolved, for example, how primordium formations at transverse
sutures (see Fig. 460) can be regarded as expressing a gradient.

A harmonious co-operation between the circular gradient
manifested in graded stripe widths and the polar gradient of mouth-
parts induction is, according to Uhlig, necessary for complete oral
development. The former guides the location and longitudinal
development of the membranellar band, later invagination of its
posterior end to form the mouthparts being induced by some
influence having its high point near the posterior end of the cell.

When a stentor divides or is cut in two, there would be, in
Uhlig's conception, readjustment to a new equilibrium in which
the original single polar gradient is converted into two. As inti-
mated above, short tail-pole fragments may therefore be slow in
regenerating because of their need for greater readjustments before
significant polar differences can be re-established. When stentors
are cut in two and rotated so that anterior and posterior stripe
systems cannot rejoin, a conflict between double but homopolar
gradients apparently ensues, which is resolved in various ways to
be described later (p. 227).

Stentors therefore may be said to bear within the structure of
every part of the cortex an antero-posterior and a left-right
polarization. In addition, there is experimental evidence for polar
and circular gradients of paramount importance in the elaboration
of major ectoplasmic organelles.

CHAPTER XII

FUSION MASSES
OF WHOLE STENTORS

Repeated and cumulative grafting of stentors made possible the
formation of relatively huge fusion masses of stentor protoplasm.
These cytoplasmic continuums made from many cells are unique
among biological phenomenon, and their potentialities for con-
tributing to our understanding of the organism have by no means
been exhausted. At present we can at least describe the response
of Stentor when confronted with the problem of organizing a far
greater than normal mass of protoplasm. The same forces of
mending, adhesion, and integration which hold the single stentor
together conspire in masses to make enduring unions, and there
is little indication that pathologies arise which would obscure or
preclude the expression of morphogenetic potentialities.

In fact, Stentor masses often live longer than single individuals
under the same circumstances, perhaps for the reason that larger
aggregates have more substance to draw upon under conditions of
relative starvation. Up to the last day or two of their life, the
masses remain active and apparently healthy, and there is no
reason to suspect that they die from any other cause than starvation.
Large masses may not even suffer from reduced surface in relation
to volume as interfering with exchange of oxygen and carbon
dioxide ; for they take the shape of pancakes about as thick as the
normal cell so that, as in the erythrocyte, every point of the interior
retains a fairly normal access to the surrounding medium. The
problems of these complexes therefore seem to be more morpho-
logical than physiological and they survive long enough to show
much of what they can do.

I. Simple masses and biotypes

We begin with the simplest combinations of only two or a few
more cells, something of the behavior of which has already been

205

206 THE BIOLOGY OF STENTOR

indicated in our previous discussions. Double animals are already
well known in ciliates (see Faure-Fremiet, 1948a) and were
encountered in Stentor cultures by early observers (Balbiani,
1891b; Johnson, 1893; Stevens, 1903; and Faure-Fremiet, 1906).
Such as these can therefore arise in nature. They probably
originate by the incomplete separation of daughter cells during
fission, tandemly joined daughter cells later shifting alongside each
other, often with apposition of feeding organelles and tail-poles. In
some ciliates, notably Colpidium (Sonneborn, 1932) and especially
Paramecium (Calkins, 191 1) growth without fission may continue
and produce monsters or very large multiple individualities. In
general, however, studies on these abnormal forms have revealed
two tendencies: first, that doublets, and to a less extent triplets,
become stable biotypes which can reproduce themselves as such,
and second, the complexes eventually become single individuaHties
again by the gradual integration of their multiple morphologies;
and it is the same in Stentor.

From the chance encounter of these forms, the next step was to
produce them at will. This can be accompHshed by a variety of
means which block the final stages of cell division — to mention
only one, the dilute formaldehyde treatments of Faure-Fremiet
(1945a). Possibilities of experiment were then greatly extended
when it was found that stentors could be fused together by grafting
in almost any number or arrangement desired (Tartar, 1941b).

In the simplest complexes, grafted pairs or 2-masses could form
I, 2, or 3 primordia on regenerating (Tartar, 1954). The number
of sets of feeding organelles produced was called the oral valency.
In the first case, the graft reverted almost at once to single indivi-
duality; in the rarest instances in which 3 anlagen were formed,
temporary triplets resulted. But the great majority of 2-masses
remained double for a long time. This corrected Weisz's (1951a)
first impression that pairs always revert to singles within 18 hours
through the dominance of one partner over the other.

All indications are that the oral valency of small masses is strictly
correlated with the number of effective primordium sites available,
as earlier intimated (Tartar, 1954). Neither total volume nor
amount of nuclear material was determinative. The number of
anlagen produced corresponded with the expected probability
with which, in random grafting, the original sites would remain

FUSION MASSES OF WHOLE STENTORS

207

intact, one would be obliterated, or grafting would produce an
extra juxtaposition of wide- and fine-stripe areas.

At first, multiple sets of feeding organelles often remained
separated by lateral striping, and such complexes were called
doubles or triples. There was a strong tendency for feeding

Fig. 57. Biotypes of S. coeruleus.

A. Doublets regenerate and divide as doublets, forming two
anlagen in correspondence to the two primordium sites or l.s.c.

B. Similarly for triplets.

C. Quadruplet formed by grafting four oral, longitudinal
halves does not persist as such but transforms into transition

disequilibrium forms.

y-

>/.

208 THE BIOLOGY OF STENTOR

organelles to associate around one frontal field as the grafted
animals shifted to produce a normal, homopolar, conical Stentor
shape as persisting doublets and triplets. The latter forms may be
called biotypes because they regenerated and reproduced as such

(Fig. 57)-

One doublet could produce thousands by multipHcation, but
after i or 2 months cultivation there was a gradual reversion to the
normal single form. Triplets also reproduced themselves and they
generally reverted to type in a shorter period, always *' stepping-
down " first to doublets and then to singles. Persistence for a long
time of these biotypes may be related to their bilateral symmetry
and unity of form as expressed, for example, in the presence of
but one tail-pole and holdfast.* Faure-Fremiet (1948a) regarded
the balance between the two halves of a doublet as imposing a
*' structural constraint " on labile transformation back to the single
type ; for in such forms as Leucophrys patiila he found that cutting
injuries or the diminution of one component led promptly to
reorganization as a single individuality. The application of this
principle to Stentor is not immediate because the removal of a
single set of feeding organelles in doublets merely leads to regenera-
tion on the cut side and reorganization on the other, producing
the doublet type again. Yet asymmetric doublets are the most
likely soon to revert spontaneously to the single type.

It is doubtless significant that the quadruplet biotype could not
be produced. This limitation has also been found in other ciliates
(Faure-Fremiet, 1945a). Grafts of 4 stentors could produce tran-
sient quadruplets but these did not persist and quickly reduced the
oral valency. Unlike triplets, quadruplets could transform at once
to giant singles (Fig. 57c) and this was the first indication of the
tendency to reduction of oral valency in relation to the number of
components grafted, which became increasingly prominent as the
size of masses was enlarged.

The problem of organic individuality is confronted when we
ask whether doublets are single or double individualities. They
swim and feed and reproduce in a co-ordinated manner like single
cells, and there is no further evidence that the two sides of a

*Uhlig (i960) reported that the one holdfast in doublets is nevertheless
doublish or larger than normal, and similarly for triplets.

FUSION MASSES OF WHOLE STENTORS

209

doublet contract independently as Balbiani (1891b) first described.
Doublets generally show the single conical shape ending in one
holdfast, but there are two contractile vacuoles, two macronuclear

Fig. 58. Conversion of doublets to singles.

A. Doublet becomes single by removing one of the primor-
dium sites or major loci of stripe contrast. On reorganizing, both
sets of original mouthparts are resorbed and the excised l.s.c. is

not reconstituted.

B. Spontaneous conversion, in which one (x) of the two
original loci of stripe contrast disappears and the specimen
reorganizes singly, also achieving a single macronuclear chain.

C. Isolated head end of doublet shows first a proportionate
shortening of the membranellar bands, then cutting out and
resorption of one of the mouths and obliteration of one primor-
dium site, becoming a normal stentor even without primordium

formation.

2IO THE BIOLOGY OF STENTOR

chains, and two complete sets of feeding organelles. Above all, the
pattern of lateral striping is double, with two primordium sites
or loci of stripe contrast ; and this is seen to be crucial, for whenever
doublets revert to singles there is always the obliteration of one
primordium site, after which all other aspects of the complex
become single. And doublets could be converted at once into
singles by excising one of the primordium sites, even if the
bistomial head was left intact (Fig. 58A).

It was difficult in cultures to catch doublets in the act of trans-
forming into singles, but something of how this occurs may be
indicated in the following. Figure 5 8b shows an asymmetrical
doublet which was in fact not a 2-mass but produced by grafting
a primordium site into a single animal. Such specimens remained
as doublets for several days, but then one of the primordium sites
disappeared as such, either the host site or that of the graft
transforming into uniform lateral striping, for there was no evidence
of stripe resorption. The transformation illustrated in (c) was
instigated in the anterior half fragment of a broad symmetrical
doublet. Reduction to half the original size resulted in the length
of the membranellar bands being greatly reduced in situ until they
became proportionate to the new cell size, but the mouthparts
remained large. One primordium site then disappeared as its
contrasting pigment stripes became of uniform width. While this
was occurring the mouth subtending these stripes separated from
the membranellar band and moved into the frontal field where it
was gradually resorbed. The two bands then joined together and
the final result was a single stentor produced even without the
formation of a reorganization primordium.

Although there is evidently a strong tendency towards unifica-
tion of shape, one may speak of a reversed propensity of sets of
lateral striping to establish separate shapes, as if a complex which
cannot achieve complete singleness then settles on a frank expres-
sion of its multiplicity. Doublets, especially when so oriented as
to have two frontal fields, become double cones or Siamese twins,
and enduring triplets also develop *' cleavages " making them
triple shaped (Fig. 59A and b).

A single animal even converted itself into a double shape when
the tail-pole was bent and directed forward (c). These examples
show again that there is no mysterious unity in the endoplasm and

FUSION MASSES OF WHOLE STENTORS

211

that cell shape is an expression of the cortical stripe pattern,
following its unity, distortion, or multiplicity. In other words, one
never finds a normal cell shape imposed on a grossly abnormal
stripe pattern.

Fig. 59. Formation of multiple cell-shapes.

A. Persistent doublets often show tendency to produce

parallel bodies.

B. The same tendency to "cleavages" shown in a triplet

(posterior end view).

C. Tail folded into wound left by removing the division
primordium. A new tail was produced at the bend (x) and each
pole organized a separate cell shape. The nuclear chain is

relocated accordingly.

2. Adjustments among formed ectoplasmic organelles

Correlated with the reconstitution of the normal stentor shape
are shifts and adjustments of formed feeding organelles and
holdfasts. Figure 60 illustrates the major tendencies.

Separated organelles migrate together, like to like, in spite of
the intervening ectoplasmic striping. In fact, the lateral striping
co-operates or may even produce these shifts by resorptive
shortening between the parts and extension elsewhere. Isolated
mouthparts and membranellar bands may travel a long way to
join with or even break into a major set of feeding organelles.
Stentors in which the left half was rotated 180° and healed securely
to the right nevertheless could sometimes gradually return to the

212

THE BIOLOGY OF STENTOR

Fig. 6o. Adjustments among formed ectoplasmic organelles.

A. Regeneration primordium sector with intact mouthparts
grafted to stentor from which mouth was excised. Graft shifts
its alignment, anlage is resorbed and mouthparts join with
membranellar band to form a complete set of feeding organelles

so that no subsequent regeneration occurred.

B. Grafted patch with wide striping and section of mem-
branellar band (x). The peristomal remnant travels all the way
to the anterior end of the host and is incorporated into the host's
band, even with resorption of a part of that band to permit

entrance.

C. Small stentor, with mouth excised, grafted to another
whose mouthparts were cut in two. Mouthparts mend as
completely normal structure, separate membranellar ring moves
to anterior end of larger animal and is incorporated, no

regeneration following.

D. In parabiotic graft of two stentors the feeding organelles
fuse in spite of intervening striping and parts of both membra-
nellar bands are resorbed to make a single frontal field.

FUSION MASSES OF WHOLE STENTORS 213

normal orientation (Tartar, 1957c) and, remarkably, the same
behavior is shown in operated early sea urchin embryos
(Horstadius, 1950). These shifts are as if like parts exert a strong
** attraction " for each other, and their coming together is an
important step in the unification of a fusion mass.

Selective resorption of parts occurs not only on the lateral stripes
but also within the joined heads. When two sets of feeding organelles
become tightly apposed, first those sections of the two membranellar
bands are resorbed which permit the formation of a single ring
and frontal field. Extra tails are resorbed or sloughed, or they may
lose their separate identities by fusion. In all these precise adjust-
ments between the parts of grafted cells we see the specific acts by
which wholeness is achieved.

3. Larger masses and reduction of oral valency

Grafts of 5 to 100 animals were necessarily of random orientation
and displayed several interesting emergent characteristics which
are shown in Fig. 61.

Most obvious is that grafts of 6 or more animals cannot attain
the unitary shape and giant individualities are not achieved.
Instead, the general impression is that of bas reUef sculpturing, as
if each set of stripes were able to make an individual hump in the
over-all contour. Although Stentor is able to make perfect forms
in tiny fragments, it is apparently unable to cope with a mass much
larger than it would ever encounter in nature. This is not because
such masses are necrotic. Their limitations seem to be morpho-
genetic rather than physiological. Either they represent simply a
self-defeating jumble or the upper size limit to form development
and regulation bears in itself important theoretical implications.
Lillie's '' minimal organization mass " seems to have lost its

E. Similar, showing integration accomplished by resorption
in only one membranellar band. Apparently, parts of the band

are resorbed when they do not subtend lateral striping.

F. Adjustment in a graft complex, showing how just those
portions of the membranellar bands are resorbed which make

for an integrated frontal field.

G. Product of graft of two stentors in early division. Acces-
sory tail-pole and holdfast move posteriorly but are eventually

resorbed (x).

214

THE BIOLOGY OF STENTOR

Fig. 6i. Large fusion masses of S. coerideus.

A. Graft of 12 stentors, heads removed, indicating bas-relief
sculpturing or partial emergence of constituent body shapes.

B. Graft of 14 stentors, regenerated, showing unusually long
garlands of membranelles without formation of mouthparts.

C. 1 5-mass, now organized into a bipolar system and with oral

valency reduced to seven. (After Tartar, 1954.)

significance from the consideration that the limit to size of regenera-
tion is simply that complete animals cannot be made of very few
parts of invariant size, but there may w^ell be a maximum organiza-
tion mass beyond which anything like the typical stentor form
cannot be realized.

Although they do not organize into single giants, larger masses
show a tendency towards unification in the reduction of their oral
valency, number of primordia formed decreasing greatly with the
number of individuals grafted. A 1 5-mass for example produced
only 7 primordia, and a 5 5-mass had between 5 and 10 anlagen in
successive reorganizations. These great reductions in the number
of oral differentiations have yet to be adequately explained.
Perhaps some of the primordium sites join together as one. Or it
may be that in larger masses there is for some reason a competition
betw^een primordium sites, with fewer becoming effective in pro-
ducing anlagen. Partly responsible, too, may be the fact that oral
differentiation favors the upper surface ; for these large masses did
not wheel about through the water but remained on the bottom
always with the same side uppermost.

FUSION MASSES OF WHOLE STENTORS 215

4. Incomplete oral differentiation

When fifteen or more stentors were grafted together there was no
longer adequate mouthparts formation. Primordia were few and
unusually long, forming extensive garlands of oral cilia stretched
across the mass (Fig. 6ib). There was some indication that the
membranelles in these bands were not completely formed, though
this has not been ascertained. But it was obvious that formation of
mouthparts was inhibited. Since induction of these parts is deter-
mined by a normal relation of the anlagen to the axis of the cell,
the presence of numerous cell axes running in random directions
and cancehng each other in their polar influences may be responsible
for the astomatous development of the feeding organelles in large
masses.

5. Absence of fission

Random masses containing more than five stentors never showed
any attempt to undergo fission. This is rather surprising for two
reasons. First, the masses are very large and, although increase in
size is not in itself invariably stimulative of division, one might
expect that a very exaggerated volume could be so. Second,
multiple fission would seem to be the easiest way for a mass to
resolve its difficulties, yet this does not occur. But when masses are
cut into pieces about the size of a normal stentor they promptly
regenerate normal singles, a test which shows that no irreversible
pathology occurs within large fusion complexes. Faure-Fremiet
(1945 a) attributed similar failure in simpler complexes to their
heteropolar arrangement, which permits the establishment of no
single plane of fission. Whatever the reason, the elimination of the
capacity to divide should make the study of fusion masses fruitful
in searching for the basis of fission. In this connection one is
reminded of an hypothesis by Berglas (1957) that cancerous
proliferation might be stopped by capitalizing on the avidity of
cancerous cells, causing their overgrowth to such a size that divi-
sion is no longer possible.

6. Tubes and ciliated vacuoles

In these unique intracellular formations the morphogenetic
capacities of Stentor seem to be extended beyond what is ever
normally expressed. The tubes extend deep into the endoplasm

2l6

THE BIOLOGY OF STENTOR

but usually open on the surface, while the vacuoles are wholly
internal though they may break through the surface later. Both
are lined with apparently normal ectoplasmic structure: pigment
stripes alternating with ciliary rows, and contractility was
sometimes noticed in the tubes.

These remarkable structures were first observed in masses of
stentors (Tartar, 1954). The tubes, at least, can occur in single
individuals. One day I isolated a very abnormal coeruleus which
was apparently the result of an incomplete fission, and on the next
day the cell was seen to be filled with elaborate internal tubules
(Fig. 62A). One tube opened where the mouth should have been
and was therefore like an exaggerated gullet. There seemed to be
other tubes with many convolutions which arose separately and

Fig. 62. Interior tubes and vesicles in S. coeruleus.
A. Front and back views of case apparently from aborted
fission which developed complex system of multiple tubules,
blue-green in color because lined with ectoplasm.

FUSION MASSES OF WHOLE STENTORS 217

opened to the outside through the ectoplasm near the posterior end.
These tubes were blue-green in color and obviously lined with
ectoplasm. In grafted pairs one or two tubes sometimes appeared
adventitiously (b). Sometimes the tubes had a neat opening through
the ectoplasm at both ends (c). Usually they opened near the
posterior pole and extended forward, suggesting gullet formation
in the normal site of oral differentiation. Their appearance may
represent acts of gullet formation entirely dissociated from anlagen
development.

Internal ciliated vacuoles are equally surprising. These were
often found in large fusion masses and may have been due to the
accumulation of water inside. The vacuoles seemed at first to have
structureless walls, but they soon became lined with typical striped,
ciliated ectoplasm, as could easily be demonstrated by slicing

B. Graft of two enucleated stentors which developed a single
tube, opening posteriorly, extending forward through the
endoplasm, and lined with ectoplasmic striping. The tube
contracted and extended with the mass, twisted through a 90°
arc autonomously, and seemed to "breathe" by independently
enlarging and narrowing. Elaboration of the structure in absence

of a nucleus is paradoxical.

C. Tube with neat opening through the ectoplasm at each

end, developed in a 3-mass.

D. Appearance of vesicle in a fusion mass. Initially the
vacuole seems to consist of a simple membrane enclosing fluid.
Later it becomes lined with ectoplasmic structure demonstrated
by ciliary circulation of mass of shed pigment granules within

and by transection to expose ectoplasmic striping.

E. Mass with two vesicles, one of which has broken through
the surface, the collapsed lining becoming continuous with

outside ectoplasm and forming deep, ear-like cavity.

F. Reorganization in Cyathodiniiim . Cortical ciliary
apparatus is resorbed and a new one formed inside as a vesicle,
lined with cilia and endosprits, which evaginates through the
lateral surface and produces a new cell axis at right angles to the
old. In division 2 endocellular ciliary anlagen are formed which

move to opposite side. (After Lucas, 1932).

G. Odd formation of tubes and vesicles lined with pigmented
ectoplasm and resembling an "archenteron" with one opening.
Differentiation of the oral anlage was incomplete. (After

Tartar, 1954).

21 8 THE BIOLOGY OF ST ENTOR

through them. Invariably there was a shedding of pigment
granules into the interior and these clumps of blue-green debris
circulated around continuously in an orderly manner by action of
the cihary lining (Fig. 62D). There was no evidence of oral cilia or
of mouthparts differentiation. Several vacuoles could be present
together in one mass, and individual vesicles sometimes increased
in size as if growing and subjecting the mass to great tension as
indicated by the spherical form assumed. After attaining consider-
able size the vacuoles often broke through the surface and their
ectoplasm became continuous with that of the outside, giving the
appearance of '' ears" because of their depth and folds (e).

In this evagination, as in their origin, the ciliated vacuoles
strikingly recall the unusual mode of cytodifferentiation in
Cyathodinium as described by Lucas (1932). During normal
reorganization and division in this ciliate one or two ciliary anlagen
arise internally, develop cilia projecting into the vacuolar space,
then evaginate to the outside in orderly manner so as to produce a
new ciliation at a different axis for the reorganized animal or the
two daughter cells (Fig. 62F). In both Cyathodinium and Stentor,
development of internal ciliation quite separate from contact with
the ectoplasm poses a test of the hypothesis of the genetic con-
tinuity of kinetosomes. But whether the basal bodies of the cilia
arise de novo^ or develop from division products of the surface
kinetosomes wandering into the interior, would be difficult to
decide.

It is also possible that tubes and vacuoles may have arisen from
bits of ectoplasm thrust into the interior during the process of
grafting stentors. In several instances (unpublished) when I tucked
pieces of ectoplasm inside the cell, tubes and ciliated vacuoles
resulted. This observation is especially interesting as suggesting
that internal ectoplasm can grow and even undergo an orderly
disposition into tubes and spheres. Growth, naturally, would be
from the morphologically inner surface of such pieces. Cannibalized
stentors, though not at first enclosed in food vacuoles, are digested
instead, since their " growth surface " never contacts the endo-
plasm of the predator. And conversely, ciliated vesicles can persist
and develop because they are " turned inside out ".

In a special case, tube and vacuole formation seemed to have
combined in a most unusual mass which showed a structure

FUSION MASSES OF WHOLE STENTORS 219

resembling an " archenteron ", with an '' appendix " and a tube
connecting to the exterior (Fig. 620). Ahhough fusion masses of
stentors become increasingly unable to reconstitute the normal
form, they seem for this very reason to be set free to express
unusual types of cytoplasmic differentiation.

CHAPTER XIII

RECONSTITUTION IN
DISARRANGED STENTORS

CiLiATES are often cited as achieving in complexity of structure
and multiplicity of function the highest elaboration of the cell as
a unit, choosing Epidinium as the ultimate. Stentors, with their
elaborate feeding organelles, complex kinetics, ribbon bundles and
M-bands in the clear stripes, and granular bands of varying width
and taper such that any part of the ectoplasm is theoretically
identifiable with reference to its position in the orderly whole, are
not far behind. Yet in spite of the cogency and high development
of the cortical pattern, stentors can sustain and recover from drastic
disruptions of this exquisitely organized ectoplasmic structure.
Nor is reconstitution accomplished by the easier way of resorbing
existing cortical differentiations and starting afresh, as in Bursaria
truncatella in which excessive injuries lead to encystment followed
by complete reconstruction, according to Lund (19 17). Instead,
the cut up and disarranged parts of stentors largely persist as such
and apparently perform remarkable shifts and reorientations and
rejoinings in a usually highly successful recovery of the normal
pattern of the cell. This performance in fact suggestively parallels
the reconstitution by dissociated sponges and disaggregated tissue
cells of organized, functional units.

I. Minced stentors

The most drastic operation conceivable with Stentor is rather
easily accomplished. The ectoplasm can be cut into as many as a
hundred separate patches by slashing deeply through the surface
of the cell with the sharp point of a glass needle. After many cuts,
large patches will have been circumscribed and ** float " free on
the endoplasm. When these in turn are repeatedly transected, the
needle not only severs the formed structures but also pushes the
patches into gross disarrangements with reference to one another

220

RECONSTITUTION IN DISARRANGED STENTORS 221

and the striped surface comes to have the appearance of fields seen
from the air. Randomness can be increased by first cutting the
cell transversely and rotating the anterior half iSo"" on the posterior;
after healing, then recutting longitudinally and rotating the left
half on the right. Quarters of the cell are thus transposed and
disoriented before the mincing.

In some of the first experiments of this type (Tartar, 1941a, b)
it was found that stentors with two heads or two tails could be
produced from singles, and an analogy was drawn between this
result and the consequence of inverting embr^^os in the 2-celled
stage, by which twins are produced. Weisz (1951a) had found that
excessive cutting injuries in grafting only resulted in death of the
specimen, but evidently the conditions of experiment were not
optimal. Further studies (Tartar, 1956c) revealed remarkable
reconstitutions and allow us to say something of how they are
brought about.

After minceration a stentor has a knobby or fascetted appearance
from the patchiness of the striping, which again substantiates that
the over-all shape of the cell is determined by the arrangement of
the ectoplasmic striping (Fig. 63A). Within a few hours the patches
begin reorientation, with their striping becoming more or less
parallel. Although this point could not be tested, it seems likely
that the arrangement of pieces becomes homopolar, like so many
tiny magnets. The gradual nature of this process suggests that the
position of the new polar axis is established statistically at first,
by any group of patches which by chance happens to be similarly
oriented and therefore can form a " field " whose influence might
then spread to adjoining sections to bring them into corresponding
orientation. With this shifting, patches soon appear much larger
than originally, and this can be attributed to their joining together
as they come into parallel and homopolar orientation. Areas
bearing wide pigment stripes do not form a continuous structural
union with fine-stripe areas, but only Hke with Hke. Although it
would be difficult to observe minor resorption of patches, it is
apparent that there is no large scale dedifferentiation.

As the cut areas move so their stripes become parallel, a visible
polarit}' appears as the mass elongates in one direction and a hold-
fast appears at the end of a projecting point. Oral regeneration
never begins until a definite locus of stripe contrast of considerable

222

THE BIOLOGY OF ST EN TOR

Fig. 63. Reconstitution in minced S. coeruleus.

A. Realignment and rejoining of pattern, a: Operation,
consisting of repeated cutting with point of a glass needle until
lateral striping is reduced to scattered patches. Holdfast and
feeding organelles were removed, b: Patches, numbering about
50 are at first separated by endoplasm. c: Patches healing
together and cell-shape knobby because of striping running in
multiple directions, d: Indication of a tail pole and axis with
patches aligning in parallel and joining when of the same type

(e.g. wide-stripe areas with wide striping).

B. Subsequent regeneration in a similar case. An oral
primordium appears as soon as a sufficient locus of stripe-width
contrast was re-established (6 hours) and the anlage follows the
course of this l.s.c. Second sketch shows nearly normal

specimen one day after operation.

C. Intact head grafted to minced mass of two stentors minus
heads and tails. Three days later the specimen became as shown,
striping normalized on ventral (oral) side, still irregular
anteriorly on dorsal side. Axis seems to be established by the
engrafted head but head and bordering stripes apparently have

RECONSTITUTION IN DISARRANGED STENTORS 223

length appears. Even if the head had previously been excised and
the animals minced just before the anlage was due to appear, the
primordium v^as still not formed until considerable reorientation
had occurred. But if a stentor is minced and the mouthparts
excised at the same time the primordium can and often does appear
within the normal time of 4 hours, so that cutting of the striping
and its subsequent rearrangement does not seem to interfere in
any way with the activation and preparation of the cell for pri-
mordium formation. All the fine-stripe patches may not aggregate
in one place and therefore two primordia may be formed producing
a double stentor from a single. Oral regeneration seems to proceed
normally whenever an l.s.c. is estabhshed to determine where the
anlage is to be placed and it was noted that without exception the
primordium does appear in an l.s.c. Such loci may be much dis-
torted due to the original disarrangement, and the anlage faithfully
follows their contorted contour (Fig. 63B). The specimen therefore
does not wait until it has reestablished perfect order in the striping
but regenerates as soon as possible and makes further adjustments
later.

When all but the primordium of regenerating stentors was
thoroughly minced there was no resorption of the anlage, which
continued to develop, though often slowly; but the membranellar
band formed was usually distorted. This indicates that the state
of activation is not nullified by severe cutting, but that orderly
striping is required for normal deployment of the developing
feeding organelles. Even if the regeneration primordium itself
was cut in two, the parts usually rejoined and development
continued to rather successful regeneration.

Specific inhibition of oral primordium formation by intact
feeding organelles occurred even though the cell was minced. This
was demonstrated by grafting intact heads to singles and 2-masses

no strong orienting influence on adjacent patches. Specimen
survived 8 days without reorganizing, and hence oral inhibition
of primordium formation was effective though lateral striping cut
into patches.
D. Mince graft of 5 whole stentors achieves axis by 6 hours,
regenerates as a doublet and begins dividing as a doublet 2 days
later. Integration of shape is better than in grafts of 5 not

minced.

224 "T^E BIOLOGY OF STENTOR

after they were minced. Often no primordium appeared (Fig. 63c),
or only days later. The grafted head became harmoniously
integrated with the minced host, and yet grafting of heads or
primordium-site sectors or large areas of intact striping did not
seem to hasten the gradual re-alignment of the patches.

Minced 2-masses like grafted pairs produced i, 2 or 3 primordia
upon regeneration. In most cases two were formed, in some cases
only one, and very rarely 3. Again, the oral valency seems to be
simply an expression of the probability of obtaining more than one
area of fine striping in the reconstituted graft complex. Mincing
a fusion mass in fact definitely favors attainment of unitary shape.
A minced 25 -mass formed a rather unified fan shape with single
axis (Tartar, 1954, Fig. 33B), though large, unminced masses
never achieved anything like the normal form. Two 5-masses,
minced, became doublets with single conical shapes, much in
contrast to the bizarre forms produced when such masses are not
minced (Fig. 63D). Minced masses, unlike minced singles, seem
to have a better chance of producing a single shape when all traces
of the original axes have been obliterated, and this inference is
substantiated by the confusion of mildly disarranged stentors,
presently to be described. The response to these operations
demonstrates an astonishing capability of thoroughly disorganized
stentors to regenerate and to reconstitute the normal, orderly
arrangement of the ectoplasmic pattern, even within a single day,
after all the complex ciliary, contractile, conductive and other
differentiations of the ectoplasm have been cut into tiny pieces
scattered at random.

Remarkable, too, is the possibility of the reverse process, in
which organization is sacrificed to autonomous disorganization.
Several instances have been found in which individual coeruleus
responded to certain treatments by spontaneously transforming the
orderly striping into a generally disarranged patchiness much as
if the cell had been minced (unpublished). The two instances from
cutting operations are shown in Fig. 64. The same effect was
sometimes produced by treatment with dilute salt solutions (see
Fig. 71). If these responses are reproducible, we have an oppor-
tunity to explore the significance of this peculiar break-up of
structure, so greatly in contrast to the general tendency of stentors
to integrate themselves into an orderly pattern. This behavior

RECONSTITUTION IN DISARRANGED STENTORS 22^

Fig. 64. Transient autonomous disorganization of shape pattern.

A. a: Left half of stage-3 divider rotated on right, h: Fission
blocked but further primordium formation, leading only to 3
incomplete oral differentiations in addition to original mouth-
parts {x). c: Reorganized singly, with fair stomatogenesis and
good striping, d: On day 4 the lateral striping except in the oral
meridian was broken into patches quite as if minced. This
condition was later corrected to normal; and the specimen
eventually divided, one of the products also then dividing,

therefore apparently an instance of postponed fission.

B. a: Sector with stage-3 regeneration primordium and 8
nuclear nodes grafted transversely onto an enucleated stage-3
regenerator. Both primordia were, paradoxically, resorbed.
Two new anlagen appeared, joined and gave fair differentiation
of feeding organelles (6) but the striping became noticeably
patchy, c: Reorganized now with striping aligned but with four

tail-poles. Further normalization occurred later.

recalls, in a possibly significant parallel, the normal fragmentation
of the cortical striping and kinetics in large forms of the ciliate
Ichthyophthirius. Patches so produced then become the ciliation
of multiple daughter cells, according to the account of Mugard
(1948). But in Stentor, the animals seemed to be able to recover
after passing through a period of self-trituration, as they do from
minceration.

226

THE BIOLOGY OF STENTOR

2. Other disarrangements of the normal cell pattern

When gross parts of the stentor cell are shifted with respect to
one another in operations much more simple than total mincing,
the effects on form are usually far more enduring and bizarre.
Original longitudinal and transverse axes are apparently retained
in the large parts and fall into conflict with each other. Gruber
(1885a) had shown that in stentors suffering a single cut the parts
could shift upon each other to produce doublish forms, and
Ishikawa (1912) produced these and large lateral flanges by slicing
into coeruleus and holding the split parts separated for a few
minutes so that they then did not heal in place. Here we shall
simply offer two new cases which are typical.

Figure 65A shows a coeruleus which was simply split longitu-
dinally yet it never recovered the normal form before it eventually
died of starvation. The other case (b) was of a stentor which had
been " quartered " with the result that each fourth of the cell was
maximally misplaced. Gross abnormality resulted, finally leading
to the formation of a double animal. In a previously cited case the
same operation produced a doublet with reversed asymmetry on

Fig. 65. Gross abnormalities of shape produced by simple shifts
of large stripe areas.

A. When stentor is split to tail-pole, and halves shift by
contraction, healing irregularly, aberrant form is produced but

later corrected.

B. Anterior half rotated 180° on posterior then left half 180°
on right. Quartered animal became very abnormal in shape,

later converting to a telobiotic double stentor.

RECONSTITUTION IN DISARRANGED STENTORS 227

one side (see Fig. 49). Also to be recalled in this connection is the
great confusion of shape when cell sectors are implanted hetero-
polar in stentors (see Fig. 54). Disarrangement of large areas of
ectoplasm therefore leads to much more confusion than mincing.
Reorientation of such areas may simply be more difficult, or their
polar " fields " may be so strong as to engender major conflicts
within the cell.

The neatest and best studied disarrangement of the pattern of
Stentor is that in which the anterior half of the stentor is rotated
180° on the posterior (Tartar, 1956a; Uhlig, 1959). When coeruleus
is selected for this operation, the pigment stripes with their varying
widths can be used to identify the cortical patterns of the two halves
and to follow the changes which occur in them.

The nature of these transformations of the striping depends in
part on whether and where the severed stripes may join. Using
these cases, stentors grafted heteropolar by the headless anterior
ends, and observations on minced animals, we can formulate
provisionally a rule for the union of lateral stripes. It will be
recalled that the complex fibrous structure Hes in the clear bands
and that the pigment stripes appear to be merely the spaces
between these which are filled in with the colored granules. Yet if
discontinuities in pigment stripes, appearing like the colorless
fission line, can be taken as a criterion that the fibers of the adjacent
clear bands are also discontinuous, then it seems that intimate
structural union between two sections of ectoplasm occurs only
when the abutting pigment stripes are of equal w^idth, approxi-
mately parallel, and homopolar. Thus in heteropolar grafts there
is no joining of pigment stripes even w^hen they are of equal width
(see Fig. 46E). In mincerates, as well as in parts of the normal
" ramifying zone " of Schuberg, it is indicated that pigment stripes
of equal widths do not join if they are at an angle to each other
(Fig. 63B). And in anterior-rotated-on-posterior grafts there is a
discontinuity where the wide stripes of one half abut the fine stripes
of the other, while in those places where stripes are of equal width
they join and become continuous (Fig. 66a). As will be noticed in
the figures, even though wide and narrow pigment stripes do not
join, there is the appearance of a strong attraction between the two.
Characteristically two fine stripes move so as to subtend one wide
stripe, although a non-pigmented line continues to separate them.

228

THE BIOLOGY OF STENTOR

Fig. 66. Consequences of rotating anterior part i8o° on the
posterior.

A. Equal halves rotated, mouthparts excised, a: Pigment
stripes of like width join and mend, wide and narrow stripes
match up, 2 to i, without joining, b: Regeneration is delayed
because membranellar band left intact, hence single primordium
arising only in relation to the extending posterior striping,
anterior stripes resorbing. c: Completion of regeneration with
anterior striping nearly replaced by growth of posterior striping.

B. Stentor transected somewhat anterior to center; anterior
striping resorbed though no primordium is formed since feeding
organelles left complete. Head therefore does not make anterior
striping dominant yet effectively inhibits anlage formation in

the posterior portion, isolated by rotation.

RECONSTITUTION IN DISARRANGED STENTORS 229

The operation of rotating anterior on posterior halves was first
used to demonstrate that primordium formation can occur in such
loci of stripe contrast (Tartar, 1956a), but it was also observed that
stripe disharmonies were resolved by resorption of the anterior
striping and extension of the posterior to take its place. If regenera-
tion was delayed by leaving the membranellar band intact, the
forward resorption of the anterior ectoplasm occurred so rapidly
that when an anlage was formed it appeared only in the primordium
site from the posterior half (Fig. 66a). Even when regeneration was
not induced, anterior striping could be dissolved as it was replaced
by the posterior (b). Alternatively, in some cases in which regenera-
tion was not induced the striping of anterior and posterior halves
appeared to interpenetrate, parts of both anterior and posterior
striping being preserved.

Later it was found that grafts with this astonishing interpenetra-
tion of stripes could be produced quite readily (Tartar, 1959b).
Figure 66c shows how the fine line zone or primordium site of
each half plows through the striping of the other half as it extends
in length and gradually reaches the opposite pole. Specimens with
two good primordium sites which reorganize as doublets are
therefore generally produced. Later readjustments, in which the
fine lines of either side become wider, then lead to eventual
recovery of the single form. In one instance the animal divided.

C. Stentor cut in half, fine-line zones interpenetrating next
day to reach opposite poles of the cell, with result that two loci
of stripe-width contrast are extended and specimen becomes a

doublet.

D. Front and back views of dividing specimen with interpene-
trating stripes, showing how fission line forms indiflferent to

suture between stripes of the rotated halves.

E. Racial difference in interpenetration of striping. Same
operation as in c, but with EUetsville race. First sketch: two
days after operation with striping still unchanged. Second : i ith
day, with striping now running pole to pole forming two primor-
dium sites, one with reversed asymrrietry and therefore producing

anlagen which gave incomplete stomatogenesis.

F. Stage-3 divider cut with anterior part larger. Specimen
reorganizes instead of dividing, with only anterior portion of the
primordium developing, the posterior part resorbed. Posterior

striping is gradually resorbed as anterior stripes extend.

230 THE BIOLOGY OF STENTOR

Division occurred while the stripes were interpenetrating and the
fission Hne did not follow the suture but cut indiscriminately
across fine and broad striping, following a course which may be
called typical (d). The latter, with similar cases, shows that abnormal
disharmonies and discontinuities in the lateral striping do not
preclude division and suggests that the fission line is determined
by some agent other than the lateral stripes themselves. Thus the
same subcortical forces which cause predivision of the carbo-
hydrate reserves in the neat manner already described may impose
a severance of the striping lying exterior to them regardless of the
nature or disposition of that striping.

Yet the most interesting questions concern how the highly
structured ectoplasm can permit stripe areas to slip by each other,
as well as the bearing of stripe extensions in limited places on the
control of growth throughout the cortex. Moreover, it appears that
races of coeruleus vary in the ease with which stripes interpenetrate
after this operation, specimens of one strain remaining as grafted (e)
long after those of another had formed doublets.

Uhlig (1959, and unpublished) has developed this type of
experiment much further, by transecting coeruleus at different levels
before rotating the two parts. He substantiated that when the cut
passes through the place of origin of the primordium producing
approximately equal halves for rotation, the anterior striping is
generally resorbed as the posterior stripes extend and take over.*
When a posterior cut produces an anterior component about four
times the size of the posterior, anterior striping now predominates
and extends posteriorly, replacing the original tail-pole striping
which is resorbed. The case shown in Fig. 66f confirms this
finding. A dividing stentor was transected across the oral end of the
division primordium and the smaller posterior part rotated on the
larger. The tip of the anlage was then resorbed ; the larger portion
continued development and led to reorganization as the original
posterior striping gradually disappeared. But these cases were not
uniform and sometimes there was an interpenetration of stripes.
Therefore it appears that in these grafts there is a delicate balance
between the two systems which may be tripped to favor the

* Uhlig (i960) claims that resorption of the anterior striping proceeds
from the anterior ends of this striping and not from the suture.

RECONSTITUTION IN DISARRANGED STENTORS 231

dominance of one or the other or may result in equiUbrium, with
the striping of both halves retained and interpenetrating. A fine
and unnoticeable difference might swing the balance one way or
the other. When the cell was so cut that the ratio of anterior half
to posterior was about 3 : 2, Uhlig found that dominance was
exerted by neither part and doublets resulted which could divide
and produce more doublets. He states that then each primordium
site '' reorganized " completely, but perhaps he also observed what
seems to me to be the case : that there is an extension of each half
of the original primordium site as it penetrates through the stripes
of the other half. He interpreted the various responses as an
interaction between the head-tail gradient and the transverse or
circumferential gradient in stripe width. For instance, when only
the posterior end of the cell is rotated, its circular gradient in the
immediate neighborhood of the steepest end of the tail-to-head
gradient is apparently obliterated.

Experiences wdth this type of operation will have a bearing on
the analysis of axial gradients in Stentor. From a more general
standpoint it is shown that stentors have still further resources, in
the selective resorption or interpenetration of stripes, for the
reconstitution of their normal form and pattern.

CHAPTER XIV

ANALYSIS OF STENTOR THROUGH ITS
RESPONSE TO EXTERNAL AGENTS

Various chemical and physical treatments of living stentors have
been used to reveal and analyze otherwise inaccessible aspects of
their structure and behavior. These studies are classified according
to objectives of the investigation, types of eflFects produced, or the
agent used.

I. Action of the membranellar band

To immediate observation, the most impressive activity of
attached stentors is the orderly beating of the large membranelles
in beautiful waves of metachronal rhythm. For the membranelles
do not all beat together in the same phase but in succession, so
that at any one instant membranelles in the eflFective beating stroke
are followed by others successively relaxed in the recovery stroke
and these are again followed by organelles in the effective stroke,
giving the impression of waves originating in the gullet and passing
along the membranellar band to its terminus. Hydrodynamically,
this type of beating is probably the most efficient, because groups
of cilia work together to move the water toward the mouth but this
action is distributed so that there is a continuous flow, whereas if
all membranelles beat in the same phase the medium would move
by starts and stops.

The types of action of which the membranelles are capable and
the variables involved are shown diagrammatically in Fig. 67.
First, the membranelles may all be stopped and pointed forward
and somewhat inward, when stentor is swimming backward or
has momentarily ceased feeding (b). When they resume beating
they do so at first individually and at random, soon falling into
metachronal rhythm. Hence each membranelle is capable of
independent beating. The number of strokes per second is the
frequency of beating. Presumably the amplitude of the eflFective

232

ANALYSIS OF STENTOR

233

Stroke may vary but this would be difficult to detect. The distance
between membranelles in the same phase is the metachronal waz'e
length. Speed with which metachronal rhythm passes along the
band is the wave velocity and is equal to the product of frequency

hme - /requAncy ot bead:
distance - ssripLiiiuLe

A

propulsive, stroke r^coveri/
l^AV£ LENGTH

stroke

■^ Wave vetoed If = ^-eqaen^i/ of beat X wavelength-

Fig. 67. Actions of the peristomal membranelles.
Analysis of successive beating or metachronal rhythm.

B. Swimming backward with ciliary beat reversed and

membranelles stopped and pointed forward.

C. Forward swimming with membranelles active and pointing

backward.

D. Coordination in transected sections of the peristome, a:
Metachronal rhythm maintained, moving distally from pace-
maker in the oral region, b: Isolated section sets up independent
rhythm, pace set by proximal membranelles. c: Beating of
membranelles still independent, as in the whole peristome when
beating recommences. Rhythm will be re-coordinated by new

pacemaker at x.

234 THE BIOLOGY OF STENTOR

of beat and wave length. Each of these factors is variable. In
addition, the membranelles can be oriented to point outward and
backward as they do in forward swimming (c).

It will be recalled that the membranelles are rooted in triangular
basal plates all of which are connected by an inner fiber. It was
natural for early microscopists to have supposed that the impulse
producing metachronal rhythm passed along this fiber, exciting
one membranelle after the other; but there are at least two argu-
ments against this supposition. The wave velocity (roughly 700 /^
per second) is slower than any known neuroid transmission (Sleigh,
vide infra). And second, on resuming their beat the membranelles
do not start at once in metachronal rhythm, which is only later
established after a brief period of irregular beating.

Coordination in the membranellar band of S. polymorphus was
the subject of astute investigations by Sleigh (1956, 1957). By
several approaches he shows that the frequency of beating of the
membranelles is dissociable from the wave velocity or rapidity of
transmission of the impulse from one membranelle to the next.
Both frequency and wave velocity decreased with lower tempera-
ture but the decrease was more rapid in the frequency of beating.
Increasing viscosity of the medium by addition of methyl cellulose
resulted in decreased frequency of beating but no change in the
wave velocity. This corresponds to expectations, for external
resistance should decrease the frequency of stroke without affecting
internal mechanisms of transmission. Magnesium chloride
increased the frequency of beating without affecting the wave
velocity ; and with aluminum chloride the trivalent ion was several
times more effective in producing the same response. If these
metal cations may be regarded as reducing the internal viscosity
of the protoplasm in cilia, increased frequency would be explained
as due to lower internal resistance. Digitoxin greatly increased the
wave velocity but only slightly increased the frequency of beat and
the shape of the effect- vs. -concentration curves was different.
Finally, cutting the membranellar band interrupted the wave
conduction but did not prevent the reappearance of metachronal
rhythm in separated sections distal to the gullet (Fig. 67D). This
experiment at once excluded that metachronal waves originate only
in the gullet region and can be stopped by cutting the fiber which
connects the basal plates of the membranelles.

ANALYSIS OF STENTOR 235

Following incision, the first membranelle distal to the cut
established a new frequency of beating, which was then taken up
by all the membranelles in the isolated section. The first mem-
branelle of a series may therefore be regarded as a pacemaker which
determines the frequency of membranelles distal to it. Being
separated from proximal membranelles the pacemaker can establish
its own intrinsic rhythm, often different in different sections.
Usually, its rate was slower than that of the membranelles on the
gullet side, but in a few cases it was more rapid, possibly due to
excitation through injury. In the intact feeding organelles, the
pacemaker would presumably be some membranelle within the
gullet. In this region. Sleigh (1957) found that the wave lengths
and wave velocity are smaller than in the distal lengths of the
membranellar band; but this discrepancy he resolved by the
observation that the membranelles are also closer together in the
gullet. Therefore the number of membranelles in one wave length
is the same throughout the band and hence the number stimulated
per second is the same regardless of their density. " The wave
velocity thus depends on the number of cilia involved in the trans-
mission, and not on the linear distance traveled by the metachronal
wave ". This is further evidence that the cilia themselves are
involved in transmission of the metachronal wave and not the
basal fiber connecting the basal plates.

Chemical and physical treatments thus indicated that there is an
intraciliary excitation which is separable from a second process,
the conduction of the impulse from membranelle to membranelle.
From these and the cutting experiments. Sleigh proposed the
hypothesis diagramed in Fig. 68. Only a single cilium in each
membranelle is shown for presumably the closely packed cilia of
each membranelle work together. Each cilium would then be
capable of spontaneous beating but at a slower frequency than
when excited by interciliary transmission. Increasing or retarding
frequency of beat would simply alter the rapidity of ciHary contrac-
tion or response to the internal state of excitation and therefore
need not affect the rate of conduction of the impulse between the
motor organelles. On the contrary, digitoxin, by decreasing the
threshold of excitability, as it does in heart muscle, might increase
the speed of excitation and therefore lead to a more rapid tripping
off of the conducted impulse so that wave velocity would be in-

236 THE BIOLOGY OF STENTOR

creased without much affecting the frequency of beat. The general
picture, then, is not of a row of effectors joined by one connecting
nerve or neuroid process, but of a series of triggers which fire each
other in succession.

CiUu-m of
mem br.anelle

ciliary roots ot
2)asBl lamella

bsLSaL
fiber

Fig. 68.

Diagram of a theory of metachronal coordination.
(After Sleigh, 1957.)

This would leave the basal fiber without defined function, and
Sleigh does not even mention it. Apparently he regards the conduc-
tion as simply spreading through the ectoplasm between the
membranelles. That such transmission without fibrous connections
is possible even at much wider dimensions is indicated by the
membranellar response of grafted stentors. Immediately after many
stentors were grafted together, and even if the membranellar bands
are cut into sections of various lengths, all the membranelles in the
mass were observed to stop and start together as the cilia simul-
taneously reversed or beat " forward ", long before any intimate
structural reconnections could have been made, as if a coordinating

ANALYSIS OF STENTOR

237

influence passed like a flash over the entire surface of the mass
(Tartar, 1954). Myonemes respond similarly, components of a
fusion mass contracting together almost from the moment of
grafting (Weisz, 1951a).

One approach to analyzing what the functions of fibrous
structures associated with the membranelles may be is provided
by the selective resorption of parts of bands during fusion of heads
in grafted stentors. In the specimen shown in Fig. 69 the resorbing
membranelles first lost their metachronal rhythm, beating irregu-
larly in a local area. This is as if structures responsible for this

Fig. 69. Loss of coordination in membranelles anticipating
resorption of portions of the membranellar band. The ends of
the two peristomes which were resorbed in a doublet stentor to
fuse the frontal fields first showed independent beating of
membranelles.

type of coordination were the first to become dediflFerentiated, for
the ectoplasm presumably retained its continuity.

Other responses of the membranelles to chemical treatments
have been observed (Tartar, 1957a). In solutions of 1% Nal the
membranelles remained continuously stopped, but in CaCl2 they
kept beating vigorously until the organelles were destroyed.
Ethanol stimulated the membranelles to keep beating even while
the remainder of the cell was being destroyed, confirming the
earlier observation of Daniel (1909). This activity was in marked
contrast to the normal avoiding response in which the membra-
nelles are stopped. In MgCl2, at much higher concentrations than
used by Sleigh, the membranelles continually started and stopped

238 THE BIOLOGY OF STENTOR

at a rate of about one change per second until the band itself was
destroyed. NiS04 in very weak solution is an effective ciliary
anaesthetic for protozoa (see Tartar, 1950); body cilia and mem-
branelles in Stentor were stopped in weak solutions, but although
not beating, the membranelles keep changing their orientation in
the two positions shown in Fig. 67B and c. This reorientation in
membranelles which were not beating was most striking to
observe — like the batting of eyelashes — and it should also be
mentioned that the body contractions of the stentor were in no
demonstrable way affected by NiS04. Hence the unstriated basal
lamellae and associated fibers of the membranelles, homologous
with the striated ciliary rootlets described in metazoa by Fawcett
and Porter (1954), rnay be contractile (like the unstriated ribbon
bundles of the clear stripes) and serve for orienting the membra-
nelles in one direction or another, a function which in this case
seems to be completely dissociable from ciliary beating.

2. Coordination of body cilia

Every part of the ciliated ectoplasm, without endoplasm or
nucleus, is a self-contained coordinating system. This was
demonstrated for Stentor and Spirostomum by Worley (1934) who
found that in isolated patches the ciha could start and stop, reverse
their effective stroke, and beat in metachronal rhythm. Treatment
with ciliary anaesthetics such as potassium chloride resulted first
in loss of the capacity to reverse, then of metachronal rhythm, and
finally of ciHary beating itself. These three kinds of ciliary action
are hence dissociable. Individual activity of a cilium and the two
types of coordinated movement of cilia therefore are probably due
to separate processes. Reversal of beating spread instantaneously
like a signal passing over the surface, uninterrupted by incisions
and therefore probably not mediated by conductile fibers. Meta-
chronal waves are much slower. Hence Worley suggested that they
are mediated by interciliary fibers, specialized structures whose
effectiveness in integrating cilia may, paradoxically, be due to
their slowing down interciliary impulses. In Spirostomum^ the
kinetics of which closely resemble those of Stentor (Randall, 1956),
Worley found that the metachronal beat could circumvent surface
cuts, indicating the presence of transverse connections between
rows of body cilia.

ANALYSIS OF STENTOR 239

Reversal of beating of the body cilia in unattached stentors is
immediately manifested in backward swimming in which the
effective stroke of the cilia is directed forward instead of backward.
Merton (1932, 1935) made an extensive study of the effects of
various salts and other substances in compelling stentors to swim
backward. The species used wcreroeseli, coeruleus, and polymorphus.

First it should be mentioned that distilled water alone produced
backward swimming, with most of the animals disintegrating in
two hours. Peters (1908) had early shown this injurious effect of
pure water on coeruleus. He transferred the animals every 15
minutes to fresh distilled water and all then died within an hour,
death occurring not by swelling of the whole cell but by the forma-
tion of internal vacuoles w^hich increased in size and led to a
blistering of the surface with final disruption. Death he attributed
to washing out of the salts of the cell, but it may just as well have
been due to other osmotic effects; for Jennings (1902) found that
sugars killed by the subtraction of water and that there is no effect
at first but only after a sudden contraction, following which the
animals crumpled and decreased in volume.

Therefore Merton made up his solutions in tap water which had
no effect on their behavior and was not immediately injurious. He
found that monovalent cations induced reversal of ciliary beating
while the bivalent cations of calcium and magnesium did not. Thus
weak solutions of KCl produced a continuous backward swimming.
Using their chlorides, the monovalent ions tested were in approxi-
mately decreasing order of effectiveness : K>Rb>Cs>Na>NH4.
Anions also had some effect on the response. Potassium compounds
were compared, and the order of decreasing effect of the anions in
promoting ciliary reversal was

C03>S04>C1>I, NO3, P04>Br>Ac.

I later confirmed these results in regard to contrasting effects of
monovalent and bivalent cations (Tartar, 1957a). In addition I
found that LiCl, which only produced disintegration of stentors
for Merton, also induced conspicuous backward swimming. And
ammonium acetate in strength of 1%, a compound not tested by
Merton, caused the most prolonged and continuous reversal of any
of the compounds used.

240 THE BIOLOGY OF STENTOR

Merton also tested hydrochloric acid, fatty acids, saponin, and
certain alkalis, which produced only injury and no ciliary reversal.
Urea and sugars also gave no reversal, and this I can confirm.

Merton regarded the induced backward movement as a specific
effect of the monovalent cations on the cilium. That osmotic effects
are not involved is obvious from the fact that the type of ion and
not its concentration is crucial. Nor was he dealing with avoiding
responses, because he found that stentors show the normal tem-
porary backing up even when they encounter calcium chloride,
but this compound does not compel the continuous backward
movement which KCl does. Also, the reversal in KCl, for example,
was not counteracted by adding an equivalent amount of CaCl2.
The bearing which these results may have for an analysis of ciliary
reversal is yet to be clarified, but to be able to produce reversal,
immediate and prolonged, is a beginning. At least we can conclude
that the mechanism of body cilia is such that it can adapt to an
abnormally continuous backward beating at increased intensity
under the influence of reversing agents.

3. Ciliary anaesthesia

It has already been mentioned that the heavy metal salt NiS04
at very low concentration causes reversible paralysis of both body
cilia and membranelles in Stentor, as I found following a suggestion
of Gelei (Tartar, 1950). After inducing ciliary reversal, NaCl and
KCl also produce partial anaesthesia of the cilia (Merton, 1935)
and the potassium salt seems to be the more effective for this use.

Following early exploratory tests of Verworn, Ishikawa (19 12)
obtained reversible narcotization of the cilia in Stentor coeruleus
with chloroform vapor. He brought a piece of filter paper soaked
in chloroform near the drop containing a stentor in order to quiet
the animal for cutting operations, but by his own account this is
not to be recommended because necrotic conditions easily develop.
At lowest concentrations the stentors showed an accelerated
activity, but at higher strengths the cilia were slowed and the
animals became semi-elongate. Wounds from cutting were slow
in healing. With nearly lethal concentration, the animals remained
quiet as if dead, though sometimes they could recover slowly, and
it was also reported that they might shed their cilia or begin
disintegrating, a small portion at a time.

ANALYSIS OF STENTOR 24I

Hofer (1890) recommended hydrochlorate of hydroxylamine
(0-25%) neutralized with sodium carbonate for slowing the cilia
and relaxing myonemes in coeruleus. Of several related compounds
Mugard and Courtney (1955) found that only KH2PO4 was a
sufficiently non-toxic immobilizer of all ciliates tested, including
stentors.

Methyl cellulose, first introduced for quieting paramecia by
Marsland (1943), remains the least noxious method for slowing
cilia in stentors. Its use will be discussed in the chapter on
techniques.

4. Anaesthesia of myonemes

Stentors, like Spirostomum and the stalked Vorticellids are
capable of very strong and instantaneous contraction. In my
experiments with stentors I have been impressed by the observa-
tion that contractility seems to be one of the last functions to
disappear, and even grossly abnormal and necrotic specimens, no
longer capable of swimming with their cilia or of regenerating,
nevertheless continue to react to poking with the needle by rapid
and vigorous contraction, almost up to the time of their final
demise.

Attempts to anaesthetize the myonemes and abolish contraction
have been pursued both for the purpose of fixing and staining
animals in the fully extended state and to test whether stentors
behave like a typical nerve-muscle preparation.

It will be recalled that Neresheimer (1903) found what he
thought to be nerve-like fibers which he called " neurophanes "
running to the myonemes, and later Dierks (1926a) described
similar fibers (" neuroids ") running exterior to the myonemes and
terminating in or sending branches to them. That these fibers with
their putative function represent specialized organelles and not
mere artifacts of fixation is still very questionable (see p. 55), yet
they led Neresheimer to an extensive study of the effects of drugs
on Stentor coeruleus which may have its merits apart from the
conclusion he drew. Control animals were first placed in a small
dish on a platform to which a graduated stick was fastened in a
vertical position so that a weight on a pulley could be dropped
from a measured height onto the platform, the vibration of which
would then stimulate the animals to contract. The minimum

242 THE BIOLOGY OF STENTOR

distance of fall to excite contraction was then used as a basis for
comparison of the reaction of stentors subjected to various drugs.

Morphine hydrochloride apparently produced the greatest
insensitivity. This relaxation was counteracted in a typical manner
by the antagonists atropine and picrotoxin. Strychnine produced
mild contractions, as would be expected from its effect on higher
animals. In curare the contraction was so energetic that both clear
and pigmented stripes were said to be torn loose in a way which
he did not describe in detail. The antagonist, physostigmin,
counteracted this effect. Neresheimer states that these results
confirmed earlier studies by Verworn on Stentor, Spirostomum^ and
Carchesium. If the myonemes are excitable only through neuroid
fibers one might have expected complete paralysis on the basis of
blockage of the neuro-muscular junction which curare produces
in higher forms. Complete relaxation of specimens which could
then be fixed in the extended position was achieved in Spirostomum
but not in Stentor. Caffein seemed to increase the sensitivity, but
in nicotine the stentors relaxed and became more insensitive.

A student of mine (N. G. Parisis, 1956, unpubHshed student
report), tested the effects of curare and strychnine, separate and
combined, on Stentor coeruleus and Spirostomum amhiguum. Both
drugs stimulated mucoid secretion, as demonstrated by the obser-
vation that the animals could be moved by an advancing needle
before the needle came near the cell. In neither substance alone
was contractility lost. In a mixture of strychnine and curare,
how ever, the ciliates lost their power of contraction completely and
could even be cut in two without responding, though the cilia
kept beating.

Neresheimer also tested one bromide (NaBr) which also made
stentors so insensitive that they could be cut in two without con-
traction, but apparently the effect was not reversible and the
animals did not survive the treatment. I have found that 1%
solutions of the iodides of sodium or potassium have the same
effect and their action is completely reversible (Tartar, 1957a).
Outstretched animals could be cut in two without a single twitch
in either half, and after returning to normal medium complete
contractility was recovered within a day.

Although they might become very insensitive, Neresheimer
found that his treated stentors always contracted when treated with

ANALYSIS OF STENTOR 243

common fixing agents. From this fact together with the general
similarity between the response of stentors and nerve preparations
to the drugs and antagonists which he tested, Neresheimer con-
cluded that the effects were not on the myonemes themselves but
on the " neurophanes " which were therefore of a neuroid
character. Dierks was, of course, of the same opinion with regard
to his '' neuroids " and he found that coeruleus became insensitive
to touch in KCl, while CaCL increased contraction and was
antagonistic to the action of potassium ; for animals made insensi-
tive in the potassium salt regained their irritability when calcium
chloride was added. Relaxed stentors still contracted when fixed
with Flemming's solution. But contraction of the cell in strongly
coagulating solutions can scarcely be taken as demonstrating that
the myonemes were not directly affected and the impression
remains that much more sophisticated studies will be needed to
demonstrate similarities and differences between the responses of
stentors and typical nerve-muscle preparations.

Merton (1932, 1935) also attempted to treat stentors so that they
could be fixed and stained in the extended form. Anticipating
Dierks, he found that KCl gives a partial anaesthesia of the
myonemes. In dilute Ringer's solution, stentors became out-
stretched but their irritability was increased. Metal salts of iron
and copper were said to produce differential contractions of the
cell and from his description it appears that the anterior end of the
stentor contracted while the tail-pole remained extended. Copper
sulphate produced a hardening of the cortex which therefore
antagonized the contraction of the myonemes and left stentors in
a semi-extended state. Fairly well extended preparations were
made by relaxing stentors for 2 to 3 hours in Ringer's solution
diluted 1:3, then applying weak copper acetate to harden the
surface, following this treatment with fixation.

Dierks (1926b) confirmed that a 0-5% solution of KCl renders
stentors insensitive to touch. Conversely, CaCl2 increased contrac-
tion and was antagonistic to the action of potassium, animals
regaining their irritability when calcium was added. A 0-04%
solution of Na2S04 paralyzed both cilia and myonemes, but stentors
relaxed in this way or with KCl still contracted on fixing.

244 "THE BIOLOGY OF STENTOR

5. Comparison of osmotic effects to cooling

Following the speculations of Jacques Loeb, Greeley (1901)
tested whether increasing the osmotic pressure of the medium
surrounding coeruleus had the same effect as decreased temperature.
Reducing the environmental temperature to 2 °C not only quieted
the animals but produced a variety of pathological conditions,
including the disappearance of the feeding organelles and the
lateral striations — to mention two of the most interesting effects
which deserve checking. On rewarming, the '' rest of the cells "
apparently survived a couple of weeks but in only a few cases did
they regenerate. Cane sugar was said to give the same effects as
cooling, though a typographical omission in the published account
prevents our ever knowing the concentration employed. As we
shall see later, sugar causes shedding of the membranellar band,
but disappearance of the lateral stripes does not occur and they
merely collapse with the cell. Full regeneration followed sugar
treatments; therefore we presume that the treatment was mild
enough to allow the animals to survive. Loeb's conjecture concern-
ing the similarity between cooling and concentration of protoplasm
by loss of water through osmosis was therefore considered to have
been confirmed.

Increasing the temperature to 25-28 °C apparently accelerated
division, as would be expected; yet we cannot give credence to
Greeley's account that within 3 to 4 hours there were many
successive divisions induced without marked decrease in size.

6. Acceleration of division

In addition to the strange account of accelerated division just
mentioned, Peters (1904) claimed that dilute solutions of KCl also
stimulate fission in coeruleus. The solution used was o-oi molar
or a bit stronger. NaCl solutions of comparable strength produced
a suppression of division instead. From the results it appears that
Peters probably did get an accelerated division. He was looking
for a specific, immediate impulse since the experiments were run
for only 6 hours. Many abnormalities were also encountered.
Apparently the peristomal band was shed in some specimens
(because regeneration occurred) although he did not say so
expHcitly. Unequal divisions and production of small blebs of
cytoplasm were reported. For a division experiment Peter's proce-

ANALYSIS OF STENTOR 245

dure was odd, as he did not count the smaller individuals ! But if
anything this should mask the strength of the results produced,
and hence we are merely left with the suggestion that possibly KCl
may supply an impulse to division. The effect, if valid, was
apparently not due to osmotic pressure, because lactose solutions
of even higher osmotic tension were without effect.

7. Changes in state of the protoplasm

An incidental observation of Prowazek (19 13) was that sodium
taurocholicum causes the endoplasm of stentors to clump into
balls and the nucleus, at first highly refractive then disappears.

Changes in the internal viscosity of stentors in relation to various
ions was studied by Heilbrunn (1928). Centrifuging the animals
in various salt solutions he observed the relative speed with which
internal granules and particles passed through the endoplasm.
Bivalent cations (calcium and magnesium) apparently decreased
the viscosity of the interior, producing liquefaction. Monovalent
ions (K, Na, NH4, Li) increased the viscosity and caused coagula-
tion. But later Heilbrunn (1943) admitted that calcium, on rapid
entrance into the cell, could produce gellation instead of liquefac-
tion. Precisely what was happening in these experiments is
therefore not clear.

Heilbrunn also studied in coeruleus and in Arbacia eggs what he
called the surface precipitation reaction, or the formation of films
over crushed cells which prevents their explosive dissolution.
Calcium appears to be necessary for this reaction, presumably a
type of coagulation, for no film formation occurred when the
calcium was removed with ammonium oxalate. (Schmitt, as quoted
by Moore (1945) states that calcium has more affinity for water
than protein polar groups and therefore desolvates these groups
which then join with others to produce a more solid state.)
Magnesium could not replace calcium in this reaction but strontium
could. That cells do not supply their own calcium for this reaction
he explained by conceiving that intracellular calcium is bound
and not free.

Noting the difficulties of studying eflFects on the endoplasm by
simple immersion of a cell. Chambers and Kao (1952) micro-
injected solutions into the interior. Among other subjects was a
*' large variety " oi Stentor, quite possibly Wrw/^w^. They injected

246 THE BIOLOGY OF STENTOR

CaCl2 and SrCL in the concentrations used by Heilbrunn and
found that there was an endoplasmic clotting at the site of injection,
hence agreeing with his addendum and general thesis that calcium
has a clotting effect. It was especially interesting to me that the
clot was moved to the surface and pinched off, as also in amoebas.
Even when as much as two-thirds of the interior had been coagu-
lated the clot was still ejected and the cortex apparently not
violated.

Swimming of stentors was normal in solutions of CaCL and
SrCL, but if the ectoplasm was torn, the wound opened and there
was a clotting of the exposed endoplasm, much as in Heilbrunn's
surface precipitation reaction. Conversely, in NaCl and KCl tears
were never repaired, the endoplasm flowing out of the cut without
any sign of coagulation. The responses of Stentor protoplasm were
therefore quite like those previously found by the senior author in
Amoeba dubia.

Chambers and Kao also injected phenol-red and bromcresol-
purple into their unnamed Stentor and found that the cytoplasm
had a pH of 6-8 while that of the macronucleus was at least 7-6.
Correspondingly, Strom (1926), using very dilute mixtures of
phosphates to obtain a varying pH without specific ionic effects,
found that stentors are only slightly influenced by changes of pH
from 6-5 to 8-0.

8. Tests for an antero-posterior metabolic gradient

In pursuing his theory of metabolic gradients in organisms.
Child (19 14) subjected coeriileiis to a respiratory depressant, KCN.
The animals were promptly disintegrated, starting from the frontal
field and membranellar band and extending posteriorly over the
lateral ectoplasm. In a few cases there was a secondary wave of
disintegration beginning at the posterior end. The species poly-
morphus gave a similar response, though difficult to follow because
of the unpigmented cortex. A number of other ciliates also showed
graded disintegration of the cell. It was concluded that there is an
antero-posterior gradient and that this is metabolic in character.

On another species (" probably roeselW) Child (1949) used a
more subtle approach in studying the intracellular reoxidation of
reduced Janus green and methylene blue. Oxidation changes the
green dye to red. This color change passed in a wave from the

ANALYSIS OF STENTOR 247

anterior to the posterior pole followed by a wave of ectoplasmic
disintegration in the same direction. The membranellar band,
especially at the level of the basal bodies, showed the sharpest
effect and therefore seemed to be a site of vigorous oxidation.
Methylene blue gave essentially the same results and Child felt
confirmed in his demonstration of a metabolic gradient in Stentor,

Confirmation also seemed to be evident in the work of Monod
(1933) who studied the differential susceptibility of different parts
of the cell in Stentor and other common ciliates to ultraviolet
radiation. Again there was an antero-posterior gradient in disinte-
gration of the ectoplasm.

Although Weisz ( 1 948d) confirmed Child in regard to the gradient
response of stentors to KCN and KMn04, he stood strongly
against Child's interpretation. This was largely because he had
found (Weisz, 1948a, c) that, other factors being the same, the rate
of oral regeneration and subsequent growth of fragments was
independent of the level of the body from which they were taken,
though of course the shape of Stentor does not admit of much
variation in this regard.

Holding that primordium formation and rate of growth are more
indicative of metaboUc state than is cellular disintegration, Weisz
denied the whole concept of metabolic gradients as applying to
Stentor and also questioned its applicability to other forms. He
gave a new twist to these experiments by studying the disintegra-
tion of fragments of stentors in KCN. Anterior pieces disintegrated
from the anterior end the same as whole animals. Posterior frag-
ments began disintegrating not at their anterior ends but at the
holdfast. And middle pieces started disintegrating first in the
region of the contractile vacuole. Hence he viewed the Stentor cell
as a heirarchy of structures which vary in their susceptibility to
external agents, and this is a function not of any cellular gradient
but of the organization of those structures, although it is stretching
the point to say that the contractile vacuole is more highly organized
than the lateral ectoplasm of middle fragments.

In answering Weisz, Child (1949) seems to be saying that
Stentor is not a good form for studying this problem an^^vay
because it does not have the long and cylindrical shape of worms
and hydroids. But we shall shortly describe that there are a great
variety of simple salts and other substances, not directly related to

R

248 THE BIOLOGY OF STENTOR

respiration or metabolism, which produce shedding of the mem-
branellar band in Stentor followed by a wave of disintegration
passing over the ectoplasm towards the posterior pole. I am
therefore inclined to agree with Weisz that the localized disintegra-
tive action of various solutions is a function of the special state of
organization of the different parts of the cell cortex.

9. Acquired tolerance to external agents

Pre-treatment of organisms with sub-lethal concentrations of
killing agents generally increases subsequent tolerance of originally
lethal concentrations of the same substances ! This adaptation has
also been demonstrated in Stentor. Davenport and Neal (1896)
succinctly summarized their studies on coeruleus. ''Stentors reared
for two days in a culture solution containing 0-00005% mercuric
chloride resist a killing solution of o -001% HgCl2 nearly four times
as long as those reared in water. Similar results were obtained by
use of quinine. " This was not due to the selection of resistant
individuals but a genuine acclimatization, because no deaths
occurred at the lower concentration and the same individuals were
carried into the higher. Nor was this a general adaptation to
increased osmotic pressure ; for the concentrations used were very
low, and NaCl solutions of the same osmotic pressure gave no
increased tolerance to the killing agents. The increased immunity
was acquired rapidly, measurable resistance developing after i or
2 hours exposure to the sub-lethal concentrations, gradually
increasing thereafter until exposure of longer than 96 hours gave
no further resistance. The stronger the acclimatizing solution the
greater the resistance developed until the strength was such that
the lethal effects were additive. Killing, by disintegration of the
cortex of the cell, occurred about three times more rapidly at 23°
than at 15 °C, indicating that death was caused by a chemical
reaction.

Similar effects were studied in the response of coeruleus to
alcohols and glycerine by Daniel (1909). Animals lived well for
weeks in 1% ethanol, were destroyed by 6 hours in 2%, and died
in 2 hours at 3%. At lethal concentrations the body cilia soon
stopped beating but the membranelles remained active up to the
time of death. Two different stocks showed notable differences in
regard to acquired tolerance. In the first stock, 1% solutions

ANALYSIS OF STENTOR 249

Stimulated the animal to great activity, accelerated division with
production of many smaller cells, and gave no acquired immunity
to higher concentrations. The resistance of this stock w^as already
high but Daniel showed that this did not obscure a fundamental
lack of acclimatization.

In the second stock, animals in i % ethanol were also excited to
increased activity but showed practically no increase in rate of
division, and they acquired a marked immunity as a result of
remaining in this weaker solution. For example, in 6% solutions
they died in 162 seconds if not acclimated but lived for 301 seconds
if pre-treated for 4 days in 1% ethanol.

If acclimated, 6% ethanol made the membranelles beat so
vigorously that the whole cell shook. The acquired tolerance was
a function of the strength of the acclimating solution and the length
of time the animals were exposed to it, appreciable immunity being
obtained by 4 hours ; with no further increases after 4 days exposure
to sub-lethal concentrations.

All stentors were killed in 8% ethanol. The membranellar band
and frontal field were the last parts to become quiet and begin
disintegration. Acquired tolerance for ethanol was not transferable
and gave no increased immunity to methanol.

In I to I /4th molar glycerine, pigment was not shed as in the
alcohols. Stentors remained motionless and then suddenly con-
tracted, whereupon the membranellar band was shed as a ribbon.
If rescued from the solution, survivors could then regenerate a
new set of feeding organelles. Otherwise the animals plasmolyzed,
beginning at the posterior end. Ethanol immunity was not trans-
ferable to glycerine and indeed only made the animals more
sensitive to the latter. Hence in general Daniel regarded his
findings as demonstrating Ehrlich's principle that immunity is
specific and non-transferable.

Daniel also found, as had Peters (1904), that stentors are not
tolerant to excess alkali or acid, and this has also been my
experience. Even very dilute solutions of hydrochloric acid pro-
duced rapid killing after the membranelles stopped beating and
pigment was shed. Apparently sodium hydroxide does not pene-
trate the cell so rapidly, and stentors could live for a remarkably
long time if the pellicle was not ruptured. The alkali caused a loss
of membranelles as in glycerine and the shed pigment became a

250 THE BIOLOGY OF STENTOR

" beautiful sea-green ". Body cilia beat as long as the ectoplasmic
structure remained intact.

10. Shedding of pigment and pellicle

The pigment of stentors is largely located in ectoplasmic
granules beneath the pellicle where it is often readily affected by
external agents. The species which has been studied is coeruleus,
observation of which indicates that pigment sloughing may even
occur under natural conditions, as was first suggested by Schuberg

(1890).

Loss of pigment occurs under three guises. A homogeneous
blue-green halo may be ejected, suggesting that the pigment
granules have been burst and their contents set free. The granules
may be cast off^ as such and appear as tiny particles, which seems
to be the case in natural sloughing. And finally, one or more layers
of the pellicle may also be shed, and in this case the outer surface
carries the granules with it where they remain in rows corresponding
to the pigmented stripes. It is surprising that the pellicle can be
sloughed without apparently interfering in any way with the cilia,
for the outer coating of the cilium is in all ciliates continuous with
the pellicle covering of the cell body. This also occurs even more
clearly in Blepharisma treated with strychnine in which the
animals swim out of the discarded pellicle (Nadler, 1929). As
already suggested, pellicular shedding may have been elaborated
as a method of case-making, both in certain species of Stentor and
in FoUiculina. In the latter, Andrews (1923) found that the form-
ing sac at first shows lines of pigment granules corresponding to
stripes on the body. An appearance very much like this can be
induced in coeruleus which forms no lorica.

In methylene blue, Neresheimer (1903) produced a separation
of the stentor ectoplasm, and it was in this way that he obtained
the pieces which he stained to demonstrate " neurophanes ".
Much later Weisz (1950a) obtained sloughing of pigment and
peUicle in Janus green.

Prowazek (1904) found that brief immersion of coeruleus in \%
NaCl caused a shedding of pigment as a homogeneous blue halo.
The coloration was then regenerated in about a day after returning
to normal medium. In the same year, Peters (1904) independently
made the same observation and carried the study much further.

ANALYSIS OF STENTOR

251

He found that pigment shedding was the immediate response of
stentors transferred to certain solutions and that animals could
even later divide in media which caused shedding. The colored
slough he described as of gelatinous consistency, a homogeneous
halo without granules. Such sloughing was produced in KCl,
NaCl, KNO3, Na2S04, (NH4)2S04, Na2HP04, NaOH, HCl,
lactose, and chloroform but no shedding occurred in CaCl2,
Ca(0H)2, CaS04, or MgS04. Hence monovalent cations which
are the ones producing reversal of ciliary beating also elicit the
sloughing response, but bivalent cations do not. The effect is
obviously not osmotic since lactose and chloroform gave a similar
result and CaCU protected the animals against the shedding effects
of Na2S04, although the osmotic pressure was correspondingly
increased. In chloroform and Na2S04 some layers of the pellicle
apparently were also shed as a " heavy coat ". Peters suggested
that the pigment is a protein which is dissolved by certain salts.

We have just noted that Daniel (1909) obtained shedding of
pigment in alcohols and NaOH but not in glycerine.

Peters' study was confirmed and extended in some of my own
investigations (Tartar, 1957a). I also found that monovalent
cations produced pigment shedding, while calcium and magnesium
salts did not. The most vigorous shedding occurred in NaHCOa,
NH4CI, and LiCl. In strychnine there was a violent shattering

\^ N

Fig. 70.

A. Shedding of pellicular layer and some pigment granules

in <S. coeruleus in i % ammonium chloride.

B. Casting off the membranellar band in 2% urea, a: Band
shed following fimbriation of membranelles, with last part to go
being the gullet lining, b: Neat removal of peristome in proper
treatment, with lateral and frontal stripe structures remaining
unaffected, c: Extended treatment produces two fronts of
disintegration, but if stopped (d) middle piece separates, survives,

and regenerates. (After Tartar, 1957a.)

252 THE BIOLOGY OF STENTOR

loose of the granules and the pellicle was shed in pieces, not as a
hull, as in the amazing demonstration with Blepharisma by Nadler.
Shedding of the pellicle was especially noticeable and clear-cut in
NH4CI, ammonium acetate, LiCl, and egg albumen (Fig. 70A).
Stentors apparently regenerate both pigment and pellicle when
returned promptly to culture medium, for their later appearance
was altogether normal. The concentrations employed were usually
1%, made up in the filtered lake water used for culturing. Attempts
by repeated treatment with salts to obtain stentors which were
completely devoid of surface pigment granules and could not
recover them later were not successful. Granules located in the
endoplasm (Weisz, 1949a) may have been mobilized (and
multiplied) to take their place.

These shedding responses might therefore be useful in tracing
the origin of the pigment granules during their rapid regenera-
tion, as well as in testing the consequences for respiration of greatly
reducing the number of cortical granules. And treatments causing
a neat shedding of the pellicle should provide a means of studying
the significance of this layer in permeability as well as in
immunological reactions.

II. Shedding of the membranellar band

In addition to producing extrusion of pigment, Prowazek's
(1904) ^% solution of table salt caused the shedding of the mem-
branellar band in coerideus; Daniel (1909) obtained such cast-offs
with glycerine. These reports were generally neglected until,
independently, I found the same effect when stentors were sub-
jected to 25% sea water (Tartar, 1957a). I then tested several
chlorides, sulfates, acetates, sugars, urea, and albumen — usually
in 1% solution. All produced sloughing of the membranellar band
with one exception. This was ethanol in which, as in the studies of
Daniel, the membranelles remained completely intact and beating
as the last part of the cell to disintegrate. In all treatments which
produced sloughing, the animals could later recover and regenerate
the feeding organelles, with the single exception of NiS04 treat-
ment. The typical response was for the stentors to swim about in
agitation, backwards in the monovalent cations, then suddenly
contracting as if the agents had succeeded in penetrating deeply.
Following this contraction, the membranelles became fimbriated

ANALYSIS OF STENTOR 253

and usually the major portion of the band was cast off, including
both membranelles and a basement ribbon (Fig. 70B). This effect
was all or none ; for although only part of the band might be shed,
there was no case in which the band was simply injured while
remaining in place. When the animals were left in these solutions,
the wave of disintegration of the ectoplasm passed over the frontal
field and proceeded posteriorly down the lateral surfaces, often
being met by a corresponding wave originating in the holdfast and
moving forward (b). Hence the appearance was just like that of the
disintegration in KCN demonstrated by Child (1914). Again,
there was an all or none effect, the ectoplasm becoming totally
disintegrated or remaining intact with cilia beating. Even after
half the ectoplasm was destroyed, disintegration stopped at once
on return to normal medium and the remaining part could still
survive and recover. The disintegrated ectoplasm as well as the
underlying endoplasm was then pinched off to leave a viable
mid-fragment (b).

It seems odd that the membranelles, with their deep-lying basal
plates, should have been the first to go, but this was clearly the case.
In fact, urea and sucrose treatments gave very neat shedding of
the membranellar band if treatment was stopped promptly,
without the frontal field or lateral ectoplasm being affected in any
way. Sea water and most of the other treatments caused a lifting
of the band first at its distal end. Specimens were often obtained
in which the mouthparts remained intact, complete with their
membranelles; but in Holtfreter's solution the membranelles fining
the gullet and bordering the oral pouch were usually the first to go.
Therefore it is possible by choosing the proper treatment to
produce specimens in which the mouthparts alone are complete,
and others in which only these parts have been subjected to dele-
tions. This technique is also convenient for producing large
numbers of animals in simultaneous regeneration (p. 353), or for
inducing primordium formation in graft complexes without the
need for cutting operations which might disturb a contrived
arrangement of the lateral striping.

Oral primordia were also shed in salts, urea, and sugars. The
more advanced its development, the more likely was the primor-
dium to be shed. Sloughing usually began at the anterior end and
proceeded posteriorly. At early stages, on the contrary, in which

254 ^^^ BIOLOGY OF STENTOR

oral cilia had not yet grown out to their definitive length, the anlage
was notably resistant and resembled in this regard the general
ectoplasm from which it presumably arises. Stage 4 is the time
when the primordium becomes susceptible to sloughing. It may
be inferred that development involves the elaboration of a certain
type of organization which is peculiarly sensitive to these external
agents.

We do not know how this shedding of the membranellar band
is brought about but we can at least exclude some possibilities.
Osmotic pressure probably plays no part because even very weak
solutions of Na2C03 produce sloughing, and sugars do likewise
even long before the cell begins to collapse. The action is not ionic
because it is shown by neutral substances like sugars. There seems
to be no relation to valency, for both NaCl and CaCl2 produce
like results. Hydrogen ion effects are ruled out by the efficacy of
neutral substances. Nor does the result appear to be due to injury
as such, since the most drastic operations with a glass needle do
not produce it. The great variety of substances producing the
effect itself poses a difficulty to analysis.

12. Morphogenetic effects

After encountering such striking and specific effects on the
stentor cell of relatively simple compounds in lethal concentrations,
I prepared sub-lethal solutions which obviously affected the animals
but allowed their indefinite survival (Tartar, 1957a). When stentors
which had shed their membranellar band in sucrose were not
washed before replacing in normal medium regeneration was
delayed from i to 2 days, doubtless because of the carry-over of
some of the sugar. Otherwise the animals were entirely normal in
their behavior. Hence sugar is in itself an effective inhibitor of
primordium formation.

Diluted sea water also gave reversible inhibition of oral regenera-
tion. Often there were graded effects, depending upon the con-
centration in the solution and the susceptibility of the individual.
Sometimes regeneration was merely delayed. In other individuals
or at different concentrations there was formation of a complete
membranellar band but with inadequate development of the
mouthparts, only a small pit being produced. A further influence
was shown when the anlage was arrested in mid-development at

ANALYSIS OF STENTOR 255

Stage 4 and remained as such. All these effects are quite like those
obtained with acriflavin by Weisz (1955). Dilute sea water also
produced cases of aborted fission without separation of the daughter
cells, as well as ectoplasmic lesions which produced a sort of
self-mincing in which the ectoplasmic striping was broken up into
irregular patches (Fig. 71 a).

Fig. 71.

A. Effects of sub-lethal dilutions of sea water on S. coeruleus.
a: Reversible inhibition (or delay) of regeneration, b: Arrest
of primordium development at stage 4. c: Incomplete stomato-
genesis. d: Incomplete division, e: Astomatous regeneration

and breaking of stripe pattern into disorganized patches.

B. Reactions to sub-lethal concentrations of lithium chloride
include the above, as well as (a) a more exaggerated self-mincing
of the stripe pattern with corresponding abnormality of cell-
shape, (b) stacking up of oral sets due to resorption failure in
repeated reorganizations, and (c) extraordinary increase in
breadth of the cell from hypertrophy of striping, leading to
spontaneous formation of self-reproducing doublets. (After

Tartar, 1957a.)

Lithium chloride in concentrations of o-i to 0-005% g^^^
reversible effects which were especially interesting (Fig. 71B).
Again, primordium formation could be reversibly inhibited for

256 THE BIOLOGY OF STENTOR

I to 4 days, with complete regeneration occurring on return to
normal medium. Regeneration was sometimes merely delayed,
astomatous, or blocked in development at stage 4, which is just
the stage at which the primordium becomes susceptible to shedding.
Abortive fissions were noted, as well as distortions of body
striping. In successive reorganizations, LiCl prevented the resorp-
tion of the old organelles, with the result that there was a stacking-
up of several sets of feeding organelles, as shown. Perhaps most
provocative of all was that the stentors became very broad, as
if multiplication of lateral striping had been stimulated much
beyond the normal bounds, and in fact some of these animals
spontaneously converted into doublets, as illustrated.

The only previous test of the eifect of LiCl on ciliates to my
knowledge was that of Faure-Fremiet (with J. Ducornet, 1949)
who found that this agent produced microstomatous forms in
Tetrahymena. This he attributed to inhibition of the multiplication
of cilia, yet it appears that, in Stentor, broadening of the cell is
accompanied by increase of kinetics and therefore considerable
multiplication of cilia and related structures.

It is well known that lithium has special eifects on developing
eggs, producing in general a vegetalization or depressing a gradient
whose maximum is at the animal pole (see Gustafson, 1954). The
precise nature of this effect is not known. It may be that lithium
alters the hydration of proteins, for it seems to produce a coarse-
ness of the cytoplasm in general and to cause proteins to become
fibrillar, coagulated, and stable. Raven (1949) states that lithium
seems to affect especially the density of the cortical cytoplasm in
the eggs of Limncea. Since the major morphogenetic events in
stentors are also located in the cortex of the cell, the effects may
be comparable and one might even regard suppression of differen-
tiation of the feeding organelles and concomitant broadening of
the lateral ectoplasm as a parallel of *' vegetalization ".At any rate
it is most interesting that lithium has unique effects upon Stentor,
as it does upon embryos.

13. Inhibition of growth by X-ray, and other effects

Kimball (1958) subjected coeruleus to X-rays and found that
when irradiated animals were returned to culture medium fission
was much delayed. Although the stentors fed and formed food

ANALYSIS OF STENTOR 257

vacuoles they grew slowly or even decreased in size. But they
could form primordia in reorganizations apparently induced by
the irradiation, or in regenerations following transection. The
nuclei were likewise apparently unaffected. At least some digestion
evidently occurred, because starvation controls decreased in size
much more rapidly than the irradiated stentors. Therefore,
Kimball concluded that X-rays inhibit growth by decreasing or
blocking the net synthesis from feeding, though not preventing
such synthesis as occurs in the building of primordia. Delayed
fission would hence be due to the animal's failing to attain fission
size or only slowly achieving the maximum volume, and not to
incapacity to form primordia or undergo the nuclear changes
which accompany division.

These effects were independent of the presence or absence of
oxygen during irradiation. But anoxia combined with X-ray
(irradiation in an atmosphere of nitrogen) resulted in deformities
such as ridges, flanges, in-pocketings, and extra tails and sets of
feeding organelles. It seems likely that these abnormalities resulted
from breaks in the ectoplasm followed by improper healing; for
the forms described resemble those obtained by disarranging the
ectoplasmic pattern mechanically so that disjunctive areas no
longer coordinate (see Figs. 65B and 66e). If so, it may be suggested
that in these operations the separated areas join but fail to achieve
intimate union because of misorientation, while in the radiation
studies the orientation is at first correct, but intimate rejoining,
say, of the fibrous structures of the clear stripes, is inhibited as an
after-effect of irradiation. This in itself would be an interesting
effect, though still leaving the question why outgrowth and
joinings of fibers, which undoubtedly occurs during oral
primordium development, is not also inhibited.

14. Effect of temperature on size

A statistical sudy of variation in dimensions with temperature
(Zingher and Fisikow% 1931) showed that mean size of stentors
increases with rising temperatures up to a certain limit. Natural
collections accordingly showed a similar enlargement, and also an
increased coefficient of variation, from winter to summer. Since
the measurement curves were unimodal, apparently only one race
of coeruleus occurred in the pond sampled. Nor was illumination a

258 THE BIOLOGY OF STENTOR

factor, because animals cultured in light and in darkness did not
differ in measurements.

All these studies attest the accessibility of stentors, as free-living
cells, to the action of external agents, and, by the often precise
selectivity of the animal's reaction, their suitability for the pursuit
of problems in cell physiology.

CHAPTER XV

METABOLISM

I. Effects of starvation

Apart from eventual death, the most conspicuous response of
protozoa to suspension of feeding is an often marked decrease in
size of the individual. That is to say, the animal consumes some of
its own substance before dying. Protozoa differ in the extent of
reduction w^hich is possible. Paramecia decrease little before they
shrivel and die, but Dileptus (Visscher, 1923) and Amoeba
(Hartmann, 1928) can persist and dwindle to i/iooth, and
Didinium (Mast, 1909) to i/6th their original volume. Bursaria
truncatella can diminish in length from 500 to 90 /x as the feeding
organelles become proportionately smaller (Lund, 19 17), and
Dembowska (1938) showed that under starvation Stylonychia
repeatedly reorganizes on a smaller scale until very tiny animals
are produced. When great latitude in size is permitted, the protozoa
do not simply become thin and emaciated like starving vertebrates ;
as with many invertebrates, including hydras and flat worms, they
become proportionately reduced in most of their parts so that they
may properly be called dwarfs. Minute forms are not only the
result of individuals consuming their own substance but may
possibly also involve so-called " hunger divisions ", or an initial
persistence of the rhythm of fission in spite of decreasing size
during the first days of starvation. The two factors are not easily
separable when dealing with large samples difficult to count. Yet
Maupas (1888) confirmed Gruber (1886) in reporting that large,
well-fed coeruleus, when isolated, continued dividing 3 or 4 times,
producing smaller than normal individuals. Division without
attaining maximum size was indicated, though on Maupas'
evidence the stentors must not have been totally without food
since division products much larger than one-eighth, say, of the
maximum volume were produced. In my experience, on the
contrary, even large stentors very seldom divide after they are

259

26o THE BIOLOGY OF STENTOR

isolated into a large drop of coarse-filtered medium on a depression
slide. The reality of hunger divisions in stentors therefore remains
still in question.

Many have observed that in the largest stentors, coeruleiis and
polymorphus, dwarf forms appear under conditions of starvation
(Maupas, 1888; Johnson, 1893; Popoflr, 1909; Prowazek, 1904
and Schulze, 1951). Stolte (1922) observed both large and small
forms in starving cultures. I have myself frequently noted a similar
range in size which is not always correlated with cannibalism.
Possibly the larger forms are animals which had recently divided,
do not then divide further after food is withheld, and therefore
would gradually diminish only through the utilization of their own
substance. To complete this historical resume Sosnowsky (1899),
as reported by Sokoloff (1923), stated that division in stentor is
stimulated by starvation, and that the macronuclear membrane
disappears under these conditions. Ivanic (1927) contributed the
equally improbable notion that, when feeding is stopped, stentors
and other protozoa actually increase in size as they use up the
remaining food but fail to divide.

Several visible changes besides decrease in cell volume occur
during starvation. In coeruleus and perhaps in other pigmented
stentors the coloration tends to disappear. This fading to nearly
white is conspicuous in single animals long isolated on slides, but
larger samples in a culture dish remain fairly green for a month or
more though starved. Weisz (1949a) thought that the pigment
granules are digested during starvation. Granular bands do seem
eventually to disappear in isolated animals, but only as death
approaches. But stentorin itself is certainly not easily assimilated
in cannibals and the pigment may even be ejected as waste.
Pigment changes are therefore enigmatic and require much more
study. Stolte (1922) emphasized that starvation produces vacuoli-
zation of the endoplasm but this pathological state, again, is
prominent only near the point of death. An important alteration
which occurs only gradually is that the macronucleus becomes
reduced. On the evidence it cannot be decided whether this is
because the substance of the nucleus is drawn upon to maintain
life or because the nucleus is adapting in size to the decreasing
volume of cytoplasm, or both. That regulation of nuclear to
cytoplasmic volume is more important than consuming the nucleus

METABOLISM 261

as a reserve is indicated by the often poor survival of hypernucleated
stentors (p. 304).

It was Johnson (1893) who first noted that dwarf stentors have
smaller and fewer macronuclear nodes than well-fed animals at
any stage, and Prowazek (1904) provided further confirmatory
observations. AUescher (19 12) made a separate study of this
phenomenon. First she found, naturally, that decrease in size of
starving stentors was greatest at higher temperatures, at which
metabolism would be expected to proceed at a higher pace; but
rate of decrease also then fell oflF rapidly, as if definite limits to
reduction in size were met. Cool and warm cultured animals
eventually shrunk to the same small size. Apparently the nucleus
decreased but little at lower temperatures, while in warm cultures
under starvation decrease in the size and especially in the number of
macronuclear nodes was conspicuous: of the order of from 20 to 5.
In this reduction some of the nodes decreased in stainabiHty as
their substance was apparently transferred to adjacent beads of
the nucleus. The reduction was therefore especially one of surface
area. Her interpretation was that the nucleus as well as the
cytoplasm was consumed during starvation and that this is possible
in ciliates with widely dispersed macronuclei, such as Stentor and
Dileptus, but not in forms with compact nuclei, like Paramecium.
I have found, indeed, that P. caudattim forced to carry two
macronuclei do eventually resorb one entirely, instead of diminish-
ing both (Tartar, 1940). Perhaps it might with equal plausibility
have been suggested that paramecia cannot decrease the nuclear
surface further, while stentors with their nodulated nucleus can
and do, in adaptation to decreasing size.

In a clone of coeruleus I made some observations on starvation
dwarfs simply by isolating abundant samples in caster dishes and
allowing them to stand for a month without added nutrients. The
size of the ciliates decreased from a maximum diameter of 376 /x
to a minimum value of 94 /z. After two weeks of starvation the
nuclear picture was varied, for the number of nodes ranged from
6 to 16 and large and small nodes were frequently found within
the same individual. This indicates that the nucleus was still in
process of adapting to decreasing size of the animals. Eventually
the dwarfs contained only 5 or 6 nodes which were still large in
proportion to the volume of the cell yet smaller than those of

262

THE BIOLOGY OF STENTOR

normal animals (Fig. 72). The dwarfs had tiny, proportionate
feeding organelles and the number of lateral stripes was about
half the normal, indicating the morphological adaptation to decrease
in size. Johnson had found that such dwarfs undergo no irrever-
sible changes and are capable of complete recovery, growing and
dividing when later fed. I demonstrated that fragments of these
tiny stentors were capable of normal oral regeneration.

Fig. 72. Largest S. coeruleus compared to smallest individual
in starvation culture. The larger had a contracted diameter of
376 ju., 18 macronuclear nodes, and c. no pigment stripes. The
smaller : 94 /x, 5 nodes, and c. 56 stripes. Note that membranelles,
colored stripes, and nodes are not proportionately smaller in the
tiny stentor.

As expected, the fat and carbohydrate reserves (further discussed
in the following section) are exhausted during periods of starvation
(Zhinkin, 1930). According to Weisz (1949a) their utiHzation is so
rapid that the endoplasm is cleared of these reserves within a day.
My impression is that at least the carbohydrate stores, which are
clearly visible as white granules in dark-field illumination are
exhausted much more slowly. Moreover, when survival of anterior
halves with little if any such reserves was compared with that of
posterior halves bearing abundant reserves no conspicuous
advantage from the stored material could be demonstrated. This
may have been due to the abnormal conditions apparently involved
in isolating animals into small drops on slides, for some specimens

METABOLISM 263

even died before the carbohydrate stores were exhausted (Tartar,
1959a).

All these findings merely confirm that stentors, like other
organisms, are able to continue living for some time by consuming
nutrient reserves or their own vital substance during periods of
starvation. More precise studies of starving cells might dissociate
factors most dependent for their maintenance on continuous inflow
of new materials, or starving protozoa may prove particularly
sensitive and discriminatory in their response to specific additives
such as certain amino acids, in contrast to well-fed cells. At present,
one is above all impressed by the adaptive morphological changes
whereby starving ciliates become not merely shrunken but re-
formed on a smaller scale, as tiny but perfectly formed dwarfs are
produced.

2. Storage and utilization of nutrient reserves

Visible reserves in Stentor take the form of glycogenoid granules
and fat droplets. The first can be demonstrated by dark-field
illumination or Lugol's iodine stain and the second by Sudan III.
These reserves were the subject of an extensive field and laboratory
study of Zhinkin (1930) on polymorphus, with incidental observa-
tions on coeruleiis.

Carbohydrate reserves are present in the form of granules which
are concentrated toward the posterior pole. Clearly revealed by
Tyndall eflFect using side illumination against a dark background,
one can observe the precise location of these granules in living
coeruleus (Tartar, 1959a). In animals which have been cleared of
food vacuoles by withholding food organisms, these reserves are
seen to occupy a subcortical band, forward from the posterior pole
and discontinuous in the oral meridian (Fig. 73). This is the
regular and preferred location, though overstepped if the carbo-
hydrates are especially abundant. Soluble in hot water and staining
red with iodine, Zhinkin identified the granules as glycogen.
Weisz thought them suflftciently diflFerent to merit the name
paraglycogen. Fat stores are present in the form of tiny droplets
throughout the endoplasm.

Zhinkin followed stentors through an annual cycle. The most
conspicuous changes were that, with decreasing temperature in the
autumn, the carbohydrate reserves increased; but when the ponds

s

264 THE BIOLOGY OF STENTOR

were covered with ice and developed an oxygen deficiency these
reserves disappeared as the abundance of fat droplets increased.
This alternation of reserve stuffs suggested correlations with
temperature and oxygen tension, as well as the possibility that the
carbohydrate was converted into fats. Zingher (1933) in fact
maintained that both starches and proteins are convertible into
fats by ciliates.

Fig. 73. Nutritional reserves in S. coeruleus.

A. Photo showing location of glycogenoid carbohydrate

granules (cf. Fig. 17A).

B. Random distribution of fat droplets revealed by Sudan III

staining. (After Zingher, 1933.)

In laboratory tests, Zhinkin found maximum increase in number
and size of glycogenoid granules at 3-5 °C, while the primary con-
dition for fat accumulation was lack of oxygen. In general it appears
from his data that, naturally, an accumulation of nutrient reserves
requires a temperature which is not so low that metabolism is
sharply curtailed nor so high that increased activity in cell multipli-
cation utilizes the food directly and may even draw upon reserves
already present. Weisz found that fat stores were not used in
regeneration, and Zingher considered them necessary to a normal
condition of the cytoplasm. My impression from long observation
of stentors in culture is that well-fed animals always have abundant

METABOLISM 265

carbyhodrate reserves and probably fat stores as well. Only with
considerable trouble could coeruleus be divested of its glycogenoid
granules (Tartar, 1959a). The seasonal cycle which Zhinkin seems
to have well documented may therefore be the consequence of a
delicate and changing equilibrium between rates of feeding and
cell growth.

3. Respiration

The oxygen requirements of Stentor have been little studied.
Stolte (1922) remarked that decreased oxygen produced vacuoliza-
tion of the endoplasm. In a Russian ecological study which I
have not seen, Zhinkin and Obraztsov (1930) observed that
polymorphus and coeruleus are found in ponds only where there is
abundant oxygen : under ice, only where bottom springs provided
enough of the dissolved gas. Sampling of an Iowa pond showed
stentors to be abundant only near the bottom under nearly
anaerobic conditions according to Sprugel (1951), a result most
paradoxical since the same animals lived well when transferred to
jars in the laboratory. Oxidation-reduction studied by means of
color indicators in protozoa, presumably including Stentor, was
pursued by Roskin and Semenov (1933) in a study which was
not available to me. It has been observed (Whiteley, personal
communication) that in some races oi coeruleus the animals remained
near the bottom while in other clones they always collected near
the surface, suggesting that there may be racial differences in
oxygen requirement.

Using the Cartesian diver technique, Whiteley (1956) discovered
a marked and unique increase in respiratory rate in halves of
starved coeruleus containing all the macronucleus and hence having
abnormally high ratios of nucleus to cytoplasm. During the first
day the rate of respiration showed increases of as much as 175%.
This acceleration was correlated solely with the nucleo-cytoplasmic
ratio and was repeated after a second removal of cytoplasm.
Appropriate controls demonstrated that neither cutting nor
regeneration were responsible for the increase. Enucleates showed
a low and gradually decreasing rate of respiration ; that of whole
animals, high and also only decreasing. In the critical macro-
nucleate halves the accelerated respiration temporarily approached
the values shown by whole cells. The whole macronucleus there-

266 THE BIOLOGY OF STENTOR

fore tends to form the respiratory system of a whole, even in a
diminished amount of cytoplasm.

Yet it is interesting that when de Terra (1959) forced all the
macronucleus into one daughter cell during division and so pro-
duced coeruleus with twice the normal macronuclear complement,
the uptake and incorporation of radiophosphorus was not different
from that of normal cells.

As relating to energy metabolism, sites of acid phosphatase in
coeruleus were determined by Weisz (1949b). Positive tests were
obtained around macronuclear nodes and other bodies in the
endoplasm as well as at the basal granules of membranelles and
body cilia. Enzyme activity decreased during starvation but not
during periods of morphogenesis, and appeared in oral primordia
only after cilia were present and active. Therefore, acid phosphatase
is probably involved in the action but not in the development of
ciHa.

4. Digestion

In Folliculina and Stentor the transit of the food vacuoles is not
in a definite track provided by cyclosis of the endoplasm, as in
Paramecium, but each is individually handled, according to
Andrews (1955). Schwartz (1935) described the normal digestion
of Colpidium by coeruleus, which required about 20 hours and
included dissolution of the nucleus of the prey. In feeding
enucleates, however, the food vacuoles from the start contained
excess fluid and at no time was digestion normal or complete. It
will also be recalled that Schwartz found indication that even a
substantial reduction in number of macronuclear nodes resulted
in abnormal metabolism with inadequate digestion.

Meissner (1888) reported that stentors (apparently ^ofymorpAw^)
take up and digest starch grains. This was confirmed by Zingher
(1933) who also noted increased fat droplets following the starch
meal, suggesting that carbohydrates are converted into fats; yet
coeruleus predominantly rejected starch grains in the feeding
studies of Schaeffer (see p. 11). Contrary to Meissner, Zingher
found that coeruleus ingests the fat droplets of milk, which he
thought were assimilated directly because rapid cell multiplication
followed. In rich cultures with little oxygen, digestion was inhibited
and stentors became stuffed with undigested food vacuoles.

METABOLISM 267

according to Stolte (1922), a condition which could be corrected
by supplying oxygen through algae added to the medium. His
interpretation was that oxygen is necessary for the elaboration
and activity of digestive enzymes.

5. Symbiosis with green algae

As in many other ciliates and in simpler metazoa, certain stentors
may bear spheroid, grass-green cells of Chlorella living within
them. These species include polymorphtis, igneiis (Balbiani, 1893),
amethystinuSy and niger (Maier, 1903). The algae reside in the
endoplasm (Johnson, 1893), where they are scattered at random.
That the relationship is symbiotic is shown by the demonstrations
that the stentors in question can live without the algae, that the
algae do not disintegrate on death of the stentor (Balbiani, 1893)
and may even continue life as free-living cells, and that both
stentors and algae receive advantages from their association.

The first experiments on symbiosis in Stentor were made by
Prowazek (1904). He reported that chlorellae can multiply within
dead and crushed polymorphus, indicating that the algae in stentors
can probably be grown in '' tissue culture " like those of Para-
mecium bursaria (Loefer, 1936). Having obtained artificial sym-
biosis by infecting Stylonychia and Euplotes with free-living
Chlorella, Prowazek tried unsuccessfully to obtain the same with
Stentor coeruleus. Failure occurred in spite of the fact that the
coeruleus digested the chlorellae only partially and ejected the
remainder. Even the enucleated stentors would not accept chlorellae
intimately into their cytoplasm, ingesting the algae but retaining
them, undigested, within food vacuoles. Prowazek concluded that
the cytoplasm of coeruleus is unfavorable for Chlorella and he there-
fore doubted Kessler's (1882) report that symbiosis can be
established between this Stentor and the chlorellae from a fresh-
water sponge. Certain experiments of mine were confirmatory
(Tartar, 1953). In interspecific grafts (a convenient method for
introducing symbionts) any substantial admixture of coeruleus
cytoplasm with polymorphus resulted in the ejection of symbionts
natural to the latter species. In Prowazek's observations neither
enucleated polymorphus nor enucleated coeruleus with ingested
chlorellae survived longer than controls without algae.

The most comprehensive study of symbiosis in polymorphus has

268 THE BIOLOGY OF STENTOR

been that of Hammerling (1946) and Schulze (1951) who began
their work together but published separately. Their findings will
be reviewed together, noting points of difference in interpretation
or observation. In the cultures of polymorphus, the stentors
collected appropriately at the lighted sides of aquaria, yet too
intense an illumination was detrimental. Hammerling remarked
that the stentors divided only at night or in the dark. This observa-
tion may be important in providing a means for obtaining
simultaneous fission of animals in well-fed cultures.

Stentors were not easily divested of their symbionts. When
grown in the dark, chlorellae decreased greatly in abundance but
a few algae were always retained, mostly toward the posterior pole.
Persisting symbionts in these pale ciliates might therefore have
been removed by cutting and culturing a number of anterior
fragments. Instead, the method employed by Pringsheim (I.e.) for
Paramecium hursaria was used: pale stentors previously grown in
the dark were cultured with abundant food (free-living algae) at
high temperature of 3o°C. Under these conditions for rapid
division, some stentors would outpace the chlorellae and emerge
entirely white. Three classes of animals from the same stock could
therefore be compared : green forms with abundant chlorellae, pale
stentors grown in the dark but always retaining some algae, and
white animals completely devoid of symbionts.

The presence of actively metabolizing chlorellae promoted the
survival of starving stentors. This was demonstrated by " feeding "
the symbionts while starving their hosts. Light and the mineral
nutrients in soil extract or Benecke's solution provided conditions
for metabolism and growth of the chlorellae, as proved by the fact
that symbionts did increase and pale animals became green when
only these factors were supplied. Conditions for starvation of the
stentors were established by withholding the free-living algae which
they had been eating and digesting, and by repeated transfers to
remove bacteria. Controls were afforded by comparing white with
green stentors and survival in darkness as well as in the light.

When kept in the light, green stentors survived twice as long
as white animals without symbionts, and pale stentors with few
but increasing chlorellae were in between. The capacity of the
stentors to undergo occasional fissions following starvation was in
the same order, green ones dividing the most. In the dark these

METABOLISM 269

differences disappeared and maximum survival times were the
same. Hence photosynthesizing chlorellae did confer an advantage
in survival of their starving hosts. Schulze further noted, hov^ever,
that in the dark the first individual to die was always a white one,
for minimal survival times were in the order : white, pale and green.
This result seems enigmatic since one would expect the catabolism
of stentor and symbiont to be additive. But according to
Hammerling, stentors receive some advantage from their
symbionts even in the dark.

What is the nature of the aid to survival provided by Chlorella ?
First, this succor was not complete, because stentors with thriving
chlorellae did eventually die. Therefore the ciHates did not become
autotrophic, or capable of living indefinitely through their
symbionts on light and mineral nutrients alone, as is possible in
the more complete symbiosis found in Paramecium hursaria
(Pringsheim, 1928). Schulze showed this deficiency to be the same
in igneiis as in polymorphus. At first Hammerling said that starving
stentors simply digest their symbionts, so that animals with
abundant chlorellae should live longer on these food reserves ; but
he later softened this conclusion. And Schulze found that the
hosts only partially and never completely digest their symbionts.
Moreover, any minor aid from partial digestion should be com-
pensated by multiplication of the chlorellae at the expense of their
hosts. Digestion of the algae was therefore probably not the basis
of longer survival. Hammerling observed that pale animals with
very few chlorellae nevertheless lived much longer in the dark than
white animals, though of course not as long as green ones, indicat-
ing that the algae wxre supplying some minor factor important to
survival. The possibility that polymorphus has lost the capacity for
the synthesis of one or more vitamins was suggested by Schulze's
finding that his white stentors could not live indefinitely unless
fed on green, free-living algae. This species of Stentor may
therefore have become dependent on plants for certain vitamins.

As stated by Hammerling the general conclusion from the tests
was that the presence of chlorellae enables stentors better to endure
a period of starvation. That such is the only advantage conferred
by the symbionts was further indicated by studies of the division
rate of well-fed stentors. He found that the fission rate was only
slightly higher if chlorellae were present and Schulze said that the

270

THE BIOLOGY OF STENTOR

multiplication rate was, if anything, slower. Evidently the sym-
biosis is of no particular advantage to stentors under optimal
conditions. In any event, rapid division led to decrease in
symbionts as the host outpaced the chlorellae.

Interesting experiments on the artificial development of sym-
biosis and exchange of chlorellae were also performed. Animals
without symbionts were put in an appropriate culture medium to
which was added a brei of crushed cells containing chlorellae.
White ciliates would then ingest the liberated symbionts, which

X ^^ aatoirophjj

hecame colorless laier

^, e/idurJn^ Sy/nbiosxs ;

autoirvphic

lahile or defeciive Symbiosis

jD^yJT

Tjor/nally \ /

-X no saiisfacwri/
symbiosis

■Makeup wihhdy^cdLty ^ hat yoocL
Symbiosis

•samajbui Chlorella £rom igneus
^ rendered capable o£

y / \ ° o* / Sipnbiosis

Symbiosis

Fig. 74. Exchange of Chlorellae and establishment of artificial
symbiosis between Stentor polymorphus, S. igneus, and Para-
mecium bursaria, diagramming data of Schulze, 1951. Letters
correspond to descriptive paragraphs in text, p. 271,

METABOLISM 271

were first encapsulated in food vacuoles but not digested and later
freed into the endoplasm. By this means normal symbiosis could
be reestablished in either Stentor polymorphus or Paramecium
bursaria provided with their own type of chlorellae. The procedure
itself was therefore adequate.

Exchanges of symbionts made by Schulze, in part confirmed by
Hammerling, are diagrammed in Fig. 74 and may be summarized
in the following propositions:

(a) Colorless P. bursaria readily acquired Chlorella from
S. polymorphus but did not become autotrophic as with their own
chlorellae. Enduring symbiosis was not established, for the ciliates
later became white. This was confirmed by Hammerling.

(b) In contrast, the paramecia established a true symbiosis and
became capable of autotrophic nutrition when chlorellae of -S". igneus
were substituted for their own.

(c) However, if chlorellae from a stock of igneus had previously
resided for 7 months in polymorphus^ they then gave no satisfactory
symbiosis when introduced into white P. bursaria. The host can
therefore alter the symbiont, and in this case the pyrenoid of the
igneus chlorellae was lost.

(d) As the preceding implies, colorless polymorphus could
establish an enduring symbiosis with chlorellae from igneus, in
spite of the fact that the algae were not readily taken up and the
hosts remained paler than polymorphus with their own chlorellae.
Using certain stocks of both species of Stentor, chlorellae from the
igneus were said to be rendered capable of free-living existence
after passage through the polymorphus, again indicating an eflrect
of host on partner.

(e) These independent algae could then quickly establish a good
symbiosis when taken up by colorless polymorphus.

(f) But when normally free-living Chlorella were offered to white
polymorphus no symbiosis developed.

(g) Colorless polymorphus estabUshed with chlorellae from
P. bursaria what Schulze called a labile partnership. Only with
difficulty was a symbiosis estabhshed, and this relationship per-
sisted only if bright light and soil extract were provided, but
Hammerling succeeded in maintaining a green culture for one year.
Notable was the fact that the chlorellae did not render the stentors
autotrophic as they do the paramecium.

272 THE BIOLOGY OF STENTOR

From these experiments it is clear that symbiosis with chlorellae
is a precise relationship involving a delicate equilibrium between
specific host and specific symbiont. Morphological differences
noted by Schulze certainly indicate that the chlorellae normally
found in polymorphus, igneus, and P. bursaria are distinct types.
Successful symbiosis was not related to the readiness with which
the ciliates took up chlorellae, for in some cases colorless animals
rapidly became green but not symbiotic, and vice versa. A first
requirement is that the chlorellae be not digested, though taken into
food vacuoles, and later infiltrated intimately into the cytoplasm.
It is not known how digestion is prevented, since the hosts seemed
to be able partially to digest symbionts when starved, and they
could also fully digest free-living chlorellae. The second require-
ment is that a harmonious equilibrium be established between
multiplication of host and symbiont such that the one does not
outpace the other in reproduction.

According to Hammerling (1946), Ohler (1922) and Pringsheim
(1928) have found that P. bursaria can take up different types of
algae, but free-living forms are not capable of substituting for the
natural symbionts. Apparently this is also the case with S. poly-
morphus, for I have found animals containing both Chlorella and

Fig. 75. Facultative multiple symbiosis, a : Stentor polymorphus
with both Chlorella and a needle-like algae free in endoplasm.
The needle forms were same as those growing in the culture.
b: On isolating onto slides, the needle algae were shed —
apparently alive — but not the Chlorella.

METABOLISM 273

bundles of a needle-shaped alga intimately within the endoplasm ;
but on isolation the latter were always eventually ejected (Fig. 75).
The transient residents were apparently picked up from those
free-living in the culture jar.

6. Parasites of stentor

A natural transition leads from symbionts to parasites, though
their effects on the metabolism of stentors have not been explored.
For the following account I have relied in part upon Kirby's
(1941a, b) excellent reviews.

Unidentified undulating filaments occurring in bundles within
vacuoles in the cytoplasm of stentors were observed by Miiller
(1856) and his students, two of whom considered these inhabitants
probably parasitic (Claparede and Lachmann, 1857). Balbiani
(1893) observed S. polymorphus with bloated nuclei in which the
macronucleus was parasitized and largely disintegrated by
''Holospora". The host was otherwise normal but its ultimate
fate was not determined. Sphcerophrya stentoris Maupas is an
unstalked suctorian which is both free-living and parasitic on
species of Stentor. In its parasitic phase this organism lives in the
cytoplasm and is without tentacles or cilia. Kalmus (1928) reported
it in S. roeseli. Hetherington (1932b) reported a cytoplasmic
invasion by bacilli in S. coeruleiis. The infection caused the animals
to become pale, but they could be ''cured" by repeated transfers
to fresh medium.

In S. coeruleiis and Spirostomum ambiguum^ Rowland (1928)
found an euglenoid with metabolic movements which appeared
commensal or endoparasitic. This she identified as Astasia captive
Beauchamp. The intruder restricted itself to the subpellicular
cortex. It was previously reported as an endoparasite in a rhabdo-
coele in France. Recently an apparently diflterent species of Astasia
has been described as a facultative parasite in S. coeruleus by
Schonfeld (1959). When well-fed stentors were presented with this
organism, grown separately in cultures, the stentors ate few and
digested those without ill effects. Starved animals gorged themselves
on the astasias, which were not digested but were liberated from
the food vacuoles and wandered about in the endoplasm. The
stentors eventually degenerated and died, a preceding vacuoliza-
tion possibly resulting from enlargement of the emptied food

274 THE BIOLOGY OF STENTOR

vacuoles. Filtered Astasia culture fluid also killed the stentors when
transferred into it.

Somewhat resembling Schonfeld's account, a student (William
I^ewis, 1959) has recently observed that coeruleus may ingest and
form food vacuoles of the flagellate Rhabdomonas incurva but the
prey passes through the stentors, emerging by defecation in a
living and active state. This could be the beginning of parasitism,
since at least many of the flagellates enter the stentor but are
protected against digestion.

We may choose this place to mention a report that stentors are
possibly toxic to other organisms. Otterstrom and Larsen (1946)
very doubtfully attributed kills of fingerling fish in hatchery ponds
to wild (but not to cultured) Stentor polymorphus producing toxins
only when irritated.

7. Abnormal stentors

Here we shall describe certain abnormalities in coeruleus relating
to the pigment granules which sometimes arise without operational
interference in cultures or stentor samples. Whether these aberrant
forms are due primarily to disturbance of metabolism or involve
other factors and even racial differences as well we do not know.
Very little is yet known about the abnormal animals themselves
or their origin, but they offer promising Unes for further study.

(a) Depigmented stentors

In starving samples of coeruleus and even rarely in normal
cultures one finds stentors which are nearly devoid of pigment
granules and appear colorless. Stolte (1922) thought that the
amount of pigmentation is a function of metabolism and decreases
with decreasing oxygen tension, but the presence of both green
and white forms in the same culture or sample implies that external
conditions are not wholly determinative. Colorless coeruleus were
seen by Schuberg (1890), Johnson (1893), and Gelei (1927).
Johnson concluded that the pigment granules had been excreted
or ejected, because green clots were found in the sample dish or
attached near the base of the stentors. In my observation white
coeruleus retain a very few pigment granules in bands which then
appear not granular but trabecular, indicating that the granular
bands do have some intrinsic structure. Hetherington (1932b)

METABOLISM 275

said that pale stentors were often infected by a great number of
bacilli and that normal ones showed none. A pathological cause
seems unnecessary, however, because every time I isolated one of
these stentors the normal coloration was eventually regenerated.
Why an occasional stentor should pass through this depigmented
stage remains a mystery and problem. A suggestive parallel is
found in the chimera studies, in which a small graft oi polymorphus
causes the depigmentation of a coeruleus host.

(b) OVER-PIGMENTED STENTORS

Sometimes one finds stentors which are very dark in color,
appearing deep red by reflected Hght. Such animals may appear in
the same culture in which depigmented stentors are found and
have also been observed to be capable of oral reorganization and
to recover the normal pigmentation. Generally these ''reds" were
smaller and thinner than normal stentors, and lacking in food
vacuoles, but the nucleus appeared normal. The pigmentation of
the granular stripes is in these specimens supplemented by an
overload of pigment granules in the endoplasm. Though some
recover, others become exceedingly abnormal and are characterized

b

Fig, 76. Over-pigmented phase of S. coeruleus.

A. Such specimens are smaller than the average (to the left)
probably by undernourishment because they show no food

vacuoles, and the coloration is very dark blue-green.

B. Later stage, showing disorganization, some extraordinarily
broad pigment stripes, and abundant pigment granules inside.

Alternatively such animals may return to normal.

276 THE BIOLOGY OF STENTOR

by extraordinarily wide pigment stripes in some areas of the
ectoplasm, which again suggests that an oversupply of granules
is present and crowding to expand the stripes (Fig. 76).

A

B

Fig. 77. Autonomously developing amorphous ^S. coeriileus.

METABOLISM 277

(c) Amorphous stentors

These bizarre specimens first appeared in crowded samples in
small dishes which were kept for two weeks or more without added
nutrient (Tartar, i959f). They have also sometimes developed in
animals isolated on slides. The usual course of changes leading to
gross abnormalities is shown in Fig. 77A. The animals are first
noticed as longer and narrower than usual. Then appears a central
mass of compact pigment granules which later becomes sharply
delimited from the surrounding cytoplasm. The form is increas-
ingly long and snake-like. Eventually much of the cytoplasm is
concentrated at the anterior end in a bouton, while a long tail
dangles behind resulting in a characteristic tadpole or vorticellid

A. Course of development, a: Possible initial stage in which
internal pigment granules accumulate among the carbohydrate
reserves, h: Narrow " snake " form with sharply bounded pig-
ment mass but normal-appearing macronucleus. c: Bulbous
* 'tadpole" form with long tail and transverse stripe arrangement
anteriorly, d: Increasing abnormality of shape and defective
stomatogenesis though nuclear nodes appear normal, e:
Recovery of one such specimen, with pigmented core lost,
normal feeding organelles and nearly normal shape. /; Usual
course tow^ard completely amorphous condition showing

projecting processes and followed by death.

B. Transmission of abnormality to normal stentor. a:
Amorphous stentor grafted in place of head of a normal, in ratio
of about 1:8 by volume, h: Gross abnormality developed
throughout by next day. c: "Tadpole" stage, with new reorgani-
zation peristome but no mouthparts. d: Amorphous by day 4
of the experiment, pigmented core now rather diffuse. Specimen

later recovered somewhat but died on the slide.

C. Similar abnormality produced by grafting normal
polymorphiis to normal coerideus. a: Enucleated polymorphus
without symbionts grafted to coerideus in proportions by volume
of 1:4. b: Reorganized as a doublet with good integration of
shape, but almost all of the cerulean pigment lost by influence
of the graft, c: Elongated form with large core of unpigmented
granules. Nuclear nodes appear normal, d: Development of

amorphous condition. Time-span shown: 7 days.

D. Photograph of amorphous coeruleus in various stages of
development: a: elongated cell; b: bulbous form; c: complete

loss of normal form.

278 THE BIOLOGY OF STENTOR

shape. Finally, the animals become completely amorphous,
showing incomplete feeding organelles and finger-like processes
extending out from the mass in all directions.

In these abnormal forms the nucleus appears quite normal.
There are several indications that the nucleus is not involved.
Pigmented cores have also appeared in enucleated stentors; but
these generally did not live long enough, apparently, to develop
the amorphous shape. When nuclei from abnormals were sub-
stituted in normal stentors no abnormality resulted; but when
abnormal cytoplasm was added by grafting to normals, the whole
fusion mass usually became and remained aberrant. Whole normal
cells were grafted to abnormal whole cells. In about half the cases
normals resulted; in the remainder, a normal stentor, even when
predominant in volume, was converted into an abnormal, and this
might even occur overnight. (Fig. 77B).

Amorphous animals could also be simulated by coeruleus to
which a minor piece of colorless polymorphus was grafted. Some-
times the central mass was of colored pigment granules, in other
cases the mass was colorless but granular (Fig. 77c). In the latter,
the mass may have been composed of depigmented coeruleus
granules or possibly of the non-pigmented ectoplasmic granules
typical oi polymorphus.

In some instances it was demonstrated that tadpole-shaped
abnormals could recover if transferred onto a slide with fresh
medium, but the introduction of normal animals into fluid dishes
in which abnormals had appeared did not result in their becoming
promptly abnormal. Therefore changes in the medium do not
seem to induce this condition. Racial differences may be more
important. Three races of coeruleus produced abnormal stentors
when starved, but five did not.

The development of such amorphous forms is much in contrast
to the usually amazing capacity of stentors to return to normal
after the most drastic operations and disturbances. A connection
with cancer or abnormal growth is suggested, first, by the stentor —
as an organism — "going wild", and second — as a cell —
transforming, among its normal fellows, into a pathological type.

Treatments which prevent or hinder primordiiim formation
may do so by disrupting the basis of protein synthesis, and these

METABOLISM 279

have already been summarized in the section on blockage of
regeneration (p. 132).

Speaking generally, metabolism studies on Stentor may be
fruitful in two directions. Demonstration of metabolic similarities
between these and other animals would prove the appropriateness
of using the special advantages of Stentor in pursuing problems of
"universal biochemistry". In addition, stentors may reveal or
provide initial discoveries of features of metabolism hitherto
unknown.

CHAPTER XVI

BEHAVIOR AND FUNCTIONS
OF THE NUCLEUS

Stentors have two kinds of nuclei, yet the nuclear story can be
simplified because the role of the micronuclei may be stated
briefly. On some signal from the cytoplasm during fission they
divide and reproduce themselves by typical ciliate endomitosis.
This also occurs during the related cycles of reorganization and
regeneration, in which the cytoplasm is similarly activated to oral
primordium formation. Although multiplication in these two
processes may be meaningless, that during fission obviously
assures the continued presence of micronuclei, available for their
single known function, namely, to produce a new macronucleus
after sexual reproduction or conjugation. In this, their performance
is indeed complex and will be described in the following chapter.
But for the vegetative life of stentors the micronuclei are
demonstrably without significance. In a few ciliates bearmg a
single, non- vesiculate, ''massive type" micronucleus the situation
may be otherwise. By massive type is meant that the micronucleus
is relatively large and appears on staining to have a consistency
much like that of the macronucleus. Thus in Ur onychia transfuga
(Calkins, 1911a), Euplotes patella (Taylor and Farber, 1924;
Reynolds, 1932), and in Paramecium caudatum and bursaria
(Schwartz, 1934, 1947; Tartar, 1940), excision of the single
micronucleus is immediately felt and manifests itself in defective
regeneration or reduced rate of fission, though later adjustments
may correct these deficiencies. But in Stentor coeruleus which has
numerous vesiculate micronuclei, a careful study by Schwartz
(1934, 1935) has shown that the micronuclei may be entirely
removed without any appreciable effect on the macronucleated
cell; and conversely, that, as in Bursaria according to Schmahl
(1926), in the absence of the macronucleus micronuclei which are
present are not only unable to carry on the life of the cell but are

280

BEHAVIOR AND FUNCTIONS OF THE NUCLEUS 281

incapable, without the stimulation accompanying the process of
conjugation, of regenerating the macronucleus. This result is
fortunate indeed, permitting us to neglect the micronuclei which
cannot be seen in the living animal and would be most difficult to
work with. Schwartz's demonstration of the ineffectiveness of the
micronuclei must have been confirmed many times in the study
of emacronucleate stentors; for although most of the adhering
micronuclei are no doubt removed with the macronucleus, a few
scattered ones probably remain, yet these specimens never
regenerate nor long survive. In what follows we shall therefore be
concerned only with the macronucleus, designating it as such or
simply by the word nucleus.

I. Location of the macronucleus

In small species of Stentor, such as igneus and multiformis , the
nucleus is a single ovoid mass near the center of the cell ; but in
forms like coeriileus and polymorphus, which are about a hundred
times larger, the nucleus consists of many parts or nodes in linear
sequence within a common nuclear membrane. In between, there
is roeseli with an elongated, nodulated nucleus and niger which is
about the same size but has the compact nucleus. It may be
significant that nodulation after division is delayed in roeseli
(Johnson, 1893) so that for a considerable time the nucleus has a
rod shape, which might be considered a transition form. The
species niger is conspicuously slow and lackadaisical in its swim-
ming movements. Phylogenetically this suggests that the former is
on the way to developing a moniliform nucleus out of one which is
rod shaped and arose by elongation of the compact form, as well
as that the latter is pushing the cell size as far as it can go on a
compact nucleus whose surface area of interaction with the cyto-
plasm is minimal with reference to the volume of nuclear material.
The chain nucleus of a form like coeriileus of course passes through
rod and spheroid phases during division and other times when
there is oral primordium formation. These changes in form led
Johnson to the conjecture that ontogeny is here repeating a
phylogeny in which the compact form of the nucleus can be
assumed to be the most primitive.

We shall confine our discussion to the well-studied chain nucleus
of coeruleus, but this description will serve fairly well for all species

282 THE BIOLOGY OF STENTOR

with moniliform nuclei. There is apparently a standard pattern
for the distribution of nuclear material in all large species. The
essential point is that with remarkable constancy the macronucleus
tends toward a definite arrangement with reference to the topo-
graphy of the cell and that if this pattern is not fulfilled or if
artificially disturbed the nodes can and do move in such a way
that they tend to recover the normal arrangement.

The typical disposition of the nuclear beads is shown in Fig. 78A.
With the possible exception of a few of the posterior nodes the
nucleus is usually entirely subcortical in its location. This is no
doubt why it can remain fixed in position; because it adheres to
the inside of the ectoplasm. We do not say attached, because one
has to allow for the movement of the nucleus during clumping and
in the correction of disarrangements. Generally, the longest part
of the nucleus is a row of nodes extending almost directly
posteriorly under the surface near the meridian which connects
the mouth with the posterior pole and considerably to the right
of the primordium site. From this row, beads extend around the
anterior end of the cell to the right for some distance, while at the
other end the row bends back on itself and terminates in several
beads which seem to be rather indefinite in their location though
they tend to place themselves on the opposite or left side. I think
this arrangement is what one would expect if he had to work with a
nucleus of minimal length for the sake of economy, to fasten it
inside the cell so that it would not "fall to the bottom", have
every point in the cytoplasm as close as possible to some nuclear
material, and place the nucleus as near as manageable to the most
active regions (the membranellar band, the mouthparts, the
primordium site, and the holdfast), while keeping it out of the way
of the migrating primordium, i.e., to assure that the ectoplasm
under which it lies will not undergo major shifts of position. The
actual deployment fulfills these assumed requirements.

Having made these statements, we now have to qualify them by
saying that the nucleus is not always in the same position, as well
as that its precise location is evidently not essential to the economy
of the cell. The first of these qualifications is documented in
Fig. 78B, which shows a number of the atypical arrangements
which have been found in stentors fished out of regular cultures
in which the majority of individuals showed a more normal

BEHAVIOR AND FUNCTIONS OF THE NUCLEUS 283

distribution of the macronucleus. Several of these forms have
already been described by Stolte (1922). He associated chains
having noticeably recurved ends with interdivisional addition of
new nodes ; and he observed that in stentors with double, parallel

Fig. 78. Location of the macronucleus in S. coeruleus.

A. Normal location, with post-oral row, right anterior wing,

and recurved terminal chain.

B. Atypical macronuclear arrangements appearing autono-
mously, including (a) secondary row, (b) forked chain, (c)
coiling and posterior shift, (d) scattered nodes (whether con-
nected was not determined), and (e) "situs inversus" in animals

of reversed asymmetry.

C. Divider with 13 nodes and posterior end excised produced
proter with 17 small nodes in normal location and opisthe with

13 in a ring, later becoming liistributed normally.

D. Doublet produced by engrafting an extra primordium
site (without nucleus) in time develops double macronuclear
chain, each deployed normally with respect to the stripe pattern.

E. When such a doublet reverts to single form the macro-

nuclear chain becomes single correspondingly.

284 THE BIOLOGY OF STENTOR

nuclear chains this condition is corrected after clumping and
renodulation in division. Stolte also made much of divergence from
the normal picture w^ith regard to the nucleus which occurs in
highly vacuolated stentors. But in pathological material one cannot
be sure that the vacuoles push the nuclear nodes out of place,
since the necrotic condition could affect the nucleus directly.
Certainly the vacuoles do not push the nucleus to the periphery
of the cell, as he said, because it is there already.

In regard to the second qualification — that the nucleus need not
be at a special place — v^e have the evidence that in stentors in
which all but one nuclear node has been removed this bead may
be variously located, yet such animals can survive and regenerate.
Still more satisfactory tests could easily be devised by shifting the
whole nucleus in such a way that it could not soon recover the
normal location, and it would be especially interesting to determine
whether the nucleus, separated by a narrow neck of cytoplasm from
the major portion of the cell, could support regeneration and growth
by diffusions across this bridge.

Certain regular changes in the distribution of the nucleus may
now be noted. When the posterior end of coeruleiis was cut off in
mid-fission it was consistently found that after renodulation the
nucleus in the opisthe had at first an abnormal arrangement, but
the typical disposition was later achieved (Fig. 78c). When graft
complexes become persisting doublets the nuclear chain is dupli-
cated even though the specimen started with but one (d), and when
doublets transform back into singles they soon achieve a normal,
single chain (e). In Fig. 59c we have a stentor developing a
secondary tail-pole which later became furnished with an extension
of the nuclear chain.

Similar deviations in the location of the nucleus in abnormal
forms have been observed in the related genus Condylostoma
(Yagiu, 1951, 1952). The simplest explanation of these cases would
be that the stripe pattern guides the location of the macronuclear
nodes. And the best substantiation thereof is that in cases of
reversed asymmetry, with the same cell shape as in normal animals,
the macronucleus assumes a reversed location (see Fig. 49).

2. Clumping of the nucleus

Typically, all nodes of the moniliform nucleus lie within a single

BEHAVIOR AND FUNCTIONS OF THE NUCLEUS 285

nuclear membrane, doubtless facilitating their coalescence into
one mass during oral redifferentiation and especially in division.
When fully deployed the strand connecting individual nodes may
be exceedingly tenuous (Fig. 79), yet rarely do terminal nodes break
off and not participate in coalescence during fission. Opinions
differ regarding whether separated nodes or sections of the nucleus
can rejoin their fellows. Stolte (1922) observed that broken nuclear
chains in vacuolated dividers condense separately, and I too have
found occasionally a double fusion mass; but he stated categori-
cally that permanent reunion was not possible because the nuclei
were within separate membranes. This supposed behavior is,
however, contrary to common experience with hypotrichs, such
as Oxytricha, in which widely separated macronuclei, for which
there is no evidence of a common boundary, nevertheless fuse at
every division. Much earlier, Prowazek (1904) had reported that
transected nuclear chains " regenerated" separately, producing two
rows of beads; but whether the nuclear volume was doubled he
did not tell us. It is probable that the normal nucleo-cytoplasmic
ratio was not upset.

Weisz (1949a) reported that broken nuclear chains can rejoin.
I severed the nuclear strand in coeruleus into five or more pieces
and its distribution was then for a while disturbed, but the animals
later became normal to all appearances, with the nodes closely
approximated in a single uniform row in the usual location

A

Fig. 79. Photographs of S. coeruleus showing macronucleus as

seen (A) in living stentor against dark field and (B) after feulgen

staining to reveal intemodal connections.

286 THE BIOLOGY OF STENTOR

(unpublished). Such nuclei have not yet been examined after
staining; all we can say is that either separated nodes fall into
perfect alignment or they are able to rejoin and become enclosed
again in the common membrane. The latter seems more probable.

Coalescence begins at both terminals and progresses toward
the mid-nodes, as Weisz (1950b) and others have noted. How is
this accomplished? Stolte suggested fusion by swelling of the nodes,
but I think what he observed was simply the coalescence of the
individual nodes into larger ones; and there certainly does not
appear to be an increase in the total volume of the nucleus during
condensation as his suggestion would imply. More probably,
internodal connections swell while the nuclear membrane contracts
and decreases greatly in area.

There remains to be conjectured why the chain nucleus should
clump at all. In both Loxodes (Balbiani, 1893) and in Dileptus
(Jones, 1 951) there is a distributed nucleus consisting of many
separate macronuclei and these do not fuse during division and
regeneration. Coalescence therefore seems not absolutely necessary
in the life of ciliates, though it may bring advantages. That no
nuclear changes occur until the oral primordium is nearly com-
pletely developed, as Balbiani (1893) first emphasized, does not
suggest that either coalescence or moving all parts of the nucleus
close to the primordium is necessary to its development. Hence
Balbiani did not share Gruber's idea that clumping is to give a
single, central guidance to morphogenetic events, because these
processes are nearly completed before the nucleus fuses; and
besides, the macronuclear chain is all one nucleus anyway.
Balbiani's suggestion was therefore that the nucleus concentrates
in order to have the greatest effect ; but the action of the nucleus is
more likely to be promoted by increasing rather than by decreasing
its surface. As already mentioned, Johnson held that coalescence
of the nuclear chain is an instance of Haeckel's law of recapitulation
and hence is performed for "historical" reasons. That the macro-
nucleus clumps in order to insure equal division through the simple
splitting of a compact mass has also been suggested (Sonneborn,
1947), but in Stentor at least, clumping does not insure this end
(see p. 71). Many hypotrichs like Eiiplotes produce reorganization
bands at every division and somehow transform or rework the
macronucleus ; yet in Stentor no one has found evidence of any

BEHAVIOR AND FUNCTIONS OF THE NUCLEUS 287

such change in the fine structure of the nucleus during clumping
and renodulation. According to certain of Weisz's findings (see
below), coalescence of the macronucleus would be necessary for
homogenization after individual nodes at its extremities had
become diverse, but this diversity could not be confirmed. It may
be that the nucleus in stentors coalesces into a single mass in order
to make possible its complete renodulation (Tartar, 1957b); for
in division the nucleus produces at one stroke about twice the
number of the original nodes which are now half-sized. This could
explain why coalescence is often not complete in regeneration and
reorganization : because the nucleus will generally return to about
the same size and number of nodes.

3. Nodulation

Following fusion into a compact mass the macronucleus extends
to a long and conspicuous rod or sausage shape which is then
renodulated, again, generally from both ends towards the middle.
An exception is roeselt, in which node formation proceeds only
from the anterior end.

As Johnson described it, nodulation seems to involve the aggre-
gation of chromatic substance into serial packets separated by clear
nuclear material where constriction then occurs; and he also
remarked that the new nodes are usually " beautifully symmetrical
and alike in size". Rarely there is produced a forked nucleus or
nuclear chain with a side branch (Fig. 78B), of which Johnson
illustrated one case in coeruleus, also noting that Stein (1867)
showed a similar case in polymorphus. The same have been observed
in Condylostoma (Yagiu, 1952) and in Spirostomum ambiguum
(Bishop, 1927). These may be ectopic rejoinings of separated
parts of the nucleus with the main strand.

Even after the period of renodulation, the number of nodes can
undergo small changes, reduction through fusion of adjacent nodes,
or increase either by the splitting of one node into two or by the
interpolation of new nodes between existing ones (Fig. 8oa). It is not
uncommon to find dumbbell-shaped nodes or one or more tiny
nodes lying between the larger. Prowazek (1904) first described
a step-wise increase in nodal number, and his account was
generally corroborated by Schwartz (1935). According to the
earlier investigator a node may either split in two or part of its

288

THE BIOLOGY OF STENTOR

substance may travel along the tube-like internodal connections,
stopping between two nodes and growing into a new one. Weisz
(1949a) stated that extra nodes are thus interpolated whenever the
strand between two nodes measures approximately 2 nodal
diameters or more. Yet it should be emphasized that major nuclear
increases occur only following primordium formation in division,
regeneration, and reorganization.

Fig. 80. Aspects of macronuclear nodulation in S. coeruleus,

A. Diagram of two means by which single nodes appear to be
added: {x) by division and {y) by interpolation between nodes.

B. When the clumped macronucleus of a stage-6 regenerator
is excised with a small amount of cytoplasm, the primordium
completes stomatogenesis and the nucleus attempts to renodulate

though isolated and confined.

C. When coalesced nucleus of a divider is sliced into several
times, the macronucleus is divided but fails to renodulate, doing
so only much later after regeneration is induced by excision of

mouthparts without injury to the nucleus.

I have found that when the clumped nucleus of a divider is
isolated into a small volume of cytoplasm it nevertheless attempts
to renodulate in spite of the confinement (Fig. 8ob). This suggests
that the impulse to nodulation is intrinsic with the nucleus itself
since the nuclear environment was so completely altered. Individual
nodes are very tough and resistant to cutting but the clumped
nucleus can easily be slashed through with a glass needle. In the

BEHAVIOR AND FUNCTIONS OF THE NUCLEUS 289

few cases in which this has been done in dividers, subsequent
nodulation was inhibited (unpubUshed). The nucleus extended as
a rod but remained as such without any further change until
regeneration was induced by excising the mouthparts, whereupon
the normal nucleus was recovered (Fig. 80c). Therefore, although
nodulation may be intrinsic, the stimulus for it is given by the
cytoplasm during the last stages of oral redifferentiation.

As previously discussed in the chapter on division, daughter
cells have about the same number of nodes as the parent (see Fig.
78c). Therefore coalescence of the macronucleus may be necessary
for its rapid renodulation into twice the number of nodes of about
half the size of the originals during fission.

4. Equivalence of macronuclear nodes

Nuclear fusions and clumping were also simply explained as a
means for recovering uniformity in parts which have become
diverse. In substantiation of this presupposition, Weisz (1949c)
claimed that an intranuclear difference does regularly develop
during divisional and reorganizational cycles in coeruleus, posterior
nodes (but not posterior cytoplasm) becoming increasingly incap-
able of supporting regeneration or even of maintaining oral and
caudal organelles as they approach the time of coalescence. Fusion
was then said to restore normal potency throughout the renodulated
macronucleus. Development and obliteration of differences seemed
to be correlated with degree of polymerization of nuclear DNA as
tested by methyl green staining (Weisz, i95o.b), though the reli-
ability of this determination was later questioned (1954).

A similar development of intranuclear heterogeneity was
proposed for Blepharisma (Weisz, 1949), but in this case the mid-
portion of the macronucleus which no longer supported regenera-
tion was destined for extrusion and dissolution anyway. The
general conception is brought further into question by Weisz's
interpretation that proximity to organelles is the basis for main-
tenance of potency in adjacent nodes; for the mouthparts in
Blepharisma are near the level of the mid-nodes which should
therefore retain their full capacities ; and in Stentor, the posterior
nodes should support maintenance and regeneration of the hold-
fast, contrary to Weisz's own account. Since even enucleate frag-
ments can regenerate the foot (Tartar, 1956c), it is indicated that

290

THE BIOLOGY OF STENTOR

the conditions in Weisz's experiments were not optimal. Further-
more, Suzuki (1957) could find no evidence for differences in the
potentiality of parts of the nucleus in Blepharisma at any stage.
Above all, I was unable by similar experiments to confirm that a
regular and recurring difference develops between the serial nodes
of the nucleus (Tartar, 1957b). I tested five different races of
coeruleus and observed not only oral regeneration but also recon-
stitution of the nuclear chain from a single node and the capacity
of fragments to reproduce indefinitely in clones. For comparison
I studied Condylostomum magnum^ which is a very long ciliate
with mouthparts far to one end and a uniform chain nucleus
running the length of the body. If posterior nodes regularly
become depotentiated, the same should be manifested in this form
even more than in Stentor. It was found that even the single,
terminal posterior node in many cases or at least the last four
could support complete oral regeneration at any stage (Fig. 81).
In stentors (where this was tested) pre-fissional and pre-reorganiza-
tional fragments with only such nodes could give rise to viable
lines with normal chain nuclei.

Fig. 81. Equivalence of macronuclear nodes in (A) Stentor

coeruleus and (B) Condylostoma magnum. Tiny fragments of early

dividers, carrying only a few of the most anterior or posterior

nodes, are capable of regeneration and more.

Evidently intranuclear differentiation is neither necessary for
nor the consequence of cytoplasmic differentiation in ciliates. We
can therefore return with confidence to the old dictum that any

BEHAVIOR AND FUNCTIONS OF THE NUCLEUS 291

portion of the macronucleUvS is sufficient for regeneration and
survival, implying that this nucleus is highly polyploid, with
representation of the complete genome throughout. The develop-
ment of internuclQSir differences as in mating type determination
in certain ciHates is of course a demonstrated fact. Had regular
intranuclear differences been confirmed, we would have been
provided with another and most excellent means for analyzing
nuclear functions. Our disappointment is however mitigated by
the consequence that we are now apparently free to continue
experiments on ciliates without having to take into consideration
an additional factor of varying capacities within the macronucleus.

5. Shape, size and number of nuclear nodes

When he removed all but a single node of the macronucleus
from coeruleuSy Prowazek (1904) found that the remaining bead
became much elongated and spindle shaped. Schwartz (1935)
observed the same, as well as that the remaining node may
become much flattened like a ribbon. Cases from my own observa-
tion are shown in Figs. 82A and 86c. These increases in the nodal
surface are as if to compensate somewhat for the great diminution
in nuclear volume.

Starting from a single node, the nuclear chain is regenerated
during episodes of primordium formation. Two reorganizations
seem to be required to recover the typical nodal number from a
single bead. This number is quite variable and in coeruleus is
between 6 and 20, with a mode around 15. K. M. Moller (unpub-
lished) has a race with a mode of lo-ii nodes and suspects that
the average number may prove to be a racial characteristic. When
de Terra (1959) implanted into enucleate coeruleus 2 macronuclear
nodes labeled with adenine-Ci4 she found that the regenerated
nuclear chain was labeled throughout, confirming the cytological
picture that this regenerative growth is not accomplished by simple
addition of new nodes. Instead, the implanted nodes increased
greatly in size, coalesced, and renodulated into many nuclear beads
from a common pool of macronuclear material.

Although the nucleus readily adapts by increase in size, the
occasions when it should decrease give an entirely different
impression. It appears difficult for the cell to take down or diminish
a too-large macronucleus. Starving stentors with decreasing

292

THE BIOLOGY OF STENTOR

cytoplasmic volume (Allescher, 19 12), or animals regenerated from
fragments with proportionately too much nucleus, tend at first
and for a long time merely to decrease the surface of the nucleus by
fusion of nodes (Fig. 82B). Nor did Hartmann (1928) observe
decrease in size of the nucleus in successive excisions of amoeba
cytoplasm. Yet indubitable decrease in nuclear volume eventually
occurs in hypernucleated Stentor fragments, as Prowazek (1904)
first reported. Stentors therefore certainly tend by nuclear increase
or decrease toward a nucleo-cytoplasmic ratio of limited range.

A

Fig. 82. Size adaptations of macronucleus in S. coeruleus.

A. Stentor with only 6 nodes, half of which are much
attenuated, as if to compensate with added surface and to make
a typically disposed nuclear chain. Mouth was excised and after
coalescence during regeneration 12 nodes were formed, normal

in shape, but in forked arrangement in this case.

B. Nucleate portion with 14 nodes is excised, regenerates
proportionate feeding organelles and reduces the number but

not the size of the nodes.

Size of nuclear beads in the row is generally quite uniform, with
the exception of interpolated nodes which are, at least initially,
very small. Daughter cells have approximately the same number

BEHAVIOR AND FUNCTIONS OF THE NUCLEUS

293

of nodes as mature animals but the nodes are at first small and all
presumably increase in size during the interdivisional period.
However, I have seen cultures of coeruleus which consistently
produced animals with non-uniform macronuclear chains, some
nodes being too large and others abnormally small. The lines
eventually succumbed.

There is some evidence that the size and number of the nodes
may vary with the conditions under which stentors are grown. At
25 °C Prowazek (1904) found that coeruleus had fewer nodes
(average of 8) than at i5°C (12). Following the earlier work of
Allescher (19 12), Stolte (1922) pursued this matter quite
thoroughly, though with what we would now call a primitive
control of culture conditions. For instance, to subject stentors to
reduced oxygen he simply grew them in tall cylinders. His results
indicated that cell size, macronuclear volume, and nodal size are
complex variables which probably both interact and are subject
to environmental influence. The results were summarized in a
table from which Fig. 83 was derived. Rich food, abundant
oxygen, and high temperature were correlated with a large number
of smaller nodes, and vice versa.

HIGH TEMPERArUKE

MUCH
FOOV

tlJTL£
fOOD

MUCH OXYGEN

large cell
many nodes
Small nodes
large arri't

ol macro n
rapid growih
& division

LIT TLB OXraEN

small cell
many nodes
large nodes
rehiivltj Ig

cutii. macron,
reduced fission

VOm T£MP.

Small cell
lav nodes
larye nodes
Small art^t.

o£ macron.
vacLLolaie
div'n . seldom

large cell
meditun no.
0/ nodes
large nodes
med- Size macron .
large vacuoles
rapid division

large cell
/eiv nodes
large nodes
smaE arri't.

of macron
large vacuotes
div'n. seldom

Fig. 83. Effects of environmental conditions on S. coeruleus,
according to and adapted from a table of Stolte's, 1922.

6. Control of nuclear behavior

Balbiani (1893) first emphasized that the macronucleus of
Stentor does not begin its major performance of coalescence until
the oral primordium is well developed (stage 5) and capable of

294 THE BIOLOGY OF STENTOR

completing itself even in the absence of the nucleus. This
chronology suggests that clumping of the nucleus has nothing to
do with primordium formation ; and in reference to fission (Johnson
(1893)) it indicates that the cytostome leads the nucleus — rather
than the reverse, as in the case of mitosis in cleaving eggs and
dividing tissue cells. We can readily suppose that the primordium
can go through its entire development in regeneration and reorgani-
zation without the act of nuclear clumping, and indeed this occurs
in experimental animals in which only one bead is left. Again,
since the products of artificially divided (transected) stentors
behave normally there seems to be no obvious reason why a
dividing stentor could not simply pinch the nuclear chain in two
where it originally lies. Perhaps we may put it this way, that if the
macronucleus is to clump, this has to occur before the cell divides,
if one daughter is not to be left without a nucleus.

Questions concerning control of nuclear behavior include the
following: What "tells" the nucleus when to fuse? Is the macro-
nucleus capable of autonomous division, or is it merely pinched in
two by the division furrow? Does the rod-shaped nucleus have to
be ''told" to renodulate? Less anthropomorphically, does the
macronucleus time its own phases or is it guided by the cytoplasm,
and if so, by what part of the cytoplasm?

Weisz (1951b) suggested that coalescence of the nucleus is
stimulated by primordium formation, for if the early division
primordium is removed or caused to be resorbed there is no division
and no compacting of the nucleus. Yet the division anlagen may be
removed at stage 4 or even stage 3 and fission can still be completed,
with the nucleus clumping normally just before furrow formation.
This implies either that it is not the primordium which gives the
stimulus for clumping or that the response of the nucleus to this
stimulus is much delayed.

Even if the primordium is not implicated, one could still
maintain as Weisz (1954) said, that ''Evidently, nuclear kinetics
depend on direct stimulus from the ectoplasm". It has already
been suggested that during primordium formation the cytoplasm,
and perhaps especially or exclusively the ectoplasm, is in a state
of activation. The ripening state of activation, or particularly when
this stage is changing over to that of inhibition at stage 5 or 6,
could therefore provide the stimulus or the means for coalescence

BEHAVIOR AND FUNCTIONS OF THE NUCLEUS 295

of nuclear nodes. This would explain Weisz's (1956) observation
that mid-division stentors grafted to non-dividers induce a
coalescence of the nuclei of both partners. Originally evolved for
the proper timing of nuclear clumping in division, this relationship
between cytoplasmic and nuclear states would also be effective in
producing a usually superfluous clumping during reorganization
and regeneration because the state of activation always accom-
panies primordium formation. On this hypothesis the nucleus
would fail to clump in dividers from which early-stage primordia
were excised or caused to resorb, not because the primordium is
missing as a stimulus to coalescence, but because the cytoplasm
again came under the inhibitive dominance of the intact feeding
organelles and the state of activation was aboHshed.

How, then, is the nucleus guided in elongating and renodulating?
Does it divide autonomously? These problems have engaged the
attention of Noel de Terra (1959) whose preliminary findings
she kindly communicated. Apparently the cytoplasm gives the cue
for elongation of the clumped nucleus, for if the compacted nucleus
of a divider was transferred to an interphase stentor no elongation
occurred. But if the condensed nucleus was transferred to a cell
with nucleus in the same condition, then the two nuclei elongated
together synchronously. Supportive are experiments already
described, indicating that if the nucleus is prevented from renodu-
lating at the close of division because of injuries suffered during its
compacted stage, then it remains as a rod in the interphase cell and
does not nodulate until the animal passes through another episode
of activation or redifferentiation (see Fig. 80c).

De Terra is also finding evidence that the macronucleus in
Stentor is generally incapable of autonomous division and therefore
has to be pinched in two at the rod stage by the division furrow,
though autonomous division of macronuclear anlagen occurs during
conjugation. This correlates with the cytological picture when
separation of daughter cells is prevented by injury to late dividers,
the compacted nucleus elongating and renodulating as a single
chain instead of two. When she caused stentors to divide very
unequally, the macronucleus was also unequally and propor-
tionately distributed to the daughter products, quite as if the
furrow cuts through the rod nucleus wherever it happens to strike.
Likewise, when the clumped nucleus of a reorganizer, not normally

296 THE BIOLOGY OF STENTOR

dividing, was used to replace that of a divider it was divided with
the cell.

These findings indicate that the cytoplasm as a whole, or its
state of activation-inhibition with reference to primordium forma-
tion, seems to be involved in guiding the nucleus in its behavior.
Thus, in some of my own unpublished studies reorganizers or
dividers were cut in two longitudinally before the nucleus had

Fig. 84. Behavior of the macronucleus of ^S. coeruleus in absence
of the oral primordium.

A. a: Stage-2 reorganizer cut in two longitudinally, b:
Non-oral half retains nucleus in nodulated state while that of
oral half clumps as reorganization proceeds to stage 5. c: But
the non-oral half soon shows a tardy coalescence of nodes and
then renodulates along with the partner half (d). e: Fragment
without mouthparts forms regeneration primordium 10 1 hours

later with nodes again fusing.

B. Macronucleus renodulates in longitudinal half of a divider
without the primordium, regenerating later. Stage-6 anlage

completes development without nucleus.

C. Nodes coalesce and renodulate in half without primordium,

while mouthparts rejoin membranellar band.

BEHAVIOR AND FUNCTIONS OF THE NUCLEUS 297

coalesced, and the nodes in the right-hand fragment without the
primordium nevertheless fused simultaneously with those of the
other fragment which carried the developing anlagen. Similarly,
if stage-6 dividers were cut in the same way so that the clumped
nucleus remained in the fragment without the primordium, the
nucleus then extended and renodulated on time (Fig. 84). Such
observations suggest that the nucleus is guided in its behavior by
something which characterizes the whole cell rather than by
stimuli emanating exclusively from the primordium.

7. Necessity of the nucleus for oral redifferentiation

No experiment demonstrates more dramatically the fundamental
duality of the cell than the failure of Stentor and other ciliates to
regenerate or survive without the nucleus. The nucleus cannot
regenerate a cytoplasm around it and is indeed so dependent upon
its cytoplasmic environment that naked nuclei soon degenerate
and cannot viably be returned to the cell. Likewise cytoplasm
alone can never produce a nucleus. Specifically in regard to Stentor,
our starting point in the study of nucleo-cytoplasmic interactions
is that oral regeneration or the formation and development of a
primordium is a cooperative effort of nucleus and cytoplasm and
does not occur in the absence of some portion of the macronucleus.

Yet the nucleus does not make the primordium in the sense of a
handicraft but remains visibly unchanged while the anlage is
elaborated at some distance from it. Hence there should be some
intermediate step through which the nucleus contributes to the
support of primordium formation in the cytoplasm. This inter-
mediary would be truly essential to redifferentiation ; the presence
of the nucleus only indirectly as its source. Such a relationship is
evident in the amazing case of the unicellular plant Acetahularia
(see Hammerling, 1953) in which both growth and the elaboration
of specific organelles continues long after enucleation, in a way
that can best be explained by supposing that the nuclear contribu-
tion is a durable substance which persists, quantitatively, in the
cytoplasm until exhausted.

If the action of the nucleus is indirect and mediated through
products which it contributes to the cytoplasm, there should also
be some evidence of this lag-effect in Stentor. Using coeruleus as
the test organism, I have found (unpublished) that the oral

298 THE BIOLOGY OF STENTOR

primordium can generally continue one or two stages further in its
development after enucleation. Stage-2 primordia could continue
to stage 4, and stage-4 anlagen to stage 6, etc. After going one or
two steps further, early primordia through stage 4 were then
resorbed. In incipient regenerators, enucleated, the primordium
could put in its initial appearance and then disappear, i.e., develop-
ment was possible from stage o to stage i. Animals with stage-6
anlagen, or even at stage 5 when there is still no sign of a gullet and
oral pouch, were able to complete oral differentiation and move
the new structures into their definitive position (see Fig. 8ob).
These performances are explainable on the assumptions, first, that
there is a nuclear contribution to cytoplasmic differentiation which
does persist for a short while or is present in small quantity at any
one time; and second, that by stage 5 the anlage has completed
most of its synthesis of new material in the form of oral cilia, etc.,
and needs only to invaginate and shift the parts already formed to
complete the elaboration of the feeding organelles.

Formation of the fission furrow after enucleation of mid-stage
dividers demonstrates its independence from the presence of the
nucleus.

The experiments on stentors with late-stage anlagen clearly
confirm many earlier observations on the completion of regenera-
tion and division of stentors in the absence of the nucleus, begin-
ning with Gruber (1883, 1885a, b). He was much impressed by
the continued normal behavior of enucleated ActinophrySj a
heliozoan, and of Stentor. For the former, he claimed " regenera-
tion" (however this may be manifest in a rhizopod) in the absence
of the nucleus, but requiring cytoplasmic chromatin of nuclear
origin. In S. coeruleus he found that cells, after removal of the
compacted macronucleus, could complete division with separation
of daughters and full development of the primordium in the
opisthe. He therefore supposed that enucleated stentors might
regenerate "under conditions not yet devised". This remark is
not very different from Morgan's (1901a) conjecture that, if the
nuclear contribution could be supplied in some other way, then
the presence of the nucleus as such should not be necessary for
regeneration in stentors. I therefore feel that Gruber has been
somewhat maligned in reviews of this subject as saying that the
nucleus is not necessary for regeneration and having to correct

BEHAVIOR AND FUNCTIONS OF THE NUCLEUS 299

this assertion later. Both Gruber and Morgan were pointing to the
high probabihty that the action of the nucleus is indirect and that
what is crucial is not the physical presence of the nucleus but of
something it produces.

Prowazek (1904) made more of a case for regeneration in
Stentor without the nucleus, though he too regarded nuclear
derivatives as indispensable. That enucleated dividing stentors can
complete fission and oral differentiation of the posterior daughters,
he also observed. In addition, he reported that elevated tempera-
tures in *' warm cultures " could supply the conditions for regenera-
tion without the nucleus. Yet he cited only one case and he did
not claim that the oral differentiation was complete. Ishikawa
(19 1 2) thought he confirmed this result in Stentor " in some cases ".
Sokoloff (1924) made similar experiments on Bursar ia truncatella
in w^arm culture and reported that eight out of thirty enucleated
animals regenerated the feeding organelles and were able to ingest
normally; but Schmahl (1926) denied this, though he did not say
specifically that he tried high temperatures. Returning to
Prowazek's studies, his strongest statement was that if stentors
are cut and recut so that they are repeatedly compelled to
regenerate, then in a few cases (3) oral regeneration could occur
in enucleate pieces, and he did not say that the regeneration was
incomplete. Regeneration in the absence of the nucleus, whether
in warm cultures or by repeated cutting, he explained as due to
the presence of chromidia, substituting for the nucleus. Before w^e
smile at this, we should remember that just as the Feulgen staining
anticipated the modern DNA doctrine, so the old chromidial
hypothesis is a sort of pre-vision of the RNA story which is
developing today; and Prowazek's exploratory study may contain
the germ of new techniques.

8. Reconstitution of shape in relation to the nucleus

Though stentors cannot redifferentiate oral structures without
the nucleus or its products, may they not at least recover their
normal form after injuries? Cutting operations usually have two
effects: the symmetrical, conical shape of a stentor is distorted,
and the lateral striping if not the membranellar band is disturbed
and misaUgned. Cutting also produces a wound, with exposure of
the endoplasm, and we can assert categorically that all investigators

300 THE BIOLOGY OF STENTOR

have found that heaHng is prompt and to all appearances as good
in enucleates as in nucleated stentors.

In regard to shape recovery, Balbiani (1891c) had observed that
enucleate aboral longitudinal halves which folded on themselves
retained the abnormal shape; nucleates recovered. From less
drastic distortions, Prowazek (1904, 1913) and Schwartz (1935)
reported that enucleates are capable of rather extensive reconstitu-
tion of the normal axis and conical form. This was my experience
also (Tartar, 1956c) and can be explained as merely the shifting
in position of parts already present without requiring new
syntheses. Schwartz observed little or no reahgnment of the striping
of the ectoplasm, however, and made the point that, since these
adjustments are gradual, the enucleated cell dies before they can
be completed. I have found that separate or separated membra-

F1G.85. Activities during survival of enucleated (minus
macronucleus) S. coeruleus.

A. Mass of 2 enucleate stentors minced and mouthparts
removed. Sections of membranellar bands come together and
join and there is considerable mending and alignment of stripe

areas. Specimen lived 5 days.

B. Stage-4 reorganizer enucleated. By day 2 of the experi-
ment the anlage had lengthened but there was no stomatogenesis
and original mouthparts were resorbed. By day 3 specimen was
about half original volume, indicating utilization of its substance
during starvation. Considerable fading of coloration occurs. On
day 4 lateral striping is present but only a few membranellar cilia
remain. Day 6 : glistening sphere without body or oral cilia but

with vestiges of striping, found dead on day 7.

BEHAVIOR AND FUNCTIONS OF THE NUCLEUS 3OI

nellar bands can unite, oral parts may migrate together, and that
there is a fair amount of reorienting and mending of stripe patches
in minced stentors in the absence of the nucleus (Fig. 85A). Yet
we still need clear-cut tests and a precise definition of capabilities
for shape and form reconstitution in enucleated stentors.

9. Functioning and re-formation of vacuole and holdfast

in enucleates

Balbiani (1889) found that the contractile vacuole of Stentor
functions without the presence of the nucleus. Its rate of pulsation
is even normal (Prowazek, 1904), as is the case in Amoeba
(Comandon and de Fonbrune, 1939b). A new contractile vacuole
appeared in enucleated posterior pieces of igneus (Balbiani, 1893)
and this was confirmed in coeruleus by Stevens (1903), Schwartz
(1935) and Tartar (1956c). The same has been known for Amoeba
since the work of Hofer (1890). As Balbiani remarked, the new
vacuole probably does not involve structural synthesis and may
arise merely by the enlargement of some feeding canal of the
existing contractile vacuole system. This is the more probable
since Schwartz observed new pulsating vacuoles in enucleated
stentors 3 minutes after the older ones had been removed.

In enucleated stentors the old holdfast is quite capable of
functioning in reattachment (Stevens, 1903). It can also be
re-formed in the absence of the nucleus, as attested by firm
reattachment by a holdfast of coeruleus from which both tail-pole
and nucleus have been removed (Tartar, 1956c). This regeneration
is understandable on the basis that Httle if any synthesis is involved,
only a modification of existing parts.

10. Behavior of enucleates

First let us note that Schwartz (1935) observed vigorous cyclosis
of the endoplasm which continued nearly up to the point of death
in enucleated coeruleus. In regard to the *' external" behavior,
investigators were impressed from the start by the sustained
activity and normal swimming behavior of enucleated ciliates.
Similarly, it is well known that amcebas can continue forming
pseudopods in the absence of the nucleus. But in Stentor the
vigorous beating of thousands of body cilia and numerous huge

302 THE BIOLOGY OF STENTOR

membranelles can continue for about one week and is dramatic
evidence of the extent of energy metabolism which continues in
the absence of the nucleus. Very likely this is to be explained by
the presence of mitochondria or specialized ectoplasmic granules
as relatively independent centers of oxidative phosphorylation.
Normal avoiding responses and searching behavior seem to be
shown by enucleates; therefore, to repeat an apercu of doubtful
brilliance, the nucleus is not a brain. As the proof of the pudding
is in the eating, so a test of effective behavior lies in the feeding.
Prowazek (1904) found that enucleated coeruleus could ingest
chlorellae and Schwartz (1935) showed that some would take up
Colpidia. Hence the general impression, which corresponds to my
own observations, is oddly indecisive: enucleated stentors with
intact feeding organelles can ingest food but, like enucleated
amoebas (Brachet, 1955), usually do not. As a rule enucleates feed
little and soon become transparent as they void the food vacuoles
which were present in them originally; for, though incapable of
further digestion, they are quite capable of normal defecation
(Balbiani, 1889; Prowazek, 1904; and Schwartz, 1935). I have
found that the presence of but one macronuclear node was
sufficient to cause stentors to gorge themselves in the presence of
abundant food.

II. Digestion in enucleates

Using vital dyes as indicators, Balbiani (1893) found that food
vacuoles do not become acidic in enucleates as they do in normal
stentors, and in this he confirmed the work of Hofer in 1890 on
amoebas. Schwartz (1935) followed the fate of Colpidia which were
ingested by some of his enucleate coeruleus. Digestion was never
complete. From the start the food vacuoles were abnormally
swollen. Staining showed no dissolution of the ingested ciliates,
as occurred in controls. In one case I noticed that a motionless
rotifer in a food vacuole within an enucleated stentor remained
without apparent change for 4 days, although rotifers are normal
food of stentors. Hence it is very probable that the macronucleus
is necessary for digestion and hence for growth of the cell.

Though apparently incapable of digestion, enucleated coeruleus
were able to utilize or cause the disappearance of their granular
carbohydrate reserves, though possibly at a slower rate than in

BEHAVIOR AND FUNCTIONS OF THE NUCLEUS 303

normal starving animals (Tartar, 1956c). This corresponds to
Brachet's (1955) results with Amoeba *

Allied to the problem of digestion is the matter of autolysis or
'' self -digestion " or the capacity of enucleated ciliates to dedifferen-
tiate and resorb existing ectoplasmic structures. As is too often
disturbingly the case, some say they do and others say they don't.
In Frofttonia, Balbiani (1889) observed the disappearance of tri-
chocysts and much of the ciliation in enucleates and he thought
this might be an autodigestion of proteins. Schmahl (1926) said that
membranelles are resorbed in enucleated Bursaria; but he
remarked that Dembowska found no resorption of parts in
Stylonychia after enucleation. Specifically in regard to Stentor
coeruleus, we have already noted that early primordia are resorbed
when the nucleus is removed, but the situation is quite different
in respect to already formed structures.

Weisz (1949c) claimed that enucleation produces prompt
dedifferentiation of existing feeding organelles and holdfast within
24 hours. He therefore believed that the nucleus is necessary not
only for the production of new parts but equally for the main-
tenance of structures already formed. My experience has been to
the contrary (Tartar, 1953). I obtained no impression that the
feeding organelles soon disappear upon withdrawal of nuclear
''support". Most frequently animals died with these organelles
intact or, at most, a bit vague. Only when survival of enucleates
was most protracted did extensive dedifferentiation finally occur
(Fig. 85B), but since dedifferentiation was then so tardy it was
probably the result rather of general necrosis. In the present
context it is significant that in the experiments in which the heads,
only, of stentors were excised, proportionality of parts, which
undoubtedly involved resorption of considerable part of the

*Regarding transport mechanisms in the cell, de Terra (i960) found
in S. coeruleus that during the later stages of division, after the macro-
nucleus has condensed, uptake of phosphate as tested by radio-active
phosphorus 32 is greatly reduced. This reduction occurred both in
nucleate and enucleated animals. Return to high uptake after division was
found only in cells with the macronucleus. Hence by inference this nucleus
is chemically inactive during late fission (which therefore does not require
the nucleus for completion), or at least the macronucleus is required for
restoration of high phosphate uptake characteristic of the interdivisional
period.

304 THE BIOLOGY OF STENTOR

original membranellar band, occurred only if nuclear beads were
present (Tartar, i959d).

12. Survival of enucleates

Gruber, Balbiani, Prowazek, and Schwartz reported survival
times of from 32 hours to 3 days. Demonstrating his main theme,
Schwartz (1935) found that survival of coeruleus in which the
macronucleus has been removed was not aided or extended by the
presence of i to 16 micronuclei, again proving the indifferent
character of these tiny nuclei with respect to vegetative functions.
His enucleates lived for a much shorter period than starved controls
which remained alive for a week. Therefore he concluded that
death was not due to starvation but to some disturbance of the
entire metabolism in the absence of the macronucleus.

I have found that the tiniest blebs of ciliated cytoplasm separated
off in abnormal division of coeruleus can live for a little more than
5 hours. Larger enucleated fragments lived for about 3 to 4 days,
and the largest enucleates generally lived for 4 days when isolated
in depression slides, though some survived for 6. It is not uncom-
mon to find enucleated stentors living as long as starved controls
(Tartar, 1956c). One wonders, then, whether enucleates may not
die merely from exhaustion of reserves rather than disturbed
metabolism.

13. Consequences of excess nucleus

Stentor fragments and fusion complexes with an unusually high
proportion of macronuclear material in relation to the cytoplasmic
volume can be produced by cutting all the nucleus into one small
fragment, forcing all the nucleus at division into one daughter cell,
or grafting together sectors of several cells bearing most of their
nuclei. Effects of this artificial alteration of the nucleo-cytoplasmic
ratio in favor of the nucleus can be followed because, though neat
studies of nuclear volume would be difficult and are lacking, there
is no evidence or impression that excess nuclear material is quickly
and adaptively resorbed in any way comparable to the speed with
which the macronuclear chain is regenerated after all but a few
nodes are removed. This statement corresponds closely to the
observations of Schwartz (1935).

Both Prowazek (1904) and Causin (1931) noted that coeruleus

BEHAVIOR AND FUNCTIONS OF THE NUCLEUS 305

fragments with too much nucleus had difficuhy in regenerating
and surviving and soon died, while Sokoloff (1924) found that
Bursaria fragments of the same sort did not regenerate. Weisz
(1948a), on the contrary, denied that excess nucleus was injurious.
The staunchest advocate of the importance of the nucleo-
cytoplasmic ratio in protozoa was Popoif (1909). But we have to
doubt, as an expression of over-exuberance for this idea, the report
that he produced dwarf lines of coeruleus from small fragments of
normal ratio, and note that his hopes for producing giant races of
stentors by the same principle were not fulfilled. In fact, Burnside
(1929) clearly demonstrated that Stentor fragments grow back to
the normal size before they divide, and this was fully confirmed
by Weisz (1948c). Yet PopoflF made a summary statement which
is probably valid : namely, that too much nucleus is not as injurious
as too little but is not without its eflFect. This remark is in part
substantiated in the following section. In the present connection,
my experience has been that hypernucleate fragments of coeruleus
often die prematurely or are notably tardy in regeneration (Tartar,
i959g) but this matter needs much more study (Fig. 86a).
Comandon and de Fonbrune (1939b) found that uninucleate
species of Amoeba carrying 3 nuclei by transplantation did not
divide though followed 2 months, but binucleates could divide.

14. Consequences of reduced nucleus

All but one macronuclear node can be removed from stentors
and the nucleo-cytoplasmic ratio can be still further shifted in
favor of the cytoplasm by grafting such an animal to one or more
completely enucleated stentors. A combination of excisions and
graftings therefore makes it possible in Stentor to produce truly
extraordinary shifts in the relative volumes of cytoplasm and
nucleus; and there is time to test the consequences because
compensatory growth of the nucleus occurs only later during
periods of oral primordium formation (Fig. 86c).

These methods were not available to earlier workers whose cut
fragments yielded disturbance of the normal nucleo-plasmic ratio
in narrow^er range and gave no eflFect. Balbiani, for example, said
repeatedly (1889, 1891c, 1893) in reference to Stentor and other
ciliates that the relative size of the nucleus is indiflFerent for the
formative processes of regeneration. This was also the conclusion

306 THE BIOLOGY OF STENTOR

of Weisz (1948a, 1954). I have found, however (Tartar, 1953,
i959g), that in whole cells with only one node, appearance of the
regeneration primordium is very much delayed, substantiating an
early remark of Popoff (1909) regarding slowness of regeneration
in animals with a decreased nucleo-cytoplasmic ratio. In grafts of
two animals carrying only one node the delay was greater ; primor-
dium formation occurred only after two days, with development

Fig. 86. Effects of marked shifts in the nucleo-cytoplasmic ratio.
A. Indication of disturbance in hypernucleates. a: Fragment
containing all the macronucleus (20 nodes) of a coeruleiis did not
regenerate until one day later and then astomatously {b). c: By
day 3, successful re-regeneration had occurred and specimen
still had 17 to 20 nodes, d: Next day there were 10 oval or
doublish nodes indicating fusion and reduction of surface, e:
Specimen became sickly and died on day 5, a day earlier than
demise of the enucleate cell remainder.

BEHAVIOR AND FUNCTIONS OF THE NUCLEUS 307

of the anlage proceeding slowly. In any case the single node was
able to support regeneration eventually, for increases in the nucleus
occurred only afterward (Fig. 86b). Occasionally no regeneration
occurred at all, although the single macronuclear node persisted
intact until death of the specimen. Therefore extreme reductions
show that the nucleo-cytoplasmic ratio does have an important
effect at least on the time required for oral regeneration.

An interesting question is posed by what happens to the single
node which enables it to support primordium formation even in
a large mass of cytoplasm. For if all but one nuclear bead is removed
from a stentor in process of regeneration, I found that early
primordia were then resorbed, though later anlagen could run
their course of development with the support of the single remain-
ing node (unpublished). In a stentor with 15 nodes each may carry
only I /15th of the burden of supporting primordium formation
and development. For one node to do so, especially within a large
mass of cytoplasm, may require an adaptation in which the output
of this single node becomes greatly accelerated.

There is some evidence that in stentors with greatly reduced
nuclear complement the general metabolism of the cell may be
upset until this disparity is redressed. The first hint of such effects

B. Delayed primordium formation in hypernucleates. a:
When a sector with stage-2 regeneration anlage and one macro-
nuclear node was grafted to non- differentiating stentor minus
nucleus and mouthparts the primordium was promptly resorbed
and no regeneration occurred until 3 days later. No nuclear
increase occurs until after anlage passes through stage 6. b: A
fusion mass of 2 stentors coerideus with mouthparts and all but 2
nodes excised did not regenerate until two days later, the 2 nodes
increasing to 7 only as the anlagen completed development.

C. Regeneration of the macronuclear chain and reconstitution
of proportionate parts, a: One macronuclear node in head
folded on itself, b: Shape regenerated and feeding organelles
reduced to proportionate size without primordium formation,
single node now spindle form (to increase its active surface
area?); reorganization primordium (to make possible nuclear
increase?), c: After reorganization the specimen has 9 small
nodes, d: Membranellar band and frontal field, again made
relatively too large through reorganization, are adaptively
decreased in size, e: Before its demise, the animal had 7 nodes

(adaptation to decreasing size in starvation on the slide?).

3o8 THE BIOLOGY OF STENTOR

is to be found in the work of Popoff (1909) who concluded from
his observations of coeruleus that stentors with too little nucleus
are sickly and in many respects like enucleates. Schwartz (1935)
was even stronger in his statement that digestion and even ''the
entire metabolism" is greatly disturbed through removal of most
of the macronucleus.

Although the old concept of nucleo-cytoplasmic ratio may not
have fulfilled the original hopes that were invested in it, it is likely
that the extreme variations in this ratio which can be produced in
Stentor will have an important bearing in newer studies of respira-
tion, synthesis, and enzyme production in single cells.

Fig. 87. Critical time for recovery after renucleation.

Macronucleus was excised from specimens on morning of
day I. If renucleated on day 4 complete regeneration of feeding
organelles and of nuclear chain occurred, with continued
survival. Renucleated on day 5, pale and murky stentor becomes
healthier, regenerates faded pigmentation, and forms primor-
dium which does not produce mouthparts. Specimen was dead
on day 10. Renucleation on day 6, the faded, murky specimen
became healthier in appearance and darker in color but died the
next day without forming an oral anlage. Fifth day therefore
seems to be transition period for recoverability of the cytoplasm.

BEHAVIOR AND FUNCTIONS OF THE NUCLEUS 309

15. Delayed renucleation

Since Verworn (1892), it has been found in Amoeba and its
relatives that the naked nucleus undergoes immediate degeneration
in the absence of its normal cytoplasmic environment and is not
viable when reimplanted (Comandon and de Fonbrune, 1939b;
Lorch and Danielli, 1953). So it is with Stentor and with
embryonic cells (Briggs and King, 1955). Cytoplasm in the
absence of the nucleus shows a very much slower deterioration.
This raises the question: Beyond what period of time is the
cytoplasm irreversibly deteriorated so that it can no longer recover
after reimplantation of a fresh nucleus? We would like to be able
to analyse what goes wrong in enucleated cells and what the
nucleus contributes to the maintenance of the cytoplasm. This
approach might be especially fruitful if renucleations were made
at a time when certain cytoplasmic functions were recoverable
and others not.

Amoebas can recover normal activity and even divide if
renucleated 2 days after enucleation (Comandon and de Fonbrune,
1939b; Lorch and DanielH, 1953); but the French workers found
that practically no recovery occurred when renucleation was
delayed to the sixth day. I have a few experiments of this type on
Stentor coeruleus (unpublished). Animals renucleated on days 3 and

4 of the experiment, or 2 and 3 days following enucleation, could
recover completely, regenerate the mouthparts, show increase in
the number of nuclear nodes following regeneration, and divide.
Very likely they would have developed into clones if the difficulties
of culturing had been surmounted. A specimen renucleated on
day 6 promptly corrected its faded coloration and necrotic turbidity
of the endoplasm, but was not able to regenerate. Another,
renucleated on day 5 recovered and regenerated incompletely, for
although the new membranellar band was normal the mouthparts
consisted only of a short tube (Fig. 87). This animal survived

5 more days on the slide. Apropos of the remarks above, the
specimen lost its contractility completely and could be cut into
without showing the slightest twich, as if the myonemes had
suffered irreversible damage or loss of response. Evidently the
fifth day without the nucleus is critical for Stentor cytoplasm and
we should have more experiments covering this period.

310 THE BIOLOGY OF STENTOR

All the evidence from Stentor points to the generally accepted
conclusion that the nucleus acts upon the cytoplasm through the
intermediation of chemical substances produced in the nucleus
and transmitted to the cytoplasm. Weisz (1949a) has even found
quite direct evidence for this transmission : fixed specimens which
were undergoing primordium development showed macronuclear
vesicles apparently breaking through the nuclear membrane to
void their contents into the endoplasm. With due allowance for
lag effects, the presence of a portion of the macronucleus is
necessary for oral regeneration, digestion and synthesis, and
survival of the cell. Energy metabolism continues unabated for a
while in enucleates, to judge by their vigorous activity, but in the
course of several days ciliary beating becomes progressively slower,
either through the impairment of this metabolism or by failure to
replace utilized substrates. The ease with which stentors carrying
much more or less of the normal proportion of nuclear material
can be prepared offers unusual opportunities for a quantitative
study of the action of the nucleus on the cytoplasm. Eventually
such studies may teach us why and how a fairly constant ratio
between nucleus and cytoplasm is maintained.

Conversely, the Stentor macronucleus is clearly affected in many
ways by the cytoplasm, in addition to the obvious fact that cyto-
plasmic environment supplies the basis for the growth and
integrity of the nucleus. The location of the macronucleus is
determined by the stripe pattern or geometry of the ectoplasm.
Cytoplasmic events evidently guide the macronucleus in its
complex behavior during coalescence and renodulation. Such
control of the nucleus by the cytoplasm is at least as extensive as
has been demonstrated in any other cell and in Stentor is capable
of being investigated by micrurgical operations.

CHAPTER XVII

TOWARD A GENETICS OF STENTOR

The collaborative functions of nucleus and cytoplasm can further
be studied by artificial exchange of nuclei and cytoplasms between
different individuals and even different species of Stentor. Poten-
tially, natural exchange of traceable genetic material between
members of the same species through conjugation would comple-
ment such investigations. The circumstance common to both
approaches is that now we deal with qualitative alterations in either
the cytoplasmic or nuclear component, or both.

I. Interspecific chimeras and nuclear transplantations

New possibilities for experimentation appeared when it was
found that different species of Stentor could be fused together by
their cytoplasms in enduring unions, and that when foreign nuclei
are transplanted they are neither immediately ejected nor destroyed
but also persist, with the expectation of revealing interactions
between species-different nuclei and cytoplasms. Many combina-
tions are possible. The macronucleus can be removed and replaced
by that of a different species. Alien cytoplasm can be grafted so
that the host nucleus now operates w ithin a mixture of cytoplasms,
or nuclei alone can be added so that now a mixed nuclear comple-
ment works with cytoplasm almost entirely of one type. And
finally, relative amounts of the two types of nuclei and cytoplasms
can be mixed together in any desired combination. The parts of
three species could also be joined in similar permutations, but we
need first to work out the simpler combinations. In comparison
with the possibilities, the work which has so far been done (Tartar,
1953, 1956c) may be regarded as only exploratory. In fact, further
development of these studies was postponed until conditions
governing the states of activation and inhibition during primordium
formation, as well as the effects of altering the nuclear-cytoplasmic
ratio, and other points could be explored ; for all these factors have

V

3"

312 THE BIOLOGY OF STENTOR

to be considered in arriving at any reliable interpretation of the
effects resulting from species differences in cytoplasm and nuclei.

What may be called the capacity of foreign macronuclear nodes
to support oral primordium formation and development in
enucleated coeruleus from which the oral region was also removed
has received a preliminary survey. We have to say preliminary,
because nuclear transfers are not easy and the number of cases
with certain combinations is still few. That there should be many
is indicated both by the circumstance that one cannot always be
sure that the last macronuclear node of the coeruleus has been
removed, as well as the experience of Lorch and Danielli (1950)
with interspecific transfers in Amoeha in which many tests gave
negative results though a small percentage could produce effective
combinations. Hence the positive result is more significant in
showing what an alien combination can do ; but the negative result,
in which no primordium formation occurs, may be merely the
result of other factors such as poor viability or insufficient number
of cases to include possible rare instances in which the combination
would work.

In the first place, controls showed that nuclei could viably be
transferred from one individual to another among similar stocks
of coeruleus leading to their subsequent complete regeneration. The
next closest combination was between typical coeruleus and an
organism I called Stentor ''X" (Tartar, 1956c). The latter was a
small, blue-green stentor like coeruleus but only about i/8th its
volume, with far fewer stripes, usually opaque cytoplasm, and tiny
nuclear nodes. This organism was at least a distinctly different
variety of coeruleus, or possibly even a different though closely
associated species. Transfers could be made in both directions and
it was found that the nucleus of either could support regeneration
in the cytoplasm of the other (Fig. 88). This may be called the
expected result, because we can assume that the nature of nuclear
support of regeneration is the same in all species and varieties of
Stentor. Size of the feeding organelles was always appropriate for
the host cytosome. Yet the chimeras did not survive as long as
controls and soon died. This is the reason it is thought that the
two were distant varieties, if not separate species. The appearance
was, therefore, that a successful interaction between nuclei and
cytoplasm of different type was later and gradually overwhelmed

TOWARD A GENETICS OF STENTOR

313

Interactions between S. coeruleus and Stentor X,

which was either a closely related species or a dwarf race.

a: S. coeruleus. b: Stentor X, of about \ volume of former,
fewer stripes, and smaller macronuclear nodes, a' : X-nuclei
implanted into enucleated coeruleus gives complete regeneration
{a") but mouthparts resorption and death soon follow {a"').

b-b'" : Similarly for the reciprocal cross. Feeding organelles
are in proportion to size of the cytosome in which they are formed.

ab : Graft of the 2 forms undergoes simultaneous reorganiza-
tion and integrates the two shapes into a doublet {ab') but the
specimen then died. (After Tartar, 1956c.)

by some subtle and possibly immunological difference which
resulted in eventual malfunctioning. The general picture suggested
is that nucleus and cytoplasm no doubt have respective roles in the
act of cytodifferentiation which are the same in all species, but that
the specific adaptation of the one to the other is developed to such
a degree that sooner or later a disharmony will almost always
emerge to nullify whatever compatibilities were at first realized.
Initially it appeared that polymorphus macronucleus could not
support regeneration in coeruleus cytoplasm (Tartar, 1953), but
further tests (1956c) showed that primordium formation at least
is possible (Fig. 89A). Typically, an anlage was produced which
progressed to stage 4 and then was resorbed. Hence there was some
effective interaction, but not as much as in the first combination
described. In regard to other combinations, it was found that
neither roeseli nor niger nucleus supported regeneration in coeruleus,
but there was one positive result with introversus nucleus which
after 5 days finally led to satisfactory oral differentiation in

314

THE BIOLOGY OF STENTOR

/%

a ■ b ^ c d

Fig. 89. Interactions in coeruleus X polymorphns chimeras.

A. a : Three polymorphiis macronuclear nodes implanted into
an enucleated coeruleus with mouthparts excised, b: Wide
primordium forming in coeruleus cytoplasm under influence of
polymorphus nucleus, 12 hours later, c: Anlage not developing,
exudate in primordium site, and original coeruleus membranelles
fimbriated, d: Specimen died after 3 days and showed 6 nodes,

indicating nuclear increase in alien cytoplasm.

B. Mutual integration of shape, a: Polymorphus grafted at
right angles to coeruleus. b : Harmonization of shapes and spread
of chlorellae throughout, c: Oral structures dediflferentiated,
coeruleus pigment lost, chlorellae clumped, d: Prompt death of

species combination. (After Tartar, 1953, 1956c.)

coeruleus cytoplasm. If pigmentation is a sign of taxonomic related-
ness, the blue-green color of both introversus and coeruleus suggests
that these species are more closely related than others, and this
could be the basis for effective interaction between their parts.
The species multiformis is also blue-green and in fact appears like
a miniature of coeruleus, with but a single macronuclear node.
These animals are so small that a whole cell had to be implanted
in the enucleated coeruleus. In 4 cases no regeneration occurred,
but we have to remember that regeneration on one node is greatly
retarded so that in this combination the chimeras may have run
into difficulties of another sort before they were able to express

TOWARD A GENETICS OF STENTOR 315

their regeneration possibilities. Nevertheless, this combination
should be attractive for further studies because of the great
difference in size together with general similarities in other respects.

Nucleated cells and cell parts were grafted to obtain mixtures
of widely varying proportions of coeruleus and of a polymorphus
strain which was grass-green with abundant symbiotic Chlorella.
The general result was that successful oral redifferentiation
occurred only when there was a preponderance of one species of
nucleus in a preponderance of its own type of cytoplasm. The
more nearly the two types of cells approached equality the less
successful w^as oral reorganization, and instead existing feeding
organelles were promptly resorbed. When both species were
represented in the cytoplasm but with the nucleus from poly-
morphus only, some oral differentiation occurred and the indications
are therefore that there is a conflict between the nuclear compo-
nents such that polymorphus nucleus is more effective when acting
on a mixture alone.

In spite of incompatibilities in regard to oral differentiation,
mixtures of coeruleus and polymorphus in any proportion showed
very good shape reconstitution as manifested by the realignment
of cells and cell parts to form a single, conical stentor shape
(Fig. 89B), and was better than that of enucleated grafts of either
species. This suggests that reorientation of the cortical pattern is
either a more generalized function in which species differences are
not prominent or makes less precise demands on nucleo-
cytoplasmic interaction. Were the cytoplasm less specific than
the nucleus, as appears, this would substantiate present-day
conceptions (Monne, 1948).

In these quantitatively varying combinations of coeruleus and
polymorphus it was also found that any considerable admixture of
coeruleus cytoplasm resulted in the ejection of chlorellae from the
fusion mass, and it should be added that uniform distribution of
the symbionts throughout showed there was complete mixture of
the endoplasm. Hence coeruleus cytoplasm appears to be anti-
thetical to the entertainment of the symbionts. Correspondingly,
admixture of polymorphus cytoplasm resulted in depigmentation
of coeruleus.

Similar results in regard to pigmentation have been found in
recent tests in which nucleated coeruleus was fused with a smaller

3l6 THE BIOLOGY OF STENTOR

portion of white polymorphus which had been grown in the dark
and contained few if any chlorellae (unpubUshed). Even when the
portion of polymorphus was relatively small there occurred an
aggregation of the pigment granules of coeruleus into splotches
which later disappeared so that the fusion complex became entirely
colorless and very much resembled in general appearance the
white polymorphus stock. These combinations could regenerate,
reorganize and even divide. They also showed several interesting
abnormalities (Fig. 90A). In some cases the form of the cell became
abnormally elongated and occasionally this also led to a complete

Fig. 90. Abnormalities from grafting a small polymorphus to
a large coeruleus,

A. Abnormal form with granular core, a: Half a polymorphus
engrafted, with cytoplasm and nucleus, b: Aggregations of
coeruleus pigment as graft causes loss of pigmentation ; mouth-
parts excised, c: Regeneration of colorless chimera resulting.
d: Snake-like form with mass of colorless granules and irregular

distribution of nuclear nodes.

B. Bleaching and failure to form membranelles. a: Feeding
organelles of grafted polymorphus resorbed, coeruleus fading, with
splotches of pigment granules, h: Chimera now colorless, with
half-length adoral cilia which start and stop together but show
no metachronal beating nor organization into membranelles. c:
Reorganized animal may have longer peristomal cilia but still no
membranelles. Eight large nuclear nodes are seen. Specimen

died on day 7 in isolation on slide.

TOWARD A GENETICS OF STENTOR 317

disruption of the normal form, resembling the "amorphous
phenotype" (see p. 276). The nuclear picture also became
abnormal, with nuclear nodes of greatly varying size, atypically
located within the cell.

One special case should be mentioned in detail for the interesting
possibilities suggested. Addition of the polymorphus component
caused the major coeruleus part to dedifferentiate its original
feeding organelles, and when regeneration occurred the oral cilia
were at first only half the normal size and never did they group into
membranelles. These cilia started and stopped together but there
was no metachronal rhythm as when membranelles are present
(Fig. 90B). An oral differentiation was obtained which was, however,
very different from the normal.

In these chimeras, the complex appeared at first as if stricken
by the shock of incompatibility, and regeneration was often at
first abortive or incomplete ; but the specimens then recovered and
generally showed good oral redifferentiation, yet with the abnor-
malities described often appearing later. One may expect, therefore,
that from small additions of one species of Stentor to another,
after the manner of transduction, and with greater skill in keeping
the specimens alive, very interesting results will emerge.

Referring again to the combination of coeruleus and niger, it was
found that the coeruleus cell was greatly affected by the addition
of a relatively small piece of niger cytoplasm (Fig. 91). In only one
case was the host able to regenerate and maintain good feeding
organelles, for in other tests regeneration was incomplete or not
even begun (Tartar, 1956c). These two species are apparently so
distantly related that even a small admixture of niger greatly affects
the behavior of coeruleus^ including the condition of the macro-
nucleus. Conversely, coeruleus caused depigmentation of the niger
graft.

The studies of Hammerling have shown that in the unicellular
and uninucleate plant, Acetabularia, the cytoplasm is relatively
passive and can elaborate cell structures with the support of any
species of nucleus of this genus whidh was tested, and the form is
that of the species contributing the nucleus or the nuclear products.
For in combinations containing two types of nuclei the form was
more like that of the species contributing the most nuclei. If the
nuclear contributions were balanced the structures produced were

3i8

THE BIOLOGY OF STENTOR

B

Fig. 91. Effects of niger graft on coeruleus.

A. Regeneration blocked, no bleaching of coeruleus — the
usual result, a: Enucleate niger grafted and mouthparts of
coeruleus excised, b: Yellow niger pigment disappears but not
the coeruleus coloration, c: In spite of regression of niger graft
no primordium formation occurs and macronucleus becomes of

5 large nodes.

B. Unusual response to same type of graft, a: Regeneration
occurs, with a secondary primordium forming in suture between
fine niger stripes and broad striping of host, b: Only the host
anlage has complete stomatogenesis. c: re-regeneration follows
with still less development of the primordium associated with
niger graft. Host nucleus abnormal, with few and unequal nodes,
two of which migrated to lie underneath the graft. (After

Tartar, 1956c.)

in character halfway between those of the two species. In some
combinations differentiation occurred but the cooperation of the
two nuclei was defective and abnormal structures resulted, as in
the case without membranelles in Stentor.

On the whole this behavior is quite different from the situation
in Stentor, although the comparison is limited by the fact that
all stentors produce the same form of oral differentiation. In Stentor
the cytoplasms seem to be much more specific and nuclei are

TOWARD A GENETICS OF STENTOR 319

effective in alien c\1:oplasm only when the taxonomic relationship
may be considered close. Another difference is that in acetabularias
both growth and differentiation can occur for some time in the
absence of any nucleus, whereas in stentors whatever the nucleus
contributes to the cvtoplasm for growth and morphogenesis is not
stored or is exceedingly short-lived.

Following methods devised by Comandon and de Fonbrune
(1939b), Danielli and his co-workers have made exchanges of
nuclei between Amoeba proteus and discoides (Lorch and DanielU,
1950, 1953; DanielU et al., 1955; reviewed to date in DanieUi,
1959). In either transfer the enucleated cell of one species recovered
its capacit}" for normal pseudopodial locomotion and its sur^ ival
was promoted after receiving a nucleus from the other. For the
most part these chimeras, as in my experience with stentors, did
not survive; but in one instance at least the implantation of a
proteus nucleus into discoides cytoplasm did produce a clone which
was kept in mass culture for over 8 years. Back transfers showed
that both nucleus and C}1:oplasm became altered in the ahen
combination, but the proteus nucleus never became functionally
or morphologically identical to that of discoides, nor did the
discoides cytoplasm become the same as that oi proteus.

The persisting influence of both the nucleus and the c\toplasm
was also evident in such characteristics of the chimeras as nuclear
size, form of the pseudopodia, growth rate, and response to
antiserum, which in general fell between those expressed by the
two species in pure form. These results led Danielli (1958) to
emphasize the irreducible importance of the cytoplasm, because
it was never completely made over into the type of the nuclear
species, and to suggest the reasonable hypothesis that the nucleus
of the cell determines the specific types of macromolecules which
are svnthesized, while the cytoplasm controls the way in which
they are organized into functional units. This conception is
certainly indicated by studies of cihates and especially of Stentor.
Faihire of regeneration and growth in the absence of the nucleus
indicates this organelle to be essential for svnthesis. But the c\-to-
plasm, and especially the ectoplasmic pattern, is obviously
intimately concerned with guiding nuclear behavior and deter-
mining the location, extent, and direction of asymmetry of the
developing feeding organelles.

320 THE BIOLOGY OF STENTOR

2. Racial differences

In the best known species, coeruleus^ definable differences between
various stocks and clones may turn out to be inherited racial
differences because they have appeared and persisted in strains
cultured by uniform methods. Since we still lack the means for
inducing conjugation at will in stentors, the most fruitful approach
would be by nuclear exchanges to test whether these traits are
determined by the nucleus or the cytoplasm. In one instance this
has been done, but for the most part we can at present merely
describe possible strain differences which might be analyzed in
this way.

State of the endoplasm. Separate clones of coeruleus may differ
in regard to the transparency of the endoplasm, a difference which
manifests itself very clearly in the ease with which macronuclear
nodes can be made out in enucleation experiments. The two
extremes of this condition are shown in my stock from Urbana,
Illinois, in which the cytoplasm is consistently transparent so that
the nucleus is clearly visible except in over-fed animals, and
another from Stella, Washington, in which the endoplasm is
notably opaque, except when the animals have been completely
starved, and this stock has also remained consistently so
characterized for 4 years.

Size. If the Stentor *'X" mentioned in the preceding section
was indeed a race of coeruleus^ then we had a dwarf variety in which
the cells never attained a volume larger than about i/8th that of
most stocks of this species. Other stocks of coeruleus seem to show
much less striking differences in the maximum size attained, but
careful measurements might reveal consistent size differences such
as have already been found in other genera of ciliates, notably
Paramecium.

^^ Astomatous". A stock of coeruleus obtained from Woods Hole
Biological Supply in 1950 was unique in producing some indivi-
duals temporarily without mouthparts (Tartar, 1957b). After
growing these organisms in the laboratory for 4 years it was noticed
that in certain subcultures about i % of the stentors were poorly
fed because they lacked ingestive organelles. In division, reorgani-
zation, and regeneration the oral primordium developed without
its posterior end invaginating and forming a gullet, and the oral
pouch was generally missing. The membranellar band itself

TOWARD A GENETICS OF STENTOR

321

appeared altogether normal, and it seemed to be of the usual length,
including the part which descends into the gullet, because this
band curved into a long spiral (Fig. 92). If this appearance is
correctly interpreted, then normal length of the row of membra-
nelles is not used as a sign that successful oral differentiation has
been completed; for regeneration or reorganization then followed.

Fig. 92. Astomatous primordium development in Woods Hole
race of S. coeruleus. Normal animal reorganizes, producing
peristomal anlage of normal length which fails to invaginate to
form mouthparts. Following regeneration now produces
complete mouthparts. Animal feeds and divides, forming
daughters with complete feeding organelles but each having one
unusually large macronuclear node.

with the production in the same individual, usually on the first try,
of good mouthparts. The animals could then feed, explaining why
the abnormality did not result in its own extinction but persisted
in the subcultures for about a year, after which only normal
animals have been found. Astomatous stentors appeared normal
in other respects except for a variability in size of the macro-
nuclear nodes, and analysis of this trait is the more difficult because
the same individual could manifest both the complete and incom-
plete development of the feeding organelles. Isolated normals
could pass through the mouthless phase while astomatous indivi-
duals later became normal. Anterior and posterior fragments did
not differ significantly in the frequency with which they regenerated
incompletely. Until this abnormality reappears and can be studied

322 THE BIOLOGY OF STENTOR

further, we can only note that since the mouthparts are induced by
the ectoplasm of the posterior pole, this induction or its gradient
basis is apparently highly labile in certain stocks.

Fluorescence. That fluorescence is characteristic of certain races
of coeruleus was discovered by MoUer (see p. 48), these animals
when killed appearing red in ultraviolet radiation.* Other races lack
the trait, although they look the same in visible light because the
major component of the pigment is the same in all coeruleus and is
not fluorescent. Whiteley and Moller (unpublished) neatly demon-
strated fluorescence to be a trait under control of the nucleus. When
animals of a fluorescent race were enucleated, the fluorescence soon
disappeared; or if the macronucleus of fluorescents was replaced
by the nucleus of a non-fluorescent race the fluorescence still
disappeared. Therefore the manifestation and maintenance of this
trait seems to depend upon the presence of a certain type of
nucleus and may be regarded as a genetic characterization of great
interest in itself and also potentially valuable in tagging cells of
diflferent origin.

Cannibalism} In the discussion of feeding reactions in stentors,
reference has already been made to the work of Gelei (1925) on
cannibalism in coeruleus. One of the main points of this study was
that the proclivity for eating one's fellows is a racial character.
This conclusion was based on the observation that some samples
from a culture, containing a natural collection of stentors which
was not a clone, showed cannibalism and others did not. Daughters
of cannibals cannibalized each other. Even when not densely con-
centrated, and regardless of whether they were well-fed or not,
cannibals seemed actively to pursue their fellows, while in other
samples the stentors simply turned away on encountering each
other. On dubious grounds Ivanic (1927) questioned that canni-
balism is a racial character in protozoa, including Stentor. More
to the point is my observation of cannibalism in all 9 stocks or
clones of coeruleus which I have under cultivation, strongly indicat-
ing that cannibalism is common to all representatives of this
species regardless of origin.

Other possible racial differences have been indicated in respect
of the following characteristics: Requirement, or not, for high

*Lately (i960) Moller reports that some races of S. coeruleus exhibit
all degrees of fluorescence.

TOWARD A GENETICS OF STENTOR 323

oxygen tension (see p. 265). Presence of 3, or of 2 rows of cilia in
the membranelles (p. 30). Acclimatization, or not, to alcohols
(p. 248). Suitability, or not, of certain culture media, like Benecke's
solution ior polymorphus (p. 268). Negative, or indifferent, response
to light, as shown by coeruleus (p. 22). Presence, or absence of
visible nucleoli; for de Terra (1959) remarked a race of coeruleus
without nucleoli, but Schwartz (1935) demonstrated their presence
in his strain. And average number of macronuclear nodes (MoUer,
unpublished).

3. Conjugation

Sexual reproduction by temporary fusion of partner animals
with cross-fertiHzation and complete renewal of the nuclear
apparatus occurs in Stentor as in other ciliates. Breeding experi-
ments towards an analysis of inheritance and the roles of nucleus
and cytoplasm in the development of racial differences, which
may well include mating types, should therefore eventually be
possible in this genus. Yet conjugation is rarely observed and
seems to be quite adventitious in its appearance, for the means
have yet to be discovered by which stentors can be induced to
conjugate as we desire. There is available, nevertheless, a com-
prehensive cytological study of sexual reproduction in Stentor
which has been generally neglected in reviews of this subject.

The general occurrence of conjugation in the genus Stentor is
attested by the observation of pairs in niger (Stein, 1867), coeruleus
(Moxon, 1869; Balbiani, 1891c), igneus (Johnson, 1893), poly-
morphus (Mulsow, 1913) and in roeseli (Balbiani). I also observed
fusion pairs in a colorless race oi polymorphus in which symbiotic
chlorellae were almost completely lacking. In seven stocks of
coeruleus, conjugation in five was observed at least once during a
period of 10 years. Schwartz (1935) remarked that he found no
evidence of autogamy in his extensive studies of coeruleus; and I,
too, have never seen any indication of nuclear renewal in unpaired
animals.

Exploratory studies of conjugation were included within the
compass of those works by Balbiani and Johnson to which we have
so often referred. Finding that the old macronucleus could be
distinguished by its no longer being clear and refractive in living
animals, Balbiani (1891c, 1893) studied the behavior of fragments

324 THE BIOLOGY OF STENTOR

in relation thereto. Pieces of exconjugants did not resorb the old
portions of the macronucleus if new nuclear anlagen were not
present in them. New macronuclei could support regeneration
from the start, but after the first stages of disintegration the old
nucleus could not. These conclusions are in line with what is
found in other ciliates, and I also have noted that coeruleus which
had begun conjugation were unable to regenerate excised oral parts.

That the proper conditions for conjugation seldom occur was
noted in the first extensive study by Hamburger (1908), who
remarked that Stentor cultures may be carried on for years without
noticing pair-formation. When conjugation does occur only about
I to 10% of the animals are involved so that no mass ** epidemics "
occur, as in Paramecium. She also found that many isolated single
animals from cultures in which conjugation was occurring did not
themselves subsequently pair. Her study was therefore of only 55
conjugants, the products of which did not live sufficiently long to
reveal the complete nuclear transformations. Later, Mulsow
obtained material from mixed pond samples which were rich in
fauna and flora. They were left unfed in the laboratory and after
a week apparently abundant conjugation of coeruleus and poly-
morphus present occurred. The experience was repeatable on fresh
samples. Bad conditions may have developed in the jars because
all the stentors died following the period of conjugation. Possibly
a putrid condition developed although he did not say so. At the
same time Paramecium and Frontonia present in the samples also
conjugated, which would indicate that methods employed for
paramecia might be applicable to stentors. However, I have made
mixtures of 5 stocks of coeruleus in the hope of providing mating
type diversity, and subjected them to feeding and starvation
routines to which paramecia are usually responsive, but this first
attempt to induce conjugation was entirely unsuccessful.

Mulsow's (191 3) study included over 2,000 conjugating pairs of
coeruleus and a smaller but substantial number of polymorphus
conjugants, incidentally confirming many of the points made by
his predecessors. He was able to keep his exconjugant animals
considerably beyond the period required for completion of all
transformations. Sectioned material was studied, for he found that
squashed preparations and total amounts were not satisfactory. I
have tried to express Mulsow's account of the course of conjugation

TOWARD A GENETICS OF STENTOR 325

in these two species in the diagrams and legends of Fig. 93 and 94
and therefore will touch only the main points in the following
paragraphs.

S. coeruleus: The size of conjugating animals is always smaller
than the maximum, as Hamburger had also noted, and Johnson
observed that conjugants were without food vacuoles apparently
from the start. Sometimes the partners are of different sizes but
they are not necessarily so and hence there was no indication of
** gamete" differentiation. All this accords with my own observa-
tion. Attachment is by the anterior rim so that the partners rest at
an angle to one another and swim together with their axes parallel.
In the conjugants I have observed there was always a special place
of attachment : a patch immediately below the membranellar band
and to the left of the mouth. In location this point corresponds to
Hamburger's figure though she said that attachment was by the
membranellar bands. That this locus of joining is not invariable,
however, is shown by the fact that Mulsow often found three
animals together in conjugation, all undergoing nuclear changes
simultaneously. He also found that the degree of union is variable,
from a small bridge to quite complete fusion of the two lateral
surfaces, and that this does not depend on the stage of conjugation.
Hence there may be endoplasmic fusion, but the migratory nucleus
always penetrates through a separating, pigmented membrane
toward the anterior end. The duration of the union is about
30 hours, which is not unusual, and nuclear renewal is not
completed until 10 days after separation.

The old rnacronucleus first breaks up into separate nodes which
then lose their orderly arrangement in a chain as well as their
adherence to the inside of the ectoplasm. Johnson had observed
that the nodes carry cytoplasmic (attachment?) threads as they
break loose from their former locations and are carried about in
the cell by a cyclosis of the endoplasm, which is unusually rapid
in conjugation. At this time. Hamburger said that the nodes lose
their amorphous character and show a honeycomb structure. The
original macronucleus so remains until its parts begin disappearing
as soon as the new macronuclear anlagen have attained considerable
size.

After breakup of the macronucleus into separate nodes, the

326

THE BIOLOGY OF STENTOR

Conjugation in S. coeruleus, largely following the
account of Mulsow, 191 3.

1. Stentors attach by areas just below membranellar bands,
maintaining a separating partition or fusing to some degree.

2. Old macronuclear nodes break apart, become spherical, and
may show honeycomb structure. (These organelles are omitted
from subsequent drawings up to stage 11.) The 50-70 micro-
nuclei are near the macronuclear nodes.

3. Micronuclei separate freely in the cytoplasm, all enlarging

and stain less intensely.

4. Up to 10 micronuclei in both cells degenerate. Others
divide mitotically but not simultaneously, giving about 60 nuclei
in each. These are probably not maturation but "multiplicative"
divisions. Stages in mitosis are shown below:

a: homogeneous, swollen micronucleus.

b: parachute stage, with chromatin at one end, single

spindle pole at the other.
c: chromatin pulling toward equator (by traction fibers?).

Degeneration may occur after this phase.
d: equatorial ring of chromatin.
e: development of second spindle pole, spindle fibers

pointing toward centrioles.
/.• polar cones flatten as nucleus becomes cylindrical; no

fibers found between chromosomes, approximately 80

in number.

TOWARD A GENETICS OF STENTOR

g: anaphase separation.

h: chromosomes reach poles as granules, and spindle body
disappears.

5. Nuclei near the partition may be the sexual nuclei, con-
siderably reduced in size from reduction of chromosomes to 20
by maturation division. Further division figures and degenerating

mitoses elsewhere.

6. Exchange of migratory pronuclei.

7. Fertilization and first 2 divisions of the amphinucleus.

a: male pronucleus becomes surrounded by halo of

cytoplasm.
b: halo carried with it as it penetrates partition into the

partner cell.
c : nucleus with chromatin massed forward breaks free from

halo.
d: union of male pronucleus with female, which is larger

and more loosely formed.
e: fertilization spindle figure, always surrounded by a

thickened cytoplasmic halo.-
/; termination of this first division.
g: beginning of second division, with parachute-form

nuclei.
h: division produces 4 similar products.

8. Partners, each with four new nuclei, now separate.

w

328 THE BIOLOGY OF STENTOR

many micronuclei depart from their location on or near the macro-
nuclear surface and become freed in the endoplasm. All become
greatly enlarged and hence less strongly staining. Then occur a
series of non-simultaneous mitotic divisions, but increase in the
number of micronuclei is overbalanced by the fact that many
degenerate even after they have begun the first stages of mitosis.
As these divisions proceed, the remaining micronuclei become
smaller and fewer in number until there is only one sexual nucleus
in each partner. Presumably the penultimate division is the first
maturation division and division of the last remaining micronucleus
is the second maturation division with reduction in number of
chromosomes in the migrating and stationary pronuclei formed.

The migrating pronucleus then becomes surrounded by a halo of
homogeneous, darkly-staining cytoplasm. Nucleus and halo both
break through the separation membrane to move across to the
partner cell so that there is always a small exchange of cytoplasm.
At early stages it appears that this cytoplasm is pulling the nucleus
along, later the reverse. Once across, the migratory pronucleus
breaks loose from its halo and unites with the partner stationary
pronucleus of different appearance. The fertiHzation nucleus then
becomes itself surrounded by a thickened halo and undergoes two
post-fertilization divisions giving rise to four simpler nuclei.

Separation of the animals occurs about this time. Two of the

9. Two nuclei form macronuclear anlagen and two produce
mitotic spindles but without showing chromosomes and will

form 4 micronuclei.

10. Day after separation i to 10 macronuclear anlagen,
depending on combinations of fusion and amitotic increase.

Four micronuclei.

11. Two-four days after separation. Anlagen with chromatin
net resembling chromosomes, and nucleoli appear. Old macro-
nuclear nodes resorb as soon as anlagen attain same size. Micro-
nuclear increase by mitosis with clearly defined chromosomes.

12. Constriction of anlagen into nodes and their attaching
together. They are not sorted out between daughter cells.
Micronuclei have final size and location but are not yet of

definitive number.

13. First fission 10 days after separation of conjugants, with
the usual vegetative division of nuclei producing 2 normal

vegetative stentors (14).

TOWARD A GENETICS OF STENTOR 329

nuclei remain heavily chromatic and will form macronuclear
anlagen, while the other two undergo mitosis, but without showing
clear chromosomes, and produce four equal micronuclei. The
number of macronuclear anlagen then changes by combined
fusions and amitoses so that there may be i to 10 masses.

Two to four days after separation the appearance is as follows :
macronuclear anlagen enlarge and show at first a chromatin net-
work resembling chromosomes, and nucleoli appear. The nodes
of the old macronucleus then begin absorbing. By mitotic division,
with appearance of definable chromosomes, the micronuclei
increase in number. Eight days after separation the macronuclear
anlagen separately constrict into nodes or chains of beads which
then attach together to form the definitive macronuclear chain
along which the micronuclei, now of their final size and number,
find their location. There is hence no sorting out of anlagen between
daughter cells, and when the first division occurs 10 days after
separation this is the fission of an animal which has in itself
regained the completely normal nuclear picture. Two variations
in macronuclear development were described but these may have
been pathological.

S. polymorphus: Conjugation in this species is of course quite
similar. Multiplicative divisions of the micronuclei and concomi-
tant degenerations occur as in coeruleus. Cross-fertilization was
established, but the amphinucleus divides three times to produce
8 products before nuclear differentiation begins. Normally 6 of
these form macronuclear anlagen by increasing in size and pro-
ducing from the karyosome a spireme, later breaking into segments
or chromosomes which seem to be in the diploid number and split
in two longitudinally, like chromosomes, before they are reduced
to chromatin granules. The anlagen then nodulate and join to form
the definitive macronucleus. The two remaining products of the
amphinucleus form the micronuclei by repeated mitotic divisions,
during which the nuclei decrease in size. Because the number of
macronuclear and micronuclear anlagen may vary, it seems likely
that the 8 products of the third division of the amphinucleus are
still equivalent, and that differentiation is not predeterminedand
might even be guided by their location in the cell, as is the case in
other ciliates.

330

THE BIOLOGY OF STENTOR

e

f

e

©■

Fig. 94. Conjugation in S. polymorphus, following the account

of Mulsow, 1913. (Early stages are like those numbered i to 6

in coeruleus, Fig. 93.)

a: Both migratory and stationary pronuclei surrounded by
halos of cytoplasm excluding chlorellae, that of the female
nucleus being stellate. Immediately preceding maturation
divisions apparently reduced the chromosomes from 56
to 28 and from 28 to 14, the latter by seemingly transverse
division. Male pronucleus flattens as a disc against the
partition.

b: Male nucleus breaks through and unites with partner's
stationary pronucleus.

c: First mitotic division of the amphinucleus.

d: Products retain cytoplasmic halos and have karyosomes
with achromatic fibers connecting to nuclear membranes.

e: Second division, halos disappeared and 56 pear-shaped
chromosomes, producing —

/; Four nuclei with karyosomes.

g: Third division, simultaneous, yielding 8 nuclei, 6 of which
usually form macronuclear anlagen and have karyosomes,
and usually 2, becoming granular, form the new
micronuclei.

h: Division and reduction in size of progenitors of
micronuclei.

/.• Separation of partners, each usually with 6 macronuclear
anlagen and 4 micronuclei.

j: Karyosome of macronuclear anlage forms chromatic
spireme.

TOWARD A GENETICS OF STENTOR 331

In the polymorphus studied by Mulsow, the micronuclei showed
larger and fewer chromosomes than his coeriileus. The latter
seemed to have about 80 in the vegetative stage, while the com-
parable number in polymorphus was close to 56. Therefore, in
polymorphus, Mulsow could demonstrate reduction of chromosomes
during maturation divisions as he could not in coeruleus. During
maturation the micronucleus is not homogeneous but has a central
chromatic body and it is surrounded by a halo of clear cytoplasm
without symbiotic chlorellae. The number of chromosomes
seemed to be at first halved to about 28 and then, when the final
division of the last remaining micronucleus into the two pronuclei
occurred, the chromosomes appeared to be further reduced to 14.
Such double reduction would be entirely anomalous, as Mulsow
noted, and would even call into question whether the bodies
observed were in fact true chromosomes. I suppose that this
paradox can be resolved on the basis that in ciliates chromosomes
are sometimes dumbbell-shaped and only apparently double. If the
chromosome number in coeruleus is reduced from 80 to 20, as
Mulsow indicates, one wonders why he did not also consider this
a case of apparent double reduction.

Conjugation accomplishes at least, and probably mainly, a
recombination of chromosomes with their varying genetic deter-
minants, both by the selection of one set of chromosomes from
two for the pronucleus of one animal, and the combination of these
with a pronucleus from a different individual. This is accomplished
through the micronuclei which alone seem to retain the capacity
to form typical chromosomes, but any recombinant difference
resulting has to be transmitted to the macronucleus which alone

k: Anlage enlarges.

/: Spireme fragments into pieces — like chromosomes — of

different length but equal thickness.
m: These chromosomes split longitudinally into 56 pairs.
n: Some chromosomes disappear (?).
o: Chromatic bodies now with nucleoli, "plastin" body with

chromatic center develops and later disappears.
p: Anlagen nodulate and attach; micronuclei multipled by

mitosis. Normal vegetative stentor produced before

first fission following separation of conjugants.

332 THE BIOLOGY OF STENTOR

can and may solely support the phenotypic expressions in the
form and functions of the cell. Hence the old macronucleus has
to be destroyed though it seems to be carried along for support of
the cell until the new nucleus is ready to take over. We know
practically nothing of how the intricate machinations of the micro-
nuclei are guided such that some degenerate and others do not,
that one should form a stationary and the other a migratory
pronucleus, and that of the products of their union some should
produce the macronucleus and others remain as micronuclei. Yet
it can be assumed that grafting, like cutting experiments, are as
possible in conjugating stentors as in vegetatives ; and this presages
a new field of experimentation in which the manipulation of
conjugating animals in a variety of ways, or the grafting of con-
jugants to non-conjugants or to other conjugants of a different
stage, or the fusion of different species of stentors during conjuga-
tion, may provide clues toward an analysis and understanding of
the forces at work during the amazing performance of fertilization
and nuclear renewal.

CHAPTER XVIII

SPECIES OF STENTOR

Brief histories of our knowledge of the kinds of stentor are
included in Kent's Manual of the Infusoria (1881) and in Johnson's
monumental study of the genus (1893). The first recorded observa-
tions of the group were made by Abraham Trembley of Hydra
fame. In a letter to the Royal Society of London (1744) he
described "funnel-like polypes" of green, blue, and white types
which would correspond to the present species called polymorphus,
coeruleus, and roeseli or muelleri. Feeding, with both rejection and
retention of particles he noted. Division was correctly described
as being oblique and this was confirmed somewhat later with much
surprise by Packard (1937). The present 5. roeseli was included in
Linnaeus' Sy sterna Naturce, tenth edition of 1767, under the
name of Hydra stentorea. Stentors were later clearly differentiated
from hydras and the first use of the generic title Stentor for this
group was made by Oken in his Lehrbuch der Naturgeschichte^
1815.

Oken's genus was not taxonomically accurate, for it included
vorticellids and rotifers and did not consistently use the binomial
nomenclature. Nevertheless, to retain a well-known name, the
genus Stentor Oken 181 5 was recently validated at the instigation
of Kirby (1956), whose account of the generic term is here
summarized. It is well that this was done because at one time the
name had been suggested for a group of howler monkeys. S.
muelleri was chosen as the type species since it was the first species
adequately described and figured, by Ehrenberg in 1831.

Descriptions of the more common and better known species
coeruleus and polymorphus have already acquainted the reader with
the general morphology of this genus. The most outstanding
features in common are the trumpet shape from which the group
derives its name, conspicuous contractility throughout the length
of the body, longitudinal rows of cilia throughout, attachment by

333

334

THE BIOLOGY OF STENTOR

a holdfast at the pointed end, and at the other end a wholly frontal
disposition of the feeding organelles, which spiral clockwise as seen
from above, and consist most obviously of an almost complete
circle of membranelles terminating in a mouth but with no
undulating membrane. The implication of variable morphology in
the names polymorphus and multiformis is misleading and erroneous.
The most complete and recent treatment of the taxonomy of
stentors is to be found in the great work of Kahl (1935) on the
classification of ciliates. Since his writing, one species has been

/eti£i

-pohfTnOt-phuS

roeseli

mzijelLeri

pi/gmseus

Fig. 95. Species of Stentor. The coerideus is about 500 /x in
length and others are approximately in scale. S. pygmceiis after
Swarczewsky, 1929; rubra and loricata after Bary, 1950; felici
after Villeneuve-Brachon, 1940; and amethystiniis after Kahl,
1935- S. introversus (contracted and expanded; after Tartar,
1958a) and others were drawn from life,

SPECIES OF STENTOR 335

transferred to a different genus and a few new ones have been
added. Stentors are perhaps most easily confused with unstalked
vorticellids, but the latter have ciliary rows which are transverse
rather than longitudinal and the oral band spirals in the opposite
direction.

A key to the species and guide to synonomy was provided by
Kahl. Drawing on the available literature and mindful that I have
not seen every one of the species, I shall attempt to give a brief but
distinguishing description of each, illustrated by the frontispiece
and Fig. 95. For all its variability, size is still a useful criterion.
Approximate average diameters of contracted animals are given,
because the degree of extension is variable and samples to be
examined are generally not in repose. New species have been
described from a single specimen but this is certainly to be
frowned upon, because stentors can easily be injured when pipetted
with filamentous algae and may retain abnormal or incompleted
shapes for some time. Moreover, in the method of cell fusion by
grafting we now have a new means for testing species differences.
When diverse forms are combined in about equal proportions they
appear as if stricken and do not produce viable clones as do fusion
complexes of like species.

The following species have blue to greenish or violet pigment
granules :

S. coeruleus Ehrenberg 1830. This is the large, cerulean blue
species, largest (350/x) of all the stentors, with a moniliform
macronucleus.

S. loricata Bary 1950. The only large, self-pigmented green
stentor which builds a case or lorica. The macronucleus is vermi-
form. So far reported only from a stream in New Zealand.

S. multiformis Miiller 1786. This is a tiny (95 /x) blue-green
stentor with an oval macronucleus. (A few further comments are
apropos. When swimming, these stentors often appear plump, with
rounded posterior ends. They have few — approximately 25 —
pigment stripes which are therefore relatively large. Kahl states
that typically there is but one micronucleus. I have found some
collections with symbiotic chlorellae. This species is reported from
brackish or salt water, but I have repeatedly found animals corres-
ponding to its description in fresh water. The fresh water form

336 THE BIOLOGY OF STENTOR

may be a different species, as Penard (1922) first suggested, and
as such may deserve his designation, S. gallinulus. Further study
is required.)

S. amethystinus Leidy 1880. A medium-sized stentor distin-
guished by its violet-blue color and the fact that it does not stretch
out but remains habitually pyriform or conical. Macronucleus is
oval — hence the only medium sized blue stentor with a compact
nucleus. Symbiotic chlorellae are present and with the pigmenta-
tion often produce a dark colored animal.

S. introversus Tartar 1958. A medium-sized (280 ^u) blue-green
stentor distinguished by a retractable head. When withdrawn the
feeding organelles and frontal field are surrounded by a lip of
folded lateral body ectoplasm. Endoplasm is brown, combining
with the pigment to give an olive-green color by transmitted Hght.
Monihform macronucleus. The holdfast is relatively large.

The following species are yellow in color:

S. niger (Miiller) Ehrenberg 1838. A medium-sized (200/x),
yellow to brownish stentor with an oval macronucleus. (Maier
(1903) states that this species has myonemes which are weaker
(narrower?) and that therefore the structure of the kinetics is
more easily studied. These animals do appear delicate as they
wheel slowly through the water.)

S. felici Villeneuve-Brachon 1940. A medium-sized yellow
stentor with moniliform macronucleus. (According to its author
the yellow color of this species is not due to the granules but
resides in the cytoplasm. I think this is to be questioned, since in
all other self-pigmented species the granules are pigmented, and
she remarked that the color is deepest in the granular stripes.
S. niger at first appears to be colored throughout the ectoplasm but
it is the granules which are yellow.)

The following species are small and pink in color:
S. igneus Ehrenberg 1838. This is a tiny (loo/x) pmk to nearly
colorless stentor with an oval macronucleus. It may have chlorellae
(Balbiani, 1893; Johnson, 1893), but all those I have seen were
without symbionts. According to Johnson there is no oral pouch.
As in other tiny stentors, the pigment stripes are few and relatively
broad.

SPECIES OF STENTOR 337

S. rubra Bary 1950. A small, pink stentor like igneus but
distinguished by a rim-like margin on the frontal disc exterior to
the membranellar band. (One wonders if this is merely a variety
of igneus.)

The following species have no pigmented granules and appear
white by reflected light, except when containing symbiotic
chlorellae :

S. polymorphus (Miiller, 1773) Ehrenberg. A large (250 /a) not
self-pigmented stentor without a case, usually grass-green with
symbiotic Chlorella. (My observations confirm Johnson's, that
this species is probably never entirely free of chlorellae unless
special steps were taken to remove them.)

S. roeseli Ehrenberg 1835. This is a small (140/x) colorless
stentor which lives in a case. Usually, but not always, the posterior
nodes of the moniliform macronucleus run together as a rod, or
are more spindle-shaped than the anterior nodes. (In both this and
the following species — muelleri — the stretched animal shows a
much attenuated stalk right up to the well-expanded frontal disc,
hence the shape of an uncoiled trombone; and both show con-
spicuous "bristles" or quiet and extended lateral body cilia near
the anterior end.)

S. muelleri (Bory St. Vincent, 1824) Ehrenberg 1838. A medium-
sized (250 /Lt) stentor without pigment granules which produces a
thick lorica. The cytoplasm is rather brownish in color. The frontal
field generally rests at an angle to the cell axis, hence cala-like in
shape. With uniform chain macronucleus.

A stentor with dark pigment of undetermined color because
described only from preserved specimens :

S. pygmceus Swarczewsky 1929. A medium-sized pigmented
stentor with an abbreviated, chitinoid case found attached to
certain Crustacea (gamarids) in the deeps of the Baikal Sea.
(Apparently the case is used chiefl)^ for attachment because most
of the animal does not withdraw into it. There is a short, monili-
form macronucleus with 4 to 6 nodes.)

This list includes species recognized by Kahl, as well as the new
species loricatUy rubra^ Felicia and introversuSy described since his

338 THE BIOLOGY OF STENTOR

publication. He also allowed the species Stentor glohator Stokes,
1885, though questioning its validity. Since glohator is very similar
to multiformis {gallinulus?) and was described from a single speci-
men, I do not think there are adequate grounds for admitting this
species. Two marine species unique in having a notched membra-
neller band, the so-called S. auriculata Kent, 1881, with a compact
macronucleus, and S. auriculatus Kahl, 1935 (auricula Gruber,
1884), with a multinodal macronucleus, are probably variations of
the same species (see Andrews, 1948a). They have been shown to
be not stentors at all but to belong to the genus Condylostoma^
because they have an undulating membrane and creep along the
bottom as well as attaching by the posterior end (Faure-Fremiet,
1936). Also with notched oral band is a non-pigmented ciUate
found in numbers on a branch of Fucus by Silen (1948) who pro-
posed the name Stentor acrobaticus. This interesting organism,
attaching by the posterior end or clinging by 2 folds of the lateral
body wall, ghdes along cross-striated filaments apparently of its
own making. Two compact macronuclei were stained. This is a
doubtful species because its contractility was not notable as in
stentors and neither feeding organelles nor lateral striping were
described.

Hence there are about 1 3 known species of Stentor. This number
includes quite recent discoveries, suggesting that still more species
are to be found. I have myself seen two or three additional forms
which do not correspond to present descriptions, but I would not
give them names until more abundant collections are available.

CHAPTER XIX

TECHNIQUES

The unique combination of qualifications of Stentor mentioned
at the beginning and displayed throughout this review, may well
have intrigued the reader with the opportunities provided by this
animal for an integration of multiple approaches to a study of the
life and structure of the cell. He will also have become aware of
the evident gaps in our knowledge, and the need for pursuing
provocative suggestions or following hints in the literature to
demonstrated conclusions, as well as the value of confirming and
thus securing as a sound foundation many points not yet well
established. Above all, the special promise of the Stentor studies
should be carried to the level of theory on which new explanations
and general principles emerge; and for this task all the many past
investigations may be regarded as establishing only the beginnings.
It remains to connect the potential student of Stentor with this
organism through an account of methods, which are themselves
doubtless capable of much further refinement and expansion.

I. Collecting

Stentors are most likely to be found in large, permanent ponds
or lakes, but they also live in streams ; and the outflows of sewage
plants are not to be neglected. The collector may equip himself
with a set of cream cans of two-quart capacity and a kitchen
strainer fastened to a long handle. Keeping in mind that stentors
are usually attached, one gathers with minimum disturbance
samples of submerged and floating vegetation such as duckweed,
Spyrogyra mats, and dead cattail leaves which are placed in the
container. More vegetation is then scooped up with the strainer
and gently wrung out into the can until it is nearly filled. Attached
and loosened stentors are likely to be gathered in this way, with
sufl[icient natural medium to start a culture. Location of each
sample as a possible source should be noted on the container

339

340

THE BIOLOGY OF STENTOR

because one can return to a source year after year and find the
same species of Stentor.

Returning to the laboratory, the Hds are removed from the cans
to expose them to the air. They are left to stand for half a day but
no longer. In this time the stentors will swim to the top. If allowed
to stand longer an injurious putrefaction may set in, or worms and
water fleas may take their toll, for stentors do not appear to be a
dominant form like parameciaor hypotrichs, and persisting cultures
are usually not obtained by simply letting the samples stand. After
a few hours, stentors, if present, will be found near the surface
where they are gathered by pipetting along the miniscus and
agitating the floating vegetation and debris. A scraping action with
the tip of the pipette when water is being sucked up will serve to
loosen stentors which have become attached. This material is

Fig. 96. Equipment for culturing stentors.

TECHNIQUES 34I

transferred to a caster dish or other shallow container and examined
for stentors under low powers of a stereomicroscope. If stentors
are found, the whole sample container may then be rotated for
gentle agitation and more samples poured out. A portion of the
original sample is then passed through filter paper of medium
porosity which will remove all large forms and pass only minute
organisms on which stentors can feed, and this natural medium
can then serve for the starting of cultures.

If stentors cannot be collected in the field they may be obtained
in mixed culture from several biological supply companies.

The next step is to select stentors out of the sample dishes,
leaving competitors and predators behind. For this purpose micro-
pipettes are necessary and the ones I use are made from narrow,
polyethylene, catheter tubing obtainable from surgical supply
companies. This and other items of culture technique are illustrated
in Fig. 96. The tubing is softened by placing it across the narrow
flame of a wing-top gas burner and pulled out to a fine point. The
degree of heating is critical. If too cool the tubing breaks when
pulled and if too hot it collapses. One can expect to spoil a dozen
pipettes before one gets the knack. When good tubes are drawn they

A. To the right: micropipette (actual size) with poly-
ethylene tip, rubber tubing "bulb", and glass rod plug; as well
as fine wire (bent) used when cleaning. To left: drawing out
polyethylene catheter tubing over wing-top burner for pipette

tips.

B. Glass block cell containing i ml in which all specimens are

clearly visible.

C. Culture in jar with hole punched in cap, examined briefly
with spotlight and magnifying glass to follow development of a

culture.

D. Development of clones. Single stentor first introduced
into one cell of deep depression slide; transferred to test tube
when multiplied to about 25 animals; transferred again from
hundreds in the test-tube to a cotton-plugged Ehrlenmeyer
flask. Filtered culture medium plus culture of food organisms

used throughout.

E. Migration tube for obtaining clean stentors. Main body
of half-inch diameter tube is covered with black plastic
electrician's tape and filled with clean water. Concentrated
S. coeruleus introduced at {x) will migrate away from lighted end

and are recovered, clean, at other end {y).

342 THE BIOLOGY OF STENTOR

are cut off to proper length and diameter with scissors and a piece
of thick-walled rubber tubing, plugged with glass rod, slipped on as
a bulb which will not be over-responsive to pressure of the fingers.
These pipettes are unbreakable and can be used for years if cleaned
out occasionally by passing a twirling fine wire through the points.

Sample dishes are now searched and individual stentors picked
up and transferred to glass block cells, one for each species if
desired. About 50 stentors of a kind should be isolated, if available,
and the isolation dish should then be surveyed, this time to
pipette out any contaminating organisms that may have been
carried over with the stentors. Block cells or their equivalent are
recommended because in them no organism escapes from view.

Enemies of Stentor and reports of their predation include the
following: the heliozoan Actinosphcerium eichhornii (Cienkowski,
1865) 5 the water plant Utrichularia which captures and kills stentors
in its unique bladders (Hegner, 1926); rhabdocoele worms
(Prowazek, 1904; Gelei 1925); oligochaete worms like Chcetogaster
diaphanus (Lankester, 1873); the giant ciliate Bursaria truncatella
(Lund, 1914) ; and the smaller ciHate, Dileptus, with its proboscidial
stinging trichocysts. I have observed that the little scavenger
ciliate, Coleps, devours injured stentors; and nematodes, water
fleas, and hypotrichous ciliates are to be excluded as predators or
otherwise undesirable.

2, Culturing

I shall now describe my method of setting up cultures, though
this is not the only nor possibly the best procedure. A half-pint,
wide-mouthed peanut butter jar is filled to a depth of about one
inch with the filtered pond water. A large pinch of absorbent cotton
is then pulled apart to form a loose mesh and dropped in. The
cotton is regarded as a purified substitute for pond vegetation.
The isolated stentors are then washed into the jar with a squirt of
filtered pond water. One drop of skimmed milk, one or two boiled
wheat or rice grains, or fragment of a rabbit-food pellet is then
added as a source of nutrients, producing a population of bacteria
and tiny flagellates and other food organisms from the original
pond water which passed through the filter paper. In this way as
many seeding stentors as obtainable are returned to the same water
from which they came. Only about 100 ml of starting culture is

TECHNIQUES 343

set Up in order that the stentors may themselves possibly regulate
the medium to their Hking; and very little nutrient is at first added,
in proportion to the few stentors present.

Progress of the starting culture can then easily be followed by
placing the jar briefly in front of a bright spotlight and examining
with a magnifying glass. At the end of a week, if the stentors are
multiplying, more nutrient is added, at first only a drop or two of
skimmed milk, but only if the water has become clear. If turbid
with uneaten flagellates and bacteria, the jar is let stand another
week before nutrifying. Since milk is a complex mixture forming
a nearly perfect food, it serves as a good basic nutrient and ionic
medium for stentors and a variety of other protozoa, including of
course the food organisms (Tartar, 1950).

As the stentors increase in number, more lake or other natural
water which has been passed through a Millipore filter to remove
all protozoa and their cysts is introduced from a stock jar, with a
little more cotton. Eventually the culture jar will be filled to the
top and can be nutrified once a week with 5 or 6 drops of skimmed
milk. (Cream content would form a film on top and exclude the
air.) From the beginning the jar is covered with its original cap, in
the center of which is punched one hole with a large nail or ice
pick, the cap preventing contamination and evaporation and the
hole allowing gaseous exchange.

Such cultures will remain in thriving condition for many months.
If removal of detrimental cohabitants was unsatisfactory, or if
hypotrichs, nematodes, etc., should later infest the culture, one
has to begin again, treating the culture as if it were a pond sample
and isolating stentors as before. A cardinal precaution is never to
over-nutrify the culture so that a distinctly putrid condition arises.
In the course of months the stentors may diminish in abundance
in spite of the regular additions of milk. When this occurs it is
assumed that the water should be changed. Since the stentors are
mostly attached to the sides and the cotton fibers, the whole jar
can be gently emptied, or the cotton can be retained, and then
immediately filled with filtered water .In the meantime the stentors
have remained attached to the sides and are protected by a fluid
film. In spite of some loss there will probably still be enough
animals to handle the large amount of new water. One may want
to add less milk now until the animals become plentiful. A con-

344 THE BIOLOGY OF STENTOR

tinued source of food organisms will of course have been retained
in the film adhering to the emptied jar. It is well to have three or
four jars of the same stock. These can be developed by splitting
the contents of one jar between two and refilling both to the top
with filtered lake water, adding more cotton as needed.

These procedures may not appear elegant but they have served
to maintain healthy stock animals in more than sufficient abundance
for my micrurgical operations continuously for 8 years, during
which not one of lo stocks has died out. The same method has
been used successfully for growing coeruleus, polymorphtis, roeseli,
and introversus. For the last named, skimmed milk must be added
very sparingly and never when clouds of uneaten colorless flagel-
lates are still present. In my experience, the cultivation of other
species like niger and multiformis is attended with great difficulties
and probably calls for exploring distinctly different methods.

Temperature at which the stentors are grown is another
important factor. Schwartz found that stentors do better at lower
temperature than at higher (io° vs. 22°C). I have found that in
winter the culture room must be thermostatically controlled
to avoid wide changes in temperature.

Genetically more uniform material is assured by developing
clones or cultures derived from a single individual. This is best
done after a good culture of the wild stock is obtained, for one can
then use filtered water from the culture itself as a starting medium
and be sure of its optimal nature. First one should develop a
separate culture of food organisms, either by nutrifying coarse-
filtered Stentor culture fluid with skimmed milk or by growing
any of the food organisms soon to be listed. Into a deep depression
sUde holding about i ml is isolated one stentor in a small drop,
checking at once that only a single animal is present. Then are
added 5 drops of the filtered parent culture and 5 drops of food
organisms. The slide is placed in a moist chamber and more food
organisms can be added as needed. One should of course start with
several such isolations to assure that at least one will be successful.
Further addition of food culture may be necessary. When about
25 stentors have developed, the contents of the slide are transferred
to a 25 ml test-tube with aluminium cap into which have previously
been added, at the bottom, 10 ml of food organisms and, on the
top, the same amount of filtered parent culture medium. When the

TECHNIQUES 345

stentors become abundant, the test tube can be emptied into a
culture jar which is carried forward as described.

Alternatively, the test tube with its clone of stentors can be
emptied into an Ehrlenmeyer flask, plugged with cotton and fed
by repeated additions of food organisms, sub-culturing when the
flask becomes filled. Growing the food organisms separately
prevents over-nutrification and is therefore recommended for
developing clones as well as for producing very abundant and clean
cultures for biochemical studies.

To obtain concentrated animals one can gently shake the flask
cultures to loosen stentors attached to the sides and pour the
contents into graduate cylinders; for at first the oxygen will be
uniform throughout and the stentors (at least coeruleus) will rapidly
sink to the bottom in mass and the overlying fluid can be decanted.
If it is now desired to free these animals from most of the food
organisms one may take advantage of the speed with which most
races of coeruleus swim away from the light — or perhaps the reverse
in the case of green polymorphus and niger. Whiteley introduces
the concentrated animals at the lighted end of a large, horizontal,
covered tube with both ends bent upward and filled with Millipore
filtered medium (Fig. 96E). Stentors soon migrate to the lighted
end, leaving the slower bacteria and food organisms behind, and
are promptly removed for study.

Other methods which have successfully been employed for the
cultivation of stentors will now be reviewed. First we have to
consider the basic fluid medium. Distilled w^ater is not used
because it is injurious and tap water is avoided because it picks up
metals in the pipes and may be chlorinated. Natural waters from
ponds and lakes are preferred. They may be freed from contaminat-
ing organisms either by previous boiling or by fine-filtering — the
latter is recommended. These waters will contain dissolved sub-
stances natural to the Stentor habitat, but many investigators
recommend the addition of a mixture of inorganic salts. (I attempt
to supply these along with organic materials in the added milk.)
Peters (1904) in particular emphasized the importance of the salt
content of the medium and suggested the following mixture,
figures representing moles of the salts: CaCU (0-0005), K2HPO4
(0-00015), NaNOa (0-00015), and MgS04 (0-00015). This solution

346 THE BIOLOGY OF STENTOR

was used in the ratio of 500 ml of salt solution to 3500 ml of
culture water. The calcium was especially important. Chalkley's
solution for amoebas (NaCl, o-ig; KCl, o-oo4g; CaCl2, 0'Oo6g,
and 1000 ml glass distilled water) has also been used (Randall and
Jackson, 1958), and Hetherington (1932b) suggested 0-06%
artificial sea water. Uhlig (unpublished) found that the addition
of soil extract and Knop's solution in equal parts produced
excellent cultures of coeruleiis. The formula for Knop's is: MgS04,
0-25 g; CaNOs, o-ig; K2PO4, o-i2g; KCl, oi2g; FeCls, trace;
and 1000 ml of distilled water. In all these additives the guiding
principle is of course that essential ions and elements should be
supplied in surplus.

Stentor polymorphus has been cultured in soil extract
(HammerHng, 1946) or o-oi% Benecke's solution (Schulze, 1951)
with the green alga Gonium tetras as food. Randall and Jackson
grew these stentors in Chalkley's solution with added gel from
wheat grains boiled in the same. Sleigh (1956) grow polymorphus
in a basic medium each liter of which contained inorganic salts
measured in millimoles as follows: NaCl (1-4), KCl (0-05),
NaHCOs (0-045), CaCl2 (0-035) and CaH4(P04)2.H20 (0-006),
made up in distilled water of about pH 6-8. This solution was
nutrified with wheat grains and the stentors were fed Chilomonas.

These cultures are to be kept in the light if the stentors bear
symbiotic chlorellae, but too bright an illumination is undesirable.
The observation that polymorphus undergoes fission only at night
(Hammerling, 1946) suggests the possibility of obtaining simul-
taneous division in well-fed mass cultures transferred from light
to darkness.

To the basic fluid medium may be added nutrient materials on
which bacteria and other food organisms can live. Hay and hay
infusions have not been found satisfactory, perhaps because the
culture becomes too acidic. The hydrogen ion concentration should
fall between 6-2 and 8-o (Strom, 1926; Belda and Bowen, 1940).
Prowazek (1904) used lettuce leaves but he found that his cultures
went through periodic depressions. The same, or lettuce extract
if cleaner cultures are desired, was recommended by Belda and
Bowen. They remarked that cultures should be grown in a darkened
place because growth is not satisfactory where abundant green

TECHNIQUES 347

algas develop, and I can confirm this. As nutrient, Stolte (1922)
used the scum from lettuce infusions, or beef extract. He found
the presence of algae useful, but I think this was because his cul-
tures were quite putrid or over-rich and therefore oxygen deficient.
Wheat and barley grains, boiled to prevent germination, are
satisfactory (Hyman, 1925, 193 1; Weisz, 1948c). The addition of
dry leaves and reeds was recommended by Peters and I have
simulated this by adding desiccated lettuce, but without
conspicuous advantage.

Stentors are to be provided with food organisms. Very likely
stentors can accumulate and ingest bacteria but the eating of larger
organisms should be more efficient. The following organisms have
been observed to be eaten and digested by stentors:

Colpidium

Blepharisma

Paramecium hursaria

Minoidium, and other colorless flagellates

Small rotifers

Chilomonas

Halteria

Tetrahymena

Glaucoma

Gonium^ and other colored flagellates.
With the least trouble, cultures can be inoculated with these
organisms, but for cleaner and more abundant cultures the food
organisms should be grown separately. Burnside (1929) fed
coeriileus on Halteria grown separately with hay or malted milk.
Constant cultures fed with Colpidium were set up by Hetherington
(1932a). Schwartz (1935) grew his animals in filtered pond water
plus soil extract and fed them on Colpidium grown separately in a
0-03% solution of yeast extract. Chilomonas with Paramecium were
recommended as food by Turner and Brandwein (1937). Stentors
cannot readily capture Paramecium caudatum so that smaller and
less vigorous species are recommended. Gerstein (1937) and
Dawson (1953) used Blepharisma grown separately and obtained
long-enduring cultures. I also found Blepharisma to be excellent
food; but the pigment of this ciliate has a photodynamic action
(Giese, 1957) which might prove damaging under bright illumina-
tion during operations on stentors which have ingested them.

348 THE BIOLOGY OF STENTOR

3. Survival on slides

Although Balbiani (1889) reported keeping one stentor alive on
a slide for nearly a month with feeding, most students since
Johnson have found that stentors isolated into a few drops or even
into watch glasses do not long survive. In fact, one gains the
impression that the results of many investigations may have been
compromised by poor survival on slides and the unfavorable
conditions this implies. We have noted, however, that stentors
will multiply and clones can be started in deep depression slides
containing only about i ml. Hetherington (1932b) was able to
maintain stentors for a year in Stender dishes, a few dozen to
the dish.

Exploring the limits of isolation culture, I kept a normal
coeruleus for 41 days in 3 drops of medium with some feeding and
two transfers to fresh fluid, together with the addition, after
3 weeks, of two squashed stentors which I thought should supply
what stentors need. Experimental animals which had been
operated on in various ways survived as long as 16 days in 3 drops,
but as a rule stentors live only about one week under these condi-
tions. It is quite possible that improvements in isolation culture
can be made, and the advantages of adding some stentor brei is
indicated. In any event, if specimens or controls do not survive
for at least two days, I regard the experiment unreliable.

4. Staining

More or less standard methods have been used for the cytological
study of stentors, and suggestions regarding fixation and staining
are given in the papers of Johnson (1893), Schwartz (1935),
Randall and Jackson (1958) and especially of Weisz (1949a, 1950a).
Diff"ering from most other ciliates, stentors have not revealed a
neat silver-line system either by the wet or dry methods of silver
staining (Villeneuve-Brachon, 1940; Weisz, 1949a). Merton (1932)
made a valiant effort to fix and stain stentors in the extended state,
but I am inclined to agree with Johnson that semi-contracted
animals are good enough for most purposes. To see the form of
extended stentors living animals are the best. Treatments which
have been used for anaesthetizing the contractile elements have
already been discussed in Chapter XIV.

TECHNIQUES 349

5. Cutting methods

The simplest way of obtaining Stentor fragments is to shake the
animals briefly but vigorously in a tiny vial. Presumably formation
and breaking of bubbles adjacent to the cells splits them into pieces.
This was the method first used by Lillie (1896) who well knew
that cleaving eggs can be separated into their blastomeres in this
manner. If a few bits of broken cover glass are added to the vial,
random cutting occurs. For more precise hand sectioning needles
are used. Steel needles trimmed and sharpened to a very fine point
were used by Prowazek, Schwartz, and Weisz. I employ only glass
needles, made by holding the ends of two glass rods in a small gas
or alcohol torch flame until they fuse together lightly and form
a ball of molten glass between, whereupon the rods are quickly
separated as they are withdrawn from the flame and one or two
good needles are produced. The puUing must be done at just the
right time, w^hen the glass is neither too fluid nor too congealed,
and this requires practice. The glass rod should be of soft glass.
I do not know the specifications, but if success is not attained one
should try a diflFerent stock. When properly made (Fig. 97A) glass
needles provide the finest points obtainable and are used like a
knife in cutting.*

In earlier days, ciliates were cut without quieting by merely
confining them to a tiny drop, beside which a large drop of
medium was placed, quickly to be flooded into the small one
immediately after the operation to prevent drying (Balbiani, 1889).
The best quieting agent is a solution of methyl cellulose which
greatly retards stentors by its high viscosity. This method was
introduced by Marsland (1943) for paramecia and later adopted
for stentor experiments by Weisz (1951a). The solution seems more
innocuous for stentors than for paramecia, and I noticed that it
quickly kills Blepharisma. Sleigh (1956), in his studies of ciliary
coordination, found that methyl cellulose is entirely reversible in
its eff"ect; and I kept coeruleus for two days in a thick solution
w^ithout apparent injury to the animals. Nevertheless, methyl
cellulose treatment sometimes showed an inhibiting influence on
early primordium formation (Tartar, 1958c). Early dividers in

*Uhlig (i960) used a Spemann pipette to remove ectoplasm of the
fission line bit by bit to produce doublets by aborted division.

THE BIOLOGY OF STENTOR

Fig. 97. Equipment for operations on stentors.

A. Glass needle drawn from soft glass rod for cutting;
eyelash fastened to handle for rolling over specimens to

examine all sides when following operated animals.

B. Moist chamber, a plastic sandwich box with wet filter
paper on the bottom and depression slides stacked on 2 bridges.

C. Bench for operating. As he bends over microscope,
operator automatically presses hinge at edge which turns on
spring switch and embryological lamp. Bench top used for arm
rests. Ordinary blue light below hole in bench is used in
searching culture samples by transmitted light. Both sources

of illumination have glass heat filters.

D. Canning funnel covered with drum-head of fine bolting
silk and immersed in culture jar, for maintaining large fusion

masses under optimum conditions.

E. Operating slide to which a square of finely woven fabric is
applied with melted paraflSn carries large drop of methyl
cellulose into which stentors have been introduced with the

micropipette.

TECHNIQUES 351

Stage I and even occasionally at stage 2 resorbed the primordium
if they were kept too long in the viscous fluid, and regenerators
either did likewise or the anlage was considerably delayed in its
appearance. After cutting in methyl cellulose the animals should
therefore be washed once by passing them through a large drop of
filtered culture medium. Old solutions had a greater inhibiting
effect so that it is well not to employ dissolved methyl cellulose
which has been kept longer than two months. A stock solution of
methyl cellulose may be prepared in the following manner: add
50 ml of dry methyl cellulose to the same amount of boiling
filtered lake water used in culturing; stir the fibers to remove air
bubbles and assure complete wetting ; remove the beaker from the
stove and allow it to stand for half an hour, after which another
50ml of cool lake water is added with stirring; let stand overnight
until the solution is cool and the methyl cellulose thoroughly
dissolved.

My method of operating is quite simple. I use a stereomicroscope
without base or mirror, because the instrument then stands low
and the bench itself can be used to give support to the arms
during deHcate operations with the needle (Fig. 97B). The 'scope
should have the highest powers available (about 150 x), which
still gives sufficient working distance between the lower lens and
the specimen to permit operating. Lower magnifications are needed
for capturing specimens. Sub-stage illumination is provided by a
hole in the bench covered by a heat filter glass with the light
underneath. But for operating, reflected light is used from an
embryological lamp, also supplied with a heat filter. It is con-
venient as well as saving of bulbs to arrange a pressure switch with
a strap hinge at the edge of the bench so that this light turns on
only when one bends over the microscope. Reflected light has the
great advantages of not silhouetting the stentor but clearly revealing
its entire surface topography, and of avoiding eye-fatiguing glare.

Using a toothpick dip, a fairly large drop of methyl cellulose
solution is placed in the center of a piece of finely woven cloth
stuck with melted paraffin to a thick slide. The slight roughness
of the cloth keeps the specimen from skidding under the needle,
as it would on glass, thus helping to hold the animal in place.
Paraffining prevents spreading of the drop. A white cloth is used
for pigmented forms like coeruleus and a black cloth for unpig-

352 THE BIOLOGY OF STENTOR

mented species like roeseli. A thick slide is easier to pick off the
microscope stage than the common thin ones. I use a depression
slide turned upside down.

With a micropipette an animal is then transferred with minimum
fluid to the center of the drop of methyl cellulose. The stentor
must not be allowed to wander to the surface, for then, oddly-
enough, the ectoplasm will adhere and tear off when the animal
is moved. With an eyelash fastened to a narrow handle the animal
is then pushed to the bottom of the drop ; reason : a glass needle
is too sharp and may impale the specimen.

The glass needle is then taken in hand and after gently moving
the stentor into position the proper cut is made. With practice the
needle can be precisely ''located" under the microscope so that
in time breakage becomes infrequent. After cutting at high magnifi-
cation, the objectives are shifted to low power and the specimen
removed with the micropipette and placed in a block cell with
several drops of filtered culture medium to wash off the methyl
cellulose. Washed specimens are then transferred to two large
drops of filtered medium in a shallow depression slide, the code
number of the experiment can be written in pencil on the frosted
edge, and the slide stacked '' pig-stye fashion " in a moist chamber.
For the latter I use plastic sandwich boxes, the bottoms of which
are covered with a thick layer of wet filter paper. One box will
accommodate about 2 dozen stacked slides. (Fig. 97c). At intervals
depending on the experiment, the slides are removed from the
chamber and examined by reflected light under the microscope,
moving them always in the same order, stacking them then in
reversed order in another moist chamber. When necessary, the
specimen can be transferred briefly to a drop of the methyl cellu-
lose for close examination under high magnification, rolling it into
position with the eyelash. On termination of an experiment, the
drop is shaken off the slide, leaving some moisture by which it may
be rubbed clean and dry with cheese cloth, and the code number
erased. I do not use elaborate cleansing methods because these are
unnecessary when control of bacterial flora is not involved.

My experience agrees with the pioneer observations of Gruber,
that healing of cut surfaces is always prompt and effective. Even
if most of the ectoplasm is removed, the remainder will stretch
and manage to cover all the endoplasm (see Fig. 25c). Defective

TECHNIQUES 353

healing therefore indicates poor material or conditions of
experiment.

Intentional disarrangement of the stripe pattern is illustrated by
rotating anterior halves i8o° on posterior halves. The cell is first
cut through halfway on one side and then gently rolled over and
cut through on the other side. The surface is thus completely
severed but the two parts remain joined by the endoplasm, the
first wound healing while the second is being made. Quickly, and
before firm rejoining of the ectoplasm occurs, the side of the needle
is spun around the anterior end causing it to be rotated until it
takes a position in which the mouthparts are now opposite the
primordium meridian on the posterior half. Within a minute, firm
healing of the parts in their new orientation will have been effected
and the specimen is ready to transfer.

Regeneration, singly or en masse, can be induced by brief salt
treatment to cause shedding of the membranellar band. For indi-
vidual specimens in a drop on a sUde, I add one drop of 4% urea
solution, rescuing the animal as soon as the membranelles are
fimbriated. Shedding is completed in a large drop of culture
medium to which the animal is transferred for washing. Regenera-
tion is easily evoked in this manner, a procedure useful when
cutting would reduce or disarrange the lateral striping.

When many stentors in simultaneous regeneration are required,
the following procedure is followed. Place 10 ml of Stentor sample
in a 25 ml graduated cylinder and, slanting the vessel, introduce
drop by drop and with minimum turbulence an equal volume of
4% urea (or other solution having the same effect; see p. 252).
Contents of the cylinder are then poured into a caster dish under a
dissecting microscope and follow^ed until membranelles begin to
be sloughed. Effective time and concentration of the salt may vary
with the condition of the animals. The dish is then emptied into a
tall olive bottle or 100 ml cylinder and quickly filled with filtered
lake water to dilute and stop the action of the chemical. As soon as
the animals settle to the bottom, the vessel is decanted and filled
again with w^ater for a second washing, after which the settled
animals are ready to be set aside or used in experiments.

It may be mentioned here that the studies of Chambers and Kao
(1952) and of de Terra (1959) demonstrate that injection and
autoradiographic techniques are quite feasible in Stentor.

354 THE BIOLOGY OF STENTOR

6. Grafting

Ciliates may be said to graft themselves in conjugation or fusion
of gametic individuals (e.g. Metopus). Heliozoa reincorporate
separated pseudopodia and may fuse together in clumps for the
purpose of digesting large food organisms. Doubtless for this
reason, heliozoa vv^ere the first protozoa artificially to be grafted,
beginning w4th Cienkoweski in 1865. The history of these experi-
ments, as w^ell as the generally futile early attempts to fuse amoeboid
forms, w^as recounted by Okada (1930) in connection with his own
experiments of this type. More recently, Daniels (1951) has been
able to fuse giant amoebas, by impaling one cell on top of the other
with one needle and breaking the cell membranes together with
another. To Gruber (1885a) belongs the credit for conceiving that
stentors can be grafted if cut surfaces are brought together quickly
before healing. In a few instances he was briefly successful in
rejoining cut stentors. Unmindful of Gruber's explorations,
Morgan (1901a) forecast Stentor grafting but was unsuccessful in
realizing it. Ignorant of both, I independently succeeded in fusing
as many as 4 stentors together (Tartar, 1941b), this possibility
being suggested at once by the ease with which two stentor halves,
left attached by a small strand of cytoplasm, fused together again.
That grafting should be successful in some other ciliates is at least
suggested by Prowazek's (1901) experience with one Glaucoma
scintillans, cut parts of diflPerent individuals being temporarily
united under a cover glass. How much may be accomplished by
cutting and shifting of parts in forms too small to graft is shown in
the excellent experiments of Suzuki (1957) on Blepharisma.

The method of grafting was explained in my first paper on this
subject (Tartar, 1953). Using the cloth-covered slide already
described, two stentors are placed in a large drop of methyl
cellulose. The stentors are moved quite close to each other. Each
animal is then cut with a sharp needle and the wound surfaces
opened widely. By using now a blunter needle, from a stock of
needles from which the fine tips have become broken, one animal
is pushed so that its gaping wound surface is brought firmly in
contact with that of the other. An extra thrust will then spread
the two wound surfaces a bit so that the temporary membranes
which had been formed over them after cutting are broken afresh,
and the two animals will then fuse firmly together (Fig. 98A).

TECHNIQUES 355

Adhesion is therefore by the naked endoplasm. Sheering of one
animal against the other promotes fusion, as if fibrous proteins
were then stretched out to expose free bonding points. Even if
firm union is achieved at only one point, this is sufficient; for
fusion will soon spread throughout the whole wound area.

Large fusion masses are produced by repeating the simple
grafting of 2 stentors. To a fresh cut in a joined pair another cut
animal is added, and so forth. If masses of 50 or more stentors are
desired, I stop occasionally to give both the mass and the operator
a rest, washing the specimen free of methyl cellulose by transfer
into culture medium where it remains until grafting is continued.
If broad adhesion is not secured and parts are left dangUng or
projecting, there is likelihood that the separate individualities will
later pull free. With a blunt needle I therefore poke protruding
heads and tails into the mass to give a compact form with uniform
surface. Large masses may be kept in a fruit-canning funnel closed
with bolting silk to permit fluid exchange with a Stentor culture
in which the receptacle is immersed (Fig. 97D). Pigmented masses
are then easily found and pipetted into block-cells for examination
under the microscope.

To graft a patch of ectoplasm onto another stentor, the desired
area is cut from one, using the granular stripe pattern as a guide,
but the patch is not entirely isolated and the remainder of the cell
is now used as a handle, impaling it with the glass needle and
carrying the patch to the host, in which a fresh incision has just
been made and opened. Adhesion is accomplished by pressing the
cell remainder into the cut opening, fusion spreading to the critical
patch; but before secure union is eflFected a tug on the cell
remainders orients the patch in place and excess parts are then
cut off. In this way the patch is grafted in the desired position
without injury from contact with the needle (Fig. 98B).

Similarly, if stentors are to be grafted as heteropolar telobiotics
the heads are first cut loose like the lids of flip-top boxes and
fused by thrusting them together, whereupon the union extends
to the cell bodies, and the heads are then excised (Fig. 98c). The
bodies of the animals have then not been touched with the needle
and their individual stripe patterns remain wholly normal. An
obvious modification of this procedure is used for making head-to-
tail tandem grafts or tail-to-tail telobiotics.

356

THE BIOLOGY OF STENTOR

Fig. 98. Grafting operations.

A. Producing a doublet, a: Two stentors are split down the
back with sharp needle and opened wide to expose the endo-
plasms. b: With a blunt needle (broken tip) animals are
orientated and one is pushed against the other, exposed endo-

plasms pressing together, c : Doublet stentor resulting.

B. Implanting a cell sector, a: Cuts made from both ends to
isolate the primordium sector with or without nuclear nodes,
leaving cell remnants at each end. b: Host split open and graft

TECHNIQUES 357

Different species of Stentor can be grafted almost as readily as
individuals of the same kind by the same methods ; and in most
cases enduring unions are produced.

7. Minceration

The striped ectoplasm of stentors can be separated into 50 or
more disarranged patches by slicing through the surface with the
tip of a needle. After repeated cutting, areas will be circumscribed
and ''float" free on the endoplasm. Further transections of these
patches not only cut them in two but drag them into random
positions. Maximum randomness is produced if, before mincing,
quarter sectors of the stentor are traded — by transverse and
longitudinal cuts, first rotating the anterior half 180° on the
posterior and then the left on the right half. The latter operation,
or beginning minceration, will render the animal incapable of
directional swimming and the operation can be continued under
optimal conditions in a drop of medium, without further recourse
to methyl cellulose.

8. Enucleation and renucleation

If stentors are to be enucleated, abundant animals are first
isolated into a caster dish and left to stand for one day. The
stentors will then have used up much of the available food material
and will be largely free of food vacuoles which might be mistaken
for nuclear nodes. A pellucid stentor is transferred to a drop of
methyl cellulose on the slide with black silk. If the position of the
embryological lamp is adjusted so that it overthrows the specimen
a bit, the nodes of the macronucleus will appear as opaque white,
or sometimes glowing bluish bodies against the dark background
(see Fig. 79 a). A slice with the glass needle from the upper right

put in place, either homopolar or heteropolar. Posterior

remnant pushed into slit to fuse, then each remainder pulled as

indicated to orient graft as fusion extends to it. Then cell

remnants excised, c: Graft in placfe; in this case its anlage will

be caused to resorb by the non-differentiating host.

C. Head-to-head telobiotic. a: Heads of two stentors cut but

left attached to cell bodies, b: Underparts of heads, with

exposed endoplasms, thrust together, then excised as fusion

spreads to the main bodies, c: Resulting telobiotic.

358 THE BIOLOGY OF STENTOR

corner of the stentor to near its base is made and the two halves,
still attached at the posterior pole, are spread out widely (Fig. 99A).
Thus exposed, the nuclear nodes stand out more clearly than ever
and are rapidly teased out with the needle or sliced off with mini-
mum cytoplasm. All this can be done while leaving the oral struc-
tures entirely intact. When all visible nodes have been removed, the
two halves of the specimen are then brought together in normal
location to aid their rejoining in normal shape. Then the posterior
end of the cell is split apart and the last nodes searched for among

Fig. 99. Enucleation and renucleation.

A. a: Incision to enucleate coeruleus without disturbing
feeding orgenelles and with minimum loss of cytoplasm, h:
After cell is laid open, margins with macronuclear nodes are
excised or nodes teased out. c: As specimen heals together,
posterior end is opened to cut out remaining nodes obscured by

carbohydrate reserves.

B. In renucleation with nodes from same or a different species,
enucleated host is split open when endoplasmic sac with nodes
is available {a); the sac is broken against host wound, the endo-
plasms fusing; and nuclear nodes are then securely inside (c).

TECHNIQUES 359

the granular carbohydrate reserves which tend to obscure them.
Soon after this operation the stentor will appear entirely normal
but lacking the macronucleus.

Comandon and de Fonbrune (1939b) devised a method and
instrument for transferring the nucleus of one amoeba into the cell
of another. Essentially, the nucleated cell is pressed against the
enucleate one and the nucleus pushed through from the donor
into the host. In Stentor the procedure is different. A healed,
enucleate specimen is returned to a drop of methyl cellulose along
with the donor of the new nucleus. The two animals are brought
adjacent with the eyelash. The nucleated specimen is then cut
open as described and one or more macronuclear nodes are teased
out of the cell without ectoplasm but within a thin halo of endo-
plasm. This endoplasm quickly forms a membrane around itself
which serves both to preserve the nucleus from exposure to the
medium and to form a means of transfer. The enucleated cell is
then cut open and considerable area of free endoplasm is exposed.
The free sac of nuclear nodes is picked up with the point of the
needle and broken against this endoplasm, whereupon fusion occurs
and the nodes are taken in (Fig. 99B). Whole macronuclei can be
transferred in the same manner with minimum endoplasm if
dividers with compacted nuclei are used as donors. Healing of the
host wound firmly traps the nucleus within the cell. Nuclear
transfers are possible between different species of Stentor, and it
can be seen that the foreign nucleus is not ejected but persists in
the alien cytoplasm.

The possibility of other techniques should be explored. For
instance, it would be desirable to find a non-toxic agent which
would permit healing but prevent intimate rejoining of cut edges
of the ectoplasm so that stentors would remain as cut; or to
discover a means of agglutinating stentors so that they adhere by
the ectoplasm without fusing. In the first case, would a mincerated
stentor express multiple individuality, and in the second could the
impulse to primordium formation be transmitted by contact from
cell to cell ? Possibly stentors could be grafted by drastic methods
not involving individual handling. I have tried forcing highly
concentrated coerideus through the finest stainless-steel screens
available. A few fusions were made in this way but not large

360 THE BIOLOGY OF STENTOR

masses as hoped ; for when the openings in the screen are smaller
than the diameters of stentor so that the animals are broken open
when passing through, the wire diameter is then wider than the
pores so that the emerging streams of stentor protoplasm are too
widely displaced to meet and fuse immediately following disrup-
tion of the cells. Yet these remarks may suggest to others more
ingenious approaches increasing the possible techniques with
Stentor which have by no means been exhaustively explored.

CHAPTER XX

EXTENSIONS

Having in hand the already considerable knowledge about Stentor,
we want to inquire into the relevance of these findings for general
problems of the organism. At present we cannot expect that these
bearings will be more than suggestive. For sound construction we
need to proceed step by step as far as we now can go. Since
physiological and biochemical studies of Stentor have only begun,
the discussion will necessarily be slanted in the direction of
epigenetics, or morphogenesis in the widest sense.

I. Stentor and other ciliates

That the performance of Stentor as revealed by experimentation
is not unique may be shown by comparing its behavior with that
of Blepharisma, the only other ciliate on which extensive studies
in experimental morphology are available. Since my investigations
and the remarkable studies of Suzuki (1957) were pursued indepen-
dently, paralleling of results on many points is the more striking.
Suzuki reported within a single publication which deserves to
become a classic a series of experiments encompassing what would
seem to be nearly the whole range of possibilities in Blepharisma.
Although these animals were not grafted as in Stentor, and are
probably too small to permit this approach, by making suitable
incisions in single and dividing animals while keeping the parts
joined by the endoplasm, Suzuki was able to shift these parts with
reference to one another and produce a variety of alterations and
disarrangements closely paralleling several of those which have
been achieved in Stentor. Cutting and enucleation experiments
completed the study. Similarities in the performance of Blepharisma
and Stentor are so numerous that these correspondences might
have been referred to repeatedly throughout our discussions, but
I have chosen to review them together, since this will allow the
special comment which is called for as well as reflecting the
independence of these studies.

361

362 THE BIOLOGY OF STENTOR

Blepharisma is also a heterotrichous ciliate, with bands of (red)
pigment granules lying between the kineties or rows of body cilia.
Otherwise stentors and blepharismas are notably different in
general aspect. Blepharisma is scarcely or not at all contractile, has
no holdfast, possesses a terminal contractile vacuole and cytopyge,
and has an undulating membrane to the right of the mouthparts,
paralleHng the peristome or row of membranellar plates. There is
no obvious gradation in pigment stripe widths around the cell, but
posterior to the mouth lies a ramifying zone where the kineties
bifurcate in multiplying, especially during the earliest stages of
division. Suzuki's drawing indicates that multiplication of clear
stripes begins in the left anterior corner of this V-area, just as in
Stentor. This region is also the site of oral primordium formation.
A groove or rift in the ectoplasm there occurs, and anlagen
development shows only two points of difference: first, the pri-
mordium separates longitudinally to place the undulating mem-
brane on one side of an oral groove and the membranellar band on
the other; and second, the anlage is apparently always parallel to
the lateral striping instead of at first cutting across the stripes. This
is understandable because Blepharisma has no frontal field and
lateral striping therefore does not need to be shifted forward. The
feeding organelles remain deployed longitudinally on one side of
the cell, extending from the anterior pole to about mid-body level
instead of being shifted entirely to the anterior end.

Major points of similarity are as follows. In division, the fission
line is seen as a clear band from which pigment granules are
missing, as in Stentor, and its position is determined late in the
division cycle. Indifferent ectoplasm blocks the division furrow.
Injury apparently causes the resorption of early division primordia,
but mid-stage dividers complete division after excision of the
original feeding organelles ; and if the division furrow is destroyed
a divider is converted into a reorganizer. The macronucleus does
not necessarily split into two equal parts, for in cut dividing cells a
smaller amount is received by smaller than normal daughter cells.
After mid-stage, division is completed even in the absence of the
nucleus or of the division primordium. If only the nucleus is
excised, completion of the opisthe shows that primordia, in what
probably correspond to stage 5 of Stentor, can complete the oral
differentiation. Removal of a substantial part of the membranellar

EXTENSIONS 363

band may incite division, as seems to be the case in stentors.

In reorganization the old mouthparts are resorbed and the new
membranellar bands join with that of the old. Micronuclear
mitoses occur during reorganization and regeneration as well as in
division. There is no evidence for intranuclear differentiation, for
all parts of the macronucleus in Blepharisma as in Stentor were
capable at all times of mediating oral regeneration.

Ablations of other than oral parts does not result in regeneration,
but the more of the peristome removed the sooner regeneration
follows. Primordium formation is therefore inhibited by existing
feeding organelles. Suzuki attributed the possibility of anlage
development during division to a release from inhibition by reason
of partial dedilferentiation of the existing feeding organelles.
Gullet and oral pouch also become vague in dividing stentors, but
in both organisms the blurring itself seems to occur only after the
primordium is well started.

Nucleate anterior fragments of Blepharisma behave like longitu-
dinal aboral halves of Stentor which also lack the original primor-
dium site. Regeneration is usually much delayed yet does occur.
Therefore the normal site is not indispensable and all parts of the
ectoplasm are equipotent as regards anlagen formation.

Rotation of the anterior on the posterior half can lead to the
formation of doublets and doublet animals could be maintained
through a series of divisions lasting a month. Like parts tend to
join, as two cytopyges coalesce or tandem membranellar bands join
together. Induced reorganization occurs on one side of a doublet
when the other side is caused to regenerate, as evident in Suzuki's
figure 38DC. Without grafting, induced resorption of primordia
could not have been demonstrated, but if one type of induction is
possible the other may be likewise, and blepharismas may also pass
through states of activation and inhibition.

Feeding organelles of reversed asymmetry can be produced in
Blepharisma. These forms were apparently the result of the
influence of a posterior pole at the " wrong " end of the primordium,
just as peristomes with mouthparts at both ends were formed when
two posterior ends were present. As in all other cases of feeding
organelles which are a mirror image of the normal, those in
Blepharisma are unable to feed and a self-reproducing biotype
with situs inversus cannot be produced. Evidently the posterior pole

364 THE BIOLOGY OF STENTOR

of the cell induces invagination and mouthparts formation in a
terminus of the primordium which lies in or near it.

Homopolar primordium sites are obviously capable of forming
anlagen though not in their normal position. Heteropolar pieces
tend to creep apart, showing that their polarities are intrinsic and
retained. As in stentors, heteropolar primordium sites may not be
activated to produce anlagen and smaller reversed patches may be
resorbed. When longitudinal halves are rotated upon each other,
an extra primordium may appear at the suture, and this indicates a
possible inductive action between these stripe areas as at loci of
stripe contrast in stentors.

The chief differences in morphogenesis in Stentor and
Blepharisma are now noted. In the latter, Suzuki speaks of an
evident ** growth zone" at the posterior end of the developing
anterior daughter cell which forms a new posterior end during
fission. Such is not obvious in stentors, in which Johnson only
indicated that something like this increase may occur. Possibly
related to the occurrence of this zone is Suzuki's finding that the
oral parts in Blepharisma induce V-areas or primordium sites in
any indifferent region lying posterior to them. Thus, when the cell
was transected and feeding organelles shifted to the side opposite
the original primordium site, a new site then developed posterior
to the displaced organelles and doublets were produced. This does
not occur in Stentor \ for if the head is rotated 180° the anlage
appears only in the original primordium site and a new site is not
generated posterior to a displaced mouth. Also, in Stentor doublets
converting to singles, a primordium site may be obliterated
posterior to one of the mouths, which remains intact.

Doublet animals behave differently in other ways. Removal of
one set of mouthparts did not result in regeneration in Blepharisma.
Apparently one set of organelles can inhibit anlagen formation in
two primordium sites; but in stentors double regeneration-
reorganization always occurred if one site was not subtended by a
set of mouthparts. Blepharisma doublets could not remodel
directly into singles, as stentors do. They achieved this end rather
by exaggerating their doubleness and splitting apart from the
anterior end. Hence these two ciliates exemplify the two types of
transformation to singles defined by Faure-Fremiet (1948b).
Finally, it may be noted that when the anterior end of a pre-

EXTENSIONS 365

divisional Blepharisma was excised, both division and the regenera-
tion of the proter proceeded simultaneously, contrasting with
Stentor in which regeneration of the anterior daughter is always
delayed until after fission is completed.

These differences are minor, though instructive, and should not
be allowed to detract from the demonstration of a remarkable
similarity between the two ciliates. In both genera, micrurgical
studies show how important is the pattern of the ectoplasm for
the course of cytodiiferentiation.

In the manner of elaboration of cytoplasmic structures Stentor
is also not remote from other ciUates (see Klein, 1932; Tartar,
1941b; Faure-Fremiet, 1948b; and Lwoff, 1950). Starting with the
flagellates from which all agree that the ciliates have evolved, the
general picture, as developed by Faure-Fremiet (1954), seems to
be as follows. The centrosome was originally developed to produce
spindle fibers for mitotic division of the nucleus. In flagellates the
centrosome also assumed the new role of producing an external,
fibrous flagellum and its associated organelles. By delegating this
function to other granules (blephoroplasts) derived from the
centrosome and also self-replicating, the number of fiagellae and
complexity of organization could be increased. In ciliates, the
fibrogenic granules lose their morphological association with the
nucleus, increase greatly in number, becoming the semi-autono-
mous kinetosomes which produce short fiagellae (i.e., cilia, with
the same fibrous structure). The transition form, Opalina, shows
uniform ciliation and there is still a certain lack of autonomy in
that the basal bodies all stem from one or two generative kinetics.
But the large population of kinetosomes and their self-reproduction
in ciliates provided the possibility of specialization of the fibers
derived from them as well as for the association of parts into
organelles. The organelles, specifically the mouthparts (and in the
case of Lichnophora^ the pedal disc), in turn become in a manner
themselves self-reproducing in that new mouthparts are developed
at least in close association with the old. But, just as the kineto-
somes become morphologically (yet not physiologically) indepen-
dent of the nucleus, so the oral anlagen became more autonomous
and, in Stentor, originate far from the preexisting mouthparts. A
vestige of the old relationship (as when in Euplotes the new
membranellar band always forms near the posterior end of the old)

366 THE BIOLOGY OF STENTOR

may be evident in Blepharisma in which the present feeding
organelles induce not primordia but sites for primordia. The
eventual evolutionary development therefore provides a cell cortex
with a persisting pattern and polarity (not labile as in flagellates)
as well as semiautonomous units of ectoplasmic structure whose
organization is apparently controlled by that pattern. Possibly the
greatest persistence and fixity of cortical pattern is to be found in
Paramecium, which cannot remodel a defective pattern as stentors
do and rounds out the contour of the cell after an end has been cut
off only by subsequent feeding and structural growth (Tartar,

1954)-

2. Hypotheses concerning morphogenesis of ciliates

A cortical pattern in ciliates is best revealed by silver staining.
By this method Klein showed that the surface of ciliates presents
a network in orderly relation to which are found the ciliary
kinetosomes, oral structures, and other ectoplasmic organelles.
Certain fibers, as others had suggested, are probably concerned
with the coordination of ciliary organelles in swimming, searching,
and feeding behavior. But Klein (1932) also conceived that the
ground network might produce the ciliary and other organelles or
at least guide their organization into specific patterns. Many have
differed with Klein, on the grounds that certain of his networks are
mere sculpturings in a " dead " pellicle and hence are an end result
rather than a possible cause of morphogenesis, as well as that fibers
do not produce kinetosomes but the reverse. Klein's work has
therefore been much neglected because of these differences of
interpretation, though Gelei (1936) made notable contributions in
a similar approach. Yet the idea of some cortical pattern which, as
a continuum, affords the basis for integrating all ectoplasmic
differentiation and, as a geometric scaffolding, provides for their
orderly deployment has endured because it fulfils a logical
requirement.

In Faure-Fremiet's (1950,1954) conception, there is a basic
cortical pattern but it is on a finer level and consists in the orienta-
tion and association of molecules in orderly arrangements. We are
therefore provided a link with the biochemical basis of the
organism, structure being successively compounded on pre-
existing structure until the visible form and differentiation of the

EXTENSIONS 367

ciliate is achieved. Important among these derived structures is
the infracihary network, and Faure-Fremiet (1948a) suggested
that this pattern could be viewed as a morphogenetic field, tending
to recover from distortions and capable of regaining its equilibrium
wholeness following excision of parts. Doublet ciliates are instruc-
tive as manifesting doubleness of fields in balance, but any dis-
equihbrium between the two sides leads in most ciliates to removal
of structural constraint and hence to remodeling of the complex
toward the single form.

According to LwoflF (1950) this cortical network or anisotropic
field would tend to become ''saturated" with kinetosomes in
certain areas where we observe the organelles. If still more
kinetosomes are produced, these would be left free to produce
their own field or be guided into a separate field which would
become a primordium.

Although evoking an ectoplasmic pattern on the basis of their
orderly arrangement, LwoflP stressed the importance of the kineto-
somes in the diflFerentiation of ciliates. This emphasis naturally
stemmed from his classical studies with Chatton on the develop-
mental cycles of apostomatous ciliates (Chatton and LwoflF, 1935a)
indicating a genetic continuity of the kinetosomes and a pluripoten-
tiality with respect to what they produce. That is to say, the
kinetosomes are self-reproducing and new ones arise only by
multiplication of others preexisting; and daughter granules
elaborate body cilia, oral cilia, trichites, or trichocysts, etc.,
depending on how they are determined to develop by their
biochemical surroundings or organizing relations with respect to
the patterned cortex. That nucleated endoplasmic spheres of
stentors entirely bereft of their ectoplasm can regenerate neither
the structured ectoplasm nor the feeding organelles (unpublished)
also suggests a genetic continuity of kinetosomes, although it
cannot be excluded that the morphogenetic failure of these spheres
may be due to the absence of normal outer membranes which
upsets the entire "metabolism" of the cell.

Perhaps the best evidence for division of kinetosomes is to be
found in the work of Hammond (1937) on Euplotes. At the level
of the division furrow the basal bodies of the dorsal bristles were
seen to increase in number and this occurred within a lengthening,
sub-pellicular tubule which would seem to exclude the migration

368 THE BIOLOGY OF STENTOR

of kinetosomes into this area from, say, the nucleus or from any
other source, save that of the preexisting kinetosomes. In oral
primordium formation in ciliates in general, a disorganized
aggregation or " anarchic field " of additional kinetosomes appears
at the site of development. If these in fact arise from multiplication
of adjacent basal bodies of the lateral cilia, this would explain the
origin of the "building blocks" or structural components of
organelles. Yet, as Lwoff said, *' . . . if kinetosomes are necessary
for morphogenesis, they seem not to 'command' but to obey
some mysterious force which is responsible for their orientation".
The alignment and organization of kinetosomes into complex
structures and determination of what type of fibrous elaborations
these granules will produce thus implies an additional agency, a
pattern of "molecular ecologies" or of some preexisting ground
structure in the cortex.

Working with Paramecium, Ehret and Powers (1959) have
challenged previous conceptions regarding the genetic continuity
of kinetosomes and the importance of fibrous networks in organiz-
ing the ciliate cortex. Briefly, they find that the cilia of the oral
primordium arise not from kinetosomes but from different entities
which might be "microsomes"; and they conceive the unit of
cortical structure as a ciliary corpuscle which usually bears one or
double cilia and associated elements, the close packing of these
spherules producing the appearance of hexagonal and rhomboidal
fibrous patterns. This interpretation is contrary to that of Yusa
(1957) and Roque (1956) who retain the postulate of the genetic
continuity of kinetosomes and agreement has not yet been reached,
yet the revolutionary conceptions of Ehret and Powers at least have
the merit of keeping the problems of ciliate morphogenesis in a
fruitfully flocculent state. The crowding of cortical granules,
apparently of internal origin, into every available space in the
ectoplasm of Stentor coeruleus would seem to offer a parallel to the
packing of ciliary corpuscles. But the unextensible, relatively thick
and solid ectoplasm of forms such as Paramecium and Frontonia
may represent a special and peculiar evolutionary development
(Tartar, 1954) and, as these investigators grant, it remains to be
seen how far their intriguing ideas are applicable to other ciliates.
The orderly packing and morphogenetic control of corpuscular
units, even in Paramecium, would seem to require, as with

EXTENSIONS 369

kinetosomal orientation, some pattern factor in addition.

Specifically in reference to his studies on Stentor (Weisz, 1951c,
1954) developed a theory of morphogenesis in ciliates involving
three postulates : first, that self- reproducing kinetosomes represent
a hierarchy, with oral granules dominant over those of a stomato-
genic kinety or primordium site and these in turn dominant over
those of other kinetics of the more generalized lateral body surface ;
second, that this hierarchy represents the degree of ascendency in
competition of the kinetosomes for their "food" or the special
metabolites supplied by the endoplasm from biochemical activities
supported by the macronucleus which they require both to grow
and to maintain themselves; and third, that the kinetosomes in
turn act back, enzymatically, on the part of the macronucleus
nearest them. How this system was thought to operate may be
illustrated by the case of regeneration. When the feeding organelles
are excised, the upper level of the kinetosomal hierarchy is vacated
so that the metabolites can flow to the kinetosomes of the next
level — those of the stomatogenic kinety or, if this was also
removed, then the next adjacent body kinety — which are then
able to multiply and produce oral cilia for the anlage of a new set
of feeding organelles. Once formed, this new set again exhausts the
special metabolites for oral cilia so that further primordium
formation is inhibited. (In division and reorganization, the oral
kinetosomes somehow lose their competitive ascendancy so that
the kinetosomes of the stomatogenic kinety are no longer held in
check.) The oral kinetosomes in place now react with adjacent
nodes of the nucleus, maintaining their capacity to produce the
special metabolites, while those far distant lose their capability,
and morphostasis is hence stabilized.

The effectiveness of this hypothesis depends upon two points :
first, that the postulated metabolites are necessary for the main-
tenance of existing feeding organelles, and second, that these
metabolites are present only in limited quantity. Only on the basis
of these assumptions would there occur that competition which
would explain the integration of the organism, e.g., that it never
has more than one set of feeding organelles. Yet I do not think
that either of these points have been substantiated by subsequent
studies. The formed feeding organelles and body cilia of enucleates
are often maintained to the point of death or at least they continue

370 THE BIOLOGY OF STENTOR

intact for a whole week and undergo dissolution only just before
the death of the specimen at which time the appearance of a general
necrosis could just as well account for structural disintegration. I
have observed nothing like an intimate nutritive relationship
between the nucleus and the feeding organelles such that removal
of the nucleus withdraws their maintenance and results in the
prompt resorption of specialized organelles. In fact, the failure of
excised heads to reduce the feeding organelles to a size propor-
tionate to the small fragment, if enucleate, indicates just the
opposite : that the nucleus is necessary for the resorption of formed
organelles. Nor is there any substantiation that the hypothetical
metabolites are present in limited quantity. Arguing against this
assumption is the fact that grafted pairs of stentor produce one,
two, or three primordia and sets of feeding organelles regardless
of the amount of nuclear material present; and grafting of an
enucleated stentor to a normal animal may lead to the production
of a doublet just as readily as when both nuclei are present
(Tartar, 1954). Similarly, if the fine-line zone of a stentor is split
by an enucleated meridional patch of wide striping, three anlagen
of normal size are usually formed instead of one (Tartar, 1956a).
On Weisz's hypothesis this would imply that the single animal is
quite able to produce three "quanta" of oral metabolites. If so,
there is no reason to suppose that an intact set of feeding organelles
would monopolize them and in this way exert inhibitive action on
the primordium site. The additional postulate — that the kineto-
somes act back on the nucleus to produce internodal differences —
was also not confirmed; for in later tests all sections of the
macronucleus were found to be equivalent (Tartar, 1957b).

Form in ciliates is still without satisfactory causal analysis; but
this is no wonder since no adequate theory of morphogenesis of
any organism has yet been achieved. In this perspective, the
progress with ciliate protozoa appears promising and we may ask
how, if eventually successful, a verifiable explanation of their
development might be pertinent to general problems of
cytodilferentiation.

3. Stentors and cells

First we shall consider whether a ciliate like Stentor is a cell, or
properly should be included in the class of those things called cells,

EXTENSIONS 371

and therefore whether Stentor studies are relevant to analysis of
the potentialities of cells in general. My opinion is already evident
from the fact that stentors have throughout this study been
referred to as cells. This follows if we define the cell as a packaged
nucleo-cytoplasmic duality capable in some degree of independent
life. We can allow that these ''packages" sometimes may have
''holes" in them connecting to other packages (cell bridges), and
that the enclosed nuclear phase may consist of one or more nuclei
of one or two types. These units may be wholly free-living, like
Stentor. They may be autonomous but not free-living as in the
case of parasitic protozoa. Or as tissue cells they may be subject to
a system of intercellular reactions leading to the morphological
and functional wholeness of cellular organs and organisms. Even
in the latter case, the cell lives a double life, both dependent and
independent, as one of the authors of the Cell Theory, Schleiden,
remarked. A tissue cell is dependent on the organism for its
sustenance and participates in multicellular interaction and
organizing relations, yet its capacity for independent life is
abundantly demonstrated by culture outside the organism; just as
its self-dependence is shown by the fact that if the long process of
a nerve cell is separated from its nucleated cell body, neither
proximity to countless nucleated fellow cells nor being continually
bathed in blood plasma can save that nerve from disintegration
after its nucleo-cytoplasmic integrity has been violated. Indeed,
the study of somatic deletions in Drosophila has shown that the
absence of a single genetic locus, which may be tagged by its
correlation with yellow color, results in the independent death of
the cells which lack it, as if all the nucleus is needed all the time
just to maintain the hfe of the cell itself (Demerec, 1934).

That larger organisms are comprised of multitudes of cells
would seem to imply the interaction between nucleus and cyto-
plasm is so intimate that no portion of the cytoplasm can be far
from an associated nucleus. The nuclear phase is not aggregated
into one "gland". Even in the neurone, in which the cytoplasm
may extend several feet from the major cell body with its nucleus,
specialized organelles — the neurofibrils — may have been
developed to allow nucleo-cytoplasmic interactions even at this
distance (Parker, 1929).

All grades are found between complete independence of cells,

372 THE BIOLOGY OF STENTOR

as in free-living protozoa, and a total dependence which might
best be exemplified by the anucleate mammalian red blood cell —
whose fellow traveler in the blood stream, the leucocyte, is
relatively autonomous and practically indistinguishable from para-
sitic amoebas. In the cellular slime molds free-living amcebas
cooperate in forming multicellular fruiting bodies. What remains
constant throughout is self-dependence of the cell as a packaged
nuclear-cytoplasmic duality capable of some degree of independent
life.

The further similarity between unicellular organisms and tissue
cells is found in the fact that the genome of protozoa is evidently
just as complex as that of metazoa and their tissue and germ cells.
Higher organisms do not have larger or longer or more numerous
chromosomes and hence, evidently, have not a correspondingly
greater number of genes, nor is the behavior of their nuclei more
complex. In present-day terms, this implies that the protozoan
nucleus contains as much "information" as the egg or tissue cell
(Elsasser, 1958). This uniformity suggests that the nucleus is
concerned first of all with the life of the individual cell, and that
in multicellular forms there is developed among the cells another
system of intercellular reactions, about which we still know
practically nothing, which provide the information for multi-
cellular differentiation. Evolution, with its teaching of the con-
tinuity of life, leads us to regard free-living and tissue cells as
basically the same, multicellular organisms arising either by the
adherence of products of cell division, as in the algae, or by a
partitioning of a multinucleated cell into a multicellular body as
seen in the Accela or in insects. A transcending unity of all cells,
not as parts but as expressing what Woodger (1929) called the cell
type of organization, certainly provides the most hopeful heuristic
principle. This does not exclude the possibility that protozoa have
taken this type of organization to extremes of multiple specializa-
tion of the cytoplasmic phase, or that we can learn as much from
them by contrast as by comparison with other cells.

4. Stentor and metazoa

Stentors elaborate themselves in only one direction, to form
another stentor. In this regard they are like eggs but lack the
multiple potentialities of embryonic cells. Repeated cleavage

EXTENSIONS 373

without feeding but also without intercellular differentiation occurs
in some ciliates, as when the large form of Ichthyophthirius pro-
duces multitudes of small forms (Mugard, 1948). The beginnings
of cellular differentiations may be seen in the formation of mating
types or the gametic differentiation of some ciliates, as well as in
multicellular stages of certain Sporozoa, the Cnidosporidea. In
clonal cell cultures, metazoa are being, as it were, reduced to
*' protozoa ". And in Chcetopterus, Lillie( 1906) was able to suppress
cleavage of the egg and yet obtained unicellular embryos of fairly
normal shape in which the elaboration of ciUa, with a particulate
contribution from the nucleus, and imprecise segregation processes
led to a fairly recognizable early embryo. Being cytoplasmic con-
tinuums, Stentor masses are not multicellular though they do show
the emergence of new capacities for morphogenesis.

Another major point of correspondence lies in the cilia and
ciliation. We now know that the basic structure of cilia and
flagella are the same in all organisms. Many animals have ciliated
epithelia and in these the joining of the cilia by fibrous connectives
does not differ fundamentally from that in ciliates. Gruber com-
mented on the remarkably similar construction of the membranelles
of Stentor and those occurring in the *' corner cells" of certain
molluscs, notably Cyclas cornea. Whitman (1893) used this corres-
pondence in argument for the inadequacy of the cell theory of
development, having to emphasize at that time the neglected and
still baffling intercellular organizing relations through which the
separate cells, regardless of their size or number, are formed into
an embryo. Stentor makes many such membranelles in one cell;
a mollusc, one in each of many cells. In their embryonic stages
many multicellular organisms are conspicuously ciliated, offering
the possibility that something like ciliate morphogenesis may play
a significant role in their early development. In the shipworm a
silver-staining material is segregated into specific blastomeres
(Faure-Fremiet and Mugard, 1948); and in the sea urchin certain
cells come to show an argentophile network with apparently a
centrosome-kinetosome in each cell which becomes part of a
ciliated structure (Mugard, 1957), but there is still no proof that
these are causal factors in development.

Embryologists are generally agreed that development implies
an initial cytoarchitecture in the cortex of the egg as a guide for

374 THE BIOLOGY OF STENTOR

orderly transformations. Stentor has such an architecture or
cortical pattern which is even visible in its heterogeneity in the
living organism and therefore can be rearranged at will. In dividing
stentors the migrations of the carbohydrate reserves mimics the
segregation of distinctive ooplasms in certain eggs, while in the
^gg coat which Holtfreter (1949) has shown to be so important in
embryogenesis we may have a direct descendant of the ciliate
pellicle. We are reminded, too, that Hthium has marked morpho-
genetic effects on stentors as it does on embryos. Truly, we do not
know which of these resemblances are superficial and which are
fundamental, but no possible correspondencies should be ignored.

5. Theoretical considerations

Before the nucleus was discovered and even after this cell
organelle was found to be present but not obviously active except
in reproducing itself at cell division, the emphasis w^as on the
cytoplasm as the basis of life. All cytoplasms were said to have a
common denominator in *' protoplasm", a semifluid substance
conceived as *' living matter". Of this view there remains today
only the fact that living organisms are intimately involved with the
colloidal state, and the hope that all living phenomena will be
explainable in terms of molecules and their interactions. With the
discovery of the nucleus and its importance in inheritance the
emphasis shifted in the other direction, and the nucleus was
regarded as ''the heart of the cell", or, currently, as ''containing
all the information for the organism". Yet both cytoplasm and
nucleus are necessary as a natural and inescapable dualism pre-
sented by the cell. Of course these two parts of the cell interact,
and Verworn (1892) early conceived a scheme embracing possible
interactions, excepting the more sophisticated modern concept of
steady states. Simply stated, we want to know what the nucleus
does and how it does it, what the cytoplasm does and how this is
accomplished, as well as how the two phases cooperate in the life
of the cell.

The nucleus seems to serve as a chemical factory for the cyto-
plasm, producing essential substances or the means of their pro-
duction, apparently coenzymes. Apart from itself growing and
replicating, the nucleus contributes substances into the cytoplasm
where reactions leading to metabolism and structural growth take

EXTENSIONS 375

place. The cytoplasm would have at any moment a store of these
substances, exhaustible if the nucleus is removed. The nucleus has
therefore been regarded as the source of the regeneration of enzyme
systems present and acting in the cytoplasm (Mazia, 1952).

Stentors emphasize the importance of this trophic function of
the nucleus. The micronuclei are significant only in genetic recom-
bination during conjugation. It is the macronucleus which supports
the Kfe of the cell, though this involves the expression of specific
genetic determinants derived from its progenitor, the micro-
nucleus, as was demonstrated for Paramecium by Sonneborn (1947).
In metazoa the metabolic function of the nucleus is cryptic and
not obvious, or is revealed only by special demonstrations as in
neurone regeneration or the somatic deletion studies of Demerec.
But in protozoa, as in studies of microbial genetics and the new
work on Neurospora^ the trophic role of the nucleus is apparent.

In stentors, the macronucleus is clearly necessary for both
digestion and synthesis which leads to growth. Therefore it should
make possible the formation of enzymes and may be a source of
RNA, via nucleolar extrusion, for protein synthesis. This nucleus
also probably sustains respiration, for though energy metabolism
long continues in enucleates it gradually diminishes. Several
instances have been cited which show that the quantitative relation-
ship between nucleus and cytoplasm is important for these
physiological or biochemical processes; and it is in the relatively
simple alterability of the nucleo-cytoplasmic ratio that stentors
should prove most fruitful in studies of cell physiology.

Some portion of the macronucleus is essential for cytodifferen-
tiation in oral regeneration. Presumably this support is either a
synthesis or a mobilization of structural proteins; but in spite of
the sameness of stentor organelles this action is fraught with
specificities, for the nucleus of one species of Stentor can rarely be
exchanged for another to yield an effective nucleo-cytoplasmic
combination. And where and how the building blocks are put
together in formed organelles is very probably the work of the
ectoplasm and its pre-formed structure.

This view was previsioned by Prowazek in the course of his
investigations on Stentor and has been enlarged upon by others up
to the present day (Tartar, 1941b; Sonneborn, 195 1; Ephrussi,
1953 ; Weisz, 1954; Danielli, 1958). Noting that the nucleus is not

376 THE BIOLOGY OF STENTOR

bound to the cytoplasm by any intimate structure and requires no
specific geometric relationship with the cytoplasm for its eflfective
action, Prowazek (1904) inferred that there is only a substance
relation between the two phases of the cell which could well be
explained in chemical terms. Neither can the form of the cell, nor
that of the organism in the case of protozoa, be related to the
flowing endoplasm which indeed has been shown in Condylostoma
(Tartar, 1941b) and in the regeneration of stentor "skins'*
(unpublished) to be dispensable. Therefore nucleus and cytoplasm
presumably interact by chemical contributions to each other, and
guidance of the elaboration of formed parts in the cytoplasm is to
be sought in neither a nucleus of unprescribed location nor a flowing
endoplasm but in the most solid portion of cell, namely, the
ectoplasm (Prowazek, 19 13).

Events in the ectoplasm of Stentor also clearly exert control
over the behavior of the nuclei. Micronuclei divide whenever an
oral primordium is formed, even when there is no cytosomal
fission. Macronuclei coalesce, re-nodulate, or are divided according
to the phase of the cytoplasm in which they find themselves.
Compensatory increase in macronuclear volume occurs only if
primordia are formed. And the disposition of the macronuclear
nodes is evidently determined in large part by the pattern of
ectoplasmic striping.

The cytoplasm may have its own replicating units, including
kinetosomes as important elements of ectoplasmic structure. Self-
reproducing entities are of limited explanatory value because they
refer us back to another of the same type of entity which we seek
to analyze and explain. Yet to establish a genetic continuity of
cytoplasmic parts would tell us whence new entities arise, i.e.,
from preexisting ones, and this is a great aid in narrowing the field
of inquiry. Moreover, replicating units and their elaborations alone
would not make an organism but merely an assemblage. '' Is
anyone willing to believe", asks Sonneborn (1951), ''that, if all
such self-duplicating components of the cell were thrown together
in a test tube in the proper portions with adequate food for their
multiplication, a Chilomonas cell or any cell at all would result? "
Something more seems to be needed and this is a cortical pattern
factor itself having genetic continuity.

In Stentor, development is organized because there is always a

EXTENSIONS 377

previous pattern. A cortical pattern of polarized, anisotropic
lateral striping is always visibly present no matter how we fragment
the cell. The location, number, length and direction of asymmetry
of the oral primordium are determined by the pattern of the ecto-
plasm, quite independently of the location or quantity of the nucleus
and endoplasm. Induction of mouthparts at the terminus of the
anlage is also to be correlated with the polar ectoplasm, for in
heteropolar implants two sets of mouthparts are produced in
correspondence with two separate posterior poles although the
endoplasms would mingle indiscriminately and the nuclear nodes
may be located anywhere. Even the smallest, nucleated pieces bear
short lengths of a few lateral ciHary bands which are separated at
graded distances by granular stripes so that healing produces a
locus of stripe-width contrast and regeneration of a whole is always
guided by this preexisting polarity and anisotropy. Even in the
relatively dedifferentiated cyst stage, striping and polarity are
evident, according to the drawings of Stein; and in other ciUates
which have been carefully studied all landmarks do not disappear
and some cortical pattern evidently persists (e.g., Garnjobst, 1937).
Empirically, Stentor possesses at all times — in all our experi-
mental situations save one: denudation of the ectoplasm — an
obviously persistent visible pattern. However the basic cortical
pattern may be conceived, to postulate it as axiomatic or as a factor
always present from the initial state in any experiment we can
perform, has an important consequence. This is that we need not
and cannot derive the final form and pattern from molecules and
their spontaneous aggregation or interactions as such. True, these
cortical cell patterns would have arisen at some time in the early
evolution of life, and phenomena such as the association of collogen
into fibrous sheets may give us some hint of how this could have
been brought about. Once developed, the relatively large scale
patterns we have in mind, like such replicating units as macro-
molecules, chromosomes, and perhaps kinetosomes, could have
been carefully conserved and never destroyed but passed on by
genetic continuity, pattern producing further pattern, and these
patterns gradually evolving in complexity, on the one hand into
very complicated unicellular flagellates and ciliates and on the
other into equally complex yet cryptic egg patterns capable of
guiding the development of the whole range of elaborate multi-

378 THE BIOLOGY OF STENTOR

cellular organisms. Therefore, in any experiment we perform today
we do not have to demonstrate the origin of life by reducing the
results to molecules and their interactions, for pattern is always
there to start with. This pattern factor may have undreamed of
capabilities, itself constituting a primary determinate of what kind
of molecules are synthesized in association with it and how they
behave, for example, in contributing to growth or increase in that
pattern.

These patterns would be in one sense "of molecules"; in pre-
cisely this sense, that the organism is obviously reducible to a
collection of identifiable molecules after chemical treatment and
destruction in a test tube. This would explain why organisms in
their functions and even in their forms are very definitely and
sometimes grossly affected by the presence of certain types of
molecules. Even a single ion like lithium exerts great influence on
morphogenesis in both embryos and stentors. For if the patterns
are ''made up of molecules" the kinds of molecules and ions
available would clearly have a substantial effect on these patterns.
The patterns themselves could have just as much influence in the
formation and behavior of the molecules. There is some evidence,
for instance, that the molecules we analyze in the test tube do not
exist as such in the organism (see Needham, 1933). Picric acid is
said to precipitate proteins in solution but not when injected into
amoebas, and sea urchin eggs do not show the characteristic
ultraviolet absorption spectrum of proteins until they are killed.

From these considerations it follows that pattern and substance
are two irreducible aspects of the organism. They may be related
in the sense of Neils Bohr's principle of complementarity, as he
has himself suggested (1958).

It would appear, therefore, that our greatest lack and most
fruitful opportunity in biology lies in conceiving and testing the
nature and capabilities of persistent supramolecular patterns. For
this task stentors should be highly appropriate because they
present us with a visible cortical geometry as an empirical reality,
and stentors as the most operable of all cells have already shown
how important this pattern is in determining form and cyto-
differentiation.

Any indication, however general, of the possible nature of basic
cortical patterns in ciliates and in eggs should help in transcending

EXTENSIONS 379

this most important gap in our understanding of organisms.
Proceeding from the Stetitor studies we may suggest that one
characteristic of the cortical pattern is that it is beyond indivi-
duahty but bears intrinsically the tendency or capacity to integrate
as one or more than one individuality. In terms of a model, the
pattern might therefore be conceived as a network which is a type
of repeat pattern, somehow capable of deriving a wholeness while
maintaining its repeat character and potentialities beyond indivi-
duality. Thus fragments of a stentor or an egg when physically
isolated can themselves become wholes, so that the original
individuality is seen to have contained multiple nascent indivi-
dualities. When a stentor or cell normally divides the original
integrated pattern is obviously converted into two. Conversely, two
or more whole eggs or stentors can be fused together to produce
only a single individuahty from the several original ones. There-
fore, whatever is operating and determinant in these experimental
situations is something which is beyond individuality but tends
to individuate. It is this characteristic which has made pattern
unmechanical and so difficult to pin down, even in the case of
mosaic eggs; for in experiments the pattern reintegrates after
disturbances and deletions so that there is no point to point
correspondence on which to base analysis until the work of deter-
mination has already been accomplished. If the pattern factor is
beyond individuality, an important consequence is that wholeness
is not, as many have maintained, an irreducible, axiomatic presup-
position about any organism but rather a result or an achievement,
as McDougall (1938) has well stated. At the moment when a
fragment of an egg or a ciliate is cut there is no wholeness except
in the sense of an object which has been physically isolated, nor is
there a wholeness at the moment when two organisms are fused.
Instead, there are molecules, replicating units, and above all a
pattern factor which is beyond individuality through which a
wholeness is later achieved. In learning how this may come about,
Stentor may be an invaluable guide. Though Stentor is a single cell
and can presumably teach us nothing x)f the complex intercellular
relations which form the multicellular organism, it may well be
that comprehending the organization of a unicellular animal is a
fruitful if not essential step towards evolving a satisfactory theory
of more complex developments.

BIBLIOGRAPHY OF STENTOR

Adolph, E. F. (1931): The Regulation of Size as Illustrated in Unicellular

Organisms. Thomas, Springfield, Illinois. (Review of size in relation to

regeneration.)
Allescher, Marie (19 12): tJber den Einfluss der Gestalt des Kernes auf

die Grossenabnahme hungernder Infusorien. Arch. Protistenk. 27,

129-171.
Alverdes, F. (1922): Zur Lokalisation des chemischen und thermischen

Sinnes bei Paramoecium und Stentor. Zool. Anz. 55, 19-21.
Andrews, E. A. (1945): Stentor's anchoring organs, jf. Morph. 77,

219-232.
Andrews, E. A. (1946): Ingestion organs in Folliculinids and in Stentors.

y. Morph. 79, 419-444.
Andrews, E. A. (1948a): Folliculinids and Stentors in British Columbia.

Trans. Am. Micros. Soc. 67, 61-65.
Andrews, E. A. (1948b): Surface parts of the contractile vesicle of

Stentor coeruleus. J. Morph. 82, 257-268.
Balamuth, W. (1940): Regeneration in protozoa: a problem of morpho-
genesis. Quart. Rev. Biol. 15, 290-337.
Balbiani, E. G. (1861): Recherches sur les phenomenes sexuelles des

Infusoires. J. de Physiol. t.IV. (Not seen).
Balbiani, E. G. (1882): Les Protozoires. Le9ons faites au College de

France. Jour, de Micrographie t.VI, p. 474 (or 1881? Not seen).
Balbiani, E. G. (1889): Recherches experimentales sur la merotomie des

infusoires cilies. Contribution a I'etude du role du noyau cellulaire.

Recueil Zoologique Suisse 5, 1-72.
Balbiani, E. G. (1891a): Sur les regenerations successives du peristome

comme caractere d'age chez les Stentors et sur le role du noyau dans ce

phenomene. Zool. Anz. 14, 312-316; 323-327.
Balbiani, E. G. (1891b): Sur la formation des monstres doubles chez les

Infusoires. Jf. de I'Anat u. de la Physiol. 27, 169-196.
Balbiani, E. G. (1891C-2): Nouvelles recherches experimentales sur la

merotomie des infusoires cilies. Annales de micrographie 4, 369-407;

449-489.
Balbiani, E. G. (1893): Nouvelles recherches experimentales sur la

merotomie des infusoires cilies. Ibid. 5, 1-25; 49-84; 113-134.
Barbier, M., E. Faure-Fremiet, and E. Lederer (1956): Sur les pig-
ments du cilie Stentor niger. C. R. Acad. Sci., Paris 242, 2 182-2 184.
Bary, B. M. (1950): Four new species of fresh-water ciliates from New

Zealand. Zoology publications from Victoria Univ. College (N.Z.),

No. 2.

" 380

BIBLIOGRAPHY OF STENTOR 381

Belda, W. H. and W. J. Bowen (1940): A tested method of growing

Stentor coeruleiis. Science 92, 206.
Bishop, Ann (1927): The cytoplasmic structure of Spirostomum

ambiguum, Ehrenberg. Quart. Jf. Micros. Sci. 71, 147-172. (Containing

review of fiber systems in Stentor.)
Bracket, J. (1957): Biochemical Cytology, Academic Press, New York.
Brauer, a. (1885): Bursaria truncatella unter Beriicksichtigung anderer

Heterotrichen und der Vorticellen. Jena. Z. Naturwiss. 19, 489-519.

(With notes on stripe multiplication in Stentor.)
BuLLiNGTON, W. E. (1925): A study of spiral movement in the ciliate

infusoria. Arch. Protistenk., 50, 219-274.
BuRNSiDE, L. H. (1929): Relation of body size to nuclear size in Stentor

coeruleus. jf. Exptl. Zool. 54, 473-483.
Causin, M. (193 i): La regeneration du Stentor coeruleus. Arch. d'Anat.

Micros. 27, 107-125.
Chambers, R. and C.-Y. Kao (1952): The effect of electrolytes on the

physical state of the nerve axon of the squid and of Stentor, a protozoon.

Exptl. Cell Research 3, 564-573.
Child, C. M. (1914): The axial gradient in ciliate Infusoria. Biol. Bull.

26, 36-54.
Child, C. M. (1949): A further study of indicator patterns in ciliate

protozoa. J'. Exptl. Zool., in, 315-347.
Claparede, E. and J. Lachmann (1857): Note sur la reproduction des

Infusoires. Ann. Sci. nat., Zool., 4, 221-244. (Not seen).
Claparede, E. and J. Lachmann (1858-61): Etudes sur les Infusoires et

les Rhizopodes. Geneve, 2 vols. Mem. Inst. nat. genev., 5; 6; 7, 1-291.

(Not seen).
Cox, J. D. (1876): Multiplication by fission of Stentor miilleri. Am.

Naturalist 10, 275-278. (Incidental).
Dabrowska, J. (1956): Tresura Paramecium caudatum, Stentor coeruleus,

Spirostomum ambiguum na budzce swietne. Folia Biologica, Polska

Akademia Nauk. 4, 77-91. (With English summary).
Daniel, J. F. (1909): Adaptation and immunity of lower organisms to

ethyl alcohol. J. Exptl. Zool. 6, 571-61 1.
Davenport, C. B. and H. V. Neal (1896). Studies in morphogenesis, V.

On the acclimatization of organisms to poisonous chemical substances.

Archiv.f. Entw.-mech. 2, 564-583.
Daw^son, J. A. (1953): The culture of Blepharisma undulans and Stentor

coeruleus. Bio. Rev. College City of Nezv York, 15, 13-15. (Not seen).
De Terra, Noel (1959): Personal communication.
De Terra, Noel (i960): Studies of nucleo-cytoplasmic interactions and

P^2 uptake during cell division in Stentor coendeus. Exptl. Cell Research.

(In press.)
Dierks, K. (1926a): Untersuchungen iiber die Morphologic und

Physiologic des Stentor coeruleus mit besonderer Briicksichtigung

seiner kontraktilen und konduktilen Elemente. Arch. Protistenk., 54,

1-9 1.

382 THE BIOLOGY OF STENTOR

DiERKS, K. (1926b): Lahmungsversuche an Stentor coeruleus durch

Kaliumionen. Zool. Anz. 67, 207-218.
Faure-Fremiet, E. (1906): Sur un cas de monstruosite chez Stentor

coeruleus. Arch. d'Anat. Micros., 8, 660-666.
Faure-Fremiet, E. (1936): Condylostoma (Stentor) auriculatus (Gruher).

Bull, de la Soc. Zool. de France 61, 51 1-5 19.
Faure-Fremiet, E. and E. Lederer (1956): Microscopie electronique de

quelque cilies. Bidl. soc. zool. France 81, 9-11.
Faure-Fremiet, E. and C. Rouiller (1955): Microscopie electronique

des structures ectoplasmiques chez les cilies du genre Stentor. C.R.

Acad. Sci., Paris 241, 678-680.
Faure-Fremiet, E., C. Rouiller, and M. Gauchery (1956): Les struc-
tures myoides chez les cilies. Etude au microscope electronique. Arch.

d'Anat. micros. Morph. exp. 45, 1 39-161.
Gelei, J. V. (1925): tjber der Kannibalismus bei Stentoren. Arch.

Protistenk. 52, 404-417.
Gelei, J. v. (1926): Sind die Neurophane von neresheimer neuroide

Elemente? Arch. Protistenk. 54, 232-242.
Gelei, J. v. (1927): Angaben zur der Symbiosefrage von Chlorella.

Biol. Zentralbl. 47, 449-461. (Including some remarks on Stentor.)
Gelfan, S. (1927): The electrical conductivity of protoplasm and a new

method of its determination. Univ. Calif. Publ. Zool. 29, 453-456.
Gerstein, J. (1937): The culture and division rate of Stentor coeruleus.

Proc. Soc. Exptl. Biol. Med. 37, 2 10-2 11.
Greeley, A. W. (1901): On the analogy between the effects of loss of

water in lowering of the temperature. Am. J. Physiol. 6, (1901-2),

122-128.
Gruber, a. (1878): Die Haftorgane der Stentoren. Zool. Anz. i, 390-391.
Gruber, a. (1883): Ueber die Einflusslosigkeit des Kerns auf die

Bewegung, die Ernahrung und das Wachstum einzelliger Tiere.

Biol. Zentralbl. 3, 580-582.
Gruber, A. (1885a): Ueber kiinstliche Teilung bei Infusorien. Ibid. 4,

717-722.
Gruber, A. (1885b): Ueber kunstliche Teilung bei Infusorien (II).

Ibid. 5, 1 37-141.
Gruber, A. (1886): Beitrage zur Kenntniss der Physiologic und Biologic

der Protozoen. Ber. naturf. Ges. Freiburg, i.B. i, 33-56.
GuTTES, E. and Sophie Guttes (1959): Regulations of mitosis in Stentor

coeruleus. Science 129, 1483.
Hamburger, Clara (1908): Zur Kenntnis der Conjugation von Stentor

coeruleus nebst einigen allgemeinen Bemerkungen liber die Con-
jugation der Infusorien. Z. wiss. Zool. 90, 421-433.
Hammerling, J. (1946): tJber die Symbiose von Stentor polymorphus.

Biol. Zentralbl. 65, 52-61.
Hartmann, M. (1922): tJber den dauernden Ersatz der ungeschlect-

lichen Fortpflantzung durch fortgesetzte Regenerationen. Ibid. 42,

364-381.

BIBLIOGRAPHY OF STENTOR 383

Hausmann, Gertrud (1927): Uber die Bewegungen einigen ciliaten

Protozoen im Wechselstrom. Biol. Generalis 3, 463-474.
Haye, a. (1930): iJber den Extretionsapparat bei den Protisten, nebst

Bemerkungen liber einige andere feinere Strukturverhaltnisse der

untersuchten Arten. Arch. Protistenk. 70, 1-86.
Hegner, R. W. (1926): The interrelations of protozoa and the utricles of

Utricularia. Biol. Bull. 50, 239-270.
Heilbrunn, L. V. (1928): The colloid chemistry of protoplasm.

Protoplasm-Monographien Vol. i, Berlin.
Heilbrunn, L. V. (1943 or 1952): An Outline of General Physiology,

either 2nd or 3rd ed. Saunders, Philadelphia.
Hetherington, a. (1932a): The constant culture of Stentor coeruleus.

Arch. Protiste?ik. 76, 1 18-129.
Hetherington, A. (1932b): On the absence of physiological regeneration

in Stentor coeruleus. Ibid. 77, 58-63.
Hofer, B. (1890): Ueber die lahmende Wirkung des Hydroxylaniines

auf die contraction Elemente. Z. zuiss. Mikr. 7, 318-326 {S. coeruleus

included).
Holt, E. B. and F. S. Lee (1901): The theory of phototactic response.

Am.y. Physiol. 4, 460-481.
Howland, Ruth B. (1928): A note on Astasia captive Beauchamp.

Science, 68, 37.
Hyman, Libbie H. (1925, 193 1): Methods of securing and cultivating

protozoa. Trans. Am. Micros. Soc. 44, 216-221; 50, 50-57.
IsHiKAWA, H. (1912): Wundheilungs- und Regenerationsvorgange bei

Infusorien. Arch. Entw.-mech. 35, 1-29.
Ivanic, M. (1926): Zur Auffassung der Kernverhaltnisse bei Stentor

coeruleus und Stentor polymorphus, nebst Bemerkung iiber einige

Kernverhaltnisse bei Infusorien im allgemeinen. Zool. Anz. 66, 55-61.

(Dubious).
Ivanic, M. (1927): Uber der Kannibahsmus bei Amoeba verrucosa

(Ehrb.), nebst Bemerkung uber den Kannibalismus bei Protozoen im

allgemeinen. Ibid. 74, 313-321. (Dubious).
Jennings, H. S. (1899): Studies on reactions to stimuli in unicellular

organisms. III. Reactions to localized stimuli in Spirostomiun and

Stentor. Am. Naturalist 33, 373-389.
Jennings, H. S. (1902): Studies on reactions to stimuli in unicellular

organisms. IX. On the behavior of fixed Infusoria {Stentor and

Vorticella), with special reference to the modifiability of Protozoon

reactions. Am. J. Physiol. 8, (1902-3), 23-60.
Jennings, H. S. and C. Jamieson (190^2): Studies on reactions to stimuli

in unicellular organisms. X. The movements and reactions of pieces

of ciliate infusoria. Biol. Bull. 3, 225-234.
Jennings, H. S. and E. M. Moore (1902): Studies on reactions to

stimuli in unicellular organisms. VIII. On the reactions of Infusoria

to carbonic and other acids, with especial reference to the causes of the

gatherings spontaneously formed. Am. J. Physiol. 6, (1901-2), 233-350.

384 THE BIOLOGY OF STENTOR

Johnson, H. P. (1893): A contribution to the morphology and biology

of the Stentors. J. Morph. 8, 467-562.
Kahl, a. (1935): Wimperthiere oder Ciliata. In Die Tierwelt Deutsch-

landsy 25th Teil, Part 3, *' Spirotricha ", 457-466. Jena.
Kalmus, H. (1928): liber den Bodenfauna der Moldau im Gebiete von

Prag. Ein Jahreszyklus. II. Protozoa, etc. Mit einem Anhang:

OkologischeBeobachtungenundVersuche./wferwa^ Rev. Hydrobiol. 19,

349-429.
Kent, W. S. (188 1-2): A Manual of the Infusoria. London, Vol. II,

pp. 588-596.
Kessler, G. (1882): Ein Beitrag zur Lehre von der Symbiose. Arch.f.

Anat. u. Physiol. 1882, 490-492. (Not seen.)
Kimball, R. F. (1958) : Experiments with Stentor coeruleus on the nature

of the radiation-induced delay in fission in the ciliates. J. Protozool. 5,

151-155-
KiRBY, H., Jr. (1941a): Relationships between certain protozoa and

other animals. In Protozoa in Biological Research (edited by Calkins

and Summers) Columbia University Press, pp. 890-1008.
KiRBY, H., Jr. (1941b): Organisms living on and in protozoa. Ibid.,

1009-1113.
KiRBY, H., Jr. (1956): In: Opinion 418. Opin. Internat. Comm. ZooL

Nom. 14, 46-68.
Lankester, E. R. (1873): Blue stentorin, the coloring matter of Stentor

coeruleus. Quart. Jf. Micros. Sci. 13, 139-142.
Lieberkuhn, N. (1857): (Muskelfasern in Stentor, etc. — fibrillar and

contractile structures in protozoa.) Arch. Anat. u. Physiol. 3, 20.

(Not seen.)
LiLLiE, F. R. (1896): On the smallest parts of stentor capable of re-
generation ; a contribution on the limits of divisibility of living matter.

jf. Morph. 12, 239-249.
Madlen, J. (1946): (The significance and occurrence of micro-organisms

in forest soils.) Lesnicka Prace 25 (1/2), 20-31. (S. polymorphus in-
correctly listed as a soil organism). (Not seen.)
Maier, N. H. (1903): tjber den feineren Bau der Wimperapparate der

Infusorien. Arch. Protistenk. 2, 71-179.
Mast, S. O. (1906): Light reactions in lower organisms. I. Stentor

coeruleus. J. Expl. Zool. 3, 359-399.
Maupas, E. (1879): Micronucleus of Stentor coerideus and Spirostomum

ambiguum. C. R. Acad. Sci., Paris (1879), 1274.
Maupas, E. (1883): Contribution a I'etude morphologique et anatomique

des Infusoires cilies. Arch. Zool. exper. et gen. i, 427-644.
Maupas, E. (1888): Recherches experimentales sur la multiplication des

infusoires cilies. Arch. Zool. exp. et gen. 6, 165-277.
Meissner, M. (1888): BeitragezurErnahrungsphysiologie derProtozoen.

Z. wiss. Zool. 46, 498-516.

BIBLIOGRAPHY OF STENTOR 385

Merton, H. (1932): Gestalterhaltende Fixierungs — versuche an

besonders kontraktilen Infusorien nebst Beobachtungen iiber das

Verhalten des lebenden Myoneme und Wimpem bei Stentor. Arch.

Protistenk. 77, 491-521.
Merton, H. (1935): Zwangsreaktionen bei Stentor als Folge bestimmter

Salzwirkung. Biol. ZenTralbL 55, 268-285.
M0LLER, K. M. (i960): On the nature of Stentorin. Compt. re?id. trav.

Lab. Carlsberg (in press).
MoNOD, J. (1933): Mise en evidence du gradient axial chez les infusoires

cilies par photolyse a I'aide des rayons ultraviolets. C. R. Acad. Set.,

Paris 196, 212-214.
Morgan, T. H. (1901a): Regeneration of proportionate structures in

Stentor. Biol. Bull. 2, 311-328.
Morgan, T. H. (1901b): Regeneration. Macmillan, London.
MoxoN, W. (1869) : On some points in the anatomy of Stentor and on its

mode of division. J. Anat. and Physiol. 3, 279-293. (Cambridge).
MuGARD, Helene and Bernadette Courtney (1955): Paralysie des In-
fusoires Cilies au moyen des phosphates alcalins. Bull. Soc. Zool.

France 80, 196-205.
MtJLLER, J. (1856): Einige Beobachtungen an Infusorien. Monatsber.

preuss. Akad. Wissensch. 1856, 389, 393 (Not seen).
MuLSOW, W. (1913): Die Conjugation von Stentor coerideus und Stentor

polyinorphus. Arch. Protistenk. 28, 363-388.
Neresheimer, E. R. (1903): Ueber die Hohe histologischen Differen-

zierung bei heterotrichen Ciliaten. Ibid. 2, 305-324.
Neresheimer, E. R. (1907): Nochmals iiber Stentor coeruleus. Ibid. 9,

137-138.
Oken, L. (18 1 5): Lehrbuch der Natiirgeschichte, 3 Teil. Zoologique.

Erste Abt. Fleischlose Thiere. Jena. (Not seen.)
Otterstrom, C. V. and K. Larsen (1946): Extensive mortality caused by

the infusorian Stentor polymorphus Ehrenb. Rept. Danish Biol. Sta. 48,

(1943-5), 53-57- (Dubious).
Packard, C. E. (1937): Oblique division in Stentor. Trans. Am. Micros.

Soc. 56, 191-192. (Incidental).
Park, O. (1929): The osmiophilic bodies of the protozoans, Stentor and

Leucophr^^s. Trans. Am. Micros. Soc. 48, 20-29.
Penard, E. (1922): Etudes sur les Infusoires d'Eau douce. Geneve.
Peters, A. W. (1904): Metabolism and division in protozoa. Proc. Am.

Acad. Arts, Sci. 39, 441-516.
Peters, A. W. (1908): Chemical studies on the cell and its medium.

III. The function of the inorganic .salts of the Protozoan cell and its

medium. Am. J. Physiol. 21, 105-125.
POPOFF, M. (1908): Experimentelle Zellstudien. Arch. Zellforsch. i,

245-379.
PopoFF, M. (1909): Experimentelle Zellstudien. II. Ober die Zellgrosse,
ihre fixierung und Vererbung. Ibid., 3, 124-180.

386 THE BIOLOGY OF STENTOR

Prowazek, S. (1901): Beitrage zur Protoplasmaphysiologie B/o/. Zentralbl.

21, 87-95; 144-155-
Prowazek, S. (1904): Beitrag zur Kenntnis der Regeneration und

Biologic der Protozoen. Arch. Protistenk. 3, 44-59.
Prowazek, S. (1913): Studien zur Biologie der Protozoen. VI. Arch.

Protistenk. 31, 47-71.
Randall, J. T. (1956): Fine structure of some ciliate protozoa. Nature^

178, 9-14.
Randall, J. T. and Sylvia Jackson (1958): Fine structure in Stentor

polymorphus. J. Biophys. Biochem. Cytol. 4, 807-830.
RoESLE, E. (1902): Die Reaktion einiger Infusorien auf einzelne In-

duktionsschlage. Z. allgem. Physiol. 2, 139-168. (Not seen.)
RoSKiN, G. (1915): La structure des myonemes contractiles de Stentor

coeruleus. Tirage a part de " Memoires Scientifiques des Chaniavsky

Universite de Moscou ", Vol. i. (Not seen.)
RosKiN, G. (1922): LJber den Bau von kontraktilen Elementen und

Stutzsubstanzen bei einigen Protozoen. (Russ. with Ger. summary.)

Arch. Soc. Russe Protist. Moscow i, 35-45. (Not seen.)
RosKiN, G. (1923): La structure des myonemes des Infusoires. Bull. Biol.,

France et Belg. 57, 143-15 1.
RosKiN, G., i. V. Semenov (1933): (Study of oxidation-reduction

processes in the cell). Arch. Russes, Anat. Hist, et Embryol. 12 (i),

Russ., 27-55; German trans. 180-182. (Not seen.)
ScHAEFFER, A. A. (1910): Selection of food in Stentor coeruleus (Ehr.).

y. Exptl. Zool., 8, 75-132.
ScHONFELD, C. (1959): Ubcr das parasitische Verhalten einer Astasia —

Art in Stentor coeruleus. Arch. Protistenk. 104, 261-264.
Schroder, O. (1907): Beitrage zur Kenntnis von Stentor coeruleus

Ehrbg. und »S^. roeselii Ehrbg. Ibid. 8, 1-16.
Schuberg, a. (1890): Zur Kenntnis der Stentor coeruleus. Zool. Jahrb.,

Abt. Anat. 4, 197-238.
Schuberg, A. (1905): Uber Cilien und Trichocysten einiger Infusorien.

Arch. Protistenk. 6, 61-110.
ScHULZE, K. L. (195 1): Experimentelle Untersuchungen iiber die

Chlorellen-symbiose bei Ciliaten. Biol. gen. {Vienna), 19, 281-298.
ScHWALBE, G. (1866): tjber die kontraktilen Behalter der Infusorien.

Arch. f. mikr. Anat. 2, 351-371.
Schwartz, V. (1934): Versuche iiber Regeneration und Kerndimorphis-

mus der Ciliaten. Nachr. Ges. Wiss. Gottingen, N.S. i, 143-155.
Schwartz, V. (1935): Versuche iiber Regeneration und Kerndimorphis-

mus bei Stentor coeruleus Ehrbg. Arch. Protistenk. 85, 100-139.
SiLEN, L. (1948): On a new stentor from the west coast of Sweden. Ark.

Zool. {Stockhohn) 40A (No. 8), i-io.
SiMROTH, H. (1876): Zur Kenntnis des Bewegungsapparates der In-

fusionsthiere. Arch. micr. Anat. 12, (Not seen).
Sleigh, M. A. (1956): Metachronism and frequency of beat in the

peristomial cilia of Stentor. J. Exptl. Biol. 33, 15-28.

BIBLIOGRAPHY^ OF STENTOR 387

Sleigh, M. A. (1957): Further observations on co-ordination and the

determination of frequency in the peristomial cilia of Stentor. Ihid. 34,

106-115.
Sleigh, M. A. (i960): The form of beat in cilia of Stentor and Opalina.

Ibid. 37, I -10.
SoKOLOFF, B. (1923): Hunger and regeneration. J. Roy. Micros. Soc,

London, (1923), 183-189.
SoKOLOFF, B. (1924): Das Regenerations problem bei Protozoen. Arch.

Protistenk. 47, 143-252.
SoKOLOFF, D. (1930a): Las particularidades del Stentor viridis de

Xochimilco. An. Inst. Biol. {Univ. Nac. Mexico) 1(1), 83-86. (Not

seen. See Kahl.)
SoKOLOFF, D. (1930b): Stentor oligomicleatiis , sp. nov. Ibid, i, 327-328.

(Not seen. See Kahl.)
SoSNOWSKi, J. (1899): Relations entre le noyeau et le Protoplasme chez

les protozoaires. Trans. Lab. Zoot. Varsovie. 20, 1-47 (Not seen).
Sprugel, G., Jr. (1951): Vertical distribution of Stentor coeruleus in

relation to dissolved oxygen levels in an Iowa pond. Ecology 32,

147-149.
Stein, F. (1867): Der Organismus der Infusionsthiere. Abt. II. Leipzig.
Stevens, Nellie M. (1903): Notes on regQwev^Xxonin Stentor coeruleus.

Arch. Entzu.-mech. 16, 461-475.
Stolte, H. a. (1922): Der Einfluss der Umwelt auf Macronucleus und

Plasma von Stentor coeruleus Ehrbg. Bin experimentelles Beitrag zur

Frage der Kernplasmabeziehungen. Arch. Protistenk. 45, 344-389.
Strom, K. M. (1926): The influence of altered H-ion concentrations on

Stentor, Diaptomus and Daphnia. Nyt. Mag. Naturvidenskab 64,

109-115. (Not seen.)
SwARCZEW^SKY, B. (1929): Zur Kenntnis der Baikalprotistenfauna. Die

an den Baikalgammariden lebenden Infusorien. VI. Stentorina. Arch.

Protistenk. 65, 38-44.
Tartar, V. (1941a): Grafting and reconstitution in a ciliate protozoan.

(Abstr.) Anat. Rec, 81 Suppl. p. 132.
Tartar, V. (1941b): Intracellular patterns: Facts and principles con-
cerning patterns exhibited in the morphogenesis and regeneration of

ciliate protozoa. Growth (suppl.) 5, 21-40.
Tartar, V. (1950): Methods for the study and cultivation of protozoa.

In, Studies Honoring Trevor Kincaid, (Edited by M. E. Hatch),

University of Washington Press, Seattle.
Tartar, V. (1953): Chimeras and nuclear transplantations in ciliates,

Stentor coeruleus X S . polymorphus . J . Exptl. Zool. 124, 63-104.
Tartar, V. (1954): Reactions of Stentor coeruleus to homoplastic

grafting. Ibid. 127, 511-576.
Tartar, V. (1956a): Grafting experiments concerning primordium

formation in Stentor coeruleus. Ibid. 131, 75-122.
Tartar, V. (1956b): Further experiments correlating primordium sites

with cytoplasmic pattern in Stentor coeruleus. Ibid. 132, 269-298.

388 THE BIOLOGY OF STENTOR

Tartar, V. (1956c): Pattern and substance in Stentor. In Cellular

Mechanisms in Differentiations and Growth (Edited by D. Rudnick),

Princeton University Press.
Tartar, V. (1957a): Reactions of Stentor coeruleus to certain substances

added to the medium. Exptl. Cell Research. 13, 317-332.
Tartar, V. (1957b): Equivalence of macronuclear nodes. J. Exptl. Zool.

135, 387-402.
Tartar, V. (1957c): Deletion experiments on the oral primordium of

Stentor coeruleus. Ibid. 136, 53-74.
Tartar, V. (1958a): Stentor introversus, n. sp. jf. Protozoal, 5, 93-95.
Tartar, V. (1958b): Induced resorption of oral primordia in regenerating

Stentor coeruleus. jf. Exptl. Zool. 139, 1-32.
Tartar, V. (1958c): Specific inhibition of the oral primordium by

formed oral structures in Stentor coeruleus. Ibid. 139, 479-505.
Tartar, V. (1959a): Equational division of carbohydrate reserves in

Stentor coeruleus. Ibid, 140, 269-280.
Tartar, V. (1959b): Effects of misorientation by rotation of the ecto-

plasmic pattern of S. coeruleus. (Abstr.) J. Protozoal. Suppl. 6, 32.
Tartar, V. (1959c): Division of the macronucleus in Stentor coeruleus

with regard to nodulation. (Abstr.) Ibid. 6, 32.
Tartar, V. (i959d): New findings on oral regeneration in Stentor

coeruleus. (Abstr.) Ibid. 6, 32.
Tartar, V. (19596): Persisting division in Stentor coeruleus regardless

of removal of parts. (Abstr.) Anat. Rec. Suppl. 134, 644.
Tartar, V. (i959f) : A reproducible lethal abnormality in Stentor coeruleus.

(Abstr.) Anat. Rec. Suppl. 134, 645.
Tartar, V. (i959g): Some eflfects of altered nucleo-plasmic ratio in

Stentor coeruleus. (Abstr.) Ibid. 134, 645.
Trembley, a. ( 1 744) : Translation of a letter from Mr. Abraham Trembley,

F.R.S., to the President, with observations upon several newly

discovered species of fresh water polypi. Phil. Trans. Royal Soc.

(London) 43, 169-183.
Tuffrau,^M. (1957): Les facteurs essentiels du phototropisme chez le

Cilie heterotriche Stentor niger. Bull. Soc. Zool., France 82, 354-356.
Turner, J. P. and P. Brandwein (1937): (Culture of Stentor). In,

Culture Methods for Invertebrates, (Edited by P. Galtsoff et al.) Ithica,

New York, pp. 60-64.
Uhlig, G. (1959): Polaritatsabhangige Anlagenentwicklung bei Stentor

coeruleus. Z. Naturforsch. Tilbigen. 14b, 353-354.
Uhlig, G. (i960): Entwicklungsphysiologische Untersuchungen zur

Morphogenese von Stentor coeruleus Ehrbg. (Doctoral dissertation,

Tiibigen Univ.) Arch. Protistenk. 105, 1-109.
Verworn, M. (1892): Die physiologische Bedeutung des Zellkerns.

Pfiiiger's Archiv. 51, 1-118.
Villeneuve-Brachon, Simone (1940): Recherches sur les cilies hetero-

triches, cinetome, argyrome, myonemes, formes nouvelles ou peu

connues. Arch. Zool. exper. et gen. 82, 1-180.

BIBLIOGRAPHY OF STENTOR 389

Weisz, p. B. (1948a): Time, polarity, size and nuclear content in the
regeneration of Stentor fragments. J. Exptl. Zool. 107, 269-287.

Weisz, P. B. (1948b): The role of carbohydrate reserves in the regenera-
tion of Stentor fragments. Ibid. 108, 263-278.

Weisz, P. B. (1948c): On the growth of regenerating fragments in Stentor
coeruleus. Ibid. 109, 427-437.

Weisz, P. B. (i948d): Regeneration in Stentor and the gradient theory.
Ibid. 109, 439-449.

Weisz, P. B. (1949a): A cytochemical and cytological study of differentia-
tion in normal and reorganizational stages oi Stentor coeruleus. Jf. Morph.

84, 335-363-
Weisz, P. B. (1949b): Phosphatases in normal and reorganizing Stentors.

Biol. Bidl. 97, 108-110.
Weisz, P. B. (1949c): The role of specific macronuclear nodes in the

differentiation and the maintenance of the oral area in Stentor.

y. Exptl. Zool. Ill, 141-156.
Weisz, P. B. (1950a): On the mitochondrial nature of the pigmented

granules in Stentor and Blepharisnta. J. Morph. 86, 177-184.
Weisz, P. B. (1950b): A correlation between macronuclear thymonucleic

acid concentration and the capacity of morphogenesis in Stentor.

Ibid. 87, 275-286.
Weisz, P. B. (1951a): Homoplastic grafting in Stentor coeruleus. Biol. Bull.

100, 1 16-126.
Weisz, P. B. (1951b): An experimental analysis of morphogenesis in

Stentor coeruleus. J. Exptl. Zool. 116, 231-257.
Weisz, P. B. (1951c): A general mechanism of differentiation based on

morphogenetic studies in ciliates. Am. Naturalist 85, 293-311.
Weisz, P. B. (1953): The embryologist and the protozoon. Scientific

American 188, 76-82.
Weisz, P. B. (1954): Morphogenesis in protozoa. Quart. Rev. Biol. 29,

207-229.
Weisz, P. B. (1955): Chemical inhibition of regeneration in Stentor

coeruleus. J. Cell. Comp. Physiol. 46, 517-527.
Weisz, P. B. (1956): Experiments on the initiation of division in Stentor

coeruleus. y. Exptl. Zool. 131, 137-162.
Wetzel, A. (1925): Vergleichende cytologische Untersuchungen an

Ciliaten. Arch. Protistenk. 51, 207-304.
Whiteley, a. H. (1956): Respiratory patterns in regenerating nucleate

and enucleate fragments of Stentor coeruleus. (Abstr.) J. Cell. Comp.

Physiol, 48, 344-345.
Whiteley, A. H. (i960): (Same titled C.R. Trav. Lab. Carlsberg.

(In press).
Worcester, G. W. (1884): Life history oi Stentor coeruleus. Proc. Centr.

Ohio Sci. Assoc, i, 97. (Dubious).
WORLEY, L. G. (1934): Ciliary metachronism and reversal in Paramecium,

Spirostomum and Stentor. J. Cell. Comp. Physiol. 5, 53-72.

390 THE BIOLOGY OF STENTOR

Zhinkin, L. (1930): Zur frage der Reservestoffe bei Infusorien (Fett

und Glykogen bei Stentor polymorphiis). Z. Morph. u. Okol. Tiere.

(Z. wissenschaftliche Biologie, Abteilung A.) 18, 217-248.
Zhinkin, L. and P. Obraztsov (1930): (Ecology of Stentor polymorphus

and Stentor coeruleus.) Russian with German summary. Hydrobiol. Z.

USSR. 9, 151-153. (Not seen.)
ZiNGHER, J. A. (1933): Beobachtung an Fetteinschliissen bei einigen

Protozoen. Arch. Protistenk. 81, 57-87.
ZiNGHER, J. A. and W. W. Fisikow (1931): Biometrische Untersuchungen

an Infusorien. i. Uber die Mittelgrosse von Stentor coeruleus Ehrbg.

Ibid. 73, 482-486.

OTHER REFERENCES CITED

Andrews, E. A. (1923): Folliculina: case-making, anatomy and trans-
formation, y. Morph. 38, 207-277.
Andrews, E. A. (1947): Temperature effect upon rate of feeding in a

folliculinid. Physiol. Zool. 20, 1-4.
Andrews, E. A. (1949): Folliculinid life-cycle.^. Morph. 84, 401-409.
Berglas, a. (1957): Cancer: Nature, Cause and Cure. Inst. Pasteur, Paris.
Bohr, N, (1958): Atomic Physics and Human Knozvledge. New York.
Bonner, J. T. (1954) : The development of cirri and bristles during binary

fission in the ciliate Euplotes eurystomus J. Morph. 95, 95-108.
Bonner, J. T. (1958): The Evolution of Development. Cambridge

University Press, England.
Bracket, J. (1955): Recherches sur les interactions biochimiques entre

le noyau et le cytoplasme chez les organismes unicellulaires. Biochim.

Biophys. Acta 18, 247-268.
Briggs, R. and T. J. King (1955): Specificity of nuclear function in

embryonic development. In, Biological Specificity and Growth (Edited

by E. G. Butler), Princeton University Press.
Bronsted, H. V. (1955): Planarian regeneration. Biol. Rev. 30, 65-126.
Calkins, G. N. (1911a): Regeneration and cell division in Uronychia.

jf. Exptl. Zool. 10, 95-116.
Calkins, G. N. (1911b): Effects produced by cutting Paramecium cells.

Biol. Bidl. 21, 36-72.
Chatton, E. and A. Lwoff (1935a): Les Cilies apostomes. I. Apergu

historique et general ; etude monographique des genres et des especes.

Arch. Zool. exp. et gen. 77, 1-453.
Chatton, E. and A. Lwoff (1935b): La constitution primitive de la

strie ciliare des infusoires. La desmodexie. C.R. Soc. Biol., Paris 118,

1068-1072.

REFERENCES 39I

Chatton, E. and Josephine Seguela (1940): La continuite genetique

des formations ciliares chez les cilies hypotriches. Le cinetome et

I'argyrome au cours de la division. Bull. Biol, de France-Belg. 74, 1-94.
Chen, Y. T. (1944): Studies on the neuromotor systems of Stylonychia

pustiilata and Stylonychia mytilus. J. Morph. 75, 335-345.
Child, C. M. (1941): Patterns and Problems of Development. University

of Chicago Press.
CiENKOWSKi, L. (1865): Beitrage zur Kenntnis der Monaden. Arch.

mikr. Anat. i, 201-232. (Not seen.)
CoMANDON, J. and P. de Fonbrune (1939a): Ablation du noyau chez une

Amibe. Reactions cinetiques a la piqure de I'amibe normal ou denucleie.

C.R. Soc. Biol., Paris 130, 740-744.
CoMANDON, J. and P. de Fonbrune (1939b): Greffe nucleaire totale,

simple, ou multiple, chez un Amibe. Ibid. 130, 744-748.
Danielli, J. F. (1958): Studies of inheritance in amoebae by the technique

of nuclear transfer. Proc. Roy. Soc. (London) B 148, 321-331.
Danielli, J. F. (1959): The cell-to-cell transfer of nuclei in Amoeba and

a comprehensive cell theory. Ann. N. Y. Acad. Sci. 78, 675-687.
Danielli, J. F., I. J. Lorch, M. J. Ord, and E. G. Wilson (1955):

Nucleus and cytoplasm in cellular inheritance. Nature 176, 1114.
Daniels, E. W. (195 i): Studies on the effect of x-irradiation upon

Pelomyxa carolinensis with special reference to nuclear division and

plasmotomy. J. Exptl. Zool. 117, 189-210.
Delbruck, M. and W. Reichardt (1956): System analysis for the light

growth reactions of Phycomyces. In, Cellular Mechanisms in Differen-
tiation and Growth (Edited by Rudnick), Princeton University Press.
Dembowska, W. S. (1938): Korperreorganisation von Stylonychia mytilus

beim Hungem. Arch. Protistenk. 91, 89-105.
Demeric, M. (1934): Biological action of small deficiencies of X-

chromosome of Drosophila melanogaster. Proc. Natl. Acad. Sci.,

U.S. 20, 354-359.

Ehret, C. F., and E. L. Pov^rs (1959): The cell surface of Paramecium.

Intern. Rev. Cytol. 8, 97-133.
Elsasser, W. M. (1958): The Physical Foundation of Biology. Pergamon

Press, London and New York.
Ephrussi, B. (1953): Nucleo-cytoplasmic Relations in Micro-organisms.

Oxford University Press.
Faure-Fremiet, E. (1932): Division et morphogenese chez Folliculina

ampulla O. F. Miiller. Bull. bioL, France-Belg. 66, 77-110.
Faure-Fremiet, E. (1945a): Symetrie et polarite chez les cilies bi-ou

multi-composites. Ibid. 79, 106-150.
Faure-Fremiet, E. (1948a): Doublets homopolaires et regulation

morphogenetique chez le cilie Leucophrys patula. Arch. Anat. micros.

et Morphol. exper. 37, 183-203.
Faure-Fremiet, E. (1948b): Les mecanismes de la morphogenese chez

les Cilies. Folia Biotheoretica 3, 25-58.

392 THE BIOLOGY OF STENTOR

Faure-Fremiet, E. (1949): Action du lithium sur la stometogenese chez

les Cilies. Belg.-NederL Cytoemhryol. Dagen. Genet. 1949, 100-102.
Faure-Fremiet, E. (1950): Probleme moleculaire de la morphogenese.

Ann. Biol. 26, 361-379.
Faure-Fremiet, E. (1954): Les problemes de la differentiation chez les

protistes. Bull. Soc. zool. France 79, 311-329.
Faure-Fremiet, E. and H. Mugard (1948): Segregation d'un materiel

cortical au cours de la segmentation chez I'oeuf de Teredo norvegica.

C.R. Acad. Sci., Paris 227, 1405-1411.
Fawcett, D. W. and K. R. Porter (1954): A study of the fine structure

of ciliated epithelia. jf. Morph. 94, 221-281.
Garnjobst, Laura (1937): A comparative study of protoplasmic re-
organization in two hypotrichous ciliates, Stylonethes sterkii and

Euplotes taylori, with special reference to encystment. Arch. Protistenk.

89, 317-381.
Gelei, J. V. (1936): Das erregungsleitende System der Ciliaten. C.R.

I2th Congres Internat. de Zool. (Lisbonne, 1935) i, 174-206.
GlESE, A. C. (1938): Cannibalism and gigantism in Blepharisma. Trans.

Am. Micros. Soc. 57, 245-255.
GiESE, A. C. (1949): A cytotoxin from Blepharisma. Biol. Bull., 87,

145-149.
GiESE, A. C. (1957): Photodynamic effect of Blepharisma pigment on

nerve. J. Cell. Comp. Physiol. 49, 295-302.
GiESE, A. C. and R. H. Alden (1938): Cannibalism and giant formation

in Stylonychia. jf. Exptl. Zool. 78, 11 7-1 34.
GOLDSCHMIDT, R. (1940): Chromosomes and genes. In, The Cell and

Protoplasm, (Edited by F. R. Moulton), Science Press, Washington,

D.C., pp. 56-66.
GuSTAFSON, T. (1954): Enzymatic aspects of embryonic differentiation.

Intern. Rev. Cytol. 3, 277-327.
HAmmerling, J. (1953): Nucleo-cytoplasmic relationships in the

development of Acetabularia. Intern. Rev. Cytol. 2, 475-498.
Hammond, D. M. (1937): The neuromotor system of Euplotes patella

during binary fission and conjugation. Quart. J. Micros. Sci. 79,

507-557.
Hartmann, M. (1928): Ueber experimentelle Unsterblichkeit von

Protozoen-Individuen. Zool. Jahrh. 45, 973-987.
HoFER, B. (1890): Experimentelle Untersuchungen iiber den Einfiuss

des Kerns auf das Protoplasma. Jena. Z. Naturwiss. 24, 105-176.
Holmes, S. J. (1907): The behaviour oi Loxophyllum and its relation to

regeneration. J. Exptl. Zool. 4, 399-430.
Holtfreter, J. (1949): Phenomena relating to the cell membrane in

embryonic processes. Exptl. Cell Research, Suppl. i (1949), 497-5 lO-
Horning, E. S. (1927): On the orientation of mitochondria in the surface

cytoplasm of Infusorians. Austral. J. Exptl. Biol. Med. 4, 187-190.

REFERENCES 393

HoRSTADius, S. (1950): Transplantation experiments to elucidate inter-
actions and regulations within the gradient system of the developing

sea urchin egg. J. Exptl. Zool. 113, 245-276.
Jones, E. E., Jr. (195 1): Encystment, excystment, and the nuclear cycle

in the ciliate Dileptus anser. J. Elisha Mitchell Sci. Soc. 67, 205-218.
Klein, B. M. (1932): Das Ciliensystem in seiner Bedeutung fiir Loko-

motion, Koordination und Formbildung mit besonderer Beriick-

sichtigung der Ciliaten. Ergebn. d. Biol. 8, 75-179.
Lewin, K. R. (1911-12): The behaviour of the infusorian micronucleus

in regeneration. Proc. Roy. Soc. {London) B84, 332-344.
LiLLiE, F. R. (1906): Observations and experiments concerning the

elementary phenomena of embryonic development in Chcetopterus.

y. Exptl. Zool. 3, 153-268.
LoEFER, J. B. (1936): Isolation and growth characteristics of the " Zooch-

lorella " of Paramecium bursaria. Am. Naturalist 70, 184-188.
LoRCH, I. Joan and J. F. Danielli (1950) : Transplantation of nuclei from

cell to cell. Nature 166, 329-330.
LoRCH, I. Jo.\N and J. F. Danielli (1953): Nuclear transplantation in

Amoeba. I. Some species characters of Amoeba proteus and Amoeba

discoides. Quart. J. Micros. Sci. 94, 445-460.
Lucas, Miriam S. (1932) : The cytoplasmic phases of rejuvenescence and

fission in Cyathodiunium piriforme. II. A type of fission heretofore

undescribed for ciliates. Arch. Protistenk. 'j'j^ 407-472.
Lund, E. E. (1935): The neuromotor system of Oxytricha. jf. Morph. 58,

257-277-
Lund, E. J. (1914): The relations of Bursaria to food. J. Exptl. Zool. 16,

1-52; 17, 1-39.
Lund, E. J. (1917-18): Reversibility of morphogenetic processes in

Bursaria. J. Exptl. Zool. 24, 1-33.
LwoFF, A. (1950): Problems of Morphogenesis in Ciliates. Wiley, New

York.
McDouGALL, W. (1938): The Riddle of Life. Methuen, London.
Marsland, D. a. (1943): Quieting Paramecium for the elementary

student. Science 98, 414.
Mast, S. O. (1909): The reactions of Didinium nasutum (Stein) with

special reference to the feeding habits and functions of the trichocysts.

Biol. Bull. 16, 91-118.
Mazia, D. (1952): Physiology of the cell nucleus. In Modern Trends in

Physiology and Biochemistry (Edited by E. S. G. Barron), New York,

pp. 77-122.
MoNNE, L. (1948): Functioning of the cytoplasm. Advances in Enzy-

mology 8, 1-65.
Moore, A. R. (1945): The Individual in Simpler Forms. University of

Oregon Press.
MuGARD, Helene (1947): Division et morphogenese chez les Ophryo-

glenes. C.R, Acad. Sci., Paris 225, 70-72.

394 THE BIOLOGY OF STENTOR

MuGARD, Helene (1948): Regulation du nombre des cineties au cours du

cycle de croissance et de division chez un cilie: Ichthyophthirsus multi-

filies Fouguet. Arch. d'Anat. micros. 37, 204-213.
MuGARD, Helene (1957): L'infraciliature chez la larve d'Oursin Paracen-

trotus lividus. Bull Soc. Zool. France 82, 81-88.
Nadler, J. E. (1929): Notes on the loss and regeneration of the pellicle

in Blepharisma iindidans. Biol. Bull. 56, 327-330.
Needham, J. N. (1933): Dissociability of the fundamental processes in

ontogenesis. Biol. Rev. 8, 180-223.
NusSBAUM, M. (1884): Ober spontane und kiinstliche Zelltheilung.

Sitzber. niederrhein. Ges. Nat.-u. Heilk., Bonn 41, 259-263.
Ohler, R. (1922): Die Zellverbindungen von Paramecium bursaria mit

Chlorella vulgaris und anderen Algen. Arb. Staatsinst. exper. Therapie

Georg Speyer Haus 15. (Not seen.)
Okada, Y. K. (1930): Transplantationsversuche an Protozoen. Arch.

Protistenk. 69, 39-94.
Parducz, B. (1953): Zur Mechanik der Zilienbewegung. Acta Biol.,

Acad. Sri. Hungari 4, 177-220.
Parker, G. H. (1929): The neurofibril hypothesis. Quart. Rev. Biol. 4,

155-178.
Pringsheim, E. G. (1928): Physiologische Untersuchungen an Para-
mecium bursaria. Ein Beitrag zur Symbioseforschung. Arch. Protistenk.

64, 289-418.
Raven, C. P. (1949): The influence of lithium on the development of the

pond snail, Limncca stagnalis L. Exptl. Cell Research, Suppl. i, 542-544.
Reynolds, Mary E. C. (1932): Regeneration in an amicronucleate

infusorian. J^. Exptl. Zool. 62, 327-361,
Robertson, T. B. (1925): The influence of certain dyes upon the

multiplication of Infusoria (Enchelys) with special reference to the

acridine dyes (Acriflavine and Proflavine). Austral, jf. Exptl. Biol. Med.

Sci. 2, 21-44.
Roque, Marie (1956): La ciliature buccale pendant I'autogamie et la

conjugaison chez Paramecium aurelia. C.R. Acad. Sci., Paris, 242, 2592.
Rose, S. M. (1957): Cellular interaction during differentiation. Biol. Revs.

32, 351-382.
Roth, L. E. (1956): Aspects of ciliary fine structure in Euplotes patella.

y. Biophys. Biochem. Cytol. 2, (Suppl.), 235-240.
ScHMAHL, O. (1926): Die Neubildung des Peristoms bie der Teilung

von Bursaria truncatella. Arch. Protistenk. 54, 359-430.
ScHV^ARTZ, V. (1947): Uber die Physiologic des Kerndimorphismus bei

Paramecium bursaria. Z. Naturforsch. 26, 369-381.
SoKOLOFF, B. (1913): Contribution au probleme de la regeneration des

Protozoaires. C.R. des Seances et Memoires de la Soc. de Biol. 75,

297-301.
SoKOLOFF, B. (1934): Vitality. Dutton, New York,

REFERENCES 395

SoNNEBORN, T. M. (1932): Experimental production of chains and its
genetic consequences in the ciliate protozoan, Colpidium covipyhim
(Stokes). Biol Bull. 63, 187-211.

SoNNEBORN, T. M. (1947): Recent advances in the genetics of Para-
mecium and Euplotes. Advances in Genetics i, 263-358,

SoNNEBORN, T. M. (1951): The role of genes in cytoplasmic inheritance.
In, Genetics in the 20th Century (Edited by L. C. Dunn). Macmillan,
New York, pp. 291-314.

Sterki, V. (1878): Beitrage zur Morphologic der Oxytrichinen. Z. wiss.
Zool. 31, 28-58.

Suzuki, S. (1957): Morphogenesis in the regeneration of Blepharisma
undulans japonicus Suzuki. Bull. Yamagata Univ., Nat. Sci. 4, 85-192.

SwANN, M. M. (1954): The control of cell division. In, Recent Develop-
ments in Cell Physiology. London.

Tartar, V. (1940): Nuclear reactions in Paraynecium. (Abstr.) Anat. Rec.
78, (Suppl.), 109.

Tartar, V. (1954) : Anomalies in regeneration oi Paramecium . J . Protozool.

1,11-17-
Taylor, C. V. (1928): Protoplasmic reorganization in Uronychia uncinata

sp. nov. during binary fission and regeneration. Physiol. Zool. i, 1-25.
Taylor, C. V. and W. P. Parser (1924): Fatal effects of the removal of

the micronucleus in Euplotes. Univ. Calif. Publ. Zool. 26, 1 31-144.
Turner, J. P. (1940) : Cytoplasmic inclusions in Tillina canalifera Turner.

Arch. Protistenk. 93, 255-272.
VisscHER, J. P. (1923): Feeding reactions in the ciliate, Dileptus gigas,

with special reference to the function of the trichocysts. Biol. Bull. 45,

113-143-
Weisz, p. B. (1949): The role of the macronucleus in the differentiation

of Blepharisma undulans. jf. Morph. 85, 503-518.
Whitman, O. C. (1893): The inadequacy of the cell theory of develop-
ment, jf. Morph. 8, 639-658.
WoODGER, J. H. (1929): Biological Principles. London.
WoRLEY, L. G. (1933): The intracellular fibre systems of Paramecium.

Proc. Natl. Acad. Sci., U.S., 19, 323-326.
Yagiu, R. (1951): Studies on Condylostoma spatiosum Ozaki and Yagiu.

III. The relationship of the quantity of the macronucleus and the

power of division. J. Sci., Hiroshima Univ., Series B, Div. i, 12, 121-130.
Yagiu, R. (1952): Studies on Condylostoma spatiosum Ozaki and Yagiu.

V. Abnormal phenomena caused by being kept in fresh water. Ibid. 13,

92-109.
Yow, F. W. (1958): A study of the regeneration pattern of Euplotes

eurystomus. J. Protozool. 5, 84-88.
YuSA, A. (1957): The morphology and morphogenesis of the buccal

organelles in Paramecium with particular reference to their systematic

significance. J. Protozool. 4, 128-142.
Zeleny, C. (1905): The relation of the degree of injury to the rate of

regeneration. J. Exptl. Zool. 2, 347-369.

*AA

AUTHOR INDEX

Adolph, 380
Allescher, 261, 292, 293
Alverdes, 24
Andresen, 48

Andrews, 13, 14, 15, 34, 36, 37,
38, 40, 41, 42, 44, 45, 47, 56,

160, 250, 266, 338

Balamuth, 147

Balbiani, 74, 84, 91, 98, 100,
102, 111, 113, 123, 196, 206,
209, 267, 273, 286, 293, 300,
301, 302, 303, 304, 305, 323,
336, 348

Barbier, Faure-Fremiet and
Lederer, 47

Bary, 335, 337

Belda and Bowen, 346

Berglas, 215

Bishop, 287

Bohr, 378

Bonner, 122, 195

Brachet, 133, 302, 303

Brauer, 32, 63, 65

Briggs and King, 309

Bronsted, 89, 201

Bullington, 1 7

BuRNSiDE, 305, 347

BuTSCHLi, 43, 58

Calkins, 206, 280
Causin, 54, 82, 100, 108, 113,
116, 121, 129, 153, 155, 156,

161, 179, 188, 196, 304
Chambers and Kao, 245, 246,

353
Chatton and Lwoff, 50, 367
Chatton and Seguela, 160
Chen, 33

Child, 117, 118, 145, 195, 201,

246, 247
Cienkowski, 354
Claparede and Lachmann, 42,

273
Comandon and De Fonbrune,

301, 305, 309, 319, 359
Cox, 70

Dabrowska, 24

Daniel, 237, 248, 249, 251, 252

Danielli, 319, 375

Danielli, Lorch, Ord and

Wilson, 319
Daniels, 3, 355
Davenport and Neal, 248
Dawson, 347

Delbruck and Reichardt, 66
Dembowska, 259, 303
Demerec, 371
De Terra, 74, 84, 85, 266, 295,

303, 323, 353
DiERKS, 13, 14, 15, 24, 29, 30, 31,

33, 34, 35, 36, 37, 38, 49, 52, 53,

55, 66, 243
Driesch, 170

Ehrenberg, 333

Ehret and Powers, 122, 180, 368

Elsasser, 372

Ephrussi, 375

Faure-Fremiet, 46, 47, 111, 160,

161,179,188,206,208,215,255,

338, 364, 365, 366, 367

Faure-Fremiet and Lederer, 382

Faure-Fremiet and Mugard, 373

Faure-Fremiet and Rouiller,

31,49, 56, 57, 58

396

AUTHOR INDEX

397

Faure-Fremiet, Rouiller and
Gauchery, 44, 45, 49, 52, 53
Fawcett and Porter, 31, 238

Garnjobst, 377

Gelei, 6, 14, 15, 16, 17, 34, 36,

47, 49, 52, 55, 56, 240, 274, 322,

342, 366
Gelfan, 58
Gerstein, 347
GiESE, 17, 48, 347
GiESE and Alden, 1 7
goldschmidt, 67
Greeley, 244
Gruber, 37, 70, 74, 105, 120, 130,

147, 152, 201, 226, 259, 286, 298,

299, 304, 352, 355, 373

GUSTAFSON, 256

GuTTES and Guttes, 114, 115, 153

Hamburger, 324, 325
Hammerling, 22, 268, 269, 271,

272, 297, 317, 346
Hammond, 367
Hartmann, 131, 259, 292
Hausmann, 24
Have, 40, 42
Hegner, 342
Heilbrunn, 245
Hetherington, 99, 100, 156, 273,

274, 346, 347, 348
Hofer, 241, 301, 302
Holmes, 21
Holt and Lee, 23
holtfreter, 374
Horning, 46
horstadius, 213

HOWLAND, 273
Hyman, 347

IsHiKAWA, 53, 226, 240, 299
IvANic, 15, 260, 322

Jennings, 6,7,12,17,19,22,239
Jennings and Jamieson, 17, 19,

21
Jennings and Moore, 22

Johnson, 7, 12, 14, 15, 16, 25, 29,
33, 37, 38,40,41,46,49, 52, 56,
58, 63, 66, 70, 72, 73, 74, 75, 77,
82, 92, 99, 102, 113, 116, 149,
152, 153, 159, 163, 164, 206, 260,
261, 262, 267, 274, 281, 286, 287,
294, 323, 325, 333, 336, 337, 348

Jones, 286

Kahl, 25, 44, 334, 335
Kalmus, 273
Kent, 333
Kessler, 267
Kimball, 256
Kirby, 273, 333
Klein, 52, 365, 366

Lankester, 46, 47, 48, 342
Lewin, 114
Lewis, 274
Lieberkuhn, 48
LiLLiE, 120, 213, 373
Loeb, 244

LOEFER, 267

LoRCH and Danielli, 309, 312,

319
Lucas, 218
Lund, E. E., 30
Lund, E. J., 13, 91, 170, 188,

220, 259
LwoFF, 3.8, 67, 90, 159, 365, 367,

368

MCDOUGALL, 379

Madlen, 384
Maier, 29, 32, 47, 267, 336
Marsland, 241, 349
Mast, 23, 259
Maupas, 41, 58, 259, 260
Mazia, 375
" Meissner, 266
Merton, 5, 19, 21, 239, 240, 243,

348
M0LLER, 48, 49, 291, 322, 323
MONNE, 315
MoNOD, 247
Moore, 245

398

Morgan, 67, 99, 107, 110, 115,
119,120,124,128,129,161,299,
355

MoxoN, 9, 30, 40, 44, 70, 161,
163, 323

MuGARD, 160, 225, 373

MuGARD and Courtney, 241

MULLER, 273

MuLSOW, 323, 324, 325

Nadler, 43, 250

Needham, 378

Neresheimer, 31, 33, 49, 54, 55,

241, 242, 243, 250
Nussbaum, 105

Ohler, 272

Okada, 355

Oken, 333

Otterstrom and Larsen, 274

Packard, 81, 333

Parducz, 19, 56

Parisis, 242

Park, 40, 57

Parker, 371

Penard, 336

Peters, 239, 249, 250, 251, 345,
347

PopoFF, 52, 63, 74, 76, 81, 82,
202, 260, 305, 306, 308

Pringsheim, 268, 269, 272

Prowazek, 12, 44, 47, 53, 81, 102,
107,113,124,129,130,145,196,
245, 250, 252, 260, 261, 267, 285,
287, 291, 292, 293, 299, 300, 301,
302, 304, 342, 346, 355, 375, 376

Randall, 49, 50, 60, 238
Randall and Jackson, 30, 31, 34,

42, 45, 49, 53, 56, 57, 58, 76, 346,

348
Raven, 256
Reynolds, 280
Robertson, 132
Roesle, 24
Roque, 368

THE BIOLOGY OF STENTOR

Rose, 145

RosKiN, 386

RosKiN and Semenov, 265

Roth, 30

Schaeffer, 11, 13, 14, 21
SchmAhl, 75, 159, 160, 169, 280,
299, 303

SCHMITT, 245

Schonfeld, 273

Schroder, 29, 30, 31, 33, 37, 38,

43, 44, 49, 52, 55
ScHUBERG, 6, 12, 15, 29, 30, 31,

33, 36, 48, 56, 66, 70, 76, 160,

164, 179, 250, 274
ScHULZE, 22, 260, 268, 269, 271,

272, 346
schwalbe, 40
Schwartz, 61, 70, 73, 75, 82, 84,

87,91,93,95,101,102,103,111,

113,130,153,160,163,164,186,

266, 280, 287, 291, 300, 301, 302,
304, 308, 323, 344, 347, 348

SiLEN, 338
SiMROTH, 30

Sleigh, 13, 57, 234, 235, 236,

346, 349
Sokoloff, B., 107, 118, 119, 120,

132, 201, 299, 305
Sokoloff, D., 387
SoNNEBORN, 206, 286, 375, 376
SosNOWSKi, 260
Sprugel, 265

Stein, 25, 26, 30, 37, 70, 287, 323
Sterki, 30
Stevens, 29, 64, 67, 76, 84, 105,

129, 161, 163, 165, 206, 301
Stolte, 74, 88, 120, 260, 265,

267, 274, 283, 285, 286, 293, 347
Strom, 246, 346

Suzuki, 84, 103, 114, 290, 355,

361 et seq.
SWANN, 89

Swarczewsky, 337

Tartar, 30, 64, 70, 74, 79, 86, 89,
91,92,98,99,100,101,107,110,

AUTHOR INDEX

399

Tartar, {continued)

111,112,116,117,118,119,121,
125,129, 131, 135,136,139,142,
145,146,148,149,152,156,161,
164,169,170,174,175,176,178,
180, 187, 188, 189, 190, 198, 203,
206, 213, 216, 221, 227, 229, 237,
238, 239, 240, 242, 252, 254, 263,
265, 267, 277, 280, 287, 290, 300,
301, 303, 304, 305, 306, 311, 312,
313,317,320,343,349,355,365,
366, 368, 370, 375, 376

Taylor, 116, 119

Taylor and Farber, 280

Trembley, 70, 333

Tuffrau, 23

Turner, 46

Turner and Brandwein, 347

Uhlig, 115, 119, 174, 176, 184,
202, 204, 208, 227, 230, 346

Verworn, 240, 242, 309, 374
Villeneuve-Brachon, 50, 52, 55,

66, 75, 138, 159, 348
ViSSCHER, 259

Weisz, 17, 38, 40, 44, 45, 47, 52,
57, 58, 59, 73, 75, 76, 77, 78, 79,
80, 82, 83, 86, 88, 91,92,95,98,
99, 100, 101, 102, 107, 108, 110,
113,114,115,116,117,118,119,
121,124,131,132,133,135,138,
142,145,146,153,161,164,170,
189, 196, 201, 202, 206, 221, 237,
247, 250, 252, 260, 262, 263, 266,
285, 286, 288, 289, 294, 295, 303,
305, 306, 310, 347, 348, 349, 369,
375

Wetzel, 389

Whiteley, 48, 133, 265, 345

Whitman, 373

WOODGER, 372

Worcester, 389
WoRLEY, 50, 238

Yagiu, 84, 87, 188, 284, 287
Yow, 113, 114

YusA, 368

Zeleny, 119
Zhinkin, 262, 263
Zhinkin and Obraztsov, 265
Zingher, 264, 266
ZiNGHER and FisiKOW, 257

SUBJECT INDEX

Abnormal stentors, 274 et seq., 316
Abortive development of primor-

dium, 137, 139
Acetahularia, 195, 297, 317, 319
Acid phosphatase, 266
acrobaticus, 338

Activation and inhibition, \ZS et
seq.

in mincerates, 223

in relation to nodulation, 295

in relation to nuclear behavior,
294, 296

site of, 157
Activation, in reorganizers and

dividers, 152
Activation, timing of, 142 et seq.
Adaptation to stimuli, 21 et seq.
Advantages of stentor, 1 et seq.
amethystinus, 46, 267, 336
Amoeba, 131, 246, 259, 292, 301,

302, 303, 305, 309, 312, 319
Amorphous stentors, 277 et seq.
'* Anarchic field ", 138
Anesthesia of stentors, 238, 240

et seq.
Antimetabolites, 132
Astasia captive, 273
Astoinatous individuals, 320
Attachment, 6, 37 et seq.

in enucleates, 301
auricula, 338
auriculata, 338
auriculatus, 338
Autogamy, 323
Autolysis, 303
Autonomous disorganization, 224,

255
Autoradiography, 303, 353

" Autotomy " of mouthparts, 98,

209, 210
Avoiding reactions, \9 et seq.

of fragments, 21

while swimming, 22
Axial gradient, 195, 201, 246

Basal lamellae, 2>\ et seq.

function of, 33, 238
Behavior, 1 1

of enucleates, 301
Biotypes, 205 et seq.
Blepharisma, 17, 43, 45, 46, 49,

84, 103, 250, 252, 289, 290,

349, 355
comparison with stentors, 361

et seq.
Border stripes of frontal field, 29,

33, 163
Bristles or stationary cilia, 25, 337
Broadening of stentors in LiCl, 256
Bursaria, 13, 75, 91, 159, 160.

169, 170, 220, 259, 280, 299,

303, 305, 342

Cancer and stentor, 3, 215, 276
Cannibalism, 14,15 et seq., 218,322
Carbohydrate reserves, 263

division of, 79

in regeneration, 107, 131

in reorganization, 92

use by enucleates, 302
Carchesium, 242
Case

building, 7, 250

in relation to avoiding reaction,
21
Cell defined, 371, 372
Cells and stentors, 59, 370

400

SUBJECT INDEX

401

Chcetopterus, 373
Chemicals, effects of
acetates, 239, 252
acriflavin, 132
adenine, 132
albumen, 252
AICI3, 234
ammonium acetate, 252
antimetabolites, 132
atropine, 242
azaguanine, 133
CaCU, 237, 240, 243, 245, 246,

251
caffein, 242
Ca(OH)2, 251
CaS04, 251

cations, bivalent, 239, 245, 251
cations, monovalent, 239, 245,

251
chlorides, 252
chloroform, 240, 251
copper salts, 243
CsCl, 239
curare, 242
digitoxin, 234
distilled water, 239
DNA, 58, 132, 289
drugs, 55, 241
ethanol, 237, 248, 252
fatty acids, 240
folic acid, 132
glycerine, 249, 252
guanine, 132
HCl, 240, 249, 251
HgCU, 248

Holtfreter's solution, 253
hydrogen ion, 246
hydroxylamine hydrochlorate,

241
iron salts, 243
Janus green, 246, 250
KAc, 239
KBr, 239
KCl, 239, 240, 243, 244, 245,

246, 251
KCN, 246, 247
K2CO3, 239

KH2PO4, 241

KI, 239, 242

KMn04, 247

KNO3, 239, 251

K2PO4, 239

K2SO4, 239

lactose, 251

LiCl, 239, 251, 252, 255, 256

methanol, 249

methyl cellulose, 241,351

effects of, 100, 145, 234, 349

et seq., 352
methyl tyrosine, 133
methylene blue, 246, 250
MgCls, 234, 237, 245
MgS04, 251
morphine, 242
NaBr, 242
NaCl, 239, 240, 244, 246, 248,

250, 251, 252
NaHCOs, 251
Na2HP04, 251
Nal, 237, 242
NaOH, 249, 251
Na2S04, 243, 251
NH4AC, 239
NH4CI, 239, 251, 252
(NH4)2S04, 251
nicotine, 242
NiS04, 238, 240, 252
oxygen, 265
physostigmin, 242
picrotoxin, 242
potassium ion, 55
quinine, 248
RbCl, 239
Ringer's solution, 243
RNA, 132, 133
RNAse, 133
saponin, 240
sea v^ater, 252, 253, 255
sodium taurocholicum, 245
SrCla, 245, 246
strychnine, 242, 251
sucrose, 253

sugars, 239, 240, 244, 252, 254
sulfates, 252

402

THE BIOLOGY OF STENTOR

Chemicals {continued)

thiocystosine, 133

thymine, 132

tyrosine, methyl, 133

uracil, 132

urea, 240, 252, 253
Chemicals, morphogenetic effects

of, 254
Chimeras, 311

abnormal differentiation in, 317

in Acetabularia, 317

in Amoeba, 319

coeruleiis X introversus, 313

coeruleus X multiformis, 314

coeruleus X niger, 313, 317

coeruleus X polymorphiis , 278,
313, 315 et seq.

coeruleus X roeseli, 313

coeruleus X "X", 312

depigmentation in, 315, 317

shape reconstitution of, 315
Chlorella, 267, 271, 272, 315

{see symbiosis with green algae)

digestion of, 12

in relation to light response, 22
Chromosomes, see conjugation
Ciha, 56 et seq.

coordination of, 238

modifiability of, 38

in relation to pseudopodia, 38

sensory, 25

shedding of, 241
Ciliary beating, 56,163,250

reversal of, 238 et seq.
Ciliary coordination, 15,19
Ciliary molting, 75
Ciliated vacuoles and tubes, 215

et seq.
Clear stripes, 49 et seq.

structure of, SO et seq.
Clones, development of, 344
Clumping of macronucleus, 284,
294

{see macronucleus, coalescence
of)

reasons for, 286

coeruleus (referred to throughout)

species defined, 335
Collecting stentors, 339
Colpidium, 48, 206, 266
Comparison with other ciliates, see
generic names of other ciliates
Concentrating stentors, 345
Condylostomum, IS, 84, 87, 91,
107, 111, 169, 284, 287, 290,
338, 376
Conjugation, 323

re induction of, 324

in coeruleus, 32S et seq.

in polymorphiis, 329 et seq.

possibilities of grafting in, 332

regeneration during, 324

of three individuals, 325
Constriction

in dividers, 72

of macronucleus, 72
Contractile vacuole, 9

in enucleates, 301

origin of, 73
Contractile vacuole system, 40 et

seq.
Contractility, 10, 14, 54, 241

irreversibly damaged, 309
Contraction of cell body, 52, 59
Control of macronuclear behavior,

293 et seq.
Coordination {see metachronal co-
ordination)

of body cilia, 238

in forming daughter cells, 74

in fusion masses, 236

in membranellar band, 232 et
seq.

in membranelles, 234 et seq.

micrurgical analysis of, 27, 33
Corrugations of pellicle, 6
Cortex

importance of, 42

structure of, 42 et seq.
Cortical pattern, 366, 373, 376
Culturing, 342

ionic media, 345 et seq.

of polymorphiis, 346

SUBJECT INDEX

403

Cutting methods, 349

Cya thodinium, 218

Cyclosis of endoplasm, 9, 91 , 301 ,

325
Cystment, 26
Cytopyge, 40
Cytostome, 36

Defecation, 9, 40

extrusion of clots, 246

through posterior pore, 40
Deformities, 226

{see amorphous stentors)

produced by x-ray, 257
Depigmented stentors, 274, 317
Deployment of primordium, 137,

163, 223
Depression in cultures, 100
Desmodexy, 50, 195
Determination

of fission line, 77, 79

of primordium, 170
Didinium, 259
Digestion, 266 et seq.

of cannibal meal, 16

in enucleates, 266, 302

of fats, 266

inhibited by anaerobiosis, 266

of pigment, 17,47

of starch, 266

of symbionts, 267, 269, 272
Dileptus, 120, 259, 286, 342
Disarrangements of pattern, 226

et seq.
Disintegration, progressive

in certain solutions, 246, 247,
248, 253

in UV, 247
Division, 67 et seq.

abnormalities of, 88

acceleration of, 244, 249

of carbohydrate reserves, 79

dispensibility of, 131

endoplasmic role in, 88

final separation, 74

herringbone pattern in dividers,
73

incitement to, SI et seq.

increase of macronuclear nodes,
74

induction of, 82

inhibition of, 257

lacking in large masses, 215

of longitudinal halves, 84

of macronucleus, 295

micronuclei in, 74

multiplication of stripes, 68

new and old parts in, 75

nuclear changes in, 72

number of stripes in relation to,
63

persistence of, 84 et seq

postponed, 88, 131, 156

primordium, 70 et seq.

proportional adjustment in, 124

in relation to differentiation, 89

in relation to size (re cannibals),
215

simultaneous, 346

in spite of injury to fission line,
87

and surface tension, 90

time, 74

unequal, 79, 244

uptake of P^^ in division, 74,
303

without macronucleus, 84, 87

without primordium, 85, 87
Division furrow, see fission line
Doubles and triples defined, 207
Doublets, 208

formation in LiCl, 256

macronuclei, 284

multiplication of, 208

number of stripes, 64

reversion to singles, 208, 210
Doublets and triplets defined, 207,

208
Drugs, effects of, 55
Dwarf stentors, 259 et seq., 305,
320

Ecological considerations, 16, 265
Ecto-endoplasmic ratio, 107

404

THE BIOLOGY OF STENTOR

Ectomyonemes, 50 et seq.
Ectoplasm

etching of, 169
necessary for regeneration, 107
reduction of, 107
totipotency of, 161
Ectoplasm inside, 218
Ectoplasmic structures, constant

size of, 121
Eggs and embryos, compared with
stentors, 59, 77, 89, 90, 213,
221, 245, 256, 294, 309, 372
et seq.
Ehrlich's principle, 249
Electrical stimulus, 24
Electrolyte concentration, 58

and specific conductance, 58
Endomyonemes, 53
Endoplasm

clumping of, 245
coagulation of, 246
composition of, 5S et seq.
endoplasmic vesicles, 58
intimate relation to M bands, 54
streaming movements, 9, 44,

91, 301, 325
transparency of, 320
unimportant in regeneration,

107, 108
vacuolization of, 54, 58, 260,
265, 273, 284
Enemies of stentor, 342
Enucleates

behavior of, 301

contractile vacuole in, 301

defecation in, 302

digestion in, 266, 302

energy metabolism in, 301

fission of, 298

healing in, 299

holdfast in, 301

maintenance of organelles in, 303

regeneration in, 299

resorption of primordium in,

144, 298
shape recovery in, 299
survival of, 304

Enucleation, 357

Epidiniuniy 220

Equivalence of macronuclear nodes,

289
Etching of ectoplasm, 169
Euplotes, 113, 267, 280, 286, 367
Exceptions to induction of primor-

dia by loci of stripe contrast,

191
Excess nucleus, effects of, 304
Exchange of symbionts, 271
Excretory pores, 40, 41
Extension of stentor cell, 10, 14,

54

Fabrea, 138, 159, 164
Fat reserves, 263
Feedback

in diflferentiation, 321

in regeneration, 116
Feeding behavior, 9, 19

{see feeding vortex)

{see food selection)

ingestion, 36
Feeding organelles, 28 et seq.
Feeding vortex, 6, 13
felici, 44, 336
Fission, see division
Fission line, 72

across irregular striping, 230

action of, 76

determination of, 77, 79

nature of, 75, 87

not from cutting the stripes, 88

shifts in location, 79

and shifts of pigment granules,
76
Fluorescent coerideiis, 48, 322

influence of, 48, 49
Fluorescent pigment, 48
Folliculina or FoUiculinids, 13,
46, 47, 160, 164, 169, 250,
266
Food organisms, 347
Food selection, \\ et seq.

basis of, 12, 13 et seq.

pre-oral, 1 3

SUBJECT INDEX

405

Food vacuoles, 8
Fragments

behavior of, 19,23,24

with head only, 125

minimum size, 120
Frontal field, 7, 29 et seq., 139,
163, 165

border stripes of, 163

origin of, 70

striping of, 52
Frontonia, 303, 368
Function of the macronucleus

{see nucleo-cytoplasmic inter-
actions)

in differentiation, 297

lag effects, 297
Funnel, oral, 7
Fusion masses, 205 et seq.

coordination in, 236

effect of orientation on, 214

incomplete oral differentiation in,
215

large, 213 et seq.

reduction of oral valency, 213
et seq.

2-masses, 206

tubes and ciliated vacuoles in,
215 et seq.

gallinulus, 336

Giants not formed, 213, 305

Glaucoma, 355

globator, 338

Gradients

axial, 195, 201, 246

circumferential, 204

in foot formation, 202

metabolic, 246 et seq.

morphogenetic, 202 et seq.
Grafting methods, 354
Granular stripes, 43 et seq., 21 A

as fill-ins, 44, 66, 169
Granules, cortical, 44 et seq.

origin of, 46
Granules, pigment, see pigment
granules

Growth, ()\ et seq.

of fragments, 247

of macronucleus, 74

of primordium, 68

spiral, 66
Gullet, 8, 14, 34 et seq.

ciliation of, 34

eversion of, 15, 34

fibers of, 36 et seq.

myonemes of, 36

pendent fibers, 15

peristalsis, 14

vacuoles, 36

Head fragments, 125
Healing, 226, 240, 246, 257, 299,
352

in enucleates, 299

of membranellar band, 129

of mouthparts, 99, 129
Heat

{see temperature)

perception of, 24

response to, 24
Henneguy-Lenhossek hypothesis,

115
Herringbone pattern in dividers,

73
Heteropolarity

disharmonies in, 199

of patches, 198

resolution of, 199

resorption in relation to, 198
Holdfast, 31 et seq.

in doublets, 208

duplication of, 111

in enucleates, 301

formation in relation to striping,
110

induction, 203

neo-formation at bend of strip-
ing, 111

regeneration, WO et seq.

time for regeneration, 110
Hunger behavior, 7, 12, 14, 15
Hunger divisions, 259

4o6

THE BIOLOGY OF STENTOR

Hypotheses

of morphogenesis, 366

of reorganization, 96 et seq.

Ichthyophthirius, 225, 373
igneus, 46, 49, 52, 267, 269, 271,

272, 281, 301, 323, 336
In situ formation of membranelles,

67, 127 et seq.
Individuality, 89, 208, 379
Induced

division, 82

reorganization, 116, 135, 149
resorption of primordium, 135,
137
Induction

of holdfast, 203

of mouthparts, 173, 174 e^ seq.,

215
of primordium sites in Ble-
pharisma, 364
Inhibition

of division, see division

of new holdfast by old, 110

of oral primordium, 119, 144

et seq.
of regeneration, 13\ et seq.
of regeneration by cold, 117
Injections, 353
Injury

causing primordium resorption,

145
not inducing regeneration, 115

et seq.
not inducing reorganization,
99, 100
Interpenetration of striping, 229,

230
introversus, 46, 52, 59, 313, 314,

336, 344
Island primordium formation, 169

Joining of striping, 227

Kinetodesma, 50
connectives, 50

Kinetosomes, 50, 57, 67, 130,
132, 138, 159, 160, 161, 218,
365, 367 et seq., 376

" Kinety, stomatogenic", 160, 161

km bands, 50 et seq.

Learning, 20 et seq., 24
Leucophrys, 208
Light, in culturing, 346
Light response

dark adaptation, 23

location of sensitivity, 23

racial variation in, 22

in relation to wave length, 23
Location of macronucleus, 281
et seq.

in relation to stripe pattern, 284

variations in, 282
Loci of stripe width contrast, 6,
179 et seq.

absence of, 189

competition between, 190

exceptions to induction of pri-
mordia by, 191

explanatory value of, 194

formation and obliteration, 190

in mincerates, 223

minor, 188 et seq.
loricata, 335
Loxodes, 286

M-bands, 53 et seq.
Macronuclear behavior

control of, 293 et seq.

determined by cell states, 143
Macronuclear chain, regeneration

of, 102, 290, 291
Macronuclear division

autonomous, 72

dependent, 74

in regeneration, 113
Macronuclear extrusions, 310
Macronuclear functions

(see nucleo-cytoplasmic inter-
actions)

in differentiation, 297

lag effects, 297

SUBJECT INDEX

407

Macronuclear increase, 102, 113,

115
{see macronucleus, growth of)
Macronuclear nodes, see nodes of

macronucleus
Macronuclear segments, joining of,

329
Macronuclei of doublets, 284
Macronucleus
{see nucleus)
addition of nodes, 287
in activation and inhibition, 143

et seq.
coalescence of, 91, 113, 114,

284 et seq., 294

in relation to activation, 294
dependence on cytoplasm, 104
division of, 295
division in regeneration, 113
effects of reduction of, 305
effect on transport, 303
elongation of, 73
forked, 287
functionally quiet in division,

303
fusion of, 72
growth of, 74

increase of nodes in division, 74
location of, 281
metabolism when reduced, 307
necessary for proportionate

adjustment of parts, 127
nodal increase in regeneration,

113
nodulation of, 287 et seq., 295
parasitized, 273
polyploidy of, 291
reduction of, 260, 291
rejoining of sections, 285
in relation to amorphous stentors,

278
resumption of typical location,

282, 284
situs inversus of, 284
structure and composition of, 57
Maintenance of organelles, 289
in enucleates, 303

Masses, see fusion masses
Medium

{see pH)

culture, 345

re effects of changes in the, 99
Membranellar band, 30 et seq.

action of, 232

autonomy in development of,
181

basal fiber, 33

base of, 30

behavior of, 249

contraction of, 31

formation of, 34

healing of, 129

in situ formation of, 67

inner lammellae of, 31 et seq

oxidation in, 247

polarity of, 33

proportional decrease in, 75,
101

shedding of, 30, 129, 249, 252

structure of, 235
Metabolism, 259

affected by X-ray, 257

in enucleates, 301, 302

with reduced macronucleus, 307
Metachronal coordination

of body cilia, 232, 238

of membranelles, 232 et seq.

passing around cuts, 50
Metazoa, comparison with, 372,

375
Methyl cellulose

effects of, 100, 349 et seq., 352

solution, 351
Micronuclear division

in fission, 74

in regeneration, 114

in reorganization, 103
Micronuclei, 58

i^see micronuclear division)

behavior and function of, 280

no effect on survival, 304

in relation to division, 82, 87
Microscope set-up, 351
Migration of organelles, 211

4o8

THE BIOLOGY OF STENTOR

Milk

in culturing, 343
ingestion of, 266
Minced stentors, 220 et seq.

activation in, 223
Minceration, the operation, 357
Mitochondria, 45, 58, 302
Molting, ciliary, 75
Morphogenetic effect of chemicals,

254
Morphogenetic gradients, 202 et

seq.
Mouthparts, 28

*' autotomy ", 98, 209, 210
direction of coiling, 174, 180,

188
formation of, 163
healing of, 99, 129
induction of, 173,174 et seq. ,215
nomenclature, 7
resorption in reorganization and

regeneration, 94 et seq.
of reversed asymmetry, 1 80
selective resorption of, 210,

213
size of, 125
Movements, unexplained proto-
plasmic, 44
Mucoid secretion stimulated, 242
muelleri, 333, 337
multiformis, 46, 59, 63, 123, 281,

314, 335, 344
Multiplication of stripes, 63, 65,
66,91,139,163,164,165,169,
179
in division, 68
Myociliary complex, 50 et seq.
Myonemes, anaesthesia of, 241 et
seq.

Necrosis, and regeneration failure,

131
** Neurophanes " and ** neuroids ",

54 et seq.
niger, 46, 47, 49, 267, 281, 313,

323, 336, 344, 345

Nodes of macronucleus

effect of conditions on, 293
equivalence of, 289
shape of, 291
size of, 292, 317
Nodulation of macronucleus, 287
et seq., 295
prevention of, 288
in relation to activation and
inhibition, 295
Nomenclature of feeding organelles,

7
Nuclear behavior

controlled by cytoplasm, 376
in relation to activation, 294,

296
in relation to primordium forma-
tion, 293 et seq.
Nuclear transplantations, 311,

359
Nuclearian, 47

Nucleo-cytoplasmic interactions,
297, 310, 311 etseq., 371, 374
et seq.
{see nucleo-cytoplasmic ratio)
Nucleo-cytoplasmic ratio, 206,
261, 265, 285, 291 etseq., 304,
305, 310
adjustment of, 113,130
in regeneration, 116, 119
in relation to division, 82
in reorganization, 102
Nucleoli, 310, 323, 329
Nucleus

{see macronucleus, micronucleus)
in activation and inhibition, 143

et seq.
degeneration without cytoplasm,

309
effects of excess, 304
effects of reduction, 103, 305,

307
{see reduction of macronucleus)
Nutrient reserves, 263
{see reserves, nutrient)

Ophryoglenids, 160

SUBJECT INDEX

409

Oral inhibition of primordium
formation, 119, 144 et seq.,
200
in mincerates, 223
Oral pouch, 7, 16, 29, 172, 173,
178, 186
closure of, 14
contraction of, 31
Oral valency
defined, 206
in mincerates, 224
reduction of, 208, 213 et seq.
Organelles

(see maintenance of organelles)
adjustment of multiple, 211,

213
migration of; 211
Osmiophilic bodies, 57
Osmotic eflfects, 239, 244
Over-pigmented stentors, 275
Oxidation in membranellar band,

247
Oxidation-reduction studies, 265
Oxytrichia, 285

Paramecium, 19, 20, 22, 40, 41,
48, 206, 259, 266, 267, 268,
269, 271, 272, 280, 324, 366,
368, 375
Parasites of stentors, 273 et seq.
Pattern

cortical, 366, 373, 376

disarrangements of, 226 et seq.

importance of, 148, 377 et seq.
Pellicle, 6, 42

regeneration of, 252

shedding of, 250 et seq.
Peristome, 7

(see membranellar band)
Persistence of division, 84 et seq.
pH

of cytoplasm, 246

effects of, 246

of food vacuoles, 302

of macronucleus, 246

of medium, 346
Phosphatase, acid, 266

Phylogeny, recapitulation of, 73,

164, 176, 281
Pigment

of coeruleus, see stentorin

depigmentation in chimeras,
315, 317

digestion of, 17, 49

fading of, 260

fluorescent, 48

of niger, see stentorol

regeneration of, 250, 252, 275

shedding, 249, 250 et seq.

types of, 46
Pigment granules, 43, 250

abnormalities, of, 274 et seq.

attempt to remove, 252

biochemistry of, 45

their displacement, 43, 44

as fill-ins, 44, 169

in metabolism, 260, 274

as mitochondria, 45

as nutritive store, 44, 107

origin of, 46

regeneration of, 252

shifts of at fission line, 76

transfer between grafted cells, 46
Pigment (granular) stripes, 43

as fill-ins, 44, 169
Pigmentation, degree of

depigmented stentors, 274

as indicating physiological state,
46

over-pigmented stentors, 275
Pinocytosis, 36
Pipette

polyethylene, 341

Spemann, 349
Plasma membrane, 43
Polarity, 195 et seq.

(see also heteropolarity)

adaptive shifts in heteropolar

- parts, 196

in bistomatous primordia, 175

cancellation of, 175

conflicts, 215, 227

fixity of, 195

in fragments, 196

410

THE BIOLOGY OF STENTOR

Polarity (continued)

of membranellar band, 33
in mincerates, 221
and rate of regeneration, 201
reversal of, 200
polymorphus, 48, 56, 120, 234,

239, 246, 259, 263, 265, 266,

267, 268, 269, 271, 272, 273,

274, 275, 278, 281, 287, 313,

323, 329, 331, 337, 344, 345,

346
nocturnal fission of, 346
Postponed division, 88
Preparatory period in primordium

formation, 117, 118, 138, 223
Primordia

abnormal, 184 et seq.
abnormal, direction of bend,

187
joining of, 178

of reversed asymmetry, 187,199
supernumerary, 1 80
synchronization of, 149 et seq.
V-shaped, loop or ring shaped,

170, 186, 191
Primordium

abnormal development, 164 et

seq.
abortive development of, 137,

139
arrested at stage 4, 132
completion without macronu-

cleus, 297
development, 1 59 et seq.
deployment, 137, 163, 223
determination of, 170
in dividers, 70 et seq.
induction of, 193, 194
lengthening of, 169, 171
partial resorption of, 153
re-formation of, 138, 139
repair after injury, 178
resorption of in enucleates, 1 44,

298
resorption due to injury, 145
resorption if nucleus reduced,

307

rerouting of, 152 et seq.

shedding of, 173, 253

stages in development, 161

synthesis in, 257
Primordium site, 160, 179 et seq.

dispensibility of, 161

obliteration of, 210
Proportionality of parts

adjustment of, 67, 75, 101, 123
et seq., 210

necessity of macronucleus for,
127
Protoplasm

changes in state of, 245 et seq.
Protrichocysts, 44
Pseudopodia in holdfast, 38
pygmceus, 337

Quadruplets, 208

Racial differences, 320
Radiophosphorus uptake, 74, 85,

266
Ramifying zone, 65, 66
Rate of regeneration, 204
of holdfast, 110
in relation to level of cut, 247
in relation to presence of hold-
fast, 202
Recapitulation of phylogeny, 73,

164, 176, 281
Recapped regenerators, 146
Reconstitution, 220 et seq.
of cell shape. 111
of minced stentors, 221
Reduction of macronucleus, 291

305
Refringent bodies, 59
Regeneration, 105

blockageof (by chemicals), 132,

254, 255
blockage of (inhibition), 107,

131 et seq., 190, 193, 198
in conjugants, 324
of contractile structures, 54,

116
without endoplasm, 107

SUBJECT INDEX

411

Regeneration {continued)

in enucleates, 299

time for holdfast, 110

inhibition by cold, 107

of macronuclear chain, 102,
290, 291

micronuclear division in, 114

minimum size for, 120 et seq.

nutritive requirements, 107,131

of pellicle, 252

of pigment, 250, 252

of pigment granules, 252

without primordium formation,
115

rate of, 204

repeated, 130 et seq.

simultaneous induction of, 353

stimulus to, 115 et seq.
Regeneration rate, see rate of re-
generation
Regeneration time, \\1 et seq. ^222>

in aboral halves, 118

in doublets, 119

effect of presence of holdfast on,
118

effect of size, 118

effect of temperature, 117

for holdfast, 110

related to extent of ablations, 119
Renucleation, 359

delayed, 309
Reorganization, 9\ et seq.

definition of, 100,154

essence of, 103

explanations of, 98 et seq.

extension of oral resorption in, 1 56

hypothesis, 96 et seq., 104

induced, 149

induction of, 116, 135

stimulus to, 98 et seq.
Reorganizers, identification of, 93
Re-regeneration, 130 et seq.
Reroutingof primordium, 152 ei^eg.
Reserves, nutrient, 263

{see fat reserves, carbohydrate
reserves)

seasonal changes in, 263

Resorption

in anterior rotated on posterior

half, 229, 230
induced, of primordium, 135,

137
of heteropolar parts, 198
of mouthparts, 94 et seq.

prevented by LiCl, 256
of primordium in enucleates,

144
of primordium through injury,

145
of primordium if nucleus
reduced, 307
Respiration, 265

{see chemicals, oxygen)
rate in fragments, 265
in relation to macronucleus, 265
Reversal of ciliary beat, 238 et seq.
Reversed asymmetry of mouthparts,

180, 187
Rhahdomonas incurva, 274
Ribbon bundles, 50 et seq.
RNA, 158, 198, 299, 375
roeseli, 239, 246, 273, 281, 287,

313, 323, 333, 337, 344
Rotation

of anterior on posterior half, 227

et seq.
of left on right half, 211
technique, 353
rubra, 337

Selective resorption of mouthparts,
210, 213

" Self-minceration ", 224, 255

Seinifolliculina, 1 60

Sensory cilia, 25

Separation of grafted stentors, 88

Shape of stentor, 211

determined bv striping. 111,
^ 221

in grafts of 2 stentors, 112
multiple shapes in grafts, 210
reconstitution of, 111, 315
recovery in enucleates, 299
unity favored by mincing, 224

412

THE BIOLOGY OF STENTOR

Shedding

of cilia, 240

of membranellar band, 129,
249, 252

of pellicle, 250 et seq.

of pigment, 249, 250 et seq.

of primordia, 173,253
Similarities in regeneration, reor-
ganization and division, 147,
152
Size

decrease in starvation, 259 et
seq.

of ectoplasmic units, constant,
121

effect of temperature on, 257

no gigantism, 17

of mouthparts, 125

racial differences, 320
Species of Stentor, 333 et seq.
(see names of species)

tested by grafting, 335
Specific gravity, 6
Sphcerophrya stentoris, 273
Spiral growth, 66
Spirostomum, 42, 50, 118, 120,

164, 238, 241, 242, 273, 287
Staining, 52, 348
Starch

digestion, 266

ingestion, 1 1
Starvation, 12, 259 et seq.

in relation to symbiosis, 268

utilization of reserves in, 262
Stenostomum, 47, 131
Stentor genus characterized, 333
Stentorin, 46 et seq.

chemical nature of, 46, 48

digestion of, 47
Stentorol, 47
Stentors as cells, 370
Stimulus to regeneration, 115

transmitted, 137, 141
** Stomatogenic kinety ", 160,

161

Stripe multiplication, 63, 64, 65,
66,91,139,163,164,165,169,
179

in division, 68
Stripes, 6

in doublets, 64

interpenetration of, 229

joining of, 227

number of, 63

in relation to division, 63
Structure, complexity of, 59, 60

in relation to morphogenesis, 60
Stylonychia, 259, 267, 303
Surface precipitation reaction, 245
Survival

of enucleates, 304

not affected by micronuclei, 304

on slides, 348
Swimming, \1 et seq.

backward, 19

of fragments, 1 9

spiral, 1 7
Symbionts

digestion of, 267, 269, 272

exchange of, 271
Symbiosis with algae, 267 et seq.y

315, 335, 336, 337
Synchronization of primordium

development, 149 et seq.
Synchronous division, 346

Techniques, 339
Temperature

and feeding rate, 13

effects on membranellar beating,
234

effect on regeneration, 107, 117

in relation to light response, 24

other effects of, 244, 248, 257,
261, 263, 268, 293, 299, 344
Tetrahymena, 255
Theoretical considerations, 374
Time (period)

for division, 74

for regeneration, 117 et seq.,
223

SUBJECT INDEX

413

Tissue cells compared with stentor,

59
Tolerance, acquired

to chemicals, 248 et seq.

to stimuh, 21, 248
Totipotency of ectoplasm, 161
Toxicity of stentors, 274
Transduction, 317
Transplantations, nuclear, 359
Transport mechanisms, 303
Triplets

multiplication of, 208

reversion to doublets and singles,
208
Tubes and ciliated vacuoles, 215
et seq.

Unequal division, 79

Uptake of radiophosphorus, 266

in division, 74, 85, 303
Ur onychia, 116, 119, 280
UV radiation, effects of, 247

Vacuolization, 239
Valency, see oral valency
Viscosity, 245
Vitamins, 269
Vorticella, 53, 241

Wholeness, 89, 379
X-rays, effect of, 256