Digitized by the IriLuriiyL Archive

in 2007 with funding from

IVIicrosoft Corporation

http://www.archive.org/details/contributionstosOOjennrich

-J

CONTRIBUTIONS

TO THE

STUDY OF THE BEHAVIOR
OF LOWER ORGANISMS

BY

HERBERT S. JENNINGS

Assistant Professor of Zoology, University of Pennsylvania
Research Assistant, Carnegie Institution of [Vashington

Published by the Carnegie Institution

OF Washington

1904

CONTRIBUTIONS

TO THE

STUDY OF THE BEHAVIOR
OF LOWER ORGANISMS

BY

HERBERT S, JENNINGS

Assistant Professor of Zoology, University of Pennsylvania
Research Assistant, Carnegie Institution of Washington

Published by the Carnegie Institution

OF Washington

1904

Carnegie Institution of Washington^
Publication No. i6.

Press of Gibson Bros.,
Washington, D. C.

>

v

zrs

PREFATORY NOTE.

The investigations the results of which are herein set forth were
carried out by the aid of certain grants from the Carnegie Institution
of Washington. The author desires to express his deep sense of obli-
gation for the aid thus rendered. The first five papers were prepared
at the Zoological Laboratory of the University of Michigan, and were
submitted to the Carnegie Institution for publication August i, 1903.
To the third paper some additions were made in February, 1904.
The sixth and seventh papers were prepared at the Naples Zoological
Station, while the writer was acting as Research Assistant of the
Carnegie Institution, and were transmitted for publication in January
and March, respectively, 1904.

LIST OF PAPERS.

1. Reactions to Heat and Cold in the Ciliate Infusoria.

2. Reactions to Light in Ciliates and Flagellates.

3. Reactions to Stimuli in Certain Rotifera.

4. The Theory of Tropisms.

5. Physiological States as Determining Factors in the Behavior of Lower

Organisms.

6. The Movements and Reactions of Amoeba.

7. The Method of Trial and Error in the Behavior of Lower Organisms.

FIRST PAPER

REACTIONS TO HEAT AND GOLD
IN THE CILIATE INFUSORIA.

5 ''

REACTIONS TO HEAT AND COLD IN THE
CILIATE INFUSORIA.

To explain the movements of organisms toward or from a source of
stimulus, we find given almost universally in one shape or another a
certain general formula. This is the schema set forth, with unessen-
tial variations, by Verworn (1899, pp. 500-502) for the orientation of
a ciliate or flagellate infusorian to a one-sided stimulus, and by Loeb
(1897, pp. 439-442) for the tropisms of organisms in general. Essen-
tially, the schema is as follows : An agent acting upon the organism
from one side causes the locomotor organs of that side to contract
either more strongly or less strongly than those of the opposite side.

Fig. I.*

In the former case (Fig. i) the animal is turned away from the source
of stimulus, till it comes into a position in which the motor organs of
the two sides are similarly affected. Then progressing straight for-
ward, it of course moves away from the source of stimulus (negative
taxis or tropism) . If the motor organs on the side most affected are
caused to contract less strongly than those on the opposite side (Fig. 2)

♦Fig. I. — Diagram of a negative reaction of an organism, according lo the
tropism schema. The motor organs which act more eflfectively are shown more
heavily drawn. The more pointed end is the anterior. A stimulus is supposed
to impinge upon the organism a from the direction indicated by arrows; this
causes the motor organs directly affected by the stimulus to beat more strongly,
as indicated by the darker shade. The result is to turn the anterior end in the
direction indicated by curved arrows. The organism thus occupies successively
the positions «, b, c, finally coming into the position d. Here the motor organs
of the two sides are equally affected by the stimulus, hence there is no further
cause for a change of position. The usual forward motion of the organism now
takes it away from the source of stimulus, as indicated by the straight arrow at d.

7

O THE BEHAVIOR OF LOWER OKGAN'ISMS.

the organism is necessarily turned with anterior end toward the source
of stimulus ; then its usual forward movements take it toward the
source of stimulus (positive taxis or tropism). Loeb lays especial
stress on the direction from which the stimulus comes, as it is this
that determines which side shall be most strongly affected by the
stimulus ; otherwise the theory as he sets it forth is essentially like that
held by Verworn. Both these authors apply this schema to the move-
ments of organisms to and from many sorts of stimuli, making it a
general formula for taxis or tropisms. Verworn says (1899, p. 503):

Thus the phenomena of positive and negative chemotaxis, thermotaxis, photo-
taxis and galvanotaxis, which are so highly interesting and important in all or-
ganic life, follow with mechanical necessity as the simple results of differences
in biotonus, which are produced by the action of stimuli at two different poles of
the free living cell.

In the present series of papers the writer proposes to examine the
behavior of a number of lower organisms, in order to determine

O:

Fig. 2.*

whether the reactions to the usual stimuli take place in accordance
with this tropism schema or not, and if not, to determine the real
nature of the reaction method. In this first paper we shall deal with
reactions to heat and cold.

In his recent series of papers on the reactions of infusoria to heat and
cold, Mendelssohn (1902, «, <5, c) develops a theory of thermotaxis in
accordance with the general theory of tropisms, above set forth. In an
earlier paper (Jennings, 1899) *^^ present author, on the other hand,

♦Fig. 2. — Diagram of a positive reaction, according to the tropism schema.
A stimulus coming from the direction indicated by the arrows to the right acts
upon the organism a. The effect of the stimulus is to cause the motor organs
directly affected by it to contract less strongly, as indicated by the lighter shade
on the right side of a. As a result the animal is turned as shown by the curved
arrows, occupying successively the positions a, b, c, d. At d the stimulus
affects the two sides alike, hence there is no cause for further turning, and the
usual forward movement of the organism takes it toward the source of stimulus.

REACTIONS TO HEAT AND COLD.

gave a brief account of the reactions of Paramecium to heat and cold,
according to which these reactions are quite inconsistent with the
tropism schema. As the matter is one of considerable interest, and the
conclusions reached by Mendelssohn and myself seem quite irrecon-
cilable, I have examined anew the phenomena in a considerable number
of infusoria, including Paramecium.

The general phenomena to be explained are well seen in the follow-
ing experiment, taken from Mendelssohn (Fig. 3). An ebonite trough
10 cm. in length and 2 cm. wide is filled with water containing Para-
mecia (a). Now, by proper methods, one end of the trough is slowly
heated to 38°, while the other is kept at the temperature 26°. The

JO' — - i>5-°

Fig. 3.*

Paramecia soon leave the heated region, traveling away from it in a
rather compact mass, and in 5 to 15 minutes they have reached the op-
posite end (6) . If now the temperature at the two ends is reversed,
the Paramecia travel back to the end from which they came. If the
temperature is lowered to 10° at one end, instead of raised, similar re-
sults are obtained ; the Paramecia leave the cold region, as before they

* Fig. 3.— General phenomena of thermotaxis in Paramecium, after Men-
delssohn (1902, a). At a the Paramecia are placed in an ebonite trough, both
ends of which have a temperature of 19°. The Paramecia are equally scattered.
At d, the temperature of one end is raised to 38"^, while at the other it is only 26°.
The Paramecia collect at the end having the lower temperature (" negative
thermotaxis"). At c, one end has a temperature of 25°, while the other is
lowered to 10°. The Paramecia now gather at the end having the higher tem-
perature ("positive thermotaxis").

lO THE BEHAVIOR OF LOWER ORGANISMS.

left the heated region (c). If one end is heated, while the other is
cooled, the Paramecia gather in the intermediate region.

How are these movements to be explained ? Mendelssohn applies
to the phenomena Verworn's schema for the orientation of a ciliate
organism to a one-sided stimulus (see Figs, i and 2). As we wish to
deal thoroughly with this schema, it will be well to set it forth here, as
applied by Mendelssohn to heat and cold, with some fullness.

The temperature being higher at one end of the trough than at the
other, that side or end of the animal directed to the heated end of the
trough has a higher temperature than has the opposite side or end
(see Fig. 4). This difference in temperature causes a difference in the
beat of the cilia. In negative thermotaxis the higher temperature
causes the cilia to contract more strongly, as indicated by the heavier
shade (on the left side) in the figure ; hence the animal is turned
toward the opposite side, or away from the source of heat, until it
comes into a position where the heat acts equally on the two sides.
The Paramecium then of course has its anterior end directed from the

heated region, and its ordinary

swimming carries it away. In
positive thermotaxis, on the
other hand, the lower tempera-
ture causes stronger contrac-
YiG. 4* tions ; hence the cilia on the

side next the cold region con-
tract more strongly, turning the anterior end in the opposite direction.
The Paramecium then swims away, as a result of its normal forward
movement.

Mendelssohn studied the subject primarily from a quantitative stand-
point, determining the optimum temperature, the rate of reaction, the
effects of different temperatures, etc. For this purpose he constructed
a very ingenious and delicate apparatus, which permitted accurate
quantitative results. Relying then upon his valuable papers for these
matters, I have devoted myself entirely to a study of the mechanism of
the reactions. For this purpose an apparatus was used that is similar

* Fig. 4. — Diagram of the thermotactic reaction of Paramecium as conceived
by Mendelssohn, after Mendelssohn (1902, b). The heavier cilia on the left side
show those contracting most strongly and hence those most effective in turning
the organism or driving it forward. In negative thermotaxis the left end would
have the higher temperature, causing the cilia of the left side of the organism
a to beat more strongly. As a result, the organism turns, occupying suc-
cessively the positions a, b, c, d. In the latter position there is no further
cause for turning, and the animal swims directly away from the heated end.
The same diagram illustrates also positive thermotaxis, if the left end is sup-
posed to be cooled below the optimum.

REACTIONS TO HEAT AND COLD.

IC

in principle to that of Mendelssohn, but more easily constructed and
permitting exact observation of the organisms with the microscope,
though otherwise much less elegant than Mendelssohn's. This ap-
paratus is shown in Fig. 5. It consists essentially of three glass tubes,
of 8 millimeters bore, which are supported in a horizontal position,
side by side, by passing them through auger holes in a block of
wood. The tubes are one inch apart and are placed exactly at
the same level, so that a glass slide rests equally on all three. To
the two ends of each of these rubber tubes are attached. The rubber
tubes from one end pass upward into vessels of water raised on a shelf
above the level of the apparatus. From the other end the rubber tubes

Fig. 5.*

pass downward into a waste pail, thus acting as overflow tubes. A
trough, or slide (5), containing infusoria, is placed on the three glass
tubes ; the water in the vessels on the shelf is heated or cooled to any
desired temperature, and is then siphoned off and allowed to flow
downward through the glass tubes. The rate of flow is controlled by
pinchcocks. In this manner heated water can be caused to flow
beneath one end of the slide, cold water beneath the other. The slide
being thus unequally warmed, the reactions of the organisms can be
observed. The rubber tubes leading from the hot and cold vessels can
be interchanged, so that the temperature at either end or the middle of
the slide can be at once changed and made high or low, without the

* Fig. 5. — Apparatus used for testing reaction to heat and cold,
scription, see text.

For de-

12 THE BEHAVIOR OF LOWER ORGANISMS.

slightest disturbance to the slide or trough containing the organisms.
The plan of this apparatus is taken from that of Mendelssohn. It can
be readily constructed in an hour or less, and gives essentially the same
results as Mendelssohn's more elaborate arrangement. With the use
of specially constructed thermometers, such as were employed by
Mendelssohn, exactly the same quantitative work could be done. The
present apparatus has the advantage that it is possible to place a
mirror beneath the glass slide or trough bearing the organisms, and
thus to observe the movements of the latter with the microscope by the
aid of reflected light. With the long-armed Braus-Driiner stand the
whole extent of the trough can be examined at ease, and the movements
of the organisms accurately observed with the stereoscopic binocular.

As a trough I usually employed a glass slide, to which strips of glass
2 mm. in diameter had been cemented, making a trough 3 inches long,
about two-thirds of an inch wide, and 2 mm. deep. In some of the
experiments the trough was covered with a glass plate ; in others it
was left open. Both methods have their advantages and disadvantages.

To realize the exact conditions under which the organisms are act-
ing it is necessary to consider a further question : What is the precise
nature of the stimulating agent in these experiments.? Are we dealing
with radiant heat or with conducted heat.? If we are dealing
primarily with radiant heat, of course currents in the water have
no effect on the distribution of the stimulating agent. If, on the other
hand, we are dealing with conducted heat, if the stimulating agent is
the heated or cooled water, then the conditions are different. Local
currents will cause local variations in the distribution of the heated
water. It is evident, I think, that the second alternative is in all
probability the correct one. Certainly in a bath-tub or in a long
vessel of any sort in which the water is heated at one end and not at
the other, it is possible by producing currents to vary the distribution
of the heated water and to perceive with the hand that it is this heated
water which acts as the stimulus.

The importance of these considerations is evident when we take into
account the fact that the ciliate infusoria are always accompanied by
currents of typical character, having a definite relation to the form and
orientation of the animal's body. As a result of these currents, the in-
fusorian becomes not a mere passive recipient of stimulations, but an
active agent, determining by its activity how and in what part of the
body it shall be affected by stimuli. This may be illustrated by a
diagram (Fig. 6) showing the typical currents produced by the cilia
of Paramecium and the effect produced by these currents upon the
distribution of the heated (or cooled) water. The temperature is con-

REACTIONS TO HEAT AND COLD.

13

ceived to be greatest to the right of the figure and to fall off regu-
larly toward the left, the lines indicating regions of equal temperature.
The last line to the left is marked 28°, this being about the threshold
temperature for the negative reaction of Paramecium, according to
Mendelssohn. The space about the Paramecium (without lines) is at
a temperature below 28° — say at the room temperature — so that it
does not act as a stimulus to cause movement. Now, as the diagram

Fig. 6.*

shows, a cone of water is drawn toward the anterior end of Para-
mecium, from a considerable distance away, necessarily therefore
including water above the threshold temperature of 28°. This cone or

*FiG. 6. — Diagram of currents produced bv the cilia in Paramecium when
the animal is nearly or quite at rest. At right of the line marked 28° the tem-
perature is above the optimum (above 28°) while at the left of this line it is at
the normal or optimum temperature. The heated water first reaches the Para-
mecium at the anterior end on the oral side, passing down the oral groove to
the mouth.

14 THE BEHAVIOR OF LOWER ORGANISMS.

vortex of water passes as a slender stream along the oral groove of
Paramecium to the mouth. Consequently, water heated above the
threshold temperature reaches the Paramecium in this region before it
touches the body elsewhere. The result is thus a stimulation on the
oral side of the body, not elsewhere.

Thus the way in which the organism is stimulated depends not ex-
clusively on the physical laws of the distribution of heat, but upon the
activity of the organism ; and the method of reaction, as we shall see,
is of a corresponding character.

It is not difficult to observe the distribution of the currents above
described if one adds to the water on one side of the nearly or quite
quiet infusorian a cloud of very finely ground India ink. The same
results are obtained with other infusoria ; in Stentor, in Bursaria, and
in some of the larger Hypotricha the results are particularly striking.
Of course if the India ink, or the surface of threshold temperature, is
advancing obliquely to the axis of the infusoria, the results are more
complicated, and a diagram such as we have in Fig. 6 is not easy to
construct. But the result is uniformly to bring the stimulating agent
to the peristome before it reaches any other part of the body. It is not
possible to observe directly the distribution of water of different tem-
peratures, but under the influence of currents this of course follows,
essentially, the same laws as do fine particles suspended in the water.

Another factor which it is important to take into consideration in
studying the effects of heat or other agents on the infusoria is the
greater sensitiveness of the anterior end and oral surface (or peristome)
as compared with the remainder of the body. This the present writer
has demonstrated for the anterior end by direct mechanical stimulation
in a considerable number of infusoria (Jennings, 1900), while Roesle
(1902) has shown a similar high comparative sensitivity for the peri-
stome region. The difference is such that in many cases where the
animal is completely enveloped by a stimulating agent (as by a chemi-
cal, or by warm or cold water) there is reason to think that the
reaction given is due to the stimulation at these regions alone. In
other words, the stimulus reaches its threshold value for the anterior
end and the region about the mouth much before it reaches this value
for the rest of the body. This consideration has an important bearing
on the theory which is frequently maintained, that the directive action
of a stimulus is due to the difference in its intensity on the two ends or
sides of the organism. Even if a stimulating agent acts, fer se^
slightly more strongly on the posterior end than on the anterior end
of an infusorian, there is reason to think that the reaction would be
conditioned entirely by the stimulus at the anterior end, this reaching

REACTIONS TO HEAT AND COLD. 1 5

its threshold value before the stimulus elsewhere produces any effect.
Corresponding statements could be made with relation to the oral and
aboral sides. Of course, owing to the course of the currents above
described, any stimulating agent whose distribution is affected by cur-
rents in the water will usually reach the anterior end and oral side
first in any case.

Summing up, we find (i) that the threshold intensity of a stimulating
agent whose distribution is affected by currents in the water will reach
the anterior end and oral side of the organism before it reaches other
parts of the body ; (2) that the anterior end and oral surface are more
sensitive than the rest of the body, so that the threshold value for^
stimuli is less here than elsewhere.

We may now proceed to an account of the observed method by
which some of the organisms react to heat and cold.

Oxytricha fallax : This is one of the most favorable of the Ciliata
for determining the method of reactions to stimuli, for two reasons,
(i) It is easily procurable in large numbers, occurring in cultures of
the same sort that produce Paramecium, and in equal abundance.
(2) It does not, as a rule, revolve rapidly on its long axis, as Para-
mecium does, but usually creeps with its oral or ventral side against a
surface, so that it is not difficult to observe the relation of the reaction
movements to the differences in the sides of the body.

When water containing a large number of Oxytrichas is placed in
the trough and one end of the trough is heated by passing warm water
through the tube which underlies it, the Oxytrichas gradually pass
toward the opposite end of the trough, forming a dense assemblage
with a rather sharply defined edge toward the heated side. If the end
at first heated is now cooled and the opposite end heated, the organisms
pass back to the end from which they first came. Similar results are
obtained by making one end very cold ; the animals gather in an
optimum region, avoiding both too great heat and too great cold.*
The phenomena are identical with what is to be observed in the case
of Paramecium, save that it requires somewhat longer for the Oxytrichas
to move from one end of the trough to the other, and the progress in a
definite direction is not so steady as we find it in Paramecium.

If the movements of the individuals are observed we find them to be
as follows : Near that end of the trough where the temperature is

* Many quantitative data for various infusoria are given in the valuable papers
of Mendelssohn. As the object of the present paper was not to obtain quantita-
tive data, but to determine just how the animals acted, absolute temperatures
are not recorded. In every case the experiments were so varied as to use at
times temperatures to which a reaction was hardly noticeable; at other times
more extreme temperatures, up to those which were destructive.

/

i6

THE BEHAVIOR OF LOWER ORGANISMS.

raised above the threshold vahie the animals begin to move about
rapidly. At first view this movement seems to be quite irregular, as
Mendelssohn describes it in Paramecium. But exact observation of

Fig. 7.*

the individuals taken separately shows that this movement is not so
entirely irregular as it at first appears. Most of the animals swim

•Fig. 7.— Method by which Oxytricha fallax reacts to heat or cold. The fig-
ure represents one end of a trough or slide, which is heated from the end x. An
Oxytricha in the position i is reached by the heat coming from the end *. The

REACTIONS TO HEAT AND COLD. 1 7

backward, circling at the same time toward the right or aboral side, as
shown in Fig. 7. This lasts but a moment ; then the animal swims
forward, at the same time turning to the right or aboral side. That is,
the individuals give the typical motor reaction, as described in the fifth
of my studies (Jennings, 1900). This reaction is repeated many
times, as long indeed as the animal remains in the heated region. But
of course this movement scatters the animals rapidly. Those that
strike against the end or sides of the trough repeat the reaction above
described, backing, turning to the right, then going forward (Fig. 7 at
8, 9, 10, II). They thus become directed in some other way. Those
that are directed away from the heated region pass into cooler water
and hence no longer give the reaction, but continue their course (Fig.
7 at 13, 14). The result is that the individuals which swim away from
the heated end continue their course, while those starting in any other
direction are stopped and turned (through the motor reaction), until
they too get started away from the heated region. Thus after a time
there is a steady stream of organisms swimming or creeping away from
the heated end, while there is no regular movement in any other
direction. In this manner arises the orientation of the animals, with
anterior ends directed away from the heated region.

The movements of the individuals are exactly as above described
even when the heat is applied some distance from the region where
the animal is found and gradually approaches it from one side. The
animal by no means turns directly away from the heated region, but
repeatedly gives the backing and turning reaction till it is finally mov-
ing in a direction which takes it out of the heated region.

How is this continued backing and turning to be accounted for on
the theory of direct action on the locomotor organs of the two sides as
maintained by Mendelssohn.? This author speaks in the case of Para-
mecium merely of "disordered" movements when the reaction first

animal reacts by turning to the right and backing (i, 2, 3), turning again (3-4),
swimming forward (4-5), backing (5-6), turning again to the right C6-7), etc.,
till it comes against the wall of the trough (8). It then reacts as before, by
backing (8-9), turning to the right (9-10). This type of reaction continues as
long as the Oxytricha is in the heated region, or as long as its movements carry
it either against the wall or into the heated region. When it finally becomes
directed away from the heated region (13), as it must in time if it continues its
reactions, it swims forward, and since it is no longer stimulated, it no longer
reacts. When large numbers of animals react in this way, in the course of
time nearly all become pointed in the same direction, as at 13 or 14, so that a
marked "orientation" is produced. Thus orientation is produced by ''ex-
clusion," due to the fact that the organism is prevented, either by the heat or
the walls of the trough, from swimming in any other direction.

l8 THE BRHAVIOR OF LOWER ORGANISMS.

begins, and thinks this is due to " individual differences, or to ill-
defined internal causes, or perhaps rather to the heterogeneity of the
medium in which they find themselves" (Mendelssohn, 1902, c, p.
492), and that it has nothing to do with thermotaxis proper. This is
typical of many of the statements made concerning the behavior of the
lower organisms ; the movements, so long as they do not agree with
the preconceived schema, are cast aside as disordered, and attention is
called only to the movements that do not conflict with the theory.
Thus Mendelssohn says that this disordered movement '' ceases im-
mediately as soon as the thermotactic action manifests itself" (/. c, p.
492). This is true merely because the thermotactic action is conceived
to begin only after the organism has, through the movements above
described, gotten itself into such a position that it moves away from
the heated region. Of course if all movements except those after
orientation has occurred are thrown out of consideration, the orientation
can be accounted for in any way desired.

In Paramecium, for which alone Mendelssohn attempts to give an
account, based on observation, of the mechanism of the thermotactic
response, the exact character of the movements is undoubtedly diflicult
to observe. This animal is nearly cylindrical in section ; the oral side
is very slightly marked, the movements are rapid, and the animal con-
tinually revolves rapidly on its long axis, so that observation of the
relation of the direction of turning to the differentiations of the body is
very difficult. In Oxytricha and other Hypotricha these difficulties
are almost absent ; the body is markedly differentiated ; the movements
are less rapid, and, most important of all, there is usually no revolu-
tion on the long axis. It is unfortunate therefore that Mendelssohn
included none of the Hypotricha among the organisms which he
studied. With careful observation of the movements of individuals
the mechanism of the reactions is in these animals absolutely clear.

A crucial test of the theory of direct orientation as maintained by
Mendelssohn is given by observation of the direction in which the
animals turn in becoming oriented. Mendelssohn (1902, c, p. 492)
says that after the disordered movements *' the movements executed to
place the body in orientation are rather movements of rotation." This
could hardly be otherwise, but the important question for deciding as
to the nature of the reaction is. How does the rotation take place? Is
it determined by the direction from which the heat comes, as required
by Mendelssohn's theory, or is it determined by the differentiations of
the animal's body ? This point is a decisive one for interpreting the
nature of the reaction. Suppose we have an Oxytricha in the position
a-a, Fig. 8, and heat is applied in such a way as to reach the organism

REACTIONS TO HEAT AND COLD.

19

from the direction indicated by the straight arrows. The heat is supra-
optimal, so that the organism moves away from it. In what direction
will the organism turn in order to reach the position of orientation
h'b} According to the theory of Mendelssohn, that the orientation
is due to an increase of the effective beat of the cilia on the side from
which the heat comes, the animal must turn in the direction indicated
by the arrow x^ and this is of course what one would naturally
expect, since this is the most direct method of becoming oriented.
But as a matter of fact the organism turns in the opposite direction, as
indicated by the arrow jk, thus demonstrating the incorrectness of the
theory that orientation is due to increase of the effective beat of the
cilia on the side from which the heat comes. I have made this obser-
vation hundreds of times, not only upon Oxytricha, but on other
Hypotricha and on infu-
soria belonging to other
groups (see below). The
direction of turning is de-
termined, under the heat
stimulus, by the differen-
tiation of the animal's
body. Oxytricha turns to
the right, without regard
to the direction from which
the heat comes. This is
very striking when the
trough is covered and part
of the animals are creep-
ing on the cover-glass with
ventral side up, while the remainder are creeping on the bottom of the
trough with ventral side down. When stimulated by heat approaching
from one side, all the members of the first group will be observed to
turn counter clock-wise, while those of the second group turn in the
same direction as the clock hands ; that is, each specimen turns toward
its right side.

For becoming completely oriented an animal in the position a-a in
Fig. 8 usually requires a number of reactions, as indicated in Fig. 7,
but the turning in every case is as indicated by the arrowy (Fig. 8).

After it has become oriented with the anterior end away from the
source of heat, Oxytricha by no means maintains this position with
rigidity ; on the contrary the individuals shoot back and forth, in a way
that might be anticipated from the method in which the reaction
occurs. They thus form groups here and there, which gradually move

Fig. 8.— Method of orientation in Oxytricha.

For details, see text.

20 THE BEHAVIOR OF LOWER ORGANISMS.

away from the heat, much as is described by Mendelssohn for Para-
tnccium bursaria. With a large number of individuals a general
orientation is evident, however, after the experiment has been some
time in progress.

If ice water is used as the stimulating agent in place of heated water,
the phenomena to be observed are practically identical with those
above described. The organisms leave the colder region, giving the
same reaction as with heated water. As the cold has the additional
effect of decreasing the movements, many individuals are immobilized
by a very low temperature before they have succeeded in escaping
from it, so that they remain in the cold region. The reaction is thus
less clearly defined than that to heat.

Oxytricka ceruginosa : This organism, though smaller, is in some
respects more favorable than O. fallax for observing the method of
reaction. This is because the individuals are more inclined to swim
freely through the water, so that their progress away from or toward
the heated region is more rapid than in O. fallax. O. ceruginosa^
further, even when moving freely through the water, either does not
revolve on the long axis at all or revolves only very slowly. In con-
sequence of this it is easy to determine the relation of the direction of
turning to the differentiations of the body.

The reaction to heat and cold is in essentials identical with that of
O. fallax^ and this reaction is repeated till the animals are carried into
a region where the temperature is not such as to cause the reaction.
Those that are carried into such a region will of course be swimming
away from the stimulating region ; hence, in a large number of individ-
uals there is an evident orientation, with anterior end directed away
from the source of heat or cold. All the conditions and details as to
the production of this orientation are as set forth above for O. fallax.

Stylonychia mytilus: This large Hypotrichan is still more favorable
for the study of the movements of individuals under the stimulus of
heat or cold coming from one side than are the two species of
Oxytricha. But I have found it less easy to obtain in large numbers,
and for this reason have not chosen it for the detailed description of
the reaction. Where comparatively few specimens are available,
the movements of individuals are easily studied, but there is little
impression of any real orientation, such as one gets clearly when large
numbers are used.

The movements of the individuals are like those described for
Oxytricha fallax. The animal in reacting always turns to its right,
without regard to the relation of this to the direction from which the
heat or cold is coming. With an organism of the large size of

REACTIONS TO HEAT AND COLD. 21

Stylonychia this is very evident. This turning to the right under the
stimulus of heat and cold in Stylonychia has already been described
by Putter (1900), incidentally to his study of the effect of contact
stimuli in this organism.

Stentor ccEruleus : Mendelssohn includes in his paper a note stat-
ing that positive and negative thermotaxis occur in some species of
Stentor, and giving the optimum ; but he made no study of the mech-
anism of the reactions in this animal. Had he done so, it seems to me
that he could not have maintained his theory of the way in which
the reaction takes place.

When one end of the trough is warmed the Stentors near that end
begin after a few seconds to move about more rapidly. In most cases
the movement is as follows : The animals swim backward some
distance, then turn toward the right aboral side and swim forward
(the typical motor reaction). Thus the general effect is as of an
irregular movement in all directions. Those individuals which swim
forward toward the other end of the slide pass out of the heated region ;
hence the motor reaction no longer takes place, and the animals con-
tinue to swim forward. Those which start in any other direction do
not escape from the heated region, and therefore soon give again the
motor reaction, backing and turning again to the right. Thus only
those that swim away from the heated region continue their course ;
the others are stopped and turned until finally they too get started in
the same direction. Therefore, after a period of apparently disordered
swimming, there is an evident orientation of many individuals, with
anterior ends away from the heated region. This orientation is caused
as it were by exclusion ; in animals swimming in any direction but one
the motor reaction is produced, so that only this direction can be main-
tained. After a time, therefore, a large proportion of the individuals
are swimming in this direction, with a common orientation.

Thus the direction in which the animals turn is determined, as in
the Hypotricha, by the structure of the body, and not by the direction
from which the heat comes.

Those outside the region where the heat has reached the threshold
temperature often swim for some distance toward the heated region ;
then arriving at a point where the heat is effective, they give the motor
reaction, backing and turning to the right. They are thus prevented
from entering the heated region.

If the temperature is rapidly raised, the animals may not succeed in
escaping from the heated region until they are injured. In this case
the specimen contracts strongly and swims backward a long time. It
becomes distorted, places the disk against the bottom or other surface,
becomes motionless, and finally dies.

12 THE BEHAVIOR OF LOWER ORGANISMS.

Fixed specimens react less readily to heat than do free-swimming
specimens. They do not orient themselves with reference to the direc-
tion from which the rise in temperature comes. They may remain
extended normally, carrying on the usual activities, after the tempera-
ture has risen beyond the point which sets the free specimens in rapid
reaction. But as the temperature rises they repeatedly bend over into
a new position (bending toward the right aboral side) , then contract
strongly, and finally free themselves from their attachment. There-
upon they behave like other free individuals.

Spirostomum ambiguum : In this large ciliate the reactions to heat
and cold take place in essentially the same manner as is described
above for Stentor and the Hypotricha, so that it is not necessary to
describe the phenomena in detail. The organism reacts to heat or
cold by backing and turning toward its aboral side ; and this whether
the change in temperature is uniform over the entire surface of the
animal or whether it approaches from one side. The movements of the
animal are slow, and under the Braus-Driiner stereoscopic microscope
its method of reaction is very clear. There is little marked common
orientation at any time, however ; this being due to the slowness of the
movements and the frequency of repetitions of the motor reaction.

Bursaria triincatella : In this very large infusorian, in which cer-
tain differentiations of the body are visible even to the naked eye, the
method of reaction to heat and cold is observed with the greatest ease.
But orientation of a large number of individuals in a common direction
is hardly to be noticed, though if Bursaria could be obtained in such
numbers as Paramecium or Oxytricha, perhaps an indication of orien-
tation would be noticeable in spite of the slowness of movement.

Bursaria is very inactive, often remaining quiet for long periods.
It swims slowly, and frequently creeps along the bottom with ventral
side down, but may also swim freely through the water, revolving to
the left. If the temperature of the trough is raised at one end, the
animals in this region that are moving freely through the water swim
backw^ard, turn to the right, and swim forward. This may be repeated
till the organism passes out of the heated region. Rather more
frequently, however, the animal, after thus reacting once or twice, sinks
to the bottom and places its ventral side against the surface. It now
conducts itself in the same manner as do the other individuals in this
situation, as will be described.

The individuals which are resting against the bottom (usually the
majority of those in the trough) react as follows : They begin to swim
backward, keeping the ventral side down and at the same time circling
toward their own right sides. They thus describe rather narrow circles.

REACTIONS TO HEAT AND COLD. 23

This continues until the heat becomes destructive — the animals cease
circling, become quiet, and finally disintegrate. The reaction of those
individuals vv^hich are resting or creeping on the bottom is thus not of
a character to save them from destruction.

Specimens which are by chance moving along the bottom from a
cool region toward the warm region do not escape ; they merely stop
and begin to circle backward to the right when they reach the heated
spot, and continue this till they die.

Thus the reaction of Bursaria to heat, while of the same general
character as that of other infusoria, must be accounted very imperfect,
since it hardly results in orientation at all, and does not preserve the
animals from destruction.

Paratnecium caudatum : * In the second of my studies (Jennings,
1899, pp. 334-336) I gave a brief account of the way in which, ac-
cording to my observations, Paramecium reacts to heat and cold.
From my more recent studies I can confirm this account. But as
Mendelssohn has recently come to different conclusions for the tempera-
ture reaction of this animal, and as he misunderstands certain points
in my brief description, it seems desirable that I should supplement the
account previously given in order to make it clear.

Paramecium reacts to heat and cold in essentially the same manner
as is described above in detail for Oxytricha. When the higher or
the lower temperature advances from one side the animals swim
backward, turn toward the aboral side, and swim forward again.
They continue this until the movement brings them into a region of
more moderate temperature. Paramecium reacts more readily than
Oxytricha, the reactions are repeated at shorter intervals, and the
movements are more rapid, so that a common orientation of many
individuals swimming away from the region of higher or lower tem-
perature is more quickly produced and is more striking to the eye. It
results farther from this more rapid movement, as well as from certain
other factors, that the method of reaction in Paramecium is much less
easily observed than in any of the other infusoria described. Indeed,
Paramecium is one of the most unfavorable forms obtainable for a
study of reaction methods, and it is, I believe, due largely to the fact
that this animal is usually employed for such study that progress has

♦The common Paramecium, which appears everywhere in immense numbers
in decaying vegetation, receives from different authors sometimes the name
Paramecium aurelta^ used by Mendelssohn ; sometimes the name given above.
I use the name caudatum because it appears to me to be the correct one, but
there is no reason for considering the animals thus differently denominated to
be really different.

34 THE BEHAVIOR OF LOWER ORGANISMS.

been so slow in appreciating the real nature of the reactions of the in-
fusoria. If Stylonychia or Oxytricha or any other of the Hypotricha
had been taken as the usual type for study on reactions, many of the
theories now maintained could never have been put forth. The body
of Paramecium is comparatively little differentiated, so that it is diffi-
cult to distinguish oral and aboral sides, and, to multiply this difficulty
many times, the animal revolves rapidly on its long axis, so that oral
and aboral sides never retain for two successive instants the same
position. It is not wonderful, therefore, that the method of reaction
by turning toward the aboral side was not observed in the first investi-
gations on Paramecium and that many still find it difficult to observe.
Nevertheless, it was on Paramecium itself that this reaction method was
first observed (Jennings, 1899), and its existence was confirmed later
on the organisms where its observation presents no difficulties. Aside
from the direct observations of the method of reaction, the following
facts throw light on the way in which the collections take place.

As described in the second of my studies
(Jennings, 1899, pp. 314, 315), the collect-
ing of Paramecia in regions of optimum
temperature may be produced in the follow-
ing manner : The infusoria are mounted in
^ water which is above the optimum temper-

ature (say 30°) on a slide beneath a cover
glass supported at its ends by glass rods. Into this slide is introduced
with the capillary pipette a little cooler water (say at 24°), which
covers a small circular area in the center of the slide. Very soon the
Paramecia have collected in this region till a dense group is formed.
The same result may be obtained by placing a drop of ice water on the
top of the cover glass of a slide of Paramecia which has been warmed
considerably above the optimum temperature. (Fig. 9.)

Are these collections due to the orienting of Paramecium by the
heat, as maintained by Mendelssohn for thermotaxis in general.? Ob-
servation shows that they are not ; that on the contrary the Paramecia
gather in the optimum region in the same manner as they gather in a
drop of weak acid, as described in my studies. The Paramecia on the
heated slide are swimming rapidly in all directions. They do not
change their course or become oriented in the least when a spot in a
certain part of the slide is cooled. But as a consequence of their

*FiG. 9. — Collection of Paramecia due to the reaction to temperature change.
The slide rests on a vessel of water at a temperature of 45"^. An elongated drop
of ice water is placed on the upper surface of the cover glass. The Paramecia
quickly collect beneath the drop of ice water.

REACTIONS TO HEAT AND COLD. 25

rapid movements many of them by chance enter the cooler region.
They do not react at all as they enter, but continue across. On
coming to the other side of the drop, however, they do react, by back-
ing and turning toward one side (the aboral). They react whenever
they come to the boundary of the cooled region ; hence they do not
leave it. In every respect their behavior is like that seen when Para-
mecia collect in a drop of weak acid, and I believe there is no longer
anyone who holds to the orientation theory for the gathering of Para-
mecium in chemicals.

As in the case of chemicals, it may be demonstrated to the eye in the
following manner that the method above described suffices to account
for the gatherings. On the upper surface of the cover glass is marked
a small ring in ink. By confining the attention to this ring it is easily
seen that in the heated preparation of Paramecia many individuals
cross the ring every instant, so that, if these could all be stopped in
the ring, a dense aggregation would soon result. Then the region
within the ring is cooled by placing a drop of ice water on the cover
above it. The Paramecia continue to swim just as before, save that
they no longer pass out of the ring after swimming in, as they did at
first. In this way a dense collection is soon formed.

Mendelssohn (1902, ^, p. 487) finds it inexplicable w^hy the Para-
mecia should form dense aggregations at the optimum temperature.
He says that they execute " only some insignificant movements " in
this region, not swimming away. On the theory of thermotaxis held
by Mendelssohn this is perhaps inexplicable, but this, it seems to me,
is only because the theory is incorrect. Such collections are due to
precisely the same factors as the rest of the reaction to heat and cold
and are clearly intelligible when the nature of the reaction, as described
above, is taken into consideration.*

In a former paper (Jennings, 1S99, p. 336), after giving a brief
account of the reaction method above described, I pointed out that this
method does not demand a sensitiveness to such minute differences in
temperature as does Mendelssohn's theory, and that therefore the sensi-
tiveness to temperature differences may have been overestimated.

* Mendelssohn (1902, b, p. 487) supposes that I would explain these gatherings
at the optimum temperature through the collection of Paramecia in CO 3 pro-
duced by themselves, and shows that this would not account for the phenomena
observed in these cases, though he confirms the fact of the collections in COg.
But I have by no means maintained that such collections can be produced only
by COg ; on the contrary, I have given an account of many different agencies
that will give rise to such collections, and have especially described the fact that
collections are formed in a warmed region through exactly the same reaction by
which they are formed in CO^. (Jennings, 1899, P* S^SO

26 THE BEHAVIOR OF LOWER ORGANISMS.

Mendelssohn (1902, a, p. 406) misunderstands my ground for this
statement. He supposes that I hold that the Paramecia do not react
to differences in temperature less than that existing in a certain illus-
trative experiment, where one end of the slide was resting on ice, while
the other was heated to 40°. This experiment was purely for the
purpose of bringing the phenomena of thermotaxis concretely before
the attention of the reader ; its details had no special significance. I
have not the slightest reason for doubting the entire accuracy of the
quantitative experimental results set forth by Mendelssohn, and consider
them a most valuable addition to our stock of exact data. But the
calculation of the sensitiveness of the organisms concerned, from these
experimental results, involves a certain interpretation as to the reaction
method, and it was this interpretation that I called in question. Men-
delssohn, in accordance with his general theory, holds that the reaction
is due to the difference in temperature between the two
ends of the organism, and he calculates that this difference
in temperature could amount, in the case of Paramecia,
to but 0.01° C. According to the reaction method which
I have described above, however, it is not the difference
in temperature between the two ends of the same indi-
vidual that causes the reaction. Consider a slide cooled
below the optimum at the end a ; above the optimum at
the end b (Fig. 10), the optimum temperature for the
Paramecia being between the lines x and y. The animal
p ^ may swim a' considerable distance from a position y^ at

one side of the optimum, to a position a?, at the other side
of the optimum, before it reacts (by backing and turning, etc.) at all.
We have no ground for maintaining then that it perceives any less differ-
erences in temperature than that between the lines x and jv, and this
difference will be much greater than that between the two ends of the
animal. A similar diagram could be made for the case where the tem-
perature is raised or lowered only at one end of the slide. It seems to
me correct, therefore, that the sensitiveness to temperature differences
has probably been much overestimated. The only way that it could
be estimated would be by observation of individuals to determine the
extent of the stretch x-y over which they pass before reacting, and to
calculate the difference in temperature between the ends of this stretch.
It would of course be very diflScult to do this with accuracy.

Mendelssohn's view that it is the difference in temperature between
the two ends of the same individual that determines the reaction is not

* Fig. 10. — Diagram illustrating conditions necessary for determining the sen-
sitiveness of Paramecia to differences in temperature. See text.

REACTIONS TO HEAT AND COLD. 27

only rendered inadmissible by the reaction method above described,
but it is rendered a priori improbable by certain other considerations.
First we have the fact that the anterior end is much more sensitive than
the posterior. Of course it is impossible to measure this difference in
sensitiveness, yet the experiments with mechanical and chemical stimuli
show that it is great. In many infusoria, while the slightest touch at
the anterior end causes a pronounced reaction, it requires a strong stroke
at the posterior end to produce even a slight reaction. (See Jennings,
1900, pp. 238, 243, 251.) Owing to the much greater sensitiveness of
the anterior end, it is probable that, with the posterior end but 0.01°
warmer than the anterior, the reaction, if any, would be due to the tem-
perature of the anterior end. In other words, there is reason to suppose
that the threshold temperature for the anterior end would be considera-
bl)' lower than that for the posterior end. If this is true the usual tem-
perature reactions would be throughout due primarily to stimulation
at the anterior end ; and the reaction, as we have seen, is of just the
character which would be expected from this. The first stage in
the reaction is to swim backward^ and this is true also when the animal
is dropped directly into water of uniformly high or low temperature, so
that the temperature of the anterior end is no greater than that of the
posterior end. There is no explanation for the swimming backward
under these circumstances on the theory that accounts for thermotaxis
by the different temperature of the two ends.

A second factor which must be taken into consideration relates to the
currents produced by the cilia of the organism itself. As shown above
(p. 13) , the water of a higher temperature (supposing that we are deal-
ing with the reaction to heat) , would as a rule first reach the anterior
end and pass at once down the oral groove, on the oral side (Fig. 6) .
The natural result therefore would be a turning toward the opposite or
aboral side, and this is exactly what we find takes place. We should
therefore not expect the organism to turn directly away from that end
of the trough from which the heat comes, for the heated water may not
reach the Paramecium from that side at all.

As will be seen, the facts adduced in the last paragraph are not incon-
sistent with the idea that the organism turns directly away from the
side stimulated. It is the oral side which is, as a rule, stimulated, and
the organism turns toward the aboral side. We seem thus to obtain
a most gratifying union of two apparently opposed views. But the
reactions to certain other stimuli do not admit of such a union. This
is notably true of the reactions to mechanical stimuli, as shown in a
previous paper (Jennings, 1900), and of the reactions to light, to be
described in the following paper.

28 THE BEHAVIOR OF LOWER ORGANISMS.

SUMMARY.

The ciliate infusoria react in the same manner to heat and cold as
to most other classes of stimuli ; the response on coming into a region
where the temperature is above or below the optimum is by backing
and turning toward a structurally defined side, followed by a movement
forward. This reaction is repeated as long as an effective supraoptimal
or suboptimal temperature continues. The result is to prevent the
organisms from entering regions of marked supraoptimal or suboptimal
temperature, and to cause them to form collections in regions of opti-
mal temperature. The common orientation of a large number of
individuals sometimes produced in this way is an indirect result of the
method of reaction. Since movement in any other direction than a
certain one is stopped, the organisms after many trials come into this
direction. Orientation is therefore by " exclusion," or by the method
of trial and error. In many of the organisms orientation is not a
noticeable feature of the reaction.

SECOND PAPER

REACTIONS TO LIGHT IN CILIATES
AND FLAGELLATES.

29

REACTIONS TO LIGHT IN CILIATES AND

FLAGELLATES. ^

In the reactions to light we are dealing with a stimulating agent
which differs in one very important respect from chemicals and from
heat or cold. The distribution of the agent with which we are con-
cerned is not affected by the currents of water produced by the organism ;
hence there is no tendency for one side or part of the organism to be
more strongly affected than the rest, as was found to be the case for
chemicals and for heat and cold. This peculiarity light shares with
the electric current and with radiant heat. The conditions demanded
for immediate orientation through direct action of the agent on the
locomotor organs, in the manner required by the general theory of
tropisms as set forth in the foregoing paper (p. 7), are therefore present.
In a recent paper Holt & Lee (1901) have attempted to show that
the reactions of organisms to light actually take place in accordance
with this theory.

We shall examine the reactions to light in Stentor ccerul^us and in
certain flagellates, in order to determine whether they take place in
accordance with the tropism schema, and, if not, just how they do occur
and on what factors they depend.

THE CILIATA.
STENTOR C^RULEUS.

As is well known, very few of the ciliate infusoria react to light.
Light reactions have been described by Engelmann (1882, a) for several
chlorophyllaceous ciliates ; by Verworn (18S9, Nachschrift) for Pleuro-
ncma chrysalis; and by Davenport (1897) ^"^ Holt & Lee (1901) for
Stentor cceruleus. In none of the ciliates have the reactions been
described in sufficient detail to enable us to determine their exact
nature.

In Stentor cceruleus the reaction to light manifests itself in the
culture dish by the usual aggregation of the organisms at the side away
from the window. If a number of Stentors are removed to a watch
glass or trough, and this is placed near a window or other source of
light, most of the Stentors are soon found on the side of the vessel away
from the light. If one-half of the glass is shaded by a screen, most of

31

32 THE BEHAVIOR OF LOWER ORGANISMS.

the Stentors are soon found in the shaded half. 6". ccEruleus thus shows
the phenomenon usually called negative phototaxis.

It is to be noted that not all the Stentors are to be found on the side
away from the light, or in the shaded half of the vessel. On the con-
trary, a considerable fraction of the whole number will usually be found
swimming about in all parts of the dish, or at rest in the lighted portion.
The light reaction is thus somewhat inconstant, and varies among
different individuals. It varies considerably with Stentors of different
cultures ; from some cultures almost all the individuals show it, while
from others it is barely noticeable. This variability and inconstancy
run through all manifestations of the light reaction in Stentor.

A word further needs to be said as to the behavior of individuals
which are not free-swimming, but are fixed by the posterior end. Such
individuals do not react at all to light. When light is thrown on them
they remain in the positions in which they are found at the beginning,
neither contracting nor in any way changing their position. No matter
whether the light is weak or strong, and without regard to the direction
from which it comes, fixed Stentors give no reaction and show no
orientation with reference to light. The contact reaction apparently
inhibits the light reaction completely. We shall therefore omit the
fixed individuals from consideration in the remainder of the account,
confining attention to the free-swimming specimens.

The typical motor reaction of Stentor, by which it responds to most
stimuli, is as follows: The Stentor stops or swims backward a short
distance, then turns toward the right aboral side, and resumes its for-
ward motion. This is the reaction which is produced by strong
mechanical stimuli, by heat, and by chemical stimuli, acting upon the
anterior end or upon the body as a whole.

How is the reaction to light brought about .'^ To answer this ques-
tion it is best to arrange experiments in such a way as to distinguish
as far as possible the effect due to unequal illumination of different
areas from the effect due to the direction from which the light is coming.

In order to produce strong differences in illumination in different
areas of the space in which the Stentors are found, a flat-bottomed
glass vessel containing many Stentors in a shallow layer of water was
placed on the stage of the microscope, in a dark room. From beneath
strong light was sent upward through the opening of the diaphragm,
by throwing the light from the projection lantern (using the electric
arc light) on the substage mirror. By this the light was directed up-
ward through the vessel containing the Stentors. Thus a small,
definitely bounded circular area was illuminated, while the rest of the
vessel remained in darkness. A black screen was usually placed over
the diaphragm opening of the microscope in such a way as to shade one-

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES.

33

half of the circular area, making a sharp line (x-y, Fig. 1 1) dividing the
light from the darkness. A mirror was placed above the microscope,
inclined in such a position as to project the image of the Stentors, very
much magnified, on the ordinary vertical screen used for receiving
lantern slide viev^s. Thus the behavior of the Stentors could be studied
with great ease on the screen.

The heat from the lantern was cut out, so far as possible, by placing
between it and the mirror of the microscope a glass cell three inches
thick, filled with cold water. In this manner the heat was excluded to
such an extent as to fall below the threshold for the stimulation of
Stentor by heat. This was demonstrated by comparing the reactions
of Stentor with those of Paramecium. Stentor is less sensitive to
changes in temperature than is Paramecium ; this was clear in my ex-
periments on the reaction to heat. Par-
amecium does not react at all on passing
into the area illuminated by the lantern,
but swims about indifferently in both the
dark and the light parts of the dish, show-
ing that the heat produced is below the
threshold for Paramecium ; it must then
be below the threshold for Stentor.

The free Stentors in the unlighted part
of the vessel swim about at random.
Many individuals thus come by chance to
the line x-y, Fig. 1 1 , where they would
pass into the lighted area. These at
once back a little, then turn toward the
right aboral side, and swim forward

again. The turning toward the right aboral side is usually through an
angle sufficient to direct the Stentor away from the lighted area (see
I, 2, 3, 4, Fig. ii) ; if it is not, the Stentor repeats the reaction until,
after one or two trials, it swims into the unlighted region.

Many of the individuals react as soon as the anterior end reaches the
lighted area, so that less than one-fourth of the body is in the light.
This shows that light falling upon the anterior end alone is sufficient
to cause the reaction.

A few specimens swim completely into the lighted area, then react

♦Fig. II. — Method of studying the manner in which Stentor reacts to light.
The figure shows a circular area, illuminated from below, with the light cut oft
from the left side hy a dark screen, the line x-y separating the light from the
dark area. The Stentors collect in the dark area. The reaction of a specimen
which comes to the line x-y is shown at i, 2, 3, 4.

34 THE BEHAVIOR OF LOWER ORGANISMS.

in the manner above described. In such cases the nature of the reac-
tion is seen with especial clearness, the entire animal being projected
on the screen and the differentiations of bodily structure (mouth, oral
and aboral sides, etc.) being conspicuous. Specimens which swim
completely into the lighted area are usually compelled to react two or
more times before they escape from the lighted region.

When the light is cut off entirely the Stentors distribute themselves
throughout the dish. If the light is now admitted from below, the
unattached Stentors in the lighted area react by swimming backwards
a certain distance, turning toward the right aboral side, then swimming
forward again. This reaction is repeated frequently until after an
interval the Stentors are carried by these movements outside the lighted
area. They then cease to give the reaction. The reaction, under these
conditions, is thus the same as that produced when Stentors or Para-
mecia are subjected to other adequate stimuli, as when they are placed
in a chemical or dropped into very warm or very cold water. The
result of the reaction is, in every case, to remove the organism from
the sphere of action of the stimulus. When the stimulus is light this
result is produced in exactly the same way as when the stimulus is
heat or cold or a chemical.

The same results may be obtained by lighting the vessel containing
the Stentors directly from above and shading one portion with a screen.
The Stentors remain in the shaded region, responding by the motor
reaction above described when they come to the lighted area. With a
favorable culture the experiment succeeds even when the source of
light is comparatively feeble, as when an ordinary incandescent electric
light is used as the source of illumination.

The results so far show that a sudden increase in the intensity of
illumination induces in Stentor a reaction which is of the same
character as the reaction to other strong stimuli. Such a sudden
increase may be due either to the passage of the Stentor from a dark
to a light region, or to a sudden increase in the brightness of the light
which falls upon the animal. The general effect of the reaction is to
prevent the Stentor from entering a brightly illuminated area, or to
remove it from such an area.

We may now arrange the conditions so that the light shall come
from one side, while at the same time differences in illumination shall
exist in different regions. This may be done by illuminating the vessel
containing the Stentors from the side, then covering one portion of the
vessel with a screen.

The organisms are placed before a lighted window, or an incandes-
cent electric light, in a vessel with a plane front (Fig. 12). One-half

REACTIONS TO LIGHT IN CII.IATES AND FLAGELLATES.

35

of the vessel is then cut off from the light by a screen (5), the shadow
of which passes across the middle of the vessel containing the Stentors.
One side of the vessel is thus in the light, the other in the shadow,
and these two regions are separated by a sharp line (Fig. 12, x y).

The Stentors are soon all collected in the shaded side of the vessel.
Here they swim about freely in all directions, but do not cross the line
into the lighted portion. Now, by focusing the Braus-Driiner on this
line, the behavior of the individuals on reaching it may be observed.

It is well to examine

the conditions in this case ».

with care, as they present
opportunities for a pre-
cise and crucial test of
the theory that the reac-
tion to light is due to a di- -*

rect orientation through
the falling of light on one
side of the organism (pho-
totaxis or phototropism

in the strict sense, as de- *■

fined by Holt & Lee).
In the lighted portion of
the vessel the rays of light
come from a certain direc-
tion, as indicated by the ^
large arrows (Fig. 12).

In the shaded region there ^

is not enough light to pro-
duce orientation, the ani-
mals swimming in every *'
direction. On passing
from the shaded region

across the line x-y into the lighted region, the animal should (according
to the tropism theory) become oriented. According to the theory of
negative phototaxis by direct orientation due to differential action on

Fig. 12.*

♦Fig. 12.— Method of testing the manner of reaction to light in Stentor. The
large arrows show the direction from which the light rays come. A screen {s)
cuts off the light from half the vessel, leaving a line {x-y) separating a shaded part
from a lighted part. The Stentors collect in the shaded part, here swimming
about without orientation. At a (i, 2, 3, 4; we see a diagram of the reaction
required by the tropism schema when the organism swims across the line x-y,
while at ^ (i, 2, 3, 4) we have a diagram of the reaction as actually given under
these conditions.

36 THE BEHAVIOR OF LOWER ORGANISMS.

the two sides, the animals on crossing the line should become oriented
by turning directly away from the source of light, as shown in the
diagram (Fig. 12) at a. The animal would then be expected to swim
in the direction x-y as shown by the specimen «, i, 2, 3, 4.

It cannot be held that the real source of light for the Stentors is that
reflected from the bottom or sides of the dish in the lighted region, and
hence coming on the whole from a direction perpendicular to the line
xy^ for the behavior of the Stentors shows that this is not the case,
A Stentor in the shaded region, close to the line x-y^ as at c, Fig. 12,
receives whatever light there may be thus reflected exactly as it does
after it has crossed the line, yet it shows no reaction and does not
orient itself in any way. On the other hand, as soon as it crosses the
line x-y^ so as to receive the light coming from the window, it reacts
strongly, as we shall see. It is thus clearly the light from the window,
coming in the direction shown by the large arrows, that causes the re-
action ; hence the Stentor ought, according to the direct orientation
theory, to orient itself in the line of these rays.

When a Stentor, swimming at random, reaches the line x-y^ it
reacts by stopping suddenly, then turning toward its aboral side,
then swimming forward. It thus swims about until its anterior end is
again within the shadow, where it continues to swim forward (Fig. 12,
3, I, 2, 3, 4). Often the first reaction is not sufficient to direct it into
the shadow ; in this case the reaction is repeated ; one to three reactions
almost invariably bring the Stentor back into the shadow. It has no
particular orientation in the shadow, but swims in whatever direction
it happens to be headed.

Very frequently the animals react when the anterior end alone has
crossed the line, so that less than the anterior half of the body is lighted.
In other cases the animal swims completely across the line, sometimes
for a distance greater than its own length, into the light, before it reacts.
In any case the reaction is that above described.

Does the Stentor, when it turns on entering the light, always turn
away from the source of light, as the theory of direct orientation
requires ?

At the moment of crossing the line into the light the Stentor may
occupy various positions. It will be well to note specifically the re-
action in certain of these positions, as we obtain here the observations
which furnish an exact and crucial test of the direct orientation theory.

I. The Stentor may reach the line with the aboral side directed
toward the source of light (Fig 12, b). It therefore turns (as usual)
toward its aboral side. It thus swings its anterior end toward the
source of lights in the direction opposite that required by the direct

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES.

37

orientation theory. This observation was made repeatedly in a very
large number of cases ; not a single exception to it was observed. The
swinging of the anterior end is continued past the point where the light
falls directly upon it until the animal is directed again into the shadow,
as illustrated in the diagram (Fig. 12, ^, i, 3, 3, 4).

2. The Stentor may reach the line with the aboral side directed away
from the source of light. In this case it turns (as usual) toward the
aboral side, thus swinging its anterior end away from the source of light.

3. The Stentor may reach the line with the aboral side directed
upward or downward or in some intermediate position. In every case
it turns toward the right aboral side, in whichever way this is directed.

Fig. 13 *

The writer wishes it to be understood that the foregoing statements
as to the direction in which the animal turns are presented, not merely
as interpretations in accordance with a certain theory, but as direct,
unequivocal observations, many times repeated. Thus, on passing from
a darker to a lighter area, even when the light comes from one side,
the Stentors react merely to the difference in illumination, without
regard to the direction from which the light comes. The direction of
turning is determined throughout by an internal factor, not by the side
of the animal on which the light falls, nor by the direction of the rays of
light. We have put the theory of orientation by direct differential

♦Fig. 13. — Another method of testing the manner in which Stentor reacts to
light. For a side view of this apparatus, see Fig. 14. Light comes from the
left side, in the direction indicated by the arrows. A screen (5) is interposed
between the source of light and the vessel containing the Stentors. This screen
is of such a height (as illustrated in Fig. 14) that it cuts off the light from the
half (/I) of the vessel next to the window, leaving the other half {B) lighted.
At c (I, 2, 3, 4, 5) is seen the reaction method of a specimen which swims across
the line x-y, separating the shaded half yl from the lighted half 5.

3S

THK BEHAVIOR OF LOWER ORGANISMS

action of the light on the two sides of the animal to a precise test, and
found it to be incorrect.

The same result is brought out in perhaps a still more striking
manner by the following method of experimentation : A vessel con-
taining Stentors is placed on a dark background near a source of light
(a window or an incandescent electric lamp). The light thus comes
from one side and a little from above. An opaque screen is placed
between the window and the vessel containing the Stentors, of such a
size and in such a position that the top of the shadow of the screen falls
across the middle of the vessel on the line x-y (Fig. 13 ; see also Fig.
14). Thus the half of the vessel next to the window {A) is darker
than the farther half {B)^ and the Stentors collect in this shaded half
After some time scarcely a specimen is found in the lighted part of the
vessel away from the window. The conditions in this case are illus-
trated in the side view (Fig. 14).

The exact behavior of the Stentors in
darkened portion of the vessel is then
3ied b}' focusing upon them the Braus-
iner microscope. The Stentors within
shaded area are not oriented nor gath-
ered in any particular region,
but swim about at random.
When one of the specimens
comes in its course to the
line AT-j^ (Fig. 13), separating
the darkened area from the

Fig. 14.*

light, it responds to the sudden light which falls upon it from the
window by giving the motor reaction, turning to the right aboral side
and swimming back into the shaded region. Often the reaction occurs
as soon as the anterior end of the Stentor has crossed the line x-y^ so
that the entire Stentor does not pass out into the lighted area. In other
cases the specimen crosses the line x-y completely before the reaction
occurs, so that the entire body is illuminated. It then reacts in the
usual manner, turning toward the right aboral side, so that it is headed
toward the shaded region ; thus shimming back across the line (Fig.
13, c). After returning into the shaded region the animals swim about
at random as before.

What is the reason for the return of the Stentor into the darkened
area after it has crossed the line into the light region.?

* Fig. 14. — Sectional view, from the side, of the conditions shown in Fig. 13.
The arrows show the direction of the light rays. The region from 5 to * is
fehaded by the screen s.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 39

By SO doing it swims toward the window, thus in the direction from
which the strongest light is coming. According to the theory of photo-
taxis as due to the direct action of the light on the motor organs of the
animal, this movement is inexplicable. Thus, in the analysis of this
theory given by Holt & Lee (1901), it is shown that in the case of a
negative organism, such as Stentor, light of supraoptimal intensity,
like that coming from the window, must be assumed to cause increased
contraction of the cilia. After the organism has passed across the line
x-y, or while it is passing across this line, it has the anterior end directed
away from the source of light ; according to the tropism theory this
is a stable position and should not be changed. For, supposing the
organism swerves a little toward either side, the cilia on that side will
be more strongly affected by the light, so that the animal will at once
be turned back into the position of equilibrium with anterior end directed
away from the light.

Nevertheless, under these circumstances the organism does turn and
swim back into the darkened area. An explanation for the apparent
movement of a negative organism against the direction of the light rays
is sometimes given in the following form : The light from the window
is said to fall upon the side or end of the dish farthest from the window
and is reflected back, so that the chief source of light for the Stentors
is not the window, but the side of the dish opposite the window. The
animal therefore becomes oriented with relation to this source of light
and swims away from it.

Comparison of the movements of the Stentors in the darkened area
A with those in the lighted area B shows that this explanation can not
possibly be correct. Consider an individual at the point 3, Fig. 13,
which turns and swims toward the window into the dark region. It
is affected by light from two sources, (i) from the window, (3) reflected
from the side opposite the window. According to the above theory
the turning is due to the fact that the light from the opposite side is of
greater strength than that from the window (in itself a most improbable
suggestion). Compare this Stentor b with an individual at «, in the
darker region. This animal receives no direct rays from the window,
yet does receive the reflected rays from the opposite side. If these
reflected rays are sufficient to cause b to become oriented in spite of
the opposing rays from the window, they must produce the same effect,
a fortiori^ on the individual a, since they are the only rays which
reach it Yet individuals in the position a do not become oriented at
all. The individuals in the shaded portion of the vessel swim about
in all directions, without relation to the direction of the light rays. It
is only when they come to the line x-y^ where they would pass into the

40 THE BEHAVIOR OF LOWER ORGANISMS.

region lighted directly from the window, that they react by turning
toward the right aboral side and passing back into the shadow. It is
thus clear that it is the light corning from the window to which they
react, not the light reflected from the sides of the dish. We have here
realized the condition concerning which there has been so much dis-
cussion, and which has been considered impossible and unrealizable by
various authors — a negative organism reaching the darker region by
swimming toward the source of strongest light.

This would of course be quite inexplicable on the tropism theory as
set forth by Holt & Lee. What does it indicate as to the real nature
of the reaction? To this inquiry there can be but one answer. The
organism reacts on passing from a darker to a lighter area, without
regard to the direction from which the light comes. It reacts to the
increase in the amount of light falling upon it as compared with the
condition an instant before it had passed into the lighted area. The
reaction takes the usual form — a backing and turning toward the right
aboral side, followed by a forward motion. The organism, therefore,
is directed again toward the shaded area, which it enters.

In all our experiments thus far there have been marked differences
in the illumination of different areas. Let us now arrange the condi-
tions so that light comes from one side, and all parts of the vessel are
equally illuminated. This may be done by placing the Stentors in a
glass vessel with plane walls at one side of a source of light, such as a
window or the bulb of an incandescent electric light. The Stentors,
after a very short interval in which the reaction seems indefinite, swim
away from the source of light, thus gathering at the side away from
the window, where they move about in a disordered way. During the
reaction the Stentors are oriented^ with the longitudinal axis in the
general direction of the light rays and with the anterior end away from
the source of light.

Thus while it is true that the direction of the rays of light has little
if any effect on the reaction when the animals are at the same time
subjected to a sudden change from dark to light, it does determine the
direction of movement when acting alone. In order to discover just
how the reaction occurs it is necessary to observe the animals at the
moment when they change from their former undirected swimming to
the movement away from the source of light.

For determining this a large number of Stentors are placed in the
dish next the window on a dark background. The light comes from
one side and a little from above. The direct rays of the sun were not
employed.

Above the glass vessel are focused the lenses of the Braus-Driiner

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 41

Stereoscopic binocular. This gives a magnification of 65 diameters,
with a working distance of 3 cm., and permits exact observation of
the movements of the individual Stentors. To one who has worked
only with the monocular microscope, the use of the stereoscopic binocu-
lar in studying the movements of small organisms will be a revelation.

The vessel containing the Stentors is first covered with a dark screen
and the Stentors are allowed to become equally distributed throughout
the dish. The screen is then raised, allowing the light from the window
to fall upon the Stentors. Those which are swimming in any other
direction than away from the window now turn and in a short time
are swimming toward the side of the dish away from the window.

With the Braus-Driiner the movements of individuals are observed
at the moment of removing the screen. Some turn at once, while most
continue for a few seconds in the direction in which they are swimming
and then turn. All turn in every case toward the right aboral side.
The turning is continued or repeated until the anterior end is directed

ki-_-------i-ii

Fig. 15.*

away from the window ; then the direct course is continued, carrying the
Stentor to the side of the dish away from the window. The direction
of turning is thus determined by an internal factor — the structure
of the body.

The behavior of the Stentors may be controlled and studied more
exactly by a different order of experimentation. The animals are
placed in a shallow rectangular glass vessel on a dark background, in
a room that is entirely dark save for two incandescent electric lights
A and B (Fig. 15). These are clamped in position, one on each side
of the dish containing the Stentors, and about eight inches from it.
Both these lights can be turned on at once ; both can be extinguished
or one can be turned on while the other is turned off. When only one
is turned on the direction of the light can be instantly reversed by
simultaneously extinguishing this one and turning on the other.

With both lights extinguished the Stentors in the vessel are allowed
to become equally distributed ; then B is illuminated. In a short time

*FiG. 15. — Method of testing the reaction of Stentor to light. A and B are
incandescent electric lights.

42 THE BEHAVIOR OF LOWER ORGANISMS.

most of the individuals have gathered at the side next to Ay as in
Fig. 15. Then ^ is extinguished, while at the same time A is illumi-
nated. The Stentors then turn and move toward B. They may be
stopped at any point in their course and the direction of swimming
reversed by simultaneously turning off one light and turning on the
other. With a sensitive culture the phenomena take place with con-
siderable precision, about four-fifths of the individuals responding
quickly to every reversal of the direction from which the light comes.

Under these circumstances it is easy to observe the individuals at the
moment of the reversal of the course. The observation already made
is confirmed ; the animals always turn at the moment of reversal toward
the right aboral side. The reaction is thus of the same sort that occurs
when there is a sudden increase in illumination. After the first reaction
the anterior end is pointed in a new direction. If this new direction
is away from the source of light the animal swims forward in the
course so laid out. If, as is usually the case, the first reaction does
not result in directing the anterior end away from the source of light,
the reaction is repeated, and this may occur several times. Thus the
anterior end becomes directed successively toward every quarter ; as
soon as it lies toward the side opposite the light the reaction ceases.
The animal now swims straight ahead (that is, in a spiral with a
straight axis) away from the source of light.

Thus while it is clear that light falling from one side produces a
well-defined orientation, this orientation does not take place in such a
way as to be in accordance with the tropism theory as set forth, for
example, by Holt & Lee. It is not the direct action of the light on
the motor organs of the side on which it impinges that determines
the direction of turning, but the latter is due to an internal factor.
This becomes still more evident when the conditions are so arranged
that the direction of turning demanded by the internal factor is the
opposite of that required by the tropism theory.

These conditions can be fulfilled in the followin'g manner: The
light to be turned on (Fig. 15) is so moved beforehand that its rays
shall fall, not directly on the anterior end of the Stentor, but obliquely
at an angle to the path they are following. The animals then react
as before, by turning toward the right aboral side. It often happens
that this involves first a direct turning toward the light, as illustrated
in Fig. 16. In such a case the turning is continued or repeated until
the anterior end is directed^ away from the source of light. We have
seen the same result produced under similar conditions in the experi-
ments illustrated in Fig. 13.

What is the real stimulus to the production of the motor reaction

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES.

43

which results in orientation? The experiments directed precisely
upon this point show that the stimulus producing the motor reaction
is an increase in the intensity of light upon the sensitive anterior end.
Now, in the reaction to a continuous light coming from one side, the
conditions are present for exactly such changes in the intensity of light
at the anterior end as would induce the observed reactions. In the
spiral course the animal swerves successively in many directions. In
certain directions the swerving subjects the anterior end to a more
intense illumination. This change acts as a stimulus to produce the
motor reaction, which carries the anterior end elsewhere. In other
directions the swerving leads to a decrease in the intensity of light
affecting the anterior end. In this case no reaction is produced, and the
organism continues to swim in that general direction. The details of
this method of reacting
will be given in the ac-
count of the reactions of
Euglena, where the mat-
ter was subjected to care-
ful analytical experimen-
tation. The evidence all
indicates that the condi-
tions in Stentor are ex-
actly parallel to those in
Euglena.

We may sum up our
results on Stentor as fol-
lows : A change from

dark to light, such as is caused by swimming from a shaded into an
illuminated region, acts as a stimulus to produce a typical motor reaction ;
the Stentor backs and turns toward the right aboral side, so that it
returns into the shaded region. A change in the illumination of the
anterior end produces the same effect as a change in the illumination
of the entire organism. The direction from which the light comes has
no observable effect on this reaction. But when the illumination is
uniform and the light comes from a definite direction, then light fall-
ing on the anterior end of the Stentor causes the reaction, while light
falling upon the posterior end causes none. The result is that the
animal turns (toward the right aboral side) until its anterior end is

*FiG. i6. — Method bj which Stentor becomes oriented to light, when the light
falls on the aboral side of the animal. Stentor turns, as shown by the arrows,
at first toward the light, but the turning is repeated or continued until the
anterior end is directed away from the light

Fig. 16*

44

THE BEHAVIOR OF LOWER ORGANISMS.

directed away from the source of light, and swims in the direction so
determined. The reaction to light is of essentially the same character
as the reaction to other usual stimuli, and takes place by what we may
/^ call the method of trial and error. When the animal comes to the
boundary of a lighted area, or when the anterior end is illuminated,
this constitutes error ; the animal tries some other direction, and repeats
the trial till the condition constituting error disappears.

Are these results in agreement with all the observed facts.? The
only point on which perhaps question might arise is in regard to the
production of a clearly marked orientation such as we find shown by
Stentor when the light falls upon it from one side. In this case, as

Fig. 17.*

we have seen, Stentor swims directly away from the source of light,
and shows thus a typical orientation. As we have had the dictum
that a motor reaction, such as I have described, " cannot account for
an orientation" (Garrey, 1900, p. 313), it will be well to examine this
matter a little farther. In a previous paper (Jennings, 1900, a) I have
shown how orientation could be produced through a motor reaction ;
the case of Stentor exactly realizes the possibility there set forth. If

♦Fig. 17. — Diagram to illustrate the difference between the method of orienta-
tion to light required by the tropism schema and that which actually takes
place. To light coming from the direction shown by the straight arrows the
tropism schema requires that an organism in the position x-y should attain the
position y-z by turning in the direction indicated by the ^ broken) arrow a-d.
The position is actually attained by turning in the direction indicated by the
long arrow c-d.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 45

the organism is at first not oriented to lines of influence coming from a
certain direction, as in Fig. 17, x-y^ and then becomes oriented, as at
Fig. 17, J/-Z, there are clearly more ways than one by which the orienta-
tion can be produced. The essential question for deciding as to the
nature of the reaction is not whether orientation occurs, but /zow the
orientation is brought about. This consideration has been too often
lost sight of in discussions of the behavior of the lower organisms.

According to the theory of tropisms, as defined by Verworn, Loeb,
and Holt & Lee, the orientation should be brought about by the differ-
ential action of the external agent on the different sides of the organism ;
the organism should turn directly into the line of action of the external
agent, and the direction of turning should be determined by an external
factor, the direction of the infalling rays, or the side on which they
strike the organism. Now this is a matter which can be settled by
direct observation. Direct observation shows us in Stentor that orien-
tation is not brought about in the manner demanded by the theory.
The direction of turning is determined by internal factors. The reac-
tion which produces orientation is identical with the typical reaction
to a mechanical shock, to chemicals, to heat and cold. The difference
between what is demanded by the theory of tropisms and what is
actually observed may be made quickly evident to the eye by Fig. 17.
According to the theory of tropisms the orientation of a negatively
phototactic organism should take place by turning in the direction of
the arrow a-b ; in a Stentor in the position shown (a:-j/), orientation
actually occurs by turning in the opposite direction, as shown by the
arrow c-d.

The further question then arises as to why the organism remains
oriented. All the facts point, in the case of Stentor, to the conclusion
that the reaction to a constant light is due to the intense illumination
on the sensitive anterior end. As soon, therefore, as the'anterior end is
turned away from the light, as is the case in the position y-z^ Fig. 17,
there is no further cause for reaction ; the animal therefore remains
with its anterior end directed away from the light ; that is, it remains
oriented. If, as a result of reaction to some other stimulus, or in any
accidental manner, the animal comes into a position such that it is no
longer oriented, the '' motor reaction" is repeated until the animal
comes again into the position of orientation in which it is no longer
stimulated.

How does the method of reaction to light here described for Stentor
agree with what we know of light reactions in other ciliates.^ As
noted in the introductory paragraphs, comparatively little is known as
to light reactions in this group of organisms. The observations of

46 THE BEHAVIOR OF LOWER ORGANISMS.

Engelmann (1882, a) on the light reactions of certain green ciliates
{^Paramecium bursaria^ Stentor virldls^ etc.) were made before the
typical motor reaction — the turning toward a certain structurally de-
fined side— had been observed in any of the infusoria. Engelmann,
therefore, paid no attention to this point. Yet there is much in his
account of the reactions to light in these organisms to suggest that
it takes place in a way similar to that which I have described above
for Stentor cceruleus. Indeed, Engelmann's account, so far as it
goes, fits precisely into the reaction method which I have described
above. He found, as I have, that the organisms react either when only
the anterior end is affected, or when the entire organism is flooded with
light from beneath. The reaction consists in a sudden turn to one side,
or a sudden start backward, just as in Stentor cceruleus. The only
point which is lacking in Engelmann's account is the observation as

to which side the organism
11 11 turned ; to this point he did

^ y \ \ not direct his attention.

It is interesting to note that
in the account given by Ver-
worn (1889, Nachschrift) of
the reaction to light in Pleu-
ronema chrysalis there is
nothing tending to support
the theory of an orienting tro-
pism. According to Verworn the reaction of Pleuronema to light is by
a sudden leap ('* Sprungbewegung"), which is repeated several times if
the light continues. This sudden leap seems identical with the " motor
reflex" which I have described as the typical reaction to stimuli in
many ciliates, and which consists usually in a leap backward, followed
by a turning toward a structurally defined side. It is in this manner,
as we have seen, that Stentor cceruleus reacts to light and the reac-
tion, as in Pleuronema, is often repeated many times.

Thus the other carefully studied accounts of reaction to light in the
Ciliata, while incomplete, agree so far as they go with that which I
have given for Stentor, and contain nothing to suggest the idea of an
orienting tropism dependent upon unequal stimulation of the motor
organs on the opposite sides of the animal.

Fig. 18.*

♦Fig. 18. — Diagram of the reaction of Stentor to light, after Holt & Lee.
Stentors are confined in a vessel behind a wedge-shaped prism containing a
substance which partly cuts off the light, so that one end of the vessel is darker
than the other. The usual course of a Stentor near the lighter end is shown by
the broken line.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES.

47

Davenport's reference (Davenport, 1897, p. 189) to the negative
light reaction of Stentor makes no attempt to explain the mechanism of
the reaction. Holt & Lee (1901) have given an account of some fea-
tures of the light reaction of Stentor cceruleus. They did not attempt
to determine directly the mechanism of the reactions, by observation
of the exact movements of the organism. Specifically, they made no
observations to determine whether Stentor becomes oriented by turn-
ing directly away from the source of light or only indirectly through
a " motor reaction " such as I have described. They did attempt, how-
ever, to show that the gross phenomena observed might be interpreted
in accordance with the prevailing theory of tropisms set forth on page
7 of the present volume. It will be
well, therefore, to examine their obser-
vations in order to determine whether
they contain anything inconsistent with
the account set forth in the present
paper.

Holt & Lee studied the behavior of
Stentor in an elongated trough which
was lighted from one side. The light
passed through a prism which con-
tained a translucent fluid (a weak solu-
tion of India ink), by means of which
a portion of the light was cut out (Figs.
18 and 19).

At the thicker end of the prism more
light was cut out, hence this end of the
trough (Fig. 19, D) was darker than
the opposite end (Z). It was found
that when Stentors were placed in the trough close behind the prism
(ato, Fig. 19) they turned and swam away from* the lighted side till the
back of the trough was reached {a to d^ Fig. 19) . This is of course ex-
actly what happens when no prism is interposed. Reaching the back of
the trough the animals give the motor reaction (by backing, then turning
toward the right aboral side), thus coming into either the position e or the
position^ (Fig. 19). They then swim forward again, strike the wall.

Fig. 19.

♦Fig. 19.— Reaction of such an infusorian as Stentor to light, under the con-
ditions shown in Fig. 18. After Holt & Lee- The animal in the position x-y^
close behind the prism, turns and swims to the position d^ where it comes against
the rear wall of the trough. It then turns either into the position e, toward the
darker end Z>, or into the position/", toward the lighter end L. In the latter case
it usually soon reacts again, and by repetition of the reaction it finally, as a rule,
becomes directed toward D. Thus, finally, most of the Stentors collect in the
dark end of the trough.

48 THE BEHAVIOR OF LOWER ORGANISMS.

and repeat the reaction. This is repeated many times until the organ-
isms are swimming either toward the end D or toward the end Z. In
course of time it is found that the preponderance of movement is toward
the dark end Z>, so that the majority of the Stentors are gathered at D.
Why this should be so is explained by Holt & Lee as follows :

The reason whj' the Stentors went eventually in greater numbers toward Z>,
and thus appeared oftener to choose e than /*, is that such Stentors as went to e
progressed farther toward D than those which went to/" could progress toward
L. These latter would soon strike the wall a second time, now pretty nearly at
right angles, and during the recoil the light stimuli would favor a return to d.
It appears then amply possible that the circumstance that the organism encoun-
ters the wall of the trough at an acute angle is sufficient to cause its farther
progress to be, in the long run, toward D.

There is evidently nothing in this account which is inconsistent with
the method of light reaction which I have described. On the contrary,
the reason why the organisms finally swim toward the dark end and
gather there becomes much more evident when the reaction method
that I have described is taken into consideration. Let us suppose that
the Stentors, after striking the back of the trough, turn in equal numbers
toward D and toward L, In those swimming toward D the anterior
end is directed away from the source of strongest light (due to reflection
from the lighted end of the dish Z), and the animals are passing into a
region of less intense light. There is thus nothing to cause the '' motor
reaction," with its accompanying change in the direction of movement.
In the Stentors swimming toward Z, on the other hand, the strongest
light falls on the anterior end, and the organisms are passing into a
region of more intense light. Either of these factors taken separately
may, as we have seen, cause the motor reaction (the turning toward
the right aboral side), thus changing the direction in which the Stentors
swim. The animals which start to swim toward Z will therefore soon
be turned, and only when the direction of movement is toward D will
there be no cause for further change.

The observations of Holt & Lee are thus quite in harmony with
the reaction method which I have described, and indeed receive
illumination when this reaction method is taken into consideration.

In the " fourth case" discussed by Holt & Lee {loc. cit.^ pp. 475-
478), the two factors mentioned as determining the turning of the
Stentors away from the end Z would work in opposite directions ; only
experience can tell which would be more effective. As Holt & Lee
do not state specially that they observed the reactions of Stentor under
these conditions no comment is required. Experiments of this character
will be further considered after we have described reactions to light
in flagellates.

REACTIONS TO LIGHT IN CII.IATES AND FLAGELLATES.

49

THE FLAGELLATA.

In the following pages we shall examine the method of reaction to
light in the flagellates Euglcna viridis^ Cryptomonas i

ovata^ and a species of Chlamydomonas.

EUGLENA VIRIDIS.

Euglena viridis swims in a spiral path, continu-
ally swerving toward that side which bears the larger
" lip " and the eye, the so-called dorsal side (Fig. 20).
Its motor reaction to most stimuli is by a sudden pro-
nounced turning toward the dorsal side ; that is, by
swerving still farther toward the same side toward
which it swerves in its normal swimming. Thus the
direction of its path is changed (Jennings, 1900).

The general features of the reaction of Euglena to
light have been well worked out by Englemann
(1882, a) and Wager (1900). These authors show
that Euglena collects in lighted regions. The organ-
isms pass into a lighted area without reaction. But
on coming to the outer boundary of such an area,
where they would pass out into the dark, they react
by turning round and passing back into the light.
The collections of Euglense in lighted areas are thus
brought about in much the same manner as the col-
lections of Paramecia in regions containing a weak
acid (Jennings, 1899). If diffuse light falls from one
side on water containing Euglenae, the organisms swim
toward the source of light. But if strong sunlight
falls upon them they swim away from the source of
light.

Engelmann showed that the colorless anterior end
is the part that is chiefly sensitive to variations of light.
Often the organism in a lighted area, on reaching the
edge, reacts by turning when only the colorless tip
has passed into the darkness.

The precise method of reaction to light, the direc-
tion of turning in becoming oriented or in passing
back into the lighted area, was not worked out by the authors named.
To this point we shall direct our attention.

When a large number of Euglenae are swimming toward the source
of light, if the illumination is suddenly decreased in any way, they give

Fig. 20*

♦ Fig. 20 shows the spiral path of Euglena in its ordinary swimming.

50 THE BEHAVIOR OF LOWER ORGANISMS.

the typical motor reaction described in my previous paper as a response
to other classes of stimuli (Jennings, 1900, p. 235). That is, they
turn at once toward the dorsal side (that bearing the larger lip and the
eye). This is very easily seen when the Euglenae are mounted in the
ordinary manner in a thin layer of water on a glass slide and observed
with the microscope in the neighborhood of a window. If the hand
is interposed between the slide and the window all the Euglenae react
in the way just described.

The reaction is a very sharp and striking one and produces a very
peculiar impression. At first all the Euglenae are swimming in parallel
lines toward the window. As soon as the shadow of the hand falls on
the slide the regularity is destroyed ; every Euglena turns strongly and
may seem to oscillate from side to side in the manner described later.

The turning is often preceded by a slight movement backward.
This was not observed in the reactions to other stimuli (Jennings, 1900,
p. 235), though it agrees with what we find in most other ciliates and
flagellates. In Euglena the reaction to variations in the intensity of
light seems more sharply defined than to most other stimuli. The fact
that the turning is always toward the dorsal side is observable with the
greatest ease. It is particularly evident when the organisms are con-
fined to a thin layer of water, so that they cannot swerve up or down,
but only to the right or left.

The reaction occurs whenever the light is suddenly decreased in any
way. Certain different conditions under which it occurs deserve special
mention, (i) As we have seen, the reaction occurs when a screen is
brought between the organisms and the source of light toward which
they are swimming. (2) It also occurs when the illumination is de-
creased by cutting off light from some other source than that toward
which they are swimming. Thus the organisms on the stage of the
microscope may be lighted from below, by the substage mirror, and at
the same time may receive light from the window at one side of the
preparation. They swim toward the window, since the light from that
quarter is much stronger than that from below. If now the light from
below is suddenly decreased by closing the iris diaphragm, the Euglenae
react as usual by turning strongly. This is notwithstanding the fact
that the proportion of light coming from the window, to which they
were oriented, is now greater than before, so that it might be supposed
that they would remain more strongly oriented than ever. For the rest,
the disturbed orientation is soon restored. (3) The reaction occurs
when the decrease in illumination is due to the movements of the
Euglenae ; that is, when the swimming organisms come to the edge of
a lighted region where they would, if the course were continued, pass
into the darkness. As a result of the reaction they return into the light.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 5 1

The reaction occurs at a decrease in illumination not only when the
organisms are oriented and swimming toward the source of light, but
also when they are not oriented and are merely scattered in a weakly
lighted area. Further, in cases where most of the Euglenae are oriented
and swimming toward a source of light, a number of specimens will
always be found that are not oriented at all, or are swimming away
from the source of light. Such individuals react to a sudden decrease
in illumination in the same manner as do the specimens that are oriented
with the anterior end toward the source of light. This result may be
observed in a curious way as a consequence of the fact that it requires
some time for the light to produce its orienting effect. Thus, if the
Euglenae are placed between a weak and a strong light they swim toward
the strong light. If, now, the strong light is cut off, they react in the
usual way and swim toward the weak light. Now the strong light
maybe restored ; the Euglenas continue for a few seconds to swim toward
the weak light, thus away from the strong light. If while they are
swimming in this manner the strong light is cutoff, the Euglenae, swim-
ming away from it, react in the usual manner, by turning strongly
toward the dorsal side.

The usual reaction may be produced by a decrease in illumination
that is not sufficient to cause a permanent change in orientation. Thus
the Euglenae on a slide or in a shallow dish may be lighted from a
window at one side. By passing a small screen in front of the window
at some distance from the preparation a portion of the light is cut off;
the Euglenae then respond in the usual way, by swerving toward the
dorsal side. The movement thus becomes very irregular. Since the
Euglenae continue to revolve on their long axes the dorsal side may lie
first to the (observer's) right, then to the left. The Euglenae all seem,
therefore, to vibrate from side to side. This is the '' Erschiitterung "
or trembling described by Strasburger (1878) as occurring in swarm-
spores when the illumination is changed ; it will be understood better
when we have considered more in detail the mechanism of the reactions.
Meanwhile the screen retains its position, but still admits more light
from the direction of the window than from any other direction. The
reaction of the Euglenae, therefore, soon ceases ; their orientation is
restored in the way to be described later, and they continue to swim
toward the window.

This experiment is an important one. It shows that the typical reac-
tion may be produced by a decrease in light that is not sufficient to
permanently destroy the orientation. Thus it is clearly the decrease
in illumination to which the organisms react ; not to a change in the
direction of the light rays. The experiment shows further that it is
not the absolute amount of light that determines the reaction. Some

53 THE BEHAVIOR OF LOWER ORGANISMS.

time after the decrease in illumination takes place the organisms behave
just as they did before, swimming in the same direction. Further,
the illumination may be decreased very slowly to the same extent
without causing a reaction. If the screen is at first far away from the
preparation and is then slowly moved to the position it occupied in
the experiment just described no reaction is produced. It is only the
sudden change that has caused the reaction. The change, however,
need not be a very marked one in order to be effective.

Our experiments thus far have shown that in a moderate light Eu-
glena reacts to a decrease in illumination. But the absolute amount of
light present has an effect on the reaction. If the light is very strongly
increased the same reaction is produced as when the light is decreased.
If while the organisms are swimming toward a moderately lighted
window direct sunlight is allowed to fall upon them, they respond in
the same way as to a sudden decrease in illumination ; that is, they
turn strongly toward the dorsal side, continuing or repeating the re-
action till the anterior end is directed away from the source of light.
They now continue to swim in that direction, the positive reaction
having been transformed into a negative one. Thus under intense
light the conditions of stimulation are the opposite of those under
moderate light. This is paralleled in the reactions of the infusoria to
chemicals ; often a strong solution of a certain chemical produces a re-
action under opposite conditions from those in which a weak solution
of the same chemical is effective.

Let us now proceed to a more careful study of the reaction itself.
The reaction which occurs when the illumination is changed is really
an accentuation of a certain feature of the usual movements. Euglena,
as we know, revolves on its long axis as it swims forward, and at the
same time it swerves toward the dorsal side. The resulting path is
therefore a spiral one (Fig. 20). The usual reaction to a stimulus is
an accentuation of this normal swerving toward the dorsal side, as com-
pared with the other factors in the swimming; the organism suddenly
swerves so much farther than usual in this direction that the path may
be completely changed. If the reaction is a very decided one the revo-
lution on the long axis and the movement forward may cease during
the swerving toward the dorsal side ; the anterior end then describes
the arc of a circle about the posterior end as a center. In a less pro-
nounced reaction the revolution on the long axis continues. The circle
described by the anterior end is then less and the whole body describes
the surface of a cone, or a frustum of a cone, as illustrated in Fig. 21.
Every gradation exists between the normal spiral course and the strong
reaction in which the anterior end swings in a circle about the
posterior end as a center.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES.

53

When oriented and swimming toward the source of light the swerv-
ing toward the dorsal side is comparatively slight. As seen from
above, the organisms seem merely to oscillate a very little from side to
side as they revolve on the long axis. Careful examination shows that
the swerving is always toward the dorsal side, as in Fig. 20, the alter-
nations in direction being due to the alternations of position of the
dorsal side. Now, when the illumination is suddenly decreased, the
Euglenae at once swing much farther than usual toward the side to

Fig. 21.*

which they are already swerving, that is, toward the dorsal side. If
the decrease in illumination is not very great, so that the stimulus is
not a strong one, the swerving is not very great, and the organism at
the same time continues to revolve on its long axis ; thus the anterior
end describes a circle and the whole body describes the surface of a

*FiG. 21. — Diagram to illustrate reaction of Euglena when the illumination
is decreased. The Euglena is swimming forward at i ; when it reaches the
position 2 the illumination is decreased. Thereupon the organism swerves
strongly toward the dorsal side. This swerving, cotnbined with the revolution
on the long axis, causes the anterior end to swing about a circle, so that the
Euglena occupies successively the positions 2, 3, 4, 5. 6, etc. From any of these
positions it may start forward, as indicated by the arrows, if the condition
causing the reaction ceases to act. In the figure the Euglena is represented as
swimming forward from the position 6.

54

THE BKMAVIOR OP LOWER ORGANISMS.

cone, or the frustum of a cone, as indicated in Fig. 21. The result, as
seen from above, is that all the specimens seem to vibrate from side to
side ; in other words, they are taken with a sudden oscillation or trem-
bling. This oscillation when the intensity of the light is suddenly
changed was observed by Strasburger (1878,
pp. 25 and 50) in flagellate swarm-spores ; he
speaks of it as " Erschiitterung " or " Zit-
tern." During this oscillation the anterior end
becomes pointed successively in many different
directions, as Fig. 21 shows. When, now, the
usual forward course is resumed (with only the
usual amount of swerving toward the dorsal
side), the animal follows one of these directions.
Thus its path is changed (Fig. 22). Strasburger
(187S, p. 25) noticed that the path followed after
the oscillation was oblique to the former path.
As a study of Figs. 21 and 22 will show, this is a
necessary consequence of the increased swerving
toward the dorsal side, to which the oscillation
itself is due. All these relations become much
clearer if a model of an actual spiral is studied ;
it is difficult to represent them upon a plane
surface.

If the stimulus is stronger, as when there is a
greater decrease in illumination, the swerving
toward the dorsal side is much greater ; the or-
ganism wheels far to that side, so that the spiral
course seems entirely interrupted. But there is
really nothing in this reaction differing in prin-
ciple from what is happening in the normal
forward swimming. If the swerving toward the

f dorsal side is long continued the specimen may

be seen to swing first far to the (observer's) right,

_ ^ then, after it has revolved on the lone axis, far

Fig. 22.* ' fc> '

to the (observer's) left; in reality it swings an
equal amount upward and downward and in intermediate directions.
It may, however, swing at once so far to the dorsal side that the new

♦ Fig. 22. — Shows the spiral path of Euglena, illustrating the effect of a
slightly marked reaction. At a the illumination is decreased; the organism
therefore swerves toward the dorsal side, causing the spiral to become wider.
At b the ordinary method of swimming is resumed; since at this point the
organism was more inclined to the axis of the spiral than before the reaction,
the new course lies at an angle with the previous one. Compare with Fig. 21.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES.

55

course forms a right angle, or a still greater angle, with the original
course; if the turning is through i8o°, the course will be squarely
reversed. Indeed, sometimes the organism swings around an entire
circle or more. When the usual method of swimming is resumed after
such reactions as those just described, the course has been completely
changed.

Strasburger (1S78, p. 25) noticed that after a decrease in illumina-
tion flagellate swarm-spores often turn strongly to one side or even
describe circles. But he did not notice that the turning was always
toward the same side of the organism,* and did not perceive the relation
between this effect and the remainder of the reaction.

Fig. 23.1

This method of reaction is particularly striking when the Euglenae
are confined to a very thin layer of water between the slide and the
cover glass, so that they cannot swerve up or down. When the light
is decreased, we will suppose that the dorsal side is to the (observer's)
left. The Euglena then swings far to the left. At the same time it

♦Naegeli (i860, p. 96) had, however, before Strasburger, observed that in such
svi^arm-spores the same side always faces the outside of the spiral path. This
observation, which really contained the germ of a correct understanding of the
reactions to stimuli, seems hardly to have been noticed by later writers.

t Fig. 23. — Diagram of the method by which Euglena becomes oriented with
anterior end toward the source of light. At i the Euglena is swimming toward
the source of light. When it reaches the position 2 the light is changed so as
to come in the direction indicated by the arrows at the right. As a consequence
of the decrease in illumination of the anterior end thus caused, the organism

56 THE BEHAVIOR OF LOWER ORGANISMS.

revolves on its long axis, bringing the dorsal side down. Since it can
not swing downward, owing to the narrow space, this has little effect
on the reaction, save to stop the movement to the left. Now, by con-
tinued rotation the dorsal side has come to lie to the (observer's)
right ; the Euglena may then be seen to swing far to the right. In each
case under these conditions it is at once evident by observing the
larger lip at the anterior end that the organism is swinging toward
the dorsal side.

This method of reaction is very effective in preventing Euglena from
passing from an illuminated region to a shaded one. As soon as the
anterior end enters the shadow, the animal swings far toward the dor-
sal side till the anterior end is brought again into the light, repeating
the reaction if necessary. There is then no further cause for reaction.
The reaction to a very strong increase of illumination is, as we have
seen, identical with that to a decrease in illumination.

In our experiments thus far we have directed attention primarily to
the effects of changes in the intensity of illumination, and have found
that such changes produce a motor reaction independently of the direc-
tion of the light rays. But it is of course well known that Euglena
does react with reference to the direction of the light rays. Euglenae
swim toward the source /^f light when weakly illuminated, away from
the source of light when strongly illuminated. If Euglenae are swim-
ming at random in a diffuse light they soon become oriented when the
light is allowed to act on them from one side, even if the intensity of
illumination remains the same. Or, if Euglenae are swimming toward
a source of very weak light and a stronger light is allowed to act upon
them from the opposite side, they become oriented, in time, with
anterior ends toward the stronger light. In examining this dependence
of the direction of swimming on the direction of the rays of light, we

swerves strongly toward the dorsal side, at the same time continuing to revolve
on the long axis. It thus occupies successively the positions 2, 3, 4, 5, 6. In
passing from 3 to 6 the illumination of the anterior end is increased; hence the
reaction nearly or quite ceases. In the next phase of the spiral, therefore, the
organism swerves but a little toward the dorsal side — from 7 to 8. But this
movement causes a decrease in the illumination of the anterior end, and this
change induces again the strong swerving toward the dorsal side. Hence in
the next phase of the spiral the organism swings through 9 and 10 to 11. In
this movement again the illumination of the anterior end is increased; hence
the reaction ceases, so that from 12 the organism swerves only as far as 13.
Then owing to the decrease in illumination caused by this movement, the
swerving increases, so that the Euglena swings from 13 through 14 and 15 to 16.
Now it is directed toward the source of light, and such swerving as takes place
in the spiral course neither increases nor decreases the illumination of the
anterior end. Hence there is no further reaction; the Euglena continues to
swim forward in the direction 16-17.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES.

57

shall have to keep in mind two questions : First, how is the position
of orientation brought about? Second, what is the real stimulus in
producing orientation ?

To answer the first question we must observe the movements of the

V-.?.-c^'

M

""^^^

<^^'

'^^^

)a

f—-4

Fig. 24.*

organism at the time orientation occurs. Observation of the individ-
uals as they are becoming oriented shows that orientation is brought
about through the same motor reaction that we have already described ;

* Fig. 24. — Path followed by Euglena when the direction of the light is
changed. From i to 2 the organism swims forward in the usual spiral path.
At 2 the position of the source of light is changed, so that it now comes from
behind. The organism then begins to swerve farther than usual toward the

58 THE BEHAVIOR OF LOWER ORGANISMS.

that is, by a turning toward the dorsal side. The simplest case is per-
haps that of the reversal of orientation, produced when strong sunlight
is allowed to fall from in front upon specimens that are swimming
toward a diffusely lighted window. Under these circumstances, as
we have seen, the Euglenas turn toward the dorsal side, changing their
course. They may turn directly through 180°, in which case they are
at once oriented with anterior ends away from the light ; but usually
the orientation is less direct than this. The reaction is generally
repeated several times. Through its continued swerving toward the
dorsal side, combined with the revolution on the long axis, the organism
directs its anterior end successively in every direction. When the
anterior end has finally come into a position where it points away from
the strong light the reaction ceases, and the organism swims forward
in the usual way. The details of the orienting reaction will be brought
out more fully in the following account of the way in which the anterior
end becomes directed toward a source of light of moderate intensity.
Let us now take a case in which the change in the direction of the
rays of light is not accompanied by a change in the intensity of illumi-
nation. Euglense are swimming about at random in a diffuse light
when all the light is allowed to fall upon them from one side. They
then become oriented, with anterior ends directed toward the source of
light Or, the organisms are swimming toward a source of light when
the direction of the light rays is changed or reversed by quickly
moving the source from which the light comes. The Euglenae then
after a time become reoriented. Under such circumstances there is no
sudden, decided reaction, such as occurs when the illumination is
suddenly decreased. The organism merely begins to swerve farther
toward the dorsal side than usual. Thus the spiral has become wider,
and the anterior end comes to be pointed successively in many dif-
ferent directions, as illustrated at 1-6 in Fig. 23. In some of these
positions the anterior end is directed farther away from the source
of light, as at 3 ; in other positions more nearly toward the source
of light, as at 6. In the latter case the swinging toward the dorsal
side becomes less marked ; hence the succeeding phase of the swing,
which carries the anterior end away from the light, is less pronounced ;

dorsal side, owing to the decrease in the illumination of the anterior end. Thus
the spiral becomes wider, a and b showing the limits of the swerving. At 3 the
normal amount of swerving is restored, so that the new path is at an angle with
the old one. Now the organism swerves at each turn of the spiral a short dis-
tance away from the source of light, as at c, e, g, and a longer distance toward
the source of light, as at d,f, h^ for the reasons shown in Fig. 23. At h it has
in this manner become directed toward the source of light, and there is no fur-
ther cause for swerving more to one side than to the other; it therefore swims
in a spiral with a straight axis toward the source of light.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 59

the anterior end therefore does not swing so far in the direction
away from the light as in the preceding phase it swung toward the
light. This is illustrated at 7-8 in Fig. 23. But as a result of such
swerving as does occur the anterior end is now (at 8) directed more
away from the source of light than before. There then follows a
new reaction, with increased swerving toward the dorsal side in the
next phase of the spiral (8-1 1, Fig. 23), which carries the dorsal side
toward the source of light. Hence the anterior end swings still further
toward the position where the light shines directly upon it. This con-
tinues. As a result of this repeated swinging of the dorsal side slightly
away from the source of light and strongly toward the source of light
the organism gradually changes its course, continuing to swim in a
spiral and to swerve toward the dorsal side, until the axis of the spiral
is in line with the light rays and the anterior end is toward the source
of light. This method of reaction will best be understood by a study
of Figs. 23 and 24 and their explanation.

Thus the orientation is gradual and for a certain stretch after the
light has begun to act the organism is not completely oriented. With
a fairly strong light, however, the period of time required for complete
orientation is very slight. Strasburger (1878, p. 24) noticed that when
Haematococcus is swimming toward a source of weak light and the
light is suddenly increased so as to reverse the orientation, there is a
period of " verschiedenen Schwankungen " before the reverse orienta-
tion is attained. He paid little attention to the behavior of the
organisms during this period, however.

Our account has been thus far purely descriptive ; we have attempted
to set forth the events as they may be observed , without trying to
indicate the causes at work. We must now inquire as to what is the
real stimulus and its method of action in producing orientation.

First, we note that in becoming oriented Euglena does not turn
directly toward the source of light. As in the reaction to other stimuli,
the turning is throughout toward a structurally defined side. This
shows that the orientation of Euglena, like that of Stentor, cannot be
accounted for on the orthodox tropism theory. In other words, the
orientation is not due to the direct effect of the light on the motor
organs of the side on which it falls. As in Stentor, orientation may
be reached by turning either toward or away from the source of light,
or in any intermediate direction. The response is a " motor reaction '*
of a definite type.

Just what is the stimulus which produces this motor reaction.? All
our experiments up to this point have shown clearly that this reaction
is produced by changes in the intensity of illumination, and that a change
in the illumination of the anterior end produces the reaction as well as

6o THE BEHAVIOR OF LOWER ORGANISMS.

does a change in the illumination of the entire body. Indeed, Engel-
mann (18S2, a) showed that a change in illumination over the remainder
of the body is ineffective in producing the reaction, so that in every case
the reaction is due to the change in illumination at the anterior end.
Now, in the orientation reaction the conditions are present for produc-
ing changes of illumination at the anterior end of precisely the character
which would, in view of our other experimental results, bring about
the reactions observed. This will best be shown by again examining
in detail from this point of view a concrete case.

In Fig. 23 we will suppose that the Euglena at i is at first swimming
toward the source of light. When it reaches the position 2 the light is
changed, so that it now comes from the direction indicated by the
arrows at the right. By this change the intensity of illumination at
the anterior end is decreased, since before the light came from directly
in front and affected the entire end, while now it falls upon but one
side. We know from other experiments that as a result of such a
change the organism reacts by swerving more toward the dorsal side,
at the same time continuing to revolve on the long axis. This is ex-
actly what happens now ; by the increased swerving the organism is
carried from position 2 to position 3. In this change the anterior end,
swinging still farther away from the source of light, is still less illumi-
nated than before. As a result of this farther decrease in illumination
the reaction is continued or increased ; combined with the revolution
on the long axis it carries the organism successively to positions 4, 5
and 6. In this part of the movement the anterior end becomes pointed
more directly toward the source of light, and is hence more strongly
illuminated ; there is therefore nothing in this movement to cause a
reaction. The strong swerving toward the dorsal side then ceases or
becomes less. But in the next phase of the spiral course (from 7 to 8),
there is necessarily at least the normal amount of swerving toward the
dorsal side, and this carries the organism to a position (8), where
the intensity of the light acting on the anterior end is decreased. As
a result of this decrease we know that the " motor reaction " must again
be induced ; the organism swings then farther toward the dorsal side*
This movement, combined with the revolution on the long axis, carries
the Euglena through 9 and 10 to 11. Here again the swerving de-
creases, because the change was from a less illuminated to a more
illuminated region. Hence after reaching 12 the Euglena swerves only
a little away from the light, to 13 ; then, as a result of the decrease in
illumination at the anterior end caused by this movement, it swerves
far toward the light, through 14 and 15 to 16. This movement causing
greater illumination, the reaction ceases. The light is now shining full
on the anterior end. The organism therefore swims forward in the

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 6l

usual spiral course, in all phases of which the illumination of the anterior
end is equal. If the light came from the rear of Euglena i instead of
from the direction indicated by the arrows, the reaction above described
would be continued in the same way until the direction of swimming
was completely reversed.

Thus the orientation of Euglena in a continuous light is due to the
production of the *' motor reaction," with its turning toward the dorsal
side, whenever there is a decrease in illumination at the anterior end.

There is no other explanation of the orientation, so far as I am able
to see, that is in agreement with all the facts. At first one is tempted
merely to say that the subjection of the anterior end to shadow pro-
duces the motor reaction, and that this is continued until the anterior
end is no longer shaded. This statement is correct if by "subjection
to shadow*' we mean an active process, involving a change from a
more illuminated condition. But if we mean that darkness as a con-
tinuous, static condition is the cause of the reaction, then considera-
tion shQws that this will not account for all the facts. It leaves out of
account the capability of the organism to become acclimatized to cer-
tain degrees of light and shade, and certain of the experimental results
are crucial against it. Thus, suppose the Euglenas are swimming
toward a source of weak light, and a stronger light is then allowed to
act upon them from another direction. The anterior end continues to
receive the same amount of light as before (since the weak light still
persists), yet the organism reacts as usual, becoming oriented toward
the stronger light. The motor reaction by which the orientation is
brought about cannot therefore be due to darkness or shade (considered
statically) at the anterior end. On the other hand, the case just men-
tioned is easily understood on applying the explanation given above.

Again, it might be held that the reaction is due in some way to the
relative amount of illumination at the two ends. It might be main-
tained, for example, that when the posterior end is more illuminated
than the anterior, this difference acts as a stimulus to cause the '* motor
reaction." There is, of course, no independent evidence in favor of this
view, and the experimental results prove it to be incorrect. We have
shown that the reaction is produced (i) when both ends are equally
stimulated, as when the light comes directly from one side ; (2) when
neither end receives light, as when the light is cut off completely. Fur-
ther, it might be held that the reaction is produced when the anterior end
is not more intensely illuminated than the posterior end. It is, of course,
a little difficult to conceive how so indefinite a condition could act as a
stimulus to a definite motor reaction, but in any case the experiments show
that this is not the real cause of the '' motor reaction." Thus certain of
the experiments show that the " motor reaction " is produced even when

63 THE BEHAVIOR OF LOWER ORGANISMS.

the light is reduced by the same amount at both ends, so that the anterior
end is still more strongly lighted than the posterior. This case is
realized in the experiment in which a small screen is interposed between
the EuglenaB and the window toward which they are swimming. The
light is thus somewhat decreased, but is still sufficient to cause orien-
tation. The anterior end is thus still lighted more than the posterior,
yet the organisms respond with the " motor reaction" at the moment
the light is decreased. The same thing is shown still more decidedly
in the experiment described on page 50, in which the " motor reaction "
is produced when the light is cut off from some other source than that
toward which the organisms are swimming. In this case the propor-
tion of light shining on the anterior end is greater after the change in
illumination than before, yet the *' motor reaction" is produced at the
moment the change takes place.

The explanation we have given is, therefore, the only one that is in
agreement with all the facts, and it accounts for every detail of the re-
actions to light. The cause of all the phenomena of light reaction in
Euglena is the fact that a sudden change in light intensity on the anterior
end induces a typical '' motor reaction." It is noticeable that the
reaction is throughout due to a dynamic factor, to some change in the
relation of the organism to the light, a change due either to an active
alteration of the environment, or to a movement of the organism. To
static conditions, if not too intense, the organism may soon become
acclimatized, so that no farther reaction is caused. The absolute in-
tensity of the light affects the reaction only in so far as it determines
whether it shall be an increase or a decrease in intensity that causes
the *' motor reaction."

To sum up, the reaction of Euglena, from beginning to end, is ex-
plained by the fact that a sudden change in illumination, even though
slight, causes a definite motor reaction, the essential feature of which
is an increased swerving toward the dorsal side. Orientation is brought
about by the increased swerving in the next phase of the spiral course
when the illumination of the anterior end is diminished, and by the
decreased swerving in the next phase of the spiral when the illumination
of the anterior end is increased. In general terms we can say that the
reaction of Euglena to light is by the method of trial and error. The
organism tries turning in many directions ; when the turning is such
as to produce a decrease in the illumination of the anterior end it
** tries" other directions; when it is such as to produce increased
illumination of the anterior end, or when no change in illumination
results, the reaction ceases and the organism continues to swim forward
in that position. The result of this method of reaction is necessarily
orientation with the anterior end toward the source of light.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 63

CRYPTOMONAS AND CHLAMYDOMONAS.

Cryptomonas ovata is one of the organisms studied by Strasburger
(187S), under the name Chilomonas ovata^ in his classical paper on
reactions to light in flagellates and swarm-spores.

The specimens studied by the present author were mostly of the
*' young" form, having pointed, curved, posterior ends. One side is
strongly convex, while the other is less curved, or is even concave near
the posterior end. It is thus very easy to distinguish the two sides of
the organism and to observe their relation to the movements.

Cryptomonas ovata swims in a rather wide spiral, with the more
convex side toward the outer surface of the spiral. In other words, the
organism swerves continually toward the more convex side. The
response to usual stimuli is a strong turn toward this convex surface ;
this is easily seen when the organism comes in contact with an
obstacle.

The Cryptomonads swim toward or away from the source of light
under the same conditions as Euglena, and gather in lighted areas in
the same manner as does the organism last named. They react to a
sudden decrease in the intensity of illumination by turning toward the
more convex side. If the decrease in intensity is marked, the organism
turns suddenly for a long distance, 90° or more, so that the course is
completely changed. If the stimulus is less the turning toward the
more convex side is not so rapid, and since the revolution on the long
axis is continued the body of the organism describes the surface of a
wide cone or frustum of a cone. When a large number of specimens
react in this way at the same time a peculiar shaking or trembling
appearance is produced; this is evidently what Strasburger (1878)
called " Erschiitterung " or " Zittern." As a consequence of the wide
swerving, when the normal method of swimming is resumed the course
lies in a new direction.

In all these respects Cryptomonas exactly resembles Euglena. Fur-
ther, the organism becomes oriented to light in precisely the same manner
as is described above for Euglena. In fact, if we substitute " more
convex side" for " dorsal side " in the account of Euglena, it will fit
almost throughout the reactions of Cryptomonas. It is therefore unnec-
essary to describe the phenomena in Cryptomonas in detail.

A study was made also of the reactions of a species of Chlamydo-
monas. The movements of Chlamydomonas and its reactions to light
resemble those of Euglena and Cryptomonas. But the organism is so
small and the differentiations of the bodily structure are so slight that
I was unable to determine the relation of its structure to the spiral
path and to the direction of turning in the reaction. The oriented

©4 THE BEHAVIOR OF LOWER ORGANISMS.

organism reacts to a decrease in illumination by a sudden turn to one
side, by an increase in the width of the spiral, and by a change in the
course, just as happens in Euglena and Cryptomonas. The unoriented
organism becomes oriented in a manner which is similar to that de-
scribed above for the two organisms just named. Since, however, I
am unable to give the precise relations of these movements to struc-
tural differentiations of the body, a further account of details would
not be of interest.

GENERAL RESULTS.

In summing up our results on reactions to light in the organisms
studied, there are two points of especial interest which should be con-
sidered separately. The first relates to the nature of the reaction
produced, the second to the nature of the agent causing the reaction.

NATURE OF REACTION PRODUCED BY LIGHT.

As to the nature of the reaction produced by light there has been
much discussion. The orthodox tropism theory is perhaps that which
has the greatest number of adherents. It is set forth in detail in the
paper of Holt & Lee (1901). According to this theory the light acts
directly on the motor organs of the side on which it impinges ; supra-
optimal light causes increase of the backward stroke (in the case of
cilia or other swimming organs) ; suboptimal light causes a decrease
in the backward stroke. The result is that the organism is turned
directly toward or from the more intensely lighted side, and hence
toward or from the source of light. The diagrams given in the pre-
ceding paper (Figs, i and 2) can be applied directly to the elucidation
of this theory.

In the experiments on the ciliates and flagellates set forth in the
present paper the precise method of reaction was determined by obser-
vation. It is not in accordance with the tropism theory above set
forth. This has been emphasized in detail in the account of the
reactions of Stentor, so that it need not be reiterated here. The reac-
tion to light is of the same character as that to other stimuli, and takes the
form of a motor reaction in which the organism performs a definite
set of actions. It first usually stops or swims backward, then turns
toward a structurally defined side, then continues forward. The
result is to change the course of the organism. As a result of the con-
tinual rotation on the long axis, together with the swerving toward a
certain side, the organism comes to be pointed successively in every
direction. In continues to swim forward in that direction which does
not induce a stimulus to further swerving. The whole reaction is a
strongly marked example of the type of behavior which may be called
the " method of trial and error."

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 65

NATURE OF AGENT CAUSING THE REACTION.

(i) The primary and essential cause of the reaction is a change of
illumination. The change of illumination must take place with some
suddenness, but need not be very great in amount. The change in
illumination acts as an effective stimulus even though the degree of
illumination preceding the change and that following it would, when
acting continuously, produce no such result. This is shown by the
experiments on Euglena, in which the light coming from one side was
decreased a certain amount. The orientation of the organisms and
their direction of movement vvas the same before and after the change,
but at the moment the change occurred there was a marked reaction.
Other experiments detailed above demonstrate the same thing. Further,
the change in illumination acts independently of the direction of the
rays of light. This is shown by the experiment just cited, in which
the effective direction of the rays of light was the same before and
after the reaction ; it is also shown in the reaction caused when the
light is decreased from below, in the case of Euglenae swimming
toward a window (p. 50), and in the reaction of Stentor on passing from
a shadow to a lighted region even when the animal is oriented with
anterior end away from the light (p. 39). The change in illumination
acts equally whether it affects the entire organism or only the anterior
end. The evidence indicates that in all cases it is really the change at
the anterior end which induces the reaction.

(2) The absolute intensity of the light affects the reaction by deter-
mining in a given case whether a reaction shall be caused by an
increase or a decrease in illumination. Through this action it also
determines, in the way to be mentioned in the next paragraph, whether
in a continuous light the sensitive anterior end shall be directed toward
or away from the source of light ; that is, whether the response shall
be *' positive" or "negative."

(3) Indirectly, and through the factor set forth in paragraph (i), the
direction from which the light comes is a determining factor in the
reactions. Through the spiral course in which the organisms swim such
conditions are furnished that in a field continuously lighted from one
side the sensitive anterior end of the unoriented organism is subjected
to repeated changes in the intensity of illumination. As a result,
organisms which respond by the motor reaction to an increase in illu-
mination at the anterior end must become oriented with anterior end
directed away from the light ; organisms which react to a decrease in
illumination must become oriented with anterior end directed toward
the light. (Details in the account of Euglena, pp. 60, 61, and Figs.
23, 24.)

66 THE BEHAVIOR OF LOWER ORGANISMS.

The results of this method of reacting may be stated correctly, though
not completely, as follows : In a negative organism light falling upon
the sensitive anterior end causes a reaction by which the anterior end
is pointed in many different directions ; the reaction ceases as soon as
a direction is reached in which the anterior end is pointed away from
the light. In a positive organism the shading of the sensitive anterior
end produces the reaction by which the anterior end is pointed in many
different directions ; the reaction ceases as soon as the anterior end is
no longer shaded.* The reaction is thus by the method of trial and
error ; when stimulated the organism tries many different positions,
till one is found in which there is no further stimulation.

Consideration will show, I think, that the factors producing reaction
to light in these lowest organisms are essentially the same as in higher
ones, if man may be taken as a type of the latter. The factors are, as
we have seen, variations in intensity of illumination, and, indirectly,
the direction from which the light comes. It is possible that in man
the latter factor works more directly than in the infusoria ; leaving this
question out of consideration, the two factors are present in both cases.
Consider a human being who reacts to light as a purely physical agent,
not with regard to the associations which it brings up. In a dark space
a gleam of light is pleasant and induces movement toward it. There
is then a positive reaction with orientation, but the orientation is not
due to the difference in intensity of light on different parts of the body,
nor to its direct effect on the motor organs. The orientation is such as
to keep the light shining on the more sensitive part of the body, the
eyes. An excessively powerful light is unpleasant and induces a nega-
tive reaction just as happens in Euglena ; the orientation is then such
as to keep the more sensitive part of the body, the eyes, away from the
light. Further, man is sensitive to a sudden change in illumination.
A strong light bursting from the darkness, or sudden darkness in the
midst of bright light, induces a marked motor reaction, and less striking
differences may produce a response. Both in man and in Euglena the
reaction likewise depends upon color ; but with this phase of the matter
we are not at present concerned.

When the factors above set forth are taken into consideration certain
peculiar experimental results that have given rise to much discussion
become clearly intelligible. I refer particularly to the experiments in
which the direction of the light and the decrease in intensity of illumi-
nation do not show the usual relations. Under ordinary conditions
movement away from a source of light is movement into a region of less

* This statement is incomplete in that it does not bring out the fact that it is
a change from light to shade or vice versa that induces the reaction; if this be
understood, the statement is correct.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 6"]

intensity ; movement toward the light into a region of greater intensity.
In the well-known experiments of Strasburger (1878) and others, this
condition is modified by passing the light through a wedge-shaped
prism filled with a solution that cuts out part of the light.

When a drop of water or a culture dish is placed beneath such a
prism, and the latter is so situated that its surface is perpendicular to
the light rays, the intensity of the illumination is greatest behind the
thin edge of the prism, and thence decreases gradually toward the
opposite end, while the rays of light all come directly from above.
Under these conditions Strasburger (1878, p. 36) found that the positive
swarm-spores remained equally distributed throughout the drop, not
collecting at the lighter end. Now, the only difference between this
experiment and the one illustrated in Fig. 1 1 of the present paper is
that in Strasburger*s experiment the decrease in illumination is very
gradual. We have seen above (p. 52) that a very gradual change in
illumination produces no reaction. Hence the organisms may wander
from one side of the drop to the other without reaction, the difference
in illumination at two successive instants never rising to the necessary
threshold of stimulation. If the relation of stimulus to reaction follows
Weber's law, the result is just what we should expect, provided the
change in illumination is sufficiently gradual. When the difference in
illumination from above is great, Strasburger's own experiments (/. c,
p. 33) show that the organisms do react.

On the other hand. Holt & Lee (1901), using a similar prism,
found, under similar conditions, that the negative organism, Stentor,
does, on the whole, tend to gather at the darker side of the drop.
This shows that the difference in illumination between neighboring
points in this particular experiment was not below the threshold of
stimulation for the organism in question. If, as Holt & Lee sup-
pose, a certain amount of light was reflected from the lighter end of
the vessel, then the inclination to go to the darker side would be rein-
forced by Stentor's tendency to turn when the light falls upon its
anterior end (see p. 43). The fact that in Strasburger's experiments
the organisms remained scattered throughout the drop seems to indi-
cate that this reflected light played no part in his results.

In another set of experiments Strasburger placed his prism over the
swarm-spores in such a way that the light came obliquely from the
direction of the thick end of the wedge. If the positive organisms now
go toward the thicker end of the wedge, they pass toward the source
of light, but into a region of decreased illumination ; if they go toward
the thin end they pass away from the source of light, but into a region
of higher illumination. Which will they choose.?

Strasburger found that the positive swarm-spores pass toward the

68 THE BEHAVIOR OF LOWER ORGANISMS.

source of light, into the region of less illumination. But is not this
exactly what we must expect? His former experiment showed us that
under the prism the change from light to darkness was so gradual that it
produced no effect on the organisms. Hence the direction from which
the rays come is left to produce its effect alone, and it produces the
usual effect. The organism reacts in the usual " trial and error" way
until the anterior end is directed toward the light ; then it moves in that
direction. Incidentally it comes into a region of less intensity of light,
though the decrease is so slight as to produce no effect on the organism.

Parallel considerations hold for the negative organism. Under
similar circumstances, if the variation in illumination is very gradual,
it directs its sensitive anterior end away from the source of light (by the
method of '* trial and error") and swims to the opposite side of the
drop, incidentally moving into a region of slightly greater (but
" unperceived ") intensity of illumination. Under similar conditions,
as we have seen in the experiment described on p. 39, if the decrease
in illumination is marked, the animal swims back into the shadow,
though in so doing it passes toward the source of light.

Thus in Strasburger's experiments with the prism the difference in
the intensity of light between neighboring regions has been made so
slight that they are unmarked by the organism and have no effect upon
it. We need not be surprised, therefore, that it reacts as if these differ-
ences did not exist ; for the organism they do not exist.

The reaction is in this case just what it would be in a higher organ-
ism under similar conditions. Let us suppose that the light stimulates
strongly the sensitive anterior end, the eyes, of a higher animal or man ;
it causes pain in the case of man. There will be a tendency (i) to
move into less illuminated regions ; (2) to turn the eyes away from the
light. Suppose that the man is enclosed in a space into which the
sun shines obliquely from above, and that the end from which it shines
is a little less illuminated than the opposite end, owing to causes similar
to those in Strasburger's experiment on the swarm-spores. Suppose that
the man is at the end next the sun. He cannot know that the other
end is more illuminated, for the only way this would be possible would
be for the greater number of rays of light to meet his eye coming from
that direction, while by hypothesis all, or a much larger number, of
the rays are coming from the opposite direction. He will, therefore,
turn his eyes away from the sun, and if he moves will move toward
the end away from the sun. After having traversed some distance he
may observe, if he is very discriminating, that he is as a matter of fact
getting into a region of somewhat greater illumination, and may perhaps
reason that the best thing he can do under the circumstances is to keep
his eyes turned away from the source of light and move backward to the

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 69

less illuminated end. But this involves the capability of making fine
distinctions, and a considerable degree of intelligence in deciding what to
do under the peculiar circumstances. The experiment with the swarm-
spores shows that they are incapable of such fine discrimination, or that
they are not sufficiently intelligent to know what to do under the circum-
stances. They give no indication that they notice the greater illumina-
tion after having passed to the end away from the light. Their action
may be considered perhaps in a certain sense as a " mistake," but it
is a mistake which even the highest organism would make, and which
could be corrected only after experience of its results.

The results of our study of the light reaction in ciliates and flagel-
lates lead to conclusions which stand in sharp contrast with certain
general conclusions in Radl's recent extensive and interesting paper on
Phototropism (Radl, 1903). Radl reaches the somewhat extraordinary
conclusion that light orients organisms by exercising an actual
mechanical pressure upon them. This pressure necessarily disturbs
the equilibrium of the body, which is then compelled to change posi-
tion until equilibrium is restored ; the organism is then oriented. The
orientation is a consequence of the interplay of two sets of forces, inner
and outer ; these cannot be in equilibrium until the body has taken a
certain position with reference to the pressure exercised by the light
(/. c, pp. 151 fT.) The actual turning which induces orientation must
be due to the action of a pair of forces (/. ^., p. 14^). One of these
forces is the pressure produced by the light.

Orientation produced in the manner described in the present paper
for the reaction of ciliates and flagellates to light, and in the preceding
paper for the reaction to heat, could of course not be brought about in
the manner supposed by Radl. One of Radl's chief arguments for his
view is that "no observation thus far shows that the final orientation
is attained by a trial or after an oscillation, but it takes place auto-
matically"* (/. c, p. 141).

The observations on ciliates and flagellates given in the present paper
show conclusively that the orientation in these cases is brought about
through repeated trials. In the statement quoted above Radl has over-
looked certain other cases. Thus Strasburger, as we have seen (p. 59),
states that after the direction of the light is changed Hsematococcus
becomes reoriented " nach verschiedenen Schwankungen" (Strasbur-
ger, 1878, p. 24). Radl himself refers on a previous page (p. 100) to
Strasburger's observation of the oscillating movement of swarm-spores
under the influence of a variation in light intensity; Rothert (1901,

*" Keine bisherige Beobachtung zeigt ferner, dass die schliessliche Orientie-
rung etwa durch eine Priifung oder nach einem Schwanken erzielt wurde,
sondern sie folgt automatisch."

70 THE BEHAVIOR OF LOWER ORGANISMS.

p. 397) has called particular attention to this as a possible factor in the
so-called phototropism. Radl also refers (p. 99) to Exner's view that in
Copilia the movements of the eyes are in the nature of a trial (" abtasten ")
of the surroundings. Radl's statement, quoted above, can then hardly
be considered strictly accurate, even leaving out of consideration the
results set forth in the present paper. In many organisms, doubtless,
the reaction to light is of that direct character assumed to be general by
Radl. But it may be strongly doubted whether this is what we may
call a primitive condition ; in other words, whether it does not involve
more complicated internal processes than the reaction by '* trial and
error." In any case, I am convinced that a similar reaction to light by
the method of " trial and error " will be shown to exist in many other
organisms ; it is demonstrated, for example, in Rotifera, in the paper
which follows the present one.

Recourse will doubtless be taken to the usual refuge when a sharp
concept has been defined to which the phenomena are not found to
correspond ; the reactions of the ciliates and flagellates will be simply
excluded from the tropisms and the definition of the latter maintained
in all its pristine purity. Indeed, it may be questioned whether the
reactions of infusoria (and Rotatoria) to light are not excluded from
phototropism through the definition given by Radl on p. 140, what-
ever the method by which they are produced. Radl says " Unter
phototropischer Orientierung ist die Fahigkeit der Organismen zu
verstehen, eine feste Einstellung der Achsen des gesamten Korpers in
dem Lichtfelde einzunehmen." Since the ciliate or flagellate (or
rotifer) revolves continually on its long axis, and swerves continually
toward a certain side, it can hardly be said that the body axes have a
" feste Einstellung" with reference to the light. In an explanatory
paragraph Radl says that in orientation "immer geht dann der
Lichtstrahl durch die (morphologische) Symmetrieebene des Korpers"
(/. c, p. 140). This is certainly not true for the ciliate or flagellate (or
rotifer), even leaving out of consideration the fact that in the former
two groups the animals are usually unsymmetrical. If it be proposed,
then, to exclude the light reactions of ciliates, flagellates, and rotifers
from the concept of *'Phototropismus," one can only agree that this is
necessary, in view of the definitions of that concept.

But what is the value of a definition which excludes some of the chief
phenomena on which the concept that we are attempting to define is
based .'* And what is the value of a theory that depends on such a
definition and that can only be correct so long as we hold to this
definition.? The phenomena themselves are, after all, the final refer-
ence for testing the correctness of any definition or theory ; it is the
observed phenomena that we are attempting to formulate and explain.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 7 1

What we desire in the study of animal behavior is (i) a correct
description of what occurs ; (2) an understanding of the relation of
what occurs here to other phenomena, this constituting their "explana-
tion" so far as an explanation is possible. Whether the phenomena
when correctly described and understood are found to fall under some-
one's definition of a tropism is comparatively unimportant ; it is only
after such correct description and understanding that final definitions
can be made. I question much if there has not been undue haste in
framing precise definitions for the phenomena of animal behavior, when
we know so little about the phenomena in any thorough way. Radl,
I believe, makes a fundamental error in attempting to separate " Pho-
totropismus" rigidly from other reactions to light. Thus, he repeat-
edly cites Euglena as an example of an organism that shows undoubted
phototropism. On page 114 he further cites the motor reaction of
Euglena when suddenly shaded * as a reaction that has nothing to do
with phototropism. As I have shown above, the two are really
closely bound up together ; the orientation in the "phototropism" is
produced through this motor reaction. When the reactions of organ-
isms to light are known in detail, I believe that many other reactions
which Radl (p. 1 14) attempts to separate sharply from " phototropism "
will be found closely connected with the reactions that go under that
name. I had occasion to point out, in the paper preceding this, on
the reactions to heat, that if everything which the organisms do, except
the orientation itself, is left out of consideration, the orientation can be
accounted for by any theory desired. A thorough study of precisely this
point — the relation of "phototropism" to the phenomena supposedly
unconnected with it — would, I believe, have saved Radl from marring
his otherwise most excellent and useful contribution to the study of
light reactions by the proposal of so fantastic a theory to account for
the reactions to light ; a theory that fiiirly produces a shock in the mind
of the reader when it is reached, coming as it does after Radl's thorough
and valuable objective study of many of the phenomena and his exceed-
ingly sane, if somewhat sharp, criticism of other theories. Definition
and precise classification are of course valuable at a certain stage of
knowledge, but when carried out without a thorough knowledge of the
phenomena dealt with they may be a hindrance rather than a help.
The thorough knowledge of the phenomena of animal behavior required
for this is far from existing at present.

♦Radl says when " beleuchtet"; this is evidently a slip of the pen.

THIRD PAPER

REACTIONS TO STIMULI IN CERTAIN
ROTIFERA.

73

REACTIONS TO STIMULI IN CERTAIN ROTIFERA.

In my series of ''Studies on Reactions to Stimuli in Unicellular
Organisms " and in the foregoing papers I have set forth the reaction
methods of many infusoria to varied stimuli. The result has been
to show that the reaction method in these organisms is of a peculiar
character, differing radically from that required by prevailing theories
of the reactions of lower organisms. The essential nature of these
reactions, with their implications as to the character of behavior in the
lower organisms, will be discussed in the following papers. Before
proceeding to this discussion it is important to determine whether the
reaction method in the Infusoria differs radically in character from
that of Metazoa. For this purpose it seems well to select a group of
Metazoa whose habitat and mode of life are similar to those of the
Infusoria. In this way differences due primarily to the different plan
of structure of the two sets of organisms may perhaps be brought out
without the complications arising from different modes of life.

A group of Metazoa much resembling the Infusoria in their mode
of life is found in the Rotatoria. As is well known, the members
of these two groups are usually found mingled together. They are
of about the same size, and both swim about by means of cilia. So
great is the resemblance in general habit and in habitat that they were
at first classed together, all being given the name of Infusoria. As
we know now, however, they are really widely separated in relation-
ship. While the Infusoria are unicellular, the Rotifera are multicellu-
lar organisms of a high degree of complexity, possessing many systems
of organs, each composed of many cells. In particular, they have a
well-developed nervous system.

A comparison of the behavior of these two groups of organisms
should show us, therefore, whether there are types of reaction having
a high degree of generality, such as is claimed for the theory of
tropisms — types that may give a key to the behavior of groups so
widely separated in relationship as the two under consideration, which
are representatives of the Protozoa and of Metazoa of a fairly high
degree of organization.

In the present paper I can attempt to give an account of the behavior
of only a few free-swimming species, and that not in an exhaustive
manner. I hope to return to an extensive study of the behavior of this
interesting group, so as to develop its implications for the theory of

75

76

THE BEHAVIOR OF LOWER ORGANISMS.

animal behavior in general. In the study here set forth observation
"^-^^ was directed primarily to the questions of howr

certain Rotifera react under the stimulus of
^ the agencies which usualh' give rise to the so-
called tropisms — light, chemicals, heat, elec-
tricity, contact, etc. — and to these questions
the present account will be devoted.

The species whose reactions were exam-
ined belong chiefly to the loricate group of
free-swimming Rotifera, and include a num-
ber of species of the Rattulidae, several species
of Cathypnadae, two or three species of Euch-
lanis, Plcesoma lenticular e^ Anurcea cochle-
ar is ^ and Brachionus pala. These were
studied as opportunity offered. In most cases
the reactions of any one species were not
determined with relation to more than two
or three classes of stimuli. The behavior of
Anurcea cochlear is was examined most fully.
This species will be used as a type in describ-
ing the reactions. I have already given a
brief account of the general reaction type in
certain species of the Rattulidag in my mono-
graph of that group (Jennings, 1903).

METHOD OF LOCOMOTION.

The free-swimming Rotifera progress

through the water in the same manner as the

ciliate infusoria. The cilia in the Rotifera

are limited to the anterior end, as they are

in the peritrichous infusoria. It is interesting

to note that the same device is adopted in the

one group as in the other, to compensate for

irregularities in the form of the body, etc.,

Fig. 25.* which might result in swerving from the

straight course. This is by revolution on the long axis, causing the

path to become a spiral with a straight axis. In the Infusoria the

* Fig. 25. — Spiral path followed in ordinary swimming by Anuria cochlearis
Gosse, showing different positions of body in different parts of the course;
a, dorsal surface; 3, left side; c, ventral surface; d, right side. The animal
revolves on its long axis over to the right, thus taking successively the positions
«, *, c, </, a, etc. The large arrow indicates the general direction of the course
followed; the smaller arrows show direction of progression in certain parts of
the course.

REACTIONS TO STIMULI IN CERTAIN ROTIFERA. 77

organism usually swerves from the straight line toward the aboral side ;

in the Rotatoria it is usually toward the dorsal side. Well-ordered
forward progression would therefore not take place, were it not for the
revolution on the long axis, converting the circular course into a spiral
one. In the Rotifera the revolution on the long axis is, so far as
observed, always over to the right. These relations have been brought
out in detail in a previous paper by the present author (Jennings, 1901).
The spiral path thus followed by most of the free-swimming Roti-
fera may be illustrated in Fig. 25, for Anurcea cochlear is Gosse. As
will be seen from the figure, the path followed depends upon three
factors : (i) the animal continually swerves toward its dorsal side ; (2)
it progresses ; (3) it revolves on its long axis. The result of these three
factors is the spiral course. In all these relations the rotifer agrees
with the infusorian.

REACTIONS TO STIMULI.

The most general reaction to a stimulus in such a free-swimming
rotifer is an accentuation of one of the factors in this course, namely,
the swerving toward the dorsal side. The result is to produce a spiral
of much greater width than previously existed. This may often be
observed when the vessel containing the rotifers is jarred. It is evi-
dent that this method of reaction is fitted to enable the rotifer to avoid
a small obstacle lying in its path, that is, in the axis of the spiral.
When the animal resumes its former method of swimming the axis of
the spiral lies in a new direction ; the course has thus been slightly
changed.

With a stronger stimulus, as when the rotifer strikes against an
object lying in its path, the swerving toward the dorsal side may be
still more pronounced, while the revolution on the long axis nearly or
quite ceases. The result is that the organism swings strongly toward
its dorsal side, and when the usual forward swimming is resumed the
axis of the spiral lies in a totally new direction (Fig. 26). It thus
avoids the obstacle, if the latter is small ; if the first reaction does not
avoid the obstacle completely the reaction is repeated until the course
is sufiiciently altered so that the rotifer no longer strikes against the
source of stimulus. In some rotifers the increased swerving toward
the dorsal side is preceded by swimming backward a short stretch.

In all these points the reaction of the rotifer agrees even to details
with that of the ciliate infusorian. There is a difference in the fact
that the Infusoria are unsymmetrical and cannot therefore be said to
swerve toward the dorsal side, as do the prevailingly symmetrical
Rotifera. In the Rattulidas, however, we have asymmetry of a char-
acter similar to that found in the Infusoria.

78

THE BEHAVIOR OF LOWER ORGANISMS.

We have dealt thus far specifically only with reactions to simple
mechanical stimuli, such as are presented by an obstacle in the path of

Fig. 26.*

the rotifer or by a simple mechanical jar. This type of reaction under
such conditions I have observed for Dlurella tigris Miiller, D. por-

* Fig. 26 is a diagram of the reaction of the rotifer Anuraea to a strong stimulus,
as when it reaches a source of mechanical stimulus or a region where some
chemical is dissolved in the water. From a to b the animal is unstimulated,
hence it follows the usual spiral course. At b it reaches the stimulating region,
whereupon it turns strongly toward the dorsal side, following the arc of a circle,
from b to d. Here it resumes the usual spiral course {d to e). The large arrow
M shows the general direction of progression before the stimulus was received;
the arrowy' shows the direction of progression after the reaction has taken place.

REACTIONS TO STIMULI IN CERTAIN ROTIFERA.

79

cellus Gosse, D, gracilis Tessin, and a number of other species of
Rattulidae ; Plcesoma lenticulare Herrick, Cathypna ungulata Gosse,
Monostyla bulla Gosse, Brachionus pala Ehr., Anurcea cochlear is
Gosse, and A. aculeata Ehr.

This method by no means exhausts the possibilities of reaction even
to a simple mechanical stimulus in these species. They may retract
the head and cease swimming, may creep over the surface of the object
with which they come in contact, or possibly may sometimes turn other-
wise than to the dorsal side when stimulated. Of this latter point I
am, however, by no means sure. It is certain that the typical reaction,
occurring in the great majority of cases, is that described above.

REACTION TO CHEMICALS.

The reaction given when the organism comes in contact with an area
containing a rather strong
diffusing chemical was
observed in Metopidia
lepadella^ Anurcea coch-
lear is ^ A. aculeata^ and
Diurella gracilis.

The method of experi-
mentation was as follows :
A drop of water contain-
ing the rotifers was placed
on a slide. Near this was
placed a drop of N/8
NaCl, and the two drops
were connected by a nar-
row neck. The behavior
of the organisms as they
came into the region of the neck and thus in contact with the salt
solution was observed with the Braus-Driiner microscope. In the
species mentioned the reaction was by a sudden turn toward the dorsal
side, by which the path of the animal was directed away from the
chemical. The reaction is thus of the same character as occurs in the
ciliate infusoria.

This manner of reaction to chemicals is in both these groups of
organisms just what might be expected when the currents caused by
the cilia are taken into consideration. In the ciliate, as I have shown

Fig. 27.*

♦Fig. 27.— Diagram of currents in a nearly quiet Anuraea, showing how a
diffusing chemical or an advancing region of warmer water (represented by shad-
ing), is drawn out by the ciliary vortex, so as to reach the mouth and the ventral
surface before affecting other parts of body.

8o THE BEHAVIOR OF LOWER ORGANISMS.

in a previous paper (Jennings, 1902, a) , the cilia cause a current coming
from the region in front of the organism to pass along the oral surface
to the mouth ; in this way the oral surface comes in contact with the
chemical before any other part is afiected. It is not surprising, there-
fore, that the organism should turn toward the opposite (aboral) side.

In the rotifera the conditions are parallel to those found in the ciliate.
The cilia cause a current, which passes to the mouth, on the ventral
surface of the body (Fig. 27) . The solution thus reaches the ventral
surface first, and the reaction is, as might be expected, a turn toward
the dorsal side.

It should be distinctly stated that this reaction method is not universal
in rotifers even toward chemical stimuli. In some of the larger species,
bearing auricles, or with the ciliary apparatus of a very complex
character in other respects, varied reactions may occur, which I hope
to analyze in another paper.

REACTION TO HEAT.

This was studied in detail only in Anurcea cochlearis. A large
number of the rotifers were mounted in a shallow trough formed of a
slide, as described on p. 12, and one end of the slide was warmed by
means of the apparatus shown in Fig. 5. The reactions were then
observed with the Braus-Driiner stereoscopic binocular.

As soon as a portion of the slide has been warmed above the optimum,
the rotifers in this region turn more strongly than usual toward the
dorsal side, so that the course followed becomes a very wide spiral and
the animals make little progress. If the heat is increased the revolu-
tion on the long axis ceases, while the animals swerve still more
strongly toward the dorsal side (Fig. 28), so that they swim in circles,
the dorsal surface being directed toward the center of the circle.
Usually after circling thus a short time the animals begin again to re-
volve on the long axis, and dart forward. The direction of this dart
forward seems purely random. If it carries the animal out of the
heated region the forward movement is continued and the animal
escapes. If it does not carry the animal out of the heated region the
circling toward the dorsal side is quickly resumed, followed by another
dart forward. This is continued either until the rotifer passes out of
the heated region or until it is overcome by the heat. Usually, if it
does not escape soon from the heated region the circling becomes
more rapid and continuous and is kept up till the animal is destroyed
by the heat.

If one end of the slide is heated and the animal approaches the
heated region from the opposite end the reaction is of the same
character as that last described. As soon as the region is reached

REACTIONS TO STIMULI IN CERTAIN ROTIFERA.

8l

where the heat acts as an effective stimulus the animal swerves strongly
toward the dorsal side, thus beginning to circle, as shown in Fig. 28.
If this swerving should continue only till the animal had described a
semicircle, then were followed by the forward dart, the animal would
of course retrace its original course (or one parallel to it) , and would
thus escape from the heated region, as happens in the reaction to the
electric current (Fig. 29) . But the reaction to heat is less precise than

f

Fig. 28.

this. ■ Usually the animal makes several com-
plete circles before darting forward, and the
direction in which it darts seems a random
one ; sometimes it is toward the heated region,
sometimes away from it, sometimes oblique to
it. If the path followed leads the animal into
the heated region the circling toward the dorsal
side, followed by the dart forward, is repeated ; while if the path leads
away from the heat no farther reaction is caused and the animal escapes.
Thus when a large number of the animals swim toward the heated
region a considerable number will be seen a little later to swim away
again. But in many cases the dart forward carries the animal still
farther into the heated region. These specimens then begin to circle
again toward the dorsal side, and if the temperature is high they may

♦Fig. 28. — Diagram of a reaction to heat in Anuraea. The unstimulated
animal at first advances in the general direction shown by the arrow x, following
thus the course a to e. The heat is supposed to be advancing from the direction
opposite the arrow x. When the rotifer reaches the point e the heat becomes
effective as a stimulus. The animal reacts by turning toward the dorsal side,
and continues this so as to describe a complete circle,/", g; A, /',/", etc. ; often it
describes such a circle several times. Finally, at some point in the circular
course, as £-, it resumes the usual spiral course, following thus the path ^, /, /.
Its original course, shown by the arrow x, has thus been exchanged for a course
having the general direction shown by the arrow v.

82 THE BEHAVIOR OF LOWER ORGANISMS.

continue this till death intervenes. In many cases they repeat the dart
forward and some escape in this way, while others do not.

The reaction of Anuraea to heat is, therefore, not very precise, and
many individuals swim into the heated region and are killed. Those
which escape do so through a reaction which is similar to that of
those which do not ; in the one case the forward movement carries the
animal out of the heated region ; in the other it does not. The essential
point to the reaction is that the animals when stimulated by heat change
their course (through a '' motor reflex"). This changed course nat-
urally is an advantage, and in accordance with the laws of probability
carries some of the organisms away from the source of danger. Others,
likewise in accordance with the laws of probability, are carried even
by the changed course toward the heated region, where they may be
killed unless a repetition of the "motor reflex" with its change of
course carries them finally away. The reaction is by the method
of" trial and error," and is not always successful.

Altogether, the reaction of the rotifer Anursea to heat is of a charac-
ter similar in principle to that of Oxytricha (Fig. 7, p. 16). The
direction of turning depends on an internal factor ; the reaction takes
the form of " a motor reflex," and is by no means compatible with the
typical tropism schema.

REACTION TO LIGHT.

In light, as I have already set forth in the account of the reactions of
Stentor, we have a stimulating agent of a different character from that
found in chemicals or in heat, since the distribution of the stimulating
agent is not affected by the currents of water produced by the motor
organs of the animal. There is thus no reason in the distribution of
the stimulating agent to favor a turning toward one side rather than the
other.

I have been able to study accurately the light reaction in but one
rotifer, Anurcea cochlearis Gosse. The conditions necessary for
precise observation of the nature of the reaction are very difficult to
fulfill, and the usual movements of the animals are such that the nature
of the reaction is obscured. As will be recalled, the organism is
normally swimming rapidly in a spiral, continually swerving toward
its dorsal side. This in itself is very confusing when one attempts to
observe just how the organism turns when stimulated. When light is
thrown upon it, or when the direction of light falling on it is changed,
the response is usually not given at once, and when it does occur, as
we shall see, it may be in the form of an accentuation of certain features
of the normal movement. From these conditions it results that it is
exceedingly difficult to tell, after a reaction to light has clearly occurred,

REACTIONS TO STIMULI IN CERTAIN KOTIFERA. S3

just how the reaction took place. Of course, only sharply defined posi-
tive observations are of value in deciding between two opposing possi-
bilities ; hence, although I have studied a number of other rotifers in
this connection, I give the results only where absolutely sure of them.
But in the two or three other rotifers I have examined in this connection
the reaction is apparently the same as that in AnurcEa cochlearis^ to
be described at once.

The specimens of Anurcea cochlearis studied had been in a small
aquarium in the laboratory some months, and were distinctly negative
to light, gathering at the side of vessel farthest from the window. The
freshly collected animals are, I believe, usually positive to light.

These negative individuals were placed in a small flat-bottomed
rectangular glass vessel, on a dark background, in a dark room. At
opposite sides of the vessel and somewhat above were clamped two
incandescent electric lights, A and B^ at a distance of about lo inches*
in the manner described for Stentor (p. 41 and Fig. 15). One of
these lights could be extinguished while the other was simultaneously
turned on. In this way the direction of the light falling on the rotifers
could be reversed.

When only one of the lights, as A^ was turned on, the Anuraeas all
collected at the opposite side of the vessel, next to B. When A was
extinguished and B turned on, they turned and swam in the opposite
direction, toward A. By reversing the direction of the light while the
animals were crossing the vessel their course could be reversed while
in full career.

Focusing the Braus-Driiner on the vessel, and reversing the lights
when the animals were well in the field of observation, the following
could be observed : Some turned at once, with some sharpness,
toward the dorsal side^ the turning continuing until the direction of
swimming was reversed and the animals were again swimming away
from the light (Fig. 29) . In these cases the direction of turning was
clear and could be observed without great difficulty.

Other individuals continued for a short time to swim in the same
direction as before, then turned, either sharply, as just described, or
more slowly, in the manner to be described.

Where the turning was sharp, as described above, there was no
great difficulty in determining with certainty the nature of the reaction.
But in many cases the turning took place more slowly, in the following
manner : Either as soon as the light was reversed, or very soon after,
the width of the spiral in which the animal was swimming became
much greater. In other words, the animal swerved more toward the
dorsal side and progressed less rapidly than usual. Thus it described
rather wide circles, and the swerving toward the dorsal side increased,

84 THE BEHAVIOR OF LOWER ORGANISMS.

while progression and revolution on the long axis had largely ceased.
After this circling had continued for some time, the swerving toward
the dorsal side apparently continually increasing, it was found that the
anterior end was directed away from the source of light; /. e.^ the
direction of swimming had been reversed, and the animal was moving
away from the light.

It is obviously very difficult to be entirely certain of all that has hap-
pened during this period of extensive circling, as a matter of direct
observation. But the evidence seems to show clearly that the essential
point in changing the course is the swerving toward the dorsal side.
The following facts all point to this conclusion : (i) In the individuals
which turn at once it is possible to be entirely certain that the turning
is toward the dorsal side. (3) In the individuals which are circling it
is entirely clear that the swerving toward the dorsal side is greatly in-
creased, and there is no evidence of turning in other directions. The
only difficulty is that one cannot follow every evolution and be certain
that nothing else has occurred. (3) Analysis of this same reaction
when given in response to other stimuli, where the conditions are
more favorable for observation, shows that it does consist of an in-
creased swerving toward the dorsal side, with a decrease, or an entire
stoppage for a time, of the forward motion. There is, then, no reason
to think that the reaction contains other factors when performed under
the influence of light. The reaction is indeed clearly the same as that
described for Euglena on p. 53, and illustrated in Fig. 21 ; a similar
analysis could be given for the reaction of Anuraea.

It may be considered certain, therefore, that in Anurcea cochlearis
the reaction to light is similar to the reaction to other stimuli, and that
the orientation is brought about by a turning toward the dorsal side.
The reaction is, therefore, not due to the direct effect of the light on the
motor organs ; the direction of turning is determined not by external
factors, but by internal factors. The reaction to light in the rotifer,
like that in the infusorian, takes place by the method of " trial and
error."

REACTION TO THE ELECTRIC CURRENT.

A considerable number of different species of the rotifera were sub-
jected to the continuous electric current without the production of any
characteristic reaction. A current was used which could be graded in
strength from practically zero to one that was destructive, but no reac-
tion comparable to that found in the ciliate infusoria was produced.
On making or breaking the current the animals frequently contracted
quickly, and if the current was very strong, the head was completely
retracted and the animal sank to the bottom. But there was no orien-
tation and the animals did not swim toward either electrode. These

REACTIONS TO STIMULI IN CERTAIN ROTIFERA. 85

negative results were obtained with several of the Philodinadse (Rotifer,
Philodina), some species of Euchlanis and Salpina, Noteus quadri-
cornis^ and a number of the Rattulidas.

In Hydatina senta there is a reaction of a peculiar character which
perhaps furnishes a clue to the cause of the more pronounced reaction
to be described for Anuraea. With a current of moderate strength,
such as that to which Paramecia react most markedly, Hydatina shows
no reaction except when the head is directed toward the anode. But
in this position the animal at once retracts its cilia and sinks to the
bottom. Thus a Hydatina may swim freely about in water through
which the current is passing, provided it swims toward the cathode, or
transversely, or obliquely ; as soon, however, as it turns its head toward
the anode it stops swimming and sinks to the bottom. Thus if an
electric current is passed through a preparation containing a large
number of specimens of Hydatina, many will be seen swimining
toward the cathode and others at all sorts of angles with the current, but
none toward the anode. This is a phenomenon akin to what I have
elsewhere called the production of orientation by exclusion. If organ-
isms are prevented from swimming in any direction but one, after a
time, provided the course is frequently changed, all that are swim-
ming will be found moving in that one direction. This condition is
realized, as I have shown in the first of these contributions, in the
reactions of infusoria to heat and cold. But in the reactions of Hydatina
to the electric current the " exclusion" is less complete than in the
cases just mentioned ; the animal may swim in any direction except
one.

The fact that the head is retracted when directed toward the anode
and not in other positions indicates that there is a greater stimulation
at the anode than elsewhere. This agrees with much that is seen in
the reactions of infusoria to the current. After Hydatina has sunk to
the bottom with anterior end to the anode, it repeatedly makes attempts
to unfold its cilia. But scarcely have they begun to operate when
they are withdrawn again. Each time that they are uncovered for an
instant, however, they turn the animal a little toward its dorsal side.
Thus, after a considerable number of attempts to unfold the cilia, the
head has become turned away from the anode ; then the cilia are spread
out and the animal goes on its way until it is so incautious as to turn
its head again toward the anode.

Anurcea cochlearis shows marked electrotaxis similar to that found
in the infusoria. When the continuous current is passed through a
preparation containing large numbers of this species, all orient quickly
and swim toward the cathode. They thus agree, so far, in their
reaction to the electric current, with the ciliate infusoria.

86 THE BEHAVIOR OF LOWER ORGANISMS.

The question as to the mechanism of the electrotactic reaction in the
rotifer is of interest when one compares the structure of these animals
with that of the ciliate infusoria. The Rotifera, in place of having cilia
scattered over the entire body, are furnished only with a group of
cilia at the anterior end. In the Ciliata it is usually possible to distin-
guish functionally two groups of cilia (i) the a do ral cilia., about the
mouth and oral groove, or at the anterior end ; (2) the body cilia,
scattered over the body. The cilia of the rotifers correspond function-
ally with the adoral cilia of the Ciliata.

Pearl (1900), Wallengren (1902-1903), and others have shown that
in the electrotactic reaction of the ciliates the two sets of cilia are in
many cases from a functional standpoint differently affected. The
adoral cilia react under the influence of the electric current in such a
way as to tend to turn the organism toward the aboral side ; that is,
they tend to produce the same reaction which the organism gives in
response to most other stimuli, a reaction not in harmony with the
tropism schema. The body cilia, on the other hand, are differently
affected on the different sides or ends of the organism ; those on the
part of the body directed toward the cathode striking in one direction ;
those on the part directed toward the anode striking in a different
direction. The result is that the organism, through the action of the
body cilia, tends to become directly oriented in a way that is in harmony
with the tropism schema. (For details, see the papers cited.) In those
ciliates in which the body cilia are much reduced, as in the Hypotricha,
the turning is determined throughout by the adoral cilia, so that the
orientation does not take place in accordance with the tropism schema,
while in some others, such as in Paramecium, the influence of the body
cilia is predominant, and the turning is in accord with the theory of
tropisms.

What conditions shall we find in the Rotifera, where the only exist-
ing cilia seem to agree functionally with the adoral cilia of the Ciliata?

As we have seen, Anuraea swims as a rule in rather wide spirals,
swerving strongly toward the dorsal side and revolving on its long axis
(Fig. 25). When the electric current suddenly acts upon it the organism
at once turns strongly toward the dorsal side, continuing the turn until
its head is brought toward the cathode, toward which it swims (Fig.
29). In some cases, as we shall see later, several reactions are neces-
sary for bringing the body in line with the current, but these are as a
rule very quickly accomplished.

If, while the animals are swimming toward the cathode, the current
is suddenly reversed, the animals again turn strongly toward the dorsal
side, continuing the turning until their position is reversed and the
heads point toward the new cathode (Fig. 29). In many cases the

REACTIONS TO STIMULI IN CERTAIN ROTIFERA.

87

turning is continued still farther, so that the head of the animal de-
scribes a complete circle ; indeed, this may continue so that the animal
whirls around several times, always towards the dorsal side. The
reaction thus far is the same as that produced by heat (Fig. 28). In
reacting to the electric current the whirling finally ceases with ante-

FiG. 29.*

rior end directed toward the new cathode. The animal then swims
forward in the direction so indicated. These turnings, even when
several times repeated, require but a moment, so that very soon prac-
tically all the specimens are swimming toward the new cathode. The

* Fig. 29. — Diagram of method by which Anuraea becomes oriented to rays of
light, or to the electric current. Taking the latter for example, the animal is at
first swimming toward the cathode, in direction indicated by arrow At; it thus
follows a spiral path from a to b. At b the electric current is reversed. The
animal thereupon swerves stronglj' toward its dorsal side, describing a semicircle,
3, c, </, until its anterior end is directed toward the new cathode, in the opposite
direction from before. It now follows the spiral path d to e., in the general direc-
tion indicated by the arrowy'. The facts are similar for the reversal of light, or
for the reaction when the current or the light is first set in operation.

S8 THE BEHAVIOR OF LOWER ORGANISMS.

reaction in AnuraBa is in a general view as striking and clear-cut as
that of Paramecium.

Thus in the rotifer Anurasa the orientation to the continuous elec-
tric current is produced through a motor reaction, the essential features
of which are determined by the structure of the organism. The organ-
ism turns always toward the dorsal side, continuing or repeating the
turning until the anterior end is directed toward the cathode. In these
respects it agrees with hypotrichous Ciliata, where the direction of
turning is determined by the action of the adoral cilia. The method
of reaction is quite incompatible with the tropism schema.

SUMMARY.

The reactions of those Rotifera of which an account is given in this
paper take place in a manner essentially similar to the reactions of the
ciliate infusoria.

In the reactions to mechanical stimuli, to chemicals, and to heat,
orientation is not a striking feature. The organism turns when stimu-
lated toward a structurally defined side — as a rule toward the dorsal
side ; in this way it avoids the source of stimulus.

In the negative reaction to light the organism becomes oriented with
anterior end directed away from the source of strongest light, but this
orientation is brought about in the same manner as in Stentor ; the
animal turns toward the dorsal side without relation to the side on
which the light strikes it, and continues the turning or repeats it until
the anterior end is directed away from the source of light.

To the continuous electric current the rotifer Anuraea orients and
swims directly toward the cathode. The reaction is brought about in
the same manner as the orientation to light. When the current is
made or reversed the animal turns toward the dorsal side and continues
the turning until the anterior end is directed toward the cathode.

Thus the direction of turning is throughout dependent on an internal
factor, not primarily on the way in which the stimulus impinges on
the organism. These reactions of the Rotifera are thus inconsistent
with a theory of tropisms which regards orientation as a primary
feature of the reactions, and which holds that the action of the stimu-
lating agent is a direct one on the motor organs of that part of the
body on which it impinges. The reactions of the Rotifera, so far as
described in the preceding pages, are brought about, like those of the
infusoria, by what may be called the method of ''trial and error."
The reaction to any stimulus is of such a nature as to head the organism
successively in many different directions. That direction is followed in
which there is no stimulus to induce further turning.

FOURTH PAPER.

THE THEORY OF TROPISMS,

CONTENTS.

PAGE.

To what Extent does the Theory of Tropisms throw Light on the Behavior

of Lower Organisms ?.......... 91

Essential Points in the Theory of Tropisms, ...... 92

Reactions to Mechanical Stimuli, ........ 94

Reactions to Chemicals, . . ........ 96

Reactions to Heat and Cold, ......... 98

Reactions to Changes in Osmotic Pressure, ...... 98

Reactions to Light, ........... 98

Reactions to Gravity, .......... 100

Reactions to Electricity, .......... 100

R^sum^ and Discussion, . . . . . . . . . .103

Summary, 106

90

THE THEORY OF TROPISMS.

TO WHAT EXTENT DOES THE THEORY OF TROPISMS THROW
LIGHT ON THE BEHAVIOR OF LOWER ORGANISMS?

The writer has been engaged for a number of years in a study, as
exact and detailed as possible, of the behavior and reactions of a num-
ber of lower organisms. While the results obtained have not, as a
rule, agreed with the view that the behavior of these organisms is
determined largely in accordance with the prevailing theory of tropisms
or taxis, he has not discussed their relation to this theory in detail.
This was because of the possibility that the reactions which he had
studied were exceptional, and that further investigation might show
after all that the behavior of the lower organisms is largely in accord-
ance with the tropism schema.

At the present time the writer feels that the work which he has
done, or which has been done by those associated with him, is of suffi-
cient extent to justify the pointing out of certain general relations.
The reactions of ciliate infusoria, which have long been used as the
types of illustration for the tropisms, have been examined in much
detail, and less extensive studies have been made on the Bacteria
(Jennings & Crosby, 1901), the Flagellata, and the Rotifera. The
reactions of a flatworm have been studied in much detail (Pearl,
1903), and researches are nearly ready for publication, by investigators
associated with the author, on the behavior of Hydra and of the leech,
and still other studies are under way. Thorough studies, directed to
the observation of the exact movements of organisms under stimuli,
have recently been given us by other observers also. It seems, there-
fore, worth while to bring out, in a preliminary way at least, the
relation of the observations made to the prevailing theories of animal
behavior. In the present paper this will be limited to a consideration
of the theory of tropisms, since this is the theory most widely held.

The great apparent value of the theory of tropisms or taxis lies in
the fact that it seems to reduce to very simple factors a large number
of the most striking activities of organisms, namely, those involved in
going toward or away from sources of stimuli of almost any character.
It is a schema, in accordance with which almost any movements of the
organism (not purely random) might be supposed to take place.

91

92 THE BEHAVIOR OF LOWER ORGANISMS.

ESSENTIAL POINTS IN THE THEORY OF TROPISMS.

The two essential features of the theory of tropisms are apparently
the following: (i) The movements of organisms toward certain
regions and their avoidance of others are due to orientation ; i. e.^ to
a certain position which the organism is forced by the external stimulus
to take, and which leads the organism toward (or away from) the
source of stimulus, without any will or desire of the organism, if we
may so express it, to approach or avoid this region. (2) The external
agent by which the movement is controlled produces its characteristic
effect directly on that part of the body upon which it impinges. It
thus brings about direct changes in the state of contraction of the
motor organs of that part of the body affected as compared with the
remainder of the body, and to these direct changes are due the changes
shown in the movements of the organism. This is brought out clearly
in the quotation from Verworn given on page 8. Loeb (1900, p. 7)
sums up the theory of tropisms as follows :

The explanation of them [the tropisms] depends first upon the specific irrita-
bility of certain elements of the body surface, and, second, upon the relations
of symmetry of the body. Symmetrical elements at the surface of the body
have the same irritability; unsymmetrical elements have a diflferent irritability.
Those nearer the oral pole possess an irritability greater than that of those
near the aboral pole. These circumstances force an animal to orient itself toward
a source of stimulus in such a way that symmetrical points on the surface of the
body are stimulated equally. In this way the animals are led without will of
their own either toward the source of stimulus or away from it.

Holt & Lee (1901) again bring out our second point, as applied
to reactions to light, with especial clearness :

The phenomena that have led to such an assumption can be satisfactorily
explained on the simpler theory that every ray of light impinging on an organism
stimulates at the point on ivhicJi it falls * and in proportion to its intensity. ♦ * *
The light operates, naturally, on the part of the animal which it reaches. The
intensity of the light determines the sense of the response, whether contractile
or expansive, and the place of the response, the part of the body stimulated,
determines the ultimate orientation of the animal." (Holt & Lee, 1901, pp.
479-480.)

The theory of tropisms as above set forth depends upon the reflex,
contractility of the motor organs when affected by certain stimuli. An
attempt has been made to give it a still simpler form in a recent paper
by Ostwald (1903). Ostwald would omit even the factor of reflex
irritability, holding that the turning which brings about orientation is
a mechanical result of differences in the internal friction of the water or

Original not italicized.

THE THEORY OF TROPISMS. 93

similar physical differences. The organism is considered to continue
to move its motor organs in exactly the same way after the external
change (usually called a stimulus) has taken place ; the reason for turn-
ing lies only in the different mechanical effect produced when the motor
organs act on a medium of greater or less internal friction than before.

It is difficult to conceive how anyone having any acquaintance with
the movements of organisms could propose such a theory as that of
Ostwald, and indeed this author states (p. 24) that his account is purely
theoretical, and that he has not attempted to test his theory by experi-
ment. We need not, therefore, dwell upon the theory, further than to
point out the fact that the reactions of many of these lower organisms
have been studied thoroughly, and the reflex movements which they
perform when subjected to directive stimuli have been fully described,
and that these movements are entirely incompatible with such a theory
as that which Ostwald sets forth.* If details are desired, it may be
pointed out that all the observations brought in the following that are
inconsistent with the theory of tropisms as dependent upon direct
stimulation of the motor organs are a fortiori inconsistent with such
a theory as that of Ostwald.

We may, then, turn to the theory of tropisms as set forth in the above
quotations from Verworn, Loeb, and Holt & Lee. Diagrams illus-
trating the method of action of a stimulus, on this theory, are given in
the first of these contributions (Figs, i and 2).

How far does this theory go in explaining the behavior of the lower
organisms.'' "Tropisms" has become the keyword everywhere in
animal behavior ; it is supposed to furnish a ready explanation of most
of the puzzles which we here encounter. How far is this justified.''

This question can be answered only by accurate observation of just
what organisms do under the influence of stimuli. The theory of

♦Some of the assumptions which Ostwald makes as a basis for his physical
analysis of the swimming of the lower organisms are so extraordinary as to
deserve mention as curiosities. He states, for example, that as a rule the lower
swimming organisms which exhibit the tropisms show active movement vertically
only upward; he thinks it probable that cases where they have been described
as swimming actively downward are errors ; that such downward movement is
really only passive falling. Yet everyone who has worked with Paramecium or
other Ciliata must know how far from the facts is this idea. In a vertical tube
Paramecia hasten as freely, and almost as frequently, downward as upward.
These infusoria by no means collect at the top in a vertical tube so regularly as
the literature on geotropism might lead one to suppose; Paramecia of this region
at least are almost as likely to collect at the bottom as at the top. And there is
little more difficulty in Paramecium in distinguishing an active movement down-
ward from a passive one than there is in man. From my own observations I
know that parallel statements could be made for many other free-swimming
organisms, including Metazoa (Rotifera and Crustacea), as well as Protozoa.

94 THE BEHAVIOR OF LOWER ORGANISMS.

tropisms says that certain definite things happen in the change of
position undergone by organisms under the influence of stimuli ; that
the organisms perform certain acts in certain ways. The problem for
the investigator is, then, Do these things happen? Does the organism
perform these acts, in these particular ways? These questions are not
metaphysical ; they can be answered by observation.

We have now before us a considerable body of exact observations
which permit us to answer these questions for a certain number of
organisms. We will here attempt to summarize these observations so
far as they bear upon the essential points in the theory of tropisms. In
particular, we will ask, (i) Is the observed behavior brought about
through orientation, in the way the theory of tropisms demands? (2)
Does the evidence show that the action of a stimulus is directly upon the
motor organs of that part of the body on which the stimulus impinges?

REACTIONS TO MECHANICAL STIMULI.

The reactions to simple mechanical stimuli, as when the organism
is touched or struck by a hard object over a certain definite area of the
body, of course do not as a rule present the conditions required for
the production of a tropism, including a definite orientation. Yet it is
important to bring out certain general relations shown in these reactions,
as they throw light on the reactions to stimuli of a different character.
f Most animals show in one way or another a tendency to avoid sources
of mechanical shock. In the higher organisms the reaction usually
takes the form of a turning away from the side stimulated. The point

\ which needs to be brought out here is that in ciliate infusoria the direc-
tion of turning depends, not upon the part of the body stimulated, but
upon an internal factor. Stylonychia turns to the right, whether stimu-

• lated on the right side, on the left side, on the dorsal surface, on the
anterior end, or by a general unlocalized mechanical shock ; and parallel
statements can be made for other infusoria. (For details see Jennings,
1900.) We have proof, therefore, that the action of the stimulus is
on the organism as a whole^ not merely upon the motor organs of

^ that region of the body stimulated. Further, it is clear that the
response is a reaction of the organism as a whole, not one brought
about as an indirect result of the fact that certain motor organs have
received a stimulus to contraction.* In these respects, therefore, the
reactions to mechanical stimuli are different in character from those
assumed to take place in the tropisms, and even in these unicellular
organisms the processes taking place must be more complex than the

♦This fact becomes still more striking when we recall that the reaction takes
place in the same way in pieces from any part of the body, from which any given
motor organs may have been removed. (Details in Jennings & Jamieson, 1902.)

THE THEORY OF TROPISMS. 95

theory of tropisms assumes. Certainly a reaction of the organism as a
unit, in response to a localized stimulus, is a phenomenon of a higher
and more complex order than would be a simple contraction or other
direct change in the motor organs at the point stimulated.

In the higher Metazoa the reaction to a slight mechanical stimulus
at one side is usually a turning either toward or away from the source
of stimulus. So long as we do not analyze the process further, this
result might be interpreted either as due to the direct response, by con-
traction, of the muscles primarily affected (thus in accordance with the
tropism theory), or as a response of the organism as a whole, depend-
ent, perhaps, on an alteration in its physiological condition brought
about by the stimulus. The former interpretation is doubtless much
the simpler. But we find in the unicellular organisms that this first
interpretation is impossible, and that we are forced to the less simple
and definite conclusion that the organism reacts as a whole. Does it ^
not then become probable that in the higher animals the very simple,
almost mechanical, explanation is likewise incorrect ; that we have in
them a phenomenon at least as complex as that found in the unicellular
animals.^ In other words, should we conclude that the reactions in
the higher Metazoa are simpler and less unified than in the Protozoa.?/

Fortunately, however, we are not forced to base our conclusions on
general considerations. These reactions have been minutely studied in
very few of the bilateral Metazoa, but Pearl (1903) has given us a
thorough analysis of the reactions of a flatworm (Planaria). This
cannot be taken up in detail here, but we may quote Pearl's con-
clusion in regard to the positive reaction. This consists in a turning
toward the point stimulated, on a superficial view a very simple reac-
tion, one especially well fitted for explanation on the theory of direct
action of the agent on the motor organs of the region stimulated. Pearl
concludes, after exhaustive study, that the processes in the reaction are
as follows :

A light stimulus, when the organism is in a certain definite tonic condition,
sets off a reaction involving (i) an equal bilateral contraction of the circular
musculature, producing the extension of the body; (2) a contraction of the
longitudinal musculature of the side stimulated, producing the turning toward
the stimulus (this is the definitive part of the reaction); and (3) contraction of
the dorsal longitudinal musculature, producing the raising of the anterior end.
In this reaction the sides do not act independently, but there is a delicately
balanced and finely coordinated reaction of the organism as a whole, depending
for its existence on an entirely normal physiological condition. (/. c, p. 619.)

Further studies carried on under the direction of the writer, and
soon to be published, will show that in certain other bilateral Metazoa
it is equally impossible to explain the simple turning toward a stimulus
as a direct reaction of the motor organs of the part stimulated.

96 THE BEHAVIOR OF LOWER ORGANISMS.

It will be important to keep in mind the nature of the reactions to
mechanical stimuli, especially in the infusoria, in considering the
reactions which are more usually classed among the tropisms.
REACTIONS TO CHEMICALS.*

The reactions to chemicals have been studied by the present author
and those associated with him in many Ciliata, in certain Flagellata,
in the Bacteria, the Rotifera, and the flatworm ; further studies, not
yet published, have been made on other organisms. Now, in regard
to our first question, as to orientation, the following must be said :
In no case has the typical reaction been found to take the form of an
orientation, such as is demanded by the theory of tropisms. In the
ciliates, flagellates, and rotifers the reaction has been found to take the
form of a " motor reflex," a backing followed by a turning toward a
certain structurally defined side, without regard to the direction from
which the chemical is diffusing. It is this motor reflex that causes the
organisms to collect in the region of certain chemicals, and to avoid
others. (Details in Jennings, "Studies," Nos. I-X.)

In the Bacteria the results of our work (Jennings & Crosby, 1901)
are in agreement with those of Rothert (1901). Here, again, in the
gatherings in certain chemical solutions, or in the avoidance of others,
there is nothing resembling an orientation in the lines of diffusion.
The phenomena are brought about through a reaction of the same
essential character as the motor reflex of the infusoria, but still simpler.
The essential point is that the Bacteria, when stimulated chemically,
reverse the direction of movement. (Details in the papers just cited.)

In the flatworm the results of the thorough study of the chemical
reaction by Pearl (1903) maybe given in that author's own words:

Planaria does not orient itself to a diffusing chemical in such a wa_y that the
longitudinal axis of the body is parallel to the lines of diffusing ions. Its reac-
tions to chemicals are motor reflexes identical with those to mechanical stimuli.
The positive reaction is an orienting reaction in the sense that it directs the
anterior end of the body toward the source of stimulus with considerable pre-
cision, but it does not bring about an orientation of the sort defined above. (Pearl
loc. cit.^ p. 701.)

For details, the original paper of the author quoted must be consulted.
It may be added that this positive reaction, by which the anterior end
is directed toward the source of stimulus, is identical with that which
takes place in response to a single mechanical stimulus. This is analyzed
above (p. 95).

Are there any precise and detailed observations which support the
idea that the reaction to chemicals is ever a typical tropism.'* Before

♦For a statement of the theory of tropisms as applied to chemicals, see Loeb
(1897, p. 442) and Garrey (1900, pp. 292, 293).

THE THEORY OF TROPISMS. 97

the method of reaction by a " motor reflex" had been described the
reactions to chemicals had been referred in a general way to the tropism
schema, but critical observations, which would differentiate between
the possibilities, have been lacking. It is necessary to use the greatest
caution in this matter, as is shown by the case of Chilomonas. Garrey
(1900), although he stated that '' a study of the mechanics by which
the organism is oriented or by which it is prevented from moving
from the ring into the stronger acid of the clear area, or the weaker
acid surrounding the ring, proved fruitless," nevertheless concluded
that the reaction in this animal was a case of typical tropism. In a
paper published in the same number of the same journal (Jennings,
1900, a), I showed that when the mechanism of the reaction is worked
out, this conclusion does not hold, but that the reaction takes place
through a motor reflex, similar to that in the Ciliata. In cases, there-
fore, where the mechanism of the reaction (that is, the exact movements
which the organism performs) has not been worked out, conclusions as
to the nature of the reaction are of little value. The only case of which
I know in which an author acquainted with the method of response by
a " motor reflex " maintains, on the basis of observation, a reaction of
unicellular organisms to chemicals in accordance with the theory
of chemotropism, is the case of Saprolegnia swarm-spores, as described
by Rothert (1901). In this case we are dealing with very minute
organisms, and Rothert has made no attempt to give an analysis of the
relation of the direction of turning to the differentiations in the body
of the organism, such as we found to be necessary above for Chilomonas
before the real nature of the reaction could be determined.

Thus it is clear that cases of true chemotropism, in accordance with
the general tropism schema, are exceedingly rare, if they exist at all.
In almost all the lower organisms in which this matter has been care-
fully studied it has been demonstrated that the reaction to chemicals is
of a different type from that demanded by the tropism theory.

In the discussion so far we have devoted attention particularly to the
question of orientation. When we examine the second question pro-
posed, as to whether the stimulus acts directly upon the motor organs
of that part of the body on which it impinges, we find the answer
somewhat less clear than it was in the case of mechanical stimuli. It
is true that in the Infusoria and Rotifera the direction of turning is, as
in the case of mechanical stimuli, always toward a structurally defined
side, without regard to the direction from which the chemical is diffus-
ing, so that at first view it seems beyond question that the reaction is
no^ due to the direct action of the stimulus on the motor organs of the
region on which it impinges. While this conclusion is highly probable,
the observed facts do not demonstrate it for chemical stimuli as they

98 THE BEHAVIOR OF LOWER ORGANISMS.

do for mechanical stimuli. This^ is owing to the fact that the organism
determines for itself the region in which it shall be stimulated by a
chemical in solution, as well as the side toward which it shall turn.
Now, it appears that the side on which, by its own activities, it is, as a
rule, first stimulated by a chemical, is (usually, at least) opposite that
toward which it turns. (For details, see Jennings, 1902, a.) It could be
contended, therefore, that the direction of turning, in the case of chemi-
cal stimuli, is a result of the direct action of the stimulating agent on
the side stimulated. Such a contention would have little general
significance, however, in view of the fact that the same reaction occurs
as a response to various other stimuli, where this explanation is quite
impossible.

In certain higher organisms, researches which were made under the
direction of the writer and are soon to appear will show (i) that chemi-
cal stimuli may produce local contractions in the part of the body with
which the chemical comes in contact ; (2) that these local contractions
have little to do with the characteristic behavior of the animals when
subjected to chemicals.

REACTIONS TO HEAT AND COLD.

Reactions to heat and cold have been fully discussed in the first of
these contributions. It is only necessary, therefore, to point out that
the results are in almost every detail parallel with those for reactions
to chemicals, and in the same way and to the same degree inconsistent
with the theory of tropisms. In organisms higher than the Infusoria
and Rotifera, the reactions to heat and cold have been very little studied
from the present point of view.

REACTIONS TO CHANGES IN OSMOTIC PRESSURE.*

In the ciliate infusoria the reactions to differences in osmotic pressure
are identical with those to chemicals, save that the organisms are much
less sensitive to osmotic changes. (Details in Jennings, 1897 and 1899.)
The bearing of these reactions on the theory of tropisms is, therefore,
the same as was brought out above in the discussion of the reactions to
chemicals.

REACTIONS TO LIGHT.

The phenomena shown in the reactions of organisms to light have per-
haps formed the chief basis for the theory of tropisms. There is usually
a definite orientation shown by the organisms ; they move with the axis
of the body parallel with the light rays either to or from the source of
light. The existence of such orientation forms the basis of the theory
of tropisms, and has been considered sufficient in itself as a proof of the

* "TonoUxis," Massart; " Osmotaxis," Rothert.

THE THEORY OF TROPISMS. 99

theory. Yet the theory makes certain definite statements as to the cause
of the orientation and the way in which it is brought about. These
statements are open to observation and experiment. In most bilateral
animals it is indeed difficult to really test the theory. This is because
these animals may turn directly toward either side under the influence
of light, and it is difficult to tell whether this turning is due to the direct
action of the light on the motor organs or to a reaction of the organisms
as a whole induced by some change in physiological condition brought
about by the light. But in the ciliate infusoria we find a set of organisms
so constituted as to permit us to bring the theory to a direct test. These
organisms are unsymmetrical, and, as we have seen, the usual reaction
is by a motor reflex involving a turning toward a structurally defined
side. We can, therefore, arrange our experiments in such a way that
the turning demanded by the theory of tropisms shall be the opposite
of that usually produced in the reaction of the organism as a whole, and
observe the results.

This is what was done with Stentor cceruleus^ as described in the
second of these contributions. The result, as we have seen, is that the
organism turns toward a structurally defined side, without regard to
what is demanded by the theory of tropisms. The same result was
obtained with a number of flagellates and with a bilateral Metazoan —
the rotifer Anurcea cochlearis.

Thus, in these cases, it is impossible to interpret the reactions as due
to the direct action of the light on the motor organs of the side on
which the light impinges. The response is as clearly a reaction of the
organism as a whole as is the reaction to mechanical stimuli.

Now that it has been shown that orientation to light does occur in
some cases in a manner quite at variance with the postulates of the
theory of tropisms, and this in organisms widely separated in structure
and classification, it can no longer be held that orientation is, per se,
a proof of the tropism theory. In other words, cases in which orien-
tation takes place, but in which the manner in which it is brought
about has not been observed, can not be assumed as cases of typical
tropism, due to the direct action of the light on the motor organs
of the side affected. The reactions of flagellates and swarm-spores
to light, as described by Strasburger (1878), have long been con-
sidered types for the tropisms. In the second of these contributions
I have shown that in Euglena and Cryptomonas (the latter being one
of the organisms studied by Strasburger) the reactions do not take
place in accordance with the tropism schema. So far as can be judged
from Strasburger's account the reactions of the swarm-spores take place
in essentially the same manner as in the flagellates. As Rothert (1901)
has pointed out, there are many details in Strasburger's account which

lOO THE BEHAVIOR OF LOWER ORGANISMS.

suggest that the explanation on the tropism schema is incorrect. These
details become intelligible as soon as we understand the real method
of reaction as set forth in the second of these contributions. The
assumption that the reaction is a typical tropism, when only the fact of
orientation is known, is as likely to fall to the ground in other cases
as in those just mentioned.

The reaction of bacteria to light, as shown by Bacterium photo-
metricum,, described by Engelmann (1882), is a typical example of a
reaction through a motor reflex not fitting the tropism schema at all.

To sum up, it is clearly shown in certain cases that the reaction to
light takes place in a way that is not consistent with the theory of
tropisms, and this is true in some cases where a pronounced orienta-
tion exists. In many cases of orientation, where it is supposed that
the theory of tropisms holds, this is an assumption, for the observations
which would decide the matter are lacking.

REACTIONS TO GRAVITY.

In no c^3e have the exact movements of unicellular organisms in
response to gravity been worked out in the manner in which this has been
done for the reactions to mechanical stimuli, chemicals, heat, light, and
electricity. We are, therefore, without the requisite data for deciding
whether these reactions agree with the theory of tropisms or do not.*

In the higher organisms in which the positive and negative reactions
to gravity have been observed (starfish, holothurians, flies, insect larvae,
etc.), the conditions are so complex that, so far as I am able to see,
observations which are crucial for deciding as to the mechanism of the
reactions have not been made and perhaps can not be made.

REACTIONS TO ELECTRICITY.

As we have seen in the third section of this paper, the reactions of
the rotifer to the continuous electric current do not take place in
accordance with the theory of tropisms. Anuraea shows a striking
orientation to the electric current, swimming directly to the cathode.
Yet this orientation is brought about in a way that is quite inconsistent
with the tropism schema. The reaction takes place through a " motor
reflex," the direction of turning is determined by an internal factor,
and not by the way in which the current strikes the organism. The
reaction can only be interpreted, therefore, as a reaction of the organism
as a whole.

♦In a forthcoming paper by the author, based on work done since the above
was written, it will be shown that the reactions of Paramecium to gravity take
place in the same way as the reactions to most other stimuli, so that they do not
agree with the theory of tropisms.

THE THEORY OF TROPISMS. lOI

In the reactions of the ciliate infusoria to the constant electric current,
however, we have, if nowhere else, phenomena which show to a
certain extent clear-cut and undoubted agreement with the theory of
tropisms. To this agreement with the theory of tropisms much of the
widespread adherence to the tropism theory for reactions in general is
doubtless due. The reaction of infusoria to the electric current is con-
sidered a type for the other reactions of organisms.

Yet, in deciding to what extent the theory of tropisms helps us to
understand the behavior of organisms, certain striking facts in regard
to the reaction to the electric current need to be taken into considera-
tion. These are the following :

(i) The reaction to the electric current never takes place in nature.
As has been repeatedly pointed out, the electric reaction is a product
of the laboratory ; it is a reaction which the organism never gives
under normal conditions. This being true, it should not be made the
type for reactions in general unless it can be shown clearly that the
characteristic features in the effects of electricity on organisms are
present also in the effects of other agents. Otherwise we may fall into
the same error that would exist if we considered the contortions of a
person who had grasped the electrodes of a powerful battery as a type
of human behavior in general.

(2) But examination shows that the characteristic features of the
effect of electricity on organisms are not present in the case of other
stimuli. The electric current, as the experiments of Kiihne (1864)
and Roux (1891) have shown, polarizes the cell. That is, it divides
it into halves, differing in chemical reaction. One half, in the case
described by Kiihne, had apparently an acid reaction, the other
half an alkaline reaction. In its effects on free-swimming organisms
a similar polarity is shown. In Paramecium, for example, the cilia
on one half of the body (where the current is entering) are caused to
take a certain position, while those on the other half (where the cur-
rent is leaving) take the opposite position. No other agent known
produces these polar effiiCis^ either chemically or in the effect on the
motor organs. Yet it is to exactly these effects that the orientation
which makes this reaction the type for the tropism theory is due.

If other agents produce these effects why are they not known and
described } There is no great difficulty in observing these effects with
the use of the electric current.^ Just as exact studies have been made
of the reactions to other stimuli ; yet, so far as I am aware, no one
has ever described any other stimulus as giving these characteristic
polar effects. On the contrary, the reactions to other stimuli are well
known not to show these characteristic phenomena.

Since, therefore, the characteristic phenomena of the reaction to the

I02 THE BEHAVIOR OF LOWER ORGANISMS.

electric current are not found in the reactions to other stimuli, it seems
a perversion to make the electric reaction a type for all others. The
reaction of the infusoria to the electric current takes in its characteristic
features a unique position among the reactions of the organism, requiring
special explanation.

(3) In the response of the infusoria to the electric current there
appears also the same type of reaction that occurs as a response to other
stimuli, but obscured by the phenomena peculiar to the effects of the
current.

This fact, that the reaction to the electric current is of a dual char-
acter, that the peculiar effects of the current are, as it were, superposed
upon the usual method of reaction, is not usually so clearly recognized
as it deserves to be.

If the constant current is passed through a preparation containing
large numbers of some species of the Hypotricha, as Stylon3xhia or
Oxytricha, it will be found that the animals, practically without excep-
tion, attain their orientation by turning toward the right side, thus
reacting as they would to any other stimulus. Further, if after they
are oriented the direction of the current is reversed, the animals all,
without exception, attain their new orientation (with anterior ends in
the opposite direction) by whirling toward their right sides. Thus, so
far, the reaction to the electric current is identical with that to other
stimuli, and the direction of turning is determined by an internal factor,
not by the way in which the current strikes the organism. In these
respects the Hypotricha agree with the Rotifera.

But exact observation shows that in the Hypotricha there is another
factor involved in the reaction. The characteristic polarizing effect of
the current appears in its action on the motor organs that are distributed
over the body surface ; those on one half of the body strike in one
direction, those on the other half in the opposite direction. Part of
these motor organs, therefore, assist in turning the organism in its usual
way (to the right) ; part oppose this turning. The result is that in
certain positions the turning to the right is opposed by the stroke of a
large number of cilia, so that the turning takes place more slowly than
usual. Nevertheless, in the Hypotricha, the determining factor in the
reaction to the electric current is almost throughout the same as in
the reaction to other stimuli ; the direction of turning is determined by
internal factors, as a reaction of the whole organism, not by the direction
in which the current strikes or passes through the organism. (Details
in the work of Pearl, 1900.)

If in place of one of the Hypotricha we experiment with an infusorian
in which the cilia cover closely the whole surface of the body, as Para-
mecium, the peculiar polarizing effect of the current on the cilia of the

THE THEORY OF TROPISMS. IO3

two halves of the body becomes much more powerful, because the
number of cilia aftected in this way is much greater. The result is that
it almost alone determines the nature of the reaction. The direction of
turning is, therefore, determined by the way in which the current strikes
the body, as required by the theory of tropisms. But it should be
recognized that this is by no means universal among the infusoria ; in
doubtless fully as many cases the direction of turning is determined,
even under the electric stimulus, by an internal factor.

This peculiar dual character of the reaction to the electric current —
one strong factor being due to the inherent tendency of the organism
to turn in a certain definite way, without regard to the way in which
the stimulating agent impinges upon it — has been studied in detail in
recent contributions by Pearl (1900), Putter (1900), and Wallengren
(1902 and 1903). We may perhaps compare it, without indicating any
similarity in details, to the behavior of a person who has taken hold of
the electrodes from a powerful induction coil. He attempts in various
ways to free himself from the electrodes. This may be compared with
the attempt of the infusorian to perform its usual reaction to strong
stimuli. He also undergoes involuntary contortions, due to the action
of the electricity on his muscles ; these may be compared with the pecu-
liar effect of the electric current on the cilia of the infusoria, causing
them to strike in opposite directions on the two halves of the body.

Putting all together, we are not justified in taking the reaction of the
infusoria to the electric current as a general type for the reactions of
the lower organisms, because in its characteristic features it differs
from all the other known reactions. Yet it is exactly these unique
features that bring it into (partial) agreement with the tropism schema.

RESUME AND DISCUSSION.

We have thus passed in review the reactions of a large number of
lower organisms to the commoner stimuli, so far as they are known
from exact observations. We have found that as a rule they do not fit
into the tropism schema. In the reactions to mechanical stimuli, to
heat and cold, to chemicals, to changes in osmotic pressure, orientation
is not a primary or striking factor of the reaction ; when a common
orientation of a large number of organisms occurs, it is a secondary
result, due to the fact that the organisms are prevented from swimming
in any other direction. In the reaction to light orientation is a strik-
ing feature ; but the orientation, in certain precisely investigated cases
at least, is brought about in a manner which is inconsistent with the
tropism schema. In the reaction to gravity the precise reaction method
has not been worked out. Only in the reaction to the constant electric
current do we have in some organisms a partial agreement in principle

104 THE BEHAVIOR OF LOWER ORGANISMS.

with the requirements of the tropism theory, and this agreement is due
to an effect on the organism in the production of which the electric
stimulus is unique, so far as known. In none of the reactions which
have been thoroughly worked out, except partially in the reaction to
the electric current, are the phenomena to be explained on the view
that the result is due to the direct action of the stimulating agent on the
motor organs of the part of the body on which it impinges. In the
reactions to mechanical stimuli and to light, and in the reactions to
the electric current in some animals, this view is absolutely disproved.
The direction in which the organism turns is, in all the well known
reactions of unicellular organisms and rotifers (except in a portion of
the reactions to the electric current) , determined by an internal factor,
and predictable from the structure of the organism without any know-
ledge of the direction from which the stimulating agent is to come.

We should perhaps consider here a modification of the original form
of the tropism theory that has been proposed by some authors. This
is in regard to the assumption that the stimulating agent acts directly
on the motor organs upon which it impinges. For this it is sometimes
proposed to substitute the view that the action of the stimulating agent
is directly on the sense organs of the side on which the stimulus im-
pinges, and only indirectly on the motor organs through their nervous
connection with the sense organs. When thus modified the theory, of
course, loses its simplicity and its direct explaining power, which made
it so attractive. In order to retain any of its value for explaining the
movements of organisms, it would have to hold at least that the connec-
tions between the sense organs and motor organs are of a perfectly
definite character, so that when a certain sense organ is stimulated a
certain motor organ moves in a certain way. When we find, as we
do in the flatworm (see the following paper), that to the same stimulus
on the same part of the body, under the same external conditions, the
animal sometimes reacts in one way, sometimes in another, the tropism
theory, of course, fails to supply a determining factor for the behavior.

But can we explain the reaction methods of the infusoria and other
animals which, as set forth above, are inconsistent with the tropism
theory in its original form, on the basis of the modification of this
theory, set forth in the last paragraph.? While in the infusoria the
assumption of nervous connections, etc., is inadmissible, we may
waive that objection and answer the question proposed from an analysis
of the observed phenomena. In Stentor or in Stylonychia, for example,
we find that the usual reaction to all classes of stimuli is by backing,
then turning toward the aboral side ; in some of the rotifers by turning
toward the aboral (dorsal) side. To simplify matters, we may take
into consideration only the turning toward the aboral side. This turn-

THE THEORY OF TROPISMS. IO5

ing is due to a certain method of movement of certain motor organs.
In the rotifers it is the coronal cilia which accomplish the turning,
while in the infusoria we know that the adoral cilia are concerned in
the movement. We may take the coronal or adoral cilia, then, as rep-
resentative of the organs active in the turning. For convenience we
may designate these active organs simply as x.

Now, when the animal is stimulated on the right side, we find that"*"
the motor organs x move in a definite way. On the tropism theory we j
would conclude, therefore, that the portion of the right side stimulated
has nervous connection with the organs x. But we find also that when
stimulated on the left side, the oral side, or the aboral side, the organs ^
X move in exactly the same manner. In other words, we find that it
does not depend on the side stimulated what organs respond, as re-
quired by the tropism theory. This theory, then, in its modified form,
is of no more service for these cases than in its original form. The
responses in the animals which we have considered must, therefore,
be conceived as reactions of the organism as a whole, and due to some
physiological change produced by the stimulus, not as the result of
direct changes in certain motor organs when they or the parts with
which they are most closely connected are locally aflfected by a stimu-
lating agent. The facts show that the parts act in the service of the
whole, not that the action of the whole is due to the more or less inde-
pendent irritability and activity of the parts.

Thus the facts brought out show that the theory of tropisms is not
of great service in helping us to understand the behavior of these lower
organisms. On the contrary, the reactions of these organisms seem
as a rule thoroughly inconsistent in principle with the fundamental
assumptions of the theory.

The facts brought out above are based on a study of what is, of course,
a comparatively small number of organisms. They rest chiefly on an
extensive study of the ciliate infusoria, with less thorough examination
of bacteria, flagellata, rotifers, and a few higher organisms. Doubtless
in organisms which are made up of many parts which are less firmly
bound together into a unified body than in those considered, we may
find greater independence of action in the parts. This seems to be the
case, for example, in the sea urchin, with its numerous independently
acting spines, pedicellariae, tube feet, etc. In this animal Von Uexktill
(1900, 1900, a) concludes from his extensive study of the reactions that
many features in the behavior which seem at first view to be activities
of the animal as an individual are really due to the independent reac-
tions of the parts, so that he can say that while in walking, in the case
of the dog, " the animal moves its legs ; in the sea urchin the legs move
the animal." This method of behavior has a general agreement with

I06 THE BEHAVIOR OF LOWER ORGANISMS.

what is demanded by the tropism schema, though when we come to
details of the behavior of the organs themselves, this theory seems
unsatisfactory, even in the sea urchin.

Such organisms as the sea urchin, composed anatomically and physio-
logically of many parts, each acting almost as an independent animal,
are certainly less common than more unified animals, such as we find
in the Infusoria, the Rotifera, the flatworms, etc. For this reason,
therefore, it has seemed worth while to sum up the real relations of the
behavior of these organisms to the tropism theory. The unicellular
animals are precisely those on which the prevailing theories of tropisms
or taxis have by many writers* been chiefly based. With the demon-
stration that the behavior of these organisms (as well as of many higher
ones), is for the greater part inconsistent with the tropism theory, per-
haps a large portion of the foundation for its acceptance as a general
formula for the chief features in the behavior of lower animals is cut
from beneath it.

In the following paper, on the part played in behavior by physio-
logical conditions of the organism, we shall find other, and, as it seems
to me, still more cogent, reasons for holding the tropism theory inade-
quate to account for the determining factors in the behavior of most
lower organisms.

SUMMARY.

The foregoing paper consists of a review of the behavior of Ciliata,
Flagellata, Bacteria ; of Rotatoria and certain other Metazoa, so far
as known from exact observation of their actions when stimulated,
with a view to determining how far the prevailing theory of tropisms
aids us in understanding the behavior of lower organisms.

The following are considered the essential points in the prevailing
theory of tropisms : (i) That orientation is the primary factor in deter-
mining the movements of organisms into or out of certain regions, or
their collection in or avoidance of certain regions ; (2) that the action
of the stimulus is directly upon the motor organs of that part of the
organism upon which the stimulus impinges, thus giving rise to changes
in the state of contraction, which produce orientation.

The reactions of the organisms above named are then reviewed to
determine in how far there is agreement with these essential points in
the theory of tropisms. The following are pointed out:

The reactions to mechanical stimuli, to chemicals, to heat and cold,
and to variations in osmotic pressure have been described in detail, and
it is found that orientation is not a primary nor a striking factor in
them. The response in all these, cases is produced through a '' motor

This, however, is not true of Loeb.

THE THEORY OF TROPISMS. I07

reaction" consisting usually of a movement backward, followed by a
turning toward a structurally defined side. The direction of turning is
thus determined by internal factors.

In the reaction to light orientation is a striking factor, but the orienta-
tion is not primary, being due to the production of the same " motor
reaction" described in the last paragraph. The method of orientation
is incompatible with the idea that orientation is due to the direct action
of the stimulus upon the motor organs of the part of the body on which
the light impinges, for orientation occurs by turning always toward a
certain structurally defined side, without regard to the part of the body
struck by the light. The turning may, therefore, be toward or away
from the source of light, or in any intermediate direction. In any case
it is continued or repeated until the anterior end is directed away from
the source of light, when it ceases.

The exact method of reaction to gravity has not been worked out by
direct observation.

In the reaction to the electric current the reaction method of the
rotifer is by a " motor reflex," and is hence inconsistent with the tro-
pism schema. In the Infusoria there is a partial (but only partial)
agreement with the requirements of the tropism theory. But this
partial agreement with the theory is due to certain peculiar effects of
the electric current which are not known to be produced by any other
stimulus. Hence the reaction to the electric current, far from being a
type for reactions in general, is a unique phenomenon, demanding
special explanation.

The general conclusion is drawn that the theory of tropisms does
not go far in helping us to understand the behavior of the lower organ-
isms ; on the contrary their reactions, when accurately studied, are, as
a rule, inconsistent with its fundamental assumptions. The responses
to stimuli are usually reactions of the organisms as wholes, brought
about by some physiological change produced by the stimulus ; they
can not, on account of the way in which they take place, be interpreted
as due to the direct effect of stimuli on the motor organs acting more
or less independently. The organism reacts as a unit, not as the sum
of a number of independently reacting organs.

FIFTH PAPER.

PHYSIOLOGICAL STATES AS DETERMINING

FACTORS IN THE BEHAVIOR OF

LOWER ORGANISMS.

109

CONTENTS.

PAGB.

Nature and Evidences of Physiological States, . . . . . . Jii

Physiological States in the Protozoa (Stentor as a Type), . . .112

Physiological States in the Lower Metazoa (the Flatworm as a Type), . 115
Changes in the Sense of "Tropisms" and other Reactions, . . • "7
Changes in the Sense of Reactions with Changes in the Intensity of the

Stimulus, . . . . . . . . . . . .118

Interference of Stimuli, . . . . . . . . . .119

Spontaneous Movements, . . . . . . . . .120

Methods of Causing Changes in Physiological States, .... 120

Nature of Reactions to Stimuli, ........ i2i

Physiological States in the Behavior of Higher Animals as compared with

those in Lower Organisms, ........ 124

Summary, . . . . . . . . . . . .126

no

PHYSIOLOGICAL STATES AS DETERMINING
FACTORS IN THE BEHAVIOR OF LOWER
ORGANISMS.

NATURE AND EVIDENCES OF PHYSIOLOGICAL STATES.

In studying the behavior of the lower organisms the units of obser-
vation, the factors to which especial attention has been paid have been
usually the tropisnis and reflexes. These factors may be considered
as determined mainly (i) by the action of external agents on the
organism ; (2) by the structure of the organism.

An examination of the results of the study of reactions in the lower
animals up to the present time shows, I believe, that we must recog-
nize another set of factors in their behavior, of equal importance with
either of those already named. This set of factors may be characterized
by the general term physiological states.

By physiological states we mean the varying internal physiological
conditions of the organism, as distinguished from permanent anatomical
conditions. Such different internal physiological conditions are not
directly perceptible to the observer, but can be inferred from their
results in the behavior of the animal. These results are of so marked
a character that the inference to different physiological conditions is
beyond question.

In the study of tropisms and reflexes a considerable number of
instances have been brought to light of changes in the reaction methods,
such as can be attributed only to changed physiological conditions.
Some of these cases will later be considered in detail in this paper.
Comparatively few investigations on the behavior of lower organisms
have been published in which attention has been consciously directed
to these physiological states, and in most of these the matter has been
taken up rather incidentally. We may mention as instances of papers
dealing more or less with this aspect of the matter that of Hodge and
Aikins (1895) on Vorticella, those of Von Uexkiill (1899, 1900, 1900, a,
1903) on the sea urchin and on Sipunculus, my own on the behavior
of fixed Infusoria (Jennings, 1902), and that of Pearl (1903) on the
flatworm. In the study of higher organisms attention has of necessity
been largely directed to the phenomena determined by varying physio-
logical states, as these play a striking part in the behavior.

In the present paper an attempt will be made to collect and analyze
a number of the known cases showing the influence of physiological
states on the behavior of the lower animals, pointing out some of their
bearings on the theories of behavior.

112 THE BEHAVIOR OF LOWER ORGANISMS.

PHYSIOLOGICAL STATES IN THE PROTOZOA (STENTOR
AS A TYPE).

We will take up first the lowest organisms in which the matter has
been studied in detail, that is, the unicellular animals. These are of
special interest in view of their entire lack of a nervous system. As
the best-known case we may take the behavior of Stentor. This has
been described in full in a former paper by the present author (Jennings,
1902, a) ; for details this paper may be consulted.

When a quiet, extended Stentor is stimulated by lightly touching it
or the support to which it is attached with a rod, it reacts by giving a
definite reflex, that is, by contracting into its tube.

After this has taken place once or twice we find that the Stentor no
longer reacts as before. All the external conditions remain the same ;
the stimulus applied is the same. Nevertheless, the Stentor does not
react. Therefore, we must conclude that the Stentor itself has changed.
Its physiological condition is now different from what it was originally.
What the nature of the change in its condition is we do not know, save
in the fact that the Stentor in this second condition does not react as
does the Stentor in the first condition. For the sake of convenience
we may number the different physiological conditions, in order that we
may determine, if possible, how varied these conditions are. We will
call the physiological condition of the undisturbed extended Stentor,
before the stimulation. No. i, or the first condition. The condition
in which the Stentor no longer responds to the slight stimulus we will
call No. 2.

We may frequently distinguish still a third condition in the behavior
under this simple stimulus. At first the Stentor reacts by contraction
(condition i). Then it no longer reacts (condition 2). Later, or in
other cases, it may react to the stimulus, but by a different method
from the first reaction. It now bends over to one side when touched
with the rod. As set forth in my previous paper, "The impression
made on the observer is very much as if the organism were at first
trying to escape a danger, and later merely trying to avoid an annoy-
ance." As conditioning this third method of behavior, when all out-
ward conditions are the same, we must postulate a third physiological
state differing from the other two ; this we may call condition No. 3.

We may thus distinguish at least three different physiological states in
the reactions to very weak stimuli, where the initial marked response
becomes a weak one or disappears. We may now analyze in the same
way the behavior under stimuli of a different character, when there is
a series of reactions which may be considered of increasing rather than

PHYSIOLOGICAL STATES AS DETERMININCJ FACTORS. II3

of decreasing intensity. Such a case is that described in my previous
paper, when water mixed with carmine particles is allowed to reach
the disk of Stentor.

The first physiological condition is again No. i— that of the undis-
turbed extended Stentor. In this condition the organism does not
respond to the stimulus at all. After the stimulus has continued
for some time, the organism does respond by turning into a new
position. We have, therefore, a new physiological condition. The
reaction in this case is the same as that given in condition No. 3,
described above. Whether the condition now existing is the same as
in the former case we do not know ; as we have no positive evidence
to the contrary, we will number it 3 also.

Next, after several repetitions of this reaction, the organism responds
in a still different manner, by momentarily reversing the ciliary current.
Since the stimulus and other external conditions remain the same, the
organism itself must have changed. We may call its physiological
condition at the present time No. 4.

Next, the animal contracts strongly and repeatedly. This is clearly
the result of a still different physiological condition which we may call
No. 5.

After thus contracting repeatedly we find that the organism remains
contracted much longer than it did at first. It is thus now in a new
physiological condition, which we may designate as No. 6.

Finally, it breaks its attachment to the bottom of the tube and swims
away through the water. Probably, therefore, we should distinguish
a seventh physiological state, corresponding to this reaction. It is
possible, however, that the breaking of the attachment is due to the
strong contractions which characterize condition No. 6, so that the
evidence for a seventh physiological condition is not unmistakable, and
it may be omitted from consideration.

We are able to distinguish clearly, therefore, in the study of these
two sets of reactions, at least six different physiological states. In each
of these states Stentor is a different organism, so far as its reactions to
stimuli are concerned. Clearly, then, the external stimuli and the
permanent anatomical configuration of the body are by no means the
deciding factors in the behavior. These factors, in the reaction series
last described, permit at least five different methods of behavior.
Which of these methods is actually realized depends not on the quality
or intensity of the stimulus, nor on the anatomical structure of the
organism, but on its physiological condition.

I do not wish to imply that I hold that the six different physiological
states above distinguished are sharply defined, separate things. On
the contrary, it is much more probable that the different physiological

114 "^"^ BEHAVIOR OF LOWER ORGANISMS.

conditions form a continuum. We can, by taking sections, as it were,
at different intervals, distinguish at least six actually different condi-
tions ; but doubtless there exists every possible gradation from one to
another, so that still other actually different conditions could be distin-
guished if we had criteria for separating them. By careful analytical
experiments the number of different physiological conditions clearly
recognizable could doubtless be increased.

In other unicellular organisms doubtless a condition of affairs may
be found similar to that set forth above for Stentor. Hodge & Aikins
(1895) showed that the reactions of Vorticella vary with its physio-
logical condition. In the same paper (Jennings, 1902, a) in which
the behavior of Stentor was described, I have shown that in various
other fixed infusoria (Carchesium, Epistylis, etc.) the behavior like-
wise depends upon physiological states of the organism. In the
free-swimming infusoria this has not been shown to be true, at least to
any such extent. There may two grounds for this. Firstly, it is proba-
ble that in the free-swimming infusoria the behavior is actually less
varied than in the fixed species. A single motor reaction usually
removes them from the action of the stimulus causing it, so that there
is no reason for a recourse to other methods of reaction, as occurs in
Stentor. Secondly, in the free-swimming infusoria it is difficult, almost
impossible, to observe continuously the reactions of a given single
individual. This difficulty could doubtless be met by proper methods
of experimentation, and if this were done it can hardly be doubted that
a dependence of the reactions on the physiological states of the organism
would be found here also. Indeed, we have indirect evidence that this
is true in the case of Paramecium, in work already published. Thus,
in one of my earlier papers (1899, a^ p. 374) I called attention to the
fact that Paramecia from different cultures often vary exceedingly in
their reaction to a given solution of a chemical. Still more pertinent
to the point under consideration is the fact, described in the first of my
studies (Jennings, 1897) , as well as in the recent paper of Putter (1900),
of the great difference in the reaction of Paramecia and other fixed
infusoria when in contact with a solid, as contrasted with their reaction
when not thus in contact. As this, however, may be interpreted as an
interference of two stimuli, a discussion of the point is reserved until
later. A study of individual specimens of some of the larger Hypo-
tricha, such as Stylonychia, from the pointof view of changes in reaction
methods with changes in physiological condition, would doubtless bring
forth interesting results.

Even in the lower unicellular organisms, the Rhizopoda, similar
dependence of the method of reaction on the physiological state of
the organism is known to exist. Thus Rhumbler (1898, p. 24:) has

PHYSIOLOGICAL STATES AS DETERMINING FACTORS. II5

observed that Amoeba verrucosa may at first begin to ingest an Alga
filament, then later, before the ingestion is complete, the filament may
be ejected. This involves a change in the physiological condition of
the Amoeba ; otherwise it would not now reject a certain object which
it before ingested.

PHYSIOLOGICAL STATES IN THE LOWER METAZOA (THE
FLAT WORM AS A TYPE).

Passing now to the Metazoa, we find in the flatworm, Planaria, as
described by Pearl (1903), a dependence of the reaction of the organism
on its physiological condition similar to that which we saw above for
Stentor. The flatworm may be considered typical of the lower bilat-
eral Metazoa, so that it will be worth while to subject some features
in its behavior to a brief analysis from our present point of view.

We may examine for a simple typical case the reactions to mechani-
cal stimuli applied to one side of the anterior part of the body. The
flatworm is touched on one side with the tip of a hair or of a fine glass
rod. The resulting response is one of two reactions — the worm turns
either toward the point stimulated (positive reaction) or away from it
(negative reaction). Typically, the positive reaction is given to a
weak stimulus, while the negative reaction results from a strong
stimulus. The words '' weak" and " strong" have, as we shall see,
only a relative meaning when used in this connection.

When, now, we ask which of these reactions shall be given as a
response to a certain stimulus, we find that this depends upon the
physiological condition of the organism. Pearl finds the reactions
determined by the following definitely marked physiological states :

1. Individuals are frequently in what may be called a resting cow-
dition. The tonus is lowered ; the animals are very inactive and do
not respond readily to stimuli. This condition is compared by Pearl
with that of sleep in higher animals. When a flatworm in this con-
dition is given a light stimulus, such as would in an active specimen
induce a positive reaction, it does not respond at all. If the strength
of the stimulus is increased, the animal finally responds with a negative
reaction, turning away from the point stimulated. We may call this
the condition of lowered tonus.

2. In the more usual active condition the flatworm gives the positive
reaction to a very light touch, a negative reaction to a stronger stimulus.
We may call this the normal condition.

3. After the animal has been repeatedly stimulated it seems to become
excited ; it moves about rapidly, and now gives always the negative
reaction to any mechanical stimulus to which it reacts at all. It behaves
much as many higher animals do under the influence of fear. We may
call this the excited condition.

Il6 THE BEHAVIOR OF LOWER ORGANISMS.

4. After the worm in the excited condition has been stimulated
repeatedly on one side, so that it turns its head steadily in the opposite
direction, after a time it suddenly changes its method of reaction. It
jerks backward,, then turns the anterior end quickly through a consider-
able arc, usually toward the side from which the stimulus is coming,
so that the head now points in an entirely new direction. Since the
stimulus and other external conditions remain the same, the organism
must have passed into a new physiological condition, or it would not
now react in a different way. We may call this for convenience the
condition of over-stimulation.

5. Sometimes individuals are found which for a brief period (two or
three hours) seem in a much more active condition than usual. They
move about rapidly, but do not conduct themselves like the excited
individuals. As they move they keep the anterior end raised and wave
it continually from side to side as if searching. Specimens in this con-
dition react to almost all mechanical stimuli, whether weak or strong,
by the positive reaction, turning toward the point stimulated. Experi-
mentation failed to show that this condition was due to hunger. We
may speak of this fifth physiological condition as the state of heightened
activity.

In addition to the effect of these five well-defined physiological states
on the method of reaction to mechanical stimuli at the anterior part of
the body. Pearl finds that other less easily definable internal conditions
affect the reactions. At times an individual will give the positive reac-
tion to a stimulus of a certain strength a few times, then cease to give
it. On account of these and other complications due to varying internal
conditions, Pearl concludes :

It is almost an absolute necessity that one should become familiar, or perhaps
better, intimate, with an organism, so that he knows it in something the same
way that he knows a person, before he can get even an approximation of the
truth regarding its behavior.

We have taken up above only the physiological conditions influ-
encing the reactions to simple mechanical stimuli in the anterior region
of the body. We find the condition of affairs even here somewhat
involved. When other more complex stimuli are taken into considera-
tion the results of the interplay of change of physiological condition
and variations in the stimuli become, of course, much more complicated.

Thus we find in the bilateral metazoan Planaria, as in the unsymmet-
rical protozoan Stentor, that we can by no means predict the behavior
of the individual from a knowledge of the anatomical structure and of
the strength of the stimulus. The anatomical structure limits the possi-
bilities of reaction to several methods, which are, however, entirely
different or opposite in their effects on the relation of the organism to

PHYSIOLOGICAL STATES AS DETERMINING I^ACTOHS. II (^

the stimulus. Just which of these reactions shall be given as a response
to any particular stimulus depends on the physiological condition of
the organism. This physiological condition depends largely, as we
shall note later, on the history of the individual. Thus no single fixed
schema, such as we have in the tropism theory, can ever possibly
explain or define the essential points in the behavior of an animal.

Stentor and Planaria may be taken as typical examples of the higher
Protozoa and of the lower Metazoa, respectively. It is true that we are
not so well informed as to changes in physiological condition in other
lower organisms as in these two cases, but this is unquestionably due
merely to the fact that investigation has not been directed especially to
this point. There are, however, many cases in the literature which
explicitly or implicitly show the importance of physiological conditions
in determining the behavior of lower organisms. A number of these
cases may be brought together here.

CHANGES IN THE SENSE OF " TROPISMS " AND OTHER
REACTIONS.

Loeb (1893) and Nagel (1894) have called attention to the fact that
certain worms and mollusks respond to a shadow by suddenly with-
drawing into their tubes, but that after the first reaction has been thus
produced the worms may no longer react. In this " after effect of the
stimulus " (Loeb) we have, of course, a case of changed physiological
condition.

Changes in the sense of "tropisms" belong here. Groom & Loeb
(1890) found that larvae of Balanus are at certain times of the day
positively phototactic ; at other times negatively phototactic. This
difference, in so far as it is independent of changed external conditions,
is, of course, due to differences in the physiological condition of the
organism. Loeb (1893) found that the larvae of the moth Porthesia
are positively phototactic when hungry ; not so after eating. Here we
have a well-defined physiological condition determining the nature of
the reaction ; hunger is one of the most important conditions in many
of the lower animals. Sosnowski (1899) and Moore (1903) show that
the geotropism of Paramecium changes from negative to positive under
various conditions. Towle (1900) and Yerkes (1900) have shown that
the sense of the phototactic reaction in Entomostraca is dependent on
the preceding treatment of the organism, mere transference with the
pipette often changing the sense of the reaction from positive to nega-
tive or vice versa. Such instances could doubtless be multiplied
indefinitely. Each case taken by itself seems perhaps of comparatively
little significance. We may look upon them, however, as indications of
an extensive dependence of behavior on physiological conditions, such

Il8 THE BEHAVIOR OF LOWER ORGANISMS.

as we found in Stentor and the flatworm. Thorough investigation of
any of these organisms from this point of view would doubtless bring
to light a variety of physiological conditions on which the reactions
depend.

CHANGES IN THE SENSE OF REACTIONS WITH CHANGES
IN THE INTENSITY OF THE STIMULUS.

Must we not bring under the same point of view the well-known
phenomenon of a change in the sense of the reaction with a change in
the intensity of the stimulus.? As a simplest case of this we may take the
reaction of Stentor to mechanical stimuli. As shown in the ninth of
my studies (Jennings, 1902), Stentor reacts to a very weak mechani-
cal stimulus on one side of the disk by bending toward the source of
stimulus; to a stronger but otherwise similar stimulus it responds by
contracting into the tube, or (later) by bending in another direction.
In the same way the flatworm reacts positively to a weak mechanical
or chemical stimulus, negatively to a stronger one. How can we ex-
plain these opposite reactions to stimuli of the same quality, differing
only in intensity."*

We have here, it seems to me, the same phenomenon shown in the
production of a change in physiological condition by a stimulus. We
know that even a single stimulus may produce a changed physiological
condition, as when after a single stimulus the organism no longer
reacts as before. We know also that the nature of the physiological
condition determines the reaction. In the present case we must con-
clude that a light stimulus throws the organism into a certain physio-
logical condition, whose concomitant reaction is turning toward the
point stimulated. A more intense stimulus induces a different
physiological condition, whose concomitant reaction is a contraction
into the tube (Stentor), or a turning in the opposite direction (flatworm).
The action of the stimulus, as we have seen in the foregoing paper
devoted to the theory of tropisms, cannot in most cases be directly on
the motor organs, so that from this point of view also we are almost
forced to the conclusion that the primary action of the stimulus is
to change the physiological condition of the organism. In any reac-
tion to stimulus we would have, therefore, the following steps : The
stimulus acting on the organism changes its physiological condition ;
this physiological condition induces a certain type of reaction. In
determining what physiological condition shall be produced, the inten-
sity of the stimulus is fully as important as its quality.

We have a similar reversal of the reaction as the intensity changes
in reactions to light. Many organisms are positive to weak light ;
negative to strong light. The cause of this reversal of the reaction as

PHYSIOLOGICAL STATES AS DETERMINING FACTORS. IIC)

the light grows stronger has given rise to much discussion (see
Hohnes, 1901 and 1903). We have here, of course, a parallel case to
the reversal of the reaction in Stentor or the flatworni under mechani-
cal stimuli of varying intensity. In the weak light we must suppose
the organism to be thrown into a certain physiological condition, the
concomitant of which is a certain type of reaction. In a more intense
light a different physiological condition is induced, corresponding to a
different reaction. The fact that different intensities of stimuli do
cause different physiological conditions and different reactions is, of
course, familiar to us, both from experimentation on animals and from
our own experience ; in the latter case we usually call the distinctive
reactions to very intense stimuli pain reactions. In the reversal of the
reaction to light as the light becomes stronger we have, it seems to me,
merely an instance of this general phenomenon, not differing in funda-
mental character from other instances.

In most of these cases we have, of course, a further problem in regard
to those features of the reaction which concern direction. Why does
the weak stimulus on the left side of Planaria cause a turning toward
that particular side ? Or, why does a weak light from a certain direc-
tion cause Volvox to swim in that particular direction ? These problems
of direction are, of course, not touched in the foregoing discussion,
which, however, loses none of its force because these problems remain.
They are simply farther problems. The tropism theory gave a simple,
direct answer to these questions ; but, as we have already shown in
the foregoing paper, this answ^er was, in many cases at least, not a
correct one. Possibly some combination of certain features of the
tropism theory with a consideration of the facts of changes in physio-
logical condition may give us a satisfactory answer to the problems
of direction.

INTERFERENCE OF STIMULI.

Again, we have in the interaction or interference of stimuli certain
phenomena which seem to fall under our present point of view. First,
we have the so-called cases of '* heterogeneous induction,'* where the
action of one stimulus reverses or modifies the reaction to another.
For example, many cases are known in which animals positively photo-
tactic become negative, or vice versa^ when the temperature is changed.
(For a collection of such cases see Loeb, 1893, and Davenport, 1897,
p. 199.) In these cases the physiological condition of the organism
seems altered by one stimulus (as heat or cold), in sucli a way that it
no longer reacts to another stimulus (light) as it did before. In the
ciliate infusoria, specimens which are in contact with solids do not
react at all to many agents which under other circumstances call
forth A marked reaction. Putter (1900) has made a special study of

I20 THE BEHAVIOR OF LOWER ORGANISMS.

this matter, describing especially the interference of contact with the
reaction to heat and to electricity. In the second of these contributions
(p. 32), we have seen that attached Stentors do not react at all to
light. Physically considered, there is no necessary opposition between
the action of contact and the action of the other stimuli named. We
must conclude then that contact with solids so alters the physiological
condition of the organism that it no longer reacts to the other stimuli.

SPONTANEOUS MOVEMENTS.

Further, we find that alterations in physiological condition may cause
definite movements, which take place without external stimulus. Such
are the movements which we call spontaneous. As an example of this
we may take the case of Hydra. If an undisturbed green Hydra is
observed continuously, it is found to contract and again to extend with-
out visible cause every i^ to 2 minutes. Thus, it remains at rest for
a period of, say, i^ minutes. Its physiological condition at this time
we may call X. At the end of this period it contracts. Since the
external conditions have not changed the Hydra itself must have
changed, otherwise it would continue at rest. The physiological condi-
tion ^passes into the condition T^ and the Hydra as a result contracts.
This contraction is, of course, exactly the reaction given as a response
to most stimuli in Hydra. In Vorticella we find similar spontaneous
contractions at intervals, essentially as in Hydra. Cases of movements
in the lower organisms that are inaugurated by internal changes in
condition could, of course, be multiplied indefinitely. For our present
point of view it is of importance to recognize clearly the fact that a
change in physiological condition may, by itself, cause exactly the same
behavior that at other times appears as a response to external stimuli.

METHODS OF CAUSING CHANGES IN PHYSIOLOGICAL
CONDITION.

Changes in physiological condition are thus evidently brought about
in a number of different ways. We may attemfft to summarize here
the diflferent methods which appear to exist in the lower organisms.

(i) A single simple stimulus may bring about a change in physiologi-
cal condition. This is proved by the fact that the organism after it has
received a single stimulus may react differently from its previous
method. Thus, Stentor reacts to a single touch, but after this single
touch it may no longer react when touched in the same way again ;
or it may react in a different manner. It is probable, further, that the
first reaction to a single simple stimulus is to be considered due to a
change in physiological condition produced by this stimulus.

(2) Repetition of the same stimulus may cause a change in physio-
logical condition such as is not produced by a single stimulus. This

PHYSIOLOGICAL STATES AS DETERMINING FACTORS. 121

again may be illustrated from the experiments on Stentor, described
above (p. 112).

(3) Internal causes, not definable, may give rise to changes in physio-
logical condition. The result is spontaneous movement at intervals,
as described above for Vorticella and Hydra. As many authors have
pointed out, rhythmical soontaneous movements may be due to a steady,
non-rhythmical, internal change of condition.

(4) The movement or reaction performed by the organism may
change the physiological condition. This is illustrated in one way by
the fact that after a single spontaneous contraction in Vorticella or
Hydra, the animal remains quiet for an interval, showing that the
original physiological condition was restored by the movement. The
fact that the reaction performed by the organism changes the physio-
logical condition of the latter is, of course, the basis of the formation of
habits in higher organisms ; in this case the performance of a reaction
once or repeatedly throws the organism into a condition where it is
more likely to react in the same way again. This particular method
of alteration of condition has perhaps not been clearly demonstrated
for unicellular organisms, though there is some indication of it in the
behavior of Vorticella, as described by Hodge & Aikins (1895), and
of Stentor as described by myself in the ninth of my studies. Thus,
Stentor responds to carmine in the water by a series of different reac-
tions, finally reaching the condition where it reacts by contracting at
once into its tube. If the stimulus is now repeated every time the
Stentor extends, it never gives its earlier method of reaction, but reacts
steadily for a long time by contracting at each stimulus. Is this persist-
ence in the contraction reaction due partly to the fact that it has begun
on this reaction method, and therefore keeps it up, or is it due only to
the fact that the stimulus has been repeated many times ? In the former
case the behavior would perhaps fall under our present point of view ;
in the latter it would not. Cases among the Protozoa where the repeated
performance of a reaction clearly makes the further performance of the
same reaction easier or more likely to occur, would be of much interest.

NATURE OF REACTIONS TO STIMULI.

The foregoing considerations evidently have a definite bearing on
the problem of the nature of reactions to stimuli. They lead, as set
forth briefly on page 118, to the following conception of the steps oc-
curring in a reaction to a stimulus: (i) The stimulus acting on the
organism causes a change in its physiological condition ; (2) this
change in physiological condition gives rise to the typical reaction.

The evidence for this view is found scattered throughout the fore-
going discussion ; its main points may be briefly summarized here as
follows :

132 THE BEHAVIOR OF LOWER ORGANISMS.

(i) We have seen that stimuli do unquestionably cause changes in
physiological condition. This is demonstrated by the fact that after a
stimulus has occurred and ceased to act, the organism reacts differently
to the same or other stimuli. (For examples see pages 112, 117.)

(2) We have seen that the changes in physiological condition do un-
questionably cause definite movements, of exactly the sort that we are
accustomed to call reactions to stimuli. (Contraction of Hydra or Vor-
ticella, etc. ; see p. 120.)

These two facts give a solid foundation for the above view of
reactions to stimuli, and, indeed, it seems to me, raise a presumption
that reactions to stimuli are, as a rule, brought about in the way
described. Further evidence in favor of this view is as follows :

(3) In the paper which precedes the present one we have demon-
strated that, in the Infusoria and Rotifera at least, the action of stimuli
is not directly on the motor organs of that part of the body on which
the stimulus impinges. The organism reacts as a whole, and in a way
that is not explicable even on the assumption of a definite plan of ner-
vous interconnection between the regions stimulated and the motor
organs, an assumption that is, of course, in any case not allowable for the
Infusoria. Such reactions cannot be explained otherwise than as due
to changes in the physiological condition of the organism as a whole.
Further, evidence was given to show that the reactions of higher organ-
isms are in many cases equally inexplicable as a result of direct action
of the stimulus on the motor organs.

Only in the reaction of some organisms to the constant electric cur-
rent did we find such conditions fulfilled as permit an explanation of a
part of the phenomena on the theory of the direct action of the agent
on that part of the body on which it impinges, in accordance with the
theory of tropisms. Other features of the reaction to this stimulus (in
many cases the determining ones) are only explicable on the theory
that they are due to the physiological state of the organisms as a
whole, induced by the stimulus (see p. 100). This shows that we may
find at any time these two methods of action mixed, or perhaps eithei
one separately. But the reaction to the electric current is the only one
out of the reactions to a multitude of agents, that, in the Infusoria, has
been shown to have this additional feature — reactions of different parts
of the body in opposed ways. The reaction to the electric current is
thus of the very greatest interest, not because it stands as type for
reactions in general, but for exactly the opposite reason, because it
presents factors which are not known to occur in other reactions.

(4) The view that reactions to stimuli take place through the inter-
mediation of changes in the physiological condition of the organism as
a whole is further reinforced by the fact, set forth above, that it is only

PHYSIOLOGICAL STATES AS DETERMINING FACTORS. 1 23

on the basis of such a view that we can understand the changes of
reaction which occur when the same stimulus is repeated ; by the facts
of the interference of stimuli, even when their direct physical action is
by no means opposed ; by the facts of heterogeneous induction, and by
the fact that organisms at different times of the day or at different sea-
sons show different methods of reaction to the same stimuli.

(5) This view is also strengthened by the fact that it brings into
relation reactions to stimuli and spontaneous movements. On this view
both are due directly to the same cause, to changes in physiological
condition, produced in one case by internal causes, in the other by
external causes.

(6) This view receives powerful support, it seems to me, in our
knowledge of what takes place in the higher animals, including man.
This point I shall attempt to develop farther on in the present paper.

This view, that reactions to stimuli in the lower organisms are pro-
duced in general through changes in physiological condition, is not, of
course, set forward as anything new or original. Many others have
doubtless taken this point of view, and it is implied, perhaps not always
consciously, in many attempted explanations of animal behavior.
The writer is merely attempting to emphasize that particular interpre-
tation, out of many existing ones, towards which the facts seem to
point strongly. He is convinced that the factor of physiological con-
dition as determining behavior has not Deen so fully and explicitly
realized and dealt with in work on the lower organisms as the facts
demand, and that many things that seem anomalous fall into their
proper places when this factor is taken fully into consideration.

What is the nature of physiological conditions, or changes in phys-
iological condition.? Of course, we are not able to answer this ques-
tion. One is tempted to think of these expressions as signifying
something like chemical states or changes in chemical states. But the
concept of physiological states is, for higher animals at least, one at
which we arrive by analysis of complex phenomena in behavior, and
this does not give us any direct evidence as to the real nature of the
change in the living substance (considered as matter) which takes
place when the physiological condition changes.

The concept '' physiological states" is a preliminary collective con-
cept, which may later be analyzed into many. Such analysis is certain,
however, to be difficult and hypothetical in character in the lowest
organisms. In man we have, of course, a basis for analysis in the sub-
jective accompaniments of physiological (here called psychological)
conditions, — in the feelings, emotions, etc.

124 THE BEHAVIOR OF LOWER ORGANISMS.

PHYSIOLOGICAL STATES IN BEHAVIOR OF HIGHER ANIMALS,
AS COMPARED WITH THOSE IN LOWER ORGANISMS.

Realization of the fact that the behavior, even in the lowest organisms,
is determined to a large degree by physiological states must be of great
service in welding into one connected whole the study of behavior in
all animals, from the lowest up to man. The attempt to divorce the
study of the behavior of man from that of the lower animals, which
has been evident in late years, seems unfortunate and unnecessary. It
is true that we are not justified in reading the subjective states of man
directly into the lower organisms. But we are not confronted with the
alternative of doing this or of separating the two subjects completely.
The behavior of man can be studied from the same objective standpoint
which we employ in investigating the behavior of animals. When this
is done, there is no reason for holding the results on man aloof from
those obtained elsewhere ; if it is proper to compare different organ-
isms of any kind from this point of view, in order to obtain general
results, as all investigators do, it is certainly proper to draw man also
into the circle of comparison. The fact that in man we can know also
the subjective accompaniments of the different physiological states and
reactions is by no means a disadvantage in this comparison ; it is merely
an additional feature, of the highest possible interest. We can even,
it seems to me, justifiably call attention to the relation between the
subjective states as found in man to certain general phenomena common
to man and other organisms. It is only when we proceed directly to
attribute to the lower animals the subjective states which we know only
in man (and, indeed, only in our own individual mind?) that we pass
the boundary of scientific procedure.

In the higher animals, and especially in man, the essential features
in behavior depend very largely on the history of the individual ; in
other words, upon the present physiological condition of the individual,
as determined by the stimuli it has received and the reactions it has
performed. But in this respect the higher animals do not differ in
principle, but only in degree, from the lower organisms, as we have seen
in our analysis of the behavior of Stentor. In this unicellular form
we were forced to distinguish at least six different physiological condi-
tions, determining in the same individual different reactions to the same
stimuli. In the higher animals, and especially in man, we can distin-
guish, as might be expected, an immensely greater number of such
conditions which induce different reactions, but there is no evident differ-
ence in principle in the two cases. Can we go farther and make a more
direct comparison of individual physiological states in the higher and
lower organisms? We find in Stentor, and again in the flatworm,
that after the organism has been repeatedly stimulated by an agent

PHYSIOLOGICAL STATES AS DETERMINING FACTORS. 1 25

which must in the long run be classed as injurious, it is thrown into a
physiological condition in which its reactions become more rapid and
powerful, and of such a nature as to remove the organism from thfe
source of stimulus. We find that in this state the organism reacts to i
any stimulus to which it reacts at all by a strong negative reaction, j
In higher animals we frequently find the same condition of affairs, andj
the animal is then commonly said to be frightened. Finally, we often\
find in man a similar condition, and here we know certain subjective i
accompaniments of the physiological condition, the most characteris-^>
tic of which is perhaps the emotion of fear. In all these cases the
objective manifestations of the physiological condition are of the same
character. Does the fact that in man we know something additional
about the matter, the subjective accompaniments, constitute grounds
for denying the essential similarity, from a physiological standpoint,
of this condition in man and that in the lower organisms ? It seems to
me that it does not ; in fact, all that is maintained in making the com-
parison is that this condition causes similar objective phenomena and
is brought about by similar conditions. Further than this our analysis
and comparison cannot go.

Another class of physiological conditions which we can distinguish ?
almost all the way through the animal series is that produced character-
istically by intense stimuli, as opposed to faint stimuli. As a rule, any
stimulus, even if it is one to which the organisms respond usually by a
positive reaction, produces, when it becomes very intense, reactions
whose general effect is to remove the organisms from the source of
stimulation (negative reactions).* This is true in Amoeba, where weak
mechanical stimuli cause spreading out and movement toward the
source of stimulus, while strong mechanical stimuli cause it to contract
and move away ; it is true for Stentor ; it is true for the stimulus of
weak and strong light in Euglena and Volvox and many other
organisms ; it is true for mechanical and chemical stimuli in the flat-
worm ; it is true in general for higher animals and man. In all these
cases the intense stimulus evidently changes the physiological con-
dition so that the organism now reacts negatively. In man we know>
that this physiological condition is accompanied subjectively by pain'
or at least discomfort, and even in higher animals such reactions are
usually spoken of as pain reactions. Objectively considered, the
phenomena are analogous throughout the animal series, so that we

* "All organisms behave in two great and opposite ways toward stimulations ;^
they approach them or they recede from them. Creatures which move as av
whole move toward some kinds of stimulations, and recede from others. Crea- •
tures which are fixed in their habitat expand toward certain stimulations, and/
contract away from others." Baldwin, 1897, p. 199.

126 THE BEHAVIOR OF LOWER ORGANISMS.

I may properly characterize the physiological condition which produces

j the negative reaction to strong stimuli, with Professor J. Mark Bald-

\ win (1897, p. 43) as the '' physiological analogue of pain." This, of

course, by no means commits us to the belief that the organisms have

a sensation of pain ; concerning this we know nothing.

It thus seems to me possible to trace some of the physiological con-
ditions which we know, from objective evidence, to exist in man and
the higher animals, back to the lowest organisms. Many conditions that
we can clearly distinguish in man will doubtless be followed back to a
common single condition in the lower organisms ; but this is exactly what
we should expect. Differentiation takes place as we pass upward in
the scale, in these matters as in others.

The most interesting and important field in which we find the
behavior of higher organisms dependent on their previous history, and,
therefore, on their present condition as influenced by previous experi-
ence, is in that group of phenomena which we call memory, or learning
by experience. Memory has as its basis the general phenomenon that
a stimulus received or a reaction performed leaves a trace on the
organism, or modifies its condition in such a way that it later reacts
differently to the same stimulus. This basis of memory is, of course,
clearly present in Stentor.

The analysis of the different physiological conditions found in the
lower organisms, the influences to which they are due, and the
reactions of these organisms as influenced by physiological conditions
certainly forms a most promising field for research, and one as yet
almost untouched.

SUMMARY.

The present paper attempts to show, by an analysis of certain
phenomena in the behavior of lower organisms, taking Stentor and
Planaria as types, that physiological states of the organism are most
important determining factors in reactions and behavior. In these
organisms, to the same stimuli, under the same external conditions,
the same individuals react at different times in radically different ways,
showing the existence of different physiological states of the organism,
which determine the nature of the reactions. In a unicellular organ-
ism (Stentor) we can distinguish at least six different physiological
states, in each of which the organism has a different reaction method,
and corresponding facts are brought out for the flatworm. Scattering
observations taken from works on tropisms, etc., are shown to indicate
that the same state of affairs is found in other lower organisms.

The conditions producing these different physiological states are
examined and their importance for the theory of behavior in the lower
organisms is brought out. The relations of these facts to " interference

PHYSIOLOGICAL STATES AS DETERMINING FACTORS. 1 27

of stimuli," "heterogeneous induction," "spontaneous movements,"
and "changes in the sense of reactions with a change of intensity in
the stimulus," are developed.

The view is set forth that in most of the lower organisms a reaction
to stimulus usually involves the following factors: (i) the stimulus
changes the physiological state of the organism as a whole ; (2) this
change in physiological state induces a certain type of reaction.
Evidence for this view is summarized.

Finally, it is pointed out that realization of the importance of
physiological states as determining factors in the behavior of the lower
organism is of service in bringing the study of these organisms into
relation with that of higher animals and man. An objective study of
the behavior of these higher animals shows the prevalence of physio-
logical states as determining factors in behavior, and in some cases, at
least, some of these states are closely analogous to what we find even
in unicellular organisms.

SIXTH PAPER

THE MOVEMENTS AND REACTIONS
OF AMCEBA.

139

CONTENTS.

Introduction: Objects of the Investigation, 131

Description of the Movements and Reactions, 132

The Movements, 132

The Movements of Amoeba as described Formation and Retraction of Pseudopodia 15a

by Rhumbler and Butschli ; Agree- Surface Currents in Formation of

ment with Currents in a Drop of Pseudopodia in contact with Sub-

Fluid Moving as a Result of a Local

Decrease in Surface Tension 132

Currents in Amoeba as studied from above;

stratum 152

Formation of Free Pseudopodia 153

Withdrawal of Pseudopodia 156

Movements at Anterior Edge 160

Lack of Backward Currents 134 Movements of Posterior Part of Body. ... 165

Movements of Upper and Lower Surfaces General View of Movements of Amoeba in

Studied Experimentally; Rolling Locomotion 169

Movement 138 Some Characteristics of the Substance of

Amceia verrucosa Sind Its Rclsitives. 140 Amoeba 173

Other Species of Amoeba 146 Fluidity 173

Historical on Rolling Movements in Rhumbler's Ento-ectoplasm Process 173

Amoeba 148

Elasticity of Form in Amoeba 175

Contractility in Ectosarc of Amoeba. 177

Reactions to Stimuli 181

Reactions to Mechanical Stimuli 181 Some Complex Activities 193

Positive Reaction 181 Activities connected with Food-taking 193

Negative Reaction 182 Taking Food ^ 193

Reaction to Chemical Stimuli 187 Pursuit of Food 196

Reaction to Heat 190 Other Amcebse as Food 198

Reactions to Other Simple Stimuli 191 Reactions to Injuries 202

Physical Theories and Physical Imitations of Amoeboid Movements, . 204

Surface Tension Theory 204 Experimental Imitation of Movements

Berthold's Theory that One-sided Adher- due to Local Contractions of Ectosarc

ence to Substratum is the Cause of and of the Roughening of Ectosarc in

Locomotion 208 Contraction 215

Experimental Imitation of Locomo- Direct or Indirect Action of External

tion in Amoeba 209 Agents in Modifying Movements 219

Formation of Free Pseudopodia 214 Direct or Indirect Action in Food-taking, 222

General Conclusion 225

Behavior of Amoeba from Standpoint of Comparative Study of Animal

Behavior 226

Habits in Amoeba 226 Relation of Different Reactions to Differ-

Classes of Stimuli to which Amoeba Re- ent Stimuli; Adaptation in Beha-

acts 227 vior of Amoeba 227

Types of Reaction 227 Reflexes and "Automatic Actions" in

Amoeba 228

Variability and Modifiability of Reactions 229

Summary, . . , , 230

130

THE MOVEMENTS AND REACTIONS OF AMCEBA.

INTRODUCTION; OBJECTS OF THE INVESTIGATION.

The present paper contains the results of an investigation which was
undertaken with two general problems in mind. The first purpose
was to determine by observation and experiment, from the standpoint
of the student of animal behavior, how far recent physical and
mechanical theories go in aiding us to explain the behavior of Amoeba.
The second object of the work was to furnish needed additional data
on the reactions of i\.moeba to stimuli, and to systematize and unify
our knowledge of its behavior.

The recent theories which would resolve the activities of Amoeba
largely into phenomena due to alterations in the surface tension of a
complex fluid seem to promise much. They are of precisely the
character from which most may be hoped ; from a study of the physics
of matter in a state similar to that found in the living substance, the
laws of action of this living substance are sought. Such theories have
been developed, as is well known, by Berthold (1886), Quincke
(1S8S), Biitschli (1892), Verworn (1892), Rhumbler (1898), Bern-
stein (1900), Jensen (1901), and others. The success of this method
of attacking the problems seems great. Activities similar, at least
externally, to those of Amoeba, are produced by physical means, and
fully analyzed from the physical and mechanical standpoint. In this
manner the movement, the control of movement by external agents,
the feeding, the choice of food, the making of the shell, and other
features of the behavior have been more or less closely imitated,* and
in a way permitting a complete analysis in accordance with chemical
and physical laws.

From the standpoint of the student of animal behavior, the resolu-
tion of the behavior of any organism into the action of known physical
laws must be a matter of the deepest interest. The actions of higher
organisms seem at present so far from such a resolution that some
investigators believe an essential difference in principle to exist
between the behavior of living things and non-living things ; between
the laws of biology and those of physics. The resolution, then, of the
behavior of even the simplest organism into known physical factors
would be an event of capital significance, affecting fundamentally the
whole theory of animal behavior. A renewed thorough study of the

* See especially Rhumbler, 1898. 131

132 THE BEHAVIOR OF LOWER ORGANISMS.

facts, with especial reference to these theories, seems, therefore, much
to be desired. The^ results of the present study will show, I believe,
that such a re-examination of the fjicts was greatly needed.

As to the second object of this investigation, stated above, it is a
somewhat remarkable fact that the observational basis for a number of
the most important reactions assumed to exist in Amoeba is exceedingly
scanty, particularly so far as control of the direction of movement is
concerned. For example, one of the reactions most often assumed to
exist in Amoeba, and most commonly selected for imitation by
physical means, is chemotaxis, the movement toward or away from a
diffusing chemical. But no account exists, so far as I have been able
to discover, of actual observation of such a reaction in Amoeba, under
experimental conditions. Again, the effects of slight or of intense
localized mechanical stimuli, in controlling the direction of movement,
has not been worked out in detail. To fill these and similar gaps in
our knowledge, and to bring the different reactions into relation with
each other, so as to make possible a connected account of the behavior
of Amoeba, is, then, the second object of this paper.

I shall first give an account of the movements and reactions of
Amoeba, as determined by observation and experiment, without enter-
ing in detail upon the theories of the subject. This will be followed
by a section dealing with the physical theories and physical imitations
of the movements and reactions, in the light of the facts set forth in the
first section. A brief final section will be devoted to a characterization
of the behavior of Amoeba from the standpoint of the student of
animal behavior.

I am compelled to give a full description of the normal movements
of Amoeba, as the course of the investigation showed that the prevalent
conception of these movements, on which many of the theories have
been based, is not correct.

DESCRIPTION OF THE MOVEMENTS AND REACTIONS.
THE MOVEMENTS.
MOVEMENTS OF AMCEBA AS DESCRIBED BY RHUMBLER AND BUTSCHLI ;
AGREEMENT WITH CURRENTS IN A DROP OF FLUID MOVING AS A RE-
SULT OF LOCAL DECREASE IN SURFACE TENSION.

There are few subjects that have been studied more than the nature
of the movements of Amoeba, but nothing final has been reached, even
from the descriptive standpoint. The first preliminary to an under-
standing of the nature of the movements must be to determine just what
movements take place.

The most extensive recent study of the movements of Amoeba has
been made by Rhumbler (1898), though the magnificent monograph of

THE MOVEMENTS AND REACTIONS OF AMOBBA.

133

the Rhizopods by Penard (1902) contains incidentally a large number
of valuable observations on this matter.

According to Rhumbler (/. c.) the movements in normal locomotion
are typically as follows : From the hinder end of the Amoeba (or of the
pseudopodium, if a single pseudopodium is under consideration) a
current of endosarc passes forward in the middle axis ; in front this flows
outward toward the sides, then backward along the surface, gradually
coming to rest. Figs. 30 and 31 , taken from Rhumbler, give diagrams
of these currents in an Amoeba moving as a whole (Fig. 30), and in
the formation of pseudopodia (Fig. 31). In an Amoeba which forms
more than one pseudopodium at once, these typical currents become
somewhat complicated (Fig. 32), but retain their main features. The
backward current shown at the sides in Figs. 30-32 is conceived to be
present also above and below, that is, over the whole surface of the
Amoeba. A diagram of the currents in side view, as given by Rhum-
bler, is shown in Fig. 33, B. An essentially similar account of the
currents is given by Biitschli (1880, 1892).

Fig. 30.*

Fig. 31. t

Fig. 32. t

Fig. 33.§

The most striking feature in the currents as above set forth is the fact
that they agree precisely with the currents produced in a drop of fluid
of any sort when the surface tension is lowered over a certain limited
area. There is always a current over the surface away from the region
where the tension is lowered, while an axial current moves toward the

♦Fig. 30. — Diagram of the currents in a progressing Amoeba Umax, after
Rhumbler (1898).

fFiG. 31. — Diagram of the ** fountain currents " in pseudopodia of Amceba,
after Rhumbler (1898).

% Fig. 32. — Diagram of complex '* fountain currents" in an Amoeba with two
large pseudopodia, after Rhumbler (1898).

§ Fig. 33. — Comparative diagrams of the currents in a rolling movement, and
in the movement of Amoeba, as conceived by Rhumbler, viewed from the side.
In A are represented what Rhumbler conceives to be the necessary currents in
a rolling movement, while B represents what Rhumbler considers the really
existing currents in Amoeba, as seen from the side. The heavier arrows in each
case represent the current on the lower surface. After Rhumbler (1898).

134 "^"^ BEHAVIOR OF LOWER ORGANISMS.

point of lowered tension. Diagrams of the movement of such drops
are given in Fig. 34. Further, the drop may elongate in the direction
of the axial current, and may move bodily in that direction, just as
happens in Amoeba.* It is most natural, therefore, to conclude as
Biitschli (1892) and Rhumbler (1898) have done, that the movements
of Amoeba are likewise due to a lowering of the surface tension at the
anterior end, provided that its movements really take -place in the
way described above,

CURRENTS IN AMCEBA AS STUDIED FROM ABOVE ; LACK OF BACKWARD

CURRENTS.

At the beginning of my work I had no doubt that the movements
occurred exactly as above described, and, therefore, did not devote
special attention to this point. But I was soon struck by the fact that
I was unable to see any backward current at the sides, as represented
in Figs. 30 and 31. Further careful study of the movements of
Amoeba limax^ A. proteus^ A. angulata^ A. verrucosa^ A, sphcero-

nucleolus,, and one or two
undetermined species
confirmed this fact, and I
may say at once that after
several months' continu-
-^ F + ^ ^"^ study of the move-

ments and reactions of
Amoeba I have never, except in one or two doubtful instances, seen
any backward movement of the substance at the sides or on the surface
of an Amoeba that was moving forward in a definite direction.

It is true that in the movements of Amoeba limax^ for example, one
receives the impression of two sets of currents, one forward in the cen-
tral axis, the other backward at the sides. But if the latter is studied
carefully it is found that there is really no current here ; the proto-
plasm is at rest, and the impression of a backward current at the sides
is produced only by contrast with the forward axial current. Amoeba

♦All these facts are easily verified bj placing a drop of clove oil on a slide in
a mixture of two parts glycerine to one part 95 per cent alcohol under a cover
supported by glass rods, as described in a previous paper by the present author
(Jennings, 1902). By mixing some soot or India ink with the clove oil the cur-
rents are made evident.

t Fig. 34. — Currents in a drop of fluid when the surface tension is decreased
on one side. A, the currents in a suspended drop, when the surface tension is
decreased at a. After Berthold (1886). ^, axial and surface currents in a drop
of clove oil, in which the surface tension is decreased at the side a. The drop
elongates and moves in the direction of a, so that an anterior (a) and a posterior
(^) end are distinguishable.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 1 35

Umax contains usually a large number of fine granules, which in many
cases extend to the very outer surface, so that it is not possible to dis-
tinguish an ectosarc, in the sense of a layer containing no granules.
By watching the movements of these particles it is possible to determine
the direction of the currents in the protoplasm. The movements in
locomotion are usually as follows : At the anterior end there pushes
forth from the interior a clear substance, which I will call the
hyaloplasm. As this moves forward it spreads out laterally, till it
reaches a position such that it forms a continuation forward of the
remainder of the lateral boundary of the animal. Into this hyaloplasm
flows then the granular endosarc. The granules flow forward, rapidly
in the middle, usually more slowly near the sides. As it reaches the
anterior end the central current spreads out in a fanlike manner, so
that some of the granules approach closely the lateral borders of the
Amoeba (Fig. 35). They then stop, while the central part of the cur-
rent passes on, following the advancing anterior end.

So long as one confines his attention to the Amoeba alone, not
observing external objects,
one receives the impres-
sion that there are two
sets of currents, an axial
current forward, marginal

currents backward. But

a !-• Fig. ^c.*

as soon as one fixes his ^^

eye upon a particular granule in the apparent backward marginal
current, and observes its relation to some external object, he dis-
covers that no such current exists. The granule remains quiet,
retaining continually its position with relation both to other granules
in the edge of the Amoeba and to objects external to the Amoeba.
Meanwhile the remainder of the substance of the Amoeba is flowing
past, so that the granule in question after a time comes to occupy
a position at the middle of the length of the Amoeba. At about
this point it usually begins to move slowly forward again, though
much less rapidly than the internal current. The nature of this
slow forward movement we shall take up later (p. 166). The main
portion of the body of the Amoeba thus continues to pass the granule,
and the latter finally reaches the posterior end. Here it usually re-
mains quiet for a time (moving forward only as the posterior end is
dragged forward). Then it is taken into the central current again,
passes to the anterior end, and comes to rest as before, while the
remainder of the Amoeba passes it by ; and this process is repeated

* Fig. 35. — Diagram of the movements of particles in an advancing Amoeba.
Each broken line represents the path of a particular particle.

136 THE BEHAVIOR OF LOWER ORGANISMS.

indefinitely. In favorable cases I have repeatedly followed a single
granule from the posterior end forward till it came to rest at the
anterior end, then watched the body of the Amoeba pass it by, until it
was again at the posterior end and started forward anew. The course
of a single granule is represented in Fig. 36. As is evident from this
figure, the granule does not travel backward in any part of its course.
Not all the granules, however, remain quiet until they have passed to
the posterior end. Many of them are taken again into the central
stream before the entire body of the Amoeba has passed them. Large
granules usually stop only a short time, starting forward again before
the middle of the Amoeba has reached them ; others are taken up at the
middle or farther back, while many smaller granules reach the posterior
end. But as a rule none show any movement backward, so far as I
have observed.

It is not only at the margins of the Amoeba, but also on the under
surface, in contact with the substratum, that the ectosarc with its
granules is at rest or moving slowly forward in the posterior half.
This is evident when the lower surface of a transparent Amoeba is
brought into focus.

^ 2 That excellent observer,

i -'x^ ..^ Dr. Wallich, saw clearly

/"^ — ,^-;?<^ — "^^ fS^ — ZXA-*- cy. J many years ago that there

Vj-^ — ~^i_^-{ — ~3p^-=r^^:^^'-f^L,^__J A is really no backward cur-

o. ^ 'c "^ e J rent, though at first view

Fig. 36.* there appears to be such.

It is only necessary to watch a specimen of Amoeba carefully to become con-
vinced that the appearance of a returning, as well as an advancing, stream of
granules is illusory. The stream, it will be observed, is invariably in the direc-
tion of the preponderating pseudopodial projections. The particles simply flow
along with the advancing rush of protoplasm. There is no return stream, but
the semblance of one is engendered by one layer of particles remaining at rest
whilst another is moving past them. (Wallich, 1863, b, p. 331.)

This statement of the facts my observations fully confirm.
In this account of the lack of backward movement in the granules
of the ectosarc on the lower surface and at the margins I find myself

♦Fig. 36. — Diagram of the movements of a single particle in Amoeba, as
seen from above. The particle begins at a, passes to b and then to c, at the
anterior edge of the Amoeba shown in the outline i. The Amoeba now passes
forward to the position 2, and thence to 3, while the particle retains the posi-
tion c; when the Amoeba has reached the position 3 the particle is thus at
its posterior end. Now the particle moves forward again, from c to d, and
thence to e and/", thus coming again to the anterior edge. Here it stops, as at
c, until the body of the Amoeba has passed it. As the figure shows, the particle
does not move backward in any part of its course.

THE MOVEMENTS AND REACTIONS OF AMCEBA. I37

at variance with certain statements of F. E. Schulze, Biitschli, and
Rhumbler. I am aware that this conflict of my observations with
those of the investigators named, who deservedly rank among the
highest in the field at present under consideration as well as elsewhere,
renders the utmost caution necessary in trusting to these results. Yet,
with this consideration in mind, and with the confident expectation in
undertaking the work that I should find the currents exactly as described
by these authors, I have been unable to come to any result save that
above set forth. The statements of Schulze (1875, pp. 344-348) deal
with Pelomyxa falustris Greef. In this animal, according to Schulze,
there are resting portions at the sides of the body, while from behind
currents pass forward through the channel enclosed by these resting
portions. At the anterior end the lateral parts of these currents turn
outward and, finally, a little backward ; any
given portion passes backward but a short dis-
tance. The currents are shown by Schulze in
a figure, a reduced copy of which is given here-
with (Fig. 37). According to Schulze the cur-
rents have this form on the upper surface as well '^^"^■***''^'^^^^^^'f^^ V^—x
as at the sides — that is, a part of the current flows
upward and backward on the upper surface.
Biitschli (1892, Anhang, p. 220) confirms this
account of the currents in Pelomyxa.

I regret that I have been unable to obtain
specimens of Pelomyxa in order to examine
these phenomena for myself. One would of
course be bold to doubt the correctness of the

KiG Vl *

observations of such investigators as Schulze

and Biitschli, and it is possible that Pelomyxa differs from Amoeba
in this respect. Yet, as we shall see later (p. 149), the account given
by these authors is certainly incorrect so far as the backward cur-
rents on the upper surface are concerned ; it is possible, then, that the
appearance of a backward current elsewhere was deceptive.

Rhumbler (1898) describes the forward axial and the backward side
currents in various species of Amoeba, and considers such movements
as typical, basing his theory of locomotion upon them. It seems
probable that slight backward, currents, such as were described by
Schulze (Fig. 37), do occur at times at the sides of the advancing
anterior end. The posterior part of the Amoeba is narrow and
rounded, the anterior part broad and thin. The current of endosarc
flows from this narrow posterior portion into the broad anterior part

♦Fig. 37. — Currents in a progressing Pelomjxa, as seen from above, after
Schulze (1875). The longer arrows represent stronger currents.

138 THE BEHAVIOR OF LOWER ORGANISMS.

and must therefore spread out ; it would not be unnatural for the
currents to flow even backward a little, as in Schulze's figure (Fig.
37), in order to fill the area just in front of the resting portion of the
protoplasm (at x, Fig. 37). As we shall see, such movement is
sometimes to be observed in inorganic fluids under similar conditions
(p. 211). Whatever the explanation of the difference between my
observations and those of the investigators named, the point of impor-
tance is that the backward current is not a constant nor an essential
part of the locomotion of Amoeba, so that it does not form a fitting
basis for a theory of locomotion. Further, as we shall see, I am able
to demonstrate conclusively the incorrectness of that conception of the
nature of amoeboid movement for which alone the account of the
currents given by Biitschli and Rhumbler is significant.

It is evident that the method of movement here described is better
adapted to the production of locomotion in a given direction than that
which Biitschli and Rhumbler describe (see Figs. 30-33), since accord-
ing to their account a portion of the substance of the body is first trans-
ported forward, then backward. In the locomotion as I observed it
there is no such useless transportation of substance in a direction
opposed to that in which the animal is traveling.

On the other hand, the movements as I have described them bear
much less resemblance to those produced in drops of fluid by local
changes in surface tension (Fig. 34) . There is only the slight turning
outward at the anterior end that can be at all compared to the backward
flow of an outer layer in the inorganic drop. Rhumbler himself notes
that in Amoeba angulata there is often no such backward current to
be seen (Rhumbler, 1898, p. 120), but bases his theory of the forward
movement entirely on the cases where it (supposedly) does occur. In
Amceba angulata^ A. verrucosa^ and A. sphceronucleolus^ according
to my observations, there is often no indication even of the turning
out of the particles in a fanlike manner ; they merely flow forward and
stop for a time. Biitschli (1892, p. 199) notes that the backward cur-
rent at the anterior end of Amoeba, required by the surface tension
theory, is very slight, but conceives it to be suflicient to fulfill the
requirements of the theory.

MOVEMENTS OF UPPER AND LOWER SURFACES STUDIED EXPERIMEN-
TALLY ROLLING MOVEMENT IN AMCEBA.

Thus far we have left out of consideration the movement of substance
on the upper surface of the Amoeba. It is usually assumed that the
condition here is the same as at the sides and on the under surface ;
thus Rhumbler gives a diagram, reproduced in my Fig. 33, B^ showing
the backward current of the upper surface. The positive observations

THE MOVEMENTS AND REACTIONS OF AMCEBA. I39

on this point are those of Schulze (1875), Berthold (1886, p. 109), and
Biitschli (1892, p. 220). These authors all agree that the backward
current visible at the sides of the anterior end in Pelomyxa are clearly
also present on the upper surface. It is not usually possible to observe
particles moving backwrard on the upper surface of Amceba, nor even
particles at rest, though this might be due to the fact that the granules
have sunken downward, leaving the upper surface clear. But to
decide whether the currents in Amoeba are essentially like those pro-
duced in a drop of fluid by a local change in surface tension, it is most
important to determine with certainty what is taking place on the upper
surface.

Evidently the most natural way of doing this is to cause, if possible,
some small object to rest upon or become attached to the upper surface
of Amceba, then to observe the movement of this object. This can be
done by mingling a considerable quantity of soot with the water in
which the Amoebae are found. Some of the soot particles settle on the

Fig. 38.* Fig. 39.!

upper surface of the Amoebae, and in some species they adhere to this
surface.

I was quite unprepared for the results of this experiment. The
upper surface of Amceba moves forward^ not backward, as required
by the surface tension theory ; nor is it at rest like the lower surface.

* Fig. 38. — Movements of a particle attached to the outer surface of Amoeba
verrucosa. When first seen the particle was at the posterior end {p) ; it then
moved forward, as shown by the arrows, until it passed around the anterior end
(a) to the under side. (The Amceba itself of course moved forward at the same
time; no attempt is made to represent its movement in the figure.)

t Fig. 39. — Diagram of the movements of a particle attached to the outer
surface o( Amoeba verrucosa, in relation to the movements of the animal. The
Amoeba is seen from above. In the position i the particle is at the anterior end
of the Amoeba. As the Amoeba moves forward, it passes over the particle,
which retains its place. Thus when the Amoeba has reached the position 2 the
particle is at the middle of its lower surface; when it reaches 3 the particle is at
its posterior end. The particle then passes upward and forward, as shown by
the arrows, so that when the Amoeba reaches the position 4 the particle is in
front of the middle, on the upper surface.

140 THE BEHAVIOR OF LOWER ORGANISMS.

AMCEBA VERRUCOSA AND ITS RELATIVES.

In giving an account of the experiments which demonstrate this, I
shall begin with species of Amoeba in which pseudopodia are, as a rule,
not formed, and the movements are uniform in character, since here the
conditions are simplest from our present standpoint. For this purpose
Amoeba verrucosa Ehr., and particularly the transparent form known
as A. sphceronucleolus Greef, are favorable. In these Amosbae particles
cling rather easily to the outer surface.

When a quantity of soot is added to the water containing Amoebae
of the species named, small masses cling to the surface of the animal.
Such a mass, attached to the upper surface, shows the following
movements : It passes slowly forward (Fig. 38) , then over the anterior
edge, and under the latter. Here it stops, while the Amoeba continues
to move forward. The mass of soot remains quiet until the entire
Amoeba has passed over it and it lies beneath the posterior end. It
now passes upward again, to the upper surface (Figs. 39, 40), then
forward once more to the anterior end. Here it goes under the Amoeba

as before, to be carried upward and forward again when the posterior
end passes over it.

These observations are made with absolute ease, and there is no pos-
sibility of mistaking internal particles for external ones. Particles lying
in the water outside the Amoeba may be seen to become attached at the
posterior end, to pass upward, lying distinctly outside the boundary of
the protoplasm (Fig. 38, posterior end), then forward, till as they double
the anterior end they are again seen sharply defined outside the boundary

♦ Fig. 40. — Diagram of the movements of a particle attached to the surface of
Amoeba verrucosa, in side view. In position i the particle is at the posterior
end; as the Amoeba progresses it moves forward, as shown at 2, and when the
Amoeba has reached the position 3 the particle is at its anterior edge, at x.
Here it is rolled under and remains in position, so that when the Amoeba has
reached the position 4 the particle is still at x, at the middle of its lower surface.
In the position 5 the particle is still in the same place, *, save that it is lifted
upward a little as the posterior end of the Amoeba becomes free from the sub-
stratum. Now as the Amoeba passes forward the particle is carried to the upper
surface, as shown at 6. (Thence it continues forward and again passes beneath
the Amoeba, etc.) The broken lines show that part of the surface of the Amoeba
which is at rest.

THE MOVEMENTS AND REACTIONS OF AMCEBA. I4I

(Fig. 38, anterior end). Further, such particles, after making one or
two revolutions, usually become detached and drop off.

It is thus clear that Amoeba verrucosa and its relatives have what
may be called a rolling motion ; a given spot on the outer pellicula
passes forward on the upper surface, downward at the anterior end,
remains quiet on the lower surface, passes upward at the posterior
end, and again forward. Its movement may be compared directly with
the movement of a given point on the circumference of a wheel that is
rolling forward. A diagram of the movement of a particle on the
surface as it would appear in a side view is given in Fig. 40.

Certain details of the movements are interesting, and may best be
brought out by description of specific observations. In one case two
small particles had become attached, a short distance apart, to the surface
of a specimen of Amoeba spkceronucleolus. They were at first side by
side and a little to the right of the middle line, one somewhat farther
to the right than the other (Fig. 41) . They moved forward in parallel
courses, and reached the anterior edge at the
same time, passing over the edge and to the
under surface. It now required two and one-
half minutes for the Amoeba to pass over p
them, during which time they remained nearly
or quite at rest. They then moved upward
to the upper surface and forward again. The ^ ^

one nearer the middle line moved a little
faster than the other, reaching the anterior edge in two and three-
quarter minutes, while the lateral one required three minutes. Both
emerged at the posterior end again at the same time, the central
one having remained quiet three and one-fourth minutes, while the
lateral one had been three minutes at rest. The next forward course
required, respectively, but one and one-half and two minutes, the central
particle movmg the more rapidly. The two particles were no longer
side by side, the central one being now a little in advance. The latter
spent after the next turn two and one-half minutes on the under surface,
while the lateral particle spent but two minutes, so that they came up
from the posterior end again at the same time.

The two particles started forward again and had reached the middle
of the upper surface when the Amoeba ceased its forward movement,
loosened its anterior end from the bottom, and became attached by its
posterior end. After five minutes it began to move again, but now in

♦Fig. 41. — Paths of two particles attached to the outer surface of Amoeba,
sphcerottucleolus as described in the text. That portion of the paths which is on
the lower surface is represented by broken lines. (No attempt is made to rep-
resent the forward movement of the Amceba in this fij^re.)

142 THE BEHAVIOR OF LOWER ORGANISMS.

the opposite direction, so that the former posterior end became anterior.
At the same time the two particles reversed their former motion and
began to travel back in the direction from which they had come — that
is, toward the new anterior end. They were observed to make several
complete turns about the Amoeba while moving in this new direction.
I will not, however, add further details, as those above recounted are
sufficient to give a conception of the main features of the movement.

Thus two definite points on the surface of an Amoeba may retain
nearly the same relation to one another for five or six complete revolu-
tions, though their distance apart and their relative position may vary
a little. The reversal of the direction of rotation when the direction of
locomotion is reversed, described in the above case, I have seen many
times.

The direction of the movement of particles on the outer surface is
the same as that of the underlying particles of endosarc. The rate is
also about the same as for the endosarc, though often, or perhaps
usually, the outer particles move a little more slowjy than those in the
endosarc.

It is not merely a thin outer layer that has the rolling movement.
This is demonstrated by the movements of bodies that are partly
embedded in the substance of the Amoeba. For example, a large
Euglena cyst had become attached to the hinder end of an Amoeba
sphceronucleolus . The cyst was carried upward and forward on the
upper surface, and at the same time it began to sink into the protoplasm,
so that when it had reached the anterior edge it was partially embedded.
It was then rolled under, remained at rest on the under surface in the
usual way, and came up at the posterior end. It was now deeply sunk
in the protoplasm, yet it moved forward in the usual way. By the
time it had reached the anterior edge again it no longer protruded
above the surface at all. After turning the anterior edge again it sank
completely into the body, still surrounded by a la3'er of ectosarc, so
that it passed to the interior of the Amoeba as a food Body. I have
repeatedly seen bodies which were thus carried forward on the upper
surface gradually taken in as food. They always continue the forward
movement even when completely embedded in the ectosarc. It is thus
evident that the whole thickness of the ectosarc partakes of the forward
movement. The forward stream in ectosarc and endosarc are one and
continuous.

The relation of the movements of the outer layer to the lines and
wrinkles seen on the upper surface of Amceba verrucosa and its rela-
tives is of interest. There are usually two sets of these wrinkles, one
set diverging from the posterior end toward the direction in which the
animal is moving, the other set forming a number of curved lines

THE MOVEMENTS AND REACTIONS OF AMCEBA. I43

parallel to the advancing edge (Fig. 38). These wrinkles and the
areas which they enclose do not change markedly as the Amoeba
advances, so that the outer surface of the body seems to be quite at
rest. It is this fact, I believe, that has prevented the true nature of
the movement in these species from being recognized before. Thus
Penard (1902, p. 118), after a thorough study, accurate so far as it
goes, of the movements of Amosba verrucosa^ notes that many facts
point to the existence of a permanent contractile outer layer, but holds
that the permanence of certain lines and patterns on the upper surface
in a moving Amoeba is crucial against the idea of a rolling movement
such as I have shown above to actually occur. In reality these
wrinkles are not static structures, but dynamic, /. e.^ the substance of
which they are formed is in continual motion ; they are like the per-
manent ri'pples on the surface of a stream where the latter crosses an
obstruction. The wrinkles indicate the direction of movement of the
substance, the longitudinal wrinkles being parallel to the lines of motion,
the others transverse to them. A particle on the upper surface may
move parallel with the longitudinal wrinkles, at a constant distance
from them, or it may move directly along one of these wrinkles, for
the whole length of the latter. On coming to one of the transverse
wrinkles the particle moves over it with a sort of jerk, as if it had
passed over a ridge or step, as indeed it has. Thus the lines and the
areas enclosed by them remain constant, while the substance of which
they are composed moves onward.

When the Amoeba changes in a marked degree its direction of
movement, so as to follow, for example, a course at right angles to the
previous one, the wrinkles on the surface usually slowly disappear, then
after the movement has become well established in the new direction,
new wrinkles appear in correspondence with the movement.

When such a change of course occurs, any particles on the upper
surface, which were moving toward the anterior edge, change their
course in correspondence with the new direction of progression. Fig.
42 represents a case of this kind, where an Amoeba verrucosa bore on
its upper surface a minute particle of debris {a) and a spherical cyst of
Euglena {b). Both moved forward over the stretch x-y (Fig. 42, A),
Now a little methyl green {iri) was allowed to diffuse against the left
side of the Amoeba. The animal changed its course, moving to the
right At the same time the two objects a and b changed their direction
of movement, traversing the stretch ^-^r (Fig. 42, B) until they reached
the new anterior edge of the Amoeba, and were carried underneath.

The free-moving (upper) surface and the resting (lower) one in
contact with the substratum may exchange roles at any time when
the contact with the substratum is changed. Thus, a specimen was

144

THE BEHAVIOR OF LOWER ORGANISMS.

creeping on the slide and bearing on its upper surface a small granule,
which was moving forward in the usual way. The Amoeba stopped
and raised its anterior edge, which came in contact with the cover
glass ; it then loosened itself entirely from the slide, while its upper
surface became attached to the cover. It now began to move forward
on the cover glass. The granule on the upper surface now remained
quiet, until it was reached by the posterior end, when it passed down-
ward to the lower free surface, there moving forward in the usual way.
Upper and lower surfaces had completely exchanged roles. In a sim-
ilar way I have seen the thin lateral edge of a specimen become the
middle of the upper moving surface.

Objects of the most varied sort cling to the surface of Amoeba verru-
cosa and its relatives. I have seen the following attached to the surface
and showing the typical movements : Particles and masses of soot,
granules of India ink, motionless bacteria, diatom shells, dead flagel-
lates, masses of debris,
cysts of Euglena, a small
Amoeba proteus (the lat-
ter was inclosed after it
had passed to the under
surface). Usually only
one or two small objects
are seen attached to any
given specimen, but to
this extent the phenomenon is very common, so that it seems rather
surprising that the movements of such particles should not have been
described before.

A number of other points must be set forth before we can form a
clear conception of the movements of these Amoebae. The species
under consideration are much flattened and have usually an oval form
as they move forward, the anterior-posterior axis being the longer,
while the posterior end is the more pointed (Fig. 38). Not the Whole
lower surface is in contact with the substratum, but only a band at the
anterior and lateral margins. In an Amoeba that was creeping on the

Fig. 42.*

* Fig. 42. — Movement of bodies attached to the surface in Amoeba verrucosa^
when the direction of locomotion is changed, a, A small granule ; b, a Euglena
cyst. In A the Amoeba is progressing to the right, as shown by the large arrow ;
the two bodies attached to the surface moved in the same direction, traversing
the stretch x-y^ as shown by the small arrows. At this point a solution of
methyl green {vt) was allowed to diffuse against the surface; the Amoeba there-
upon changed its course, as indicated by the large arrow oi B. At the same
time the bodies a and b changed their course, traversing the stretch y-z. The
stretch x-y-z in B shows thus the path of the attached bodies before and after
the reaction.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 1 45

lower surface of the cover glass I was able to define with some accuracy
the parts that were attached and those that were not. A small flagel-
late was moving briskly about between the Amoeba and the cover glass,
but its excursions were limited by a visible line running parallel with
the anterior edge of the Amoeba and extending at the sides back to
about one-third the animal's length from the rear (Fig. 43, a-a-a) .
The zone between this and the margin was pressed close to the glass,
and was evidently attached to it. The more pointed posterior end was
held quite away from the glass, leaving a broad passageway through
which the flagellate finally escaped.

The results of this observation were confirmed by another. An
Amceba verrucosa in full career was suddenly turned on one lateral
edge by a strong current from a rotifer, and its upper edge coming in
contact with the cover glass, it remained in that position some time
without change of form. It could be seen that the under surface was
concave, the edges very thin and flat, while the
posterior portion was thick and arched (Fig. 44).

It is clearly at the advancing edge of the ani-
mal that the most active movements are taking
place. Here the hyaloplasm may be seen to
push forward in a series of short waves, the
anterior edge of each becoming attached to the
substratum. At the same time, of course, an
equivalent amount of protoplasm becomes de-
tached from the substratum along the line a~a-a^ ^ --^ss^?-^
Fig. 43, though this does not take place in
waves, so far as observable. The anterior wave must in some way
pull upon the upper surface of the Amoeba, bringing it forward, and
dragging with it the elevated sac-like posterior end. A certain feature
of the advance of the anterior edge seems of much significance. Each
wave seems to arise just behind the previous anterior boundary line
and overlaps it, leaving it buried. This line often remains visible for
a short time after the new wave has been formed. The new wave
rolls over the preceding one in such a way that its original upper
surface becomes applied to the substratum. This is demonstrated by
the rolling under of small objects on the upper surface of the advanc-
ing wave. A diagram of the movement at the anterior edge is given
in Fig. 45. The movement can be imitated roughly by making a
cylinder of cloth, laying it flat on a plane surface, and pulling forward

♦ Fig. 43. — Attached surface oi Amoeba verrucosa^ creeping on the lower surface
of the cover glass. The unshaded portion in front of the line a-a-a is attached
to the substratum, while the shaded portion is free and raised slightly above
the substratum.

146 THK BEHAVIOR OF I.OWER ORGANISMS.

the anterior edge in a series of waves. The entire cylinder then rolls
forward just as the Amoeba does.

The essential features of the movement seem to be (i) the advance of
the wave from the upper surface at the anterior edge ; (2) the pull exer-
cised by this wave on the remainder of the upper surface of the body,
bringing it forward. Most of the other phenomena follow as conse-
quences of these two. The flowing forward of the granules of the
endosarc seems to demand no special explanation, since a fluid con-
taining granules within a rolling sac must necessarily flow forward as
the sac rolls. By the movement forward of the anterior end a space is
left free ; by the rolling forward of the posterior end the fluid is piled
up and pressed upon, and must flow forward into the empty space in
front. Possibly there may be other causes at work in producing the
endosarcal currents, but such currents would be produced without
other cause in a sac moving as Amoeba does.

3:4

a o

Fig. 44.* Fig. 45.!

OTHER SPECIES OF AMGBBA.

Thus far we have dealt only with Amoebae of rather constant form,
which do not produce pseudopodia, or only rarely do so. We must now
take up species in which the form is changeable and the movements
varied. Of such species I have studied chiefly Amoeba limax^ A.
proteus^ and a smaller Amoeba, which I take to be Amoeba angulata
Meresch. In these species the outer surface is not viscid, except at the
posterior end, so that small objects rarely cling to it. It is, therefore,
much more difficult to determine the direction of movement of the
upper surface than in Afjioeba verrucosa and its relatives. Yet, by
mixing soot with the water, and devoting a sufficient amount of time
and patience to the work, one can obtain as many observations as he
desires. The soot settles upon the upper surface in particles or masses

*FiG. 44. — Side view (partly an optical section) of a creeping Amceba verru-
cosa, showing the thin anterior edge {A) attached to the substratum, and the
high posterior portion (/*) with a cavity beneath it.

t Fig. 45. — Diagram of the movement at the anterior edge of Amceba verrucosa.
The region b-c pushes out, taking up the position b'-c\ and pulling forward the
region c-d, so that it comes to occupy the position c'-d'. The point a remains
in its place.

THE MOVEMENTS AND REACTIONS OF AMCEBA. I47

and its movements can be followed ; at times, also, objects actually
cling to the surface, as in the other species.

The results are essentially the same as in the species already described ;
foreign particles resting upon or clinging to the upper surface are car-
ried forward to the anterior edge. Here they roll over the edge, passing
beneath the Amceba, which now moves across them. As a rule in
these species particles do not cling to the surface after passing to the
lower side, so that they are left behind when the posterior end passes
over them. Sometimes they do thus cling, however, and in such cases
I have seen them pass upward at the posterior end and again forward,
exactly as in ^. verrucosa and its relatives. In order that my state-
ments may not remain abstract and general, I copy a few observations
from my notebook, all relating to Amoeba proteus.

1 . A large particle of debris with bits of soot attached to it was seen
lying on the upper surface just behind the middle. It was carried
forward to the anterior end and over the edge. Then it came to rest
on the bottom, and the Amoeba crept over it till it was passed by the
posterior end and left behind.

2. A number of soot particles on the upper surface just in front of
the middle were carried forward, changing their direction as the proto-
plasmic currents beneath them changed direction. They were finally
carried over the anterior edge.

3. A small mass of soot was lying on the middle of the upper surface.
It moved forward in the same way as the endosarcal granules under-
neath. The latter changed their direction of movement several times ;
the soot mass changed correspondingly at the same time. It was
finally carried over the anterior edge, where it could be seen clearly
separate from the Amoeba.

4. A large mass of soot one-quarter the size of the Amoeba was
carried forward on the upper surface for a distance, but fell off' at the
side before reaching the anterior end.

5. Two small masses of soot lying on the upper surface of the
posterior end were carried forward over the anterior edge.

6. Several small particles were clinging to the lower surface of the
posterior end. They passed upward, one of them around the very
middle of the posterior end, to the upper surface ; here they were
carried forward and over the anterior edge.

I could add a large number of such observations.

On the under surface the particles are quiet, as I have shown before
(p. 136). At the lateral margins the edges of this quiet lower surface
are seen, so that particles situated here are usually likewise quiet, until
they have reached the posterior part of the Amoeba (see p. 135).

As to the details of the movement of the upper surface, the following

148 THE BEHAVIOR OF LOWER ORGANISMS.

are important. Particles situated on the upper surface move usually
at the same, or nearly the same, rate as the granules beneath them, in
the endosarc. The movement of the surface particles follows exactly
that of the endosarc beneath them, changing in direction when the latter
changes. Two particles close together on the upper surface may thus
diverge or even flow in opposite directions, carried by two different
currents which are visible in the endosarc. Any portion of the ecto-
sarc, like any portion of the interior, may stop at any time, while other
parts flow onward. One may thus see at times a particle at rest on the
upper surface of a moving Amoeba. Isolated observations of this kind
might lead one to suppose that the upper surface, like the lower, remains
at rest while the endosarc passes forward. "But when a particle on the
surface is at rest, one will usually find, by a proper change of focus,
that the endosarc beneath it is likewise at rest. It is, of course, well
known that certain portions of the endosarc may be at rest while the
remainder is in movement (see Rhumbler, 1898, p. 122). In the same
way a portion of the outer layer may sometimes be at rest while the
adjacent endosarc is in motion ; this, however, is rather unusual.

We may sum up our results thus far in the following statements :
In an advancing Amoeba substance flows forward on the upper sur-
face^ rolls over at the anterior edge^ coming in contact with the
substratum^ then remains quiet until the body of the Amoeba has
passed over it. It then moves 7ipward at the posterior end., and
forward again on the upper surface., continuing in rotation as long
as the Amoeba continues to progress. The motion of the upper
surface is congruent with that of the endosarc^ the two forming a
single stream.

HISTORICAL, ON ROLLING MOVEMENTS IN AMCEBA.

The possibility that Amceba progresses by a rolling movement was
discussed by Claparede & Lachmann (1858). In Amoeba Umax and
Amoeba quadrilineata (= A. verrucosa) ., according to these authors,
the general appearance of locomotion is in many respects in favor of
this view : " On croit positivement voir I'animal rouler sur lui-meme "
(p. 435). But this correct view is rejected (in the text) because of the
(supposed) permanence in the position of the contractile vacuole.
Claparede & Lachmann insist that the contractile vacuole is situated
in the ectosarc ; hence, they argue, if there were a rolling movement
of the ectosarc, the vacuole would necessarily partake of the movement !
In a note on p. 437 it is stated, however, that Lachmann personally
believed the motion to be of this rolling character. " II croit s'etre
assure que I' A. quadrilineata roule sur elle-meme." According to
Claparede & Lachmann, Perty held this view also.

TtlE AfOVEXfENTS And RHACtlONS Oi^^ AMCEBA. 149

Dr. Wallich shared the correct opinion of Lachmann and Perty.
This excellent observer unfortunately often gave his results in the form
of mere brief general statements, so that one cannot judge how much
evidence he had for them, and little attention has, therefore, been paid
them. But it is singular how many of these statements show them-
selves to be correct, even in opposition to later work. Concerning
the matter in question, Wallich has the following :

In short the effect is similar to that which would be produced were an empty
and transparent bladder or caoutchouc sac, containing granular bodies of greater
specific gravity than the viscid fluid within which they were sustained, to be
rolled along a plain surface. (Wallich, 1863, 3, p. 331.)

He makes no attempt to demonstrate the truth of this correct com-
parison, and does not develop the matter beyond the mere statement
given above.

Schulze (1875), Berthold (i8S6), and Butschli (1892), as we have
seen, agree in stating that the currents on the upper surface at the
anterior end in Pelomyxa are backward. In view of the great authority
of these writers we should be compelled to suppose that the movement
in this animal is of an entirely different character from that found in
the various species of Amoeba, but for a most fortunate circumstance.
The only previous demonstrated observation of the forward movement
of the upper surface in the Rhizopoda relates precisely to Pelomyxa,
and was made by an investigator closely associated with Biitschli.
It was not until my work was finished and the present paper written
that I came across the note of Blochmann (1S94) on the movements of
Pelomyxa. Blochmann shows that the movement of substance on the
upper surface of Pelomyxa is forward, just as we have found it to be
in Amoeba. The outer surface of Pelomyxa is covered with fine cilia-
like projections. By observing these projections Blochmann had no
diflSculty in seeing that they move forward on the upper surface. The
rate of movement was the same as that of the internal forward current.
Biitschli (1892, Appendix, p. 220) had already observed, greatly to his
surprise, that there is a forward current in the water next to the sur-
face of an advancing Pelomyxa, this current being exactly the reverse
of that called for by the theory that the motion is due to a lowering of
the surface tension at the anterior end.

Biitschli (/. c.) attempted to save the surface tension theory by sug-
gesting that it was only a thin outer layer that moves forward. The
currents in the moving animal would then be as follows : A forward
current within, a backward current just beneath the surface, a forward
current in a thin layer on the surface. It is possible that this compli-
cated arrangement of currents might be brought into harmony in some
way with the surface-tension theory of the movement, though it is

150 THE BEHAVIOR OF LOWER ORGANISMS.

rather difficult to see how the forward movement of the outer layer
would be produced. As we have seen, Blochmann found that the
internal and external forward currents move at the same rate and in
the same direction. It is difficult to explain how this should occur if
the two are separated by a layer moving in the opposite direction.
But Blochmann accepted Biitschli's suggestion, and attempted to give
some evidence in its favor. He says that one sees the outer current,
with the movement of the projections, in places where the marginal
current has come to rest, and that the outer and internal currents then
move at the same rate, separated by the resting marginal layer. Now
one can receive exactly this impression in Amceba in the following
manner : The margins where they are pressed against the substratum
are at rest. Just above this region, and often visible in the same focus,
there is the forward current, which is visible on the one hand on the
surface (through the movements of the projections) ; on the other hand,
in the interior (through the movements of the granules of the endosarc).
Unless one is on his guard as to the slight difference in level, one might
seem to see two currents separated by a resting layer, particularly if
the probability that this were true had been suggested beforehand. It
is notable that Blochmann describes nowhere outer and inner forward
currents, separated by a marginal backward current, as would be
required by the modified surface-tension theory.

We have demonstrated above, for Amoeba at least, that the forward
movement is not confined to a thin outer layer, but extends from the
outer surface to the endosarc (p. 142) ; in other words, that the outer
surface moves in continuity with the internal substance.

Rhumbler (1898, pp. 126-130) discussed at length the possibility of
explaining the movement of Amoeba by means of a rolling sac of
ectoplasm, only to come to the conclusion that it was impossible.
Rhumbler's discussion of this matter is an excellent example of the
fact that acumen and excellent reasoning may lead one astray in scien-
tific matters when the observational basis for the reasoning is not
secure. What chiefly misled him was an incorrect idea as to the
direction of the currents in the substance of Amoeba, particularly his
assumption that there is a backward current on the upper surface.
The diagram which he gives of the currents as they must occur in an
Amoeba moving in a rolling manner (Fig. 33, A, p. 133) is, therefore,
much more nearly correct than that in which he shows what he con-
siders the really existing currents (Fig. 33, B).

Rhumbler's conception as to the necessary movements in the sub-
stance of an Amoeba progressing by rotation is, however, incorrect in
one particular, so that his diagram (Fig. 33, A) does not correspond
to the facts in this point. He assumes that there must be a backward

THE MOVEMENTS AND REACTIONS OF AMCEBA. I5I

current on the lower surface, as indicated by the lower (heavier) arrows
in his diagram (Fig. 33, A). This backward current does not exist,
and is theoretically unnecessary, as may be seen by making a cylinder
of cloth and moving it in the manner described above (p. 145). The
under surface remains at rest until it passes upward at the posterior
end {cf. Fig. 40). Rhumbler held that this backward current below,
with the forward current above (Fig. 33, A)^ must set the endosarc
in rotation ; " the endoplasma granules would themselves necessarily
all move, like the particles of the ectosarc, in circular or elliptical
courses" (p. izS). The absence of such circular or elliptical paths for
the granules of the endosarc would then speak against the method of
movement by rotation of the ectoplasm. But since there is no such
backward current as Rhumbler assumes, and not even the particles of
the ectosarc move in circular or elliptical courses, this objection falls
to the ground.

Further, Rhumbler seems to assume that for locomotion by a rotary
movement of the ectosarc, the latter must necessarily be a "sharply
defined persistent organ," and that its contractions could only be due
to preformed, permanent fibers, in a definite arrangement. Rhumbler
is able to show of course that these two assumptions are probably incor-
rect, and considers that this weighs against the possibility of movement
in the manner characterized. But both these assumptions are unneces-
sary. The rotation demonstrably does occur, yet the permanent, sharply
defined ectosarc with definitely arranged persistent fibers does not exist,
as Rhumbler has set forth, and as must be evident to anyone who
studies for a long time the changes of form and movement in Amoeba.
As we shall see later, a simple drop of fluid, with no differentiated outer
layer, may move in the same manner.

Penard (1902) also discusses the possibility of movement by rotation
of the ectosarc in Amoeba verrucosa (= A. terricold). His study of
the movements is excellent and he gives as a possibility on p. 115 what
is really in its main features a nearly accurate statement of the method
in which locomotion actually occurs, only to reject this possibility later.
The ground for this rejection is as follows : The posterior end of the
Amoeba often bears an irregular saclike projection (what Penard calls
the " houppe" ) ; this may be much wrinkled or covered with projec-
tions. This wrinkled sac retains its position ; in Amoeba verrucosa it
is covered, like the rest of the body, with a resistant cuticula, which can
be dissolved only with great difficulty and very slowly.

If the Amoeba rolled on itself in progressing, the posterior part of this mem-
brane would necessarily follow the movement and pass little by little forward,
which is contrary to the facts. The best manner of assuring one's self of the
immobility of the pellicle is to look very attentively at the surfiice of the con-

152

THE BEHAVIOR OF LOVv'ER ORGANISMS.

tractile vacuole; there one sees almost always very fine folds, forming angles
and varied patterns; these angles and these patterns remain for a long time
absolutely the same, which shows that nothing has changed place. (Penard,
1902, p. 118.)

In all the specimens oi Amoeba verrucosa and A. sphceronucleolus
in which I have studied the matter, the posterior part of the outer mem-
brane does follow the movement. Particles clinoring to the outer surface
of the hinder part of the ectosarc pass upward over the wrinkled sac-
like posterior end and forward on the upper surface. In so doing they
pass directly across the wrinkles on the body sur-
face, as set forth on p. 143. Had Penard chanced to
^ - ^ see the movements of a particle attached to the outer

surface of the body he could not have been misled
by the apparent permanence of the surface wrinkles .

FORMATION AND RETRACTION OF PSEUDOPODIA.

Thus far the phenomena in Amoeba proteus
and its relatives are essentially like those found
in Amoeba verrucosa. At times Atnoeba proteus
flows forward as a single simple mass ; then its
locomotion may be compared directly in its chief
features to that of A?nceba verrucosa. But in
Amoeba .proteus and its relatives the movement
is, of course, usually much complicated by the for-
mation of pseudopodia. In considering the way
in which these are formed we must deal separately
with two different cases, depending on whether the
pseudopodium when sent out is or is not in contact
with the substratum.

SURFACK CURRENTS IN THE FORMATION OF PSEUDOPODIA
IN CONTACT WITH THE SUBSTRATUM.

Fig. 46.* When the pseudopodium is sent out in contact

with the substratum, the phenomena accompany-
ing its formation are essentially the same as those which take place at
the anterior end of an advancing Amoeba ; the latter may indeed be
considered as merely a large pseudopodium. Even when the pseudo-

♦FiG. 46.— Movement of a particle attached to the outer surface of a pseudo-
podium that is extending in contact with the substratum. At a the particle is
at the middle of the upper surface; at b it has nearly reached the tip. When
the pseudopodium has reached the length shown at c the particle has passed
over its tip. Here it remains, so that at d, when the pseudopodium has become
longer, the particle is still at the same level, but on the under surface of the
pseudopodium, some distance behind the tip.

THE MOVEMENTS AND REACTIONS OF AMCEBA.

53

podia are slender and pointed, the protoplasnn flows forward (toward
the point) on the free upper surface and in the interior, while on the
side which is in contact with the substratum the protoplasm is at rest.
I have often seen small particles which had been brought forward on
the surface of the Amoeba carried out to the tip of a pseudopodium on
its upper surface, finally rolling over the point and becoming covered
by the advancing protoplasm (Fig. 46).

FORMATION OF FREE PSEUDOPODIA.

When a pseudopodium is sent out directly into the water, so that its
surface is free on all sides, it is much more difficult to determine the
nature of the movement. Particles rarely cling to the surface of such
a pseudopodium, and without this aid one cannot be certain what the
movement of the outer layer is. However, by devoting several entire
days under most favorable conditions to the determination of this point,
I collected a number of observations which demonstrate clearly the
nature of the move-
ment. The point \ a "b ^
of special interest
is, here as else-
whe re , whether
there is a back-
ward current on
the surface of the
advancing pseudo-
podium, as repre-
sented in the dia-

FiG. 47.^

gram from Rhumbler, Fig. 31. To this the observations give a negative
answer. Particles clinging to the surface of a pseudopodium, whether
at the tip or at the base or at any intermediate point, are uniformly
carried outward, in the same direction as the tip. Particles situated at
a certain distance from the tip of a short pseudopodium maintain the
same distance as a rule when the pseudopodium is lengthened, though
in so doing they are carried far out from the body. Sometimes the tip
moves outward a little faster than the parts behind it, the pseudopodium
thus becoming more slender as it extends, but all parts agree in being
carried outward. A number of examples of actual observations will
make this point clear.

I. Amoeba angulata : When first observed there was a short pseudo-
podium in front, projecting freely into the water. A small particle was
attached to the surface at about the middle of its length (Fig. 47, a).

* Fig. 47. — Successive stages in the formation of a free pseudopodium, showing
the movement of a particle attached to its surface. The particle moves outward,
keeping at approximately the same distance from the tip.

154 THE BEHAVIOR OF LOWER ORGANISMS.

The pseudopodium lengthened, carrying the particle with it, the latter
maintaining its distance from the tip nearly or quite constant, but being
carried far from the body (Fig. 47, b). The pseudopodium finally became
very long and slender (c), the particle remaining attached near the tip.

2. Amoeba proteus : A particle clinging to the surface of one side,
near the anterior end. A pseudopodium was formed at exactly this
point, extending freely into the water, so that the particle was borne
on the tip of the pseudopodium. It maintained this position while the
pseudopodium was extending, and was still found at the tip after the
pseudopodium had become long and slender (Fig. 48).

The third example which I give is one of much interest, because it
shows the movements of a given point on the surface in both the retrac-
tion and extension of pseudopodia, as well as in transference from the
posterior to the anterior region of the body.

3. Afnceba proteus: When first observed the animal was rather

slender, creeping in a certain direc-
tion, and with two long pseudo-
podia at the posterior end, extend-
ing, one on each side, at right
angles to the axis of progression
(Fig. 49, a). The left pseudo-
podium was the longer, and bore
at about one-fourth its length from
its base a small particle {x) at-

FiG. 48.* tached to its surface by a very

short stalk in such a way that it
was seen in profile (Fig. 49, a). The pseudopodium was not in
contact with the bottom, and was slowly retracting, its internal con-
tents flowing into the body, while the pseudopodium itself shortened.
As this occurred the particle approached the body and finally passed
on to its surface (<$, c) while the pseudopodium was yet of considerable
length. It was evident that the shortening of the pseudopodium took
place chiefly at its base, since the part between the base and the particle
X had become incorporated with the body when the portion between a;
and the tip had changed only a little in length (3, c) . This distal
portion apparently did become somewhat shorter at the same time,
while its surface became slightly wrinkled. By the time the tip of the
pseudopodium had united with the body {d) the particle x had moved
a considerable distance forward on the latter. The posterior portion
of the body was here thick, and the particle was still seen in profile,
though it was some distance above the substratum. It now moved

♦ Fig. 48. — Movement of a particle attached to surface of an Amoeba at point
where a free pseudopodium is pushed forth. The particle remains at the tip.

THE MOVEMENTS AND REACTIONS OF AMCEBA.

155

forward very slowly (c, d^ e) till at f it passed to the upper surface.
It then moved rapidly forward, occupying successively the positions
indicated by the line of circles in g. (The Amoeba itself was, of course,
progressing ; no attempt is made in the diagram to represent its change
of position.) Finally the particle x had nearly reached the anterior
end, when the latter forked, sending two pseudopodia upward and
forward into the water (^, h). The particle x was at first at the base
of the right pseudopodium. This was now projected forward as a very
long, slender pseudopodium bearing the particle x. The latter was
carried steadily out from the body, maintaining almost exactly its
original distance from the tip of the pseudopodium (Z^, /, 7). It is

Fig. 49.*

possible that as the tip became very slender its distance from x became
slightly greater as if, by a circular contraction of the intervening part,
the tip were forced further out ; but there was no movement backward
o{ x\ on the contrary, it moved steadily forward, its distance from the
base of the pseudopodium continually increasing. Unfortunately at
this point the animal passed under a mass of debris, so that I was
unable to trace further the history of that point on the body surface
marked by the particle x.

I have, altogether, about a dozen observations showing this outward
movement of particles on the surface of free pseudopodia. The three

♦ Fig 49. — Movements of a particle (a?) attached to the surface of Amoeba in
passing from a pseudopodium at the posterior end over the body to a pseudopo-
dium at the anterior end. For explanation see text.

156 TMK Behavior of lower orc^anIsMS.

examples above given are typical for all. They show^ the following as
to the manner in which the pseudopodia are formed when they are
projected freely into the water.

1. The pseudopodium grows in length chiefly from the base, so that
any part on the surface retains nearly its original distance from the tip.

2. The increase in surface as the pseudopodium grows is not pro-
duced by the flowing outward and backward of the endosarc at the tip
with its transformation into ectosarc (as represented by Fig. 31), but
by the transference of a portion of the surface layer of the body to the
pseudopodium. The same substance remains at the tip of the pseu-
dopodium from the beginning (observation 2, p. 154 ; I have other
observations showing the same thing).

3. Thus the movement of the free pseudopodium is like that of the
pseudopodium in contact with a surface, save that in the latter case one
side is held back by attachment to the substratum. In the free pseudo-
podium all sides move outward ; in the attached one, all sides but one.

The outer layer of the body in its transference to the pseudopodium
may doubtless become thicker or thinner or be otherwise modified. As
will be shown later, I am not at all inclined to deny the possibility of
the transformation of endosarc into ectosarc, and vice versa. The
observations show, however, that this transformation of substance does
not, as a rule, take place in pseudopodia by means of the ''fountain
currents" represented in the diagrams from Rhumbler (Figs. 30-32).

Further, the surface of the pseudopodium may be increased by the
flowing into it of the endosarc, producing a sort of stretching of the
outer layer, involving, of course, the appearance at the surface of por-
tions of substance which were before covered.

WITHDRAWAL OF PSEUDOPODIA.

In the withdrawal of pseudopodia the process is the reverse of that
occurring in the formation of pseudopodia, as is shown in case 3, above
(p. 154, Fig. 49). The basal parts of the pseudopodial surface first
pass on to the body, followed by the distal portions.

The withdrawal of pseudopodia shows certain other features that are
of importance for the understanding of the mechanism of amceboid
movement. The process differs somewhat in different cases, depending
on whether the withdrawal is slow or rapid. When the pseudopodium is
slowly withdrawn, its surface may remain perfectly smooth, the decrease
in surface keeping pace with the decrease in volume, until the pseudo-
podium has quite disappeared. But when the withdrawal is more rapid
the surface becomes thrown into folds or warty prominences of various
sorts. This is more common than retraction without the formation of
such prominences. Evidently the volume decreases so fast that the

THE MOVEMENTS AND REACTIONS OF AMCEBA. I57

decrease in surface cannot keep pace with it^ so that the surface is
thrown into folds. This phenomenon is particularly interesting in its
bearing on the theory that would account for the retraction of pseudo-
podia by the action of surface tension. On this theory we should
naturally expect the surface to remain smooth, and by no means to be
thrown into folds, since it is by the tendency of the surface to decrease
that the decrease in volume is accounted for ; the decrease in volume
should not, therefore, precede the decrease in surface. This matter will
be taken up later.

As the pseudopodium decreases in size, the fluid endosarc, of course,
flows out of it and joins the endosarc of the body. The backward
current begins at the mouth or inner end of the pseudopodium, and
gradually extends backward to near the tip ; the current is most rapid
in the central axis of the pseudopodium, and in this axis it is most
rapid at the inner end.*

Where is the impelling force in the outflow of the endosarc and the
decrease in size of the pseudopodium ? The observations seem to sug-
gest several factors here. The fact that the ectosarc of the pseudo-
podium passes on to the body when the pseudopodium shortens, as is
shown in Fig. 49, «, b^ c, indicates that the ectosarc of the body exercises
a pull on the outer layer of the pseudopodium, drawing it inward.
This would, of course, force the fluid endosarc into the body. But this
would not account for the wrinkling and roughening of the outer surface
of the pseudopodium, which is so prominent a feature in the withdrawal.
For this there are two conceivable causes, (i) The ectosarc itself may
contract actively, driving out the endosarc. If the real contractile por-
tion of the ectosarc is not on the outer surface (in the cuticula, as it has
sometimes been called), but in a deeper layer, then the outer surface
would be thrown into folds or prominences as contraction occurs. (2) On
the other hand, it is conceivable that the endorsarc might be drawn out
of the pseudopodium, the latter collapsing and becoming wrinkled as
a result. This is the explanation given by Biitschli (1S92, p. 201).
This view would have to assume some force pulling on the endosarc at
the mouth of the pseudopodium, and sufficient viscosity in the endosarc
so that a pull thus exercised would draw out the whole mass contained
within the pseudopodium. Thus, in Fig. 49, a, the general advancing
current within the bod}^ of the Amoeba might be thought to exercise a
pull at the point jK in the direction of the arrow ; if the endosarc were

♦This account differs from that given b_y Biitschli (1880, p. 116), according to
whom the withdrawal of the pseudopodium begins at the tip. The observations
present no difficulty, and I am unable to understand how Biitschli came to this
result. In a large pseudopodium the method of retraction described above is
evident.

158

THE BEHAVIOR OF LOWER ORGANISMS.

sufficient!}' viscous the entire mass of endosarc would be withdrawn,
and the pseudopodium would collapse.

There are certain facts that speak against this second view. Thus,
the endosarc often passes out when there is no current away from the
mouth of the pseudopodium, so that there can be nothing pulling upon
the endosarc. A pseudopodium may be withdrawn when the animal
is otherwise quiet ; or, when the animal is stimulated strongly, all the
pseudopodia may be withdrawn at the same time, while there is no
endosarcal current in the body of the animal. A striking case that
belongs here is sometimes to be observed in Amoeba radiosa. This
animal frequently floats in the water, with many long, pointed pseudopo-
dia radiating in all directions from the body. Now, if the pseudopodia

are stimulated with a rod,
they begin to contract. The
endosarc first passes inward,
but the resistance of the body
is so great that the fluid stops
at the base of the pseudo-
podia. These, therefore,
swell up in a bulbous
fashion, as illustrated in
Fig. 50. Such cases, indeed
all the numerous cases in
which the endosarc passes
out of a pseudopodium and
comes to rest as soon as it has
left the latter, can only be ex-
plained on the assumption
that the endosarc is forced
out by the contraction of the ectosarc, or by some active movements of
the endosarc itself, of a character not understood.

Further, there are certain facts which speak positively in favor of
the view that the production of the wrinkles is due to a contraction of the
inner layer of the ectosarc. Thus, when an Amoeba is strongly stimu-
lated and withdraws all its pseudopodia quickly, the whole surface of
the body becomes rough and wrinkled. The endosarc has not passed
out of it, so that it cannot be considered in a state of collapse ; on the con-
trary, it is clearly contracted as strongly as possible. Again, if a large
pseudopodium is cut from the body, it contracts strongly, showing the
rough, wrinkled contour, though the endosarc has not passed out of it.

Fig. 50.*

*FiG. 50. — Specimen o{ Amoeba radiosa in which the endosarc has passed out
of the distal portions of the pseudopodia into the basal parts, causing them to
swell up in a bulbous fashion.

THE MOVEMENTS AND REACTIONS OF AMCEBA.

159

The processes occurring in a retracting pseudopodium are the same
as those taking place at the posterior end in a moving animal. In fact,
the cases are really identical ; the posterior portion of the body may be
considered merely a large retracting pseudopodium. Often, as we shall
see, it is impossible, from its method of formation, to consider it any-
thing else.

The pseudopodia are, of course, usually formed in the anterior part
of the Amoeba, in contact with the substratum, and are directed in the
line of progression, or
at a slight angle with
the line of progression.
As they are withdrawn,
their surface, as we have
seen, usually becomes
folded or roughened,
and the small roughened
project ion resulting
from the withdrawal
lasts, as a rule, for a long
time. As the Amoeba
continues its forward
course, the base of the
retracting pseudopo-
dium retains nearly its
original position, as do
the other parts of that
layer of the ectosarc that
is against the substra-
tum, so that the body
of the Amoeba gradually
passes the pseudopo-
dium, and the latter
finally becomes united with the posterior end. During this transference to
the rear, the pseudopodium usually changes its position (Fig. 51 ). At
first it is directed nearly forward (a) , then it takes a position at right
angles to the body {^), and finally swings around with its point directed

Fig. 5

* Fig. 51. — Successive stages in the retraction of a pseudopodium. At a the
pseudopodium extends forward at the anterior edge ; at 3 it has partly withdrawn
and stands at right angles to the body, which has partly passed it; at c the
pseudopodium is directed backward, and is in the posterior part of the body; at
d the small roughened remnant of the pseudopodium has nearly united with the
tail. At X the successive positions occupied by the withdrawing pseudopodium
are shown in a single diagram.

l6o THE BEHAVIOR OF LOWER ORGANISMS.

nearly backward (Fig. 51, c). This change of position is clue to the
contraction of the posterior part of the AmcEba. The ectosarc just
behind the base of the pseudopodium contracts toward the middle, as
described on page 171. As a result the pseudopodium must swing
around till it points nearly backward. The mechanism of the process
will be best understood by an examination of Fig. 51, a*.

Finally, what remains of the pseudopodium reaches the posterior
end or tail (Fig. 51, d). By this time usually all that is left of it is a
small roughened projection, its surface being of essentially the same
character as that of the tail. This projection fuses completely with the
tail, its projections taking up a portion of the surface of the latter. The
tail is in fact nothing but the fused remnants of all the pseudopodia
that have been formed, together with the contracted outer layer of the
body of the Amoeba (the latter cannot be distinguished in any essential
way from a pseudopodium). A roughened tail is formed de novo when-
ever an Amoeba suddenly changes its direction of movement. The
previous anterior end then becomes roughened in contracting and forms
a typical tail. This latter unites with the old tail if any of the latter
remains. The substance of the tail gradually passes forward into the
rest of the body, as we have seen.

MOVEMENTS AT THE ANTERIOR EDGE.

As in Amoeba verrucosa and its relatives, so in the species of more
changeable form, the most active movements are taking place at the
anterior edge. In Amceba proteus ?ind A. Umax oxiq sees still more
distinctly than in the species before named the pushing forward of a
series of waves of hyaloplasm which become attached to the substratum
in front. In Amoeba Umax and its relatives especially such a wave
may be very pronounced, extending forward at times one-fifth the
length of the body or more, though usually much less.

At first the advancing wave, as it moves rapidly forward, is usually
free from granules, and may be spoken of, therefore, as hyaloplasm.
If the motion is less rapid, however, it contains granules, and is not
distinguishable in any way from the interior endosarc. Where it is at
first free from granules, it is nevertheless highly fluid in character, as
is shown by the fact that it flows and spreads out swiftly, and that the
granules of the endosarc pass into it rapidly. The freedom of the
advancing hyaloplasm from granules is not due to its greater density
or solidity as a result of the action of water upon it, as has sometimes
been maintained, for it is at first free from granules ; then, after the
water has acted longer upon it, it becomes filled. Apparently the
reason for its freedom from granules is merely the fact that it moves
forward faster than the granules and leaves them behind. This view

THE MOVEMENTS AND REACTIONS OP' AMCEBA. IDI

is supported by the variations to be observed in the relation of the clear
substance to the granular substance at the anterior end. These varia-
tions arise chiefly from the different rates of movement of the two
substances, and may be summarized as follows :

I . The clear substance moves fastest at first, and therefore becomes
separated from the granules as a broad band in front (Fig. 52, a); this
is then immediately filled completely by the granules. Even large
granules or vacuoles pass to the very anterior edge, so that one sees
but a line between these and the outer water.

•i4-*--t;«:v'?:f>.»-»'-.o', 7/

Fig. 52.*

2. The clear substance advances fastest, and so continues to do, so that
it remains a long time as a broad, clear band in front of the granules.

3. The two substances advance at the same rate, so that there is no
separation between them. The granules and vacuoles are then found at

* Fig. 52. — Distribution of hyaloplasm and granules at the anterior end in
Amceba Umax and its relatives : a, hyaloplasm without granules at the anterior
end ; b, granules and vacuole at the anterior edge ; c and d, two successive
instants in the locomotion ; at c the anterior half of the body is free from gran-
ules, the latter being heaped up behind a well-marked barrier; at d the barrier
has burst at a certain point {x), allowing a stream of granules to flow forward to
the anterior edge; e and/", two successive stages at the advancing anterior end;
in e the clear hyaloplasm has stopped at the line x-y, at/" the hyaloplasm has
advanced, while the granules are heaped up behind the same line x-y.

l62 THE BEHAVIOR OF LOWER ORGANISMS.

the advancing edge. This condition is not at all rare. In such cases
there is at the anterior end less clear space between the granular region
and the water than in any other part of the body. In fact, there is
typically no space at all. I have seen large vacuoles come so close to
the anterior edge at such times that it was not possible to distinguish
between the boundary of the vacuole and the boundary of the Amoeba
(Fig. 52, ^).

4. Finally, either hyaloplasm or endosarc or both may stop in any
of the positions mentioned above. Thus the hyaloplasm may stop,
whereupon the endosarc flows into it and fills it, or both may stop, so
that the hyaloplasm remains empty, as a clear band, for a long time.

The line separating hyaloplasm and endosarc is at times very sharply
defined, as has often been pointed out. A number of unusually favorable
specimens gave me the opportunity of determining the reason for this,
in many cases at least. The Amoeba in question (Fig. 52, c-f) was an
elongated, rapidly moving form, much resembling A, limax^ but having
usually two contractile vacuoles, one very large, in the fore part of the
body, the other smaller and in the rear. The body contained many
fine granules, which, when the animal was at rest, were scattered almost
uniformly through the body; the peripheral, more solid zone (usually
called ectosarc) contained as many of these as did the endosarc.

In moving this Amoeba usually sends out first a large amount of clear
fluid containing no granules ; this at times extends so far as to constitute
half the length of the Amoeba (Fig. 52, c). There is a perfectly sharp
line between the clear hyaloplasm and the granular endosarc, and behind
this line the granules of the endosarc are heaped up, as if under pressure.
Suddenly this line gives way over a small area (at a*), and the granules
pour through it in a thin stream nearly or quite to the anterior tip of
the Amoeba (Fig. 52, d)* Gradually the whole barrier gives way, and
nothing is left to mark the position it occupied. If after its first out-
flow the hyaloplasm has stopped, the whole Amoeba is filled with
granules. But if, as is usually the case, after a pause the hyaloplasm
has started forward again, the granules of the endosarc stream forward
not to the anterior tip, but only to the line which forfned the anterior
boundary of the hyaloplasm at its pause {x-y^ Fig. 52, e^f). Here
the granules stop and are heaped up again, until finally the barrier
breaks as before, and the granules rush forward again, to be stopped
at a new line.

The explanation of these phenomena becomes evident on careful
examination. It is to be noted that the line x-y which stops the
flow of the granules of endosarc is always identical with one which

Similar phenomena have been observed by Prowazek (1900, p. 17; Fig. 18).

THE MOVEMENTS AND REACTIONS OF AMCEBA. 163

formed the anterior boundary of the hyaloplasm (that is, of the Amoeba
as a whole) at a previous pause. The reason for this is as follows :
The lower surface of the Amoeba, as we know, is at rest ; here the
protoplasm has become modified to form a sort of membrane. This
membrane extends up a very little distance at the sides and ends (as
is shown by the fact that the protoplasm at the sides is at rest). Thus
at the anterior end there is, after a pause, a low barrier formed by this
membrane. The next wave of advancing hyaloplasm arises just behind
this barrier, overleaps it, and pushes forward (the conditions being
essentially the same as in Amoeba verrucosa^ already described). This
advancing wave when first formed is very thin, forming a mere sheet
lying on the substratum. This is shown by the fact that when the out-
line of the remainder of the Amoeba is sharply in focus, the anterior
portion is often undefined, and one is compelled to focus lower to get
its outline sharply. The thin sheet of hyaloplasm which has just pushed
forward is bounded behind by a low wall, formed from the membrane
which previouslylimit-
ed it in front (Fig. 53,
X). Now the granules
of the endosarc flow for-
ward until they reach
this boundary ; there
they stop and become
heaped up against it (Fig. 53). After a time the membrane forming
this barrier, since it is now completely enveloped by protoplasm, be-
comes dissolved and gives way in the manner described above ; the
granules then flow forward. Meanwhile, a new partial boundary
has been left in front by the hyaloplasm ; this again stops the endosarc,
and the whole process is repeated many times.

Of course, when the anterior boundary advances uniformly, without
pauses, no anterior membrane is formed, and there is nothing to hold
back the granules of the endosarc ; hence there is no reason for a separa-
tion of hyaloplasm and endosarc, and we find that none exists. On the
other hand, when the Amoeba has paused for a long time the anterior

* Fig. 53.— Diagram of a longitudinal section of the anterior edge of Amoeba,
to show the cause of the stopping of the granules of the endosarc some distance
behind the anterior margin. The line beneath represents the substratum to
which the Amoeba is attached. The anterior hyaloplasm at first moves forward to
the line x-x ; stopping there it becomes covered with a firmer wall, as represented
by the heavy black line. Now the hyaloplasm pushes forward from above the
anterior edge x-x, forming a thin sheet closely applied to the substratum, as
shown in the figure. The endosarcal granules flow forward, but are stopped by
the barrier x-x (the former anterior boundary of the Amoeba) ; they cannot flow
forward till this boundary is liquefied.

164 THE BEHAVIOR OF LOWER ORGANISMS.

membrane seems especially well developed, for the hyaloplasm pushes
then a long way ahead and may form half the length of the Amceba
before the endosarc has burst through the membrane.

The phenomena described above are very general in creeping Amoebae,
both in those with usually but a single pseudopodium (as A. Umax),
and in those with many pseudopodia (as A. angulata) .

Of course, the general fact that there is a separation between hyalo-
plasm and endosarc is not explained by these observations ;" thus we
know that they are often separated even in pseudopodia that are pro-
jected freely into the water. But the phenomena are much less marked
in such cases ; it is exactly the observed difference between them and
the phenomena to be seen in a creeping Amoeba that the above obser-
vations explain.

In all these details it is important not to lose sight of the essential
point in the movement at the anterior end. This is as follows : The new
wave begins on the upper surface just behind the former boundary line,
and rolls forward, so that its former upper surface is now in contact
with the substratum. *

This method of movement explains a peculiar fact which one observes
frequently, and which I found it difficult to understand before this
movement was demonstrated. The advancing edge in Amoeba usually
does not push forward fine, loose particles lying on the substraturti in
front of it, but overlaps them instead, so that the Amoeba creeps over
them. This is especially noticeable when the water contains many
particles of India ink or of soot. In view of the rolling movement with
the series of waves, each coming from behind the previous anterior
edge and thus descending on the substratum from above, this burying
of loose movable particles becomes intelligible.

In Amoeba proteus and its relatives the advancing anterior edge does
not move forward in a single uniform curve, as it does in A. verru-
cosa, and, as a rule, in A. Umax, On the contrary, the anterior edge
may show the most varied form and modeling (Fig. 54) ; narrow points
may be sent out here ; a broad rounded lobe there ; a rectangular projec-
tion elsewhere. Pseudopodia may rise above the general level, pro-
jecting freely into the water and later coming in contact with the
substratum, or be withdrawn without Coming thus in contact. The
anterior portion of the body may divide into two lobes, or may become
hollowed out so as to contain a cavity bounded by thin walls (see
later, Figs. 75, 76). These facts show clearly that the method of
advance of the anterior edge cannot possibly be determined by general
pressure from behind, such as would be produced, for example, by a
contraction of the posterior part of the body. Such pressure could not
produce the cavity shown in Fig. 75 or the thin edge bearing numerous
points shown in Fig. 54.

THE MOVEMENTS AND REACTIONS OF AMCEBA.

65

MOVEMENTS OF THE POSTERIOR PART OF THE BODY.

In an Amoeba moving as a simple, elongated mass, the anterior por-
tion of the body is broadest and very thin, being flattened against the
substratum, while the posterior part is narrower and much thicker. In
many cases the posterior end rises to an actual hump, the body being
thickest at the posterior edge, or a little in front of this edge (Figs. 44,
58). This is true as well of the Amoebse of nearly constant form (A.
verrucosa, etc.. Fig. 44), as of
those related to A. proteus. From
this hump the upper surface slopes
forward to the thin anterior edge.
The margins in the posterior part of
the body are not thin, but rounded
like the surface of a cylinder.

The anterior portion of the
Amoeba is attached to the substra-
tum. This attachment of the ante-
rior portion has been clearly demon-
strated by Rhumbler (1898), and I
can confirm his results throughout. t
The attachment is probably by a
mucus-like secretion ; at least such
a secretion exists, as Rhumbler and

others have shown and I can confirm. I have sometimes been able to
pull an Amoeba about by using a glass rod to which a thread of this
mucus had become attached (Fig. 55). The animal then seems to
follow^ the rod at a distance, the thread of mucus not being visible.
In virtue of this attachment the Amoeba resists currents of water,
or the impinging of solid bodies against it. The posterior portion
of the body is not thus attached, but is entirely free from the bottom.

In many cases the most posterior part of the body forms a more or
less distinctly marked off portion, the surface of which is wrinkled or
warty or villous, or otherwise irregular. This is variously known as
the tail, the villous patch, or appendage (Wallich, Leidy), houppe
(Penard), Schopf (Rhumbler), etc. The occurrence of this appendage
is variable. In some species it is usually present, in others less com-
mon. Its occurrence and degree of development vary, indeed, in the
same individual.

Fig. 54.^

♦Fig. 54. — Amosba angulata in locomotion, showing the numerous points in
the anterior region, some attached to the substratum, others projecting freely into
the water, a is the "antenna-like" pseudopodium, described on p. 177.

tit is rather curious that Butschli (1892), in his discussion of the movements
of Amoeba, is inclined to deny that there is any attachment to the substratum.

l66 THE BEHAVIOR OF LOWER ORGANISMS.

In an advancing Amoeba the posterior end moves forward at about
the same rate as does the anterior, since the distance between the two
remains about the same. Leaving out of account for the moment
specimens with the wrinkled appendage, there is a continual current
forward from the posterior end. Nevertheless, the latter remains on
the average of about the same size. The material which flows out of

it above is supplied from
beneath. As we have
seen, a layer of material
at the under surface of
the Amoeba is at rest.
The main portion of the
body passes over this
layer, dragging the pos-
^^^' ^^* terior end. The latter

takes up as it passes the resting layer which was against the
substratum. This gradually becomes fluid, and passes for-
ward again in the advancing current. All this may be
clearly seen by observing the course of individual particles
in the protoplasm and on the surface, and is fully set
forth in the preceding pages of this paper.

The unattached posterior portion steadily contracts as it
moves forward. Particles on its upper surface are moving forward, as
we have seen in detail. But this is not all. Particles on its sides and
under surface likewise move forward ; there is an actual contraction
independent of the streaming already described. The movement of
substance due to this contraction is more striking and rapid as we
approach the posterior end. As this contraction is an important fact,
it will be well to give some details of the observations which show it
to exist.

Particles attached to the lower surface, or to the lateral margins of
the Amoeba, next to the substratum, in the anterior part of the body,
remain quiet for a long time. But this lasts only till they have reached
that portion of the body which is free from the substratum ; then they
begin to move slowly forward as a result of the contraction just
described. Of two such objects, one nearer the posterior end, the
hinder one moves the more rapidly, so that the distance between the
two slowly but distinctly decreases. Though such objects on the bot-
tom or sides move forward, they do so less rapidly than does the
posterior end. The latter, therefore, in time overtakes them, and they
are finally pulled around the posterior end to the upper surface, where
they pass forward, as we have already seen in detail.

♦ Fig. 55. — An Amoeba drawn backward by a thread of its viscid secretion.

THE MOVEMENTS AND REACTIONS OF AMCEBA.

167

Fig. 56.

The contraction of the posterior part of the body is further shown by
the behavior of retracted pseudopodia. When a pseudopodium con-
tracts it usually produces, as we have seen, a
small wart-like excrescence, which persists
for some time. Such wart-like remains of
pseudopodia behave like foreign bodies at-
tached to the margin of the Amoeba. In the
anterior half they remain quiet, while in the
free posterior half they move slowly forward,
as a result of the contraction of this part of
the body. When two or more of these rem-
nants of pseudopodia are formed at once,
with an interval between them, this interval
becomes less as a result of the more rapid
movement of the hinder one.

The contraction of this posterior region is
sometimes very striking, especially when the
posterior end becomes attached to some for-
eign object and is drawn out longer than
usual ; when it finally becomes free it con-
tracts suddenly and rapidly. Thus, for example, an Amoeba hav-
ing the form shown in Fig. 56, a, began to move in the direction
shown by the arrow, when it became evident that the posterior end was
attached to a diatom shell, which was fast to the substratum. As the
Amoeba crept away the posterior end was drawn out, as shown at 6.
Finally the diatom became detached from the bottom, when the stretched
posterior end at once contracted, shortening up rapidly, so that the
Amoeba had the form shown at c. Such observations are often made.

This contraction does not occur in that part of the Amoeba which is
attached to the bottom, but begins at once as soon as the attachment
ceases. One might compare the outer layer to a stretched sheet of India
rubber that is attached to a surface by means of some adhesive sub-
stance. As soon as the adhesion gives way the sheet contracts. There
is no definite point at which the attachment to the substratum must
cease ; sometimes it is farther forward, sometimes farther back. The
gradual freeing of the posterior portion can be clearly observed in many
cases, particularly in Amoeba angulata^ and may be seen to go hand
in hand with the contraction of the ectosarc. As soon as a certain part
of the body becomes free, its contraction takes place with some sudden-
ness, and the contraction is the more noticeable the greater the part of
the body that is freed at once. Often the process of becoming freed

* Fig. 56. — Successive forms of an Amoeba, showing the marked contraction
of the posterior end. (See text).

1 68 THE BEHAVIOR OF LOWER ORGANISMS.

takes place in a series of jerks, and there is a corresponding jerkiness
in the contractions of the posterior part of the body. When a large
amount of surface is freed at once there is a sudden forward rush of the
fluid portion of the Amoeba, with a striking contraction of the posterior
part of the body. Such a case as is shown in Fig. 56 is only an unusu-
ally strong contraction of this sort, due to the fact that the hinder part
on the body had remained locally attached longer than usual.

As a result of this contraction, the ectosarc of the posterior part of
the body becomes thickened and wrinkled, or warty. The change from
a flat plate to a rounded form involves a decrease in the amount of exter-
nal surface, and as the amount of material in the surface layer is not at
once decreased, this layer is compelled to fold and become wrinkled and
warty. When this process is very pronounced we have produced at
the posterior end the wrinkled, warty appendage so often described.
Such a roughened structure may be produced in any part of the AmcEba
by rapid contraction, as we have seen above (p. 160). The rough,
warty appendage at the posterior end is the common product of all the
contractions which have taken place.

In Amoeba verrucosa and its relatives the current forward on the upper
surface extends backward to the posterior end ; the outer surface of the
latter seems not markedly different in texture from that of the rest of
the body. In other species of Amoeba there is a greater difference
between the texture of the surface layer of the anterior part of the body
and that of the posterior end, and this may involve some differences in
the movements. Often, in even these species, the forward current
extends backward to the very posterior end ; particles on the under side
pass up over the posterior end and forward, just as in A. verrucosa
(see p. 147). But in other cases, in A. limax^ A. froteus^ etc., the
surface material at the posterior end is so stiffened that it is temporarily
excluded from the current. There is then produced the distinct, rough-
ened appendage, which is for a time dragged passively behind the
Amoeba. In such a case the currents from beneath pass upward on
either side of this appendage, meeting in the middle line (Fig. 57).
Particles attached to the under surface on either side of the appendage,
therefore, soon pass to the upper surface and are carried forward, while
those on the under surface of the appendage itself may remain in position
and be dragged forward for a considerable time.

But I have rarely found this posterior appendage so completely cut
off'from the general circulation as is often supposed. Usually there is
a very slow current forward on the upper surface of the appendage,
involving also its internal parts. Into this current particles attached
to the posterior end, and even to the under surface of the appendage,
are in course of time drawn. I have thus seen particles of soot dragged

THE MOVEMENTS AND REACTIONS OF AMCEBA. 1 69

about for ten minutes at the posterior end, then finally pass upward into
the surface current, where they were carried to the anterior end. Even
when this slow current from the posterior appendage is completely
suspended the suspension is only temporary. The currents after a period
begin again, and a strongly marked warty posterior appendage may in
time completely disappear, its substance hav-
ing become mingled with that of the remain-
der of the Amoeba.

Other parts of the Amoeba may become
temporarily immobile and thus excluded from
the general circulation. One often sees certain Fig. 57.*

parts of Amoeba proteus and other similar species thus quiet, while the
restof thebody is in active motion (see Rhumbler, 1S9S, p. 122).

In the species with ^-eruptive" pseudo podia this process seems to
have gone farthest. Here the whole outer layer apparently becomes
hardened at times, so that movement occurs only when the inner sub-
stance bursts through this, forming "eruptive" pseudopodia. In this
case the pseudopodia formation apparently differs from that in A?noeba
proteus and its relatives, as described above, in the fact that the ectosarc
of thebody is not transferred to the surface of the pseudopodium. I
have not been able to study this process in detail further than to deter-
mine that there is no backward current on such pseudopodia. The
matter is worthy of further examination. In Amoeba verrucosa^ the
species which has been hitherto supposed to have the most immobile
ectosarc, we have shown that the outer layer is in continual rotary
motion in the progressing animal.

GENERAL VIEW OF THE MOVEMENTS OF AMCEBA IN LOCOMOTION.

Let us now, with the aid of a diagram, attempt to form a clear con-
ception of the movements occurring in an Amoeba that is progressing
in a definite direction, with* a view to determining the nature of the
forces at work. Fig. 58 may represent a longitudinal section of such
an Amoeba as seen in a side view. The anterior end, A^ is, as we have
seen, very thin, and is applied closely to the substratum, while the
posterior end, /*, is high and rounded, forming a sort of pouch. It is
free from the substratum, beginning at the point a*, at about the middle
of the body.

At the anterior end waves of hyaloplasm are pushed forward one
after the other, so that the anterior end successively occupies the posi-
tions a, 3, c. As we know, it is the upper surface which thus pushes
out ; it rolls over, so that a point which was originally on the upper

*FiG. 57. — Diagram of the surface currents when the posterior appendage ic
excluded from the general stream.

lyO THE BEHAVIOR OF LOWER ORGANISMS.

surface becomes applied to the substratum. It is evident that the sur-
face of the Amoeba is increased in extent by the pushing out of these
waves.

On the upper surface of the Amoeba there is a forward current of the
outer layer, as indicated by the arrows. This current extends back-
ward to the posterior end, where it is continued as a movement upward
from the lower surface. This upward movement stops, at any given
movement, at about the point _y, though, of course, the point where
it ceases cannot be precisely fixed. That part of the lower surface
which is in contact with the substratum, from the anterior end to x^ is
quiet. That part of the lower surface which is not in contact with the
substratum (from x io y) is moving slightly forward owing to the con-
traction in this region, as described on pages 166-168. This movement
is comparatively slight, as indicated by the small arrows. Within the

Fig. 58.*

Amoeba are currents moving forward in the same direction as the
current on the upper surface.

The posterior end as a whole moves forward, so that it comes to
occupy later the position shown by the dotted line. The point x^
where the lower surface of the Amoeba becomes free from the sub-
stratum, moves forward an equivalent amount to x. The entire
Amoeba thus moves forward in the direction indicated by the large
arrow above.

Thus far our account has been purely descriptive, containing only
what has been demonstrated by observation and experiment, and intro-
ducing nothing hypothetical. We must now endeavor to form a con-
ception of the location and direction of action of the forces at work in
producing the movements. Discussion of the ultimate character of
the forces will be reserved for the section on the theories of amoeboid
movement.

One of the primary phenomena is evidently the pushing forward of

* Fig. 58.— Diagram of the movements of Amoeba, as seen in side view. A^
anterior end; P, posterior end; a, b, c, successive positions occupied by the an-
terior edge. The large arrow above shows the direction of locomotion; the
other arrows show the direction of the protoplasmic currents, the longer arrows
indicating the more rapid currents. For further explanation see text.

THE MOVEMENTS AND REACTIONS OF AMCEBA. I7I

the anterior edge. The form of the Amoeba shows that this cannot be
due to pressure from behind, for if the pressure were greatest behind
and less in front, the mass of internal fluid would, of course, be forced
forward, and the Amoeba would be thickest in front instead of behind ;
the form of anterior and posterior ends would be reversed. In the
varied modeling of the anterior end we have seen another proof of the
impossibility of accounting for the action here by pressure from behind
(p. 164). Further, the forward current on the upper surface of the
Amoeba could not possibly be produced by pressure from behind.
The impossibility of accounting for the form and movement by pres-
sure from behind has been recognized by most investigators, though
on other grounds than those here set forth.

On the other hand, if we take the view that the anterior wave, after
attaching itself to the substratum, exercises a pull on the parts behind
it, the rest of the phenomena follow most naturally. Such a pull
would draw forward the tenacious upper layer of the body, and we
find that this is actually moving forward. The mass of inactive inter-
nal contents would drag behind as far as possible, so that the thickest
part of the Amceba would be at the rear, and this is exactly what we
find to be true. The posterior end would be dragged forward. This,
also, is true. In its movement forward it would be partly rolled ; that
is, its lower surface would gradually pass upward and become the upper
surface. This, also, we know to be true.* Finally, the internal fluid
contents would be compelled to stream forward as the anterior end
advanced and the posterior end followed it. This streaming is, of
course, one of the striking features of the Amoeba. The character-
istics of the endosarcal streaming are, I believe, exactly what might
be expected from the method of origin just set forth.

We have one other more or less independent factor of the movement
in the contraction of the posterior part of the body that is not in con-
tact with the substratum. As we have seen, the substance of the sides
and bottom, as well as of the upper surface, are moving forward in
this region, as indicated by the small arrows of the diagram (between
^and^). Moreover, we know that there is a lateral contraction as
well as an antero-posterior one, for the wide, flat anterior portion
shrinks together as soon as it is released from the substratum. This
contraction should probably be brought into relation with the previous
increase of surface at the anterior end. As the anterior wave is sent
forth the surface at the anterior end is much increased. We might com-
pare the action with the stretching of a sheet of India rubber. This
tense portion then becomes attached to the substratum, as we might

* Large bodies within the Amoeba close against the posterior surface are often
rolled over in this process, as I have several times observed.

172 THE BEHAVIOR OF LOWER ORGANISMS.

fasten the stretched sheet of India rubber by means of a strong cement
to a plate of glass. Upon being released the tense layer of protoplasm
contracts again, just as the rubber sheet would do. In the protoplasm
the contraction takes place rather slowly, causing the steady pulling
forward of the posterior half of the body, as exhibited by objects attached
to its lower surface.

In connection with the contraction of the lower surface as just de-
scribed we must consider also certain properties of the upper surface.
When this is pulled upon by the advancing anterior wave it does not
respond like an inelastic membrane, but like an elastic, contractile one.
If it were inelastic its motion would follow that of the anterior end
exactly, and thus take place in a series of jerks. But this does not
occur. When the anterior wave* pushes forward, thus extending the
upper surface, there is no immediate increase in the movement of this
surface, nor of the posterior end ; the movement forward is a steady
one. The property of contractility is further shown very directly in the
phenomena which take place when a large portion of the lower surface
of the Amoeba is suddenly released, as described on page 167. It seems
clear that the entire surface of the Amoeba is in a state of tension and that
this tension is directed toward the advancing anterior end. The condi-
tion would be imitated by partly filling a rubber sack with a heavy fluid,
causing one surface to adhere to the substratum and pulling on one side.

As a result of this tension on the surface the internal contents of the
Amoeba must, of course, be under a certain slight amount of pressure.
As we have seen (p. 171) even without this pressure the internal con-
tents must flow forward, since the posterior surface, against which
they are resting, is moved forward. But the pressure accounts for
certain details of the movements of the endosarc. Thus, when a pseudo-
podium is sent forth, or one of the anterior waves moves forward, it is
usually soon filled by endosarc. In its pushing forward the pseudo-
podium forms a region where the tension is relieved ; the fluid contents,
under pressure elsewhere, therefore flow into it.

It is to be noted that this pressure is a mere consequence of the tension
due to the pushing forward of the anterior edge, and is by no means a
cause of the pushing forward ; it is always, therefore, subordinate to and
dependent upon the latter, and is not a matter of primary significance.

Altogether, then, our results lead us to look upon Amoeba as an elastic
and contractile sac, containing fluid. In locomotion one side of this sac
actively stretches out, becomes attached to the substratum, and draws
the remainder of the sac after it in a rolling movement. The primary
phenomena are the stretching out of one side, the elasticity, and the
contractility of the outer layer.

Whether this elasticity and contractility should not be considered

THE MOVEMENTS AND REACTIONS OF AMOEBA. 1 73

properties, not merely of the outer layer, but of the entire substance of
the Amoeba, may be a question. The fact that ectosarc and endosarc
are mutually interconvertible would seem to imply an affirmative answ^er
to this question, and I believe other evidence could be adduced looking
in the same direction. But the locomotion itself seems to require these
properties only in the ectosarc, so that we shall for the present leave out
of consideration the question as to their existence in the endosarc.

To the three primary phenomena above mentioned we must devote
further attention. It has been maintained by certain writers that the
ectosarc is not an elastic and contractile membrane, as above set forth ;
hence we must examine the evidence on that point. There then remain
the questions : What is the cause of the pushing out at the anterior edge ?
and. What is the essential nature of the contractility of the ectosarc .!*
These questions will be reserved for a special section on the theories
of amoeboid movement. We will at this point investigate certain
general properties of the substance of Amceba, with a view espe-
cially to determining whether we are justified in considering the
ectosarc elastic and contractile, though not limiting our attention to
these properties alone.

SOME CHARACTERISTICS OF THE SUBSTANCE OF AMCEBA.
FLUIDITY.

It is, of course, not necessary to dwell upon this point ; it has been
treated in detail recently by Rhumbler (189S, 1902) and Jensen (1900).
For anyone who is familiar with the movements of Amoeba from per-
sonal observation, doubt cannot exist that its protoplasm has at least
some of the most striking properties of fluids, notably the property of
flowing, with the freedom of movement of the particles with reference to
each other that this implies. This applies most strongly to the endosarc ;
for the ectosarc, as we shall see, there are decided limitations to the
fluidity. Nevertheless, the particles of the ectosarc have, to a considera-
ble degree, that freedom of movement with relation to each other that is
characteristic of fluids. This is shown, for example, by the fact that any
portion of the ectosarc may be temporarily excluded from the advanc-
ing stream (especially common at the posterior end, p. 169), and in the
fact that neighboring portions of the ectosarc may flow in opposite
directions (p. 148) . But the characteristics of the ectosarc are much
more those of a tough, rather persistent skin than has sometimes been
supposed. This point is brought out in the following sections.

rhumblkr's " knto-ectoplasm process."
The movements of the outer layer of the body described in this paper
have an important bearing on the transformation of endosarc into ecto-
sarc and vice versa^ of which so much is said in Rhumbler's recent

1^4 THE BEHAVIOR OF LOWER ORGANISMS.

extensive paper (1898). My observations show that this transforma-
tion is confined within much narrower limits than Rhumbler supposed.
In the account which he gives of the movements of Amoeba this trans-
formation (the " ento-ectoplasm process," as he calls it) plays a very
large part, and is essential to locomotion. At the anterior end, accord-
ing to Rhumbler, endosarc constantly flows out to the surface and is
there transformed into ectosarc, flowing back as such along the surface
of the body. Somewhere in the posterior part of the body or at the base
of a pseudopodium this ectosarc passes inward and is retransformed
into endosarc. These supposed processes are indicated in the diagrams
from Rhumbler, Figs. 30-33.

My observations show that this view of the constant inter-transforma-
tion of the two layers is incorrect, and that we must attribute to the
ectosarc a much higher degree of permanence than Rhumbler supposed.
There is no regular transformation of endosarc into ectosarc at the
anterior end. On the contrary, the ectosarc here retains its continuity
unbroken, moving across the anterior end in the same manner as across
other parts of the body. In the same way, the ectosarc is not regularly
transformed into endosarc in the hinder part of the body. We can trace
a single definite point on the ectosarc (or a complex of such points
having a definite relation to each other) continually until it has passed
completely around the Amoeba ; several complete rotations of this sort
are described from actual observation on page 141 . In the species having
very changeable forms a single point on the ectosarc may be traced,
for example, from the surface of a pseudopodium at the posterior end
across the whole length of the body to near the tip of a long pseudo-
podium at the anterior end (Fig. 49, p. 155) ; there is no reason to
suppose that it could not be traced indefinitely but for the difficulties
of observation.

On the other hand, there is no doubt that ectosarc may be trans-
formed into endosarc, and vice versa, under certain conditions. This
process was apparently first clearly seen by Wallich (1863, a, p. 370).
Wallich saw the formation of " eruptive pseudopodia " by the outflow
of the endosarc through a small perforation in the ectosarc. A portion
of the latter was thus covered by the endosarc, and gradually resorbed.
Rhumbler figures a similar case (1S9S, p. 152), and I have repeatedly
seen such. Further, as Rhumbler has shown, and as I can confirm, in
A. verrucosa food bodies are enclosed in a layer of thick ectosarc, which
passes with the food into the endosarc, there to be resorbed (seep. 195).

Thus, it is clear that there may be a transformation of one layer into
the other under special circumstances, but such transformation is much
less general than Rhumbler supposed, and is by no means a regular
accompaniment of locomotion.

THE MOVEMENTS AND REACTIONS OF AMCEBA.

^1S

ELASTICITY OF FORM IN AMCEBA.

For a real understanding of the phenomena shown by amoeboid
protoplasm it is important to realize that it has, besides some of the
chief characteristics of fluids, a number of properties that are usually
considered characteristic of solids. This came out clearly in certain
of my experiments. They show that Amoeba has elasticity of form to
a considerable degree.

These experiments consisted in changing the shapes of Amoebae with
a fine capillary glass rod, under the microscope, in the open drop.
From numerous experiments of this character the following may be
selected as typical :

An Amoeba had sent out one rather long, thick pseudopodium, as
shown in Fig. 59, a. With the capillary glass rod this pseudopodium,
«, was pulled loose from the bottom and bent over into the position
shown by the dotted outline b. On being released it rather quickly

a

c f

Fig. 59.*

Fig. 6o.t

sprang back into its original position, a. This experiment was repeated
on different Amoebae many times.

An elongated Amoeba {a-a^ Fig. 60) was bent with the rod at about
its middle, so that the anterior half was pushed far to one side of the
original median axis (to b) . This anterior half at once attached itself
to the bottom, whereupon the posterior half, which was not attached,
immediately swung round into line with it, so that the Amoeba occu-
pied the position b-c. Thus the original straight Amoeba on being
bent immediately straightens itself out again. On repeating this experi-
ment with many elongated individuals it was found that frequently the
straightening out was not quite complete, so that after it had occurred
there was still a slight bend in the middle.

An Amoeba had a long pseudopodium curved over to one side, as in
Fig. 61, a. This pseudopodium was loosened from the bottom with

♦Fig. 59. — A straight pseudopodium, a, is bent into the position b with a rod.
It at once returns to the position a.

tFiG. 60. — A narrow Amoeba, a-a, is bent with the rod into the position a-b ,
the end, b, then becomes attached, and the animal at once straightens into the
position b-c.

176 THE BEHAVIOR OF LOWER ORGANISMS.

the rod and straightened out (6). On being released it at once swung
back to its original form and position at a.

An Amoeba had many long, slender, free pseudopodia standing out
radially from the body. These could be pushed repeatedly to one side
or the other, or bent at a marked angle. In every case they returned
at once to the radial position.

An indefinite number of such experiments could be detailed. They

show clearly that Amoeba has, to a certain degree at least, one of the

most distinctive properties of solids, a tendency to resist deformation of

shape, and to restore the shape when changed. It will be observed

that the cases are not such as can be accounted for on the assumption

that Amoeba is a simple fluid mass which tends to take a certain form

in accordance with the principle of least surfaces. A small sphere of

fluid when deformed returns to its original shape in conformity with

the principle just mentioned. But such returns to the original form as

^~,^ ^ are illustrated in Fig. 59

( \ and Fig. 61, for example,

y^ ^ """""""^--v. \ \ are not required by the

(]^ 1 \ principle of least surfaces

^\^ V.^ X -' "^^ , so long as the Amoeba is

^v^....^^^ ) J ^ considered a simple fluid

— -"^^ mass.

^^''' ^'■* On the other hand, if

we consider Amoeba not as a simple fluid, but as a fluid mass of
a foamlike or alveolar structure, composed of a tense meshwork of one
fluid enclosing minute drops of another, then the results above set
forth might be explained without assuming that the protoplasm has
in any part passed from a liquid to a solid state. This follows
from the considerations and experiments recently set forth by Rlium-
bler in a most important and suggestive paper (1902). Rhumbler
shows that a fluid mass having alveolar structure must react to tran-
sient pressure from without like an elastic body ; in other words,
that it must have elasticity of form. The results which I have set
forth above might almost seem, then, to have been predicted in Rhum-
bler's summing up: "Transient tensions or pressures produce an
elastic reaction of the cell body; longer action, on the other hand,
produces a plastic reaction " (/. c, p. 371). For the detailed demonstra-
tion of this principle the reader must be referred to the original paper
of Rhumbler. It will suffice to note here that the result is due to the
fact that deformations of the body as a whole produce deformations of
the alveoli, and that the surface tension of the alveolar walls tends to

♦ Fig. 61. — A large, curved pseudopodium, a, is straightened out into the posi-
tion d with a rod. On being released it at once returns to the position a.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 1 77

restore them at once to their original form, resulting in a return of the
whole body to its original form.

If, then, we hold that the substance of Amoeba is composed of such
an alveolar structure, the above observations are intelligible, even
though no part of the substance of the organism is in the solid state.
But whatever the cause, we must recognize that the protoplasm of
Amoeba shows in the gross some of the characteristics .of solids.

Leaving out of account the minute structure of the protoplasm, most
or all of the observations detailed above could be accounted for on the
assumption that the body of Amoeba consists of a sac of fluid, the outer
wall of the sac being tough and elastic, and, as is shown later, contrac-
tile. In this case only the outer layer, the ectosarc, would show the
characteristics of matter in the solid state of aggregation. Certain
observations made by the writer indicate that this is the true state of the
case. These observations are as follows : In several cases a long, slender
pseudopodium, formed of both endosarc and ectosarc, was stimulated at
the tip, causing the endosarc to be withdrawn, and leaving the pseudo-
podium formed of ectosarc alone, as illustrated in Fig. 50, page 158. Such
pseudopodia could with the glass rod be bent sharply at an angle, and
would often remain thus for some time. If, while thus sharply bent,
the endosarc, as sometimes happens, begins to flow back into the
pseudopodium, the latter straightens out with a sort of jerk as soon as
the endosarc begins to fill it. The ectosarc thus acts like an empty
glove-finger, which might bend over when empty but which would
straighten out on becoming filled with a fluid. A tough skin could
perhaps be formed by an alveolar fluid in accordance with the princi-
ples developed by Rhumbler as above set forth. Whatever the explana-
tion, the experiments indicate the existence of this tough skin-like layer
on the outer surface of the body.

CONTRACTILITY IN THE ECTOSARC OF AMCEBA.

Besides elasticity of form, the outer layer of Amoeba clearly has the
power of contracting locally. This is a fact which is omitted from
consideration in many of the theories in which amoeboid movement is
referred to local changes in the surface tension of a fluid mass. It will
be well, therefore, to set forth some of the observations on this point.
I transcribe here some of my notes, with the corresponding sketches.

1. Specimen with a single long, prominent, curved pseudopodium.
This rather quickly swings around bodily toward its concave side, and
unites with the protoplasm of the body (Fig. 62, a).

2. A specimen sends out a long, curved pseudopodium (Fig. 62, 6).
This slowly straightens out, passing from position i to position 2.

3. A. angulata usually sends out at the anterior end a single
pointed pseudopodium obliquely upward into the water (Fig. 62, c).
This point frequently waves from side to side like an antenna.

178

THE BEHAVIOR OF LOWER ORGANISMS.

4. An Amoeba was creeping on the surface film, with a very long,
slender pseudopodium trailing behind down into the water and bent to
one side. This pseudopodium suddenly swung far over to the other side.

5. Amoeba with many pseudopodia extending in all directions freely
into the water. Just as withdrawal begins, a given pseudopodium
bends over to one side, becomes curved to form a half circle, or waves
back and forth from one side to the other.

An indefinite number of such observations could be adduced, showing
that with its other movements Amoeba has the power of bending and

straightening its pseudopodia and waving
them from side to side. Such movements
have, of course, been described by many
authors ; in the magnificent monograph of
the Rhizopoda by Penard (1902) many
still more striking cases than those I have
described are set forth. Some of them
should be quoted. In Amoeba radiosa the
pseudopodia '' maybe displaced as a whole
in the liquid, and I have seen them de-
scribe in this manner a quarter of a full
circle in a second, like the handsof a watch
which one pushes forward suddenly by fif-
teen minutes. On two or three occasions
also I have noticed in the very sharp point
of a pseudopodium a rapid movement of
wave-like vibration, so that one could com-
pare it with a flagellum" (/.c.,p.88). Simi-
lar phenomena are described for Amoeba
Umax (p. 36), A. gorgonia (p. 79), and
especially for A, ambulacralis (pp. 91 , 92),
in which the pseudopodia act like tentacles.
In other rhizopods, relatives of Amoeba,
Penard describes similar phenomena. Thus
in Pamphagus mutabilis (p. 439) the pseudopodia are said to move
as a whole in the water " almost as quickly as flagella." Similar
facts are described for Difflugla pristis (p. 2^^)^Cystodifflugia sac-
culus (p. 429), Pamphagus granulatus (p. 436), Nadinella tenella
(p. 462), and various other rhizopods. Penard compares the movements

Fig. 62.*

*FiG. 62. — Movements of pseudopodia: a, a pseudopodium in the position i
bends quickly in the direction shown by the arrow, and unites with the body;
*, a curved pseudopodium, i, straightens into the position 2; c, the antenna-like
anterior pseudopodium of ^w^k^^j angulata; it vibrates from i to 2, thence back
through I to 3, etc.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 1 79

of the pseudopodia in many cases to the vibrations of flagella. Similar
movements of pseudopodia have, of course, been described by other
authors, incUiding Butschli (1878, p. 272 ; 1880, p. 123). The strik-
ing resemblance of the movements of the pseudopodia in some cases to
those of flagella (see especially the account of Podostoma, Clapar^de &
Lachmann, 1858, p. 441) seems to indicate that the motion in these
two classes of structures must be essentially similar in character, and
that no theory of amoeboid movement is likely to be correct that is
inconsistent w^ith the movements of flagella. Certain suggestions as to
the possibility of bringing the two in relation are given in the theoretical
portion of the present paper (p. 218).

The whole body is sometimes moved rapidly by such movements of
the pseudopodia. This happens especially when the body is suspended
in the water and bears many long pseudopodia, one of which comes
in contact with the substratum. This pseudopodium spreads out and
extends along the surface for a distance, the part along the surface
forming nearly a right angle with the free portion. Suddenly the
pseudopodium straightens ; since the distal end is attached, the body
is thrown almost violently against the substratum.

Somewhat similar movements take place frequently in Amoeba ver-
rucosa and its relatives, without the formation of pseudopodia. The
course of events is usually as follows : A specimen is creeping in a
certain direction in the usual manner with the anterior border attached,
while the posterior end is raised a slight distance from the substratum.
As a reaction to stimulus, or for some other reason, the anterior end
releases itself from the bottom. The posterior end thereupon sinks
down and becomes attached. Then its ectosarc contracts slightly, in
such a way as to lift the anterior end suddenly. The animal thus stands
upon what was its posterior end. Now, by varied contractions of the
parts of the ectosarc in contact with the substratum, the animal may
jerk from side to side rapidly and repeatedly, reminding one of the
movements of certain caterpillars which jerk their anterior ends about
in a similar manner. These movements are very striking and are much
more rapid than any that occur in other species of Amoeba, so far as
I have observed.

The animal may even move from place to place in this manner.
Standing on one end, it jerks its body suddenly over to one side, so that
the previously upper end comes close to the substratum. This end now
becomes attached, while the other is released. Next a new sudden con-
traction brings the released end upward, so that the animal now occupies
a new location, one body's length from that previously occupied. I have
never seen the movement go any farther than what I have just described,
so that there is no evidence that this method is employed for bringing
about orderly locomotion.

i8o

THE BEHAVIOR OF LOWER ORGANISMS.

These movements remind one of the '' rolling motion " described by
Rhumbler(i898, p. 115) for these species, though they take place with-
out any noticeable change of form and in a manner entirely different
from the movements described by Rhumbler. As we have seen above
(p. 140), the normal locomotion of these species is, in a certain sense,
of a *' rolling" character, so that the phenomena described by Rhumb-
ler as the rolling movements, perhaps really presented nothing different
in principle from the usual motion, though occurring in a diflferent way
because the organism was unattached.

In addition to movements of the character above described, certain
other phenomena show in a different way the contractility of the ecto-
sarc. Thus, I stimulated sharply with a glass rod one side of an elon-
gated moving specimen of Amoeba
Umax about one-third its length from
the posterior end (Fig. 63, a). The
body at once contracted rapidly, in a
ring-like manner (^), at this point,
and in about i J seconds the posterior
portion was cut off* completely, save
by a fine thread (c), by which it
hung to the anterior portion for a
minute or two. Later this broke,
and the posterior piece finally under-
went degeneration.

Penard, in his great work on the
Rhizopoda, describes similar phe-
nomena in Amoeba terricola {ver-
rucosa) after injury to the ectosarc.
After a small injury the injured region is invaginated, forming a small
tube passing inward, which is later resorbed. But if the injury is large
the part surrounding it contracts strongly, forming a deep constriction
between it and the remainder of the body (Fig. 64), and this injured
portion is finally constricted off* completely (Penard, 1902, p. 109).!

Altogether, then, we may consider it thoroughly demonstrated that
the ectosarc has the power of contracting in definitely limited regions
in such a way as (i) to produce movements of entire pseudopodia com-
parable to those of flagella ; (2) to produce ringlike contractions which
may even progress so far as to cut the body in two completely.

We need not, therefore, hesitate to admit the existence of contrac-
tions of the ectosarc in ordinary locomotion ; these are, for the rest,
as clearly observed as those just described.

*FiG. 63. — An A. Umax is stimulated strongly near the posterior end at a; the
stimulated part thereupon constricts (^, c), separating off the posterior end {d),
+ For other observations on reactions to injuries see pp. 302-204.

Fig. 63.*

THE MOVEMENTS AND REACTIONS OF AMCEBA. l8l

REACTIONS TO STIMULI.

Of particular importance for the understanding of the behavior of
organisms are those reactions which determine the direction of locomo-
tion. Experiments show that the stimuli to such reactions must, in a
slow-moving organism like Amoeba, affect only one side of the body,
or at least affect different parts of the body differently. Owing to the
minute size of Amoeba, it is difficult to apply stimuli in such a way as
to fulfill this condition. Heat or cold, or a chemical in solution, for
examples, when applied to one side are likely
to extend to the other side, and far beyond,
before the slow reaction of Amoeba has taken
place ; the reaction when it occurs is then to a
general, and not to a local stimulation. For ^

this reason the reactions of Amoeba to such
general stimulation are much better known than those to stimuli locally
applied. I have devoted myself to the reactions to localized stimuli,
and have succeeded in overcoming the experimental difficulties for a
number of different classes of agents, though not for all.

In examining the reactions to stimuli, it will be necessary to keep in
mind the method of locomotion (set forth briefly on p. 169 ; diagram
in Fig. 58, p. 170) . The factors to which special attention must be paid
are : (i) the sending out (or rolling over, as perhaps it would be better
to say) of waves of the ectosarc on one side, determining the anterior
end in the locomotion ; (2) the atttichment to the substratum ; (3) the
contraction of parts of the body.

The reactions were studied chiefly in Amoeba proteus and A. angu-
lata; where other species were used, they are specifically mentioned.

REACTIONS TO MECHANICAL STIMULI.

The reaction to mechanical stimuli may be either positive or negative.

POSITIVE REACTION.

An Amoeba floating in the water frequently takes a starlike form,
with many long pseudopodia projecting in all directions. If one of
these pseudopodia comes in contact with a solid object or the surface
film (which may always be considered a solid for these purposes) , the
portion in contact flattens out, attaches itself to the object, and its proto-
plasm begins to flow out in a sheet over the latter. The other pseudo-
podia are now slowly withdrawn and the entire animal spreads out on
the solid, moving usually in the direction inaugurated by the first
pseudopodium which came in contact. Often in passing to the surface
of the solid there are a number of rapid jerking movements, due to

*FiG. 64. — A. verrucosa constricting ofFan injured region, after Penard (1902).

l82

THE BEHAVIOR OF LOWER ORGANISMS.

Straightening out or bending of pseudopodia as described on p. 179.
After becoming completely transferred to the surface of the solid the
form may differ much from that of the floating Amoeba. Fig. 65 illus-
trates such a reaction. A floating Amoeba will thus spread out on the
substratum, on the surface film, or, so far as possible, on small masses
of debris suspended in the water.

An Amoeba which is moving along a surface also shows at times a
positive reaction to mechanical stimuli by turning toward small objects
with which it comes in contact at one side of the anterior end. This reac-
tion takes place very frequently in the normal locomotion of Amoeba,
but I have not been able to produce it experimentally by touching one
side of the animal with a glass rod. This is because it is difficult to
give a touch so light that it shall not induce the negative reaction. I
shall give a detailed account of reactions that probably belong here in
connection with the account of food reactions (pp. 196-202, and Figs.

73-76). As will there be shown, the reaction is often long continued
and rather complicated.

Le Dantec (1895) gave a good account of the positive reaction of
Amoeba, as shown in its spreading out on solids.

NEGATIVE REACTION.

I have studied the negative reaction to mechanical stimuli by touch-
ing a spot on one side or end of the animal with the tip of a fine glass
rod. A glass rod may easily be so drawn out that its tip is as fine as
the tip of a pseudopodium, and with some practice it is possible to give,
under the microscope, in the open drop, very precisely localized stim-
uli with this.

We will first examine the reaction to a rather strong stimulus at the
anterior edge of an Amoeba that is creeping forward with outspread
anterior end and contracted posterior end in the usual way. The tip
of the glass rod is thrust sharply against the anterior edge, producing

* Fig. 65. — Positive reaction to a mechanical stimulus in Amoeba, in side view.
A floating Amoeba comes in contact by one of its pseudopodia with a solid («);
it thereupon passes to the solid, withdrawing the other pseudopodia {b and c).
See text.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 183

a depression in the ectosarc that may last for some time (Fig. 66).
(We will suppose that the thrust does not detach the Amoeba from the
surface, as sometimes happens.) At once the anterior portion of the
Amoeba ceases to advance. It remains quiet for a definite interval,
which I should judge to be about a second, while the current from behind
continues to move forward. As a result of the stoppage at the anterior
edge there is a heaping up of the protoplasm in the middle of the body.
After about a second the part stimulated begins to contract and, currents
start backward from it. Thus the currents from the two ends meet in
the middle, often producing a further heaping up in this region. Usu-
ally, however, the ectosarc of one side of the Amoeba quickly gives
way and a new pseudopodium starts out laterally. As a rule this new
pseudopodium is formed near the original anterior margin, often at the
very edge of the area directly affected by the stimulus (Fig. 66). The
reason for this is evident. Only this anterior half of the Amoeba is
expanded and attached to the substratum, the posterior half being free
and contracted. It is, therefore, much easier to continue locomotion
by sending out pseudopodia somewhere in the attached region than
behind it. If sent out in the unattached region, the original contraction
would have to be overcome, and no locomotion could occur until the
pseudopodium had (by chance ?) come in contact with the substratum
and become attached to it. By sending out pseudopodia thus in some
portion of the attached region, the movement is, in a certain sense, a
continuation of that which was taking place before stimulation, though
in a different direction. The Amoeba follows a path which forms an
angle with its previous one.

The course of the reaction may vary considerably from that above
described. If the stimulus is weak the reaction may consist merely in
a stoppage at the point stimulated without any contraction there. The
current from behind continues ; a pseudopodium breaks out at one side
of the region stimulated, and the Amoeba moves in the direction so
indicated. If the stimulus is very weak the current may cease only
for an instant in the region stimulated, then continue as before ; the
direction of progress thus remains unchanged.

If the stimulus is very strong the contraction which takes place at
the region stimulated may be very marked, resulting in the formation
of strong folds in this region. The contraction may include the entire
anterior end of the Amoeba. Such a contraction destroys the attach-
ment to the substratum, and the new pseudopodium now bursts out in
some part of what was the posterior end of the body. The new course
followed may then be at right angles to the old one, or at any greater
angle, or the course may be exactly reversed, the new pseudopodium
being formed at the posterior end. If the posterior end was much

iS4

THE BEHAVIOR OF LOWER ORGANISMS.

wrinkled or bore a pronounced roughened *' tail," it is to be noticed
that the new pseudopodium does not flow out directly from this, but to
one side of it or above it (Fig. 67). Then as the AmcEba moves in the
reverse direction the body passes the old '* tail," which finally brings
up the rear again, fusing with the rough area produced by contraction of
the region stimulated (Fig. 67, b). Of course, the new pseudopodium
formed must come in contact with the substratum and become attached
to it before locomotion in the new direction can occur. Sometimes
the new pseudopodium formed is sent directly upward into the water ;
then there is no locomotion until the Amoeba topples over, bringing
the new pseudopodium in contact with the substratum.

At times when the anterior end is stimulated, two new pseudopodia
are sent out in opposite directions on each side of the region stimulated.

Fig. 66.*

Fig. 67.1

Both evidently pull on the Amoeba, which becomes drawn out to form
a narrow isthmus between them. Finally one end pulls the other away
from its attachment to the bottom ; the latter then contracts, and loco-
motion continues in the direction of the prevailing pseudopodium.
There is at times a peculiar additional feature of the reaction to

* Fig. 66. — Negative reaction to a mechanical stimulus in Amoeba. An Amoeba
advancing in the direction shown by the arrows is stimulated strongly with the
glass rod at the anterior end (at a). Thereupon the currents are changed and a
new pseudopodium sent out as at b.

fFiG. 67. — Negative reaction to a mechanical stimulus when the anterior end
is strongly stimulated. The arrow, «r, shows the original direction of motion ;
the arrows in a show the currents immediately after the stimulation. A large
pseudopodium is sent out from above and to one side of the former tail (/), as is
shown by the broken outline. In b this pseudopodium has come in contact with
the bottom ; the arrows show the direction of the currents and of locomotion at
this time; /, the original tail; t\ the new tail formed by the contraction of the
anterior end.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 185

strong stimuli. In some cases there is for an instant after a strong
stimulation at the anterior end a sudden rush of protoplasm toward
the region stimulated ; this is immediately followed by the stoppage
and contraction above described. Apparently this sudden rush toward
the point stimulated is produced as follows : The first effect of the
additional contraction caused by the stimulus is to release a certain
amount of surface at the posterior edge of the attached area from its
attachment to the substratum . This portion was nearly ready to become
released in the ordinary course of events, so that probably a very slight
shock would release it at once. Now, as I have shown on p. 167? when
a portion of the lower surface of the Amoeba is suddenly released from
the substratum, it contracts, causing a strong forward current. This
is what happens in the case under consideration. Later this current is
stopped by the effect of the stimulus in the anterior region.

The surface currents in the reaction are changed in the same way as
are the internal currents, and are throughout congruent with them.
Particles moving forward on the upper surface of the Amoeba stop
after the stimulus, then move in the direction of the new forward
current. This has been illustrated in detail for Amoeba verrucosa
(p. 143), so that we need not go into the matter further here.

The essential features of the negative reaction to a mechanical stim-
ulus are, then, a contraction of the region stimulated, with the formation
of a new pseudopodium in what may be considered the region of least
resistance, followed by a change in the direction of the currents of
protoplasm, thus altering the course of the Amoeba.

By repeated mechanical stimuli it is possible to drive the Amoeba in
any desired direction. I have at times made use of this possibility in
order to bring into contact two Amoebae or two pieces of an Amoeba
whose courses lay in different directions. Such driving of an Amoeba
requires considerable skill and a rather high tension on the part of the
operator. The new^ pseudopodium formed is stimulated to withdraw
as often as it is formed, until it finally starts out in the desired direction.
If it were possible to stimulate all of one side of an Amoeba at once it
would, of course, be driven directly toward the opposite side, even
though the stimulus were weak. With chemical and some other stimuli,
as we shall see, this is possible. With mechanical stimuli it is usually
possible only when the stimulus is very strong. By drawing the tip
of the glass rod along one side of a moving Amoeba, it is often possible
to make it flow directly toward the opposite side, as illustrated in Fig.
(i^. This point is important for an understanding of the effects of such
stimuli as chemicals, heat, and light.

When a single pseudopodium is stimulated, it is merely withdrawn,
wrinkling and becoming warty in the usual way ; there may be no
other effect on the movement of the animal.

i86

THE BEHAVIOR OF LOWER ORGANISMS.

Among the sweeping statements that one finds current in regard to
the behavior of these low organisms is one to the effect that the Protist
does not avoid an obstacle in its path. This statement is made for
example by Ziehen, in his excellent Leitfaden der physiologischen
Psychologies* It is worth while, therefore, to describe in connection
with the reactions to mechanical stimuli just how Amoeba avoids an
obstacle. Let us take a concrete case. An Amoeba creeping with a
broad, flat anterior end came in contact at the middle of its anterior
edge with the end of a long filament of some sort (Fig. 69). The
particular spot touched {c) ceased to move forward, becoming entirely
quiet (reaction to a weak mechanical stimulus). On each side of it
motion continued as before, so that after a time the filament projected
into a notch in the middle of the anterior edge. Then gradually the

Fig. 68. t

Fig. 69. J

forward movement ceased on the side x and increased at jk, the pseudo-
podium X contracted, and its endosarc passed into^. The animal then
continued its course in the direction indicated by y. It had thus
changed its path so as to avoid the obstacle presented by the filament.
Such cases are often seen.

* " Hindernissen weichen dieselben nicht aus " (/. c, p. 11^.

t Fig. 68. — An AmcEba moving in the direction shown by the arrows in the
unbroken outline is stimulated by drawing the tip of a glass rod along one side,
from a to b. Thereupon a pseudopodium bursts out of the opposite side, as
shown by the broken outline, and the Amceba continues locomotion in the direc-
tion so indicated.

X Fig. 69. — Method by which Amceba avoids an obstacle. The Amceba a-b-c-d
comes in contact at c with the end of a filament. Thereupon motion at c ceases,
while elsewhere it continues, so that after a time the Amoeba has the position
shown by the broken outline. Then the currents become changed in x; its sub-
stance passes into the pseudopodium y, and the Amoeba continues to move in the
direction indicated by the arrows in the lower figure.

THE MOVEMENTS AND REACTIONS OF AMCKBA. 187

Much more striking cases of the regulation of the movement in
accordance with the position and changes in position of outward things
(" automatic acts," Zielien, /. c.) than are found in such a reaction, or
even than in the possibility of driving an Amoeba, will be described
in the account of the food reactions (p. 196).

REACTIONS TO CHEMICAL STIMULI.

By analogy with the effects of mechanical stimuli we might expect
to find a positive reaction to chemical stimuli. Such reactions doubt-
less occur, but I have not been able to demonstrate them under experi-
mental conditions, and, so far as I am aware, no one else has succeeded
in doing this. Stahl (18S4) has shown the corresponding reaction to
take place in the myxomycete plasmodium. Verworn (1S90, p. 456)
records an observation which he refers with much probability to a
positive reaction to a chemical stimulus in Difflugia. If two conjuga-
ting Difflugias were separated, they crept directly together again, and
it is difficult to see how the movements could have been directed save
by some chemical. But I believe there is no instance of positive chemo-
taxis in Rhizopoda where the nature of the active substance is known
and the reaction was controlled experimentally. A number of striking
positive reactions, which should probably be attributed partly to chemi-
cal stimuli, are described later in connection with the attempts of
Amoeba to obtain food (p. 196).

The same state of affairs has existed hitherto with regard to our
knowledge of a negative reaction to chemicals. In Amoeba it is, how-
ever, not difficult to produce such reactions experimentally.

For this purpose only a small amount of the chemical must be used,
so that it can act on but a limited portion of the body of the animal.
If there is a considerable amount of the solution, diffusing over a large
area, it reaches a strength sufficient to cause a reaction at about the same
time over the whole body of the Amoeba ; thus the reaction is a general
one, not involving movement in a definite direction. To produce a
directed reaction there must be a decided difference in the strength of
the solution on two sides of the organism.

The easiest method of producing the reaction, and the one giving at
the same time the most striking results, is to dip the moistened tip of
a capillary glass rod into powdered methyl green or methyline blue ;
then to bring this near one side or end of the Amoeba, in an open drop
of water. The chemical diffuses in a colored cloud ; the reaction takes
place when the edge of this cloud comes in contact with the Amoeba.

The reaction is essentially the same as that to mechanical stimuli.
The region stimulated stops suddenly, and about a second later con-
tracts, while a current moves away from the side stimulated. This
may meet the previously existing current coming from the original

l88 THE BEHAVIOR OF LOWER ORGANISMS.

posterior end ; the two turn to one side, and a pseudopodium starts out

in a new place. If the stimulation took place at the anterior end and

was limited to a small area, the new pseudopodium starts out at one

side of the original anterior end ; the new course followed, therefore,

forms only a slight angle with the former one (Fig. 71, a). But if the

stimulus affects all of one side of the body, or a still greater portion of

its area, pseudopodia are sent out on the opposite side ; the Amcsba

then creeps directly away from the source of diffusion of the chemical.

This gives a typical case of negative *' chemotaxis," the longitudinal

axis being in the line of diffusion of the ions or molecules, with the

anterior end directed away from the source of diffusion (Fig. 70). It

will be noted that the reaction is exactly the same as that produced by

mechanical stimuli ; in " chemotaxis," where the animal is '' oriented,"

we have the same process as in '' driving" the Amoeba in a definite

direction by means of mechanical stimuli. All movement toward the

chemical is inhibited, because this brings the protoplasm into a region

-«>'?^lr**"?.'- where it is stimulated. Pseudopodia can be sent out,

.jM^^^0^0/;i^ therefore, only on the side away from the chemical,

Mi^s^p^Ma^i^c and movement can occur only m that direction.

^00 The surface currents are changed exactly as are the

internal currents; the facts here are identical with

those described for mechanical stimuli (p. 185). The

^ ^ surface currents are thus awav from a chemical which

Fig. 70.* . . , - ^.

causes a negative reaction (see Fig. 43, p. 144).

A number of variations in the reactions to chemicals are shown in
Fig. 71, all of them taken from actual experiments. As the figure
shows, after stimulation frequently two pseudopodia start out in oppo-
site directions, one finally prevailing over the other (Fig. 71, d).

The contraction due to the chemical is often very marked, the ecto-
sarc against which the chemical impinges shrinking sharply together
and becoming covered with folds (Fig. 71, 6). With methyl green as
the stimulus, the surface touched by the chemical is sometimes stained,
so that the shrinkage in area is very precisely definable. With a solu-
tion of NaCl the shrinkage is extreme, while the opposite side spreads
out widely, compensating, or more than compensating, for the decrease
in surface caused by the shrinkage (Fig. 71, 6).

The effect of substances not in the form of powder was tried in the
following manner : A glass tube was drawn out to a very fine point,
and into it was introduced some of the solution to be tested. The fine
point of the tube was then brought close to the Amoeba. Some of the

* Fig. 70. — Diagram of " negative chemotaxis " in Amoeba. A chemical diffuses
from a center, as indicated by the radii; the Amoeba reacts in such a way as to
creep directly away from the 80urceofdiffusion,ina line with the radii of diffusion.

THE MOVEMENTS AND REACTIONS OF AMCEBA.

189

chemical flowed slowly out, and its action on the Amoeba could be
observed. The results obtained by this method were very clear. A
negative reaction, as described above, was observed in this way for
solutions of the following substances : Methyl green, methyline blue,
sodium chloride, potassium nitrate, potassium hydroxide, sodium car-
bonate, acetic acid, hydrochloric acid, cane sugar. Of these substances
relatively strong solutions were used (usually about i per cent). Of
course, the solution which came in contact with the Amoeba was much
weaker than this, being diluted by the surrounding water. Emphasis
was not laid on the quantitative aspect of the matter ; the question pro-
posed was. How does the animal react .^ and not, How much is required
to produce the reaction ? Therefore, different strengths were employed

:.•.v.^r_•;.•.■:v■^;••.'.•• .-. •• -

Fig. 71.*

till one was found that was effective. In any case, I do not know of
any way in which one could determine the exact strength of the solution
which comes in contact with the surface of the Amoeba.

* Fig. 71. — Variations in reactions of Amoeba to chemicals. The dotted area
represents in each case the diffusing chemical. The arrows show the direction
of the protoplasmic currents.

a. The chemical (methjl green) diffuses against the anterior end of an advanc-
ing Amoeba; the latter reacts by sending out a new pseudopodium at one side of
the anterior end and moving in the direction so indicated.

6. A solution of NaCl diffuses against the right side of a moving Amoeba (i).
The side affected contracts and wrinkles strongly, while the opposite side expands
(2), the currents flowing in the direction indicated by the arrows.

c. A solution of NaCl diffuses against the anterior end of an advancing Amoeba.
The course is thereupon reversed, a broad pseudopodium, shown by the dotted
line, pushing out from the upper surface of the posterior end above the tail.

d. Asolution of methyline blue diffuses against the anteriorendof an advancing
Amoeba (i); thereupon a pseudopodium is sent out on each side of the posterior
end at right angles with the original course (2). Into these pseudopodia are
drawn the body and the tail (3).

190 THE BEHAVIOR OF LOWER ORGANISMS.

As a control experiment, distilled water was used in the tube in
place of a chemical in solution. Amoeba was found to react negatively
to this also, though the reaction was less marked than with most of the
chemicals. But this result, of course, rendered the experiments with
solutions of chemicals in the tube indecisive, as the Amoeba may have
reacted to the distilled water in which the* solutions were made up.
The solutions were, therefore, made up with culture water, and the
same results were obtained as before.

The results with the chemicals show merely that Amoeba responds
negatively to almost any solution differing markedly from that in which
the animal is immersed, the precise chemical composition of the solu-
tion being of little consequence. The animal responded negatively not
only to distilled water and to the chemicals mentioned, but also to tap
water, and to water taken from other cultures than that in which the
specimens occurred.

REACTIONS TO HEAT.

Verworn (1889, pp. 64-67) studied the directive influence of heat on
the locomotion of Amoeba by concentrating the sunlight on a small por-
tion of the slide and leaving the rest dark, then observing the behavior
of the Amoeba on coming to the boundary of this lighted and heated area.
The effects of the light proper were excluded by control experiments.
It was found that on coming to the heated area Amoeba remained quiet
a moment, then contracted on the heated side, and sent out a pseudopo-
dium on the opposite side. It then crept away in the direction indicated
by this pseudopodium (" negative thermotropism").

My experiments differed from those of Verworn in employing con-
ducted heat in place of radiant heat ; thus there was no possibility of
a complication from the effects of light. As Verworn sets forth, it is
diflScult to warm only one side of so small an object as an Amoeba. I
succeeded^ however, in doing this in a very sinfple manner. For each
experiment an Amoeba was selected that was creeping on the under
surface of the cover glass. This was placed in focus under an objective
of a considerable focal distance, yet of high enough power so that the
internal movements could be seen. A needle was then heated in a flame
and its point was brought against the cover glass a short distance in
advance of the Amoeba. Control experiments had shown that the use
of a needle at room temperature had no effect.

If the heated needle was placed at a proper distance from the Amoeba,
the phenomena follow as described by Verworn (/. c). There was a
short pause, then the side next to the needle contracted. A current
of protoplasm passed toward the opposite side, at times meeting the
current already in existence. A new pseudopodium was sent out, either
on the side opposite the needle, or, in many cases, in a direction inter-

THE MOVEMENTS AND REACTIONS OF AMCEBA. I9I

mediate between this and the original one. The phenomena are in all
respects identical with those seen in the negative reaction to mechanical
stimuli. If the needle is brought a little nearer, so that the heat acts
more strongly, there is a sudden strong contraction of the side affected.

Simultaneously with this there is often, as in the case of strong
mechanical stimuli, a sudden rush of the internal fluid toward the side
stimulated. This lasts but an instant and is succeeded by a current
away from the stimulated side, the formation of a pseudopodium on the
unstimulated side, and locomotion in that direction. The sudden rush
of internal contents toward the side affected is, I think, clearly due to
the cause suggested under mechanical stimuli. Part of the posterior
portion of the attached area of the Amoeba is loosened from the substra-
tum by the sudden contraction at the front end ; this portion, therefore,
contracts quickly and sends a current forward, as described on p. 168.

When the heat is still more powerful the entire Amoeba is affected.
It contracts and at the same time loses its attachment to the substratum.
There is a strong momentary rush of the internal fluid toward the end
which had been anterior, due to the cause set forth in the preceding
paragraph. This ceases and the body becomes very irregular and ceases
to move.

The reaction to local stimulation by heat is thus of essentially the
same character as the reaction to mechanical stimuli and to chemicals.

Like Verworn (1889, P* ^7)' I ^^^^e been unable to obtain a reaction
to cold in Amoeba.

REACTIONS TO OTHER SIMPLE STIMULI.

The reactions of Amoeba to electricity and to light have been thor-
oughly studied by other authors, so that it will not be necessary to treat
them in detail here. Only certain especially important points will be
touched upon.

The reactions of Amoeba to the continuous electric current have
been studied in detail by Verworn (1890, a; 1897). I have repeated the
experiments in order to determine by observation the direction of the
surface currents of protoplasm during the reaction. For this purpose
soot was mingled with the water containing the Amoebae, and the elec-
tric current was passed through the preparation. The typical reac-
tion as described by Verworn was observed in many cases, but the
surface currents, of course, cannot be seen unless soot is resting upon or
is attached to the surface of the animal, which happens only rarely.
Finally a specimen of Amceba proteus was observed with a string of
soot particles attached to one side (Fig. 73). The electric current was
then passed through the preparation in such a way that the side bearing
the soot was next to the anode (Fig. 72, a). The Amoeba thereupon

192

THE BEHAVIOR OF LOWER ORGANISMS.

turned and began to move toward the cathode (6) , dragging the parti-
cles of soot behind it for a short distance. Then the string of soot began
to pass forward on the upper surface, in the usual way (c, d). This
continued until the soot reached the anterior edge and dropped off the
surface of the Amoeba (e) . The currents on the upper surface of Amoeba
are, then, forward (toward the cathode) in the reaction to the electric
current, as well as in other cases.

On reversing the current the specimen described above began to
move in the opposite direction toward the new cathode. In this and
many other observed cases of the reversal of movement under the in-
fluence of the electric current, the reversal occurred in the same manner
as when induced by other stimuli (see p. 183). That is, the new pseudo-
podium was sent out from one side of the attached (anterior) half of the
body, changing the course a certain amount. From this new portion
another new pseudopodium was sent out on the side toward the anode,

Fig. 72.*

and this continued until the direction of movement had been, by a
gradual process, completely reversed. Verworn (/. c.) describes cases
in which the reversal takes place suddenly, the new pseudopodium
appearing at the original posterior end. This happens also at times, as
we have seen, in the reactions to other stimuli (p. 183). It is to be noted
that the reaction to the electric current is exactly that which would
occur if the animal were strongly stimulated on the anode side.

Verworn (1889), Davenport (1S97), and Harrington & Leaming
(1900) have studied the reaction of Amoeba to light. Verworn (/. c,
p. 97) found that light falling perpendicularly on one-half of the Amoeba
produced no reaction. Davenport (/. c, pp. 186, 188) confirmed this
result, but showed that when the light falls obliquely from one side on
Amoeba the animal reacts negatively. Harrington & Leaming (/. c.)
found that when white light falls upon the Amoeba from above the

*FiG. 72. — Movement of particles attached to the outer surface of Amoeba in
the reaction to the electric current. Anode and cathode are represented by the
plus (-}-) and minus ( — ) signs, a, Form and direction of movement of the Amoeba
before the current is made ; x, a chain of soot particles attached to one side ; 6, c,
d, e, successive stages during the reaction. The chain of soot particles (x) passes
to the upper surface and forward, reaching at e the anterior edge.

THE MOVEMENTS AND REACTIONS OF AMCEBA. I93

movements cease, while in red light they begin again ; lights of other
colors have various intermediate effects.

It seems to the writer that further experimentation is desirable on the
results of a perpendicular illumination of one-half the animal. The
difference between the results thus far obtained from such illumination
and those from Davenport's experiments where light is admitted from
one side is very remarkable. If this difference is constant, it is of much
significance for the theory of light reactions. Possibly the lack of re-
action when but one-half the animal is illuminated may be accounted
for as follows : When one end of an Amoeba is illuminated from below,
as in Verworn's experiments, it is difficult or impossible to keep this
difference of illumination constant for any considerable period. If the
Amoeba does not react at once it passes completely into the lighted area,
where there is no cause for changing the direction of movement. On
the other hand, in the case of light falling obliquely from one side, the
different action of the light on the two sides is constant, so that in time
a reaction is produced. The slowness of Amoeba in reacting is such
as to make this possibility worth considering. For further work from
this point of view a source of powerful artificial light is needed. This
I have not had at command during the present investigation.

It is evident that the reaction of Amoeba to light falling from one side
is exactly that which would be produced were the Amoeba strongly
stimulated on the side on which the light impinges.

For an account of the reactions of Amoeba to general (not localized)
stimuli, see Verworn, 1888, and the Allgenieine Physiologic of the
same author.

SOME COMPLEX ACTIVITIES.

Under this heading I propose to describe certain striking phenomena
in the behavior of Amoeba, the stimuli to which are complex or not
sharply definable. These concern the reactions of Amoeba to food and
to injuries, and the relations of one Amoeba to another.

ACTIVITIES CONNECTED WITH FOOD-TAKING.

The behavior of Amoeba in taking food or in attempting to take food
shows many features of great interest for one attempting to understand
the behavior of these organisms. I have observed the process of food-
taking many times, and will describe it, together with a number of
related activities.

Let us take a concrete case. A specimen oi Amceha proteus was creep-
ing about on a slide which contained many spherical cysts of Euglena
viridis. One of these, which was not attached to the bottom (as most
of them are), was lying in the path of the Amoeba. The latter in its
forward movement came against the cyst and pushed it forward a short
distance. There was no evidence of a tendency of the cyst to adhere to

194 THE BEHAVIOR OF LOWER ORGANISMS.

the Amoeba ; on the contrary, it was pushed ahead as fast as the Amoeba
moved. The Amoeba now put out a pseudopodium on each side of the
cyst, while that part of the protoplasm immediately behind it stopped
moving. Thus the cyst was enclosed in a little bay. On bringing the
upper surface of the cyst into focus it could be seen that a thin layer of
protoplasm was also sent over the cyst. The two pseudopodia enclosing
the cyst now bent over at their free ends, so that the cyst could not be
pushed away by movement of the Amoeba. The two free ends finally
met, leaving only a sort of transparent seam to show the place of con-
tact. Later this disappeared, and the two pseudopodia fused completely.
At the end of two minutes from the time that the Amoeba first came in
contact with it the cyst was completely enclosed. The Amoeba now
remained perfectly quiet for one minute, then crept away, carrying the
cyst with it With the cyst had been taken in some water, so that it
was enclosed in a vacuole a little larger than itself. The walls of the
vacuole had exactly the appearance of the ectosarc on the outer surface
of the Amoeba.

This is essentially the method of food taking that I have observed in
a large number of cases in Amoeba proteus and its relatives. The
essential points are the sending out of pseudopodia on each side of
and above the food body and the fusion of these pseudopodia at their
free ends or edges, thus enclosing the food. In no case, in these species,
was there any evidence that the Amoeba was aided by the adherence
of the food body to its protoplasm. On the contrary, there was a
decided tendency for the food body to be pushed away, and an essential
part of the process is the overcoming of this mechanical difficulty by
sending out a pseudopodium on each side of the body and bending the
ends of them together, so as to prevent slipping on the part of the
food.* That this difficulty is no imaginary one will be shown later,
in the description of cases where the Amoeba was unable, after many
efforts, to enclose the food.

It is commonly said that the posterior rough, tail-like portion of the
body is especially important in the taking of food, though it is sometimes
added that one rather more often sees the partly ingested food given out
again in this region (see Leidy, 1879, p. 45 ; Penard, 1890, p. 81 ; 1902,
p. 16). t I have never seen food taken in at this part of the body, though,
as noted above, I have many times seen it taken at the anterior end.
While it may be true that food is at times taken at the posterior end, I

*This lack of adherence between the protoplasm and the food substance is
emphasized by Le Dantec (1894, p. 68; as a result of his careful studies on food-
taking in Amoeba.

t The references to food-taking at the posterior end seem all to go back to a
paper by P. M. Duncan, " Studies amongst Amoebae," in the Popular Science
Review for 1877. I regret that I have been unable to sec this paper.

THE MOVEMENTS AND REACTIONS OF AMCEBA. I95

believe that the supposed prevalence of this method of food-taking and
of the giving off of incompletely ingested food here are really due to
incorrect interpretation of another very common process. In its loco-
motion Amoeba frequently comes in contact with diatoms, desmids,
encysted Protozoa, etc. These it usually creeps over, so that they lie
beneath it As the Amoeba progresses the objects come in contact with
the posterior portion of the body, which is raised from the bottom and
covered with a viscid secretion. Owing to this viscid substance the
objects often cling to the under surface of this part of the body and are
carried along with it. If observed at this time one cannot tell whether
these objects have been ingested or not. But as a result of the method
of movement of the Amceba they gradually pass to the posterior end,
and are usually finally left behind. When such an object separates from
the Amoeba, becoming detached from its under surface, it appears ex-
actly as if it were given off from within ; it is only by observing the
whole process from beginning to end that one can be sure of its exact
nature. I am convinced that many of the supposed cases of the inges-
tion of food and of the ejection of food previously ingested at the poste-
rior end are to be explained in this way. In all the detailed descriptions
of food-taking in forms related to Amoeba proteus that I have found in
literature, the food was taken at the anterior end in a way similar to
that which I have described above (see Carter, 1863, p. 45 ; Wallich,
1863, c, p. 453 ; Leidy, 1879, p. 49 ; Le Dantec, 1894, p. (i^) ; also Biit-
schli, 1880, p. 117).

In Amoeba verrucosa and the other species which do not often form
pseudopodia food is taken in a somewhat different manner. Food-
taking in Amoeba verrucosa has been described by Rhumbler (189S,
p. 205). Penard (1902), though he spent much time studying this
species, says that he has not observed the taking of food, and thinks it
must occur only rarely. In my own cultures specimens of this species
taking food were positively abundant. I have seen the process much
oftener than in other species. In A. verrucosa and its relatives food-
taking is greatly aided by the tendency of foreign particles to cling to
the surface of the body, a tendency which we found so convenient for
determining the movement of points on the body surface (see pp. 140-
146). This adhesiveness of the outer surface compensates for the lack of
formation of pseudopodia in these species. The outer surface gradually
sinks in at the point where the food body is attached to it. The latter
is thus carried to the inside of the body, surrounded by a pouch of
ectosarc. This pouch becomes separated from the outer ectosarc. The
food is thus completely enveloped and later digested. Not only large
objects, but often very small ones, spores of algae, small diatoms, flagel-
lates, etc., are taken in in this way. Rliumbler (/. c, p. 20S) has

196

THE BEHAVIOR OF LOWER ORGANISMS.

given a very interesting account of the rolling up and ingestion of threads
of Oscillaria by this species (see p. 223).*

PURSUIT OF FOOD.

Ammba froteus does not always succeed in ingesting its food so
easily as in the case just described (p. 194). There is, as noted above,
a tendency for the food body to be pushed away by the forward move-
ment at the anterior end of the Amoeba, and this sometimes gives
serious difficulty. In such cases Amoeba may show what would be
called in higher organisms remarkable pertinacity in continuing its

Y

^

Fig. 73. t

attempts to ingest the food. This will be illustrated from a concrete
case (Fig. 73).

An Amoeba proteus was creeping toward an encysted Euglena.
The latter ^as perfectly spherical and very easily moved, so that when
the anterior edge of the Amoeba came in contact with it the cyst merely
moved forward a little and slipped to one side (the left). The Amoeba

♦Leidy (1879, P- 86) gives a very similar account of the ingestion of filaments
of algae in Dinamcaba.

tFiG. 73. — Amceba following a rolling Euglena cyst. Nos. 1-9 show successire
positions occupied by Amoeba and cyst. See text for explanation.

THE MOVEMENTS AND REACTIONS OP AMCEBA. I97

thereupon altered its course so as to follow the cyst (Fig. 73, i). The
cyst was shoved forward again and again, a little to the left ; the
Amoeba continued to follow. This continued until the two had tra-
versed about one-fourth the circumference of a circle ; then (at 3) the
cyst, when pushed forward, rolled to the left quite out of contact with the
Amoeba. The latter then continued forward with broad anterior edge
in a direction which would have taken it past the cyst. But a small
pseudopodium on its left side came in contact with the cyst. The
Amoeba thereupon turned again and followed the rolling cyst. At
times it sent out two pseudopodia, one on each side of the cyst (as at
4), as if trying to inclose the latter, but the ball-like cyst rolled so easily
that this did not succeed. At other times a single very long, slender
pseudopodium was sent out, only the tip of which remained in contact
with the cyst (5). Then the body of the Amoeba was brought up from
the rear and the cyst pushed farther. This continued until the rolling
cyst and the following Amoeba had described almost a complete circle,
returning nearly to the point where the Amoeba had first come in con-
tact with the cyst. At this point, owing to the form of the anterior
end of the Amoeba (7) the cyst rolled to the right instead of to the left
as it was pushed forward. The Amoeba followed (8, 9). This new
path was continued for two or three times the length of the Amoeba.
The direction in which the ball was rolling would soon have brought
it against an impediment, *and I thought it possible that the Amoeba
might succeed in ingesting it after all. But at this point one of those
troublesome disturbers of the peace in microscopic work, a ciliate infu-
sorian, came near and whisked the ball away in its ciliary current.
After the ball was carried away the Amoeba continued to follow in the
same direction for only a very short distance, about one-fifth its length,
then reversed its course and went elsewhere.

The movements of Amoeba are, of course, very slow, and the behavior
described required a considerable period of time — 10 or 15 minutes, I
should judge. The whole scene made really an extraordinary impres-
sion on the observer, and it is difficult in describing it to refrain from
the use of words that imply a great deal of resemblance between Amoeba
and immensely higher organisms. One seems to see that the Amoeba is
trying to obtain this cyst for food, that it puts forth efforts to accom-
plish this in various ways, and that it shows remarkable pertinacity
in continuing its attempts to ingest the food when it meets with diffi-
culty. Indeed, the scene could be described in a much more vivid
and interesting way by the use of terms still more anthromorphic in
tendency.

I have seen a large number of cases like that above described ; in
some of my cultures containing many specimens of Amoeba proteus

198

THE BKHAVIOR OF LOWER ORGANISMS.

and many Euglena cysts it was not at all rare to find the animals
engaged in thus following a rolling ball of food. I have made full
notes and sketches of a number of other cases, but they show nothing
different in principle from that above described, so that it is not worth
while to enter into details. One further point is, however, worthy of
special note. Often a single pseudopodium comes in contact with such
a cyst and stretches out toward it, while the remainder of the Amoeba
continues on its course, away from the cyst. The pseudopodium in
contact then stretches out as far as possible, keeping in contact with
the cyst and often pushing it ahead (Fig. 74), until it is finally pulled
bodily away by the movements of the whole Amoeba. Apparently this
one pseudopodium reacts to the stimulus quite independently of the
remainder of the body. Again, two pseudopodia on opposite sides of
the body may each come in contact with a cyst. Each then stretches

Fig. 74.*

out, pulling a portion of the body with it, and follows its cyst, until the
body forms two lobes, connected only by a narrow isthmus. Finally,
one half succeeds in pulling the other away from the attachment to the
substratum, and the entire Amoeba follows the victorious pseudopodium.
Mechanical stimuli are, of course, involved in the above reactions ;
perhaps also chemical stimuli from the cyst. It is important from the
theoretical standpoint to note that the movement of particles on the
surface of the Amoeba is toward the object causing the reaction. This
I have been so fortunate as to have opportunity of observing in several
cases.

OTHER AMCEB^ AS FOOD.

Amoebae frequently prey upon each other, as Leidy has already de-
scribed and figured (1879, p. 94 ; plate 7, Figs. 12-19) • ^"t the victim
does not always conduct itself so passively as in the case described by
Leidy, and sometimes finally escapes from its pursuer. A description

♦Fig. 74. — A single pseudopodium (:v) reacts positively to a Euglena cjst, the
protoplasm flowing in direction of cyst and pushing it forward, while remainder
of the Amceba moves in another direction ; 1—4, successive forms taken. At 4 the
reacting pseudopodium is pulled away from the cyst, and then contracts.

THE MOVEMENTS AND REACTIONS OF AMCEBA.

199

of two or three concrete cases among those which I have observed in
A??iceba angulata will bring out the nature of the behavior under such
conditions. Penard (1902, p. 700) mentions that he has seen one
Amoeba pursue and finally capture another, but does not give a detailed
account of the process.

(i) Two Amoebae were observed, a large one and a small one, the
former apparently attempting to swallow the latter (Fig. 75). The
small Amoeba was creeping rapidly forward, while its wrinkled pos-
terior portion was enveloped by the anterior part of the larger Amoeba.
The large Amoeba had the anterior portion of its body quite hollowed
out, so as to form a cavity large enough to contain the entire small

/ _

2

Fig. 75.*

Amoeba, and in the anterior portion of this cavity was inclosed the
hinder portion of the body of the smaller Amoeba. Whether this
cavity was bounded below as well as above and at the sides by pro-
toplasm I could not determine with certainty. The large Amoeba was
following the small one, moving at about the same rate. There was
no union between the protoplasm of the two ; on the contrary the
boundaries of both were clearly defined, and they seemed to be only
slightly in contact, the posterior end of the small specimen moving
easily within the cavity of the other. As they moved forward, some-
times the posterior specimen flowed a little faster, and then a little
more of the smaller one became enveloped ; at other times the smaller
Amoeba moved a little faster, and then withdrew a part of its inclosed

* Fig. 75. — Pursuit of one Amceba by another. See text for explanation.

aOO THE BEHAVIOR OF LOWKK ORGANISMS.

tail. They progressed in this fashion for a long distance, many times
their own length. I watched them thus for more than lo minutes.
The smaller, anterior specimen frequently altered its course ; the
posterior one followed. I stimulated the anterior end of the small
specimen with the tip of a glass rod (Fig. 75, 3) ; it turned at a right
angle, and the posterior specimen followed. After about 12 minutes it
could be seen that the smaller specimen was moving slightly faster
than the other and was slowly withdrawing its posterior end. Finally
it pulled completely away from the large Amoeba, which was still fol-
lowing as rapidly as possible. After the small Amoeba had completely
escaped the large one stopped and remained entirely quiet for a few
seconds. The large cavity in its anterior portion, which it had pre-
pared for the reception of the small Amoeba, and which extended
back behind the middle of the body, was still very evident. After a
time the Amoeba began to change form and sent out pseudopodia
irregularly in all directions. The smaller Amoeba continued its for-
ward locomotion as long as observed. The performance is illustrated
in Fig. 75, from sketches made while it was in progress.

(2) In a second case I was able to observe the beginning as well as
the end of this microscopical drama (Fig. 76). I had attempted to
cut an Amoeba in two with the tip of a glass rod, in the manner described
later. The posterior third of the Amoeba, in the form of a wrinkled
ball, remained attached to the body only by a slender cord, the remains
of the ectosarc. The Amoeba began to creep away, dragging with it
this ball. I will call this Amoeba a, while the ball will be designated 3.
A larger Amoeba (c) approached, moving at right angles to the path
of the first Amoeba ; its course accidentally brought it into contact
with the ball <5, which was dragging past its front. Amoeba c there-
upon turned, followed Amoeba a, and began to engulf the ball d. A
large cavity was formed in the anterior end of Amoeba c, reaching back
nearly or quite to its middle, and much more than sufficient to contain
the ball d. Amoeba a now turned into a new path ; Amoeba c followed
(Fig. 76 at 4). After the pursuit had lasted for some time the ball 6
had become completely enveloped by Amoeba c; the cord connecting
it with Amoeba a broke, and the latter went on its way (at 5) and dis-
appears from our account. Now the anterior opening of the cavity in
Amoeba c became partly closed, leaving a slender canal (5) . The ball d
was thus completely inclosed, together with a quantity of water.
There was no union or adhesion of the protoplasm of 6 and c; on the
contrary (as the sequel will show clearly) both remained quite sepa-
rate, c merely inclosing 5.

Now the large Amoeba c stopped, then began to move in another
direction (Fig. 76, 5-6), carrying with it its meal. But the meal, tlie

THE MOVEMENTS AND REACTIONS OF AMCEBA.

20 1

ball 3, now began to show signs of life, sent out pseudopodia, and,
indeed, became very active. We shall henceforth, therefore, speak of it
as Amoeba b. It began to creep out through the still open canal, send-
ing forth its pseudopodia to the outside (Fig. 76, 7). Thereupon
Amoeba c sent forth its pseudopodia in the same direction, and after
creeping in that direction several times its own length, again completely

Fig. 76.*

inclosed b (7-8). The latter again partly escaped (9), and was again
engulfed completely (10). Amoeba c now started again in the opposite
direction (i i), whereupon Amoeba b, by a few rapid movements, escaped
entirely from the posterior end of c, and was free, being completely
separated from c (11-12). Thereupon c reversed its course (12), crept
up to 3, engulfed it completely again (13), and started away. Amoeba b

♦Fig. 76. — Pursuit, capture, and ingestion of one Amoeba by another; escape
of captured Amoeba and its recapture ; final escape. See text for detailed account.

a02 THE BEHAVIOR OF LOWER ORGANISMS.

now contracted into a ball, its protoplasm clearly set off from the pro-
toplasm of its captor, and remained quiet for a time. Apparently the
drama was over. Amoeba c went on its way for about five minutes,
without any sign of life in b. In the movements of the Amoeba c the
ball h gradually became transferred to the posterior end of c, until
finally there was only a thin layer between b and the outer water.
Now b began to move again, sent out pseudopodia to the outside
through the thin wall, and then passed bodily out into the water (14).
This time Amoeba c did not return and recapture b. The two Amoebae
moved in diflierent directions and remained completely separated. The
whole performance occupied, I should judge, about 12 to 15 minutes
(the time was not taken till several minutes after the beginning).

After working with simple stimuli and getting always direct simple
responses, so that one begins to feel that he understands the behavior
of the animal, it is somewhat bewildering to become a spectator of so
striking and complicated a drama. If we attempt an analysis of the
observed behavior of the Amoeba c into stimuli and reactions, we ob-
tain some such a result as follows : At first the stimulus of contact with
^, and perhaps a chemical stimulus from the same source, causes the
Amoeba c to react by flowing toward 3, and at the same time to change
form, so as to hollow out the anterior end. Later, every change in the
direction of movement of a and b induces a corresponding change in
the direction of movement of c; there is a finely co-ordinated adapta-
tion of the latter to the movements of the former. After the separation
of b from <z, the movement of c (at 5-6) in a different direction may
have been due to some external stimulus. But what is the stimulus for
the change of direction of locomotion in the Amoeba c at 7 when b has
begun to escape ? And why does Amoeba c go in that direction only
long enough to get b firmly inclosed again, then reverse its course.?
And, finally, why does Amoeba c reverse its course at 1 1-13, when b has
entirely escaped, and continue in this reversed direction till it reaches
and recaptures b? The action is remarkably like that of a higher
animal. Doubtless we must assume chemical and mechanical stimuli
as directives for each of the movements of c, but the analysis so obtained
seems not very complete or satisfactory.

RBACTIONS TO INJURIES.

Certain cases that belong under the heading of reactions to injuries
have already been described as evidences of the contractility of the
ectosarc ; for these page 180 should be consulted. The cases which
we take up here are of a difl^erent character. They concern Amoeba
angulata.

Jensen (1896) has shown in the case of certain Foraminifera that
two pieces of protoplasm from the same individual will readily unite

THE MOVEMENTS AND REACTIONS OF AMCEBA.

303

if brought in contact, while pieces from different individuals will not
thus unite. I was interested in the question as to whether this would
hold true also for Amoeba, and for that purpose undertook to cut speci-
mens in two. With the fine tip of a glass rod it is possible, under the
microscope, in the open drop, to cut in two an elongated Amoeba at
almost any desired point. The sharp point of the rod is simply drawn
across the Amoeba as it lies outspread on the substratum.

In this operation it was found that the Amoeba was not, as a rule, at
first completely cut in two by the stroke itself. The endosarc is divided
completely, but the two halves are still connected by a thin layer of
ectosarc, which resists the cutting, and shows fine longitudinal stria-
tions ; these may be merely longitudinal folds (Fig. 77, 2). This thin
layer of ectosarc seems very tenacious.

2

Fig. 77.*

The Amoeba is thus left in the condition shown in Fig. 77, 2. The
two halves usually both contract strongly. Now ensues a very pecu-
liar process. One of the two halves begins to send out pseudopodia
in such a way as to partly inclose the other (3) ; the second half is
thus drawn as a narrow wedge-shaped mass inside of the other, as at 4.
It seems to be usually the half that contains the nucleus that envelopes
the other, though, as will be shown later, the nucleus is not necessary
for this reaction. If the piece thus embraced is considerably smaller
than the other, it may become completely inclosed, and is then carried
away, appearing like a mass of food. It does not become fused with
the remainder of the protoplasm, but there is a sharp boundary between
it and that which envelopes it. Specimens were followed for 10 min-
utes after thus inclosing a piece of their own bodies ; during this time
no marked change was seen to occur in the inclosed piece.

In the much more common cases where the two pieces are nearly

*FiG. 77.— Reaction of Amceba to injury. See text.

204 THE BEHAVIOR OF LOWER ORGANISMS.

equal in size, or when it is the smaller piece that begins to envelop
the larger, the process results differently. After one piece has been
drawn far into the other (Fig. 77, 4), both seem to contract strongly,
whereupon the connecting band of ectosarc breaks, the partly inclosed
piece is squeezed out of the other ; and the two separate. Usually
each retains its form for a few seconds after separation ; one bearing
a slender truncate pyramid or cone, while the other shows a deep
depression corresponding to this pyramid (Fig. 77, 5). After a time
both halves change form and move away. Usually the half which
partly inclosed the other becomes active long before the other, but
this is not invariably true.

In a large number of cases observed it was the part which contained
the nucleus that attempted to envelop the other half. In order to
determine whether the nucleus plays a necessary part in this perform-
ance, I tried the following experiment: After the half which had
no nucleus had again become active and was moving about, I cut it in
two, as before. Now one half of this piece partly enveloped the other
in the usual way, thus showing that the nucleus is not necessary for
this reaction.

These results should be compared with Fenard's observations on
injured specimens of A* terricola^ noted on page 180 of the present
paper.

As to the question which these experiments were originally intended
to answer, whether two pieces of a single Amoeba would reunite
after separation, my results were negative. After the two pieces had
begun to move about freely they were induced to come in contact, or
sometimes they came in contact accidentally ; but in no case was there
any union. Prowazek (1901, p. 93) obtained the same result with
small species of Amoeba, but in a larger undetermined species he suc-
ceeded in bringing about a union of pieces not only from the same
individual, but from different individuals.

PHYSICAL THEORIES AND PHYSICAL IMITATIONS OF
AMCEBOID MOVEMENTS.

THE SURFACE TENSION THEORY.

The movements of Amoeba as presented by Biitschli (1880, 1892)
and Rhumbler (1898) (see Figs. 30-33) are exactly those of a drop of
fluid moving as a result of a local change in surface tension (Fig. 34).
It was, therefore, natural to assume that the cause of the movements is
the same in the two cases. This is the view taken by the two authors
named. According to Biitschli, Rhumbler, and many other authors,
Amoeba is a drop of complex fluid which moves about as a result of
local changes in surface tension.

THE MOVEMENTS AMD REACTIONS OF AMCEBA. 205

In the foregoing investigation it has been shown that the movements
in Amoeba are not of the character supposed by Biitschli and Rhumbler.
There is, indeed, very little resemblance betw^een the movements of
Amoeba and those of an inorganic drop moving as a result of a local
change in surface tension. The difference is clearly brought out by a
comparison of Fig. 58, showing the currents in Amoeba, with Fig. 34,
showing those in the inorganic drop. The more striking differences
are as follows :

(i) In the drop moving as a result of a local change in surface ten-
sion the currents on the surface are (and must be) away from the side
on which a projection is formed and toward which the drop is moving ;
in the Amoeba the surface current is toward this side.

(2) In the drop the surface currents are in a direction opposed to
that of the axial current ; in Amoeba surface currents and axial current
are in the same direction.

(3) The movement of Amoeba is in the nature of rolling, the upper
surface passing continually around the anterior end and becoming the
lower surface. In the inorganic drop there is no such rolling move-
ment, but the interior portions of the fluid are continually passing to
the surface at the anterior end.

Clearly, then, the forces producing the moyements in the two cases
are not acting in the same manner. The locomotion of Amoeba is
demonstrably not due to a local decrease in surface tension on the side
toward which the animal is moving.

This becomes still clearer when we consider in detail the method by
which the movements are produced in a drop of inorganic fluid as a
result of a local decrease in its surface tension.

The phenomena of surface tension are usually considered to be the
result of the uncompensated attractions of those particles of the fluid
which are near to the surface. Such particles are attracted inward and
laterally, but not outward (or less strongly outward) . The resulting
forces may be considered as resolvable into two components, one acting
tangent to the surface, the other acting perpendicular to the surface.
The former is what may be called surface tension proper ; the latter
is often spoken of as normal pressure. The result is very much as if
the fluid were covered with a stretched India rubber membrane.
These relations are well set forth in a recent paper of Jensen (1901).
It is further to be noted that these two components, surface tension and
normal pressure, are two aspects of one and the same thing, and, there-
fore, vary together and from the same causes ; they can not be separated
either theoretically or experimentally. Whenever one of these factors
increases or decreases, the other shows a corresponding change. The
two are often spoken of (together with the pressure due to curvature of
the surface) as surface tension (see Jensen, /. c).

2o6 THE BEHAVIOR OF LOWER ORGANISMS.

Now, when the interattraction of the particles at a certain region of
the surface of a mass of fluid is decreased, the pressure inward and the
tension along the surface are decreased. As the pressure remains the
same elsewhere, fluid tends to be pressed out at the point where the
pressure is lowered ; thus a projection may be formed here. As the
tension remains the same elsewhere, the remainder of the surface of
the drop pulls harder than that of the region under consideration ;
hence it pulls the surface of the fluid away from the region where the
tension is lowered. The eft'ect is similar to that which would be
produced if one portion of a stretched sheet of India rubber w^ere
weakened or cut ; the remainder of the sheet would pull away from this
region. Thus there are produced the currents characteristic of such a
drop of fluid — an axial current toward the region of lowered tension,
surface currents away from the region of lowered tension (Fig. 34).
An increase in the tension at the opposite side would produce exactly
the same currents, as Rhumbler (1898, p. 188) has set forth, the axial
current being always toward the region of lowest tension, the surface
currents in the opposite direction.

In the moving Amoeba, as we have seen, the currents are by no
means of this character. The axial current and the surface current are
congruent, and both are in the direction of locomotion. Such move-
ment could not be produced by a local decrease in the surface tension
of some part of the body surface.

The formation of pseudopodia is, as we have seen, essentially the
same process as the forward movement at the anterior end of the
Amoeba. On the upper surface of a pseudopodium that is in contact
with the substratum there is a forward movement, so that particles
clinging to the upper surface are carried over the tip ; the currents
which must result from a decrease in surface tension are not present.
On the contrary, there is a current on the surface in the opposite direc-
tion from that required. The formation of such a pseudopodium can
not, then, be due to a local decrease in surface tension.

The same is true, essentially, when a pseudopodium is sent out into
the water, not coming in contact with a surface. In such a case, as
we have seen, the entire surface moves outward, in the same direction
as the tip ; there is no such backward movement as the theory requires.

Altogether, it is clear that the supposed resemblance between the
movements and internal currents of Amoeba and those of a drop of fluid
moving as a result of a local increase or decrease of surface tension
does not exist. We must conclude that the movements of Amoeba are
not due to local changes in surface tension.

One might be tempted to inquire whether the movement of Amoeba
could not be explained by considering separately the action of the two

THE MOVEMENTS AXD REACTIONS OF AMCEBA. 207

factors in surface tension, the ''surface tension proper" and the
" normal pressure.'* If the normal pressure, directed inward, were
decreased in a certain region, while the tangential factor, the ''surface
tension proper," were not decreased, were perhaps even increased,
could not pseudopodia be formed as actually occurs, without any back-
ward current on the surface? Jensen seems to lean toward the possi-
bility of such action when he speaks of the variation of one of these
factors without the other (Jensen 1901, p. 374).*

But with such an inquiry should we not leave the field of realities to
wander among abstractions ? One who is not a physicist can, of course,
not speak positively on such a point. Yet, so far as I am able to dis-
cover from the results of experiments and from the theories of surface
tension, the state of the case is about as follows : The tangential tension
and the normal pressure are not two different things ; they are only
different aspects of one and the same thing. Viewed from the stand-
point of " energetics," what the physical experiments show liquids to
possess is surface energy, in virtue of which they tend to decrease their
surface and resist an increase of surface, however these changes are
brought about (see Ostwald, 1902, p. 197). When the " surface ten-
sion is decreased" in a certain region (x) of a fluid mass, this signifies that
the tendency to a decrease of surface and the resistance to an increase
of surface is lessened in this region. As a result, the remainder (y) of
the fluid decreases its surface at the expense of the region x; the latter
is thus compelled to increase its surface. This takes place by simulta-
neous passage of the contents of y into x and of the surface of a: on to
y, the two operations being essentially one, and both having the result
of decreasing the surface of ^ and increasing that of x. There would
seem to be no ground, theoretical or experimental, for supposing that
in a fluid one of these operations could take place without the other.
An attempted explanation of this sort would be, if these considerations
are correct, not a physical explanation, but a purely hypothetical one,
working with conditions not known to exist. The whole value of the
surface tension theory lies in its direct reference back to the results
of physical experiments — in its fidelity to the results of such experi-
ments. As soon as it leaves this ground it becomes of no more value
than the thousand and one other hypotheses that have been constructed
for the explanation of contractility.

Further, even this purely hypothetical explanation could not account
for the forward currents on the upper surface of Amoeba, nor for the
transference of portions of the body surface to the surface of a pseudo-
podium. In any form we can give it, the theory that the movement is
due to local changes in surface tension is not in agreement with the
observed phenomena.

* Though elsewhere he speaks of the necessity of their varying together.

2o8 THE BEHAVIOR OF LOWER ORGANISMS.

BERTHOLD's theory that ONE-SIDED ADHERENCE TO THE SUB-
STRATUM IS THE CAUSE OF LOCOMOTION.

If, then, the movements of Amoeba are not due to local decrease in sur-
face tension, with the formation of " Ausbreitungscentren " (Biitschli),
is it possible to find a physical explanation for them ? In taking up
this question we must consider separately (i) locomotion, and (2) the
formation of pseudopodia.

Observation and experiment indicate, as we have seen in the obser-
vational portion of this paper, that Amoeba is a drop of fluid which
becomes attached to the substratum in front and pulls itself forward,
the pull extending backward from the attached region over the upper
surface, and producing a rolling motion.

Now, a drop of inorganic fluid under the influence of similar
forces moves in precisely the same manner. There is no great difl[i-
culty in causing a drop of inorganic fluid to adhere more strongly to
the substratum on one side than elsewhere. When this is brought
about the drop moves toward the more adherent side by a rolling
motion, precisely like that of Amoeba. By a proper arrangement of
the conditions almost every detail of amoeboid locomotion may be
closely imitated.

That this is the method of movement in Amoeba was the theory
maintained by Berthold (1886), though it is rather curious that the
supposed facts on which he based this view were incorrect. Berthold
confirmed on the basis of observations on Amoeba verrucosa ( ! ) and
other species the account of the currents in Amoeba given by Schulze
(see p. 137 and Fig. 37) ; that is, such currents as would be consistent
with the theory of local decrease in surface tension, but are quite incon-
sistent with his own theory. He rejected the theory that locomotion is
due to a decrease in surface tension at the anterior end, on the ground
that no currents are to be observed in the surrounding water, as this
theory demands. Berthold held that the locomotion is due to the
spreading out of the anterior end of the fluid mass on the surface of a
solid, this spreading out being due to adhesion between the fluid and
the solid. Unfortunately for the understanding of his theory, he tried
to bring this into relation with many other much less simple phenom-
ena. In particular he compared the movements to those of a drop of
water on a glass plate, which flees when a rod wet with ether is brought
near one side. This was an unfortunate comparison, as the movements
in a drop of water under such circumstances are of a character entirely
different from those produced when a mass of fluid adheres by one side
to a solid. The movement in a drop of water fleeing from the ether-
ized rod is a result of the currents produced by the lowering of the

THE MOVEMENTS AND REACTIONS OF AMCEBA. 2O9

surface tension on one side, as Biitschli (1S92, pp. 191 and 194), has
shown. This comparison, and Berthold's discussion of the relations
between spreading out on the surface of solids and on the surface of
liquids, together with his incorrect idea of the currents in such spread-
ing out on a solid, have served to distract the attention of investigators
from the really simple essential features of such a theory. Berthold
did not attempt to study directly the currents and other movements of
a drop of fluid moving as a result of one-sided adherence to a solid.
We may, therefore, leave his account and examine for ourselves the
phenomena in question.

EXPERIMENTAL IMITATION OF THE LOCOMOTION OF AMCEBA.

The experiments with inorganic fluids may be performed as follows :
A piece of smooth cardboard, such as the Bristol board used for drawing,
is placed on the level bottom of a shallow vessel, such as a Petrie dish,
and soaked with bone oil by spreading the latter over its surface. A
small area on the surface of the board is protected from the oil by
placing upon it a drop of water. After the board has become well
soaked, the drop of water is removed with a pipette, leaving this spot
merely damp, while a layer of oil some millimeters deep is poured into
the vessel, covering the cardboard completely. A drop of glycerine or
of water is then introduced ; this settles to the bottom, but adheres to
it only slightly. A drop of glycerine is in some respects preferable, as
its movements are slower. To the drop should be added beforehand a
quantity of soot, in order to make its internal movements visible.
Some of the soot remains on the surface, projecting out into the oil,
thus making it possible to observe the surface currents.

If the drop is brought close to the spot on the cardboard that was
protected from the oil, so that one side comes in contact with this
region, the edge of the glycerine or water drop spreads out over this
area. Thereupon the remainder of the drop is pulled in that direction,
till the whole drop takes up its position over the protected spot. In
the movement of the drop toward the area to wliich one side adheres,
it rolls exactly as Amoeba does. The currents on the upper surface
and within the drop are forward. Toward the sides the currents are
somewhat less marked, and on the under surface they cease entirely;
particles within the drop but in contact with the lower surface are not
moved at all. The forward current is most rapid in front, becoming
slower at the rear, exactly as in Amoeba. At the posterior end the
surface rolls upward ; particles on the surface which were at first on
the bottom may be seen to pass upward around the posterior end and
then forward, as in Amoeba. The form of the drop may become much
elongated; the anterior edge is thin, the posterior end thick and
rounded. In all these respects the drop resembles the moving Amoeba.

2IO THE BEHAVIOR OF LOWER ORGANISMS.

By inclining the vessel the drop may be made to roll away from the
attractive spot; then when the level is restored it moves back again.
By repeating this process the movements may be studied in detail.
For studying the movements in all parts except at the anterior edge
another more convenient method may be employed. A small piece of
wood may be brought against one side of the drop ; toward this it
moves in the manner just described. If the piece of wood is moved
continually in a certain direction, the drop follows, and its movements
may be examined with ease. In this case the anterior edge, of course,
is not thin and pressed against the surface, but otherwise the move-
ments are the same.

By proper modifications further details of the movement of Amoeba
are exactly imitated. Thus a quantity of sand grains or other heavy
objects may be added to the drop. In the movement these collect at
the posterior end, as happens with the coarse internal contents in
Amoeba. A large, spherical bubble of oil may be introduced into the
drop, in imitation of the contractile vacuole ; this likewise stays near
the posterior end. When a considerable quantity of heavy material is
collected at the posterior end, the latter becomes drawn out into a sort
of pouch, which is dragged along, its substance not partaking of tlie
currents shown by the remainder of the drop. It thus plays the same
part as the well-known posterior appendage of Amoeba. Material
passes up from the bottom to the upper surface on each side of this
posterior pouch, just as happens in Amoeba (see p. i68). Particles
clinging to the outer surface at the posterior end are often dragged
along for a considerable time, then finally pass upward to the upper
surface and so forward, exactly as described for Amoeba (p. 169).*

In another detail the movements of the drop of water or glycerine
are strikingly like those of Amoeba. As we have seen, the current is
most rapid at the anterior end in both cases, becoming as a rule slow
toward the rear. But I have pointed out that in Amoeba the move-
ments at the posterior end are not uniform. Sometimes the under sur-
face remains attached to the bottom longer than usual, then, when it
becomes detached over a considerable area at once, there is a sudden
rush of the fluid forward from the posterior region (p. 16S). Exactly
the same thing is to be observed in the inorganic drop. The bottom
is not uniform, so that sometimes the posterior end clings to it longer
than usual ; this end is then drawn out, and when it is finally released

*It maybe worthwhile to state that these experiments on inorganic fluids
were performed after the work on the movements of Amoeba had been completed
and the description entirely written in the form given in the preceding pages.
No details of the movements of Amoeba were added after the behavior of the
inorganic drops had been studied.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 211

there is a sudden rush forward of the internal fluid from this region,
giving the movement a jerky character.

Still another resemblance in detail between the movements of the
inorganic drop and of Amoeba may be noted at times. As we have
seen, in the posterior part of Amoeba that is detached from the bottom
there is a movement forward not only on the upper surface, but also a
slow movement on the lower surface ; the entire posterior region is
contracting. The same thing may be seen in the inorganic drop. The
phenomenon in question is not so regular here, because the posterior
half usually still clings to the surface to a certain extent, while in Amoeba
it is as a rule entirely free. But when the posterior half of the inor-
ganic drop does become entirely free, it is seen to contract as a whole,
with a forward movement on both upper and lower surfaces, exactly
as in Amoeba.

One may even see at times, under special conditions, a slight turning
backward of the current at the sides of the anterior end, such as has
been described by a number of authors for Amoeba (see p. 137). This
occurs when the drop is slender and elongated, and the area on which
it spreads out is broad. On coming in contact with the area, the end
of the drop rushes forward and spreads out. If the whole width of the
area is not covered at first, some of the particles that have moved for-
ward curve outward and a little backward till the area is quite covered.

Altogether, the resemblance between the movements of the inor-
ganic drop and those of Amoeba is extraordinary, extending even to
details. What are the forces at work in such a drop, and in how far
may they be supposed to be active also in Amoeba?

The spreading out of the drop of glycerine or water at the anterior
end is due to its adherence here to the substratum. The remainder of
the movements of the inorganic drop are due to the interplay of surface
tension and adhesion to the substratum. As a result of surface tension
the drop seeks to regain its spherical form ; hence the posterior part is
pulled forward, the force required to accomplish this being less than
would be demanded for freeing the anterior edge from the substratum.
In the pulling forward of the posterior portion the adherence of the
lower surface to the bottom keeps this surface from moving ; hence the
upper surface moves forward while the lower surface remains quiet
or moves forward only very slowly ; the movement is thus converted
into a rolling motion. The details given above depend merely upon
the relative part played by adherence and surface tension, with the
resistance offered by the weight and inertia of particles inclosed in the
drop.

In Amoeba, so far as the evidence of observation goes, the conditions
are similar. Amoeba adheres to the substratum and spreads out in a

212 THE BEHAVIOR OF LOWER ORGANISMS.

similar manner. In one respect there is, of course, a striking differ-
ence between the two cases. Amoeba does not require that the sub-
stratum should be different on its two sides in order that there should
be movement ; on the contrary, it may move steadily in a certain direc-
tion on a uniform surface. The mechanism of the movement might,
nevertheless, be the same as in the inorganic drop. Chemically different
substances show different degrees of adherence to the same surface. It
may be supposed, therefore, that there is a chemical difference between
the anterior and posterior regions of Amoeba of such a nature that the
anterior region clings to the surface while the posterior region does not.
This chemical difference must, of course, be continually renewed, since
new parts of the body continually come in contact with the substratum.

It may, however, be questioned whether the adhesion of Amoeba to
the substratum is of the same character as the adhesion of a drop of
water to glass ; in other words, whether Amoeba really plays here the
part of a fluid, and " wets" the substratum. This was the view taken
by Berthold (1886) and, if I understand him correctly, Le Dantec
(1895). Apparently opposed to such a view is the fact that Amoeba
may creep on the under side of the surface film of water, as I have
often observed. This surface film is, of course, fluid ; if in adhesion
Amoeba itself also plays the part of a fluid, we should have two fluids
in contact, having the same relation of attraction or adhesion that a
fluid has for a solid that it " wets" ; that is, the particles of each fluid
have a greater attraction for those of the other fluid than for each other.
This, it would appear, could result only in the formation of diffusion
currents in the two fluids ; the two would mix. This result does not
follow, so that it would appear that in adhesion Amoeba does not play
simply the part of a fluid which wets the substratum. As we have
seen (p. 165), there is evidence that the adhesion takes place through
the mediation of a viscid secretion.

Whatever the nature of the adhesion, we know it exists at one pole
of the Amoeba and not at the other. Given such chemical differences
between the two poles as would produce this difference in adhesion,
then locomotion would follow essentially as we find it to occur in
Amoeba Umax or A. verrucosa. No further properties except those
common to fluids would be required.* For the determination of the
direction and rate of locomotion, the distribution of these chemical
differences would be the essential factor.

Caution is necessary, however, in transferring the results of these and
other similar experiments to Amoeba. The resemblances between the
movements of the inorganic drops and those of Amoeba show merely

♦It will be noted that this statement is made for simple locomotion, and does
not refer to the formation of pseudopodia.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 2I3

that the forces acting upon the two have a similar localization and
direction, not necessarily that the forces themselves are identical. This
caution is emphasized by the fact that drops moving down an inclined
plane as a result of the action of gravity have a similar rolling motion.
This is well shown in the drops of glycerine or water on the oiled
surface, in the experiments just described. Most (though not all) of
the details mentioned above, in which the movements of the inorganic
drop resemble those of Amoeba, may be observed also in drops moving
under the influence of gravity. The essential difference between the
two sets of experiments is that the action of adhesion, pulling the drop
in a certain direction in the one case, is replaced by the action of
gravity, pulling in the same direction, in the other case. Correspond-
ing with this difference is the chief difference to be observed in the
movements of the drops under the different conditions. In those mov-
ing as a result of greater adhesion on one side, the anterior edge is
thin and flat, as in Amoeba, while in those moving from the action of
gravity this is not true.

In Amoeba observation shows that we have the one-sided greater
adhesion, and the tendency of the lower surface to cling slightly to the
substratum, as in the first set of experiments with the inorganic drops.
There remains only something corresponding to the surface tension
factor, common to both sets of inorganic experiments, to be accounted
for in Amoeba. Since Amoeba acts like a fluid in many respects, there
is no a -priori reason to deny it surface tension, and nothing further is
required to produce locomotion. To this there is, however, one objec-
tion. This is found in the roughening and wrinkling of the surface
at the posterior end as it contracts, and in the similar roughening of a
contracting pseudopodium (pp. i6o, i68) . This is exactly the opposite
of what should take place in a fluid contracting as a result of surface ten-
sion. In such a case the primary phenomenon is the decrease in sur-
face ; the latter should, therefore, remain perfectly smooth, and as small
as possible. Of course, surface tension might be replaced in Amoeba
by a specific property of contractility of some sort, having its seat a
little beneath the surface. Locomotion would then take place as in
the inorganic drop, and the wrinkling of the outer surface would be
accounted for. On the other hand, if we can account for the contrac-
tility by a known property of fluids, such as surface tension, our expla-
nation will, of course, be simpler and more probable. By taking into
consideration the apparent fact that the outer layer of Amoeba is partly
fluid, partly solid, I believe that such an explanation, accounting for
the roughening as well as the contractility, can be given ; this I shall
attempt in the next section of this paper (p. 215).

The formation of projections at the anterior edge or side of the inor-
ganic drop, comparable to the formation of pseudopodia in contact with

214 '^^^ BftHAVlOR OF LOWER ORGANISMS.

the substratum in Amoeba, may also be induced in the oil drops. For
this purpose it is necessary to produce a greater adhesion on a small
area at one side. A projection is at once sent out here. The move-
ment in sending out such a projection is the same as that to be observed
in the formation of a pseudopodium under such circumstances. The
projection is thinnest at the tip ; its upper surface moves forward and
rolls over at the point, while the lower surface is at rest.

FORMATION OF FREE PSEUDOPODIA.

On the other hand, the projection of free pseudopodia into the water
cannot be imitated under these conditions. As we have seen, the move-
ment in the formation of a free pseudopodium differs from that in form-
ing a pseudopodium along a surface merely in the fact that in the latter
case the contact surface is at rest, while in the free pseudopodium all
surfaces move equally, a given point on the surface remaining approx-
imately at the same distance from the 'tip. In the inorganic drop pro-
jections can indeed be formed in which the surface moves in exactly
the same manner as in the free pseudopodia of Amoeba, but under con-
ditions that are essentially different. If some small object to which the
fluid adheres, such as a sliver of wood, is brought into contact with one
side of the drop, the fluid flows out over it, and may form thus a long,
slender projection. The surface of -this projection moves in the same
manner as the surface of a pseudopodium in Amoeba (p. 153). Thus
the surface of a free pseudopodium shows such movements as it would
if drawn out by an object to which it adheres at its tip. Since no such
object is present, it is clear that the formation of free pseudopodia is not
explicable in this manner.

As we have seen above, the locomotion and the formation of pseudo-
podia in contact with the substratum could, if they stood alone, be con-
sidered due to the adherence and spreading out of a fluid on a solid, as
maintained by Berthold (1886). But they do not stand alone ; we have
the additional fact that pseudopodia may be sent out which are not in
contact with the substratum. The anterior edge of an Amoeba, further,
may be pushed out freely into the water as a single pseudopodium.
This may frequently be seen in Ammba Umax. As a reaction to a
stimulus the protoplasm may push upward freely as a thick lobe, till
the greater part of the substance is transferred upward and the Amoeba
topples over (p. 184) . In the formation of such a lobe, the protoplasm
may flow from both ends toward the middle, producing the *' heaping
up'* from which the thick upward projection results (see p. 183).
Currents flowing in this manner could not possibly be produced by
adherence to the substratum. Again, in an advancing Amoeba angzi-
lata^ short triangular pseudopodia are constantly pushed forward, some
in contact with the substratum, others not thus in contact, but raised a

THK MOVEMENTS AND REACTIONS OF AMCEBA. 315

little above it (Fig. 54). Some, which are at first free, later come in
contact. Clearly, Amceifa is able to -perform all the activities con-
cerned in locomotion without adherence to a solid. The adherence
is only necessary that there may be a movement from place to place.
The case is quite parallel to that of higher organisms, where contact
with the substratum is necessary in order that progression may occur,
though all the movements concerned in locomotion may be performed
without such contact.

We are compelled to conclude, therefore, that in the advancing end of
an Amoeba or the projecting pseudopodium there is an active move-
ment of the protoplasm, of a sort which has not been physically
explained. This involves the general conclusion that no physical
explanation is at present possible of the locomotion and projection of
pseudopodia in Amoeba.

To account for the contraction of the posterior part of the body, on
the other hand, possibly the properties common to Amoeba with other
fluids are sufficient. If surface tension may be considered the cause of
the contraction of the posterior part of the body, it is notable that it
acts as a constant factor, tending always to decrease the surface as much
as possible, not as a variable factor. In other words, there is no indi-
cation that local increase or decrease in surface tension (see note, p.
325) plays any part in the production of the movements, as is main-
tained in the prevailing theories. The part played by surface tension
is thus a ver}' subordinate one.

EXPERIMENTAL IMITATION OF MOVEMENTS DUE TO LOCAL CON-
TRACTIONS OF THE ECTOSARC, AND OF THE ROUGHENING OF
THE ECTOSARC IN CONTRACTION.

Besides the sending out of pseudopodia, there are certain other phe-
nomena in the movements of Amoeba for which we lack, so far as I am
aware, any attempt at a physical explanation. These are the swinging
movements, vibrations, and local contractions of pseudopodia, described
on pages 177-179? and the roughening of the ectosarc in the contraction
of pseudopodia or other parts of the body (p. 16S).

These phenomena are in certain details so similar to some that I
have observed in inorganic fluids that I believe it worth while to
analyze the latter ; possibly they give an indication of the direction in
which an explanation of the phenomena in Amoeba above mentioned
may lie.*

*In view of the repeated failures of physical explanations in attempting to
account for vital phenomena, one does not approach a new attempt of this sort
with great confidence. Yet it is desirable that any possibility of this kind should
be worked out and submitted to criticism, in order that its truth or lack of truth
may be demonstrated.

2l6 THE BEHAVIOR OF LOWER ORGANISMS.

Swinging or bending movements take place with special frequency
while the pseudopodia are withdrawing ; in some Amoebse such move-
ments are an almost constant accompaniment of withdrawal. As it is
withdrawn the pseudopodium becomes roughened or warty on its sur-
face, as we have seen, and at the same time bends to one side or the
other, or swings back and forth. The impression given is that the
outer layer of the pseudopodium has become partially solid. In with-
drawing, the solid substance seems to melt gradually away, in a some-
what irregular manner, so as to leave solid masses connected by liquid
protoplasm, the projecting solid masses forming the wart-like roughen-
ings of the surface. When this melting away occurs more strongly on
one side, the pseudopodium bends at that point, toward the side which
has apparently become more fluid.

We have in such a case, if appearances may be trusted, a mass com-
posed partly of solid, partly of fluid. While it is usually admitted that
parts of the protoplasm may become solid at times, little attempt has
been made to understand protoplasmic movements by studying the
physics of such mixtures of solids and fluids.* In certain experiments
with inorganic mixtures of this kind, in which movements were pro-
duced that resembled those just referred to in Amoeba, the writer
became convinced of the possible importance of the physics of such
mixtures for the understanding of protoplasmic activities.

The experiments in question were concerned with the movements
under the action of surface tension of oil drops to which soot had been
added for the purpose of rendering the currents visible. When a large
quantity of soot was added, the drops became somewhat stiffened, and
now showed to a marked degree a mingling of the characteristic proper-
ties of fluids and solids.

In one set of experiments clove oil was thus mixed with soot and
introduced as drops into a mixture of three parts glycerine and one part
95 per cent alcohol. The drops move about, as a result of local
decrease in surface tension, in the same manner as the olive-oil emul-
sion in Butschli's celebrated experiments. Much of the soot collects
next to the surface of the drop, and becomes massed in certain regions,
as a result of the currents, covering these regions with a sort of crust,
this crust being formed of separate solid particles. The particles are
crowded together as closely as possible, owing to the surface tension of
the fluid in which they are floating.f If the particles are not too minute
they may project above the surface of the drop, giving it a rough appear-

♦ Some of the experiments of Rhumbler (1898, 1902) deal with such mixtures,
though not with a view to an understanding of the movements, but of certain
other processes.

t According to the principles set forth b_y Rhumbler, 1898, p. 332.

THE MOVEMENTS AND REACTIONS OF AMOEBA. 21 7

ance, similar to that of the withdrawing pseudopodium of Amoeba.
Such parts of drops or whole drops, so covered, show peculiar proper-
ties. The form may be changed by lowering the surface tension locally
or by mechanical action from outside, exactly as in a fluid, but there
is an inclination to hold a form once received. The tendency to take
the spherical form is still somewhat marked, and if the irregular drop
is strongly disturbed it frequently slowly becomes spherical. On the
other hand, if not strongly disturbed, it may retain almost any form
impressed upon it — cylindrical, flattened, irregular, or with long, slender
projections. In this power of receiving and retaining an irregular form,
yet with a tendency to become spherical, these drops, of course, resem-
ble Amoeba. These properties vary with the amount of soot present in
the oil ; if this is less, the drops slowly return to the spherical shape
when deformed ; if greater, they retain the irregular shape indefinitely.
Such irregular masses nevertheless flow together if brought in contact,
will quickly gather together into a close mass if strongly deformed,
and in many other ways they show the characteristics of fluids.

The reason for their tendency to retain irregular forms is obvious.
The surface is covered with small solid particles that are in contact.
The projection of these particles above the surface may cause a rough-
ening of the surface. Any change of form, such as surface tension
would produce, causes much friction between these particles. The
form taken is, then, a resultant of the action of surface tension and the
resistance of these particles to movements.* Similar forms are pro-
ducible by mixing soot with bone oil and studying drops of the mixture
in a vessel of glycerine.

For our purpose the phenomena which occur when the soot particles
are unequally distributed are of special interest. Consider an elon-
gated projection, as in Fig. 78, A, with the surface entirely covered
with closely crowded soot particles except in a certain region, x-y^ on
one side. In this region x-y surface tension will have free play, tend-
ing to draw the points x and y together, while elsewhere the tendency
to contraction will be resisted by the friction of the particles. The
result is that the points x-y are drawn together, and the projection
bends toward that side. Since this bending does not tend to crowd
the particles on other parts of the surface closer together they do not
resist it.

In the oil drops mixed with soot the bending of projections In the
manner described is often to be observed under the appropriate condi-
tions. They should be compared with the bending of pseudopodia as

* Similar forms of fluids have been produced by Rhumbler in a parallel man-
ner in his imitations of the formation of Difflugia shells (1898, p. 287) and of
the shapes of the shells of Foraminifera (1902, p. 265).

3l8 THE BEHAVIOR OF LOWER ORGANISMS.

described for Amoeba (p. 177, and Fig. 62, «) . The chief experimental
difficulty in producing such bending is to arrange the conditions in
such a way that there is less soot on one side of a projection than on
the other. This occurs somewhat frequently when two irregular drops
are allowed to fuse, or when a drop is mechanically deformed with a
rod, as described in the next paragraph. One can sometimes bring it
about by placing with the capillary pipette a minute drop of oil on one
side of a projection, though this method is not as a rule very effective.
'» Such drops may also show another prop-

,, erty in common with Amoeba, namely,
elasticity of form, or a phenomenon pro-
ducing similar results. Consider a curved
projection, as in Fig. 78, B. It is com-
pletely covered with soot particles and re-
tains its form in vi rtue of their resistance to a
change in position. Now, suppose we forcibly straighten out the pro-
section by pushing it to one side with a rod. By so doing the side a-6
is lengthened ; the soot particles on this side are, therefore, separated,
leaving certain areas of free fluid surface. When the projection is
released, surface tension can act on these areas, and the projection is
drawn back at once to its original form. I have often observed such
immediate returns to the original form after bending a projection of
one of the oil drops.

In Amoeba we have exactly the conditions most favorable for the
production of movements of this sort, and we actually find numerous
movements of just this character. It is generally admitted that the
outer layer becomes partially solidified ; as a pseudopodium is with-
drawn the solid portions evidently become liquefied in an irregular
way, some of them projecting above the surface and making it rough.
If the liquid substance produced shows surface tension, the movements
described must follow in the manner set forth above. It seems possible
that many of the observed movements are thus produced by local lique-
faction, with the intervention of surface tension, in the liquefied area.
In view of the apparently unlimited possibilities of partial solidifica-
tion and liquefaction in the protoplasmic body, with the resulting varied
action of surface tension, shall we not go a step farther and inquire
whether there may not be an outlook for an explanation of vibratory
movements, such as we find in flagella, along this line.'' In an elon-
gated structure like a flagellum, a limited liquefaction of one side would
result in a bending toward this side. By regular alternation of lique-
factions in different regions, a regular vibration could be produced.
The chief difficulty in the way of such a theory would seem to lie in
*FiG. 78. — Diagrams illustrating phenomena in mixtures of oil and soot.

THE MOVEMENTS AND IlEACTlOXS OF AMCEBA, 2X9

the restoration of the original length in a given side after liquefaction
and consequent contraction had occurred. This could, perhaps, be
brought about by an elastic rod in the axis of the structure, such as
many cilia and flagella are known to possess.

The above is merely a suggestion made tentatively ; its justification
as a suggestion lies in the following facts: (i) The swinging move-
ment of pseudopodia in Amoeba in some cases strikingly resembles
movements of the character above set forth in inorganic fluids, and
precisely the conditions for such movements are present in Amoeba ;
(2) Swinging movements of pseudopodia seem to grade almost insen-
sibly into the vibratory movements of flagella.

DIRECT OR INDIRECT ACTION OF EXTERXAL AGENTS IN
MODIFYING THE MOVEMENTS.

Is the effect of external agents in modifying the movements of Amoeba
due to the direct physical action of the agent on that part of the fluid
substance with which it comes in contact? Or is its action indirect,
in that it serves merely as a stimulus to certain internal changes, the
latter bringing about the modifications in the behavior? Both views
find adherents. The difference between them is fundamental, for they
lead to essentially different conceptions as to the nature of behavior in
these lower organisms.

In higher animals we know that the movements and changes of
movement are not produced in a direct way, but the effect of external
agents is to cause internal alterations which result in changes of move-
ment. It is not, therefore, possible to predict the movements of the
organism from a knowledge of the direct physical changes produced in
its substance by the agent in question. If the theory of direct action is
correct for Amoeba, we have in these animals a condition of affairs incom-
parably simpler, for here we can resolve the behavior directly into its
physical factors. If, on the other hand, the theory of indirect action is
correct, then there appears to be nothing fundamentally different in
principle between the behavior of Amoeba and that of higher organisms.

How the form and movement of a fluid mass might be determined
by the direct action of external agents on its surface may be simply
illustrated by certain experiments which I have described elsewhere
(Jennings, 1902). A mixture of 2 parts glycerine and i part 95 per
cent alcohol is placed on a slide and covered with a cover glass sup-
ported by glass rods. Into this is introduced with a capillary pipette
a drop of clove oil. The clove-oil drop, at first circular in form, soon
changes shape, shows internal currents, sends out projections in various
directions, moves about from place to place, and may divide into two
drops. The alcohol, not being uniformly distributed throughout the
glycerine, acts more strongly on some parts of the clove-oil drop than

220 THE BEHAVIOR OF LOWER ORGANISMS.

on others, thus lowering the surface tension in the region acted upon.
Thereupon the drop sends out a projection on the side affected, and may
follow this up by moving in that direction. We may cause the drop to
send out projections and move in a certain direction by placing a
minute drop of alcohol near one side, or by heating one side ; move-
ment takes place toward the side affected. These agents act directly
on the surface of the drop, lowering the tension ; the movements are a
direct consequence of this change. Parallel phenomena maybe pro-
duced, as Bernstein (1900) has shown, with a drop of quicksilver. If
such a drop is placed in 10 per cent nitric acid, in which bichromate
of potash has been dissolved, the drop changes form and moves about.
If we place near such a drop, in vessel of 10 per cent nitric acid, a
crystal of potassium bichromate, the mercury drop moves rapidly over
to the crystal. Here, again, the chemical acts directly on the mercury,
lowering the surface tension at the region where it comes in contact
with it, thus producing the movement.

Certain authors have held that the movements of Amoeba are pro-
duced in this way by the direct action of external agents decreasing
or increasing the surface tension of certain parts of the fluid mass. As
an example of the theory of direct action of external agents in control-
ling the behavior, we may take the view of the reaction of Amceba to
chemicals recently given by Rhumbler (1902, p. 384). According to
Rhumbler, it is evident that when an Amoeba moves toward or away
from a certain chemical, the side directed toward the chemical has, in
the first case, a lessened surface tension, in the second case an increased
surface tension, as compared with the remainder of the body.

The necessary differences of tension on the positive and negative sides may be
easily understood from our present standpoint, by holding that a positively
acting chemical decreases the surface tension both in the living alveolar system
of the cell and especially on the cell surface, upon which it must work most
strongly; that a negatively acting chemical, on the other hand, produces an
increase of surface tension in the alveoli of the cell, and especially, again, on the
cell surface; this increase is the greater, and from a physical standpoint must be
the greater, the more the molecules of the chemical affect or modify the tension
of the different cell alveoli or different parts of the cell surface (/. c, p. 384).

The explanation of thermotaxis and electrotaxis would be, according
to Rhumbler, "exactly the same as for chemotaxis" (/. c, p. 385) ;
thus also as a result of the direct action of external agents. A fuller
explanation of the " tropisms " on this basis is given by Rhumbler in
an earlier paper (1898, pp. 183, 188).*

* Rhumbler emphasizes in the paper just cited (1898, p. 184) the importance
of " inner disposition " in deciding what effect shall be produced by external
agents, but in the tropisms, at least, he considers the action of the external agent
to be direct, the inner disposition deciding merely whether the substance of the
Amceba is of such a character as to admit of the production of a given definite
change in surface tension by the outer agent.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 221

Based similarly on a direct action of external agents is the theory of
amoeboid movements proposed by Verworn in 1891, and extended in
his paper on Die Bewegung der lebendigen Substanz (1892) and in
the Allgetneine Physiologie. In its original form Verworn's theory
considers the movements and changes of form to be brought about
directly through chemical attraction (Verworn, 1891, p. 105), but in
later publications (1892, 1895) the effect of the chemical is considered,
as in Rhumbler's theory, to be that of increasing or decreasing the
surface tension.

Thus it is evident that it must be the chemical affinity of certain parts of the
protoplasm for oxygen that decreases the surface tension in definite regions and
thus leads to the formation of pseudopodia. But it will be possible for the same
effect to be produced by other substances of the surrounding medium, if they
have chemical affinity for certain components of the protoplasm. In the case
where the substance acts from one side, this principle must lead to positive
chemotropism. (Vervvrorn, 1895, p. 545.)

It is evident that the method of movement of Amoeba, as described
in this paper, has an immediate bearing on the question of direct or
indirect action of external agents. If the action of an external agent
is to increase or decrease directly the surface tension, as set forth by
Rhumbler and Verworn, this effect must be shown in the characteris-
tic currents which appear in any fluid when the surface tension is thus
locally changed. In the case of negative chemotaxis we should have
an axial current away from the side affected, with surface currents
toward the chemical, as indicated in the figure given by Rhumbler
(1898, p. 188). In positive chemotaxis both sets of currents should be
the reverse of that just indicated.

In the account of the movements set forth by Biitschli and Rhumbler,
the currents agreed with the scheme for direct action above set forth.
But this account of the movements was erroneous, as we have seen.
The internal currents and the surface currents are forward, away from
the region stimulated, in a negative reaction ; toward the region stim-
ulated in a positive reaction ; the movement is of a rolling character.
There is thus no evidence that the action of the stimulus is to cause a
change in the surface tension of the parts directly affected ; on the con-
trary, the direction of the currents is quite inconsistent with this view.*
We must conclude, then, that the theory of the direct action of exter-
nal agents in causing or changing the movements of Amoeba is nega-
tived by the character of the movements produced ; these are not such as
would follow from the direct physical action of the agents in question.

* A description of the forward surface currents in negative chemotaxis is given
on p. 143; in the reaction to a mechanical stimulus on p. 185; to the electrical
stimulus on p. 192 ; in a positive food reaction (chemical and mechanical stimuli ?)
on p. 198.

222 THE BEHAVIOR OF LOWER ORGANISMS.

There is much indirect evidence that points in the same direction,
particularly in the fact that the activities of the animal may remain con-
stant while the environment is continually chanj^^ing. Some of these
lines of evidence are summarized by Rhumbler (1898, p. 185). They
still leave open the possibility that when the environment does modify
the movements its action is direct. But even this is shown to be
excluded by the nature of the currents produced, as described above.

The currents as they actually occur in the movements are equally
opposed to certain theories of the indirect action of stimuli. Bernstein
(1900), Jensen (1901), and others, have expressed the opinion that the
effect of stimuli is to change the surface tension, but that this effect is
not due to the direct physical action of the agent on the protoplasm,
but rather to some change in the internal physiological processes of the
cell produced by the agent acting as a stimulus. Jensen (1901, 1902)
has developed this view into a detailed theory, according to which
stimuli that increase the normal assimilatory processes of the cell lead
to a reduction of surface tension, and hence to expansion and move-
ment toward the agent in question, while stimuli that increase the
dissimilatory processes have the opposite effect.

The currents in the moving Amoeba lend no support to this view.
There is no evidence in the movement that the effect of a stimulus is to
alter the surface tension in any way. In view of the facts given in the
body of this paper as to the nature of the movements, we are forced to
give up the idea that the effect of stimuli is to modify the tension* of
the surface of the protoplasmic mass, either directly or indirectly.
Alterations in the tension of the surface can no longer be considered
the prime factor in the behavior of Amoeba.

DIRECT OR INDIRECT ACTION IN THE TAKING OF FOOD.

Rhumbler, in his most interesting and suggestive paper (189S), has
attempted to give a physical analysis of food-taking and the choice of
food in Amoeba. According to Rhumbler, the taking of food is due
to adhesion between the protoplasm and the food substance, and may
be compared with the pulling inward of a splinter of wood by a drop of
water, or of a bit of shellac by a drop of chloroform. The selection of
food is explained as due to the fact that the protoplasm, as might be
expected from physical considerations, tends to adhere to some sub-
stances and not to others. Parallel phenomena are shown, in a most
ingenious experiment, to be demonstrable for the chloroform drop
(/. c, p. 248). It takes in certain substances, while others are refused
or thrown out if introduced.

♦See note, p. 235.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 223

This theory is a most attractive one, and seems a priori probable.*
It is conceivable that there may be, or may have been, organisms where
it is applicable throughout. But an objective study of the behavior of
Amoeba show^s that it gives by no means an adequate explanation of
food-taking in this animal. As I have shown in the descriptive por-
tion, in Ammba proteus and A. angulata the food in most cases is far
from adhering to the protoplasm ; on the contrary, it rolls away when
the Amoeba comes in contact with it, and it is often only as a result of
long-continued effort that the animal succeeds in ingesting it. The
first act in ingestion consists in sending out pseudopodia on each side of
the mass to overcome the mechanical difficulty resulting from the fact
that the body does ;?d?/ adhere to the protoplasm, but tends to roll away.

Further, a quantity of water is usually, or frequently, taken in with
the food, and the latter floats about in a cavity after it is ingested, show-
ing no tendency to adhere to the protoplasm (see Leidy, 1879, numer-
ous figures of food vacuoles, etc., and Le Dantec, 1S94). A similar
condition of affairs is shown in the account of the feeding of one
Amoeba on another, given on page 201 of the present paper. Here the
prey does not adhere to the protoplasm of its captor, but moves about
within the latter and escapes repeatedly.

Thus, in these species, the taking of food and the choice of food can-
not be explained by the adherence of the protoplasm to the food sub-
stance, for the lack of such adherence is strikingly evident.

Rhumbler's studies of food taking were made chiefly on Animba
verrucosa. In this species and its close relatives there is much more
tendency for foreign objects to cling to the surface than in the other
species. But this adhesiveness applies to other objects as well as to
food. It is of special aid, as we have seen, in tracing surface move-
ments (p. 140). Particles of soot and various other indifierent bodies
stick to the surface, rendering its movements apparent. Not all such
adhering bodies are taken into the interior, so that the ingestion
involves an additional reaction, and is not fully accounted for by the
adhesion even in these species.

Rhumbler has given an ingenious physical analysis of the rolling up
and taking in of filaments of Oscillaria by Amoeba verrucosa^ and has
illustrated the process as he conceives it to occur by a very striking
experiment (1S98, p. 230). A chloroform drop brought in contact
with the middle of a filament of shellac rolls the filament together and
encloses it. Rhumbler conceives the forces at work in rolling up the
filament to be essentially the same in the chloroform drop and in the
Amoeba. In both cases, according to Rhumbler, the surface tension of

♦See Jennings, 1902, where I adopted this view before having investigated for
nnyself the behavior of Amoeba.

224 THE BEHAVIOR OF LOWER ORGANISMS.

the drop pulls on the filament, tending to force it inward from both
directions. That part of the filament within the drop becomes softened
by the action of the fluid ; it, therefore, yields to the thrust from both
directions, and bends, permitting more of the filament to be brought
into the drop by the action of surface tension. (For full explanation,
see the original.)

It is necessary to point out that the explanation of the rolling up of
the shellac filament given by Rhumbler is erroneous. The surface
tension of the drop, with its inward thrust, has nothing to do with the
process, for the filament is rolled up in exactly the same manner when
it is completely submerged in a large vessel of chloroform, so that it is
not in contact with the surface film at all. The rolling up is evidently
due in some way to the strains within the shellac filament, produced
when it was drawn out, and to the adhesiveness of its surface when
acted upon by the chloroform. The process thus loses all similarity
to the rolling up of the alga filament by Amoeba. The coil formed is
just as small and close, and the filament remains a filament just as long
when the experiment is tried in a large vessel of chloroform as when
only a drop is used, as in Rhumbler's experiments.

Rhumbler's explanation of the way in which Amoeba rolls up the
Oscillaria filament may, of course, still be correct, though the physical
experiment by which he attempted to illustrate it has nothing to do
with the matter. There are certain points in his description of the pro-
cess as it occurs in Amoeba, however, that might easily be interpreted
in another manner. Such a bending over of the pseudopodium as is
shown in Rhumbler's Fig. 58 (/. c, p. 233) is not called for by the
surface-tension theor)-. Rhumbler holds that this bending of the pseu-
dopodium is passive, and due to the bending of the filament within the
body (/.<:., p. 233). In view of what we have shown above (pp. 177-179)
as to the power of active bending in the pseudopodia, and as to active
contractions of parts of the ectosarc in this same species (pp. 179, iSo),
one might be inclined to believe rather that this bending of the pseu-
dopodium is active and plays an important part in bringing the fila-
ment into the body. Rhumbler's figures (Fig. 50) would support this
view fully as strongly as his own theory, though this would, of course,
not give us a simple physical explanation of the ingestion of the filament.

Altogether, we must conclude that adhesion between the protoplasm
and the food substance cannot by any means give us a general explana-
tion of food-taking in Amoeba. In some cases the ingestion of food
is aided by such adhesion, but in other cases the adhesion is conspicu-
ously absent.*

* For further confirmation of last stated fact, see paper of Le Dantec ('1894).

THE MOVEMENTS AND REACTIONS OF AMCEBA. 225

GENERAL CONCLUSION.

Putting all our results together, we must conclude that the move-
ments and reactions of Amoeba have as yet by no means been resolved
into their physical components. Amoeba is a drop of fluid which
moves in its usual locomotion in much the same way as inorganic drops
move under the influence of similarly directed forces. But what these
forces are is by no means clear. When we take into considera-
tion the currents as they actually exist, local decrease in surface tension
breaks down completely as an explanation for the locomotion and other
movements. The locomotion taken by itself might be explained as due
to the adhesion of the fluid protoplasm to solids, taken in connection
with the surface tension of the fluid, but this explanation fails when we
consider the formation of free pseudopodia, and discover that all the
processes concerned in locomotion can take place without adhesion to
the substratum.

For the reactions to stimuli we find a parallel condition of affairs.
The currents in the protoplasm in the positive and negative reactions
are not similar to those produced in the attraction or repulsion of drops
of fluid by the direct action of external agents. Therefore we cannot
consider these reactions as due to the increase or decrease of surface
tension* produced by the direct (or even indirect) action of the external
agents. The taking and choice of food cannot be physically explained
in any general way by the physical adherence of the protoplasm as a
substance to the food as a substance, for food is taken in many cases
(usually, in some species) where it is demonstrable that no such
adherence exists.

While we must agree that Amoeba, as a drop of fluid, is a marvel-
lously simple organism, we are compelled, I believe, to hold that it has
many traits which are comparable to the ''reflexes" or "habits" of
higher organisms. f VVe may, perhaps, have faith that such traits are

♦It should be pointed out that this and other statements concerning surface
tension in Amoeba apply to the tension of the actual body surface, comparing
Amoeba thus to a drop of simple fluid. This is the basis on which rest the pre-
vailing theories that would explain the movements of AmcBba by surface tension.
It is these theories which I desired to test. There remains untouched, of course,
the possibility that the movements of all sorts of protoplasmic masses may be
explained by changes in the surface tension of the meshes of Butschli's honey-
comb structure, in the manner indicated by Biitschli (1892, p. 208). But this is
at present merely a hypothesis, not worked out and not controllable by observa-
tion. To attempt to maintain it for Amoeba would be to relegate the movements
of this animal to the same obscure category as the movements of cilia and of
muscles, possibly a correct proceeding, but removing the matter at present from
the field of experimental observation.

t See the next division of this paper, where this point is developed.

226 THE BEHAVIOR OF LOWER ORGANISMS.

finally resolvable into the action of chemical and physical laws, but we
must admit that we have not arrived at this goal even for the simpler
activities of Amoeba.

THE BEHAVIOR OF AMCEBA FROM THE STANDPOINT OF THE
COMPARATIVE STUDY OF ANIMAL BEHAVIOR.

HABITS IN AMCEBA.

Although in general Amoeba has the rolling movement of a drop of
fluid, yet this statement by no means brings out all the characteristics
of the movement in any given species of Amoeba. Different kinds of
Amoebae move differently, and the differences are in many cases not
such as can be accounted for by differences in the state of aggregation
of the body substance. Some Amoebae, as is well known, form many
pseudopodia, others few or none. Different Amoebae have different
characteristic forms in locomotion. But more striking than these gen-
erally recognized peculiarities are certain others of a more special char-
acter. A creeping Amoeba angulata^ as we have seen above, frequently
pushes upward and forward at the anterior end a short, acute pseudo-
podium, which waves slightly from side to side like an antenna (p. 177
and Fig. 62, c). This peculiar habit is much more pronounced in
Amoeba velata Parona, according to Penard (1902). In this animal
the free anterior pseudopodium may extend for a length greater than
the diameter of the body ; Penard compares it directly to a tentacle.
Some other species of Amoeba never send forward such an antenna-like
pseudopodium. The great work of Penard (/. c.) contains innumer-
able instances of such peculiarities of form, movement, and function
among the different species of Amoeba and other Rhizopods ; some of
them are collected in that author's interesting section on the pseudo-
podia (/. c, pp. 625-629). It is not necessary to take these up in detail
here. \ The point of interest is that different sorts of Amoebae have dif-
ferent customary methods of action, such as are commonly spoken of
as " habits"* in higher animals, and that these *' habits" are no more
easily explicable on direct physical grounds in Amoeba than in higher
animals. Let anyone attempt, for example, to explain from the
physics of viscous fluids why Amoeba velata or A. angulata push
out an antenna-like pseudopodium at the anterior end and wave it from
side to side, while Amoeba proteus and A, Umax do not.

* The word habit is, of course, not used here of a method of action acquired
during the life of the individual, but merely of a fixed method of behavior. At
all events, it is difficult to distinguish between these two things where individual
organisms, as in Amoeba, have lived as long as the race.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 227

CLASSES OF STIMULI TO WHICH AMCEBA REACTS.

The simple naked mass of protoplasm reacts to all classes of stimuli
to which higher animals react (if we consider the auditory stimulus
merely a special case of the mechanical stimulus). Mechanical stimuli,
chemical stimuli, temperature differences, light, and electricity — all
control the direction of movement, as they do in higher animals.

TYPES OF REACTION.

Amoeba has two chief types of reaction, by one or the other of which
it responds to most stimuli. These we may call the positive and the
negative reactions. As a third type we must distinguish the food
reaction, which cannot be brought completely under either of the two
chief types of reaction above mentioned.

(i) The positive reaction consists in pushing out the body substance
toward the source of stimulus and rolling in that direction.

(2) The negative reaction consists in withdrawal of body substance
and rolling in some other direction — not necessarily in the opposite
direction.

(3) The food reaction is not sharply definable. Its most character-
istic features consist in the hollowing out of the anterior end and in
the pushing out of pseudopodia at each side of and over the food body.
It involves also the positive reaction above characterized.

RELATION OF THE DIFFERENT REACTIONS TO DIFFERENT STIMULI ;
ADAPTATION IN THE BEHAVIOR OF AMCEBA.

(a) The positive reaction is known to be produced by weak mechanical
stimuli ; it is probably produced also by weak chemical stimuli (in the
reactions to food, pp. 193-202). The positive reaction to weak mechan-
ical stimuli serves the purpose of bringing the floating animal to a sur-
face on which it can creep. The positive reaction to food substances
(mechanical and chemical stimuli), of course, serves to obtain food.
The positive reaction is thus, as a rule, performed under such circum-
stances as to be beneficial to the organism ; z. e., it is directly adaptive.

(6) The negative reaction is produced by powerful stimuli of all
sorts. Such powerful stimuli are, as a rule, injurious, and the nega-
tive reaction tends to remove the Amoeba from their action ; it is, there-
fore, directly adaptive. This is true of the negative reaction to light
as well as to other stimuli, for light is known to interfere with the activi-
ties of Amoeba. \The reaction to the electric current is of exactly the
character that would be produced by a strong stimulus on the anode
side, but owing to the peculiarity of the current the reaction does not
assist the Amoeba to escape. The reaction to the electric current can
not then be considered adaptive ; this stimulus forms, of course, no
part of the normal environment of an Amoeba.,

22^ THE BEHAVIOR OF LOWER ORGANISMS.

The method by which, througfh the negative reaction, Amoeba avoids
an injurious agent is of interest. The animal does not go directly away
from the injurious agent, as by moving toward the side opposite that
stimulated. It moves in any direction except toward the region stimu-
lated. There is no system of conduction such that strong stimulation in
a given spot involves movement directed toward the opposite side. If
the movement in the new direction induces a new stimulation, the direc-
tion is again changed. This may be continued until the original direction
of movements is squarely reversed. The method is akin to that of trial
and error in higher organisms (see Morgan, 1S94, pp. 241-242).;

(c) The food reaction is directly adaptive in that it procures food.
As a rule this reaction occurs only when the source of stimulation is
fitted to serve as food. Empty diatom shells, sand grains, debris, etc.,
are, as a rule, not taken into the body, as many observers have pointed
out. Sometimes material is taken into the body that is not useful, as
is described by Rhumbler (1898, p. 236). In such cases there is no
evidence of a food reaction in the sense characterized above ; the
material is ingested accidentally, as it were, through its adherence to
the protoplasm. The food reaction, as a definite form of behavior, is
always adaptive so far as known.

(d) Some of the habits of Amoeba, characterized above, are clearly
adaptive. The use of the antenna-like pseudopodium sent out by
Amoeba velata and A, a7tgulata is evident. Penard describes in detail
how Amoeba velata uses it in passing from one substratum to another.

The habit which some Amoebae have, when suspended freely in the
water, of sending out pseudopodia in all directions (p. 181) is, of course,
useful in that it increases the chances of coming in contact with some
solid object, without which the Amoeba cannot move from place to
place.

REFLEXES AND " AUTOMATIC ACTIONS " IN AMCEBA.

In the behavior of Amoeba we can distinguish factors directly com-
parable to the reflexes and '* automatic activities " of higher organisms.
The responses of Amoeba to stimuli have the nature of reflexes in the
fact that they are not direct efl^ects of the physical action of the stimulus
(see p. 219), but are determined by the internal conditions of the
organism. They may be called reflexes, unless we propose, as certain
writers do, to restrict the term reflex to processes involving difleren-
tiated nerves. The precise designation is unimportant ; the essential
point is that the responses agree with the reflexes of higher animals in
being indirect.

Ziehen, in his JLeitfaden der physiologischen Psychologie (sixth
edition, p. 10), defines as automatic acts " motor reactions, which do

THE MOVEMENTS AND REACTIONS OF AMCEBA. 229

not, like the reflexes, follow unchangeably upon a definite stimulus,
but which are modified in their course by new, intercurrent stimuli."
In this sense Amoeba, of course, shows automatic behavior. Its
responses are by no means unchangeably fixed ; on the contrary, its
behavior is often modified by the slightest change in the stimulus to
which it is reacting. For examples of this see the chase of one Amoeba
by another (p. 200), the following of a rolling ball of food (p. 196), the
account of the driving of Amoeba (p. 185), and the description of the
method by which Ama^ba avoids an obstacle (p. 186).

Whether these actions agree with the accepted idea of an automatic
action in being unconscious we have, of course, no means of knowing.

VARIABILITY AND MODIFIABILITY OF REACTIONS.

JThere is little that can be said on this point. Verworn (1890, «, p.
271) says that when an electric current is passed through a prepara-
tion containing many Amoebae, some respond strongly, while others
do not ; thus difterent individuals vary in their responsiveness. Fur-
ther, a given individual may become accustomed to the current, at first
responding to it, later not responding. Doubtless such phenomena of
acclimation are common in the reactions to all sorts of stimuli.

Rhumbler (1898, p. 203) shows that when Amoebas are engaged in
taking Oscillaria filaments as food, light thrown upon them modifies
them physiologically in such a way that they eject the food. The
nature of the reaction is thus shown to depend partly on the physio-
logical condition of the animal.

There is no direct experimental evidence as yet, so far as I am aware,
that Amoeba shows memory.* Experimental evidence as to whether
the reactions of a given Amoeba to a given stimulus are modified by
previous stimuli received is very difficult to obtain, principally because
it is practically impossible to make succeeding stimuli alike, so that
one cannot tell whether a difference in the reaction is due to a difier-
ence in the present stimulus or not. Possibly there is a faint indication
of something akin to memory shown in the facts described on page 201.
Here a smaller Amoeba which had been ingested as prey escaped from
the posterior end of the captor ; the latter thereupon reversed its move-
ments, came up with the escaping prey, and again ingested it. In the
interval between the complete separation of the prey from its captor
and its recapture, the behavior of the captor would seem to have been
determined by some trace left within it by the former possession of the

♦The word memorj is, of course, used here of the objective phenomenon that
in many animals present behavior is modified in accordance with past stimuli
received, or past reactions given. Of possible subjective accompaniments of this
objective phenomenon we, of course, know nothing directlj to far as the lower
organisms are concerned.

23© THE BEHAVIOR OF LOWER ORGANISMS.

prey. But, possibly, this trace was merely of a gross physical char-
acter, acting as a direct stimulus to produce the observed behavior. If
this is true, the behavior shows no indication of memory.

SUMMARY.

OBSERVATIONS.
THE USUAL MOVEMENTS.

(i) Locomotion in Amoeba is a process that may be compared with
rolling, the upper and lower surfaces continually interchanging posi-
tions. This is shown by observation of the movements of particles
attached to the outer surface or embedded in the ectosarc of the animal.
Such attached particles move forward on the upper surface and over
the anterior edge, remain quiet on the under surface till the body of the
Amoeba has passed, then pass upward at the posterior end and forward
on the upper surface again. Single particles may thus be observed to
make many complete revolutions. (See p. 170, Fig. 58, and Figs. 38, 39,
40, 41.) *

(2) Thus the upper surface moves forward in the same direction as
the internal currents, while the lower surface is at rest. There is char-
acteristically no backward current anywhere in Amoeba, though at
times some of the endoplasmic particles, spreading out laterally at the
anterior end, may move a slight distance backward at the sides. This
is rare (see p. 134). The forward current on the upper surface is not
confined to a thin layer, but extends inward to the endosarc ; the endo-
sarcal and surface currents are one (p. 142).

(3) In the formation of pseudopodia that are in contact with the
substratum the movement of protoplasm is identical with that at the
anterior end of the Amoeba. The upper surface and internal contents
flow toward the tip, while the surface in contact with the substratum is
quiet. Particles adhering to the upper surface are carried out to the tip
and rolled under to the lower surface, where they remain quiet (p. 152).

(4) In the formation of pseudopodia projecting freely into the water,
the movements of substance are the same as in pseudopodia that are in
contact, save that there is no part of the surface at rest. The whole

* Of anyone who is inclined to reject these results on the basis of previous
observations, or of their supposed incompatibility with other known facts, let
me make the following request: Before taking ground against the results, pro-
cure some specimens of Amoeba verrucosa or one of its relatives. This is usually
easily done. Then mix thoroughly with the water containing them some fine
soot, and observe carefully the movements of the animals. The particles attach
themselves to the outside, and the movements of the surface are then observable
with the greatest ease. It is such a simple matter to determine certain of the
chief points for one's self in this manner that it would be regrettable for contro-
versy to arise through neglect of the needed observations.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 231

surface thus moves outward, and new parts of the surface of the body
continually pass on to the pseudopodium. An object adhering to the
surface of the pseudopodium remains at approximately the same dis-
tance from the tip, both when the pseudopodium is short and when it
has become very long (pp. 153-156, and Figs. 47-49).

(5) In the withdrawal of a free pseudopodium {a) a process occurs
that is the reverse of that described in (4), the surface of the pseudo-
podium passing from its base on to the surface of the body (p. 156,
and Fig. 49) ; (3) the surface of the pseudopodium becomes wrinkled
and shrunken ; {c) the endosarc flows back into the body.

(6) Any part of the protoplasm may be excluded temporarily from
the forward currents. In many Amoebae there is usually a region at
the posterior end which is thus temporarily excluded (the posterior
appendage, tail). In such cases the lower surface of the Amoeba
passes upward on each side of this appendage to become part of the
upper surface, then passes forward (p. 169, and Fig. 57). The sub-
stance of the posterior appendage is itself gradually drawn into the
forward current.

(7) The anterior portion of the advancing Amoeba is attached to the
substratum, while the posterior portion isnot (p. 165). There is a viscid
secretion produced on the outer surface of the Amoeba, to which the
attachment may be due.

(8) The attached anterior portion of the body is spread out and
usually very thin. The unattached posterior portion becomes rounded
and thick, and is contracting, so that there is a slight forward move-
ment on the lower surface, as well as on the upper surface, in this part
(p. 166).

(9) All the activities concerned in locomotion can be performed when
the animal is not attached to the substratum. (But for progression such
attachment is necessary, p. 215.)

(10) The locomotion of Amoeba is similar even in details to the
movements of a drop of inorganic fluid which adheres strongly to the
substratum at one edge and spreads out upon it here, while the other
edge is free (pp. 209-214). It is similar in most respects (except in the
thinness of the anterior edge) to the movements under the influence of
gravity of a drop of fluid along an inclined surface to which it adheres
but slightly.

(11) The currents in a moving Amoeba are not similar to those of a
drop of inorganic fluid that is moving or elongating as a result of a local
increase or decrease in surface tension. The surface currents away
from the region of least tension and in the opposite direction to the
axial currents that are characteristic of such a drop are lacking in
Amoeba. Here surface and axial currents have the same direction
(p. 205).

2^2 THE BEHAVIOR OF LOWER ORGANISMS.

(12) Similarly, the movements of material in a forming pseudo-
podium are not like those in a projection which is produced in a drop
of inorganic fluid as a result of a local decrease in surface tension. The
surface currents away from the tip in the inorganic projection are lack-
ing in Amoeba, the surface here moving in the same direction as the
tip (p. 206).

(13) The body of Amoeba shows under some conditions elasticity
of form, such as is characteristic of solids (pp. 175-177)-

(14) Besides the movements directly, concerned in the usual loco-
motion, limited local contractions may occur, resulting in swinging or
vibratory motions of the pseudopodia (p. 177), or in sudden, jerky
movements of the body as a whole (p. 179), or in the constricting off
of parts of the body (p. 180).

(15) The roughness of surface in a retracting pseudopodium, the
retention of irregular forms by Amoeba, and the movements mentioned
in the foregoing paragraph, are similar to certain phenomena to be
observed in mixtures of solids and fluids, as a result of the interaction
of surface tension and the friction of the solid particles (p. 215).

REACTIONS TO STIMULI.

(16) The following reactions are described : Positive and negative
reactions to mechanical stimuli (pp. 1S1-187) ; negative reactions to
chemical stimuli (pp. 187-190); negative reaction to heat (pp. 190-191);
reaction to the constant electric current (pp. 191-192); complex reac-
tions connected with food-taking (pp. 193-202); reactions to injuries
(pp. 202-204).

(17) In most reactions modifying the direction of motion a new
advancing wave of protoplasm is sent out from some part of that por-
tion of the body which is already attached to the substratum. The
internal and surface currents then flow in that direction, thus changing
the course of the animal. Thus, when the animal changes its course in
a reaction, the surface currents change their course in a corresponding
way, as shown by the movements of particles on the surface (pp. 143,
185, 192).

(18) Sometimes the course is squarely reversed as a result of a
stimulus. In this case the original anterior region becomes detached
from the bottom, while a new pseudopodium projects freely into the
water from the former posterior (unattached) region, settles down, and
becomes attached ; the internal and surface currents then follow it.
This process requires usually a considerable interval of time (pp. 183,

1S4).

(19.) Both surface currents and internal currents are toward the
source of stimulation in a positive reaction, away from the source of
stimulation in a negative reaction.

THE MOVEMENTS ANt> REACTIONS OF AMCEBA. 2^'^

(20) The currents in the positive and negative reactions are not
similar to the currents in a drop of inorganic fluid moving toward or
away from an agent which causes a local decrease or increase in the
surface tension. In Amoeba the currents on the surface and in the
interior are congruent : in the inorganic fluid they are opposed.

(21) In the taking of food the protoplasm and the food body in
many cases do not tend to adhere, so that the Amoeba is compelled to
overcome considerable mechanical difliculty before the food can be
inclosed. Frequently the food body rolls away from the animal as soon
as it is touched (pp. 193, 196). The difliculty is overcome by sending
out pseudopodia on each side of the body and inclosing it, together with
a certain amount of water. In Amoeba verrucosa and its relatives food-
taking is aided by the tendency of foreign bodies to adhere to the body
surface. Amoebas frequently prey upon each other, and this often gives
rise to a long and complex train of reactions (pp. 198-202, and Fig. 76).

CONCLUSIONS.

(22) The chief factors in locomotion seem to be as follows: (i) At
the anterior edge of the Amoeba a wave of protoplasm pushes out, rolls
over, and becomes attached to the substratum ; (2) This pulls on the
upper surface of the Amoeba, causing it to move forward ; (3) The
hinder portion of the Amoeba becomes released from the substratum,
and contracts slowly ; (4) As a result of the strong pull from in front
and the slight contraction from behind the posterior end moves forward ;
(5) The internal substance must flow forward as a result of the pull on
the upper surface, the movement forward of the posterior end, and the
pressure due to the pulling from in front and the contraction behind.
The movement of the internal fluid is comparable to that in a sac or
bladder half filled with water and rolled along a surface (pp. 146, 149,

.7.).

(23) There is no continuous transformation of endosarc into ectosarc
at the anterior end, and of ectosarc into endosarc behind this (Rhum-
bler's ento-ectoplasm process), as a necessary feature of locomotion,
since the ectosarc of the upper surface rolls over to the under surface
at the anterior end (p. 14S). Nevertheless, ectosarc and endosarc are
mutually interconvertible when need arises for the change of one into
the other.

(24) It results from paragraphs (i), (2), (11), above, that the loco-
motion of Amoeba cannot, with fidelity to the results of the physical
experiments, be accounted for by a decrease in surface tension at the
anterior end.

(25) From (3), (4), (12), above, we must conclude that the sending
out of pseudopodia cannot, without violence to the results of the phys-
ical experiments, be accounted for as due to a local decrease in surface
tension at the point of the pseudopodium.

334 "^^^ BEHAVIOR OF LOWER ORGANISMS.

(26) From (i) and (10) above it results that the simple locomotion
on a substratum could, taken by itself, be accounted for on Berthold's
theory that the movement is due to the spreading out of a fluid on a
solid. But this theory fails when we take into account the formation
of free pseudopodia (p. 214), and the fact that all the processes con-
cerned in locomotion can be performed without adherence to a solid
(paragraph (9) above).

(27) From (3), (4), (9), (11), (12), (34), (25), (26), we must con-
elude that the formation of pseudopodia and the sending out of waves
of protoplasm at the anterior end of a moving Amoeba are due to a
local activity of the protoplasm for which no physical explanation has
been given. Since these are the essential features in locomotion, we
must conclude that locomotion in Amoeba has not been physically
explained.

(28) From (17), (19), (20), it follows that we cannot, with fidelity
to the results of physical experimentation, hold that the effects of stimuli
in modifying the movements of Amoeba are due to their direct (or even
indirect) action in changing the surface tension of the parts affected.

(29) From (21) we must conclude that adherence between the proto-
plasm and the food substance does not furnish an adequate explanation
of food-taking and the choice of food in Amoeba.

(30) From (24), (25), (27), (28), we must conclude that changes in
the surface tension of the body are not the primary factors in the move-
ments and reactions of Amoeba. (See note, p. 225).

(31) All the results taken together lead to the conclusion that neither
the usual movements nor the reactions of Amoeba have as yet been
resolved into known physical factors. There is the same unbridged
gap between the physical effect of the stimulus and the reaction of the
organism that we find in higher animals.

(32) In the behavior of Amoeba we may distinguish factors compar-
able to the habits, reflexes, and automatic activities (Ziehen) of higher
organisms (pp. 228-229). Its reactions as a rule are adaptive (pp.
227-228).

SEVENTH PAPER

THE METHOD OF TRIAL AND ERROR IN THE
BEHAVIOR OF LOWER ORGANISMS.

235

THE METHOD OF TRIAL AND ERROR IN THE
BEHAVIOR OF LOWER ORGANISMS.

A certain type of behavior in higher animals has been characterized
by Lloyd Morgan as the method of trial and error. The nature of such
behavior is well brought to mind by an example from Morgan (1894,
p. 257). His dog was required to bring a hooked walking stick
through a narrow gap in a fence. The dog did not pause to consider
that the stick would pass through the narrow opening only if taken
by one end and pulled lengthwise. On the contrary, he simply seized
the stick in the way that happened to be most convenient, near its
middle, and tried to carry it through the gap in the fence in that man-
ner. Of course, the stick would not pass, and after some effort the
dog was forced to drop it. Then he seized it again at random, and
made a new effort. Again the stick was stopped by the fence ; again
the dog dropped it, took a new hold, and tried again. After several
repetitions of this performance, the dog seized the stick by the hooked
end. This time it passed through the gap in the fence easily.

The dog had "tried" all possible methods of pulling the stick
through the fence. Most of the attempts showed themselves to be
*' error.'* Then the dog tried again, till he finally succeeded. Thus
he worked by the method of trial and error.

This method of reaction has been found by Lloyd Morgan, Thorn-
dike (1S9S), and others, to play a large part in the development of
intellieence in hijjher animals. Intelligfent action arises as follows:
The animal works by the method of trial and error till it has come
upon the proper method of performing an action. Thereafter it begins
with the proper way, not performing the trials anew each time. Thus
intelligent action has its basis in the method of " trial and error," but
does not abide indefinitely in that method.

Behavior having the essential features of the method of " trial and
error" is widespread among the lower and lowest organisms, though
it does not pass in them so immediately to intelligent action. But like
the dog bringing the stick through the fence the first time, they try all
ways, till one shows itself practicable.

This is the general plan of behavior among the lowest organisms
under the action of the stimuli which pour upon them from the sur-
roundings. On receiving a stimulus that induces a motor reaction,
they try going ahead in various directions. When the direction fol-
lowed leads to a new stimulus, they try another, till one is found which
does not lead to efi'ective stimulation.

237

238

THE BEHAVIOR OF LOWER ORGANISMS.

>»^:

This method of trial and error is especially well developed in free-
swimming single-cell organisms — the flagellate and ciliate infusoria —
and in higher animals living unde similar conditions, as in the Roti-
fera. In these creatures the structure and the
method of locomotion and reaction are such
as to seem cunningly devised for permitting
behavior on the plan of trial and error in the
simplest and yet most effective way.

These organisms, as they swim through the
water, typically revolve on the long axis, and at
the same time swerve toward one side, which
is structurally marked. This side we will call
X. Thus the path becomes a spiral. The or-
ganism is, therefore, even in its usual course,
successively directed toward many different
points in space. It has opportunity to try suc-
cessively many directions though still progress-
ing along a definite line which forms the axis
of the spiral (see Fig. 79). At the same time
the motion of the cilia by which it swims is
pulling toward the head or mouth a little of the
water from a slight distance in advance (Fig.
79). The organism is, as it were, continually
taking "samples" of the water in front of it.
This is easily seen when a cloud of India ink
is added to the water containing many such
organisms.

At times the sample of water thus obtained is
of such a nature as to act as a stimulus for a
motor reaction. It is hotter or colder than
usual, or contains some strong chemical in solu-
tion, perhaps. Thereupon the organism reacts
in a very definite way. At first it usually
stops or swims backward a short distance, then
it swings its anterior end farther than usual
toward the same side X to which it is al-
ready swerving. Thus its path is changed.
After this it begins to swim forward again. The amount of backing
and of swerving toward the side X is greater when the stimulus i^ more
intense.

Fig.

79-

*FiG. 79. — Spiral path in the ordinary swimming of Paramecium, showing
how the anterior end is pointed successively in different directions, and how a
sample of water is drawn to the anterior end and mouth from each of these
directions.

Volume XI, Part 2

CONTRIBUTIONS

TO THE

STUDY OF THE BEHAVIOR
OF LOWER ORGANISMS

BY

HERBERT S. JENNINGS

Assistant Professor of Zoology, University of Pennsylvania
Research Assistant, Carnegie Institution of Washington

Published by the Carnegie Institution

OF Washington

1904

CONTRIBUTIONS

TO THE

STUDY OF THE BEHAVIOR
OF LOWER ORGANISMS

BY

HERBERT S. JENNINGS

Assistant Professor of Zoology, University of Pennsylvania
Research Assistant, Carnegie Institution of Washington

Published by the Carnegie Institution

OF Washington

1904

Carnegie Institution of Washington,
Publication No. i6.

Press of Gibson Bros.
Washington, D. C.

PREFATORY NOTE.

The investigations the results of which are herein set forth were
carried out by the aid of certain grants from the Carnegie Institution
of Washington. The author desires to express his deep sense of obli-
gation for the aid thus rendered. The first five papers were prepared
at the Zoological Laboratory of the University of Michigan, and were
submitted to the Carnegie Institution for publication August i, 1903.
To the third paper some additions were made in February, 1904.
The sixth and seventh papers were prepared at the Naples Zoological
Station, while the writer was acting as Research Assistant of the
Carnegie Institution, and were transmitted for publication in January
and March, respectively, 1904.

LIST OF PAPERS.

1. Reactions to Heat and Cold in the Ciliate Infusoria.

2. Reactions to Light in Ciliates and Flagellates.

3. Reactions to Stimuli in Certain Rotifera.

4. The Theory of Tropisms.

5. Physiological States as Determining Factors in the Behavior of Lower

Organisms.

6. The Movements and Reactions of Amoeba.

7. The Method of Trial and Error in the Behavior of Lower Organisms.

FIRST PAPER.

REACTIONS TO HEAT AND COLD
IN THE CILIATE INFUSORIA.

REACTIONS TO HEAT AND COLD IN THE
CILIATE INFUSORIA.

To explain the movements of organisms toward or from a source of
stimulus, we find given almost universally in one shape or another a
certain general formula. This is the schema set forth, with unessen-
tial variations, by Verworn (1899, pp. 500-502) for the orientation of
a ciliate or flagellate infusorian to a one-sided stimulus, and by Loeb
(1S97, PP' 439~442) for the tropisms of organisms in general. Essen-
tially, the schema is as follows : An agent acting upon the organism
from one side causes the locomotor organs of that side to contract
either more strongly or less strongly than those of the opposite side.

Fig. I.*

In the former case (Fig. i) the animal is turned away from the source
of stimulus, till it comes into a position in which the motor organs of
the two sides are similarly affected. Then progressing straight for-
ward, it of course moves away from the source of stimulus (negative
taxis or tropism). If the motor organs on the side most affected are
caused to contract less strongly than those on the opposite side (Fig. 2)

*FiG. 1. — Diagram of a negative reaction of an organism, according to the
tropism schema. The motor organs which act more effectively are shown more
heavily drawn. The more pointed end is the anterior. A stimulus is supposed
to impinge upon the organism a from the direction indicated by arrows; this
causes the motor organs directly affected by the stimulus to beat more strongly,
as indicated by the darker shade. The result is to turn the anterior end in the
direction indicated by curved arrows. The organism thus occupies successively
the positions «, b, c, finally coming into the position d. Here the motor organs
of the two sides are equally affected by the stimulus, hence there is no further
cause for a change of position. The usual forward motion of the organism now
takes it away from the source of stimulus, as indicated by the straight arrow at d.

7

THK BEHAVIOR OF LOWER ORGANISMS.

the organism is necessarily turned witii anterior end toward the source
of stimulus ; then its usual forward movements take it toward the
source of stimulus (positive taxis or tropism). Loeb lays especial
stress on the direction from which the stimulus comes, as it is this
that determines which side shall be most strongly affected by the
stimulus ; otherwise the theory as he sets it forth is essentially like that
held by Verworn. Both these authors apply this schema to the move-
ments of organisms to and from many sorts of stimuli, making it a
general formula for taxis or iropisms. Verworn says (1899, p. 503):

Thus the phenomena of positive and negative chemotaxis, thermotaxis, photo-
taxis and galvanotaxis, which are so highly interesting and important in all or-
ganic life, follow with mechanical necessity as the simple results of differences
in biotonus, which are produced by the action of stimuli at two different poles of
the free living cell.

In the present series of papers the writer proposes to examine the
behavior of a number of lower organisms, in order to determine

Fig. 2.*

whether the reactions to the usual stimuli take place in accordance
with this tropism schema or not, and if not, to determine the real
nature of the reaction method. In this first paper we shall deal with
reactions to heat and cold.

In his recent series of papers on the reactions of infusoria to heat and
cold, Mendelssohn (1902, «, <5, c) develops a theory of thermotaxis in
accordance with the general theory of tropisms, above set forth. In an
earlier paper (Jennings, 1899) ^^^ present author, on the other hand.

♦Fig. 2. — Diagram of a positive reaction, according to the tropism schema.
A stimulus coming from the direction indicated by the arrows to the right acts
upon the organism a. The effect of the stimulus is to cause the motor organs
directly affected by it to contract less strongly, as indicated by the lighter shade
on the right side of a. As a result the animal is turned as shown by the curved
arrows, occupying successively the positions a, by c, d. At d the stimulus
affects the two sides alike, hence there is no cause for further turning, and the
usual forward movement of the organism takes it toward the source of stimulus.

REACTIONS TO HEAT AND COLD.

gave a brief account of the reactions of Paramecium to heat and cold,
according to which these reactions are quite inconsistent with the
tropism schema. As the matter is one of considerable interest, and the
conclusions reached by Mendelssohn and myself seem quite irrecon-
cilable, I have examined anew the phenomena in a considerable number
of infusoria, including Paramecium.

The general phenomena to be explained are well seen in the follow-
ing experiment, taken from Mendelssohn (Fig. 3). An ebonite trough
10 cm. in length and 2 cm. wide is filled with water containing Para-
mecia (a). Now, by proper methods, one end of the trough is slowly
heated to 38°, while the other is kept at the temperature 26°. The

a

T — : 5—7*": — y - "< ■-— ■» — r-=

IS'

13

i

26'- - —

38

c

L

10

25'

Fig. 3.*

Paramecia soon leave the heated region, traveling away from it in a
rather compact mass, and in 5 to 15 minutes they have reached the op-
posite end (3) . If now the temperature at the two ends is reversed,
the Paramecia travel back to the end from which they came. If the
temperature is lowered to 10° at one end, instead of raised, similar re-
sults are obtained ; the Paramecia leave the cold region, as before they

* Fig. 3. — General phenomena of thermotaxis in Paramecium, after Men-
delssohn (1902, a). At a the Paramecia are placed in an ebonite trough, both
ends of which have a temperature of 19°. The Paramecia are equallj scattered.
At 3, the temperature of one end is raised to 38°, while at the other it is only 26°.
The Paramecia collect at the end having the lower temperature ("negative
thermotaxis"). At c, one end has a temperature of 25°, while the other is
lowered to 10°. The Paramecia now gather at the end having the higher tem-
perature ("positive thermotaxis").

lO THE BEHAVIOR OF I.OWKR ORGANISMS.

left the heated region (c). If one end is heated, while the other is
cooled, the Paramecia gather in the intermediate region.

How are these movements to be explained? Mendelssohn applies
to the phenomena Verworn's schema for the orientation of a ciliate
organism to a one-sided stimulus (see Figs, i and 2). As we wish to
deal thoroughly with this schema, it will be well to set it forth here, as
applied by Mendelssohn to heat and cold, with some fullness.

The temperature being higher at one end of the trough than at the
other, that side or end of the animal directed to the heated end of the
trough has a higher temperature than has the opposite side or end
(see Fig. 4). This difference in temperature causes a difference in the
beat of the cilia. In negative thermotaxis the higher temperature
causes the cilia to contract more strongly, as indicated by the heavier
shade (on the left side) in the figure ; hence the animal is turned
toward the opposite side, or away from the source of heat, until it
comes into a position where the heat acts equally on the two sides.
The Paramecium then of course has its anterior end directed from the

heated region, and its ordinary

swimming carries it away. In
positive thermotaxis, on the
other hand, the lower tempera-
ture causes stronger contrac-
FiG.A* tions ; hence the cilia on the

side np xt the cold region con-
tract more strongly, turning the anterior end in the opposite direction.
The Paramecium then swims away, as a result of its normal forward
movement.

Mendelssohn studied the subject primarily from a quantitative stand-
point, determining the optimum temperature, the rate of reaction, the
effects of different temperatures, etc. For this purpose he constructed
a very ingenious and delicate apparatus, which permitted accurate
quantitative results. Relying then upon his valuable papers for these
matters, I have devoted myself entirely to a study of the mechanism of
the reactions. For this purpose an apparatus was used that is similar

♦ Fig. 4. — Diagram of the thermotactic reaction of Paramecium as conceived
by Mendelssohn, after Mendelssohn (1902, d). The heavier cilia on the left side
show those contracting most strongly and hence those most effective in turning
the organism or driving it forward. In negative thermotaxis the left end would
have the higher temperature, causing the cilia of the left side of the organism
a to beat more strongly. As a result, the organism turns, occupying suc-
cessively the positions a, if, c, d. In the latter position there is no further
cause for turning, and the animal swims directly away from the heated end.
The same diagram illustrates also positive thermotaxis, if the left end is sup-
posed to be cooled below the optimum.

REACTIONS TO lEEAT AND COLD.

II

in principle to that of Mendelssohn, but more easily constructed and
permitting exact observation of the organisms with the microscope,
though otherwise much less elegant than Mendelssohn's. This ap-
paratus is shown in Fig. 5. It consists essentially of three glass tubes,
of 8 millimeters bore, which are supported in a horizontal position,
side by side, by passing them through auger holes in a block of
wood. The tubes are one inch apart and are placed exactly at
the same level, so that a glass slide rests equally on all three. To
the two ends of each of these rubber tubes are attached. The rubber
tubes from one end pass upward into vessels of water raised on a shelf
above the level of the apparatus. From the other end the rubber tubes

pass downward into a waste pail, thus acting as overflow tubes. A
trough, or slide (5), containing infusoria, is placed on the three glass
tubes ; the water in the vessels on the shelf is heated or cooled to any
desired temperature, and is then siphoned off and allowed to flow
downward through the glass tubes. The rate of flow is controlled by
pinchcocks. In this manner heated water can be caused to flow
beneath one end of the slide, cold water beneath the other. The slide
being thus unequally warmed, the reactions of the organisms can be
observed. The rubber tubes leading from the hot and cold vessels can
be interchanged, so that the temperature at either end or the middle of
the slide can be at once changed and made high or low, without the

♦Fig. 5. — Apparatus used for testing reaction to heat and cold. For de-
scription, see text.

12 THE BEHAVIOR OF LOWER ORGANISMS.

slightest disturbance to the slide or trough containing the organisms.
The plan of this apparatus is taken from that of Mendelssohn. It can
be readily constructed in an hour or less, and gives essentially the same
results as Mendelssohn's more elaborate arrangement. With the use
of specially constructed thermometers, such as were employed by
Mendelssohn, exactly the same quantitative work could be done. The
present apparatus has the advantage that it is possible to place a
mirror beneath the glass slide or trough bearing the organisms, and
thus to observe the movements of the latter with the microscope by the
aid of reflected light. With the long-armed Braus-Driiner stand the
whole extent of the trough can be examined at ease, and the movements
of the organisms accurately observed with the stereoscopic binocular.

As a trough I usually employed a glass slide, to which strips of glass
2 mm. in diameter had been cemented, making a trough 3 inches long,
about two-thirds of an inch wide, and 2 mm. deep. In some of the
experiments the trough was covered with a glass plate ; in others it
was left open. Both methods have their advantages and disadvantages.

To realize the exact conditions under which the organisms are act-
ing it is necessary to consider a further question : What is the precise
nature of the stimulating agent in these experiments.'* Are we dealing
with radiant heat or with conducted heat.f* If we are dealing
primarily with radiant heat, of course currents in the water have
no effect on the distribution of the stimulating agent. If, on the other
hand, we are dealing with conducted heat, if the stimulating agent is
the heated or cooled water, then the conditions are different. Local
currents will cause local variations in the distribution of the heated
water. It is evident, I think, that the second alternative is in all
probability the correct one. Certainly in a bath-tub or in a long
vessel of any sort in which the water is heated at one end and not at
the other, it is possible by producing currents to vary the distribution
of the heated water and to perceive with the hand that it is this heated
water which acts as the stimulus.

The importance of these considerations is evident when we take into
account the fact that the ciliate infusoria are always accompanied by
currents of typical character, having a definite relation to the form and
orientation of the animal's body. As a result of these currents, the in-
fusorian becomes not a mere passive recipient of stimulations, but an
active agent, determining by its activity how and in what part of the
body it shall be affected by stimuli. This may be illustrated by a
diagram (Fig. 6) showing the typical currents produced by the cilia
of Paramecium and the effect produced by these currents upon the
distribution of the heated (or cooled) water. The temperature is con-

REACTIONS TO HEAT AND COLD.

13

ceived to be greatest to the right of the figure and to fall off regu-
larly toward the left, the lines indicating regions of equal temperature.
The last line to the left is marked 28°, this being about the threshold
temperature for the negative reaction of Paramecium, according to
Mendelssohn. The space about the Paramecium (without lines) is at
a temperature below 28° — say at the room temperature — so that it
does not act as a stimulus to cause movement. Now, as the diagram

Fig. 6.*

shows, a cone of water is drawn toward the anterior end of Para-
mecium, from a considerable distance away, necessarily therefore
including water above the threshold temperature of 28°. This cone or

♦Fig. 6. — Diagram of currents produced by the cilia in Paramecium when
the animal is nearly or quite at rest. At right of the line marked 28° the tem-
perature is above the optimum (above 28°) while at the left of this line it is at
the normal or optimum temperature. The heated water first reaches the Para-
mecium at the anterior end on the oral side, passing down the oral groove to
the mouth.

14 THE BEHAVIOR OF I.OWER ORGANISMS.

vortex of water passes as a slender stream along the oral groove of
Paramecium to the mouth. Consequently, water heated above the
threshold temperature reaches the Paramecium in this region before it
touches the body elsewhere. The result is thus a stimulation on the
oral side of the body, not elsewhere.

Thus the way in which the organism is stimulated depends not ex-
clusively on the physical laws of the distribution of heat, but upon the
activity of the organism ; and the method of reaction, as we shall see,
is of a corresponding character.

It is not difficult to observe the distribution of the currents above
described if one adds to the water on one side of the nearly or quite
quiet infusorian a cloud of very finely ground India ink. The same
results are obtained with other infusoria ; in Stentor, in Bursaria, and
in some of the larger Hypotricha the results are particularly striking.
Of course if the India ink, or the surface of threshold temperature, is
advancing obliquely to the axis of the infusoria, the results are more
complicated, and a diagram such as we have in Fig. 6 is not easy to
construct. But the result is uniformly to bring the stimulating agent
to the peristome before it reaches any other part of the body. It is not
possible to observe directly the distribution of water of different tem-
peratures, but under the influence of currents this of course follows,
essentially, the same laws as do fine particles suspended in the water.

Another factor which it is important to take into consideration in
studying the effects of heat or other agents on the infusoria is the
greater sensitiveness of the anterior end and oral surface (or peristome)
as compared with the remainder of the body. This the present writer
has demonstrated for the anterior end by direct mechanical stimulation
in a considerable number of infusoria (Jennings, 1900), while Roesle
(1902) has shown a similar high comparative sensitivity for the peri-
stome region. The difference is such that in many cases where the
animal is completely enveloped by a stimulating agent (as by a chemi-
cal, or by warm or cold water) there is reason to think that the
reaction given is due to the stimulation at these regions alone. In
other words, the stimulus reaches its threshold value for the anterior
end and the region about the mouth much before it reaches this value
for the rest of the body. This consideration has an important bearing
on the theory which is frequently maintained, that the directive action
of a stimulus is due to the difference in its intensity on the two ends or
sides of the organism. Even if a stimulating agent acts, per se,
slightly more strongly on the posterior end than on the anterior end
of an infusorian, there is reason to think that the reaction would be
conditioned entirely by the stimulus at the anterior end, this reaching

REACTIONS TO HEAT AND COLD. I5

its threshold value before the stimulus elsewhere produces any effect.
Corresponding statements could be made with relation to the oral and
aboral sides. Of course, owing to the course of the currents above
described, any stimulating agent whose distribution is affected by cur-
rents in the water will usually reach the anterior end and oral side
first in any case.

Summing up, we find (i) that the threshold intensity of a stimulating
agent whose distribution is affected by currents in the water will reach
the anterior end and oral side of the organism before it reaches other
parts of the body ; (2) that the anterior end and oral surface are more
sensitive than the rest of the body, so that the threshold value for
stimuli is less here than elsewhere.

We may now proceed to an account of the observed method by
which some of the organisms react to heat and cold.

Oxytricha fallax : This is one of the most favorable of the Ciliata
for determining the method of reactions to stimuli, for two reasons,
(i) It is easil}' procurable in large numbers, occurring in cultures of
the same sort that produce Paramecium, and in equal abundance.
(2) It does not, as a rule, revolve rapidly on its long axis, as Para-
mecium does, but usually creeps with its oral or ventral side against a
surface, so that it is not difficult to observe the relation of the reaction
movements to the differences in the sides of the body.

When water containing a large number of Oxytrichas is placed in
the trough and one end of the trough is heated by passing warm water
through the tube which underlies it, the Oxytrichas gradually pass
toward the opposite end of the trough, forming a dense assemblage
with a rather sharply defined edge toward the heated side. If the end
at first heated is now cooled and the opposite end heated, the organisms
pass back to the end from which they first came. Similar results are
obtained by making one end very cold ; the animals gather in an
optimum region, avoiding both too great heat and too great cold.*
The phenomena are identical with what is to be observed in the case
of Paramecium , save that it requires somewhat longer for the Oxytrichas
to move from one end of the trough to the other, and the progress in a
definite direction is not so steady as we find it in Paramecium.

If the movements of the individuals are observed we find them to be
as follows : Near that end of the trough where the temperature is

♦ Many quantitative data for various infusoria are given in the valuable papers
of Mendelssohn. As the object of the present paper was not to obtain quantita-
tive data, but to determine just how the animals acted, absolute temperatures
are not recorded. In every case the experiments were so varied as to use at
times temperatures to which a reaction was hardly noticeable; at other times
more extreme temperatures, up to those which were destructive.

i6

THE BEHAVIOR OF LOWER ORGANISMS.

raised above the threshold vahie the animals begin to move about
rapidly. At first view this movement seems to be quite irregular, as
Mendelssohn describes it in Paramecium. But exact observation of

Fig. 7.*

the individuals taken separately shows that this movement is not so
entirely irregular as it at first appears. Most of the animals swim

♦Fig. 7.— Method by which Oxytrtcka fallax reacts to heat or cold. The fig-
ure represents one end of a trough or slide, which is heated from the end x. An
Oxytricha in the position i is reached by the heat coming from the end x. The

REACTIONS TO HEAT AND COLD. 1 7

backward, circling at the same time toward the right or aboral side, as
shown in Fig. 7. This lasts but a moment ; then the animal swims
forward, at the same time turning to the right or aboral side. That is,
the individuals give the typical motor reaction, as described in the fifth
of my studies (Jennings, 1900). This reaction is repeated many
times, as long indeed as the animal remains in the heated region. But
of course this movement scatters the animals rapidly. Those that
strike against the end or sides of the trough repeat the reaction above
described, backing, turning to the right, then going forward (Fig. 7 at
8, 9, 10, 11). They thus become directed in some other way. Those
that are directed away from the heated region pass into cooler water
and hence no longer give the reaction, but continue their course (Fig.
7 at 13, 14). The result is that the individuals which swim away from
the heated end continue their course, while those starting in any other
direction are stopped and turned (through the motor reaction), until
they too get started away from the heated region. Thus after a time
there is a steady stream of organisms swimming or creeping away from
the heated end, while there is no regular movement in any other
direction. In this manner arises the orientation of the animals, with
anterior ends directed away from the heated region.

The movements of the individuals are exactly as above described
even when the heat is applied some distance from the region where
the animal is found and gradually approaches it from one side. The
animal by no means turns directly away from the heated region, but
repeatedly gives the backing and turning reaction till it is finally mov-
ing in a direction which takes it out of the heated region.

How is this continued backing and turning to be accounted for on
the theory of direct action on the locomotor organs of the two sides as
maintained by Mendelssohn.? This author speaks in the case of Para-
mecium merely of "disordered" movements when the reaction first

animal reacts by turning to the right and backing (i, 2, 3), turning again (3-4),
swimming forward (4-5), backing (5-6), turning again to the right ('6-7), etc.,
till it comes against the wall of the trough (8). It then reacts as before, by
backing (8-9), turning to the right (9-10). This type of reaction continues as
long as the Oxytricha is in the heated region, or as long as its movements carry
it either against the wall or into the heated region. When it finally becomes
directed away from the heated region (13), as it must in time if it continues its
reactions, it swims forward, and since it is no longer stimulated, it no longer
reacts. When large numbers of animals react in this way, in the course of
time nearly all become pointed in the same direction, as at 13 or 14, so that a
marked "orientation" is produced. Thus orientation is produced by "ex-
clusion," due to the fact that the organism is prevented, either by the heat or
the walls of the trough, from swimming in any other direction.

l8 THE BEHAVIOR OF LOWER ORGANISMS.

begins, and thinks this is due to " individual differences, or to ill-
defined internal causes, or perhaps rather to the heterogeneity of the
medium in which they find themselves" (Mendelssohn, 1902, c, p.
492), and that it has nothing to do with thermotaxis proper. This is
typical of many of the statements made concerning the behavior of the
lower organisms ; the movements, so long as they do not agree with
the preconceived schema, are cast aside as disordered, and attention is
called only to the movements that do not conflict with the theory.
Thus Mendelssohn says that this disordered movement '' ceases im-
mediately as soon as the thermotactic action manifests itself" (/. c, p.
492). This is true merely because the thermotactic action is conceived
to begin only after the organism has, through the movements above
described, gotten itself into such a position that it moves away from
the heated region. Of course if all movements except those after
orientation has occurred are thrown out of consideration, the orientation
can be accounted for in any way desired.

In Paramecium, for which alone Mendelssohn attempts to give an
account, based on observation, of the mechanism of the thermotactic
response, the exact character of the movements is undoubtedly difficult
to observe. This animal is nearly cylindrical in section ; the oral side
is very slightly marked, the movements are rapid, and the animal con-
tinually revolves rapidly on its long axis, so that observation of the
relation of the direction of turning to the differentiations of the body is
very difficult. In Oxytricha and other Hypotricha these difficulties
are almost absent ; the body is markedly differentiated ; the movements
are less rapid, and, most important of all, there is usually no revolu-
tion on the long axis. It is unfortunate therefore that Mendelssohn
included none of the Hypotricha among the organisms which he
studied. With careful observation of the movements of individuals
the mechanism of the reactions is in these animals absolutely clear.

A crucial test of the theory of direct orientation as maintained by
Mendelssohn is given by observation of the direction in which the
animals turn in becoming oriented. Mendelssohn (1902, c, p. 492)
says that after the disordered movements '' the movements executed to
place the body in orientation are rather movements of rotation." This
could hardly be otherwise, but the important question for deciding as
to the nature of the reaction is. How does the rotation take place? Is
it determined by the direction from which the heat comes, as required
by Mendelssohn's theory, or is it determined by the differentiations of
the animal's body } This point is a decisive one for interpreting the
nature of the reaction. Suppose we have an Oxytricha in the position
a-a, Fig. 8, and heat is applied in such a way as to reach the organism

REACTIONS TO HEAT AND COLD. I9

from the direction indicated by the straight arrows. The heat is supra-
optimal, so that the organism moves away from it. In what direction
will the organism turn in order to reach the position of orientation
h-b'> According to the theory of Mendelssohn, that the orientation
is due to an increase of the effective beat of the cilia on the side from
which the heat comes, the animal must turn in the direction indicated
by the arrow x^ and this is of course what one would naturally
expect, since this is the most direct method of becoming oriented.
But as a matter of fact the organism turns in the opposite direction, as
indicated by the arrow j)/, thus demonstrating the incorrectness of the
theory that orientation is due to increase of the effective beat of the
cilia on the side from which the heat comes. I have made this obser-
vation hundreds of times, not only upon Oxytricha, but on other
Hypotricha and on infu-
soria belonging to other ,!^
groups (see below). The ^-'^ ^^^W"^^^
direction of turning is de- / '^ 3\ yS^\
termined, under the heat / '^f ^^ \

stimulus, by the differen- "If ^yyy'*^^^ v^l^ ^^ "* ~

tiation of the animal's 7)^^^^"^ -^^^-^i "^^Stz.' -•

body. Oxytricha turns to ^^h'<iS^^M'" • • '1 o>^ ' '

the right, without regard \V *^ Z ^Kv ^^^^ I! ^

to the direction from which ^^\ \^ v\'|)^ /^' ^

the heat comes. This is \^ %mv /

very striking when the ^'"^^li---"^''

trough is covered and part

of the animals are creep- ^^°- 8.— Method of orientation in Oxytricha.

,1 1 .,1 For details, see text.

mg on the cover-glass With

ventral side up, while the remainder are creeping on the bottom of the
trough with ventral side down. When stimulated by heat approaching
from one side, all the members of the first group will be 9bserved to
turn counter clock-wise, while those of the second group turn in the
same direction as the clock hands ; that is, each specimen turns toward
its right side.

For becoming completely oriented an animal in the position a-a in
Fig. 8 usually requires a number of reactions, as indicated in Fig. 7,
but the turning in every case is as indicated by the arrow j/ (Fig. 8).

After it has become oriented with the anterior end away from the
source of heat, Oxytricha by no means maintains this position with
rigidity ; on the contrary the individuals shoot back and forth, in a way
that might be anticipated from the method in which the reaction
occurs. They thus form groups here and there, which gradually move

20 THE BEHAVIOR OF LOWER ORGANISMS.

away from the heat, much as is described by Mendelssohn for Para-
mecium bursaria. With a large number of individuals a general
orientation is evident, however, after the experiment has been some
time in progress.

If ice water is used as the stimulating agent in place of heated water,
the phenomena to be observed are practically identical with those
above described. The organisms leave the colder region, giving the
same reaction as with heated water. As the cold has the additional
effect of decreasing the movements, many individuals are immobilized
by a very low temperature before they have succeeded in escaping
from it, so that they remain in the cold region. The reaction is thus
less clearly defined than that to heat.

Oxytricha ceruginosa : This organism, though smaller, is in some
respects more favorable than O. fallax for observing the method of
reaction. This is because the individuals are more inclined to swim
freely through the water, so that their progress away from or toward
the heated region is more rapid than in O. fallax. O. ceruginosa^
further, even when moving freely through the water, either does not
revolve on the long axis at all or revolves only very slowly. In con-
sequence of this it is easy to determine the relation of the direction of
turning to the differentiations of the body.

The reaction to heat and cold is in essentials identical with that of
O. fallax, and this reaction is repeated till the animals are carried into
a region where the temperature is not such as to cause the reaction.
Those that are carried into such a region will of course be swimming
away from the stimulating region ; hence, in a large number of individ-
uals there is an evident orientation, with anterior end directed away
from the source of heat or cold. All the conditions and details as to
the production of this orientation are as set forth above for O. fallax.

Stylonychia mytilus: This large Hypotrichan is still more favorable
for the study of the movements of individuals under the stimulus of
heat or cold coming from one side than are the two species of
Oxytricha. But I have found it less easy to obtain in large numbers,
and for this reason have not chosen it for the detailed description of
the reaction. Where comparatively few specimens are available,
the movements of individuals are easily studied, but there is little
impression of any real orientation, such as one gets clearly when large
numbers are used.

The movements of the individuals are like those described for
Oxytricha fallax. The animal in reacting always turns to its right,
without regard to the relation of this to the direction from which the
heat or cold is coming. With an organism of the large size of

REACTIONS TO HEAT AND COLD. 21

Stylonychia this is very evident. This turning to the right under the
stimulus of heat and cold in Stylonychia has already been described
by Putter (1900), incidentally to his study of the effect of contact
stimuli in this organism.

Stentor cceruleus : Mendelssohn includes in his paper a note stat-
ing that positive and negative thermotaxis occur in some species of
Stentor, and giving the optimum ; but he made no study of the mech-
anism of the reactions in this animal. Had he done so, it seems to me
that he could not have maintained his theory of the way in which
the reaction takes place.

When one end of the trough is warmed the Stentors near that end
begin after a few seconds to move about more rapidly. In most cases
the movement is as follows : The animals swim backward some
distance, then turn toward the right aboral side and swim forward
(the typical motor reaction). Thus the general effect is as of an
irregular movement in all directions. Those individuals which swim
forward toward the other end of the slide pass out of the heated region ;
hence the motor reaction no longer takes place, and the animals con-
tinue to swim forward. Those which start in any other direction do
not escape from the heated region, and therefore soon give again the
motor reaction, backing and turning again to the right. Thus only
those that swim away from the heated region continue their course ;
the others are stopped and turned until finally they too get started in
the same direction. Therefore, after a period of apparently disordered
swimming, there is an evident orientation of many individuals, with
anterior ends away from the heated region. This orientation is caused
as it were by exclusion ; in animals swimming in any direction but one
the motor reaction is produced, so that only this direction can be main-
tained. After a time, therefore, a large proportion of the individuals
are swimming in this direction, with a common orientation.

Thus the direction in which the animals turn is determined, as in
the Hypotricha, by the structure of the body, and not by the direction
from which the heat comes.

Those outside the region where the heat has reached the threshold
temperature often swim for some distance toward the heated region ;
then arriving at a point where the heat is effective, they give the motor
reaction, backing and turning to the right. They are thus prevented
from entering the heated region.

If the temperature is rapidly raised, the animals may not succeed in
escaping from the heated region until they are injured. In this case
the specimen contracts strongly and swims backward a long time. It
becomes distorted, places the disk against the bottom or other surface,
becomes motionless, and finally dies.

ii THE BEHAVIOR OF LOWER ORGANISMS.

Fixed specimens react less readily to heat than do free-swimming
specimens. They do not orient themselves with reference to the direc-
tion from which the rise in temperature comes. They may remain
extended normally, carrying on the usual activities, after the tempera-
ture has risen beyond the point which sets the free specimens in rapid
reaction. But as the temperature rises they repeatedly bend over into
a new position (bending toward the right aboral side), then contract
strongly, and finally free themselves from their attachment. There-
upon they behave like other free individuals.

Spirostomum ambiguum : In this large ciliate the reactions to heat
and cold take place in essentially the same manner as is described
above for Stentor and the Hypotricha, so that it is not necessary to
describe the phenomena in detail. The organism reacts to heat or
cold by backing and turning toward its aboral side ; and this whether
the change in temperature is uniform over the entire surface of the
animal or whether it approaches from one side. The movements of the
animal are slow, and under the Braus-Driiner stereoscopic microscope
its method of reaction is very clear. There is little marked common
orientation at any time, however ; this being due to the slowness of the
movements and the frequency of repetitions of the motor reaction.

Bursaria truncatella : In this very large infusorian, in which cer-
tain differentiations of the body are visible even to the naked eye, the
method of reaction to heat and cold is observed with the greatest ease.
But orientation of a large number of individuals in a common direction
is hardly to be noticed, though if Bursaria could be obtained in such
numbers as Paramecium or Oxytricha, perhaps an indication of orien-
tation would be noticeable in spite of the slowness of movement.

Bursaria is very inactive, often remaining quiet for long periods.
It swims slowly, and frequently creeps along the bottom with ventral
side down, but may also swim freely through the water, revolving to
the left. If the temperature of the trough is raised at one end, the
animals in this region that are moving freely through the water swim
backward, turn to the right, and swim forward. This may be repeated
till the organism passes out of the heated region. Rather more
frequently, however, the animal, after thus reacting once or twice, sinks
to the bottom and places its ventral side against the surface. It now
conducts itself in the same manner as do the other individuals in this
situation, as will be described.

The individuals which are resting against the bottom (usually the
majority of those in the trough) react as follows : They begin to swim
backward, keeping the ventral side down and at the same time circling
toward their own right sides. They thus describe rather narrow circles.

REACTIONS TO HEAT AND COLD. 2^

This continues until the heat becomes destructive — the animals cease
circling, become quiet, and finally disintegrate. The reaction of those
individuals which are resting or creeping on the bottom is thus not of
a character to save them from destruction.

Specimens which are by chance moving along the bottom from a
cool region toward the warm region do not escape ; they merely stop
and begin to circle backward to the right when they reach the heated
spot, and continue this till they die.

Thus the reaction of Bursaria to heat, while of the same general
character as that of other infusoria, must be accounted very imperfect,
since it hardly results in orientation at all, and does not preserve the
animals from destruction.

Paramecium caudatum : * In the second of my studies (Jennings,
1899, pp. 334-336) I gave a brief account of the way in which, ac-
cording to my observations, Paramecium reacts to" heat and cold.
From my more recent studies I can confirm this account. But as
Mendelssohn has recently come to different conclusions for the tempera-
ture reaction of this animal, and as he misunderstands certain points
in my brief description, it seems desirable that I should supplement the
account previously given in order to make it clear.

Paramecium reacts to heat and cold in essentially the same manner
as is described above in detail for Oxytricha. When the higher or
the lower temperature advances from one side the animals swim
backward, turn toward the aboral side, and swim forward again.
They continue this until the movement brings them into a region of
more moderate temperature. Paramecium reacts more readily than
Oxytricha, the reactions are repeated at shorter intervals, and the
movements are more rapid, so that a common orientation of many
individuals swimming away from the region of higher or lower tem-
perature is more quickly produced and is more striking to the eye. It
results farther from this more rapid movement, as well as from certain
other factors, that the method of reaction in Paramecium is much less
easily observed than in any of the other infusoria described. Indeed,
Paramecium is one of the most unfavorable forms obtainable for a
study of reaction methods, and it is, I believe, due largely to the fact
that this animal is usually employed for such study that progress has

♦The common Paramecium, which appears everywhere in immense numbers
in decaying vegetation, receives from different authors sometimes the name
Paramecium aurelia, used by Mendelssohn ; sometimes the name given above.
I use the name caudatum because it appears to me to be the correct one, but
there is no reason for considering the animals thus differently denominated to
be really different.

24 THE BEHAVIOR OF LOWER ORGANISMS.

been so slow in appreciating the real nature of the reactions of the in-
fusoria. If Stylonychia or Oxytricha or any other of the Hypotricha
had been taken as the usual type for study on reactions, many of the
theories now maintained could never have been put forth. The body
of Paramecium is comparatively little differentiated, so that it is diffi-
cult to distinguish oral and aboral sides, and, to multiply tliis difficulty
many times, the animal revolves rapidly on its long axis, so that oral
and aboral sides never retain for two successive instants the same
position. It is not wonderful, therefore, that the method of reaction
by turning toward the aboral side was not observed in the first investi-
gations on Paramecium and that many still find it difficult to observe.
Nevertheless, it was on Paramecium itself that this reaction method was
first observed (Jennings, 1S99), and its existence was confirmed later
on the organisms where its observation presents no difficulties. Aside
from the direct observations of the method of reaction, the following
facts throw light on the way in which the collections take place.

As described in the second of my studies
(Jennings, 1899, pp. 314, 315), the collect-
ing of Paramecia in regions of optimum
temperature may be produced in the follow-
ing manner : The infusoria are mounted in
„ ^ water which is above the optimum temper-

ature (say 30°) on a slide beneath a cover
glass supported at its ends by glass rods. Into this slide is introduced
with the capillary pipette a little cooler water (say at 24°), which
covers a small circular area in the center of the slide. Very soon the
Paramecia have collected in this region till a dense group is formed.
The same result may be obtained by placing a drop of ice water on the
top of the cover glass of a slide of Paramecia which has been warmed
considerably above the optimum temperature. (Fig. 9.)

Are these collections due to the orienting of Paramecium by the
heat, as maintained by Mendelssohn for thermotaxis in general.? Ob-
servation shows that they are not ; that on the contrary the Paramecia
gather in the optimum region in the same manner as they gather in a
drop of weak acid, as described in my studies. The Paramecia on the
heated slide are swimming rapidly in all directions. They do not
change their course or become oriented in the least when a spot in a
certain part of the slide is cooled. But as a consequence of their

*FiG. 9. — Collection of Paramecia due to the reaction to temperature change.
The slide rests on a vessel of water at a tetnperature of 45". An elongated drop
of ice water is placed on the upper surface of the cover glass. The Paramecia
quickly collect beneath the drop of ice water.

REACTIONS TO HEAT AND COLD. 35

rapid movements many of them by chance enter the cooler region.
They do not react at all as they enter, but continue across. On
coming to the other side of the drop, however, they do react, by back-
ing and turning toward one side (the aboral). They react whenever
they come to the boundary of the cooled region ; hence they do not
leave it. In every respect their behavior is like that seen when Para-
mecia collect in a drop of weak acid, and I believe there is no longer
anyone who holds to the orientation theory for the gathering of Para-
mecium in chemicals.

As in the case of chemicals, it may be demonstrated to the eye in the
following manner that the method above described suffices to account
for the gatherings. On the upper surface of the cover glass is marked
a small ring in ink. By confining the attention to this ring it is easily
seen that in the heated preparation of Paramecia many individuals
cross the ring every instant, so that, if these could all be stopped in
the ring, a dense aggregation would soon result. Then the region
within the ring is cooled by placing a drop of ice water on the cover
above it. The Paramecia continue to swim just as before, save that
they no longer pass out of the ring after swimming in, as they did at
first. In this way a dense collection is soon formed.

Mendelssohn (1902, ^, p. 487) finds it inexplicable why the Para-
mecia should form dense aggregations at the optimum temperature.
He says that they execute " only some insignificant movements" in
this region, not swimming away. On the theory of thermotaxis held
by Mendelssohn this is perhaps inexplicable, but this, it seems to me,
is only because the theory is incorrect. Such collections are due to
precisely the same factors as the rest of the reaction to heat and cold
and are clearly intelligible when the nature of the reaction, as described
above, is taken into consideration.*

In a former paper (Jennings, 1S99, p. 336), after giving a brief
account of the reaction method above described, I pointed out that this
method does not demand a sensitiveness to such minute differences in
temperature as does Mendelssohn's theory, and that therefore the sensi-
tiveness to temperature differences may have been overestimated.

* Mendelssohn (1902, A, p. 487) supposes that I would explain these gatherings
at the optimum temperature through the collection of Paramecia in COg pro-
duced by themselves, and shows that this would not account for the phenomena
observed in these cases, though he confirms the fact of the collections in COg.
But I have by no means maintained that such collections can be produced only
by CO2 ; on the contrary, I have given an account of many different agencies
that will give rise to such collections, and have especially described the fact that
collections are formed in a warmed region through exactly the same reacti©n by
which they are formed in CO^. (Jennings, 1899, P* S^SO

26 THE BEHAVIOR OF LOWER ORGANISMS.

Mendelssohn (1902, a, p. 406) misunderstands my ground for this
statement. He supposes that I hold that the Paramecia do not react
to differences in temperature less than that existing in a certain illus-
trative experiment, where one end of the slide was resting on ice, while
the other was heated to 40°. This experiment was purely for the
purpose of bringing the phenomena of thermotaxis concretely before
the attention of the reader ; its details had no special significance. I
have not the slightest reason for doubting the entire accuracy of the
quantitative experimental results set forth by Mendelssohn, and consider
them a most valuable addition to our stock of exact data. But the
calculation of the sensitiveness of the organisms concerned, from these
experimental results, involves a certain interpretation as to the reaction
method, and it was this interpretation that I called in question. Men-
delssohn, in accordance with his general theory, holds that the reaction
is due to the difference in temperature between the two
ends of the organism, and he calculates that this difference
in temperature could amount, in the case of Paramecia,
to but 0.01° C. According to the reaction method which
I have described above, however, it is not the difference
in temperature between the two ends of the same indi-
vidual that causes the reaction. Consider a slide cooled
below the optimum at the end a ; above the optimum at
the end b (Fig. 10), the optimum temperature for the
Paramecia being between the lines x and y. The animal
p ^ ^ may swim a considerable distance from a position y^ at
one side of the optimum, to a position at, at the other side
of the optimum, before it reacts (by backing and turning, etc.) at all.
We have no ground for maintaining then that it perceives any less differ-
erences in temperature than that between the lines x and y^ and this
difference will be much greater than that between the two ends of the
animal. A similar diagram could be made for the case where the tem-
perature is raised or lowered only at one end of the slide. It seems to
me correct, therefore, that the sensitiveness to temperature differences
has probably been much overestimated. The only way that it could
be estimated would be by observation of individuals to determine the
extent of the stretch x-y over which they pass before reacting, and to
calculate the difference in temperature between the ends of this stretch.
It would of course be very difficult to do this with accuracy.

Mendelssohn's view that it is the difference in temperature between
the two ends of the same individual that determines the reaction is not

♦ Fig. id. — Diagram illustrating conditions necessarj' for determining the sen-
sitiveness of Paramecia to differences in temperature. See text.

REACTIONS TO HEAT AND COLD. 2*J

only rendered inadmissible by the reaction method above described,
but it is rendered a priori improbable by certain other considerations.
First we liave the fact that the anterior end is much more sensitive than
the posterior. Of course it is impossible to measure this difference in
sensitiveness, yet the experiments with mechanical and chemical stimuli
show that it is great. In many infusoria, while the slightest touch at
the anterior end causes a pronounced reaction, it requires a strong stroke
at the posterior end to produce even a slight reaction. (See Jennings,
1900, pp. 23S, 243, 251.) Owing to the much greater sensitiveness of
the anterior end, it is probable that, with the posterior end but 0.01°
warmer than the anterior, the reaction, if any, would be due to the tem-
perature of the anterior end. In other words, there is reason to suppose
that the threshold temperature for the anterior end would be considera-
bly lower than that for the posterior end. If this is true the usual tem-
perature reactions would be throughout due primarily to stimulation
at the anterior end ; and the reaction, as we have seen, is of just the
character which would be expected from this. The first stage in
the reaction is to swim backward^ and this is true also when the animal
is dropped directly into water of uniformly high or low temperature, so
that the temperature of the anterior end is no greater than that of the
posterior end. There is no explanation for the swimming backward
under these circumstances on the theory that accounts for thermotaxis
by the different temperature of the two ends.

A second factor which must be taken into consideration relates to the
currents produced by the cilia of the organism itself. As shown above
(p. 13) , the water of a higher temperature (supposing that we are deal-
ing with the reaction to heat), would as a rule first reach the anterior
end and pass at once down the oral groove, on the oral side (Fig. 6).
The natural result therefore would be a turning toward the opposite or
aboral side, and this is exactly what we find takes place. We should
therefore not expect the organism to turn directly away from that end
of the trough from which the heat comes, for the heated water may not
reach the Paramecium from that side at all.

As will be seen, the facts adduced in the last paragraph are not incon-
sistent with the idea that the organism turns directly away from the
side stimulated. It is the oral side which is, as a rule, stimulated, and
the organism turns toward the aboral side. We seem thus to obtain
a most gratifying union of two apparently opposed views. But the
reactions to certain other stimuli do not admit of such a union. This
is notably true of the reactions to mechanical stimuli, as shown in a
previous paper (Jennings, 1900), and of the reactions to light, lo be
described in the following paper.

a8 THE BEHAVIOR OF LOWER ORGANISMS.

SUMMARY.

The ciliate infusoria react in the same manner to heat and cold as
to most other classes of stimuli ; the response on coming into a region
where the temperature is above or below the optimum is by backing
and turning toward a structurally defined side, followed by a movement
forward. This reaction is repeated as long as an effective supraoptimal
or suboptimal temperature continues. The result is to prevent the
organisms from entering regions of marked supraoptimal or suboptimal
temperature, and to cause them to form collections in regions of opti-
mal temperature. The common orientation of a large number of
individuals sometimes produced in this way is an indirect result of the
method of reaction. Since movement in any other direction than a
certain one is stopped, the organisms after many trials come into this
direction. Orientation is therefore by " exclusion," or by the method
of trial and error. In many of the organisms orientation is not a
noticeable feature of the reaction.

SECOND PAPER

REACTIONS TO LIGHT IN CILIATES
AND FLAGELLATES.

29

REACTIONS TO LIGHT IN CILIATES AND
FLAGELLATES.

In the reactions to light we are dealing with a stimulating agent
which differs in one very important respect from chemicals and from
heat or cold. The distribution of the agent with which we are con-
cerned is not affected by the currents of water produced by the organism ;
hence there is no tendency for one side or part of the organism to be
more strongly affected than the rest, as was found to be the case for
chemicals and for heat and cold. This peculiarity light shares with
the electric current and with radiant heat. The conditions demanded
for immediate orientation through direct action of the agent on the
locomotor organs, in the manner required by the general theory of
tropisms as set forth in the foregoing paper (p. 7) , are therefore present.
In a recent paper Holt & Lee (1901) have attempted to show that
the reactions of organisms to light actually take place in accordance
with this theory.

We shall examine the reactions to light in Stentor cceruleus and in
certain flagellates, in order to determine whether they take place in
accordance with the tropism schema, and, if not, just how they do occur
and on what factors they depend.

THE CILIATA.

STENTOR C^RULEUS.

As is well known, very few of the ciliate infusoria react to light.
Light reactions have been described by Engelmann (1882, a) for several
chlorophyllaceous ciliates ; by Verworn (18S9, Nachschrift) for Pleuro-
nema chrysalis; and by Davenport (1897) and Holt & Lee (1901) for
Stentor cceruleus. In none of the ciliates have the reactions been
described in sufficient detail to enable us to determine their exact
nature.

In Stentor cceruleus the reaction to light manifests itself in the
culture dish by the usual aggregation of the organisms at the side away
from the window. If a number of Stentors are removed to a watch
glass or trough, and this is placed near a window or other source of
light, most of the Stentors are soon found on the side of the vessel away
from the light. If one-half of the glass is shaded by a screen, most of

31

32 THE BEHAVIOR OF LOWER ORGANISMS.

the Stentors are soon found in the shaded half. ^S. cceruleus thus shows
the phenomenon usually called negative phototaxis.

It is to be noted that not all the Stentors are to be found on the side
away from the light, or in the shaded half of the vessel. On the con-
trary, a considerable fraction of the whole number will usually be found
swimming about in all parts of the dish, or at rest in the lighted portion.
The light reaction is thus somewhat inconstant, and varies among
different individuals. It varies considerably with Stentors of different
cultures ; from some cultures almost all the individuals show it, while
from others it is barely noticeable. This variability and inconstancy
run through all manifestations of the light reaction in Stentor.

A word further needs to be said as to the behavior of individuals
which are not free-swimming, but are fixed by the posterior end. Such
individuals do not react at all to light. When light is thrown on them
they remain in the positions in which they are found at the beginning,
neither contracting nor in any way changing their position. No matter
whether the light is weak or strong, and without regard to the direction
from which it comes, fixed Stentors give no reaction and show no
orientation with reference to light. The contact reaction apparently
inhibits the light reaction completely. We shall therefore omit the
fixed individuals from consideration in the remainder of the account,
confining attention to the free-swimming specimens.

The typical motor reaction of Stentor, by which it responds to most
stimuli, is as follows : The Stentor stops or swims backward a short
distance, then turns toward the right aboral side, and resumes its for-
ward motion. This is the reaction which is produced by strong
mechanical stimuli, by heat, and by chemical stimuli, acting upon the
anterior end or upon the body as a whole.

How is the reaction to light brought about.? To answer this ques-
tion it is best to arrange experiments in such a way as to distinguish
as far as possible the effect due to unequal illumination of different
areas from the effect due to the direction from which the light is coming.

In order to produce strong differences in illumination in different
areas of the space in which the Stentors are found, a flat-bottomed
glass vessel containing many Stentors in a shallow layer of water was
placed on the stage of the microscope, in a dark room. From beneath
strong light was sent upward through the opening of the diaphragm,
by throwing the light from the projection lantern (using the electric
arc light) on the substage mirror. By this the light was directed up-
ward through the vessel containing the Stentors. Thus a small,
definitely bounded circular area was illuminated, while the rest of the
vessel remained in darkness. A black screen was usually placed over
the diaphragm opening of the microscope in such a way as to shade one-

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES.

33

half of the circular area, making a sharp line {x-y, Fig. ii) dividing the
light from the darkness. A mirror was placed above the microscope,
inclined in such a position as to project the image of the Stentors, very
much magnified, on the ordinary vertical screen used for receiving
lantern slide views. Thus the behavior of the Stentors could be studied
with great ease on the screen.

The heat from the lantern was cut out, so far as possible, by placing
between it and the mirror of the microscope a glass cell three inches
thick, filled with cold water. In this manner the heat was excluded to
such an extent as to fall below the threshold for the stimulation of
Stentor by heat. This was demonstrated by comparing the reactions
of Stentor with those of Paramecium. Stentor is less sensitive to
changes in temperature than is Paramecium ; this was clear in my ex-
periments on the reaction to heat. Par-
amecium does not react at all on passing
into the area illuminated by the lantern,
but swims about indifferently in both the
dark and the light parts of the dish, show-
ing that the heat produced is below the
threshold for Paramecium ; it must then
be below the threshold for Stentor.

The free Stentors in the unlighted part
of the vessel swim about at random.
Many individuals thus come by chance to
the line x-y, Fig*. 1 1 , where they would
pass into the lighted area. These at
once back a little, then turn toward the
right aboral side, and swim forward

again. The turning toward the right aboral side is usually through an
angle sufficient to direct the Stentor away from the lighted area (see
I, 2, 3, 4, Fig. ii) ; if it is not, the Stentor repeats the reaction until,
after one or two trials, it swims into the unlighted region.

Many of the individuals react as soon as the anterior end reaches the
lighted area, so that less than one-fourth of the body is in the light.
This shows that light falling upon the anterior end alone is sufficient
to cause the reaction.

A few specimens swim completely into the lighted area, then react

♦Fig. II. — Method of studying the manner in which Stentor reacts to light.
The figure shows a circular area, illuminated from below, with the light cut oft
from the left side by a dark screen, the line x-y separating the light from the
dark area. The Stentors collect in the dark area. The reaction of a specimen
which comes to ihe line x-y is shown at i, 2, 3, 4.

34 THE BEHAVIOR OF LOWER ORGANISMS.

in the manner above described. In such cases the nature of the reac-
tion is seen with especial clearness, the entire animal being projected
on the screen and the differentiations of bodily structure (mouth, oral
and aboral sides, etc.) being conspicuous. Specimens which swim
completely into the lighted area are usually compelled to react two or
more times before they escape from the lighted region.

When the light is cut off entirely the Stentors distribute themselves
throughout the dish. If the light is now admitted from below, the
unattached Stentors in the lighted area react by swimming backwards
a certain distance, turning toward the right aboral side, then swimming
forward again. This reaction is repeated frequently until after an
interval the Stentors are carried by these movements outside the lighted
area. They then cease to give the reaction. The reaction, under these
conditions, is thus the same as that produced when Stentors or Para-
mecia are subjected to other adequate stimuli, as when they are placed
in a chemical or dropped into very warm or very cold water. The
result of the reaction is, in every case, to remove the organism from
the sphere of action of the stimulus. When the stimulus is light this
result is produced in exactly the same way as when the stimulus is
heat or cold or a chemical.

The same results may be obtained by lighting the vessel containing
the Stentors directly from above and shading one portion with a screen.
The Stentors remain in the shaded region, responding by the motor
reaction above described when they come to the lighted area. With a
favorable culture the experiment succeeds even when the source of
light is comparatively feeble, as when an ordinary incandescent electric
light is used as the source of illumination.

The results so far show that a sudden increase in the intensity of
illumination induces in Stentor a reaction which is of the same
character as the reaction to other strong stimuli. Such a sudden
increase may be due either to the passage of the Stentor from a dark
to a light region, or to a sudden increase in the brightness of the light
which falls upon the animal. The general effect of the reaction is to
prevent the Stentor from entering a brightly illuminated area, or to
remove it from such an area.

We may now arrange the conditions so that the light shall come
from one side, while at the same time differences in illumination shall
exist in different regions. This may be done by illuminating the vessel
containing the Stentors from the side, then covering one portion of the
vessel with a screen.

The organisms are placed before a lighted window, or an incandes-
cent electric light, in a vessel with a plane front (Fig. 12). One-half

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES.

35

of the vessel is then cut off from the light by a screen (j), the shadow
of which passes across the middle of the vessel containing the Stentors.
One side of the vessel is thus in the light, the other in the shadow,
and these two regions are separated by a sharp line (Fig. 12, x y).

The Stentors are soon all collected in the shaded side of the vessel.
Here they swim about freely in all directions, but do not cross the line
into the lighted portion. Now, by focusing the Braus-Driiner on this
line, the behavior of the individuals on reaching it may be observed.

It is well to examine

the conditions in this case >

with care, as they present
opportunities for a pre-
cise and crucial test of
the theory that the reac-
tion to light is due to a di- ^

rect orientation through
the falling of light on one
side of the organism (pho-
totaxis or phototropism

in the strict sense, as de- *

fined by Holt & Lee).
In the lighted portion of
the vessel the rays of light
come from a certain direc-
tion, as indicated by the *"
large arrows (Fig. 12).

In the shaded region there ^

is not enough light to pro-
duce orientation, the ani-
mals swimming in every *"
direction. On passing
from the shaded region

across the line x-y into the lighted region, the animal should (according
to the tropism theory) become oriented. According to the theory of
negative phototaxis by direct orientation due to differential action on

Fig. 12.*

* Fig. 12. — Method of testing the manner of reaction to light in Stentor. The
large arrows show the direction from which the light rajs come. A screen (s)
cuts off the light from half the vessel, leaving a line (x-y) separating a shaded part
from a lighted part. The Stentors collect in the shaded part, here swimming
about without orientation. At a (i, 2, 3, 4) we see a diagram of the reaction
required by the tropism schema when the organism swims across the line x-y,
while at d (i, 2, 3, 4) we have a diagram of the reaction as actually given under
these conditions.

^6 THE BEHAVIOR OF LOWER ORGANIf?MS.

the two sides, the animals on crossing the line should become oriented
by turning directly away from the source of light, as shown in the
diagram (Fig. 12) at a. The animal would then be expected to swim
in the direction x-y as shown by the specimen «, i, 2, 3, 4.

It cannot be held that the real source of light for the Stentors is that
reflected from the bottom or sides of the dish in the lighted region, and
hence coming on the whole from a direction perpendicular to the line
xy^ for the behavior of the Stentors shows that this is not the case.
A Stentor in the shaded region, close to the line x-y^ as at c. Fig. 12,
receives whatever light there may be thus reflected exactly as it does
after it has crossed the line, yet it shows no reaction and does not
orient itself in any way. On the other hand, as soon as it crosses the
line x-y., so as to receive the light coming from the window, it reacts
strongly, as we shall see. It is thus clearly the light from the window,
coming in the direction shown by the large arrows, that causes the re-
action ; hence the Stentor ought, according to the direct orientation
theory, to orient itself in the line of these rays.

When a Stentor, swimming at random, reaches the line a:-j/, it
reacts by stopping suddenly, then turning toward its aboral side,
then swimming forward. It thus swims about until its anterior end is
again within the shadow, where it continues to swim forward (Fig. 12,
3, 1,2, 3, 4). Often the first reaction is not sufficient to direct it into
the shadow ; in this case the reaction is repeated ; one to three reactions
almost invariably bring the Stentor back into the shadow. It has no
particular orientation in the shadow, but swims in whatever direction
it happens to be headed.

Very frequently the animals react when the anterior end alone has
crossed the line, so that less than the anterior half of the body is lighted.
In other cases the animal swims completely across the line, sometimes
for a distance greater than its own length, into the light, before it reacts.
In any case the reaction is that above described.

Does the Stentor, when it turns on entering the light, always turn
away from the source of light, as the theory of direct orientation
requires .''

At the moment of crossing the line into the light the »Stentor may
occupy various positions. It will be well to note specifically the re-
action in certain of these positions, as we obtain here the observations
which furnish an exact and crucial test of the direct orientation theory.

I. The Stentor may reach the line with the aboral side directed
toward the source of light (Fig 12, b). It therefore turns (as usual)
toward its aboral side. It thus swings its anterior end toward the
source of lights in the direction opposite that required by the direct

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES.

37

orientation theory. This observation was made repeatedly in a very
large number of cases ; not a single exception to it was observed. The
swinging of the anterior end is continued past the point where the light
falls directly upon it until the animal is directed again into the shadow,
as illustrated in the diagram (Fig. 12, ^, i, 3, 3, 4).

2. The Stentor may reach the line with the aboral side directed away
from the source of light. In this case it turns (as usual) toward the
aboral side, thus swinging its anterior end away from the source of light.

3. The Stentor may reach the line with the aboral side directed
upward or downward or in some intermediate position. In every case
it turns toward the right aboral side, in whichever way this is directed.

The writer wishes it to be understood that the foregoing statements
as to the direction in which the animal turns are presented, not merely
as interpretations in accordance with a certain theory, but as direct,
unequivocal observations, many times repeated. Thus, on passing from
a darker to a lighter area, even when the light comes from one side,
the Stentors react merely to the difference in illumination, without
regard to the direction from which the light comes. The direction of
turning is determined tln-oughout by an internal factor, not by the side
of the animal on which the light falls, nor by the direction of the rays of
light. We have put the theory of orientation by direct differential

♦Fig. 13. — Another method of testing the manner in which Stantor reacts to
light. For a side view of this apparatus, see Fig. 14. Light comes from the
left side, in the direction indicated by the arrows. A screen (5) is interposed
between the source of light and the vessel containing the Stentors. This screen
is of such a height (as illustrated in Fig. 14) that it cuts off the light from the
half (/4) of the vessel next to the window, leaving the other half {B) lighted.
At c (i, 2, 3, 4, 5) is seen the reaction method of a specimen which swims across
the line x-y, separating the shaded half ^ from the lighted half ^.

3S

THE BEHAVIOR OF LOWER ORGANISMS,

action of the light on the two sides of the animal to a precise test, and
found it to be incorrect.

The same result is brought out in perhaps a still more striking
manner by the following method of experimentation : A vessel con-
taining Stentors is placed on a dark background near a source of light
(a window or an incandescent electric lamp). The light thus comes
from one side and a little from above. An opaque screen is placed
between the window and the vessel containing the Stentors, of such a
size and in such a position that the top of the shadow of the screen falls
across the middle of the vessel on the line x-y (Fig. 13 ; see also Fig.
14). Thus the half of the vessel next to the window (-4) is darker
than the farther half (-5), and the Stentors collect in this shaded half
After some time scarcely a specimen is found in the lighted part of the
vessel away from the window. The conditions in this case are illus-
trated in the side view (Fig. 14).

The exact behavior of the Stentors in
the darkened portion of the vessel is then
studied by focusing upon them the Braus-
Driiner microscope. The Stentors within
the shaded area are not oriented nor gath-
ered in any particular region,
but swim about at random.
When one of the specimens
comes in its course to the
line :v-j/ (Fig. 13), separating
the darkened area from the
light, it responds to the sudden light which falls upon it from the
window by giving the motor reaction, turning to the right aboral side
and swimming back into the shaded region. Often the reaction occurs
as soon as the anterior end of the Stentor has crossed the line x-y^ so
that the entire Stentor does not pass out into the lighted area. In other
cases the specimen crosses the line x-y completely before the reaction
occurs, so that the entire body is illuminated. It then reacts in the
usual manner, turning toward the right aboral side, so that it is headed
toward the shaded region ; thus swimming back across the line (Fig.
13, c) . After returning into the shaded region the animals swim about
at random as before.

What is the reason for the return of the Stentor into the darkened
area after it has crossed the line into the light region.?

♦Fig. 14. — Sectional view, from the side, of the conditions shown in Fig. 13.
The arrows show the direction of the light rays. The region from 5 to ;« is
shaded by the screen s.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 39

By SO doing it swims toward tlie window, thus in the direction from
which the strongest light is coming. According to the theory of photo-
taxis as due to the direct action of the light on the motor organs of the
animal, this movement is inexplicable. Thus, in the analysis of this
theory given by Holt & Lee (1901), it is shown that in the case of a
negative organism, such as Stentor, light of supraoptimal intensity,
like that coming from the window, must be assumed to cause increased
contraction of the cilia. After the organism has passed across the line
x-y, or while it is passing across this line, it has the anterior end directed
away from the source of light ; according to the tropism theory this
is a stable position and should not be changed. For, supposing the
organism swerves a little toward either side, the cilia on that side will
be more strongly affected by the light, so that the animal will at once
be turned back into the position of equilibrium with anterior end directed
away from the light.

Nevertheless, under these circumstances the organism does turn and
swim back into the darkened area. An explanation for the apparent
movement of a negative organism against the direction of the light rays
is sometimes given in the following form : The light from the window
is said to fall upon the side or end of the dish farthest from the window
and is reflected back, so that the chief source of light for the Stentors
is not the window, but the side of the dish opposite the window. The
animal therefore becomes oriented with relation to this source of light
and swims away from it.

Comparison of the movements of the Stentors in the darkened area
A with those in the lighted area B shows that this explanation can not
possibly be correct. Consider an individual at the point <5, Fig. 13,
which turns and swims toward the window into the dark region. It
is affected by light from two sources, (i) from the window, (2) reflected
from the side opposite the window. According to the above theory
the turning is due to the fact that the light from the opposite side is of
greater strength than that from the window (in itself a most improbable
suggestion). Compare this Stentor b with an individual at at, in the
darker region. This animal receives no direct rays from the window,
yet does receive the reflected rays from the opposite side. If these
reflected rays are sufficient to cause b to become oriented in spite of
the opposing rays from the window, they must produce the same effect,
a fortiori^ on the individual «, since they are the only rays which
reach it Yet individuals in the position a do not become oriented at
all. The individuals in the shaded portion of the vessel swim about
in all directions, without relation to the direction of the light rays. It
is only when they come to the line x-y.^ where they would pass into the

40 THE BEHAVIOR OF LOWER ORGANISMS.

region lighted directly from the window, that they react by turning
toward the right aboral side and passing back into the shadow. It is
thus clear that it is the light coming from the window to which they
react, not the light reflected from the sides of the dish. We have here
realized the condition concerning which there has been so much dis-
cussion, and which has been considered impossible and unrealizable by
various authors — a negative organism reaching the darker region by
swimming toward the source of strongest light.

This would of course be quite inexplicable on the tropism theory as
set forth by Holt & Lee. What does it indicate as to the real nature
of the reaction? To this inquiry there can be but one answer. The
organism reacts on passing from a darker to a lighter area, without
regard to the direction from wliich the light comes. It reacts to the
increase in the amount of light falling upon it as compared with the
condition an instant before it had passed into the lighted area. The
reaction takes the usual form — a backing and turning toward the right
aboral side, followed by a forward motion. The organism, therefore,
is directed again toward the shaded area, which it enters.

In all our experiments thus far there have been marked differences
in the illumination of different areas. Let us now arrange the condi-
tions so that light comes from one side, and all parts of the vessel are
equally illuminated. This may be done by placing the Stentors in a
glass vessel with plane walls at one side of a source of light, such as a
window or the bulb of an incandescent electric light. The Stentors,
after a very short interval in which the reaction seems indefinite, swim
away from the source of light, thus gathering at the side away from
the window, where they move about in a disordered way. During the
reaction the Stentors are oriented^ with the longitudinal axis in the
general direction of the light rays and with the anterior end away from
the source of light.

Thus while it is true that the direction of the rays of light has little
if any effect on the reaction when the animals are at the same time
subjected to a sudden change from dark to light, it does determine the
direction of movement when acting alone. In order to discover just
how the reaction occurs it is necessary to observe the animals at the
moment when they change from their former undirected swimming to
the movement away from the source of light.

For determining this a large number of Stentors are placed in the
dish next the window on a dark background. The light comes from
one side and a little from above. The direct rays of the sun were not
employed.

Above the glass vessel are focused the lenses of the Braus-Driiner

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 4I

stereoscopic binocular. This gives a magnification of 6^ diameters,
with a working distance of 3 cm., and permits exact observation of
the movements of the individual Stentors. To one who has worked
only with the monocular microscope, the use of the stereoscopic binocu-
lar in studying the movements of small organisms will be a revelation.

The vessel containing the Stentors is first covered with a dark screen
and the Stentors are allowed to become equally distributed throughout
the dish. The screen is then raised, allowing the light from the window
to fall upon the Stentors. Those which are swimming in any other
direction than away from the window now turn and in a short time
are swimming toward the side of the dish away from the window.

With the Braus-Driiner the movements of individuals are observed
at the moment of removing the screen. Some turn at once, while most
continue for a few seconds in the direction in which they are swimming
and then turn. All turn in every case toward the right aboral side.
The turning is continued or repeated until the anterior end is directed

tes

Fig. 15.*

away from the window ; then the direct course is continued, carrying the
Stentor to the side of the dish away from the window. The direction
of tur7iing is thus determined by an internal factor — the structure
of the body.

The behavior of the Stentors may be controlled and studied more
exactly by a diflferent order of experimentation. The animals are
placed in a shallow rectangular glass vessel on a dark background, in
a room that is entirely dark save for two incandescent electric lights
A and B (Fig. 15). These are clamped in position, one on each side
of the dish containing the Stentors, and about eight inches from it.
Both these lights can be turned on at once ; both can be extinguished
or one can be turned on while the other is turned off. When only one
is turned on the direction of the light can be instantly reversed by
simultaneously extinguishing this one and turning on the other.

With both lights extinguished the Stentors in the vessel are allowed
to become equally distributed ; then B is illuminated. In a short time

♦Fig. 15. — Method of testing the reaction of Stentor to light. A and B are
incandescent electric lights.

42 THE BEHAVIOR OP LOWER ORGANISMS.

most of the individuals have gathered at the side next to A^ as in
Fig. 15. Then B is extinguished, while at the same time A is illumi-
nated. The Stentors then turn and move toward B. They may be
stopped at any point in their course and the direction of swimming
reversed by simultaneously turning off one light and turning on the
other. With a sensitive culture the phenomena take place with con-
siderable precision, about four-fifths of the individuals responding
quickly to every reversal of the direction from which the light comes.

Under these circumstances it is easy to observe the individuals at the
moment of the reversal of the course. The observation already made
is confirmed ; the animals always turn at the moment of reversal toward
the right aboral side. The reaction is thus of the same sort that occurs
when there is a sudden increase in illumination. After the first reaction
the anterior end is pointed in a new direction. If this new direction
is away from the source of light the animal swims forward in the
course so laid out. If, as is usually the case, the first reaction does
not result in directing the anterior end away from the source of light,
the reaction is repeated, and this may occur several times. Thus the
anterior end becomes directed successively toward every quarter ; as
soon as it lies toward the side opposite the light the reaction ceases.
The animal now swims straight ahead (that is, in a spiral with a
straight axis) away from the source of light.

Thus while it is clear that light falling from one side produces a
well-defined orientation, this orientation does not take place in such a
way as to be in accordance with the tropism theory as set forth, for
example, by Holt & Lee. It is not the direct action of the light on
the motor organs of the side on which it impinges that determines
the direction of turning, but the latter is due to an internal factor.
This becomes still more evident when the conditions are so arranged
that the direction of turning demanded by the internal factor is the
opposite of that required by the tropism theory.

These conditions can be fulfilled in the following manner : The
light to be turned on (Fig. 15) is so moved beforehand that its rays
shall fall, not directly on the anterior end of the Stentor, but obliquely
at an angle to the path they are following. The animals then react
as before, by turning toward the right aboral side. It often happens
that this involves first a direct turning toward the light, as illustrated
in Fig. 16. In such a case the turning is continued or repeated until
the anterior end is directed away from the source of light. We have
seen the same result produced under similar conditions in the experi-
ments illustrated in Fig. 13.

What is the real stimulus to the production of the motor reaction

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES.

43

which results in orientation? The experiments directed precisely
upon this point show that the stimulus producing the motor reaction
is an increase in the intensity of light upon the sensitive anterior end.
Now, in the reaction to a continuous light coming from one side, the
conditions are present for exactly such changes in the intensity of light
at the anterior end as would induce the observed reactions. In the
spiral course the animal swerves successively in many directions. In
certain directions the swerving subjects the anterior end to a more
intense illumination. This change acts as a stimulus to produce the
motor reaction, which carries the anterior end elsewhere. In other
directions the swerving leads to a decrease in the intensity of light
affecting the anterior end. In this case no reaction is produced, and the
organism continues to swim in that general direction. The details of
this method of reacting
will be given in the ac-
count of the reactions of
Euglena, where the mat-
ter was subjected to care-
ful analytical experimen-
tation. The evidence all
indicates that the condi-
tions in Stentor are ex-
actly parallel to those in
Euglena.

We may sum up our
results on Stentor as fol-
lows : A change from

dark to light, such as is caused by swimming from a shaded into an
illuminated region, acts as a stimulus to produce a typical motor reaction ;
the Stentor backs and turns toward the right aboral side, so that it
returns into the shaded region. A change in the illumination of the
anterior end produces the same effect as a change in the illumination
of the entire organism. The direction from which the light comes has
no observable effect on this reaction. But when the illumination is
uniform and the light comes from a definite direction, then light fall-
ing on the anterior end of the Stentor causes the reaction, while light
falling upon the posterior end causes none. The result is that the
animal turns (toward the right aboral side) until its anterior end is

Fig. i6.*

♦Fig. i6. — Method by which Stentor becomes oriented to light, when the light
falls on the aboral side of the animal. Stentor turns, as shown by the arrows,
at first toward the light, but the turning is repeated or continued until the
anterior end is directed away from the light

44 THE BEHAVIOR OF LOWER ORGANISMS.

directed away from the source of light, and swims in the direction so
determined. The reaction to light is of essentially the same character
as the reaction to other usual stimuli, and takes place by what we may
call the method of trial and error. When the animal comes to the
boundary of a lighted area, or when the anterior end is illuminated,
this constitutes error ; the animal tries some other direction, and repeats
the trial till the condition constituting error disappears.

Are these results in agreement with all the observed facts.? The
only point on which perhaps question might arise is in regard to the
production of a clearly marked orientation such as we find shown by
Stentor when the light falls upon it from one side. In this case, as

Fig. 17.*

we have seen, Stentor swims directly away from the source of light,
and shows thus a typical orientation. As we have had the dictum
that a motor reaction, such as I have described, " cannot account for
an orientation" (Garrey, 1900, p. 313), it will be well to examine this
matter a little farther. In a previous paper (Jennings, 1900, a) I have
shown how orientation could be produced through a motor reaction ;
the case of Stentor exactly realizes the possibility there set forth. If

♦Fig. 17. — Diagram to illustrate the difference between the method of orienta-
tion tp light required by the tropism schema and that which actually takes
place. To light coming from the direction shown by the straight arrows the
tropism schema requires that an organism in the position x-y should attain the
position y-z by turning in the direction indicated by the (^broken) arrow a-i>.
The position is actually attained by turning in the direction indicated by the
long arrow c-d.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 45

the organism is at first not oriented to lines of influence coming from a
certain direction, as in Fig. 17, x-y^ and then becomes oriented, as at
Fig. 17, y-z^ there are clearly more ways than one by which the orienta-
tion can be produced. The essential question for deciding as to the
nature of the reaction is not whether orientation occurs, but how the
orientation is brought about. This consideration has been too often
lost sight of in discussions of the behavior of the lower organisms.

According to the theory of tropisms, as defined by Verworn, Loeb,
and Holt & Lee, the orientation should be brought about by the differ-
ential action of the external agent on the different sides of the organism ;
the organism should turn directly into the line of action of the external
agent, and the direction of turning should be determined by an external
factor, the direction of the infalling rays, or the side on which they
strike the organism. Now this is a matter which can be settled by
direct observation. Direct observation shows us in Stentor that orien-
tation is not brought about in the manner demanded by the theory.
The direction of turning is determined by internal fiictors. The reac-
tion which produces orientation is identical with the typical reaction
to a mechanical shock, to chemicals, to heat and cold. The difference
between what is demanded by the theory of tropisms and what is
actually observed may be made quickly evident to the eye by Fig. 17.
According to the theory of tropisms the orientation of a negatively
phototactic organism should take place by turning in the direction of
the arrow a-b\ in a Stentor in the position shown {x-y')^ orientation
actually occurs by turning in the opposite direction, as shown by the
arrow c-d.

The further question then arises as to why the organism remains
oriented. All the facts point, in the case of Stentor, to the conclusion
that the reaction to a constant light is due to the intense illumination
on the sensitive anterior end. As soon, therefore, as the anterior end is
turned away from the light, as is the case in the position y-z^ Fig. 17,
there is no further cause for reaction ; the animal therefore remains
with its anterior end directed away from the light ; that is, it remains
oriented. If, as a result of reaction to some other stimulus, or in any
accidental manner, the animal comes into a position such that it is no
longer oriented, the '* motor reaction" is repeated until the animal
comes again into the position of orientation in which it is no longer
stimulated.

How does the method of reaction to light here described for Stentor
agree with what we know of light reactions in other ciliates.? As
noted in the introductory paragraphs, comparatively little is known as
to light reactions in this group of organisms. The observations of

1 i I I

46 THE BEHAVIOR OF LOWER ORGANISMS.

Engelmann (1882, a) on the light reactions of certain green ciliates
(^Paramecium bursaria^ Stentor viridis^ etc.) were made before the
typical motor reaction — the turning toward a certain structurally de-
fined side— had been observed in any of the infusoria. Engelmann,
therefore, paid no attention to this point. Yet there is much in his
account of the reactions to light in these organisms to suggest that
it takes place in a way similar to that which I have described above
for Stentor cceruleus. Indeed, Engelmann's account, so far as it
goes, fits precisely into the reaction method which I have described
above. He found, as I have, that the organisms react either when only
the anterior end is aflfected, or when the entire organism is flooded with
light from beneath. The reaction consists in a sudden turn to one side,
or a sudden start backward, just as in Stentor cceruleus. The only
point which is lacking in Engelmann's account is the observation as

to which side the organism
turned ; to this point he did
not direct his attention.

It is interesting to note that
in the account given by Ver-
worn (1889, Nachschrift) of
the reaction to light in Pleu-
ronema chrysalis there is
p, T,„ "" nothing tending to support

the theory of an orienting tro-
pism. According to Verworn the reaction of Pleuronema to light is by
a sudden leap (" Sprungbewegung"), which is repeated several times if
the light continues. This sudden leap seems identical with the *' motor
reflex" which I have described as the typical reaction to stimuli in
many ciliates, and which consists usually in a leap backward, followed
by a turning toward a structurally defined side. It is in this manner,
as we have seen, that Stentor cceruleus reacts to light and the reac-
tion, as in Pleuronema, is often repeated many times.

Thus the other carefully studied accounts of reaction to light in the
Ciliata, while incomplete, agree so far as they go with that which I
have given for Stentor, and contain nothing to suggest the idea of an
orienting tropism dependent upon unequal stimulation of the motor
organs on the opposite sides of the animal.

♦Fig. 18. — Diagram of the reaction of Stentor to Hght, after Holt & Lee.
Stentors are confined in a vessel behind a wedge-shaped prism containing a
substance which partly cuts off the light, so that one end of the vessel is darker
than the other. The usual course of a Stentor near the lighter end is shown by
the broken line.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES.

47

Davenport's reference (Davenport, 1897, p. 189) to the negative
light reaction of Stentor makes no attempt to explain the mechanism of
the reaction. Holt & Lee (1901) have given an account of some fea-
tures of the light reaction of Stentor cceruleus. They did not attempt
to determine directly the mechanism of the reactions, by observation
of the exact movements of the organism. Specifically, they made no
observations to determine w^hether Stentor becomes oriented by turn-
ing directly aw^ay from the source of light or only indirectly through
a " motor reaction " such as I have described. They did attempt, how-
ever, to show that the gross phenomena observed might be interpreted
in accordance with the prevailing theory of tropisms set forth on page
7 of the present volume. It will be
well, therefore, to examine their obser-
vations in order to determine whether
they contain anything inconsistent with
the account set forth in the present
paper.

Holt & Lee studied the behavior of
Stentor in an elongated trough which
was lighted from one side. The light
passed through a prism which con-
tained a translucent fluid (a weak solu-
tion of India ink), by means of which
a portion of the light was cut out (Figs.
18 and 19).

At the thicker end of the prism more
light was cut out, hence this end of the
trough (Fig. 19, D) was darker than
the opposite end (Z). It was found
that when Stentors were placed in the trough close behind the prism
(ato, Fig. 19) they turned and swam away from the lighted side till the
back of the trough was reached {a to </, Fig. 19) . This is of course ex-
actly what happens when no prism is interposed. Reaching the back of
the trough the animals give the motor reaction (by backing, then turning
toward the right aboral side), thus coming into either the position e or the
positiony (Fig. 19). They then swim forward again, strike the wall,

Fig. 19.

* Fig. 19.— Reaction of such an infusorian as Stentor to light, under the con-
ditions shown in Fig. 18. After Holt & Lee. The animal in the position *-y,
close behind the prism, turns and swims to the position </, where it comes against
the rear wall of the trough. It then turns either into the position e, toward the
darker end Z>, or into the position/, toward the lighter end L. In the latter case
it usually soon reacts again, and by repetition of the reaction it finally, as a rule,
becomes directed toward D. Thus, finally, most of the Stentors collect in the
dark end of the trough.

4$ THE BEHAVIOR OF LOWER ORGANISMS.

and repeat the reaction. This is repeated many times until the organ-
isms are swimming either toward the end D or toward the end L. In
course of time it is found that the preponderance of movement is toward
the dark end Z>, so that the majority of the Stentors are gathered at D.
Why this should be so is explained by Holt & Lee as follows :

The reason why the Stentors went eventually in greater numbers toward D,
and thus appeared oftener to choose e than /", is that such Stentors as went to e
progressed farther toward D than those which went to/" could progress toward
L,. These latter would soon strike the wall a second time, now pretty nearly at
right angles, and during the recoil the light stimuli would favor a return to d.
It appears then amplj' possible that the circumstance that the organism encoun-
ters the wall of the trough at an acute angle is sufficient to cause its farther
progress to be, in the long run, toward D.

There is evidently nothing in this account which is inconsistent with
the method of light reaction which I have described. On the contrary,
the reason why the organisms finally swim toward the dark end and
gather there becomes much more evident when the reaction method
that I have described is taken into consideration. Let us suppose that
the Stentors, after striking the back of the trough, turn in equal numbers
toward D and toward Z. In those swimming toward D the anterior
end is directed away from the source of strongest light (due to reflection
from the lighted end of the dish /,), and the animals are passing into a
region of less intense light. There is thus nothing to cause the " motor
reaction," with its accompanying change in the direction of movement.
In the Stentors swimming toward Z, on the other hand, the strongest
light falls on the anterior end, and the organisms are passing into a
region of more intense light. Either of these factors taken separately
may, as we have seen, cause the motor reaction (the turning toward
the right aboral side), thus changing the direction in which the Stentors
swim. The animals which start to swim toward L will therefore soon
be turned, and only when the direction of movement is toward D will
there be no cause for further change.

The observations of Holt & Lee are thus quite in harmony with
the reaction method which I have described, and indeed receive
illumination when this reaction method is taken into consideration.

In the " fourth case" discussed by Holt & Lee {loc. cit.^ pp. 475"
478), the two factors mentioned as determining the turning of the
Stentors away from the end L would work in opposite directions ; only
experience can tell which would be more eflective. As Holt & Lee
do not state specially that they observed the reactions of Stentor under
these conditions no comment is required. Experiments of this character
will be further considered after we have described reactions to light
in flagellates.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES.

49

THE FLAGELLATA.

In the following pages we shall examine the method of reaction to
light in the flagellates Euglena viridis^ Cryptomonas i

ovata^ and a species of Chlamydomonas.

EUGLENA VIRIDIS.

Euglena viridis swims in a spiral path, continu-
ally swerving toward that side which bears the larger
'' lip " and the eye, the so-called dorsal side (Fig. 20).
Its motor reaction to most stimuli is by a sudden pro-
nounced turning toward the dorsal side ; that is, by
swerving still farther toward the same side toward
which it swerves in its normal swimming. Thus the
direction of its path is changed (Jennings, 1900).

The general features of the reaction of Euglena to
light have been well worked out by Englemann
(1882, a) and Wager (1900). These authors show
that Euglena collects in lighted regions. The organ-
isms pass into a lighted area without reaction. But
on coming to the outer boundary of such an area,
where they would pass out into the dark, they react
by turning round and passing back into the light.
The collections of Euglenae in lighted areas are thus
brought about in much the same manner as the col-
lections of Paramecia in regions containing a weak
acid (Jennings, 1899). If diffuse light falls from one
side on water containing Euglenas, the organisms swim
toward the source of light. But if strong sunlight
falls upon them they swim away from the source of
light.

Engelmann showed that the colorless anterior end
is the part that is chiefly sensitive to variations of light.
Often the organism in a lighted area, on reaching the
edge, reacts by turning when only the colorless tip
has passed into the darkness.

The precise method of reaction to light, the direc-
tion of turning in becoming oriented or in passing
back into the lighted area, was not worked out by the authors named.
To this point we shall direct our attention.

When a large number of Euglenae are swimming toward the source
of light, if the illumination is suddenly decreased in anyway, they give

Fig. 20*

♦ Fig. 20 shows the spiral path of Euglena in its ordinary swimming.

50 THE BEHAVIOR OF LOWER ORGANISMS.

the typical motor reaction described in my previous paper as a response
to other classes of stimuli (Jennings, 1900, p. 235). That is, they
turn at once toward the dorsal side (that bearing the larger lip and the
eye). This is very easily seen when the Euglenae are mounted in the
ordinary manner in a thin layer of water on a glass slide and observed
with the microscope in the neighborhood of a window. If the hand
is interposed between the slide and the window all the Euglenae react
in the way just described.

The reaction is a very sharp and striking one and produces a very
peculiar impression. At first all the Euglenae are swimming in parallel
lines toward the window. As soon as the shadow of the hand falls on
the slide the regularity is destroyed ; every Euglena turns strongly and
may seem to oscillate from side to side in the manner described later.

The turning is often preceded by a slight movement backward.
This was not observed in the reactions to other stimuli (Jennings, 1900,
p. 235), though it agrees with what we find in most other ciliates and
flagellates. In Euglena the reaction to variations in the intensity of
light seems more sharply defined than to most other stimuli. The fact
that the turning is always toward the dorsal side is observable with the
greatest ease. It is particularly evident when the organisms are con-
fined to a thin layer of water, so that they cannot swerve up or down,
but only to the right or left.

The reaction occurs whenever the light is suddenly decreased in any
way. Certain different conditions under which it occurs deserve special
mention, (i) As we have seen, the reaction occurs when a screen is
brought between the organisms and the source of light toward which
they are swimming. (2) It also occurs when the illumination is de-
creased by cutting off light from some other source than that toward
which they are swimming. Thus the organisms on the stage of the
microscope may be lighted from below, by the substage mirror, and at
the same time may receive light from the window at one side of the
preparation. They swim toward the window, since the light from that
quarter is much stronger than that from below. If now the light from
below is suddenly decreased by closing the iris diaphragm, the Euglenae
react as usual by turning strongly. This is notwithstanding the fact
that the proportion of light coming from the window, to which they
were oriented, is now greater than before, so that it might be supposed
that they would remain more strongly oriented than ever. For the rest,
the disturbed orientation is soon restored. (3) The reaction occurs
when the decrease in illumination is due to the movements of the
Euglenae ; that is, when the swimming organisms come to the edge of
a lighted region where they would, if the course were continued, pass
into the darkness. As a result of the reaction they return into the light.

REACTIONS TO LIGHT IN CII.IATES AND FLAGELLATES. 5 1

The reaction occurs at a decrease in illumination not only when the
organisms are oriented and swimming toward the source of light, but
also when they are not oriented and are merely scattered in a weakly
lighted area. Further, in cases where most of the Euglense are oriented
and swimming toward a source of light, a number of specimens will
always be found that are not oriented at all, or are swimming away
from the source of light. Such individuals react to a sudden decrease
in illumination in the same manner as do the specimens that are oriented
with the anterior end toward the source of light. This result may be
observed in a curious way as a consequence of the fact that it requires
some time for the light to produce its orienting effect. Thus, if the
Euglenae are placed between a weak and a strong light they swim toward
the strong light. If, now, the strong light is cut off, they react in the
usual way and swim toward the weak light. Now the strong light
may be restored ; the Euglenae continue for a few seconds to swim toward
the weak light, thus away from the strong light. If while they are
swimming in this manner the strong light is cutoff, the Euglenae, swim-
ming away from it, react in the usual manner, by turning strongly
toward the dorsal side.

The usual reaction may be produced by a decrease in illumination
that is not sufficient to cause a permanent change in orientation. Thus
the Euglenae on a slide or in a shallow dish may be lighted from a
window at one side. By passing a small screen in front of the window
at some distance from the preparation a portion of the light is cut off;
the Euglenae then respond in the usual way, by swerving toward the
dorsal side. The movement thus becomes very irregular. Since the
Euglenae continue to revolve on their long axes the dorsal side may lie
first to the (observer's) right, then to the left. The Euglenae all seem,
therefore, to vibrate from side to side. This is the *' Erschiitterung "
or trembling described by Strasburger (187S) as occurring in swarm-
spores when the illumination is changed ; it will be understood better
when we have considered more in detail the mechanism of the reactions.
Meanwhile the screen retains its position, but still admits more light
from the direction of the window than from any other direction. The
reaction of the Euglenae, therefore, soon ceases ; their orientation is
restored in the way to be described later, and they continue to swim
toward the window.

This experiment is an important one. It shows that the typical reac-
tion may be produced by a decrease in light that is not sufficient to
permanently destroy the orientation. Thus it is clearly the decrease
in illumination to which the organisms react ; not to a change in the
direction of the light rays. The experiment shows further that it is
not the absolute amount of light that determines the reaction. Some

53 THE BEHAVIOR OF LOWER ORGANISMS.

time after the decrease in illumination takes place the organisms behave
just as they did before, swimming in the same direction. Further,
the illumination may be decreased very slowly to the same extent
without causing a reaction. If the screen is at first far away from the
preparation and is then slowly moved to the position it occupied in
the experiment just described no reaction is produced. It is only the
sudden change that has caused the reaction. The change, however,
need not be a very marked one in order to be effective.

Our experiments thus far have shown that in a moderate light Eu-
glena reacts to a decrease in illumination. But the absolute amount of
light present has an effect on the reaction. If the light is very strongly
increased the same reaction is produced as when the light is decreased.
If while the organisms are swimming toward a moderately lighted
window direct sunlight is allowed to fall upon them, they respond in
the same way as to a sudden decrease in illumination ; that is, they
turn strongly toward the dorsal side, continuing or repeating the re-
action till the anterior end is directed away from the source of light.
They now continue to swim in that direction, the positive reaction
having been transformed into a negative one. Thus under intense
light the conditions of stimulation are the opposite of those under
moderate light. This is paralleled in the reactions of the infusoria to
chemicals ; often a strong solution of a certain chemical produces a re-
action under opposite conditions from those in which a weak solution
of the same chemical is effective.

Let us now proceed to a more careful study of the reaction itself.
The reaction which occurs when the illumination is changed is really
an accentuation of a certain feature of the usual movements. Euglena,
as we know, revolves on its long axis as it swims forward, and at the
same time it swerves toward the dorsal side. The resulting path is
therefore a spiral one (Fig. 20). The usual reaction to a stimulus is
an accentuation of this normal swerving toward the dorsal side, as com-
pared with the other factors in the swimming; the organism suddenly
swerves so much farther than usual in this direction that the path may
be completely changed. If the reaction is a very decided one the revo-
lution on the long axis and the movement forward may cease during
the swerving toward the dorsal side ; the anterior end then describes
the arc of a circle about the posterior end as a center. In a less pro-
nounced reaction the revolution on the long axis continues. The circle,
described by the anterior end is then less and the whole body describes
the surface of a cone, or a frustum of a cone, as illustrated in Fig. 21.
Every gradation exists between the normal spiral course and the strong
reaction in which the anterior end swings in a circle about the
posterior end as a center.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES.

53

When oriented and swimming toward the source of light the swerv-
ing toward the dorsal side is comparatively slight. As seen from
above, the organisms seem merely to oscillate a very little from side to
side as they revolve on the long axis. Careful examination shows that
the swerving is always toward the dorsal side, as in Fig. 20, the alter-
nations in direction being due to the alternations of position of the
dorsal side. Now, when the illumination is suddenly decreased, the
Euglenas at once swing much farther than usual toward the side to

Fig. 21.*

which they are already swerving, that is, toward the dorsal side. If
the decrease in illumination is not very great, so that the stimulus is
not a strong one, the swerving is not very great, and the organism at
the same time continues to revolve on its long axis ; thus the anterior
end describes a circle and the whole body describes the surface of a

*FiG. 21. — Diagram to illustrate reaction of Euglena when the illumination
is decreased. The Euglena is swimming forward at i ; when it reaches the
position 2 the illumination is decreased. Thereupon the organism swerves
strongly toward the dorsal side. This swerving, combined with the revolution
on the long axis, causes the anterior end to swing about a circle, so that the
Euglena occupies successively the positions 2, 3, 4, 5. 6, etc. From any of these
positions it may start forward, as indicated by the arrows, if the condition
causing the reaction ceases to act. In the figure the Euglena is represented as
swimming forward from the position 6.

54

THE feKHAVIOR OF LOWER ORGANISMS.

cone, or the frustum of a cone, as indicated in Fig. 21. The result, as
seen from above, is that all the specimens seem to vibrate from side to
side ; in other words, they are taken with a sudden oscillation or trem-
bling. This oscillation when the intensity of the light is suddenly
changed was observed by Strasburger (1878,
pp. 25 and 50) in flagellate swarm-spores ; he
speaks of it as *' Erschiitterung " or " Zit-
tern." During this oscillation the anterior end
becomes pointed successively in many different
directions, as Fig. 21 show^s. When, now, the
usual forward course is resumed (with only the
usual amount of swerving toward the dorsal
side), the animal follows one of these directions.
Thus its path is changed (Fig. 22). Strasburger
(1878, p. 25) noticed that the path followed after
the oscillation was oblique to the former path.
As a study of Figs. 21 and 22 will show, this is a
necessary consequence of the increased swerving
toward the dorsal side, to which the oscillation
itself is due. All these relations become much
clearer if a model of an actual spiral is studied ;
it is difficult to represent them upon a plane
surface.

If the stimulus is stronger, as when there is a
greater decrease in illumination, the swerving
toward the dorsal side is much greater ; the or-
ganism wheels far to that side, so that the spiral
course seems entirely interrupted. But there is
really nothing in this reaction differing in prin-
ciple from what is happening in the normal
forward swimming. If the swerving toward the

f dorsal side is long continued the specimen may

be seen to swing first far to the (observer's) right,

^ ^ then, after it has revolved on the long: axis, far

Fig. 22.* ^ '

to the (observer's) left; in reality it swings an

equal amount upward and downward and in intermediate directions.

It may, however, swing at once so far to the dorsal side that the new

* Fig. 22. — Shows the spiral path of Euglena, illustrating the eifect of a
slightly marked reaction. At a the illumination is decreased; the organism
therefore swerves toward the dorsal side, causing the spiral to become wider.
At b the ordinary method of swimming is resumed; since at this point the
organism was more inclined to the axis of the spiral than before the reaction,
the new course lies at an angle with the previous one. Compare with Fig. 21.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES.

55

course forms a right angle, or a still greater angle, with the original
course ; if the turning is through iSo°, the course will be squarely
reversed. Indeed, sometimes the organism swings around an entire
circle or more. When the usual method of swimming is resumed after
such reactions as those just described, the course has been completely
changed.

Strasburger (1S78, p. 25) noticed that after a decrease in illumina-
tion flagellate swarm-spores often turn strongly to one side or even
describe circles. But he did not notice that the turning was always
toward the same side of the organism,* and did not perceive the relation
between this effect and the remainder of the reaction.

Fig. 23. t

This method of reaction is particularly striking when the Euglenae
are confined to a very thin layer of water between the slide and the
cover glass, so that they cannot swerve up or down. When the light
is decreased, we will suppose that the dorsal side is to the (observer's)
left. The Euglena then swings far to the left. At the same time it

♦Naegeli (i86o, p. 96) had, however, before Strasburger, observed that in such
swarm-spores the same side always faces the outside of the spiral path. This
observation, which really contained the germ of a correct understanding of the
reactions to stimuli, seems hardly to have been noticed by later writers,

t Fig. 23. — Diagram of the method by which Euglena becomes oriented with
anterior end toward the source of light. At i the Euglena is swimming toward
the source of light. When it reaches the position 2 the light is changed so as
to come in the direction indicated by the arrows at the right. As a consequence
of tlie decrease in illumination of the anterior end thus caused, the orijanism

56 THK BEHAVIOR OF LOWER ORGANISMS.

revolves on its long axis, bringing the dorsal side down. Since it can
not swing downward, owing to the narrow space, this has little effect
on the reaction, save to stop the movement to tlie left. Now, by con-
tinued rotation the dorsal side has come to lie to the (observer's)
right ; the Euglena may then be seen to swing far to the right. In each
case under these conditions it is at once evident by observing the
larger lip at the anterior end that tiie organism is swinging toward
the dorsal side.

This method of reaction is very effective in preventing Euglena from
passing from an illuminated region to a shaded one. As soon as the
anterior end enters the shadow, the animal swings far toward the dor-
sal side till the anterior end is brought again into the light, repeating
the reaction if necessary. There is then no further cause for reaction.
The reaction to a very strong increase of illumination is, as we have
seen, identical with that to a decrease in illumination.

In our experiments thus far we have directed attention primarily to
the effects of changes in the intensity of illumination, and have found
that such changes produce a motor reaction independently of the direc-
tion of the light rays. But it is of course well known that Euglena
does react with reference to the direction of the light rays. Euglenae
swim toward the source of light when weakly illuminated, away from
the source of light when strongly illuminated. If Euglenae are swim-
ming at random in a diffuse light they soon become oriented when the
light is allowed to act on them from one side, even if the intensity of
illumination remains the same. Or, if Euglenae are swimming toward
a source of very weak light and a stronger light is allowed to act upon
them from the opposite side, they become oriented, in time, with
anterior ends toward the stronger light. In examining this dependence
of the direction of swimming on the direction of the rays of light, we

swerves strongly toward the dorsal side, at the same time continuing to revolve
on the long axis. It thus occupies successively the positions 2, 3, 4, 5, 6. In
passing from 3 to 6 the illumination of the anterior end is increased; hence the
reaction nearly or quite ceases. In the next phase of the spiral, therefore, the
organism swerves but a little toward the dorsal side — from 7 to 8. But this
movement causes a decrease in the illumination of the anterior end, and this
change induces again the strong swerving toward the dorsal side. Hence in
the next phase of the spiral the organism swings through 9 and 10 to 11. In
this movement again the illumination of the anterior end is increased; hence
the reaction ceases, so that from 12 the organism swerves only as far as 13.
Then owing to the decrease in illumination caused by this movement, the
swerving increases, so that the Euglena swings from 13 through 14 and 15 to 16.
Now it is directed toward the source of light, and such swerving as takes place
in the spiral course neither increases nor decreases the illumination of the
anterior end. Hence there is no further reaction; the Euglena continues to
swim forward in the direction 16-17.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES.

SI

shall have to keep in mind two questions : First, how is the position
of orientation brought about? Second, what is the real stimulus in
producing orientation?

To answer the first question we must observe the movements of the

■^^^^

-^-

)a

^

\h

V

Fig. 24.*

organism at the time orientation occurs. Observation of the individ-
uals as they are becoming oriented shows that orientation is brought
about through the same motor reaction that we have already described ;

* Fig 24. — Path followed by Euglena when the direction of the light is
changed. From i to 2 the organism swims forward in the usual spiral path
At 2 the position of the source of light is changed, so that it now comes from
behind. The organism then begins to swerve farther than usual toward the

58 THE BEHAVIOR OF LOWER ORGANISMS.

that is, by a turning toward the dorsal side. The simplest case is per-
haps that of the reversal of orientation, produced when strong sunlight
is allowed to fall from in front upon specimens that are swimming
toward a diffusely lighted window. Under these circumstances, as
we have seen, the Euglenae turn toward the dorsal side, changing their
course. They may turn directly through 180°, in which case they are
at once oriented with anterior ends away from the light ; but usually
the orientation is less direct than this. The reaction is generally
repeated several times. Through its continued swerving toward the
dorsal side, combined with the revolution on the long axis, the organism
directs its anterior end successively in every direction. When the
anterior end has finally come into a position where it points away from
the strong light the reaction ceases, and the organism swims forward
in the usual way. The details of the orienting reaction will be brought
out more fully in the following account of the way in which the anterior
end becomes directed toward a source of light of moderate intensity.
Let us now take a case in which the change in the direction of the
rays of light is not accompanied by a change in the intensity of illumi-
nation. Euglense are swimming about at random in a diffuse light
when all the light is allowed to fall upon them from one side. They
then become oriented, with anterior ends directed toward the source of
light. Or, the organisms are swimming toward a source of light when
the direction of the light rays is changed or reversed by quickly
moving the source from which the light comes. The Euglenae then
after a time become reoriented. Under such circumstances there is no
sudden, decided reaction, such as occurs when the illumination is
suddenly decreased. The organism merely begins to swerve farther
toward the dorsal side than usual. Thus the spiral has become wider,
and the anterior end comes to be pointed successively in many dif-
ferent directions, as illustrated at 1-6 in Fig. 23. In some of these
positions the anterior end is directed farther away from the source
of light, as at 3 ; in other positions more nearly toward the source
of light, as at 6. In the latter case the swinging toward the dorsal
side becomes less marked ; hence the succeeding phase of the swing,
which carries the anterior end away from the light, is less pronounced ;

dorsal side, owing to the decrease in the illumination of the anterior end. Thus
the spiral becomes wider, a and b showing the limits of the swerving. At 3 the
normal amount of swerving is restored, so that the new path is at an angle with
the old one. Now the organism swerves at each turn of the spiral a short dis-
tance away from the source of light, as at c, e, g^ and a longer distance toward
the source of light, as at </,/", A, for the reasons shown in Fig. 23. At h it has
in this manner become directed toward the source of light, and there is no fur-
ther cause for swerving more to one side than to the other; it therefore swims
in a spiral with a straight axis toward the source of light.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 59

the anterior end therefore does not swing so far in the direction
away from the light as in the preceding phase it swung toward the
light. This is illustrated at 7-8 in Fig. 23. But as a result of such
swerving as does occur the anterior end is now (at 8) directed more
away from the source of light than before. There then follows a
new reaction, with increased swerving toward the dorsal side in the
next phase of the spiral (8-1 1, Fig. 23), which carries the dorsal side
toward the source of light. Hence the anterior end swings still further
toward the position where the light shines directly upon it. This con-
tinues. As a result of this repeated swinging of the dorsal side slightly
away from the source of light and strongly toward the source of light
the organism gradually changes its course, continuing to swim in a
spiral and to swerve toward the dorsal side, until the axis of the spiral
is in line with the light rays and the anterior end is toward the source
of light. This method of reaction will best be understood by a study
of Figs. 23 and 24 and their explanation.

Thus the orientation is gradual and for a certain stretch after the
light has begun to act the organism is not completely oriented. With
a fairly strong light, however, the period of time required for complete
orientation is very slight. Strasburger (1878, p. 24) noticed that when
Haematococcus is swimming toward a source of weak light and the
light is suddenly increased so as to reverse the orientation, there is a
period of >' verschiedenen Schwankungen " before the reverse orienta-
tion is attained. He paid little attention to the behavior of the
organisms during this period, however.

Our account has been thus far purely descriptive ; we have attempted
to set forth the events as they may be observed, without trying to
indicate the causes at work. We must now inquire as to what is the
real stimulus and its method of action in producing orientation.

First, we note that in becoming oriented Euglena does not turn
directly toward the source of light. As in the reaction to other stimuli,
the turning is throughout toward a structurally defined side. This
shows that the orientation of Euglena, like that of Stentor, cannot be
accounted for on the orthodox tropism theory. In other words, the
orientation is not due to the direct effect of the light on the motor
organs of the side on which it falls. As in Stentor, orientation may
be reached by turning either toward or away from the source of light,
or in any intermediate direction. The response is a '* motor reaction "
of a definite type.

Just what is the stimulus which produces this motor reaction.'* All
our experiments up to this point have shown clearly that this reaction
is produced by changes in the intensity of illumination, and that a change
in the illumination of the anterior end produces the reaction as well as

6o THE BEHAVIOR OF LOWER ORGANISMS.

does a change in the illumination of the entire body. Indeed, Engel-
mann (18S2, a) showed that a change in illumination over the remainder
of the body is inetTective in producing the reaction, so that in every case
the reaction is due to the change in illumination at the anterior end.
Now, in the orientation reaction the conditions are present for produc-
ing changes of illumination at the anterior end of precisely the character
which would, in view of our other experimental results, bring about
the reactions observed. This will best be shown by again examining
in detail from this point of view a concrete case.

In Fig. 23 we will suppose that the Euglena at i is at first swimming
toward the source of light. When it reaches the position 2 the light is
changed, so that it now comes from the direction indicated by the
arrows at the right. By this change the intensity of illumination at
the anterior end is decreased, since before the light came from directly
in front and affected the entire end, while now it falls upon but one
side. We know from other experiments that as a result of such a
change the organism reacts by swerving more toward the dorsal side,
at the same time continuing to revolve on the long axis. This is ex-
actly what happens now ; by the increased swerving the organism is
carried from position 2 to position 3. In this change the anterior end,
swinging still farther away from the source of light, is still less illumi-
nated than before. As a result of this farther decrease in illumination
the reaction is continued or increased ; combined with the revolution
on the long axis it carries the organism successively to positions 4, 5
and 6. In this part of the movement the anterior end becomes pointed
more directly toward the source of light, and is hence more strongly
illuminated ; there is therefore nothing in this movement to cause a
reaction. The strong swerving toward the dorsal side then ceases or
becomes less. But in the next phase of the spiral course (from 7 to 8),
there is necessarily at least the normal amount of swerving toward the
dorsal side, and this carries the organism to a position (8), where
the intensit}' of the light acting on the anterior end is decreased. As
a result of this decrease we know that the "' motor reaction " must again
be induced ; the organism swings then farther toward the dorsal side*
This movement, combined with the revolution on the long axis, carries
the Euglena through 9 and 10 to 11. Here again the swerving de-
creases, because the change was from a less illuminated to a more
illuminated region. Hence after reaching 12 the Euglena swerves only
a little away from the light, to 13 ; then, as a result of the decrease in
illumination at the anterior end caused by this movement, it swerves
far toward the light, through 14 and 15 to 16. This movement causing
greater illumination, the reaction ceases. The light is now shining full
on the anterior end. The organism therefore swims forward in the

REACTIONS TO LIGHT IN CILTATES AND FLAGELLATES. 6l

usual spiral course, in all phases of which the illumination of the anterior
end is equal. If the light came from the rear of Euglena i instead of
from the direction indicated by the arrows, the reaction above described
would be continued in the same way until the direction of swimming
was completely reversed.

Thus the orientation of Euglena in a continuous light is due to the
production of the " motor reaction," with its turning toward the dorsal
side, whenever there is a decrease in illumination at the anterior end.

There is no other explanation of the orientation, so far as I am able
to see, that is in agreement with all the facts. At first one is tempted
merely to say that the subjection of the anterior end to shadow pro-
duces the motor reaction, and that this is continued until the anterior
end is no longer shaded. This statement is correct if by "subjection
to shadow" we mean an active process, involving a change from a
more illuminated condition. But if we mean that darkness as a con-
tinuous, static condition is the cause of the reaction, then considera-
tion shows that this will not account for all the facts. It leaves out of
account the capability of the organism to become acclimatized to cer-
tain degrees of light and shade, and certain of the experimental results
are crucial against it. Thus, suppose the Euglenae are swimming
toward a source of weak light, and a stronger light is then allowed to
act upon them from another direction. The anterior end continues to
receive the same amount of light as before (since the weak light still
persists), yet the organism reacts as usual, becoming oriented toward
the stronger light. The motor reaction by which the orientation is
brought about cannot therefore be due to darkness or shade (considered
statically) at the anterior end. On the other hand, the case just men-
tioned is easily understood on applying the explanation given above.

Again, it might be held that the reaction is due in some way to the
relative amount of illumination at the two ends. It might be main-
tained, for example, that when the posterior end is more illuminated
than the anterior, this difference acts as a stimulus to cause the " motor
reaction." There is, of course, no independent evidence in favor of this
view, and the experimental results prove it to be incorrect. We have
shown that the reaction is produced (i) when both ends are equally
stimulated, as when the light comes directly from one side ; (2) when
neither end receives light, as when the light is cut off completely. Fur-
ther, it might be held that the reaction is produced when the anterior end
is not more intensely illuminated than the posterior end. It is, of course,
a little difficult to conceive how so indefinite a condition could act as a
stimulus to a definite motor reaction, but in any case the experiments show
that this is not the real cause of the " motor reaction." Thus certain of
the experiments show that the " motor reaction " is produced even when

62 THE BEHAVIOR OF LOWER ORGANISMS.

the light is reduced by the same amount at both ends, so that the anterior
end is still more strongly lighted than the posterior. This case is
realized in the experiment in which a small screen is interposed between
the Euglenae and the window toward which they are swimming. The
light is thus somewhat decreased, but is still sufficient to cause orien-
tation. The anterior end is thus still lighted more than the posterior,
yet the organisms respond with the '' motor reaction" at the moment
the light is decreased. The same thing is shown still more decidedly
in the experiment described on page 50, in which the '' motor reaction "
is produced when the light is cut ofi' from some other source than that
toward which the organisms are swimming. In this case the propor-
tion of light shining on the anterior end is greater after the change in
illumination than before, yet the '' motor reaction" is produced at the
moment the change takes place.

The explanation we have given is, therefore, the only one that is in
agreement with all the facts, and it accounts for every detail of the re-
actions to light. The cause of all the phenomena of light reaction in
Euglena is the fact that a sudden change in light intensity on the anterior
end induces a typical " motor reaction." It is noticeable that the
reaction is throughout due to a dynamic factor, to some change in the
relation of the organism to the light, a change due either to an active
alteration of the environment, or to a movement of the organism. To
static conditions, if not too intense, the organism may soon become
acclimatized, so that no farther reaction is caused. The absolute in-
tensity of the light affects the reaction only in so far as it determines
whether it shall be an increase or a decrease in intensity that causes
the *' motor reaction."

To sum up, the reaction of Euglena, from beginning to end, is ex-
plained by the fact that a sudden change in illumination, even though
slight, causes a definite motor reaction, the essential feature of which
is an increased swerving toward the dorsal side. Orientation is brought
about by the increased swerving in the next phase of the spiral course
when the illumination of the anterior end is diminished, and by the
decreased swerving in the next phase of the spiral when the illumination
of the anterior end is increased. In general terms we can say that the
reaction of Euglena to light is by the method of trial and error. The
organism tries turning in many directions ; when the turning is such
as to produce a decrease in the illumination of the anterior end it
** tries" other directions; when it is such as to produce increased
illumination of the anterior end, or when no change in illumination
results, the reaction ceases and the organism continues to swim forward
in that position. The result of this method of reaction is necessarily
orientation with the anterior end toward the source of light.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 63

CRYPTOMONAS AND CHLAMYDOMONAS.

Cryptomonas ovata is one of the organisms studied by Strasburger
(1878), under the name Chilomonas ovata^ in his classical paper on
reactions to light in flagellates and swarm-spores.

The specimens studied by the present author were mostly of the
'* young" form, having pointed, curved, posterior ends. One side is
strongly convex, while the other is less curved, or is even concave near
the posterior end. It is thus very easy to distinguish the two sides of
the organism and to observe their relation to the movements.

Cryptomonas ovata swims in a rather wide spiral, with the more
convex side toward the outer surface of the spiral. In other words, the
organism swerves continually toward the more convex side. The
response to usual stimuli is a strong turn toward this convex surface ;
this is easily seen when the organism comes in contact with an
obstacle.

The Cryptomonads swim toward or away from the source of light
under the same conditions as Euglena, and gather in lighted areas in
the same manner as does the organism last named. They react to a
sudden decrease in the intensity of illumination by turning toward the
more convex side. If the decrease in intensity is marked, the organism
turns suddenly for a long distance, 90° or more, so that the course is
completely changed. If the stimulus is less the turning toward the
more convex side is not so rapid, and since the revolution on the long
axis is continued the body of the organism describes the surface of a
wide cone or frustum of a cone. When a large number of specimens
react in this way at the same time a peculiar shaking or trembling
appearance is produced; this is evidently what Strasburger (1878)
called " Erschiitterung " or *' Zittern." As a consequence of the wide
swerving, when the normal method of swimming is resumed the course
lies in a new direction.

In all these respects Cryptomonas exactly resembles Euglena. Fur-
ther, the organism becomes oriented to light in precisely the same manner
as is described above for Euglena. In fact, if we substitute '' more
convex side " for " dorsal side " in the account of Euglena, it will fit
almost throughout the reactions of Cryptomonas. It is therefore unnec-
essary to describe the phenomena in Cryptomonas in detail.

A study was made also of the reactions of a species of Chlamydo-
monas. The movements of Chlamydomonas and its reactions to light
resemble those of Euglena and Cryptomonas. But the organism is so
small and the differentiations of the bodily structure are so slight that
I was unable to determine the relation of its structure to the spiral
path and to the direction of turning in the reaction. The oriented

64 THE BEHAVIOR OF LOWER ORGANISMS.

organism reacts to a decrease in illumination by a sudden turn to one
side, by an increase in the width of the spiral, and by a change in the
course, just as happens in Euglena and Cryptomonas. The unoriented
organism becomes oriented in a manner which is similar to that de-
scribed above for the two organisms just named. Since, however, I
am unable to give the precise relations of these movements to struc-
tural differentiations of the body, a further account of details would
not be of interest.

GENERAL RESULTS.

In summing up our results on reactions to light in the organisms
studied, there are two points of especial interest which should be con-
sidered separately. The first relates to the nature of the reaction
produced, the second to the nature of the agent causing the reaction.

NATURE OF REACTION PRODUCED BY LIGHT.

As to the nature of the reaction produced by light there has been
much discussion. The orthodox tropism theory is perhaps that which
has the greatest number of adherents. It is set forth in detail in the
paper of Holt & Lee (1901). According to this theory the light acts
directly on the motor organs of the side on which it impinges ; supra-
optimal light causes increase of the backward stroke (in the case of
cilia or other swimming organs) ; suboptimal light causes a decrease
in the backward stroke. The result is that the organism is turned
directly toward or from the more intensely lighted side, and hence
toward or from the source of light. The diagrams given in the pre-
ceding paper (Figs, i and 2) can be applied directly to the elucidation
of this theory.

In the experiments on the ciliates and flagellates set forth in the
present paper the precise method of reaction was determined by obser-
vation. It is not in accordance with the tropism theory above set
forth. This has been emphasized in detail in the account of the
reactions of Stentor, so that it need not be reiterated here. The reac-
tion to light is of the same character as that to other stimuli, and takes the
form of a motor reaction in which the organism performs a definite
set of actions. It first usually stops or swims backward, then turns
toward a structurally defined side, then continues forward. The
result is to change the course of the organism. As a result of the con-
tinual rotation on the long axis, together with the swerving toward a
certain side, the organism comes to be pointed successively in every
direction. In continues to swim forward in that direction which does
not induce a stimulus to further swerving. The whole reaction is a
strongly marked example of the type of behavior which may be called
the *' method of trial and error."

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 65

NATURE OF AGENT CAUSING THE REACTION.

(i) The primary and essential cause of the reaction is a change of
illumination. The change of illumination must take place with some
suddenness, but need not be very great in amount. The change in
illumination acts as an effective stimulus even though the degree of
illumination preceding the change and that following it would, when
acting continuously, produce no such result. This is shown by the
experiments on Euglena, in which the light coming from one side was
decreased a certain amount. The orientation of the organisms and
their direction of movement was the same before and after the change,
but at the moment the change occurred there was a marked reaction.
Other experiments detailed above demonstrate the same thing. Further,
the change in illumination acts independently of the direction of the
rays of light. This is shown by the experiment just cited, in which
the effective direction of the rays of light was the same before and
after the reaction ; it is also shown in the reaction caused when the
light is decreased from below, in the case of Euglenae swimming
toward a window (p. 50), and in the reaction of Stentor on passing from
a shadow to a lighted region even when the animal is oriented with
anterior end away from the light (p. 39). The change in illumination
acts equally whether it affects the entire organism or only the anterior
end. The evidence indicates that in all cases it is really the change at
the anterior end which induces the reaction.

(2) The absolute intensity of the light affects the reaction by deter-
mining in a given case whether a reaction shall be caused by an
increase or a decrease in illumination. Through this action it also
determines, in the way to be mentioned in the next paragraph, whether
in a continuous light the sensitive anterior end shall be directed toward
or away from the source of light ; that is, whether the response shall
be ''positive" or "negative."

(3) Indirectly, and through the factor set forth in paragraph (i), the
direction from which the light comes is a determining factor in the
reactions. Through the spiral course in which the organisms swim such
conditions are furnished that in a field continuously lighted from one
side the sensitive anterior end of the unoriented organism is subjected
to repeated changes in the intensity of illumination. As a result,
organisms which respond by the motor reaction to an increase in illu-
mination at the anterior end must become oriented with anterior end
directed away from the light ; organisms which react to a decrease in
illumination must become oriented with anterior end directed toward
the light. (Details in the account of Euglena, pp. 60, 61, and Figs.
23, 24.)

66 THE BEHAVIOR OF LOWER ORGANISMS.

The results of this method of reacting may be stated correctly, though
not completely, as follows : In a negative organism light falling upon
the sensitive anterior end causes a reaction by which the anterior end
is pointed in many different directions ; the reaction ceases as soon as
a direction is reached in which the anterior end is pointed away from
the light. In a positive organism the shading of the sensitive anterior
end produces the reaction by which the anterior end is pointed in many
different directions ; the reaction ceases as soon as the anterior end is
no longer shaded.* The reaction is thus by the method of trial and
error ; when stimulated the organism tries many different positions,
till one is found in which there is no further stimulation.

Consideration will show, I think, that the factors producing reaction
to light in these lowest organisms are essentially the same as in higher
ones, if man may be taken as a type of the latter. The factors are, as
we have seen, variations in intensity of illumination, and, indirectly,
the direction from which the light comes. It is possible that in man
the latter factor works more directly than in the infusoria ; leaving this
question out of consideration, the two factors are present in both cases.
Consider a human being who reacts to light as a purely physical agent,
not with regard to the associations which it brings up. In a dark space
a gleam of light is pleasant and induces movement toward it. There
is then a positive reaction with orientation, but the orientation is not
due to the difference in intensity of light on different parts of the body,
nor to its direct effect on the motor organs. The orientation is such as
to keep the light shining on the more sensitive part of the body, the
eyes. An excessively powerful light is unpleasant and induces a nega-
tive reaction just as happens in Euglena ; the orientation is then such
as to keep the more sensitive part of the body, the eyes, away from the
light. Further, man is sensitive to a sudden change in illumination.
A strong light bursting from the darkness, or sudden darkness in the
midst of bright light, induces a marked motor reaction, and less striking
differences may produce a response. Both in man and in Euglena the
reaction likewise depends upon color ; but with this phase of the matter
we are not at present concerned.

When the factors above set forth are taken into consideration certain
peculiar experimental results that have given rise to much discussion
become clearly intelligible. I refer particularly to the experiments in
which the direction of the light and the decrease in intensity of illumi-
nation do not show the usual relations. Under ordinary conditions
movement away from a source of light is movement into a region of less

* This statement is incomplete in that it does not bring out the fact that it is
a change from light to shade or vice versa that induces the reaction ; if this be
understood, the statement is correct.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 67

intensity ; movement toward the light into a region of greater intensity.
In the well-known experiments of Strasburger (1878) and others, this
condition is modified by passing the light through a wedge-shaped
prism filled with a solution that cuts out part of the light.

When a drop of water or a culture dish is placed beneath such a
prism, and the latter is so situated that its surface is perpendicular to
the light rays, the intensity of the illumination is greatest behind the
thin edge of the prism, and thence decreases gradually toward the
opposite end, while the rays of light all come directly from above.
Under these conditions Strasburger (1878, p. 36) found that the positive
swarm-spores remained equally distributed throughout the drop, not
collecting at the lighter end. Now, the only difference between this
experiment and the one illustrated in Fig. 11 of the present paper is
that in Strasburger*s experiment the decrease in illumination is very
gradual. We have seen above (p. 52) that a very gradual change in
illumination produces no reaction. Hence the organisms may wander
from one side of the drop to the other without reaction, the difference
in illumination at two successive instants never rising to the necessary
threshold of stimulation. If the relation of stimulus to reaction follows
Weber's law, the result is just what we should expect, provided the
change in illumination is sufficiently gradual. When the difference in
illumination from above is great, Strasburger's own experiments (/. c,
p. 33) show that the organisms do react

On the other hand, Holt & Lee (1901), using a similar prism,
found, under similar conditions, that the negative organism, Stentor,
does, on the whole, tend to gather at the darker side of the drop.
This shows that the difference in illumination between neighboring
points in this particular experiment was not below the threshold of
stimulation for the organism in question. If, as Holt & Lee sup-
pose, a certain amount of light was reflected from the lighter end of
the vessel, then the inclination to go to the darker side would be rein-
forced by Stentor's tendency to turn when the light falls upon its
anterior end (see p. 43). The fact that in Strasburger's experiments
the organisms remained scattered throughout the drop seems to indi-
cate that this reflected light played no part in his results.

In another set of experiments Strasburger placed his prism over the
swarm-spores in such a way that the light came obliquely from the
direction of the thick end of the wedge. If the positive organisms now
go toward the thicker end of the wedge, they pass toward the source
of light, but into a region of decreased illumination ; if they go toward
the thin end they pass away from the source of light, but into a region
of higher illumination. Which will they choose.^

Strasburger found that the positive swarm-spores pass toward the

6S THE BEHAVIOR OF LOWER ORGANISMS.

source of light, into the region of less illumination. But is not this
exactly what we must expect? His former experiment showed us that
under the prism the change from light to darkness was so gradual that it
produced no effect on the organisms. Hence the direction from which
the rays come is left to produce its effect alone, and it produces the
usual effect. The organism reacts in the usual " trial and error'* way
until the anterior end is directed toward the light ; then it moves in that
direction. Incidentally it comes into a region of less intensity of light,
though the decrease is so slight as to produce no effect on the organism.

Parallel considerations hold for the negative organism. Under
similar circumstances, if the variation in illumination is very gradual,
it directs its sensitive anterior end away from the source of light (by the
method of "trial and error") and swims to the opposite side of the
drop, incidentall}- moving into a region of slightly greater (but
" unperceived ") intensity of illumination. Under similar conditions,
as we have seen in the experiment described on p. 39, if the decrease
in illumination is marked, the animal swims back into the shadow,
though in so doing it passes toward the source of light.

Thus in Strasburger's experiments with the prism the difference in
the intensity of light between neighboring regions has been made so
slight that they are unmarked by the organism and have no effect upon
it. We need not be surprised, therefore, that it reacts as if these differ-
ences did not exist ; for the organism they do not exist.

The reaction is in this case just what it would be in a higher organ-
ism under similar conditions. Let us suppose that the light stimulates
strongly the sensitive anterior end, the eyes, of a higher animal or man ;
it causes pain in the case of man. There will be a tendency (i) to
move into less illuminated regions ; (2) to turn the eyes away from the
light. Suppose that the man is enclosed in a space into which the
sun shines obliquely from above, and that the end from which it shines
is a little less illuminated than the opposite end, owing to causes similar
to those in Strasburger's experiment on the swarm-spores. Suppose that
the man is at the end next the sun. He cannot know that the other
end is more illuminated, for the only way this would be possible would
be for the greater number of rays of light to meet his eye coming from
that direction, while by hypothesis all, or a much larger number, of
the rays are coming from the opposite direction. He will, therefore,
turn his eyes away from the sun, and if he moves will move toward
the end away from the sun. After having traversed some distance he
may observe, if he is very discriminating, that he is as a matter of fact
getting into a region of somewhat greater illumination, and may perhaps
reason that the best thing he can do under the circumstances is to keep
his eyes turned away from the source of light and move backward to the

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 69

less illuminated end. But this involves the capability of making fine
distinctions, and a considerable degree of intelligence in deciding what to
do under the peculiar circumstances. The experiment w^ith the swarm-
spores shows that they are incapable of such fine discrimination, or that
they are not sufficiently intelligent to know what to do under the circum-
stances. They give no indication that they notice the greater illumina-
tion after having passed to the end away from the light. Their action
may be considered perhaps in a certain sense as a " mistake," but it
is a mistake which even the highest organism would make, and which
could be corrected only after experience of its results.

The results of our study of the light reaction in ciliates and flagel-
lates lead to conclusions which stand in sharp contrast with certain
general conclusions in Radl's recent extensive and interesting paper on
Phototropism (Radl, 1903). Radl reaches the somewhat extraordinary
conclusion that light orients organisms by exercising an actual
mechanical pressure upon them. This pressure necessarily disturbs
the equilibrium of the body, which is then compelled to change posi-
tion until equilibrium is restored ; the organism is then oriented. The
orientation is a consequence of the interplay of two sets of forces, inner
and outer ; these cannot be in equilibrium until the body has taken a
certain position with reference to the pressure exercised by the light
(/. c, pp. 151 fT.) The actual turning which induces orientation must
be due to the action of a pair of forces (/. c, p. 148). One of these
forces is the pressure produced by the light.

Orientation produced in the manner described in the present paper
for the reaction of ciliates and flagellates to light, and in the preceding
paper for the reaction to heat, could of course not be brought about in
the manner supposed by Radl. One of Radl's chief arguments for his
view is that ''no observation thus far shows that the final orientation
is attained by a trial or after an oscillation, but it takes place auto-
matically"* (/. c, p. 141).

The observations on ciliates and flagellates given in the present paper
show conclusively that the orientation in these cases is brought about
through repeated trials. In the statement quoted above Radl has over-
looked certain other cases. Thus Strasburger, as we have seen (p. 59),
states that after the direction of the light is changed Haematococcus
becomes reoriented " nach verschiedenen Schwankungen" (Strasbur-
ger, 1878, p. 24). Radl himself refers on a previous page (p. loo) to
Strasburger's observation of the oscillating movement of swarm-spores
under the influence of a variation in light intensity; Rothert (190J,

*"Keine bisherige Beobachtung zeigt ferner, dass die schliessliche Orientie-
rung etwa durch eine Priifung oder nach einem Schwanken erzielt wurde,
sondern sie folgt automatisch."

7© THE BEHAVIOR OF LOWER ORGANISMS.

p. 397) has called particular attention to this as a possible factor in the
so-called phototropism. Radl also refers (p. 99) to Exner's view that in
Copilia the movements of the eyes are in the nature of a trial (" abtasten")
of the surroundings. RadPs statement, quoted above, can then hardly
be considered strictly accurate, even leaving out of consideration the
results set forth in the present paper. In many organisms, doubtless,
the reaction to light is of that direct character assumed to be general by
Radl. But it may be strongly doubted whether this is what we may
call a primitive condition ; in other words, whether it does not involve
more complicated internal processes than the reaction by " trial and
error." In any case, I am convinced that a similar reaction to light by
the method of " trial and error " will be shown to exist in many other
organisms ; it is demonstrated, for example, in Rotifera, in the paper
which follows the present one.

Recourse will doubtless be taken to the usual refuge when a sharp
concept has been defined to which the phenomena are not found to
correspond ; the reactions of the ciliates and flagellates will be simply
excluded from the tropisms and the definition of the latter maintained
in all its pristine purity. Indeed, it may be questioned whether the
reactions of infusoria (and Rotatoria) to light are not excluded from
phototropism through the definition given by Radl on p. 140, what-
ever the method by which they are produced. Radl says " Unter
phototropischer Orientierung ist die Fahigkeit der Organismen zu
velstehen, eine feste Einstellung der Achsen des gesamten Korpers in
dem Lichtfelde einzunehmen." Since the ciliate or flagellate (or
rotifer) revolves continually on its long axis, and swerves continually
toward a certain side, it can hardly be said that the body axes have a
" feste Einstellung" with reference to the light. In an explanatory
paragraph Radl says that in orientation ''immer geht dann der
Lichtstrahl durch die (morphologische) Symmetrieebene des Korpers"
(/. c, p. 140) . This is certainly not true for the ciliate or flagellate (or
rotifer), even leaving out of consideration the fact that in the former
two groups the animals are usually unsymmetrical. If it be proposed,
then, to exclude the light reactions of ciliates, flagellates, and rotifers
from the concept of *'Phototropismus," one can only agree that this is
necessary, in view of the definitions of that concept.

But what is the value of a definition which excludes some of the chief
phenomena on which the concept that we are attempting to define is
based ? And what is the value of a theory that depends on such a
definition and that can only be correct so long as we hold to this
definition.? The phenomena themselves are, after all, the final refer-
ence for testing the correctness of any definition or theory ; it is the
observed phenomena that we are attempting to formulate and explain.

REACTIONS TO LIGHT IN CILIATES AND FLAGELLATES. 7 1

What we desire in the study of animal behavior is (i) a correct
description of what occurs ; (2) an understanding of the relation of
what occurs here to other phenomena, this constituting their *' explana-
tion" so far as an explanation is possible. Whether the phenomena
when correctly described and understood are found to fall under some-
one's definition of a tropism is comparatively unimportant ; it is only
after such correct description and understanding that final definitions
can be made. I question much if there has not been undue haste in
framing precise definitions for the phenomena of animal behavior, when
we know so little about the phenomena in any thorough way. Radl,
I believe, makes a fundamental error in attempting to separate " Pho-
totropismus" rigidly from other reactions to light. Thus, he repeat-
edly cites Euglena as an example of an organism that shows undoubted
phototropism. On page 114 he further cites the motor reaction of
Euglena when suddenly shaded * as a reaction that has nothing to do
with phototropism. As I have shown above, the two are really
closely bound up together ; the orientation in the "phototropism" is
produced through this motor reaction. When the reactions of organ-
isms to light are known in detail, I believe that many other reactions
which Radl (p. 1 14) attempts to separate sharply from " phototropism "
will be found closely connected with the reactions that go under that
name. I had occasion to point out, in the paper preceding this, on
the reactions to heat, that if everything which the organisms do, except
the orientation itself, is left out of consideration, the orientation can be
accounted for by any theory desired. A thorough study of precisely this
point — the relation of "phototropism" to the phenomena supposedly
unconnected with it— would, I believe, have saved Radl from marring
his otherwise most excellent and useful contribution to the study of
light reactions by the proposal of so fantastic a theory to account for
the reactions to light ; a theory that fairly produces a shock in the mind
of the reader when it is reached, coming as it does after Radl's thorough
and valuable objective study of many of the phenomena and his exceed-
ingly sane, if somewhat sharp, criticism of other theories. Definition
and precise classification are of course valuable at a certain stage of
knowledge, but when carried out without a thorough knowledge of the
phenomena dealt with they may be a hindrance rather than a help.
The thorough knowledge of the phenomena of animal behavior required
for this is far from existing at present.

* Radl says when " beleuchtet " ; this is evidentlj a slip of the pen.

THIRD PAPER

REACTIONS TO STIMULI IN CERTAIN
ROTIFERA.

73

REACTIONS TO STIMULI IN CERTAIN ROTIFERA,

In my series of " Studies on Reactions to Stimuli in Unicellular
Organisms " and in the foregoing papers I have set forth the reaction
methods of many infusoria to varied stimuli. The result has been
to show that the reaction method in these organisms is of a peculiar
character, differing radically from that required by prevailing theories
of the reactions of lower organisms. The essential nature of these
reactions, with their implications as to the character of behavior in the
lower organisms, will be discussed in the following papers. Before
proceeding to this discussion it is important to determine whether the
reaction method in the Infusoria differs radically in character from
that of Metazoa. For this purpose it seems well to select a group of
Metazoa whose habitat and mode of life are similar to those of the
Infusoria. In this way differences due primarily to the different plan
of structure of the two sets of organisms may perhaps be brought out
without the complications arising from different modes of life.

A group of Metazoa much resembling the Infusoria in their mode
of life is found in the Rotatoria. As is well known, the members
of these two groups are usually found mingled together. They are
of about the same size, and both swim about by means of cilia. So
great is the resemblance in general habit and in habitat that they were
at first classed together, all being given the name of Infusoria. As
we know now, however, they are really widely separated in relation-
ship. While the Infusoria are unicellular, the Rotifera are multicellu-
lar organisms of a high degree of complexity, possessing many systems
of organs, each composed of many cells. In particular, they have a
well-developed nervous system.

A comparison of the behavior of these two groups of organisms
should show us, therefore, whether there are types of reaction having
a high degree of generality, such as is claimed for the theory of
tropisms — types that may give a key to the behavior of groups so
widely separated in relationship as the two under consideration, which
are representatives of the Protozoa and of Metazoa of a fairly high
degree of organization.

In the present paper I can attempt to give an account of the behavior
of only a few free-swimming species, and that not in an exhaustive
manner. I hope to return to an extensive study of the behavior of this
interesting group, so as to develop its implications for the theory of

75

76

THE BEHAVIOR OF I.OWER ORGANISMS.

animal behavior in general. In the study here set forth observation
— -^^ was directed primarily to the questions of how

certain Rotifera react under the stimulus of
the agencies which usually give rise to the so-
called tropisms — light, chemicals, heat, elec-
tricity, contact, etc. — and to these questions
the present account will be devoted.

The species whose reactions were exam-
ined belong chiefly to the loricate group of
free-swimming Rotifera, and include a num-
ber of species of the Rattulidae, several species
of Cathypnadae, two or three species of Euch-
lanis, Plcesoma lenticulare^ Anurcea cochle-
aris^ and Brachionus pala. These were
studied as opportunity offered. In most cases
the reactions of any one species were not
determined with relation to more than two
or three classes of stimuli. The behavior of
AnurcEa cochlear is was examined most fully.
This species will be used as a type in describ-
ing the reactions. I have already given a
brief account of the general reaction type in
certain species of the Rattulidae in my mono-
graph of that group (Jennings, 1903).

METHOD OF LOCOMOTION.

The free-swimming Rotifera progress

through the water in the same manner as the

ciliate infusoria. The cilia in the Rotifera

are limited to the anterior end, as they are

in the peritrichous infusoria. It is interesting

to note that the same device is adopted in the

one group as in the other, to compensate for

irregularities in the form of the body, etc.,

Fig. 25.* which might result in swerving from the

straight course. This is by revolution on the long axis, causing the

path to become a spiral with a straight axis. In the Infusoria the

♦Fig. 25. — Spiral path followed in ordinary swimming by Anurcea cochlearis
Gosse, showing different positions of body in different parts of the course;
a, dorsal surface; ^, left side; c, ventral surface; d, right side. The animal
revolves on its long axis over to the right, thus taking successively the positions
«, b, c, d, a, etc. The large arrow indicates the general direction of the course
followed; the smaller arrows show direction of progression in certain parts of
the course.

REACTIONS TO STIMULI IN CERTAIN ROTIFERA. 77

organism usually swerves from the straight line toward the aboral side ;
in the Rotatoria it is usually toward the dorsal side. Well-ordered
forward progression would therefore not take place, were it not for the
revolution on the long axis, converting the circular course into a spiral
one. In the Rotifera the revolution on the long axis is, so far as
observed, always over to the right. These relations have been brought
out in detail in a previous paper by the present author (Jennings, 1901).
The spiral path thus followed by most of the free-swimming Roti-
fera may be illustrated in Fig. 25, for Anurcea cochlear is Gosse. As
will be seen from the figure, the path followed depends upon three
factors : (i) the animal continually swerves toward its dorsal side ; (2)
it progresses ; (3) it revolves on its long axis. The result of these three
factors is the spiral course. In all these relations the rotifer agrees
with the infusorian.

REACTIONS TO STIMULI.

The most general reaction to a stimulus in such a free-swimming
rotifer is an accentuation of one of the factors in this course, namely,
the swerving toward the dorsal side. The result is to produce a spiral
of much greater width than previously existed. This may often be
observed when the vessel containing the rotifers is jarred. It is evi-
dent that this method of reaction is fitted to enable the rotifer to avoid
a small obstacle lying in its path, that is, in the axis of the spiral.
When the animal resumes its former method of swimming the axis of
the spiral lies in a new direction ; the course has thus been slightly
changed.

With a stronger stimulus, as when the rotifer strikes against an
object lying in its path, the swerving toward the dorsal side may be
still more pronounced, while the revolution on the long axis nearly or
quite ceases. The result is that the organism swings strongly toward
its dorsal side, and when the usual forward swimming is resumed the
axis of the spiral lies in a totally new direction (Fig. 26). It thus
avoids the obstacle, if the latter is small ; if the first reaction does not
avoid the obstacle completely the reaction is repeated until the course
is sufficiently altered so that the rotifer no longer strikes against the
source of stimulus. In some rotifers the increased swerving toward
the dorsal side is preceded by swimming backward a short stretch.

In all these points the reaction of the rotifer agrees even to details
with that of the ciliate infusorian. There is a difference in the fact
that the Infusoria are unsymmetrical and cannot therefore be said to
swerve toward the dorsal side, as do the prevailingly symmetrical
Rotifera. In the Rattulidas, however, we have asymmetry of a char-
acter similar to that found in the Infusoria.

78

THE BEHAVIOR OF LOWER ORGANISMS.

We have dealt thus far specifically only with reactions to simple
mechanical stimuli, such as are presented by an obstacle in the path of

Fig. 26.*

the rotifer or by a simple mechanical jar. This type of reaction under
such conditions I have observed for Diurella tigris Miiller, D. por-

♦ Fig. 26 is a diagram of the reaction of the rotifer Anuraea to a strong stimulus,
as when it reaches a source of mechanical stimulus or a region where some
chemical is dissolved in the water. From a to ^ the animal is unstimulated,
hence it follows the usual spiral course. At b it reaches the stimulating region,
whereupon it turns strongly toward the dorsal side, following the arc of a circle,
from b to d. Here it resumes the usual spiral course {d to e). The large arrow
X shows the general direction of progression before the stimulus was received;
the arrow J shows the direction of progression after the reaction has taken place.

REACTIONS TO STIMULI IN CERTAIN ROTIFERA.

79

cellus Gosse, D. gracilis Tessin, and a number of other species of
Rattulidae ; Plcesoma lenticulare Herrick, Cathypna ungulata Gosse,
Monostyla bulla Gosse, Brachionus pala Ehr., Anurcea cochlear is
Gosse, and A. aculeata Ehr.

This method by no means exhausts the possibilities of reaction even
to a simple mechanical stimulus in these species. They may retract
the head and cease swimming, may creep over the surface of the object
vv^ith which they come in contact, or possibly may sometimes turn other-
wise than to the dorsal side when stimulated. Of this latter point I
am, however, by no means sure. It is certain that the typical reaction,
occurring in the great majority of cases, is that described above.

REACTION TO CHEMICALS.

The reaction given when the organism comes in contact with an area
containing a rather strong
diffusing chemical was
observed in Aletopidia
lepadella^ Anurma coch-
lear is .^ A. aculeata^ and
Diurella gracilis.

The method of experi-
mentation was as follows :
A drop of water contain-
ing the rotifers was placed
on a slide. Near this was
placed a drop of N/8
NaCl, and the two drops
were connected by a nar-
row neck. The behavior
of the organisms as they
came into the region of the neck and thus in contact with the salt
solution was observed with the Braus-Driiner microscope. In the
species mentioned the reaction was by a sudden turn toward the dorsal
side, by which the path of the animal was directed away from the
chemical. The reaction is thus of the same character as occurs in the
ciliate infusoria.

This manner of reaction to chemicals is in both these groups of
organisms just what might be expected when the currents caused by
the cilia are taken into consideration. In the ciliate, as I have shown

Fig. 27.

♦Fig. 27.— Diagram of currents in a nearly quiet Anuraea, showing how a
diffusing chemical or an advancing region of warmer water (represented by shad-
ing), is drawn out by the ciliary vortex, so as to reach the mouth and the ventral
surface before affecting other parts of body.

So THE BEHAVIOR OF LOWER ORGANISMS.

in a previous paper (Jennings, 1902, a) , the cilia cause a current coming
from the region in front of the organism to pass along the oral surface
to the mouth ; in this way the oral surface comes in contact with the
chemical before any other part is aflected. It is not surprising, there-
fore, that the organism should turn toward the opposite (aboral) side.

In the rotifera the conditions are parallel to those found in the ciliate.
The cilia cause a current, which passes to the mouth, on the ventral
surface of the body (Fig. 27). The solution thus reaches the ventral
surface first, and the reaction is, as might be expected, a turn toward
the dorsal side.

It should be distinctly stated that this reaction method is not universal
in rotifers even toward chemical stimuli. In some of the larger species,
bearing auricles, or with the ciliary apparatus of a very complex
character in other respects, varied reactions may occur, which I hope
to analyze in another paper.

REACTION TO HEAT.

This was studied in detail only in Anurcea cochlearis. A large
number of the rotifers were mounted in a shallow trough formed of a
slide, as described on p. 12, and one end of the slide was warmed by
means of the apparatus shown in Fig. 5. The reactions were then
observed with the Braus-Driiner stereoscopic binocular.

As soon as a portion of the slide has been warmed above the optimum,
the rotifers in this region turn more strongly than usual toward the
dorsal side, so that the course followed becomes a very wide spiral and
the animals make little progress. If the heat is increased the revolu-
tion on the long axis ceases, while the animals swerve still more
strongly toward the dorsal side (Fig. 28), so that they swim in circles,
the dorsal surface being directed toward the center of the circle.
Usually after circling thus a short time the animals begin again to re-
volve on the long axis, and dart forward. The direction of this dart
forward seems purely random. If it carries the animal out of the
heated region the forward movement is continued and the animal
escapes. If it does not carry the animal out of the heated region the
circling toward the dorsal side is quickly resumed, followed by another
dart forward. This is continued either until the rotifer passes out of
the heated region or until it is overcome by the heat. Usually, if it
does not escape soon from the heated region the circling becomes
more rapid and continuous and is kept up till the animal is destroyed
by the heat.

If one end of the slide is heated and the animal approaches the
heated region from the opposite end the reaction is of the same
character as that last described. As soon as the region is reached

REACTIONS TO STIMULI IN CERTAIN ROTIFERA.

8l

wliere the heat acts as an effective stimulus the animal swerves strongly
toward the dorsal side, thus beginning to circle, as shown in Fig. 28.
If this swerving should continue only till the animal had described a
semicircle, then were followed by the forward dart, the animal would
of course retrace its original course (or one parallel to it), and would
thus escape from the heated region, as happens in the reaction to the
electric current (Fig. 29). But the reaction to heat is less precise than

f

Fig. 28.*

this. Usually the animal makes several com-
plete circles before darting forward, and the
direction in which it darts seems a random
one ; sometimes it is toward the heated region,
sometimes away from it, sometimes oblique to
it. If the path followed leads the animal into
the heated region the circling toward the dorsal
side, followed by the dart forward, is repeated ; while if the path leads
away from the heat no farther reaction is caused and the animal escapes.
Thus when a large number of the animals swim toward the heated
region a considerable number will be seen a little later to swim away
again. But in many cases the dart forward carries the animal still
farther into the heated region. These specimens then begin to circle
again toward the dorsal side, and if the temperature is high they may

* Fig. 28. — Diagram of a reaction to heat in Anuraea. The unstimulated
animal at first advances in the general direction shown by the arrow x, following
thus the course a to e. The heat is supposed to be advancing from the direction
opposite the arrow x. When the rotifer reaches the point e the heat becomes
effective as a stimulus. The animal reacts by turning toward the dorsal side,
and continues this so as to describe a complete circle,/", ^, A, /,/", etc. ; often it
describes such a circle several times. Finally, at some point in the circular
course, as _^, it resumes the usual spiral course, following thus the path ^, /, /.
Its original course, shown by the arrow x, has thus been exchanged for a course
having the general direction shown by the arrow v.

$2 THE BEHAVIOR OF LOWER ORGANISMS.

continue this till death intervenes. In many cases they repeat the dart
forward and some escape in this way, while others do not.

The reaction of Anuraea to heat is, therefore, not very precise, and
many individuals swim into the heated region and are killed. Those
which escape do so through a reaction which is similar to that of
those which do not ; in the one case the forward movement carries the
animal out of the heated region ; in the other it does not. The essential
point to the reaction is that the animals when stimulated by heat change
their course (through a " motor reflex"). This changed course nat-
urally is an advantage, and in accordance with the laws of probability
carries some of the organisms away from the source of danger. Others,
likewise in accordance with the laws of probability, are carried even
by the changed course toward the heated region, where they may be
killed unless a repetition of the '* motor reflex" with its change of
course carries them finally away. The reaction is by the method
of" trial and error," and is not always successful.

Altogether, the reaction of the rotifer Anuraea to heat is of a charac-
ter similar in principle to that of Oxytricha (Fig. 7, p. 16). The
direction of turning depends on an internal factor ; the reaction takes
the form of " a motor reflex," and is by no means compatible with the
typical tropism schema.

REACTION TO LIGHT.

In light, as I have already set forth in the account of the reactions of
Stentor, we have a stimulating agent of a different character from that
found in chemicals or in heat, since the distribution of the stimulating
agent is not afl^ected by the currents of water produced by the motor
organs of the animal. There is thus no reason in the distribution of
the stimulating agent to favor a turning toward one side rather than the
other.

I have been able to study accurately the light reaction in but one
rotifer, Anurcsa cochlearis Gosse. The conditions necessary for
precise observation of the nature of the reaction are very difficult to
fulfill, and the usual movements of the animals are such that the nature
of the reaction is obscured. As will be recalled, the organism is
normally swimming rapidly in a spiral, continually swerving toward
its dorsal side. This in itself is very confusing when one attempts to
observe just how the organism turns when stimulated. When light is
thrown upon it, or when the direction of light falling on it is changed,
the response is usually not given at once, and when it does occur, as
we shall see, it may be in the form of an accentuation of certain features
of the normal movement. From these conditions it results that it is
exceedingly difficult to tell, after a reaction to light has clearly occurred,

REACTIONS TO STIiVlUI.I IN CERTAIN HOTIFERA. S3

just how the reaction took place. Of course, only sharply defined posi-
tive observations are of value in deciding between two opposing possi-
bilities ; hence, although I have studied a number of other rotifers in
this connection, I give the results only where absolutely sure of them.
But in the two or three other rotifers I have examined in this connection
the reaction is apparently the same as that in Anurcea cochlearis^ to
be described at once.

The specimens of AnurcBa cochlearis studied had been in a small
aquarium in the laboratory some months, and were distinctly negative
to light, gathering at the side of vessel farthest from the window. The
freshly collected animals are, I believe, usually positive to light.

These negative individuals were placed in a small flat-bottomed
rectangular glass vessel, on a dark background, in a dark room. At
opposite sides of the vessel and somewhat above were clamped two
incandescent electric lights, A and B^ at a distance of about lo inches*
in the manner described for Stentor (p. 41 and Fig. 15). One of
these lights could be extinguished while the other was simultaneously
turned on. In this way the direction of the light falling on the rotifers
could be reversed.

When only one of the lights, as A^ was turned on, the Anuraeas all
collected at the opposite side of the vessel, next to B. When A was
extinguished and B turned on, they turned and swam in the opposite
direction, toward A. By reversing the direction of the light while the
animals were crossing the vessel their course could be reversed while
in full career.

Focusing the Braus-Driiner on the vessel, and reversing the lights
when the animals were well in the field of observation, the following
could be observed : Some turned at once, with some sharpness,
toward the dorsal side^ the turning continuing until the direction of
swimming was reversed and the animals were again swimming away
from the light (Fig. 29) . In these cases the direction of turning was
clear and could be observed without great difficulty.

Other individuals continued for a short time to swim in the same
direction as before, then turned, either sharply, as just described, or
more slowly, in the manner to be described.

Where the turning was sharp, as described above, there was no
great difficulty in determining with certainty the nature of the reaction.
But in many cases the turning took place more slowly, in the following
manner : Either as soon as the light was reversed, or very soon after,
the width of the spiral in which the animal was swimming became
much greater. In other words, the animal swerved more toward the
dorsal side and progressed less rapidly than usual. Thus it described
rather wide circles, and the swerving toward the dorsal side increased,

84 THE BEHAVIOR OF LOWER ORGANISMS.

while progression and revolution on the long axis had largely ceased.
After this circling had continued for some time, the swerving toward
the dorsal side apparently continually increasing, it was found that the
anterior end was directed away from the source of light; i. c, the
direction of swimming had been reversed, and the animal was moving
away from the light.

It is obviously very difficult to be entirely certain of all that has hap-
pened during this period of extensive circling, as a matter of direct
observation. But the evidence seems to show clearly that the essential
point in changing the course is the swerving toward the dorsal side.
The following facts all point to this conclusion : (i) In the individuals
which turn at once it is possible to be entirely certain that the turning
is toward the dorsal side. (2) In the individuals which are circling it
is entirely clear that the swerving toward the dorsal side is greatly in-
creased, and there is no evidence of turning in other directions. The
only difficulty is that one cannot follow every evolution and be certain
that nothing else has occurred. (3) Analysis of this same reaction
when given in response to other stimuli, where the conditions are
more favorable for observation, shows that it does consist of an in-
creased swerving toward the dorsal side, with a decrease, or an entire
stoppage for a time, of the forward motion. There is, then, no reason
to think that the reaction contains other factors when performed under
the influence of light. The reaction is indeed clearly the same as that
described for Euglena on p. 53, and illustrated in Fig. 21 ; a similar
analysis could be given for the reaction of Anurasa.

It may be considered certain, therefore, that in Anurcea cochlearis
the reaction to light is similar to the reaction to other stimuli, and that
the orientation is brought about by a turning toward the dorsal side.
The reaction is, therefore, not due to the direct effect of the light on the
motor organs ; the direction of turning is determined not by external
factors, but by internal factors. The reaction to light in the rotifer,
like that in the infusorian, takes place by the method of '' trial and
error."

REACTION TO THE ELECTRIC CURRENT.

A considerable number of different species of the rotifera were sub-
jected to the continuous electric current without the production of any
characteristic reaction. A current was used which could be graded in
strength from practically zero to one that was destructive, but no reac-
tion comparable to that found in the ciliate infusoria was produced.
On making or breaking the current the animals frequently contracted
quickly, and if the current was very strong, the head was completely
retracted and the animal sank to the bottom. But there was no orien-
tation and the animals did not swim toward either electrode. These

REACTIONS TO STIMULI IN CERTAIN ROTIFERA. 85

negative results were obtained with several of the Philodinadae (Rotifer,
Philodina), some species of Euchlanis and Salpina, Noteus quadri-
cornls^ and a number of the Rattulidae.

In Hydatina senta there is a reaction of a peculiar character which
perhaps furnishes a clue to the cause of the more pronounced reaction
to be described for Anuraea. With a current of moderate strength,
such as that to which Paramecia react most markedly, Hydatina shows
no reaction except when the head is directed toward the anode. But
in this position the animal at once retracts its cilia and sinks to the
bottom. Thus a Hydatina may swim freely about in water through
which the current is passing, provided it swims toward the cathode, or
transversely, or obliquely ; as soon, however, as it turns its head toward
the anode it stops swimming and sinks to the bottom. Thus if an
electric current is passed through a preparation containing a large
number of specimens of Hydatina, many will be seen swimming
toward the cathode and others at all sorts of angles with the current, but
none toward the anode. This is a phenomenon akin to what I have
elsewhere called the production of orientation by exclusion. If organ-
isms are prevented from swimming in any direction but one, after a
time, provided the course is frequently changed, all that are swim-
ming will be found moving in that one direction. This condition is
realized, as I have shown in the first of these contributions, in the
reactions of infusoria to heat and cold. But in the reactions of Hydatina
to the electric current the "exclusion" is less complete than in the
cases just mentioned ; the animal may swim in any direction except
one.

The fact that the head is retracted when directed toward the anode
and not in other positions indicates that there is a greater stimulation
at the anode than elsewhere. This agrees with much that is seen in
the reactions of infusoria to the current. After Hydatina has sunk to
the bottom with anterior end to the anode, it repeatedly makes attempts
to unfold its cilia. But scarcely have they begun to operate when
they are withdrawn again. Each time that they are uncovered for an
instant, however, they turn the animal a little toward its dorsal side.
Thus, after a considerable number of attempts to unfold the cilia, the
head has become turned away from the anode ; then the cilia are spread
out and the animal goes on its way until it is so incautious as to turn
its head again toward the anode.

AnurcBa cochlearis shows marked electrotaxis similar to that found
in the infusoria. When the continuous current is passed through a
preparation containing large numbers of this species, all orient quickly
and swim toward the cathode. They thus agree, so far, in their
reaction to the electric current, with the ciliate infusoria.

86 THE BEHAVIOR OF LOWER ORGANISMS.

The question as to the mechanism of the electrotactic reaction in the
rotifer is of interest when one compares the structure of these animals
with that of the ciliate infusoria. The Rotifera, in place of having cilia
scattered over the entire body, are furnished only with a group of
cilia at the anterior end. In the Ciliata it is usually possible to distin-
guish functionally two groups of cilia (i) the adora/ ciUa^ about the
mouth and oral groove, or at the anterior end; (2) the body cilia,
scattered over the body. The cilia of the rotifers correspond function-
ally with the adoral cilia of the Ciliata.

Pearl (1900), Wallengren (1902-1903), and others have shown that
in the electrotactic reaction of the ciliates the two sets of cilia are in
many cases from a functional standpoint differently affected. The
adoral cilia react under the influence of the electric current in such a
way as to tend to turn the organism toward the aboral side ; that is,
they tend to produce the same reaction which the organism gives in
response to most other stimuli, a reaction not in harmony with the
tropism schema. The body cilia, on the other hand, are differently
affected on the different sides or ends of the organism ; those on the
part of the body directed toward the cathode striking in one direction ;
those on the part directed toward the anode striking in a different
direction. The result is that the organism, through the action of the
body cilia, tends to become directly oriented in a way that is in harmony
with the tropism schema. (For details, see the papers cited.) In those
ciliates in which the body cilia are much reduced, as in the Hypotricha,
the turning is determined throughout by the adoral cilia, so that the
orientation does not take place in accordance with the tropism schema,
while in some others, such as in Paramecium, the influence of the body
cilia is predominant, and the turning is in accord with the theory of
tropisms.

What conditions shall we find in the Rotifera, where the only exist-
ing cilia seem to agree functionally with the adoral cilia of the Ciliata ?

As we have seen, Anuraea swims as a rule in rather wide spirals,
swerving strongly toward the dorsal side and revolving on its long axis
(Fig. 25). When the electric current suddenly acts upon it the organism
at once turns strongly toward the dorsal side, continuing the turn until
its head is brought toward the cathode, toward which it swims (Fig.
29). In some cases, as we shall see later, several reactions are neces-
sary for bringing the body in line with the current, but these are as a
rule very quickly accomplished.

If, while the animals are swimming toward the cathode, the current
is suddenly reversed, the animals again turn strongly toward the dorsal
side, continuing the turning until their position is reversed and the
heads point toward the new cathode (Fig. 29). In many cases the

REACTIONS TO STIMULI IN CERTAIN ROTIFERA.

87

turning is continued still farther, so that the head of the animal de-
scribes a complete circle ; indeed, this may continue so that the animal
whirls around several times, always towards the dorsal side. The
reaction thus far is the same as that produced by heat (Fig. 28). In
reacting to the electric current the whirling finally ceases with ante-

FiG. 29.*

rior end directed toward the new cathode. The animal then swims
forward in the direction so indicated. These turnings, even when
several times repeated, require but a moment, so that very soon prac-
tically all the specimens are swimming toward the new cathode. The

* Fig. 29. — Diagram of method by which Anuraea becomes oriented to rays of
light, or to the electric current. Taking the latter for example, the animal is at
first swimming toward the cathode, in direction indicated by arrow x ; it thus
follows a spiral path from a to d. At i> the electric current is reversed. The
animal thereupon swerves strongly toward its dorsal side, describing a semicircle,
bj c, d, until its anterior end is directed toward the new cathode, in the opposite
direction from before. It now follows the spiral path d to e, in the general direc-
tion indicated by the arrowy. The facts are similar for the reversal of light, or
for the reaction when the current or the light is first set in operation.

S8 THE BEHAVIOR OF LOWER ORGANISMS.

reaction in Anuraea is in a general view as striking and clear-cut as
that of Paramecium.

Thus in the rotifer Anuraea the orientation to the continuous elec-
tric current is produced through a motor reaction, the essential features
of which are determined by the structure of the organism. The organ-
ism turns always toward the dorsal side, continuing or repeating the
turning until the anterior end is directed toward the cathode. In these
respects it agrees with hypotrichous Ciliata, where the direction of
turning is determined by the action of the adoral cilia. The method
of reaction is quite incompatible with the tropism schema.

SUMMARY.

The reactions of those Rotifera of which an account is given in this
paper take place in a manner essentially similar to the reactions of the
ciliate infusoria.

In the reactions to -mechanical stimuli, to chemicals, and to heat,
orientation is not a striking feature. The organism turns when stimu-
lated toward a structurally defined side — as a rule toward the dorsal
side ; in this way it avoids the source of stimulus.

In the negative reaction to light the organism becomes oriented with
anterior end directed away from the source of strongest light, but this
orientation is brought about in the same manner as in Stentor ; the
animal turns toward the dorsal side without relation to the side on
which the light strikes it, and continues the turning or repeats it until
the anterior end is directed away from the source of light.

To the continuous electric current the rotifer Anuraea orients and
swims directly toward the cathode. The reaction is brought about in
the same manner as the orientation to light. When the current is
made or reversed the animal turns toward the dorsal side and continues
the turning until the anterior end is directed toward the cathode.

Thus the direction of turning is throughout dependent on an internal
factor, not primarily on the way in which the stimulus impinges on
the organism. These reactions of the Rotifera are thus inconsistent
with a theory of tropisms which regards orientation as a primary
feature of the reactions, and which holds that the action of the stimu-
lating agent is a direct one on the motor organs of that part of the
body on which it impinges. The reactions of the Rotifera, so far as
described in the preceding pages, are brought about, like those of the
infusoria, by what may be called the method of " trial and error."
The reaction to any stimulus is of such a nature as to head the organism
successively in many different directions. That direction is followed in
which there is no stimulus to induce further turning.

FOURTH PAPER.

THE THEORY OF TROPISMS.

89

CONTENTS.

PAGE.

To what Extent does the Theory of Tropisms throw Light on the Behavior

of Lower Organisms ?.......... 91

Essential Points in the Theory of Tropisms, 92

Reactions to Mechanical Stimuli, 94

Reactions to Chemicals, . . ........ 96

Reactions to Heat and Cold, 98

Reactions to Changes in Osmotic Pressure, ...... 98

Reactions to Light, ........... 98

Reactions to Gravity, .......... 100

Reactions to Electricity, 100

R6sum6 and Discussion, 103

Summary, 106

90

THE THEORY OF TROPISMS.

TO WHAT EXTENT DOES THE THEORY OF TROPISMS THROW
LIGHT ON THE BEHAVIOR OF LOWER ORGANISMS?

The writer has been engaged for a number of years in a study, as
exact and detailed as possible, of the behavior and reactions of a num-
ber of lower organisms. While the results obtained have not, as a
rule, agreed with the view that the behavior of these organisms is
determined largely in accordance with the prevailing theory of tropisms
or taxis, he has not discussed their relation to this theory in detail.
This was because of the possibility that the reactions which he had
studied were exceptional, and that further investigation might show
after all that the behavior of the lower organisms is largely in accord-
ance with the tropism schema.

At the present time the writer feels that the work which he has
done, or which has been done by those associated with him, is of suffi-
cient extent to justify the pointing out of certain general relations.
The reactions of ciliate infusoria, which have long been used as the
types of illustration for the tropisms, have been examined in much
detail, and less extensive studies have been made on the Bacteria
(Jennings & Crosby, 1901), the Flagellata, and the Rotifera. The
reactions of a flatworm have been studied in much detail (Pearl,
1903), and researches are nearly ready for publication, by investigators
associated with the author, on the behavior of Hydra and of the leech,
and still other studies are under way. Thorough studies, directed to
the observation of the exact movements of organisms under stimuli,
have recently been given us by other observers also. It seems, there-
fore, worth while to bring out, in a preliminary way at least, the
relation of the observations made to the prevailing theories of animal
behavior. In the present paper this will be limited to a consideration
of the theory of tropisms, since this is the theory most widely held.

The great apparent value of the theory of tropisms or taxis lies in
the fact that it seems to reduce to very simple factors a large number
of the most striking activities of organisms, namely, those involved in
going toward or away from sources of stimuli of almost any character.
It is a schema, in accordance with which almost any movements of the
organism (not purely random) might be supposed to take place.

91

92 THE BEHAVIOR oF LOWER ORGANISMS.

ESSENTIAL POINTS IN THE THEORY OF TROPISMS.

The two essential features of the theory of tropisms are apparently
the following: (i) The movements of organisms toward certain
regions and their avoidance of others are due to orientation ; i. e.^ to
a certain position which the organism is forced by the external stimulus
to take, and which leads the organism toward (or away from) the
source of stimulus, without any will or desire of the organism, if we
may so express it, to approach or avoid this region. (2) The external
agent by which the movement is controlled produces its characteristic
effect directly on that part of the body upon which it impinges. It
thus brings about direct changes in the state of contraction of the
motor organs of that part of the body affected as compared with the
remainder of the body, and to these direct changes are due the changes
shown in the movements of the organism. This is brought out clearly
in the quotation from Verworn given on page 8. Loeb (1900, p. 7)
sums up the theory of tropisms as follows :

The explanation of them [the tropisms] depends first upon the specific irrita-
bilitj of certain elements of the bodj surface, and, second, upon the relations
of symmetry of the body. Symmetrical elements at the surface of the body
have the same irritability; unsymmetrical elements have a different irritability.
Those nearer the oral pole possess an irritability greater than that of those
near the aboral pole. These circumstances force an animal to orient itself toward
a source of stimulus in such a v^^ay that symmetrical points on the surface of the
body are stimulated equally. In this way the animals are led without will of
their own either toward the source of stimulus or away from it.

Holt & Lee (1901) again bring out our second point, as applied
to reactions to light, with especial clearness :

The phenomena that have led to such an assumption can be satisfactorily
explained on the simpler theory that every ray of light impinging on an organism
stimulates at the -point on -which it falls^* and in proportion to its intensity. ♦ * *
The light operates, naturally, on the part of the animal which it reaches. The
intensity of the light determines the sense of the response, whether contractile
or expansive, and the place of the response, the part of the body stimulated,
determines the ultimate orientation of the animal." (Holt & Lee, 1901, pp.
479-480.)

The theory of tropisms as above set forth depends upon the reflex
contractility of the motor organs when affected by certain stimuli. An
attempt has been made to give it a still simpler form in a recent paper
by Ostwald (1903). Ostwald would omit even the factor of reflex
irritability, holding that the turning which brings about orientation is
a mechanical result of differences in the internal friction of the water or

♦Original not italicized.

THE THEORY OF TROPISMS. 93

similar physical differences. The organism is considered to continue
to move its motor organs in exactly the same way after the external
change (usually called a stimulus) has taken place ; the reason for turn-
ing lies only in the different mechanical effect produced when the motor
organs act on a medium of greater or less internal friction than hefore.

It is difficult to conceive how anyone having any acquaintance with
the movements of organisms could propose such a theory as that of
Ostwald, and indeed this author states (p. 34) that his account is purely
theoretical, and that he has not attempted to test his theory by experi-
ment. We need not, therefore, dwell upon the theory, further than to
point out the fact that the reactions of many of these lower organisms
have been studied thoroughly, and the reflex movements which they
perform when subjected to directive stimuli have been fully described,
and that these movements are entirely incompatible with such a theory
as that which Ostwald sets forth.* If details are desired, it may be
pointed out that all the observations brought in the following that are
inconsistent with the theory of tropisms as dependent upon direct
stimulation of the motor organs are a fortiori inconsistent with such
a theory as that of Ostwald.

We may, then, turn to the theory of tropisms as set forth in the above
quotations from Verworn, Loeb, and Holt & Lee. Diagrams illus-
trating the method of action of a stimulus, on this theory, are given in
the first of these contributions (Figs, i and 2).

How far does this theor}' go in explaining the behavior of the lower
organisms. -^ "Tropisms" has become the keyword everywhere in
animal behavior ; it is supposed to furnish a ready explanation of most
of the puzzles which we here encounter. How far is this justified.?

This question can be answered only by accurate observation of just
what organisms do under the influence of stimuli. The theory of

♦Some of the assumptions which Ostwald makes as a basis for his physical
analysis of the swimming of the lower organisms are so extraordinary as to
deserve mention as curiosities. He states, for example, that as a rule the lower
swimming organisms which exhibit the tropisms show active movement vertically
only upward; he thinks it probable that cases where they have been described
as swimming actively downward are errors ; that such downward movement is
really only passive falling. Yet everyone who has worked with Paramecium or
other Ciliata must know how far from the facts is this idea. In a vertical tube
Paramecia hasten as freely, and almost as frequently', downward as upward.
These infusoria by no means collect at the top in a vertical tube so regularly as
the literature on geotropism might lead one to suppose ; Paramecia of this region
at least are almost as likely to collect at the bottom as at the top. And there is
little more difficulty in Paramecium in distinguishing an active movement down-
ward from a passive one than there is in man. From my own observations I
know that parallel statements could be made for many other free-swimming
organisms, including Metazoa (Rotifera and Crustacea), as well as Protozoa.

^ THE BEHAVIOR OF LOWER ORGANISMS.

tropisms says that certain definite things happen in the change of
position undergone by organisms under the influence of stimuli ; that
the organisms perform certain acts in certain ways. The problem for
the investigator is, then, Do these things happen? Does the organism
perform these acts, in these particular ways? These questions are not
metaphysical ; they can be answered by observation.

We have now before us a considerable body of exact observations
which permit us to answer these questions for a certain number of
organisms. We will here attempt to summarize these observations so
far as they bear upon the essential points in the theory of tropisms. In
particular, we will ask, (i) Is the observed behavior brought about
through orientation, in the way the theory of tropisms demands? (2)
Does the evidence show that the action of a stimulus is directly upon the
motor organs of that part of the body on which the stimulus impinges?

REACTIONS TO MECHANICAL STIMULI.

The reactions to simple mechanical stimuli, as when the organism
is touched or struck by a hard object over a certain definite area of the
body, of course do not as a rule present the conditions required for
the production of a tropism, including a definite orientation. Yet it is
important to bring out certain general relations shown in these reactions,
as they throw light on the reactions to stimuli of a difl^erent character.

Most animals show in one way or another a tendency to avoid sources
of mechanical shock. In the higher organisms the reaction usually
takes the form of a turning away from the side stimulated. The point
which needs to be brought out here is that in ciliate infusoria the direc-
tion of turning depends, not upon the part of the body stimulated, but
upon an internal factor. Stylonychia turns to the right, whether stimu-
lated on the right side, on the left side, on the dorsal surface, on the
anterior end, or by a general unlocalized mechanical shock ; and parallel
statements can be made for other infusoria. (For details see Jennings,
1900.) We have proof, therefore, that the action of the stimulus is
on the organism as a whole^ not merely upon the ?notor organs of
that region of the body stimulated. Further, it is clear that the
response is a reaction of the organism as a whole, not one brought
about as an indirect result of the fact that certain motor organs have
received a stimulus to contraction.* In these respects, therefore, the
reactions to mechanical stimuli are different in character from those
assumed to take place in the tropisms, and even in these unicellular
organisms the processes taking place must be more complex than the

♦This fact becomes still more striking when we recall that the reaction takes
place in the same way in pieces from any part of the body, from which any given
motor organs may have been removed. (Details in Jennings & Jamieson, 1902.)

THE THEORY OF TROPISMS. 95

theory of tropisms assumes. Certainly a reaction of the organism as a
unit, in response to a localized stimulus, is a phenomenon of a higher
and more complex order than would be a simple contraction or other
direct change in the motor organs at the point stimulated.

In the higher Metazoa the reaction to a slight mechanical stimulus
at one side is usually a turning either toward or away from the source
of stimulus. So long as we do not analyze the process further, this
result might be interpreted either as due to the direct response, by con-
traction, of the muscles primarily affected (thus in accordance with the
tropism theory), or as a response of the organism as a whole, depend-
ent, perhaps, on an alteration in its physiological condition brought
about by the stimulus. The former interpretation is doubtless much
the simpler. But we find in the unicellular organisms that this first
interpretation is impossible, and that we are forced to the less simple
and definite conclusion that the organism reacts as a whole. Does it
not then become probable that in the higher animals the very simple,
almost mechanical, explanation is likewise incorrect ; that we have in
them a phenomenon at least as complex as that found in the unicellular
animals.'' In other words, should we conclude that the reactions in
the higher Metazoa are simpler and less unified than in the Protozoa.?

Fortunately, however, we are not forced to base our conclusions on
general considerations. These reactions have been minutely studied in
very few of the bilateral Metazoa, but Pearl (1903) has given us a
thorough analysis of the reactions of a flatworm (Planaria). This
cannot be taken up in detail here, but we may quote Pearl's con-
clusion in regard to the positive reaction. This consists in a turning
toward the point stimulated, on a superficial view a very simple reac-
tion, one especially well fitted for explanation on the theory of direct
action of the agent on the motor organs of the region stimulated. Pearl
concludes, after exhaustive study, that the processes in the reaction are
as follows :

A light stimulus, when the organism is in a certain definite tonic condition,
sets off a reaction involving (i) an equal bilateral contraction of the circular
musculature, producing the extension of the body; (2) a contraction of the
longitudinal musculature of the side stimulated, producing the turning toward
the stimulus (this is the definitive part of the reaction); and (3) contraction of
the dorsal longitudinal musculature, producing the raising of the anterior end.
In this reaction the sides do not act independently, but there is a delicately
balanced and finely coordinated reaction of the organism as a whole, depending
for its existence on an entirely normal physiological condition. (/. c, p. 619.)

Further studies carried on under the direction of the writer, and
soon to be published, will show that in certain other bilateral Metazoa
it is equally impossible to explain the simple turning toward a stimulus
as a direct reaction of the motor organs of the part stimulated.

go THE BEHAVIOR OF LOWER ORGANISMS.

It will be important to keep in mind the nature of the reactions to
mechanical stimuli, especially in the infusoria, in considering the
reactions which are more usually classed among the tropisms.

REACTIONS TO CHEMICALS.*

The reactions to chemicals have been studied by the present author
and those associated with him in many Ciliata, in certain Flagellata,
in the Bacteria, the Rotifera, and the flatworm ; further studies, not
yet published, have been made on other organisms.' Now, in regard
to our first question, as to orientation, the following must be said :
In no case has the typical reaction been found to take the form of an
orientation, such as is demanded by the theory of tropisms. In the
ciliates, flagellates, and rotifers the reaction has been found to take the
form of a " motor reflex,'* a backing followed by a turning toward a
certain structurally defined side, without regard to the direction from
which the chemical is diffiising. It is this motor reflex that causes the
organisms to collect in the region of certain chemicals, and to avoid
others. (Details in Jennings, "Studies," Nos. I-X.)

In the Bacteria the results of our work (Jennings & Crosby, 1901)
are in agreement with those of Rothert (1901). Here, again, in the
gatherings in certain chemical solutions, or in the avoidance of others,
there is nothing resembling an orientation in the lines of diffusion.
The phenomena are brought about through a reaction of the same
essential character as the motor reflex of the infusoria, but still simpler.
The essential point is that the Bacteria, when stimulated chemically,
reverse the direction of movement. (Details in the papers just cited.)

In the flatworm the results of the thorough study of the chemical
reaction by Pearl (1903) maybe given in that author's own words:

Planaria does not orient itself to a diffusing chemical in such a way that the
longitudinal axis of the body is parallel to the lines of diffusing ions. Its reac-
tions to chemicals are motor reflexes identical with those to mechanical stimuli.
The positive reaction is an orienting reaction in the sense that it directs the
anterior end of the body toward the source of stimulus with considerable pre-
cision, but it does not bring about an orientation of the sort defined above. (Pearl
loc. cit.^ p. 701,)

For details, the original paper of the author quoted must be consulted.
It may be added that this positive reaction, by which the anterior end
is directed toward the source of stimulus, is identical with that which
takes place in response to a single mechanical stimulus. This is analyzed
above (p. 95).

Are there any precise and detailed observations which support the
idea that the reaction to chemicals is ever a typical tropism } Before

♦For a statement of the theory of tropisms as applied to chemicals, see Loeb
(1897, p. 442) and Garrey (1900, pp. 292, 293).

THE THEORY OF TROPISMS. 97

the method of reaction by a '' motor reflex" had been described the
reactions to chemicals had been referred in a general way to the tropism
schema, but critical observations, which would differentiate between
the possibilities, have been lacking. It is necessary to use the greatest
caution in this matter, as is shown by the case of Chilomonas. Garrey
(1900), although he stated that "a study of the mechanics by which
the organism is oriented or by which it is prevented from moving
from the ring into the stronger acid of the clear area, or the weaker
acid surrounding the ring, proved fruitless," nevertheless concluded
that the reaction in this animal was a case of typical tropism. In a
paper published in the same number of the same journal (Jennings,
1900, a), I showed that when the mechanism of the reaction is worked
out, this conclusion does not hold, but that the reaction takes place
through a motor reflex, similar to that in the Ciliata. In cases, there-
fore, where the mechanism of the reaction (that is, the exact movements
which the organism performs) has not been worked out, conclusions as
to the nature of the reaction are of little value. The only case of which
I know in which an author acquainted with the method of response by
a " motor reflex " maintains, on the basis of observation, a reaction of
unicellular organisms to chemicals in accordance with the theory
of chemotropism, is the case of Saprolegnia swarm-spores, as described
by Rothert (1901). In this case we are dealing with very minute
organisms, and Rothert has made no attempt to give an analysis of the
relation of the direction of turning to the differentiations in the body
of the organism, such as we found to be necessary above for Chilomonas
before the real nature of the reaction could be determined.

Thus it is clear that cases of true chemotropism, in accordance with
the general tropism schema, are exceedingly rare, if they exist at all.
In almost all the lower organisms in which this matter has been care-
fully studied it has been demonstrated that the reaction to chemicals is
of a different type from that demanded by the tropism theory.

In the discussion so far we have devoted attention particularly to the
question of orientation. When we examine the second question pro-
posed, as to whether the stimulus acts directly upon the motor organs
of that part of the body on which it impinges, we find the answer
somewhat less clear than it was in the case of mechanical stimuli. It
is true that in the Infusoria and Rotifera the direction of turning is, as
in the case of mechanical stimuli, always toward a structurally defined
side, without regard to the direction from which the chemical is diffiis-
ing, so that at first view it seems beyond question that the reaction is
no^ due to the direct action of the stimulus on the motor organs of the
region on which it impinges. While this conclusion is highly probable,
the observed facts do not demonstrate it for chemical stimuli as they

98 THE BEHAVIOR OF LOWER ORGANISMS.

do for mechanical stimuli. This is owing to the fact that the organism
determines for itself the region in which it shall be stimulated by a
chemical in solution, as well as the side toward which it shall turn.
Now, it appears that the side on which, by its own activities, it is, as a
rule, first stimulated by a chemical, is (usually, at least) opposite that
toward which it turns. (For details, see Jennings, 1902, a.) It could be
contended, therefore, that the direction of turning, in the case of chemi-
cal stimuli, is a result of the direct action of the stimulating agent on
the side stimulated. Such a contention would have little general
significance, however, in view of the fact that the same reaction occurs
as a response to various other stimuli, where this explanation is quite
impossible.

In certain higher organisms, researches which were made under the
direction of the writer and are soon to appear will show (i) that chemi-
cal stimuli may produce local contractions in the part of the body with
which the chemical comes in contact ; (2) that these local contractions
have little to do with the characteristic behavior of the animals when
subjected to chemicals.

REACTIONS TO HEAT AND COLD.

Reactions to heat and cold have been fully discussed in the first of
these contributions. It is only necessary, therefore, to point out that
the results are in almost every detail parallel with those for reactions
to chemicals, and in the same way and to the same degree inconsistent
with the theory of tropisms. In organisms higher than the Infusoria
and Rotifera, the reactions to heat and cold have been very little studied
from the present point of view.

REACTIONS TO CHANGES IN OSMOTIC PRESSURE.*

In the ciliate infusoria the reactions to differences in osmotic pressure
are identical with those to chemicals, save that the organisms are much
less sensitive to osmotic changes. (Details in Jennings, 1S97 and 1899.)
The bearing of these reactions on the theory of tropisms is, therefore,
the same as was brought out above in the discussion of the reactions to

chemicals.

REACTIONS TO LIGHT.

The phenomena shown in the reactions of organisms to light have per-
haps formed the chief basis for the theory of tropisms. There is usually
a definite orientation shown by the organisms ; they move with the axis
of the body parallel with the light rays either to or from the source of
light. The existence of such orientation forms the basis of the theory
of tropisms, and has been considered sufficient in itself as a proof of the

♦ " Tonotaxis," Massart ; " Osmotaxis," Rothert.

THE THEORY OF TROPISMS. 99

theory. Yet the theory makes certain definite statements as to the cause
of the orientation and the way in which it is brought about. These
statements are open to observation and experiment. In most bilateral
animals it is indeed difficult to really test the theory. This is because
these animals may turn directly toward either side under the influence
of light, and it is difficult to tell whether this turning is due to the direct
action of the light on the motor organs or to a reaction of the organisms
as a whole induced by some change in physiological condition brought
about by the light. But in the ciliate infusoria we find a set of organisms
so constituted as to permit us to bring the theory to a direct test. These
organisms are unsymmetrical, and, as we have seen, the usual reaction
is by a motor reflex involving a turning toward a structurally defined
side. We can, therefore, arrange our experiments in such a way that
the turning demanded by the theory of tropisms shall be the opposite
of that usually produced in the reaction of the organisni as a whole, and
observe the results.

This is what was done with Stentor cceruleus^ as described in the
second of these contributions. The result, as we have seen, is that the
organism turns toward a structurally defined side, without regard to
what is demanded by the theory of tropisms. The same result was
obtained with a number of flagellates and with a bilateral Metazoan —
the rotifer AnurcEa cochlearis.

Thus, in these cases, it is impossible to interpret the reactions as due
to the direct action of the light on the motor organs of the side on
which the light impinges. The response is as clearly a reaction of the
organism as a whole as is the reaction to mechanical stimuli.

Now that it has been shown that orientation to light does occur in
some cases in a manner quite at variance with the postulates of the
theory of tropisms, and this in organisms widely separated in structure
and classification, it can no longer be held that orientation is, fer se,
a proof of the tropism theory. In other words, cases in which orien-
tation takes place, but in which the manner in which it is brought
about has not been observed, can not be assumed as cases of typical
tropism, due to the direct action of the light on the motor organs
of the side affected. The reactions of flagellates and swarm-spores
to light, as described by Strasburger (1878), have long been con-
sidered types for the tropisms. In the second of these contributions
I have shown that in Euglena and Cryptomonas (the latter being one
of the organisms studied by Strasburger) the reactions do not take
place in accordance with the tropism schema. So far as can be judged
from Strasburger's account the reactions of the swarm-spores take place
in essentially the same manner as in the flagellates. As Rothert (1901)
has pointed out, there are many details in Strasburger's account which

lOO THE BEHAVIOR OF LOWER ORGANISMS.

suggest that the explanation on the tropism schema is incorrect. These
details become intelligible as soon as we understand the real method
of reaction as set forth in the second of these contributions. The
assumption that the reaction is a typical tropism, when only the fact of
orientation is known, is as likely to fall to the ground in other cases
as in those just mentioned.

The reaction of bacteria to light, as shown by Bacterium photo-
metricuni^ described by Engelmann (1882), is a typical example of a
reaction through a motor reflex not fitting the tropism schema at all.

To sum up, it is clearly shown in certain cases that the reaction to
light takes place in a way that is not consistent with the theory of
tropisms, and this is true in some cases where a pronounced orienta-
tion exists. In many cases of orientation, where it is supposed that
the theory of tropisms holds, this is an assumption, for the observations
which would decide the matter are lacking.

REACTIONS TO GRAVITY.

In no case have the exact movements of unicellular organisms in
response to gravity been worked out in the manner in which this has been
done for the reactions to mechanical stimuli, chemicals, heat, light, and
electricity. We are, therefore, without the requisite data for deciding
whether these reactions agree with the theory of tropisms or do not.*

In the higher organisms in which the positive and negative reactions
to gravity have been observed (starfish, holothurians, flies, insect larvae,
etc.), the conditions are so complex that, so far as I am able to see,
observations which are crucial for deciding as to the mechanism of the
reactions have not been made and perhaps can not be made.

REACTIONS TO ELECTRICITY.

As we have seen in the third section of this paper, the reactions of
the rotifer to the continuous electric current do not take place in
accordance with the theory of tropisms. Anuraea shows a striking
orientation to the electric current, swimming directly to the cathode.
Yet this orientation is brought about in a way that is quite inconsistent
with the tropism schema. The reaction takes place through a " motor
reflex," the direction of turning is determined by an internal factor,
and not by the way in which the current strikes the organism. The
reaction can only be interpreted, therefore, as a reaction of the organism
as a whole.

♦ In a forthcoming paper by the author, based on work done since the above
was written, it will be shown that the reactions of Paramecium to gravity take
place in the same way as the reactions to most other stimuli, so that they do not
agree with the theory of tropisms.

THE THEORY OF TROPISMS. lOI

In the reactions of the ciliate infusoria to the constant electric current,
however, we have, if nowhere else, phenomena which show to a
certain extent clear-cut and undoubted agreement with the theory of
tropisms. To this agreement with the theory of tropisms much of the
widespread adherence to the tropism theory for reactions in general is
doubtless due. The reaction of infusoria to the electric current is con-
sidered a type for the other reactions of organisms.

Yet, in deciding to what extent the theoi*y of tropisms helps us to
understand the behavior of organisms, certain striking facts in regard
to the reaction to the electric current need to be taken into considera-
tion. These are the following :

(i) The reaction to the electric current never takes place in nature.
As has been repeatedly pointed out, the electric reaction is a product
of the laboratory ; it is a reaction which the organism never gives
under normal conditions. This being true, it should not be made the
type for reactions in general unless it can be shown clearly that the
characteristic features in the effects of electricity on organisms are
present also in the effects of other agents. Otherwise we may fall into
the same error that would exist if we considered the contortions of a
person who had grasped the electrodes of a powerful battery as a type
of human behavior in general.

(2) But examination shows that the characteristic features of the
effect of electricity on organisms are not present in the case of other
stimuli. The electric current, as the experiments of Kiihne (1864)
and Roux (1891) have shown, polarizes the cell. That is, it divides
it into halves, differing in chemical reaction. One half, in the case
described by Kiihne, had apparently an acid reaction, the other
half an alkaline reaction. In its effects on free-swimming organisms
a similar polarity is shown. In Paramecium, for example, the cilia
on one half of the body (where the current is entering) are caused to
take a certain position, while those on the other half (where the cur-
rent is leaving) take the opposite position. No other agent known
produces these polar effects^ either chemically or in the effect on the
motor organs. Yet it is to exactly these effects that the orientation
which makes this reaction the type for the tropism theory is due.

■If other agents produce these effects why are they not known and
described } There is no great difficulty in observing these effects with
the use of the electric current. Just as exact studies have been made
of the reactions to other stimuli ; yet, so far as I am aware, no one
has ever described any other stimulus as giving these characteristic
polar effects. On the contrary, the reactions to other stimuli are well
known not to show these characteristic phenomena.

Since, therefore, the characteristic phenomena of the reaction to the

I02 THE BEHAVIOR OF LOWER ORGANISMS.

electric current are not found in the reactions to other stimuli, it seems
a perversion to make the electric reaction a type for all others. The
reaction of the infusoria to the electric current takes in its characteristic
features a unique position amongthereactionsof the organism, requiring
special explanation.

(3) In the response of the infusoria to the electric current there
appears also the same type of reaction that occurs as a response to other
stimuli, but obscured by the phenomena peculiar to the effects of the
current.

This fact, that the reaction to the electric current is of a dual char-
acter, that the peculiar effects of the current are, as it were, superposed
upon the usual method of reaction, is not usually so clearly recognized
as it deserves to be.

If the constant current is passed through a preparation containing
large numbers of some species of the Hypotricha, as Stylonychia or
Oxytricha, it will be found that the animals, practically without excep-
tion, attain their orientation by turning toward the right side, thus
reacting as they would to any other stimulus. Further, if after they
are oriented the direction of the current is reversed, the animals all,
without exception, attain their new orientation (with anterior ends in
the opposite direction) by whirling toward their right sides. Thus, so
far, the reaction to the electric current is identical with that to other
stimuli, and the direction of turning is determined by an internal factor,
not by the way in which the current strikes the organism. In these
respects the Hypotricha agree with the Rotifera.

But exact observation shows that in the Hypotricha there is another
factor involved in the reaction. The characteristic polarizing effect of
the current appears in its action on the motor organs that are distributed
over the body surface ; those on one half of the body strike in one
direction, those on the other half in the opposite direction. Part of
these motor organs, therefore, assist in turning the organism in its usual
way (to the right) ; part oppose this turning. The result is that in
certain positions the turning to the right is opposed by the stroke of a
large number of cilia, so that the turning takes place more slowly than
usual. Nevertheless, in the Hypotricha, the determining factor in the
reaction to the electric current is almost throughout the same as in
the reaction to other stimuli ; the direction of turning is determined by
internal factors, as a reaction of the whole organism, not by the direction
in which the current strikes or passes through the organism. (Details
in the work of Pearl, 1900.)

If in place of one of the Hypotricha we experiment with an infusorian
in which the cilia cover closely the whole surface of the body, as Para-
mecium, the peculiar polarizing effect of the current on the cilia of the

THE THEORY OF TROPISMS. IO3

two halves of the body becomes much more powerful, because the
number of cilia affected in this way is much greater. The result is that
it almost alone determines the nature of the reaction. The direction of
turning is, therefore, determined by the way in which the current strikes
the body, as required by the theory of tropisms. But it should be
recognized that this is by no means universal among the infusoria ; in
doubtless fully as many cases the direction of turning is determined,
even under the electric stimulus, by an internal factor.

This peculiar dual character of the reaction to the electric current —
one strong factor being due to the inherent tendency of the organism
to turn in a certain definite way, without regard to the way in which
the stimulating agent impinges upon it — has been studied in detail in
recent contributions by Pearl (1900), Putter (1900), and Wallengren
(1902 and 1903). We may perhaps compare it, without indicating any
similarity in details, to the behavior of a person who has taken hold of
the electrodes from a powerful induction coil. He attempts in various
ways to free himself from the electrodes. This may be compared with
the attempt of the infusorian to perform its usual reaction to strong
stimuli. He also undergoes involuntary contortions, due to the action
of the electricity on his muscles ; these may be compared with the pecu-
liar effect of the electric current on the cilia of the infusoria, causing
them to strike in opposite directions on the two halves of the body.

Putting all together, we are not justified in taking the reaction of the
infusoria to the electric current as a general type for the reactions of
the lower organisms, because in its characteristic features it differs
from all the other known reactions. Yet it is exactly these unique
features that bring it into (partial) agreement with the tropism schema.

RESUME AND DISCUSSION.

We have thus passed in review the reactions of a large number of
lower organisms to the commoner stimuli, so far as they are known
from exact observations. We have found that as a rule they do not fit
into the tropism schema. In the reactions to mechanical stimuli, to
heat and cold, to chemicals, to changes in osmotic pressure, orientation
is not a primary or striking factor of the reaction ; when a common
orientation of a large number of organisms occurs, it is a secondary
result, due to the fact that the organisms are prevented from swimming
in any other direction. In the reaction to light orientation is a strik-
ing feature ; but the orientation, in certain precisely investigated cases
at least, is brought about in a manner which is inconsistent with the
tropism schema. In the reaction to gravity the precise reaction method
has not been worked out. Only in the reaction to the constant electric
current do we have in some organisms a partial agreement in principle

104 "T"^ BEHAVIOR OF LOWER ORGANISMS.

with the requirements of the tropism theory, and this agreement is due
to an effect on the organism in the production of which the electric
stimulus is unique, so far as known. In none of the reactions which
have been thoroughly worked out, except partially in the reaction to
the electric current, are the phenomena to be explained on the view
that the result is due to the direct action of the stimulating agent on the
motor organs of the part of the body on which it impinges. In the
reactions to mechanical stimuli and to light, and in the reactions to
the electric current in some animals, this view is absolutely disproved.
The direction in which the organism turns is, in all the well known
reactions of unicellular organisms and rotifers (except in a portion of
the reactions to the electric current) , determined by an internal factor,
and predictable from the structure of the organism without any know-
ledge of the direction from which the stimulating agent is to come.

We should perhaps consider here a modification of the original form
of the tropism theory that has been proposed by some authors. This
is in regard to the assumption that the stimulating agent acts directly
on the motor organs upon which it impinges. For this it is sometimes
proposed to substitute the view that the action of the stimulating agent
is directly on the sense organs of the side on which the stimulus im-
pinges, and only indirectly on the motor organs through their nervous
connection with the sense organs. When thus modified the theory, of
course, loses its simplicity and its direct explaining power, which made
it so attractive. In order to retain any of its value for explaining the
movements of organisms, it would have to hold at least that the connec-
tions between the sense organs and motor organs are of a perfectly
definite character, so that when a certain sense organ is stimulated a
certain motor organ moves in a certain way. When we find, as we
do in the flatworm (see the following paper), that to the same stimulus
on the same part of the body, under the same external conditions, the
animal sometimes reacts in one way, sometimes in another, the tropism
theory, of course, fails to supply a determining factor for the behavior.

But can we explain the reaction methods of the infusoria and other
animals which, as set forth above, are inconsistent with the tropism
theory in its original form, on the basis of the modification of this
theory, set forth in the last paragraph.? While in the infusoria the
assumption of nervous connections, etc., is inadmissible, we may
waive that objection and answer the question proposed from an analysis
of the obsei-ved phenomena. In Stentor or in Stylonychia, for example,
we find that the usual reaction to all classes of stimuli is by backing,
then turning toward the aboral side ; in some of the rotifers by turning
toward the aboral (dorsal) side. To simplify matters, we may take
into consideration only the turning toward the aboral side. This turn-

THE THEORY OF TROPISMS. IO5

ing is due to a certain method of movement of certain motor organs.
In the rotifers it is the coronal cilia which accomplish the turning,
while in the infusoria we know that the adoral cilia are concerned in
the movement. We may take the coronal or adoral cilia, then, as rep-
resentativ^e of the organs active in the turning. For convenience we
may designate these active organs simply as x.

Now, when the animal is stimulated on the right side, we find that
the motor organs x move in a definite way. On the tropism theory we
would conclude, therefore, that the portion of the right side stimulated
has nervous connection with the organs x. But we find also that when
stimulated on the left side, the oral side, or the aboral side, the organs
X move in exactly the same manner. In other words, we find that it
does not depend on the side stimulated what organs respond, as re-
quired by the tropism theory. This theory, then, in its modified form,
is of no more service for these cases than in its original form. The
responses in the animals which we have considered must, therefore,
be conceived as reactions of the organism as a whole, and due to some
physiological change produced by the stimulus, not as the result of
direct changes in certain motor organs when they or the parts with
which they are most closely connected are locally affected by a stimu-
lating agent. The facts show that the parts act in the service of the
wliole, not that the action of the whole is due to the more or less inde-
pendent irritability and activity of the parts.

Thus the facts brought out show that the theory of tropisms is not
of great service in helping us to understand the behavior of these lower
organisms. On the contrary, the reactions of these organisms seem
as a rule thoroughly inconsistent in principle with the fundamental
assumptions of the theory.

The facts brought out above are based on a study of what is, of course,
a comparatively small number of organisms. They rest chiefly on an
extensive study of the ciliate infusoria, with less thorough examination
of bacteria, flagellata, rotifers, and a few higher organisms. Doubtless
in organisms which are made up of many parts which are less firmly
bound together into a unified body than in those considered, we may
find greater independence of action in the parts. This seems to be the
case, for example, in the sea urchin, with its numerous independently
acting spines, pedicellariae, tube feet, etc. In this animal Von Uexkiill
(1900, 1900, a) concludes from his extensive study of the reactions that
many features in the behavior which seem at first view to be activities
of the animal as an individual are really due to the independent reac-
tions of the parts, so that he can say that while in walking, in the case
of the dog, '' the animal moves its legs ; in the sea urchin the legs move
the animal." This method of behavior has a general agreement with

I06 THE BEHAVIOR OP LOWER ORGANISMS.

what is demanded by the tropism schema, though when we come to
details of the behavior of the organs themselves, this theory seems
unsatisfactory, even in the sea urchin.

Such organisms as the sea urchin, composed anatomically and physio-
logically of many parts, each acting almost as an independent animal,
are certainly less common than more unified animals, such as we find
in the Infusoria, the Rotifera, the flatworms, etc. For this reason,
therefore, it has seemed worth while to sum up the real relations of the
behavior of these organisms to the tropism theory. The unicellular
animals are precisely those on which the prevailing theories of tropisms
or taxis have by many writers* been chiefly based. With the demon-
stration that the behavior of these organisms (as well as of many higher
ones), is for the greater part inconsistent with the tropism theory, per-
haps a large portion of the foundation for its acceptance as a general
formula for the chief features in the behavior of lower animals is cut
from beneath it.

In the following paper, on the part played in behavior by physio-
logical conditions of the organism, we shall find other, and, as it seems
to me, still more cogent, reasons for holding the tropism theory inade-
quate to account for the determining factors in the behavior of most
lower organisms.

SUMMARY.

The foregoing paper consists of a review of the behavior of Ciliata,
Flagellata, Bacteria ; of Rotatoria and certain other Metazoa, so far
as known from exact observation of their actions when stimulated,
with a view to determining how far the prevailing theory of tropisms
aids us in understanding the behavior of lower organisms.

The following are considered the essential points in the prevailing
theory of tropisms ; (i) That orientation is the primary factor in deter-
mining the movements of organisms into or out of certain regions, or
their collection in or avoidance of certain regions ; (2) that the action
of the stimulus is directly upon the motor organs of that part of the
organism upon which the stimulus impinges, thus giving rise to changes
in the state of contraction, which produce orientation.

The reactions of the organisms above named are then reviewed to
determine in how far there is agreement with these essential points in
the theory of tropisms. The following are pointed out:

The reactions to mechanical stimuli, to chemicals, to heat and cold,
and to variations in osmotic pressure have been described in detail, and
it is found that orientation is not a primary nor a striking factor in
them. The response in all these cases is produced through a " motor

♦This, however, is not true of Loeb.

THE THEORY OF TROPISMS. IO7

reaction" consisting usually of a movement backward, followed by a
turning toward a structurally defined side. The direction of turning is
thus determined by internal factors.

In the reaction to light orientation is a striking factor, but the orienta-
tion is not primary, being due to the production of the same " motor
reaction" described in the last paragraph. The method of orientation
is incompatible with the idea that orientation is due to the direct action
of the stimulus upon the motor organs of the part of the body on which
the light impinges, for orientation occurs by turning always toward a
certain structurally defined side, without regard to the part of the body
struck by the light. The turning may, therefore, be toward or away
from the source of light, or in any intermediate direction. In any case
it is continued or repeated until the anterior end is directed away from
the source of light, when it ceases.

The exact method of reaction to gravity has not been worked out by
direct observation.

In the reaction to the electric current the reaction method of the
rotifer is by a " motor reflex," and is hence inconsistent with the tro-
pism schema. In the Infusoria there is a partial (but only partial)
agreement with the requirements of the tropism theory. But this
partial agreement with the theory is due to certain peculiar effects of
the electric current which are not known to be produced by any other
stimulus. Hence the reaction to the electric current, far from being a
type for reactions in general, is a unique phenomenon, demanding
special explanation.

The general conclusion is drawn that the theory of tropisms does
not go far in helping us to understand the behavior of the lower organ-
isms ; on the contrary their reactions, when accurately studied, are, as
a rule, inconsistent with its fundamental assumptions. The responses
to stimuli are usually reactions of the organisms as wholes, brought
about by some physiological change produced by the stimulus ; they
can not, on account of the way in which they take place, be interpreted
as due to the direct effect of stimuli on the motor organs acting more
or less independently. The organism reacts as a unit, not as the sum
of a number of independently reacting organs.

FIFTH PAPER

PHYSIOLOGICAL STATES AS DETERMINING

FACTORS IN THE BEHAVIOR OF

LOWER ORGANISMS.

109

CONTENTS.

PAGE.

Nature and Evidences of Physiological States, . . . . . . iii

Physiological States in the Protozoa (Stentor as a Type), . . .112

Physiological States in the Lower Metazoa (the Flatworm as a Type), . 115
Changes in the Sense of " Tropisms " and other Reactions, . . . 117

Changes in the Sense of Reactions with Changes in the Intensity of the

Stimulus, 118

Interference of Stimuli, . . . . . . . . . .119

Spontaneous Movements, ......... 120

Methods of Causing Changes in Physiological States, .... 120

Nature of Reactions to Stimuli, 121

Physiological States in the Behavior of Higher Animals as compared with

those in Lower Organisms, ........ 124

Summary, ............ 126

no

PHYSIOLOGICAL STATES AS DETERMINING
FACTORS IN THE BEHAVIOR OF LOWER
ORGANISMS.

NATURE AND EVIDENCES OF PHYSIOLOGICAL STATES.

In studying the behavior of the lower organisms the units of obser-
vation, the factors to which especial attention has been paid have been
usually the iropisms and reflexes. These factors may be considered
as determined mainly (i) by the action of external agents on the
organism ; (2) by the structure of the organism.

An examination of the results of the study of reactions in the lower
animals up to the present time shows, I believe, that we must recog-
nize another set of factors in their behavior, of equal importance with
either of those already named. This set of factors may be characterized
by the general term physiological states.

By physiological states we mean the varying internal physiological
conditions of the organism, as distinguished from permanent anatomical
conditions. Such different internal ph3^siological conditions are not
directly perceptible to the observer, but can be inferred from their
results in the behavior of the animal. These results are of so marked
a character that the inference to different physiological conditions is
beyond question.

In the study of tropisms and reflexes a considerable number of
instances have been brought to light of changes in the reaction methods,
such as can be attributed only to changed physiological conditions.
Some of these cases will later be considered in detail in this paper.
Comparatively few investigations on the behavior of lower organisms
have been published in which attention has been consciously directed
to these physiological states, and in most of these the matter has been
taken up rather incidentally. We may mention as instances of papers
dealing more or less with this aspect of the matter that of Hodge and
Aikins (1895) on Vorticella, those of Von Uexkiill (1S99, 1900, 1900, a,
1903) on the sea urchin and on Sipunculus, my own on the behavior
of fixed Infusoria (Jennings, 1902), and that of Pearl (1903) on the
flatworm. In the study of higher organisms attention has of necessity
been largely directed to the phenomena determined by varying physio-
logical states, as these play a striking part in the behavior.

In the present paper an attempt will be made to collect and analyze

a number of the known cases showing the influence of physiological

states on the behavior of the lower animals, pointing out some of their

bearings on the theories of behavior.

Ill

112 THE BEHAVIOR OF LOWER ORGANISMS.

PHYSIOLOGICAL STATES IN THE PROTOZOA (STENTOR
AS A TYPE).

We will take up first the lowest organisms in which the matter has
been studied in detail, that is, the unicellular animals. These are of
special interest in view of their entire lack of a nervous system. As
the best-known case we may take the behavior of Stentor. This has
been described in full in a former paper by the present author (Jennings,
1902, a) ; for details this paper may be consulted.

When a quiet, extended Stentor is stimulated by lightly touching it
or the support to which it is attached with a rod, it reacts by giving a
definite reflex, that is, by contracting into its tube.

After this has taken place once or twice we find that the Stentor no
longer reacts as before. All the external conditions remain the same ;
the stimulus applied is the same. Nevertheless, the Stentor does not
react. Therefore, we must conclude that the Stentor itself has changed.
Its physiological condition is now difierent from what it was originally.
What the nature of the change in its condition is we do not know, save
in the fact that the Stentor in this second condition does not react as
does the Stentor in the first condition. For the sake of convenience
we may number the different physiological conditions, in order that we
may determine, if possible, how varied these conditions are. We will
call the physiological condition of the undisturbed extended Stentor,
before the stimulation. No. i, or the first condition. The condition
in which the Stentor no longer responds to the slight stimulus we will
call No. 2.

We may frequently distinguish still a third condition in the behavior
under this simple stimulus. At first the Stentor reacts by contraction
(condition i). Then it no longer reacts (condition 2). Later, or in
other cases, it may react to the stimulus, but by a different method
from the first reaction. It now bends over to one side when touched
with the rod. As set forth in my previous paper, ''The impression
made on the observer is very much as if the organism were at first
trying to escape a danger, and later merely trying to avoid an annoy-
ance." As conditioning this third method of behavior, when all out-
ward conditions are the same, we must postulate a third physiological
state differing from the other two ; this we may call condition No. 3.

We may thus distinguish at least three different physiological states in
the reactions to very weak stimuli, where the initial marked response
becomes a weak one or disappears. We may now analyze in the same
way the behavior under stimuli of a different character, when there is
a series of reactions which may be considered of increasing rather than

PHYSIOLOGICAL STATES AS DETEKMININCi FACTORS. II3

of decreasing intensity. Such a case is that described in my previous
paper, when water mixed with carmine particles is allowed to reach
the disk of Stentor.

The first physiological condition is again No. i — that of the undis-
turbed extended Stentor. In this condition the organism does not
respond to the stimulus at all. After the stimulus has continued
for some time, the organism does respond by turning into a new
position. We have, therefore, a new physiological condition. The
reaction in this case is the same as that given in condition No. 3,
described above. Whether the condition now existing is the same as
in the former case we do not know ; as we have no positive evidence
to the contrary, we will number it 3 also.

Next, after several repetitions of this reaction, the organism responds
in a still different manner, by momentarily reversing the ciliary current.
Since the stimulus and other external conditions remain the same, the
organism itself must have changed. We may call its physiological
condition at the present time No. 4.

Next, the animal contracts strongly and repeatedly. This is clearly
the result of a still different physiological condition which we may call
No. 5.

After thus contracting repeatedly we find that the organism remains
contracted much longer than it did at first. It is thus now in a new
physiological condition, which we may designate as No. 6.

Finally, it breaks its attachment to the bottom of the tube and swims
away through the water. Probably, therefore, we should distinguish
a seventh physiological state, corresponding to this reaction. It is
possible, however, that the breaking of the attachment is due to the
strong contractions which characterize condition No. 6, so that the
evidence for a seventh physiological condition is not unmistakable, and
it may be omitted from consideration.

We are able to distinguish clearly, therefore, in the study of these
two sets of reactions, at least six different physiological states. In each
of these states Stentor is a different organism, so far as its reactions to
stimuli are concerned. Clearly, then, the external stimuli and the
permanent anatomical configuration of the body are by no means the
deciding factors in the behavior. These factors, in the reaction series
last described, permit at least five different methods of behavior.
Which of these methods is actually realized depends not on the quality
or intensity of the stimulus, nor on the anatomical structure of the
organism, but on its physiological condition.

I do not wish to imply that I hold that the six different physiological
states above distinguished are sharply defined, separate things. On
the contrary, it is much more probable that the different physiological

114 THE BEHAVIOR OF LOWER ORGANISMS.

conditions form a continuum. We can, by taking sections, as it were,
at different intervals, distinguish at least six actually different condi-
tions ; but doubtless there exists every possible gradation from one to
another, so that still other actually different conditions could be distin-
guished if we had criteria for separating them. By careful analytical
experiments the number of different physiological conditions clearly
recognizable could doubtless be increased.

In other unicellular organisms doubtless a condition of affairs may
be found similar to that set forth above for Stentor. Hodge & Aikins
(1895) showed that the reactions of Vorticella vary with its physio-
logical condition. In the same paper (Jennings, 1902, a) in which
the behavior of Stentor was described, I have shown that in various
other fixed infusoria (Carchesium, Epistylis, etc.) the behavior like-
wise depends upon physiological states of the organism. In the
free-swimming infusoria this has not been shown to be true, at least to
any such extent. There may two grounds for this. Firstly, it is proba-
ble that in the free-swimming infusoria the behavior is actually less
varied than in the fixed species. A single motor reaction usually
removes them from the action of the stimulus causing it, so that there
is no reason for a recourse to other methods of reaction, as occurs in
Stentor. Secondly, in the free-swimming infusoria it is difficult, almost
impossible, to observe continuously the reactions of a given single
individual. This difficulty could doubtless be met by proper methods
of experimentation, and if this were done it can hardly be doubted that
a dependence of the reactions on the physiological states of the organism
would be found here also. Indeed, we have indirect evidence that this
is true in the case of Paramecium, in work already published. Thus,
in one of my earlier papers (1899, a, p. 374) I called attention to the
fact that Paramecia from different cultures often vary exceedingly in
their reaction to a given solution of a chemical. Still more pertinent
to the point under consideration is the fact, described in the first of my
studies (Jennings, 1897), as well as in the recent paper of Putter (1900),
of the great difference in the reaction of Paramecia and other fixed
infusoria when in contact with a solid, as contrasted with their reaction
when not thus in contact. As this, however, may be interpreted as an
interference of two stimuli, a discussion of the point is reserved until
later. A study of individual specimens of some of the larger Hypo-
tricha, such as Stylonychia, from the point of view of changes in reaction
methods with changes in physiological condition, would doubtless bring
forth interesting results.

Even in the lower unicellular organisms, the Rhizopoda, similar
dependence of the method of reaction on the physiological state of
the organism is known to exist. Thus Rhumbler (1898, p. 241) has

PHYSIOLOGICAL STATES AS DETERMINING FACTORS. II5

observed that Atnoeba verrucosa may at first begin to ingest an Alga
filament, then later, before the ingestion is complete, the filament may
be ejected. This involves a change in the physiological condition of
the Amoeba ; otherwise it would not now reject a certain object which
it before ingested.

PHYSIOLOGICAL STATES IN THE LOWER METAZOA (THE
FLATWORM AS A TYPE).

Passing now to the Metazoa, we find in the flatworm, Planaria, as
described by Pearl (1903), a dependence of the reaction of the organism
on its physiological condition similar to that which we saw above for
Stentor. The flatworm may be considered typical of the lower bilat-
eral Metazoa, so that it will be worth while to subject some features
in its behavior to a brief analysis from our present point of view.

We may examine for a simple typical case the reactions to mechani-
cal stimuli applied to one side of the anterior part of the body. The
flatworm is touched on one side with the tip of a hair or of a fine glass
rod. The resulting response is one of two reactions — the worm turns
either toward the point stimulated (positive reaction) or away from it
(negative reaction). Typically, the positive reaction is given to a
weak stimulus, while the negative reaction results from a strong
stimulus. The words " weak" and " strong" have, as we shall see,
only a relative meaning when used in this connection.

When, now, we ask which of these reactions shall be given as a
response to a certain stimulus, we find that this depends upon the
physiological condition of the organism. Pearl finds the reactions
determined by the following definitely marked physiological states :

1. Individuals are frequently in what may be called a resting con-
dition. The tonus is lowered ; the animals are very inactive and do
not respond readily to stimuli. This condition is compared by Pearl
with that of sleep in higher animals. When a flatworm in this con-
dition is given a light stimulus, such as would in an active specimen
induce a positive reaction, it does not respond at all. If the strength
of the stimulus is increased, the animal finally responds with a negative
reaction, turning away from the point stimulated. We may call this
the condition of lowered tonus.

2. In the more usual active condition the flatworm gives the positive
reaction to a very light touch, a negative reaction to a stronger stimulus.
We may call this the normal condition.

3. After the animal has been repeatedly stimulated it seems to become
excited; it moves about rapidly, and now gives always the negative
reaction to any mechanical stimulus to which it reacts at all. It behaves
much as many higher animals do under the influence of fear. We may
call this the excited condition.

1X6 THB BEHAVIOR OF LOWER ORGANISMS.

4. After the worm in the excited condition has been stimulated
repeatedly on one side, so that it turns its head steadily in the opposite
direction, after a time it suddenly changes its method of reaction. It
jerks backward, then turns the anterior end quickly through a consider-
able arc, usually toward the side from which the stimulus is coming,
so that the head now points in an entirely new direction. Since the
stimulus and other external conditions remain the same, the organism
must have passed into a new physiological condition, or it would not
now react in a different way. We may call this for convenience the
condition of over-stimulation.

5. Sometimes individuals are found which for a brief period (two or
three hours) seem in a much more active condition than usual. They
move about rapidly, but do not conduct themselves like the excited
individuals. As they move they keep the anterior end raised and wave
it continually from side to side as if searching. Specimens in this con-
dition react to almost all mechanical stimuli, whether weak or strong,
by the positive reaction, turning toward the point stimulated. Experi-
mentation failed to show that this condition was due to hunger. We
may speak of this fifth physiological condition as the state of heightened
activity.

In addition to the effect of these five well-defined physiological states
on the method of reaction to mechanical stimuli at the anterior part of
the body. Pearl finds that other less easily definable internal conditions
affect the reactions. At times an individual will give the positive reac-
tion to a stimulus of a certain strength a few times, then cease to give
it. On account of these and other complications due to varying internal
conditions. Pearl concludes :

It is almost an absolute necessity that one should become familiar, or perhaps
better, intimate, with an organism, so that he knows it in something the same
way that he knows a person, before he can get even an approximation of the
truth regarding its behavior.

We have taken up above only the physiological conditions influ-
encing the reactions to simple mechanical stimuli in the anterior region
of the body. We find the condition of affairs even here somewhat
involved. When other more complex stimuli are taken into considera-
tion the results of the interplay of change of physiological condition
and variations in the stimuli become, of course, much more complicated.

Thus we find in the bilateral metazoan Planaria, as in the unsymmet-
rical protozoan Stentor, that we can by no means predict the behavior
of the individual from a knowledge of the anatomical structure and of
the strength of the stimulus. The anatomical structure limits the possi-
bilities of reaction to several methods, which are, however, entirely
different or opposite in their effects on the relation of the organism to

PHYSIOLOGICAL STATES AS DETERMINING JtACTOIIS. II 7

the stimulus. Just which of these reactions shall be given as a response
to any particular stimulus depends on the physiological condition of
the organism. This physiological condition depends largely, as we
shall note later, on the history of the individual. Thus no single fixed
schema, such as we have in the tropism theory, can ever possibly
explain or define the essential points in the behavior of an animal.

Stentor and Planaria may be taken as typical examples of the higher
Protozoa and of the lower Metazoa, respectively. It is true that we are
not so well informed as to changes in physiological condition in other
lower organisms as in these two cases, but this is unquestionably due
merely to the fact that investigation has not been directed especially to
this point. There are, however, many cases in the literature which
explicitly or implicitly show the importance of physiological conditions
in determining the behavior of lower organisms. A number of these
cases may be brought together here.

CHANGES IN THE SENSE OF " TROPISMS " AND OTHER
REACTIONS.

Loeb (1S93) and Nagel (1894) have called attention to the fact that
certain worms and mollusks respond to a shadow by suddenly with-
drawing into their tubes, but that after the first reaction has been thus
produced the worms may no longer react. In this " after effect of the
stimulus" (Loeb) we have, of course, a case of changed physiological
condition.

Changes in the sense of "tropisms" belong here. Groom & Loeb
(1890) found that larvae of Balanus are at certain times of the day
positively phototactic ; at other times negatively phototactic. This
difference, in so far as it is independent of changed external conditions,
is, of course, due to differences in the physiological condition of the
organism. Loeb (1893) found that the larvae of the moth Porthesia
are positively phototactic when hungry ; not so after eating. Here we
have a well-defined physiological condition determining the nature of
the reaction ; hunger is one of the most important conditions in many
of the lower animals. Sosnowski (1899) and Moore (1903) show that
the geotropism of Paramecium changes from negative to positive under
various conditions. Towle (1900) and Yerkes (1900) have shown that
the sense of the phototactic reaction in Entomostraca is dependent on
the preceding treatment of the organism, mere transference with the
pipette often changing the sense of the reaction from positive to nega-
tive or vice versa. Such instances could doubtless be multiplied
indefinitely. Each case taken by itself seems perhaps of comparatively
little significance. We may look upon them, however, as indications of
an extensive dependence of behavior on physiological conditions, such

IlS THE BEHAVIOR OF LOWER ORGANISMS.

as we found in Stentor and the flatworm. Thorough investigation of
any of these organisms from this point of view would doubtless bring
to light a variety of physiological conditions on which the reactions
depend.

CHANGES IN THE SENSE OF REACTIONS WITH CHANGES
IN THE INTENSITY OF THE STIMULUS.

Must we not bring under the same point of view the well-known
phenomenon of a change in the sense of the reaction with a change in
the intensity of the stimulus ? As a simplest case of this we may take the
reaction of Stentor to mechanical stimuli. As shown in the ninth of
my studies (Jennings, 1902), Stentor reacts to a very weak mechani-
cal stimulus on one side of the disk by bending toward the source of
stimulus ; to a stronger but otherwise similar stimulus it responds by
contracting into the tube, or (later) by bending in another direction.
In the same way the flatworm reacts positively to a weak mechanical
or chemical stimulus, negatively to a stronger one. How can we ex-
plain these opposite reactions to stimuli of the same quality, differing
only in intensity.^

We have here, it seems to me, the same phenomenon shown in the
production of a change in physiological condition by a stimulus. We
know that even a single stimulus may produce a changed physiological
condition, as when after a single stimulus the organism no longer
reacts as before. We know also that the nature of the physiological
condition determines the reaction. In the present case we must con-
clude that a light stimulus throws the organism into a certain physio-
logical condition, whose concomitant reaction is turning toward the
point stimulated. A more intense stimulus induces a different
physiological condition, whose concomitant reaction is a contraction
into the tube (Stentor), or a turning in the opposite direction (flatworm).
The action of the stimulus, as we have seen in the foregoing paper
devoted to the theory of tropisms, cannot in most cases be directly on
the motor organs, so that from this point of view also we are almost
forced to the conclusion that the primary action of the stimulus is
to change the physiological condition of the organism. In any reac-
tion to stimulus we would have, therefore, the following steps : The
stimulus acting on the organism changes its physiological condition ;
this physiological condition induces a certain type of reaction. In
determining what physiological condition shall be produced, the inten-
sity of the stimulus is fully as important as its quality.

We have a similar reversal of the reaction as the intensity changes
in reactions to light. Many organisms are positive to weak light;
negative to strong light. The cause of this reversal of the reaction as

i»MYSIOLOGICAL STATES AS DETERMINII^G FACTOkS. 119

the light grows stronger has given rise to muph discussion (see
Hoh-nes, 1901 and 1903). We have here, of course, a parallel case to
the reversal of the reaction in Stentor or the flatworm under mechani-
cal stimuli of varying intensity. In the weak light we must suppose
the organism to be thrown into a certain physiological condition, the
concomitant of which is a certain type of reaction. In a more intense
light a different physiological condition is induced, corresponding to a
different reaction. The fact that different intensities of stimuli do
cause different physiological conditions and different reactions is, of
course, familiar to us, both from experimentation on animals and from
our own experience ; in the latter case we usually call the distinctive
reactions to very intense stimuli pain reactions. In the reversal of the
reaction to light as the light becomes stronger we have, it seems to me,
merely an instance of this general phenomenon, not differing in funda-
mental character from other instances.

In most of these cases we have, of course, a further problem in regard
to those features of the reaction which concern direction. Why does
the weak stimulus on the left side of Planaria cause a turning toward
that particular side? Or, why does a weak light from a certain direc-
tion cause Volvox to swim in that particular direction ? These problems
of direction are, of course, not touched in the foregoing discussion,
which, however, loses none of its force because these problems remain.
They are simply farther problems. The tropism theory gave a simple,
direct answer to these questions ; but, as we have alreadv shown in
the foregoing paper, this answer was, in many cases at least, not a
correct one. Possibly some combination of certain features of the
tropism theory with a consideration of the facts of changes in physio-
logical condition may give us a satisfactory answer to the problems
of direction.

INTERFERENCE OF STIMULI.

Again, we have in the interaction or interference of stimuli certain
phenomena which seem to fall under our present point of view. First,
we have the so-called cases of " heterogeneous induction," where the
action of one stimulus reverses or modifies the reaction to another.
For example, many cases are known in which animals positively photo-
tactic become negative, or vice versa^ when the temperature is changed.
(For a collection of such cases see Loeb, 1893, and Davenport, 1897,
p. 199.) In these cases the physiological condition of the organism
seems altered by one stimulus (as heat or cold), in such a way that it
no longer reacts to another stimulus (light) as it did before. In the
ciliate infusoria, specimens which are in contact with solids do not
react at all to many agents which under other circumstances call
forth a marked reaction. Putter (1900) has made a special study of

I20 THE BEHAVIOR OF LOWER ORGANISMS.

this matter, describing especially the interference of contact with the
reaction to heat and to electricity. In the second of these contributions
(p. 32), we have seen that attached Stentors do not react at all to
light. Physically considered, there is no necessary opposition between
the action of contact and the action of the other stimuli named. We
must conclude then that contact with solids so alters the physiological
condition of the organism that it no longer reacts to the other stimuli.

SPONTANEOUS MOVEMENTS.

Further, we find that alterations in physiological condition may cause
definite movements, which take place without external stimulus. Such
are the movements which we call spontaneous. As an example of this
we may take the case of Hydra. If an undisturbed green Hydra is
observed continuously, it is found to contract and again to extend with-
out visible cause every i J to 2 minutes. Thus, it remains at rest for
a period of, say, ij minutes. Its physiological condition at this time
we may call X. At the end of this period it contracts. Since the
external conditions have not changed the Hydra itself niust have
changed, otherwise it would continue at rest. The physiological condi-
tion A^ passes into the condition T^ and the Hydra as a result contracts.
This contraction is, of course, exactly the reaction given as a response
to most stimuli in Hydra. In Vorticella we find similar spontaneous
contractions at intervals, essentially as in Hydra. Cases of movements
in the lower organisms that are inaugurated by internal changes in
condition could, of course, be multiplied indefinitely. For our present
point of view it is of importance to recognize clearly the fact that a
change in physiological condition may, by itself, cause exactly the same
behavior that at other times appears as a response to external stimuli.

METHODS OF CAUSING CHANGES IN PHYSIOLOGICAL
CONDITION.

Changes in physiological condition are thus evidently brought about
in a number of different ways. We may attempt to summarize here
the different methods which appear to exist in the lower organisms.

(i) A single simple stimulus may bring about a change in physiologi-
cal condition. This is proved by the fact that the organism after it has
received a single stimulus may react differently from its previous
method. Thus, Stentor reacts to a single touch, but after this single
touch it may no longer react when touched in the same way again ;
or it may react in a different manner. It is probable, further, that the
first reaction to a single simple stimulus is to be considered due to a
change in physiological condition produced by this stimulus.

(2) Repetition of the same stimulus may cause a change in physio-
logical condition such as is not produced by a single stimulus. This

PHYSIOLOGICAL STATES AS DETERMINING FACTORS. 121

again may be illustrated from the experiments on Stentor, described
above (p. 112).

(3) Internal causes, not definable, may give rise to changes in physio-
logical condition. The result is spontaneous movement at intervals,
as described above for Vorticella and Hydra. As many authors have
pointed out, rhythmical SDontaneous movements may be due to a steady,
non-rhythmical, internal change of condition.

(4) The movement or reaction performed by the organism may
change the physiological condition. This is illustrated in one way by
the fact that after a single spontaneous contraction in Vorticella or
Hydra, the animal remains quiet for an interval, showing that the
original physiological condition was restored by the movement. The
fact that the reaction performed by the organism changes the physio-
logical condition of the latter is, of course, the basis of the formation of
habits in higher organisms ; in this case the performance of a reaction
once or repeatedly throws the organism into a condition where it is
more likely to react in the same way again. This particular method
of alteration of condition has perhaps not been clearly demonstrated
for unicellular organisms, though there is some indication of it in the
behavior of Vorticella, as described by Hodge & Aikins (1895), and
of Stentor as described by myself in the ninth of my studies. Thus,
Stentor responds to carmine in the water by a series of different reac-
tions, finally reaching the condition where it reacts by contracting at
once into its tube. If the stimulus is now repeated every time the
Stentor extends, it never gives its earlier method of reaction, but reacts
steadily for a long time by contracting at each stimulus. Is this persist-
ence in the contraction reaction due partly to the fact that it has begun
on this reaction method, and therefore keeps it up, or is it due only to
the fact that the stimulus has been repeated many times ? In the former
case the behavior would perhaps fall under our present point of view ;
in the latter it would not. Cases among the Protozoa where the repeated
performance of a reaction clearly makes the further performance of the
same reaction easier or more likely to occur, would be of much interest.

NATURE OF REACTIONS TO STIMULI.

The foregoing considerations evidently have a definite bearing on
the problem of the nature of reactions to stimuli. They lead, as set
forth briefly on page iiS, to the following conception of the steps oc-
curring in a reaction to a stimulus: (i) The stimulus acting on the
organism causes a change in its physiological condition ; (2) this
change in physiological condition gives rise to the typical reaction.

The evidence for this view is found scattered throughout the fore-
going discussion ; its main points may be briefly summarized here as
follows :

t%± THE SKHAVIOft OP LOWER ORGANISAtS.

(i) We have seen that stimuli do unquestionably cause changes in
physiological condition. This is demonstrated by the fact that after a
stimulus has occurred and ceased to act, the organism reacts differently
to the same or other stimuli. (For examples see pages 112, 117.)

(2) We have seen that the changes in physiological condition do un-
questionably cause definite movements, of exactly the sort that we are
accustomed to call reactions to stimuli. (Contraction of Hydra or Vor-
ticella, etc. ; see p. 120.)

These two facts give a solid foundation for the above view of
reactions to stimuli, and, indeed, it seems to me, raise a presumption
that reactions to stimuli are, as a rule, brought about in the way
described. Further evidence in favor of this view is as follows :

(3) In the paper which precedes the present one we have demon-
strated that, in the Infusoria and Rotifera at least, the action of stimuli
is not directly on the motor organs of that part of the body on which
the stimulus impinges. The organism reacts as a whole, and in a way
that is not explicable even on the assumption of a definite plan of ner-
vous interconnection between the regions stimulated and the motor
organs, an assumption that is, of course, in any case not allowable for the
Infusoria. Such reactions cannot be explained otherwise than as due
to changes in the physiological condition of the organism as a whole.
Further, evidence was given to show that the reactions of higher organ-
isms are in many cases equally inexplicable as a result of direct action
of the stimulus on the motor organs.

Only in the reaction of some organisms to the constant electric cur-
rent did we find such conditions fulfilled as permit an explanation of a
part of the phenomena on the theory of the direct action of the agent
on that part of the body on which it impinges, in accordance with the
theory of tropisms. Other features of the reaction to this stimulus (in
many cases the determining ones) are only explicable on the theory
that they are due to the physiological state of the organisms as a
whole, induced by the stimulus (see p. 100). This shows that we may
find at any time these two methods of action mixed, or perhaps eithei
one separately. But the reaction to the electric current is the only one
out of the reactions to a multitude of agents, that, in the Infusoria, has
been shown to have this additional feature — reactions of different parts
of the body in opposed ways. The reaction to the electric current is
thus of the very greatest interest, not because it stands as type for
reactions in general, but for exactly the opposite reason, because it
presents factors which are not known to occur in other reactions.

(4) The view that reactions to stimuli take place through the inter-
mediation of changes in the physiological condition of the organism as
a whole is further reinforced by the fact, set forth above, that it is only

PHYSIOLOGICAL STATES AS DETERMINING FACTORS. 1 23

on the basis of such a view that we can understand the changes of
reaction which occur when the same stimulus is repeated ; by the facts
of the interference of stimuli, even when their direct physical action is
by no means opposed ; by the facts of heterogeneous induction, and by
the fact that organisms at different times of the day or at different sea-
sons show different methods of reaction to the same stimuli.

(5) This view is also strengthened by the fact that it brings into
relation reactions to stimuli and spontaneous movements. On this view
both are due directly to the same cause, to changes in physiological
condition, produced in one case by internal causes, in the other by
external causes.

(6) This view receives powerful support, it seems to me, in our
knowledge of what takes place in the higher animals, including man.
This point I shall attempt to develop farther on in the present paper.

This view, that reactions to stimuli in the lower organisms are pro-
duced in general through changes in physiological condition, is not, of
course, set forward as anything new or original. Many others have
doubtless taken this point of view, and it is implied, perhaps not always
consciously, in many attempted explanations of animal behavior.
The writer is merely attempting to emphasize that particular interpre-
tation, out of many existing ones, towards which the facts seem to
point strongly. He is convinced that the factor of physiological con-
dition as determining behavior has not oeen so fully and explicitly
realized and dealt with in work on the lower organisms as the facts
demand, and that many things that seem anomalous fall into their
proper places when this factor is taken fully into consideration.

What is the nature of physiological conditions, or changes in phys-
iological condition ? Of course, we are not able to answer this ques-
tion. One is tempted to think of these expressions as signifying
something like chemical states or changes in chemical states. But the
concept of physiological states is, for higher animals at least, one at
which we arrive by analysis of complex phenomena in behavior, and
this does not give us any direct evidence as to the real nature of the
change in the living substance (considered as matter) which takes
place when the physiological condition changes.

The concept '' physiological states" is a preliminary collective con-
cept, which may later be analyzed into many. Such analysis is certain,
however, to be difficult and hypothetical in character in the lowest
organisms. In man we have, of course, a basis for analysis in the sub-
jective accompaniments of physiological (here called psychological)
conditions, — in the feelings, emotions, etc.

124 "^"^ BEHAVIOR OF LOWER ORGANISMS.

PHYSIOLOGICAL STATES IN BEHAVIOR OF HIGHER ANIMALS,
AS COMPARED WITH THOSE IN LOWER ORGANISMS.

Realization of the fact that the behavior, even in the lowest organisms,
is determined to a large degree by physiological states must be of great
service in welding into one connected whole the study of behavior in
all animals, from the lowest up to man. The attempt to divorce the
study of the behavior of man from that of the lower animals, which
has been evident in late years, seems unfortunate and unnecessary. It
is true that we are not justified in reading the subjective states of man
directly into the lower organisms. But we are not confronted with the
alternative of doing this or of separating the two subjects completely.
The behavior of man can be studied from the same objective standpoint
which we employ in investigating the behavior of animals. When this
is done, there is no reason for holding the results on man aloof from
those obtained elsewhere ; if it is proper to compare diflerent organ-
isms of any kind from this point of view, in order to obtain general
results, as all investigators do, it is certainly proper to draw man also
into the circle of comparison. The fact that in man we can know also
the subjective accompaniments of the different physiological states and
reactions is by no means a disadvantage in this comparison ; it is merely
an additional feature, of the highest possible interest. We can even,
it seems to me, justifiably call attention to the relation between the
subjective states as found in man to certain general phenomena common
to man and other organisms. It is only when we proceed directly to
attribute to the lower animals the subjective states which we know only
in man (and, indeed, only in our own individual minds) that we pass
the boundary of scientific procedure.

In the higher animals, and especially in man, the essential features
in behavior depend very largely on the history of the individual ; in
other words, upon the present physiological condition of the individual,
as determined by the stimuli it has received and the reactions it has
performed. But in this respect the higher animals do not differ in
principle, but only in degree, from the lower organisms, as we have seen
in our analysis of the behavior of Stentor. In this unicellular form
we were forced to distinguish at least six different physiological condi-
tions, determining in the same individual different reactions to the same
stimuli. In the higher animals, and especially in man, we can distin-
guish, as might be expected, an immensely greater number of such
conditions which induce difterent reactions, but there is no evident differ-
ence in principle in the two cases. Can we go farther and make a more
direct comparison of individual physiological states in the higher and
lower organisms? We find in Stentor, and again in the flatworm,
that after the organism has been repeatedly stimulated by an agent

PHYSIOLOGICAL STATES AS DETERMINING FACTORS. 1 25

which must in the long run be classed as injurious, it is thrown into a
physiological condition in which its reactions become more rapid and
powerful, and of such a nature as to remove the organism from the
source of stimulus. We find that in this state the organism reacts to
any stimulus to which it reacts at all by a strong negative reaction.
In higher animals we frequently find the same condition of affairs, and
the animal is then commonly said to be frightened. Finally, we often
find in man a similar condition, and here we know certain subjective
accompaniments of the physiological condition, the most characteris-
tic of which is perhaps the emotion of fear. In all these cases the
objective manifestations of the physiological condition are of the same
character. Does the fact that in man we know something additional
about the matter, the subjective accompaniments, constitute grounds
for denying the essential similarity, from a physiological standpoint,
of this condition in man and that in the lower organisms.? It seems to
me that it does not ; in fact, all that is maintained in making the com-
parison is that this condition causes similar objective phenomena and
is brought about by similar conditions. Further than this our analysis
and comparison cannot go.

Another class of physiological conditions which we can distinguish
almost all the way through the animal series is that produced character-
istically by intense stimuli, as opposed to faint stimuli. As a rule, any
stimulus, even if it is one to which the organisms respond usually by a
positive reaction, produces, when it becomes very intense, reactions
whose general effect is to remove the organisms from the source of
stimulation (negative reactions).* This is true in Amoeba, where weak
mechanical stimuli cause spreading out and movement toward the
source of stimulus, while strong mechanical stimuli cause it to contract
and move away ; it is true for Stentor ; it is true for the stimulus of
weak and strong light in Euglena and Volvox and many other
organisms ; it is true for mechanical and chemical stimuli in the flat-
worm ; it is true in general for higher animals and man. In all these
cases the intense stimulus evidently changes the physiological con-
dition so that the organism now reacts negatively. In man we know
that this physiological condition is accompanied subjectively by pain
or at least discomfort, and even in higher animals such reactions are
usually spoken of as pain reactions. Objectively considered, the
phenomena are analogous throughout the animal series, so that we

***All organisms behave in two great and opposite ways toward stimulations;
they approach them or they recede from them. Creatures which move as a
whole move toward some kinds of stimulations, and recede from others. Crea-
tures which are fixed in their habitat expand toward certain stimulations, and
contract away from others." Baldwin, 1897, p. 199.

ia6 THE BEHAVIOR OF LOWER ORGANISMS.

may properly characterize the physiological condition which produces
the negative reaction to strong stimuli, with Professor J. Mark Bald-
win (1897, p. 43) as the " physiological analogue of pain." This, of
course, by no means commits us to the belief that the organisms have
a sensation of pain ; concerning this we know nothing.

It thus seems to me possible to trace some of the physiological con-
ditions which we know, from objective evidence, to exist in man and
the higher animals, back to the lowest organisms. Many conditions that
we can clearly distinguish in man will doubtless be followed back to a
common single condition in the lower organisms ; but this is exactly what
we should expect. Differentiation takes place as we pass upward in
the scale, in these matters as in others.

The most interesting and important field in which we find the
behavior of higher organisms dependent on their previous history, and,
therefore, on their present condition as influenced by previous experi-
ence, is in that group of phenomena which we call memory, or learning
by experience. Memory has as its basis the general phenomenon that
a stimulus received or a reaction performed leaves a trace on the
organism, or modifies its condition in such a way that it later reacts
differently to the same stimulus. This basis of memory is, of course,
clearly present in Stentor.

The analysis of the different physiological conditions found in the
lower organisms, the influences to which they are due, and the
reactions of these organisms as influenced by physiological conditions
certainly forms a most promising field for research, and one as yet
almost untouched.

SUMMARY.

The present paper attempts to show, by an analysis of certain
phenomena in the behavior of lower organisms, taking Stentor and
Planaria as types, that physiological states of the organism are most
important determining factors in reactions and behavior. In these
organisms, to the same stimuli, under the same external conditions,
the same individuals react at different limes in radically different ways,
showing the existence of different physiological states of the organism,
which determine the nature of the reactions. In a unicellular organ-
ism (Stentor) we can distinguish at least six different physiological
states, in each of which the organism has a different reaction method,
and corresponding facts are brought out for the flatworm. Scattering
observations taken from works on tropisms, etc., are shown to indicate
that the same state of affairs is found in other lower organisms.

The conditions producing these different physiological states are
examined and their importance for the theory of behavior in the lower
organisms is brought out. The relations of these facts to '' interference

PHYSIOLOGICAL STATES AS DETERMINING FACTORS. 1 27

of stimuli," ''heterogeneous induction," "spontaneous movements,"
and "changes in the sense of reactions with a change of intensity in
the stimulus," are developed.

The view is set forth that in most of the lower organisms a reaction
to stimulus usually involves the following factors: (i) the stimulus
changes the physiological state of the organism as a whole ; (2) this
change in physiological state induces a certain type of reaction.
Evidence for this view is summarized.

Finally, it is pointed out that realization of the importance of
physiological states as determining factors in the behavior of the lower
organism is of service in bringing the study of these organisms into
relation with that of higher animals and man. An objective study of
the behavior of these higher animals shows the prevalence of physio-
logical states as determining factors in behavior, and in some cases, at
least, some of these states are closely analogous to what we find even
in unicellular organisms.

SIXTH PAPER.

THE MOVEMENTS AND REACTIONS
OF AMCEBA.

"9 1

CONTENTS.

Introduction: Objects of the Investigation 131

Description of the Movements and Reactions, 132

The Movements, 132

The Movements of Amoeba as described Formation and Retraction of Pseudopodia 15a

by Rhumbler and Butschli ; Agree- Surface Currents in Formation of

ment with Currents in a Drop of Pseudopodia in contact with Sub-

Fluid Moving as a Result of a Local

Decrease in Surface Tension 13a

Currents in Amoeba as studied from above;

stratum 152

Formation of Free Pseudopodia 153

Withdrawal of Pseudopodia 156

Movements at Anterior Edge i6o

Lack of Backward Currents 134 Movements of Posterior Part of Body. ... 165

Movements of Upper and Lower Surfaces General View of Movements of Amceba in

Studied Experimentally; Rolling Locomotion 169

Movement 138 Some Characteristics of the Substance of

A mceia verrucosa siTid Its Rdsitivcs. 140 Amoeba 173

Other Species of Amoeba 146 Fluidity 173

Historical on Rolling Movements in Rhumbler's Ento-ectoplasm Process 173

Elasticity of Form in Amoeba........... 175

Contractility in Ectosarc of Amoeba. 177

Amoeba 148

Reactions to Stimuli 181

Reactions to Mechanical Stimuli 181 Some Complex Activities 193

Positive Reaction 181 Activities connected with Food-taking 193

Negative Reaction 182 Taking Food „ 193

Reaction to Chemical Stimuli 187 Pursuit of Food 196

Reaction to Heat 190 Other Amoeba as Food 198

Reactions to Other Simple Stimuli 191 Reactions to Injuries aoa

Physical Theories and Physical Imitations of Amoeboid Movements, . 204

Surface Tension Theory 204 Experimental Imitation of Movements

Berthold's Theory that One-sided Adher- due to Local Contractions of Ectosarc

ence to Substratum is the Cause of and of the Roughening of Ectosarc in

Locomotion 208 Contraction 215

Experimental Imitation of Locomo- Direct or Indirect Action of External

tion in Amoeba 209 Agents in Modifying Movements 219

Formation of Free Pseudopodia 214 Direct or Indirect Action in Food-taking, 232

General Conclusion 225

Behavior of Amoeba from Standpoint of Comparative Study of Animal

Behavior, 226

Habits in Amoeba - 226 Relation of Different Reactions to Differ-

Classes of Stimuli to which Amoeba Re- ent Stimuli; Adaptation in Beha-

j^cts 227 vior of Amoeba aa;

Types of Reaction 227 Reflexes and "Automatic Actions" in

Amoeba 228

Variability and Modifiability of Reactions 229

Summary 230

130

THE MOVEMENTS AND REACTIONS OF AMCEBA.

INTRODUCTION: OBJECTS OF THE INVESTIGATION.

The present paper contains the results of an investigation which was
undertaken with two general problems in mind. The first purpose
was to determine by observation and experiment, from the standpoint
of the student of animal behavior, how far recent physical and
mechanical theories go in aiding us to explain the behavior of Amoeba.
The second object of the work was to furnish needed additional data
on the reactions of Amoeba to stimuli, and to systematize and unify
our knowledge of its behavior.

The recent theories which would resolve the activities of Amoeba
largely into phenomena due to alterations in the surface tension of a
complex fluid seem to promise much. They are of precisely the
character from which most may be hoped ; from a study of the physics
of matter in a state similar to that found in the living substance, the
laws of action of this living substance are sought. Such theories have
been developed, as is .well known, by Berthold (1886), Quincke
(1S88), Biitschli (1892), Verworn (1892), Rhumbler (1898), Bern-
stein (1900), Jensen (1901), and others. The success of this method
of attacking the problems seems great. Activities similar, at least
externally, to those of Amoeba, are produced by physical means, and
fully analyzed from the physical and mechanical standpoint. In this
manner the movement, the control of movement by external agents,
the feeding, the choice of food, the making of the shell, and other
features of the behavior have been more or less closely imitated,* and
in a way permitting a complete analysis in accordance with chemical
and physical laws.

From the standpoint of the student of animal behavior, the resolu-
tion of the behavior of any organism into the action of known physical
laws must be a matter of the deepest interest. The actions of higher
organisms seem at present so far from such a resolution that some
investigators believe an essential difierence in principle to exist
between the behavior of living things and non-living things ; between
the laws of biology and those of physics. The resolution, then, of the
behavior of even the simplest organism into known physical factors
would be an event of capital significance, affecting fundamentally the
whole theory of animal behavior. A renewed thorough study of the

* See especially Rhumbler, 1898. 131

132 THE BEHAVIOR OF LOWER ORGANISMS.

facts, with especial reference to these theories, seems, therefore, much
to be desired. The results of the present study will show, I believe,
that such a re-examination of the facts was greatly needed.

As to the second object of this investigation, stated above, it is a
somewhat remarkable fact that the observational basis for a number of
the most important reactions assumed to exist in Amoeba is exceedingly
scanty, particularly so far as control of the direction of movement is
concerned. For example, one of the reactions most often assumed to
exist in Amoeba, and most commonly selected for imitation by
physical means, is chemotaxis, the movement toward or away from a
diffusing chemical. But no account exists, so far as I have been able
to discover, of actual observation of such a reaction in Amoeba, under
experimental conditions. Again, the effects of slight or of intense
localized mechanical stimuli, in controlling the direction of movement,
has not been worked out in detail. To fill these and similar gaps in
our knowledge, and to bring the different reactions into relation with
each other, so as to make possible a connected account of the behavior
of Amoeba, is, then, the second object of this paper.

I shall first give an account of the movements and reactions of
Amoeba, as determined by observation and experiment, without enter-
ing in detail upon the theories of the subject. This will be followed
by a section dealing with the physical theories and physical imitations
of the movements and reactions, in the light of the facts set forth in the
first section. A brief final section will be devoted to a characterization
of the behavior of Amoeba from the standpoint of the student of
animal behavior.

I am compelled to give a full description of the normal movements
of Amoeba, as the course of the investigation showed that the prevalent
conception of these movements, on which many of the theories have
been based, is not correct.

DESCRIPTION OF THE MOVEMENTS AND REACTIONS.
THE MOVEMENTS.

MOVEMENTS OF AMCEBA AS DESCRIBED BY RHUMBLER AND BUTSCHLI ;
AGREEMENT WITH CURRENTS IN A DROP OF FLUID MOVING AS A RE-
SULT OF LOCAL DECREASE IN SURFACE TENSION.

There are few subjects that have been studied more than the nature
of the movements of Amoeba, but nothing final has been reached, even
from the descriptive standpoint. The first preliminary to an under-
standing of the nature of the movements must be to determine just what
movements take place.

The most extensive recent study of the movements of Amoeba has
been made by Rhumbler (1898), though the magnificent monograph of

THE MOVEMENTS AND REACTIONS OF AMGBBA.

133

the Rhizopods by Penard (1902) contains incidentally a large number
of valuable observations on this matter.

According to Rhumbler (/. c.) the movements in normal locomotion
are typically as follows : From the hinder end of the Amoeba (or of the
pseudopodium, if a single pseudopodium is under consideration) a
current of endosarc passes forward in the middle axis ; in front this flows
outward toward the sides, then backward along the surface, gradually
coming to rest. Figs. 30 and 31 , taken from Rhumbler, give diagrams
of these currents in an Amoeba moving as a whole (Fig. 30), and in
the formation of pseudopodia (Fig. 31). In an Amoeba which forms
more than one pseudopodium at once, these typical currents become
somewhat complicated (Fig. 32), but retain their main features. The
backward current shown at the sides in Figs. 30-32 is conceived to be
present also above and below, that is, over the whole surface of the
Amoeba. A diagram of the currents in side view, as given by Rhum-
bler, is shown in Fig. 33, B. An essentially similar account of the
currents is given by Biitschli (1880, 1892).

Fig. 30.*

Fig. 3i.t

Fig. 324

Fig. 33. §

The most striking feature in the currents as above set forth is the fact
that they agree precisely with the currents produced in a drop of fluid
of any sort when the surface tension is lowered over a certain limited
area. There is always a current over the surftice away from the region
where the tension is lowered, while an axial current moves toward the

* Fig. 30. — Diagram of the currents in a progressing Amoeba Umax, after
Rhumbler (1S98).

fFiG. 31. — Diagram of the ** fountain currents " in pseudopodia of Amoeba,
after Rhumbler (1898).

J Fig. 32. — Diagram of complex '* (puntain currents" in an Amoeba with two
large pseudopodia, after Rhumbler (1898).

§ Fig. 33. — Comparative diagrams of the currents in a rolling movement, and
in the movement of Amoeba, as conceived by Rhumbler, viewed from the side.
In A are represented what Rhumbler conceives to be the necessary currents in
a rolling movement, while B represents what Rhumbler considers the really
existing currents in Amoeba, as seen from the side. The heavier arrows in each
case represent the current on the lower surface. After Rhumbler (1898).

134 "^^^ BEHAVIOR OF LOWER ORGANISMS.

point of lowered tension. Diagrams of the movement of such drops
are given in Fig. 34. Further, the drop may elongate in the direction
of the axial current, and may move bodily in that direction, just as
happens in Amoeba.* It is most natural, therefore, to conclude as
Biitschli (1892) and Rhumbler (1898) have done, that the movements
of Amoeba are likewise due to a lowering of the surface tension at the
anterior end, provided that its movements really take place in the
way described above,

CURRENTS IN AMCEBA AS STUDIED FROM ABOVE ; LACK OF BACKWARD

CURRENTS.

At the beginning of my work I had no doubt that the movements

occurred exactly as above described, and, therefore, did not devote

special attention to this point. But I was soon struck by the fact that

I was unable to see any backward current at the sides, as represented

in Figs. 30 and 31. Further careful study of the movements of

Amoeba limax^ A. proteus^ A. angulata^ A. verrucosa^ A. sphcero-

nucleolus^ and one or two

undetermined species

confirmed this fact, and I

may say at once that after

several months' continu-

^ T^ J. ous study of the move-

FiG.34.t ^ A .' c

ments and reactions of

Amoeba I have never, except in one or two doubtful instances, seen
any backward movement of the substance at the sides or on the surface
of an Amoeba that was moving forward in a definite direction.

It is true that in the movements of Amoeba Umax, for example, one
receives the impression of two sets of currents, one forward in the cen-
tral axis, the other backward at the sides. But if the latter is studied
carefully it is found that there is really no current here ; the proto-
plasm is at rest, and the impression of a backward current at the sides
is produced only by contrast with the forward axial current. Amoeba

* All these facts are easily verified by placing a drop of clove oil on a slide in
a mixture of two parts glycerine to one part 95 per cent alcohol under a cover
supported by glass rods, as described in a previous paper by the present author
(Jennings, 1902). By mixing some soot or India ink with the clove oil the cur-
rents are made evident.

t Fig. 34. — Currents in a drop of fluid when the surface tension is decreased
on one side. A, the currents in a suspended drop, when the surface tension is
decreased at a. After Berthold (1886). B, axial and surface currents in a drop
of clove oil, in which the surface tension is decreased at the side a. The drop
elongates and moves in the direction of a, so that an anterior (a) and a posterior
{p) end are distinguishable.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 1 35

Umax contains usually a large number of fine granules, which in many
cases extend to the very outer surface, so that it is not possible to dis-
tinguish an ectosarc, in the sense of a layer containing no granules.
By watching the movements of these particles it is possible to determine
the direction of the currents in the protoplasm. The movements in
locomotion are usually as follows ; At the anterior end there pushes
forth from the interior a clear substance, which I will call the
hyaloplasm. As this moves forward it spreads out laterally, till it
reaches a position such that it forms a continuation forward of the
remainder of the lateral boundary of the animal. Into this hyaloplasm
flows then the granular endosarc. The granules flow forward, rapidly
in the middle, usually more slowly near the sides. As it reaches the
anterior end the central current spreads out in a fanlike manner, so
that some of the granules approach closely the lateral borders of the
Amoeba (Fig. 35). They then stop, while the central part of the cur-
rent passes on, following the advancing anterior end.

So long as one confines his attention to the Amoeba alone, not
observing external objects,
one receives the impres-
sion that there are two
sets of currents, an axial
current forward, marginal

currents backward. But

a u- Fig. 35.*

as soon as one nxes his •'^

eye upon a particular granule in the apparent backward marginal
current, and observes its relation to some external object, he dis-
covers that no such current exists. The granule remains quiet,
retaining continually its position with relation both to other granules
in the edge of the Amoeba and to objects external to the Amoeba.
Meanwhile the remainder of the substance of the Amoeba is flowing
past, so that the granule in question after a time comes to occupy
a position at the middle of the length of the Amoeba. At about
this point it usually begins to move slowly forward again, though
much less rapidly than the internal current. The nature of this
slow forward movement we shall take up later (p. 166). The main
portion of the body of the Amoeba thus continues to pass the granule,
and the latter finally reaches the posterior end. Here it usually re-
mains quiet for a time (moving forward only as the posterior end is
dragged forward). Then it is taken into the central current again,
passes to the anterior end, and comes to rest as before, while the
remainder of the Amoeba passes it by ; and this process is repeated

* Fig. 35. — Diagram of the movements of particles in an advancing Amoeba.
Each broken line represents the path of a particular particle.

136 THE BKHAVIOIl OF LOWER ORGANISMS.

indefinitely. In favorable cases I have repeatedly followed a single
granule from the posterior end forward till it came to rest at the
anterior end, then watched the body of the Amoeba pass it by, until it
was again at the posterior end and started forward anew. The course
of a single granule is represented in Fig. 36. As is evident from this
figure, the granule does not travel backward in any part of its course.
Not all the granules, however, remain quiet until they have passed to
the posterior end. Many of them are taken again into the central
stream before the entire body of the Amoeba has passed them. Large
granules usually stop only a short time, starting forward again before
the middle of the Amoeba has reached them ; others are taken up at the
middle or farther back, while many smaller granules reach the posterior
end. But as a rule none show any movement backward, so far as I
have observed.

It is not only at the margins of the Amoeba, but also on the under
surface, in contact with the substratum, that the ectosarc with its
granules is at rest or moving slowly forward in the posterior half.
This is evident when the lower surface of a transparent Amoeba is
brought into focus.

2 That excellent observer,

/ <^^ __^ Dr. Wallich, saw clearly

r^ 7^*^^~-— '''' /^TTin^r-^ ) "^^ny years ago that there

V_£^^~~,^v_^ ^Illja^^rrrr^^:^ j vf is really no backward cur-

O. g 'c "^ e J rent, though at first view

Fig. 36.* there appears to be such.

It is only necessary to watch a specimen of Amoeba carefully to become con-
vinced that the appearance of a returning, as well as an advancing, stream of
granules is illusory. The stream, it will be observed, is invariably in the direc-
tion of the preponderating pseudopodial projections. The particles simply flow
along with the advancing rush of protoplasm. There is no return stream, but
the semblance of one is engendered by one layer of particles remaining at rest
whilst another is moving past them. (Wallich, 1863, ^> P* 33i-)

This statement of the facts my observations fully confirm.
In this account of the lack of backward movement in the granules
of the ectosarc on the lower surface and at the margins I find myself

*FiG. 36. — Diagram of the movements of a single particle in Amoeba, as
seen from above. The particle begins at a, passes to b and then to c, at the
anterior edge of the Amoeba shown in the outline i. The Amoeba now passes
forward to the position 2, and thence to 3, while the particle retains the posi-
tion c; when the Amoeba has reached the position 3 the particle is thus at
its posterior end. Now the particle moves forward again, from c to <f, and
thence to e and/", thus coming again to the anterior edge. Here it stops, as at
c, until the body of the Amoeba has passed it. As the figure shows, the particle
does not move backward in any part of its course.

THE MOVEMENTS AND REACTIONS OF AMCEBA. I37

at variance with certain statements of F. E. Schulze, Biitschli, and
Rhumbler. I am aware that this conflict of my observations with
those of the investigators named, who deservedly rank among the
highest in the field at present under consideration as well as elsewhere,
renders the utmost caution necessary in trusting to these results. Yet,
with this consideration in mind, and with the confident expectation in
undertaking the work that I should find the currents exactly as described
by these authors, I have been unable to come to any result save that
above set forth. The statements of Schulze (1875, pp. 344-348) deal
with Pelomyxa palustris Greef. In this animal, according to Schulze,
there are resting portions at the sides of the body, while from behind
currents pass forward through the channel enclosed by these resting
portions. At the anterior end the lateral parts of these currents turn
outward and, finally, a little backward ; any
given portion passes backward but a short dis-
tance. The currents are shown by Schulze in
a figure, a reduced copy of which is given here-
with (Fig. 37). According to Schulze the cur-
rents have this form on the upper surface as well l'+^*^'*^^V\\lTui/^//'^ Vr}— :v
as at the sides — that is, a part of the current flows
upward and backward on the upper surface.
Biitschli (1892, Anhang, p. 220) confirms this
account of the currents in Pelomyxa.

I regret that I have been unable to obtain
specimens of Pelomyxa in order to examine
these phenomena for myself. One would of

course be bold to doubt the correctness of the

Fig. ^7 *
observations of such investigators as Schulze

and Biitschli, and it is possible that Pelomyxa differs from Amoeba
in this respect. Yet, as we shall see later (p. 149), the account given
by these authors is certainly incorrect so far as the backward cur-
rents on the upper surface are concerned ; it is possible, then, that the
appearance of a backward current elsewhere was deceptive.

Rhumbler (1898) describes the forward axial and the backward side
currents in various species of Amoeba, and considers such movements
as typical, basing his theory of locomotion upon them. It seems
probable that slight backward currents, such as were described by
Schulze (Fig. 37), do occur at times at the sides of the advancing
anterior end. The posterior part of the Amoeba is narrow and
rounded, the anterior part broad and thin. The current of endosarc
flows from this narrow posterior portion into the broad anterior part

*FiG. 37.— Currents in a progressing Pelomyxa, as seen from above, after
Schulze (1875). The longer arrows represent stronger currents.

138 THE BEHAVIOR OF LOWER ORGANISMS.

and must therefore spread out ; it would not be unnatural for the
currents to flow even backward a little, as in Schulze's figure (Fig.
37), in order to fill the area just in front of the resting portion of the
protoplasm (at x, Fig. 37). As we shall see, such movement is
sometimes to be observed in inorganic fluids under similar conditions
(p. 211). Whatever the explanation of the difference between my
observations and those of the investigators named, the point of impor-
tance is that the backward current is not a constant nor an essential
part of the locomotion of Amoeba, so that it does not form a fitting
basis for a theory of locomotion. Further, as we shall see, I am able
to demonstrate conclusively the incorrectness of that conception of the
nature of amoeboid movement for which alone the account of the
currents given by Biitschli and Rhumbler is significant.

It is evident that the method of movement here described is better
adapted to the production of locomotion in a given direction than that
which Biitschli and Rhumbler describe (see Figs. 30-33), since accord-
ing to their account a portion of the substance of the body is first trans-
ported forward, then backward. In the locomotion as I observed it
there is no such useless transportation of substance in a direction
opposed to that in which the animal is traveling.

On the other hand, the movements as I have described them bear
much less resemblance to those produced in drops of fluid by local
changes in surface tension (Fig. 34) . There is only the slight turning
outward at the anterior end that can be at all compared to the backward
flow of an outer layer in the inorganic drop. Rhumbler himself notes
that in Amoeba angulata there is often no such backward current to
be seen (Rhumbler, 1898, p. 120), but bases his theory of the forward
movement entirely on the cases where it (supposedly) does occur. In
Amoeba angulata^ A. verrucosa^ and A. sphceronucleolus^ according
to my observations, there is often no indication even of the turning
out of the particles in a fanlike manner ; they merely flow forward and
stop for a time. Biitschli (1892, p. 199) notes that the backward cur-
rent at the anterior end of Amoeba, required by the surface tension
theory, is very slight, but conceives it to be sufliicient to fulfill the
requirements of the theory.

MOVEMENTS OF UPPER AND LOWER SURFACES STUDIED EXPERIMEN-
TALLY ROLLING MOVEMENT IN AMCEBA.

Thus far we have left out of consideration the movement of substance
on the upper surface of the Amoeba. It is usually assumed that the
condition here is the same as at the sides and on the under surface ;
thus Rhumbler gives a diagram, reproduced in my Fig. 33, B^ showing
the backward current of the upper surface. The positive observations

THE MOVEMENTS AND REACTIONS OF AMCEBA.

139

on this point are those of Schulze (1875), Berthold (1886, p. 109), and
Biitschli (1892, p. 220). These authors all agree that the backward
current visible at the sides of the anterior end in Pelomyxa are clearly
also present on the upper surface. It is not usually possible to observe
particles moving backward on the upper surface of Amoeba, nor even
particles at rest, though this might be due to the fact that the granules
have sunken downward, leaving the upper surface clear. But to
decide whether the currents in Amoeba are essentially like those pro-
duced in a drop of fluid by a local change in surface tension, it is most
important to determine with certainty what is taking place on the upper
surface.

Evidently the most natural way of doing this is to cause, if possible,
some small object to rest upon or become attached to the upper surface
of Amoeba, then to observe the movement of this object. This can be
done by mingling a considerahjle quantity of soot with the water in
which the Amoebae are found. Some of the soot particles settle on the

Fig. 38.* Fig. 39. t

upper surface of the Amoebae, and in some species they adhere to this
surface.

I was quite unprepared for the results of this experiment. The
upper surface of Amoeba moves forward,, not backward, as required
by the surfiice tension theory ; nor is it at rest like the lower surface.

* Fig. 38, — Movements of a particle attached to the outer surface of Amoeba
verrucosa. When first seen the particle was at the posterior end (/) ; it then
moved forward, as shown by the arrows, until it passed around the anterior end
{a) to the under side. (The Amoeba itself of course moved forward at the same
time; no attempt is made to represent its movement in the figure.)

t Fig. 39. — Diagram of the movements of a particle attached to the outer
surface oi Amoeha verrucosa^ in relation to the movements of the animal. The
Amoeba is seen from above. In the position i the particle is at the anterior end
of tlie Amoeba. As the Amoeba moves forward, it passes over the particle,
which retains its place. Thus when the Amoeba has reached the position 2 the
particle is at the middle of its lower surface; when it reaches 3 the particle is at
its posterior end. The particle then passes upward and forward, as shown by
the arrows, so that when the Amoeba reaches the position 4 the particle is in
front of the middle, on the upper surface.

140 THE BEHAVIOR OF LOWER ORGANISMS.

AMCEBA VERRUCOSA AND ITS RELATIVES.

In giving an account of the experiments w^hich demonstrate this, I
shall begin with species of Amoeba in which pseudopodia are, as a rule,
not formed, and the movements are uniform in character, since here the
conditions are simplest from our present standpoint. For this purpose
Amoeba verrucosa Ehr., and particularly the transparent form known
as A. spkceronucleolus Greef, are favorable. In these Amoebae particles
cling rather easily to the outer surface.

When a quantity of soot is added to the water containing Amoebae
of the species named, small masses cling to the surface of the animal.
Such a mass, attached to the upper surface, shows the following
movements : It passes slowly forward (Fig. 38) , then over the anterior
edge, and under the latter. Here it stops, while the Amoeba continues
to move forward. The mass of soot remains quiet until the entire
Amoeba has passed over it and it lies beneath the posterior end. It
now passes upward again, to the upper surface (Figs. 39, 40), then
forward once more to the anterior end. Here it goes under the Amoeba

Fig. 40.*

as before, to be carried upward and forward again when the posterior
end passes over it.

These observations are made with absolute ease, and there is no pos-
sibility of mistaking internal particles for external ones. Particles lying
in the water outside the Amoeba may be seen to become attached at the
posterior end, to pass upward, lying distinctly outside the boundary of
the protoplasm (Fig. 38, posterior end), then forward, till as they double
the anterior end they are again seen sharply defined outside the boundary

♦ Fig. 40. — Diagram of the movements of a particle attached to the surface of
Amoeba verrucosa, in side view. In position i the particle is at the posterior
end; as the Amoeba progresses it moves forward, as shown at 2, and when the
Amoeba has reached the position 3 the particle is at its anterior edge, at x.
Here it is rolled under and remains in position, so that when the Amoeba has
reached the position 4 the particle is still at x, at the middle of its lower surface.
In the position 5 the particle is still in the same place, x, save that it is lifted
upward a little as the posterior end of the Amoeba becomes free from the sub-
stratum. Now as the Amoeba passes forward the particle is carried to the upper
surface, as shown at 6. (Thence it continues forward and again passes beneath
the Amoeba, etc.) The broken lines show that part of the surface of the Amoeba
which is at rest.

THE MOVEMENTS AND REACTIONS OF AMCEBA. I4I

(Fig. 38, anterior end) . Further, such particles, after making one or
two revolutions, usually become detached and drop off.

It is thus clear that Amoeba verrucosa and its relatives have what
may be called a rolling motion ; a given spot on the outer pellicula
passes forward on the upper surface, downward at the anterior end,
remains quiet on the lower surface, passes upward at the posterior
end, and again forward. Its movement may be compared directly with
the movement of a given point on the circumference of a wheel that is
rolling forward. A diagram of the movement of a particle on the
surface as it would appear in a side view is given in Fig. 40.

Certain details of the movements are interesting, and may best be
brought out by description of specific observations. In one case two
small particles had become attached, a short distance apart, to the surface
of a specimen of Amoeba sfhceronucleolus. They were at first side by
side and a little to the right of the middle line, one somewhat farther
to the right than the other (Fig. 41). They moved forward in parallel
courses, and reached the anterior edge at the
same time, passing over the edge and to the
under surface. It now required two and one-
half minutes for the Amoeba to pass over J)
them, during which time they remained nearly
or quite at rest. They then moved upward
to the upper surface and forward again. The
one nearer the middle line moved a little
faster than the other, reaching the anterior edge in two and three-
quarter minutes, while the lateral one required three minutes. Both
emerged at the posterior end again at the same time, the central
one having remained quiet three and one-fourth minutes, while the
lateral one had been three minutes at rest. The next forward course
required, respectively, but one and one-half and two minutes, the central
particle movmg the more rapidly. The two particles were no longer
side by side, the central one being now a little in advance. The latter
spent after the next turn two and one-half minutes on the under surface,
while the lateral particle spent but two minutes, so that they came up
from the posterior end again at the same time.

The two particles started forward again and had reached the middle
of the upper surface when the Amoeba ceased its forward movement,
loosened its anterior end from the bottom, and became attached by its
posterior end. After five minutes it began to move again, but now in

♦Fig. 41. — Paths of two particles attached to the outer surface of Amoeba
sphcsr 01111 cleolus as described in the text. That portion of the paths which is on
the lower surface is represented by broken lines. (No attempt is made to rep-
resent the forward movement of the Amoeba in this figure.)

142 THE BEHAVIOR OF LOWER ORGANISMS.

the opposite direction, so that the former posterior end became anterior.
At the same time the two particles reversed their former motion and
began to travel back in the direction from which they had come — that
is, toward the new anterior end. They were observed to make several
complete turns about the Amoeba while moving in this new direction.
I will not, however, add further details, as those above recounted are
sufficient to give a conception of the main features of the movement.

Thus two definite points on the surface of an Amoeba may retain
nearly the same relation to one another for five or six complete revolu-
tions, though their distance apart and their relative position may vary
a little. The reversal of the direction of rotation when the direction of
locomotion is reversed, described in the above case, I have seen many
times.

The direction of the movement of particles on the outer surface is
the same as that of the underlying particles of endosarc. The rate is
also about the same as for the endosarc, though often, or perhaps
usually, the outer particles move a little more slowly than those in the
endosarc.

It is not merely a thin outer layer that has the rolling movement.
This is demonstrated by the movements of bodies that are partly
embedded in the substance of the Amoeba. For example, a large
Euglena cyst had become attached to the hinder end of an Amoeba
sphceronucleolus. The cyst was carried upward and forward on the
upper surface, and at the same time it began to sink into the protoplasm,
so that when it had reached the anterior edge it was partially embedded.
It was then rolled under, remained at rest on the under surface in the
usual way, and came up at the posterior end. It was now deeply sunk
in the protoplasm, yet it moved forward in the usual way. By the
time it had reached the anterior edge again it no longer protruded
above the surface at all. After turning the anterior edge again it sank
completely into the body, still surrounded by a la3'er of ectosarc, so
that it passed to the interior of the AmcEba as a food body. I have
repeatedly seen bodies which were thus carried forward on the upper
surface gradually taken in as food. They always continue the forward
movement even when completely embedded in the ectosarc. It is thus
evident that the whole thickness of the ectosarc partakes of the forward
movement. The forward stream in ectosarc and endosarc are one and
continuous.

The relation of the movements of the outer layer to the lines and
wrinkles seen on the upper surface of Amoeba verrucosa and its rela-
tives is of interest. There are usually two sets of these wrinkles, one
set diverging from the posterior end toward the direction in which the
animal is moving, the other set forming a number of curved lines

THE MOVEMENTS AND REACTIONS OF AMOEBA. I43

parallel to the advancing edge (Fig. 38). These wrinkles and the
areas which they enclose do not change markedly as the Amoeba
advances, so that the outer surface of the body seems to be quite at
rest. It is this fact, I believe, that has prevented the true nature of
the movement in these species from being recognized before. Thus
Penard (1902, p. 118), after a thorough study, accurate so far as it
goes, of the movements of Amoeba verrucosa^ notes that many facts
point to the existence of a permanent contractile outer layer, but holds
that the permanence of certain lines and patterns on the upper surface
in a moving Amoeba is crucial against the idea of a rolling movement
such as I have shown above to actually occur. In reality these
wrinkles are not static structures, but dynamic, /. ^., the substance of
which they are formed is in continual motion ; they are like the per-
manent ripples on the surface of a stream where the latter crosses an
obstruction. The wrinkles indicate the direction of movement of the
substance, the longitudinal wrinkles being parallel to the lines of motion,
the others transverse to them. A particle on the upper surface may
move parallel with the longitudinal wrinkles, at a constant distance
from them, or it may move directly along one of these wrinkles, for
the whole length of the latter. On coming to one of the transverse
wrinkles the particle moves over it with a sort of jerk, as if it had
passed over a ridge or step, as indeed it has. Thus the lines and the
areas enclosed by them remain constant, while the substance of which
they are composed moves onward.

When the Amoeba changes in a marked degree its direction of
movement, so as to follow, for example, a course at right angles to the
previous one, the wrinkles on the surface usually slowly disappear, then
after the movement has become well established in the new direction,
new wrinkles appear in correspondence with the movement.

When such a change of course occurs, any particles on the upper
surface, which were moving toward the anterior edge, change their
course in correspondence with the new direction of progression. Fig.
42 represents a case of this kind, where an Amoeba verrucosa bore on
its upper surface a minute particle of debris {a) and a spherical cyst of
Euglena {b). Both moved forward over the stretch x-y (Fig. 42, A),
Now a little methyl green (m) was allowed to diffuse against the left
side of the Amoeba. The animal changed its course, moving to the
right At the same time the two objects a and b changed their direction
of movement, traversing the stretch ^-^r (Fig. 42, B) until they reached
the new anterior edge of the Amoeba, and were carried underneath.

The free-moving (upper) surface and the resting (lower) one in
contact with the substratum may exchange roles at any time when
the contact with the substratum is changed. Thus, a specimen was

144

THE BEHAVIOR OF LOWER ORGANISMS.

creeping on the slide and bearing on its upper surface a small granule,
which was moving forward in the usual way. The Amoeba stopped
and raised its anterior edge, which came in contact with the cover
glass ; it then loosened itself entirely from the slide, while its upper
surface became attached to the cover. It now began to move forward
on the cover glass. The granule on the upper surface now remained
quiet, until it was reached by the posterior end, when 'it passed down-
ward to the lower free surface, there moving forward in the usual way.
Upper and lower surfaces had completely exchanged roles. In a sim-
ilar way I have seen the thin lateral edge of a specimen become the
middle of the upper moving surface.

Objects of the most varied sort cling to the surface of Amoeba verru-
cosa and its relatives. I have seen the following attached to the surface
and showing the typical movements : Particles and masses of soot,
granules of India ink, motionless bacteria, diatom shells, dead flagel-
lates, masses of debris,
cysts of Euglena, a small
Amoeba proteus (the lat-
ter was inclosed after it
had passed to the under
surface). Usually only
one or two small objects
are seen attached to any
given specimen, but to
this extent the phenomenon is very common, so that it seems rather
surprising that the movements of such particles should not have been
described before.

A number of other points must be set forth before we can form a
clear conception of the movements of these Amoebae. The species
under consideration are much flattened and have usually an oval form
as they move forward, the anterior-posterior axis being the longer,
while the posterior end is the more pointed (Fig. 38). Not the whole
lower surface is in contact with the substratum, but only a band at the
anterior and lateral margins. In an Amoeba that was creeping on the

Fig. 42.*

* Fig. 42. — Movement of bodies attached to the surface in Amoeba verrucosa,
when the direction of locomotion is changed, a, P^ small granule ; b, a Euglena
cyst. In A the Amoeba is progressing to the right, as shown bj the large arrow ;
the two bodies attached to the surface moved in the same direction, traversing
the stretch x-y, as shown bj the small arrows. At this point a solution of
methyl green (;«) was allowed to diffuse against the surface ; the Amoeba there-
upon changed its course, as indicated by the large arrow of ^. At the same
time the bodies a and b changed their course, traversing the stretch y-z. The
stretch x-y-z in B shows thus the path of the attached bodies before and after
the reaction.

THE MOVEMENTS AND REACTIONS OF AMCEBA. I45

lower surface of the cover glass I was able to define with some accuracy
the parts that were attached and those that were not. A small flagel-
late was moving briskly about between the Amoeba and the cover glass,
but its excursions were limited by a visible line running parallel with
the anterior edge of the Amoeba and extending at the sides back to
about one-third the animal's length from the rear (Fig. 43, a-a-a).
The zone between this and the margin was pressed close to the glass,
and was evidently attached to it The more pointed posterior end was
held quite away from the glass, leaving a broad passageway through
which the flagellate finally escaped.

The results of this observation were confirmed by another. An
Amoeba verrucosa in full career was suddenly turned on one lateral
edge by a strong current from a rotifer, and its upper edge coming in
contact with the cover, glass, it remained in that position sometime
without change of form. It could be seen that the under surface was
concave, the edges very thin and flat, while the
posterior portion was thick and arched (Fig. 44).

It is clearly at the advancing edge of the ani-
mal that the most active movements are taking
place. Here the hyaloplasm may be seen to
push forward in a series of short waves, the
anterior edge of each becoming attached to the
substratum. At the same time, of course, an
equivalent amount of protoplasm becomes de-
tached from the substratum along the line a-a-a^
Fig. 43, though this does not take place in
waves, so far as observable. The anterior wave must in some way
pull upon the upper surface of the Amoeba, bringing it forward, and
dragging with it the elevated sac-like posterior end. A certain feature
of the advance of the anterior edge seems of much significance. Each
wave seems to arise just behind the previous anterior boundary line
and overlaps it, leaving it buried. This line often remains visible for
a short time after the new wave has been formed. The new wave
rolls over the preceding one in such a way that its original upper
surface becomes applied to the substratum. This is demonstrated by
the rolling under of small objects on the upper surface of the advanc-
ing wave. A diagram of the movement at the anterior edge is given
in Fig. 45. The movement can be imitated roughly by making a
cylinder of cloth, laying it flat on a plane surface, and pulling forward

* Fig. 43. — Attached surface oiAmceba verrucosa, creeping on the lower surface
of the cover glass. The unshaded portion in front of the line a-a-a is attached
to the substratum, while the shaded portion is free and raised slightly above
the substratum.

146 THE BEHAVIOR OF LOWER ORGANISMS.

the anterior edge in a series of waves. The entire cylinder then rolls
forward just as the Amoeba does.

The essential features of the movement seem to be (i) the advance of
the wave from the upper surface at the anterior edge ; (2) the pull exer-
cised by this wave on the remainder of the upper surface of the body,
bringing it forward. Most of the other phenomena follow as conse-
quences of these two. The flowing forward of the granules of the
endosarc seems to demand no special explanation, since a fluid con-
taining granules within a rolling sac must necessarily flow forward as
the sac rolls. By the movement forward of the anterior end a space is
left free ; by the rolling forward of the posterior end the fluid is piled
up and pressed upon, and must flow forward into the empty space in
front. Possibly there may be other causes at work in producing the
endosarcal currents, but such currents would be produced without
other cause in a sac moving as Amoeba does.

^

\

a

h

Fig. 44.* Fig. 45.!

other species of amoeba.

Thus far we have dealt only with Amoebae of rather constant form,
which do not produce pseudopodia, or only rarely do so. We must now
take up species in which the form is changeable and the movements
varied. Of such species I have studied chiefly Amoeba limax^ A.
proteus^ and a smaller Amoeba, which I take to be Amoeba angulata
Meresch. In these species the outer surface is not viscid, except at the
posterior end, so that small objects rarely cling to it. It is, therefore,
much more difficult to determine the direction of movement of the
upper surface than in Amoeba verrucosa and its relatives. Yet, by
mixing soot with the water, and devoting a sufficient amount of time
and patience to the work, one can obtain as many observations as he
desires. The soot settles upon the upper surface in particles or masses

♦Fig. 44. — Side view (partly an optical section) of a creeping Amosba verru-
cosa^ showing the thin anterior edge (/I) attached to the substratum, and the
high posterior portion (/*) with a cavity beneath it.

t Fig. 45. — Diagram of the movement at the anterior edge of Amoeba verrucosa.
The region h-c pushes out, taking up the position b'-c\ and pulling forward the
region c-</, so that it comes to occupy the position c'-d' . The point a remains
in its place.

THE MOVEMENTS AND REACTIONS OF AMOEBA. I47

and its movements can be followed ; at times, also, objects actually
cling to the surface, as in the other species.

The results are essentially the same as in the species already described ;
foreign particles resting upon or clinging to the upper surface are car-
ried forward to the anterior edge. Here they roll over the edge, passing
beneath the Amoeba, which now moves across them. As a rule in
these species particles do not cling to the surface after passing to the
lower side, so that they are left behind when the posterior end passes
over them. Sometimes they do thus cling, however, and in such cases
I have seen them pass upward at the posterior end and again forward,
exactly as in A. verrucosa and its relatives. In order that my state-
ments may not remain abstract and general, I copy a few observations
from my notebook, all relating to Amoeba proteus.

1. A large particle of debris with bits of soot attached to it was seen
lying on the upper surface just behind the middle. It was carried
forward to the anterior end and over the edge. Then it came to rest
on the bottom, and the Amoeba crept over it till it was passed by the
posterior end and left behind.

2. A number of soot particles on the upper surface just in front of
the middle were carried forward, changing their direction as the proto-
plasmic currents beneath them changed direction. They were finally
carried over the anterior edge.

3. A small mass of soot was lying on the middle of the upper surface.
It moved forward in the same way as the endosarcal granules under-
neath. The latter changed their direction of movement several times ;
the soot mass changed correspondingly at the same time. It was
finally carried over the anterior edge, where it could be seen clearly
separate from the Amoeba.

4. A large mass of soot one-quarter the size of the Amoeba was
carried forward on the upper surface for a distance, but fell off' at the
side before reaching the anterior end.

5. Two small masses of soot lying on the upper surface of the
posterior end were carried forward over the anterior edge.

6. Several small particles were clinging to the lower surface of the
posterior end. They passed upward, one of them around the very
middle of the posterior end, to the upper surface ; here they were
carried forward and over the anterior edge.

I could add a large number of such observations.

On the under surface the particles are quiet, as I have shown before
(p. 136). At the lateral margins the edges of this quiet lower surface
are seen, so that particles situated here are usually likewise quiet, until
they have reached the posterior part of the Amoeba (see p. 135).

As to the details of the movement of the upper surface, the following

I4S THE BEHAVIOR OF LOWER ORGANISMS.

are important. Particles situated on the upper surface move usually
at the same, or nearly the same, rate as the granules beneath them, in
the endosarc. The movement of the surface particles follows exactly
that of the endosarc beneath them, changing in direction when the latter
changes. Two particles close together on the upper surface may thus
diverge or even flow in opposite directions, carried by two different
currents which are visible in the endosarc. Any portion of the ecto-
sarc, like any portion of the interior, may stop at any time, while other
parts flow onward. One may thus see at times a particle at rest on the
upper surface of a moving Amoeba. Isolated observations of this kind
might lead one to suppose that the upper surface, like the lower, remains
at rest while the endosarc passes forward. But when a particle on the
surface is at rest, one will usually find, by a proper change of focus,
that the endosarc beneath it is likewise at rest. It is, of course, well
known that certain portions of the endosarc may be at rest while the
remainder is in movement (see Rhumbler, 1898, p. 122). In the same
way a portion of the outer layer may sometimes be at rest while the
adjacent endosarc is in motion ; this, however, is rather unusual.

We may sum up our results thus far in the following statements :
In an advancing Amoeba substance Jloivs forward on the upper sur-
face^ rolls over at the anterior edge^ coming in contact with the
substratum^ then remains quiet until the body of the Amoeba has
passed over it. It then moves upward at the posterior end., and
forward again on the upper surface., continuing in rotation as long
as the Amoeba continues to progress. The motion of the upper
surface is congruent with that of the endosarc, the two forming a
single stream.

HISTORICAL, ON ROLLING MOVEMENTS IN AMOEBA.

The possibility that Amoeba progresses by a rolling movement was
discussed by Claparede & Lachmann (1858). In Amoeba Umax and
Amoeba quadrilineata (== A. verrucosa) ., according to these authors,
the general appearance of locomotion is in many respects in favor of
this view : " On croit positivement voir Tanimal rouler sur lui-meme "
(p. 435). But this correct view is rejected (in the text) because of the
(supposed) permanence in the position of the contractile vacuole.
Claparede & Lachmann insist that the contractile vacuole is situated
in the ectosarc ; hence, they argue, if there were a rolling movement
of the ectosarc, the vacuole would necessarily partake of the movement !
In a note on p. 437 it is stated, however, that Lachmann personally
believed the motion to be of this rolling character. " II croit s'etre
assur^ que V A. quadrilineata roule sur elle-meme." According to
Claparede & Lachmann, Perty held this view also.

THE MOVEMENTS AND REACTIONS Oi^ AMCEBA. t^g

Dr. Wallich shared the correct opinion of Lachmann and Party.
This excellent observer unfortunately often gave his results in the form
of mere brief general statements, so that one cannot judge how^ much
evidence he had for them, and little attention has, therefore, been paid
them. But it is singular how many of these statements show them-
selves to be correct, even in opposition to later work. Concerning
the matter in question, Wallich has the following:

In short the effect is similar to that which would be produced were an empty
and transparent bladder or caoutchouc sac, containing granular bodies of greater
specific gravity than the viscid fluid within which they were sustained, to be
rolled along a plain surface. (Wallich, 1863, b, p. 331.)

He makes no attempt to' demonstrate the truth of this correct com-
parison, and does not develop the matter beyond the mere statement
given above.

Schulze (1S75), Berthold (1886), and Biitschli (1892), as we have
seen, agree in stating that the currents on the upper surface at the
anterior end in Pelomyxa are backward. In view of the great authority
of these writers we should be compelled to suppose that the movement
in this animal is of an entirely different character from that found in
the various species of Amoeba, but for a most fortunate circumstance.
The only previous demonstrated observation of the forward movement
of the upper surface in the Rhizopoda relates precisely to Pelomyxa,
and was made by an investigator closely associated with Biitschli.
It was not until my work was finished and the present paper written
that I came across the note of Blochmann (1S94) on the movements of
Pelomyxa. Blochmann shows that the movement of substance on the
upper surface of Pelomyxa is forward, just as we have found it to be
in Amoeba. The outer surface of Pelomyxa is covered with fine cilia-
like projections. By observing these projections Blochmann had no
difficulty in seeing that they move forward on the upper surface. The
rate of movement was the same as that of the internal forward current.
Biitschli (1893, Appendix, p. 220) had already observed, greatly to his
surprise, that there is a forward current in the water next to the sur-
face of an advancing Pelomyxa, this current being exactly the reverse
of that called for by the theory that the motion is due to a lowering of
the surface tension at the anterior end.

Biitschli (/. c.) attempted to save the surface tension theory by sug-
gesting that it was only a thin outer layer that moves forward. The
currents in the moving animal would then be as follows : A forward
current within, a backward current just beneath the surface, a forward
current in a thin layer on the surface. It is possible that this compli-
cated arrangement of currents might be brought into harmony in some
way with the surface-tension theory of the movement, though it is

150 THE BEHAVIOR OF LOWER ORGANISMS.

rather difficult to see how the forward movement of the outer layer
would be produced. As we have seen, Blochmann found that the
internal and external forward currents move at the same rate and in
the same direction. It is difficult to explain how this should occur if
the two are separated by a layer moving in the opposite direction.
But Blochmann accepted Biitschli's suggestion, and attempted to give
some evidence in its favor. He says that one sees the outer current,
with the movement of the projections, in places where the marginal
current has come to rest, and that the outer and internal currents then
move at the same rate, separated by the resting marginal layer. Now
one can receive exactly this impression in Amoeba in the following
manner : The margins where they are pressed against the substratum
are at rest. Just above this region, and often visible in the same focus,
there is the forward current, which is visible on the one hand on the
surface (through the movements of the projections) ; on the other hand,
in the interior (through the movements of the granules of the endosarc).
Unless one is on his guard as to the slight difference in level, one might
seem to see two currents separated by a resting layer, particularly if
the probability that this were true had been suggested beforehand. It
is notable that Blochmann describes nowhere outer and inner forward
currents, separated by a marginal backward current, as would be
required by the modified surface-tension theory.

We have demonstrated above, for Amoeba at least, that the forward
movement is not confined to a thin outer layer, but extends from the
outer surface to the endosarc (p. 142) ; in other words, that the outer
surface moves in continuity with the internal substance.

Rhumbler (1898, pp. 126-130) discussed at length the possibility of
explaining the movement of Amoeba by means of a rolling sac of
ectoplasm, only to come to the conclusion that it was impossible.
Rhumbler's discussion of this matter is an excellent example of the
fact that acumen and excellent reasoning may lead one astray in scien-
tific matters when the observational basis for the reasoning is not
secure. What chiefly misled him was an incorrect idea as to the
direction of the currents in the substance of Amoeba, particularly his
assumption that there is a backward current on the upper surface.
The diagram which he gives of the currents as they must occur in an
Amoeba moving in a rolling manner (Fig. 33, A, p. 133) is, therefore,
much more nearly correct than that in which he shows what he con-
siders the really existing currents (Fig. 33, jS),

Rhumbler's conception as to the necessary movements in the sub-
stance of an Amoeba progressing by rotation is, however, incorrect in
one particular, so that his diagram (Fig. 33, A) does not correspond
to the facts in this point. He assumes that there must be a backward

THE MOVEMENTS AND REACTIONS OF AMCEBA. 15I

current on the lower surface, as indicated by the lower (heavier) arrows
in his diagram (Fig. 33, A). This backward current does not exist,
and is theoretically unnecessary, as may be seen by making a cylinder
of cloth and moving it in the manner described above (p. 145). The
under surface remains at rest until it passes upward at the posterior
end {cf. Fig. 40) . Rhumbler held that this backward current below,
with the forward current above (Fig. 33, A), must set the endosarc
in rotation ; " the endoplasma granules would themselves necessarily
all move, like the particles of the ectosarc, in circular or elliptical
courses" (p. 128). The absence of such circular or elliptical paths for
the granules of the endosarc would then speak against the method of
movement by rotation of the ectoplasm. But since there is no such
backward current as Rhumbler assumes, and not even the particles of
the ectosarc move in circular or elliptical courses, this objection falls
to the ground.

Further, Rhumbler seems to assume that for locomotion by a rotary
movement of the ectosarc, the latter must necessarily be a "sharply
defined persistent organ," and that its contractions could only be due
to preformed, permanent fibers, in a definite arrangement. Rhumbler
is able to show of course that these two assumptions are probably incor-
rect, and considers that this weighs against the possibility of movement
in the manner characterized. But both these assumptions are unneces-
sary. The rotation demonstrably does occur, yet the permanent, sharply
defined ectosarc with definitely arranged persistent fibers does not exist,
as Rhumbler has set forth, and as must be evident to anyone who
studies for a long time the changes of form and movement in Amoeba.
As we shall see later, a simple drop of fluid, with no differentiated outer
layer, may move in the same manner.

Penard (1902) also discusses the possibility of movement by rotation
of the ectosarc in Amceda verrucosa (== A. terricold) . His study of
the movements is excellent and he gives as a possibility on p. 115 what
is really in its main features a nearly accurate statement of the method
in which locomotion actually occurs, only to reject this possibility later.
The ground for this rejection is as follows : The posterior end of the
Amoeba often bears an irregular saclike projection (what Penard calls
the " houppe" ) ; this may be much wrinkled or covered with projec-
tions. This wrinkled sac retains its position ; in Afnoeba verrucosa it
is covered, like the rest of the body, with a resistant cuticula, which can
be dissolved only with great difficulty and very slowly.

If the Amoeba rolled on itself in progressing, the posterior part of this mem-
brane would necessarily follow the movement and pass little by little forward,
which is contrary to the facts. The best manner of assuring one's self of the
immobility of the pellicle is to look very attentively at the surface of the con-

152

THE BEHAVIOR OF LOWER ORGANISMS.

tractile vacuole; there one sees almost always very fine folds, forming angles
and varied patterns; these angles and these patterns remain for a long time
absolutely the same, which shows that nothing has changed place. (Penard,
1902, p. 118.)

In all the specimens of Amceda verrucosa and A. sphceronucleolus
in which I have studied the matter, the posterior part of the outer mem-
brane does follow the movement. Particles clins^ing to the outer surface
of the hinder part of the ectosarc pass upward over the wrinkled sac-
like posterior end and forward on the upper surface. In so doing they
pass directly across the wrinkles on the body sur-
face, as set forth on p. 143. Had Penard chanced to
^ - ^ see the movements of a particle attached to the outer

surface of the body he could not have been misled
by the apparent permanence of the surface wrinkles .

FORMATION AND RETRACTION OF PSEUDOPODIA.

Thus far the phenomena in Afrioeba froteus
and its relatives are essentially like those found
in Amceba verrucosa. At times Amoeba proteus
flows forward as a single simple mass ; then its
locomotion may be compared directly in its chief
features to that of Amoeba verrucosa. But in
Amoeba proteus and its relatives the movement
is, of course, usually much complicated by the for-
mation of pseudopodia. In considering the way
n which these are formed we must deal separately
with two different cases, depending on whether the
pseudopodium when sent out is or is not in contact
with the substratum.

SURFACE CURRENTS IN THE FORMATION OF PSEUDOPODIA
IN CONTACT Vi^ITH THE SUBSTRATUM.

Fig. 46.* When the pseudopodium is sent out in contact

with the substratum, the phenomena accompany-
ing its formation are essentially the same as those which take place at
the anterior end of an advancing Amoeba ; the latter may indeed be
considered as merely a large pseudopodium. Even when the pseudo-

* Fig. 46. — Movement of a particle attached to the outer surface of a pseudo-
podium that is extending in contact with the substratum. At a the particle is
at the middle of the upper surface; at b it has nearly reached the tip. When
the pseudopodium has reached the length shown at c the particle has passed
over its tip. Here it remains, so that at d, when the pseudopodium has become
longer, the particle is still at the same level, but on the under surface of the
pseudopodium, some distance behind the tip.

THE MOVEMENTS AND REACTIONS OF AMCEBA.

'53

podia are slender and pointed, the protoplasm flows forward (toward
the point) on the free upper surface and in the interior, while on the
side which is in contact with the substratum the protoplasm is at rest.
I have often seen small particles which had been brought forward on
the surface of the Amoeba carried out to the tip of a pseudopodium on
its upper surface, finally rolling over the point and becoming covered
by the advancing protoplasm (Fig. 46) .

FORMATION OF FREE PSEUDOPODIA.

When a pseudopodium is sent out directly into the water, so that its
surface is free on all sides, "it is much more difficult to determine the
nature of the movement. Particles rarely cling to the surface of such
a pseudopodium, and without this aid one cannot be certain what the
movement of the outer layer is. However, by devoting several entire
days under most favorable conditions to the determination of this point,
I collected a number of observations which demonstrate clearly the
nature of the move-
ment. The point \ a ^ ^
of special interest
is, here as else-
whe re , whether
there is a back-
ward current on
the surface of the
advancing pseudo-
podium, as repre-
sented in the dia-

FiG. 47.

gram from Rhumbler, Fig. 31. To this the observations give a negative
answer. Particles clinging to the surface of a pseudopodium, whether
at the tip or at the base or at any intermediate point, are uniformly
carried outward, in the same direction as the tip. Particles situated at
a certain distance from the tip of a short pseudopodium maintain the
same distance as a rule when the pseudopodium is lengthened, though
in so doing they are carried far out from the body. Sometimes the tip
moves outward a little faster than the parts behind it, the pseudopodium
thus becoming more slender as it extends, but all parts agree in being
carried outward. A number of examples of actual observations will
make this point clear.

I. Amoeba angulata : When first observed there was a short pseudo-
podium in front, projecting freely into the water. A small particle was
attached to the surface at about the middle of its length (Fig. 47, a),

* Fig. 47. — Successive stages in the formation of a free pseudopodium, showing
the movement of a particle attached to its surface. The particle moves outward,
keeping at approximately the same distance from the tip.

154 THE BEHAVIOR OF LOWER ORGANISMS.

The pseudopodium lengthened, carrying the particle with it, the latter
maintaining its distance from the tip nearly or quite constant, but being
carried far from the body (Fig. 47, b). The pseudopodium finally became
very long and slender (c), the particle remaining attached near the tip.

2. Afiiceba proteus : A particle clinging to the surface of one side,
near the anterior end. A pseudopodium was formed at exactly this
point, extending freely into the water, so that the particle was borne
on the tip of the pseudopodium. It maintained this position while the
pseudopodium was extending, and was still found at the tip after the
pseudopodium had become long and slender (Fig. 48).

The third example which I give is one of much interest, because it
shows the movements of a given point on the surface in both the retrac-
tion and extension of pseudopodia, as well as in transference from the
posterior to the anterior region of the body.

3. A^nceba proteus: When first observed the animal was rather

slender, creeping in a certain direc-
tion, and with two long pseudo-
podia at the posterior end, extend-
ing, one on each side, at right
angles to the axis of progression
(Fig. 49, a). The left pseudo-
podium was the longer, and bore
at about one-fourth its length from
its base a small particle {x) at-

Pjq .3 * tached to its surface by a very

short stalk in such a way that it
was seen in profile (Fig. 49, a). The pseudopodium was not in
contact with the bottom, and was slowh^ retracting, its internal con-
tents flowing into the body, while the pseudopodium itself shortened.
As this occurred the particle approached the body and finally passed
on to its surface {b, c) while the pseudopodium was yet of considerable
length. It was evident that the shortening of the pseudopodium took
place chiefly at its base, since the part between the base and the particle
X had become incorporated with the body when the portion between x
and the tip had changed only a little in length {b, c) . This distal
portion apparently did become somewhat shorter at the same time,
while its surface became slightly wrinkled. By the time the tip of the
pseudopodium had united with the body {d) the particle x had moved
a considerable distance forward on the latter. The posterior portion
of the body was here thick, and the particle was still seen in profile,
though it was some distance above the substratum. It now moved

♦ Fig. 48. — Movement of a particle attached to surface of an Amoeba at point
where a free pseudopodium is pushed forth. The particle remains at the tip.

THE MOVEMENTS AND REACTIONS OF AMCEBA.

155

forward very slowly (c, d^ e) till at f it passed to the upper surface.
It then moved rapidly forward, occupying successively the positions
indicated by the line of circles in g: (The Amoeba itself was, of course,
progressing ; no attempt is made in the diagram to represent its change
of position.) Finally the particle x had nearly reached the anterior
end, when the latter forked, sending two pseudopodia upward and
forward into the water (^, h). The particle x was at first at the base
of the right pseudopodium. This was now projected forward as a very
long, slender pseudopodium bearing the particle x. The latter was
carried steadily out from the' body, maintaining almost exactly its
original distance from the tip of the pseudopodium {h^ i^j)- It is

I

Fig. 49*

possible that as the tip became very slender its distance from x became
slightly greater as if, by a circular contraction of the intervening part,
the tip were forced further out ; but there was no movement backward
of :v; on the contrary, it moved steadily forward, its distance from the
base of the pseudopodium continually increasing. Unfortunately at
this point the animal passed under a mass of debris, so that I was
unable to trace further the history of that point on the body surface
marked by the particle x.

I have, altogether, about a dozen observations showing this outward
movement of particles on the surface of free pseudopodia. The three

* Fig 49. — Movements of a particle (») attached to the surface of Amoeba in
passing from a pseudopodium at the posterior end over the body to a pseudopo-
dium at the anterior end. For explanation see text.

156 THE BEHAVIOR OF LOWER ORGANISMS.

examples above given are typical for all. They show the following as
to the manner in which the pseudopodia are formed when they are
projected freely into the water.

1. The pseudopodium grows in length chiefly from the base, so that
any part on the surface retains nearly its original distance from the tip.

2. The increase in surface as the pseudopodium grows is not pro-
duced by the flowing outward and backward of the endosarc at the tip
with its transformation into ectosarc (as represented by Fig. 31), but
by the transference of a portion of the surface layer of the body to the
pseudopodium. The same substance remains at the tip of the pseu-
dopodium from the beginning (observation 2, p. 154 ; I have other
observations showing the same thing).

3. Thus the movement of the free pseudopodium is like that of the
pseudopodium in contact with a surface, save that in the latter case one
side is held back by attachment to the substratum. In the free pseudo-
podium all sides move outward ; in the attached one, all sides but one.

The outer layer of the body in its transference to the pseudopodium
may doubtless become thicker or thinner or be otherwise modified. As
will be shown later, I am not at all inclined to deny the possibility of
the transformation of endosarc into ectosarc, and vice versa. The
observations show, however, that this transformation of substance does
not, as a rule, take place in pseudopodia by means of the "fountain
currents" represented in the diagrams from Rhumbler (Figs. 30-32).

Further, the surface of the pseudopodium may be increased by the
flowing into it of the endosarc, producing a sort of stretching of the
outer layer, involving, of course, the appearance at the surface of por-
tions of substance which were before covered.

WITHDRAWAL OF PSEUDOPODIA.

In the withdrawal of pseudopodia the process is the reverse of that
occurring in the formation of pseudopodia, as is shown in case 3, above
(p. 154, Fig. 49). The basal parts of the pseudopodial surface first
pass on to the body, followed by the distal portions.

The withdrawal of pseudopodia shows certain other features that are
of importance for the understanding of the mechanism of amoeboid
movement. The process differs somewhat in different cases, depending
on whether the withdrawal is slow or rapid. When the pseudopodium is
slowly withdiawn, its surface may remain perfectly smooth, the decrease
in surface keeping pace with the decrease in volume, until the pseudo-
podium has quite disappeared. But when the withdrawal is more rapid
the surface becomes thrown into folds or warty prominences of various
sorts. This is more common than retraction without the formation of
such prominences. Evidently the volume decreases so fast that the

THE MOVEMENTS AND REACTIONS OF AMCEBA. 1 57

decrease in surface cannot keep pace with it^ so that the surface is
thrown into folds. This phenomenon is particularly interesting in its
bearing on the theory that would account for the retraction of pseudo-
podia by the action of surface tension. On this theory we should
naturally expect the surface to remain smooth, and by no means to be
thrown into folds, since it is by the tendency of the surface to decrease
that the decrease in volume is accounted for ; the decrease in volume
should not, therefore, precede the decrease in surface. This matter will
be taken up later.

As the pseudopodium decreases in size, the fluid endosarc, of course,
flows out of it and joins the endosarc of the body. The backward
current begins at the mouth or inner end of the pseudopodium, and
gradually extends backward to near the tip ; the current is most rapid
in the central axis of the pseudopodium, and in this axis it is most
rapid at the inner end.*

Where is the impelling force in the outflow of the endosarc and the
decrease in size of the pseudopodium } The observations seem to sug-
gest several factors here. The fact that the ectosarc of the pseudo-
podium passes on to the body when the pseudopodium shortens, as is
shown in Fig. 49, a, b^ c, indicates that the ectosarc of the body exercises
a pull on the outer layer of the pseudopodium, drawing it inward.
This would, of course, force the fluid endosarc into the body. But this
would not account for the wrinkling and roughening of the outer surface
of the pseudopodium, which is so prominent a feature in the withdrawal.
For this there are two conceivable causes, (i) The ectosarc itself may
contract actively, driving out the endosarc. If the real contractile por-
tion of the ectosarc is not on the outer surface (in the cuticula, as it has
sometimes been called), but in a deeper layer, then the outer surface
would be thrown into folds or prominences as contraction occurs. (2) On
the other hand, it is conceivable that the endorsarc might be drawn out
of the pseudopodium, the latter collapsing and becoming wrinkled as
a result. This is the explanation given by Biitschli (1893, p. 201).
This view would have to assume some force pulling on the endosarc at
the mouth of the pseudopodium, and sufficient viscosity in the endosarc
so that a pull thus exercised would draw out the whole mass contained
within the pseudopodium. Thus, in Fig. 49, a, the general advancing
current within the body of the Amoeba might be thought to exercise a
pull at the point j^ in the direction of the arrow ; if the endosarc were

♦This account differs from that given by Biitschli (1880, p, 116), according to
whom the withdrawal of the pseudopodium begins at the tip. The observations
present no difficulty, and I am unable to understand how Biitschli came to this
result. In a large pseudopodium the method of retraction described above is
evident.

158

THE BEHAVIOR OF LOWER ORGANISMS.

sufficiently viscous the entire mass of endosarc would be withdrawn,
and the pseudopodium would collapse.

There are certain facts that speak against this second view. Thus,
the endosarc often passes out when there is no current away from the
mouth of the pseudopodium, so that there can be nothing pulling upon
the endosarc. A pseudopodium may be withdrawn when the animal
is otherwise quiet ; or, when the animal is stimulated strongly, all the
pseudopodia may be withdrawn at the same time, while there is no
endosarcal current in the body of the animal. A striking case that
belongs here is sometimes to be observed in Amceba radiosa. This
animal frequently floats in the water, with many long, pointed pseudopo-
dia radiating in all directions from the body. Now, if the pseudopodia

are stimulated with a rod,
they begin to contract. The
endosarc first passes inward,
but the resistance of the body
is so great that the fluid stops
at the base of the pseudo-
podia. These, therefore,
swell up in a bulbous
fashion, as illustrated in
Fig. 50. Such cases, indeed
all the numerous cases in
which the endosarc passes
out of a pseudopodium and
comes to rest as soon as it has
left the latter, can only be ex-
plained on the assumption
that the endosarc is forced
out by the contraction of the ectosarc, or by some active movements of
the endosarc itself, of a character not understood.

Further, there are certain facts which speak positively in favor of
the view that the production of the wrinkles is due to a contraction of the
inner layer of the ectosarc. Thus, when an Amoeba is strongly stimu-
lated and withdraws all its pseudopodia quickly, the whole surface of
the body becomes rough and wrinkled. The endosarc has not passed
out of it, so that it cannot be considered in a state of collapse ; on the con-
trary, it is clearly contracted as strongly as possible. Again, if a large
pseudopodium is cut from the body, it contracts strongly, showing the
rough, wrinkled contour, though the endosarc has not passed out of it.

Fig. 50.*

♦Fig. 50. — Specimen oi Amoeba radiosa in which the endosarc has passed out
of the distal portions of the pseudopodia into the basal parts, causing them to
swell up in .1 bulbous fashion.

THE MOVEMENTS AND REACTIONS OF AMCEBA.

159

The processes occurring in a retracting pseudopodium are the same
as those taking place at the posterior end in a moving animal. In fact,
the cases are really identical ; the posterior portion of the body may be
considered merely a large retracting pseudopodium. Often, as we shall
see, it is impossible, from its method of formation, to consider it any-
thing else.

The pseudopodia are, of course, usually formed in the anterior part
of the Amoeba, in contact with the substratum, and are directed in the
line of progression, or
at a slight angle with
the line of progression.
As they are withdrawn,
their surface, as we have
seen, usually becomes
folded or roughened,
and the small roughened
projection resulting
from the withdrawal
lasts, as a rule, for a long
time. As the Amoeba
continues its forward
course, the base of the
retracting pseudopo-
dium retains nearly its
original position, as do
the other parts of that
layer of the ectosarc that
is against the substra-
tum, so that the body
of the Amoeba gradually
passes the pseudopo-
dium, and the latter
finally becomes united with the posterior end. During this transference to
the rear, the pseudopodium usually changes its position (Fig. 51 ). At
first it is directed nearly forward («), then it takes a position at right
angles to the body (^), and finally swings around with its point directed

Fig. 5

♦Fig. 51. — Successive stages in the retraction of a pseudopodium. At a the
pseudopodium extends forward at the anterior edge ; at 3 it has partly withdrawn
and stands at right angles to the body, which has partly passed it; at c the
pseudopodium is directed backward, and is in the posterior part of the body; at
</the small roughened remnant of the pseudopodium has nearly united with the
tail. At X the successive positions occupied by the withdrawing pseudopodium
are shown in a single diagram.

|6o THE BEHAVIOR OF LOWER ORGANISMS.

nearly backward (Fig. 51, c). This change of position is due to the
contraction of the posterior part of the Amoeba. The ectosarc just
behind the base of the pseudopodium contracts toward the middle, as
described on page 171. As a result the pseudopodium must swing
around till it points nearly backward. The mechanism of the process
will be best understood by an examination of Fig. 51, at.

Finally, what remains of the pseudopodium reaches the posterior
end or tail (Fig. 51, d). By this time usually all that is left of it is a
small roughened projection, its surface being of essentially the same
character as that of the tail. This projection fuses completely with the
tail, its projections taking up a portion of the surface of the latter. The
tail is in fact nothing but the fused remnants of all the pseudopodia
that have been formed, together with the contracted outer layer of the
body of the Amoeba (the latter cannot be distinguished in any essential
way from a pseudopodium). A roughened tail is formed de novo when-
ever an Amoeba suddenly changes its direction of movement. The
previous anterior end then becomes roughened in contracting and forms
a typical tail. This latter unites with the old tail if any of the latter
remains. The substance of the tail gradually passes forward into the
rest of the body, as we have seen.

MOVEMENTS AT THE ANTERIOR EDGE.

As in Amceba verrucosa and its relatives, so in the species of more
changeable form, the most active movements are taking place at the
anterior edge. In Amceba proteus ?ind A. Umax one sees still more
distinctly than in the species before named the pushing forward of a
series of waves of hyaloplasm which become attached to the substratum
in front. In Amceba Umax and its relatives especially such a wave
may be very pronounced, extending forward at times one-fifth the
length of the body or more, though usually much less.

At first the advancing wave, as it moves rapidly forward, is usually
free from granules, and may be spoken of, therefore, as hyaloplasm.
If the motion is less rapid, however, it contains granules, and is not
distinguishable in any way from the interior endosarc. Where it is at
first free from granules, it is nevertheless highly fluid in character, as
is shown by the fact that it flows and spreads out swiftly, and that the
granules of the endosarc pass into it rapidly. The freedom of the
advancing hyaloplasm from granules is not due to its greater density
or solidity as a result of the action of water upon it, as has sometimes
been maintained, for it is at first free from granules ; then, after the
water has acted longer upon it, it becomes filled. Apparently the
reason for its freedom from granules is merely the fact that it moves
forward faster than the granules and leaves them behind. This view

THE MOVEMENTS AND REACTIONS OF AMCEBA.

l6l

is supported by the variations to be observed in the relation of the clear
substance to the granular substance at the anterior end. These varia-
tions arise chiefly from the different rates of movement of the two
substances, and may be summarized as follows :

I . The clear substance moves fastest at first, and therefore becomes
separated from the granules as a broad band in front (Fig. 52, a); this
is then immediately filled completely by the granules. Even large
granules or vacuoles pass to the very anterior edge, so that one sees
but a line between these and the quter water.

Fig. 52.*

2. The clear substance advances fastest, and so continues to do, so that
it remains a long time as a broad, clear band in front of the granules.

3. The two substances advance at the same rate, so that there is no
separation between them. The granules and vacuoles are then found at

♦ Fig. 52. — Distribution of hyaloplasm and granules at the anterior end in
Amoeba Umax and its relatives : a, hyaloplasm without granules at the anterior
end ; b, granules and vacuole at the anterior edge ; c and d, two successive
instants in the locomotion ; at c the anterior half of the body is free from gran-
ules, the latter being heaped up behind a well-marked barrier; at d the barrier
has burst at a certain point (a;), allowing a stream of granules to flow forward to
the anterior edge; e andy, two successive stages at the advancing anterior end;
in e the clear hyaloplasm has stopped at the line x-y\ at/" the hyaloplasm has
advanced, while the granules are heaped up behind the same line x-y.

l62 THE BEHAVIOR OF LOWER ORGANISMS.

the advancing edge. This condition is not at all rare. In such cases
there is at the anterior end less clear space between the granular region
and the water than in any other part of the body. In fact, there is
typically no space at all. I have seen large vacuoles come so close to
the anterior edge at such times that it was not possible to distinguish
between the boundary of the vacuole and the boundary of the Amoeba
(Fig. 52, <5).

4. Finally, either hyaloplasm or endosarc or both may stop in any
of the positions mentioned above. Thus the hyaloplasm may stop,
whereupon the endosarc flows into it and fills it, or both may stop, so
that the hyaloplasm remains empty, as a clear band, for a long time.

The line separating hyaloplasm and endosarc is at times very sharply
defined, as has often been pointed out. A number of unusually favorable
specimens gave me the opportunity of determining the reason for this,
in many cases at least. The Amoeba in question (Fig. 52, c-f) was an
elongated, rapidly moving form, much resembling A, limax^ but having
usually two contractile vacuoles, one very large, in the fore part of the
body, the other smaller and in the rear. The body contained many
fine granules, which, when the animal was at rest, were scattered almost
uniformly through the body; the peripheral, more solid zone (usually
called ectosarc) contained as many of these as did the endosarc.

In moving this Amoeba usually sends out first a large amount of clear
fluid containing no granules ; this at times extends so far as to constitute
half the length of the Amoeba (Fig. 52, c). There is a perfectly sharp
line between the clear hyaloplasm and the granular endosarc, and behind
this line the granules of the endosarc are heaped up, as if under pressure.
Suddenly this line gives way over a small area (at x)^ and the granules
pour through it in a thin stream nearly or quite to the anterior tip of
the Amoeba (Fig. 52, d).* Gradually the whole barrier gives way, and
nothing is left to .mark the position it occupied. If after its first out-
flow the hyaloplasm has stopped, the whole Amoeba is filled with
granules. But if, as is usually the case, after a pause the hyaloplasm
has started forward again, the granules of the endosarc stream forward
not to the anterior tip, but only to the line which formed the anterior
boundary of the hyaloplasm at its pause {x-y^ Fig. 52, e^f). Here
the granules stop and are heaped up again, until finally the barrier
breaks as before, and the granules rush forward again, to be stopped
at a new line.

The explanation of these phenomena becomes evident on careful
examination. It is to be noted that the line x-y which stops the
flow of the granules of endosarc is always identical with one which

* Similar phenomena have been observed by Prowazek (1900, p. 17 ; Fig. 18).

THE MOVEMENTS AND REACTIONS OF AMCEBA. 1 63

formed the anterior boundary of the hyaloplasm (that is, of the Amoeba
as a whole) at a previous pause. The reason for this is as follows :
The lower surface of the Amoeba, as we know, is at rest ; here the
protoplasm has become modified to form a sort of membrane. This
membrane extends up a very little distance at the sides and ends (as
is shown by the fact that the protoplasm at the sides is at rest). Thus
at the anterior end there is, after a pause, a low barrier formed by this
membrane. The next wave of advancing hyaloplasm arises just behind
this barrier, overleaps it, and pushes forward (the conditions being
essentially the same as in Amoeba verrucosa^ already described). This
advancing wave when first formed is very thin, forming a mere sheet
lying on the substratum. This is shown by the fact that when the out-
line of the remainder of the Amoeba is sharply in focus, the anterior
portion is often undefined, and one is compelled to focus lower to get
its outline sharply. The thin sheet of hyaloplasm which has just pushed
forward is bounded behind by a low wall, formed from the membrane
which previously limit-
ed it in front (Fig. 53,
x). Now the granules
of the endosarc flow for-
ward until they reach
this boundary ; there
they stop and become
heaped up against it (Fig. 53). After a time the membrane forming
this barrier, since it is now completely enveloped by protoplasm, be-
comes dissolved and gives way in the manner described above ; the
granules then flow forward. Meanwhile, a new partial boundary
has been left in front by the hyaloplasm ; this again stops the endosarc,
and the whole process is repeated many times.

Of course, when the anterior boundary advances uniformly, without
pauses, no anterior membrane is formed, and there is nothing to hold
back the granules of the endosarc ; hence there is no reason for a separa-
tion of hyaloplasm and endosarc, and we find that none exists. On the
other hand, when the Amoeba has paused for a long time the anterior

* Fig. 53.-— Diagram of a longitudinal section of the anterior edge of Amoeba,
to show the cause of the stopping of the granules of the endosarc some distance
behind the anterior margin. The line beneath represents the substratum to
which the Amoeba is attached. The anterior h_yaloplasm at first moves forward to
the line x-x ; stopping there it becomes covered with a firmer wall, as represented
by the heavy black line. Now the hyaloplasm pushes forward from above the
anterior edge x-x, forming a thin sheet closely applied to the substratum, as
shown in the figure. The endosarcal granules flow forward, but are stopped by
the barrier x-x (the former anterior boundary of the Amoeba) ; they cannot flow
forward till this boundary is liquefied.

164 THE BEHAVIOR OF LOWER ORGANISMS.

membrane seems especially well developed, for the hyaloplasm pushes
then a long way ahead and may form half the length of the Amoeba
before the endosarc has burst through the membrane.

The phenomena described above are very general in creeping Amoebae,
both in those with usually but a single pseudopodium (as A. Umax),
and in those with many pseudopodia (as A. angulata).

Of course, the general fact that there is a separation between hyalo-
plasm and endosarc is not explained by these observations ; thus we
know that they are often separated even in pseudopodia that are pro-
jected freely into the water. But the phenomena are much less marked
in such cases ; it is exactly the observed difference between them and
the phenomena to be seen in a creeping Amoeba that the above obser-
vations explain.

In all these details it is important not to lose sight of the essential
point in the movement at the anterior end. This is as follows : The new
wave begins on the upper surface just behind the former boundary line,
and rolls forward, so that its former upper surface is now in contact
with the substratum.

This method of movement explains a peculiar fact which one observes
frequently, and which I found it difficult to understand before this
movement was demonstrated. The advancing edge in Amoeba usually
does not push forward fine, loose particles lying on the substratum in
front of it, but overlaps them instead, so that the Amoeba creeps over
them. This is especially noticeable when the water contains many
particles of India ink or of soot. In view of the rolling movement with
the series of waves, each coming from behind the previous anterior
edge and thus descending on the substratum from above, this burying
of loose movable particles becomes intelligible.

In Amoeba proteus and its relatives the advancing anterior edge does
not move forward in a single uniform curve, as it does in A. verru-
cosa, and, as a rule, in A. Umax. On the contrary, the anterior edge
may show the most varied form and modeling (Fig. 54) ; narrow points
may be sent out here ; a broad rounded lobe there ; a rectangular projec-
tion elsewhere. Pseudopodia may rise above the general level, pro-
jecting freely into the water and later coming in contact with the
substratum, or be withdrawn without coming thus in contact. The
anterior portion of the body may divide into two lobes, or may become
hollowed out so as to contain a cavity bounded by thin walls (see
later. Figs. ^^, 76). These facts show clearly that the method of
advance of the anterior edge cannot possibly be determined by general
pressure from behind, such as would be produced, for example, by a
contraction of the posterior part of the body. Such pressure could not
produce the cavity shown in Fig. 75 or the thin edge bearing numerous
points shown in Fig. 54.

THE MOVEMENTS AND REACTIONS OF AMCEBA.

•65

MOVEMENTS OF THE POSTERIOR PART OF THE BODY.

In an Amoeba moving as a simple, elongated mass, the anterior por-
tion of the body is broadest and very thin, being flattened against the
substratum, while the posterior part is narrower and much thicker. In
many cases the posterior end rises to an actual hump, the body being
thickest at the posterior edge, or a little in front of this edge (Figs. 44,
58). This is true as well of the Amoebae of nearly constant form (A.
verrucosa^ etc.. Fig. 44), as of
those related to A, proteus. From
this hump the upper surface slopes
forward to the thin anterior edge.
The margins in the posterior part of
the body are not thin, but rounded
like the surface of a cylinder.

The anterior portion of the
Amoeba is attached to the substra-
tum. This attachment of the ante-
rior portion has been clearly demon-
strated by Rhumbler (1898), and I
can confirm his results throughout. f
The attachment is probably by a
mucus-like secretion ; at least such
a secretion exists, as Rhumbler and

others have shown and I can confirm. I have sometimes been able to
pull an Amoeba about by using a glass rod to which a thread of this
mucus had become attached (Fig. 55). The animal then seems to
follow the rod at a distance, the thread of mucus not being visible.
In virtue of this attachment the Amoeba resists currents of water,
or the impinging of solid bodies against it. The posterior portion
of the body is not thus attached, but is entirely free from the bottom.

In many cases the most posterior part of the body forms a more or
less distinctly marked oflT portion, the surface of which is wrinkled or
warty or villous, or otherwise irregular. This is variously known as
the tail, the villous patch, or appendage (Wallich, Leidy), houppe
(Penard), Schopf (Rhumbler), etc. The occurrence of this appendage
is variable. In some species it is usually present, in others less com-
mon. Its occurrence and degree of development vary, indeed, in the
same individual.

Fig. 54.

♦Fig. 54. — Amoeba angulata in locomotion, showing the numerous points in
the anterior region, some attached to the substratum, others projecting freely into
the water, a is the "antenna-like" pseudopodium, described on p. 177.

tit is rather curious that Biitschli (1892), in his discussion of the movements
of Amoeba, is inclined to deny that there is any attachment to the substratum.

1 66 THE BEHAVIOR OF LOWER ORGANISMS.

In an advancing Amoeba the posterior end moves forward at about
the same rate as does the anterior, since the distance between the two
remains about the same. Leaving out of account for the moment
specimens with the wrinkled appendage, there is a continual current
forward from the posterior end. Nevertheless, the latter remains on
the average of about the same size. The material which flows out of

it above is supplied from
beneath. As we have
seen, a layer of material
at the under surface of
the Amoeba is at rest.
The main portion of the
body passes over this
layer, dragging the pos-
^^^' ^^* terior end. The latter

takes up as it passes the resting layer which was against the
substratum. This gradually becomes fluid, and passes for-
ward again in the advancing current. All this may be
clearly seen by observing the course of individual particles
in the protoplasm and on the surface, and is fully set
forth in the preceding pages of this paper.

The unattached posterior portion steadily contracts as it
moves forward. Particles on its upper surface are moving forward, as
we have seen in detail. But this is not all. Particles on its sides and
under surface likewise move forward ; there is an actual contraction
independent of the streaming already described. The movement of
substance due to this contraction is more striking and rapid as we
approach the posterior end. As this contraction is an important fact,
it will be well to give some details of the observations which show it
to exist.

Particles attached to the lower surface, or to the lateral margins of
the Amoeba, next to the substratum, in the anterior part of the body,
remain quiet for a long time. But this lasts only till they have reached
that portion of the body which is free from the substratum ; then they
begin to move slowly forward as a result of the contraction just
described. Of two such objects, one nearer the posterior end, the
hinder one moves the more rapidly, so that the distance between the
two slowly but distinctly decreases. Though such objects on the bot-
tom or sides move forward, they do so less rapidly than does the
posterior end. The latter, therefore, in time overtakes them, and they
are finally pulled around the posterior end to the upper surface, where
they pass forward, as we have already seen in detail.

* Fig. 55. — An Amoeba drawn backward by a thread of its viscid secretion.

THE MOVEMENTS AND REACTIONS OF AMCEBA.

167

Fig. 56.*

The contraction of the posterior part of the body is further shown by
the behavior of retracted pseudopodia. When a pseudopodium con-
tracts it usually produces, as we have seen, a
small wart-like excrescence, which persists
for some time. Such wart-like remains of
pseudopodia behave like foreign bodies at-
tached to the margin of the Amoeba. In the
anterior half they remain quiet, while in the
free posterior half they move slowly forward,
as a result of the contraction of this part of
the body. When two or more of these rem-
nants of pseudopodia are formed at once,
with an interval between them, this interval
becomes less as a result of the more rapid
movement of the hinder one.

The contraction of this posterior region is
sometimes very striking, especially when the
posterior end becomes attached to some for-
eign object and is drawn out longer than
usual ; when it finally becomes free it con-
tracts suddenly and rapidly. Thus, for example, an Amoeba hav-
ing the form shown in Fig. 56, a^ began to move in the direction
shown by the arrow, when it became evident that the posterior end was
attached to a diatom shell, which was fast to the substratum. As the
Amoeba crept away the posterior end was drawn out, as shown at b.
Finally the diatom became detached from the bottom, when the stretched
posterior end at once contracted, shortening up rapidly, so that the
Amoeba had the form shown at c. Such observations are often made.

This contraction does not occur in that part of the Amoeba which is
attached to the bottom, but begins at once as soon as the attachment
ceases. One might compare the outer layer to a stretched sheet of India
rubber that is attached to a surface by means of some adhesive sub-
stance. As soon as the adhesion gives way the sheet contracts. There
is no definite point at which the attachment to the substratum must
cease ; sometimes it is farther forward, sometimes farther back. The
gradual freeing of the posterior portion can be clearly observed in many
cases, particularly in Amoeba a7igulata^ and may be seen to go hand
in hand with the contraction of the ectosarc. As soon as a certain part
of the body becomes free, its contraction takes place with some sudden-
ness, and the contraction is the more noticeable the greater the part of
the body that is freed at once. Often the process of becoming freed

*FiG. 56. — Successive forms of an Amoeba, showing the marked contraction
of the posterior end. (See text).

l68 THE BEHAVIOR OF LOWER ORGANISMS.

takes place in a series of jerks, and there is a corresponding jerkiness
in the contractions of the posterior part of the body. When a large
amount of surface is freed at once there is a sudden forward rush of the
fluid portion of the Amoeba, with a striking contraction of the posterior
part of the body. Such a case as is shown in Fig. 56 is only an unusu-
ally strong contraction of this sort, due to the fact that the hinder part
on the body had remained locally attached longer than usual.

As a result of this contraction, the ectosarc of the posterior part of
the body becomes thickened and wrinkled, or warty. The change from
a flat plate to a rounded form involves a decrease in the amount of exter-
nal surface, and as the amount of material in the surface layer is not at
once decreased, this layer is compelled to fold and become wrinkled and
warty. When this process is very pronounced we have produced at
the posterior end the wrinkled, warty appendage so often described.
Such a roughened structure may be produced in any part of the Amoeba
by rapid contraction, as we have seen above (p. 160). The rough,
warty appendage at the posterior end is the common product of all the
contractions which have taken place.

In Amoeba verrucosa and its relatives the current forward on the upper
surface extends backward to the posterior end ; the outer surface of the
latter seems not markedly different in texture from that of the rest of
the body. In other species of Amoeba there is a greater difference
between the texture of the surface layer of the anterior part of the body
and that of the posterior end, and this may involve some differences in
the movements. Often, in even these species, the forward current
extends backward to the very posterior end ; particles on the under side
pass up over the posterior end and forward, just as in A. verrucosa
(see p. 147). But in other cases, in A. lt?nax^ A. froteus^ etc., the
surface material at the posterior end is so stiffened that it is temporarily
excluded from the current. There is then produced the distinct, rough-
ened appendage, which is for a time dragged passively behind the
Amoeba. In such a case the currents from beneath pass upward on
either side of this appendage, meeting in the middle line (Fig. 57).
Particles attached to the under surface on either side of the appendage,
therefore, soon pass to the upper surface and are carried forward, while
those on the under surface of the appendage itself may remain in position
and be dragged forward for a considerable time.

But I have rarely found this posterior appendage so completely cut
off' from the general circulation as is often supposed. Usually there is
a very slow current forward on the upper surface of the appendage,
involving also its internal parts. Into this current particles attached
to the posterior end, and even to the under surface of the appendage,
are in course of time drawn. I have thus seen particles of soot dragged

THE MOVEMENTS AND REACTIONS OF AMCEBA. 169

about for ten minutes at the posterior end, then finally pass upward into
the surface current, where they were carried to the anterior end. Even
when this slow current from the posterior appendage is completely
suspended the suspension is only temporary. The currents after a period
begin again, and a strongly marked warty posterior appendage may in
time completely disappear, its substance hav-
ing become mingled with that of the remain-
der of the Amoeba.

Other parts of the Amoeba may become
temporarily immobile and thus excluded from
the general circulation. One often sees certain Fig. 57.*

parts of Amoeba proteus and other similar species thus quiet, while the
restof thebod3MS in active motion (see Rhumbler, 1S98, o. 122).

In the species with "eruptive" pseudo podia this process seems to
have gone farthest. Here the whole outer layer apparently becomes
hardened at times, so that movement occurs only when the inner sub-
stance bursts through this, forming "eruptive" pseudopodia. In this
case the pseudopodia formation apparently differs from that in Amoeba
proteus and its relatives, as described above, in the fact that the ectosarc
of the body is not transferred to the surface of the pseudopodium. I
have not been able to study this process in detail further than to deter-
mine that there is no backward current on such pseudopodia. The
matter is worthy of further examination. In Amoeba verrucosa^ the
species which has been hitherto supposed to have the most immobile
ectosarc, we have shown that the outer layer is in continual rotary
motion in the progressing animal.

GENERAL VIEW OF THE MOVEMENTS OF AMCEBA IN LOCOMOTION.

Let us now, with the aid of a diagram, attempt to form a clear con-
ception of the movements occurring in an Amoeba that is progressing
in a definite direction, with a view to determining the nature of the
forces at work. Fig. 58 may represent a longitudinal section of such
an Amoeba as seen in a side view. The anterior end, A^ is, as we have
seen, very thin, and is applied closely to the substratum, while the
posterior end, P^ is high and rounded, forming a sort of pouch. It is
free from the substratum, beginning at the point at, at about the middle
of the body.

At the anterior end waves of hyaloplasm are pushed forward one
after the other, so that the anterior end successively occupies the posi-
tions «, ^, c. As we know, it is the upper surface which thus pushes
out ; it rolls over, so that a point which was originally on the upper

♦Fig. 57. — Diagram of the surface currents when the posterior appendage is
excluded from the general stream.

ifjO THE BEHAVIOR OF LOWER ORGANISMS.

surface becomes applied to the substratum. It is evident that the sur-
face of the Amoeba is increased in extent by the pushing out of these
waves.

On the upper surface of the Amoeba there is a forward current of the
outer layer, as indicated by the arrows. This current extends back-
ward to the posterior end, where it is continued as a movement upward
from the lower surface. This upward movement stops, at any given
movement, at about the point jj/, though, of course, the point where
it ceases cannot be precisely fixed. That part of the lower surface
which is in contact with the substratum, from the anterior end to x^ is
quiet. That part of the lower surface which is not in contact with the
substratum (from x to y) is moving slightly forward owing to the con-
traction in this region, as described on pages 166-168. This movement
is comparatively slight, as indicated by the small arrows. Within the

Jtr JC' a o c

Fig. 58.*

Amoeba are currents moving forward in the same direction as the
current on the upper surface.

The posterior end as a whole moves forward, so that it comes to
occupy later the position shown by the dotted line. The point x^
where the lower surface of the Amoeba becomes free from the sub-
stratum, moves forward an equivalent amount to x. The entire
Amoeba thus moves forward in the direction indicated by the large
arrow above.

Thus far our account has been purely descriptive, containing only
what has been demonstrated by observation and experiment, and intro-
ducing nothing hypothetical. We must now endeavor to form a con-
ception of the location and direction of action of the forces at work In
producing the movements. Discussion of the ultimate character of
the forces will be reserved for the section on the theories of amoeboid
movement.

One of the primary phenomena is evidently the pushing forward of

♦Fig. 58. — Diagram of the movements of Amoeba, as seen in side view. A^
anterior end; P, posterior end; a, b, c, successive positions occupied by the an-
terior edge. The large arrow above shows the direction of locomotion ; the
other arrows show the direction of the protoplasmic currents, the longer arrows
indicating the more rapid currents. For further explanation see text.

THE MOVEMENTS AND REACTIONS OF AMCEBA. I7I

the anterior edge. The form of the Amoeba shows that this cannot be
due to pressure from behind, for if the pressure were greatest behind
and less in front, the mass of internal fluid would, of course, be forced
forward, and the Amoeba would be thickest in front instead of behind ;
the form of anterior and posterior ends would be reversed. In the
varied modeling of the anterior end we have seen another proof of the
impossibility of accounting for the action here by pressure from behind
(p. 164). Further, the forward current on the upper surface of the
Amoeba could not possibly be produced by pressure from behind.
The impossibility of accounting for the form and movement by pres-
sure from behind has been recognized by most investigators, though
on other grounds than those here set forth.

On the other hand, if we take the view that the anterior wave, after
attaching itself to the substratum, exercises a pull on the parts behind
it, the rest of the phenomena follow most naturally. Such a pull
would draw forward the tenacious upper layer of the body, and we
find that this is actually moving forward. The mass of inactive inter-
nal contents would drag behind as far as possible, so that the thickest
part of the Amoeba would be at the rear, and this is exactly what we
find to be true. The posterior end would be dragged forward. This,
also, is true. In its movement forward it would be partly rolled ; that
is, its lower surface would gradually pass upward and become the upper
surface. This, also, we know to be true.* Finally, the internal fluid
contents would be compelled to stream forward as the anterior end
advanced and the posterior end followed it. This streaming is, of
course, one of the striking features of the Amoeba. The character-
istics of the endosarcal streaming are, I believe, exactly what might
be expected from the method of origin just set forth.

We have one other more or less independent factor of the movement
in the contraction of the posterior part of the body that is not in con-
tact with the substratum. As we have seen, the substance of the sides
and bottom, as well as of the upper surface, are moving forward in
this region, as indicated by the small arrows of the diagram (between
;vandjv). Moreover, we know that there is a lateral contraction as
well as an antero-posterior one, for the wide, flat anterior portion
shrinks together as soon as it is released from the substratum. This
contraction should probably be brought into relation with the previous
increase of surface at the anterior end. As the anterior wave is sent
forth the surface at the anterior end is much increased. We might com-
pare the action with the stretching of a sheet of India rubber. This
tense portion then becomes attached to the substratum, as we might

* Large bodies within the Amoeba close against the posterior surface are often
rolled over in this process, as I have several times observed.

172 THE BEHAVIOR OF LOWER ORGANISMS.

fasten the stretched sheet of India rubber by means of a strong cement
to a plate of glass. Upon being released the tense layer of protoplasm
contracts again, just as the rubber sheet would do. In the protoplasm
the contraction takes place rather slowly, causing the steady pulling
forward of the posterior half of the body, as exhibited by objects attached
to its lower surface.

In connection with the contraction of the lower surface as just de-
scribed we must consider also certain properties of the upper surface.
When this is pulled upon by the advancing anterior wave it does not
respond like an inelastic membrane, but like an elastic, contractile one.
If it were inelastic its motion would follow that of the anterior end
exactly, and thus take place in a series of jerks. But this does not
occur. When the anterior wave pushes forward, thus extending the
upper surface, there is no immediate increase in the movement of this
surface, nor of the posterior end ; the movement forward is a steady
one. The property of contractility is further shown very directly in the
phenomena which take place when a large portion of the lower surface
of the Amoeba is suddenly released, as described on page 167. It seems
clear that the entire surface of the Amoeba is in a state of tension and that
this tension is directed toward the advancing anterior end. The condi-
tion would be imitated by partly filling a rubber sack with a heavy fluid,
causing one surface to adhere to the substratum and pulling on one side.

As a result of this tension on the surface the internal contents of the
Amoeba must, of course, be under a certain slight amount of pressure.
As we have seen (p. 171) even without this pressure the internal con-
tents must flow forward, since the posterior surface, against which
they are resting, is moved forward. But the pressure accounts for
certain details of the movements of the endosarc. Thus, when a pseudo-
podium is sent forth, or one of the anterior waves moves forward, it is
usually soon filled by endosarc. In its pushing forward the pseudo-
podium forms a region where the tension is relieved ; the fluid contents,
under pressure elsewhere, therefore flow into it.

It is to be noted that this pressure is a mere consequence of the tension
due to the pushing forward of the anterior edge, and is by no means a
cause of the pushing forward ; it is always, therefore, subordinate to and
dependent upon the latter, and is not a matter of primary significance.

Altogether, then, our results lead us to look upon Amoeba as an elastic
and contractile sac, containing fluid. In locomotion one side of this sac
actively stretches out, becomes attached to the substratum, and draws
the remainder of the sac after it in a rolling movement. The primary
phenomena are the stretching out of one side, the elasticity, and the
contractility of the outer layer.

Whether this elasticity and contractility should not be considered

THE MOVEMENTS AND REACTIONS OF AMCEBA. 1 73

properties, not merely of the outer layer, but of the entire substance of
the Amoeba, may be a question. The fact that ectosarc and endosarc
are mutually interconvertible would seem to imply an affirmative answer
to this question, and I believe other evidence could be adduced looking
in the same direction. But the locomotion itself seems to require these
properties only in the ectosarc, so that we shall for the present leave out
of consideration the question as to their existence in the endosarc.

To the three primary phenomena above mentioned we must devote
further attention. It has been maintained by certain writers that the
ectosarc is not an elastic and contractile membrane, as above set forth ;
hence we must examine the evidence on that point. There then remain
the questions : What is the cause of the pushing out at the anterior edge ?
and. What is the essential nature of the contractility of the ectosarc?
These questions will be reserved for a special section on the theories
of amoeboid movement. We will at this point investigate certain
general properties of the substance of Amoeba, with a view espe-
cially to determining whether we are justified in considering the
ectosarc elastic and contractile, though not limiting our attention to
these properties alone.

SOME CHARACTERISTICS OF THE SUBSTANCE OF AMCEBA.
FLUIDITY.

It is, of course, not necessary to dwell upon this point ; it has been
treated in detail recently by Rhumbler (1898, 1902) and Jensen (1900).
For anyone who is familiar with the movements of Amoeba from per-
sonal observation, doubt cannot exist that its protoplasm has at least
some of the most striking properties of fluids, notably the property of
flowing, with the freedom of movement of the particles with reference to
each other that this implies. This applies most strongly to the endosarc ;
for the ectosarc, as we shall see, there are decided limitations to the
fluidity. Nevertheless, the particles of the ectosarc have, to a considera-
ble degree, that freedom of movement with relation to each other that is
characteristic of fluids. This is shown, for example, by the fact that any
portion of the ectosarc may be temporarily excluded from the advanc-
ing stream (especially common at the posterior end, p. 169), and in the
fact that neighboring portions of the ectosarc may flow in opposite
directions (p. 148). But the characteristics of the ectosarc are much
more those of a tough, rather persistent skin than has sometimes been
supposed. This point is brought out in the following sections.

rhumblkr's " knto-ectoplasm process."
The movements of the outer layer of the body described in this paper
have an important bearing on the transformation of endosarc into ecto-
sarc and vice versa^ of which so much is said in Rhumbler's recent

174 "^"^ BEHAVIOR OF LOWER ORGANISMS.

extensive paper (1898). My observations show that this transforma-
tion is confined within much narrower limits than Rhumbler supposed.
In the account which he gives of the movements of Amoeba this trans-
formation (the " ento-ectoplasm process," as he calls it) plays a very
large part, and is essential to locomotion. At the anterior end, accord-
ing to Rhumbler, endosarc constantly flows out to the surface and is
there transformed into ectosarc, flowing back as such along the surface
of the body. Somewhere in the posterior part of the body or at the base
of a pseudopodium this ectosarc passes inward and is retransformed
into endosarc. These supposed processes are indicated in the diagrams
from Rhumbler, Figs. 30-33.

My observations show that this view of the constant inter-transforma-
tion of the two layers is incorrect, and that we must attribute to the
ectosarc a much higher degree of permanence than Rhumbler supposed.
There is no regular transformation of endosarc into ectosarc at the
anterior end. On the contrary, the ectosarc here retains its continuity
unbroken, moving across the anterior end in the same manner as across
other parts of the body. In the same way, the ectosarc is not regularly
transformed into endosarc in the hinder part of the body. We can trace
a single definite point on the ectosarc (or a complex of such points
having a definite relation to each other) continually until it has passed
completely around the Amoeba ; several complete rotations of this sort
are described from actual observation on page 141 . In the species having
very changeable forms a single point on the ectosarc may be traced,
for example, from the surface of a pseudopodium at the posterior end
across the whole length of the body to near the tip of a long pseudo-
podium at the anterior end (Fig. 49, p. 155) ; there is no reason to
suppose that it could not be traced indefinitely but for the difficulties
of observation.

On the other hand, there is no doubt that ectosarc may be trans-
formed into endosarc, and vice versa^ under certain conditions. This
process was apparently first clearly seen by Wallich (1863, a, p. 370).
Wallich saw the formation of " eruptive pseudopodia " by the outflow
of the endosarc through a small perforation in the ectosarc. A portion
of the latter was thus covered by the endosarc, and gradually resorbed.
Rhumbler figures a similar case (1898, p. 152), and I have repeatedly
seen such. Further, as Rhumbler has shown, and as I can confirm, in
A. verrucosa food bodies are enclosed in a layer of thick ectosarc, which
passes with the food into the endosarc, there to be resorbed (seep. 195).

Thus, it is clear that there may be a transformation of one layer into
the other under special circumstances, but such transformation is much
less general than Rhumbler supposed, and is by no means a regular
accompaniment of locomotion.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 1 75

ELASTICITY OF FORM IN AMCEBA.

For a real understanding of the phenomena shown by amoeboid
protoplasm it is important to realize that it has, besides some of the
chief characteristics of fluids, a number of properties that are usually
considered characteristic of solids. This came out clearly in certain
of my experiments. They show that Amoeba has elasticity of form to
a considerable degree.

These experiments consisted in changing the shapes of Amoebae with
a fine capillary glass rod, under the microscope, in the open drop.
From numerous experiments of this character the following may be
selected as typical :

An Amoeba had sent out one rather long, thick pseudopodium, as
shown in Fig. 59, a. With the capillary glass rod this pseudopodium,
a^ was pulled loose from the bottom and bent over into the position
shown by the dotted outline b. On being released it rather quickly

Fig. 6o.t

sprang back into its original position, a. This experiment was repeated
on different Amoebae many times.

An elongated Amoeba («-«, Fig. 60) was bent with the rod at about
its middle, so that the anterior half was pushed far to one side of the
original median axis (to b). This anterior half at once attached itself
to the bottom, whereupon the posterior half, which was not attached,
immediately swung round into line with it, so that the Amoeba occu-
pied the position b-c. Thus the original straight Amoeba on being
bent immediately straightens itself out again. On repeating this experi-
ment with many elongated individuals it was found that frequently the
straightening out was not quite complete, so that after it had occurred
there was still a slight bend in the middle.

An Amoeba had a long pseudopodium curved over to one side, as in
Fig. 61, a. This pseudopodium was loosened from the bottom with

♦Fig. 59. — A straight pseudopodium, a, is bent into the position b with a rod.
It at once returns to the position a.

tFiG. 60. — A narrow Amoeba, a-a, is bent with the rod into the position a-b ,
the end, b, then becomes attached, and the animal at once straightens into the
position b-c.

176 THE BEHAVIOR OF LOWER ORGANISMS-

the rod and straightened out {b) . On being released it at once swung
back to its original form and position at a.

An Amoeba had many long, slender, free pseudopodia standing out
radially from the body. These could be pushed repeatedly to one side
or the other, or bent at a marked angle. In every case they returned
at once to the radial position.

An indefinite number of such experiments could be detailed. They

show clearly that Amoeba has, to a certain degree at least, one of the

most distinctive properties of solids, a tendency to resist deformation of

shape, and to restore the shape when changed. It will be observed

that the cases are not such as can be accounted for on the assumption

that Amoeba is a simple fluid mass which tends to take a certain form

in accordance with the principle of least surfaces. A small sphere of

fluid when deformed returns to its original shape in conformity with

the principle just mentioned. But such returns to the original form as

^-^^^ ^ are illustrated in Fig. 59

( \ and Fig. 61, for example,

y^^ '"'^' N. \ \ are not required by the

Q^ 1 A principle of least surfaces

\^^ V^ /.— . . ,- s^ so long as the Amoeba is

^^^^^^^ J J ^ considered a simple fluid

-— ^"^'^ mass.

^''^- ^'•* On the other hand, if

we consider Amoeba not as a simple fluid, but as a fluid mass of
a foamlike or alveolar structure, composed of a tense meshwoik of one
fluid enclosing minute drops of another, then the results above set
forth might be explained without assuming that the protoplasm has
in any part passed from a liquid to a solid state. This follows
from th^ considerations and experiments recently set forth by Rhum-
bler in a most important and suggestive paper (1902). Rhumbler
shows that a fluid mass having alveolar structure must react to tran-
sient pressure from without like an elastic body ; in other words,
that it must have elasticity of form. The results which I have set
forth above might almost seem, then, to have been predicted in Rhum-
bler's summing up: "Transient tensions or pressures produce an
elastic reaction of the cell body; longer action, on the other hand,
produces a plastic reaction " (/. <:., p. 371). For the detailed demonstra-
tion of this principle the reader must be referred to the original paper
of Rhumbler. It will suffice to note here that the result is due to the
fact that deformations of the body as a whole produce deformations of
the alveoli, and that the surface tension of the alveolar walls tends to

* Fig. 61. — A large, curved pseudopodium, a, is straightened out into the posi-
tion b with a rod. On being released it at once returns to the position a.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 1 77

restore them at once to their original form, resulting in a return of the
whole body to its original form.

If, then, we hold that the substance of Amoeba is composed of such
an alveolar structure, the above observations are intelligible, even
though no part of the substance of the organism is in the solid state.
But whatever the cause, we must recognize that the protoplasm of
Amoeba shows in the gross some of the characteristics of solids.

Leaving out of account the minute structure of the protoplasm, most
or all of the observations detailed above could be accounted for on the
assumption that the body of Amoeba consists of a sac of fluid, the outer
wall of the sac being tough and elastic, and, as is shown later, contrac-
tile. In this case only the outer layer, the ectosarc, would show the
characteristics of matter in the solid state of aggregation. Certain
observations made by the writer indicate that this is the true state of the
case. These observations are as follows : In several cases a long, slender
pseudopodium, formed of both endosarc and ectosarc, was stimulated at
the tip, causing the endosarc to be withdrawn, and leaving the pseudo-
podium formed of ectosarc alone, as illustrated in Fig. 50, page 158. Such
pseudopodia could with the glass rod be bent sharply at an angle, and
would often remain thus for some time. If, while thus sharply bent,
the endosarc, as sometimes happens, begins to flow back into the
pseudopodium, the latter straightens out with a sort of jerk as soon as
the endosarc begins to fill it. The ectosarc thus acts like an empty
glove-finger, which might bend over when empty but which would
straighten out on becoming filled with a fluid. A tough skin could
perhaps be formed by an alveolar fluid in accordance with the princi-
ples developed by Rhumbler as above set forth. Whatever the explana-
tion, the experiments indicate the existence of this tough skin-like layer
on the outer surface of the body.

CONTRACTILITY IN THE ECTOSARC OF AMCEBA.

Besides elasticity of form, the outer layer of Amoeba clearly has the
power of contracting locally. This is a fact which is omitted from
consideration in many of the theories in which amoeboid movement is
referred to local changes in the surface tension of a fluid mass. It will
be well, therefore, to set forth some of the observations on this point.
I transcribe here some of my notes, with the corresponding sketches.

1. Specimen with a single long, prominent, curved pseudopodium.
This rather quickly swings around bodily toward its concave side, and
unites with the protoplasm of the body (Fig. 62, a).

2. A specimen sends out a long, curved pseudopodium (Fig. 62, 6),
This slowly straightens out, passing from position i to position 2.

3. A. angulata usually sends out at the anterior end a single
pointed pseudopodium obliquely upward into the water (Fig. 62, c).
This point frequently waves from side to side like an antenna.

178

THE BEHAVIOR OF LOWER ORGANISMS.

4. An Amceba was creeping on the surface film, with a very long,
slender pseudopodium trailing behind down into the water and bent to
one side. This pseudopodium suddenly swung far over to the other side.

5. Amoeba with many pseudopodia extending in all directions freely
into the water. Just as withdrawal begins, a given pseudopodium
bends over to one side, becomes curved to form a half circle, or waves
back and forth from one side to the other.

An indefinite number of such observations could be adduced, showing
that with its other movements Amceba has the power of bending and

straightening its pseudopodia and waving
them from side to side. Such movements
have, of course, been described by many
authors ; in the magnificent monograph of
the Rhizopoda by Penard (1902) many
still more striking cases than those I have
described are set forth. Some of them
should be quoted. In Amozba radiosa the
pseudopodia " maybe displaced as a whole
in the liquid, and I have seen them de-
scribe in this manner a quarter of a full
circle in a second, like the handsof a watch
which one pushes forward suddenly by fif-
teen minutes. On two or three occasions
also I have noticed in the very sharp point
of a pseudopodium a rapid movement of
wave-like vibration, so that one could com-
pare itwithaflagellum" (/.<:.,p.88). Simi-
lar phenomena are described for Amoeba
Umax (p. 36), A. gorgonia (p. 79), and
especially for A. ambulacralis (pp. 91 , 92),
in which the pseudopodia act like tentacles.
In other rhizopods, relatives of Amoeba,
Penard describes similar phenomena. Thus
in Pamfhagus mutabilis (p. 439) the pseudopodia are said to move
as a whole in the water "almost as quickly as flagella." Similar
facts are described for Difflugia fristis (p. 255), Cystodifflugia sac-
cuius (p. 429), Pamphagus granulatus (p. 436), Nadinella tenella
(p. 462), and various other rhizopods. Penard compares the movements

Fig. 62.*

♦ Fig. 62. — Movements of pseudopodia : a, a pseudopodium in the position i
bends quickly in the direction shown by the arrow, and unites with the body;
*, a curved pseudopodium, i, straightens into the position 2; c, the antenna-like
anterior pseudopodium of ./l/«ce3« angulata; it vibrates from I to 2, thence back
through I to 3, etc.

THE MOVEMENTS AND REACTIONS OF AMCEBA. l^g

of the pseiidopodia in many cases to the vibrations of flagella. Similar
movements of pseudopodia have, of course, been described by other
authors, including Biitschli (1878, p. 272; 1880, p. 123). The strik-
ing resemblance of the movements of the pseudopodia in some cases to
those of flagella (see especially the account of Podostoma, Clapar^de &
Lachmann, 1858, p. 441) seems to indicate that the motion in these
two classes of structures must be essentially similar in character, and
that no theory of amoeboid movement is likely to be correct that is
inconsistent with the movements offlagella. Certain suggestions as to
the possibility of bringing the two in relation are given in the theoretical
portion of the present paper (p. 218).

The whole body is sometimes moved rapidly by such movements of
the pseudopodia. This happens especially when the body is suspended
in the water and bears many long pseudopodia, one of which comes
in contact with the substratum. This pseudopodium spreads out and
extends along the surface for a distance, the part along the surface
forming nearly a right angle with the free portion. Suddenly the
pseudopodium straightens ; since the distal end is attached, the body
is thrown almost violently against the substratum.

Somewhat similar movements take place frequently in Amoe6a ver-
rucosa and its relatives, without the formation of pseudopodia. The
course of events is usually as follows : A specimen is creeping in a
certain direction in the usual manner with the anterior border attached,
while the posterior end is raised a slight distance from the substratum.
As a reaction to stimulus, or for some other reason, the anterior end
releases itself from the bottom. The posterior end thereupon sinks
down and becomes attached. Then its ectosarc contracts slightly, in
such a way as to lift the anterior end suddenly. The animal thus stands
upon what was its posterior end. Now, by varied contractions of the
parts of the ectosarc in contact with the substratum, the animal may
jerk from side to side rapidly and repeatedly, reminding one of the
movements of certain caterpillars which jerk their anterior ends about
in a similar manner. These movements are very striking and are much
more rapid than any that occur in other species of Amoeba, so far as
I have observed.

The animal may even move from place to place in this manner.
Standing on one end, it jerks its body suddenly over to one side, so that
the previously upper end comes close to the substratum. This end now
becomes attached, while the other is released. Next a new sudden con-
traction brings the released end upward, so that the animal now occupies
a new location, one body's length from that previously occupied. I have
never seen the movement go any farther than what I have just described,
so that there is no evidence that this method is employed for bringing
about orderly locomotion.

i8o

THE BEHAVIOR OF LOWER ORGANISMS.

These movements remind one of the " rolling motion " described by
Rhumbler(i898, p. 115) for these species, though they take place with-
out any noticeable change of form and in a manner entirely different
from the movements described by Rhumbler. As we have seen above
(p. 140), the normal locomotion of these species is, in a certain sense,
of a " rolling" character, so that the phenomena described by Rhumb-
ler as the rolling movements, perhaps really presented nothing different
in principle from the usual motion, though occurring in a different way
because the organism was unattached.

In addition to movements of the character above described, certain
other phenomena show in a different way the contractility of the ecto-
sarc. Thus, I stimulated sharply with a glass rod one side of an elon-
gated moving specimen of Amoeba
Umax about one-third its length from
the posterior end (Fig. 63, a). The
body at once contracted rapidly, in a
ring-like manner (3), at this point,
J and in about \\ seconds the posterior
portion was cut off completely, save
by a fine thread (c), by which it
hung to the anterior portion for a
minute or two. Later this broke,
and the posterior piece finally under-
went degeneration.

Penard, in his great work on the
Rhizopoda, describes similar phe-
nomena in Amoeba terricola {ver-
rucosa^ after injury to the ectosarc.
After a small injury the injured region is invaginated, forming a small
tube passing inward, which is later resorbed. But if the injury is large
the part surrounding it contracts strongly, forming a deep constriction
between it and the remainder of the body (Fig. 64), and this injured
portion is finally constricted off completely (Penard, 1902, p. 109).!

Altogether, then, we may consider it thoroughly demonstrated that
the ectosarc has the power of contracting in definitely limited regions
in such a way as (i) to produce movements of entire pseudopodia com-
parable to those of flagella ; (2) to produce ringlike contractions which
may even progress so far as to cut the body in two completely.

We need not, therefore, hesitate to admit the existence of contrac-
tions of the ectosarc in ordinary locomotion ; these are, for the rest,
as clearly observed as those just described.

* Fig. 63. — An A. Umax is stimulated strongly near the posterior end at a; the
stimulated part thereupon constricts (*, c), separating off the posterior end (rf).
t For other observations on reactions to injuries see pp. 202-204.

Fig. 63.^

THE MOVEMENTS AND REACTIONS OF AMCEBA. l8l

REACTIONS TO STIMULI.

Of particular importance for the understanding of the behavior of
organisms are those reactions which determine the direction of locomo-
tion. Experiments show that the stimuli to such reactions must, in a
slow-moving organism like Amoeba, affect only one side of the body,
or at least affect different parts of the body differently. Owing to the
minute size of Amoeba, it is difficult to apply stimuli in such a way as
to fulfill this condition. Heat or cold, or a chemical in solution, for
examples, when applied to one side are likely
to extend to the other side, and far beyond,
before the slow reaction of Amoeba has taken
place ; the reaction when it occurs is then to a

general, and not to a local stimulation. For „

. . Fig. 64.*

this reason the reactions of Amoeba to such

general stimulation are much better known than those to stimuli locally

applied. I have devoted myself to the reactions to localized stimuli,

and have succeeded in overcoming the experimental difficulties for a

number of different classes of agents, though not for all.

In examining the reactions to stimuli, it will be necessary to keep in
mind the method of locomotion (set forth briefly on p. 169 ; diagram
in Fig. 58, p. 170) . The factors to which special attention must be paid
are : (i) the sending out (or rolling over, as perhaps it would be better
to say) of waves of the ectosarc on one side, determining the anterior
end in the locomotion ; (2) the attachment to the substratum ; (3) the
contraction of parts of the body.

The reactions were studied chiefly in Amoeba proteus and A. angu-
lata; where other species were used, they are specifically mentioned.

REACTIONS TO MECHANICAL STIMULI.

The reaction to mechanical stimuli may be either positive or negative.

POSITIVE REACTION.

An Amoeba floating in the water frequently takes a starlike form,
with many long pseudopodia projecting in all directions. If one of
these pseudopodia comes in contact with a solid object or the surface
film (which may always be considered a solid for these purposes) , the
portion in contact flattens out, attaches itself to the object, and its proto-
plasm begins to flow out in a sheet over the latter. The other pseudo-
podia are now slowly withdrawn and the entire animal spreads out on
the solid, moving usually in the direction inaugurated by the first
pseudopodium which came in contact. Often in passing to the surface
of the solid there are a number of rapid jerking movements, due to

* Fig. 64. — A. verrucosa constricting ofFan injured region, after Penard (1902).

l82

THE BEHAVIOR OF LOWER ORGANISMS.

Straightening out or bending of pseudopodia as described on p. 179.
After becoming completely transferred to the surface of the solid the
form may differ much from that of the floating Amoeba. Fig. 65 illus-
trates such a reaction. A floating Amoeba will thus spread out on the
substratum, on the surface film, or, so far as possible, on small masses
of debris suspended in the water.

An Amoeba which is moving along a surface also shows at times a
positive reaction to mechanical stimuli by turning toward small objects
with which it comes in contact at one side of the anterior end. This reac-
tion takes place very frequently in the normal locomotion of Amoeba,
but I have not been able to produce it experimentally by touching one
side of the animal with a glass rod. This is because it is difficult to
give a touch so light that it shall not induce the negative reaction. I
shall give a detailed account of reactions that probably belong here in
connection with the account of food reactions (pp. 196-202, and Figs.

d c

Fig. 65.*

73-76). As will there be shown, the reaction is often long continued
and rather complicated.

Le Dantec (1895) gave a good account of the positive reaction of
Amoeba, as shown in its spreading out on solids.

NEGATIVE REACTION.

I have studied the negative reaction to mechanical stimuli by touch-
ing a spot on one side or end of the animal with the tip of a fine glass
rod. A glass rod may easily be so drawn out that its tip is as fine as
the tip of a pseudopodium, and with some practice it is possible to give,
under the microscope, in the open drop, very precisely localized stim-
uli with this.

We will first examine the reaction to a rather strong stimulus at the
anterior edge of an Amoeba that is creeping forward with outspread
anterior end and contracted posterior end in the usual way. The tip
of the glass rod is thrust sharply against the anterior edge, producing

* Fig. 65. — Positive reaction to a mechanical stimulus in Amoeba, in side view.
A floating Amoeba comes in contact by one of its pseudopodia with a solid (a);
it thereupon passes to the solid, withdrawing the other pseudopodia {b and c).
See text.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 183

a depression in the ectosarc that may last for some time (Fig. 66),
(We will suppose that the thrust does not detach the Amoeba from the
surface, as sometimes happens.) At once the anterior portion of the
Amoeba ceases to advance. It remains quiet for a definite interval,
which I should judge to be about a second, while the current from behind
continues to move forward. As a result of the stoppage at the anterior
edge there is a heaping up of the protoplasm in the middle of the body.
After about a second the part stimulated begins to contract and currents
start backward from it. Thus the currents from the two ends meet in
the middle, often producing a further heaping up in this region. Usu-
ally, however, the ectosarc of one side of the Amoeba quickly gives
way and a new pseudopodium starts out laterally. As a rule this new
pseudopodium is formed near the original anterior margin, often at the
very edge of the area directly affected by the stimulus (Fig. 66). The
reason for this is evident. Only this anterior half of the Amoeba is
expanded and attached to the substratum, the posterior half being free
and contracted. It is, therefore, much easier to continue locomotion
by sending out pseudopodia somewhere in the attached region than
behind it. If sent out in the unattached region, the original contraction
would have to be overcome, and no locomotion could occur until the
pseudopodium had (by chance ?) come in contact with the substratum
and become attached to it. By sending out pseudopodia thus in some
portion of the attached region, the movement is, in a certain sense, a
continuation of that which was taking place before stimulation, though
in a different direction. The Amoeba follows a path which forms an
angle with its previous one.

The course of the reaction may vary considerably from that above
described. If the stimulus is weak the reaction may consist merely in
a stoppage at the point stimulated without any contraction there. The
current from behind continues ; a pseudopodium breaks out at one side
of the region stimulated, and the Amoeba moves in the direction so
indicated. If the stimulus is very weak the current may cease only
for an instant in the region stimulated, then continue as before ; the
direction of progress thus remains unchanged.

If the stimulus is very strong the contraction which takes place at
the region stimulated may be very marked, resulting in the formation
of strong folds in this region. The contraction may include the entire
anterior end of the Amoeba. Such a contraction destroys the attach-
ment to the substratum, and the new pseudopodium now bursts out in
some part of what was the posterior end of the body. The new course
followed may then be at right angles to the old one, or at any greater
angle, or the course may be exactly reversed, the new pseudopodium
being formed at the posterior end. If the posterior end was much

.84

THE BEHAVIOR OF LOWER ORGANISMS.

wrinkled or bore a pronounced roughened " tail," it is to be noticed
that the new pseudopodium does not flow out directly from this, but to
one side of it or above it (Fig. 6"]). Then as the Amceba moves in the
reverse direction the body passes the old " tail," which finally brings
up the rear again, fusing with the rough area produced by contraction of
the region stimulated (Fig. 67, b). Of course, the new pseudopodium
formed must come in contact with the substratum and become attached
to it before locomotion in the new direction can occur. Sometimes
the new pseudopodium formed is sent directly upward into the water ;
then there is no locomotion until the Amoeba topples over, bringing
the new pseudopodium in contact with the substratum.

At times when the anterior end is stimulated, two new pseudopodia
are sent out in opposite directions on each side of the region stimulated.

Fig. 66.*

Fig. 67.1

Both evidently pull on the Amoeba, which becomes drawn out to form
a narrow isthmus between them. Finally one end pulls the other away
from its attachment to the bottom ; the latter then contracts, and loco-
motion continues in the direction of the prevailing pseudopodium.
There is at times a peculiar additional feature of the reaction to

♦ Fig. 66. — Negative reaction to a mechanical stimulus in Amoeba. An Amoeba
advancing in the direction shown by the arrows is stimulated strongly with the
glass rod at the anterior end (at a). Thereupon the currents are changed and a
new pseudopodium sent out as at b.

tFiG. 67. — Negative reaction to a mechanical stimulus when the anterior end
is strongly stimulated. The arrow, a;, shows the original direction of motion ;
the arrows in a show the currents immediately after the stimulation. A large
pseudopodium is sent out from above and to one side of the former tail (/), as is
shown by the broken outline. In b this pseudopodium has come in contact with
the bottom ; the arrows show the direction of the currents and of locomotion at
this time; t, the original tail; /', the new tail formed by the contraction of the
anterior end.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 1 85

Strong stimuli. In some cases there is for an instant after a strong
stimulation at the anterior end a sudden rush of protoplasm toward
the region stimulated ; this is immediately followed by the stoppage
and contraction above described. Apparently this sudden rush toward
the point stimulated is produced as follows : The first effect of the
additional contraction caused by the stimulus is to release a certain
amount of surface at the posterior edge of the attached area from its
attachment to the substratum. This portion was nearly ready to become
released in the ordinary course of events, so that probably a very slight
shock would release it at once. Now, as I have shown on p. 167, when
a portion of the lower surface of the Amoeba is suddenly released from
the substratum, it contracts, causing a strong forward current. This
is what happens in the case under consideration. Later this current is
stopped by the effect of the stimulus in the anterior region.

The surface currents in the reaction are changed in the same way as
are the internal currents, and are throughout congruent with them.
Particles moving forward on the upper surface of the Amoeba stop
after the stimulus, then move in the direction of the new forward
current. This has been illustrated in detail for Amoeba verrucosa
(p. 143), so that we need not go into the matter further here.

The essential features of the negative reaction to a mechanical stim-
ulus are, then, a contraction of the region stimulated, with the formation
of a new pseudopodium in what may be considered the region of least
resistance, followed by a change in the direction of the currents of
protoplasm, thus altering the course of the Amoeba.

By repeated mechanical stimuli it is possible to drive the Amoeba in
any desired direction. I have at times made use of this possibility in
order to bring into contact two Amoebae or two pieces of an Amoeba
whose courses lay in different directions. Such driving of an Amoeba
requires considerable skill and a rather high tension on the part of the
operator. The new pseudopodium formed is stimulated to withdraw
as often as it is formed, until it finally starts out in the desired direction.
If it were possible to stimulate all of one side of an Amoeba at once it
would, of course, be driven directly toward the opposite side, even
though the stimulus were weak. With chemical and some other stimuli,
as we shall see, this is possible. With mechanical stimuli it is usually
possible only when the stimulus is very strong. By drawing the tip
of the glass rod along one side of a moving Amoeba, it is often possible
to make it flow directly toward the opposite side, as illustrated in Fig.
6^, This point is important for an understanding of the effects of such
stimuli as chemicals, heat, and light.

When a single pseudopodium is stimulated, it is merely withdrawn,
wrinkling and becoming warty in the usual way ; there may be no
other effect on the movement of the animal.

i86

THE BEHAVIOR OF LOWER ORGANISMS.

Among the sweeping statements that one finds current in regard to
the behavior of these low organisms is one to the effect that the Protist
does not avoid an obstacle in its path. This statement is made for
example by Ziehen, in his excellent Leitfaden der fhysiologischen
Psychologic* It is worth while, therefore, to describe in connection
with the reactions to mechanical stimuli just how Amoeba avoids an
obstacle. Let us take a concrete case. An AmcEba creeping with a
broad, flat anterior end came in contact at the middle of its anterior
edge with the end of a long filament of some sort (Fig. 69). The
particular spot touched (c) ceased to move forward, becoming entirely
quiet (reaction to a weak mechanical stimulus). On each side of it
motion continued as before, so that after a time the filament projected
into a notch in the middle of the anterior edge. Then gradually the

Fig. 694

forward movement ceased on the side x and increased at j, the pseudo-
podium X contracted, and its endosarc passed into jk- The animal then
continued its course in the direction indicated by y. It had thus
changed its path so as to avoid the obstacle presented by the filament.
Such cases are often seen.

♦ " Hindernissen weichen dieselben nicht aus " (/. c, p. 11).

tFiG. 68. — An Amceba moving in the direction shown by the arrows in the
unbroken outline is stimulated by drawing the tip of a glass rod along one side,
from a to b. Thereupon a pseudopodium bursts out of the opposite side, as
shown by the broken outline, and the Amoeba continues locomotion in the direc-
tion so indicated.

\ Fig. 69. — Method by which Amoeba avoids an obstacle. The Amoeba a-b-c-d
comes in contact at c with the end of a filament. Thereupon motion at c ceases,
■while elsewhere it continues, so that after a time the Amoeba has the position
shown by the broken outline. Then the currents become changed in x; its sub-
stance passes into the pseudopodium jk> and the AmcBba continues to move in the
direction indicated by the arrows in the lower figure.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 187

Much more striking cases of the regulation of the movement in
accordance with the position and changes in position of outward things
C automatic acts/' Ziehen, /. c.) than are found in such a reaction, or
even than in the possibility of driving an Amoeba, will be described
in the account of the food reactions (p. 196).

REACTIONS TO CHEMICAL STIMULI.

By analogy with the effects of mechanical stimuli we might expect
to find a positive reaction to chemical stimuli. Such reactions doubt-
less occur, but I have not been able to demonstrate them under experi-
mental conditions, and, so far as I am aware, no one else has succeeded
in doing this. Stahl (18S4) has shown the corresponding reaction to
take place in the myxomycete plasmodium. Verworn (1S90, p. 456)
records an observation which he refers with much probability to a
positive reaction to a chemical stimulus in Difflugia. If two conjuga-
ting Difflugias were separated, they crept directly together again, and
it is difficult to see how the movements could have been directed save
by some chemical. But I believe there is no instance of positive chemo-
taxis in Rhizopoda where the nature of the active substance is known
and the reaction was controlled experimentally. A number of striking
positive reactions, which should probably be attributed partly to chemi-
cal stimuli, are described later in connection with the attempts of
Amoeba to obtain food (p. 196).

The same state of affairs has existed hitherto with regard to our
knowledge of a negative reaction to chemicals. In Amoeba it is, how-
ever, not difficult to produce such reactions experimentally.

For this purpose only a small amount of the chemical must be used,
so that it can act on but a limited portion of the body of the animal.
If there is a considerable amount of the solution, diffusing over a large
area, it reaches a strength sufficient to cause a reaction at about the same
time over the whole body of the Amoeba ; thus the reaction is a general
one, not involving movement in a definite direction. To produce a
directed reaction there must be a decided difference in the strength of
the solution on two sides of the organism.

The easiest method of producing the reaction, and the one giving at
the same time the most striking results, is to dip the moistened tip of
a capillary glass rod into powdered methyl green or methyline blue ;
then to bring this near one side or end of the Amoeba, in an open drop
of water. The chemical diffuses in a colored cloud ; the reaction takes
place when the edge of this cloud comes in contact with the Amoeba.

The reaction is essentially the same as that to mechanical stimuli.
The region stimulated stops suddenly, and about a second later con-
tracts, while a current moves away from the side stimulated. This
may meet the previously existing current coming from the original

^^■■^-■■- i

l88 THE BEHAVIOR OF LOWER ORGANISMS.

posterior end ; the two turn to one side, and a pseudopodium starts out

in a new place. If the stimulation took place at the anterior end and

was limited to a small area, the new pseudopodium starts out at one

side of the original anterior end ; the new course followed, therefore,

forms only a slight angle with the former one (Fig. 71, a). But if the

stimulus affects all of one side of the body, or a still greater portion of

its area, pseudopodia are sent out on the opposite side ; the Amceba

then creeps directly away from the source of diffusion of the chemical.

This gives a typical case of negative " chemotaxis,'* the longitudinal

axis being in the line of diffusion of the ions or molecules, with the

anterior end directed away from the source of diffusion (Fig. 70) . It

will be noted that the reaction is exactly the same as that produced by

mechanical stimuli ; in " chemotaxis," where the animal is *' oriented,"

we have the same process as in '' driving" the Amoeba in a definite

direction by means of mechanical stimuli. All movement toward the

chemical is inhibited, because this brings the protoplasm into a region

-^•''**'*>''^X'- where it is stimulated. Pseudopodia can be sent out,

,^^^^^j^^._ therefore, only on the side away from the chemical,

:;^^ and movement can occur only in that direction.

The surface currents are changed exactly as are the

internal currents ; the facts here are identical with

those described for mechanical stimuli (p. 185). The

„ ^ surface currents are thus awav from a chemical which

Fig. 70.* / " T^

causes a negative reaction (see Fig. 42, p. 144).

A number of variations in the reactions to chemicals are shown in
Fig. 71, all of them taken from actual experiments. As the figure
shows, after stimulation frequently two pseudopodia start out in oppo-
site directions, one finally prevailing over the other (Fig. 71, d).

The contraction due to the chemical is often very marked, the ecto-
sarc against which the chemical impinges shrinking sharply together
and becoming covered with folds (Fig. 71, 6). With methyl green as
the stimulus, the surface touched by the chemical is sometimes stained,
so that the shrinkage in area is very precisely definable. With a solu-
tion of NaCl the shrinkage is extreme, while the opposite side spreads
out widely, compensating, or more than compensating, for the decrease
in surface caused by the shrinkage (Fig. 71, d).

The effect of substances not in the form of powder was tried in the
following manner : A glass tube was drawn out to a very fine point,
and into it was introduced some of the solution to be tested. The fine
point of the tube was then brought close to the Amoeba. Some of the

* Fig. 70. — Diagram of *' negative chemotaxis " in Amoeba. A chemical diffuses
from a center, as indicated b_y the radii ; the Amoeba reacts in such a way as to
creep directly away from the sourceofdifFusion,ina line with the radii of diffusion.

THE MOVEMENTS AND REACTIONS OF AMCEBA.

189

chemical flowed slowly out, and its action on the Amoeba could be
observed. The results obtained by this method were very clear. A
negative reaction, as described above, was observed in this way for
solutions of the following substances : Methyl green, methyline blue,
sodium chloride, potassium nitrate, potassium hydroxide, sodium car-
bonate, acetic acid, hydrochloric acid, cane sugar. Of these substances
relatively strong solutions were used (usually about i per cent). Of
course, the solution which came in contact with the Amceba was much
weaker than this, being diluted by the surrounding water. Emphasis
was not laid on the quantitative aspect of the matter ; the question pro-
posed was, How does the animal react .'^ and not. How much is required
to produce the reaction ? Therefore, different strengths were employed

Fig. 71.*

till one was found that was eff^ective. In any case, I do not know of
any way in which one could determine the exact strength of the solution
which comes in contact with the surface of the Amoeba.

♦Fig. 71. — Variations in reactions of Amceba to chemicals. The dotted area
represents in each case the diifusing chemical. The arrows show the direction
of the protoplasmic currents.

a. The chemical (methyl green) diffuses against the anterior end of an advanc-
ing Amoeba; the latter reacts by sending out a new pseudopodium at one side of
the anterior end and moving in the direction so indicated.

b. A solution of NaCl diffuses against the right side of a moving Amoeba (i).
The side affected contracts and wrinkles strongly, while the opposite side expands
(2), the currents flowing in the direction indicated by the arrows.

c. A solution of NaCl diffuses against the anterior end of an advancing Amceba.
The course is thereupon reversed, a broad pseudopodium, shown by the dotted
line, pushing out from the upper surface of the posterior end above the tail.

d. Asolution of methyline blue diffuses against the anteriorendof an advancing
Amoeba (i); thereupon a pseudopodium is sent out on each side of the posterior
end at right angles with the original course (2). Into these pseudopodia are
drawn the body and the tail (3).

igO THE BEHAVIOR OF LOWER ORGANISMS.

As a control experiment, distilled water was used in the tube in
place of a chemical in solution. Amoeba was found to react negatively
to this also, though the reaction was less marked than with most of the
chemicals. But this result, of course, rendered the experiments with
solutions of chemicals in the tube indecisive, as the Amoeba may have
reacted to the distilled water in which the solutions were made up.
The solutions were, therefore, made up with culture water, and the
same results were obtained as before.

The results with the chemicals show merely that Amoeba responds
negatively to almost any solution differing markedly from that in which
the animal is immersed, the precise chemical composition of the solu-
tion being of little consequence. The animal responded negatively not
only to distilled water and to the chemicals mentioned, but also to tap
water, and to water taken from other cultures than that in which the
specimens occurred.

REACTIONS TO HEAT.

Verworn (1889, pp. 64-67) studied the directive influence of heat on
the locomotion of Amoeba by concentrating the sunlight on a small por-
tion of the slide and leaving the rest dark, then observing the behavior
of the Amoeba on coming to the boundary of this lighted and heated area.
The effects of the light proper were excluded by control experiments.
It was found that on coming to the heated area Amoeba remained quiet
a moment, then contracted on the heated side, and sent out a pseudopo-
dium on the opposite side. It then crept away in the direction indicated
by this pseudopodium (" negative thermotropism").

My experiments differed from those of Verworn in employing con-
ducted heat in place of radiant heat ; thus there was no possibility of
a complication from the effects of light. As Verworn sets forth, it is
difficult to warm only one side of so small an object as an Amoeba. I
succeeded, however, in doing this in a very simple manner. For each
experiment an Amoeba was selected that was creeping on the under
surface of the cover glass. This was placed in focus under an objective
of a considerable focal distance, yet of high enough power so that the
internal movements could be seen. A needle was then heated in a flame
and its point was brought against the cover glass a short distance in
advance of the Amoeba. Control experiments had shown that the use
of a needle at room temperature had no effect.

If the heated needle was placed at a proper distance from the Amoeba,
the phenomena follow as described by Verworn (/. c). There was a
short pause, then the side next to the needle contracted. A current
of protoplasm passed toward the opposite side, at times meeting the
current already in existence. A new pseudopodium was sent out, either
on the side opposite the needle, or, in many cases, in a direction inter-

THE MOVEMENTS AND REACTIONS OF AMCEBA. I9I

mediate between this and the original one. The phenomena are in all
respects identical with those seen in the negative reaction to mechanical
stimuli. If the needle is brought a little nearer, so that the heat acts
more strongly, there is a sudden strong contraction of the side affected.

Simultaneously with this there is often, as in the case of strong
mechanical stimuli, a sudden rush of the internal fluid toward the side
stimulated. This lasts but an instant and is succeeded by a current
away from the stimulated side, the formation of a pseudopodium on the
unstimulated side, and locomotion in that direction. The sudden rush
of internal contents toward the side affected is, I think, clearly due to
the cause suggested under mechanical stimuli. Part of the posterior
portion of the attached area of the Amceba is loosened from the substra-
tum by the sudden contraction at the front end ; this portion, therefore,
contracts quickly and sends a current forward, as described on p. 168.

When the heat is still more powerful the entire Amoeba is affected.
It contracts and at the same time loses its attachment to the substratum.
There is a strong momentary rush of the internal fluid toward the end
which had been anterior, due to the cause set forth in the preceding
paragraph. This ceases and the body becomes very irregular and ceases
to move.

The reaction to local stimulation by heat is thus of essentially the
same character as the reaction to mechanical stimuli and to chemicals.

Like Verworn (1889, p. 67), I have been unable to obtain a reaction
to cold in Amoeba.

REACTIONS TO OTHER SIMPLE STIMULI.

The reactions of Amoeba to electricity and to light have been thor-
oughly studied by other authors, so that it will not be necessary to treat
them in detail here. Only certain especially important points will be
touched upon.

The reactions of Amoeba to the continuous electric current have
been studied in detail by Verworn (1890, a; 1897). I have repeated the
experiments in order to determine by observation the direction of the
surface currents of protoplasm during the reaction. For this purpose
soot was mingled with the water containing the Amoebae, and the elec-
tric current was passed through the preparation. The typical reac-
tion as described by Verworn was observed in many cases, but the
surface currents, of course, cannot be seen unless soot is resting upon or
is attached to the surface of the animal, which happens only rarely.
Finally a specimen of Amoeba froteus was observed with a string of
soot particles attached to one side (Fig. 72). The electric current was
then passed through the preparation in such a way that the side bearing
the soot was next to the anode (Fig. 72, a). The Amoeba thereupon

192

THE BEHAVIOR OF LOWER ORGANISMS,

turned and began to move toward the cathode (<$) , dragging the parti-
cles of soot behind it for a short distance. Then the string of soot began
to pass forward on the upper surface, in the usual way (c, d). This
continued until the soot reached the anterior edge and dropped off the
surface of the Amoeba {e). The currents on the upper surface of Amoeba
are, then, forward (toward the cathode) in the reaction to the electric
current, as well as in other cases.

On reversing the current the specimen described above began to
move in the opposite direction toward the new cathode. In this and
many other observed cases of the reversal of movement under the in-
fluence of the electric current, the reversal occurred in the same manner
as when induced by other stimuli (see p. 183). That is, the new pseudo-
podium was sent out from one side of the attached (anterior) half of the
body, changing the course a certain amount. From this new portion
another new pseudopodium was sent out on the side toward the anode,

Fig. 72.*

and this continued until the direction of movement had been, by a
gradual process, completely reversed. Verworn (/. c.) describes cases
in which the reversal takes place suddenly, the new pseudopodium
appearing at the original posterior end. This happens also at times, as
we have seen, in the reactions to other stimuli (p. 183). It is to be noted
that the reaction to the electric current is exactly that which would
occur if the animal were strongly stimulated on the anode side.

Verworn (1889), Davenport (1897), and Harrington & Leaming
(1900) have studied the reaction of Amoeba to light. Verworn (/. c.^
p. 97) found that light falling perpendicularly on one-half of the Amoeba
produced no reaction. Davenport (/. c, pp. 186, 188) confirmed this
result, but showed that when the light falls obliquely from one side on
Amoeba the animal reacts negatively. Harrington & Leaming (/. c.)
found that when white light falls upon the Amoeba from above the

♦Fig. 72. — Movement of particles attached to the outer surface of Amceba in
the reaction to the electric current. Anode and cathode are represented by the
plus C-}-) and minus ( — ) signs, a, Form and direction of movement of the Amoeba
before the current is made ; at, a chain of soot particles attached to one side ; b, c,
d, e, successive stages during the reaction. The chain of soot particles {x) passes
to the upper surface and forward, reaching at e the anterior edge.

J

THE MOVEMENTS AND REACTIONS OF AMCEBA. I93

movements cease, while in red light they begin again ; lights of other
colors have various intermediate effects.

It seems to the writer that further experimentation is desirable on the
results of a perpendicular illumination of one-half the animal. The
difference between the results thus far obtained from such illumination
and those from Davenport's experiments where light is admitted from
one side is very remarkable. If this difference is constant, it is of much
significance for the theory of light reactions. Possibly the lack of re-
action when but one-half the animal is illuminated may be accounted
for as follows : When one end of an Amoeba is illuminated from below,
as in Verworn's experiments, it is difficult or impossible to keep this
difference of illumination constant for any considerable period. If the
Amoeba does not react at once it passes completely into the lighted area,
where there is no cause for changing the direction of movement. On
the other hand, in the case of light falling obliquely from one side, the
different action of the light on the two sides is constant, so that in time
a reaction is produced. The slowness of Amoeba in reacting is such
as to make this possibility worth considering. For further work from
this point of view a source of powerful artificial light is needed. This
I have not had at command during the present investigation.

It is evident that the reaction of Amoeba to light falling from one side
is exactly that which would be produced were the Amoeba strongly
stimulated on the side on which the light impinges.

For an account of the reactions of Amoeba to general (not localized)
stimuli, see Verworn, 1888, and the Allgemeine Physiologie of the
same author.

SOME COMPLEX ACTIVITIES.

Under this heading I propose to describe certain striking phenomena
in the behavior of Amoeba, the stimuli to which are complex or not
sharply definable. These concern the reactions of Amoeba to food and
to injuries, and the relations of one Amoeba to another.

ACTIVITIES CONNECTED WITH FOOD-TAKING.

The behavior of Amoeba in taking food or in attempting to take food
shows many features of great interest for one attempting to understand
the behavior of these organisms. I have observed the process of food-
taking many times, and will describe it, together with a number of
related activities.

Let us take a concrete case. A specimen oi Amoeba proteus was creep-
ing about on a slide which contained many spherical cysts of Euglena
viridis. One of these, which was not attached to the bottom (as most
of them are), was lying in the path of the Amoeba. The latter in its
forward movement came against the cyst and pushed it forward a short
distance. There was no evidence of a tendency of the cyst to adhere to

194 '^"^ BEHAVIOR OF LOWER ORGANISMS.

the Amoeba ; on the contrary, it was pushed ahead as fast as the Amoeba
moved. The Amoeba now put out a pseudopodium on each side of the
cyst, while that part of the protoplasm immediately behind it stopped
moving. Thus the cyst was enclosed in a little bay. On bringing the
upper surface of the cyst into focus it could be seen that a thin layer of
protoplasm was also sent over the cyst. The two pseudopodia enclosing
the cyst now bent over at their free ends, so that the cyst could not be
pushed away by movement of the Amoeba. The two free ends finally
met, leaving only a sort of transparent seam to show the place of con-
tact. Later this disappeared, and the two pseudopodia fused completely.
At the end of two minutes from the time that the Amoeba first came in
contact with it the cyst was completely enclosed. The Amoeba now
remained perfectly quiet for one minute, then crept away, carrying the
cyst with it. With the cyst had been taken in some water, so that it
was enclosed in a vacuole a little larger than itself. The walls of the
vacuole had exactly the appearance of the ectosarc on the outer surface
of the Amoeba.

This is essentially the method of food taking that I have observed in
a large number of cases in Amoeba proteus and its relatives. The
essential points are the sending out of pseudopodia on each side of
and above the food body and the fusion of these pseudopodia at their
free ends or edges, thus enclosing the food. In no case, in these species,
was there any evidence that the Amoeba was aided by the adherence
of the food body to its protoplasm. On the contrary, there was a
decided tendency for the food body to be pushed away, and an essential
part of the process is the overcoming of this mechanical difficulty by
sending out a pseudopodium on each side of the body and bending the
ends of them together, so as to prevent slipping on the part of the
food.* That this difficulty is no imaginary one will be shown later,
in the description of cases where the Amoeba was unable, after many
efforts, to enclose the food.

It is commonly said that the posterior rough, tail-like portion of the
body is especially important in the taking of food, though it is sometimes
added that one rather more often sees the partly ingested food given out
again in this region (see Leidy, 1879, p. 45 ; Penard, 1890, p. 81 ; 1902,
p. 16). t I have never seen food taken in at this part of the body, though,
as noted above, I have many times seen it taken at the anterior end.
While it may be true that food is at times taken at the posterior end, I

♦This lack of adherence between the protoplasm and the food substance is
emphasized by Le Dantec (1894, p. 68; as a result of his careful studies on food-
taking in Amoeba.

tThe references to food-taking at the posterior end seem all to go back to a
paper by P. M. Duncan, " Studies amongst Amoebae," in the Popular Science
Review for 1877. I regret that I have been unable to sec this paper.

THE MOVEMENTS AND REACTIONS OF AMCEBA. I95

believe that the supposed prevalence of this method of food-taking and
of the giving off of incompletely ingested food here are really due to
incorrect interpretation of another very common process. In its loco-
motion Amoeba frequently comes in contact with diatoms, desmids,
encysted Protozoa, etc. These it usually creeps over, so that they lie
beneath it. As the Amoeba progresses the objects come in contact w^ith
the posterior portion of the body, w^hich is raised from the bottom and
covered with a viscid secretion. Owing to this viscid substance the
objects often cling to the under surface of this part of the body and are
carried along with it. If observed at this time one cannot tell whether
these objects have been ingested or not. But as a result of the method
of movement of the Amoeba they gradually pass to the posterior end,
and are usually finally left behind. When such an object separates from
the Amoeba, becoming detached from its under surface, it appears ex-
actly as if it were given off from within ; it is only by observing the
whole process from beginning to end that one can be sure of its exact
nature. I am convinced that many of the supposed cases of the inges-
tion of food and of the ejection of food previously ingested at the poste-
rior end are to be explained in this way. In all the detailed descriptions
of food-taking in forms related to Afnoeba proteus that I have found in
literature, the food was taken at the anterior end in a way similar to
that which I have described above (see Carter, 1S63, p. 45 ; Wallich,
1863, c, p. 453 ; Leidy, 1879, p. 49 ; Le Dantec, 1894, p. 6% ; also Biit-
schli, 1880, p. 117).

In Amoeba verrucosa and the other species which do not often form
pseudopodia food is taken in a somewhat different manner. Food-
taking in Amoeba verrucosa has been described by Rhumbler (1898,
p. 205). Penard (1902), though he spent much time studying this
species, says that he has not observed the taking of food, and thinks it
must occur only rarely. In my own cultures specimens of this species
taking food were positively abundant. I have seen the process much
oftener than in other species. In A. verrucosa and its relatives food-
taking is greatly aided by the tendency of foreign particles to cling to
the surface of the body, a tendency which we found so convenient for
determining the movement of points on the body surface (see pp. 140-
146). This adhesiveness of the outer surface compensates for the lack of
formation of pseudopodia in these species. The outer surface gradually
sinks in at the point where the food body is attached to it. The latter
is thus carried to the inside of the body, surrounded by a pouch of
ectosarc. This pouch becomes separated from the outer ectosarc. The
food is thus completely enveloped and later digested. Not only large
objects, but often very small ones, spores of algae, small diatoms, flagel-
lates, etc., are taken in in this way. Rhumbler (/. c, p. 20S) has

196

THE BEHAVIOR OF LOWER ORGANISMS.

given a very interesting account of the rolling up and ingestion of threads
of Osciliaria by this species (see p. 223).*

PURSUIT OF FOOD.

Amoeba froteus does not always succeed in ingesting its food so
easily as in the case just described (p. 194). There is, as noted above,
a tendency for the food body to be pushed away by the forward move-
ment at the anterior end of the Amoeba, and this sometimes gives
serious difficulty. In such cases Amceba may show what would be
called in higher organisms remarkable pertinacity in continuing its

Fig. 73.t

attempts to ingest the food. This will be illustrated from a concrete
case (Fig. 73).

An Amoeba froteus was creeping toward an encysted Euglena.
The latter was perfectly spherical and very easily moved, so that when
the anterior edge of the Amoeba came in contact with it the cyst merely
moved forward a little and slipped to one side (the left). The Amoeba

♦Leidy C1879, p. 86) gives a very similar account of the ingestion of filaments
of algae in Dinamoeba.

tFiG. 73. — Amoeba following a rolling Euglena cyst. Nos. 1-9 shovr successive
positions occupied by Amoeba and cyst. See text for explanation.

I

THE MOVEMENTS AND REACTIONS OF AMCEBA. I97

thereupon altered its course so as to follow the cyst (Fig. 73, i). The
cyst was shoved forward again and again, a little to the left ; the
Amoeba continued to follow. This continued until the two had tra-
versed about one-fourth the circumference of a circle ; then (at 3) the
cyst, when pushed forward, rolled to the left quite out of contact with the
Amoeba. The latter then continued forward with broad anterior edge
in a direction which would have taken it past the cyst. But a small
pseudopodium on its left side came in contact with the cyst. The
Amoeba thereupon turned again and followed the rolling cyst. At
times it sent out two pseudopodia, one on each side of the cyst (as at
4), as if trying to inclose the latter, but the ball-like cyst rolled so easily
that this did not succeed. At other times a single very long, slender
pseudopodium was sent out, only the tip of which remained in contact
with the cyst (5). Then the body of the Amoeba was brought up from
the rear and the cyst pushed farther. This continued until the rolling
cyst and the following Amoeba had described almost a complete circle,
returning nearly to the point where the Amoeba had first come in con-
tact with the cyst. At this point, owing to the form of the anterior
end of the Amoeba (7) the cyst rolled to the right instead of to the left
as it was pushed forward. The Amoeba followed (8, 9). This new
path was continued for two or three times the length of the Amoeba.
The direction in which the ball was rolling would soon have brought
it against an impediment, and I thought it possible that the Amoeba
might succeed in ingesting it after all. But at this point one of those
troublesome disturbers of the peace in microscopic work, a ciliate infu-
sorian, came near and whisked the ball away in its ciliary current.
After the ball was carried away the Amoeba continued to follow in the
same direction for only a very short distance, about one-fifth its length,
then reversed its course and went elsewhere.

The movements of Amoeba are, of course, very slow, and the behavior
described required a considerable period of time — 10 or 15 minutes, I
should judge. The whole scene made really an extraordinary impres-
sion on the observer, and it is diflScult in describing it to refrain from
the use of words that imply a great deal of resemblance between Amoeba
and immensely higher organisms. One seems to see that the Amoeba is
trying to obtain this cyst for food, that it puts forth efforts to accom-
plish this in various ways, and that it shows remarkable pertinacity
in continuing its attempts to ingest the food when it meets with diffi-
culty. Indeed, the scene could be described in a much more vivid
and interesting way by the use of terms still more anthromorphic in
tendency.

I have seen a large number of cases like that above described ; in
some of my cultures containing many specimens of Amoeba proteus

98

THK BEHAVIOR OF LOWKR ORGANISMS.

and many Euglena cysts it was not at all rare to find the animals
engaged in thus following a rolling ball of food. I have made full
notes and sketches of a number of other cases, but they show nothing
diflferent in principle from that above described, so that it is not worth
while to enter into details. One further point is, however, worthy of
special note. Often a single pseudopodium comes in contact with such
a cyst and stretches out toward it, while the remainder of the Amoeba
continues on its course, away from the cyst. The pseudopodium in
contact then stretches out as far as possible, keeping in contact with
the cyst and often pushing it ahead (Fig. 74)1 until it is finally pulled
bodily away by the movements of the whole Amoeba. Apparently this
one pseudopodium reacts to the stimulus quite independently of the
remainder of the body. Again, two pseudopodia on opposite sides of
the body may each come in contact with a cyst. Each then stretches

Fig. 74.*

out, pulling a portion of the body with it, and follows its cyst, until the
body forms two lobes, connected only by a narrow isthmus. Finally,
one half succeeds in pulling the other away from the attachment to the
substratum, and the entire Amoeba follows the victorious pseudopodium.
Mechanical stimuli are, of course, involved in the above reactions ;
perhaps also chemical stimuli from the cyst. It is important from the
theoretical standpoint to note that the movement of particles on the
surface of the Amoeba is toward the object causing the reaction. This
I have been so fortunate as to have opportunity of observing in several
cases.

OTHER AMCEB^ AS FOOD.

Amoebae frequently prey upon each other, as Leidy has already de-
scribed and figured (1879, p. 94; plate 7, Figs. 12-19). ^^t the victim
does not always conduct itself so passively as in the case described by
Leidy, and sometimes finally escapes from its pursuer. A description

♦Fig. 74. — A single pseudopodium {x) reacts positively to a Euglena cyst, the
protoplasm flowing in direction of cjst and pushing it forward, while remainder
of the Amoeba moves in another direction ; 1—4, successive forms taken. At 4 the
reacting pseudopodium is pulled awaj from the cjst, and then contracts.

THE MOVEMENTS AND REACTIONS OF AMCEBA.

199

of two or three concrete cases among those which I have observed in
Amoeba angulata will bring out the nature of the behavior under such
conditions. Penard (1902, p. 700) mentions that he has seen one
Amoeba pursue and finally capture another, but does not give a detailed
account of the process.

(i) Two Amoebae were observed, a large one and a small one, the
former apparently attempting to swallow the latter (Fig. "j^^). The
small Amoeba was creeping rapidly forward, while its wrinkled pos-
terior portion was enveloped by the anterior part of the larger Amoeba.
The large Amoeba had the anterior portion of its body quite hollowed
out, so as to form a cavity large enough to contain the entire small

1 _

2

Fig. 75.*

Amoeba, and in the anterior portion of this cavity was inclosed the
hinder portion of the body of the smaller Amoeba. Whether this
cavity was bounded below as well as above and at the sides by pro-
toplasm I could not determine with certainty. The large Amoeba was
following the small one, moving at about the same rate. There was
no union between the protoplasm of the two ; on the contrary the
boundaries of both were clearly defined, and they seemed to be only
slightly in contact, the posterior end of the small specimen moving
easily within the cavity of the other. As they moved forward, some-
times the posterior specimen flowed a little faster, and then a little
more of the smaller one became enveloped ; at other times the smaller
Amoeba moved a little faster, and then withdrew a part of its inclosed

* Fig. 75. — Pursuit of one Amoeba by another. See text for explanation.

200 THE BEHAVIOR OF LOWFiR ORGANISMS.

tail. They proo^ressed in this fashion for a long distance, many times
their own length. I watched them thus for more than lo minutes.
The smaller, anterior specimen frequently altered its course ; the
posterior one followed. T stimulated the anterior end of the small
specimen with the tip of a glass rod (Fig. 75, 3) ; it turned at a right
angle, and the posterior specimen followed. After about 12 minutes it
could be seen that the smaller specimen was moving slightly faster
than the other and was slowly withdrawing its posterior end. Finally
it pulled completely away from the large Amoeba, which was still fol-
lowing as rapidly as possible. After the small Amoeba had completely
escaped the large one stopped and remained entirely quiet for a few
seconds. The large cavity in its anterior portion, which it had pre-
pared for the reception of the small Amoeba, and which extended
back behind the middle of the body, was still very evident. After a
time the Amoeba began to change form and sent out pseudopodia
irregularly in all directions. The smaller Amoeba continued its for-
ward locomotion as long as observed. The performance is illustrated
in Fig. 75, from sketches made while it was in progress.

(2) In a second case I was able to observe the beginning as well as
the end of this microscopical drama (Fig. 76). I had attempted to
cut an Amoeba in two with the tip of a glass rod, in the manner described
later. The posterior third of the Amoeba, in the form of a wrinkled
ball, remained attached to the body only by a slender cord, the remains
of the ectosarc. The Amoeba began to creep away, dragging with it
this ball. I will call this Amoeba a^ while the ball will be designated b.
A larger Amoeba {c) approached, moving at right angles to the path
of the first Amoeba ; its course accidentally brought it into contact
with the ball /$, which was dragging past its front. Amoeba c there-
upon turned, followed Amoeba a, and began to engulf the ball b, A
large cavity was formed in the anterior end of Amoeba c, reaching back
nearly or quite to its middle, and much more than sufficient to contain
the ball b. Amoeba a now turned into a new path ; Amoeba c followed
(Fig. 76 at 4) . After the pursuit had lasted for some time the ball b
had become completely enveloped by Amoeba c; the cord connecting
it with Amoeba a broke, and the latter went on its way (at 5) and dis-
appears from our account. Now the anterior opening of the cavity in
Amoeba c became partly closed, leaving a slender canal (5) . The ball b
was thus completely inclosed, together with a quantity of water.
There was no union or adhesion of the protoplasm of b and c; on the
contrary (as the sequel will show clearly) both remained quite sepa-
rate, c merely inclosing b.

Now the large Amoeba c stopped, then began to move in another
direction (Fig. 76, 5-6), carrying with it its meal. But the meal, the

THE MOVEMENTS AND REACTIONS OF AMCEBA.

20I

ball 3, now began to show signs of life, sent out pseudopodia, and,
indeed, became very active. We shall henceforth, therefore, speak of it
as Amoeba b. It began to creep out through the still open canal, send-
ing forth its pseudopodia to the outside (Fig. 76, 7). Thereupon
Amoeba c sent forth its pseudopodia in the same direction, and after
creeping in that direction several times its own length, again completely

Fig. 76.*

inclosed b (7-8). The latter again partly escaped (9), and was again
engulfed completely (10). Amoeba c now started again in the opposite
direction (i i), whereupon Amoeba 3, by a few rapid movements, escaped
entirely from the posterior end of c, and was free, being completely
separated from c (11-12). Thereupon c reversed its course (12), crept
up to 3, engulfed it completely again (13), and started away. Amoeba b

*FiG. 76. — Pursuit, capture, and ingestion of one Amoeba bv another; escape
of captured AmcBba and its recapture ; final escape. See text for detailed account.

202 THE BEHAVIOR OF LOWER ORGANISMS.

now contracted into a ball, its protoplasm clearly set off from the pro-
toplasm of its captor, and remained quiet for a time. Appai'ently the
drama was over. Amoeba c went on its way for about five minutes,
without any sign of life in b. In the movements of the Amoeba c the
ball b gradually became transferred to the posterior end of c, until
finally there was only a thin layer between b and the outer water.
Now b began to move again, sent out pseudopodia to the outside
through the thin wall, and then passed bodily out into the water (14).
This time Amoeba c did not return and recapture b. The two Amoebse
moved in different directions and remained completely separated. The
whole performance occupied, I should judge, about 12 to 15 minutes
(the time was not taken till several minutes after the beginning).

After working with simple stimuli and getting always direct simple
responses, so that one begins to feel that he understands the behavior
of the animal, it is somewhat bewildering to become a spectator of so
striking and complicated a drama. If we attempt an analysis of the
observed behavior of the Amoeba c into stimuli and reactions, we ob-
tain some such a result as follows : At first the stimulus of contact with
^, and perhaps a chemical stimulus from the same source, causes the
Amoeba c to react by flowing toward ^, and at the same time to change
form, so as to hollow out the anterior end. Later, every change in the
direction of movement of a and b induces a corresponding change in
the direction of movement of c/ there is a finely co-ordinated adapta-
tion of the latter to the movements of the former. After the separation
of b from ^, the movement of c (at 5-6) in a different direction may
have been due to some external stimulus. But what is the stimulus for
the change of direction of locomotion in the Amoeba c at 7 when b has
begun to escape .'' And why does Amoeba c go in that direction only
long enough to get b firmly inclosed again, then reverse its course .^^
And, finally, why does Amoeba c reverse its course at 1 1-13, when b has
entirely escaped, and continue in this reversed direction till it reaches
and recaptures b? The action is remarkably like that of a higher
animal. Doubtless we must assume chemical and mechanical stimuli
as directives for each of the movements of c, but the analysis so obtained
seems not very complete or satisfactory.

REACTIONS TO INJURIES.

Certain cases that belong under the heading of reactions to injuries
have already been described as evidences of the contractility of the
ectosarc; for these page 180 should be consulted. The cases which
we take up here are of a different character. They concern Amoeba
angulaia.

Jensen (1896) has shown in the case of certain Foraminifera that
two pieces of protoplasm from the same individual will readily unite

THE MOVEMENTS AND REACTIONS OF AMCEBA.

JO3

if brought in contact, while pieces from different individuals will not
thus unite. I was interested in the question as to whether this would
hold true also for Amoeba, and for that purpose undertook to cut speci-
mens in two. With the fine tip of a glass rod it is possible, under the
microscope, in the open drop, to cut in two an elongated Amoeba at
almost any desired point. The sharp point of the rod is simply drawn
across the Amoeba as it lies outspread on the substratum.

In this operation it was found that the Amoeba was not, as a rule, at
first completely cut in two by the stroke itself. The endosarc is divided
completely, but the two halves are still connected by a thin layer of
ectosarc, which resists the cutting, and shows fine longitudinal stria-
tions ; these may be merely longitudinal folds (Fig. 77, 2). This thin
layer of ectosarc seems very tenacious.

Fig. 77.*

The Amoeba is thus left in the condition shown in Fig. 77, 2. The
two halves usually both contract strongly. Now ensues a very pecu-
liar process. One of the two halves begins to send out pseudopodia
in such a way as to partly inclose the other (3) ; the second half is
thus drawn as a narrow wedge-shaped mass inside of the other, as at 4.
It seems to be usually the half that contains the nucleus that envelopes
the other, though, as will be shown later, the nucleus is not necessary
for this reaction. If the piece thus embraced is considerably smaller
than the other, it may become completely inclosed, and is then carried
away, appearing like a mass of food. It does not become fused with
the remainder of the protoplasm, but there is a sharp boundary between
it and that which envelopes it. Specimens were followed for 10 min-
utes after thus inclosing a piece of their own bodies ; during this time
no marked change was seen to occur in the inclosed piece.

In the much more common cases where the two pieces are nearly

204 THE BEHAVIOR OF LOWER ORGANISMS.

equal in size, or when it is the smaller piece that begins to envelop
the larger, the process results differently. After one piece has been
drawn far into the other (Fig. 77, 4), both seem to contract strongly,
whereupon the connecting band of ectosarc breaks, the partly inclosed
piece is squeezed out of the other ; and the two separate. Usually
each retains its form for a few seconds after separation ; one bearing
a slender truncate pyramid or cone, while the other shows a deep
depression corresponding to this pyramid (Fig 77, 5). After a time
both halves change form and move away. Usually the half which
partly inclosed the other becomes active long before the other, but
this is not invariably true.

In a large number of cases observed it was the part which contained
the nucleus that attempted to envelop the other half. In order to
determine whether the nucleus plays a necessary part in this perform-
ance, I tried the following experiment: After the half which had
no nucleus had again become active and was moving about, I cut it in
two, as before. Now one half of this piece partly enveloped the other
in the usual way, thus showing that the nucleus is not necessary for
this reaction.

These results should be compared with Penard's observations on
injured specimens of A, terricola^ noted on page 180 of the present
paper.

As to the question which these experiments were originally intended
to answer, whether two pieces of a single Amoeba would reunite
after separation, my results were negative. After the two pieces had
begun to move about freely they were induced to come in contact, or
sometimes they came in contact accidentally ; but in no case was there
any union. Prowazek (1901, p. 93) obtained the same result with
small species of Amoeba, but in a larger undetermined species he suc-
ceeded in bringing about a union of pieces not only from the same
individual, but from different individuals.

PHYSICAL THEORIES AND PHYSICAL IMITATIONS OF
AMCEBOID MOVEMENTS.

THE SURFACE TENSION THEORY.

The movements of Amoeba as presented by Biitschli (18S0, 1893)
and Rhumbler (1S98) (see Figs. 30-33) are exactly those of a drop of
fluid moving as a result of a local change in surface tension (Fig. 34).
It was, therefore, natural to assume that the cause of the movements is
the same in the two cases. This is the view taken by the two authors
named. According to Biitschli, Rhumbler, and many other authors,
Amoeba is a drop of complex fluid which moves about as a result of
local changes in surface tension.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 205

In the foregoing investigation it has been shown that the movements
in Amoeba are not of the character supposed by Biitschli and Rhumbler.
There is, indeed, very little resemblance betw^een the movements of
Amoeba and those of an inorganic drop moving as a result of a local
change in surface tension. The difference is clearly brought out by a
comparison of Fig. 58, showing the currents in Amoeba, with Fig. 34,
showing those in the inorganic drop. The more striking differences
are as follows :

(i) In the drop moving as a result of a local change in surface ten-
sion the currents on the surface are (and must be) away from the side
on which a projection is formed and toward which the drop is moving ;
in the Amoeba the surface current is toward this side.

(2) In the drop the surface currents are in a direction opposed to
that of the axial current ; in Amoeba surface currents and axial current
are in the same direction.

(3) The movement of Amoeba is in the nature of rolling, the upper
surface passing continually around the anterior end and becoming the
lower surface. In the inorganic drop there is no such rolling move-
ment, but the interior portions of the fluid are continually passing to
the surface at the anterior end.

Clearly, then, the forces producing the movements in the two cases
are not acting in the same manner. The locomotion of Amoeba is
demonstrably not due to a local decrease in surface tension on the side
toward which the animal is moving.

This becomes still clearer when we consider in detail the method by
which the movements are produced in a drop of inorganic fluid as a
result of a local decrease in its surface tension.

The phenomena of surface tension are usually considered to be the
result of the uncompensated attractions of those particles of the fluid
which are near to the surface. Such particles are attracted inward and
laterally, but not outward (or less strongly outward) . The resulting
forces may be considered as resolvable into two components, one acting
tangent to the surface, the other acting perpendicular to the surface.
The former is what may be called surface tension proper ; the latter
is often spoken of as normal pressure. The result is very much as if
the fluid were covered with a stretched India rubber membrane.
These relations are well set forth in a recent paper of Jensen (1901).
It is further to be noted that these two components, surface tension and
normal pressure, are two aspects of one and the same thing, and, there-
fore, vary together and from the same causes ; they can not be separated
either theoretically or experimentally. Whenever one of these factors
increases or decreases, the other shows a corresponding change. The
two are often spoken of (together with the pressure due to curvature of
the surface) as surface tension (see Jensen, /. c).

2o6 THE BEHAVIOR OF LOWER ORGANISMS.

Now, when the interattraction of the particles at a certain region of
the surface of a mass of fluid is decreased, the pressure inward and the
tension along the surface are decreased. As the pressure remains the
same elsewhere, fluid tends to be pressed out at the point where the
pressure is lowered ; thus a projection may be formed here. As the
tension remains the same elsewhere, the remainder of the surface of
the drop pulls harder than that of the region under consideration ;
hence it pulls the surface of the fluid away from the region where the
tension is lowered. The effect is similar to that which would be
produced if one portion of a stretched sheet of India rubber w^ere
weakened or cut ; the remainder of the sheet would pull away from this
region. Thus there are produced the currents characteristic of such a
drop of fluid — an axial current toward the region of lowered tension,
surface currents away from the region of lowered tension (Fig. 34).
An increase in the tension at the opposite side would produce exactly
the same currents, as Rhumbler (1898, p. 188) has set forth, the axial
current being always toward the region of lowest tension, the surface
currents in the opposite direction.

In the moving Amoeba, as we have seen, the currents are by no
means of this character. The axial current and the surface current are
congruent, and both are in the direction of locomotion. Such move-
ment could not be produced by a local decrease in the surface tension
of some part of the body surface.

The formation of pseudopodia is, as we have seen, essentially the
same process as the forward movement at the anterior end of the
AmcEba. On the upper surface of a pseudopodium that is in contact
with the substratum there is a forward movement, so that particles
clinging to the upper surface are carried over the tip ; the currents
which must result from a decrease in surface tension are not present.
On the contrary, there is a current on the surface in the opposite direc-
tion from that required. The formation of such a pseudopodium can
not, then, be due to a local decrease in surface tension.

The same is true, essentially, when a pseudopodium is sent out into
the water, not coming in contact with a surface. In such a case, as
we have seen, the entire surface moves outward, in the same direction
as the tip ; there is no such backward movement as the theory requires.

Altogether, it is clear that the supposed resemblance between the
movements and internal currents of Amoeba and those of a drop of fluid
moving as a result of a local increase or decrease of surface tension
does not exist. We must conclude that the movements of Amoeba are
not due to local changes in surface tension.

One might be tempted to inquire whether the movement of Amoeba
could not be explained by considering separately the action of the two

THE MOVEMENTS AND REACTIONS OF AMCEBA. 207

factors in surface tension, the "surface tension proper" and the
" normal pressure." If the normal pressure, directed inward, were
decreased in a certain region, while the tangential factor, the "surface
tension proper," were not decreased, were perhaps even increased,
could not pseudopodia be formed as actually occurs, without any back-
ward current on the surface ? Jensen seems to lean toward the possi-
bility of such action when he speaks of the variation of one of these
factors without the other (Jensen 1901, p. 374).*

But with such an inquiry should we hot leave the field of realities to
wander among abstractions ? One who is not a physicist can, of course,
not speak positively on such a point. Yet, so far as I am able to dis-
cover from the results of experiments and from the theories of surface
tension, the state of the case is about as follows : The tangential tension
and the normal pressure are not two different things ; they are only
different aspects of one and the same thing. Viewed from the stand-
point of " energetics," what the physical experiments show liquids to
possess is surface energy, in virtue of which they tend to decrease their
surface and resist an increase of surface, however these changes are
brought about (see Ostwald, 1902, p. 197). When the " surface ten-
sion is decreased " in a certain region {x) of a fluid mass, this signifies that
the tendency to a decrease of surface and the resistance to an increase
of surface is lessened in this region. As a result, the remainder (j/) of
the fluid decreases its surface at the expense of the region x; the latter
is thus compelled to increase its surface. This takes place by simulta-
neous passage of the contents of y into x and of the surface of at on to
y, the two operations being essentially one, and both having the result
of decreasing the surface of y and increasing that of x. There would
seem to be no ground, theoretical or experimental, for supposing that
in a fluid one of these operations could take place without the other.
An attempted explanation of this sort would be, if these considerations
are correct, not a physical explanation, but a purely hypothetical one,
working with conditions not known to exist. The whole value of the
surface tension theory lies in its direct reference back to the results
of physical experiments — in its fidelity to the results of such experi-
ments. As soon as it leaves this ground it becomes of no more value
than the thousand and one other hypotheses that have been constructed
for the explanation of contractility.

Further, even this purely hypothetical explanation could not account
for the forward currents on the upper surface of Amoeba, nor for the
transference of portions of the body surface to the surface of a pseudo-
podium. In any form we can give it, the theory that the movement is
due to local changes in surface tension is not in agreement with the
observed phenomena.

* Though elsewhere he speaks of the necessity of their varying together.

J08 THE BEHAVIOR OF LOWER ORGANISMS.

BERTHOLD's THEORY THAT ONE-SIDED ADHERENCE TO THE SUB-
STRATUM IS THE CAUSE OF LOCOMOTION.

If, then, the movements of Amoeba are not due to local decrease in sur-
face tension, with the formation of *'Ausbreitungscentren " (Biitschli),
is it possible to find a physical explanation for them ? In taking up
this question we must consider separately (i) locomotion, and (2) the
formation of pseudopodia.

Observation and experiment indicate, as we have seen in the obser-
vational portion of this paper, that Amoeba is a drop of fluid which
becomes attached to the substratum in front and pulls itself forward,
the pull extending backward from the attached region over the upper
surface, and producing a rolling motion.

Now, a drop of inorganic fluid under the influence of similar
forces moves in precisely the same manner. There is no great difli-
culty in causing a drop of inorganic fluid to adhere more strongly to
the substratum on one side than elsewhere. When this is brought
about the drop moves toward the more adherent side by a rolling
motion, precisely like that of Amoeba. By a proper arrangement of
the conditions almost every detail of amoeboid locomotion may be
closely imitated.

That this is the method of movement in Amoeba was the theory
maintained by Berthold (1886), though it is rather curious that the
supposed facts on which he based this view were incorrect. Berthold
confirmed on the basis of observations on A?noeba verrucosa ( ! ) and
other species the account of the currents in Amoeba given by Schulze
(see p. 137 and Fig. 37) ; that is, such currents as would be consistent
with the theory of local decrease in surface tension, but are quite incon-
sistent with his own theory. He rejected the theory that locomotion is
due to a decrease in surface tension at the anterior end, on the ground
that no currents are to be observed in the surrounding water, as this
theory demands. Berthold held that the locomotion is due to the
spreading out of the anterior end of the fluid mass on the surface of a
solid, this spreading out being due to adhesion between the fluid and
the solid. Unfortunately for the understanding of his theory, he tried
to bring this into relation with many other much less simple phenom-
ena. In particular he compared the movements to those of a drop of
water on a glass plate, which flees when a rod wet with ether is brought
near one side. This was an unfortunate comparison, as the movements
in a drop of water under such circumstances are of a character entirely
different from those produced when a mass of fluid adheres by one side
to a solid. The movement in a drop of water fleeing from the ether-
ized rod is a result of the currents produced by the lowering of the

THE MOVEMENTS AND REACTIONS OF AMCEBA. 2O9

surface tension on one side, as Biitschli (1892, pp. 191 and 194), has
shown. This comparison, and Berthold's discussion of the relations
between spreading out on the surface of solids and on the surface of
liquids, together with his incorrect idea of the currents in such spread-
ing out on a solid, have served to distract the attention of investigators
from the really simple essential features of such a theory. Berthold
did not attempt to study directly the currents and other movements of
a drop of fluid moving as a result of one-sided adherence to a solid.
We may, therefore, leave his account and examine for ourselves the
phenomena in question.

EXPERIMENTAL IMITATION OF THE LOCOMOTION OF AMCEBA.

The experiments with inorganic fluids may be performed as follows :
A piece of smooth cardboard, such as the Bristol board used for drawing,
is placed on the level bottom of a shallow vessel, such as a Petrie dish,
and soaked with bone oil by spreading the latter over its surface. A
small area on the surface of the board is protected from the oil by
placing upon it a drop of water. After the board has become well
soaked, the drop of water is removed with a pipette, leaving this spot
merely damp, while a layer of oil some millimeters deep is poured into
the vessel, covering the cardboard completely. A drop of glycerine or
of water is then introduced; this settles to the bottom, but adheres to
it only slightly. A drop of glycerine is in some respects preferable, as
its movements are slower. To the drop should be added beforehand a
quantity of soot, in order to make its internal movements visible.
Some of the soot remains on the surface, projecting out into the oil,
thus making it possible to observe the surface currents.

If the drop is brought close to the spot on the cardboard that was
protected from the oil, so that one side comes in contact with this
region, the edge of the glycerine or water drop spreads out over this
area. Thereupon the remainder of the drop is pulled in that direction,
till the whole drop takes up its position over the protected spot. In
the movement of the drop toward the area to which one side adheres,
it rolls exactly as Amoeba does. The currents on the upper surface
and within the drop are forward. Toward the sides the currents are
somewhat less marked, and on the under surface they cease entirely ;
particles within the drop but in contact with the lower surface are not
moved at all. The forward current is most rapid in front, becoming
slower at the rear, exactly as in Amoeba. At the posterior end the
surface rolls upward ; particles on the surface which were at first on
the bottom may be seen to pass upward around the posterior end and
then forward, as in Amoeba. The form of the drop may become much
elongated ; the anterior edge is thin, the posterior end thick and
rounded. In all these respects the drop resembles the moving Amoeba.

no TIfE BEHAVIOR OF I.OWKB ORGANISMS.

By inclining the vessel the drop may be made to roll away from the
attractive spot ; then when the level is restored it moves back again.
By repeating this process the movements may be studied in detail.
P*or studying the movements in all parts except at the anterior edge
another more convenient method may be employed. A small piece of
wood may be brought against one side of the drop ; toward this it
moves in the manner just described. If the piece of wood is moved
continually in a certain direction, the drop follows, and its movements
may be examined with ease. In this case the anterior edge, of course,
is not thin and pressed against the surface, but otherwise the move-
ments are the same.

By proper modifications further details of the movement of Amccba
are exactly imitated. Thus a quantity of sand grains or other heavy
objects may be added to the drop. In the movement these collect at
the posterior end, as happens with the coarse internal contents in
Amoeba. A large, spherical bubble of oil may be introduced into the
drop, in imitation of the contractile vacuole ; this likewise stays near
the posterior end. When a considerable quantity of heavy material is
collected at the posterior end, the latter becomes drawn out into a sort
of pouch, which is dragged along, its substance not partaking of the
currents shown by the remainder of the drop. It thus plays the same
part as the well-known posterior appendage of Amoeba. Material
passes up from the bottom to the upper surface on each side of this
posterior pouch, just as happens in Amoeba (see p. i6S). Particles
clinging to the outer surface at the posterior end are often dragged
along for a considerable time, then finally pass upward to the upper
surface and so forward, exactly as described for Amoeba (p. 169).*

In another detail the movements of the drop of water or glycerine
are strikingly like those of Amoeba. As we have seen, the current is
most rapid at the anterior end in both cases, becoming as a rule slow
toward the rear. But I have pointed out that in Amoeba the move-
ments at the posterior end are not uniform. Sometimes the under sur-
face remains attached to the bottom longer than usual, then, when it
becomes detached over a considerable area at once, there is a sudden
rush of the fluid forward from the posterior region (p. 16S). Exactly
the same thing is to be observed in the inorganic drop. The bottom
is not uniform, so that sometimes the posterior end clings to it longer
than usual ; this end is then drawn out, and when it is finally released

*It may be worth while to state that these experiments on inorganic fluids
were performed after the work on the movements of Amoeba had been completed
and the description entirely written in the form given in the preceding pages.
No details of the movements of Amoeba were added after the behavior of the
inorganic drops had been studied^

THE MOVEMENTS AND REACTIONS OF AMCEBA. 211

there is a sudden rush forward of the internal fluid from this region,
giving the movement a jerky character.

Still another resemblance in detail between the movements of the
inorganic drop and of Amoeba may be noted at times. As we have
seen, in the posterior part of Amoeba that is detached from the bottom
there is a movement forward not only on the upper surface, but also a
slow movement on the lower surface ; the entire posterior region is
contracting. The same thing may be seen in the inorganic drop. The
phenomenon in question is not so regular here, because the posterior
half usually still clings to the surface to a certain extent, while in Amoeba
it is as a rule entirely free. But when the posterior half of the inor-
ganic drop does become entirely free, it is seen to contract as a whole,
with a forward movement on both upper and lower surfaces, exactly
as in Amoeba.

One may even see at times, under special conditions, a slight turning
backward of the current at the sides of the anterior end, such as has
been described by a number of authors for Amoeba (see p. 137). This
occurs when the drop is slender and elongated, and. the area on which
it spreads out is broad. On coming in contact with the area, the end
of the drop rushes forward and spreads out. If the whole width of the
area is not covered at first, some of the particles that have moved for-
ward curve outward and a little backward till the area is quite covered.

Altogether, the resemblance between the movements of the inor-
ganic drop and those of Amoeba is extraordinary, extending even to
details. What are the forces at work in such a drop, and in how far
may they be supposed to be active also in Amoeba.'*

The spreading out of the drop of glycerine or water at the anterior
end is due to its adherence here to the substratum. The remainder of
the movements of the inorganic drop are due to the interplay of surface
tension and adhesion to the substratum. As a result of surface tension
the drop seeks to regain its spherical form ; hence the posterior part is
pulled forward, the force required to accomplish this being less than
would be demanded for freeing the anterior edge from the substratum.
In the pulling forward of the posterior portion the adherence of the
lower surface to the bottom keeps this surface from moving ; hence the
upper surface moves forward while the lower surface remains quiet
or moves forward only very slowly ; the movement is thus converted
into a rolling motion. The details given above depend merely upon
the relative part played by adherence and surface tension, with the
resistance offered by the weight and inertia of particles inclosed in the
drop.

In Amoeba, so far as the evidence of observation goes, the conditions
are similar. Amoeba adheres to the substratum and spreads out in a

212 THE BEHAVIOR OF LOWER ORGANISMS.

similar manner. In one respect there is, of course, a striking differ-
ence between the two cases. Amoeba does not require that the sub-
stratum should be different on its two sides in order that there should
be movement ; on the contrary, it may move steadily in a certain direc-
tion on a uniform surface. The mechanism of the movement might,
nevertheless, be the same as in the inorganic drop. Chemically different
substances show different degrees of adherence to the same surface. It
may be supposed, therefore, that there is a chemical difference between
the anterior and posterior regions of Amoeba of such a nature that the
anterior region clings to the surface while the posterior region does not.
This chemical difference must, of course, be continually renewed, since
new parts of the body continually come in contact with the substratum.

It may, however, be questioned whether the adhesion of Amoeba to
the substratum is of the same character as the adhesion of a drop of
water to glass ; in other words, whether Amoeba really plays here the
part of a fluid, and " wets" the substratum. This was the view taken
by Berthold (1886) and, if I understand him correctly, Le Dantec
(1895). Apparently opposed to such a view is the fact that Amoeba
may creep on the under side of the surface film of water, as I have
often observed. This surface film is, of course, fluid ; if in adhesion
Amoeba itself also plays the part of a fluid, we should have two fluids
in contact, having the same relation of attraction or adhesion that a
fluid has for a solid that it " wets" ; that is, the particles of each fluid
have a greater attraction for those of the other fluid than for each other.
This, it would appear, could result only in the formation of diffusion
currents in the two fluids ; the two would mix. This result does not
follow, so that it would appear that in adhesion Amoeba does not play
simply the part of a fluid which wets the substratum. As we have
seen (p. 165), there is evidence that the adhesion takes place through
the mediation of a viscid secretion.

Whatever the nature of the adhesion, we know it exists at one pole
of the Amoeba and not at the other. Given such chemical difl!erences
between the two poles as would produce this difference in adhesion,
then locomotion would follow essentially as we find it to occur in
Amoeba Umax or A. verrucosa. No further properties except those
common to fluids would be required.* For the determination of the
direction and rate of locomotion, the distribution of these chemical
diflJerences would be the essential factor.

Caution is necessary, however, in transferring the results of these and
other similar experiments to Amoeba. The resemblances between the
movements of the inorganic drops and those of Amoeba show merely

*It will be noted that this statement is made for simple locomotion, and does
not refer to the formation of pseudopodia.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 213

that the forces acting upon the two have a similar localization and
direction, not necessarily that the forces themselves are identical. This
caution is emphasized by the fact that drops moving down an inclined
plane as a result of the action of gravity have a similar rolling motion.
This is well shown in the drops of glycerine or water on the oiled
surface, in the experiments just described. Most (though not all) of
the details mentioned above, in which the movements of the inorganic
drop resemble those of Amoeba, may be observed also in drops moving
under the influence of gravity. The essential difference between the
two sets of experiments is that the action of adhesion, pulling the drop
in a certain direction in the one case, is replaced by the action of
gravity, pulling in the same direction, in the other case. Correspond-
ing with this difference is the chief difference to be observed in the
movements of the drops under the different conditions. In those mov-
ing as a result of greater adhesion on one side, the anterior edge is
thin and flat, as in Amceba, while in those moving from the action of
gravity this is not true.

In Amoeba observation shows that we have the one-sided greater
adhesion, and the tendency of the lower surface to cling slightly to the
substratum, as in the first set of experiments with the inorganic drops.
There remains only something corresponding to the surface tension
factor, common to both sets of inorganic experiments, to be accounted
for in Amoeba. Since Amoeba acts like a fluid in many respects, there
is no a priori reason to deny it surface tension, and nothing further is
required to produce locomotion. To this there is, however, one objec-
tion. This is found in the roughening and wrinkling of the surface
at the posterior end as it contracts, and in the similar roughening of a
contracting pseudopodium (pp. i6o, i68) . This is exactly the opposite
of what should take place in a fluid contracting as a result of surface ten-
sion. In such a case the primary phenomenon is the decrease in sur-
face ; the latter should, therefore, remain perfectly smooth, and as small
as possible. Of course, surface tension might be replaced in Amoeba
by a specific property of contractility of some sort, having its seat a
little beneath the surface. Locomotion would then take place as in
the inorganic drop, and the wrinkling of the outer surface would be
accounted for. On the other hand, if we can account for the contrac-
tility by a known property of fluids, such as surface tension, our expla-
nation will, of course, be simpler and more probable. By taking into
consideration the apparent fact that the outer layer of Amoeba is partly
fluid, partly solid, I believe that such an explanation, accounting for
the roughening as well as the contractility, can be given ; this I shall
attempt in the next section of this paper (p. 215).

The formation of projections at the anterior edge or side of the inor-
ganic drop, comparable to the formation of pseudopodia in contact with

214 '^"^ BEHAVIOR OF LOWER ORGANISMS.

the substratum in Amoeba, may also be induced in the oil drops. For
this purpose it is necessary to produce a greater adhesion on a small
area at one side. A projection is at once sent out here. The move-
ment in sending out such a projection is the same as that to be observed
in the formation of a pseudopodium under such circumstances. The
projection is thinnest at the tip ; its upper surface moves forv^'ard and
rolls over at the point, while the lower surface is at rest.

FORMATION OF FREE PSKUDOPODIA.

On the other hand, the projection of free pseudopodia into the water
cannot be imitated under these conditions. As we have seen, the move-
ment in the formation of a free pseudopodium differs from that in form-
ing a pseudopodium along a surface merely in the fact that in the latter
case the contact surface is at rest, while in the free pseudopodium all
surfaces move equally, a given point on the surface remaining approx-
imately at the same distance from the tip. In the inorganic drop pro-
jections can indeed be formed in which the surface moves in exactly
the same manner as in the free pseudopodia of Amoeba, but under con-
ditions that are essentially different. If some small object to which the
fluid adheres, such as a sliver of wood, is brought into contact with one
side of the drop, the fluid flows out over it, and may form thus a long,
slender projection. The surface of this projection moves in the same
manner as the surface of a pseudopodium in Amoeba (p. 153). Thus
the surface of a free pseudopodium shows such movements as it would
if drawn out by an object to which it adheres at its tip. Since no such
object is present, it is clear that the formation of free pseudopodia is not
explicable in this manner.

As we have seen above, the locomotion and the formation of pseudo-
podia in contact with the substratum could, if they stood alone, be con-
sidered due to the adherence and spreading out of a fluid on a solid, as
maintained by Berthold (18S6). But they do not stand alone ; we have
the additional fact that pseudopodia may be sent out which are not in
contact with the substratum . The anterior edge of an Amoeba, further,
may be pushed out freely into the water as a single pseudopodium.
This may frequently be seen in Amoeba Umax. As a reaction to a
.stimulus the protoplasm may push upward freely as a thick lobe, till
the greater part of the substance is transferred upward and the Amoeba
topples over (p. 184) . In the formation of such a lobe, the protoplasm
may flow from both ends toward the middle, producing the '* heaping
up" from which the thick upward projection results (see p. 183).
Currents flowing in this manner could not possibly be produced by
adherence to the substratum. Again, in an advancing Afiioeba avgu-
lata^ short triangular pseudopodia are constantly pushed forward, some
in contact with the substratum, others not thus in contact, but raised a

THE MOVEMKVTS AND REACTIONS OF AMCEBA. 21^

little above it (Fig. 54). Some, which are at first free, later come in
contact. Clearly, Amoeba is able to perform all the activities con-
cerned in locomotion without adherence to a solid. The adherence
is only necessary that there may be a movement from place to place.
The case is quite parallel to that of higher organisms, where contact
with the substratum is necessary in order that progression may occur,
though all the movements concerned in locomotion may be performed
without such contact.

We are compelled to conclude, therefore, that in the advancing end of
an Amoeba or the projecting pseudopodium there is an active move-
ment of the protoplasm, of a sort which has not been physically
explained. This involves the general conclusion that no physical
explanation is at present possible of the locomotion and projection of
pseudopodia in Amoeba.

To account for the contraction of the posterior part of the body, on
the other hand, possibly the properties common to Amoeba with other
fluids are sufficient. If surface tension may be considered the cause of
the contraction of the posterior part of the body, it is notable that it
acts as a constant factor, tending always to decrease the surface as much
as postiible, not as a variable factor. In other words, there is no indi-
cation that local increase or decrease in surface tension (see note, p.
235) plays any part in the production of the movements, as is main-
tained in the prevailing theories. The part played by surface tension
is thus a very subordinate one.

EXPERIMENTAL IMITATION OF MOVEMENTS DUE TO LOCAL CON-
TRACTIONS OF THE ECTOSARC, AND OF THE ROUGHENING OF
THE ECTOSARC IN CONTRACTION.

Besides the sending out of pseudopodia, there are certain other phe-
nomena in the movements of Amceba for which we lack, so far as I am
aware, any attempt at a physical explanation. These are the swinging
movements, vibrations, and local contractions of pseudopodia, described
on pages 1 77-1 79, and the roughening of the ectosarc in the contraction
of pseudopodia or other parts of the body (p. 16S).

These phenomena are in certain details so similar to some that I
have observed in inorganic fluids that I believe it worth while to
analyze the latter ; possibly they give an indication of the direction in
which an explanation of the phenomena in Amoeba above mentioned
nay lie.*

♦In view of the repeated failures of physical explanations in attempting to
account for vital phenomena, one does not approach a new attempt of this sort
with great confidence. Yet it is desirable that any possibility of this kind should
be worked out and submitted to criticism, in order that its truth or lack of truth
mav be demonstrated.

2l6 THE BEHAVIOR OF LOWER ORGANISMS.

Swinging or bending movements take place with special frequency
while the pseudopodia are withdrawing ; in some Amoebae such move-
ments are an almost constant accompaniment of withdrawal. As it is
withdrawn the pseudopodium becomes roughened or warty on its sur-
face, as we have seen, and at the same time bends to one side or the
other, or swings back and forth. The impression given is that the
outer layer of the pseudopodium has become partially solid. In with-
drawing, the solid substance seems to melt gradually away, in a some-
what irregular manner, so as to leave solid masses connected by liquid
protoplasm, the projecting solid masses forming the wart-like roughen-
ings of the surface. When this melting away occurs more strongly on
one side, the pseudopodium bends at that point, toward the side which
has apparently become more fluid.

We have in such a case, if appearances may be trusted, a mass com-
posed partly of solid, partly of fluid. While it is usually admitted that
parts of the protoplasm may become solid at times, little attempt has
been made to understand protoplasmic movements by studying the
physics of such mixtures of solids and fluids.* In certain experiments
with inorganic mixtures of this kind, in which movements were pro-
duced that resembled those just referred to in Amoeba, the writer
became convinced of the possible importance of the physics of such
mixtures for the understanding of protoplasmic activities.

The experiments in question were concerned with the movements
under the action of surface tension of oil drops to which soot had been
added for the purpose of rendering the currents visible. When a large
quantity of soot was added, the drops became somewhat stiffened, and
now showed to a marked degree a mingling of the characteristic proper-
ties of fluids and solids.

In one set of experiments clove oil was thus mixed with soot and
introduced as drops into a mixture of three parts glycerine and one part
95 per cent alcohol. The drops move about, as a result of local
decrease in surface tension, in the same manner as the olive-oil emul-
sion in Biitschli's celebrated experiments. Much of the soot collects
next to the surface of the drop, and becomes massed in certain regions,
as a result of the currents, covering these regions with a sort of crust,
this crust being formed of separate solid particles. The particles are
crowded together as closely as possible, owing to the surface tension of
the fluid in which they are floating-t If the particles are not too minute
they may project above the surface of the drop, giving it a rough appear-

♦Some of the experiments of Rhumbler (1898, 1902) deal with such mixtures,
though not with a view to an understanding of the movements, but of certain
other processes.

t According to the principles set forth by Rhumbler, 1898, p. 332.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 21 7

ance, similar to that of the withdrawing pseudopodium of Amoeba.
Such parts of drops or whole drops, so covered, show peculiar proper-
ties. The form may be changed by lowering the surface tension locally
or by mechanical action from outside, exactly as in a fluid, but there
is an inclination to hold a form once received. The tendency to take
the spherical form is still somewhat marked, and if the irregular drop
is strongly disturbed it frequently slowly becomes spherical. On the
other hand, if not strongly disturbed,, it may retain almost any form
impressed upon it — cylindrical, flattened, irregular, or with long, slender
projections. In this power of receiving and retaining an irregular form,
yet with a tendency to become spherical, these drops, of course, resem-
ble Amoeba. These properties vary with the amount of soot present in
the oil ; if this is less, the drops slowly return to the spherical shape
when deformed ; if greater, they retain the irregular shape indefinitely.
Such irregular masses nevertheless flow together if brought in contact,
will quickly gather together into a close mass if strongly deformed,
and in many other ways they show the characteristics of fluids.

The reason for their tendency to retain irregular forms is obvious.
The surface is covered with small solid particles that are in contact.
The projection of these particles above the surface may cause a rough-
ening of the surface. Any change of form, such as surface tension
would produce, causes much friction between these particles. The
form taken is, then, a resultant of the action of surface tension and the
resistance of these particles to movements.* Similar forms are pro-
ducible by mixing soot with bone oil and studying drops of the mixture
in a vessel of glycerine.

For our purpose the phenomena which occur when the soot particles
are unequally distributed are of special interest. Consider an elon-
gated projection, as in Fig. 7S, A, with the surface entirely covered
with closely crowded soot particles except in a certain region, x-y, on
one side. In this region x-y surface tension will have free play, tend-
ing to draw the points x and y together, while elsewhere the tendency
to contraction will be resisted by the friction of the particles. The
result is that the points x-y are drawn together, and the projection
bends toward that side. Since this bending does not tend to crowd
the particles on other parts of the surface closer together they do not
resist it.

In the oil drops mixed with soot the bending of projections in the
manner described is often to be observed under the appropriate condi-
tions. They should be compared with the bending of pseudopodia as

* Similar forms of fluids have been produced by Rhumbler in a parallel man-
ner in his imitations of the formation of Difllugia shells (1898, p. 287) and of
the shapes of the shells of Foraminifera (1902, p. 265).

2l8 THK BKHAVIOR OF LOWER ORGANISMS.

described for Amoeba (p. 177, and Fig. 62, a) . The chief experimental
difficulty in producing such bending is to arrange the conditions in
such a way that there is less soot on one side of a projection than on
the other. This occurs somewhat frequently when two irregular drops
are allowed to fuse, or when a drop is mechanically deformed with a
rod, as described in the next paragraph. One can sometimes bring it
about by placing with the capillary pipette a minute drop of oil on one
side of a projection, though this method is not as a rule very effective.
*^, - Such drops may also show another prop-

, erty in common with Amoeba, namely,
elasticity of form, or a phenomenon pro-
ducing similar results. Consider a curved
projection, as in Fig. 78, B. It is com-
P ^^^ pletely covered with soot particles and re-

tains its form in virtue of their resistance to a
change in position. Now, suppose we forcibly straighten out the pro-
section by pushing it to one side with a rod. By so doing the side a-6
is lengthened ; the soot particles on this side are, therefore, separated,
leaving certain areas of free fluid surface. When the projection is
released, surface tension can act on these areas, and the projection is
drawn back at once to its original form. I have often observed such
immediate returns to the original form after bending a projection of
one of the oil drops.

In Amoeba we have exactly the conditions most favorable for the
production of movements of this sort, and we actually find numerous
movements of just this character. It is generally admitted that the
outer layer becomes partially solidified ; as a pseudopodium is with-
drawn the solid portions evidently become liquefied in an irregular
way, some of them projecting above the surface and making it rough.
If the liquid substance produced shows surface tension, the movements
described must follow in the manner set forth above. It seems possible
that many of the observed movements are thus produced by local lique-
faction, with the intervention of surface tension, in the liquefied area.
In view of the apparently unlimited possibilities of partial solidifica-
tion and liquefaction in the protoplasmic body, with the resulting varied
action of surface tension, shall we not go a step farther and inquire
whether there may not be an outlook for an explanation of vibratory
movements, such as we find in flagella, along this line.? In an elon-
gated structure like a flagellum, a limited liquefaction of one side would
result in a bending toward this side. By regular alternation of lique-
factions in different regions, a regular vibration could be produced.
Th^ chief difficulty in the way of such a theory would seem to lie in

*FiG, 78. — Diagrams illustrating phenomena in mixtures of oil and soot.

THE MOVEXfENTS AND REACTIONS OF AMCEBA. 2lC)

the restoration of the original length in a given side after liquefaction
and consequent contraction had occurred. This could, perhaps, be
brought about by an elastic rod in the axis of the structure, such as
many cilia and flagella are known to possess.

The above is merely a suggestion made tentatively ; its justification
as a suggestion lies in the following facts: (i) The swinging move-
ment of pseudopodia in Amoeba in some cases strikingly resembles
movements of the character above set forth in inorganic fluids, and
precisely the conditions for such movements are present in Amoeba ;
(2) Swinging movements of pseudopodia seem to grade almost insen-
sibly into the vibratory movements of flagella.

DIRECT OR INDIRECT ACTION OF EXTERNAL AGENTS IN
MODIFYING THE MOVEMENTS.

Is the efl*ect of external agents in modifying the movements of Amoeba
due to the direct physical action of the agent on that part of the fluid
substance with which it comes in contact? Or is its action indirect,
in that it serves merely as a stimulus to certain internal changes, the
latter bringing about the modifications in the behavior? Both views
find adherents. The difference between them is fundamental, for they
lead to essentially different conceptions as to the nature of behavior in
these lower organisms.

In higher animals we know that the movements and changes of
movement are not produced in a direct way, but the effect of external
agents is to cause internal alterations which result in changes of move-
ment. It is not, therefore, possible to predict the movements of the
organism from a knowledge of the direct physical changes produced in
its substance by the agent in question. If the theory of direct action is
correct for Amoeba, we have in these animals a condition of affairs incom-
parably simpler, for here we can resolve the behavior directly into its
physical factors. If, on the other hand, the theory of indirect action is
correct, then there appears to be nothing fundamentally different in
principle between the behavior of Amoeba and that of higher organisms.

How the form and movement of a fluid mass might be determined
by the direct action of external agents on its surface may be simply
illustrated by certain experiments which I have described elsewhere
(Jennings, 1902). A mixture of 2 parts glycerine and i part 95 per
cent alcohol is placed on a slide and covered with a cover glass sup-
ported by glass rods. Into this is introduced with a capillary pipette
a drop of clove oil. The clove-oil drop, at first circular in form, soon
changes shape, shows internal currents, sends out projections in various
directions, moves about from place to place, and may divide into two
drops. The alcohol, not being uniformly distributed throughout the
glycerine, acts more strongly on some parts of the clove-oil drop than

220 THE BEHAVIOR OF LOWER ORGANISMS.

on others, thus lowering the surface tension in the region acted upon.
Thereupon the drop sends out a projection on the side affected, and may
follow this up by moving in that direction. We may cause the drop to
send out projections and move in a certain direction by placing a
minute drop of alcohol near one side, or by heating one side ; move-
ment takes place toward the side affected. These agents act directly
on the surface of the drop, lowering the tension ; the movements are a
direct consequence of this change. Parallel phenomena may be pro-
duced, as Bernstein (1900) has shown, with a drop of quicksilver. If
such a drop is placed in 10 per cent nitric acid, in which bichromate
of potash has been dissolved, the drop changes form and moves about.
If we place near such a drop, in vessel of 10 per cent nitric acid, a
crystal of potassium bichromate, the mercury drop moves rapidly over
to the crystal. Here, again, the chemical acts directly on the mercury,
lowering the surface tension at the region where it comes in contact
with it, thus producing the movement.

Certain authors have held that the movements of Amoeba are pro-
duced in this way by the direct action of external agents decreasing
or mcreasing the surface tension of certain parts of the fluid mass. As
an example of the theory of direct action of external agents in control-
ling the behavior, we may take the view of the reaction of Amoeba to
chemicals recently given by Rhumbler (1902, p. 384). According to
Rhumbler, it is evident that when an Amoeba moves toward or away
from a certain chemical, the side directed toward the chemical has, in
the first case, a lessened surface tension, in the second case an increased
surface tension, as compared with the remainder of the body.

The necessary differences of tension on the positive and negative sides may be
easily understood from our present standpoint, by holding that a positively
acting chemical decreases the surface tension both in the living alveolar system
of the cell and especially on the cell surface, upon which it must work most
strongly; that a negatively acting chemical, on the other hand, produces an
increase of surface tension in the alveoli of the cell, and especially, again, on the
cell surface; this increase is the greater, and from a physical standpoint must be
the greater, the more the molecules of the chemical affect or modify the tension
of the different cell alveoli or different parts of the cell surface (/. c, p. 384).

The explanation of thermotaxis and electrotaxis would be, according
to Rhumbler, ''exactly the same as for chemotaxis" (/. c, p. 385) ;
thus also as a result of the direct action of external agents. A fuller
explanation of the " tropisms " on this basis is given by Rhumbler in
an earlier paper (1898, pp. 183, 188).*

♦ Rhumbler emphasizes in the paper just cited (1898, p. 184) the importance
of *' inner disposition " in deciding what effect shall be produced by external
agents, but in the tropisms, at least, he considers the action of the external agent
to be direct, the inner disposition deciding merely whether the substance of the
Amoeba is of such a character as to admit of the production of a given definite
change in surface tension by the outer agent.

THE MOVEMEr^TS AND REACTIONS OF AMCEBA. 221

Based similarly on a direct action of external agents is the theory of
amoeboid movements proposed by Verworn in 1891, and extended in
his paper on Die Bewegung der lebendtgen Substanz (1892) and in
the Allgemeinc Physiologie, In its original form Verworn^s theory
considers the movements and changes of form to be brought about
directly through chemical attraction (Verworn, 1891, p. 105), but in
later publications (1892, 1895) the effect of the chemical is considered,
as in Rhumbler's theory, to be that of increasing or decreasing the
surface tension.

Thus it is evident that it must be the chemical affinity of certain parts of the
protoplasm for oxygen that decreases the surface tension in definite regions and
thus leads to the formation of pseudopodia. But it will be possible for the same
effect to be produced by other substances of the surrounding medium, if they
have chemical affinity for certain components of ihe protoplasm. In the case
where the substance acts from one side, this principle must lead to positive
chemotropism. (Verworn, 1895, p. 545.)

It is evident that the method of movement of Amoeba, as described
in this paper, has an immediate bearing on the question of direct or
indirect action of external agents. If the action of an external agent
is to increase or decrease directly the surface tension, as set forth by
Rhumbler and Verw^orn, this effect must be shown in the characteris-
tic currents which appear in any fluid when the surface tension is thus
locally changed. In the case of negative chemotaxis we should have
an axial current away from the side affected, with surface currents
toward the chemical, as indicated in the figure given by Rhumbler
(1898, p. 188). In positive chemotaxis both sets of currents should be
the reverse of that just indicated.

In the account of the movements set forth by Biitschli and Rhumbler,
the currents agreed with the scheme for direct action above set forth.
But this account of the movements was erroneous, as we have seen.
The internal currents and the surface currents are forward, away from
the region stimulated, in a negative reaction ; toward the region stim-
ulated in a positive reaction ; the movement is of a rolling character.
There is thus no evidence that the action of the stimulus is to cause a
change in the surface tension of the parts directly affected ; on the con-
trary, the direction of the currents is quite inconsistent with this view.*
We must conclude, then, that the theory of the direct action of exter-
nal agents in causing or changing the movements of Amoeba is nega-
tived by the character of the movements produced ; these are not such as
would follow from the direct physical action of the agents in question.

* A description of the forward surface currents in negative chemotaxis is given
on p. 143; in the reaction to a mechanical stimulus on p. 185; to the electrical
stimulus on p. 192 ; in a positive food reaction (chemical and mechanical stimuli ?)
on p. 198.

223 THE BEHAVIOR OF LOWER ORGANISMS.

There is much indirect evidence that points in the same direction,
particularly in the fact that the activities of the animal may remain con-
stant while the environment is continually changing. Some of these
lines of evidence are summarized by Rhumbler (1898, p. 185). They
still leave open the possibility that when the environment does modify
the movements its action is direct. But even this is shown to be
excluded by the nature of the currents produced, as described above.

The currents as they actually occur in the movements are equally
opposed to certain theories of the indirect action of stimuli. Bernstein
(1900), Jensen (1901), and others, have expressed the opinion that the
effect of stimuli is to change the surface tension, but that this eflect is
not due to the direct physical action of the agent on the protoplasm,
but rather to some change in the internal physiological processes of the
cell produced by the agent acting as a stimulus. Jensen (1901, 1902)
has developed this view into a detailed theory, according to which
stimuli that increase the normal assimilatory processes of the cell lead
to a reduction of surface tension, and hence to expansion and move-
ment toward the agent in question, while stimuli that increase the
dissimilatory processes have the opposite effect.

The currents in the moving Amceba lend no support to this view.
There is no evidence in the movement that the efi'ect of a stimulus is to
alter the surface tension in any way. In view of the facts given in the
body of this paper as to the nature of the movements, we are forced to
give up the idea that the efiect of stimuli is to modify the tension* of
the surface of the protoplasmic mass, either directly or indirectly.
Alterations in the tension of the surface can no longer be considered
the prime factor in the behavior of Amceba.

DIRECT OR INDIRECT ACTION IN THE TAKING OF FOOD.

Rhumbler, in his most interesting and suggestive paper (1S9S), has
attempted to give a physical analysis of food-taking and the choice of
food in Amceba. According to Rhumbler, the taking of food is due
to adhesion between the protoplasm and the food substance, and may
be compared with the pulling inward of a splinter of wood by a drop of
water, or of a bit of shellac by a drop of chloroform. The selection of
food is explained as due to the fact that the protoplasm, as might be
expected from physical considerations, tends to adhere to some sub-
stances and not to others. Parallel phenomena are shown, in a most
ingenious experiment, to be demonstrable for the chloroform drop
(/. c, p. 248). It takes in certain substances, while others are refused
or thrown out if introduced.

♦See note, p. 225.

THE iMOVEMENTS AND REACTIONS OF AMCEBA. 223

This theory is a most attractive one, and seems a priori probable.*
It is conceivable that there may be, or may have been, organisms where
it is applicable throughout. But an objective study of the behavior of
Amoeba shows that it gives by no means an adequate explanation of
food-taking in this animal. As I have shown in the descriptive por-
tion, in Amoeba proteus and A. angidata the food in most cases is far
from adhering to the protoplasm ; on the contrary, it rolls away when
the Amoeba comes in contact with it, and it is often only as a result of
long-continued effort that the animal succeeds in ingesting it. The
first act in ingestion consists in sending out pseudopodia on each side of
the mass to overcome the mechanical difficulty resulting from the fact
that the body does 7iot adhere to the protoplasm, but tends to roll away.

Further, a quantity of water is usually, or frequently, taken in with
the food, and the latter floats about in a cavity after it is ingested, show-
ing no tendency to adhere to the protoplasm (see Leidy, 1S79, numer-
ous figures of food vacuoles, etc., and Le Dantec, 1S94). A similar
condition of affairs is shown in the account of the feeding of one
Amoeba on another, given on page 201 of the present paper. Here the
prey does not adhere to the protoplasm of its captor, but moves about
within the latter and escapes repeatedly.

Thus, in these species, the taking of food and the choice of food can-
not be explained by the adherence of the protoplasm to the food sub-
stance, for the lack of such adherence is strikingly evident.

Rhumbler's studies of food taking were made chiefly on Amoeba
verrucosa. In this species and its close relatives there is much more
tendency for foreign objects to cling to the surface than in the other
species. But this adhesiveness applies to other objects as well as to
food. It is of special aid, as we have seen, in tracing surface move-
ments (p. 140). Particles of soot and various other indifferent bodies
stick to the surfoce, rendering its movements apparent. Not all such
adhering bodies are taken into the interior, so that the ingestion
involves an additional reaction, and is not fully accounted for by the
adhesion even in these species.

Rhumbler has given an ingenious physical analysis of the rolling up
and taking in of filaments of Oscillaria by Amceba verrucosa^ and has
illustrated the process as he conceives it to occur by a very striking
experiment (1S98, p. 330). A chloroform drop brought in contact
with the middle of a filament of shellac rolls the filament together and
encloses it. Rhumbler conceives the forces at work in rolling up the
filament to be essentially the same in the chloroform drop and in the
Amceba. In both cases, according to Rhumbler, the surface tension of

♦See Jennings, 1902, where I adopted this view before having investigated for
myself the behavior of Amoeba.

224 '^^^ BEHAVIOR OF LOWER ORGANISMS.

the drop pulls on the filament, tending to force it inward from both
directions. That part of the filament within the drop becomes softened
by the action of the fluid ; it, therefore, yields to the thrust from both
directions, and bends, permitting more of the filament to be brought
into the drop by the action of surface tension. (For full explanation,
see the original.)

It is necessary to point out that the explanation of the rolling up of
the shellac filament given by Rhumbler is erroneous. The surface
tension of the drop, with its inward thrust, has nothing to do with the
process, for the filament is rolled up in exactly the same manner when
it is completely submerged in a large vessel of chloroform, so that it is
not in contact with the surface film at all. The rolling up is evidently
due in some way to the strains within the shellac filament, produced
when it was drawn out, and to the adhesiveness of its surface when
acted upon by the chloroform. The process thus loses all similarity
to the rolling up of the alga filament by Amoeba. The coil formed is
just as small and close, and the filament remains a filament just as long
when the experiment is tried in a large vessel of chloroform as when
only a drop is used, as in Rhumbler's experiments.

Rhumbler's explanation of the way in which Amoeba rolls up the
Oscillaria filament may, of course, still be correct, though the physical
experiment by which he attempted to illustrate it has nothing to do
with the matter. There are certain points in his description of the pro-
cess as it occurs in Amoeba, however, that might easily be interpreted
in another manner. Such a bending over of the pseudopodium as is
shown in Rhumbler's Fig. 58 (/. c, p. 233) is not called for by the
surface-tension theor}'. Rhumbler holds that this bending of the pseu-
dopodium is passive, and due to the bending of the filament within the
body(/. c.,p. 233). In view of what we have shown above (pp. 177-179)
as to the power of active bending in the pseudopodia, and as to active
contractions of parts of the ectosarc in this same species (pp. 179, iSo),
one might be inclined to believe rather that this bending of the pseu-
dopodium is active and plays an important part in bringing the fila-
ment into the body. Rhumbler's figures (Fig. 50) would support this
view fully as strongly as his own theory, though this would, of course,
not give us a simple physical explanation of the ingestion of the filament.

Altogether, we must conclude that adhesion between the protoplasm
and the food substance cannot by any means give us a general explana-
tion of food-taking in Amoeba. In some cases the ingestion of food
is aided by such adhesion, but in other cases the adhesion is conspicu-
ously absent.*

♦ For further confirmation of last stated fact, see paper of Le Dantec ('1894).

THE MOVEMENTS AND REACTIONS OF AMCEBA. 225

GENERAL CONCLUSION.

Putting all our results together, we must conclude that the move-
ments and reactions of Amoeba have as yet by no means been resolved
into their physical components. Amoeba is a drop of fluid which
moves in its usual locomotion in much the same way as inorganic drops
move under the influence of similarly directed forces. But what these
forces are is by no means clear. When we take into considera-
tion the currents as they actually exist, local decrease in surface tension
breaks down completely as an explanation for the locomotion and other
movements. The locomotion taken by itself might be explained as due
to the adhesion of the fluid protoplasm to solids, taken in connection
with the surface tension of the fluid, but this explanation fails when we
consider the formation of free pseudopodia, and discover that all the
processes concerned in locomotion can take place without adhesion to
the substratum.

For the reactions to stimuli we find a parallel condition of aflJairs.
The currents in the protoplasm in the positive and negative reactions
are not similar to those produced in the attraction or repulsion of drops
of fluid by the direct action of external agents. Therefore we cannot
consider these reactions as due to the increase or decrease of surface
tension* produced by the direct (or even indirect) action of the external
agents. The taking and choice of food cannot be physically explained
in any general way by the physical adherence of the protoplasm as a
substance to the food as a substance, for food is taken in many cases
(usually, in some species) where it is demonstrable that no such
adherence exists.

While we must agree that Amoeba, as a drop of fluid, is a marvel-
lously simple organism, we are compelled, I believe, to hold that it has
many traits which are comparable to the "reflexes" or *' habits" of
higher organisms. f We may, perhaps, have faith that such traits are

*It should be pointed out that this and other statements concerning surface
tension in Amceba apply to the tension of the actual body surface, comparing
Amoeba thus to a drop of simple fluid. This is the basis on which rest the pre-
vailing theories that would explain the movements of Amoeba by surface tension.
It is these theories which I desired to test. There remains untouched, of course,
the possibility that the movements of all sorts of protoplasmic masses may be
explained by changes in the surface tension of the meshes of Blitschli's honey-
comb structure, in the manner indicated by BUtschli (1892, p. 208). But this is
at present merely a hypothesis, not worked out and not controllable by observa-
tion. To attempt to maintain it for Amoeba would be to relegate the movements
of this animal to the same obscure category as the movements of cilia and of
muscles, possibly a correct proceeding, but removing the matter at present from
the field of experimental observation.

t See the next division of this paper, wheie this point is developet}.

226 THE BEHAVIOR OF LOWER ORGANISMS.

finally resolvable into the action of chemical and physical laws, but we
must admit that we have not arrived at this goal even for the simpler
activities of Amoeba.

THE BEHAVIOR OF AMCEBA FROM THE STANDPOINT OF THE
COMPARATIVE STUDY OF ANIMAL BEHAVIOR.

HABITS IN AMCEBA.

Although in general Amoeba has the rolling movement of a drop of
fluid, yet this statement by no means brings out all the characteristics
of the movement in any given species of Amoeba. Different kinds of
Amoebae move differently, and the differences are in many cases not
such as can be accounted for by differences in the state of aggregation
of the body substance. Some Amoebae, as is well known, form many
pseudopodia, others few or none. Different Amoebae have different
characteristic forms in locomotion. But more striking than these gen-
erally recognized peculiarities are certain others of a more special char-
acter. A creeping Amoeba angulata^ as we have seen above, frequently
pushes upward and forward at the anterior end a short, acute pseudo-
podium, which waves slightly from side to side like an antenna (p. 177
and Fig. 62, c). This peculiar habit is much more pronounced in
Amoeba velata Parona, according to Penard (1902). In this animal
the free anterior pseudopodium may extend for a length greater than
the diameter of the body ; Penard compares it directly to a tentacle.
Some other species of Amoeba never send forward such an antenna-like
pseudopodium. The great work of Penard (/. c.) contains innumer-
able instances of such peculiarities of form, movement, and function
among the different species of Amoeba and other Rhizopods ; some of
them are collected in that author's interesting section on the pseudo-
podia (/. c, pp. 625-629). It is not necessary to take these up in detail
here. The point of interest is that different sorts of Amoebae have dif-
ferent customary methods of action, such as are commonly spoken of
as *' habits"* in higher animals, and that these *' habits" are no more
easily explicable on direct physical grounds in Amoeba than in higher
animals. Let anyone attempt, for example, to explain from the
physics of viscous fluids why Afnoeba velata or A, angulata push
out an antenna-like pseudopodium at the anterior end and wave it from
side to side, while Amoeba proteus and A. Umax do not.

♦The word habit is, of course, not used here of a method of action acquired
during the life of the individual, but merely of a fixed method of behavior. At
all events, it is difficult to distinguish between these two things where individual
organisms, as in Amoeba, have lived as long as the race.

THE MOVEMENTS AND REACTIONS OF AMCEBA.

227

CLASSES OF STIMULI TO WHICH AMCEBA REACTS.

The simple naked mass of protoplasm reacts to all classes of stimuli
to which higher animals react (if we consider the auditory stimulus
merely a special case of the mechanical stimulus). Mechanical stimuli,
chemical stimuli, temperature differences, light, and electricity — all
control the direction of movement, as they do in higher animals.

TYPES OF REACTION.

Amoeba has two chief types of reaction, by one or the other of which
it responds to most stimuli. These we may call the positive and the
negative reactions. As a third type we must distinguish the food
reaction, which cannot be brought completely under either of the two
chief types of reaction above mentioned.

(i) The positive reaction consists in pushing lout the body substance
toward the source of stimulus and rolling in that direction.

(2) The negative reaction consists in withdrawal of body substance
and rolling in some other direction — not necessarily in the opposite
direction.

(3) The food reaction is not sharply definable. Its most character-
istic features consist in the hollowing out of the anterior end and in
the pushing out of pseudopodia at each side of and over the food body.
It involves also the positive reaction above characterized.

RELATION OF THE DIFFERENT REACTIONS TO DIFFERENT STIMULI ;
ADAPTATION IN THE BEHAVIOR OF AMCEBA.

(a) The positive reaction is known to be produced by weak mechanical
stimuli ; it is probably produced also by weak chemical stimuli (in the
reactions to food, pp. 193-202). The positive reaction to weak mechan-
ical stimuli serves the purpose of bringing the floating animal to a sur-
face on which it can creep. The positive reaction to food substances
(mechanical and chemical stimuli), of course, serves to obtain food.
The positive reaction is thus, as a rule, performed under such circum-
stances as to be beneficial to the organism ; i. e., it is directly adaptive.

(3) The negative reaction is produced by powerful stimuli of all
sorts. Such powerful stimuli are, as a rule, injurious, and the nega-
tive reaction tends to remove the Amoeba from their action ; it is, there-
fore, directly adaptive. This is true of the negative reaction to light
as well as to other stimuli, for light is known to interfere with the activi-
ties of Amoeba. The reaction to the electric current is of exactly the
character that would be produced by a strong stimulus on the anode
side, but owing to the peculiarity of the current the reaction does not
assist the Amoeba to escape. The reaction to the electric current can
not then be considered adaptive ; this stimulus forms, of course, no
part of the normal environment of an Amoeba.

228 THE BEHAVIOR OF LOWER ORGANISMS.

The method by which, throuj^h the negative reaction, Amoeba avoids
an injurious agent is of interest. The animal does not go directly away
from the injurious agent, as by moving toward the side opposite that
stimulated. It moves in any direction except toward the region stimu-
lated. There is no system of conduction such that strong stimulation in
a given spot involves movement directed toward the opposite side. If
the movement in the new direction induces a new stimulation, the direc-
tion is again changed. This may be continued until the original direction
of movements is squarely reversed. The method is akin to that of trial
and error in higher organisms (see Morgan, 1894, pp. 241-242).

(c) The food reaction is directly adaptive in that it procures food.
As a rule this reaction occurs only when the source of stimulation is
fitted to serve as food. Empty diatom shells, sand grains, debris, etc.,
are, as a rule, not taken into the body, as many observers have pointed
out. Sometimes material is taken into the body that is not useful, as
is described by Rhumbler (189S, p. 236). In such cases there is no
evidence of a food reaction in the sense characterized above; the
material is ingested accidentally^ as it were, through its adherence to
the protoplasm. The food reaction, as a definite form of behavior, is
always adaptive so far as known.

{d) Some of the habits of Amoeba, characterized above, are clearly
adaptive. The use of the antenna-like pseudopodium sent out by
Ammba velata and A. angulata is evident. Penard describes in detail
how Ammba velata .uses it in pafssing from one substratum to another.

The habit which some Amoebae have, when suspended freely in the
water, of sending out pseudopodia in all directions (p. iSi) is, of course,
useful in that it increases the chances of coming in contact with some
solid object, without which the Amoeba cannot move from place to
place.

REFLEXES AND '^ AUTOMATIC ACTIONS " IN AMCEBA.

In the behavior of Amoeba we can distinguish factors directly com-
parable to the reflexes and *' automatic activities " of higher organisms.
The responses of Amoeba to stimuli have the nature of reflexes in the
fact that they are not direct effects of the physical action of the stimulus
(see p. 219), but are determined by the internal conditions of the
organism. They may be called reflexes, unless we propose, as certain
writers do, to restrict the term reflex to processes involving differen-
tiated nerves. The precise designation is unimportant ; the essential
point is that the responses agree with the reflexes of higher animals in
being indirect.

Ziehen, in his Leitfaden der physiologischen Psychologic (sixth
edition, p. 10), defines as automatic acts '• motor reactions, which do

THE MOVEMENTS AND REACTIONS OF AMCEBA. 229

not, like the reflexes, follow unchangeably upon a definite stimulus,
but which are modified in their course by new, intercurrent stimuli."
In this sense Amoeba, of course, shows automatic behavior. Its
responses are by no means unchangeably fixed ; on the contrary, its
behavior is often modified by the slightest change in the stimulus to
which it is reacting. For examples of this see the chase of one Amoeba
by another (p. 200), the following of a rolling ball of food (p. 196), the
account of the driving of Amoeba (p. 185), and the description of the
method by which Amoeba avoids an obstacle (p. 186).

Whether these actions agree with the accepted idea of an automatic
action in being unconscious we have, of course, no means of knowing.

VARIABILITY AND MODIFIABILITY OF REACTIONS.

There is little that can be said on this point. Verworn (1890, «, p.
271) says that when an electric current is passed through a prepara-
tion containing many Amoebae, some Vespond strongly, while others
do not ; thus different individuals vary in their responsiveness. Fur-
ther, a given individual may become accustomed to the current, at first
responding to it, later not responding. Doubtless such phenomena of
acclimation are common in the reactions to all sorts of stimuli.

Rhumbler (1898, p. 203) shows that when Amoebae are engaged in
taking Oscillaria filaments as food, light thrown upon them modifies
them physiologically in such a way that they eject the food. The
nature of the reaction is thus shown to depend partly on the physio-
logical condition of the animal.

There is no direct experimental evidence as yet, so far as I am aware,
that Amoeba shows memory.* Experimental evidence as to whether
the reactions of a given Amoeba to a given stimulus are modified by
previous stimuli received is very difficult to obtain, principally because
it is practically impossible to make succeeding stimuli alike, so that
one cannot tell whether a difference in the reaction is due to a differ-
ence in the present stinnilus or not. Possibly there is a faint indication
of something akin to memory shown in the facts described on page 201.
Here a smaller Amoeba which had been ingested as prey escaped from
the posterior end of the captor ; the latter tliereupon reversed its move-
ments, came up with the escaping prey, and again ingested it. In the
interval between the complete separation of the prey from its captor
and its recapture, the behavior of the captor would seem to have been
determined by some trace left within it by the former possession of the

♦The word memory is, of course, used here of the objective phenomenon that
in many animals present behavior is modified in acconlance with past stimuli
received, or past reactions given. Of possible subjective accompaniments of this
objective phenomenon we, of course, know nothing directly so far as the lower
organisms are concerned.

230 THE BEHAVIOR OF LOWER ORGANISMS.

prey. But, possibly, this trace was merely of a gross physical char-
acter, acting as a direct stimulus to produce the observed behavior. If
this is true, the behavior shows no indication of memory.

SUMMARY.

OBSERVATIONS.
THE USUAL MOVEMENTS.

(i) Locomotion in Amoeba is a process that may be compared with
rolling, the upper and lower surfaces continually interchanging posi-
tions. This is shown by observation of the movements of particles
attached to the outer surface or embedded in the ectosarc of the animal.
Such attached particles move forward on the upper surface and over
the anterior edge, remain quiet on the under surface till the body of the
Amoeba has passed, then pass upward at the posterior end and forward
on the upper surface again. Single particles may thus be observed to
make many complete revolutions. (See p. 170, Fig. 58, and Figs. 38, 39,
40, 41.) *

(2) Thus the upper surface moves forward in the same direction as
the internal currents, while the lower surface is at rest. There is char-
acteristically no backward current anywhere in Amoeba, though at
times some of the endoplasmic particles, spreading out laterally at the
anterior end, may move a slight distance backward at the sides. This
is rare (see p. 134). The forward current on the upper surface is not
confined to a thin layer, but extends inward to the endosarc ; the endo-
sarcal and surface currents are one (p. 142).

(3) In the formation of pseudopodia that are in contact with the
substratum the movement of protoplasm is identical with that at the
anterior end of the Amoeba. The upper surface and internal contents
flow toward the tip, while the surface in contact with the substratum is
quiet. Particles adhering to the upper surface are carried out to the tip
and rolled under to the lower surface, where they remain quiet (p. 152).

(4) In the formation of pseudopodia projecting freely into the water,
the movements of substance are the same as in pseudopodia that are in
contact, save that there is no part of the surface at rest. The whole

♦ Of anyone who is inclined to reject these results on the basis of previous
observations, or of their supposed incompatibility with other known facts, let
me make the following request : Before taking ground against the results, pro-
cure some specimens of Amoeba verrucosa or one of its relatives. This is usually
easily done. Then mix thoroughly with the water containing them some fine
soot, and observe carefully the movements of the animals. The particles attach
themselves to the outside, and the movements of the surface are then observable
with the greatest ease. It is such a simple matter to determine certain of the
chief points for one's self in this manner that it would be regrettable for contro-
versy to arise through neglect of the needed observations.

THE MOVEMENTS AND REACTIONS OF AMCEBA. 23 1

surface thus moves outward, and new parts of the surface of the body
continually pass on to the pseudopodium. An object adhering to the
surface of the pseudopodium remains at approximately the same dis-
tance from the tip, both when the pseudopodium is short and when it
has become very long (pp. 153-156, and Figs. 47-49).

(5) In the withdrawal of a free pseudopodium (a) a process occurs
that is the reverse of that described in (4), the surface of the pseudo-
podium passing from its base on, to the surface of the body (p. 156,
and Fig. 49) ; (6) the surface of the pseudopodium becomes wrinkled
and shrunken ; (c) the endosarc flows back into the body.

(6) Any part of the protoplasm may be excluded temporarily from
the forward currents. In many Amoebae there is usually a region at
the posterior end which is thus temporarily excluded (the posterior
appendage, tail). In such cases the lower surface of the Amoeba
passes upward on each side of this appendage to become part of the
upper surface, then passes forward (p. 169, and Fig. 57). The sub-
stance of the posterior appendage is itself gradually drawn into the
forward current.

(7) The anterior portion of the advancing Amoeba is attached to the
subvStratum, while the posterior portion is not (p. 165). There is a viscid
secretion produced on the outer surface of the Amoeba, to which the
attachment may be due.

(8) The attached anterior portion of the body is spread out and
usually very thin. The unattached posterior portion becomes rounded
and thick, and is contracting, so that there is a slight forward move-
ment on the lower surface, as well as on the upper surface, in this part
(p. 166).

(9) All the activities concerned in locomotion can be performed when
the animal is not attached to the substratum. (But for progression such
attachment is necessary, p. 215.)

(10) The locomotion of Amoeba is similar even in details to the
movements of a drop of inorganic fluid which adheres strongly to the
substratum at one edge and spreads out upon it here, while the other
edge is free (pp. 209-214). It is similar in most respects (except in the
thinness of the anterior edge) to the movements under the influence of
gravity of a drop of fluid along an inclined surface to which it adheres
but slightly.

(11) The currents in a moving Amoeba are not similar to those of a
drop of inorganic fluid that is moving or elongating as a result of a local
increase or decrease in surface tension. The surface currents away
from the region of least tension and in the opposite direction to the
axial currents that are characteristic of such a drop are lacking in
Amoeba. Here surface and axial currents have the same direction
(p. 205).

332 THE BEHAVIOR OF LOWKR ORGAKISMS.

(12) Similarly, the movements of material in a forming pseudo-
podium are not like those in a projection which is produced in a drop
of inorganic fluid as a result of a local decrease in surface tension. The
surface currents away from the tip in the inorganic projection are lack-
ing in Amoeba, the surface here moving in the same direction as the
tip (p. 206).

(13) The body of Amoeba shows under some conditions elasticity
of form, such as is characteristic of solids (pp. 175-177).

(14) Besides the movements directly concerned in the usual loco-
motion, limited local contractions may occur, resulting in swinging or
vibratory motions of the pseudopodia (p. 177), or in sudden, jerky
movements of the body as a whole (p. 179), or in the constricting off"
of parts of the body (p. 180).

(15) The roughness of surface in a retracting pseudopodium, the
retention of irregular forms by Amoeba, and the movements mentioned
in the foregoing paragraph, are similar to certain phenomena to be
observed in mixtures of solids and fluids, as a result of the interaction
of surface tension and the friction of the solid particles (p. 215).

REACTIONS TO STIMULI.

(16) The following reactions are described : Positive and negative
reactions to mechanical stimuli (pp. 1S1-187) ; negative reactions to
chemical stimuli (pp. 1S7-190); negative reaction to heat (pp. 190-191);
reaction to the constant electric current (pp. 1 91-192); complex reac-
tions connected with food-taking (pp. 193-202); reactions to injuries
(pp. 202-204).

(17) In most reactions modifying the direction of motion a new
advancing wave of protoplasm is sent out from some part of that por-
tion of the body which is already attached to the substratum. The
internal and surface currents then flow in that direction, thus changing
the course of the animal. Thus, when the animal changes its course in
a reaction, the surface currents change their course in a corresponding
way, as shown by the movements of particles on the surface (pp. 143,
185, 192).

(18) Sometimes the course is squarely reversed as a result of a
stimulus. In this case the original anterior region becomes detached
from the bottom, while a new pseudopodium projects freely into the
water from the former posterior (unattached) region, settles down, and
becomes attached ; the internal and surface currents then follow it.
This process requires usually a considerable interval of time (pp. 183,
1S4).

(19.) Both surface currents and internal currents are toward the
source of stimulation in a positive reaction, away from the source of
stimulation in a negative reaction.

THE MOVEMENtS AND tlEACTlONS OF AMOEBA. 2'^^'^

(20) The currents in the positive and negative reactions are not
similar to the currents in a drop of inorganic fluid moving toward or
away from an agent which causes a local decrease or increase in the
surface tension. In Amoeba the currents on the surface and in the
interior are congruent ; in the inorganic fluid they are opposed.

(21) In the taking of food the protoplasm and the food body in
many cases do not tend to adhere, so that the Amoeba is compelled to
overcome considerable mechanical, difliculty before the food can be
inclosed. Frequently the food body rolls away from the animal as soon
as it is touched (pp. 193, 196). The difficulty is overcome by sending
out pseudopodia on each side of the body and inclosing it, together with
a certain amount of water. In A?nceba verrucosa and its relatives food-
taking is aided by the tendency of foreign bodies to adhere to the body
surface. Amoebae frequently prey upon each other, and this often gives
rise to a long and complex train of reactions (pp. 198-202, and Fig. 76).

CONCLUSIONS.

(22) The chief factors in locomotion seem to be as follows: (i) At
the anterior edge of the Amoeba a wave of protoplasm pushes out, rolls
over, and becomes attached to the substratum ; (2) This pulls on the
upper surface of the Amoeba, causing it to move forward ; (3) The
hinder portion of the Amoeba becomes released from the substratum,
and contracts slowly ; (4) As a result of the strong pull from in front
and the slight contraction from behind the posterior end moves forward ;
(5) The internal substance must flow forward as a result of the pull on
the upper surface, the movement forward of the posterior end, and the
pressure due to the pulling from in front and the contraction behind.
The movement of the internal fluid is comparable to that in a sac or
bladder half filled with water and rolled along a surface (pp. 146, 149,
171).

(23) There is no continuous transformation of endosarc into ectosarc
at the anterior end, and of ectosarc into endosarc behind this (Rhum-
bler's ento-ectoplasm process), as a necessary feature of locomotion,
since the ectosarc of the upper surface rolls over to the under surface
at the anterior end (p. 148). Nevertheless, ectosarc and endosarc are
mutually interconvertible when need arises for the change of one into
the other.

(24) It results from paragraphs (i), (2), (11), above, that the loco-
motion of Amoeba cannot, with fidelity to the results of the physical
experiments, be accounted for by a decrease in surface tension at the
anterior end.

(25) From (3), (4), (12), above, we must conclude that the sending
out of pseudopodia cannot, without violence to the results of the phys-
ical experiments, be accounted for as due to a local decrease in surface
tension at the point of the pseudopodium.

334 '^^^ BEHAVIOR OF LOWER ORGANISMS.

(26) From (i) and (10) above it results that the simple locomotion
on a substratum could, taken by itself, be accounted for on Berthold's
theory that the movement is due to the spreading out of a fluid on a
solid. But this theory fails when we take into account the formation
of free pseudopodia (p. 214), and the fact that all the processes con-
cerned in locomotion can be performed without adherence to a solid
(paragraph (9) above).

(27) From (3), (4), (9), (11), (12), (24), (25), (26), we must con-
clude that the formation of pseudopodia and the sending out of waves
of protoplasm at the anterior end of a moving Amoeba are due to a
local activity of the protoplasm for which no physical explanation has
been given. Since these are the essential features in locomotion, we
must conclude that locomotion in Amoeba has not been physically
explained.

(28) From (17), (19), (20), it follows that we cannot, with fidelity
to the results of physical experimentation, hold that the effects of stimuli
in modifying the movements of Amoeba are due to their direct (or even
indirect) action in changing the surface tension of the parts affected.

(29) From (21) we must conclude that adherence between the proto-
plasm and the food substance does not furnish an adequate explanation
of food-taking and the choice of food in Amoeba.

(30) From (24), (25), (27), (28), we must conclude that changes in
the surface tension of the body are not the primary factors in the move-
ments and reactions of Amoeba. (See note, p. 225).

(31) All the results taken together lead to the conclusion that neither
the usual movements nor the reactions of Amoeba have as yet been
resolved into known physical factors. There is the same unbridged
gap between the physical effect of the stimulus and the reaction of the
organism that we find in higher animals.

(32) In the behavior of Amoeba we may distinguish factors compar-
able to the habits, reflexes, and automatic activities (Ziehen) of higher
organisms (pp. 228-229). Its reactions as a rule are adaptive (pp.
227-228).

SEVENTH PAPER

THE METHOD OF TRIAL AND ERROR IN THE
BEHAVIOR OF LOWER ORGANISMS.

235

THE METHOD OF TRIAL AND ERROR IN THE
BEHAVIOR OF LOWER ORGANISMS.

A certain type of behavior in higher animals has been characterized
by Lloyd Morgan as the method of trial and error. The nature of such
behavior is well brought to mind by an example from Morgan (1894,
p. 357). His dog v^as required to bring a hooked walking stick
through a narrow gap in a fence. The dog did not pause to consider
that the stick would pass through the narrow opening only if taken
by one end and pulled lengthwise. On the contrary, he simply seized
the stick in the way that happened to be most convenient, near its
middle, and tried to carry it through the gap in the fence in that man-
ner. Of course, the stick would not pass, and after some effort the
dog was forced to drop it. Then he seized it again at random, and
made a new effort. Again the stick was stopped by the fence ; again
the dog dropped it, took a new hold, and tried again. After several
repetitions of this performance, the dog seized the stick by the hooked
end. This time it passed through the gap in the fence easily.

The dog had '"tried" all possible methods of pulling the stick
through the fence. Most of the attempts showed themselves to be
*' error." Then the dog tried again, till lie finally succeeded. Thus
he worked by the method of trial and error.

This method of reaction has been found by Lloyd Morgan, Thorn-
dike (1S9S), and others, to play a large part in the development of
intelligence in higher animals. Intelligent action arises as follows:
The animal works by the method of trial and error till it has come
upon the proper method of performing an action. Thereafter it begins
with the proper way, not performing the trials anew each time. Thus
intelligent action has its basis in the method of '' trial and error," but
does not abide indefinitely in that method.

Behavior having the essential features of the method of '* trial and
error" is widespread among the lower and lowest organisms, though
it does not pass in them so immediately to intelligent action. But like
the dog bringing the stick through the fence the first time, they try all
ways, till one shows itself practicable.

This is the general plan of behavior among the lowest organisms
under the action of the stimuli which pour upon them from the sur-
roundings. On receiving a stimulus that induces a motor reaction,
they try going ahead in various directions. When the direction fol-
lowed leads to a new stimulus, they tr}' another, till one is found which
does not lead to effective stimulation,

237

238

THE BEHAVIOR OF LOWER ORGANISMS.

This method of trial and error is especially well developed in free-
swimming single-cell organisms — the flagellate and ciliate infusoria —
and in higher animals living under similar conditions, as in the Roti-
fera. In these creatures the structure and the
method of locomotion and reaction are such
as to seem cunningly devised for permitting
behavior on the plan of trial and error in the
simplest and yet most effective way.

These organisms, as they swim through the
water, typically revolve on the long axis, and at
the same time swerve toward one side, which
is structurally marked. This side we will call
X. Thus the path becomes a spiral. The or-
ganism is, therefore, even in its usual course,
successively directed toward many different
points in space. It has opportunity to try suc-
cessively many directions though still progress-
ing along a definite line which forms the axis
of the spiral (see Fig. 79). At the same time
the motion of the cilia by which it swims is
pulling toward the head or mouth a little of the
water from a slight distance in advance (Fig.
79). The organism is, as it were, continually
taking "samples" of the water in front of it.
This is easily seen when a cloud of India ink
is added to the water containing many such
organisms.

At times the sample of water thus obtained is
of such a nature as to act as a stimulus for a
motor reaction. It is hotter or colder than
usual, or contains some strong chemical in solu-
tion, perhaps. Thereupon the organism reacts
in a very definite way. At first it usually
stops or swims backward a short distance, then
it swings its anterior end farther than usual
toward the same side X to which it is al-
ready swerving. Thus its path is changed.
After this it begins to swim forward again. The amount of backing
and of swerving toward the side X is greater when the stimulus is more
intense.

3

Fig. 79.*

♦Fig. 79. — Spiral path in the ordinary swimming of Paramecium, showing
how the anterior end is pointed successively in different directions, and how a
sample of water is drawn to the anterior end and mouth from each of these
directions.

THE METHOD OF TRIAL AND ERROR.

239

This method of reaction seems very set and simple when considered
by itself. It is almost like that of a muscle which reacts by the same
contraction to all effective stimuli. The behavior of these animals
seems, then, of the very simplest character. To practically all strong
stimuli they react in a single definite way.

But if we look closely at this simple method of reacting, we find it,
after all, marvelously effective. The organism, as we have seen, is

i^

revolving on its long axis. When, as a consequence of stimulation, it
swings its anterior end toward the side X^ this movement is combined

♦ Fig. 80 — Diagrams of the movements in a reaction to a stimulus in an infu-
sorian, Paramecium C>4), and in a rotifer, Anursea(5), The anterior end swings
about in a circle (turning continually toward the aboral or dorsal side). It thus
tries many different directions, at the same time receiving samples of water from
each of these directions, i, 2, 3, 4, 5, the successive positions taken, with the
currents of water at the anterior ends. If the stimulus ceases the organism
may stop in any of these positions, and swim forward in the direction so indi-
cated. (The backward swimming, which precedes or accompanies the turning,
is not represented.)

240 THE BEHAVIOR OF I.OWER ORGANISMS.

with the revolution on the long axis. As a consequence, the anterior
end is swung about in a wide circle ; the organism tries successively
many widely differing directions (Fig. 80). From each of these direc-
tions, as we have seen, a sample of water is brought to the sensitive
anterior end or mouth. Thus the reaction in itself consists in trying
the water in many different directions. As long as the water coming
from these various directions evinces the qualities which caused the
reaction — the greater heat or cold or the chemical — the reaction, with
its swinging to one side, continues. When a direction is reached from
which the water no longer shows these qualities, there is no further
cause for reaction ; the strong swerving toward the side A' ceases, and
the organism swims forward in the direction toward which it is now
pointed. It has thus avoided the region where the conditions were
such as to produce stimulation.

While the account just given shows the essential features of the reac-
tion, the actual series of events appears in many cases more complicated,
though there is nothing differing in principle from what was just set
forth. The apparently greater complication arises from the repetition
of that feature of the reaction which consists in swimming backward,
and in the cessation of the reaction at intervals, with an attempt to
swim forward. After the organism has swung the anterior end to one
side, if it still receives the stimulus it may begin the reaction anew ;
that is, it may swim backward a distance, and again begin turning
toward the side X. This may be repeated several times. Each time
it is repeated the organism swings its anterior end through a new series
of positions, thus increasing the chances of finding one in which there
is no farther stimulation. Again, the organism, after reacting in the
way described in the last paragraph, may begin to swim forward, only
to find that it receives the stimulus again ; it then repeats the whole
reaction, thus supplying itself with a completely new set of directions
and of samples of water from those directions. In some cases the reac-
tion is thus repeated many times before any direction is found toward
which the organism can swim without receiving stimulation.

This is the method of behavior which the present author has been
describing in detail in many organisms in his series of ten Studies on
Reactions to Stimuli in Unicellular Organisms,* and in the foregoing
Contributions to the Study of the Behavior of the Lower Organisms.
Not until recently, it must be confessed, has the real significance of
this type of behavior been fully perceived. The results seemed to a
large degree negative ; the reaction method clearly did not agree with
the prevailing tropism theory, nor with any other of the commonly

♦Journ. of Physiol., 1897, vol. 21; Amer. Journ. of Physiol., vols. 2 to 8,
1899 to 1902; Amer. Naturalist, vol. 33, 1899; Biol, Bull., vol. 3, 190;.

THE METHOD OF TRIAL AND ERROR. 24I

held theories as to the reactions of lower organisms. Just what the
organism did was, indeed, fairly clear, but the plan of it all, the gen-
eral relations involved in all the details, was not clear. This was
partly due, perhaps, to overemphasis of certain phases of the reaction
and to a tendency to consider other features unimportant. The beha-
vior under stimuli is a unit ; each factor must be considered in connec-
tion with all the others ; then the general method running through it
all becomes strikingly evident.

Let us now return to the organisms. Sometimes stimuli are received
of such a nature that their distribution is not affected by the currents
produced by the cilia ; in other words, they cannot be sampled in the
currents of water brought to the anterior end or mouth, as shown in
Figs. 79 and 80. This is true, for example, of stimulation by light,
and of stimulation by contact with solid objects. Under such stimula-
tion the behavior is nevertheless still by the method of trial and error.
Let us consider first the reaction to a mechanical stimulus.

When the organism comes in contact with a mechanical obstacle the
reaction is exactly the same as that already described. It swims back-
ward, swings toward the side X^ and this, with the revolution on the
long axis, points the anterior end successively in many different direc-
tions. The organism then follows one of these directions. If this leads
against the obstacle, the reaction is repeated, till finally a direction is
found in which the obstacle is avoided.

In the reaction to light, as it occurs in Stentor or Euglena, experi-
ment shows that changes in the intensity of illumination at the sensi-
tive anterior end are the agents causing reaction (see the second of these
contributions) . The reaction produced is that already described ; by
turning toward the side A^and revolving on its long axis, the organism
tries many directions.

When a negative organism, such as Stentor, comes in its swimming
to an area that is more brightly illuminated, or when a positive organ-
ism, such as Euglena, comes to an area that is less brightly illuminated,
the change in intensity acts as a stimulus. The organism responds in
the way already described ; it backs away, then tries many different
directions by swinging its anterior end about in a circle. It then
starts forward in one of these directions. If this does not lead into the
area causing stimulation, well and good ; if it does, the organism
repeats the reaction, trying a new set of directions, till it finds one that
does not carry it to the area causing stimulation.

When light coming from a certain direction falls upon one side of a
swimming infusorian, the spiral path followed, of course, causes the
anterior end to be pointed successively in different directions. As a
result, the illumination of the anterior end is repeatedly changed, since

242 THE BEHAVIOR OF LOWER ORGANISMS.

in some directions it is pointed more nearly toward the light, in others
more away from the light, so as to be partly shaded by the rest of the
body. These changes in illumination cause the reaction ; the organism
tries pointing in various directions. When it comes into such a posi-
tion that the anterior end is no longer subjected to changes in intensity
of illumination, it continues to swim forward in that direction. Such
a position is found only when the axis of the spiral path is in the direc-
tion of the light rays. In an organism which reacts when the intensity of
illumination is decreased, such a position is stable only when the anterior
end is directed toward the source of light ; in an organism which reacts
when the intensity of illumination is increased, only when the anterior
end is directed away from the source of light (details in second of these
contributions). Thus the organism tries various directions till one is
found which does not subject it to changes in intensity of illumination
at the anterior end ; in this direction it swims forward.

The reaction which produces orientation to light can be stated more
simply, but less completely and accurately, as follows : When light
coming from a certain direction falls upon the sensitive anterior end of
a negative organism, such as Stentor, this causes the reaction above
described. The animal, after backing, tries a new set of directions,
by whirling its anterior end about in a circle. It continues or repeats
this until a direction is found in which the light no longer falls on the sen-
sitive anterior end. It is then oriented with anterior end away from the
source of light. In the positive organisms, such as Euglena, the method
of reaction is the same, save that it is the shading of the anterior end
that causes the reaction. When the anterior end is shaded the organ-
ism reacts in the usual way. It tries successively many different direc-
tions, by whirling its anterior end about in a wide circle ; when the
anterior end becomes pointed toward the source of light, the organism
continues forward in that direction.

In those infusoria which creep along the bottom, as Stylonychia or
Oxytricha, the reaction method is of a slightly simpler character,
though identical in principle. These animals when creeping do not
rotate on the long axis. When stimulated in any of the ways described,
they dart back, then turn to their right. They thus keep in contact
with the bottom, and may turn through any number of degrees up to
360 or more (Fig. 81). The reaction places them thus successively
in every position with reference to the source of stimulus that is possi-
ble so long as they remain on the bottom, and in each position the
adoral cilia are bringing samples of water to the anterior end and
mouth, as in Fig. 81. When a position is reached where the stimulus
no longer acts, the reaction ceases, and the animal moves forward in
that direction. The reaction is sometimes repeated several times before
the definitive position is attained.

THE METHOD OF TRIAL AND ERROR.

243

In no other group of organisms does the method of trial and error
so completely dominate behavior, perhaps, as in the infusoria. In this
group the entire organization seems based on this method. But reac-
tions on this plan are found abundantly elsewhere. In Amceba the
present author has shown that many of the reactions are of this charac-
ter. (See the preceding paper on the movements and reactions of
Amoeba.) When stimulated mechanically or by a chemical, the
Amoeba does not move directly away from the source of stimulus, but
merely in some other direction than that toward the side stimulated.
If this leads to a new stimulus, the animal tries another direction. By
continued stimuli Amoeba may be driven in a definite direction. The
conditions necessary for this are that movement in any other direction
shall lead to stimulation.

Reaction on the plan of trial and error is, perhaps, best seen in
Amoeba in the method by which
a specimen suspended in the water
finds and attaches itself to a solid
object. The suspended Amoeba
sends out pseudopodia in all di-
rections. If the tip of one of these
pseudopodia comes in contact with
a solid object it becomes attached ;
the protoplasm begins to flow in
that direction, and all the other
pseudopodia are withdrawn. The
Amoeba then passes to the solid
and creeps over its surface. Thus
the Amoeba has tried sending out pseudopodia in all directions ; that
which has been successful in finding a solid determines the direction of
movement.

In bacteria the reactions to light, to chemicals, and to mechanical
stimuli are essentially like those of the infusoria. The details as to the
direction of turning, etc., are not known, owing to the minuteness of
these organisms. But the essential point is that when the bacteria are
stimulated effectively they change the direction of movement. vSuch
change is repeated until the organisms are brought into a position
where there is no effective stimulation. The behavior is clearly that
of trial and error.

Fig. 81.*

* Fig. 81.— Different positions occupied in the usual reaction to stimuli in
Oxytricha. The animal swings its anterior end in a circle, occupying succes-
sively positions i, 2, 3, 4, 5, 6, and receiving a sample of water from each direc-
tion in which the anterior end is pointed. When the stimulus ceases the animal
may swim forward in any of these directions, (The backward swimming which
precedes or accompanies the turning toward the right side is not represented.)

244 THE BEHAVIOR OF LOWER ORGANISMS.

In the Metazoa behavior is usually not that of trial and error in
so elementary a form as is found in the organisms thus far considered.
The higher animals, with the development of a nervous system and
other bodily differentiations, have usually acquired the power of react-
ing more precisely with reference to the localization of the source of
stimulus. They more often, therefore, turn directly toward or away
from the source of stimulus, a preliminary trial of different directions
being unnecessary. But with the acquirement of many reaction possi-
bilities, the field for the operation of the method of trial and error is
greatly broadened. This could be amply illustrated from the behavior
of certain of the lower Metazoa. I hope to develop this point in detail
at some later time. Since at present we are interested chiefly in the
lowest organism, I shall mention here only a few cases in the Metazoa.

In the Rotifera behavior is, under many conditions, precisely similar
to that which I have described above for the infusoria (details in the
third of these contributions) ; that is, the behavior is an example of the
method of trial and error in a very pure form.

In Hydra we find the method of trial and error in a number of
features of the behavior. Since a general paper on the behavior of
Hydra is in preparation,* I will mention only one or two points that
have already been described. If we observe a living, unstimulated
green Hydra, we find that it does not remain at rest. If the Hydra is
extended in a certain direction, after one or two minutes it contracts,
bends over to a new position, then extends in a new direction. After
about two minutes it contracts again, bends into a still different posi-
tion, and again extends. This process is repeated at fairly regular
intervals, so that after a time the Hydra has tried every position possi-
ble in its present place of attachment. This exploration of all parts of
the surrounding region, of course, aids greatly in finding food.

Mast (1903) finds that when Hydra is heated from one side it does
not move directly away from the source of heat, but merely moves in
some random direction. In other words, the animal when heated
merely tries a new position.

Moebius (1873) describes the reaction of a large mollusk (Nassa) to
chemical stimuli, as shown when a piece of meat is thrown into the
aquarium containing them. They do not orient themselves in the lines
of diffusion and travel directly toward the meat, but move " now to the
right, now to the left, like a blind man who guides himself forward by
trial with his stick. In this way they discover whether they are com-
ing nearer or going farther away from the point from which the attrac-
tive stimulus arises" (/. c, p. 9, translation). The reaction is thus a
clear case of the method of trial and error. Experiments on the leech,

* By Mr. George Wagner.

THE METHOD OF TRIAL AND ERROR. 245

by Miss Frances Dunbar, which I hope may soon be published, show
that that animal finds its food in a similar manner. All searching is,
of course, behavior on the plan of trial and error, and many organisms
are known to search for food.

The righting reactions of organisms are among the most striking
examples of trial and error in behavior. In the starfish, for example,
when the animal is laid on its back " the tube feet of all the arms are
stretched out and are moved hither and thither, as if ieeling for some-
thing, and soon the tips of one or more arms turn over and touch the
underlying surface with their ventral side " (Loeb, 1900, p. 62). As
soon as these one or two arms have been successful, the others cease
their eflforts ; the attached arms then turn the body over. If all the
arms attempted to turn the animal at the same time, in other words, if
there were no way of recognizing " success " in the trial, the animal
could not right itself.

The righting reaction of the starfish shows much resemblance to the
method, described above, by which a suspended Amoeba passes to a
solid. It is probable that a Difflugia, turned with the opening of the
shell upward, would show a righting reaction essentially similar to
that of the starfish.

The righting reaction of the flatworm Planaria, as described by Pearl
(1903), is not so evidently brought about through the method of trial
and error. Yet there are certain facts that indicate that this method is
really essentially present here. Thus, Pearl shows that when the flat-
worm is prevented from righting itself in the usual way, its rights itself
in another manner. Probably various reactions are tried ; if the first
does not succeed another may.

This peculiar form of the method of trial and error, in which several
different reactions are tried even under but a single stimulus, is brought
out by Mast (1903) in the behavior of Planaria under other conditions.
If the water containing the flatworms is heated, the animals give, as the
temperature rises, practically all the reactions that they ever give under
any conditions.

We have thus, as the temperature rises and the stimulation increases, the fol-
lowing reactions given consecutively: positive, negative, crawling, righting,
and final (all the reactions described by Pearl, with the exception of the food
reactions, and the final reaction in addition). (Mast, 1903, p. 185.)

We shall have occasion to inquire as to the significance of this
responding to the same stimulus by many different reactions when we
take up, in another connection, certain similar phenomena in Stentor.

In the higher vertebrates, as we have mentioned at the beginning,
the method of trial and error plays a very large part. It is here espe-
cially that it has been recognized as a definite type of behavior in the

246 THE BEHAVIOR OF LOWER ORGANISMS.

work of Lloyd Morgan, Thorndike, and others. With the details of its
manifestations in higher animals we need not concern ourselves here.

Let us now return to the method of trial and error in the infusoria.
Here it is most strongly marked, and certain general problems which
arise from it are sharply defined. It is possible to formulate the reac-
tion method of the infusoria as follows : When effectively stimulated
by agents of almost any sort the organism moves backward, and turns
toward a structurally defined side X^ while at the same time it may
continue to revolve on its long axis. As thus stated, the reaction
method seems exceedingly simple and stereotyped, and as such I pre-
sented the behavior of these organisms in a former general paper.* If
we limit ourselves to a consideration of the reaction itself, this seems
inevitable, although certain additional reactions have been described
since that paper was written. But when we study the relation of this
reaction method to the environmental conditions, the results are most
remarkable, and a totally new set of problems appears. Whether the
behavior is to be called simple and stereotyped, or complicated and
flexible, is not so easy to decide. The relations of the reaction of the
environmental conditions are, perhaps, the really essential point in
animal behavior. What are the relations which we find in the organ-
isms reacting in the way set forth above }

In general terms we find that through this reaction by trial and error
the organisms are kept in conditions favorable to their existence, and
prevented from entering unfavorable regions. Through it they keep
out of hot and cold regions and collect in regions of moderate temper-
ature. Through it they tend to keep out of strong or injurious chemi-
cals and out of regions where the osmotic pressure is much above or
below that to which they are accustomed. Through it they gather in
regions containing small amounts of certain chemicals, not leaving
them for regions where there is either more or less of these chemicals.
When oxygen is needed they collect through this reaction in regions
containing oxygen ; when the oxygen pressure is high, they do not
react with reference to oxygen, or through this reaction they avoid
regions containing much oxygen. Through this reaction organisms
which contain chlorophyll, and therefore need light, gather in lighted
regions or move toward the source of light ; through the same reaction
the same organisms avoid very powerful light.f In all these cases.

* Psychology of a Protozoan, Amer. Journ. Psychol., vol. 10, 1899. pp. 503-515.

t For details, see the author's Studies, already referred to, and the preceding
contributions. In papers by Engelmann (1882, a) and Rothert (1901) the reac-
tion method involved is also described for certain organisms, though these
writers, like the present author in his earlier papers, did not bring out the rela-
tions to a general method of trial and error. Engelmann, however, character-
ized the reaction of Euglena to light directly as a " Probiren" — a " trial."

THE METHOD OF TRIAL AND ERROR. 247

when there is error the organism goes back and tries a new direction,
or a whole series of new directions.

But what constitutes "error" ? This is a fundamental question for this
method of behavior. Why does the organism react to some things by
turning away and trying new directions, to others not? Why do they
react thus on coming to certain chemicals, and on leaving others.?
Why do they react thus on coming to a strong chemical, and also on
leaving a weak solution of the same chemical? Why does the same
organism react thus to strong light, and also to darkness? To heat
and also to cold? What decides whether a certain condition is '* er-
ror '* or not? A list of all the different agents that must be considered
" error" from the standpoint of this reaction method reveals, so far as
chemical or physical classification is concerned, a most heterogeneous
and even contradictory collection. What is the common factor which
makes them all error?

Examination shows that error from the standpoint of this behavior
is as a rule error also from the standpoint of the general interests of
the organism, considering as the interests of the organism the perform-
ance of its normal functions, the preservation of its existence, and the
production of posterity. In general the organism reacts as error to
those things which are injurious to it, while in those conditions which
are beneficial it continues its normal activities. There are some excep-
tions to this, but in a general view it is clearly evident. There is no
common thread running through all the different agents which consti-
tute '* error" in the reactions, save this one, that they are error from
the standpoint of the general interests of the organism.

How can we account for the fact that these lowest organisms react
to all sorts of things that are injurious to them by a reaction which
tends to remove them from the action of the agent, — by a negative reac-
tion? The first response to this question must be another question.
How can we account for the fact that in man we have the same condi-
tion of affairs? How does it happen that we respond by drawing back
both from flame and from ice, though these act physically in opposite
ways? Why do we seek light, but avoid a blinding glare? Why do
we receive without opposition certain chemical stimuli (odors and
tastes) and avoid others? The facts are quite parallel in man and in
the lowest organisms in these respects. In man certain stimuli cause
reactions which tend to remove the organism from the source of the
stimulus (negative reactions), while others have the opposite effect;
this is true also of Euglena and Paramecium. In both cases the stimuli
which produce the negative reaction form a heterogeneous collection
from the chemical or physical standpoint. In both cases the stimuli
producing the negative reaction are in general injurious to the organism.
The problem is one for the highest and for the lowest organisms.

248 THE BEHAVIOR OF LOWER ORGANISMS.

In ourselves the stimuli which induce the negative reaction bring
about the subjective state known as pain (in varying degrees, from dis-
comfort or dislike to anguish), and popularly we consider that the draw-
ing back is due to the pain. Is there ground for this view? Or is the
reaction entirely accounted for by the chemical and physical processes
involved? When the burnt child draws its hand back from the flame,
does the state of consciousness called pain have anything to do with
the reaction ?

Without attempting to answer this question, we wish to point out
the bearing of possible answers on our problem in the lower organisms.
If we hold that in man we cannot account for the reaction without tak-
ing into consideration the pain, then we must hold to the same view
for the lower organisms. If we maintain that in man we cannot
account for the selection of such a heterogeneous group of conditions for
the negative response — conditions seeming to have nothing in common
save that they cause pain — without taking into consideration the pain,
then we are forced to the same conclusion in the unicellullar organisms,
for here we have a precisely parallel series of phenomena. Anyone,
then, who holds that pain is a necessary link in the chain of events in
man must consider that we are undertaking a hopeless task in trying to
account for the reactions of the lower organisms purely from the chem-
ical and physical conditions. And the converse is also true. Anyone
who holds that we can account fully for the reactions of Euglena or
Paramecium, purely from the physico-chemical conditions, without
taking into account any states of consciousness, must logically hold that
we can do the same in man. The method of trial and error implies
some way of distinguishing error ; the problem is : How is this done?
The problem is one, so far as objective evidence goes, throughout the
animal series.

We can, of course, know nothing of pain in any organism except
the self, and we can, in a purely formal way at least, solve our problem
equally well (or equally ill) without taking pain directly into consider-
ation. Even in man we must hold that pain is preceded or accom-
panied in every case by a certain physiological condition. And if there
is something common to all states of pain, it would appear that there
must be something common to all the physiological states accompany-
ing or preceding pain. We thus get a common basis for all the nega-
tive reactions ; if they are preceded or accompanied by a common
physiological state, this state will serve formally as an explanation for
the common reaction, fully as well as would a common state of con-
sciousness. The facts could be formulated as follows : In any animal,
from the lowest up to man, a certain heterogeneous set of agents, which
are in general injurious, produce a certain physiological state, common

THE \CETHOD OF TRIAL AND ERROR. 249

to all ; as a consequence or concomitant of this physiological state a
negative reaction follows. In man this physiological state is accom-
panied by pain. The common physiological state might then properly
receive a name which brings out its relation to the state accompanied
in man by pain ; for example, J. Mark Baldwin's *' organic analogue
of pain.''

We have left under this formulation the fundamental question as to
how a set of agencies that are quite heterogeneous from a chemical or
physical standpoint can produce a common physiological state. This
question is, of course, of precisely the same order of difficulty as that
which asks how the same heterogeneous set of agents can produce a
common state of consciousness, namely, pain. We therefore lose noth-
ing, so far as this problem is concerned, by substituting '•'• a common
physiological state" for "pain" in dealing with the subject. But to
attempt to deal with the problem of negative reactions in the lower
organisms without recognizing that they are conditioned in the same
way as the negative reactions of man — without admitting the existence
of some physiological state analogous to that which is accompanied by
pain in man, is, I believe, to close one's eyes to patent realities.

We have seen above that the method of trial and error involves some
way of distinguishing error. But do not some of the facts indicate
that it involves, at least sometimes, also some way of distinguishing the
opposite of error ; that is, what we may call success.^ For most of the
reactions of the infusoria this seems not necessary, for what the organ-
ism does when successful is merely to continue the condition in which
it finds itself at the time. There is then no objective evidence that a
stimulus is acting at this time. In these organisms it seems to be chiefly
the injurious or negative stimuli that induce a motor reaction. But
consider the floating Amceba, which sends forth pseudopodia in all
directions. Finally one of these pseudopodia comes in contact with
a solid, and to this stimulus the Amceba reacts positively. Now all
the other pseudopodia, though the external conditions directly affecting
them remain the same, become retracted, and the whole Amceba moves
toward the pseudopodium in contact. This withdrawal of the other
pseudopodia requires for its explanation a change in physiological state
which can be due only to the success of the pseudopodium that has
come in contact with a solid. There is certainly no basis here for con-
sidering the reaction as due to an " organic analogue of pain" ; possi-
bly a case could be made out, on the other hand, for a physiological
state corresponding to that which conditions pleasure in ourselves.
Possibly similar considerations hold for the positive reactions of other
organisms — infusoria, etc. — to solids.

There appears to be a similar state of affairs in the righting reaction.

250 THE BEHAVIOR OP LOWER ORGANISMS.

When the starfish is placed on its back, the physiological state existing
induces all the arms to initiate feeling movements ; the animal tries to
reach a solid with its tube feet. As soon as one or two arms have suc-
ceeded, this success is recognized by the cessation of effort on the part
of the other arms. Their physiological state has changed to one cor-
responding to success.

We may sum up our discussion on these points as follows : The
method of trial and error involves some way of distinguishing error,
and also, in some cases at least, some method of distinguishing success.
The problem as to how this is done is the same for man and for the
infusorian. We are compelled to postulate throughout the series cer-
tain physiological states to account for the negative reactions under
error, and the positive reactions under success. In man these physio-
logical states are those conditioning pain and pleasure.

The *' method of trial and error," as this phrase is used in the
present paper, is evidently the same as reaction by " selection of over-
produced movements," which plays so large a part in the theories of
Spencer and Bain and especially in the recent discussions of behavior
by J. Mark Baldwin. To this aspect of the matter the present writer
will return in the future.

This method of trial and error, which forms the most essential feature
of the behavior of these lower organisms, is in complete contrast with
the tropism schema, which has long been supposed to express the
essential characteristics of their behavior. The tropism was conceived
as a fixed way of acting, forced upon the organism by the direct action
of external agents upon its motor organs. Each class of external agents
had its corresponding tropism ; under its action the organism performed
certain forced movements, usually resulting in its taking up a rigid
position with reference to the direction from which the stimulus came.
Whether it then moved toward or away from the source of stimulus
was determined by accidental conditions, and played no essential part
in the reaction. There was no trial of the conditions ; no indication
of anything like what we call choice in the higher organisms ; the
behavior was stereotyped. Doubtless such methods of reaction do
exist. In the reactions of infusoria to the electric current (an agent
with which they never come into relation in nature), there are cer-
tain features which fit the tropism schema, and in the instincts — the
'' Triebe" — of animals there are features of this stereotyped character.
The behavior of animals is woven of elements of the most diverse kind.
But certainly in the lower organisms which we have taken chiefly into
consideration the behavior is not typically of the stereotyped character
expressed in the tropism schema. The method of trial and error is
flexible ; indeed, plasticity is its essential characteristic. Working in

THE METHOD OF TRIAL AND ERROR. 25 1

the lowest organisms with very simple factors, it is nevertheless capable
of development ; it leads upward. The tropism leads nowhere ; it is
a fixed, final thing, like a crystal. The method of trial and error on
the other hand has been called the ''method of intelligence" (Lloyd
Morgan, 1900, p. 139) ; it involves in almost every movement an
activity such as we call choice in higher organisms. With the acquire-
ment of 2i finer perception of differences the organism acting on the
method of trial and error rises at'once to a higher grade in behavior.
Combining this with the development of sense organs and the diflferen-
tiation of motor apparatus, the path of advancement is wide open
before it.

The most important step in advance is that shown when the results
of one reaction by trial and error become the basis for a succeeding
reaction. The method of reacting which leads to success is determined
by trial ; after it is once or several times thus determined, the trials are
omitted, and the organism at once performs the successful action.
This is intelligence^ according to Lloyd Morgan (1900, p. 138), and
it is as leading to this result that the method of trial and error can be
characterized also as the method of intelligence. In intelligent action,
while the organism must react the first time by the method of trial and
error, it need not begin all over again each time the same circumstances
are presented. Do we find any indication of such action among uni-
cellular organisms?

In Stentor we find action of this character to a certain extent. It
does not continue reacting strongly to a stimulus that is not injurious,
but after a time, when such a stimulus is repeated, it ceases to react,
or reacts in some less pronounced way than at first. To an injurious
stimulus, on the other hand, it does continue to react, but not through-
out in the same manner. When such a stimulus is repeated, Stentor
tries various different ways of reacting to it. If the result of reacting
by bending to one side is not success, it tries reversing the ciliary cur-
rent, then contracting into its tube, then leaving its tube, etc. (details
in Jennings, 1902, a). This is clearly the method of trial and error
passing into the method of intelligence, but the intelligence lasts for
only very short periods. To really modify the life of the organism in
any permanent way, as happens in higher animals, the method of
reacting discovered to be successful by the method of trial and error
should persist for a long time. Apparently this is not the case for
unicellular organisms, but further work is needed on this point.

An application of the method of trial and error similar to that of
Stentor is found under certain circumstances in the flatworm. Pearl
(1903) found that after the animal had reacted to a repeated mechanical
stimulus for a long time by turning away from it, it suddenly reversed

252 THE BEHAVIOR OF LOWER ORGANISMS.

the reaction, and turned far toward the side on which the stimulus was
acting. Mast (1903) showed, as we have seen, that when the flatworm
is heated it tries successively almost every form of reaction which it
has at command. Such results have a most important bearing on the
problem of the relation of the reaction method to the stimulus. Neither
direct action of the stimulus on the motor organs as separate entities,
nor a typical fixed interconnection of sense organs and motor organs
can explain such results. As a result of continued strong stimulation
the organism passes from one physiological state to another, and each
physiological state has its concomitant method of reaction.

The present paper may be considered as the summing up of the
general results of several years' work by the author on the behavior of
the lowest organisms. This work has shown that in these creatures
the behavior is not as a rule on the tropism plan — a set, forced method
of reacting to each particular agent — but takes place in a much more
flexible, less directly machine-like way, by the method of trial and
error. This method involves many of the fundamental qualities which
we find in the behavior of higher animals, yet with the simplest possi-
ble basis in ways of action ; a great portion of the behavior consisting
often of but one or two definite movements, movements that are stereo-
typed when considered by themselves, but not stereotyped in their
relation to the environment. This method leads upward, offering at
every point opportunity for development, and showing even in the
unicellular organisms what must be considered the beginnings of intel-
ligence* and of many other qualities found in higher animals. Tropic
action doubtless occurs, but the main basis of behavior is in these
organisms the method of trial and error.

♦Throughout this paper a number of terms are used whose significance as
they are commonly employed is determined by our subjective experience. But
all these terms (save those directly characterized as " subjective states," or
"states of consciousness") will be found susceptible also of definition from cer-
tain objective manifestations, and it is in this objective sense that they are used
in the present paper. Thus *' perception " of a stimulus means merely that the
organism reacts to it in some way ; " discrimination " of two stimuli means that
the organism reacts differently to them ; " intelligence " is defined by the objective
manifestations mentioned in the text, etc. These terms are employed because it
would involve endless circumlocution to avoid them; they are the vocabulary
that has been developed for describing the behavior of men, and if we reject them,
it is almost impossible to describe behavior intelligibly. When their objective
significance is kept in mind there is no theoretical objection to them, and they
have the advantage that they bring out the identity of the objective factors in the
behavior of animals with the objective factors in the behavior of man.

LITERATURE CITED.

Baldwin, J. Mark.

1897. Mental development in the child and in the race. Methods and pro-
cesses. Second Edition. 4^ pp. New York.
Bernstein, J.

1900. Chemotropische Bewegungen eines Quecksilbertropfens. Arch. f. d.
ges. Physiol., Bd. 80, pp. 628-637.
Berthold, G.

1886. Studien iiber Protoplasmamechanik. 332 pp., 7 pi. Leipzig.
Blochmann, F.

1894. Kleine Mitteilungen ilber Protozoen. Biol. Centralb., Bd. 14, pp.

82-91.

BiJTSCHLI, O.

1878. Beitrage zur Kenntniss der Flagellaten und einiger verwandten Organ-

ismen. Zeitschr. f. wiss. Zool., Bd. 30, pp. 205-281, pis. 11-15.
1880. Protozoa, I Abth. Bronn's Klassen und Ordnungen des Thierreichs.

Leipzig & Heidelberg.
1892. Untersuchungen iiber mikroskopische Schaume und das Protoplasma.
234 pp., 6 pi. Leipzig.
Carter, H. J.

1863. On Amoeba princess and its reproductive cells, compared with ./^tha-
liiitn, Pythium, Mucor, and Achlya. Ann. and Mag. of Nat. Hist. (3),
Vol. 12, pp. 30-52, pi. 3.
Claparede, Ed., et Lachmann, J.

1858-9. Etudes sur les infusoires et les rhizopodes. T. I. 482 pp., 24 pi.
Geneve.
Davenport, C. B.

1897, Experimental Morphology. Vol. I. 280 pp. New York.
Engelmann, T. W.

1882. Bacterium photometricum, ein Beitrag zur vergleichenden Physiologic
des Licht- und Farbensinnes. Arch. f. d. ges. Physiol., Bd. 30, pp.
95-124. PI. I.
1882 a. Ueber Licht- und Farbenperception niederster Organismen. Arch.
f. d. ges. Physiol., Bd. 29, pp. 387-400.
Garrey, W. E.

1900. The effect of ions upon the aggregation of flagellated infusoria. Amer.
Journ. Physiol., Vol. 3, pp. 291-315.
Groom, T. T. , & Loeb, J.

1890. Der Heliotropismus der Nauplien von Balanus perforatus und die pe-
riodischen Tiefenwanderungen pelagischer Tiere. Biol. Centralb.,
Bd. 10, pp. 160-177.
Harrington, N. R., and Leaming, E.

1900. The reaction of Amoeba to light of different colors. Amer. Journ. of

Physiol., Bd. 3, pp. 9-18.
Hodge, C. F., & Aikins, H. A.

1895. The daily life of a Protozoan ; a study in comparative psychophysi-

ology. Amer. Journ. Psychology, Vol. 6, pp. 524-533.
Holmes, S. J.

1901. Phototaxis in the Amphipoda. Amer. Journ. Physiol., Vol. 5, pp.

211-234.
1903. Phototaxis in Volvox. Biol. Bulletin, Vol. 4, pp. 319-326.
Holt, E. B., and Lee, F. S.

1901. The theory of phototactic response. Amer. Journ. Physiol., Vol. 4,
pp. 460-481.
Jennings, H. S.

1897. Studies on reactions to stimuli in unicellular organisms. L Reactions
to chemical, osmotic, and mechanical stimuli in the ciliate infusoria.
Jour, of Physiol., Vol. 21, pp. 258-322.

253

254 "^"^ BEHAVIOR OF LOWER ORGANISMS.

Jennings, H. S.

1899. Studies, etc. II. The mechanism of the motor reactions of Parame-

cium. Amer. Journ. Physiol., Vol. 2, pp. 311-341.

1899 a. Studies, etc. IV. Laws of chemotaxis in Paramecium. Amer.

Journ. Physiol., Vol. 2, pp. 355-379.

1900. Studies, etc. V. On the movements and reactions of the Flagellata

and Ciliata. Amer. Journ. Physiol., Vol. 3, pp. 229-260.

1900 a. Reactions of infusoria to chemicals. A criticism. Amer. Natural.,

Vol. 34, pp. 259-265.

1901. On the significance of the spiral swimming of organisms. Amer. Nat-

uralist, Vol. 35, pp. 369-378.

1902. Artificial imitations of protoplasmic activities and methods of demon-

strating them. Journ. of Appl. Microsc. and Lab. Methods, Vol. 5,
pp. 1597-1602.
1902 rt. Studies, etc. IX. On the behavior of fixed infusoria {Stenior and
Vorticella)^ with special reference to the modifiability of protozoan
reactions. Amer. Journ. Physiol., Vol. 8, pp. 23-60.

1903. Rotatoria of America. II. A monograph of the i?a^/«//e/<F. Bull. U. S.

Fish Commission for 1902, pp. 273-352.
Jennings, H. S., and Crosby, J. H.

1901. Studies, etc. VII. The manner in which bacteria react to stimuli,

especially to chemical stimuli. Amer. Journ. Physiol., Vol. 6, pp.

31-37-
Jennings, H. S., and Jamieson, Clara.

1902. Studies, etc. X. The movements and reactions of pieces of ciliatc

infusoria. Biol. Bull., Vol. 3, pp. 225-334.
Jensen, P.

1896. Ueber individuelle physiologische Unterschiede zwischen Zellen der

gleichen Art. Arch. f. d. ges. Physiol., Bd. 62, pp. 172-200.

1900. Ueber den Aggregatzustand des Muskels und der lebendigen Substanz

Uberhaupt. Arch. f. d. ges. Physiol., Bd. 80, pp. 176-228.

1901. Untersuchungen iiber Protoplasmamechanik, Arch. f. d. ges. Physiol.,

Bd. 87, pp. 361-417-

1902. Die Protoplasmabewegung. Asher & Spiro's Ergebnisse der Physiol.,

erster Jahrgang. 42 pp.

KiJHNE, W.

1864. Untersuchungen iiber das Protoplasma und die Contractilitat. 158
pp., 8 pi. Leipzig.
Le Dantkc, F.

1894. Etudes biologiques comparatives sur les Rhizopodes lob^s at r^ticulds

d'eau douce. Bull, scient. de la France et de la Belgique. T. 26, pp.
56-99.

1895. Sur I'adh^rence des Amibesaux corps solides. C. R. Acad. Sci., Paris.

T. 120, pp. 210-213.
Leidy, J.

1879. Freshwater Rhizopods of North America. Report of the U. S. Geol.
Survey of the Territories. Vol. 12, 324 pp., 48 pi.
Loeb, J.

1893. Ueber kiinstlicheUmwandlung positiv heliotropischer Thiere in nega-
tiv und umgekehrt. Arch. f. d. ges. Physiol., Bd. 54, pp. 81-107.

1897. Zur Theorie der physiologischen Licht- und Schwerkraftwirkungen.

Arch. f. d. ges. Physiol., Bd. 66, pp. 439-466.
1900. Comparative physiology of the brain and comparative psjchologj.
309 pp. New York.
Mast, S. O.

1903. Reactions to temperature changes in Spirillum, Hydra, and fresh-water

Planarians. Amer. Journ. Physiol., Vol. 10, pp. 165-190.
Mendklssohn, Maurice.

1902 a. Recherches sur la thermotaxie des organismes unicellulaires. Journ.
de Physiol, et de Path, gen., T. 4, pp. 393-410.

1902 b. Recherches sur I'interfdrence de la thermotaxie avec d'autres tac-
tismes et sur le mecanisme du mouvement thermotactique. Ibid., do.
475-488.

1902 c. Quelques considerations sur la nature et la role biologique de la ther-
motaxie. Ibid., pp. 489-496.

LITERATURE CITED. 255

MOEBIUS, K.

1873. Die Bewegungen der Thiere und ihr psychischer Horizont. 20 pp.
Kiel.
Moore, Anne.

1903. Some facts concerning the geotropic gatherings of Paramecia. Amer.
Journ. Physiol., Vol. 9, pp. 238-244.
Morgan, C. Lloyd.

1894. Introduction to Comparative Psychology, 382 pp. London.
T900. Animal Behavior. 344 pp. London.
Naegeli, C.

i860. Ortsbewegungen der Pflanzenzellen und ihren Theile (Stromungen).
Beitrage zur wissenschaftlichen Botanik, Heft 2, pp. 59-108.
Nagel, W. a.

1894. Beobachtungen Uber den Lichtsinn augenloser Muscheln. Biol. Cen-
tralb., Bd. 14, pp. 385-390.

OSTWALD, WiLHELM.

1902. Vorlesungen uber Naturphilosophie. 457 pp. Leipzig.
OsTWALD, Wolfgang.

1903. Zur Theorie der Richtungsbewegungen schwimmender niederer Organ-

ismen. Arch. f. d. ges. Physiol., Bd. 95, pp. 23-65.
Pearl, R.

1900. Studies on electrotaxis. I. On the reactions of certain infusoria to

the electric current. Amer. Journ. Physiol., Vol. 4, pp. 96-123.
1903. The movements and reactions of fresh-water Planarians. A study in
animal behavior. Quart. Journ, Micr. Sci., Vol. 46, pp. 509-714.
Penard, Eug,

1890. Etudes sur les Rhizopodes d'eau douce. M6m. de la Soc. de Physique

et d'Histoire naturelle de Geneve, T. 31, pp. 1-230. Pis. i-ii.

1902. Faune rhizopodique du bassin du Leman. 714 pp. Geneve.
Provi'azek, S.

1900, Protozoenstudien, IL Arb, a, d. zool. Inst. Wien, Bd. 12, pp. 243-297,

298-300.

1901, Beitrage zur Protoplasma-Physiologie. Biol. Centralb., Bd. 21, pp.

87-95.
Putter, Aug.

1900. Studien iiber Thigmotaxis bei Protisten. Arch. f. Anat. u. Physiol.,

physiol. Abth., Supplementband, 1900, pp. 243-302.
Quincke, G.

1888. UeberperiodischeAusbreitung von Fliissigkeitsoberflachen und dadurch

hervorgerufene Bewegungserscheinungen. Sitzb. d. kgl. preuss.

Akad. d. Wiss. zu Berlin, Bd, 34, pp. 791-804.
Radl, Em,

1903. Untersuchungen iiber den Phototropismus der Tiere, Leipzig. 188 pp.

ROESLE, E.

1902, Die Reaktion einiger Infusorien auf einzelne InduktionsschlJlge.

Zeitschr, f, allg. Physiol., Bd. 2, pp. 138-168.

ROTHERT, W.

1901. Beobachtungen und Betrachtungen iiber tactische Reizerscheinungen.

Flora, Bd. 88, pp. 371-421.
Roux, W.

1891. Ueber die " morphologische Polarisation " von Eiern und Embryonen

durch den electrischen Strom. Sitz.-Ber. d. k. Akad. d. Wissensch.
zu Wien, Math. u. Naturw. Classe, Bd. ioi,pp. 27-228 (Ges. Abhdlg.,
Bd, II, pp, 540-765).
Rhumbler, L.

1898. Physikalische Analyse von Lebenserscheinungen der Zelle. I. Bewe-
gung, Nahrungsaufnahme, Defakation, Vacuolen- Pulsation, und
Gehausebau bei lobosen Rhizopoden. Arch. f. Entw.-mech. der
Organismen, Bd. 7, pp. 103-350.

1902. Der Aggregatzustand und die physikalischen Besonderheiten des

lebenden Zellinhaltes. Zeitschr. f. allg. Physiol., Bd. i, pp. 279-388;
Bd. 2, pp. 183-340.
Schulze, F. E.

1875. Rhizopodenstudien. IV. Arch. f. Mikr. Anat., Bd. n, pp. 329-353,
pis. 18-19.

256 THE BEHAVIOR OF LOWER ORGANISMS.

SOSNOWSKI, J.

1899. Untersuchungen uber die Veranderungen der Geotropismus bei Para-

mecium aurelia. Bull. Internat. de I'Acad. d. Sci. de Cracovie, Mars,

1899, pp. 130-136-
Stahl, E.

1884. Zur Biologie der Myxomjceten. Bot Zeitung, Jhrg. 40, pp. 146-155 ;

162-175; 187-191.
Strasburger, E,

1S78. Wirkung des Lichtes und der Warme auf Schwarmsporen. Jenaische

Zeitschr., N. F., Bd. 12, pp. 551-625. Also separate, Jena. 75 pp.

(The citations are to pages of the separate reprint.)
Thorndike, E. L.

1898. Animal intelligence. An experimental study of the associative pro-

cesses in animals. The Psj^cho logical Review, Monograph Supple-
ment 2. 109 pp.
Tov^'LE, Eliz. W.

1900. A study in the heliotropism of Cypridopsis. Amer. Journ. Physiol.,

Vol. 3, pp. 345-365.
Uexkull, J. V.

1899. Die Physiologie der Pedicellarien. Zeitschr. f. Biol., Bd. 37, pp. 334-

403.

1900. Die Phjsiologie des Seeigelstachels. Ibid., Bd. 39, pp. 73-112.

1900 a. Die Wirkung von Licht und Schatten auf die Seeigel. Ibid., Bd. 40,

pp. 447-476.
1903. Studien iiber den Tonus. I. Der biologische Bauplan von Sipunculus
nudus. Ibid., Bd. 44, pp. 269-344.
Verworn, M.

1889. Psycho-physiologische Protistenstudien. Experimentelle Untersuch-

ungen. 219 pp., 6 pi. Jena.

1890. Biologische Protistenstudien, II. Zeitschr. f. Wiss. Zool., Bd. 50, pp.

443-468, pi. 18.
1890 a. Die polare Erregung der Protisten durch den galvanischen Strom.
Arch. f. d. ges. Physiol., Bd. 46, pp. 281-303.

1891. Die physiologische Bedeutung des Zellkernes. Arch. f. d. ges. Physiol.,

Bd. 51, pp. 1-118, pis. 1-6.

1892. Die Bewegung der lebendigen Substanz. 103 pp. Jena.

1895. Allgemeine Physiologie. Ein Grundriss der Lehre vom Leben. 584

pp. Jena.
1897. Die polare Erregung der Protisten durch den galvanischen Strom. IV.

Mittheilung. Arch. f. d. ges. Physiol., Bd. 65, pp. 47-62.

1899. General Physiology. An Outline of the Science of Life. Transl. by

F. S. Lee. New York. 615 pp.
Wager, H.

1900. On the eye-spot and flagellum of Englena viridis. Journ. Lin. Soc.

London, Zool., Vol. 27, pp. 463-481.
Wallengren, Hans.

1902. Zur Kenntnis der Galvanotaxis. Zeitschr. f. allg. Physiol., Bd. 2, pp.

341-384-

1903. Zur Kenntnis der Galvanotaxis. II. Eine Analyse der Galvanotaxis

bei Spirostomum. Ibid., Bd. 2, pp. 516-555.
Wallich, G. C.

1863 a. Further observations on an undescribed indigenous Amoeba, with
notes on remarkable forms of Actinophrys and Difflugia. Ann. and
Mag. of Nat. Hist. (3), Vol. 11, pp. 365-371, pi. 9.

1863 b. Further observations on the distinctive characters and reproductive
phenomena of the amoeban Rhizopods. Ann. and Mag. of Nat. Hist.
(3), Vol. 12, pp. 329-337.

1863 c. Further observations on the distinctive characters, habits, and repro-
ductive phenomena of the amoeban Rhizopods. Ann. and Mag. of
Nat. Hist. (3), Vol. 12, pp. 448-468, pi. 8.
Yerkes, R. M.

1900. Reactions of Entomostraca to stimulation by light. II. Reactions of
Daphnia and Cypris. Amer. Journ. Physiol., Vol. 4, pp. 405-423.

THE UNIVERSITY LIBRARY

UNIVERSITY OF CALIFORNIA, SANTA CRUZ

SCIENCE LIBRARY

This book is due on the last DATE stamped below.

To renew by phone, call 459-2050.

Books not returned or renewed within 14 days

after due date are subject to billing.

Series 2477

nr'n ht"

ni r

i

I, 3 2106 00256 1139