Gi7ésl
BF 3

New York
State College of Agriculture

At Cornell University
Ithaca, N. B.

Library

~

Cornell University Libra

The development of certain instincts and

Cornell University

The original of this book is in
the Cornell University Library.

There are no known copyright restrictions in
the United States on the use of the text.

http://www.archive.org/details/cu31924002908410

BEHAVIOR MONOGRAPHS

Volume 1, Number 1, 1911 Serial Number |!

Edited by JOHN B. WATSON
The Johns Hopkins University

The Development of Certain Instincts and

Habits in Chicks

FREDERICK S. BREED

Published
at Cambridge, Boston, Mass.
HENRY HOLT & COMPANY
34 West 33d Street, New York
G. E. STECHERT & CO., London, Paris and Leipzig, Foreign Agents

The Journal of Animal Behavior

An organ for the publication of studies concerning the instincts,
habits and intelligence of organisms '

The Journal contains a Department of Notes in which appear brief accounts of
especially interesting and valuable observations of behavior.

Published bi-monthly at Cambridge, Boston, Mass., by Henry
Holt and Compatiy, New York. Each volume contains at least 450
pages with plates and text-figures. :

“The subscription price is $3.00 per volume (foreign subscription
$3.50) postpaid.

Manuscripts, subscriptions, and all business correspondence should
be addressed to

THE JOURNAL OF ANIMAL BEHAVIOR
; Emerson Hall, Cambridge, Massachusetts,

;

Behavior Monographs

For the publication of studies in behavior and intelligence which are
too lengthy or too expensive for acceptance
‘by The Journal of Animal Behavior

‘Published at, irregular intervals at Cambridge, Boston, Mass., in
-connection with the Journal of Animal Behavior, by Henry Holt and
Company, New York. Each volume contains approximately 450
pages with plates and text-figures.

The subscription price is. $3.00 per volume (foreign subscription
$3.50) postpaid.

Monographs may be purchased separately Be prices varying with
_the cost of manufacture.

Manuscripts and inquiries concerning terms of publication should
be addressed to the Editor of the Behavior Monographs,
JOHN B. WATSON,
The Johns Hopkins University,
: Baltimore, Md.

Subscription: to The Journal of Animal Behavior and the Behavior

Monographs should be sent to Emerson Hall, Cambridge, Massa-
chusetts.

BEHAVIOR MONOGRAPHS

Volume 1, Number |, 1911 Serial Number |

Edited by JOHN B. WATSON
The Johns Hopkins University

The Development of Certain Instincts and
Habits in Chicks

FREDERICK S. BREED

Assistant Professor of Education

The University of Michigan

From the Harvard Psychological Laboratory

Wa
X

Published
at Cambridge Boston, Mass.
HENRY HOLT & COMPANY

24 West 33d Street, New York
G. E. STECHERT & CO., London, Paris and Leipzig, Foreign Agents
loos

Mn

CONTENTS

PAGE
PART I. INSTINCTIVE REACTIONS. .............cccce cece en eeees 1
I. Introduction and Statement of the Problem ................. 1
II. Previous Experimental Work on Chicks..................0005. 1
IW. (General Apparatus... c0ccsnnevcia gra ncveenae ke ps een ey 4
IV. Observations and Experiments............... 0.00 e cece eee 5
A, First Activities::5.. 43% .en¢5 nace tee. cede e Sse ebaes pen RE 5
Bex PURI OIRAM Bee 2 act oe evs dctncatsst dod cmon ens eh Geer gaa eg ad Sereda tan adcci dhs 8
Gi, STPPODICHG toc esd aitentle atanune tia ba Ca oee aaah S 8
b. Method and tests............ 0. ccc eee eee 8
ce. Discussion of results............ 00 sce e eee ee eens 11
CG. sBeckings vcezuvex unrated cea seman an gine eeaeN seen eS 14
a. Apparatus and method...................0.0000- 14
b. Pecking artificially deferred................. Sanince 16
c. The natural development of the pecking reaction, with
J a collateral study of the effects of social influence. 21
1, ‘Sources: of errorscs wewicinguleseeagnres aeictiaues 21
2. Data and their significance................. 26
PART II. ACQUIRED REACTIONS............. 00.0... 0 cee ee eens 41
I. Introduction and Statement of the Problem.................... 41
ED, AAP PATA tS i occ cettis ances nears eee ries nti oareance in dar ann oe Moa Big ated 42
III. Method.......... Dae Paces ible cone uy 4s sit dana eines 45
TV. “Byperimentsy ys mc ye0 Ses ages ue kias peerey eee hoe heme eys 48
Ay. (Colonia ss tues suis cemabael ee. cae tod oho aMe puke thoes 48
MB oe SORA sic sah se siesnt Meecha uenaes Shands alone aad a a aes De 53
Os.d SUZ ise ean sitaseun ie aeste phere einlans mele taste’ eke Eh hash obinha eth tien 57
D. To What Extent Has Training General Value?........... 59
E. Some Effects of Modifications.......................--. 67
F. The Persistence of Modifications........................ 71
Summary and Conclusions... . 6.06.66. 75

PART I. INSTINCTIVE REACTIONS
I. Introduction and Statement of the Problem

The experimental work reported in this monograph was car-
ried on during the years 1907, 1908, and 1909 in the Harvard
Psychological Laboratory. The variety of chick used in all the
work was the Barred Plymouth Rock. It is usually considered
by poultry breeders a hardy chick, and for this reason more
than any other was selected for these tests. The program of
research began with a cursory study of the first activities of
the animals, preliminary to a more detailed study of certain
instinctive modes of response such as drinking, pecking, and
“imitating.” In the case of drinking, interest centered in the
nature of the stimulus; in the case of pecking, in the accuracy
of the reaction; as for imitation, the question was asked, In
how far is social influence a means to improvement in the accu-
racy of pecking? After some knowledge of the natural tenden-
cies of the chicks had been gained, an attempt was made to
trace the course of development of certain habits of response
to optical stimuli, as well as to study the interrelation and
persistence of these habits. Quantitative methods were devised
and applied wherever they suggested themselves to the ex-
perimenter.

The investigation was pursued under the immediate direction
of Professor Robert M. Yerkes. What I owe to his searching
criticism and fertile suggestions, a mere word of acknowledgment
will not suffice to say. I shall be-only too glad if my immature
effort reveal some trace of his admirable scientific spirit. I am
also much indebted to my colleague, Professor Charles Scott Berry,
for a critical reading of the manuscript, and to Dr. William F.
Hauhart, Instructor in German, University of Michigan, for
valuable assistance in reading the proof.

II. Previous Experimental Work on Chicks

Attention is limited in this résumé to the conclusions of inves-
tigators in regard to the accuracy of the pecking response. The
results of some previous studies of the drinking reaction will be

2 FREDERICK 8. BREED

discussed in connection with the results of a study of the reac-
tion reported later in the paper.

Spalding’s observations on the pecking reaction have been
quoted widely. Concerning the pecking of his chicks he says:
“They did not attempt to seize things beyond their reach, as
babies are said to grasp at the moon; and they may be said to
have invariably hit the objects at which they struck—they never
missed by more than a hair’s breadth, and that too, when the
specks at which they aimed were no bigger, and less visible,
than the smallest dot of an 7. To seize between the points of
the mandibles at the very instant of striking seemed a more
difficult operation. I have seen a chicken seize and swallow an
insect at the first attempt; most frequently, however, they
struck five or six times, lifting once or twice before they suc-

ceeded in swallowing their first food.” (Italics mine.)
On the question of the accuracy of the pecking of chicks,
Preyer ? disagrees with Spalding. ‘‘I cannot admit,” he says,

“the supposed infallibility to within a hair’s breadth. They
miss in pecking by as much as two millimeters, though seldom.
On the other hand, the attempts at swallowing frequently fail.
Here it should be considered that even grown fowls are not sure
in their pecking, seizing, and swallowing, as any one that ob-
serves closely may easily perceive. The accuracy is, however,
marvelous at the very beginning.’ (Italics mine.)

Romanes,? though reporting no experimental work of his
own, deserves notice on account of his wide influence. He
quotes Spalding extensively and with approval. For Romanes,
perfection as applied to instinct means perfect adaptation in-
dependent of individual experience. He illustrates such perfec-
tion “by considering the wonderful accuracy of many among the
highly refined and complex adjustments which are manifested by
the newly-born young of the higher animals,” (italics mine),
citing first in his list of examples of perfection the pecking reac-
tion of chicks as reported by Spalding.

Eimer * experimented on chicks during two different years.
The impression he got of the accuracy of pecking appears in the

1 Spalding, D. A.: Instinct. With original observations on young animals.
Macmillan’s Mag., 1873, vol. 27, p. 284.

° Preyer, W.: The senses and the will. (1881.) Tr., New York, 1890, p. 67.

5 Romanes, G. J.: Mental evolution in animals. New York, 1884, p. 161.

‘Eimer, G. H. T.: Organic evolution. Tr., London, 1890, p. 246.

INSTINCTS AND HABITS IN CHICKS 3

following report of the reactions of one of his animals: ‘‘At the
first attempt it touched a grain with as much certainty as if it
had pecked at millet for eversolong. But it did not succeed in
taking up the grain in its beak. A second attempt immediately
afterwards likewise failed, but at the third, which succeeded
without a pause, the grain was grasped and swallowed, and now
the little creature went on feeding as though it had done so for
years!’ (Italics mine.)

Mills + thinks that some of Spalding’s statements exaggerate.
“Instinct is not the hard and fast thing it is sometimes sup-
posed to be.” On the basis of his own experiments he con-
cludes: ‘‘ Thus one (chick) may strike a crumb accurately every time
it pecks, and pick it up on the first attempt; another misses,
or shows great difficulty in getting it into the mouth.” (Italics
mine.)

Morgan ? concludes that the complicated process of striking,
seizing, and swallowing ‘‘is performed so soon, and with so
few trials—often at the third or fourth attempt—that one must
regard the whole as essentially congenital in its definiteness, and
look upon the few preparatory efforts as merely the steadying
of the inherited organic apparatus to its work.’’ (Italics mine.)
By congenital response is meant one that is not the result of
acquired skill. ;

Writing earlier on the same topic, Morgan ® had said the
chicks “required a little intelligence, acting by and through
experience, to perfect their (instinctive) activities. The in-
stincts were very nearly, but not quite, perfect.”

Thorndike‘ criticises Morgan’s conclusions on the improve-
ment in accuracy of the pecking reaction. ‘‘ Lloyd Morgan,
for instance, has chosen a dubious example of perfecting through
habit in the seizing of bits of food by chicks. They often do
fail to seize in their first experiences, as he observed, but they
often, perhaps just as often, fail even after long expertence.”’
(Italics mine.)

Finally, a further matter, aside from the question of the
_-! Mills, W.: The psychic development of young animals and its physical correla-
‘tion. Trans. R. 8. Can., Sec. IV, 1895; or, The nature and development of animal

‘intelligence, New York, 1898, p. 262.
2 Morgan, C. L.: Habit and instinct. London and New York, 1896, p. 37.
3 Morgan, C. L.: Introduction to comparative psychology. London, 1894, p. 207.

4Thorndike, E. L.: The instinctive reaction of young chicks. Psych. Rev.,
VI, 1899, p. 282. Also, Instinct. Biol. Lectures, Woods Hole, 1899, p. 61.

4 FREDERICK 8. BREED

accuracy of pecking. Katz and Révész' support Hess ? in the
view that the hungry chick does not peck at food which it does
not see. The former assert, for example, that a chicken in the
dark-room, even if placed directly in the food, will not peck
at it. The point happens to be vital to their method in some
important work.

It has seemed best to report the essential conclusions of the
writers mentioned, so far as possible, in their own words. It
may not be amiss, however, to note two general impressions
one gets from these studies of the pecking instinct: (1) that
the reaction has a high degree of congenital accuracy, and (2)
that it has less of this than was formerly supposed.

III. General Apparatus

The room used for these experiments had a western exposure.
It was equipped with an incubator? and a brooder* so that
chicks for the experiments might be hatched at any time during
the year in the laboratory. Such an arrangement also provided
a good opportunity for complete knowledge of the history of
the animals. For the sake of safety, economy, and convenience,
both brooder and incubator were heated with gas. In addition
to the regular thermostat on the incubator, a gas regulator °
was attached to insure a more constant temperature. No trouble
was experienced at any time in securing eggs of the special
kind desired. The incubator and ‘brooder were easily operated
after a careful study of the directions that came with the
machines. The percentage of fertile eggs hatched during the
first year was 65.6. At times two settings were run simulta-
neously in the same incubator.

"Katz, D. und Révész, G.: Experimentell-psychologische Untersuchungen mit
‘ Hihnern. Zeitschrift f. Psych. u. Physiol. d. Sinnesorgane, 1909, Bd. 50, S. 93.
~ 2 Hess, C.: Ueber Dunkeladaptation und Sehpurpur bei Htthnern und Tauben.
Archiv 7. Augenheilkunde, 1907, Bd. 57, S. 298-316.

——————: Untersuchungen tber Lichtsinn und Farbensinn der Tagvogel.
Ibid., 1907, Bd. 57, 8. 317-327.

————-: Untersuchungen uber die Ausdehnung des pupillomotorisch
wirksamen Bezirkes der Netzhaut und tiber die pupillomotorischen Aufnahme-
organe. Ibid., 1908, Bd. 58, 8. 182.

—————: Untersuchungen tber das Sehen und tber die Pupillenreaktion
von Tag-und von Nachtvégeln. Ibid., 1908, Bd. 59, S. 143-167.

§ Junior No. 2, sold by the Prairie State Incubator Co., Homer City, Pa.

4Indoor Brooder, No. 5, sold by the same firm.

5Single Float Standard Automatic Gas Regulator, secured from the Gas Con-
sumers’ Association, 18 Boylston St., Boston, Mass.

INSTINCTS AND HABITS IN CHICKS 5

Greater freedom of movement was provided for the birds by
the use of two cages that were connected with the brooder. For
litter, white sand and shredded alfalfa were found satisfactory.
Sand is more suitable than ordinary earth where it is desirable
to avoid dust. A chick food‘ that is regularly on the market
served our purposes very well. With the above simple equip-
ment the chicks for the experiments were hatched and kept in
the animal wing of the laboratory.

IV. Observations and Experiments
A. First Activities

If one would have the ontogenetic history of a chick, he must
begin his study of the development of activity while the chick
is still within the egg. A movement of the egg tray was some-
times found to be the occasion of a chirp within the egg hours
before the shell was chipped. Or, chirping at times occurred
so loud that it could be heard distinctly in the room in which
the incubator was situated, when the incubator was tightly
closed and no external stimuli, such as noises and jars, were
noticed. Before the shell was broken at any point, a continuous
tapping sound could often be heard within the egg, somewhat
as if the chick were pecking against the inside of the shell. The
movements of chicks just after the shell was chipped were
.observed through the glass door of the incubator and also seen
to good advantage when the eggs were removed from the incu-
bator and put in an appropriate receptacle under an electric
light. Upon these latter occasions the ragged edges of the egg
shell near the beak of the chicken were removed to give a better
view of the animal’s movements. Upon the observations follow-
ing the transfer of a chipped egg to a nest warmed by an electric
light, I find the following note: ‘‘I broke away much of the shell.
After bill of chick was entirely out of reach of the shell, the
chick continued to make a sound somewhat like pecking, and
that, too, without opening and closing the bill. The ‘ clicking ’
sounds seemed to accompany the rhythmical heaving of the
animal, which I took to be breathing. . . Eyes were
open, apparently, as soon as chick had unfolded itself.’ At
the next hatch another chick was observed under an electric

1 Cyphers Chick Food, sold by the Cyphers Incubator Co., Boston, Mass.

6 FREDERICK 8S. BREED

light in the same way. I follow my notes: ‘‘ Made at times
a regular clicking noise with breathing. Frequently clapped
mandibles together. Repeatedly raised bill with a

lifting motion of the head. This reaction might bring tip of
bill, and hence the bill scale, in contact with the shell. Opened
mouth and lifted bill. Eyes closed while in egg. Sometimes
pushed bill forward, following this movement with a clapping
together of the mandibles.”

While the chicks were still in the egg the legs were folded
well up on the breast and the head was turned down toward
the breast. After the egg was chipped so that the behavior
of the animal could be observed, the lifting movement of head
and beak above referred to was more frequent than any other.
One much less often saw anything like a forward thrust of
the bill.

The chicks came out of the eggs wet and, in a temperature
of 103° F., remained several hours before drying off. Spells of
vigorous activity alternated with periods of passivity during the
chicks’ struggles just prior to hatching, as well as for some
hours after hatching. The post-embryonic life of the chick in
its early hours seemingly prolonged without interruption the
life in the embryo. Chicks appeared to break the shell in two
by a lifting, struggling movement of the head accompanied by
a stretching, straightening movement of the legs. At least I
have observed this combination of movements at the moment
of hatching. Now if one watch carefully the behavior of a
chick for some time after exclusion, he finds that there are these
similarities between the later life within the shell and the earlier
life outside the shell. The positions of the head and legs within
the shell correspond to the positions assumed by these members
during the passive states shortly after hatching. In both cases
activity and passivity alternate, the periodicity of which alter-
nation seems to be largely determined by intra-organic stimuli.
If it be true that the legs participate in the action that finally
breaks the shell, as I think they do, the lifting of the head and
the pushing with the legs would be represented in the early
post-embryonic life by those pulses of activity in which the
chick lifts its head, rises to its feet, staggers a few steps, or
struggles a few moments, and lapses again into the passive state.
And just as the chick in the shell often claps its mandibles

INSTINCTS AND HABITS IN CHICKS 7

together without marked activity in other regions of the organ-
ism, so the chick during the first few hours after exclusion may
be seen to work its mandibles in a similar way without disturb-
ing its sitting posture. The pulses of activity, embryonic and
post-embryonic alike, are often accompanied by loud chirping.

One might expect to find more unmistakable evidencé
pecking reaction while the chick is yet within the egg. There
is an impression abroad that the chick pecks its way out of the
shell. When the position of the head in the embryo is taken
into consideration, one can see two reasons why most of the
movements of the head are lifting movements, and not pecking
reactions: (1) the lifting movement tends to free the head and
neck from their folded position, and (2) pecking would seem
to be a difficult matter with the head folded down on the breast
in this wise. Besides, the lifting reaction does actually break
shell and tear confining membranes, and so is effective in releas-
ing the chick. It may be found that the chick does not peck
its way out of the shell. However, the matter will need more
careful study. For hours I have watched chicks laboring in
the egg to discover a clear case of the pecking reaction. At
times, before the egg has been broken in two, one does see short,
quick, forward thrusts of the bill followed by working of the
mandibles. And chicks only a few hours out of the egg may
be observed repeatedly executing what might be called a pecking
reaction ‘‘into the air,’’ followed by clapping together of the
mandibles. The following note describes the reaction more
adequately: ‘“‘ Chicks after exclusion, on becoming aroused from
a dormant period, often open bill with a chewing motion and
sometimes thrust the bill forward sharply into the air even
without fully opening the eyes, no object, apparently, being
pecked at. Noticed a number of times.”

The following activities, generally recognized as instinctive,
were observed within the incubator on the first day, before
the chicks were completely dry: Preening the down of the
neck, wings, and breast; flapping of wings; chirping; walking;
pecking; lying on side and stretching out legs (in the rays of a
16 c.p. electric lamp); following a moving object with a motion
of the head; and chirring. Scratching, twittering, and wiping
of the bill have been noticed on the second day when the chicks
were taken out of the incubator for the first time and placed on

8 FREDERICK 8. BREED

a black cardboard for their pecking tests. Our knowledge of
the time of appearance of the various instincts should not be
left to depend entirely upon chance stimuli. Who shall say
that a given reaction might not have occurred much earlier
if the appropriate stimulus had been provided?

The chicks, while still in the incubator, are known to be posi-
tively phototropic. The incubator trays that have near the
glass door a trap through which the chicks fall to a screen below,
depend for their effectiveness on the fact that the chicks crowd
toward the light.

B. Drinking

a. Problem—In the study of instincts from the objective
point of view, interest naturally centers first in function, second
in structure. The activities which are known as instinctive
must be analyzed into component units of behavior, of which
they are nearly always complexes. Furthermore, no account of
instinct will be satisfactory, no explanation complete, until we
understand the structure of the machinery involved in each
action. But so much accomplished, this is not all. These
structures are not of such a nature that they in some way get
themselves into action. So far as we know, they have no in-
herent principle of spontaneity. Intra- or extra-organic stimuli
are necessary to touch them off. Environment in the form of
energies external to the structures and additional to the func-
tions seems to be a sine qua non. Hence a complete under-
standing of instinctive actions will include a detailed knowledge
of the ‘objects’? in conjunction with which the particular
activities manifest themselves. In the following bit of work
on the instinct of drinking, consideration of the problems of
function and structure is made secondary to an inquiry into
the nature of the extra-organic stimulus.

b. Method and Tests—-Chicks no. 69 to no. 81, inclusive,
were hatched during the afternoon and evening of Dec. 2 and the
morning of Dec: 3, 1908. Beginning with Dec. 3, their pecking
had been tested in the regular manner each day. As the animals
finished the first pecking tests they were marked, numbered,
and transferred from the incubator to the brooder. Although
given food and freedom to run about in the litter, they were
allowed nothing at all to drink until tested as described below,
neither were they permitted to see other chicks drink. A pos-

INSTINCTS AND HABITS IN CHICKS 9

sible chance of coming in contact with something like water
was afforded through deposits of waterish excrement. Even if
this as a stimulus were able to elicit the drinking reaction, the
fact that the chicks in the incubator rested on an elevated wire
screen lessened the probability of the occurrence of such a
stimulus. In the brooder the floor was constantly covered
with the regular litter from one to two inches deep. On Dec.
5, from 3 to 5:30 p. m., when the chicks varied from 2.5 to 3
days of age, they were brought one at a time to the experiment
table, just as if they were to be given a pecking test, and were
returned to the brooder each as its drinking test was completed.
In front of each animal was set a clean watch glass containing
fresh water devoid at the start of bubbles or sediment. The
watch glass rested on a square piece of smooth, white, plain
note paper. As the observations on each chick were completed,
the soiled paper was replaced by a clean piece, the dish was
washed, the water renewed, and bits of food or drops of water
carefully brushed from the table. The report of the observations
follows in the order in which the chicks were tested.

While no. 70 was eating on the table, the watch glass, in the
manner set forth above, was presented. The chick pecked grains
which had been scattered over the white paper and then ran
its bill in a forward direction along the paper in the drinking
reaction. This it did repeatedly. Later I helped it to find
the water in the dish. It seems unnecessary to state that pre-
caution was taken to see that the paper was dry.

No. 72 was tried next. It pecked the plain, clean, white paper.
Before the water was found by no. 72, no. 70 was brought out
and allowed to drink in the presence of no. 72. No. 72 followed
no. 70 about, performing the drinking reaction along the edge
of the dish. One of the chicks stepped into the dish and car-
ried some water to the paper surrounding it. No. 72 got its
bill into this and forthwith responded with the drinking reac-
tion. Then it wandered about the table and ran its bill along
the black leather of my watch fob, 30 cm. distant from the
dish, giving the drinking reaction. This does not mean that
it touched the fob with its bill and then lifted its head in the
manner so well known. The reaction to the fob was just what
one sees when he watches a chick gathering water into its bill
and throat while the bill is inserted in the liquid. This reaction is

10 FREDERICK 8. BREED

not easily confused with pecking. It was repeated on the silver
charm of the fob.

No. 71 learned to drink by pecking at a bit of excrement
that chanced to get into the water. Time after time it pecked
about the edge of the dish. It was not observed to dip its bill
into clear water. It secured water first by pecking and persisted
in getting it this way throughout these first trials.

No. 74, when set on the table, pecked the edge of the white
paper. First contact of the bill with water came from pecking
a drop of water deposited on the paper. I put its bill into the
water in the dish, whereupon it reacted to the situation by
pecking the edge of the dish.

No. 77. By catching the paper on which the dish rested,
the water was caused to wave slightly in the presence of the
chick. Its head went down hard in a pecking reaction and hit
the bottom of the dish, following which the head was lifted in
the manner characteristic of drinking. On the second trial the
head approached the water not in the manner of pecking but
of drinking. The pecking approach to an object is decidedly
different from the drinking approach. In contrast to the sharp
descent upon the object in the former, the low gentle reaching
movement in drinking, accompanied by a straightening of the
neck, opening of the mouth, and a peculiar motion of the throat,
is most marked.

No. 69. The surface of the water was agitated as for no.
44. No. 69 pecked directly into the water. Shortly afterward
it pecked the water without my shaking it and at a point where
I could discern no special stimulus like a bubble or particle of
food. The positions of chick and dish were such that light
reflected from the water might have been a factor.

No. 75. Pecked edge of the white paper. Pecked about the
edge of the dish and its bill slipped into the water. The chick
then pecked the edge of the dish twelve times, the bill getting
into the water during some of these reactions. Pecked twice
into the dish and got a little water. Of the next twenty reac-
tions all were pecks and all but two were directed toward the
side and edge of the dish.

No. 80. I thought I observed the chick performing the drink-
ing reaction on a piece of glazed kymograph record paper about
3 cm. square that I vibrated between two fingers in front of

INSTINCTS AND HABITS IN CHICKS 11

the chick. But the reaction came out unmistakably a moment
later on the side of the glass dish. The water was still, the glass
was dry, and the chick ran its bill along the edge of the glass
exhibiting plainly the drinking movements without touching the
water. It later found the water and began to drink without
assistance.

No. 81. On the first attempt it dipped its bill into the still
water and drank from a point at which I could distinguish no
special object in the water. While crouching by the dish drink-
ing, as its bill was coming down slowly at the termination of
a drinking reaction, it turned its head to the left, touched its
bill gently to a bit of dry chick food lying within reach, and per-
formed the drinking reaction upon it. It did not eat the grain.

No. 79. Pecked side of dish. Ran beak along edge of dish.
Its beak (accidentally) slipped into the water. After this the
chick began to drink. Frequently ran beak down the outside
of the dish.

No. 73. Did not peck at dish nor find water. No. 74 was
placed on the table with no. 73. No. 74 pecked five times at
the edge of the dish, not touching the water, when no. 73 began
to peck near the same place. The latter’s bill slipped into the
water and it began to drink energetically.

No. 76. I shook the water in its presence and it pecked into
it. It began to drink and twitter. Pecked the edge of the dish.
Ran its bill along the outside and inner edge of it.

c. Discussion of results—Spalding,; commenting on the
drinking of chicks, remarks: ‘‘It also appeared that, though
thirsty, they did not recognize water by sight, bes
and they had to some extent to learn to drink.” Discussing
the same instinct, Morgan? says: ‘‘ The statement of fact (so
far as my observations go) that I made was this: That the
sight of still water evoked no instinctive response; but that
the touch of water in the bill at once evoked the characteristic
instinctive behavior.’’ Mills * expresses his opinion thus: “ It
is not primarily so much the sight, but rather the touch of water

that in the very first instance leads to drinking.”

1 Spalding, D. A.: Loc. cit., p. 288.

2 Morgan, C. L.: The habit of drinking in young birds. Science, N. 8., 1896,

vol. 3, p. 900. ; .
3 Mills, W.: The nature and development of animal intelligence. New York,

1898, p. 281.

12 FREDERICK 8. BREED

On the basis of these and similar observations, it has been
asserted that a chick swallows water instinctively, but must
learn to drink by imitation or accident; that is to say, the drink-
ing instinct requires supplementation. A passage from Bald-
win! will illustrate: ‘‘In the case of the fowl’s drinking, it
is not the mere fact that drinking and eating may differ in the
degree to which the performance is congenital; the reports
seem to show that this varies in different fowl; but that instincts
(in this case drinking) may be only half congenital, and may
have to be supplemented by imitation, accident, intelligence,
instruction, etc., in order to act, even when the actions are so
necessary to life that the creature would certainly die if the
function were not performed. That is the interesting point.”

For the sake of clearness in the discussion of drinking, the
parts of the drinking complex must be more sharply distin-
guished. There is (1) the approach to the object, elicited evi-
dently by optical stimulation, (2) a sort of rhythmic movement
of mandibles and throat, which brings the object within the
mouth, and (3) the swallowing activity, which is evoked by
stimuli resulting from the contact of water with the mouth,
and which is marked by an elevating movement of the head
and neck. We find an exactly parallel series upon analysis
of the pecking reaction: (1) striking, (2) seizing, and (3) swal-
lowing. Any full account of the drinking instinct must include
the approach to the object as well as the subsequent manipu-
lation.

No one, I think, will deny that the touch of water in the
bill evokes reaction 3 of the above series. And we know that
chicks may get this appropriate contact-stimulus indirectly by
pecking. Furthermore, the drinking of one chick in the presence
of another often stimulates this other to become active about
the water and thereby leads to its drinking. That is, drinking
usually does begin as the result of a contact stimulation mediated
by the prior activity of the pecking and imitating instincts.
But this at once suggests the further question, Are imitating
and feeding necessary precursors to reaction 1 of the series,
the movement of approach? The results of the foregoing experi-
ments seem to show clearly that the drinking instinct is self-
dependent in so far as its relation to these other instincts is

' Baldwin, J. M.: Instinct. Science, N. 8., 1896, vol. 3, p. 669.

INSTINCTS AND HABITS IN CHICKS 13

concerned. When chicks, without having previously drunk,
respond with the drinking reaction to the surface of smooth
white note paper, the edge of white glazed kymograph paper,
or the edge of a glass dish, all these objects must be supposed
to have some quality or qualities, undoubtedly visual, which
evoke the drinking reaction. Here are stimuli that call forth a
first drinking reaction independently of other instincts. If
further evidence were needed to substantiate the conclusion
that the instinct is self-dependent, it is found in certain other
Observations. After the first actual drinking, the drinking
response was made to a grain of food, a piece of black leather,
a silver ornament, and, in the case of chicks other than those
studied in the special experiment, to a line in some dust on a
smooth surface, a white spot on the experiment table, the
clean surface of black cardboard, and the polished surface of
a table.

Thus the feeding and drinking instincts are more similar
than writers have hitherto supposed. In the case of feeding,
the fact seems to be that newly hatched chicks respond to a
great variety of objects indifferently, and only later come to
select those which are food from the rest; in the case of drinking,
the observations show that, if the need be sufficiently urgent, a
large variety of objects in like manner elicit the action, and
apparently with a like result. The indiscriminate use of reac-
tion 1 of the drinking instinct may bring the appropriate stimulus
for reaction 3, without the co-operation of the pecking and
imitative activities. The drinking instinct, therefore, does not
‘have to be supplemented by imitation, accident, intelligence,
instruction, etc., in order to act.”

Finally, is still water a sufficient stimulus for the act of drink-
ing? I am by no means ready to say it is not. I regret that
my experiments are not so complete on this point as they might
have been. There is general agreement at present that chicks do
not begin to drink in response to this optical stimulus. But de-
prive them of water fora sufficient length of time after hatching
and perhaps the “ sight of still water ’’ will evoke this instinctive
reaction. It is not improbable that the effective element or
elements in the objects which have been observed to draw forth
the reaction are common also to water.

14 FREDERICK 8. BREED

C. Pecking

a. Apparatus and method.—A detailed and prolonged study
of the pecking reaction was now attempted. The literature on
this topic reveals the fact that. the interest in the accuracy of
this instinctive activity has been the central one. As suggested
before, the accuracy of the reaction became the central interest
also in the investigation that is reported in the following pages.

The apparatus used in these experiments was very simple.
A table with a hard polished surface was set neara window where
there was good light. To this table the chicks were each morn-
ing brought, one at a time, and permitted to eat in a natural
way from the surface of a piece of black cardboard about 20
cm. wide and 25 cm. long. Carried daily to and from the ex-
periment table, the chicks became so habituated to the transfer
that the fear response did not enter in to mar the value of the
results. From the first the chicks ate from my hand, and soon
many of them energetically followed the hand from point to
point, gathering up the bits of food as they were dropped. A
little later many even gave the “food twitter”? while in my
hand on the way from the brooder to the table. But it may
not be said with accuracy that the chicks became “ habituated ”’
to the operation, if at anyage without previous trials they sub-
mitted, without signs of being disturbed, to the conditions of
the experiment. As a fact, they did not thus submit. Animals
that had not been used in the pecking tests nor been handled
previously in any other experiment were brought to the table
for a control test. They usually struggled when picked up and
seemed so disturbed by the situation, when set on the table,
that they would not eat at all. This was especially true if they
had been allowed to live in the brooder unmolested for two or
three weeks.

For the first tests, which were conducted on the second day,
because of the physical weakness of the chicks and their indis-
position to eat on the first day, slightly moistened bread was
used, of such a consistency that it could be rolled between the
fingers into food particles of suitable size. After the second
day Cyphers Chick Food, slightly moist, replaced the bread
pellets in the tests. This, which was the regular food of the
chicks, is a mixture of whole wheat, Kaffir corn, cracked corn,
millet, etc.

INSTINCTS AND HABITS IN CHICKS 15

The food dish was kept out of the chicks’ field of view and
the bits of food to be pecked at were dropped by hand upon the
cardboard in such a way that a particle would not be in motion
when a chick pecked at it. From one to three grains only were
dropped at a time. This proved advisable because the number
of reactions elicited by more than three grains, considering also
the possible variety and rapidity of the reactions, made it
difficult at times to secure an accurate record.

In a letter published by Mills; Bumpus? has suggested that
the different aspects of the pecking reaction be distinguished.
For these separate parts he proposed the terms seizure, mouth-
ing or mulling, and deglutition. The terms used by Morgan *
seem much more appropriate—striking, seizing, and swallow-
‘ing. It is interesting to. note that Spalding + distinguished these
phases of the reaction and employed the same terms. These
we have adopted, using in addition the term missing for failure
to hit the object. As they appear in our records, these terms
have the following definite meanings: (1) Missing denotes all
cases of the pecking reaction in which the bill fails to hit the
particular object supplied. by the experimenter; (2) striking,
those cases in which the bill hits the object without seizing it;
(3) seizing, cases in which the object is grasped momentarily
in the bill and then dropped; and (4) swallowing denotes what
may be termed the perfect or complete reaction, the object being
struck, seized, and swallowed in an errorless series or chain of
movements. To facilitate the taking of records, the numerals
I, 2, 3, and 4 were used to represent muissed, struck, serzed, and
swallowed, respectively. Note was taken, of course, of the reac-
tions independently of the number of food particles pecked at,
for a single grain might call forth a half dozen reactions in suc-
cession. For example, suppose one millet seed brought the
result 1-2-3-4. In this case the chick first missed the grain,
on the second reaction it struck it but did not get hold of it
between its mandibles, on the third attempt it caught it in
its bill but dropped it, and on the fourth, it struck, seized, and
swallowed the grain without error.

1 Mills, W.: The nature and development of animal intelligence. New York,
1898, p. 296.

?Bumpus, H. C.: Instinct and education in birds. Science, N. 8., 1896, vol.
4, p. 213.

‘a Morgan, C. L.: Habit and instinct. London and New York, 1896, p..37-- -

4Spalding, D. A.: Loe. cit.

16 FREDERICK 8. BREED

For this, as well as for all other work, the chicks were marked
by colored yarns tied on their legs. Each animal thus iden-
tified was assigned a number by which it was afterwards known
and referred to. .

b. Pecking artificially deferred—The experiments of Spald-
ing,! in which chicks on leaving the shell were blindfolded with
little hoods or kept in a flannel bag, have attracted much atten-
tion from students of instinct. Spalding sought to.ascertain the
facts in regard to instinct. In his time the skeptical were hold-
ing ‘that all the supposed examples of instinct may be—for
anything that has yet been observed to the contrary—nothing
more than cases of rapid learning, imitation, or instruction.”
The hooding device was intended to permit the chicks to acquire
“enough control over their muscles to enable them to give
evidence as to their instinctive power.’’ In other words, the
aim was to test the activity of pecking when the factor of acqui-
sition was eliminated. The conclusions of Spalding as to the
accuracy of pecking, previously referred to, are based on the
results of these tests.

In the similar tests that are reported below, the purpose
was to measure the accuracy of the pecking response under
circumstances like the above. Not much success was achieved
by hooding the chicks. Other means of excluding the light were
relied upon. After being tested the chicks were placed in the
brooder with the rest of the flock.

On Dec. 10, 1907, chick no. 8, immediately upon hatching
at 1:30 a. m., was blindfolded and transferred to the brooder
where it was kept until 10 a. m., Dec. 12, under the curtained
hover in a black-lined box at a temperature of 103° F. This
box, open at the top, was enclosed in a green flannel bag.
At 10 a. m., Dec. 12, the chick was brought carefully to the
experiment table for a pecking test. Almost immediately it
twice pecked the board upon which it stood. The particular
stimulus to the reaction was not apparent. Then followed, in
a series, seven reactions, in all of which the chick missed the
pellet of bread (rolled as previously described) which elicited
them. Seven more pecks at the same crumb followed without
hitting it. Then the board on which it stood drew forth two
more reactions, when the pecking was interrupted by an awk-

1Spalding, D. A.. Loc. cit., p. 282.

INSTINCTS AND HABITS IN CHICKS 17

ward attempt to bite its wing. After a spell of violent chirping,
it missed twice, and hit as often, a spot on the board upon which
it stood, following which it lifted a foot to scratch its head and
lost its balance. Then four pecks at a crumb missed, but a
fifth was successful and the crumb was swallowed—the first
perfect reaction. Thereupon came a series of 79 pecking reac-
tions in no one of which a particle of food was hit, seized, and
swallowed in a chain reflex. In many of these reactions the
head wobbled from side to side as the bill moved slowly toward
the object pecked at. On the twentieth reaction in this series
of 79, the object was seized in the bill and apparently rejected
by the chick. Prior to this, excluding the one perfect reaction,
bits of food had been seized on two occasions but dropped through
seeming lack of skill. Just preceding an interval occupied in
preening the feathers of its wings, an interval which marked the
end of the 79-reactions series, there occurred a group of five
reactions, in three of which the object was struck and in the other
two seized. It is but fair to state that many of the pecking
reactions were not in the direction of food particles, there being
something about the bare surface upon which the chick stood
that drew the reactions forth. In the summary of the chick’s
record these reactions have been classified with the 1’s. At
this point it was incidentally noticed that the chick followed
with both head and eyes a movement of the experimenter’s
hand. After the above mentioned preening diversion, twelve
more pecking reactions ensued: seven 1’s, one 2, and four 3’s.
Then the animal hesitated long enough to scratch its bill with
its foot. The next nine reactions, of which three were 1’s, five
2’s, and one a 3, came out in straggling order, interrupted by
gaping, pecking of toes, scratching the bill, and preening the
feathers of the breast.

When tested by a moving object, a moist bit of bread swing-
ing by a black silk thread, the chick’s first impulse was mani-
festly toward it, but later it acted as if afraid. During these
tests the chick appeared in no danger, except by accident, of
stepping off the edge of the piece of cardboard upon which it
was placed, when the cardboard was so arranged that the outer
edge of it coincided with an edge of the experiment table. The
animal could easily be pushed away from the edge but, when
near the edge, resisted strongly if pushed toward it. This same

18 FREDERICK 8. BREED

behavior was noticed in other chicks. For example, no. 14,
the day after it was hatched, when on the experiment table
for the first time, was pushed toward the edge. It resisted by
bracing its legs in front of itself and hurried back from the
edge as soon as it was released.

The summarized record of no. 8 is as follows :—

I-1—-I-I-I~I—I—I—I~—I~—I-1-I-I1~I~-I-I-I—1—2—2-I-I—I-I—4-1-1
2-I-I-I-I-I—1~—I—I—2—2—3-I-2-I-1—3—-3-I-I-I-I-I-I-1-I-1-I-
J-I-I~-I-I-I-I~I~I-I-I-I-I—I-I-I-I—2—2-1-I-I-I-I-I-I-I-1—
{-I-I~-I—I—I—I~I~I—I~—2—I-I—I1-2—-2-2—3-2-2-3-I-I—-3-I-I-I-I-I

—2—3-3-3-I—2-2-2~-1~2-2-1-3.

Attention has been called to the fact that the chick pecked
the cardboard at a place where no particular spot or grain was
discerned by the experimenter. Not only that, but, like a child
repeating da—da, it sometimes continued with a rapid succession
of reactions at or near the same place. Twenty-five of the long
series of 1’s, immediately following the third 3 in the record
above, represent such pecks at the bare card. They came out
in three groups, first 2, then 18, and after an interval the other
5. To see a chick hammering in this way 18 times in rapid
succession, and hard, too, at apparently “ nothing,” reminds
one of the mere exercise of a function, and suggests what Bald-
win has termed the circular reaction.

A comparison, in regard to accuracy, of no. 8’s record, with
the numerous records which will be presented later, will show
that this chick was below the average efficiency in its first test.
It is not unusual to find normally kept chicks that do not make
a single swallowing reaction in the first twenty trials on their
second day, and chicks sometimes fail at this age to seize a
single grain in the same number of trials, but I have found only
one in the regular tests that failed to strike at least one grain
in the first twenty reactions. There were instances in these first
tests in which the side of the chick’s bill touched the side of
the object as the bill slowly passed and hit the cardboard. These
were recorded as x’s, here and elsewhere.

Naturally, after being kept 56.5 hours prior to the tests under
the conditions previously described, the chick’s down was ruffled
and it lacked the general sleek appearance of its unconfined
mates. Whether it was deficient in anything else but the results

INSTINCTS AND HABITS IN CHICKS 19

of practice, I cannot say. The confinement did not do it any
permanent injury, it seems, for it continued in the pecking tests,
lived to make a perfect record after an average number of trials
in later black-blue color selection tests, and survived after that
till it was killed and dissected for sex determination on March
14, 1908. Its recordin the later pecking tests is submitted in
table 1. This table shows the accuracy of its pecking during
the nine days after Dec. 12, on the basis of twenty reactions
a day.

TABLE 1
Cuick no. 8. Dec. 13-21, 1907. Prcxine
Date os sen.ve ncaa das 13 14 16 17 18 19 20 21
MISSES ais sieccus Sain oe eecdsadesnaetts 2 1 2
Dirck: secs eve ates g cas 3 4 7 2 2 2 1
Beizede oc eds de seh dd et 1 1 3 1 1
Swallowed................ 14 14 13 17 15 18 18 18

The record of this chick in its first test is not presented as
typical, but is submitted to show how a chick may act under
conditions like those set forth above.

Two other chicks, no. 7 and no. 14, of the same brood as no.
8, were enclosed at the same time in the dark-box with no. 8.
These chicks were helped out of their shells just as they were
hatching and transferred at once to the dark-box. They were
not hooded, but even if they had been securely blindfolded,
little would have been gained, for the light that reached the
interior of the black-lined box was negligible in amount.

No. 7 was brought to the light forits tests at 11:05 a. m,,
Dec. 11, after 33 hours, 35 min. in the dark-box. My notes on
its behavior are as follows: ‘‘ Followed moving object (hand)
with head. Pecked wood over edge of black cardboard. Wiped
bill on board. Pecked three times at a pellet of bread moistened
with milk. Missed it each time. Chick pushed bill slowly toward
bread crumb on the black surface—a slow peck. Missed crumb,
weak peck. Bit toes. Touched small white speck with bill.
(The experimenter did not point at objects.) Rested under
hand with its eyes closed. Missed a crumb twice. Missed another.
Moved bill toward it slowly. Chick able to stand. Pecked at
feet and lost balance. Bit wing twice. Pecked black card-
board. . . Chick seemed cold and was returned to the
brooder.’’ The record of no. 7’s pecking at food particles and

20 FREDERICK 8. BREED

spots on the card runs thus in its first test, as detailed above:
I-I-1-1-2-1-1-1. Table 2 shows the accuracy of its pecking
from Dec. 13 to Dec. 26.

TABLE 2
Cuick No. 7. Duc. 13-26, 1907. Prcxine
Dates sa c4 ics 13 14 16 17 18 19 20 21 24 26
Missed........ 1
Struck........ 7 5 2 3 1 7 1 2
Seized........ 3 3 1 2 1 1 1
Swallowed.... 17 13 12 17 14 19 13 18 19 17

Chick no. 14, as just stated, had been placed in the same
enclosure with no. 7 at the same hour on the same date, and was
kept there till 4:25 p.m., Dec. 11. After 38 hours and 55 min.
thus enclosed, it was tested in the same way as no. 7. On the
first pecking attempt it pushed its bill, mandibles closed, into
a bread crumb. Then followed six pecks that missed. Attempted
to scratch its head and lost its balance. Seemed weaker than
the other chicks. Pecked toes and bare black cardboard;
scratched bill. A second time it pushed its bill slowly into a
bread crumb. Then followed a reaction in which it missed, etc.
The summarized record for this test appears thus: 2-1-1-1-1-
I—I—2—I—2—2-I—-2—2—2-2-I—3—-2—2—2—-2—3-2-2-2. The accuracy of
its later pecking, Dec. 13 to Dec. 26, is shown in the follow-
ing table:

TABLE 3
Cuick no. 14. Dec. 18-26, 1907. Prcxrne
Date......... 13 14 16 17 18 19 20 21 24 26
Missed........ 1 1
Struck........ 9 6 12 6 5 3 4 1
Seized........ 2 1 1 1 1 2
Swallowed.... 9 13 8 13 14 20 19 16 14 18

A comparison of the records of these three chicks for Dec.
13 with the corresponding records of five others of the same
age which were not subjected to confinement in darkness shows
that there was no noteworthy difference in the development at
this time. The averages of the three chicks on the above
date were .7, 4, 2, and 13.3 for reactions 1, 2, 3, and 4, respect-
ively, while the other five chicks averaged .8, 1.4, 4, and 13.8
for the same reactions, in a total of twenty reactions. On the

INSTINCTS AND HABITS IN CHICKS 21
; 4

other hand, the development of no. 8 on Dec. 12 was far behind
that of any chick in the above group of five on the same date.

It looks very much as if (1) the development of the instinct
was retarded by disuse, and (2) the retardation was quickly
overcome with use. Several attempts to repeat these dark-box
tests with other chicks failed. On these later occasions the plan
was conceived of placing the dark-box inside the incubator just
before the expiration of the period of incubation, depositing
therein a few of the eggs, and allowing the prospective chicks
to hatch and remain in the dark enclosure till needed. For
some unforeseen reason the chicks did not appear.

We turn now to a more thorough investigation of the ‘‘ con-
genital definiteness ’’ of the pecking reaction, with a view to
a quantitative determination of its course of development.

c. The natural development of the pecking reaction, with a
collateral study of the effects of social influence. 1. Sources of
error.—In addition to what has already been said of the method
of taking records, the important sources of error might be men-
tioned. Investigators have noted the fact that the chick, if
it misses the grain at all, usually comes within a hair’s breadth
of it. Hence, when the animal is working rapidly, it is sometimes
difficult to determine whether the bill hits the grain or not. In
the work herewith reported, a chick was credited with reaction
2 if there was any doubt between 1 and 2. In other words a
i was recorded only when the miss was positively observed.
The fact that more missing reactions seem to be recorded among
the later results may be due rather to modification of the ex-
perimenter than to variation in the groups of chicks. I feel
quite confident that I became a more efficient observer of the
missing reactions as the work progressed.

Again, there are times when reactions 2 and 3 may be easily
confounded. A grain seized simultaneously with the strike may
almost immediately fly from the bill—like a cherry pit squeezed
between the fingers. This reaction, properly 3, may be confused
with the reaction where the grain is driven from its position by
the mere impact of the bill. Still, as one becomes familiar with
such cases, the difficulty of distinguishing them almost wholly
vanishes.

Occasionally fewer than the regular number of reactions in
the daily series are reported for a given date. In such cases

‘

22 FREDERICK 8. BREED

the cause was usually the refusal of the chick to eat. This
difficulty was soon obviated by carefully controlling the amount
of food allowed the chicks, and compelling them to scratch for
it in the litter. The objection also was thereby forestalled that
many chicks under an improperly regulated food supply might
react the required number of times but below their highest
efficiency.

Not only did these difficulties threaten to affect the reliability
of the results, but an additional reaction came in to complicate
the record. When the food supply was improperly controlled,
the chicks would quite often reject grains, that is, lift them from
the ground with the bill and then either drop them or throw
them some distance without attempt to recover. Forcible rejec-
tion is distinctly a reaction in addition to reaction 3. It involves
control of the grain by the mandibles—sufficient, one should
suppose, for manipulation without error prior to swallowing.
If this be true, it is manifestly unfair to charge the chick with
a degree of imperfection by recording reaction 2. Yet, one is
not at liberty to credit it with reaction 4. A clear instance of
rejection should simply not figure in the records. But this
disturbing feature was practically eliminated when the chicks
were confined in the brooder at night and were given the proper
amount of food only after the morning tests. I say practically
eliminated, because, even after taking the precaution men-
tioned, some chicks in time came to eat one kind of grain and
not another. For example, I have found an occasional chick
that discriminated in this fashion between millet seeds with
the shell on and those without the shell, to say nothing of prefer-
ences for wheat as against corn, millet as against wheat, and
the like. In such cases the chick was fed what it readily ate.

Another variation, which we may term switching of the bill,
occasionally appeared. It is similar in character to the forcible-
rejecting movement, and was observed in some very young
chicks. It is executed while the food particle is held firmly
between the mandibles. It is mentioned here rather as an
additional type of reaction than as a source of error, for it did
not noticeably affect the records.

I have so many times observed chicks, especially very young
ones, peck where I could see nothing but bare wood or card-
board, that I am compelled to believe that the stimulus to the

INSTINCTS AND HABITS IN CHICKS 23

pecking reaction need not be some object of a size convenient
for eating. In line with this consideration is the further fact
that the bill of the chick is not only a feeding but a testing organ.
If anywhere, the ‘oral sense ’’ has its organ here. The chick
often digs in the litter with its bill when the reaction is distinctly
not the pecking reaction. While trying to escape from the
reaction box of a discrimination apparatus in a dark-room,
chicks often approached a sheet of clean, smooth, lighted opal
glass and pecked it. In the same box they pecked the black
walls and especially the black cardboard closing the exit, when
they happened to enter the wrong side and found their egress
barred. But, except where specially noted, no missing re-
actions were recorded when no object to be pecked was supplied.

Besides the occasional variation in the missing reaction in
which the animal pecked the bare cardboard when no other
stimulus seemed to be present, there were two other reactions
that were easily distinguishable from each other and from the
regular missing reaction, but which were, nevertheless, put down
in the records as reaction 1. These variations in the missing
reaction were altogether so few in number, compared with the
many hundreds of reactions recorded, that the averages are not
materially affected by the classification of them with reaction 1.
I refer here (1) to cases in which the chick pecked in the direc-
tion of the grain but did not reach it, and (2) to cases in which the
bill struck the cardboard almost exactly midway between two
grains. Of the first of these I have records of thirteen instances;
of the second, thirty-six instances. Case one needs no special
comment. The bill each time was going in the right direction,
but the innervation seemed to be insufficient. Case two is, if
anything, more interesting. When the typical instance of it
occurred, the grains were usually about 1 cm. apart, of about
equal size, and the line connecting the two ran about normal
to the direction in which the chick was oriented, the chick stand-
ing directly before them. One could say with little certainty that
this was an instance of pure missing. We have already agreed
with other observers that the chick from the first rarely missed
by more than a hair’s breadth an object pecked at. But here
was a case where a chick from one to four weeks old missed a
grain by 4to 6mm. Not only that, but I have five records where
in a situation like that described above the chicks pecked twice

24 FREDERICK 8. BREED

in succession between the two grains, and one record in which
there were four such successive pecks. There are two other
observations that help to explain this reaction. On Dec. 23,
no. 70 and on Dec. 29, no. 87, were working rapidly in their
regular pecking tests. The pecking activity was interrupted
momentarily while the bill poised in air above and about mid-
way between two grains approximately 1 cm. apart. On Dec.
24, no. 76 was on the experiment table going through its
pecking test. In the presence of two grains lying as those in
the preceding cases, its pecking reaction started in the direction
of one of the grains and was completed by a perfect reaction
upon the other. I concluded, therefore, that case two was not
an ordinary instance of missing, but that the reaction in response
to one of the stimuli exerted an inhibitory effect on a reaction
in response to the other. The poising of the bill in air, the zig-
zagging first toward one grain and then toward the other during
a continuous forward movement of the bill, and the striking of
the cardboard fairly between the grains, have at bottom, I
believe, the above common principle of explanation.

In view of the recent important work of Hess and that of
Katz and Révész, I feel impelled to report some activities of
the chicks in the dark-room. The assumption that hungry
chicks do eat when they can see the food is ordinarily true, but
the second assumption, that chicks do not eat when they cannot
see it, surely is not borne out by observations of the behavior
of chicks in the dark-room. During some studies of individually
acquired reactions, the chicks were placed in a dark-basket in
the dark-room for ten minutes prior to the regular daily tests.
The basket which was used for this adaptation work was a
willow waste-paper basket lined with black cloth and covered
with the same material. Food was placed in the bottom of the
basket before the basket and chicks were taken into the dark-
room. The animals were thus allowed to eat in the -basket
a short time before the period of darkness adaptation began.
Again and again, after the basket was taken into the dark-room,
the door closed, all lights turned off, and the basket covered, the
animals continued to eat the chick food in the bottom of the
basket. I have also gone out of the dark-room and returned at
the end of the ten minute period to find the chicks eating. The
darkness was so nearly absolute in the room that I could not

INSTINCTS AND HABITS IN CHICKS 25

see the dark basket while sitting in a chair nearby. The peck-
ing sounds, as the bills of the chicks struck the bottom of the
basket, could be distinctly heard. The usual mouthing noise
made by the mandibles could be heard also when one held his
ear near the side of the basket. And when I put my hand into
the basket, I could feel the heads of the chicks rising and falling
in the pecking reaction.

Before formulating a conclusion let me describe the behavior
of no. 71. This chick, in the pecking reaction tests, made the
following record from its second to its eleventh day, and then
ceased to peck at the grains when brought to the experiment
table:

TABLE 4
Cuick no. 71. Dec. 3-12, 1908. Prckine
Dates. a4) ex se 3 4 5 6 VG 8 9 10 11 12
Missed........ 13 2 3 2 2 2
Struck........ 6 25 32 26 22 8 17 6 12 43
Seized........ 9 5 3 2 3 3 3 1 1
Swallowed.... 1 14 10 19 24 39 30 39 37 6

On Dec. 13 it staggered about in a peculiar way, but did not
eat. It would walk directly off the table if not prevented. When
tested with water in a watch glass, it drank when the water was
held to its bill, but after the dish was moved three or four centi-
meters away, it dipped its bill about where the dish had been.
Such indications as these convinced me that the animal was
blind. For a day or two it was fed by having grain introduced
into -its bill, Then it was placed in a large dish partly filled
with chick food, and on Dec. 16 it began to scratch and eat.
For many days thereafter it was regularly placed in the food
dish at the feeding hour and procured its meal without assistance.
When I entered the room and the rest of the flock ran toward
the wires of the cage, no. 71 was often seen approaching, run-
ning into obstacles in its path, bobbing up and down in a peculiar
variation of the pecking movement, and taking time intermit-
tently to scratch in the litter. In this fashion it lived until it
was found dead one morning in the drinking vessel. The beha-
vior of the chicks in the dark basket, as well as that of this
supposedly blind chick, furnish quite clear evidence in refuta-
tion of the assumption that chicks do not peck at or eat food
when they cannot see it.

26 FREDERICK 8. BREED

.But happily the reaction in which we are most interested,
reaction'4, can be distinguished from all the others with practi-
cally no liability of error. Even supposing a few 1’s, 2’s and
3’s were to become interchanged, the use of reaction 4 as a
quantitative index of development would not be invalidated.
The main interest is to discover the ratio of the number of per-
fect reactions to the whole number of legitimate pecking reac-
tions in the daily series, and to plot the variation of this ratio.
The necessary data for this can be gathered with precision.

2. Data and their significance. The first animals tested
were numbers 1 to 6, inclusive, which will be referred to as

Trias
20
18 = =
pes =i ~
16 > =
eee [ae “ ar
14 a :
12
10 HitAy
\
8
el tta x
; /| hnK
ea Ng ieee ee |
Psd ieee a ne en la Oe eS 74
ste —
Gays! 2 3 4 5 6 7 8 B 10 11 12 13 14 15 16 17 18 189 20 2t 22 23 24 25

Ficure 1—The development of the pecking instinct in the chicks of Group A, num-
bers 1 to 6, inclusive. Distances along axis of abscissae represent days after
hatching; distances along axis of ordinates, the number of a given kind of re-
action in a series of twenty pecking reactions. Curve IJ-A shows the rate of
decrease in the daily number of reaction 2; Curve III-A, the same for reaction 3;
Curve IV-A shows the improvement in accuracy of reaction 4 for the group.

Group A. They were hatched on Dec. 2, 1907. Their pecking
records cover the period from Dec. 3 to 30, inclusive. Curve
IV-A, fig. 1, shows the development of reaction 4 for the group.
For the calculations only those totals were used in which every
chick had a complete series of twenty reactions. Hence a very
interesting part of this curve, namely, the detail for the third,
fourth, and fifth days, is lacking. The blanks are due to the
failure of some of the chicks on each of these days to respond
to the food. Improper regulation of the food supply was the
cause. Nevertheless, the results were such as to encourage a
continuance of the experiment. Barring the vital omissions at

INSTINCTS AND HABITS IN CHICKS 27.

the most critical stage of development, the curve on the whole
not only indicates that the co-ordination of movements that
constitute reaction 4 has a certain period of development, but
it roughly marks out that course. The rate of improvement is
not uniform from day to day, but is very much more rapid during
the first few days than later.

Before discussing reactions 1, 2, and 3 of this group, let us
be clear as to the meaning of these three reactions and their
relation to 4. In the first place, it should be noticed that the
records under 1, 2, and 3 may be regarded as error records.
That is, under 1 are listed the errors of striking, under 2 the

Trials
1 Oo ee.
16 Ses ee al ep =
ee oe
“ 1a: a oe
[ 7
12 -
1°TTFAR, 7 A
8 as
6 A
‘| RIS.
|| Zia
vA \
A
SPSS SPSS Sea docdae donde

Qasi 2 3 4 § 6 7 8 8 10 ML 12 13 14 15 16 17 18 19 2 2 22 23 24 25

Figure 2—A comparison of curves of development of the pecking instinct to show
the possible effect of social influence. Distances along axis of abscissae repre-
sent days of age; distances along axis of ordinates, the number of a given kind
of reaction in a daily series of twenty pecking reactions. Curve IV-A shows
the improvement in accuracy of reaction 4, Group A; Curve IV-B, the im-
provement in accuracy of the same reaction for Group B. Curve III-A shows
the course of reaction 3, Group A; Curve III-B, the course of reaction 3,
Group B.

errors of seizing, and under 3 the errors of swallowing, if swal-
lowing here may designate the manipulation of the object after
it is seized. Although for convenience the term swallowing has
been applied to reaction 4, it is clear that reaction 4 is really
striking-plus-seizing-plus-swallowing in an errorless train.

There are always three regular chances of error, then, at the
beginning of any pecking reaction. I shall aim to show the
rate of decrease of each kind of error separately, and, correla-
tively, the improvement in the co-ordination of the three
reactions.

28 FREDERICK 8. BREED

Curve II-A, fig. 1, shows the course of reaction 2, and Curve
III-A of reaction 3, for the chicks of Group A. The errors of
swallowing tend to vanish more rapidly and more completely
than those of seizing. The data on reaction 1 are too meager
to justify consideration here.

Curve IV-B, fig. 2, is the curve of development of reaction
4 in Group B, chicks nos. 9 to 13, inclusive. These chicks were
hatched just eight days after those of Group A. The two groups
were segregated until chicks nos. 9 to 13 were beginning their
seventh day, and then were allowed to mingle. The pecking
tests were made throughout in the regular way, both groups being
treated as nearly alike as possible. Curves IV-A and IV-B, fig.
2, are matched for corresponding days in the lives of the chicks—
that is, for corresponding ages, and not for the dates on which
the records were made. Although it was not part of the plan,
when the experiment was conceived, to study social influence,
the higher curve for Group B indicates that. possibly association
with the older chicks had the effect of facilitating improve-
ment in the younger ones, and suggests a method of measuring
this influence. A glance at Curve IV-A, fig. 2, shows that the
efficiency of the chicks of Group A on their fifteenth day, when
those of Group B joined them, was numerically 16 on a scale of
20, and that the efficiency of Group B at the same time, the
seventh day for the latter, was 14.6. But the difference in the
curves may be due to unsuspected irregularities in method, or
to variability in the chicks. It seemed hazardous, therefore, to
rest such an important conclusion on one test, and a repetition
of the experiment was planned and executed.

Curves IV-A and IV-B indicate, then, that there is a very
rapid improvement in the integration of the components of the
complete reaction within the first three days, followed by a
slower but fairly steady improvement after that for some time.
It is interesting to note just how nearly the development approx-
imates perfection,—18.4 on a scale of 20 representing the degree
of approximation at the highest point.

Curves III-A and III-B, fig. 2, trace the course of reaction
3 for Groups A and B respectively.

The study of the pecking reaction was resumed almost a year
later with chicks nos. 57 to 65, inclusive, which we shall style
Group C. These chicks were hatched Oct. 25, 1908. With a

INSTINCTS AND HABITS IN CHICKS 29

desire to continue the study of the effect of association of younger
with older chicks, eggs were set with the intention of having
another brood eight days after Group C, so that the difference
between the ages should correspond exactly with the difference
in age between the chicks of Groups A and B. But the plan
was frustrated to this extent, that the chicks of Group C came
out of the eggs one day before the date on which they were
scheduled to arrive, and a sudden change of temperature so
chilled the second setting that only two healthy chicks arrived
to constitute Group D. It was decided to make the most of
the opportunity, however, and records were taken of the pecking

Trrats
20
18 =
18 z hs eo st--heT |
16 rab: —S == SL
i — a
“4 ;
F
12 j
/ i
10 / L
3 T
li
8 T
aiwel
d
* EB
ati
VD
Qysi2 3 4 5 6 7 86 9 10 11 12 13 14 15 16 17 18 19 2 21 22 23 24 25

Figure 3—A comparison of curves of development of the pecking instinct to show
the possible effect of social influence. Distances along axis of abscissae repre-
sent days of age; distances along axis of ordinates, the number of ‘ perfect ”
reactions in a daily series of twenty pecking reactions. Curve IV-C shows the
improvement in accuracy of pecking in Group C; Curve IV-D, the improve-
ment in Group D.

of the nine chicks in one group and the two in the other. When
the members of Group D were beginning their eighth day, the
cages of the two groups were thrown together and the chicks
allowed to mingle. For three days, beginning with ‘this day,
100 pecking reactions per day were recorded for each chick in
both groups; and, continuing for a week after this, 50 reactions
a day were taken. It was thought that by this means the influ-
ence, if any, of one group upon the other might be more easily
and accurately detected. The results are plotted in fig. 3, and
the curves matched for corresponding ages as before. The
chicks of Group C had a pecking efficiency of 17.1 at the begin-
ning of their seventeenth day when those of Group D joined

30 FREDERICK 8. BREED

them. On this same date, when the individuals of Group D
were eight days old, they had an efficiency of 16.5. Very little
change occurred in the accuracy of pecking of Group D after
the commingling, nor indeed could a heightening of their curve
after the eighth day be well expected, for they had by that
time practically attained the maximum.

Nothing very conclusive having accrued here, it was deemed
desirable to arrange a test so that younger chicks could be
enclosed with older ones immediately after hatching. Thus, it
was thought, the influence might be measured, if the effect of
social influence be such as to affect the rate of improvement of
the pecking reaction. The most active opponents of inferential,
ideational, or voluntary imitation seem agreed that the presence
of one animal, under the proper conditions, may have the effect
of stimulating another of its kind to greater activity. There is
no difficulty whatever in establishing this fact. My observations
have proved it to my satisfaction many times. It may be that
the change in the mode of functioning of the organism due to
the presence of others is in the direction of increasing the inten-
sity and rapidity of the reactions, without increasing their rate of
improvement in accuracy. Such a variation would have selective
value, inasmuch as the animal would get more food in a given
time, even though the pecking were no more accurate. But
such a speculation does no more for us than open up wider
vistas of experimental research in which, exact quantitative
studies can no doubt be made.

Mention has previously been made of the rapidity with which
reaction 4 improves. The curves all show very rapid improve-
ment on the second day. To make a special measurement of
the rate of this change, the pecking of six chicks in Group C
was tested twice on the second day. The first record was taken
from 10-12 a.m. and the second from 2-4 p.m. In the interim
the chicks were committed to the brooder and allowed the
freedom of the litter of the cage for the first time. From fore-
noon to afternoon, within the limits just stated, the improve-
ment was 82%, assuming, as we have, that the development of
reaction 4 is an index of the improvement of the chicks’ peck-
ing efficiency.

The sudden drop in Curve IV-C, fig. 3, on the fifth day is
due mostly to the unusually poor records. of no. 5.7 and no. 58

INSTINCTS AND HABITS IN CHICKS 31

on that date. Twelve of no. 58’s striking reactions were at one
particle. It followed this up, hitting it time after time until the
grain was knocked off the table.

The final test of social influence in its relation to pecking
was made on two comparable groups of six chicks each, which
will be referred to as Group Eand Group F. The chicks of Group
E were hatched on Dec. 2, 1908; those of Group F on Dec. 12,
1908. The members of Group F, thus ten days younger, were
transferred from the incubator to the cages and brooder occu-
pied by Group E. Records were taken as before. Still greater
precaution was observed as to food supply. The food was not

Trials
be]

0 VE A <I> eo ae eee
A :

3 wel] N

TOK, AF

ul
We, 7S = i a ee a a=
aa Eee ad Fs
7 alt a =]t>4 ete

Gysi2 3 4 5 6 7 8 8 10 it 12 13 15 16 17 18 {8 20 21 2 23 24 25

Figure 4—A comparison of curves of development of the pecking instinct to show
the possible effect of social influence. Distances along axis of abscissae repre-
sent days of age; distances along axis of ordinates, the number of a given type
of reaction in a daily series of fifty pecking reactions. Curves IIIJ-E and III-F
show respectively the rate of decrease in the number of reaction 3 for Groups
E and F; Curves IV-E and IV-F show the improvement in accuracy of reac-
tion 4 for the same groups.

unnecessarily stinted, however. The chicks chirped loudly when
IT came into the room in the morning and crowded toward me
when I approached the side of the brooder. In the tests they
picked up the grains very energetically.

The records for Group E are given in full in table 5, the sum-
mary for Group F in table 6. The curves for reactions 3 and
4 of both groups are plotted in fig. 4; for reactions 1 and 2 in
fig. 5. The efficiency of the pecking of the older group is repre-
sented numerically by 42.2 on a scale of 50 on their eleventh
day, when the chicks of Group F entered the cage. The twelve
chicks were confined in the brooder each evening together, were

FREDERICK 8S. BREED

32

SO6I ‘9Z-E “OFM “A anowy ‘oz anv ‘Oy ‘ey ‘ZL ‘OL ‘69 “SON SHOIHD JO ONIMONT

¢ ATaVL

6% FZ 8S WZ IZ OF GI SI ZI OL SI FI et Bt Tl or é6 8 4 9 F & B MY

ch Sh Ib S& OF Gh bh Fh IP OF Sh Sh Sb OF Zh 6I 6 ce Ie OF LE GE OF PeMO[[EMS

€& T @ @ € € € T G6 T @ 9 T 9 9 € € € F L E “*"paztag

Pr FP 2 OFT §¢ € & 4 & 9 @ Tt € @ te 1 VE If OT OL 4 ES SE ©) PMS
T 9 a | bp passin, «64 “ON

eb 98 Zh Sb &b Ib Sb Sh Sh br PR 68 EF Sh OF 6& BP Gh OF LE SE FZ ET 8 pemolfeag

9 + F €& FF € &@ &F F EE F tT € £2 € T 4 T OL FI TE GF *'**p9zteg

1 of * ¢ € FG € € @ E€ GY L 4 tT & FG F E& IT 8 BS a éE “YONAS
T b- b&b °° pass, 94 “ON

2g l@ 88 OF 8& FE Ch SE Ih IF SE Sr OF OF FE 62 OF BE FE OE Le IT poMolTeag

1 8 9 € € € ¢ € F € 6 9 € @ 8 F€ @ TT F G 9 LT 9 **""pezteg

c ¢ 9 9 8 ef € 8 F € 8 T £4 8 8S GT 8 LE OL PT 8T TE 2 L ©) FONTS
T oT T T & T T G @ T 2 ~*"' pesstl €2 “ON

ch 6F Lb Sb 6h Sb Fr OF Lb FF OF Sb Gh Sh Le OF Fr OG Ch OF Sh 8E LE ET PemoleEag

€ 4 G G 6G G¢ @ ¢ @ F €& F F L T € OT @ °"*"peztag

¢ tT T @ £T @ & © € T BG E€ G G T &F T 6 & FF 7° ¥eNI4g
T T 7" passtf, 32 ‘ON

ch ch OF OF IF cb OF 8&.8E IF OF 9E OF LE CF LE Sb FR 6E Gh 6E 66 FI PeMOT[eAS

4 € T b&b F&F FF & 8 ¢ 9 8 € £9 € € &@ 9 F FH FF C S ****paztag

IT F- 6 GF € € £ 8 € F € € 9 9 @ 8S FF F FG FLL F OT 7 HENS
T @ Tt yout I Tt T ; G ° pesstwl OL “ON

OF Zh It OF &F OF OF OF 68 SF OF EF 8E SE Zh Sh SE Sh FF 68 Th 9F Th F PpeMmoTeag

9 € T € 8° 6 bP ee @ cB Pw Se RE L F LT oe cb: ("* {peZles

- G¢ G FF @ ¢ @ 6 9 F< € F LL F€¢ € 8 BF 4 9 8S FE L =" HONIG
& GG T Be SBT T é oT 8 ‘pass, 69 “ON
9% SZ 2 &% GS 1% OZ GT ST AL OL ST FI SI ZI IT OL 6 8 4 9 F$¢ F E MT syoryp

t

16 19 36

32 29 14
7 224 230 230 248 216 213 253 246 249 251 245 259 249 251 258 242 260 254 249 252 250

5 2 1 3 38 6 38 5 2 3
4 24 32 18 32
9 18 23 17 12

4

1

6 1
50 57 34 39 58 61 18 33 31 20 27 18 26 3

46 24 11 28 13 20 25 28 17 15 26 27 20 22

eo
N
N
io
OSn
re
Yo ia)
OoOr
re
Filta I A Rs
mona
Om ow
Na eta
De .
mac tH
a
a4
3
Ser
Oo
a

INSTINCTS AND HABITS IN CHICKS

TABLE 6
Averacr DarLy NuMBER oF Eacu Kinp or Pecxina Reaction For THE S1x Cuicks or Group F

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

8

Age

Lu
oo

Cor ON
LD HD
oD

HIER 00. OF
Nr
tH
QQ
uD 4 00
on
onmwte
~~

Swallowed. .

34 FREDERICK 8. BREED

given their tests at the same period in the morning, and of course
were fed together in the same litter, both groups being com-
pelled to scratch for their food. Yet the similarity of the curves
for the first eight days, when modification progresses most rap-
idly, is conspicuous, the curve for the younger brood running
slightly lower. These young chicks, in spite of the examples of
more accurate pecking furnished them by the behavior of the
chicks of Group E, began less accurately than their elders,
remained behind them by about the same margin during the
critical period of development, and hardly equalled them while
the experiment continued.

Trias

so
48
40
as
wo
25
” eller
is ASL.
a \ \
10 Nethe ASA 4
a ¥ AS > Ley ;
HEACF “ b= ies es Gas aa <= ls
oS :
Gays: 2 3 4 5 6 7 8 9 10 MM 12 13 14 [5 16 17 18 19 20 21 22 23 2 2

Figure 5—A comparison of the curves of development of the pecking instinct to
show the possible effect of social influence. Distances along axis of abscissae
represent days of age; distances along axis of ordinates, the number of a given
type of reaction in a daily series of fifty reactions. Curves I-E and I-F show
the rate of decrease in the number of reaction 1 for Groups E and F, respect-
ively; Curves II-E and II-F, the rate of decrease in the number of reaction 2
for the same groups.

A feature worthy of notice is the fact that development of
reaction 4 in both groups seemed to halt on the third day, with
very rapid improvement preceding and following. A search for
the cause of this retardation of development in the chain of
actions is interesting. A detailed examination of the relations
of Curves I, II, II, and IV for both groups brings out the fol-
lowing points:

1st day, the course of development is unknown;

and day, striking and seizing improve rapidly, while swallow-
ing (in the restricted sense) declines considerably in effectiveness ;

INSTINCTS AND HABITS IN CHICKS 35

3rd_ day, .swallowing: improves rapidly, striking improves
slightly, and seizing suffers a reversal of form.

In other words, from the beginning of the second day the accu-
racy of striking improves uninterruptedly, the accuracy of swal-
lowing declines temporarily on the second day, and that of
seizing temporarily on the third day. It is evident that the
rate of improvement of reaction 4 depends upon the rate of
decrease of the total, number of errors in these reaction types.
In neither Group E nor Group F was improvement in all three
types of reaction uninterrupted. The retardation of reaction
4 on the third day is due specifically to a decline in the accuracy

of seizing on that day.

Wats
50
4S “7. “Z hs ms
p-~-L a ee 2 recalls ie a
40 t I | | et * N
i Neo
Ul
. Ries
30 f
| F
2s
| }
2 +
est
1S t
IN-75| !
10 f
1
i
5
veil!

Gysi 2 3 4 5 86§ 7 8 8 10 th 12 13 4 15 16 17 18 1H MOA RBwBea 3

Figure 6—Curves showing the rate of improvement in pecking accuracy of two
chicks kept in isolation. Distances along axis of abscissae represent days;
distances along axis of ordinates, the number of perfect reactions in the daily

series of pecking reactions.

In regard to the tests for social influence, the objection can
be made that social influence, if it affect the accuracy of reaction
at all, may work as well.between members of the same group
as between members of different groups. The objection has
force. As a control test, chicks nos. 75, 80, and 81 were kept’in
individual apartments and tested. The results are exhibited§in
fig. 6. The record of no. 80 is omitted because valueless. This
chick soon became: physically unfit. No. 75, although it did
not thrive under the conditions, made a good pecking record.
No. 81 not only pecked with unusual accuracy but was lively
and vigorous after the first few days. The curves do not
furnish convincing evidence in support of the view that social

36 FREDERICK 8. BREED

influence accelerates improvement in the accuracy of the re-
action studied.

The curve for no. 81 clearly suggests retardation of develop-
ment—-a very interesting fact, if due to the isolation. At first
this animal seemed quite indifferent to food. Besides, its toes
were bent under in such a way that it looked crippled. In a
few days, however, neither of these failings was observable.
They may indicate that the chick was slow in the process of
unfoldment in more respects than in pecking. It would hardly
be fair, for example, to attribute the slow expansion of the
toes to tHe isolation. No. 81 turned out to be a fine animal.
The rapidity with which the components of reaction 4 became
integrated, once the process set in, is instructive. The maxi-
mum of accuracy attained, as well as the high average main-
tained, mark the nicety of adjustment in no. 81’s pecking
mechanism.

If chicks that are isolated peck with a normal degree of accu-
racy within the natural time limit, and chicks stimulated by
the presence of others much more efficient than themselves fail
to show a supernormal rate of improvement, one has good
grounds for believing that the social influence, whatever else its
effect, does not appreciably modify the natural course of develop-
ment of the pecking concatenation.

In fig. 7 are presented the curves of development of the peck-
ing reaction, based on the averages obtained from the records
of the twenty-one chicks in Groups C, E, and F during their
first twenty-four days. The data for these curves are given in
table 7. Curves I, II, and III, standing for reaction 1, 2,
and 3, respectively, represent, as explained before, the distribu-
tion of errors. Curve IV represents reaction 4. The sum of
the heights of the error curves above the base line on any given
day will equal the distance of Curve IV below its limit for the
same day.

“Tn nearly all cases, as one might expect,’’ says Morgan,!
“the simple process of striking is more accurate than the more
complicated process of striking and seizing; and this, again,
than the yet more elaborate process of striking, seizing, and
swallowing.’’ From the data in table 7 it is an easy matter
not only to substantiate this assertion, but to show the quanti-

‘Morgan, C. L.: Habit and instinct. London and New York, 1896, p. 37.

37

INSTINCTS AND HABITS IN CHICKS

L9'Ob L9°0F IL 1h Sb OF FIG OL OF LG IF 06 IF SO'Sh LO°Zh 98" IF 29°F L9°OF FZ IF OT OF 61'6E OO'8E GI'8E L9°9E N8'SE BE TE 9L'6Z 60'6Z 49'S “SedBIEAY
$98 $98 948 SSF S88 9S8 E48 O88 E88 68 618 48 FSS 998 ZEB E28 862 GOB OLL SSL 699 SZ9 SI9 ZBI ~*“SsTBIOL

986 SEZ SZ IS ho GHZ BES BZ GHG LFZ GPS LEZ OSS GES FEZ OFS OG 62% 8IS IIS SIS I9l OMI Ge
092 292 6FZ 9S 09 ShS BGG 194 6S G6SZ SbZ ISS GFZ OFS ESS LIZ 91S 8S OFS OFZ FES LLT OLI ZY
g9e 498 918 ose 69€ L448 I88 S88 zZ6E se GSE cce GLE G9E OLE SHE GZE SE SIE GZS LBS G8S G6

yp NOLLOVaY

IL’ 14°S BSS BGS 94'S LS GI'S BFL GIS SES IB FSS OLS 98'S BE 06'S G6'S LG'S OG'E IL'F ILS BFS OL IL6S'8 ‘sedvieay
82 L2¢ 8S & 8 8S 9F OF OF 6 O8 Le G9 O89 8 19 2 FF CB 66 OL BLT LIS FLT *"'S[BIOL

af Nd Ke te 246 6 6 St 2 6t 86 9 S@ 1% 86 66 4 1&8 LE 8 IE 4 FF BE
T
Il 6 s tI Il 6 8 4t OL Gé ¢ sc 2 4t 2 Sst ot Zt OF $9 SOT ZcOT O08

€ NOWOVay

8F°S 62°9 8E8°S 89°G IL’b G09 G09 06'S ZS'S 98°F LG°S Eh'9 ILS GO'G ILS BFL 298 —*6 G68 FIG ILS SH OL Sh'9 GOT ‘SsedBIOAY
SIL Z&l fIt 49 66 221 242E FZT QIT BOT GL Set OCI 9OT OSI AST I8l 68 S88 261 249% 6IS Gel Lee ~°“S[BIOL

Te 8 Ff se of 4 Gh ec iH j%sLe Te 8 CG S& SV IT 18 Ge FH OF GS OTT 9L GET
Fl #62 we f& St se 8 6 O64 SL 4246 OG ITE $8 SE 19 8S 68 Fe Le Of FZ FE OL

Oo. 69 29 1¢ 89 19 8S OF Le 4 28 %L9 OF O98 G9 C6 GIT OTT G6 Zt GE Ge SIT
Z NOWMOVaAY
bl’ ee 88 oF Be Sh FI’ 39 GI’ 6I° IL 29 8% 8 29° 8&8 8E GI’ Sh FZ FL’ 8B Shs SF OT ‘sedvI0AYy
Std 8 ¢ 8 6 € el FF ‘4 ST St Of 2 FT 8 8 % 6 g € 46 TG GE ~""S[BIOL
z é g g 4 T 9 I 4 % € T 9 z ¥ I I I z 6 OZT “dT o»
€ 6 g g 9 € € T rd S ¥ T Tt 9 8 z z & S s9 “Hon
T s & € rd 6 I T 4 z a L G 2 6L&)~—SCOOOT: CQ doin

[ NOLLOVaY

oe oe 8% we Te © 81 St 21 8 SE FE St at It OT 6 8 Z 9 ¢ v $ rd aay
a ONV ‘q ‘9 SanoUy JO F ANV ‘g “Z ‘T SNOILOVAY WOd SIVLOY, ATIV] FHL WONT SHOIHD [Z HOM SHDVUAAY ONIN

ZL ATavL

38 ' FREDERICK 8. BREED

tative relations involved. E.g., on the second day, out of fifty
trials the object was struck, on an average, 33.48 times on the
first attempt. Of this number only 16.96 included seizing; and
of the latter only 8.67 were reactions in which the object was
struck, seized, and swallowed. As the chicks advanced in age,
each of these sets of figures approximated the same limit, but
naturally they preserved the original order as to size, for with
each additional reaction in the series there was an added source
of error.

Reverting to the earlier consideration that Curves I, II, and
III may be regarded as error curves, an examination of these

:
Ey

D
4
= mT ee
6 2
==
Bs Veo
ca
* LL 0
co) é
i
fs 7
1s ui
10 an a |---|
“ Ww" PN he ieee A ee ee ees epg ee eee ee
sy Se —_— —
Sede | ee ON Sa en it eas ae = ae
Qyst 2 3 4 5 G6 7 8 9 10 12 13 14 IS 16 17 18 19 20 21 22 23 2% 25

Figure 7—Curves of development of the pecking instinct, based on the averages
obtained from the records of twenty-one chicks. Data in table 7. Distances
along axis of abscissae represent days of age; distances along axis of ordi-
nates, the number of occurrences of a given type of reaction in a daily series
of fifty pecking reactions. Curves I, II, and III trace the course of reactions
1, 2, and 3, respectively. Curve IV shows the average improvement in
accuracy of reaction 4.

curves shows that the improvement of peckingRaccuracy is
retarded more by failures in seizing than by failures in swallow-
ing; and more by failures in swallowing than by failures in strik-
ing. After the third day, the difficulty of seizure remains greater
than that of the other two reactions combined. The striking
reaction improves in accuracy rapidly and without relapse,
closely approximating perfection by the fifth day—a degree of
accuracy that might easily inspire belief in the perfection of
instinct, for, indeed, this is the reaction which has no doubt
been central in the earlier discussions when accuracy has come
into question.

INSTINCTS AND HABITS IN CHICKS 39

Of course these averages hide the unusual performances of
individuals. Chick no. 72, for instance, missed only once out of
fifty reactions on its second day, on the fifth day it struck,
seized, and swallowed forty-eight grains in fifty trials, and on
the eighth day devoured every one of the fifty without a slip.
The complete record of this animal is given in table 5. A com-
parison of its record with that of no. 70 brings out nicely the
point of individual differences in accuracy. No. 72 developed
more rapidly than no. 70, and also attained a much higher aver-
age degree of efficiency. The records of both chicks are con-
sistent throughout. They bespeak a more finely adjusted mechan-
ism in the one case than in the other. If we were seeking points
of difference instead of identity in this investigation, we should
dwell more at length on variations in other respects. Chicks,
as one might expect, differ definitely and consistently from
day to day in the accuracy of their reactions, in the rapidity of
their movements, in their food preferences, and the like, much
as men do,

To return to fig. 7. The curve representing the course of
development of the complete co-ordination, Curve IV, reaches
its highest elevation during the twenty-four days at 42.5 on a
scale of 50. Each of these curves is intended to represent an
aspect of the pecking of the chicks under the conditions described.
Whether or not the reactions in the litter were more accurate
than those on the harder cardboard, I cannot say, not having
made any measurements. To be sure, the long periods of prac-
tice were spent in pecking under conditions quite different from
those that obtained in the tests. .

It has been the aim of this investigation to discover the mean-
ing that should attach to such expressions as the accuracy, the
perfection, or the congenital definiteness of the instinctive action
that has been generally considered one of the most perfect,
namely, the pecking reaction of chicks. Morgan and Thorndike
are correct in asserting that this reaction is imperfect at birth—
not ‘‘ very nearly ”’ perfect as Morgan thinks, but very imperfect.
But Thorndike 1 doubts the fact of improvement. ‘‘As a matter
of fact,” he writes, ‘‘ the pecking reaction may be as perfect at
birth as it is after 10 or 12 days’ experience.’ I shall have to

1 Thorndike, E. L.: The instinctive reaction of young chicks. Psych. Rev., 1899,
vol. VI, p. 285.

40 FREDERICK 8. BREED

side with Morgan here and insist that the pecking reaction
improves in accuracy after birth. Whether, now, this is a case
of perfecting through habit, involves a still further problem,
that of the relation of habit to the instinct. One sometimes
speaks of the modzfiability of an instinctive action like that
of pecking, but wherever this term has been employed in this
paper in connection with instinct no more has been implied than
the objective fact of improvement in accuracy, an increasingly
successful adjustment of parts in a more comprehensive func-
tion. The problem still remains, Is this development dependent
upon practice, or is it the natural functional correlate of struc-
tural maturation independent of practice? Swallows are reported
to be able to fly without previous practice. If the pecking of
chicks could be successfully inhibited for a week’s time without
doing violence to the normal physical condition of the animals,
would the accuracy of the reactions at the end of that time
average 36.67 on a scale of 50, the average for our lot of twenty-
one? There is evidence in support of the belief that such chicks
would very quickly be pecking with average efficiency, without
anything like the amount of practice chicks would have had
by this time when growing naturally. In other words, improve-
ment does not depend entirely upon practice. How much of
the improvement does depend upon practice? All of it, one is
led to believe from Morgan’s pages. ‘‘ Steadying of the inherited
organic apparatus ”’ through preparatory efforts means improve-
ment through habit. Besides, overestimating, as he did, the
degree of perfection of the instinct at birth, he has left less
room for maturation and the effects of practice than there really
is. So far as the facts are concerned, the most one can say
is that the development of the pecking instinct proceeds some-
what without practice and is hastened by it. Maturation and
use run along in time together. No means has yet been devised
of measuring the amount either factor apart from the other
contributes to the development of the pecking reaction.

The importance attached to individual acquisition as a factor
in development seems on the whole to become increasingly
restricted. The theory of the non-transmission of acquired
characters enormously narrows its scope in phylogenetic develop-
ment. Perhaps we shall discover its lesser importance in onto-
genetic development. The drinking reaction, for example, is

INSTINCTS AND HABITS IN CHICKS 4}

more self-dependent than we have hitherto supposed. The peck-
ing reaction suffers the supplementation of habit, one may well
believe, when it does not demand it. It remains to be shown
to what extent these instincts are typical of instincts in general.
But even if it should be established that acquisition contributes
telatively little, environment would still remain a powerful
factor in development. The animal begins life with an hereditary
endowment in interaction with the environment. One is neces-
sary to the other. In the economy of an organism there is no
reaction without stimulation. And not only is the environment
a system of energies without which natural tendencies of the
organism cannot be realized, but it is a selective system. What
tendencies shall be realized will be determined by what stimuli
are provided. Heredity and environment are not opposites, but
complements. ,

PART II. ACQUIRED REACTIONS
I. Introduction and Statement of the Problem

From the study of the modifiability of reactions that are
made prior to experience and are supposed to be based on
inherited neuro-muscular co-ordinations, our interest now shifts
to a study of the development of reactions that depend upon
no such hereditary dispositions in the nervous system, but are
individually acquired. It was one of the aims of this investiga-
tion to discover not only whether reactions to certain optical
stimuli or stimulus-complexes are modifiable, but also to deter-
mine the rate of modification; in other words, to describe the
progress of habit-formation in quantitative terms.

Modifiability was tested by the so-called discrimination method.
Never more than two possibilities of selection were offered to
the animal at one time. The conditions were so arranged that
the chick’s natural tendency to react to confinement and soli-
tude by efforts to escape furnished the necessary random activity.
When, after a certain number of trials, the animal reacted
selectively to one of two constant form or size or color stimuli,
for neither of which it displayed a preference before training
began, the process of habit-formation was adjudged completed.
By selective reaction is here meant the ability to react without
error to a constant stimulus for an arbitrarily fixed number

of times.

42 FREDERICK 8. BREED

The adoption of the discrimination method followed directly
from the results of Yerkes’ work with the dancing mouse. I
have appropriated his method, my contribution being merely
an adaptation of the method to the study of chicks. So much
for the problem in a general way.

Ii. Apparatus

The apparatus used was a fan-shaped reaction box built of
wood and painted brown. Fig. 8 shows the ground plan, fig. 9
the perspective. The method of construction was as follows.
On a base 88 by 95.5 cm. a convenient point near one edge was
selected from which as a center, with radii of 38.64 cm. and
69 cm., respectively, arcs of circles were described. This center
later became the point from which the chick started at each
trial, and around which the entrance box was built. The nearer
arc marked the location of the cards first approached by the
chick; the more distant one outlined a boundary of the apparatus.
The box was made with four compartments, but only the two
inner ones were used in these experiments. Removable parti-
tions (J, fig. 8) made possible this use of only a limited por-
tion of the box. The height of the apparatus was 20 cm., inside
measurement. The covering consisted of two lifting doors, one
over the entrance box A and the other over the rest of the appa-
ratus. These were made of light wooden frames and wire netting,
and were so arranged that they lifted on their hinges directly
away from the experimenter as he stood at the entrance box.
The mesh of the netting was 3 cm.—as large as conditions would
permit, so as to interfere with the light as little as possible.
The exits H were closed with removable galvanized sheet iron
sliding doors that fitted from above in vertical grooves. The
end of the box about G was closed with wire netting of .86 cm.
mesh. Grooves at this end also provided for wooden screens
inside the wire of sufficient height to conceal the flock of chicks
in a cage just beyond and below the experiment box from the
chick reacting. Two sets of cards were used, one at D and
another at F. The card-carriers at F projected 2.55 cm. from
the walls, making it impossible for the exit H to be seen from
the chamber C.

Following the ground plan in fig. 8 and the path of a chick
through the apparatus from the entrance box A, the remaining

INSTINCTS AND HABITS IN CHICKS 43

dimensions and details may be clearly supplied. From the
entrance box A, 16 by 20 cm., the chick was released by a wire
screen door of .86 cm. mesh that was raised and lowered in its
grooves at B by the experimenter. This retaining screen was
shaped so that its lower edge coincided with an arc of a circle
constructed from the center above referred to. The width of
the opening at B was 10 cm. The first cardholders were at
D, the passage ways through which were 9.33 cm. wide. Through

8—Ground plan of apparatus for testing habit-formation. A, entrance

Ba B, i en ae door; D and F, location of card-holders; E,E, electric

chambers; G, H, I, way out of box; K, electric key; L, electric battery; M,
inductorium.

D the chick passed into the electric chamber E, the floor of
which was crossed with copper wires 1 cm. apart, so arranged
that they formed an interrupted circuit which could be closed
by the chick when the current was on by its stepping simul-
taneously on two adjoining wires. Out of E the chick: passed
through the opening F, 8.5 cm. wide, the position of the second
card. From G there was an. outlook through the wire netting

44 FREDERICK 8. BREED

upon the flock of chicks in the cage below and also an outlet
through H which was 8.5 cm. wide. The door of galvanized
sheet iron could be inserted here to close the exit when desired.
Once in I, the chick joined the rest of the flock by passing down
over the door of the cage that was turned up toward this end
of the apparatus at an angle of about 45 degrees. The dimen-
sions and details were the same by either of the routes a chick
might take from A to the cage. K indicates a key by which
the circuit was made and broken. The current was supplied by

u r

Figure 9—Perspective of apparatus for testing habit-formation, ground plan of
which is presented in fig. 8.

a no. 6 Columbia dry cell which was connected with the primary
coil of a Porter inductorium, the secondary coil of which was
in circuit with the ends of the two wires that were wrapped
parallel to each other about floors that were fitted into the
electric chambers E. During the tests the apparatus was so
placed that the entrance box was nearest the source of light
and the sides of the apparatus were symmetrically situated with
reference to the two windows that admitted the light. The
cards used at D were 14.25 x 19.5 cm. and those at F 13.5 x

@
INSTINCTS AND HABITS IN CHICKS 45

18.75 cm. In all color work the openings in the two sets of
cards were the same in size, 8.5 cm. wide and rr. 5 cm. high.
These rectangular openings were cut in the lower part of the
card in such a way as to leave equal margins of color on either
side of the entry way.

III. Method

For a test of the reactions to color, form or size the appro-
priate cards were placed in the card-holders. For example, if
reactions to black and blue were to be tested, one might start
with a black card in each of the holders at D and F, right, and
two blue cards in the corresponding holders on the left. The
cage in which the animals lived was carried to a position with
reference to the table upon which the apparatus sat such that
one of the wire netting doors of the cage would turn back toward
the exit-end of the apparatus as above described, and form an
easy passage-way from the reaction box to the cage.

The chick to be experimented upon, marked and numbered,
was carried from the cage to the entrance box. When placed
in A, it soon acquired the habit of crowding up near -the wire
netting door, standing at a point equidistant from the two
cards exposed at D. It also soon ceased pecking at the wire
covering or flying against it, if it happened at first to try this
way of escape. The work had not gone very far before the
plan was rigidly adhered to of not releasing a chick from A
until it made an effort to escape by urging toward the screen.
Sometimes this meant a delay of several minutes, but it seemed
useless to spend time with a chick that did not react negatively
to the confinement or isolation. Further, care had at all times
to be exercised that a chick be not kept struggling in A. Results
were vitiated if this occurred. I feel sure that discriminating
ability deteriorated as the response to the isolation of the reac-
tion box approached too near in one direction the passivity
of indifference, or in the other the activity of excitement. I
have watched the record of indifferent chicks recover from a
lapse when the response to isolation was accentuated by suffi-
ciently prolonged solitary confinement between trials; on the
other hand it is easy to point to numerous instances where a
record has been marred by undue excitement. I concluded from
experience, that much the wisest plan was to watch for an
incipient struggle to escape, and to lift the door just when the

2

46 FREDERICK 8. BREED.

impulse had begun to gather force. There is a nice middle
ground here that gave the best results. There were great indi-
vidual differences in the animals. Some stood about for a while
before the impulse to escape seized them, others rushed imme-
diately toward the screen door. The experimenter stood back
of the entrance box between the two windows of the room, and,
after releasing the chick from A, remained as quiet and motion-
less as possible’ until the animal found its way out of the appa-
ratus. I had practically no success with the device Yerkes
used with the dancing mouse, the card by which he gradually
forced the animals toward the electric box. Chicks that won’t
work without force won’t work with it, is a rule that can be
generally relied upon. After the animal had escaped to the cage,
the experimenter secured it again, brought it back to the. en-
trance box, and the procedure as described above was repeated.

At the beginning of a set of experiments the animal was
familiarized with the apparatus, as well as tested for its prefer-
ence of the stimuli presented in the card-holders, by a number
of “ preference trials,’’ usually one series, which consisted of ten
trials. Sometimes a second series was given. These series are
denoted in the tables by A and B, respectively. During the
preference trials the exits at H were both left open, the two
sets of cards were alternated right and left after each trial, and
no electric shock was used. To show the readiness with which
the chicks reacted in the preference trials as well as to suggest
the availability of the animals for certain kinds of labyrinth
training, I present below as typical the average time in seconds
for each preference trial, from one to ten, of chicks nos. 32, 33,
38, and 39, Group I, at the beginning of their black-blue work:
69.2, 24.1, 18.1, 3.8, 3.4, 2.6, 2.1, 1.9, 2.2, 1.9. Chicks thus
tested for an experiment were continued in that experiment, un-
less they were found to have a preference for the stimulus to
which it was desired they should form the habit of positive
reaction. A chick that revealed a preference for this stimulus
prior to training was rejected.

After the preference trials had proved a chick satisfactory, it
was subjected to the regular conditions of training. The ‘“‘wrong”’
side of the apparatus was closed at the exit and the animal was
given a shock if it entered the electric chamber on that side.
This was done by quietly pressing the key previously referred

INSTINCTS AND HABITS IN CHICKS 47

to. The shock was not severe enough to cause the chick to
jump from the wires or chirp on receiving it. It was found
that a shock from a current value of about 1.69 amperes with
the secondary coil set at 5.5 according to the scale on the induc-
torium, proved most generally satisfactory. Here again there
were individual differences in sensitiveness among the chicks,
but in the main the above shock brought results without being
injurious, and had no more effect on the chicks than to start
them from their position, occasionally causing one to lift a
foot from the wires. The floor of the entrance box was kept
wet so that the feet of the chicks would be in a condition to
receive the shock with regularity.

At first a chick usually ran right over the wires and on to
the front netting, from which there was a full view of the other
chicks in the cage. With the chick in this position on the
“wrong ”’ side, it was found best to wait until it happened
upon the right road out by its own random activity. Some
times the wait on the first training trial was ten minutes, but
such a delay was exceptional. The young chicks, as a rule, were
active and soon acquired the habit of withdrawing from the
“wrong” side promptly. The draft on the experimenter’s
time after the training was well under way, averaged about
I.5 minutes per reaction, including everything. No trouble was
experienced with hesitancy on the part of the chicks to return
over the electric wires. Of course they were never shocked
on the return passage.

In a series of trials the positive card was shifted right and
left in this order: r,1,r,1,r,r,1,1,r, 1. Thus each card in each
series was displayed the same number of times on the right-
and left-hand sides. When the chick had made a series of error-
less reactions with the above changes of cards, it was tested
for another perfect series with a different order and an equal
number of rights and lefts. If no errors were made this time, a
third series with still a different sequence of changes was given,
when, if the chick reacted without error, the process of habit-
formation was pronounced completed.

In none of the work with this apparatus was it necessary
to starve the chick or even to keep it excessively hungry. It
is my opinion that hunger did not play a more important part
than the reaction to confinement and solitude. True, it is easy

48 FREDERICK 8. BREED

to show that, as a rule, a chick with a full crop is not so good
material for an experiment like this as one not so well fed, but
these animals while being experimented upon did not always
rush about for food as soon as they had reached the cage.

IV. Experiments

In these experiments on modification thirty-eight chicks were
used. The work was done from Dec. 15, 1907 to June 30, 1908.
To test the interrelation of habits of response to optical stimuli,
the necessity arose of discovering at least two methods of measur-
ing modifiability in this sense mode. It was desirable in the
case of any one method that the stimuli bear such a relation
to each other that the average chick could complete the training
within a reasonable time, and that the time should be as nearly
as possible constant for different chicks. The feasibility of the
electric shock and satisfactory conditions and devices for train-
ing the chicks were determined upon after several hundred
preliminary trials in a crude reaction box. The apparatus
shown in figures 8 and 9 was constructed on the basis of experi-
ence in these preliminary tests, combined with the suggestions
afforded by the apparatus of Yerkes‘: and that of Hamilton?
used in similar experiments. Means for retaining the animal in
the entrance box until it made movements to escape, permitting
it meanwhile to be within range of the stimuli; the location
of these stimuli at a suitable distance from the point of release;
the open screen overlooking the cage—these features were
planned with characteristic activities of the chicks in mind.
Releasing the chick too soon or too late at any given trial pro-
duced unsatisfactory results. If the cards were too near, the
pulse of activity often carried the animal into the electric box
before the card stimuli had wrought the proper effect. And when
the animal had made an error and had arrived at the closed exit,
the chicks in the cage below acted as an additional stimulus to
escape. The animals readily found the exit through H when the
door was open.

A. Color

In all the color work the reflection method with the Bradley
standard colored papers was employed. Hue, tint, and shade

1'Yerkes, R. M.: The dancing mouse. New York, 1907, p. 92.
? Hamilton, G. V.: A study of trial and error reactions in mammals. The Journal
of Animal Behavior, 1911, vol. I, p. 33.

INSTINCTS AND HABITS IN CHICKS 49

are described only by reference to the catalogue name or num-
ber of the papers. Measurement of tint (or brightness) was
attempted in only a few special cases. The purpose of these
experiments at their inception was not to investigate the color
reception of chicks, but to test modification of béhavior. How-
ever, the reactions to reflected colored light proved so interesting
that a slight digression from the central interest of the research
was later made in the effort to gather some new facts in this
closely related field.

Chicks were tried first with Bradley black and white. Most
of the first tests involved many elements of irregularity. In
some the right and left exchanges of cards were not balanced
in the series. In others, chicks had been shocked too severely
and either hesitated about entering the electric boxes or balked
completely. Again, work was carried over from the preliminary
to the new apparatus. The best of these first results are pre-
sented below. Chick no. 7, after being given preference trials
which were of questionable value, was regularly trained to
avoid white and go to black.

TABLE 8
Buack-WHITE REACTIONS
Cuick No. 7. Hatcuep 12/10, 07. Sex, F.?

Series? Date Right | Wrong
SDs sSte a ecoatis cece ie ree auc fh a a Pate na Rano Jan. 8 11 9
De ise Sy batieed eicaiacdachd Gantt h DA Ske DRE A “10 13 7
Disaotinint wet Clk MAA h na SeaMauheu deg asm ei Subsea eS 1 13 a
a ge Ne Pa en el RNa is once aN ha nth ae Sse “13 10 10
DSi cast tehury tpn att altey.g ec Sendotenuan idk tece GE an seeded ha “14 12 8
6. cee ae She rein Meee ee eee ee «15 9 1l
Tee Gis St 6 EEE SSE ED EEE SEI ES «16 7 13
By ee td DEL MR ERUCR EEO LES ines © 17. 18 2
Qh uid eS da hinkae ene ee Osea Baie ie . 18 2
TOs ap oxccat nai ete eas baer ee oe “« 20 17 3
DS ots veg css sides cdc Sottha texas aot RN suo aey Dia Sahn ME be 7 3
Vie ects escecon iAbasny Ree ONS TA AER AAO ES «21 17 3
UBt e lrceekwaaiig ee ee ae eet s Sau sne t's Bones BS D2 19 1
TAocccak thas MG BARES HEE Ae ea coer “ 23 17 3
PD cc ssc esetacereie cag Mack abana Re Baan Bee Pema as ee 2: 9 1
DBs oo Gaay a Abe acne b Hier tees eee eG Us SR ay eee “ 10 0
TG i dosed cua nalts Rare hel ohaaens Hien ton be eae eee ae ae “25 20 0

1In all the tables of results in discrimination work the stimulus to which a chick
was forming the habit of positive reaction is each time mentioned first. It is to
be assumed that the animals were without previous training unless a definite state-
ment to the contrary is made. ; ;

2 All sex determinations were made by dissection. __ :

3JIn some of the preliminary work the regular 10-trial series was doubled.

50 FREDERICK 8. BREED

In table 9 are presented the results obtained from chick
no. g under conditions exactly similar to those under which
no. 7 was tested. No. 9, however, was trained to accept white
and reject black.

TABLE 9
Wuirr-Buiack REACTIONS

Cuick No.9. Harcuep 12/10, 07. Sex F.

Series Date Right Wrong
Ma icin Se Babe ue etetiels nctatiat At chs wibent tebe. Gentacjun tase Jan. 9 11 9
ED a recat nH web a? BR ag Si 15 gah ite Sel a ee OE RAEN “10 9 11
De Bh 28S las eases BANA ara ia aa ban eae dando dim cate aetl hs «13 13 7
A ed ase, fares wr oh ncra ares Sean Bop aah a Bante “14 12 8
Doig in eas PEA RA SER EEE kee e Pes eats «15 13 7
Cie d casi on hate awd ene WP) eae aa ali . © 36 14 6
Discs bids ges Bit ceite Ae liste GER AYE Re EE SO) or ARE 20 0
Bc Fea HAGA NGC a aN AO OG endear cae ON . s 10 0

Now, if some animals were going to require, as no. 8, 310
tests besides preference trials to complete the black-white train-
ing, working with this habit would consume too much of the
experimenter’s time. An objection to the white-black habit lay
in the preference prior to training that chicks showed for the
white. If a combination of visual stimuli could be discovered
for neither of which the chicks showed any marked preference,
such a combination would commend itself for use, provided the
time required to form the desired habit of reaction should be
fairly uniform among different chicks and conveniently short.
The combination, orange and white, was tried with the result
shown in table 10, The work of chick no. 12 was devoid of
the irregularities spoken of above.

TABLE 10
ORANGE-WHITE REACTIONS
Cuick no. 12. Harcuep 12/10, 07. Sex, F.
Series Date Right Wrong

Ks)
COCRFrORKRSO

- INSTINCTS AND HABITS IN CHICKS

51

This record was the cleanest and most promising so far, but
was offset by the results obtained from experiments conducted
concurrently upon no. 10. The latter chick,in a period of training
twice as long as that given to no. 12, failed to acquire a perfect
habit of response, to orange. The preference trials in each case

were satisfactory.
TABLE 11

ORANGE- WHITE REACTIONS

Cuick no. 10. Harcuep 12/10, 07. Sex, M.

Series Date Right Wrong

pHa Dey Ree wae E Raa Jan. 29 10 10
W Stitt AOS spel ba raunete eee Minagawerarég “ 30 5 5
cists Rk Ee Se Sek Made dk eel BG oA Rea SB a “31 4 6
2 Gidea Meg eis Be NE ep eS UE a LAR a re Feb. 1 5 5
wis dete aod Enath B avant Sw Pigs Sa ne aoa dee eieenen ae + 4 6
Bk Sect Sel sed eae | MON Nee h nbaak aa RENE Raia “4 5 5
alcibe ances tce nen arn eG mies oe matine AoeWalh se qanteee sy ethan fe 5 4 6
BN ah whee a cinta Gra ahi dart entad uN aN feseneis re ae © 6 4 6
silat uae on et aves wml tba eal enalene are wise aesenstee ys i 7 5 5
Aus cy IS SED ONS HEH ES BA ee 8 6 4
Best indeed icaue dnt BA seers) Sue tes Geico go ayactine Bae “« —10 9 1
Seseedaes Nh auch gate ints BION OAR AROS “il 8 2)
5 Sys tous ieh aeoa oy Ee Soa Os AB DAMS Anca M ode iarreneeensinaes es \GE2 8 2
se Na a HA Voge ecatp CE. Sshata ae hE Mh GMa ema oT 9 1
Lag ncn gates acd Ray nan nel a ule asta a eG “14 7 3
‘ofc! eh Da wees be ees mide devlaken Liste Say ealanace 15 9 1
= falonstasen 0G DUA Wee ee St RS EE eS Ne 9 1

Chick no. 8 was given more searching preference tests and

then trained on black-blue.
TABLE 12

Buiack-BLUE REACTIONS

Cuicx no. 8. Harcuep 12/10, 07. Sex, M.

‘Series Date Right | Wrong
Orange Blue
IN ta elds Bd ihe a dal acne BE ae OR ERR eR Jan, 27 13 27
Black White
BR ose eaten ea eee te oe a es Baas eae asta Jan. 28 6 14
Yellow Blue
Oh os. aise rhea aee sheen nes Jan. 28 7 13
Black Yellow
Dito cee eae seers ae et wees Jan. 29 6 14

52 FREDERICK 8S. BREED.

Black Blue

alesse dasa! areegathitnt a tatabl dca ead anse areas Age taunts Jan. 29 2 8
DN fe ir uate oh hateal haces ROE Ac eg Baas Loe SAD a fe -Bl 1 9
Bike cides ciate hcg os Gore dn ana caeaasesuee sealer ee Feb. 1 1 9
Reiley ngeh SA 4 PRS A EER Ge ROARS a Pie pete es 3 2 8
Decors ay cade epic che SeelRented a ahh ataadttles lla A 4 3 7
Gila rnd Aoly qe Tk oe ales beth a hey ee ne £ 5 6 4
Tiskoieketiumetbitie Gl hire dee gine be Galan ce reas . 6 4 6
Shier Derielian Buna hgh emer eA enn ey eae esc ae 7 8 2
Os en see eathid actiele A ad, OOO DUNE e Wtialy canta S “ 8 7 3
sae dss dee! hd cect a Ste oh eda ch A a Nh & oy La «10 8 2
BE asic a aah ak ee PS ch Rae we W Abale G Se isle aN one bf 8 2
IB gischa Noss encabth Ge 8 tere Nah acta ch we GENTS AO te tess ee eR ac a s 12 10 0
Tee tegatana wed age ewkes BES ae ee es me - 13) 10 0
Lt ere eared ee ree et ee re ene Tene Sew ene “14 10 0

Preference tests B and D, table 12, indicate a natural pref-
erence for the brighter of the two cards. The orange-blue and
yellow-blue preference tests brought to light interesting data.
The yellow here was very light, much brighter than the com-
paratively dark blue, not only as judged directly by the human
eye, but as tested by the flicker method. Particular attention
is called to these first reactions to blue for, in so far as this
paper has anything to contribute to the study of color vision,
the interest will largely center in and results will turn upon the
reactions of the chicks to blue.

The training of no. 8 on the black-blue was now carried through
with the result detailed in table 12. Within a reasonable
number of trials, 140, including the final three perfect series,
the chick nicely developed the habit of selecting the black.
It was thereupon decided to adopt the black-blue combination
for a more searching test, with a view to its use in the determina-
tion of the rate of modification, the permanence of habit, and
some of the effects of training.

Chicks nos. 16, 17, 18, 19, and 20 were selected for this work.
From all appearances they were normal chicks, and were chosen
at random from the healthy specimens in a flock of seventeen.
They were twelve days old when given their preference tests.
In all tests prior to those reported in table 13 but one set
of cards was used, the cards at D, figure 8. In this set of ex-
periments, after series 2, an additional pair of cards was placed
at F, whereby shock and color stimuli were presented simul-
taneously and not successively only, as in the previous experi-
ments. In table 13 the rate of decrease of error or the rapidity
of modification is shown in the column, “ average number of

INSTINCTS AND HABITS.IN CHICKS 53

errors.”’ On the basis of these results with black-blue there
seemed to have been developed one habit of response to optical
stimuli that would serve both for a measurement of the rate of
modification in the chick and, when used along with another
method yet to be discovered, as a test of the interrelation of
modifications.
TABLE 13
Buack-BuuE Reactions
Cuicks Nos. 16 ro 20. Hatcuep 2/5, 08.

No. 16 No. 17 No. 18 No. 19 No. 20 Av.
Sex, F. Sex,F. Sex,F. Sex,F. Sex, M. No. '

of
Series Date R W R W R W R W RB OW errors
Bitte eseonde siete deintrat Feb. 17 5 5 5 5 3 7 4 6 5 5 5.6
Mera rociamieuih dg else Sade “18 3.7 iD) 1th 19 6 4 4 6 6.6
De a nis deen toen peck «19 3.47 3.7 5 5 5 5 6 4 5.6
Socciaeyesiee wees «20 6 4 7 3 2 8 3.7 6 4 5.2
oer eae ee er aoe 4 | 9 1 8 2 8 2 6 4 8 2 2532
Destrawinsiiena RIN oO % 22. 7 3 #10 0 5 5 6 4 8 2 2.8
_ ea eee eee “24 9 1 9 t 9 1 6 4 9 1 1.6
y COR ore eee © 25 6 4 10 0 7 3 #410 0 #10 0 1.4
Gisicenues on eke cerning «26 10 0 10 0 9 1 10 0 10 O 0.2
D sapttaty oieielal Ba Mae a “27 10 0 100 100 100 10 0 0.
AQ fey Riecueiin dake sone «28 9 1 9 1 0.4
ih) eeenaee Seererene ce «29 +10 0 10 0 0.
DD ae tech stetety coves sae oe bose Mar. 1 10 O 10 0 0.
VB iyaie sea wa wee gels be 2 10 0 10 0 0.

The persistence of habits and their effects, as well as the
bearing of the results upon the problem of color vision, will be
discussed separately in a later section of the paper. For the
present, interest is confined to certain aspects of the process
of modification, and the question as to whether the reaction
of the chick is determined by the quality as well as the intensity
of the color stimulus is not a matter of special concern. The
term color is used simply to denote a certain stimulus, with no
implication that the chick’s reactions are affected by variations
in wave-length of the ether vibrations.

Let us turn now to the search for a second method of measur-
ing modification.

B. Form

To test the animals’ responses to forms, black cards, cut as
previous cards, were used in the regular card-holders of the appar-
atus. The cards bore two-dimensional forms above the openings
through which a chick had to pass on its way out of the appar-

54 FREDERICK 8. BREED

atus. That is, forms were cut out of white paper and pasted
midway between the two sides and a little below the top of
the card. The chicks were first tested with a square and a
triangle of equal area. The triangle had a base of 8.13 cm. and
an altitude of 6.35 cm. The square was 5.08 x 5.08cm. No. 11,
a chick that had succeeded in acquiring the black-white habit
to the point of making one error in the last twenty trials, reacted
in the square-triangle test as shown in table 14.

TABLE 14
Reactions TO ForM
Cuick No. 11. Hatcuep 12/10, 07. Sex, F.

Square-Triangle

Series Date Right Wrong
ie cssacstsuilat a Suma ABA Bratt! deh os pete Regu Reetbgbe Jan, 31 5 5
Dist ara hs elect a a eure Hay wet gata are ai ve “ 1 9
Dias tia Bonn Dia Laks 2 abe categorie rma Feb. 1 3 7
bee Ew Beene Uh dees SAE Ct lees ih gan < 3 4 6
As te Si cyfi a IAed Sapsn 8 Blank adr Hane aad Aeon aaa A sos - 4 6 4
ID ih eakabsnaie Rot Pald-sake tha aia teeph Aue caehee avaineutbie Pacts fe 5 6 4
OA ee eo toabeehs Oe alee 5 Matha ta Geis Goce eI NAT ae cas 6 4 6
Mie eit UE Cob rE a Noh apatite dnt tink nate iet “ 7 7 3
Bic haiciehnateo Au and eemaarh @-aatviinlc hank Suse bcuuaae ach Monts 8 5 5
aac din soe aie Sele arseh teh eRe aoe loca Ae “10 2 8
LO vigks ace atonal, pains Saaikln ae Aas wa hla dh haa AS «il 4 6
Lisa hing Sucenbis So ean dow teria bn BAL vier chit ada dat le 6 4
DD ag raed area nye a pun plainer a aad pte ata « 138 2 8
TOia cece: tug slows ek eae aie ae alee eg. Sih ales Nes “14 3 7
Do eer eneee eto re neh eae an ea a rere “15 6 4
Ds oeasie rsa act pti duané Seat Ma od Glas Satsantnis Sue Des “17 4 6

Not only were the results here purely negative, but one had
only to watch the chick to become convinced that the prob-
ability of its forming the habit of selecting the square was
almost nil.

No. 13, a chick that had acquired the white-black (darkened ')
habit perfectly, was tried at the same time by the same method
in this form test with the result shown in table rs.

No. 13 showed no signs of improvement and the experiment
was discontinued. The square-triangle combination was dis-
carded for a “ design” that was placed at the top of the cards.
This time the cards in the apparatus were white and the forms
upon them black. On one set of cards were pasted, vertically

1 The electric chamber on the side of black was covered with cardboard.

INSTINCTS AND HABITS IN CHICKS 55

and 2.5 cm. apart, two parallel bands.of black paper, each band
being rectangular in form and measuring 4.2 x .9 cm. On the
other set of cards two black bands of equal area and the same
material were fastened at right angles to each other and in
this wise (A). Chick no. 21 was tried first with this combination
of forms, but with negative results. No. 15 was tested at the
same time. It showed no improvement after 60 trials and
from that on the continuous training plan was used. On Mar. 1
it was given 160 trials. The regular system of card exchanges
was disregarded and one setting of cards was allowed to remain
during several successive trials to see if the chick might thus
hit upon the habit of selecting the vertical bars. Sixty trials
were given in a similar way on Mar. 2 and as many more on
Mar. 3, with negative results.

TABLE 15
Reactions To Form
Cuick no. 18. Hatcuep 12/10, 07. Sex, F.
h Triangle-Square
Series ‘Date Right Wrong

ABONANNMAPAWAAwDWPwWH
TR WOW ROOANROAN ONS

The plan of having forms at the tops of the display cards
was then abandoned and another scheme employed. This
time display cards were made of gray cardboard and the forms
were presented as openings in the cards through which the
chick had to pass on its way out of the apparatus. In one set
of cards a circular opening was cut the diameter of which was
8.47+cm. In the other set of cards a rectangular opening was
cut whose base was 6.35 cm. and altitude 8.89. The circle and

56 FREDERICK 8. BREED

the rectangle were constructed as nearly as possible of equal
areas. The regular manner of training was resumed and the
reactions of chicks 15, 22,and 23 to the circle-rectangle com-
bination were tested. At the risk of being tedious I present
the results in detail, for whether or not the ability to react
selectively in this case would have demonstrated the depend-
ence of the reactions on the element of form, it did not seem
unreasonable to expect the chick to develop a habit of reacting
differently to stimuli as different as these.

A glance at tables 16, 17, and 18 reveals only negative results.
The chicks soon came to a point in these form tests where
they did not make an effort to escape from the box. Pre-
caution was taken to see that the electric shock was not
too severe. Previous experience in the color experiments had
shown that a chick occasionally received a shock of too great
intensity and thereafter hesitated about entering the electric
chambers. But the shock for nos. 15 and 23 ran as low as 6
to 6.5 on the scale of the inductorium, and did not reach a greater
intensity than 6.5 for no.22. The regular shock, as stated before,
was 5.5. Just prior to the discontinuance of the tests, no. 22
would approach the display cards, hesitate, at times shake its
bill, and go no further. The expedients of moving a card over
the top of the apparatus from the entrance box toward the

TABLE 16
Reactions To Form
Caick no. 15. Hatcuep 2/5, 08. Sex, UNDETERMINED

Cirele-Square

Date Right Wrong
Mar. 6 4 6
“ “ 2 8
a“ “ 2 8
“ “ 3 7
i“ “ 4 6
“ 7 4 6
a“ “ 4 6
“ “ 2 8
“ 8 2 8
“ ccs 4 6
i : :
“9 5 5
“ « 4 6
“10 3 7

INSTINCTS AND HABITS IN CHICKS 57

TABLE 17
ReEactTions To Form
Cuick No. 22, Harcuep 2/5, 08. Sex, F.
Circle-Square

Series Date Right Wrong
NA eats tein a tiaras ois See Gad eo RTA AA assed mtn Mar. 9 4 6
Ti Josovanion tie saben napa teen eet esl ae tenet oe pag eae “10 4 6
Diacice ccna wetely ou eutaceaeh oat mintaahiia a am ela aese ey oe 5 5
iD eckssdeae ar saa Meas aoe aag hia ou ea ea ee eS AD 5 5
4 News nse wine SA MR EEE OE CORED Os HRS RS a ve 5 5
Diigitedht, Pe sina wrasse ta wvalire-4 lean bie evens oa «13 4 6
Gs isecintarss ta Sesyr eieraunen aiansusur eB avai da @hepahn ar ianaeid aacil ae 2 8
Ms ssoinaeaes 8 rt enna cab RaaM De € the cb Sadan 2 oper reeset “16 6 4
Dikcisiualt ny bastard aanered danaseaman aun aeeanne tre oe 4 6
TABLE 18

ReEacTIONSs TO ForM
Cuick No. 23. HatcHep 2/5, 08. Sex, UNDETERMINED
Circle-Square

Series Date Right Wrong
ANS ask Ge atcee es eee uteh atta ag is Aiea SRE US tee Mar. 9 4 6
Loca csicd Sten. clnheniaias ceed acuta edt iano anc yatee Gast . “ 9 1
7 RE OEE REE a Oe Te ane eae « 10 2 8
CB eeciac eh ance ead ae caer tatt aid Roce ae Hf “ 6 4
iis Sl coi Abts, Sus ae Gta H laa PHM NE ei “12 2 8
BS econ whims eid neseccuiahy have Metal haute Pia sass AN x « 13 6 4
occas sap SEE EES EES NEE ETN os 6 4
Toschai 058 sper 86S HPAES Ee EOS GS BEALS “« 16 8 2
Bic EE dat cols tual Sn canes SAepi sax neh WoeSp ei Alec WP Selo ARON “17 5 5
Oia aencav iin a oe US Eg URS Teak safe cated ane MBarG fe Li) 4 6
LOE degeveiagelelercels cas Steed ihe aed cael s Gdn a asaesh «20 5 5

chick and of forcing him gently from behind were tried without
effect. The chick simply ‘ gave up,’ as my note has it.

C. Size

The search for a second method of measuring modification of
behavior was continued in an investigation of the reactions of
the chick to size differences in two dimensions, difference in
form being excluded. The plan was adopted of having openings
in gray display-cards through which the chick had to pass. The
openings in one set of cards were 6.35 cm. wide and 8.89 cm.
high; those in the other set 8.89 cm. wide and 12.44 cm. high.

1 Later, during some work as yet unpublished, I found a chick that acquired
the circle-square habit when the experiment was conducted in a dark-room. Opal

glass was electrically lighted from behind and the above forms, cut in black tin
screens, permitted illuminated areas of these shapes to be exposed to the chick.

58 FREDERICK 8S. BREED

The openings were the same in shape and one was approximately
twice the size of the other. The experiment was undertaken
with chicks no. 24 and no. 25, and completed with the result
detailed in table 19. The training was finished quite satis-

TABLE 19
Reactions To SizE: SMALL-LarGE OPENINGS

Curcks nos. 24 anp 25. Harcuep 2/5, ’08

No. 24 No. 25

Sex, M. Sex, F.
Series Date R WwW R WwW
eer ssnsh ea uae dada Manin Spann An anes Mar. 17 5 5 3 7
Dea tara toncnalys esr santa same dvoet aan gear aa « 18 0 10 3 7
De whe ents Bad ern ce yulampeniahar us FB Rtien e Gte mo 3 7 ~~ (frightened?)
rats toimaieet-s ty gediw ein Heeb an wahonoe aun «19 5 5 4 6
Bes Lak oe AE Sed BONES ESE EI SES ee « 8 2 6 4
95S. sign Guetnl Sosa ede ie ee oe FNS Ge «20 7 3 7 3
Bik one cnera Ein See cds a tech eee ates CP ae 7 3 5 5
TT iccsidigh igi Battbceen he table Nee hole tere tat eae Eh Sn “21 8 2 3 re
sch tases oalsge ea te pba ak sly PED RR 8 etic sf 6 10 0 6 4
Ol atat fete ayy Sone ditocs Smrcneate nes ated acee gael an “ 283 10 0 10 0
OIC 8 etary rae mre oe cara ie ene rae ne eS 10 0 10 0
Di soos ica ese Papen gM ba ioral” AEN 'e Ones ane, Pen es “25 9 J

ConrroL TEsts

Card of small opening brighter, and its electric box brighter:
Alf no hS Breas logue tattaler ick lduutysl 4 Ssh azets aiceNnte gs Mar. 24 9 1

Card of large opening brighter, its electric box brighter:
Mar. 24 9 1

Cards equally illuminated, electric box of small opening made the brighter by
illumination of electric light:

Mar. 24 9 1

Card of small opening brighter, electric box of large opening made much darker
by overhead cardboard:

Mar. 24 10 0

Cards equally illuminated, electric box of large opening made the darker by over-
head cardboard:
Da caddie aside Bah RO ABR estas Ne een Mar. 25 10 0

factorily and within a reasonable time limit, so the method
seemed available for further use. Five control series were given
to check out the influence of brightness. It is manifest that
with equal illumination on the faces of the display-cards, which
was roughly secured by adjustment of the shades of the two
windows that admitted light to the room, the electric box back

INSTINCTS AND HABITS IN CHICKS 59

of the large opening would be brighter than that back of the
small one. The angle at which the cards stood to each other,
together with the position of the windows with reference to the
cards, made it possible to vary the brightness of the two cards
independently, and that of the electric boxes so that the electric
box on the side of the brighter card could be made also the
brighter box. A cardboard over an electric box was also used
to screen off light. Within the degrees of variation employed,
the chicks continued to select the smaller opening regardless
of the relative brightness of the faces of the cards or the electric
boxes. When both electric chambers were covered with a ground
glass plate and a 2 c.p. electric light was adjusted over the
electric box on the side of the small opening, no. 24 approached
as usual the small opening, hesitated a moment, glanced about
the apparatus in the vicinity of the card, and then slowly en-
tered. (See control test 3, table 19.) Control 5 exhibits the
same general relations as 3, but without the use of electric light.
The pronounced variation, even to a different quality of light,
did not inhibit the selective reaction. In control series 4 the
electric box on the side of the card having the large opening
was darkened considerably by cutting off the light from above
with a piece of black cardboard. The figures for this series of
reactions, Io—-o, in comparison with the 9-1 records in the other
series, represent but do not express fully the difference in readi-
ness with which the chick responded. This series repeated
control series 1 with greater difference in the relative brightness
of the electric boxes. The conclusion seems to be justified,
on the basis of these results, that the chicks were responding
selectively to one of two objects of different size. Of course
the control of other possible determinants of reaction, such as
odor, would have made this conclusion more certain. The
small-large habit was now adopted to be used in conjunction
with the black-blue habit for an experimental test of the effects
and interrelation of habits.

D. To What Extent Has Training General Value?

After an experimental test of the influence of one labyrinth
habit upon the formation of another, Yerkes‘! concludes that
“the acquisition of one form of labyrinth habit may facilitate

1 Yerkes, R. M.: Op. cit., p. 261.

60 FREDERICK S. BREED

the acquisition of others.’”’ ‘‘ For the student of animal behavior,
as for the human educator, it is of importance to learn,” he
writes, ‘‘ whether one kind of training increases the efficiency
of similar forms of training.’”’ The question for which an answer
was sought in the work about to be reported is closely related
to the problem which Yerkes set for himself, but is concerned
with the more general value of training: How does acquired
ability to react perfectly to one kind of element or complex
affect the acquisition of ability to react to a quite different
element or complex?

Among certain organisms a particular act once acquired is
more easily re-acquired. A particular act once acquired may
facilitate the acquisition of a similar act, similar meaning here
a relation of partial identity. Now the question naturally
arises, As two acts are less and less similar, what effect does the
acquisition of one have on the later acquisition of the other?
This raises the problem of formal discipline, capable, surely,
of solution by experiment, but not yet solved. It has been
suggested that there may be “two kinds or aspects of organic
modification in connection with training; those which constitute
the basis of a definite form of motor activity, and those which
constitute the bases or dispositions for the acquirement of
certain types of behavior.’”’! The data upon which this sugges-
tion is based would allow us to say nothing definite about “ the
acquirement of certain types of behavior” not of the original
labyrinth form, which Yerkes used. But at that the general
value of training has been experimentally demonstrated, in so
far as different labyrinths are different objects, even different
objects of the same kind. Although there is a degree of generality
here, it is so slight that the training value involved would no
doubt ordinarily be classified as specific. If no two particular
stimuli are identical, then, practically speaking, much specific
training has general value, which does not mean, however, that
this generality of value may not rest at bottom on specific
modifications. This must be our conclusion when we look at
the matter from the side of content as opposed to function,
stimulus as opposed to reaction. Practically, it cannot be denied
that in some organisms certain kinds of training do have value
for certain other kinds. The question awaiting answer in the

1Yerkes, R. M.: Op. cit., p. 261.

INSTINCTS AND HABITS IN CHICKS 61

present state of our knowledge is, Between what kinds of modifi-
actions and to what degree is there transfer? Few will dispute
the assertion that there may be dispositions preceding and under-
lying the acquisition of a particular act which are not identical
with those that are at the basis of its actual performance. In-
deed the former may no doubt be congenitally present, with or
without the actual ability to perform. In a similar manner, of
course, one may think of an acquired neurological disposition
“ predisposing ’’ the individual to certain new activities. And
within what range? That is our problem.

To bring a particular case to a test, two comparable groups
of four chicks each were chosen, Group I consisting of chicks
nos. 32, 33, 38, and 39, Group II of chicks nos. 34, 35, 36, and
40. All the chicks were of the same brood, hatched April 13,
1908. None of them had any other training prior to or during
the course of these experiments. Uniformity in method was
strictly adhered to. The chicks of Group I were given the
black-blue training to the point where they made no errors for
three successive days, ten trials per day. Then the chicks of
both grours entered together into the size training. During the
course of the development of each habit one setting of the cards
was used for the corresponding trial of all the animals. For
example, in the black-blue work, the black card occupied the
positions r) 1, r, 1, r) 7 1. 1, 7 Lm each series until an. erroriess
series was reached. At the first setting of the cards—black
right, blue left—the chicks of Group I went through one at a
time until each animal had to its credit one reaction. Then
the cards were exchanged to the second position and the chicks
again taken as before.

As a precaution which seemed advisable, the secondary coil
of the inductorium was adjusted at 6 to begin with in the black-
blue training, and it was continued thus throughout. In the
later size training a shock of the same intensity was employed
at the start and at the beginning of the ninth series this was
raised to 5.5 on all the chicks alike.

The order of training was this. When the chicks of Group
I were twelve days old, work was begun with them, and they
were trained until they responded errorlessly with a positive
reaction to black. On May 11, by which time this group had
finished the color work, these animals with those of Group II

62 FREDERICK 8. BREED

were introduced into the size training. Table 20 shows the
course of modification of response to the color stimuli. Only
the errors appear in this summary.

TABLE 20
Errors in Biack-Biug REACTIONS
Cuicks Nos. 32, 33, 38, anp 39. Harcuep 4/13, ’08

Group Digsac oe gas She es een Set 32 33 38 39
Sex, F. Sex,M. Sex, M. Sex, F.

Series * Date WwW WwW Ww Ww
ipisgacth achg-. te Birth AE NAY Bigg valde seein BRRALY Apr. 25 5 5 5 5
Do murs, wie devel wire mien oh aeeh “26 6 7 7 7
Ue i mais NS orien mat Oy in aa Wale tal MT: a 3 5 4
Biton it as FORE RS HEGRE EASE «28 1 3 38 2
A rete Or ndenn Sind ute GOA Se Ur tiesed odds eae 29 0 3 2: 4
Dh hhc Ramen neh APLAR ba Seip bc diel Feltoe «30 0 2 1 2
Oise debits RAUB Se Meh ees a 8 May i 0 1 2 0
Theta es ies ers eet tegisauesnpaae cera cae shal eraeh nd “ 2 1 1 1
So es raaetdicel SS Alaeiina eh gee eRe ed e 3 0 1 0
han aces at Deeb ahve enig td ake ern ees s 4 0 0 0
TQ ecetnig ancien have mene mires relat tae e 5 0 3 0
DM caged ala 263 ono las ts aie aR adi led it “ 6 0
Dee es 2k a ee ee heed arn 7 a 0
DE retcitiis Bad ye BAe ese Poe Se “ 8 0

These chicks really do not vary so much in the modifiability
of their behavior as the figures might indicate. No. 38 became
excited during the tenth series, which may explain its lapse.
The rapidity with which no. 32 finished the work is noticeable.
No chick before or afterwards in my experiments excelled this
record of 30 trials exclusive of preference and perfect series.
But it is sufficient for our present purpose simply to show clearly
the perfection of the training that preceded the formation of
the second habit, without extended discussion of the other
characteristics of the modification.

Following the completion of this training of Group I, both
groups were introduced into a size experiment, as described
above. Table 2: reveals the progressive development of the
second modification for each group in the column of average
daily number of errors. The total of these averages for Group
I amounts to 66.3 for the sixteen days training; for Group II,
70.1 during the same period. Curves I and II, fig. 10, exhibit
graphically the rate of decrease of error for the corresponding
groups. The record of each group in the preference tests is

_ INSTINCTS AND HABITS IN CHICKS 63

included in the curves. In a comparison of results for the two
groups two things especially must be taken into consideration,
(1) the natural preferences before training began and (2) pre-
vious training. In regard to the first, Group II had to overcome
a stronger preference for the larger of the two openings. The
chicks of Group I took the smaller opening more often than
those of Group I, perhaps because in general the black-blue

TABLE 21
Recorp or Errors
REACTIONS To Size: SmMaut-LarcE OPENINGS
HartcuHep 4/13, ’08

Group I Group II
Series Date 32 33 38 39 Av 34 35 386 40 = Av.
Bip tescerieiiasised ota May 11 6 7 5 38 5.3 5 5 7 8 6.3
1 “12 8 5 9 FTF 7.8 8 9 5 9 7.8
2 “3 6 7 6 7 6.5 6 9 10 5 7.5
Biv wak gua eee “14 6 5 6 8 6.3 6 7 5 7 6.3
1 eee eee 5 7 7 4 7 6.8 ob fF £6,
5 “16 5 fb 6 FT Bed 7 8 6 7 7%.
Gievcneire neha “18 5 4 6 3 4.5 6 6 4+ 6 5.5
i erence ee “19 7 7 2 J 4.8 3 4 3 6 4.
Sh conciaeniw srs “20 5 8 2 8 4.5 2 6 1 56 38.5
9 21 3 6 4 2 3:5 5 4 1 8 4.5
10 «22 4 7 38 2 4. 3 3 3 6 3.8
ED cath ti bold "23 6 38 2 2 23.3 12 2 6 2.8
dD ere ene «25 4 6 2 4 4. 2 3 0 4 2:3
tO ea «26 1 1 1 1 ifs 1 2 4 5 8.
TAs ees sie DE: 1 2 1 2 1.5 Ty 2 3: 4 DS
1D: es eyecuaces «28 0 1 4 1 1.5 1 1 1 4 1.8
WGzseccinare: 29 2 EF 8 DL. 228 EO) ie Bb 18
Total av...... 66.3 Total av...... 70.1

training inclined chicks to accept the stimulus of lower bright-
ness value, which in this case has been shown to be on the side
of the smaller opening. The previous brightness training here
came into conflict, however, with the natural(?) tendency to
pass out by the larger opening, which may be something more
than a brightness preference, and thus may be explained the
small difference between the two groups in the results of the
preference trials. As to the second point above, Group I, besides
acquiring the black-blue habit, had become habituated to the
general conditions of the apparatus in such a way that adjust-
ment to the new conditions of the size experiment must have

64 FREDERICK 8. BREED

been somewhat facilitated—through a lessening of the tendency
to general excitement during the first tests, if in no other way.
In view of these considerations there is no conclusive evidence
that the previous training of Group I with black-blue facilitated
the formation of the small-large habit, albeit Group I averaged
5.4% less errors per day than Group II.

It should be stated in connection with the above size work
that the results indicate difficulty of acquisition. The behavior
of the chicks pointed in the same direction. So, any more general
conclusion than we have ventured would have to be made sub-

Lrrors
10

ai

3 . X ~
Se a
a ~N
Sel

Series A 1 2 3 4 & @ ¥ 8 8 W it &@ 8B we &

Figure 10—Rate of modification in size training, arranged to test the effect of
previous color training. Curve I shows the rate of acquisition by Group I of
the small-large habit after having acquired the black-blue habit; Curve II
represents the rate of acquisition of the same habit by Group II, the same
number and the same age as the chicks in Group I, but without previous
color training. Preference tests on first day. Distances along axis of abscissae
represent series; along axis of ordinates, average number of errors.

ject to this further fact. It ought also to be pointed out that
chicks of the same age are not always of the same size. The
members within each group were of unequal size when six weeks
old. Yet, fortunately for our experiment, the groups remained
comparable as to size, individual for individual, throughout the
tests. This matter is vital to the experiment, for naturally
chicks of different sizes could pass through the small opening
with different degrees of ease.

INSTINCTS AND HABITS IN CHICKS 65

But the conclusion as to the general efficiency of the black-
blue training should not rest alone upon a comparison of the
rates of modification in the two groups of chicks. Having
learned in the course of the experimentation with the black-
blue habit, from data which will appear later, that this modifi-
cation persists for some time with relatively slight change, it
seemed of value to investigate the relation of the size training
to the color training by means of the so-called memory test.

TABLE 22

THE NON-INTERFERENCE OF MopIFICATIONS

Persistence of

Training in Training in the Black-Blue
Black-Blue Small-Large Modification
First
Inter- selec-

No. Age Sex Trials Errors} Age Trials Errors| val R W tion

32 12 F 30 10 28 160 70 32 10 0 Black
M 70 20 28 160 74 30 10. O Black
38 12 M 100 25 28 160 62 31 4 6 Blue
F 70 20 28 160 58 30 9 1 Blue

Table 22 exhibits the results of this test. The table shows the
age of the chicks of Group I when the work on black-blue was
begun, 12 days; the number of trials necessary to complete the
training, exclusive of preference and final perfect series; the
number of errors made in the course of this training; and also
the corresponding age, trials and errors for the size training.
The interval, thirty days or more in each case, is the number of
days between the date when the black-blue modification was
pronounced perfect and the date upon which the memory test
was given. Of course sixteen days of this interval were occupied
with the development of the small-large habit.

The reactions of the chicks to black-blue in the persistence
tests are detailed in the columns R and W. Chicks nos. 32
and 33 made perfect records, no. 39 made an error on its first
trial, being thereafter perfect, while the reactions of no. 38
seemed to show no trace of the original modification. The
results altogether weigh in favor of the view that the modifi-
cations with which we have dealt proceed quite independently

66 FREDERICK 8. BREED

of each other. By consulting table 26 it will be seen that the
persistence of the black-blue habit in these chicks after 30 days
measures fully up to the average made by other chicks that had
not had intervening training. If it be objected at this point
that one should expect the small-large training to exert an
influence in the direction of strengthening the black-blue modifi-
cation, I agree, in so far as brightness is concerned, but the
inference is that something more than brightness was involved
in each habit. That color and not brightness was the effective
element in these tests of persistence I have not shown. The
results are offered in their present stage for their possible value
as a suggestion for method. It is not too much to expect relative
permanency of the black-blue habit even if the brightness rela-
tions of the first training were reversed in the second. Though
both forms of training are habits of response to optical stimuli,
they are probably quite separate from each other in their develop-
ment. Physiologically, this likely means the formation of
separate neural bases. For our explanations we look to the central
nervous system. Judd! observes that ‘‘ Discrimination is not a
process of impression.”’ ‘‘ The raw materials for adaptation, the
impressions, are there very early, waiting for the individual
gradually to adjust himself to them.’ Discrimination depends
upon the ability of the individual to react specifically. Capacity
to discriminate is linked with the capacity to react. If this
be true, the animal’s scope of reaction will determine the limits
of its modifiability. According to this view, the grouping of
elements to form a single object of reference in reaction derives
its unity from the co-ordination in a single active brain process.
This does not ignore the fact that the structure of the end organs
will also define in advance the range of objects that may be dis-
criminated. Impressions to be distinct must be accompanied by
distinctive neurological dispositions. ‘‘In short,” says Judd
again, “so far as this raw material for development which
presents itself in the form of impressions is concerned, it must
all be worked over and connected with individual reactions
before it can be regarded as really assimilated by the develop-
ing individual.”
But let us return from theory to facts.

1 Judd, C. H.: Genetic psychology for teachers. New York, 1903, p. 153 ff.

INSTINCTS AND HABITS IN CHICKS 67

E. Some Effects of Modifications

In table 12 were pointed out the natural color preferences
of chick no. 8: white in preference to black, blue to orange,
blue to yellow, and yellow to black. The peculiarity of the reac-
tions to blue was mentioned. If blue be supposed to have a
higher brightness value for chicks than for humans, the chick’s
behavior might be satisfactorily explained on the ground only
of brightness of stimulus. Further experimentation was under-
taken with a view to an explanation of the above apparent
irregularity. No. 11, a chick which had been trained on black-
white and had progressed so far that in its last series of twenty
successive trials it made but one error, was tested for its pref-
erence on the combination of colors listed in table 23. Eight
days had elapsed between the date of the 19-1 series and these
color preference tests.

TABLE 23
PREFERENCE Tests Arrer Buack-WHITE TRAINING

Cuick no. 11. Harcuep 12/10, 07. Sex, F.

Series Date Right Wrong

Black White

Jan, 21 19 1
Orange White

Pra si Boh Dao eR Behe Ge A ee Sts Jan. 29 8 2
Black Blue

Bente cea ctg ee epee se eke SHA Uae Ra oe ee Jan. 29 10 0
Blue White

ig Padas raaihece Sawa ted ties hiaw aged oes aber bab ge sae Jan. 29 6 4
is Yellow White

YS A hed sic teh a st Regt oe Sere ha or etna aE gs Jan. 29 8 2

This record reports the tests in the order in which they were
given, as do other records. Again the reactions to blue seemed
somewhat surprising. That blue should be rejected as freely
as white, when appearing with black, and that blue and white
should be accepted almost indifferently, are not results that one
should expect. These facts seem to confirm the hypothesis that
blue has a special brightness value for the chick. 7

Further investigation followed with six other chicks, inci-
dental to the direct line of investigation. Table 24 shows the
first preference reactions of chicks nos. 17, 18, 20, 32, 33, and

68 FREDERICK 8. BREED

39 to black-blue. The preference tests are not representative
of our chicks in general, for those chicks that already had a
preference for black were rejected because they were not suitable
for the special purposes of the experiment of which this black-
blue training formed a part. There were a few of these. But
this does not affect in the least the point that I wish to make
in connection with the effects of training. The state in which
the black-blue modification persisted at the time of the later
color preference tests is shown in the columns under “ Per-

sistence.”” The tests on blue-white, black-white, blue-vellow,
TABLE 24
Some Errects or TRAINING
Prefer- Persist-
ence ence Preference after Training
a Sy
rs g z
iS =I 3
5 op = ~ =/38 a
a 3 a ais]. 3
s|_| 2 lalolElelelzlelglaiglelBlelclzlel:
Zia oa) AlMIiA/AlA;a;A MIF Ola le aia lala
17 | 12|Black- |
Blue} 5} 5] 30) 10; 0; 2; 4; 6) 10; O} 3] 7.
18 | 12 ee 3) 7} 30} 8} 2) OF 3) 7 8} 2) 2 8
20 | 12 ne 5} 5) 30] 10) O;} 2; 3) 7) 9} 1 1 9
32 | 12 . 5] 5] 32) 10} Of O} O; 10} 3] 7 O;| 10 «38; TY 9] I) 9
33 | 12 a 5) 5] 30} 10} O; O; O] 10; 6 4) O| 10 1) OF 10) 5&5 5
39 | 12 € 5} 5} 30) 9} 1] OO} 2) 8 S&F 5} 1 9 Ly 5) 5
49 | 39/None 1) 9 1) 10} 0
51 | 39 « 1] 9 | 1] 7 3

blue (tint no. 1) -yellow, blue-blue (tint no. 1) were given in
the order named and, in all but two cases, immediately fol-
lowing the persistence tests. The most striking feature of these
results is again the character of the reactions to blue. If bright-
ness were the only factor, a chick trained to black-blue would
be expected thereafter to react positively to the less bright of
two stimuli, other things being equal, when a different combina-
tion were presented. But chicks that one minute accepted
black and rejected blue, that is, selected the card of lower bright-
ness value, the next minute accepted white and rejected blue.
How shall we explain this behavior if we consider merely the

INSTINCTS AND HABITS IN CHICKS 69

brightness values of the stimuli? After black-blue training the
ehicks quite commonly rejected this blue no matter with what
other color it was in combination. It may be suggested that
after the long period of training the chicks respond to a par-
ticular brightness value, the blue amounting to a certain degree
of gray. But no. 32 and no. 33 rejected blue (tint no. 1) when
it was used in combination with the much brighter yellow.

The method of training with the electric shock frequently
makes the stimulus to which it is attached the emphatic one.
In a word, the chicks thus trained may not be guided in their
reactions by the color through which they escape, but by the
color in connection with which they are shocked. This reaction
is one primarily of rejecting blue, rather than of accepting black.
It is indicated by such reactions as those of nos. 32, 33, and 39
to black-white after having been trained to black-blue. When
white replaced the blue of the original training combination,
the stimuli became almost indifferent to the chicks. Again, the
same chicks seemed indifferent to the color presented with blue.
When required to react to blue-white there was no hesitancy,
the response being one of getting-away-from-the-blue-no-mat-
ter-what-the-other-color. For data see table 24.

All the chicks did not behave so, however. A comparison of
the corresponding reactions of nos. 17, 18, and 20 is instructive.
These animals, when white replaced blue as above, responded
definitely and positively to black, and were quite lost when
white accompanied blue. With these chicks black seemed to
be the guiding stimulus.

It is sometimes said that negation is affirmation, but is this
true in any proper psychological sense? Naturally, avoidance
of one object means the acceptance of some other object—in
the sense that leaving one place means going somewhere else.
But does the ‘‘ somewhere else ’’ need to be defined as a con-
stituent element of the avoiding reaction? Or, upon the occur-
rence of a positive response, is it necessary to consider the point
of departure, which is objectively definable, of course, as a part
of the process? The point of departure in the latter or of arrival
in the former case does not appear to be essential.

Passing now from the standpoint of stimulus to that of reac-
tion, why should a positive response be regarded as a simul-
taneous negative response? As antagonists may they not mutu-

70 FREDERICK 8. BREED

ally exclude each other? But only in so far as a positive reaction
essentially includes a negative one and vice versa, 1s:affirmation
psychologically negation.

There still remains another aspect of the relation between
rejecting and accepting, their possible necessary concomitance.
But if, with innervation of a given motor center, there is, accord-
ing to the Muensterbergian action theory, a simultaneous inhibi-
tion of the antagonistic motor center, we are bound to think
of real affirmation and real negation as distinct functions, for
inhibition is not negation.

What, then, shall we say of our method as a “ discrimina-
tion method?’”’ From the point of view of behavior discrimi-
nation is selective reaction. In this sense the Yerkes method
is a discrimination method.

It is interesting to observe the result when ‘neither black nor
blue are in the test combination presented after the training.
No. 44 on May 22 completed the black-blue habit with three
errorless series. On June 6 its persistence test was black, 10;
blue, o. The same day its preference tests on orange-white
resulted 5~5. The daily training on orange-white, continued from
this date, progressed as follows: 7-3, 8-2, 7-3, I0-0, I0-0,
to-o. No. 43, the same age, began black-blue training on the
same date as no. 44 and completed the work the day after no.
44 with three perfect series. Its persistence test on June 6
resulted also 10-0. But when tested on the same day for its
orange-white preference prior to the training on this combina-
tion, its record was orange, 10, white, o. The figures thereafter
were 9Q-I, 9-I, 9-I, 10-0, 10-0, 10-0. It is apparent in both
these cases that the black-blue training strikingly predisposed
the chicks to react positively to the darker of the two stimuli
in the second combination. It cannot be urged in regard to no.
44 that, either on account of ready modifiability or an easy
combination, this was simply a case of rapid learning unassisted
by the previous acquisition, first, because orange-white, as indi-
cated by earlier tests, is not so easy a combination; and secondly;
the rate of acquisition by this chick was only average, nine
days’ training having been required for it to complete the black-
blue habit.

The important exception to the ase rule for brightness
training, namely, that when the color which: the chick was

INSTINCTS AND HABITS IN CHICKS 71

trained to avoid was presented in combination with a new
color, the chick continued its specific avoiding reaction regard-
less of the relative brightness values of the stimuli, demands,
I believe, that color quality as well as intensity be assumed as
a determining factor in the reactions of these chicks.

F. The Persistence of Modifications

Hints of the material about to be presented have necessarily
been dropped in the discussion of previous topics. Yerkes '
publishes in his work on the dancing mouse a valuable table
of results on “‘ Measurements of the Duration of a Habit.”
Commenting on this data, he says, ‘‘It is safe, therefore, to
conclude from the results which have been obtained that a
white-black or black-white discrimination habit may persist
during an interval of from two to eight weeks of disuse, but that
such a habit is seldom perfect after more than four weeks.’
If one should assume that the chicks depended upon the element
of brightness only, the results obtained for them in our black-
blue tests would be comparable with the black-white and white-
black records of the mice, provided one allowed for age differ-
ences, as well as for the fact that the discrimination, assuming
the human standard, was made more difficult for the chicks on
account of the smaller difference in the brightness of the two
stimuli.

The chicks in these tests (see table 25) were placed in the
entrance box in the manner followed in the original experiments
and allowed to escape from the apparatus to the cage. The
cards were exchanged after each trial, that is, regularly alter-
nated right and left. No shock was given and both exits remained
open throughout. There was a noticeable difference in the
behavior of the chicks in the apparatus after the interval.
Instead of smoothly going to work they usually hesitated a
few moments and craned their necks about before starting for
an opening. The same kind of phenomenon was noticeable,
ordinarily, in the behavior of a previously trained chick when
lifted in the hand after an interval of some weeks during which
it had not been handled. Whereas the animal had come to
submit freely and without struggle in the course of the original
training, now, when grasped over the back, it sometimes kicked,

1 Yerkes, R. M.: Op. cit., p. 256.

72 FREDERICK 8. BREED

chirped, or attempted to make flapping movements with the
wings. For the tests of persistence, excitement of this kind
was avoided as much as possible.

In estimating the value of the results printed in table 25,
it must be remembered that chicks nos. 32, 33, 38, and 39 had
16 days’ size training in the interval. Also, nos. 43 and 44 had
orange-white training to perfection between the first and the
second persistence tests.

TABLE 25
THE PERSISTENCE OF HaBiT
Days’ First
Chick Sex Training interval R W selection

UG) Sock aes Same e Female.......... Black-Blue...... 30 10 0
Tatas aoe ties MO poup sre leaban a gab aah 30 10 0
[Geass eityewe Oe Sykes aapredies Be esha 30 8 2, Black
TQ oes ee sone SO eis emg Se 2 Sy gw ages 30 7 3 e
20 bce ne tess hes Male............ ee re eer 30 10 0
DA wept da hes Gia Biss Sha AI ee Small-Large..... 19 6 4 Large
Aa Sie aae Female.......... We * Tee ead 17 6 4 Small
oie inane AE ep Me aneiaeagees Black-Blue...... 32, 10 0
PE eo hearers et ee eee irre 55 8 2 Black
BOs eq aseeee Males oo sccaccniacs Ebel 30 10 0
ee ee ee ee Bod. (5 Sea eas 51 9 1 ~
Of rales HE be aun waseia nba § Small-Large..... 71 9 1 Small
S8e2s uae Gkw es aaa nea gais s Black-Blue...... 31 4 6 Blue
Mog Pewee mace ee eee Mee eeest 48. 5 5 Black
B9o cere eee Female.......... MO a ag 30 9 1 Blue
ca a eee ee toteaet chase pee sioaruishe 14 10 0
fy poids deNe Gat feane Bet ASB dread iat et arastets) 21 10 0
a ite giaaAs a cn BEY sh detiriel deh Ss gbeirieeig 77 10 0
BAe cst he ctipte ais Maléy sce ea a 15 10 0
ere een ee Pe Sashes Mee aot eo slate Asked 22 10 0

Of eight mice which Yerkes trained in black-white or white-
black, only two showed perfect persistence of the habit after an
interval of four weeks. Of nine chicks trained in black-blue and
tested in a similar way after 30 days, five obtained perfect records.

The behavior of no. 34 in the reaction box equipped with the
size cards 71 days after the completion of the original training
furnished a nice commentary on the ‘‘intelligence”’ of the
animal. This chick, standing naturally, was 19 cm. high, meas-
ured to the highest point of its back. So large that the larger
of the two openings would permit its passing out with difficulty,
it crouched at the smaller opening and attempted to go through
it at every trial but the last. Each time, as soon as it got its
head through the small opening and was just beginning to

INSTINCTS AND HABITS IN CHICKS 73

struggle, it was brought back to the entrance box for the next
trial. The record of its behavior found in table 25 is made sub-
ject to the irregularity reported below. The result of this
persistence test is the same as that in its last training series.
It would be difficult to reach any trustworthy conclusion in
regard to the persistence of the small-large habit from the
data at hand, for these data have to be considered together
with a certain defect in the adjustment of the apparatus. When
the persistence tests of nos. 24 and 25 were made, these chicks
were so large that they had to squeeze through the smaller
Opening. For the test of no. 34 the small opening was made
slightly larger and a dimension of the larger opening altered
to preserve similarity of shape.

TABLE 26
Tue Rapipiry or MopIFICATION AND ITS PERMANENCE

Black-Blue Training

Persistence

Chick Age Sex Trials Errors Interval R WwW
VG isd scire duets diggaba cies 12 F 100 28 30 10 0
Eisen seteeds fone onniee 1 12 F 60 20 30 10 0
ED Ses tea cata ace penta, ee 12 F 100 35 30 8 2
NG cca ih atrnarern wagidianses 12 F 60 28 30 7 3
QO i vies sragic tae noe 12 M 60 19 30 10 0
Bo ahh opus Se ane eS 12 F° 30 10 32 10 0
Bo acces Gynec Noone ae 12 M 70 20 30 10 0
Bios cAaiaataseiameaontas 12 M 100 25 31 4 6
DOE iat 8 stort letpostaecane 12 F 70 20 30 9 1

Index of modifiability. . 2.0.0.0... eee eee ee 72.2

For the assistance it may be to others in estimating the value
of these tests of the persistence of modifications, it should be
stated that some of the chicks were in poor physical condition
when the tests were made. In the confinement of the laboratory
few chicks remained in normal physical condition after eight to
twelve weeks.

It is interesting to study the above data on the persistence of
habit in connection with the number of trials and errors in the
original training series to see if any correlation exists between
the rapidity of acquisition and its permanence or retention. In
table 26 the necessary items are set side by side. One often
hears the pedagogical byword, ‘Come easy, go easy.”’ This

74 FREDERICK 8. BREED

certainly does not apply to the chicks tested. Of three animals
that required 100 trials to perfect the black-blue modification,
one made a perfect persistence test; of two that required 70
trials, one made a perfect record; of three that required 60 trials,
two made perfect tests; and the chick that completed its work
in the shortest time also made a perfect record. The number
of trials given in the table is exclusive of preference and final
perfect series. In the error column is given for each chick the
total number of errors in this number of trials. Of five chicks
that made twenty or less errors, four had perfect persistence

Successes
10

P Va
; yal

SeriesA 1 2 3 64 5 6 7 8 9 10 11 12 13 14 15 16

Figure 11—The success curve for nine chicks of the same age in black-blue training,
plotted from averages based on data given in tables 13 and 20. Distances
along axis of abscissae represent series; along axis of ordinates, right or suc-
cessful trials.

tests; of four that made more than twenty errors, one had a
perfect persistence test.

One may well question the appropriateness of the expression
“trial and error’ intended to be descriptive of this method of
learning. If the number of errors is included in the number of:
trials, as for example in this paper, then the original term is
awkward if not illogical. These animals learn through successes
as well as failures, and the sum of the number of both equals
the whole number of trials. In this sense the method is more
accurately described as a method of success and failure. Again,

INSTINCTS AND HABITS IN CHICKS 75

if the word “‘ trial” in the above expression is meant to refer to
the principle of random activity, and “‘ error’ to learning based
on the avoidance of harmful stimuli, still ‘trial and error” is
an inadequate descriptive term. One could as well say the
general law is, “‘ Prove all things; hold fast that which is good.”
If the term selection be construed broadly enough to include
both the positive and negative aspects, then trial and selection
appears on the whole to be a more satisfactory name for this
type of modification.

If the “index of modifiability,” to borrow Yerkes’ term and
definition, be considered ‘‘ that number of tests after which no
errors occur for at least thirty tests,’’ the index of modifiability
of these Barred Plymouth Rock chicks, under the conditions
described, is 72.2.

The rate of modification for the same nine chicks is set forth
graphically in fig. 11. This is a success curve plotted from
data in tables 13 and 20. The averages upon which the curve
is based, including the preference test, are as follows: 4.7, 3.3,
5.2, 6.1, 7.8, 7.9, 8.8, 8.9, 9.8, 10, 9.4, 10, 10, 10. It is quite
possible that valuable accretions to our knowledge of habit
and instinct may come through a careful study of the course
of development of reactions of both types and a comparison
of-the quantitative relations involved. I offer the curve of
the development of the complete pecking co-ordination (fig. 7)
and that of the development of the black-blue modification
(fig. 11) as suggestions for the beginning of a more specialized
investigation.

SUMMARY AND CONCLUSIONS.

The early post-embryonic life of the chicks continued the
scope of activities already begun in the egg. The alternations
of passivity and activity, the lifting movements of the head
combined with stretching movements of the legs, the occasional
reflex forward thrust of the bill followed usually by movements
of the mandibles, loud chirping along with other violent activity,
—all these were common aspects of the behavior of chicks just
before -as well as immediately after hatching. The pecking
reaction might have assisted in the process of exclusion, but
by far the most common reaction while the chick was struggling
in the egg was the lifting movement of the head and bill. This
reaction actually broke shell and tore confining membranes.

76 FREDERICK 8. BREED

The chicks, left to develop naturally in the vicinity of food
and water, usually found the water by fortuitous pecking or
by performing the drinking movement in imitation of other
chicks. By imitation here I mean that the performance of
drinking by one chick in the presence of another sometimes
stimulated that other itself to perform the drinking act. I
reach this conclusion on the basis of incidental observations in
the course of the study of other problems. I have no doubt
that a special study of this point will bear out the above
conclusion. Chicks kept without drink until the third day
did not perform ‘the drinking reaction by accident or imita-
tion only. The reaction was elicited by a variety of objects
before drink of any kind had touched the bill of the chicks,
showing that, when the need of the organism became sufficiently
imperative, the drinking reaction appeared in response to many
of the same kind of stimuli as the unpracticed pecking reaction.
The drinking instinct does not ‘‘ have to be supplemented by
imitation, accident, intelligence, instruction, etc., in order to
act.”” And it is not improbable that the ‘sight of still water ”’
will be found to evoke this reaction,

The few tests that were made to determine the effect on the
pecking reaction of disuse during the first two days of the life
of the chick seemed to show (1) that the development of the
instinct was retarded by disuse, and (2) that the retardation was
quickly overcome with use.

The stimulus for the pecking reaction did not have to be
some object of a size convenient for eating. The bill of the
chick was used as a testing organ,

Pecking situations were found in which the reaction to one
stimulus exerted an inhibitory effect on the reaction to another.

The assumption that chicks do not peck at or eat food when
they cannot see it is not supported by the experimental data
submitted in this paper. The chicks did peck at and eat food
when the light was excluded.

The pecking instinct was investigated primarily to discover
the meaning that should attach to such terms as perfection,
accuracy, or congenital definiteness, when applied to this re-
sponse. Pecking is a co-ordination of three reactions,—striking,
seizing, and swallowing. The amount of improvement from day
to day of the complete co-ordination, not only, but of the com-

INSTINCTS AND HABITS IN CHICKS 77

ponents separately, was measured. Records were taken up to
the twenty-fifth day. The pecking improved in accuracy very
rapidly during the first two days, reaching by the beginning of
the third day an efficiency numerically represented by 29.29
on a scale of 50; by the beginning of the eleventh day an effi-
ciency of 40.10 was attained; and during the rest of the period
of measurement the degree of accuracy ran no higher than
42.57. The improvement was retarded more by errors in seiz-
ing than by errors in swallowing, and more by errors in swal-
lowing than by errors in striking. After the third day the-
imperfection of seizure remained greater than that of the other
two reactions combined. The striking reaction, seldom widely
erroneous, improved rapidly and without relapse, ‘closely approx-
imating perfection by the fifth day,—a degree of accuracy that
might easily have inspired belief in the perfection of the peck-
ing instinct.

It did not appear that the effect of social influence was such
as to increase the rate of improvement in accuracy of the peck-
ing reaction. It may be that the change in the mode of func-
tioning of the organism due to the presence of others was in
the direction of increasing the intensity and rapidity of the re-
actions, without increasing their rate of improvement in accuracy.
Such a variation would have selective value, inasmuch as the
animal would get more food in a given time, even though the
pecking were no more accurate.

The important exception to the general rule for brightness
training, namely, that when the color which the chick was
trained to avoid was presented in combination with a color
other than that used in the original training, the chick con-
tinued its specific avoiding reaction regardless of the relative
brightness values of the stimuli, demands, I believe, that color :
quality as well as intensity be assumed as a determining factor
in the reactions of the chicks.

The chicks, without much doubt, responded selectively to
one of two objects of different size.

The results of the form tests reported were purely nega-
tive.

There was no conclusive evidence that previous formation of
the black-blue habit facilitated the formation of the small-large

habit.

78 FREDERICK 8. BREED

The tests of retention of the black-blue habit after size train-
ing, which intervened between the original training and the
retention tests, at least suggest a method that may be useful
in studies of the interrelation of habits.

The method of training with the electric shock frequently
made the stimulus to which it was attached the emphatic one.
In such a case the chicks were not guided in their reactions
by the color through which they escaped, but by the color in
connection with which they were shocked. That is, the reaction
was negative to blue, not positive to black, when this occurred
in black-blue training. Psychologically, it seems, negation is
not affirmation.

Of nine chicks perfectly trained in black-blue, five made
perfect persistence tests after an interval of thirty days.

In the case of the same tests of retention, rapidity of modifi-
cation was positively correlated with permanence of modification.

For these nine chicks the “index of modifiability ” was 72.2.

“Trial and error’? is an unsatisfactory name for this method
of learning.

Behavior Monographs

Volume 1

1, The development of certain instincts and habits in chicks, By
Frederick $. Breed. Pp, iv+78. $1.00

In Preparation

Methods of studying’ vision in animals, as Robert M. Yerkes and Jolin
B. Watson.

An experimental study on the death- feigning of Belostoma (-zaitha aucct.)
flumineum say and nepa apiculata’ uhler. By Henry H. P. Severin and
Harry C. Severin. ,

The Biology of Physa. By Jean Dawson.