II. PROBLEMS AND METHODS

My study of the pigeon's emotional life had for its object certain respiratory "expressions." These were investigated by means of a pneumographic tracing, secured while the animal was comfortably fastened in a shallow nest, partially open below. A small box was placed over the bird, and apparatus was so arranged that the time of giving various stimuli was recorded automatically on the smoked paper of the kymograph drum, below the breathing-curve. A third line indicated rate of drum movement. Although some interesting results were obtained, the chief significance of the research consists in its demonstration of the fact that this method of studying animal mind is valuable.

In the study of association I sought to determine the sense-data which the process involves, its method of formation (with due regard to social conditions), its rapidity, permanence, and modifiability, and also its probable degree of complexity. Material contributing to the subject was secured by observing the behavior of the animal when seeking to obtain food by overcoming such obstacles as labyrinths with wire passages, and latches, when the food was left in view, or by finding it when out of sight. In the latter case it was placed in a box occupying a customary place in a group of exactly similar boxes, or else in a box of color or form unlike the other members of the group and variously arranged, from time to time, with respect to them. When the animals were learning the labyrinth habits, various stimulations were given them; later the character of some of these was altered, and the resulting changes in behavior were noted. After the habits had been thoroughly learned, the birds were given a rest for some weeks, and then tested again under the old conditions. A few trials were arranged with special reference to the study of imitation; the animals here were tested as to their ability to execute simple but unfamiliar acts, after having only seen them performed by an animal previously trained. Throughout the associational tests the animals very seldom received food in their cages; but as they were tested daily and allowed to satisfy their hunger completely at the last test, they were never in a state of "utter hunger"—a condition which most experimenters think best to avoid. A series of tests, given at the conclusion of the investigation, indicated that the odor of the food had not assisted the animals in reaching it.

In the two series of experiments (emotion and association) thirty-five animals in all were used. They were confined in large cages in a fairly well lighted and ventilated room, and were fed wheat, cracked corn, and occasionally fruit, and kept well supplied with fresh water and sand. They generally remained in a healthy condition throughout the tests, especially during the winter. To exclude, as far as possible, the disturbing influence of fear, they were usually handled only after the room had been darkened. As the noise made by the curtains was objectionable, the birds were tested with the room illuminated by incandescent lamps; the light was turned off before the birds were placed in position for the trials, and again before they were removed from the apparatus to the cages. It is generally agreed that an experimenter should be out of sight when giving a test. I am convinced that it is important to avoid being seen by the animals at any time. This involves great inconvenience, especially when one employs the pneumographic method, but better results are thus obtained.

For practical suggestions as to apparatus and methods I am greatly indebted to Dr. Robert MacDougall, at the beginning of my investigation, and to Dr. Robert M. Yerkes, throughout. I also owe much to the researches of Zoneff and Meumann,[188] Thorndike,[189] Mills,[190] Small,[191] and Kinnaman.[192] Porter's[193] interesting study of sparrows was made almost simultaneously with the investigation here reported. Fewer animals were used by him, but in some instances more tests were given.

III. INVESTIGATION OF EMOTION[194]

1. Respiration in general. The normal breathing-curve in pigeons is quite similar in contour to that of the human subject, although the rhythm is more rapid and the pauses are less pronounced. When acoustical, visual, olfactory, or tactual stimuli are given, various modifications appear, for example, quickening, deepening, and minor irregularities. It was noticed that meaningless stimuli (pistol-shots) quickly lose their disturbing influence, whereas the breathing remains sensitive to those of a significant character, such as the noises made by other birds. It was also found that a stimulus which no longer affects the breathing will sometimes occasion disturbance if accompanied by a second stimulus of another order, although of a weak intensity (summation).

2. Respiratory reactions to light. As the easy control of conditions makes vision an excellent field in which to work, light reactions were investigated in detail. Two distinct series of tests were given. One sought to determine the relation between quality of light and reaction; the other, between intensity of light and reaction. Four colors of one intensity and three intensities of one color, respectively, were used. In the first series four stimuli, one for each of the colors, red, yellow, green, and blue, were given daily; in the second series five daily stimuli were given, of the same intensity for any one day, and one minute apart; this made it possible to observe also the effect of repetition. Each stimulus was given at the beginning of a respiration and continued two seconds. When the tracings were studied, various modifications were noted, but special attention was paid to alterations in rate of breathing. In the case of both sets of trials an immediate quickening usually occurred after stimulation, and occasionally shallowing[195] and minor irregularities of contour.

In the first set of tests ten animals were used for twenty-five days. Average results indicated that red and yellow are less stimulating than green and blue. To secure data that would assist in the interpretation of these results, an investigation was made of the animals' color-preference. This was done by recording, at thirty-minute intervals, the position of the birds when confined, singly, in a box one half of which was illuminated (from the side) by light of one color, and one half by light of different color but of the same intensity. A water screen excluded the heat rays. After nine records had been taken the colored glasses were interchanged, and the animal's position relatively to the two colors was observed as before. This was repeated with the other colors until each of the four had been used with each of the other three. There were far more choices of green and blue than of red and yellow, though none of the colors was avoided. It seemed a question of degree of liking, rather than of liking or disliking. As stated, Graber's experiments indicated that pigeons have no color-preference, but his results are probably untrustworthy, since he tested several animals at once and apparently was not careful to change the colored glasses regularly. Putting together our two sets of data (the latter stated first) we have the following comparison:

RYGB
Color choices of 5 animals72129167172
Breathing-rise of 10 animals9.94%10.39%10.41%12.11%

Although the proportions do not hold, there is a direct correspondence between the two series of responses; hence it would seem that increased respiratory activity is an expression of agreeable feeling in pigeons, and this especially since the breathing, when varying at all in amplitude, usually became shallower, and also showed certain minor irregularities of contour, as often occurs in human respiration during moderate stimulation of a pleasant character.[196]

In the second series of respiratory tests four animals were used for fifteen days. Average results showed nothing as to the relation between intensity of stimulus and amount of quickening, since the three intensities used, 1, 2, and 4, produced reactions, respectively, as follows: 6.6%, 4.3%, and 6.4%. This may have been because the three intensities were employed each on different days. When the reactions are averaged according to daily succession, without regard to the intensities of the stimuli, we get the following results: first reaction, 8.0% rise in rate; second, 3.7%; third, 4.1%; fourth, 5.7%; and fifth, 6.9%. We should have expected the second daily response to be less vigorous than the first, since the animals were perhaps better prepared for the second stimulation. That the reactions increased thereafter was probably due, partially to summation, and partially to the fact that the short illuminations occasioned mental action (perception of interior of box, increased desire to escape, etc.) which involved heightened, rather than depressed, breathing activity, and thus worked directly against the dulling tendency of repetition.[197]