Abstract
Four tomtits were trained to pick up their food from under either one of two white covers of the same size, the form of the one coer being a circle (r=2 cm) and that of the other an ellipse (a=2.41 cm, b=1.66 cm). For two of the tomtits the positive stimulus was the circle and the negative the ellipse, and for the others the positive was the ellipse, the negative ine circle.The covers were placed side by side, separated from each other by a glass screen, right under the perch, so that the birds approached to the cover almost perpendicularly. The birds could get their food under the positive stimulus only, and no special punishment was used excepting the detour they had to make in older to get food when they went to the negative stimulus.
This training was considered to have been completed when the birds selected twenty times or more consecutively the positive stimulus. After the training was completed, the direction of ellipse to the circle, the relative size of the two forms, and finally the forms themselves were changed systematically to a certain degree, and it was asked whether these variations influence the result of the training or not. The answer was negative in general. The change which influenced most was that in the direction of ellipse and the change which influenced least was that of the size. But in each one of them the percentage of selection-of positive stimulus never wen down below 91.25% in average. Accordingly it is clear that there is a certain substitution-zone for these variations in the selective response of the tomtit for form.
In the following series of experiments two birds only among four were used. The stimuli were now removed to a point which was seen by the bird on the perch at an angle of 45° and they were supported by a wooden stand which had also an inclination of 45° to the floor, so that the stimuli were still placed at a r;ght angle to the visual line of the bird. After a training experiment was executed in this condition as before, the stand was removed and the stimuli were laid down on the floor, so that the forms were now seen by the bird at an inclination of 45°. According to the retinal images ot the bird, the circle should now appear as an ellipse. of the approximately same form as the ellipse of the stimuli, and the ellipse as a circle . Then it was observed toward which of the two stimuli the tomtit would approach first in this situation.
The result showed that the number of selection of the positive stimulus surpassed by a small majority that of the negative stimulus. It is obvious from this fact that the selective response for form on the part of the tomtit depends not so much upon the geometrical-optical condition of tha stimuli as upon the form of real objects. In tins sense we can acknowledge the existence of a certain tendency for constancy of form.
However, when the inclination of .stimuli to the visual line of the bird was changed to 30° by bringing down the perch. the perceniage of selective response for the positive stimulus suddenly fell and the tomtit which had been trained to go to the circle now began to go to the ellipse in most of the cases and vice versa. We can know from this fact that the tendency for constancy of form in the tomtit is not very strong.
