The present paper is the second report on the experimental studies, which were designed to elucidate animal's behavior in relation to the goal. Approach behavior: The Ss used in the first experiment were ten guinea pigs. They were thoroughly trained to run for food from the right end to the left end of the alley. After the training trials, tests were made to determine the strength of pull they would exert when food was removed from them at different points in the alley, under 23-hour food deprivation. In this situation, the right end of the alley does not acquire the character of the starting point. The results are shown in Fig. 5. It was found that the approach gradient was a negative growth function of the distance from the goal, that is, the animals pulled harder the nearer they were to it. But, a question rises whether it is really a goal gradient. In order to make this point clear, an additional experiment was designed under the situation under which the number of passages was equalized at each point in the alley. The results are presented in Fig. 6. A comparison of the curves of Fig. 5 and Fig. 6 indicates that the gradient of Fig. 5 was not based upon the difference in the number of passages at each point, but the effect of the goal as a function of the distance between the subject and the goal. The second experiment was designed to investigate the approach gradient as a function of drive level, under 4-, 12-, 23- and 36- hour food deprivation. After five animals had been trained to run down the alley to secure the food separated from them, they were tested for strength of pull under different drive levels. The heights of the gradients, as can be seen in Fig. 8, vary with the strength of drive. Escape behavior: The third experiment was designed to investigate the escape gradient as a function of the distance from the point at which the animals received a shock and the effect of the strength of the shock upon the gradient. The Ss used in this experiment were fifteen guinea pigs, and they were divided into three groups by shock intensity given, that is, 8-volt group, 20-volt group and 32-volt group. They were tested for strength of pull at different distances from the point at which the shock was administered. The results are presented in Fig. 13. The animals pulled harder the nearer they were to the shock point. That is, the escape gradient is a negative growth function of the distance from the shock. The heights of the gradients vary with the strength of the shock. Besides, a comparison of the results of approach and escape behaviors indicates that the escape gradient is steeper than the approach gradient.
The fact that the visual threshold of one stimulus figure is influenced by another one presented in spatio-temporal proximity to it has usually been regarded as products of some psychophysiological interaction or “induction”. The strength of the induction, or of the “field” influenced by one figure at a certain point around it, is usually determined in terms of the change, in general heightening, of the threshold of a small test spot (TS) presented there. For instance, Yokose, following Köhler's isomorphism, has studied this effect systematically, and has proved that the properties of psychophysiological “potential field” could exactly be predicted in terms of the formula derived from these of electromagnetic field. On the other hand, some have insisted that this effect is to be attributed to the stray light in the eye, and have called it the “glare effect”. In the present experiment, it was attempted to decide which of the two theories was correct. Four experiments were performed. (1) The threshold of TS was measured under two conditions: First, TS was exposed at the center of a dimmly illuminated disc, and second, at the center of the enclosed dark area by a bright ring. (In the former, the stimulus situation was that of direct adaptation, and in the latter, of indirect adaptation.) A value of the illuminance of the disc and of the ring respectively was determined experimentally so that the thresholds of TS was the same under both conditions. The ratio of illuminance of the disc to the ring thus obtained being kept constant, five levels of illuminance of these two figures were set in equal logarithmic interval. Thresholds of TS were measured at each of the five levels of illuminance for two figures. If the glare effect theory was correct, the threshold of TS at each of the five levels should be identical under both conditions. Our results confirmed the prediction. (2) Area-intensity interchangeability of glare effect was examined. As stimuli, contour figures of a ring and a square, or segmental parts of each were used. Illuminance of the stimuli was determined such as to increase in proportion to the decrease of the area. The test spot was always exposed at the center of the ring or square, or at the corresponding position for segment figures. No remarkable differences were found among these conditions. (3) Boynton has recently studied the nature of stray light in an excised eye, and obtained a regular functional relationship between the strength of stray light and distance. We have tried to compare following two conditions. One was to vary the distance of the disc figure from the centrally fixated TS at the fovea (glare stimulation). The other was to vary the illuminance of the disc, presented always at the central retina (figure-ground relationship). Illuminance conditions in our experiment were determined by calculation from Boynton's result, in order to show the strength of stray light at each distance. If Boynton's result was right, it would be expected that in the two conditions the adaptative liluminance at the position of TS should be equal. Results showed that this expectation was correct and the threshold of TS was quite the same under two conditions. (4) Optic disc is receptor-free. Even if light is casted there, no physiological changes occur. In this experiment, a light stimulus was casted to optic disc and to the neighbouring positions, while the thresholds of TS were measured always at the fovea. Results showed that the threshold of TS in this case was not lower than when a light stimulus was casted to any other positions. The threshold was merely a function of the distance from the light stimulus to TS. It was concluded that these findings could. be explained more plainly in terms of stray light rather than of physiological induction.
The “assimilation-contrast” phenomenon accompanying the illusion of concentric circles may briefly be described as follows: With the increase in size of the outer circle, the perceived size of inner circle increases gradually to a certain maximum, beyond which it decreases and even gives negative illusion. The fact that the maximum illusion is given when the ratio of the outer circle to the inner circle is 3/2 is what we are concerned with in this study. We consider that this ratio should be given a functional significance in theoretical considerations of the illusion. We have carried out experiments with a series of modified figures: various part or parts of the original double circles pattern (Stimulus Figures A-I), those accompanied with additional lines (J, M, N), two pairs of parallel straight lines of equal length (K, L, O), and those with different lengths (P). Results can be itemized as follows: 1) Figures with inner and outer arcs (A-E) have phenomenal characteristics distinct from those of the original pattern. 2) Such a distinction also holds in the case of figures having inner arcs and outer circles (F). 3) Dimensional differences of the size and distance in the figures may be of little significance in consideration of functional relations. It follows that two (physical) circles are not always necessary to characterize the illusion of concentric circles (I). 4) In the case of figures having inner circles and outer arcs, the arcs are of such significance as. to play the role of circles. 5) In perceiving the figures with additional lines (J), the illusion is the maxima when the ratio reaches the value, which is given in the case of full circles. 6) The difference between pairs of arcs and pairs of lines is not merely of geometrical nature but supposedly of functional or configurational one. 7) In the perception of figures having pairs of straight lines with distances corresponding to those in original circles, the ratio for the maximum illusion changes gradually from 3.5/2 through 3/2 to 2.5/2 as the length of lines increases. This fact means that the distance is not the exclusive factor for determining the illusion and also that the length has some special functional meaning in the case of part figures such as arcs and lines. These results, as a whole, present us the problem of how to consider illusion in terms of configurational factors. However, in explaning functionally the above mentioned phenomena, the concept of Figur itself can not always be a positive factor, for it could not be concluded from above experiments that the space between inner lines in P figure is easier to be perceived as Zusammengefasste than the space between inner arcs in J figure.