§1. Studies of the so-called ‘constancy’ of visual velocity have originally been made by J. F. Brown (19 8, 1931), and followed by J. Ogasawara (1938) and H. Wallach (1939). Brown performed experiments on this phenomenon on the assumption that if the constancy of velocity was a mere consequence of that of size, velocity ought to be quite as constant as size is. His results however designated that the constancy of speed was more reduced than that of size at the same distance, and he concluded that the former could not be deduced from the latter. Ogasawara, on the other hand, finding from his experimental results that speed constancy was considerably reduced with the decrease of the inhomogeneity of the moving field indicated that as this was the common feature of all other constancy phenomena, speed constancy must rest on the same psychophysical bases as other common constancy phenomena are. As for Wallach he came to the conclusion theoretically that speed constancy could be explained with no reference to the constancy of visual size but by the ‘transposition principles’ (in Brown's sense) alone, if those principles were applied to retinal images and displacements.§2. Let us discuss on the above three theories. Brown's outcome that speed constancy is more reduced than size constancy is inconsistent with Ogasawara's in which the former constancy is either the same as or even larger than the latter. But the assertion of Ogasawara does not seem valid partly because his experimental set-up is incomplete and partly because it conflicts with Brown's results, in which the constancy of speed does not differ when the experiment is conducted in the dark room from when performed in the light room. I tried a series of experiments in order to examine this point. Both experiments on speed constancy and on size constancy were performed under the same conditions (the same objects subjects, distances, etc.) both in the highly articulated and less articulated visual fields. Table I shows the results of three subjects. The figures represent constancy indices computed by Brunswik's logarithmic formula.Table ISIZE CONSTANCYSPEED CONSTANCY(a) under the heterogeneous conditions in the light room.(b) under the less heterogeneous conditions in the light room.It can be stated from the table that size constancy is greatly influenced by the field condition, whereas speed constancy is not, though both experiments were tried under the same situations. I performed further experiments in order to find any relationship between speed constancy, on one hand, and the distance and the visul field on the other hand. The results lead to conclude that (1) even in the case of the completely dark room and of the monocular observation, constancy indices are nearly 1 ranging from .94 to .89, (2) the homogeneity of the moving field, however, reduces speed constancy just a little and (3) constancy indices have no correlation to distance, but visual speed is decreased with distance. This leads me to question the validity of Qgasawara's conclusion.Though quite plausible Wallach's theory is, stating that the phenomena of speed constancy and the ‘transposition’ of visual speed are the same on the retina, how can it explain the fact that while speed constancy is greater in the daylight illumination, the transposition principles are impaired under the same conditions?§3. Suppose the two moving fields A and B are set at 1m and 10m distance respectively in front of the observer in a completely dark room and B is seen, ceteris parlbus, as nearly a tenth of A in size and both are seen on the same frontal parallel plane. This situation is phenomenally nothing but that in which two objects A′ and B′ in the relation of B′=1/10A′ are placed equidistant. Therefore if we assume that the transposition principles work with reference to the phenomenal variables, we must
Read full abstract