In an earler paper I have reported that the structure of visusl space when we bend forward and look backward between the spread legs, that is, when we assume ‘matanozoki-posture’ (shortened as M-P infra. ‘Matanozoki’ is a Japanese word) is different from that when we assume a normal erect posture (shortened as N-P infra) (3). Now, why does this change in the visual space take place? The following factors may be taken into consideration as relevant to the solution of this problem: 1. the framework-effect of the spread legs, 2. raising or lowering of the eyes according as the bearing of the position of the eyes upon the object varies (2), 3. the change of relations between the directions within the process of visual perception and the gradient within the tissue of the visual center (3), 4. the change of the field of visual perception according as the physical Ego varies. It is assumed thad this change in the visual space under M-P is due to the change in the organization of the field of visual perception which is caused by one or the combination of some or all of these factors. However, it was shown in a previous work of mine that the factors 1 and 2 are not essential in this respect. (4).
The present article is an outcome of my attempt to clarify which of the factors 3 and 4 is essential, or whether both of them are essential. I made this study using the slight variation in the degree of size constancy (shortened as c infra) as a clue. In the first place, I made a comparison of two c's, c when the whole visual field is in a normal erect position (at 0°) and c when the whole visual field was turned upside-down (at 180°). In either case the body of the subject was kept in a normal erect posture. Further, in order to see more clearly the relative degree of c's at 0° and at 180°, I also measured c's in three different situations, 45°, 90° and 135°, as the whole visual field was beig turned to the right from 0° to 180°. The turning of the visual field was effected by means of a trapezoidal prism, and monocular regard was adopted. As a result of these experiments, it was found that the transitional relation of the degree of the c's in different situations took the form of c at 0°C>……>c at 90°C<……>c at 180° and that c at 0°=c at 180°. Thus it may be concluded that the factor 3 is not essentiasl in the determination of the structure of the visual space under M-P.
Next, I compared c in M-P. with c in N-P. Under various conditions (binocular regard, monocular regard, use of a redudtion-tunnel, etc.). The results showed invariably that in each case c in N-P>c in M-P, suggesting the factor 4 might be the only essential cause of the change in the structure of the visual space under M-P.
The change of the physical Ego (factor 4), however, takes place in two directions, (a) the change in the relation between the body and gravity and (b) the change in the bending of the body and thus we have to determine which of the two is the more essntial in our case. I compared, in the first place, two c's, c under the normal erect posture (at 0°) and c when the body of the subject was turned upside-down (at 180°). In addition to this, I measured c's in three different siuations, at 45°, 90° and 135° as the body was being turned to the left (so that the whole visual field was being turned to the right) from 0° to 180°. Here also I found that the transitional relation of the degree of c's in those different situations took, as in the case of the factor 3, the form of c at 0°>……<c at 180° and that c at 0°=c at 180°. This indicatcs that quite likely the factor 4-a may not play any important role in the determination of the nature of the field of visual perception under consideration. Thus, there remains only the factor 4-b to be tested. I have compared c when the
