感應理論の研究 (第九報告)

光滲の研究

抄録

Problem: It is well known that a black square on a white background appears smaller than a white square of the same size on a black background. This is the irradiation phenomenon (the positive irradiation). In 1863, however, Volkmann reported a new kind of irradiation which is of a contrary nature to the ordinary one; he found that very fine black threads on a white background looked wider than they really are (the negative irradiation after Volkmann). In order to measure precisely the extent of irradiation, he made use of the following procedure. By means of a micrometer-screw, two fine threads were gradually brought together until the interval was just the same as the width of the threads. The degree to which the threads had to be moved in order to procure the desired effect served as a measure for irradiation.
In the present investigation we proceeded from the experiments made by Volkmann and tried to make a further thorough analysis of the conditions which determine the occurence of the negative as well as the positive irradiation. For this purpose special apparatus, which enable us to change the brightness-difference between the object and its background, and also the width and length of the threads (test-objects), was constructed (Fig.5, Fig.6 in the Japanese text p.6).
Results: The first problem settled was the dependence of the irradiation on the width of the test-objects. The average outcome of the experiment for 10 observers is indicated graphically by Fig.7 in the Japanese text p.8, which shows how the irradiation changes with the increase in the width of the test-objects, the particular course of the change depending upon the brightness-difference between the object and its background. It is interesting to note in Fig.7, the course of the change in irradiation as the width of the test-objects increases; the negative irradiation decreases with the increase in the width of the objects, and the irradiation is reversed when the test-objects are sufficiently wide.
Next, the length of the test-objects is another factor which governs the change of the irradiation. It follows from Fig.8 that the irradiation increases up to a certain point with the increase in the length of the test-objects, but it decreases when their size exceeds a certain limit.
Lastly, an investigation of the relationship between brightness-difference and irradiation has been made following the same technique as described above. In this work, we used a number of test-objects of various brightness-difference which was obtained by illuminating the background, the brightness of the test-object being kept constant (black). The results of these experiments are plotted on logarithmic scale in Fig. 9 in the Japanese text p. 10. The continuous line indicates the calculated values; the discontinuous line the observed values. The agreement between the calculated and observed values is on the whole satisfactory. The equation found to express this relationship is
y=A+nlogx
in which y= amount of irradiation
x = brightness-difference between object and background
A and n are constants.
That is the relationship is exactly the same as obtains between stimulus and sensation in the Weber-Fechner law.