Supplement to “figural after-effects: An examination of Köhler's theory”

Abstract

Towards the end of the previous paper (3), a series of experiments, called the “Second Set of Experiments” was reported, in which lines were used as the I- and the T-figures. The responses obtained in these experiments were divided into two classes, positive and negative, for calculations. Due to certain theoretical and experimental considerations, a new statistical treatment of these responses without dividing them into any such classes has become necessary. The new treatment of the responses shows that the figural after-effects are very small and their occurrence is uncertain. Besides, the direction of the effect of the I-line on the T-line in the experiments, in which the I- and the T-lines were shown simultaneously is opposite to that in the experiments, in which they were shown successively. The rest of the results obtained on the basis of the new treatment are substantially the same as obtained in the previous paper. Consequently, they lead to the same conclusion which has already been derived at the end of the previous paper, namely that more unimpressive I-lines cause greater deviations of the subjective horizontal than the less unimpressive I-lines.