MANY AFTER-EFFECT PHENOMENA have been reported since Purkinje (1825) and Addams (1834) independently rediscovered the after-effect of seen movement. After-effects were described before 1930 for such diverse perceptual spaces as auditory location (Flugel, 1921;Bartlett & Mark, 1922), repetition rate (Cathcart & Dawson, 1928), pitch (von Bekesy, 1929Cathcart & Dawson, 1929), colour (Cathcart & Dawson, 1929, and geometric size (Cathcart & Dawson, 1929). However, it awaited the now classic paper by Kohler and Wallach (1944) for interest in the problem to become widespread. Kohler and Wallach presented an ingenious theory based on the spread of currents which build up resistance to their own passage through the cortex. Many studies have tested predictions based on this theory of cortical satiation, and prediction has often proved satisfactory. However, some crucial experiments have given results at variance with the theory. Drastic changes in the gross electrical characteristics of the visual cortex have been produced, without apparent alteration in the visual function (Lashley, et al., 1951;. Ikeda and Obonai (1955a) reported die existence of a retroactive effect, in which the inspection figure follows the test figure in time. The displacements in the retroactive effect are of the same character as those in the ordinary after-effect. It is difficult to see how either of these observations can be reconciled with the basic premises of satiation theory.Other theories of the figural after-effects have been proposed (for example, Gibson, 1933;Osgood & Heyer, 1952;Obonai, 1954;Obonai, 1957;Yokose, 1954). Nevertheless, the recent reviews by Spitz (1958), McEwan (1958), andDay, et al. (1959), who did not consider the Japanese theories, agree that no satisfactory explanation exists for the figural after-effects. Both Japanese theories are restricted in application. That of Yokose appears, from the temporal relations found by Nozawa (1958), to deal only with neutralization effects, and that of Obonai seems not at present to be capable of quantitative prediction. Either or both of these theories may eventually prove compatible with the theory to be presented below. It is not suggested that the present paper gives a iDefence Research Medical Laboratories Project no. 250, DRML Report no. 250-3, H.R. no. 234.