The principle of stereopsis, that crossed disparity causes a convex perception and uncrossed disparity a concave one, has for a long time been considered to depend on a very rigid neural mechanism not affected by experience. Experiments are reported here which show that this relationship between disparity and perceived depth can be reversed by experience. An observer wore a pair of left-right reversing spectacles continuously for nine days. The spectacles also reversed the relation between the direction of perceived depth and the direction of binocular depth cues, ie disparity and vergence. For a period starting two days before wearing the spectacles and continuing until seventy-nine days after their removal the observer was examined with a haploscope and an electrooculograph. All the stereoscopic experiments were carried out without spectacles in order to examine some aftereffects of wearing spectacles. For the stereograms with linear contours not only the adaptive reversal of the relation between disparity and perceived depth, but also some abnormal depth perceptions and long-lasting aftereffects were found. For Julesz's random-dot stereograms, however, in which contours can be seen only after binocular combination, no adaptive change or reversal occurred. These results suggest that the process of stereopsis consists of two concurrent subprocesses.
