The main principles of stereoscopy as known experimentally with plane pictures as subject matter have been fairly well understood since the days of Wheatstone and Brewster. Thus it is common knowledge that two essential conditions for obtaining the full stereoscopic effect are: First, that there must be dissimilar pictures registered from two discrete locations and singly presented, one to one eye and the other to the other. And secondly, that when fixating analogous points in these pictures the eyes must be converged to that optimum extent which for the given individual will give a unitary image having full solidity. In experimental stereoscopy these two conditions, single presentation of dissimilar pictures and optimum convergence of the eyes, can be met by a variety of instrumental means, the most familiar being the lenticular stereoscope of Brewster. This serviceable instrument has the disadvantage, however, of limiting one to small pictures viewed at close range. These limitations of the prismatic stereoscope have recently been overcome in the Toronto laboratory through a special type of stereoscope with flexible prisms which will enable one to fuse at will juxtaposed pictures of any size and shown at any distance from the observer. The facility of experimental control thus obtained has brought to light some points about stereoscopy which I believe are not generally appreciated. With one of these this paper will deal. (PsycINFO Database Record (c) 2006 APA, all rights reserved)