Stereoscopy re-stated

Abstract

While the phenomenon of stereoscopy has been to a limited extent developed in a practical manner, there appears to have been but little actual research work done within recent years, the impression being a general one that the particular field had been already thoroughly explored. If an attempt be made to reproduce the experiments described by Wheatstone in his original paper, in which his great discovery of stereoscopy was published, it will be found that the results are somewhat different in important respects from those described. From experiments recently carried out by the author, it appears that stereoscopic vision is possible only within certain limits. For certain pairs of objects, whether on the same or different horizons, there are generally two extreme critical points beyond which stereoscopic vision breaks down. For certain pairs of objects on the same horizon there are two inner critical points. Objects of dissimilar form but approximately the same average angular dimensions can frequently be combined stereoscopically: thus a circle can be combined with a triangle if their average angular dimensions are about equal. When the angular dimensions are very different, combination is generally impossible: thus, for example, a thin line cannot be combined with a thick line or triangle or other form of considerably larger size. When the pairs of objects are dissimilar in size there is only one pair of outer and inner critical points if the objects are on the same horizon and one outer critical point if they are on different horizons. For pairs of objects any of which can be combined together, there are two pairs of such critical points. Data regarding the positions of the critical points under various conditions are contained in the paper, which also deals generally with the effects of variation of the conditions of observation and particularly variations of the illumination.