When there is relative motion between an observer and a textured surface such as the ground, motion-related visual information is available about the orientation of the surface through vergence of the paths of the elements of the texture, change in their sizes and in their velocities. This experiment determined the perceptual effectiveness of each variable in foveal and peripheral viewing for producing perceptions of three-dimensional motion. The three variables were electronically separated and displayed singly, in all possible pairs, and all together. Subjects communicated their perceptions of degree of perceived surface tilt at the top, bottom, and middle of the display for the eight different combinations of variables because, during pilot work perceived surface bendings were frequently noted where top, bottom, and middle appeared tilted to different extents. All three variables can lead to relatively reliable perceptions of perceived orientation of the plane of motion, with change in velocity being the most powerful determiner. Change in size was the weakest. Certain combinations can lead to perceptions of extreme warping and should be avoided in motion displays. Subjects consistently underestimate the amount of simulated tilt. Foveal viewing was more accurate than peripheral viewing but peripheral performance was adequately consistent as an input channel for some orientation tasks.

A means of conceptualizing and generating visual displays that are "self-stereoptic manifolds" is described. First, single patterns that can replace pairs of stereograms to produce illusions of depth are defined and an example is shown. Patterns such as these produce illusory three-dimensional objects hanging in space before or behind the display surface. It is further demonstrated geometrically that such a display actually has three-dimensional information embedded in it peculiar to each of a family, or manifold, of objects that can be experienced one at a time. Each object of the family appears when the viewer looks in space where it "exists." The others remain invisible unless their locations are fixated. If any member of a specific manifold of three-dimensional illusory objects is physically duplicated as a real object and textured in the same way that the illusory object appeared to be, then this new real object will, in turn, generate an illusion of each of the other objects of the manifold when the observer fixates in space where each "exist." Also, if then the viewer looks where the original display previously was, the newly constructed object will disappear and the display will reappear. The geometry and the advantages and disadvantages in relation to a stereoptic pair are discussed.