The main object of this study was a possible induction effect, in the sense meant, for example, by Gogel (1972). More precisely, it was a depth induction effect, since the crucial aspects that we considered in both its inducing and test components were related to perceptual depththe third dimension of perceptual space. In studying the hypothesized effect, we focused on certain positional relations between the inducing and the test components, which we will call optical contacts. The stimuli we used were static images displayed on a monitor. Figure 1 shows an example.The inducing component for the effect was a set of four regions covered by a random checkerboard texture, as illustrated in Figure 1. We will call this the frame (in full form: spatial reference frame) in the stimulus. Three spatial structures could be distinguished for it. First, the frame was endowed with an optical structure, which was its spatial organization as a part of the pictorial stimulusthat is, the image on the monitor. This structure was 2-D in character. 1 Second, the frame had a simulated structure. We produced our stimuli with computer graphics software, by defining a program so that the resulting image represented a corridor (a 3-D virtual structure). Lastly, the frame had a perceptual structure. This was the spatial organization really perceived by an observer looking at the stimulus, as we presented it in the experiments. On the frame, three definite cues to depth were in action-linear perspective, texture, and illumination perspective-which consistently supported a 3-D perceptual rendering. For this reason, we presumed that, with high probability, the perceptual structure of the frame would be 3-D in character. Of course, there might be differences between the simulated and the perceptual structures of the frame, in spite of their common 3-D character.The test component was a distinct linear unit, which we will call the target (as displayed on the monitor, it was red). When the frame perceptually organized as a corridor, the target might be interpreted as a pole, with a certain position and orientation inside the corridor. By the slant of a pole in the visual field, we will mean its inclination relative to the vertical in that field; more precisely, we will define it as the angle between the vertical and the projection of the pole on the sagittal plane. We will use positive numbers for upward slants (the frontal plane containing the upper end of the pole is farther from the observer than that containing the lower end), negative numbers for downward slants (the two frontal planes stand in the opposite relation), and the number zero for no slant (the pole lies in a frontal plane). Now, regarding the target in a visual context like that shown in Figure 1, we were specifically interested in its perceptual slant-that is, its apparent inclination in depth when it was perceptually interpreted as a pole inside a corridor. In its optical specification, the target was a straight-linear component having uniform thickness and color over its e...