Abstract
This article reports on three experiments in viewing illusions held at different slants to the line of sight. In Experiment 1, we used six geometrical illusions and found that they were illusory when viewed with the line of sight normal to the display surface and that the illusion was dispelled when the figures were viewed from glancing angles. In Experiment 2, questions were put to the viewer about the real state of affairs in two of the illusions in Experiment 1. Viewers who suffered from the illusion from one vantage point and who had the illusory effect dispelled from the other vantage point were unable to discern the correct state of affairs. Experiment 3 found that aspects of shape in a geometrical illusion could be accurately discerned by viewers given a set of vantage points spaced 360' around the figure.