The time required to compare extents in various orientations

In Experiment I, Ss compared the lengths of lines that varied in absolute and relative orientation. Their reaction times showed a constant error pattern of which the familiar horizontal-vertical illusion was a special case, but horizontals were underestimated (relative to other slopes) much more than verticals were overestimate. The effect of difference inorientation was not great, though RT did increase significantly as the difference neared 90 deg. Experiment II was similar except that the lengths of lines were compared withthe widths of bars in a grating. In this case, no consistent pattern of constant error emerged, and there was no suggestion that difference in orientation affected RT.