Fig. 1.

Double-step memory saccade task: the updating of the second target across the first saccade. (a) Craniotopic (screen) representation. After subjects fixate on the cross, the cross disappears, and the square and diamond are flashed successively. Subjects then sequentially saccade to the remembered square and diamond positions. (b) Retinotopic representation across the first saccade (back projected from the retina to the screen for comparison with a). The cross and square are superimposed as they correspond to the same retinal position, the fovea. The magenta and green arrows indicate the diamond’s retinotopic positions before and after the first saccade (the rightward black arrow in a), respectively. The leftward black arrow indicates that the diamond’s retinotopic position needs to be updated backward by subtracting the saccade vector.