It is well known that a stimulus flashed in the dark before, during or after saccade tends to be mislocalised. Size of the mislocalisation depends on the difference between stimulus onset and saccade onset. This phenomenon has been frequently used to examine the time course of internal representation of eye position (e.g. Honda, 1990, Attention and Performance, XIII: 567-582). In such studies, it is assumed that the size of mislocalisation is equal to the deviation of internal representation of eye position from real eye position at the moment of stimulus onset. This assumption is generally accepted as far as stimulus is flashed in the dark, however, is that really true? To examine the validity of this assumption, we executed two experiments.

In experiment 1, we measured the time course of mislocalisation of a dot flashed for 1 ms before leftward saccade, and then compared it with the perception of a flickering dot (200 Hz, 1 ms-on 4 ms-off) presented for 80 ms before leftward saccade. From the measured time course of mislocalisation it was suggested that the flickering dot should be perceived as an array of dots drawn from right to left, but subjects never saw such arrays. In experiment 2, two dots was flashed at 80 ms interval before leftward saccade, and subjects reported whether the latter dot was perceived at right or left of the former one. From the time course of mislocalisation it was suggested that PSE of the latter dot should shift to the right, but PSE did not shift.

These results suggest that perceived location of brief stimulus is not determined solely by the internal representation of eye position and the retinal information at the moment of stimulus onset. We must be careful in discussing the relationship between the time course of mislocalisation and the time course of internal representation of eye position.