It is well known that the temporal gap from extinction of a fixation point (FP) until appearance of a saccade target produces saccades having an extremely short reaction time (express saccade, Fischer, 1987, Rev. Physiol. Biolchem. Pharmacol., 106: 1-35). This gap effect can be explained as that the attentional disengagement begins at saccade target appearance in no gap condition (or overlap condition), whereas the disengagement begins at FP extinction in the gap condition. Thus, the FP is a factor responsible for the saccade reaction time (SRT). In the present study, the effects of the FP size and the attended-area size on the FP was investigated in the gap and overlap conditions.

Four male students participated in the experiment as the subjects. Their heads were immobilised in complete darkness. Visual stimuli were produced using a 15-inch non-interlaced computer display. Horizontal eye positions were recorded using an infrared limbus reflection system. The experiment consisted of the gap and overlap paradigms. In the gap condition, a central FP was turned off 200 ms before appearance of a peripheral saccade target. In the overlap condition, the fixation stimulus remained visible throughout each trial. For each condition, the following four tasks were conducted: small stimulus (SS), large stimulus (LS), small attended (SA), and large attended (LA) tasks. In the SS and LS tasks, the FP diameter was 0.2 deg and 3.2 deg, respectively. In the SA and LA tasks, the FP was a concentric disc: a small disc (dia. 0.2 deg) superimposed upon a larger disc (dia. 3.2 deg). The inside and outside discs were distinguishable by their colour difference. In the SA and LA tasks, during FP fixation the subject was asked to concentrate the inside and outside discs of the FP, respectively. The saccade target (dia. 0.8 deg) was presented at one of the four positions: 4 and 8 deg to the right and left.

The results showed that the SRT's of almost all saccades elicited in the gap and overlap conditions were within the SRT range of the express saccade (100 - 125 ms) and the regular saccade (150 - 350 ms), respectively. For both gap and overlap conditions, the mean SRT in the SS task was significantly longer than that in the LS task for all subjects. In addition, the four subjects showed the common tendency that the mean SRT in the SA task was longer than that in the LA task. These results suggest that the attended-area size during fixation influences the SRT.

No researcher has described the mechanism of the disengagement of the attention before the saccade in detail. We propose the hypothesis concerning the disengagement as that the attention focused on the small area during the fixation broadens the expanse of attention during the disengagement. That is, in the case of the smaller FP or smaller attended area, more time is necessary for the saccade to be able to occur because it takes more time to widen the attended area.