We studied the trajectories of self-paced saccades in two experimental conditions. Saccades were made between two visual targets in one condition and between the same two, not visible, positions in the other condition. Target pairs were presented which required oblique saccades of 20 or 40 deg. At least 200 saccades were made between each pair of targets. Horizontal and vertical eye movements were measured of the right eye with a scleral coil technique. We computed the angle between starting and end point of each primary saccade (effective direction). We also computed the angle between starting point and eye position when the saccade had covered a distance of 2.5 deg (initial direction). We found that variability in initial directions was two to seven times larger than variability in the effective directions. This effect was found in both experimental directions for saccades made in all tested directions. We conclude that curvedness of saccades is the result of a purposeful control strategy. The saccadic trajectories show that, initially, the eye is accelerated roughly in the direction of the target and subsequently is guided to the target. This behavior cannot be described by present models of saccade generation. We suggest that the coupling between saccadic pulse and step signals is not as tight as generally is accepted in the literature.
|Title of host publication||Eye movement research : mechanisms, processes and applications|
|Editors||J.M. Findlay, R. Walker, R.W. Kentridge|
|Place of Publication||Amsterdam|
|Publication status||Published - 1995|
|Name||Studies in Visual Information Processing|