Can the oculomotor control system compensate for initial velocity when making a saccade?

It is generally considered - but unproven - that the oculomotor control system takes initial velocity immediately preceding saccades into account when planning and executing saccadic eye movements. This is probably so that degradation of vision during a saccade can be minimized by achieving an optimal velocity profile which balances conflicting requirements of getting to a target as quickly as possible while not overshooting or ringing once eye has reached its target. Simulations, based on a linear homeomorphic mechanical model of the eyeball and its musculature, showed that the peak velocity of a saccade is strongly dependent on the velocity of the eye just prior to the saccade. An experiment involving 4 normal adult subjects (3M/1F, 20-24 yr) was subsequently undertaken in which each subject was asked to visually track either a stationary or a sinusoidal target. At unexpected times during this smooth oculomotor pursuit, and while the subject's eyes were stationary or moving at various velocities up to /spl plusmn/40/spl deg//s, the target switched to a step to induce a saccade. The target steps were in both the same and opposite directions to that of the pre-saccade smooth pursuit velocities. Results from this experiment showed that, for both forward and reverse saccades, an increase in initial velocity increases the peak velocity of same-size saccades. For forward saccades, the brain is only able to partially compensate for the initial velocity of the eye. In contrast, for reverse saccades, the brain overcompensates for the initial velocity. This is the first study to demonstrate experimentally that peak velocities during same-size saccades are dependent upon initial eye velocity.