Oblique saccadic eye movements of primates.

The objective of these experiments was to determine whether the trajectories of the horizontal and vertical components of oblique saccades in primates were coupled. Human and monkey eye movements were recorded during a visual tracking task that jumped a small visible target spot to different locations on a tangent screen. For oblique saccades larger than ca. 3 deg, there was coupling between the horizontal and vertical components so that the duration of the smaller component was longer ("stretched") than would have been expected from its amplitude-duration relationship. The duration of a stretched component of an oblique saccade was linearly related to the vector amplitude of the eye movement but not to the amplitude of the stretched component. Stretched components of oblique saccades had lower peak and average velocities than would have occurred with pure horizontal or vertical saccades of the same size. Decreased component velocity was not caused by low-velocity eye movement components inserted at the beginning or end of the saccade, but was a function of the saccade's direction and component amplitude. For any saccade, there was a linear relationship between peak and average component velocity. We compared the discharge of monkey abducens neurons with the characteristics of the on-direction horizontal components of oblique saccades. The burst duration of an abducens neuron was lengthened when the horizontal component of an oblique saccade was stretched. Intraburst firing frequency was also decreased in correspondence with a decrease in horizontal component velocity. For an oblique saccade, the duration of the neuron's burst was correlated with the duration of the horizontal component and with the vector amplitude of the saccade, but was not correlated with the amplitude of the horizontal component itself. The duration of the smaller component of an oblique saccade was proportional but not always equal to the duration of the larger component. Usually, the smaller component began later and ended earlier than the larger component. These results show that the horizontal and vertical components of oblique saccades are coupled centrally so that the velocity of the smaller component is decreased and its duration is increased. For oblique saccades, larger than ca. 3 deg, amplitude-duration and amplitude-velocity relationships based on pure horizontal or vertical saccade data are not applicable. These findings are discussed in relation to three recently proposed models of coupled saccadic burst generators.