Abstract
Experiments have been conducted on human subjects to assess the role of the vestibulo-ocular reflex during co-ordinated head and eye movements to acquire stationary visual targets. During such manoeuvres there is a characteristic pattern of eye movement consisting of an initial large amplitude saccade in the direction of head movement followed by a slower return towards orbital centre which compensates for the remaining head movement. Experiments indicated that a similar pattern of eye movement could be elicited when the head was moved either voluntarily or passively in the complete absence of any visual target information. A comparison, within subjects, of the oculomotor response to head movements in the dark with those executed during visual target acquisition indicated that the important saccadic component of the eye movement was quantitatively similar in the two conditions. However, the observed differences in the temporal and spatial characteristics of the eye movement led to the postulation of a dual mode of control for head-eye co-ordination. In one mode, normally associated with low amplitude target offsets (< 45°), control is mediated by retinal information which frequently initiates eye movement before head movement. In the other mode, associated with larger target offsets, the gaze displacement is generated as an automatic response to head turning probably under the control of the vestibulo-ocular system. In support of this hypothesis a mathematical model of the mechanism of saccadic generation in the vestibulo-ocular reflex has been developed which gives responses compatible with the experimental results. In addition the model is able to predict the oculomotor response to many other types of vestibular stimulation. An attempt has been made, following an investigation of current experimental evidence, to suggest neurophysiological substrates for the model components, and further speculations have been made concerning the interaction of visual and vestibular mechanisms in oculomotor control.