Ocular vergence under natural conditions. II. Gaze shifts between real targets differing in distance and direction

Abstract

Horizontal binocular eye movements of three subjects were recorded with the scleral sensor coil - revolving magnetic field technique during vol­untary shifts of gaze between pairs of stationary, real, continuously visible targets. The target pairs were located either along the median plane (requiring symmetrical vergence), or on either side of the median plane (requiring asymmetrical vergence). Symmetrical vergence was primarily smooth, but it was often assisted by small, disjunctive saccades. Peak vergence speeds were very high; they increased from about 50° s -1 for vergence changes of 5° to between 150 and 200° s -1 for vergence changes of 34°. Differences between con­vergence and divergence were idiosyncratic. Asymmetrical vergence, requiring a vergence of 11° combined with a version of 45°, was largely saccadic. Unequal saccades mediated virtually all (95%) of the vergence required in the divergent direction, whereas 75% of the vergence required in the convergent direction was mediated by unequal saccades, with the remaining convergence mediated by smooth vergence, following completion of the saccades. Peak divergence speeds during these saccades were very high (180° s -1 for a change of vergence of 11°); much faster than the smooth, symmetrical vergence change of comparable size (14°). Peak convergent saccadic speeds were about 20% lower. This difference in peak speed was caused by an initial, transient divergence, observed at the beginning of all horizontal saccades. The waveform of disjunctive saccades did not have the same shape as the waveform of conjugate saccades of similar size. The smaller saccade of the disjunctive pair was stretched out in time so as to have the same duration as its larger, companion saccade. These results permitted the conclusion that the subsystems controlling saccades and vergence are not independent. Vergence responses were relatively slow and incomplete with monocular viewing, which excluded disparity as a cue. Monocularly stimulated vergence decreased as a function of the increasing presbyopia of our three subjects. Subjects were able to generate some vergence in darkness towards previously seen and remembered targets. Such responses, however, were slow, irregular and evanescent. In conclusion, vergence shifts between targets, which provided all natural cues to distance, were fast and accurate; they appeared adequate to provide effective binocular vision under natural conditions. This result could not have been expected on the basis of previous observations, all of which had been made with severely reduced cues to depth. We also found that asymmetric vergence was largely saccadic and conclude that the generation of saccades of unequal sizes in each of the eyes is a normal feature of oculomotor performance whenever gaze is shifted between targets that differ in distance as well as direction.