Contribution of the otoliths to the calculation of linear displacement.

1. The present work is a quantitative study of the eye movements induced by linear translation when the subject is instructed to stabilize his gaze on a memorized earth-fixed target. These experiments may allow a better understanding of the central processing of otolithic signals. 2. Human subjects were submitted to either sinusoidal or step-like horizontal linear displacements along the interaural (Y)-axis in darkness, seated in a cart moving along a linear track. Each subject's head was fixed by a helmet secured to the cart. They were asked to keep their eyes on an earth-fixed memorized target at 63 cm from them on the X-axis. 3. During sinusoidal motion, a combination of low smooth compensatory eye movements and of compensatory saccades allowed the subjects to track the memorized target. The linear model of the responses of five subjects (seven sessions) exhibited a near-ideal slope of 1.14 (range 0.84-1.58). Two subjects did not compensate properly for their displacement. The mean "vestibular-saccadic" (VS) gain (ratio of overall eye movement peak-peak amplitude versus head displacement amplitude) was 1.52 +/- 0.80 (SD), showing an overestimation of head displacement. 4. The otolith-ocular reflex (OOR) mean gain values (ratio of slow phase cumulated peak-peak amplitude versus head displacement amplitude) were about 0.13 degrees/cm. This value is 5 times higher than what has been reported in the literature, probably due to the fact that the target was at a short distance. 5. The number of saccades occurring during sinusoidal stimulations varied according to the different subjects. They were obviously compensatory saccades and not quick phases. They indicate that although the gain of the OOR was small, the brain has computed the adequate desired eye position. 6. During steplike head displacements in darkness, although the OOR gain was also small, seven of the eight subjects could stabilize their gaze with a mean VS gain of 1.01 +/- 0.70. The linear model for the pooled responses of these subjects exhibited a slope of 0.82. 7. When subjects were instructed not to move their eyes during the translation, three of the five examined could still correctly reproduce the head movement amplitude with saccades, even as late as 50 s after motion had stopped. This indicates that head displacement was stored with the adequate metrics and could be used to drive the saccadic system. 8. Bilabyrinthectomized patients could not perform any adequate gaze stabilization. This shows that the observed performance was of vestibular origin.