Stabilization of gaze during circular locomotion in light. I. Compensatory head and eye nystagmus in the running monkey

1. A rhesus and cynomolgus monkey were trained to run around the perimeter of a circular platform in light. We call this 'circular locomotion' because forward motion had an angular component. Head and body velocity in space were recorded with angular rate sensors and eye movements with electrooculography (EOG). From these measurements we derived signals related to the angular velocity of the eyes in the head (Ė(h)), of the head on the body (Ḣ(b)), of gaze on the body (Ġ(b)), of the body in space (Ḃ(s)), of gaze in space (Ġ(s)), and of the gain of gaze (Ġ(b)/Ḃ(s)). 2. The monkeys had continuous compensatory nystagmus of the head and eyes while running, which stabilized Ġ(s) during the slow phases. The eyes established and maintained compensatory gaze velocities at the beginning and end of the slow phases. The head contributed to gaze velocity during the middle of the slow phases. Slow phase G(b) was as high as 250°/s, and targets were fixed for gaze angles as large as 90-140°. 3. Properties of the visual surround affected both the gain and strategy of gaze compensation in the one monkey tested. Gains of Ė(h) ranged from 0.3 to 1.1 during compensatory gaze nystagmus. Gains of Ḣ(b) varied around 0.3 (0.2-0.7), building to a maximum as Ė(h) dropped while running past sectors of interest. Consistent with predictions, gaze gains varied from below to above unity, when translational and angular body movements with regard to the target were in opposite or the same directions, respectively. 4. Gaze moved in saccadic shifts in the direction of running during quick phases. Most head quick phases were small, and at times the head only paused during an eye quick phase. Eye quick phases were larger, ranging up to 60°. This is larger than quick phases during passive rotation or saccades made with the head fixed. 5. These data indicate that head and eye nystagmus are natural phenomena that support gaze compensation during locomotion. Despite differential utilization of the head and eyes in various conditions, Ġ(b) compensated for Ḃ(s). There are various frames of reference in which an estimate of angular velocity that drives the head and eyes could be based. We infer that body in space velocity (Ḃ(s)) is likely to be represented centrally to provide this signal.