An internal clock for predictive saccades is established identically by auditory or visual information.

Previously we have shown that repetitive predictive saccades to alternating visual targets are mediated by an internal clock. That is, when subjects track a periodic visual stimulus alternating at a high rate (a small inter-stimulus interval, ISI, of 500 or 833 ms), they use an internal estimate of stimulus timing to pre-program the eye movement timing. Auditory pacing tones at the same rate also generate predictive saccades. It is natural to ask if an identical internal clock is used to generate the predictive saccades in each case. We hypothesized that if subjects can use auditory information to establish an internal estimate of stimulus timing--as we demonstrated can be done with visual targets--then the distributions of predictive inter-saccade intervals should demonstrate the well-known "Scalar Property" for either Auditory Cued or Visual Cued stimuli: inter-saccade interval histograms should be almost identical when each is divided by its mean. However, when making reactive saccades to a pacing stimulus (at a low rate), there should be a difference in the timing statistics between Auditory and Visual pacing, due to differences in sensory processing. We report here that the variances of inter-saccade intervals at three predictive pacing rates (ISIs of 500, 833, and 1000 ms) are equivalent, whereas the variance for Auditory Cued Pacing was greater than that for Visual Cued Pacing during reactive saccades at two reactive pacing rates (ISIs of 1667 and 2500 ms). When the inter-saccade interval histograms at the predictive pacing rates were normalized, the distributions were nearly identical for both Visual and Auditory Cued Pacing, which means that the Scalar Property holds for predictive saccades from either pacing stimulus. These results suggest that (1) an internal timing reference (clock) can be established by either auditory or visual information and (2) during predictive tracking the variability in saccade timing is due to the variability in the internal timing representation, while during reactive tracking the variability in saccade timing depends on the sensory modality used to trigger the saccades.