Fig 1.
Each trace represents 3 seconds of horizontal (top) and vertical (bottom) eye position recordings containing SWJs. Primarily horizontal (left), vertical (middle), and oblique (right) SWJs are displayed. All time scales are as in the bottom trace.
Fig 2.
Peak velocity-magnitude relationship for SWJ saccades and non-SWJ saccades.
Main panel: Each dot represents a saccade with the peak velocity indicated on the y-axis and the magnitude on the x-axis. Color determines whether the saccade was part of a SWJ (red) or not (blue). Bottom panel: Average saccade magnitude distribution across monkeys (n = 5). Left panel: Average peak velocity distribution across monkeys (n = 5). Saccade magnitude and peak velocity were greater for SWJ saccades than for non-SWJ saccades (Z-values = 2.02; p-values = 0.04). The slopes of the peak velocity-magnitude relationships for SWJ saccades and non-SWJ did not differ statistically (Z-value = 0.14; p = 0.89).
Table 1.
Characteristics of SWJs and non-SWJ saccades.
Fig 3.
Direction of SWJ saccades and non-SWJ saccades.
A) Polar histogram of saccade directions for SWJ saccades and non-SWJ saccades. We show the average saccade direction distribution across monkeys (n = 5). Both SWJ saccades and non-SWJ saccades are biased vertically. Whereas downward-directed component is prominent for both SWJ saccades and non-SWJ saccades, the upward-directed component is more pronounced for SWJ saccades than non-SWJ saccades. Median Euclidean distance values were significantly lower for SWJ saccades than for non-SWJ saccades, indicating higher symmetry (Z-value = 2.02; p = 0.04; see Methods). B) Saccade magnitude and vertical component. The vertical component remained constant across saccadic magnitudes for both SWJ saccades (Friedman's test (5, 9) = 5.50; p = 0.787) and non-SWJ saccades (Friedman's test (5, 9) = 3.19; p = 0.956). Error bars represent the s.e.m. across monkeys (n = 5).
Fig 4.
Relationship between fixation error and subsequent SWJ generation.
A) Schematic illustration of a fixation error. The fixation target location, the magnitude of the first saccade in the SWJ, and the distance to the fixation target after the first SWJ saccade are indicated. Relationship between a saccade’s likelihood of being part of a SWJ and B) the post-saccadic distance to the fixation target and C) the saccade’s magnitude. Both relationships follow a logistic regression (p-values < 0.05), where a saccade’s likelihood of being part of a SWJ increases with both the distance to the fixation target after the saccade [Friedman's test (5, 4) = 12.32; p = 0.015] and the saccadic magnitude [Friedman's test (5, 4) = 16.64; p = 0.002)] up to 1 deg, and then plateaus for values > 1 deg [p-values >0.9]. Grey shadows indicate the s.e.m. across monkeys (n = 5).
Fig 5.
Relationship between fixation error and Intra-SWJ ISI.
A) The larger the distance to fixation target, the shorter the intra-SWJ ISI [Friedman's test (5, 9) = 17.55; p = 0.047)]. B) Likewise, larger saccades tend to be more quickly followed by return saccades [Friedman's test (5, 9) = 18.80; p = 0.026)]. Shadows indicate the s.e.m. across monkeys (n = 5).