Fig. 3. Stimulus size experiment. (A) 180° fisheye view of the celestial hemisphere taken by a camera positioned in a meadow. A considerable part of the sky is obstructed by vegetation. (B–D) Polarotactic response as a function of stimulus size. The radius (r) of a zenithal stimulus was reduced from 2r=92° to 1° with a degree of polarization (d) of either 100% or 0%. Tests (2r=1° to 48°, d=100%) are indicated by black, motivation controls (2r=92°, d=100%) by gray, and zero controls (2r=1° or 92°, d=0%) by white (16 series of 11 individuals). (B) Survey of results. The relative strength of the polarotactic response (S/S_mot, mean ± s.d.) is plotted against stimulus size. (C,D) Comparison between the largest (92°, top row) and the smallest (1°, bottom row) stimulus. (C) Distribution of S-values. (D) Walking direction of the crickets given by the number of dots that passed the detector (mean ± s.d.; positive and negative values indicate right and left turns, respectively) plotted versus stimulus orientation. Prior to averaging, data were standardized, i.e. the runs were phase-adjusted and corrected for an overall deviation from a straight walking path by subtraction of the inherent turning tendency. Note: a reduction in stimulus size to a diameter as low as 1° did not impair the polarotactic response.