Fig. 1. Data recording and evaluation. (A) Top view of a cricket walking on a Styrofoam^TM ball under a slowly rotating, polarized stimulus. The animal is kept on the spot by a balanced arm (not shown). Its walking movements are transferred to the ball and registered by detecting the moving dots on the surface of the ball. (B) Rotational movements of the cricket recorded during two full revolutions of the stimulus (4x180°). Abscissa: walking direction (rotational component of the run) given by the number of dots that passed the detector; positive and negative values indicate right and left turns, respectively. Ordinate: stimulus orientation. Provided that the translation (forward movement) of the cricket was constant, the resulting curve also reflects the virtual walking path. Note the bias in walking direction caused by the inherent turning tendency of the animal. (C) Fourier spectrum of turning speed per degree. Data shown in B were differentiated to remove the bias and then analyzed by a fast Fourier transform (FFT). Abscissa: period of modulation of walking direction. Ordinate: amplitude of FFT signal. Because of the 180° periodicity of the polarized signal, the amplitude at 180° (S) was taken as a measure of the strength of the polarotactic response.