Resolution of the locust eye measured by rotation of radial striped patterns

Abstract

The resolving power of the compound eye of the locust was tested by using plane radial patterns of stripes rotated about their axes at various distances from the eye. Nerve-cord spike discharges were used to indicate a response to the stimulus. Both large and small spikes were present in the response, but only the large spike response was evoked by stripe angles less than about 2°, and this response was used in the present experiments. Resolving power was found to be affected by a number of factors in the stimulus situation. Resolving power was increased by ( a ) increase in intensity of pattern illumination, ( b ) increase in size of pattern used, ( c ) extending the exposed area of the eye in masked preparations, ( d ) using a black centre and surround, ( e ) increasing the rotational velocity up to about 300°/s, ( f ) using patterns in which black stripes were wider than the intervening white stripes. Under conditions optimal for all these factors the highest resolution found was 10′ of stripe angle and a mean threshold of 20′. This result is closely comparable with that obtained earlier using linear patterns moved behind a rectangular window (Burtt & Catton 1962). Reasons are offered for supposing that the value for resolution was not dependent on unwanted periodicities in the patterns. An independent physical check showed these to be of a low order.