204 THE PHYSIOLOGY OF INSECT SENSES 



faces in a different direction. As a result of these differences in align- 

 ment and the fact that the lenses may differ in size from one part of the 

 eye to another, the angle subtended by adjacent ommatidia is not 

 constant for any given eye (Baumgartner, 1928; del Portillo, 1936). 

 Measurements of visual acuity in insects in which various areas of the 

 eye have been blacked out have shown that the eye is not uniform 

 (Hecht and Wolf, 1929). When the centre of an eye, where the minimal 

 ommatidial angle is 1 degree as compared with 4 degrees at the 

 periphery, is blacked out the visual acuity decreases (Fig. 97). It also 

 decreases as the number of functional units is reduced (Fig. 97). 



Another technique for measuring visual acuity was employed by 

 Burtt and Catton (1954) following their discovery in Locusta, Phormia, 

 and Calliphora that the movement of illuminated objects in the visual 

 field was accompanied by specific electrical discharges in the ventral 

 nerve cord. By employing these discharges as a criterion of movement 

 perception, Burtt and Catton were able to ascertain that the overlap of 

 ommatidial fields, at least in Locusta, is much greater than heretofore 

 realized. The angle of the visual field is 20 degrees; the average angle 

 subtended by adjacent ommatidia is about 2-4 degrees in the longi- 

 tudinal and 1 degree in the vertical meridian. Responses to movement 

 were obtained to angular displacements in the visual field as small as 

 0-16 degree of arc. The threshold of acuity was independent of light- 

 and dark-adaptation. Burtt and Catton suggested that the extensive 

 overlapping makes possible the perception of an angle smaller than 

 that subtended by adjacent ommatidia; however, the experiments of 

 de Vries (1956) suggesting independent action of individual retinal 

 cells offer an alternative explanation. 



Whatever their visual acuity may be, there is no doubt that insects 

 can resolve patterns. A series of field experiments with honeybees has 

 shown them to be fairly efficient in this respect (Hertz, 1929 a, 1929 b, 

 1931, 1933a, 1934, 1935a, 1935b, 1935c, 1937a). By training bees to 

 associate specific black figures on a white background with food. Hertz 

 was able to demonstrate that contrast, contour, and the degree of sub- 

 division of the form are perceived. She concluded that the most greatly 

 divided pattern is preferred, but that the pattern is recognized as such. 

 The results of Zerrahn (1933) and Wolf and Zerrahn-Wolf (1937), 

 based on field and laboratory experiments in which bees were per- 

 mitted to choose patterns spontaneously, are in essential agreement ; 

 however, these workers derived different conclusions. They contended 

 that the pattern as such has no meaning. The number of choices of each 

 pattern was proportional to the length of its contours, that is, the 



