202 THE PHYSIOLOGY OF INSECT SENSES 



If one recalls that the ommatidium is a multi-innervated sensillum 

 and that the neurons of a sensillum may possess quite different 

 characteristics (e.g., the chemoreceptive hair of flies), the indepen- 

 dence of retinal cells is not unreasonable. On the other hand, the 

 significance of the pattern of responses from single retinal cells cannot 

 be simple because of the intricacy of synaptic connexions in the lamina 

 ganglionaris. For every retinal cell whose axon traverses the area there 

 are many that synapse together on giant monopolar neurons. Further- 

 more, there are many local chiasmata each involving several ommati- 

 dia. Some success towards understanding the extent to which an 

 ommatidium is a functional unit of the eye has been achieved by the 

 sophisticated functional analyses of movement perception by Hassen- 

 stein (1951, 1959) and Hassenstein and Reichardt (1956). 



Behavioural measurements of visual acuity (usually derived from 

 optomotor responses) have not always corresponded to the minimal 

 ommatidial angle (Hecht and Wolf, 1929; Hecht and Wald, 1934; 

 Wolf, 1933 a, 1933 b). Compared with man, whose visual acuity is from 

 2 to 2-5, the maximal visual acuity of the bee is 0-017 and ofDrosophila, 

 0-0018. In the case of the bee, the value obtained corresponds exactly 

 to the smallest ommatidial angle (0-90-1-00 degrees) measured by 

 Baumgartner (1928). In other cases, however, the correspondence has 

 been less exact, the acuity being either greater or less. 



Where the visual acuity has been less than expected in terms of 

 ommatidial angles a number of explanations have been offered. Von 

 Buddenbrock and Schulz (1933) proposed that ommatidia might be 

 acting not as units but as pairs or groups. This hypothesis would also 

 explain observed differences of visual acuity at high and low light 

 intensities. As with man, the relation between visual acuity and the 

 logarithm of intensity is described by a sigmoid curve (Fig. 96) 

 (Hecht and Wald, 1934) which has been interpreted as an integral 

 population curve denoting the number of retinal elements active at 

 each intensity. In both the bee and Drosophila visual acuity and in- 

 tensity discrimination, as measured by optomotor reactions, begin at 

 approximately the same intensity and change rapidly within two log 

 units (Wolf, 1933 a, 1933 b; Hecht and Wald, 1934). 



It is also possible that visual-acuity measurements obtained from 

 behavioural studies depend upon what part of the eye is being stimu- 

 lated at the time. Viewed from the outside, the lens surface of a com- 

 pound eye is an approximate and imperfect hemisphere. The relative 

 area with respect to the head, and the number of ommatidia, vary from 

 species to species and often between the sexes. Each of the ommatidia 



