INVERTEBRATE PHOTORECEPTORS G49 



tapetum analogous to that in the eyes of such vertebrates as the fish 

 Abrmnis. A tapetum associated with the basement membrane can be 

 called a retinal tapetum, in contrast to an '4ris tapetum" of reflecting 

 pigments in cells distal to the receptors. Iris tapeta are known in various 

 crustaceans and insects as reflecting caps to pigment cells surrounding 

 the receptors. These caps can be shifted distally or drawn basally, and 

 their position alters radically the outward appearance of the eye (Zimmer- 

 mann, 1913; Welsh, 1930; Uchida, 193-1). In terrestrial arthropods the 

 iris tapetum may assist the organism by rejecting ultraviolet or infrared 

 radiations that would produce excessive heating at the receptor level. 



According to Notthaft (1881), each ommatidium operates on an all- 

 or-none principle. Either a target is included in its visual field enough 

 to stimulate the receptor system, or it is not. This dichotomy is cer- 

 tainly too severe a view, but under certain circumstances a close approxi- 

 mation is reached. If the visual field consists of upright dark bands 

 alternating with pale (as in a caterpillar's view of trees against the sky) 

 or consists of pale flowers against a background of dark foliage, relative 

 movement between a compound eye and the complex target will sweep 

 each feature of the target across one ommatidial field after another, 

 inducing on-off responses in the receptor cells. The characteristic swing- 

 ing of a caterpillar's head or the normal movements of a flying bee pro- 

 vide all the movement required for a pattern to produce flickering 

 stimulation in each ommatidium. The compound eye seems particu- 

 larly efficient in detecting flicker. As in the human eye, there is a close 

 relationship between the maximum rate of alternation in which the visual 

 mechanism can detect dissimilarities between the bright phase and the 

 dark phase, and the intensity of illumination. Flicker-fusion curves, like 

 visual-acuity curves, are essentially straight lines when plotted on a prob- 

 ability grid (Wolf, 1933a,b). This may be due to a normal distribution 

 of sensitivities among the ommatidia. Or it may arise through the con- 

 vexity of the compound eye in that, the more intense the stimulation, 

 the more ommatidia not facing directly toward the target area are 

 obliquely illuminated at intensities above their thresholds; higher intensi- 

 ties would then recruit more ommatidia. Crozier and Wolf (1939) 

 believed that they demonstrated that the latter was the limiting factor 

 in the convex eye of the crayfish Cambarus. 



Two very different ranges of flicker detection were found in insects by 

 Autrum and Stoecker (1952). In the fly Calliphora, the wasp Vespa, 

 and the bee Apis response to flickering could be detected at rates as high 

 as 200-220 per second. In the cockroach Periplaneta and the grasshopper 

 Tachycines any flickering rate higher than 5 or 10 per second was appar- 

 ently fused into a constant stimulus. These authors postulated that in 

 the orthopterans an afterimage extended temporal summation, whereas 

 in the fly and hymenopterans examined afterimaging was lacking, per- 



