PHOTOSENSITIVITY IN INVERTEBRATES 



635 



200 per sec, better than five times the performance 

 of the human eye. In the cockroach Periplaneta and 

 the grasshopper Tachycines, by contrast, any flickering 

 rate higher than 5 or 10 per sec. was e\idently fused 

 and interpretated as a constant stimulus. The authors 

 postulated that in the orthopterans an after-image 

 was present, a phenomenon lacking in the dipterans 

 and hymenopterans. 



For the fly Calliphora the electroretinograms show 

 that the effective angle of view of each ommatidium 

 is about twice as great in the horizontal plane as in 

 the vertical (6). Hence a target remains for a longer 

 time within the visual field if it is moving horizontally; 

 summation can permit its detection at a lower thresh- 

 old than would be found in the same target moving 

 vertically. The structural basis for astigmatism of this 

 kind can be found in the dimensions and divergence 

 of ommatidia. Ommatidia facing downward com- 

 monly are relatively shorter and have larger lenses 

 than those facing upward; usually they diverge from 

 one another more strongly. Antrum (6) generalized 

 that in all insects which fly well the angle of view of 

 each ommatidium in the horizontal direction is 

 about twice that in the vertical. 



In Apis the situation is somewhat more complex 

 (i i). The radius of curvature of the bee eye is smaller 

 in the transverse plane than in the frontal, with a 

 ratio near 2.5 to i. The angle between ommatidia is 

 regularly greater in the transverse than in the frontal 

 plane, with a ratio of difference reaching about 2 to i . 

 In consequence maximum acuity lies in a plane in- 

 clined 65 degrees to the sagittal, and in this plane 

 only in an arc from 47 degrees behind the anterior 

 margin of the eye to 49 degrees ahead of the pos- 

 terior margin. 



The extremes of difference in dimensions and 

 angular separation among ommatidia in a single 

 compound eye are met in some deep-sea crustaceans 

 and in insects belonging to the orders Homoptera 

 (287), Ephemeroptera and Diptera (57, 58). In most 

 of these the region with short ommatidia and large 

 lenses is confined to one part of the eye, and the por- 

 tion with long ommatidia, slight divergences and 

 fine lenses forms a sort of 'turban' toward the top of 

 the head. In many instances the owner of such a 

 'divided' eye is a predator. However, Radl (221) 

 concluded that it indicated a duality of embryonic 

 origin and proposed a "duplicity theory.' Zavfel 

 (308) extended this into a triplicity theory, but later 

 workers have not supported either hypothesis. 



de Serres (53) appears to have been the earliest to 

 experiment with arthropod vision by painting over 



all or part of a compound eye with black varnish. He 

 found many of the postural changes which became 

 classic demonstrations of the 'muscle tonus' theory in 

 rather recent texts of physiology. Light intensity, 

 interpreted through the compound eyes, w-as be- 

 lieved to control the tonus of muscles involved in 

 posture and locomotion. 'Circus' movements of uni- 

 laterally blinded arthropods were explained on this 

 basis. 



At the same time de Serres pointed out the 'false 

 pupil' seen as a shifting dot or line in many living 

 compound eyes. Ewing (66) described it more fully 

 and concluded correctly that it represents ommatidial 

 pigment visible in those few ommatidia facing an 

 observer's eye. In cylindrical compound eyes, such as 

 the stalked ones of the crab Ocypoda, it takes the form 

 of a vertical line which follows the observer or camera 

 lens through as much as a 360-degree field of view. 



Duges (60) noted that a false pupil can be seen 

 simultaneously in the two compound eyes of many 

 insects and suggested that they must have binocular 

 vision. Demoll (47), working from sections of com- 

 pound eyes, showed the extent of these binocular 

 fields. It is easy to assume that binocular vision is 

 important to predaceous arthropods and that they 

 snatch for prey when the proper ommatidia in the 

 two eyes are stimulated simultaneously by an object 

 placed symmetrically in the binocular field. Distance 

 estimation is evidently good in both naiad and adult 

 stages of most members of the insect order Odonata 

 (i, 8, 50), among the dipteran family Asilidae (185) 

 and in tiger beetles. 



The adaptaiaility of the neural components in the 

 eye-brain team appears to have been underestimated. 

 In a matter of hours or days, the postural peculiarities 

 and circus movements of unilaterally blinded arthro- 

 pods often disappear entirely. Partially blinded odona- 

 tan naiads can adapt their behavior to approach 

 prey monocularly, pivot and seize at the appropriate 

 instant (i, 8). Whether marginal ommatidia can 

 participate in this versatility has been questioned (9). 

 It would be interesting to know whether variations in 

 adaptability correspond to the zones found in the 

 Notonecta eye (172), since in this heteropteran insect 

 kinetic responses seem related to specific areas of the 

 compound eye. Certainly ommatidia can serve in un- 

 usual reflexes (2191 220) but limitations may still be 

 present. 



Many, perhaps most, arthropod eyes show a sensi- 

 tivity to the plane of polarization of light from the 

 sky. Something of the kind has been suspected for 

 many years to account for the homing ability of vari- 



