VISUAL SYSTEM: STATE OF THE ART 49 



during daylight (photopic conditions). Such a division of labor is expected 

 and reasonable in view of the day-night dichotomy of the photic environ- 

 ment, and, all else being equal, visual systems should adapt to this duplex 

 ecological factor. This is precisely what has occurred, if we accept the 

 Duplexity Theory. 



Rather detailed accounts of the evidence and inconsistencies of the 

 Duplexity Theory are given by Pedler (1965), Willmer (1965), Graham 

 (1965), Cohen (1972), and Crescitelli (1972). The important point for this 

 discussion is that until recently the elasmobranch retina was said to be 

 cone-free, except in one or two species. Actually there were several early 

 reports of duplex elasmobranch retinas, but these were not effective in 

 changing the viewpoint expressed first by Schultze (1866); later by Verrier 

 (1930), Walls (1942), Rochon-Duvigneaud (1943), Prince (1956), Duke-Elder 

 (1958), and Gilbert (1963); and more recently by Wolken (1975), namely, 

 that sharks possess a pure rod retina entirely devoid of cones and are thus 

 extremely specialized for nocturnal vision. Since 1963, when Gruber et al. 

 presented unequivocal histological evidence of cone photoreceptors in the 

 retina of the lemon shark, Negaprion, many authors have confirmed this 

 general finding in all taxonomic orders of elasmobranchs (Table 2). This 

 finding is all the more reasonable in view of Cohen's (1972) statement that 

 electron microscopists have failed to find a retina that is free of the type of 

 output terminal— the pedicle— characteristic of cones. The one exception 

 appears to be the pure rod retina of the skate Raja. Cohen himself (cited in 

 Green and Siegel 1975) failed to find any evidence of cone structure, and 

 Dowling and Ripps (1970) have looked exhaustively for conelike structures 

 or physiological activity in Raja, with negative results. 



Although anomalous and overlapping characters exist (Pedler 1965, 

 Underwood 1968, Cohen 1972, Crescitelli 1972), the differences between 

 rods and cones appear to be great, as can be seen in Table 3. The major 

 identifying features of the two receptor types are the shape of the outer 

 segment relative to the inner segment and the output terminal. 



Gilbert (1961), Kato (1962), and Kobayashi (1962) have investigated the 

 elasmobranch retina with conventional histological techniques under the 

 light microscope. Gilbert reported results from 15 species, mostly car- 

 charhinids, while Kato studied two carcharhinid species and Kobayashi 

 investigated Mustelus and six Rajaformes. These investigators reported that 

 all retinas were cone-free. However, Kobayashi presented evidence leading to 

 the conclusion that the photoreceptors of Mustelus, Holorhinus, Dasyatis, 

 and Urolophus (but not Narke or Raja) should be physiologically differ- 

 entiated into rods and cones even though there were no morphological dif- 

 ferences. Why these authors were not able to describe the cone photorecep- 

 tors present in most of the species studied is not clear. One problem could 

 have been thickness of the paraffin sections. Histological detail in sections 

 thicker than about 5 urn is obscured because one receptor overlies and 

 interferes with the visualization of others. Another possible difficulty is the 

 ph«*tolytic tendency of elasmobranch visual cells upon exposure to bright 

 light (Hamasaki et al. 1967). 



