288 



THE EYE IN EVOLUTION 



usual, of a single layer of hexagonal cells, but when a tapetum is 

 i^resent these are unjiigmented until the ora is reached in order to 

 allow the passage of light to tliis structure (Fig. 315). The architecture 

 of the retina itself is simple with the usual layering, but a considerable 

 scattering of cells outside the confines of their layers may occur. The 

 horizontal cells are unusually massive (like those of the lamprey) and 

 ganglion cells are sparse. Characteristically the retina, is pure -rod, the 



Fig. 315. — The Selachian Retina. 



The retina of the ray, Raja maculata. 1, pigment of 

 choroid ; 2, (non-pigniented) retinal epitheUum ; 3, laj'er 

 of rods ; 4, external limiting meinbrane ; 5, outer 

 nuclear layer ; 6, inner nuclear layer ; 7, nerve fibre layer 

 (Mallory's trijjle stain) (Katharine Tansley). 



Fig. 316. — The 

 Cone and Rod 

 OF THE Dogfish, 



JVffsri.;7.f.s'(xlOOO) 

 (Gordon Walls). 



Squatina 



cells being thin and long ; the ratio of visual to ganglion cells varies 

 (152 : 1 in Efmoptenis, 14 : 1 in Myliohatis, 12 : 1 in Raja miraletus — 

 Verrier, 1930). There is no area centralis, although in some species, 

 jiarticularly the dogfish, Mustelus, the density of the visual elements 

 is increased in a round central area so as to suggest an elementary 

 precursor of this characteristic of the higher Vertebrates (Franz, 1905) 

 (Fig. 317). Only in a few particularly active species are cones found — 

 the dogfish, Mustelus, the eagle-ray, Myliohatis, and the angel-shark, 

 Squatina (Franz, 1905 ; Verrier, 1930 ; Rochon-Duvigneaud, 1943) 

 (Fig. 316). 



The optic nerve has various septal patterns and in some species 

 an cpendymal core, as in lampreys (Prince, 1955) ; like the retina it is 

 avas lar. The optic disc is small and flat and a lamina cribrosa is 



