252 THE EYE IN EVOLUTION 



possible in a uniform retina by extraction of the photopigments, the 

 method is inapphcable in a duplex retina since the concentration of 

 pigment is not sufficiently great to allow the histological demonstration 

 of vitamin A even by methods so delicate as fluorescence -microscopy 

 (Stern, 1905 ; Hopkins, 1927 ; Walls, 1935 ; Stenius, 1940 ; Greenberg 

 and Popper, 1941 ; see Saxen, 1954 ; and others). 



There is no doubt, of course, that fundamentally the two elements 

 are alike and it is obvious that within the vertebrate phylum many 

 transitional forms between the two exist ; between these, wherein the 

 anatomical difficulties of differentiation occur, a sharp distinction may 

 be illegitimate. Both are probably derived from the same primitive 

 ancestral cells, and it has been suggested that cones are transformed 

 into rods during development (Steinlin, 1868 ; Bernard, 1900-3 ; 

 Cameron, 1911), a theory, however, which later evidence has questioned 

 (Detwiler, 1943 ; Birukow, 1949 ; Saxen, 1954) ; similar criticism has 

 been directed to the theory of Walls (1934) that the one may be 

 transmuted into the other in phylogeny. 



Walls's theory — ingenious, attractive, fanciful and mvich criticized — is that 

 the primitive visual cell of Vertebrates was a cone and that therefrom rods were 

 evolved as a transmvitation-form with a view to increasing sensitivity with the 

 development of rhodopsin — presumably first in deep-sea types. The brilliance 

 of illumination on land allowed many reptiles (diurnal lizards) to retain a pure- 

 cone retina ; their adoption of nocturnality as a protective measure forced some 

 species (Xavtusia) to develop a transitional rod-like element, and the adoption 

 of complete nocturnality by most geckos led to the transmutation into rods. 

 The visual elements of many snakes are similarly interpreted, the cones of some 

 secretive nocturnal types showing a structvnal or a complete transmutation 

 into rods, in the first case withovit, in the second with rhodopsin. 



It is interesting that recent research has to a considerable extent 

 confirmed this somewhat revolutionary view. That such a trans- 

 mutation had in fact occurred is suggested by the finding of Crozier and 

 Wolf (1939^ that the rod-retina of the gecko, Sphcerodactylus, has a 

 critical fusion frequency similar to that obtained in the turtle with 

 its predominantly pure-cone retina. The same conclusions could be 

 said to follow the finding of Underwood (1951) that some primitive 

 Jamaican geckos had oil-droplets in their rod-like receptors. The 

 peculiar pigment with its unusual absorption curve for a substance 

 based on vitamin A^ (maximum at 524 m^ti) described in certain 

 geckos by Denton (1953) {Gekko gekko) and Crescitelli (1956) {Phyl- 

 lurus) again could perhaps be interpreted as an attempt to transform 

 ancestral cones into rods, as if they were unable to re-invent rhodopsin 

 for lack of the suitable protein, and had thus been forced to conjugate 

 their ret;;,enei as a chromophore and produce a pigment with an 

 absorpti intermediate in spectral position between those generally 



