THE VERTEBRATE EYE 253 



accepted as typical of rod-pigments and cone-pigments. Finally, the 

 observations of Bellairs and Underwood (1951) support the view that 

 snakes were derived from burrowing lizard-like ancestors. 



In the present state of our knowledge the problem, which raises 

 questions as difficult as they are interesting, is unsolved. 



Combinations of these visual elements are frequently encountered 

 in several classes of the vertebrate phylum. Double rods are rare 

 (geckos and some nocturnal snakes). A second rarity is the twin 

 cones — a fusion of identical elements — which are found only in Teleo- 

 steans (Figs. 347-8). Double cones are more common, appearing first 

 in Holosteans and occurring in every other class. ^ Typically they 

 represent the fusion of two unlike elements, the principal resembling 

 the single cones in the same retina and the accessory, generally of a 

 simpler type, rarely containing an oil-droplet but frequently an 

 unusually large paraboloid. In Amphibians, Saxen (1954-56) has 

 brought forward evidence that the double visual elements represent 

 not the fusion of two cones as has generally been thought but the fusion 

 of a rod with a cone. Triple " cones " (perhaps two cone-like com- 

 ponents with a third rod-like component, Saxen, 1953) occur in some 

 Teleosts (trout) and Anurans, while quadruple elements have also 

 been described in the minnow, Phoxinus (Lyall, 1956).- 



The origin of these double cells has given rise to some controversy. The 

 sceptical view that they were histological artefacts was put forward by Koganei 

 (1884) and has been maintained by such writers as Cameron (1911) and Roze- 

 meyer and Stolte (1930). There seems no doubt, however, that they do exist. 

 Dobrowolsky (1871) put forward the hypothesis that they resulted from the 

 incomplete division of single cones, a view upheld by Howard (1908) and Franz 

 (1913). On the other hand, Detwiler and Laurens (1921), finding that double 

 cones appeared during development at a stage when no further cell-divisions 

 took place, suggested that they were produced by the fusion of two separate 

 progenitors ; this view has been well substantiated in Amphibians by Saxen 

 (1954-56). 



The physiological significance of the association of more than one visual 

 cell is not understood. The fact that the dendrites of the two components sink 

 to different depths in the outer plexiform layer suggests some difference in func- 

 tion (Cajal, 1893 ; Greeff, 1898), while the observation of v. Genderen-Stort 

 (1887) that photomechanical reactions are confined to the principal elements 

 points to the probability that the accessory element has a subsidiary function. 

 Whether this is visual or metabolic, the two elements living in symbiosis (Howard, 

 1908 ; Franz, 1913), is conjectural. 



Apart from the fundamental structure of the retina the other 

 ocular tissues, although in general conforming to the vertebrate plan 

 seen in man (Fig. 268), show considerable variations depending upon 



^ Many Teleosts, Protopterus, Amjjhibians, Reptiles except some snakes, Birds, the 

 platypus, and Marsupials. 



2 See also footnote, p, 364. 



