THE PECTEN AND ITS ANALOGUES 653 



with a network* of 'hyaloid' vessels at the vireo-retinal interface, or with 

 a falciform process — an obvious physiological counterpart of the avian 

 pecten (the pecten being essentially an ectodermal imitation of the older 

 structure!). It is not too much to hope that someone will sometime 

 determine whether the retinae of the hyaloid-vessel teleosts have greater 

 requirements of glucose and oxygen than those of the falciform-process 

 species. The retention of one S N D or the other in even the most strongly 

 nocturnal teleosts may seem disturbing to our thesis; but their nocturn- 

 alities are probably all secondary, and there has been no such urgent 

 need to eliminate an S N D from a secondarily nocturnal eye as to evolve 

 one in a secondarily diurnal one. The presence of vitreal vessels in 

 Polypterus, Protopterus, and Lepidosiren may have such an explanation. 



Among the anuran amphibians, the ranid frogs alone exhibit the 

 primitive arhythmic or diurnal habit of the group, as is evidenced by the 

 persistence of yellow cone oil-droplets in those forms alone. All known 

 anurans have vitreal vessels, whose presence (in as full development?) 

 in the secondarily nocturnal toads and tree-frogs is thus a failure-to- 

 discard. It is only natural that the urodeles and cxcilians have never 

 developed such vessels. 



Turning to the reptiles, we are confronted by the paradox that neither 

 the diurnal turtles nor the nocturnal crocodilians have preserved the 

 ancestral conus papillaris in a useful condition. Its loss in the croco- 

 dilians (and in Sphenodon) makes good sense; but the turtles all have 

 many cones — some, perhaps, only cones — in their retinae. The turtle 

 rates as 'sluggish' alongside the average lizard. The latter has the conus, 

 of course ; and it would be interesting to know whether the requirements 

 of the relatively crude (though cone-rich) turtle retina are sufficiently 

 lower than those of lizard retinze to explain the difference with regard to 

 the conus. Again, among the lizards themselves, the relative size of the 

 conus does not go perfectly with diurnality-versus-nocturnality : it is 

 smallest in certain geckoes and other nocturnal lizards (Pygopus, etc,) ; 

 but there are geckoes with large coni, and the chameleons have very 

 small ones. Anyone who has ever watched the 'slow-motion' performance 

 of a true chameleon, however, should be willing to imagine that its 

 retinal metabolism may be little if any higher than that of a tortoise. 

 As for the geckoes, there are reasons for thinking that their peculiarly 

 pure-rod retinae have a physiology much like that of a pure-cone one. 



* Richer in Lepisoiteus than in Amij, according to Virchow. 



