THE STRUCTURE OF VISUAL CELLS 45 



present in small numbers. Their numbers vary greatly in 

 different portions of the retina. In dual retinas they are 

 typically less numerous in the fundus and more numerous 

 peripherally. Thus the rod-cone ratio varies in different 

 portions of the retina. Rods are entirely lacking in the region 

 of the fovea. 



The Basis for Rod and Cone Classification 



The question as to what constitutes a rod or a cone 

 would seem to be a relatively simple matter, but the great 

 variety in the form of these cells in different animals makes 

 a general classification difficult. Structural differentiation 

 is based usually upon the form of the outer segment (cone- 

 shape vs. rod-shape). However, this does not meet all the re- 

 quirements for, as pointed out above, the cones at the fovea 

 are long cylindrical structures and look more like rods. 



Putter (1912) stressed the importance of the mode of con- 

 nection with the bipolar cell as a real basis for classification. 

 According to him cones have centripetal dendritic endings 

 whereas rods have knob-like or molecular endings, regardless 

 of their external form. This is the condition usually depicted 

 in textbooks, which, of course, is correct for the mammalian 

 and human eye, as well as for the eyes of teleost fishes. 

 However, as will be seen in Figures 17 to 22, the rods of 

 selachians, amphibians, reptiles, and diurnal birds (according 

 to Franz) also possess dendritic endings. When we find 

 visual cells, which from their general form (outer and inner 

 segments) should be called rods, possessing typical cone 

 connections (e.g. frogs, diurnal birds) there seems to be no 

 reason for calling them cones because they terminate in 

 dendrites. There is considerable reason to designate them as 

 rods on functional as well as on structural grounds. 



Diurnal lizards have cones only, but in the crocodilia, for 

 example, both rods and cones occur, structurally as well as 

 functionally. According to Putter, they would all have to 

 be regarded as cones because their centripetal terminations 

 are similar to those found in the cones of man. 



