168 S. R. DETWILER 



remain for several days. Further dehydration, consuming at 

 least 2 days, the lens being removed after the eyes had been in 

 95 per cent alcohol for several hours. For infiltration the chloro- 

 form parafin method was employed, paraffin melting at 52° C. 

 being used for imbedding. Sections were cut 8 to 10/^ thick, 

 stained in Ehrlich's haematoxylin, followed by eosin. In a few 

 cases, in an attempt to secure a quicker and perhaps more per- 

 fect fixation, the lens was removed before the eyes were dropped 

 into the fixing fluid. It was found, however, that this method 

 caused shrinkage and folding of the retina to such an extent, at 

 the same time producing no better fixation, that it had to be 

 given up. 



ANATOMICAL 



Before we proceed to consider the results of the comparison 

 of light and dark eyes it will be best to give a brief account of 

 the anatomical relations of the species with which we are work- 

 ing. Considerably more has been done by previous investiga- 

 tors on the morphology of the reptile retina than on its physi- 

 ology, and it will be well to review briefly the results of this 

 work. 



Schultze ('66 and '67) noted that in Emys europaea, L. viridis, 

 L. agilis and L. muralis there were no rods. From Hulke's ('67, 

 p. 94) incomplete description one might assume that rods were 

 to be found in the retinae of, among others, Testudo graeca, 

 Emys europaea, Chelone midas, Lacerta viridis and Anguis fra- 

 gilis, though, as Krause pointed out later, these 'rods' are more 

 correctly to be considered as cones without oil drops. Heine- 

 mann ('77, p. 423) who examined the retinae of several Mexican 

 species of tortoises concluded that, if the form of the outer seg- 

 ment be taken as the criterion, then rods as well as cones can be 

 distinguished in the retina of Chelonians. That in lizards, how- 

 ever (p. 431) no elements with rod-like outer segments can be 

 distinguished. In the Geckos, however, (p. 434) it is doubtful 

 whether the visual elements are rods or cones. 



Angelucci ('78, p. 371) found that rods are entirely lacking in 

 Testudo graeca, L. agilis, L. muralis and in L. viridis. Boll 



