36 H. M. BERNARD. 



limbs of the rods quite distorted by tlie intrusion of a cone, 

 the base of which is swollen into a large clear vesicle. Such 

 figures as these might be multiplied indefinitely, and in each 

 case with some slight but instructive variation in detail. 



In the toad I have found outer limbs of rods, in stage r^. 

 not much more than 3 ;u in thickness, and in striking contrast 

 to the thick outer limbs of the fully developed rod. 



Uro deles. — The conical shape attributed to the rods 

 of the Triton by Max Schultze led him to believe that, 

 in the eyes of gilled forms, still more conical and probably 

 transition forms between rods and cones would be dis- 

 covered. In Triton he found the conditions such that if 

 the rods had been only a little thinner and the cones a 

 little thicker, the two would pass into one another.^ He 

 would not at all accept Steinlin's suggestion that, in general, 

 the two structures hardly admitted of sharp distinction.^ 

 Admitting their common origin, transition forms were, he 

 thought, rare, and the rods were certainly the primary 

 simpler structures, out of which the cones were developed 

 later as more highly specialised organs. 



My own observations have led me (perhaps erroneously) to 

 lay no great stress upon the slightly conical shape sometimes 

 noticed in the thick rods of Urodeles.^ The undoubted tran- 

 sitional stages which can be found between the rods and 

 cones speak for themselves, whether this very slight form- 

 feature has any value or not. Numerous as they are, earlier 

 investigators were not prepared to see them, and missed 

 them accordingly. For example, diagram fig. 8 shows a series 

 of elements chosen from a few sections through the retina 

 of an adult axolotl. They might be multiplied manifold. 

 They have been arranged according to the lengths of the 

 different parts. At a glance they can be seen to form a 

 developmental series. We thus see that longitudinal measure- 



1 ' Archiv f. mikro. Anat.,' Bd. iii, 1867, p. 238. 

 « Ibid., Bd. iv, 1868, pp. 10 and 22. 



^ In microscopic sections it may be due to the plane of the section not 

 passing strictly longitudinally through the thick rod. 



