THE DEKMAL FIN-RAYS OP FISHES. 471 



It is not till some time after the ceratotricbia have begun 

 to develop that the overlying denticles make their appear- 

 ance. In this paper it is not proposed to enter into a 

 detailed description of the development of these structures; 

 but I may point out that in its initial stages the denticle 

 appears as a mere thickening of the basement membrane 

 secreted by the mesoblast cells^ which group themselves 

 below its place of origin (figs. 11 and 7 cZ). The continuity 

 of the denticle (the dentine cone) with the membrane is 

 retained throughout its development. 



By the time the denticles begin to develop, the cerato- 

 trichia are separated from the epidermis by a considerable 

 layer of fibrous connective tissue; audj moreover, since new 

 rays are laid down at the growing edge of the fin, we may 

 find several layers of ceratotricbia. As the fish grows older, 

 the proximal thick ends of the ceratotricbia come to lie rela- 

 tively closer to the median endo-skeletal fin-supports. Fig. 1 

 shows the relations of these structures in the adult. 



Ceratotricbia occur in all living Elasmobrancbs, although 

 they may be much reduced in size and number in the highly 

 specialised fins of tbe Rajidte. So far as is known, no other 

 kind of dermal ray whatever exists in these fish. The fossil 

 Elasmobranchs, in this particular, seem to resemble in every 

 essential their modern relatives. 



Holocephali. 

 Chimsera and Callorhynchus are provided with dermal rays 

 of exactly the same structure as those of the Elasniobranchs. 

 The presence of typical ceratotricbia iu the Holocephali con- 

 firms the view that this group is merely a highly specialised 

 offshoot from the Elasmobranch stem. 



Teleostei. 

 Passing now to the higher bony fish, we find quite a 

 different state of things. The general structure and relations 

 of the dermal rays of these forms are so well known that only 

 a brief description is here necessary. 



