[39] THE EVOLUTION OF THE FINS OF FISHES. 1019 



or that next the epidermis, or simply between the latter and the under- 

 lying mesoblast, especially the distal and least modified portions of the 

 rays at the margins of the fins. This relation of the rays to the primary 

 embryonic layers has led writers to speak of them as exoskeletal in con- 

 tradistinction to the other portions of the skeleton, named endoskeletal, 

 for the reason that the stratum which they at first overlie subsequently 

 becomes the corium and a part of the adult skin. But it so happens 

 that very young or larval fishes have no corium or true deep layer of 

 the skin developed, and that proximal i)ortions of the rays soon become 

 quite as deeply imbedded in the mesoblast, even below what can ever 

 become corium, as a considerable part of the clavicle, for example, the 

 membrane of which also lies at first very close to the epidermis in lar- 

 val fishes. It is thus made evident that the distinction between the 

 rays as exoskeletal as distinguished from the other hard parts called 

 endoskeletal is really a forced one. Furthermore, phylogeny and em- 

 bryology both indicate that the embryonic fin-rays are in reality the 

 membranous representatives of the more numerous rays once possessed 

 by the most ancient fishes. 



The composition of the embryonic rays is the same chemically as that 

 of the membranes in which membrane bones are formed, as is shown by 

 the way in which they stain. A good many of the embryonic rays also 

 atrophy, especially the ends of those lying at the bases of the fin-folds, 

 in the intervals between the points where the permanent rays appear. 

 This last fact is a further illustration of the common occurrence of the 

 reduction of the number of homologous parts with the advance in 

 specialization or evolution. The embryonic fin-rays are never distinctly 

 developed as fibrils till the mesoblast has proliferated into the epiblas- 

 tic folds from which the fins are developed, which is quite enough evi- 

 dence to prove that they have as much right to be called endoskeletal 

 as any other portion of the skeleton which develops in membrane im- 

 bedded in mesoblast. 



The superficial or dermal position of the rays is, however, an impor- 

 tant character, and one which is in striking contrast with the position 

 of the axial or cartilaginous skeleton, but not so important when con- 

 trasted with the position of the points of origin of many membrane 

 bones, which develop in a quite superficial position, such as the dentary 

 and the ujiper j)art of the clavicle. 



Another character, which is of some importance, is the proximal ex- 

 tension of the true rays. They uniformly terminate near the edge of 

 the fins distally and proximally just over either side of the distal end of 

 the axial skeleton in both the median and paired fins. Their mode of 

 connection with the axial skeleton is also approximately the same in 

 both kinds of fins ; that is, the distal nodular elements or actinophores 

 of the axial or basal cartilages become'invested on their right and left 

 sides by the basal ends of the right and left halves of the rays when 

 the latter have been completely differentiated. In the case of those 



