THE EVOLUTION OF FINS. 



29 



arches (Fig. i), while those of the abdominal region are exactly twice 

 as numerous as the arches (Figs, i, 2). 



Again, it is a well-ascertained fact in Palaeontology that the 

 median fin-supports of the more primitive fishes are all divided by 

 tranverse joints into two or three pieces, as shown in the Pleuracan- 

 thidae (Figs, i, 2). The distal piece supports the delicate dermal hair- 

 like rays (when present), which are eventually replaced in higher 

 fishes by typical fin-rays. As the dermal fin-rays are developed, 

 the two or three vertical rows of fin-supports gradually become 

 reduced to one ; and if not fused together the remnant of this series, 

 instead of retaining its association with the neural arches, finally 

 becomes correlated numerically with the series of dermal rays it 

 supports. The onl}' exception to this rule known to the writer occurs 



Fig. I.— Portion of Trunk of Xcnacanthus decheni, from the Lower Permian of Bohemia 

 ch. Position of notocliord; /. Anal fin; p, r. Dorsal and ventral portions of caudal fin. 



After A. Fritsch.^ 



in the caudal fin of some Crossopterygii {e.g., the Coelacanthidae), 

 where the fin-supports are equal in number both to the neural and 

 haemal arches and to the dermal rays. 



The atrophy, or " shortening up," of the endoskeletal fin-supports 

 seems to be the invariable result of specialisation in the median fins ; 

 and all the higher types of fishes with well-developed fin-rays exhibit 

 the numerical correspondence of the supports, except where 

 secondary fusion of these supports has taken place. The dorsal and 

 anal median fins of such fishes, when not a continuous fold, owe 

 their actual extent to the disappearance, from some unknown cause, 

 of the rest of the fin-membrane, of which the rudimentary series of 



^ We are indebted to Dr. Anton Fritsch for the loan of Figs, i, 2, 3, 7 and 8, from his well-known 

 work on "The Fauna of the Bohemian Gas Coal." 



