[41] THE EVOLUTION OF THE FINS OF FISHES. 1021 



luent of spines exaggerated in thickness by synostosis with su(;h supor- 

 iicia] hard or bony "Structures. 



The term liorn or horny as applied to the embryonic tin-rays is not 

 justified even in the case of Chimcera and Geratodus, or in the other 

 instances where the embryonic rays retain their primitive cliaractei', 

 as, for example, in the atlipose fins of Sahnonoids and Nemutognatld. 

 In Chimcera and Ceratodus we might naturally expect to find that the 

 fin-rays exhibited embryonic features, when we bear in mind how embry- 

 onic and cartilaginous the skeleton has remained. Geratodus seeujs, in 

 fact, as if it were an embryo Teleost which had been permitted to lose 

 its yelk-sack and grow large, meanwhile losing none of its embry- 

 onic skeletal characters, though acquiring others, such as a more dif- 

 ferentiated resj)iratory system, but with the development of its tail 

 airested at a point nearly coincident with the end of the lophocercal 

 the stage of the Salmon. Such an arrest of development gives to 

 fin-skeletons of Dipnoi and Holocephali their peculiar traits, which 

 so closely parallel the embryonic condition of Teleosts, so that their 

 fin-rays, while partaking apparently of a horny nature, in reality have 

 no very close resemblance to horn. x\s already stated, the horny fibers 

 (embryonic rays) are similar to perichondrium histologically 5 it follows 

 that the rays of Dipnoi and Holocephali are simply this perichoudrium- 

 like substance, which has grown in volume and suffered little or no 

 calcification. Another reason why the rays of these forms, as well as 

 those of embryos, are not comparable to horn is that the epidermis has 

 little or nothing to do with their formation, while truly horny struc- 

 tures, such as nails and hairs, arise, in great part at least, by the di- 

 rect cornification of the epidermis 



The persistence of embryonic rays in the adipose fins of Physostomes 

 is doubtless correlated with that other more embryonic condition of 

 theirs, namely, the possession of an open pnenmatic duct, which shows 

 them to be less highly difiterentiated than the Physoclists, a group 

 which, it should also be borne in mind, embraces the most specialized 

 Teleostei, such as the Lophobranchii^ Hemihranchii^ and Plectof/nathi, 

 the latter embracing the Moloidea, in which the caudal skeleton is more 

 highly specialized than in any other fishes. 



VIII. — Special 3iodifications of the development of the fins. 



In order to bring forward some of the forms of development of the 

 fins or fin -like organs which it is difiicult to account for, it has been 

 thought advisable to devote a few paragraphs to certain modifications 

 which are characterized by their exaggeration or extremely si)ecialized 

 mode of development. 



The development of Fierasfer, which has been Avorked out by Emery, 

 illustrates in a striking manner the statement that an extremel}' special- 

 ized but transitory structure may be developed in connection with the 

 anterior part of the dorsal fin-fold. On first hatching an interruption 



