[11] THE EVOLUTION OF THE FINS OF FISHES. 991 



etoii. It is i)iobable that we have no remains of the larvae of PalaBo- 

 zoic lishes preserved in the rocks, so that we have no means of con- 

 trasting their early phases with those of existing forms, but it is certain 

 that none of the most simple forms of the Palaeozoic fishes, in respect 

 to their skeletal structure, even approximate such a primary condition as 

 the lophocercal stage. Of modern forms, the only trait which they possess 

 in common are continuous median fins, in the first, containing perma- 

 nent rays, the latter without them, but at most provided with embryonic 

 fin rays only. When we know the larvae of Ceratodns, Polyptents, Lep- 

 idosiren, and Protopterus as well as we know that of LepidosteuH we 

 may have a more coujprehensive understanding of the main features 

 of the larvae of Palaeozoic fishes. 



The evidence in favor of degeneration of the caudal region is the ex- 

 istence of a permanent archicercal opisthure in Chimcera monsfrosa and 

 Stylephorus chordatus ; the extensive development of a temporary opis- 

 thure in Lepidosteus ; the concrescence of the hypural pieces; the ven 

 trally diplacanthous and even triplacanthous caudal vertebrae or their 

 coalesced representative, the urostyle ; the existence of hypaxial opis- 

 thural elements ; the abortion of the epaxial spines of the caudal verte- 

 brae, and finally the abortion or extreme modification of the last muscu- 

 lar somites of the caudal region. 



(5.) Homocercy. — This merely expresses the condition of epaxial and 

 hypaxial symmetry presented by the fan-shaped caudal of Teleosts, and 

 is the final term in the evolution of the growth of the rays of that fin, in 

 consequence of which the archaic symmetry of perfect diphycercy be- 

 comes again restored though the structure of the tail is heterocercal. 



(C.) GephyrocQrcy. — This type of tail appears to be normally met with 

 in only two forms, of Teleo.sts, namely, Mola and Fierasfer. The prim- 

 itive opisthure or end of the urosome in these forms is apparently 

 aborted in the first case, in the course of larval existence : in the other, 

 during jiost-larval life. As a result of this a hiatus is left between the 

 epaxial and hypaxial rudiments of the median fins, and in the center of 

 this hiatus the axial column ends abruptly as if cut or bitten ofi". The 

 hinder hypaxial and epaxial tissues concerned in the formation of rays 

 and their snp])orts are then approxmiated and developed later than 

 tbe other iiicdian fin-rays, and the interval so bridged by a secondary 

 process of development leads to the lormation of what we may call a 

 gephyrocercal* tail. 



HI. — On the degeneration of the posterior part of jhe 



AXIAL SKELETON OF FISHES. 



There is undonbred evidence of the extensive degeneration and con- 

 sequent modification of the caudal part of the vertebral axis of fishes. 

 This is apparent upon making a study of the tail of the entire series of 



" yecpvpa, bridge; and uipKoi, tail.) 



