[13] THE EVOLUTION OF THE FINS OF FISHES. 993 



Shaw's figure shows that there are absolutely no hypaxial rays devel- 

 oped, nor does the accompanying description mention them. The form 

 therefore contrasts morphologically with Oymnotus, which has no dor- 

 sal or epaxial system of fin-rays. 



Dohrn* has lately published some interesting and important specu- 

 lations upon the phylogenetic origin of the paired and unpaired fins. 

 He lays particular stress upon the presence of a postanal gut in fish 

 embryos as determining the abortion of the preanal unpaired median 

 fin-fold, and its division into two ventro lateral folds, which he thinks it 

 conceivable were derived from the parapodia of a worm-like ancestor. 

 The abortion of the postanal section of the gut in vertebrate embryos 

 has permitted the approximation or coalescence medially of these primi- 

 tive lateral folds or metameric series of papillae from which parapodia 

 have been evolved in worms, and which led to the diflerentiatiou of 

 paired fins in their vertebrate offshoots. Since Dohrn's investigations 

 have undoubtedly shown the archi pterygium and cheiropterygium 

 hypotheses of Gegeubaur and Huxley to be utterly untenable, as had 

 been pointed out before by Balfour, there are still some difiiculties in 

 the way of Dohrn's own conclusions, which seem to me to be serious 

 enough to be worthy of passing notice. What is here referred to is the 

 presence of preanal unpaired fin-folds in certain forms of fish embryos, 

 as in Alosa, FomolobuSj Salmo, Coregonus, Cybiiim, Lepidostens, &c.,t 

 and which would indicate that the presence of the intestine in these 

 instances had not exerted the influence which he has suggested. In 

 Alosa, Pomolobus, and Leindosteus, this i)reanal fold, contrary to what 

 is found in Elasmobranchs, is especially well developed. It is there- 

 fore evident that Dohrn's view of the origin of the median folds does 

 not hold fo^ all of the Ichtliyopsida without some modifications. 



Dohrn is right, however, in the view that what I have called the uro. 

 some is a structure secondarily developed, for, as he points out, we 

 know from the researches of Goette on Bomhinator, Kowalewsky and 

 Hatschek on Amphioxus, and Balfour on the Elasmobranchs, that 

 there exists an open communication between the primitive neural tube 

 and the intestine by way of a neurenteric canal and postanal gut, and 

 that while such a relation is obscured somewhat in Marsipobranch and 

 Teleostean embryos, traces of it nevertheless exist even in those types. 

 The tendency to develop an anal opening considerably anterior to the 

 end of the tail after the archicercal or worm-like condition is passed 

 over, while the postanal section of the gut atrophies instead of devel- 

 oping a quite terminal anus after the manner of many worms, is very 

 significant, and would indicate that we have here an instance of degen- 

 eration manifesting itself at a very early period. This process of de- 

 generation by which in reality an advance is made upon the more prim- 



* Studien zur Urgeschichte des Wirbeltliierkorpers, Pt. VI. Mitth. zool. Sta. zu 

 IJeapel. V. Hft. I, 1884, pp. 174-189. 



+ A contribution to the Embryography of Osseous Fishes, § 18, p. 67. 

 H. Mis. 68 63 



