1918.] R. H. WniTEiTousF. : The Cdmhil Fins of Fishr.'i. i:>.9 



during beterncerc)^ in orrlcr again to restore an external syriTmetrv of a 

 more efficient type. 



There is still another morphological question that the above evolu- 

 tionary process involves. What is the true nature of the homocercal 

 caudal fin ? Is it a true modified caudal or is it an anal fin which has 

 come to occupy a relatively posterior position ? If we examine almost 

 any heterocercal caudal fin, such as that of Acipenser or Polyodon, we 

 see that the greater part of the fin is supported by hyjDurals some con- 

 siderable distance from the end of the chordal axis. Should the axis 

 again be straightened, this portion would certainly be regarded as anal 

 fin from its very position. That the vertebral axis of Teleosts, and it 

 may be of all fishes, has been very much reduced in length is certain ; 

 the continuation of the spinal cord beyond the last centrum is sufficient 

 proof of this, and with this reduction the original caudal rays have gone 

 too, unless a few dorsal rays associated with the opisthure are the last 

 remnants. The upturning of the chorda therefore almost certainly 

 involved the bringing of a more anterior fin, an anal, into a relatively 

 posterior terminal position. 



The question as to whether separate median fins had been yet differ- 

 entiated need not seriously affect this view, for during development 

 it can be seen that the skeletal elements of the differentiated fins are 

 laid down before heterocercy sets in, and thus differentiated fins might 

 be regarded as having been established during the protocercal condition 

 in some cases at least. Differentiation, however, may not have been 

 developed in all cases, and then it is only a matter of extended growth 

 of a part of a continuous fin. Thus this question is not one of primary 

 importance since certainly dorsal, caudal and anal fins are only names 

 for portions of a once continuous series. 



The supporting elements for the caudal dermotrichia are of three 

 kinds : firstly hypurals ; secondly epurals ; and thirdly radials, either 

 dorsal or ventral. These terms are here used according to the defini- 

 tions given by me in an earlier work,^ and it may be useful to quote 

 them briefly. A hypural is defined as any hypaxial element having 

 direct connection with the chordal axis, and bearing one or more caudal 

 fin-rays distally ; an epural is the corresponding epaxial element ; a 

 radial is synonymous with ' somactid ' and ' interspinous bone.' 



One constantly sees statements which refer to hypurals as haemal 

 arches ; Sedgwick writes ^ " In all fishes the ventral part of the caudal 

 differs from the other median fins in the fact that the dermotrichia 

 (fin-rays) are supported directly by the haemal arches," Such a refer- 

 ence is typical of text-books in general concerning caudal fins, but the 

 matter is one deserving of discussion. As far back as 1851 Stannius 

 in his text-book on the Vertebrata clearly stated that the fin-ray sup- 

 ports were compound structures, consisting of arch and radial combined. 

 Ryder^ in 1884 seems to have come to the same conclusion but this inter- 

 pretation seems to have been ignored in more recent works. 



^ Loc. cit. 



2 Stvrlent.'i" TexlhooJc of Zoology, 190.5. 



^ " Evolution of tlie Fins of Fislies," Jiej). Cowm. Fish and Fis/trries, Washington, 

 1884, imblished 188(5. 



