1052 REPORT OF COMMISSIONER OF FISH AND FISHERIES. [72] 



opithural piece has apparently been left behind, as the chorda has 

 bent upwards, and the seven hypural cartilages have been carried with 

 it, their proximal ends having been at the same time crowded forwards, 

 so that seven of them correspond to but three indistinctly marked seg- 

 ments of the urostyle. 



The total snppression of the caudal rays of Jlippovampun, as shown 

 in the figure of the tail of its larva in Fig. 3, PI. IV, is due to the new 

 function acquired by the tail as a prehensile organ. How marked the 

 influence of the atrophy of a part in the adult may be in determining 

 its failure to develop in the young imij be gathered from Fig. 2, Tl- U? 

 of the tail of an embryo of Alosa. Here traces of five hypural carti- 

 lages have shown themselves, while iniront and behind them none have 

 yet ajjpeared. On the upper side of the chorda no epurals have ap- 

 peared, even in relation to the same segments which bear the rudiments 

 of hypural cartilages, but the archaic gephyrocercal condition is clearly 

 expressed even here, for we may note that the embryonic rays are con- 

 tinued round the end of the chorda as a confluent series. This condi- 

 tion we may denominate protopterygian r/ephyroccrcy. 



Turning now to the form in which gephyroeercy has been expressed 

 in the most pronounced way, we need not do more here than call atten- 

 tion to the development which it attains in Mola, as shown on Plate 

 Vin, which illustrates how the hiatus between tlie dorsal and anal 

 seems to be filled up by a new outgrowth of rays, as indicated by the 

 stages covered by Figs. 1 to 5, while a no less remarkable metamor- 

 phosis occurs iu the passage from the condition shown in Fig. 5 to that 

 of Fig. 9, or that of the adult. Xo evidence of the existence of hetero- 

 cercy at any stage of the development of 3[ola is in existence, unless 

 we may infer that its Plectognath ancestry was heterocercal. Gephyro- 

 eercy, here as iu JiJchiodon, was directly deve]oi)ed, if we may legitimately 

 infer that Fig. 1 is really the young of this or a closely allied form. 



The constant tendency in the evolution of the caudal skeleton of 

 fishes has been to either indirectly or directly abort the termination of 

 the chorda or axial skeleton, and in the first instance to develop an 

 upbent urochord, as in Elasmobranchs, with epural and hypural pieces 

 resting on the whole of its upper and under faces ; this same urostyle 

 in higher forms became more and more shortened, reduced, and in- 

 cluded bj* adjacent structures ; while, in the last instance, the hinder 

 terminus of the chordal or vertebral axis is wholly suppressed either in a 

 very early larval stage {Mola), or persists for a longer or shorter post- 

 larval period, as in EcModon and Fierasfer. The tendency of all of 

 these processes, however, is to bring about an approach towards a con- 

 fluent series of epural and hypural arches or spines. The development 

 of heterocercy always tends also to suppress the last epaxial arches, as 

 iu Amiurus and Salmo, so that if gephyrocercj' were iu these cases de- 

 veloped by an absorption of the urostyle, the supports of the caudal 

 finrays and the rays themselves would be hypaxial in their jnorphQ- 

 logical relation to the caudal axis. 



