﻿40 



GANOID FISHES OF THE CARBONIFEROUS FORMATION. 



rather a prominent projection over the mouth. I have already alluded to the remarkable 

 similarity in the form of the palato-quadrate apparatus in Palceoniscus and in Polyodon, 

 and to the fact that the levator muscle of the lower jaw must have pursued exactly the 

 same strange course in the former as in the latter. 



The opercular and branchiostegal apparatus is indeed feebly developed in Polyodon, 

 and even more so in Acipenser, though the latter has a cranial shield of closely fitting 

 ganoid plates. But here a connecting link is seen in Chondrosteus, in which, though 

 the squamation of the body is apparently in the same condition as in Polyodon, we 

 have nevertheless a well-developed opercular apparatus and a set of branchiostegal rays 

 in the form of strong, imbricating, osseous plates. The cranial buckler of Chondrosteus 

 also consists of strong closely fitting plates, as in the sturgeon. 1 



The resemblances between P alceoniscus and Acipenser are of course much less 

 prominent. But it seems impossible to avoid the conclusion that Palceoniscus must 

 accompany Polyodon in whatsoever group the latter is placed, and that therefore the 

 Palceoniscidce must be accepted, not as a family of the Lepidosteid, but of the Acipen- 

 seroid series. For certainly they belong neither to the Crossopterygii nor to the Amioidei: 

 and the reasons supposed to indicate for them a place among the Lepidosteoidei may be 

 disposed of as follows. The Palceoniscidce have rhombic scales as in the Lepidosteoids, 

 but rhombic scales are also found in Polyodon, though confined to a very limited portion 

 of the body ; the form of the scales can also hardly be taken as a subordinal characteristic, 

 seeing that in such closely allied Crossopterygians as Megalichthys and Rhizodus the 

 scales are in the former rhomboidal, in the latter cycloidal. The fins are fulcrated, but in 

 some Palceoniscidce {Thrissonotus, Cosmolepis) fulcra have not been discovered on any of 

 the fins save the caudal, and they are, though exceptional in form, undoubtedly present in 

 some Crossopterygii [Gyroptychius, Osteolepis). The paired fins are non-lobate, so are 

 they also in the Acipenseroids and in the Teleostei. But in my opinion outweighing these 

 resemblances with the Lepidosteoids, the Palceoniscidce have, in common with the Acipense- 

 roids, the completely heterocercal tail, the excess in number of the rays of the median fins 

 over their supporting ossicles, the well-developed infraclavicular elements in the shoulder- 

 girdle. The opercular and branchiostegal apparatus resembles that of Chondrosteus as 

 much as that of the Lepidosteoids. It is true that these characters of the tail, of the fin- 

 rays, and of the infraclaviculars, occur also (especially the two latter) in the Crossopterygii ; 

 but from this group the Palceoniscidce are widely separated by the structure of the paired 

 fins, and the replacement of the jugular plates by imbricating branchiostegal rays. 



Are the Palceoniscidce, however, the only group of fossil fishes which it is advisable 

 to transfer to the Acipenseroidei ? There is one other curious series of extinct fishes, 

 which have been sometimes grouped along with the Palceoniscidce, at other times removed 

 from them, the details of whose structure I hope to treat of in another part of this work, 



1 See the description of Chondrosteus by Sir Philip Grey Egerton, ' Phil. Trans.,' 1858. 



