FOSSIL FISHES. 303 



The osseous fishes proper appear for the first time in Cretaceous 

 strata, being immediately preceded by the teleostoid group of ga- 

 noid fishes of the family Leptolepidas, which effects a passage to 

 them. Indeed, by many authors the genus Leptolepis and its 

 Jurassic allies (Caturus, Thrissops, &c.) are classed with the for- 

 mer, and placed near the herring, with which they appear to have 

 been most nearly related. Although both the physoclist and phy- 

 sostome types, or those in which the swimming-bladder is closed 

 off from, or remains connected with, the gullet, appear very nearly 

 simultaneously in the same deposits, and consequently by their 

 occurrence give no evidence as to their respective antiquity, there 

 can be no question that the physostome is the more ancient type, 

 the severance of the bladder in the physoclists being the result of 

 the disuse of parts. This is further j^roved by the existence of a 

 connecting air-bladder among the recent (and, doubtless, also among 

 the ancient) ganoids. The Cretaceous teleosts belong largely to 

 existing types — Clupea (herring), Osmei-us (smelt), Esox (pike), 

 Beryx ; but it is not until the Eocene period that we find a repre- 

 sentative modern ichthyic fauna. To this, and the succeeding 

 Tertiary period, most of the more prominent existing types date 

 their first appearance. In addition to a very large representation 

 of both the arthropterous and anarthropterous forms, the Eocene 

 deposits contain remains of the Lophobranchii (Syngnathus, pipe- 

 fish), Plectognathi (Diodon, porcupine-fish; Ostracion, trunk-fish), 

 and Apodes (Anguilla, eel). 



The study of the distribution of fossil fishes renders evident 

 two important facts : First, that there has been a progressive modi- 

 fication and evolution from less to more highly organised types; 

 and, secondly, that among the almost innumerable forms of com- 

 paratively recent origin, our existing ichthyic fauna still holds the 

 v^reck of a past fauna, whose period of decline belongs already to 

 the earlier part of the earth's history. It, moreover, reveals an 

 extraordinary persistence on the part of some of the individual 

 types. The occurrence of the recent genus Ceratodus in deposits as 

 ancient as the Permian is certainly very remarkable, but it does 

 not argue, as some would lead us to believe, that this particular 

 fish has undergone no modification since the period of its first 

 introduction. The fact that its remains have been found fossil in 

 the deposits of Europe and America, whereas at the present time it 



