1 843-44.] POISSONS FOSSfLES. 215 



reduced, as I have said above, to only two genera, then 

 formed the majority of the population of the seas. 

 These same fishes present in their history a very re- 

 markable fact. Until the Lias epoch, all the Ganoids 

 possessed on the superior part of the tail a lobe formed 

 by a prolongation of the vertebrate column. But from 

 the Lias, on the contrary, all had a tail formed as that 

 of the osseous fishes of the present time ; that is 

 to say, the vertebrate column stops at the base of 

 the tail. 



"It is, therefore, possible to divide the palaeonto- 

 logical history of the fishes into three periods or epochs. 

 During the first, extending from the Silurian to the 

 Trias, the faunas are composed of Placoids and Ganoids 

 with the vertebrate column prolonged to the upper lobe 

 of the tail. In the second, which corresponds to the 

 Jurassic epoch, we find the Placoids and Ganoids, with 

 the ordinary tail. In the third, which began during the 

 Cretaceous epoch and continued in our modern period, 

 the Placoids, the Ctenoids, and the Cycloids form almost 

 entirely the ichthyological population of the world. 

 Hence, if a geologist found a Ganoid with a prolonged 

 tail, he could conclude that the strata in which he found 

 it belonged to the first period ; a Ganoid with an ordi- 

 nary tail would indicate, with sufficient certainty, that 

 the group of strata belonged to the second epoch ; and 

 so on. 







" It is easy to understand the interest created by a 

 work, the aim and result of which are to demonstrate 

 such remarkable laws, more especially when the proofs 

 are based on an incredible number of facts and observa- 



