Recherches sur les Poissons Fossiles, 



69 



sent, and consequently presented a more uniform character 

 than we find to be the case at present. 



The following are M. Agassiz's observations on the clas- 

 sifications of fishes, a subject which he proposes to enter 

 into in more detail, in a more advanced stage of his work on 

 fossil fishes : — 



All naturalists are agreed as to the imperfection of the 

 various classifications that have been hitherto proposed for 

 the class of fishes. I shall not enter into a criticism of 

 the subject here, but confine myself to an indication of some 

 of the peculiarities of the classification observed in this 

 work, leaving to ichthyologists to appreciate their value. The 

 only thing I now ask is, that it may be remembered that 

 the sketch here presented is incomplete, and that it is 

 intended hereafter, at the end of the first volume, to enter 

 into the necessary details on the subject. 



I have established in this class four orders, which I believe 

 to be the representatives of one another, and that each en- 

 joyed a particular reign at different geological epochs. Each 

 of these orders contains fishes possessed of cartilaginous 

 skeletons,* each of them contain genera with spinous rays 

 in the dorsal,f as well as other genera of which all the rays 

 are soft;]; finally, each of these orders have the apodal ge- 

 nera§ and abdominal genera, || and in each of these orders, 

 there are also thoracic and jugular fishes. We thus see 

 the divisions here proposed are not unfounded, even to the 

 ossification of the skeleton, and the structure of the vertical, 

 and the position of the double fins, or those that are placed 

 in pairs ; but recognise all these characters as in those 

 systems which have up to this time been followed. I have 

 endeavoured to find in the differences presented by the 



* As Sharks and Skaits. t As Siluridse. 



X As the sea Gobies. § As the Eels. 



II As the Carps. Thus each order represents the various leading 

 types of the entire class. — Ed. 



