AND ON THE CLASSIFICATION OF FISH. 503 
facilitates their distinction, it may be advantageously used. On 
the other hand, an artificial distinction becomes doubtful to us, 
if we deduce from it details of the age and development of fami- 
lies, as, for instance, that no fish belonging to the family of the 
Lepidoidei could last until the present period. The Lepidoidet 
are also disordered by the genus Lepidotus, the teeth of which 
are far removed from the characters given to the families. It 
also differs from the other Lepidoidei in possessing a perfectly 
bony vertebral column, but it does not appear to belong to the 
Pycnodontes from the teeth. It is related to the Lepisosleus at 
present existing, both in the double rows of fulcra of the fins, 
and in the bony vertebral column. 
The differences of the living Ganoids alone are perfectly ac- 
cessible. This is so much the more important, as the two still 
existing ones, Lepisosteus and Polypterus, which are placed 
among the Sauroids, are so totally different from one another 
in external and internal structure, that they merit more than 
any one of the fossil ganoid genera to serve as types of particu- 
lar families, as will appear from their anatomical structures. 
Finally, M. Agassiz has himself observed this difference in the 
osteological analysis of these fishes, and remarked that he was 
inclined to place them in separate families. Considering the 
perfect examination which these two fish admit of, and the marked 
differences which they present, I think that no two ganoid ge- 
nera, possessing the structure of their scales, are more certainly 
distinct from one another. 
In the conclusion of his great work, and in his recent mono- 
graph of the fossil fish of the old red sandstone, M. Agassiz has 
removed a large number of genera from the Lepidoidei, and 
formed them into separate families, as the Cephalaspides, Acan- 
thodei, the true Lepidoidet and the Dipterous Sauroidei; this 
appears to me a decided advance. I have no doubt that the 
greater number of the fossil fish described and figured by Agas- 
siz, and considered as Ganoids, really belong to a large order 
which includes Lepisosteus and Polypterus, and ranks as a group 
with the rest of the osseous fishes, the Selachit and the Cyclo- 
stomi; but I have never been able to convince myself that the 
other families of existing fishes enumerated among the Ganoids, 
the Loricarini, Siluroidei, Lophobranchii, Sclerodermi and Gym- 
nodontes, belong to the Ganoids. Agassiz has in some degree 
perceived the difference of these fishes from the Ganoids of 
