AND ON THE CLASSIFICATION OF FISH. 529 
at present are separated from these families. Heterocercy, how- 
ever, anatomically, insensibly passes into Homocercy. Although 
many Ganoids which form one limit have no fin-rays above the 
extremity of the vertebral column, still these occur in the Stur- 
geons, for before the elongated upper pieces of the tail-fin 
reach its extremity, inarticulate cylindrical spines, which form 
the summit of this tail-piece, join on to other unarticulated fin- 
rays, which are placed upon the chorda, and are of a similar 
structure to the lower rays of this piece. No sharply-defined 
character can be drawn from this structure, for as the upper 
tail-piece becomes successively shortened, it passes into a homo- 
cercous tail. The heterocercy of the Plagiostomi disappears just 
as imperceptibly. If we examine a heterocercous Shark, we tind 
beneath the skin, above the vertebral column, just such a beard 
of a fin, consisting of filiform cartilaginous threads as beneath 
the vertebral column, only shorter. 
In concluding these remarks on the Ganoids, the question 
occurs as to what subdivisions, sub-classes or orders are to be 
formed of the other fish. 
Cuvier, at the conclusion of the first volume of his Hist. Nat. 
des Poissons, in his remarks on the methodic division of fishes, 
comes to the conclusion, that the formation of families presented 
few more difficulties, but that there was still a want of important 
characters by which to arrange the families satisfactorily into 
large sections. “ But to arrange these genera and families in 
some order, it would be necessary to take a few important cha- 
racters whence large divisions would result, which, without dis- 
turbing their natural relations, might be sufficiently precise not 
to allow of there being any doubt as to the place of each fish; 
and we are not yet in a condition to do this in a sufficiently de- 
tailed manner.” I think that we have at present arrived at this 
point of our knowledge, and I will now attempt to develope the 
great subdivisions of fish according to their internal and external 
characters, and to limit them by acute definitions. 
The section of the Chondropterygii, first instituted by Artedi, 
confirmed by Gronovius and adopted by Cuvier, appears at first 
sight to be an unnatural combination of the most varied families ; 
in it we find the Sturgeons, the Chimere, the Plagiostomi, and 
Cyclostomi united. It cannot be doubted that in this section 
the most perfectly organized fish, which hence approximate to 
the Reptiles, and the most imperfect, the Cyclostomi, as the Petro- 
