AND ON THE CLASSIFICATION OF FISH. 527 
among the Acanthoidei, from which it appears to differ both in 
the absence of the spines as also in the possession of the fulcra. 
I consider the separation of the Lepidoidet and the Sauroidei 
as artificial. But among the number of genera contained in 
them, there are several which have evident affinities to one 
another, and might afford grounds for separation. Agassiz 
himself has lately commenced so doing, as the Acanthoidei, 
Cephalaspides and Dipteri are principally taken from the Lepi- 
doidei. But I do not know how it is possible to distinguish 
the remaining Lepidoidei from the Sauroidei by essential cha- 
racters. It appears to me that if the Ganoids be formed into one 
family, they must agree in the condition of the spinal column as 
to whether it is osseous or its central portion cartilaginous. It 
then appears to me that those fossil Ganoids which are always 
evidently without fulcra to the fins belong to one group, and 
those in which they are constantly present to another. Among 
those Ganoids which are furnished with fulcra at the anterior 
margin of some or of all the fins, there are again essential dif- 
ferences in the structure of the fulcra, and such as appear to me 
systematically important. In the examination of well-preserved 
specimens I obtained the following results. 
When the summit of the elongated upper tail-piece is covered 
with spines, they always appear to form a single row as far as 
the extremity; this is the case in the Sturgeons, and also in 
Paleoniscus and Acrolepis. The appearance of the fulcra in the 
first tail-fin of a heterocercous Ganoid does not necessarily im- 
ply that the anterior margin of the inferior piece and other fins 
should possess fulcra, for they are absent in the Sturgeons. The 
fulcra on the entire summit of the fin, where there are no rays, 
must be considered merely as a covering of scales in general, 
not as the fulcra of the fin-rays; hence a heterocercous Ganoid, 
which possesses simple fulcra on the summit of the elongated 
upper tail-piece, may be furnished with a double row of fulcra 
at the anterior margin of the inferior piece, as I think I have 
found in Paleoniscus and Acrolepis (A. asper). There are genera 
of fossil Ganoids the anterior margins of the fins in which are 
covered with a single row of fulcra as far as the extremity, in 
which case they are bifid fulcra with simple spiny points. Ac- 
cording to Agassiz’s description of a row of acute pieces placed 
along the upper and lower margin of the caudal fins in Dapedius 
punctatus (p. 194), Dapedius helongs to this place. I find a single 
