522 MULLER ON THE STRUCTURE OF THE GANOIDSs, 
to the Reptiles in one or two points of their organization. I have 
never been able to convince myself that they are related to them 
more than any other fish, and that they form a transition to the 
Saurians. I find merely combinations of the properties of the 
osseous fishes and of Plagiostomi united in a third distinct 
form. The double vomer in Lepisosteus (Agassiz), and the con- 
nexion of its vertebrze by articular heads and sockets (Blainville), 
are decidedly remarkable in fish, and this certainly is a combina- 
tion of forms which are next met with in Reptiles. Nor do 
these less frequently exhibit the peculiar spinal formation of fish 
with double excavated facets, as the Ichthyosauri, Plesiosauri, and 
the fish-like Amphibia, &c., and the Proteidea, Derotreta and 
Cecilia. 'The composition of the lower jaw of Lepisosteus of as 
many pieces as in the Reptiles (Geoffroy St. Hilaire), which is 
not repeated in Polypterus, I find in a decidedly osseous fish, 
Osteoglossum. 'The reception of the apophyses of the vertebrze 
into cavities in Lepidotus, M. Agassiz considers as characteristic, 
and with this exception peculiar to the Placoids, reminding us 
of the Ichthyosauri. It'would be superfluous to compare this 
structure with that of the spine of the Cycloidei and Ctenoidei, 
as this insertion never occurs in the latter. I must remark here, 
that it actually occurs in many families of osseous fishes, as in 
the Cyprinoidei, Salmones, Esox and Elops. The only fishes 
which decidedly approximate to the Reptiles, are those which 
possess at the same time lungs and gills and perforated nares : 
Lepidosiren is to fish what the pisciform Protei are to the Am- 
phibia. Isolated affinities always occur; but we likewise meet 
with them in other orders: in the sexual organs the Plagiostomi 
agree most with the other Vertebrata, 7. e. come nearest in 
this respect to the Reptiles, and differ completely from the type 
of the ordinary osseous fishes, in their oviduct and accessory 
testicles. 
By separating the Lophobranchii, Gymnodontes, Sclerodermi, 
Goniodontes and Siluroidei, the present section of the Ganoids 
becomes very considerably reduced, perhaps to a half; neverthe- 
less the name Ganoid must be retained for the remainder as a 
subclass order of fish, not merely because this residue still con- 
tains the greater portion of the fossil Ganoids, and the separate 
families are but little, and partly not at all represented in the 
strata of our globe, but still more on account of the great merit 
which Agassiz has earned by his foundation of the Ganoids 
