518 MULLER ON THE STRUCTURE OF THE GANOIDS, 
from a pseudo-branchia, and may occur at the same time with 
it; many have also blowing-holes, as in the Plagiostomi. 
They have several valves in the arterial trunk like the latter; 
their ova are conveyed from the abdominal cavity by tubes; 
their optic nerves do not pass over one another in a cruciform 
manner; their intestine often contains the spiral valve of the 
Plagiostomi; they have a swimming-bladder with an air tube 
like many osseous fishes; their skeleton is either bony or partly 
cartilaginous ; and their ventral fins are abdominal. 
When, however, we form those characters only which are ab- 
solute and never absent into a definition, the Ganoids are in fact 
those fishes which have numerous valves in the arterial trunk, no 
decussation of the optic nerves, free gills and opercula, with ab- 
dominal ventral fins. The skin and scales with which we com- 
menced our investigation are not included. I consider the cha- 
racter of the abdominal ventral fins merely as a temporary link. 
Among the fishes considered by Agassiz as Ganoids, there are 
fortunately but few from families which are now certainly known 
to be common osseous fishes. Acanthoderma, Pleuracanthus, 
Diodon, Ostracion, and Calamostoma, positively belong to the 
true osseous fishes, the last genus as a Lophiobranchiate, the 
others as Plectognathi. 
As the fossil genera Blochius, Dercetis and Rhinellus, have 
little or no resemblance to the Sclerodermi, to which they are 
referred in the Poissons Fossiles, it becomes a question whether 
they should not be retained among the Ganoids. Blochius, 
according to Agassiz, has rhomboidal enamelled scales, but the 
probable position of its ventral fin near the thoracic fin renders 
their ganoid nature doubtful. Rhomboidal scales alone afford 
no positive character, for Balistes have them without being 
Ganoids. As regards the enamel, I consider this of value on 
small scales only, when no other characters are opposed to 
the ganoid structure; the existence of enamel was also ascribed 
to Balistes, but this I cannot admit. A more accurate know- 
ledge of the position of the ventral fin in Blochius will assist us 
materially on this point. The osseous plates of Dercetis and 
Rhinellus would not be sufficient to prove them to be Ganoids ; 
for such plates exist in many osseous fishes, and in others 
which do not possess them in the adult state they sometimes 
occur in the young, as in the Sword-fish. 
Be this, however, as it may, whether Blochius, Dercetis and 
