SYSTEMATIC RELATIONS OF FISHES. 581 
II. SPECIAL ON THE GANOIDS. 
Recurring to Miiller’s system, the writer adopts, as characterized 
beyond dispute, his sub-classes, or orders of Leprocarpu, DER- 
MOPTERI, SELACHII and Drpnor, and confines himself at present to 
the recent Ganoidea and Teleostei. I have shared in the doubts 
occasionally expressed by icthyologists, as to the essential dis- 
tinction of these latter divisions, and an examination into the 
osteology, with reference to this point, confirms the doubts raised 
by a study of the soft parts. As is well known, Miiller distin- 
guished the Ganoidea by the muscular bulbus arteriosus contain- 
ing numerous valves, and the connection of the optic nerves by 
commissure rather than by decussation. He added several other 
characters, knowing them, however, to be shared by various other 
orders and sub-classes, and I have selected the only two which 
seemed to be restricted to the division. Their restriction to it, 
however, is only apparent, and Kner points out that the peculiar- 
ity of the optic commissure is shared by some Physostomi, and 
that the difference between the number and character of the valves 
of the bulbus in Lepidosteus and Amia, is quite as great as that 
existing between Amia and some of the Physostomi. After an 
examination of the skeleton it is obvious that in this part of the 
organism also, there is nothing to distinguish this division from 
the Teleostei of Müller. It is true that each of the genera re- 
ferred to it possesses marked skeletal peculiarities, but they are 
either not common to all of them, or are shared by some of the 
Physostomi. If, on the other hand, we compare these genera with 
each other, differences of the greatest importance are observable, 
which at once distinguish two gehen ine one represented by Polyp- 
terus, the other by Lepidosteus and 
In the first place the basal radii of sh paws fins of Polypte- 
rus are observed to be excluded from articulation with the scapu- 
lar arch by the intervention of three elements, which form a 
pedicel or veritable arm for the fin. In Lepidosteus and Amia the 
radii are sessile on the scapular arch as in ordinary fishes. The 
ventral fins present a like difference ; the basal radii are long and 
four in number in Polypterus. In the other two genera they are 
absent, excepting one rudimental ossicle on the inner basis of the 
fin (two in Lepidosteus) precisely as in the Physostomous families 
Mormyride, Catostomide, etc. If we examine the branchial ap- 
