348 THE AMERICAN NATURALIST. (VoL. XXXII. 
I do not see how, in the pycnodonts, the expansions of the 
neural and hæmal arches can be regarded as essentially different 
from the epicentra and hypocentra of the tail of the young 
Amia. They are indeed continuous with the corresponding 
arches; but so, too, are the vertebral centra of many other 
fishes. We do not know that the ossifications of the arches of 
Lepisosteus are at any time distinct from the central ring. 
That the superior expansions lying upon the notochord of the 
pycnodonts are continuous with the upper arches is evidence 
that these expansions do not originate from pleurocentra. The 
intercalated elements appear to have been suppressed or nearly 
so, since the bases of the arches in some genera are suturally 
connected, thus contributing to the rigidity of body of these 
remarkable fishes. 
I conclude, therefore, that we have here two distinct series 
of fishes to deal with. In the one series the intercalated 
elements are never of special importance and do not form 
distinct ossifications. The vertebral centra, when developed, 
arise principally or altogether from coalescence of the bases of 
the upper and the lower arches. This series will include the’ 
Semionotidz, the Pycnodontide, the Lepisosteidæ, and the 
Aspidorhynchide. The other series will contain those fishes 
in which there is an evident tendency for the pleurocentrum to 
usurp the place and function of the ossifications that should 
arise in the bases of the upper arches. This series will embrace 
the Macrosemiidz, the Eugnathide, the Amiide, and the 
Pachycormide. That is, I believe that Mr. Woodward has 
gone too far in his transference of genera from the Lepisosteoid 
to the Amioid series. 
As regards the systematic value and kinship of the two 
groups, it seems to me that they rank no higher than suborders 
of a distinct order of Actinopterygia. The two suborders are 
more closely related to each other than to the Chondrostei, or 
even to the Isospondyli, although the latter order has probably 
taken its origin from the Amia-like fishes. 
I believe that the characters on which my groups of these 
fishes have been founded indicate two very distinct lines of 
development, and lines of such a nature that when once entered 
