Dr. R. H. Traquair—Devsnian Fishes of Canada. 265 
Mr. Smith Woodward, however, in the second part of his Catalogue 
of the Fossil Fishes in the British Museum, retains the family of 
Phaneropleurida, in which he also still includes Uronemus. That 
Uronemus must, on account of its non-ctenodont dentition, be placed 
in a different family from Phaneropleuron is, as I have above stated, 
perfectly clear to my mind at least ; for the rest, the Phaneropleuridz 
are separated from the Ctenodontidz, according to Mr. Woodward’s 
definition, by the possession of marginal teeth and of jugular plates. 
As regards the former character I believe it to be founded on a 
mistaken interpretation; as to the latter, are we quite sure that we 
know the structure of Ctenodus sufficiently well to be able to decide 
with certainty as to whether it had jugular plates or not? 
Mr. Woodward also erects a separate family for Dipterus—that of 
the Dipterid, the essential characters of which, as distinguished from 
those of the Phaneropleuride, are: the absence of marginal teeth, 
and the heterocercal configuration of the tail, this form of caudal fin 
along with the possession of jugular plates distinguishing it from the 
Ctenodontide. Itis not necessary to repeat what has been said as to 
the marginal tooth question; as regards the heterocercal tail in 
opposition to the diphycercal form, if we accepted that as a family 
character in the Crossopterygian fishes we should separate genera 
which otherwise are most closely allied. Take the Rhizodontide 
for example,—in Rhizodopsis the tail is heterocercal, in Gyroptychius 
as nearly as possible diphycercal ; so in like manner the most marked 
distinction between Tristichopterus and Eusthenopteron is the decided 
heterocercy of the former and the nearly perfect diphycercy of the 
latter. But the median fin-system of Scaumenacia is connected by 
so perfect a transition through Dipterus macropterus to that of the 
typical D. Valenciennesii, that I cannot see how it can be placed in a 
different family on characters based on those organs. 
Consequently I must for the present still consider it the safest thing 
to include Phaneropleuron, Scaumenacia, Dipterus, Ctenodus and their 
immediate allies in one comprehensive family of Ctenodontide, 
characterised by the possession of numerous dermal cranial roof-plates 
and a ctenodont dentition. 
CoccosTEIDz#. 
Coccosteus Canadensis, A. S. Woodward, Grou. Mac. (3), Vol. IX. (1892), p. 483. 
One plate referable to Coccosteus and presumably to the above- 
named species occurs in the present collection. It is the posterior 
ventro-lateral of the left side, and is peculiar in being narrow 
posteriorly, and in having the stellate based tubercles with which it 
is ornamented much coarser than in the typical C. decipiens of the 
Scottish Old Red. 
Ho.LoprycHtiipZ&. 
Glyptolepis Quebecensis, Whiteaves, Trans. Roy. Soc. Canada, vol. vi. sect. iv. p. 77, 
pl. v. fig. 4. A.S. Woodward, Cat. Foss. Fishes Brit. Mus. pt. ii. p. 336. 
The specimen here referred to is the head and anterior part of the 
body of a large Glyptolepis, the only example of the genus in the 
present consignment of Canadian fishes. It measures nine inches 
