80 Mr. T. Atthey on various Species of Ctenodus 
single piece, are equally distinct; they are thick and well- 
rounded, being only slightly elongated transversely, and have 
the upper or hinge-margin flattened, the surface irregularly 
granulated, punctured, and enamelled. The tail is in a con- 
fused state ; but there can be little doubt that it is heterocerecal, 
of the rhomboidal type. And there are traces of an anal and 
a ventral fin immediately before the caudal; but it is impos- 
sible to say whether or not they are lobed. There are a few 
slender, curved bones scattered about behind the head, which 
have the appearance of ribs; these and the bones of the head 
alone have been ossified; the vertebre have all disappeared. 
Such is the description of C. elegans, so far as it can be 
determined by the imperfect remains of the specimen in m 
possession. That it is specifically distinct from the Old-Red- 
Sandstone species is evident enough; and it seems quite 
probable that it even differs from them generically. It is 
much smaller than most of the Dipterti—though it is apparently 
not full-grown, as the teeth are considerably smaller than the 
largest specimens of them belonging to this species in my 
possession ; and, on account of its minuteness when com- 
pared with the largest species of Ctenodus from the coal- 
measures, there is some doubt whether it may not be generi- 
cally different also from them. It is not more than three 
inches long, as has been already stated, while three at least of 
the larger species cannot be less than four or five feet from 
head to tail. The sphenoid bones previously mentioned of 
C. elegans are together about half an inch long; the same 
bones of the large species alluded to must have been at least 
eleven inches in length. 
From the fragments of a large sphenoid and palatal bones 
in my possession, it is possible to restore this portion of a 
skull that probably belonged to C. eristatus, which is not the 
largest species ; and we are thus enabled to form an approxi- 
mate estimate of the size of the head, which must have been 
upwards of twelve inches long and nine broad. Therefore, 
assuming the proportions to be similar to those of C. elegans, 
we arrive at the conclusion that these large Ctenodi were not 
less than four or five feet in length. 
This estimate of the size of these animals receives some 
support from the fact that very large opercular plates occur in 
the same shales in which the palatal armature is found. I 
have in my possession six or seven different kinds of opercula, 
all having the essential characters of those found in connexion 
with the head of C. elegans ; that is, they agree with them in 
the roundness of their form, in being composed of a single 
solid piece, and in the surface-granulation and enamel. So 
