354 Messrs. Hancock & Atthey on Reptile- and Fish- ins 
A fragment of a maxillary bone of Rhizodopsis has, it is im- 
possible to doubt, served for the establishment of the so-called 
Characodus (pl.13). Here there is not one tooth left; they are 
all broken away; but the form of the fragment itself, tapering 
at one extremity and suddenly expanding at the other, as like- 
wise the columnar structure of the bone for the support of the 
teeth, prove this to be an imperfect maxillary of Rihizodopsis 
saurotdes. 'These peculiar as of bone supporting the teeth 
are very characteristic of the jaw-bones of this fish; but in the 
premaxilla they are most developed. Some of our specimens 
(Pl. XVI. fig. 5) are precisely similar to that figured as 
Characodus, the teeth having been all broken away, with the 
exception of three or four. The display of this curious struc- 
ture depends much on the plane of the sections; it is possible 
to cut it nearly all away, leaving merely the external layer of 
bone on one side; and it is never developed to the same 
extent in the preemaxilla and mandible. 
The premaxilla is the basis of the genus Grastrodus (pls. 14 
& 15) the supposed Batrachian, as is evinced by the aap of 
the fragment, the size, form, character, and disposition of the 
teeth ; nor is there any important difference in the minute 
structure of the teeth in this so-called genus. According to 
Prof. Owen’s measurements, the dentinal tubules in Déttodus 
parallelus have a diameter of +5455 of an inch, in Characodus 
sates Of an inch, and in Gastrodus +>455 3 While in Rhizo- 
dopsis we have ascertained that they are likewise about+y4e5 ~ 
of an inch in diameter. The teeth of the so-called Gastrodus 
are certainly represented to be without enamel; but we have 
seen that it is frequently absent in Rhizodopsis; and many of 
the teeth, as exhibited in the figure, are cut diagonally short, 
so that their form and proportions are destroyed. ‘The ap- 
pearance thus presented is very common in sections of minute 
jaws, and, unless clearly understood, may readily lead to error. 
The diagonal section of a quill illustrates this very well. . 
The bone-cells of the jaw of Rhizodopsis are quite as Ba- 
trachian as are those figured of the pseudo-Gastrodus ; and so 
are those of Megalichthys and many other sauroidal fishes. 
There is, then, no evidence in the paper referred to of a 
minute air-breathing Batrachian of the age of the lower seams 
of the Northumberland coal-field, the so-called genus Gastrodus 
being resolvable into Rhizodopsis sauroides, a Ganoid fish. 
Ctenodus cristatus. 
Since the publication of the paper on Ctenodus*, the matrix 
* Ann. & Mag. Nat. Hist. Feb. 1868, 
