FROM SOUTH AFRICA AND AUSTRALIA 125 
upper jaw seem to be completely anchylosed with the bony walls of. 
the jaw, so as to look like mere processes of it. I would be under- 
stood to speak with considerable hesitation on these points, however, 
the parts being but very imperfectly preserved. 
It is at once obvious that the skull which I have just described 
could have belonged to no true Reptile, but is either that of an 
Amphibian or that of a Fish. 
The composition of the lower jaw, the characters of the teeth, the 
well-developed nasal apertures, and the arrangement of the bones in 
the temporal region leave no doubt in my mind as to which of the 
latter alternatives is to be preferred, and satisfactorily prove the 
amphibian affinities of the fossil. 
Such being the case, there is but one order of the Amphibia, as 
they are at present arranged, to which it can be referred—the 
Labyrinthodonta,—with a knowledge of whose characteristic pecu- 
liarities, so much of the structure of the skull as can be made out 
becomes readily intelligible. Thus, in shape and in the position of 
its orbital and nasal apertures, the African fossil presents a certain” 
resemblance to the German Labyrinthodont Zeéopzas, and to the 
imperfectly known Russian RAznosaurus. The arrangement of the 
cranial and facial bones, and their ornamentation, coincide very well, 
so far as they go, with the corresponding features of those Labyrin- 
thodonts which have been best studied ; and the peculiarities of the 
jugal, postorbital, and squamosal bones are especially characteristic. 
Again, I should hardly have ventured to interpret so confidently 
the appearances presented by the mandible, had I not recently had 
an opportunity of studying the composition of its articular moiety in 
some portions of very large mandibles of Labyrinthodon, or Masto- 
donsaurus, from Warwickshire. I find, from these fossils, that the 
articular element of the Labyrinthodont jaw (in these genera at any 
rate) sends a hollow bony prolongation (at first probably a mere 
osseous sheath around Meckel’s cartilage) for a long distance towards 
the symphysial end of the jaw; and I suspect that the cone of 
matrix which I have described above is nothing but the cast of a 
similar prolongation. In the Warwickshire Labyrinthodont, a strong 
process, formed partly by the angular and partly by the articular 
bone, is given off inwards and forwards from the posterior part of the 
inner surface of the ramus; and this is perhaps represented by the 
inward production of the posterior part of the ramus of the mandible 
in the African fossil. On the other hand, the great Labyrinthodonts 
have a very distinct angular process prolonged backwards behind the 
articulare, and composed in great measure of a process of that 
