MAMMAL-LIKE REPTILES. 769 
suggest that the teeth may have been like those of Dewtero- 
SMUTUS. 
There is a single canine, which is large, strongly inclined 
forwards, and circular in section. Behind this there are three 
molars, shown in transverse section to be circular and relatively 
small; it is probable that there was one more only. 
One important difference between this type and the preceding 
three is that whilst they all have very numerous replacing teeth, 
which are not in the same stage and were thrust up indefinitely 
like those of a crocodile, in Lamiasaurus there is no trace of 
successional teeth. 
The other types of South African Tapinocephaloids which have 
been described are :— 
Delphinognathus conocephalus, a much smaller form without 
large canines, of which the details of the dentition are not 
known. 
Moschops capensis, also a small form with no canines and with 
no details of the dentition. 
Taurops macrodon, a large form apparently with teeth some- 
what like those of Tapinocephalus, but not sufficiently described 
or figured to be definitely identified at present. 
Eccasaurus priscus, founded on a humerus of rather unusual 
type, to which is referred a tooth of the general plan of Tapino- 
cephalus but of a narrower oval form and more cuspidate. 
Moschognathus whaitsi, founded on a figured lower jaw which, 
as it seems to contain no teeth, is indeterminable. 
The Titanosuchid branch of the Deinocephalia is represented by 
the following skull-material in the British Museum (Natural 
History) :— 
An imperfect skull (Pl. V., & text-figs. 17, 18) consisting of 
the dorsal surface from behind the nostrils to the back, with the 
squamosals and part of the zygomatic arch; and a block which 
has a rather poor fit with the upper part, but contains a good 
deal of the maxilla and prevomers and a small fragment of the 
palatine. 
This specimen (R. 3595) was collected by Prof. Seeley at 
Tamboer Fontein. The only other skull-materials I have used 
are the premaxille of the type-specimen (49370), which itself 
consists of the associated remains of two individuals of identically 
the same size. The maxille of these agree exactly, so far as can 
be seen, with that of the specimen mentioned above. 
Nothing is known of the basicranial part of the skull, but the 
supraoccipital above the post-temporal fossz is present in the 
Tamboer specimen. It is a flat wide plate with rather large post- 
temporal fossze bounding it below, and the upper part of the 
deep and narrow brain-cavity impressed on its lower surface. Its 
posterior surface is overlapped by the interparieta 1 and tabulares, 
and its upper edge is in contact and perhaps fused with the 
parietals. 
