AND CLASSIFICATION OF THR FOSSIL REPTILIA. 
221 
height there is no indication of its existence in the figure ; but in the specimen the 
bone projects forward a little, and is well defined by a groove above it. Its inner 
border is convex, and the pterygoid unites with tliat border, just above the condyle, 
by an attachment which is 2 centirns. deep. The attachment was not firm, for in 
D.'pardice'ps and other species the quadrate bones are lost from the otherwise perfect skull. 
The Dicynodon 'pardiceps (Owen), South African Cat., No. 70 ; Brit. Mus., 
No. 47,045, in many respects the most instructive of all the specimens, has never 
been adequately described. It shows details of structure, owing to the softness of 
the matrix, which manifest the union between many of the bones. The remarkably 
elongated zygomatic arch I find formed chiefly by the squamosal bone, which extends 
forward to the orbit. The maxillary bone tdso contributes to form it. extending below 
and behind the squamosal backward, almost to the descending process. The malar 
bone forms the lower border of the orbit. It rests externally upon the squamosal, 
and internally upon the maxillary; its posterior extremity supports the post-frontal 
bone, and its anterior extremity appears to extend forward to the lachrymal, and 
inward to meet what I take to be an outwardly directed process of the palatine. The 
post-frontal is a slender transverse bar, and meets the frontal by a well defined suture. 
The small parietal lies in front of the parietal foramen; but I do not feel certain that 
the long oblique posterior processes are rightly referred to that bone, and it would 
seem as though the analogous structure in Lizards had suggested an explanation 
which has not been questioned. These bones seem to me to diverge anteriorly to 
expose the parietal, and they divei'ge posteriorly to support the squamosals. If they 
are separate ossifications, they may represent external elements in the Amphibian 
skull which have remained after a deeper seated ossification was developed, just as in 
some types basi-temporals remain after the sphenoid is ossified- 
In Dicynodon pardiceps a groove connects the upper posterior corner of the nasal 
aperture with the orbit, much as in Pareiasaurus, though the depression is but 
slightly marked. In D. leoniceps the region behind the nares is impressed over the 
depth of the apertures. 
In Dicynodon tigriceps (Owen) tlie configuration of the zygoma and temporal 
fossa is stated to be most nearly paralleled in Chelydra, though the difterence is 
considerable. In another specimen referred to the same species, the inter-orbital 
space is said to be more completely ossified than in modern Crocodllia, Chelonia, or 
Lacertilia, and is an approximation to Mammalian structure. The upper anterior 
angle of the pterygoid is said to join the anterior extension of a cranial bone which 
may correspond with the pre-sphenoid of Crocodiles, or the orbito-sphenoid of 
Chelonians. The reference of this specimen to D. tigriceps is on several grounds 
open to question. But the bone which the author regards as pre-sphenoid (5) seems 
to me to be the median plate of the pterygoid. It is possible that a small ossifica¬ 
tion above this, which extends obliquely forward and upward, may be the pre¬ 
sphenoid, for it is in the position which the pre-sphenoid should occupy. The hone 
