J38 
TERTIARY VERTEBRATA OF THE FAYCtM. 
succession. Tlic (lifFercuces between the palatines and pterygoids in the earlier and 
later types are of course dependent on the same causes.. 
'I'he anterior ])ortiou of the skull, particularly the upper portion, is completely 
wanting in the large specimen on wliich the above description is for the most part 
founded ; but another specimen (PI. XIII.) supplies the necessary information 
concerning this region. It is the anterior portion of a skull of a very young 
individual in which the milk-teeth are still in use ; it includes the whole of the 
frontals as far back as considerably behind the postorbital processes, the nasals, 
premaxillaries, lachrymals, and most of the maxillae, including the zygomatic 
process, also the palate as far back as the level of the middle of the zygoma. 
There are two teeth in situ on one side and three on the other ; they are probably 
mm. 2, 3, 4, and will be described below. The crushed base of one of the milk- 
tusks is preserved in the socket. As will be shown below, this specimen in the 
shortening of the nasals and the shifting backwards of the narial opening supplies 
a beautiful illustration of the approximation towards the later Proboscidean type found 
in this genus. In the relations of the bones to one another it precisely resembles 
Elcphas. 
I'he right frontal (Jr.) is mostly broken away, but the left is better preserved, and 
although its actual junction with the parietal cannot be observed, nevertheless 
little can be wanting from its hinder border. Together the upper part of the 
combined frontals in the interorbital and part of the postorbital regions forms the 
gently convex skull-roof. In this young specimen scarcely any trace of the supra- 
temporal ridges (the divided sagittal crest) can be seen, but it is evident that they 
ran on to the postorbital processes. External to these ridges in the postorbital 
region the frontals run down into the temporal fossm, and are strongly convex from 
above downw'ards. The postorbital processes are blunt prominences, from each 
of wdiich a ridge runs downwards and backw’ards, separating the orbit from the 
temporal fossa. Ventrally this ridge must have been continuous wdth that formed 
by the free edge of the alisjdienoid described above. In front of the postorbital 
processes the frontals form a w’ell-defined upper rim to the orbit, and anteriorly 
they unite in suture with the maxillm, lachrymals, premaxillae, and nasals, their 
general relationshi])s to the neighbouring bones being exactly as in Elephas. 
The lachrymal (Jac.) is a small bone wedged beUveen the frontal and maxilla, 
and grooved below by the upper surface of the antorbital canal. It is ])erforatcd 
by a large foramen w-hich lies 'within the border of tlie orbit ; above the foramen 
and on the rim of the orbit there is a small but prominent tubercle. 
The nasals {na.) are short bones wliich already a])proxiniate very nearly in form to 
those of Elephas. Together they ])roject in a blunt point over tlie nasal aperture. 
Posteriorly they are together roughly semicircular in outline and unite for the greater 
part of their width w'ith the frontals, but externally to these with the jiremaxillai. 
