12 
TliliTIARY VERTEBRATA OF THE FAYtlxM. 
measure at least, for tlio mucli greater height of the outer side of the crown than of 
tlie inner side. 'I’lie same diagram also shows in section the junction of the maxillae 
with the ])r('maxill;e {ntx.pmx.s.), and the probably mesethmoid groove {mes.g.) 
which lies between the upper ends of the premaxilloe and was no doubt occupied 
in life by the lower edge of a vertical cartilaginous nasal septum. Another possible 
explanation of the deeply concave palate is, that Arsinoitli erium may have possessed 
a tongue of peculiar form, probably extensible and ])rehensile ; for it is difficult to 
understand how an animal of such great size could grasp sufficient food with so 
extremely narrow an incisor region if it had not some additional means of prehension 
in the form of a mobile up])er lip or tongue. 
Further back the palatal surface of the maxilhe becomes more flattened, and opposite 
the second molar these bones unite with the palatines in a broad transverse suture ; 
behind this point they only appear on the palate as a narrow strip on either side, 
llehind the last alveolus the maxilla is produced backwards into a short pointed 
])rocess which forms the outer wall of a foramen (PI. II. fig. I, m.p.f.)^ the rest of 
which is a deep notch in the edge of the palatine. Above this foramen the hinder 
border of the maxilla forms the outer edge of a groove {p-P-g.), running up towards 
the anterior opening of the alisphenoid canal (text-fig. 4 , al.c.). This groove and 
foramen, which seem to be homologous with the posterior palatine foramen of the human 
skull, also occur in Ilgrax (Procavia) and possibly other Ungulates, and no doubt 
transmitted branches of the jjalatine nerve and blood-vessels. The foramina to which 
the name posterior palatine is applied in most mammals lie on the palate at or near 
the junction of the anterior end of the palatines with the maxillae ; and these openings 
are also present in Arsinoitherium, though very small. Probably both pairs of 
openings simply mark the points at which the palatine vessels and nerves ])assed 
between the maxillas and the palatines, when these bones were still separate in the 
young ; and the presence or absence of the anterior or posterior of the foramina 
mentioned depends on whether the palatine nerves and vessels passed out iu one or 
more groups on either side. In Arsinoitherini)i the posterior openings are large, and 
those near the anterior end of the palatines small, owing probably to some structural 
peculiarity which rendered necessary a larger nerve- and blood-sup])ly for the ])osterior 
region of the palate. In some animals there is a series of small loramina on either 
side, the anterior one usually being much the largest. 
In front of the posterior edge above mentioned the maxilla widens into a large 
mass, which in the adult lodges the roots of the posterior molars, and in the young, in 
which it is relatively still larger, contains the develoj)ing germs of the same teeth. 
This alveolar mass to a considerable extent forms a floor to tlie orbit. Although the 
maxilla is excluded from the actual border of the orbit by the union of the lachrymal 
and jugal, it nevertheless forms a large part of its anterior wall iu the neighbourliood 
of the posterior opening of the antorbital canal, where it unites in suture with the 
