JDr. C. W. Andrews — Notes on Egyptian Eocene Mammals. Ill 



through the development of sinuses which communicate with the 

 outside of the skull by a number of irregularly arranged foramina. 

 The frontals are comparatively short ; they are slightly marked by 

 «upra-orbital ridges, which run along the upper borders of the 

 orbits ; anteriorly they are cut off from the premaxillee by the union 

 of the maxillge with the nasals. These latter bones are short, and 

 overhang the narial opening to a very small extent ; from behind 

 forwards they form sutures with the frontals, maxillae, and pre- 

 maxillee ; in fact, their relations to the surrounding bones are exactly 

 -as in the later Proboscidea. 



The nasal opening is large and looks forward, not upward as in 

 SJlephas. It is not at the end of the snout, and the upper surface of 

 the premaxillEe in front of it is deeply grooved, probably indicating 

 the presence of a short proboscis as in the Tapir. The orbit is very 

 small, and there is a mere trace of supra-orbital processes. 



In the young specimen above referred to, the last molar is still 

 uncut, and from its position it is clear that the rest of the cheek 

 teeth must move considerably forward in order to allow it to come 

 into position. 



The occurrence of a species of Moeritherium, probably identical 

 with M. lyonsi, in the Upper Eocene beds in association with 

 PalcBomastodon raises the question of whether Moeritherium can bo 

 ancestral to Palceomastodon. If it is not, at least it must be extremely 

 similar and very closely I'elated to the actual ancestor-, for it presents 

 all the proboscidean chai^acters in exactly the more generalised 

 condition that one would expect to find. Moreover, it may be 

 pointed out that PalcBomastodon does not occur in the Middle Eocene 

 beds in which McBritlierium is abundant, while in the upper beds 

 Palceomastodon is common, and but few Moeritherium remains have 

 been found. 



From the vertebral column of a large, and at present unnamed, 

 species of Mcnritherium the number of the vertebrse in the different 

 regions can be ascertained with reasonable certainty. These are : 

 cervical, 7 ; thoracic, 20 ; lumbar, 3 ; sacral, 3 ; caudal, number 

 unknown, but the tail must have been of moderate length. The 

 axis has a peg-like odontoid, the lower surface of which bears 

 a large facet for articulation with the atlas ; the neural spine is 

 high. The centra of the other cervicals are rather short, but longer 

 in proportion to their size than in the later proboscideans. 



The scapula is oval in outline ; the coracoid process is large, and 

 the glenoid surface for the humerus is continued on to its lower 

 face. The humerus is in some respects more like that of some 

 oarnivores than that of an ungulate. Its most remarkable feature 

 is the extreme compression of the shaft from side to side. The 

 ent-epicondyle is very large, and there is no ent-epicondylar foramen. 

 The supinator ridge is well developed. The femur is flattened from 

 before backwards ; the great trochanter rises a little above the 

 head, and there is no third trochanter. The distal articular end 

 is relatively small, and the condyles ai'e as in Elephas. The feet are 

 at present quite unknown. 



