Dr. C. W. Andrews — Extinct Vertebrates from Egypt. 295 



incisors (i^) are modified to form a pair of downwardly directed 

 tusks, and there are also one or two pairs of small laterally placed 

 incisors {i-, P) near the premaxillo-maxillary suture. Canines (c) 

 were present; both these and the posterior incisors were probably 

 shed early in life. The upper cheek-teeth are eight in number, con- 

 sisting of four single-rooted teeth (premolars) and four double-rooted 

 molars, the crowns of which are bilophodont, each ridge being 

 distinctly composed of two tubercles. (Figs. 1, 2.) 



The downwardly deflected symphyseal region of the mandible 

 is greatly thickened and very massive ; it bears the alveoli of four 

 pairs of teeth, apparently the incisors and canines. The anterior 

 pair of these alveoli is comparatively shallow, and in all probability 

 the teeth were lost early, this region being covered by a horny 

 sheath. There were probably three single-rooted and five double- 

 rooted cheek-teeth ; but this is uncertain owing to the want of 

 well-preserved specimens. There are indications that a replacement 

 of the molars from behind was already established in this early form, 

 at least in the mandible. (Fig. 3.) 



The vertebral column is much like that of Halitherium, and the 

 scapula is characteristically Sirenian in form, having a backwardly 

 curved, sickle-shaped blade. The rudimentary os innominatum is 

 similar to that of Halitherium Schinzi, but the acetabulum is more 

 strongly defined. The species from the Fayum may be called 

 Eosiren libyca. The dimensions of the type skull are : — 



mm. 



Extreme length iu straight line 290 



Width at occipital condyles 75 



"Width of foramen magnum ... ... ... ... 34 



Height of foramen magnum ... ... ... ... 22 



"Width between supra -orbital processes 87 



"Width of cranial roof between the temporal fossae ... 47 



"Width of snout in front of external nares .. . ... ... 45 



Length of dental series from canine to last molar... ... 114 



Length of molar series ... ... ... ... ... 56 



Length of external narial opening 64 



Width of external narial opening 30 



The mandible figured was not associated with the skull, but was 

 found on the same horizon and in the same locality. 



It is interesting to note that the alveoli for the posterior incisors 

 and the canines are not borne on the alveolar border of the upper 

 jaw, but seem to have been thrust somewhat outwards on to the 

 facial surface, probably through the development of the horny plate 

 which formed the actual biting surface of the anterior portion of 

 the jaw. It is remarkable that, except in the presence of posterior 

 incisors and canines, this early (Middle Eocene) Sirenian is scarcely 

 at all more generalized than the later Halitherium, and it appears 

 that the Sirenia must have branched off from their parent stock 

 at an extremely early period. In some respects, particularly in the 

 structure of the teeth and of the humerus, there is a certain similarity 

 with Moeritherium, and it seems not improbable, therefore, that the 

 relationship between the Sirenia and the Proboscidea suggested by 

 Blainville and others may have a real existence. 



