EOSIEEN LIBTCA. 203 



might be expected in a land-animal, the olfactory lobes are more developed than in 

 Eosiren, but this does not affect the general similarity, which is the more striking when 

 taken in conjunction with the many other points of likeness existing between the Sirenia 

 and the Proboscidea. It is interesting to note further that in the Middle Eocene forms 

 of both these orders the brain is relatively large for mammals of that early date. In the 

 terrestrial Proboscidea living exposed to many enemies, and in a complex and changing 

 environment, this large development of the brain persists even in the modern types, 

 and, indeed, their survival may be due to this very character. On the other hand, in 

 the Sirenia, which early adopted a sheltered and sluggish mode of life, the conditions 

 of which must have remained nearly constant, although there is some tendency towards 

 increase in the size of the brain, there is no need for increased complexity, the result 

 being the peculiar type of structure which Dr. Elliot Smith states can only be 

 parallelled by the structure " occasionally presented by the brains of idiots in which 

 the process of elaboration has ceased in the earlier months of intra-uterine life, and 

 the organ has simply grown in size without becoming perfected in structure." 



Upper Dentition. — The dental formula of the upper jaw appears to be i. 2 or 3, 

 c. 1, m.-\-pm. 8. At the anterior end of the downwardly turned rostrum there are 

 the alveoli for a pair of enlarged incisors (^. 1), which were oval in section and directed 

 downwards and forwards ; the sockets are comparatively shallow, these teeth being 

 apparently rooted. 



The only trace of the second incisors (i. 2) seems to be a small alveolar pit situated a 

 little below and just in front of the third incisors (^. 3), which also are represented in this 

 specimen by their alveoli only, these being placed immediately in front of the maxillo- 

 premaxillary suture and on the side of the rostrum abou,t 5 mm. above the sharp 

 edge of the premaxillae. Both the second and third incisors are clearly on the way 

 to disappearance, and seem to have been thrust outwards to the side of the jaw 

 by the development of a horny plate covering the palatal and sharp lateral edges 

 of the premaxillary region and replacing the teeth functionally. The canine 

 alveolus (c.) is also small and on the side of the rostrum ; it is placed immediately 

 behind the maxillo-premaxillary suture, and is almost in contact with the alveolus 

 of i. 3. 



Behind the canine, and separated both from it and from one another by nearly 

 equal intervals (about I'S cm.), are the sockets for four single-rooted teeth. In the 

 type specimen only the last of these is in situ ; it consists of a comparatively small 

 single cone with a very strongly developed internal shelf or cingulum. Immediately 

 behind this come the relatively small quadrate two-rooted molars. There are four of 

 these teeth, of approximately equal size. Each molar crown consists of two 

 transverse ridges, each composed of two pointed cusps without any secondary 

 tubercles ; there is also a small anterior ridge or cingulum with a median cusp, which 

 in wear becomes connected with the antero-internal main cusp. All the specimens 



2d2 



