36 ANNALS NEW YORK ACADE2IY OF SCIENCES 



and the roots of the fourth (milk-molar, dp^) indicating a smaller tooth 

 than the corresponding tooth in H. scliinzi. It agrees better with Abel's 

 diagnosis (Abel, 1904, p. 16, 25) of H. cJiristoli Fitz., from the upper 

 marine Molasse of Linz (Middle Miocene), but Pitzinger's (Fitzinger,. 

 1842) figures of the jaw in this species are not accessible. 



Probable Affinities. — Accepting provisionally the reference of the Porto 

 Eico sirenian to Halitlierium, it may be of interest to note where it stands 

 in the evolutionary history of the Sirenians. 



It is generally accepted at present that this group is descended from a 

 common stock with the Proboscidea — that is to say, they are derived 

 from terrestrial ungulates with short five-toed plantigrade feet, a com- 

 plete series of teeth, bunodont molars, four or five cusped, the posterior 

 premolars partly molarif orm, the anterior ones simple, canines not notably 

 enlarged, but a tendency to enlargement of a pair of upper and lower 

 incisors — and a variety of other characters which I need not notice. 

 Moeritherium, of the Upper Eocene and Lower Oligocene of Egypt, stands 

 not very far from this common stock; but whether or not it be really 

 ancestral to the Proboscidea it has gone a short distance in that direction, 

 the limbs being somewhat long and straight and the teeth and skull ap- 

 proaching in some degree the Proboscidean specialties more clearly shown 

 in Palceomastodon of the Egyptian Oligocene. 



Prorastomus, on the other hand, may be taken as representing the 

 primitive Sirenian. Unfortunately we do not know its skeleton charac- 

 ters. But being found in a marine limestone it probably was already 

 adapted to aquatic life. The long narrow skull, rather slender jaws, 

 teeth conforming to the primitive type indicated and not widely different 

 from those of Mceritherium, all point to its ancestral position. 



From this primary stock we find three or four diverse lines of speciali- 

 zation. In the Manatee the front teeth disappear and the cheek teeth all 

 become molariform and appear to increase in number, pushing upward 

 and forward in the jaw to replace those lost by wear. This increase in 

 number of the cheek teeth is supposed to be due to reduplication of the 

 molars from behind, a fourth, fifth, sixth true molar etc. appearing 

 de novo (Thomas and Lydekker, 1897). 



In the Dugongs, on the other hand, one pair of upper incisor teeth is 

 retained and enlarged into tusks, while the cheek teeth are progressively 

 reduced in number, the premolars becoming smaller and simpler and the 

 anterior ones disappearing, while there is no tendency to increase in 

 number of the true molars. The skull in both Manatee and Dugong is 

 much shortened and widened, the jaws deepened and the front of muzzle 

 and jaw bent downwards and covered with horny plates for triturating 



