82 MR. R. LYDEKKER ON A StRENlAN JAW. [Feb. 2, 



the milk-molars were not like those of the present specimen, since 

 the permanent molars were of a more Hippopotamus-like structure 

 than are those of H. veronense, and show no indications whatever of 

 selenodontism. 



Conclusive evidence of the Sirenian nature of our fossil is, however, 

 afforded by the orbital region, which is almost exactly the same as 

 in the Sirenian from Jamaica described by Sir H. Owen^ as Pro- 

 rastomus sirenoides. In both we have the same peculiar eversion 

 and projection of the lower border of the orbit which is absolutely 

 characteristic of the Sirenia. In both, again, we notice the extra- 

 ordinary size of the foramen for the exit of the fifth nerve, and its 

 immediate proximity to the anterior border of the orbit, these 

 being also distinctive Sirenian features. 



Then, again, a comparison of the last milk-molar with the upper 

 molars oi Halitherium schinzi {e.g. B.M. No. 40859) clearly shows 

 the ordinal identity of the two forms. 



I take it, therefore, that the Siienian nature of the Vicenza 

 specimen is certain; and since its milk-molars agree in general 

 structure and relative size with the permanent molars of Halitherium 

 veronense, which occurs in the same country and probably on the 

 same geological horizon, the evidence appears to be very strongly in 

 favour of the reference of the specimen in question to that animal. 



Apart, however, from any question of specific reference, the 

 specimen before us undoubtedly throws a flood of light on the origin 

 of the Sirenia, and points clearly to their derivation from an ancestor 

 belonging to an Artiodactyle Ungulate with short-crowned and 

 selenodont molar teeth. It is, indeed, no new idea that the Sirenians 

 show Ungulate affinities, this presumed origin having been very 

 strongly urged by many zoologists ; although Professor Flower, 

 writing in the article " Manatee " in the ' Encylopeedia Britan- 

 nica,' expresses his opinion that the few facts at present known 

 relating to the ancestry of the Sirenians " lend no countenance to 

 their association with the Cetacea, and on the other hand their 

 supposed aflanity with the Ungulata, so much favoured by modern 

 zoologists, receives no very material support from them." If, 

 however, my interpretation of the affinities of the present specimen 

 be accepted, it will go a very long way towards solving the problem 

 of the Sirenian genealogj'. 



So far as I am aware, the component elements of the molar teeth 

 of the Sirenians have not hitherto been homologized with those of 

 mere typical teeth. The molars of the present specimen clearly show 

 us, however, the homology of the elements of the simple and con- 

 tinuous transverse ridges found in Manatus and Halitherium schinzi, 

 such ridges being clearly only one step more in the degeneration from 

 a selenodont type exhibited in the molars of the specimen before us. 



I may add that although the upper permanent molars of H. ver- 

 onense differ considerably from those of more typical species of 

 Halitherium, while there is no evidence that the latter had milk- 

 molars of the type of the present specim.en, yet I should not oa 

 ' Quart. Journ. Geol. Soc. vol. xxxi. p. 559, pis. xxxviii., xxxix. (1875). 



