1892.] FROM THE OLIGOCENE OF ITALY. 81 



apparently degenerate modification of the brachydont selenodont 

 molars of certain extinct Artiodactyle Ungulates. 



We may, however, go a step further than this. It is, I believe, 

 an attribute of all Artiodactyle Ungulates, whether their cheek-teeth 

 be of the bunodont or of the selenodont type, that while the last 

 upper deciduous or milk-molar resembles the permanent molars in 

 form, the penultimate milk -molar is an elongated tooth of a more 

 complex type than either the last milk-molar or the two succeeding 

 permanent molars. In the adult dentition, on the other hand, the 

 last upper premolar of Artiodactyles is nearly always simpler than 

 the first molar, and in none is it more complex or longer. 



Now the two teeth of the specimen under consideration present 

 precisely the same structural relationship to one another as is pre- 

 sented by the penultimate and last (third and fourth) upper milk- 

 molars of Artiodactyles. I take it, therefore, not only that these 

 teeth are third and fourth upper milk-molars, but likewise that they 

 belonged to an animal showing decided indications of Artiodactyle 

 affinities — these affinities being with an Artiodactyle that had 

 assumed selenodont molars more or less closely approaching the 

 Merycopotamus type. 



This being so, the question arises whether the specimen under 

 consideration may not have belonged to an actual Artiodactyle. To 

 this it may be replied that, so far as I am aware, no Artiodactyle has 

 hitherto been described possessing molar teeth of the type under 

 consideration ; so that if the specimen were really Artiodactylate, it 

 would indicate an entirely new form. 



Apart, however, from this, the structure of the second tooth in 

 the specimen presents such a marked Sirenian facies that, as I have 

 said, one is prompted to at once refer the specimen to that group of 

 mammals. If, moreover, it be compared with Baron Zigno's figure 

 of the upper molars of Halitherium veronense, it will be seen that 

 the last milk-molar of the present specimen accords in almost all 

 respects with these teeth. Both have two interrupted transverse 

 ridges, with a large anterior and a small posterior talon ; and in 

 both there is a tubercle a little to the inner side of the middle of the 

 transverse valley. Moreover, both teeth agree in the shape of the 

 crown ; while the excess in the size of the teeth of Baron Zigno's 

 specimen over those of the one under consideration is precisely such 

 as we should expect to find between the milk and permanent molars 

 of one and the same animal. It is true, indeed, that in Zigno's 

 figure what I may call the masked selenodontism of the teeth under 

 consideration is not apparent. This may, however, be due to the 

 circumstance that the teeth of the type of H. veronense are con- 

 siderably worn ; while it may also be in part owing to the difficulty 

 of expressing such features in a lithograph. Moreover, there is the 

 possibility that the masked selenodontism of the milk-molars may 

 not have been retained in the permanent dentition. 



In regard to the existence of a deciduous dentition in the Sirenia, 

 it is already known that milk-molars and premolars were developed 

 in Halitherium scMngi. It is, however, probable that in that species 

 Proc. Zool. Soc— 1892, No. VI. 6 



