ITS EARLIEST HEAD-QUARTERS. 249 



as M. Lartet has approximative!}- referred the Sevastopol 

 remains to E. meridionalis, coupled with the occurrence of 

 M. Borsoni either in Siberia or the Crimea, is strongly pre- 

 sumptive of Pliocene beds, yielding Elephants of a much 

 more ancient date than the Mammoth-yielding gravels of 

 the Urals. The dentition of E. meridionalis, in the ridge- 

 formula, is identical with that of the Miocene fossil Elephas 

 planifrons of the Sewalik hills, and the characters yielded 

 by the enamel-plates and discs of wear are also closely con- 

 formable ; while E. Armeniacus, as stated above, approaches 

 nearer to the existing Indian species. 



There is another point connected with distribution of fossil 

 Elephants over the European area, to which I am desirous 

 of directing the attention of palaeontologists. I now enter- 

 tain a strong suspicion that remains of E. Armeniacus, or of a 

 form closely allied to it, occur in Italy. This impression is 

 founded upon specimens which I observed in the Natural 

 History Museum of Turin, in the University Museum of Pisa, 

 in the private collection of the Marchese Carlo Strozzi at 

 Leghorn, and in those of Professor Ponzi and Signor Ceselli 

 at Rome, the satisfactory specific identification of which 

 puzzled me greatly. They certainly cannot be referred either 

 to E. meridionalis or to E. antiquus, from the high numerical 

 expression of their ridge-formula, nor do they appear suscep- 

 tible of identification with E. primigenius, without strainino- 

 the distinctive characters of that species to a degree which is 

 not warranted by our experience of it elsewhere. The first 

 which I shall adduce in illustration are a series of molars, 

 discovered in the Astigiano, during the excavation of the 

 railway section between Alexandria and Turin. One of them 

 is a huge last upper molar, right side, in the finest pre- 

 servation and half-worn. The crown is not quite perfect in 

 front, the portion borne upon the large anterior fang havino- 

 been worn down and broken off. What remains of it presents 

 no fewer than twenty-four ridge-plates, including the hind 

 talon ; and of these the twelve anterior ones are more or less 

 worn, the rest being intact. The crown is very broad in 

 front, and the plates, where unworn, are very high, as will 

 be seen by the dimensions annexed. The discs of wear are 

 transverse, without mesial expansion ; they are not so open 

 as in the Indian Elephant, but wider than in the Mammoth, 

 except in specimens of the latter worn low down ; and they 

 exhibit nothing of the retroflexion of the lateral cornua, 

 commonly seen in E. antiqmis. The enamel-plates are flex- 

 uose in the middle with decided crimping there, which does 

 not extend to the sides ; they are thicker than in E. primi- 

 genius, but less so than in E. antiquus. In this respect they 



