E. (LOXODON) PRISCUS. 103 



almost equally rare. A single molar of that species was dis- 

 covered by Abbe Borson in 1820, in the Pliocene deposits 

 of the Astesan, since which date, up to 1856, not a single 

 additional specimen had been acquired for tbe collections of 

 Florence, Pisa, Turin, Milan, or Pavia, although the ossiferous 

 Pliocene strata of the Astesan had subsequently been largely 

 laid open by railway-cuttings. The detailed proofs have been 

 obtained from deposits in Auvergne and elsewhere in France. 

 In 1845, I was unable to reconcile the characters yielded by 

 the Grays Thurrock specimen with those of any recognized 

 species of fossil elephant, except E. prisms ; and after an 

 interval of twelve years, with a large addition of experience 

 in the investigation of the subject, and with more materials, 

 my conviction of its being distinct is as strong as that in 

 favour of any other species in the genus. In Mammalian 

 pakeontology, when the evidence furnished by the teeth can 

 be crucially tested, by means of the varied characters of the 

 cranium and of the bones of the extremities, a safe and 

 satisfactory conclusion as to the distinctness or otherwise of 

 the species can generally be attained. But when a few teeth 

 only are available, the area of the evidence becomes very 

 limited, and there is a constant unperceived tendency in the 

 observer either to magnify the value of the differential 

 remarks, or to underrate them, as the case may be, according 

 to his inclination, from some extraneous influence, to make 

 the species distinct or merely a variety of some other form. 

 In this case I have tried to guard myself against a bias either 

 way, and the evidence has appeared to me to be conclusive of 

 the distinctness of the species. Although so little is known 

 of the details of the different teeth, the ridge-formula is 

 inferred to be 7 : 7 •+ 8 + 11 in the last milk molars and 

 three true molars, as in the African Elephant. The known 

 limits to the dimensions of the molars in the Elephants, 

 coupled with the average great antero-posterior extent of the 

 ridges, in this form, namely, one inch to each, necessarily 

 involves a limited number of the latter. The distinction 

 from the African species is founded upon the characters that 

 the lozenges are regularly rhomboidal in the one, and some- 

 what crescentic, with the angular expansions more apart, in 

 the other. Both species belong to the ' Stenocoronine ' type 

 of Loxodon. The distinction from the fossil form to be next 

 described, E. (Loxodon) meridionalis, is borne out by well- 

 marked characters, the crowns of the molars in the latter 

 being constantly very broad, the digitations thick and distinct, 

 and the discs of wear free from mesial rhomboidal expansion. 

 If E. (Loxod.) prisons could be reconciled with any other 

 fossil species, it would be with E. ffiielep>has) antiquus, in 



