THREE EXTINCT SPECIES OF ELEPHANT. 299 



in E. antiquus the number is ten, and in E. primigenius and E. indicus it amounts to 

 twelve. 



" (2) True molars. — Of the antepenultimate true molar (m. 1) there is no perfect speci- 

 men in the collection*. But as in all the species of Elephant and Mastodon this tooth 

 invariably repeats the composition of the last milk-molar, we have no difficulty in fixing 

 the normal number of its ridges at eight, besides talons. In E. antiquus the number is 

 ten, and in the Mammoth and Indian Elephant twelve. 



" Of the penultimate true molar (m. 2) there is no entire specimen of a lower tooth ; 

 but we have the upper beautifully preserved, as shown in fig. 9 {vide note p. 296). It 

 exhibits a crown distinctly composed of nine ridges, besides a front and hind talon. 

 In the African Elephant the same tooth is commonly made up of nine lidges. In 

 E. antiquus the normal number is twelve, while in E. primigenius and E. indicus it 

 amounts to sixteen. 



" Of the last true molar (m. 3) there are fortunately specimens belonging to both the 

 upper and lower jaws; and although the portion supported on the anterior fang is 

 wanting in both, as that constantly corresponds in all the species of Elephant with what 

 is home upon the same fang of the penultimate, we have little difficulty in restoring the 

 missing part of the teeth. 



" The .upper molar exhibits the remains of ten ridges; and adding two for the part 

 corresponding with the anterior fang, we get a complement of twelve ridges for the 

 crown of the last molar. In the African Elephant the same tooth in the upper jaw 

 ranges with from nine to ten ridges, and in the lower from nine to twelve. In E. anti- 

 quus the number is sixteen ; and in E. primigenius the number of plates reaches twenty- 

 four f . 



" From the above data the ridge-formula of the molar series is deduced to have been 



Milk-molars. True molars. 

 3 + 5 + 8 . 8 + 9 + 12 

 3+0+8 ' 8+9+1/ 



This formula at once brings the small Zebbug species within the subgeneric group of 

 the Elephants which I have called Loxodon, along with E. africanus. The affinity of 

 the fossil to the existing species is further clearly indicated by the narrow crown and 

 mesial expansion of the disks of wear, together with the point already alluded to of the 

 milk-incisors being invested at the crown with a layer of enamel. But, though allied, the 

 two forms are specifically very distinct. Besides the signal diflFerence" of size, the forms 

 of the disks of wear, although belonging to a common pattern, present broad marks of 

 distinction. In the African species the disks are angularly dilated into rhombs in the 

 middle, and the angles terminate in a round loop caused by a single outlying digital 

 element, which in the progress of abrasion becomes confluent with the disk of the ridge 



* Unless, as I believe, we may regard fig. 9 as such, 

 t The numbers of plates were not filled in in Dr. Falconer's MS. 

 VOL. VI. PART v. 2 T 



