FOSSIL MAMMALS. 279 



Inchea. 

 II — 6. De Blainville, PI. xii, Bueuos Ayres - - - - - 9 

 T. De Blainville, PI. xii, Chile ------- 9 



8. Falconer, PL xxsv. Figs. 3 and 3°', (lower jaw) - - - 9 



9. Falconer, PI. xl, Fig. 14 ------- 8Jg 



From the above measurements it will be seen that these nine molars may be arranged in two 

 distinct groups : those in one measuring between six and seven inches in length, and those in 

 the other between eight and nine. The ninth specimen is nearly an inch shorter than the 

 other specimens of the same group, which may be attributed to the circumstance of its belong- 

 ing to the upper, while the others belong to the lower jaw. 



Not only does there exist this difference in the dimensions of the teeth, but there is good 

 evidence for the belief that a corresponding one exists in those of the lower jaw. Of these, De 

 Blainville gives the dimensions of three fragments, Ouvier of one fragment, and Falconer 

 of an entire mandible ; to these should be added the fragment described in this notice, the dimen- 

 sions of all of which are, respectively, recorded in the following table, an allowance having 

 been made for the last portions. 



Inches. 



1. Falconer, PL xxxv, Figs. 3 and 3*; entire - - - - 30 



2. De Blainville, Chile ; broken off at angle, length more - - 30 



3. De Blainville ; broken at symphysis and angle, about - - 20 



4. Cuvier, PL xxviii. Fig. 4 ; broken at angle, about - - - 20 



5. Specimen from Tagua-Tagua, aboxit ----- 20 



The estimated length of the broken specimens is based upon the proportions of the entire 

 mandible figured and described by Falconer, as above. The broken mandibles are all fractured 

 just behind the first molar, which corresponds very nearly with the base of the coronoid pro- 

 cess. The length of the jaw behind the base of the coronoid is a little more than one third of 

 the whole length. Although the above estimates have not the accuracy that is desirable, yet 

 they clearly indicate the existence of jaws which acquire quite different dimensions, viz : of 

 thirty inches and of about twenty inches. The ultimate molars contained in those of thirty 

 inches in length were nine inches, while in those of twenty they were six inches in length ; not 

 only do the shorter ones contain ultimate molars, but in one instance the tooth is ground quite 

 to its base, so that only traces of the transverse ridges remain. 



From the facts which have just been mentioned, we have strong evidence, in confirmation 

 of the opinion of Cuvier, that there exists a large as well as a small species of Mastodon in 

 South America. The lower jaw from Tagua-Tagua corresponds with those of the smaller 

 dimensions, and which Cuvier recognised as affording the basis for a distinct species. If the 

 existence of the second species— viz : 31. Andmm — be not admitted, the only alternative which 

 remains is to suppose that an ultimate molar may range in its length, in different individuals, 

 from six to nine inches, and the lower jaw from twenty to thirty inches. The existence of two 

 species — viz : M. Andium and M. Humboldtii — distinguished, as Cuvier stated, by difference in 

 size, seems by far the more probable view. 



Plate XIH, Pigs. 1 and 2. 



II. Sixth molar of Mastodon Humboldtii. — The single broken tooth which was sent in com- 

 pany with the lower jaw presents some peculiarities of structure which render it desirable that 

 it should be described separately. It is an upper molar, of which the anterior portion is 

 broken off; but a slight abrasion of some of the anterior points which remain shows that it had 

 come into use. Its greatest breadth is in front, and becomes, as is usual in ultimate molars, 

 gradually more narrow posteriorly. Its length is six inches, and its breadth three and a 

 half; it has four ridges remaining, and a conical nipple which forms the posterior talon. If 

 it had five ridges, which is the case generally in ultimate molars, its entire length must 



