12 PROCEEDINGS OF THE NATIONAL MUSEUM Vol. G8 



down on its front face three strong ridges, two of them crossing 

 over to the buttress of the crest in front. 



The talon is composed of three conules, the largest on the outer 

 side, the smallest on the inner. One may satisfy himself that he 

 finds the same elements as in the fifth crest. 



The lower left hindmost molar is present and in fine condition, 

 except that the roots are injured. Few differences between this and 

 the one of the right side are observed. The length is 225 mm. At 

 the first crest the width is 91 mm. ; at the second, 95 mm. ; at the 

 third, 101 mm. ; at the fourth, 93 mm. ; at the fifth, 83 mm. Most of 

 the great root supporting the front crest is preserved. Measured 

 from the crown the height is 140 mm., and some part is missing. 

 The width is 70 mm. 



Of the upper tusks of the animal above described nothing is 

 known. The lower teeth of this species show relationships of their 

 possessor to Cope's Dihelodon troyicus^ a species based on H. von 

 Meyer's Mastodon huniboJdtii f^ found in Mexico and referred 

 by Freudenberg to the Pleistocene. Von Meyer's plate was repro- 

 duced by Dr. W. Freudenberg in 1922.^ The tooth from Aransas 

 River appears not to belong to that species for these reasons : 1, The 

 lower teeth of Anancus tropicus (Cope) appear to have possessed a 

 well-defined cingulum; nothing of the kind appears in the Aransas 

 tooth, except large tubercles in the ends of some of the valleys ; also 

 on the pretrite end of the fourth crest of the right lower tooth a row 

 of small tubercles represents the cingulum. 2, The pretrite cones 

 appear to have been less elevated. 3, The buttresses producing 

 trefoils were evidently weaker than they are in the Aransas tooth, 

 and the lobes of these are of a different form. 4, The fifth crest of 

 A. tropicus is far simpler in structure than it is in A. orai^us. 5, 

 The talon of the former consists of a single conule, in A. orarius of 

 three. This species is evidently in a higher stage of development 

 than that named and described by Cope. 



Cope in 1893'^° referred remains found in the Blanco beds to 

 Dibelodon tropicus^ first called by him Mastodon successor; but Os- 

 born has restored the name last mentioned. These teeth appear to 

 have trefoils on both halves of the two anterior crests. 



In 1911 " the writer characterized a large upper left hindmost 

 molar of a mastodon from Fort Worth, Texas, and referred it to 

 Anancus gratus. With a cast of this tooth at hand and a third 

 molar of the type of A. gratus and the Aransas River upper third 

 molar, the writer is inclined to refer the Fort Worth tooth (pi. 4, 



' Proc. Amer. Phllos. Soc, vol. 22, p. 7. 



« Palaeontgr., vol. 17, 1867, p. 64, pi. 6. 



• Geol. Palaeont. Abhandl., vol. 14, Heft. 3, p. 29. 



>" Vert. Palaeont. Llano Estac, p. 62, pis. 16, 17. 



"Publ. 322 A, Carnegie Institution, Washington, p. 13. 



