338 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [April 8, 



mens from the Crag have been as yet figured or described ; but the 

 characters presented by these teeth in 31. (Tetraloph.) Arvernensis 

 are well known, l)oth in the upper and lower jaws, through the spe- 

 cimens discovered by Croizet and Jobert, Bravard, and others in Au- 

 vergne or the Velay. The ante-penultimate presents two ridges, and 

 the penultimate three ridges, with the usual talon-complications. 

 They are readily distinguishable — the upper from (he lower — when 

 met with detached, from the circumstance that the milk-molars of 

 the lower jaw are narrower and more compressed, the ante-penulti- 

 mate being reduced to a single cusp. Figures of these teeth are given 

 by Croizet and Jobert in the work already referred to'''. 



The ridge-formula in the molar teeth of the Crag Mastodon, in- 

 cluding milk- and true molars, but exclusive of premolars, as inferred 

 from the various data detailed in the previous pages, is — 



Milk-molars. True molars. 



2+^3 + 4 . 4 + 4 + 5 



2+3+4 * 4+4+5 



The assigned numbers have not been verified in every instance upon 

 Crag remains ; but they are all founded on an examination of 

 specimens, of which some were of foreign origin, when materials were 

 not available from the Cragf . 



* Oss. Foss. du Puy-de-Dome, pi. 1. figs. 1-3 and p]. 2. fig. 7. 



i' In tlie descriptions of the various teeth throughout this memoir, the terms 

 ante-penidlhvafe, penultimate, and la^>t have heen used, instead of the uiuneral 

 expressions oi first, second, and third, when dcsigriating the position cither of the 

 milU- or of the permanent mohu's. This would seem indispensable when syndiols 

 are not employed, to avoid confusion in the designation of the milk-molars, since 

 the typical first or most anterior of the milk-molar series, which is present in 

 many other pachydermatous genera, is constantly suppressed in the Mastodons 

 and generally also in the Elephants. When, therefore, the terras first, second, 

 and third milk-molars are applied to Mastodon, they are not the equivalents of 

 the same numerals applied to Rhinoceros or Hippopotamus, in which all the four 

 milk-molars are developed ; whereas the terms aate-vcmdtiynate, pemdtimate, 

 and last in every case represent homologous teeth in the milk-molars of all the 

 ungulate genera. This is the more necessary, as the theoretical j'zr^^ or pre-ante~ 

 penultimate milk-molar is occasionally met with in the African Elephant. De 

 Blainville (Osteographie : Des Elephants, tab. ix. p. 81) has given an illustration 

 of its presence in the latter species, on one side of the lower jaw, regarding it as 

 a " supernumerary" tooth ; and a corresponding occurrence in the lower jaw of 

 the same species is rej)resented in the ' Fauna Antiqua Sivalensis ' (pi. 14. fig. 4 a). 

 It is usually restricted to one side ; and I regard it as not very uncommon. As 

 the true molars never exceed, nor are below, three in number in the Pachydermata 

 and Kuminantia, the same terms may be conveniently used in describing them. 

 The inconvenience of designating the molars in Mastodon and Elephas by success- 

 ive numbers ranging from 1 to (i or 7, which include both milk- and true molars 

 without distinction in the same numerical category, is exhibited in the descrip- 

 tions of the Elephants given throughout by De Blainville in his * Osteographie ; * 

 and moi*e recently in the otherwise excellent descriptions by Gervais of the den- 

 tition of Mastodon Andium in the ' Expedition de Castelnau.' The penultimate 

 and last milk-molars are there figured and described as the " third and fourth " 

 molars, involving a confusion of the ridge-formula, which is seen to be of so 

 much importance in the subgeneric distinctions (Recherches sur les Mammiferes 

 Fossiles de I'Amerique Meridionale, 1855, pp. 20-22, pi. 5. figs. 1-5). 



