106 FAUNA ANTIQUA SIVALEXSIS. 



transverse pairs,' ^ with a narrow subordinate basal ridge in 

 front, and a qiiadritubercular talon behind. A pair of small 

 tubercles is placed in each of the valleys, in the long axis, 

 forming a connection between the transverse ridges. The 

 dimensions are stated to be 4 inches 10 lines in length, 

 width behind 2 inches 11 lines, and height of the middle 

 ridge from the base of the crown 2 inches 6 lines. According 

 to Professor Owen, the specimen presents a generic and 

 nearly specific identity with Cuvier's rej)resentation of M. 

 angustidens, as exhibited in PI. II. fig. 11, PI. III. fig. 2, and 

 PL I. fig. 4 of the ' Ossemens Possiles.' He considers that it 

 differs as much from M. anciustidens as the latter does from 

 M. Andium, and from the M. longirostris of Kaup in the 

 principal transverse ridges being more compressed antero- 

 posteriorly in proportion to the height, while they taper to 

 sharper summits. The dimensions correspond very nearly 

 with those assigned above to the penultimate of the lower 

 molar of 31. Andium. 



The occurrence in the Australian continent of fossil 

 marsupial tj^^es of the large Pachydermata, the dentition of 

 which so closely assimilates that of the latter that Dipro- 

 todon was in the fii'st instance referred to Dinotherium, 

 might raise a question whether the sj^ecimen under con- 

 sideration might not be a marsupial rej)resentative of Mas- 

 todon ; but Professor Owen, with good reason, believes it to 

 belong to a true Mastodon, founding his o^jinion upon the 

 complex form of the molar. A question of more weight 

 arises respecting the authenticity of the specimen as really 

 of Australian origin. It was not found in situ by the emi- 

 nent traveller, but prociu'ed by him from a native of New 

 Holland, at some distance from the asserted locality. The 

 tooth itself does not farnish characters sufficient to disting-uish 

 it from M. Andium ; and luitil further evidence is adduced, 

 the specific independence of M. Australis and its Australian 

 origin must, under these doubts, remain in abeyance. The 

 almost cosmopolitan range of Mastodon makes the presump- 

 tion more probable of a species of this genus extending to 

 New Holland than perhaps of any other genus among the 

 Pachydermata. The specimen belongs to the TrilopJiodon 

 section of Mastodons. 



ISTo species of this group has yet been discovered in the 

 fossil state in India. 



Mastodon, Section Tetralopliodon. — The group of species 

 for which we propose this sectional name is defined by 

 characters, in the numerical disposition of the crown ridges 



' Owen, Ann. of Nat. Hist. toI. xiv. 1 844, p. 269. 



