154—6 INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA. 
however, to have rather more open valleys than those of M. pandionis, hut the 
resemblance is greater between the lower teeth' : the Pikermi species has no 
premolars. Closer comparisons between the two cannot be made until some of the 
true molars of M. pentelici are known ; but it seems probable that both the latter 
and 31. pandionis are branches of the older M. angustidens stock. 
The associated milk-molars from the cranium of a young Siwalik mastodon 
figured in plate XXII., figs. 2, 3, and XXXIII., fig. 2 of the first volume of this 
work, under the name of M. faleoneri, but which have been shown on page 32 of the 
present volume to be distinct, differ from those of M. pandionis in the more open 
valleys of the intermediate molars, and in this res2)ect agree more nearly with M. 
pentelici ; which the Siwalik form also resembles in the absence of premolars. The 
resemblance between the penultimate milk-molars of the Siwalik and Pikermi forms^ 
is indeed extremely close ; and it appears by no means improbable that the two may 
really be specifically the same. 
1 Compare Gaudry, op. cit., pi. XXII., fig. 3, m. 3, with plate V., figs. 1, la of the present volume. 
2 Compare supra, vol. I , pi. XXXII., fig. 2, and XXXIII., fig. 2, with m. 2 of figs. 2 and 3 of plate XXII. of 
Professor Gaudry’ s work. 
