18 
INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA. 
dinotheres and trilophodons of the lower Siwaliks of the west of India were, if not 
the actual progenitors, at least closely connected with the ancestors, of the tetralo- 
phodons, stegodons, and true elephants of the upper Siwaliks of the eastern side of 
that country, although from the existence of M. latidens in the lower Siwaliks of 
Sind some of these transitions may have taken place in the regions to the west of 
India; the higher forms alone having migrated eastward/ It is quite possible to 
imagine that the tetralophodont M. latidens, with its short mandibular symphysis, 
may have taken origin from the trilophodont M. angustidens,^ with its long 
symphysis ; and the transition from the former to the stegodons and higher 
elephants has been shown in the first volume to be so extremely gradual and 
complete, that it is highly probable that the one may be the ancestor of the other. 
If then, the true elephants took their origin from the stegodons characteristic 
of Eastern Asia there must either have been a subsequent western migration towards 
Europe ; or, which is equally possible, the stream went continuously eastward via 
China, Japan, and America.^ T]iat there should be difficulties in some of the 
above-mentioned views (which are put forth merely in the light of suggestions for 
further thought) is but natural. Thus if the tetralophodons of India took their 
origin in that country from the trilophodons, how did the iqDper miocene tetra- 
lophodon of Europe ( Jf. longirostrisy arise ? Could the different tetralophodons 
have had separate origins from the trilo^Dhodons ? If moreover true elejDhants 
originated in India during some part of the Siwalik period, there is some difficulty 
in seeing hoAv they should have reached Europe by the time of the iqiper pliocene. 
These difficulties, however, though great, are not insurmountable ; but before these 
views can be put into definite shape it requires a further knoAvledge of the 
proboscidians of China and Japan ; and some information regarding those of Persia 
and Asia Minor. 
Genealogy of the elephants . — A few observations on the genealogy of the 
elephantine Proboscidia arising from the study of the sjjecimens described in the 
sequel may be here recorded. The presence in some of the mastodons of simple 
tetraconodont premolars like those represented in pi. Y., figs. 2 and 6, suggests the 
probability of the descent of these animals from some primal ungulate, furnished 
with teeth of this simple structure, in which the premolars were as fully deAmloped 
as the molars. The replacement of the three-ridged last lower milk-molar of the 
pigs by a much simpler premolar, is very similar to the replacement of the 
corresponding upper tooth in the trilophodont mastodons by a tetraconodont 
1 The stegodons are known to extend to China and Japan ; Elephaa namadiem also occurring in that country. 
2 The ridges and valleys in this species are not quite so simple as in M. laiidens ; hut there seems no reason why the 
modification in one direction may not have tended to simple, and in another to complex ridges. 
* The association of E. primigenhn with E. namadicus in Japan and the occurrence of the former in N. America is in 
favour of this view : and it has been shown in the previous volume that the migration of the machaerodonts not improbably 
followed a similar course. 
4 The European tetralophodons (df. lovgirostri!:, M. arvernensik), from the very irregular ridges of their molars, cannot 
be looked upon as on the direct ancestral line of the true elephants. 
