the rigors of the ice-age, may have enabled it to survive that period, which was apparently so fatal to elephant life at large.
Summing up the details we have thus collated, from the geological side, we may now face the problem of the origin of the elephant race. Not that the problem itself is fully answerable, for our knowledge of the elephant race in the past is yet of comparatively limited extent; but the main lines of the biological argument are clear enough to those who will consider, even casually, the evidence already at hand. It is thus clear that the true elephants, which belong to the Pliocene period, are ushered into existence, so to speak, by forms that are less typical elephants mastodon and dinotherium when judged by the standard of existent elephantine structure. There are various species of mastodons known to geologists, which exhibit a gradation in the matter of their teeth, and presumably in other structural aspects as well, toward the ordinary elephant type. As the mastodons precede the ordinary elephants in time, we shall not be deducing an unwarrantable inference if we maintain that the origin of the true elephants, both fossil and living forms, may safely be regarded as arising from the mastodon stock. The elephants of to-day are connected by links of obvious nature with the Pliocene and Post-Pliocene forms; and, when the "ice-age" cleared the earth of the vast majority of the species, the progenitors of our living elephants must have escaped destruction and have survived the cold, possibly in the regions wherein they now exist, just as the mammoth, in its turn, survived the rigors of the ice-period, through the presence of its woolly coating and its hardier constitution. There seems thus to be no special difficulty, either of purely geological or of intellectual nature, in conceiving that the elephants of to-day are simply survivals of that elephantine host, whose existence was well-nigh terminated by the "ice-age," and which left the mammoth, and the progenitors of our living elephants, to replenish the earth after a catastrophe as sweeping and fatal in its nature as any deluge.
But if the origin of the modern and later elephants may thus be accounted for, and if their geographical birthplace may be assumed to exist within the confines of the Old World, a more fundamental and anterior query may be put with reference to the origin of the mastodon stock, which we have supposed, and with reason, is the founder of the true elephant races. From what stock, in other words, did the mastodons themselves arise? The chain of organic causation, to be perfect and complete, can not assume the mysterious origin of the mastodon. That stock must, in its turn, have originated in an ancestry less like the elephants than itself. It is not improbable that the evolutionist of the future will seek and find the mastodon ancestry in the dinotherium group, or in some nearly related forms. For, as we have seen, the dinotherium exhibits a structure which appears to relate the elephants to other and lower quadrupeds, such as the sea-cows