[Sidenote:  The Atlantic abyss in historic movements of peoples.]

This deep, unbridged chasm of the Atlantic, closed only four hundred years ago, must be taken into account in all investigations of the geographical distribution of races, whether in prehistoric or historic times.  The influences of those ages when it formed an impassable gulf are still operative in directing the movements of the peoples to-day inhabiting its shores, because that barrier maintained the continents of America as a vast territorial reserve, sparsely inhabited by a Stone Age people, and affording a fresh field for the superior, accumulated energies of Europe.

[Sidenote:  Races and continents.]

Australia and the double continent of America show each the coincidence of an ethnic realm with an isolated continent.  In contrast, when we come to the Old World triad of Europe, Asia and Africa, we find three races, to be sure, but races whose geographical distribution ignores the boundaries of the continents.  The White race belongs to all three, and from time immemorial has made the central basin of the Mediterranean the white man’s sea.  The Mongolian, though primarily at home in Asia, stretches along the coast of the Arctic Ocean to the Atlantic shores of Norway, and in historical times has penetrated up the Danube to the foot of the Alps.  Nor was the Negroid stock confined to Africa, though Africa has always been its geographical core.  The Indian Peninsula and Malay Archipelago, once peopled by a primitive Negroid race, but now harboring only remnants of them in the Deccan, Malacca, the Philippines and elsewhere, bridge the distance to the other great Negroid center in Melanesia and the derivative or secondary Negroid area of Australia.[768] The Negroid race belongs essentially to the long southern land pendants of the Eastern Hemisphere; and wherever it has bordered on the lighter northern stocks, it has drawn a typical boundary zone of mingled tints which never diverges far from the Equator, from the Atlantic shores of the Sudan to Pacific Fiji.[769] [See map page 105.]

The effort of the old ethnology, as represented by Blumenbach, to make a five-fold division of the races in agreement with the five continents was a mistake.  To distinguish between the continents is one thing and to distinguish between the races is another.  Neither bio-geography nor anthropo-geography can adopt the continents as geographical provinces, although floras, faunas and races the world over give evidence of partial or temporary restriction to a certain continent, whence they have overflowed to other lands.  A ground-plan for the geographical classification of races is to be found, as Tylor says, in the fact that they are not found scattered indiscriminately over the earth’s surface, but that certain races belong to certain regions, in whose peculiar environment they have developed their type, and whence they have spread to other lands, undergoing modifications from race intermixture and successive changes of environment on the way.[770]