(2)
may imagine the predecessors of all the present inhabitants of the Province to have from time immemorial moved down from the north, following the line of least resistance along the hill ridges or the river valleys, as the case might be, till they found a final resting place for their feet. By watching the movement of the Kachins we can guess how the Burmans, and indeed all the indigenous inhabitants of Burma, came.
As to the time and order of their coming we can form but the very roughest idea. There are chronicles that give us a general conception of how the ethnical elements in Burma were disposed at the beginning of the historical period. So far as they go they merely show a distribution of tribes, much as it exists now–Burmans and Talaings in the plains, Chins and Karens in the hills—a distribution, moreover, that is such that proximity cannot be looked upon as any test of relationship. Here and there, too, there has been such fusion of different tribes that even custom and legend is shared in common. What geography and history tell us is too often fallacious. It is language alone that shows relatively few anomalies and gaps and exhibits a development along the surest lines. So it is that if we are to attempt a classification of the peoples of Burma, we must look for our guide, not to chronicles or custom or folklore or propinquity on the map, but to speech, and only employ the other tests to check the criterion of language.
Now research has shown that, with the solitary exception of Salon, (the speech of the sea-gypsies of the Mergui Archipelago in the far south) all the languages spoken in Burma belong either to one or the other of two main language families, the Mon Khmer and the Tibeto-Chinese. Of these the Mon Khmer has comparatively few, the Tibeto-Chinese a large number of representatives in the Province. Of groups and sub-groups there are many, but all the Tibeto-
