(16)
developing for hundreds, probably for thousands, of years along different lines. What relationship there is is based not only on similarity of vocabularies, but also on the fact that Karen, like Shan and Chinese, is a highly tonal language. The degree to which tones figure in the vernaculars of the hill people of Burma is a matter that has not yet been at all exhaustively gone into. All that need be said here on a somewhat vexed question is that, even granted that tones exist elsewhere, Karen is the only hill vernacular of the Province in which, so far as has been at present ascertained, they are a feature marked enough to have attracted attention in the past, Their presence points incidentally to another inference. So far as Chinese and its sister languages are concerned, it has by now been clearly shown that tones do not mark a very primitive but a comparatively advanced stage of linguistic development, and the fact that the Karens are in possession of a vernacular that has as many tones as either Chinese or Shan affords reasonable grounds for the presumption that they did not begin to occupy their present seats at any extraordinarily remote period of the world's history—probably not till after a good many of the other hill tribes who now inhabit Burma had entered into theirs. They were however certainly pre-Shan. At whatever time they may have come, it is clear that, in coming, both Tais and Karens followed a path lying midway between that of the Tibeto-Burmans and that of the Mon Khmers and bearing generally south-westwards. Crossing the Mekong the Karens probably entered wrhat is now the Shan States somewhat north of the neighbourhood of Karenni, and from thence spread westwards and southwards along the lower reaches of the Salween and Sittang into the Irrawaddy delta and the southern portion of the Tenasserim Division. In the Shan States the descendants of these early immigrants are now for the most part known by special names, such as
