South Mongol Domain.

The Mongolian stock may be divided into two main branches[375]: the Mongolo-Tatar, of the western area, and the Tibeto-Indo-Chinese of the eastern area, the latter extending into a secondary branch, Oceanic Mongols. These two, that is, the main and secondary branch, which jointly occupy the greater part of south-east Asia with most of Malaysia, Madagascar, the Philippines and Formosa, will form the subject of the present and following chapters. Allowing for encroachments and overlappings, especially in Manchuria and North Tibet, the northern "divide" towards the Mongolo-Tatar domain is roughly indicated by the Great Wall and the Kuen-lun range westwards to the Hindu-Kush, and towards the south-west by the Himalayas from the Hindu-Kush eastwards to Assam. The Continental section thus comprises the whole of China proper and Indo-China, together with a great part of Tibet with Little Tibet (Baltistan and Ladakh), and the Himalayan uplands including their southern slopes. This section is again separated from the Oceanic section by the Isthmus of Kra—the Malay Peninsula belonging ethnically to the insular Malay world. "I believe," writes Warington Smyth, "that the Malay never really extended further south than the Kra isthmus[376]."

Tibet, the Mongol Cradle-land.

From the considerations advanced in Ethnology, Chap. XII., it seems a reasonable assumption that the lacustrine Tibetan tableland with its Himalayan escarpments, all standing in pleistocene times at a considerably lower level than at present, was the cradle of the Mongol division of mankind. Here were found all the natural conditions favourable to the development of a new variety of the species moving from the tropics northwards—ample space such as all areas of marked specialisation seem to require; a different and cooler climate than that of the equatorial region, though, thanks to its then lower elevation, warmer than that of the bleak and now barely inhabitable Tibetan plateau; extensive plains, nowhere perhaps too densely wooded, intersected by ridges of moderate height, and diversified by a lacustrine system far more extensive than that revealed by the exploration of modern travellers[377].

Under these circumstances, which are not matter of mere speculation, but to be directly inferred from the observations of intelligent explorers and of trained Anglo-Indian surveyors, it would seem not only probable but inevitable that the pleistocene Indo-Malayan should become modified and improved in his new and more favourable Central Asiatic environment.

Stone Age in Tibet.

Later, with the gradual upheaval of the land to a mean altitude of some 14,000 feet above sea-level, the climate deteriorated, and the present somewhat rude and rugged inhabitants of Tibet are to be regarded as the outcome of slow adaptation to their slowly changing surroundings since the occupation of the country by the Indo-Malayan pleistocene precursor. To this precursor Tibet was accessible either from India or from Indo-China, and although few of his implements have yet been reported from the plateau, it is certain that Tibet has passed through the Stone as well as the Metal Ages. In Bogle's time "thunder-stones" were still used for tonsuring the lamas, and even now stone cooking-pots are found amongst the shepherds of the uplands, although they are acquainted both with copper and iron. In India also and Indo-China palaeoliths of rude type occur at various points—Arcot, the Narbada gravels, Mirzapur[378], the Irawadi valley and the Shan territory—as if to indicate the routes followed by early man in his migrations from Indo-Malaysia northwards.

Thus, where man is silent the stones speak, and so old are these links of past and present that amongst the Shans, as in ancient Greece, their origin being entirely forgotten, they are often mounted as jewellery and worn as charms against mishaps.

The Primitive Mongol Type.

Usually the Mongols proper, that is, the steppe nomads who have more than once overrun half the eastern hemisphere, are taken as the typical and original stem of the Mongolian stock. But if Ch. de Ujfalvy's views can be accepted this honour will now have to be transferred to the Tibetans, who still occupy the supposed cradle of the race. This veteran student of the Central Asiatic peoples describes two Mongol types, a northern round-headed and a southern long-headed, and thinks that the latter, which includes "the Ladakhi, the Champas and Tibetans proper," was "the primitive Mongol type[379]."