CHAPTER XV

ETHNOLOGY OF THE CHINESE FAR WEST

In the foregoing chapters I have attempted to give some account of a portion of that wild border country which constitutes the Far West of China, and most readers will perhaps agree that of all its striking features it possesses no peculiarity so remarkable and so puzzling as the number and diversity of the races by which it is peopled. At a meeting of the Royal Geographical Society in 1904 a well-known Oriental scholar, Sir George Scott, K.C.I.E., made the remark that "the country north of Tali-fu is the place where we shall find, if we ever do find, the solution of a great many of the puzzling questions of the different races who inhabit the frontier hills." Those secluded ravines and icy mountains have served as both the cradle and the death-bed of nations. From that region have issued vigorous and ambitious tribes, bent on a career of glory and conquest; and back to it the shattered remnants of decaying races have crept home to die.

The preceding chapters may have revealed something of the nature of the country from the geographical point of view. Fathomless chasms, towering cliffs and gloomy river-gorges are to be found throughout its length and breadth; and we need hardly be surprised to find that the strange variations which characterise its climate, scenery, fauna and flora, are faithfully reproduced in the vivid contrasts that exist among its many-tongued peoples.

The comparatively short journey from Tachienlu to Tali-fu has introduced us to some of these peoples, but these do not by any means exhaust the tribal varieties of this remote part of the Chinese empire. North of Tachienlu, up to the borders of Chinese Turkestan and Mongolia, there are semi-independent races, some of which are perhaps only remotely connected with any of those mentioned in this book; while the wilder parts of the extreme north and east of Upper Burma and the Shan States contain further ethnological problems of their own, which scholars have so far indicated rather than solved.

Most of these tribes disclaim any connection with each other, but it is impossible to believe that they are all of independent origin. Probably the scattered threads of their history will never be gathered up until scholars have found time and opportunity to study their social, physical and linguistic peculiarities by prolonged residence among them. Libraries cannot assist us much when historical records are entirely wanting or are obviously unreliable, and travellers who move rapidly from place to place can do little more.

[Sidenote: MIXTURE OF RACES]

Professor E. H. Parker sums up his remarks on the subject by expressing the opinion that most of the far western tribes of China "will be found to range themselves either under the Shan or the Tibetan head."[284] This is probably true enough if we give a very wide interpretation to the word "Tibetan," and also to the word "Shan." Our knowledge of Tibet has till recently been of a very fragmentary nature, and the veil of mystery that hung over that secluded land until it was

## partially torn away by Lord Curzon's Lhasa expedition, and by such

enterprising travellers as Sven Hedin, has prevented us from realising how far the Tibetans are from being a homogeneous race. No one who has come across the people of eastern Tibet and has also read the descriptions of western and central Tibet given us by recent writers and travellers, can fail to see that in spite of all that they possess in common the inhabitants of Tibet are a mixed people. "Long-heads" and "broad-heads," swarthy faces, white faces and yellow faces, long noses and flattened noses, oblique eyes and straight eyes, coal-black hair and brown hair, and many other physical peculiarities differentiate the people of one Tibetan district from those of another, just in the same way as they differentiate the various races of India and Indo-China. Nearly all the people of eastern Tibet have adopted the peculiar form of Buddhism which as Lamaism we have learned to associate with that country, and their languages and customs are saturated with Tibetan influences. In spite of many dialectical peculiarities I found that the people of the Yalung watershed were nearly always able to speak and understand Tibetan. Yet many of them are bi-lingual, and their own languages--as may be seen from the vocabularies in the Appendix--appear to be nearly as distinct from Tibetan as they are from Chinese.

Who are entitled to be called pure Tibetans--if such people exist--and who should be regarded as of hybrid or alien race, are therefore questions not very easily determined. All the country west of the Ta Tu river and the Chien-ch'ang valley may rightly enough be designated Tibetan Ssuch'uan or Chinese Tibet, if the name does not mislead us into supposing that the natives of the king of Chala's state, for instance, are of the same type as the inhabitants of Lhasa. But as these tribes are certainly not Chinese, what are we to call them if Tibetan is, strictly speaking, a misnomer?

[Sidenote: MAN-TZŬ]

By the Chinese many of the western tribes are more or less indiscriminately known as Man-tzŭ, Man-chia, Hsi Fan and T'u Fan. Now the words Hsi Fan and T'u Fan[285] appear at first sight to mean simply western Barbarians and aboriginal (or perhaps agricultural as distinct from pastoral or nomadic) Barbarians; but, as is now well known,[286] the old pronunciation of the second Chinese character in T'u Fan was not _fan_, but something like _po_ or _p'o_, which is simply the Tibetan word Bod,[287] pronounced Bö or Beu, the name by which the Tibetans describe their own country and people, and from which we derive the second syllable of our word Tibet. The character _t'u_ was similarly a Chinese approximation to the Tibetan word _teu_, meaning _upper_ or _superior_.[288] T'u Fan is then simply the Chinese equivalent of our own word Tibetan, and means the Bö(d) of the Uplands; and I assume that the name _Turfan_ (the oasis on the borders of Turkestan and Sungaria, where some remarkable discoveries have recently been made by exploring expeditions) has the same origin, though Turfan itself happens to lie in a very deep depression. The combination Hsi Fan, as Mr T. W. Kingsmill gives good reason for supposing, is in like manner derived from the sound Shar-bar, the name of a tribe of eastern Tibet still found near the town of Sung-p'an in north-western Ssuch'uan.

The word Man-tzŭ (蠻子)[289] is of very wide application, and at the present time conveys the meaning of "Savage Fellows" or "Sons of Savages," though it is not impossible that here too we have a rough attempt at imitating the sound of a non-Chinese word. When the Chinese had spread themselves over all northern China they used this term to describe all the uncivilised tribes whose habitations lay to the south; just as they described the "barbarians" beyond their western frontier as Jung (戎), those beyond their northern frontier as Ti (狄), and those of the east as I (夷). These terms all appear again and again in the ancient Chinese classics, such as the _Shu Ching_ and _Shih Ching_, and the references show that at the time to which they refer, that is to say as far back as the third millennium B.C., the Chinese were constantly at war with their less civilised neighbours, and by no means met with uniform success in contending with them. In the _Shu Ching_, for example, we read that "the invading barbarous tribes of the west (Jung) have greatly injured our empire."[290] The _Man_ tribes are in several places described as eight in number,[291] and we learn that in the reign of the more or less mythical king Yu (2204-2197 B.C.) their country was known as the Wild Domain,[292] and that Chinese criminals were transported thither when sentenced to exile,[293] much as the Russians send their convicts at the present day to Siberia. Sometimes in the Chinese classics the name _Man_ is combined with _I_ or _Jung_ as a definition of barbarians in general,[294] and sometimes the word _Nan_ (South) is prefixed to make the definition more specific. In Mencius we hear of "this shrike-tongued barbarian of the south (_Nan Man_), whose doctrines are not those of the ancient kings."[295] Naturally enough they are often spoken of contemptuously. "I have heard of men," says Mencius, "using the doctrine of our great land to change barbarians, but I have never yet heard of any being changed by barbarians."[296] Yet it is interesting to notice that Confucius was liberal-minded enough to admit that even a "barbarian" might--if he were truthful and honourable--be regarded as a gentleman.[297]

[Sidenote: MANZI AND CATHAY]

The name Man-tzŭ clung to the inhabitants of what is now southern China long after the Chinese had themselves begun to spread over that country, and no doubt many of the early Chinese immigrants and their descendants writhed under the derogatory epithet. About Marco Polo's time (in the second half of the thirteenth century) the southern portion of the empire--in fact, the greater part of China south of the Yellow River--was ruled by the emperors of the expiring Sung dynasty, who, owing to the successful invasions of the Chin Tartars, had been expelled from north China and had created a new capital for themselves at Hangchow (Marco's Kinsay) in the maritime province of Chekiang. The whole of their empire was known to the Venetian traveller as the land of the _Manzi_ or Man-tzŭ, as distinct from the northern (Chin-Tartar and afterwards Mongol) empire which he calls Cathay. He has handed down a circumstantial account of the splendour and wealth of the so-called Manzi capital, and he was an eye-witness of many of the stirring episodes in that long series of campaigns which overwhelmed the Sung imperial house and the Shan and Lolo princes of Tien, and established the descendants of Genghis Khan on the throne of a united China as emperors of the Yüan or Mongol dynasty.

In the Yunnanese Shan and so-called Lolo states, which were reduced to obedience by the Mongol prince Kúblái (afterwards emperor of China, and known to history as Kúblái Khan), the native tribes were too powerful and numerous to be exterminated. Great numbers, disdaining the Chinese yoke, migrated southward to Siam; some of those who remained behind were allowed to retain their tribal organisations under Chinese suzerainty, and to a limited extent they have retained it ever since. But in other parts of southern China the "barbarians" were much more harshly dealt with, for they were gradually broken up into small bands and forced to find for themselves a scanty subsistence in the rugged and mountainous regions of Kuang-si, Kuei-chou and Ssuch'uan, and multitudes seem to have fallen back on Annam. The term Man-tzŭ, as applied to all inhabitants of south China irrespective of race or descent, was then gradually dropped, but a curious instance of its survival in quite recent times is mentioned by Professor Parker,[298] who found, in an official proclamation, the word used to describe the Chinese of the Canton province.

[Sidenote: MAN-CHIA AND MAN-TZŬ]

Various notices of the Man-tzŭ and other hill-tribes are to be found in the monumental work of the historian Ssŭ-ma Ch'ien (about B.C. 100), and in the later dynastic and provincial records; but in none of them do we find anything like a clear statement of the history and origin of these tribes. The fact that they were all barbarians was sufficient, in the Chinese mind, to justify their being left severely alone or lumped together under some meaningless designation made applicable to them all.[299] At Tachienlu we come in contact with representatives of all the various tribes of western China and eastern Tibet, but they are nearly all labelled either Man-chia or Man-tzŭ. The former term means "barbarian families," and in practice is applied to the people whom the Chinese choose to regard as true Tibetans as distinct from the wilder denizens of the hills and forests. The Tibetan language is _Man-hua_ ("the language of the barbarians"), and the Chinese language is _Han-hua_ ("the language of the men of Han"). The term Man-tzŭ may now for practical purposes be restricted to certain of the western hill-tribes to whom both Tibetan and Chinese are foreign languages, and who preserve distinct customs of their own in the matters of dress, religion and social intercourse. A considerable proportion of the people who inhabit the scattered villages of the kingdom of Chala, through which lay my route to the Yalung, are Man-tzŭ, not Man-chia. M. Bonin, who has travelled widely in western Ssuch'uan, identifies the Man-tzŭ (using the term in the narrower sense) with the Lolos. In common with many other Europeans he has observed that the word Lolo, whatever it may mean, is an opprobrious epithet, which is not used by the Lolos themselves and should never be used in their presence. He considers that the word Lolo should be dropped altogether, and that we should substitute Man-tzŭ as the designation of both peoples. This word, he says, has the advantage of comprehending Mo-so, Hsi Fan, Ku-tsung, Menia and Li-so, who are, he considers, all of the same origin.[300] I venture to express a doubt whether we should gain much by classing under one such designation a number of peoples who, whatever their origin, have been so long separated from one another that they refuse to acknowledge any mutual connection, and to some extent have different customs and speak different languages.[301]

[Sidenote: LOLOS, MAN-TZŬ AND SHAN]

As regards the identification of the Lolos with the Man-tzŭ, however, there is good ground for believing that it is justified. Probably no one has a better acquaintance with the Lolos than the Catholic missionary, M. Paul Vial. He has lived for many years among the Nyi (or Ngi) Lolos of Yunnan, and has come to the conclusion that "Man-tzŭ et Lolos ne sont qu'une seule et même race."[302] His historical sketch is unfortunately too brief to be of much value. It would appear that in his opinion the great ruling power in Yunnan up to the thirteenth century was not Shan but Lolo. Indeed, his little book almost ignores the Shans altogether, though he states that, judging from linguistic evidence (which should always be accepted with very great caution) the Lolos are "brothers of the Burmese and cousins of the T'ai"--who, of course, include the Shans. Here, however, he seems to have gone a little astray, as his remarks would imply a closer relationship between Burmese and Shans than can be proved to exist; and when he says that the Lolo language has no relationship with the Chinese--which seems to be true--he overlooks the fact that the language of the Shans, whom he claims as cousins of the Lolos, is generally recognised as being related to Chinese.[303] He concludes that the Lolos, Shans and Burmese all belong to the same stock and came originally from the unexplored regions between the upper Mekong and the Brahmaputra; but he does not account for the subsequent divergence of languages, customs and traditions.

M. Vial has also some interesting notes on a Kuei-chou tribe called the Chung-chia-tzŭ.[304] Their own tradition, he says, declares that they came from the province of Kiangsi more than eight hundred years ago, and conquered the Ke-lao. These latter are, perhaps, Marco Polo's Koloman or Toloman, who Yule thought might have been a tribe of Lolos.[305] The Chung-chia-tzŭ are different from any of the hill-tribes round them, and are apparently related neither to Lolos nor to Miao-tzŭ. According to Terrien de Lacouperie, they speak a dialect "much resembling the Siamese, of whom they are undoubtedly the elder brothers."[306] If that is so, the Chung-chia-tzŭ must be related to the Shans, for both Shans and Siamese belong to the widely-spread T'ai family. Mr Warry, it should be noticed, identifies the Chung-chia-tzŭ with the Miao-tzŭ,[307] yet from M. Vial we learn that they differ in manners, customs and language.[308]

[Sidenote: MIAO-TZŬ, MO-SO AND LI-SO]

Mr F. S. A. Bourne, a first-rate authority, classes the Miao-tzŭ by themselves, for he believes that exclusive of the Tibetans (embracing Hsi Fan, Ku-tsung and others) there are three great non-Chinese races in south China: Lolo, Shan and Miao-tzŭ. Whatever their origin may be, the Miao-tzŭ have succeeded in planting representatives of their race in various widely-separated localities. We have seen traces of them in Muli-land, west of the Yalung,[309] and they are also to be found as far south as the Lao states and as far east as Kuang-tung. They call themselves Mung, Hmung or Hmêng, and it has been suggested that they are an Indo-Chinese race connected with the Môns, Peguans or Talaings.[310] M. Vial says that the Miao-tzŭ of Kuei-chou believe that they came "from the East," which is vague. In all probability no surviving race has been settled in southern China for a longer period than the Miao-tzŭ, and no attempt to connect them with the surrounding races has yet been successful.

As regards the Mo-so and Li-so, the people of those tribes whom I met between Yung-ning and Li-chiang denied there was any connection between them, and both were strenuously opposed to the idea that they were in any way related to the Lolos. Such denials, however, do not go for much, especially in the case of people who are totally lacking in any historical sense. The Mo-so of Yung-ning told me that they were an immigrant race and originally came from Mongolia, but this may be the result of confused reminiscences of their relations with the Mongol armies between six and seven hundred years ago. It is a well-ascertained fact that the Mo-so once occupied a large portion of south-western Tibet, and indeed there is a kind of national epic celebrating their wars with the Tibetans. At Li-chiang, as stated above,[311] they founded a capital which was the centre of a powerful principality, and they still have a prince near the Mekong river, south of Tse-ku. At times under weak rulers they were subject to the suzerainty of the great Shan kingdom of Nan Chao, the capital of which was generally at Tali-fu or not far from it; but at other times they were practically independent of any external control. It was not till Kúblái brought his Mongol troops to Yunnan in order to break up the Nan Chao kingdom as a preliminary to the overthrow of the Sung dynasty in south China that the political power of the Mo-so was laid low. Kúblái, in order to avert the possibility of being taken in the rear by hostile tribes, turned aside from his direct march to Tali-fu in order to reduce the Mo-so. He captured Li-chiang and broke up the Mo-so power about the year 1253. He subsequently besieged and took Tali-fu. The pacification of the newly-conquered province was entrusted by Kúblái to his great Mongol general Uriangkadai, and was successfully accomplished. The Mo-so, Lolos and Shans were never again able, with any hope of success, to defy the power of the emperor of China.

[Sidenote: MO-SO]

The origin of the word Mo-so is unknown.[312] They call themselves Lashi or Nashi (the _l_ and _n_ being interchangeable), and the Tibetans call them Djiung.[313] Perhaps they are the descendants of the Jung tribes which, as stated above,[314] are mentioned in the Chinese classics as having frequently menaced the western frontier of China; though it seems more probable that the Jung were the ancestors of the Hiung-nu. In a recent geographical work on China[315] the Mo-so are not referred to with much appreciation. They are described as deceitful and shifty, and a proverbial saying is quoted to the effect that three Chinese are necessary to deceive one Tibetan, and three Tibetans to deceive one Mo-so. Most of the eastern Mo-so speak Chinese as well as their own language, which bears various resemblances to Lolo. When I pointed out to some Yung-ning Mo-so that many common words in their language were identical with Lolo words conveying the same meaning, they admitted the fact but vehemently denied that it betokened any racial affinity. This attitude may be due to the fact that the Mo-so, once a warlike race, have settled down quietly under Chinese rule as peaceful tillers of the ground, while the Lolos have earned the reputation of being lawless freebooters. The Mo-so resents being taken for a Lolo, just as a sturdy Dumfriesshire farmer--whose ancestor may have been an expert cattle-lifter--would resent being described as the scion of a race of highway robbers.

The Yung-ning district, as we have seen, still enjoys a measure of independence under a native prince on whom the Chinese long ago conferred the hereditary rank of prefect.[316] The Li-chiang district is now more directly under Chinese rule, but even there a Mo-so official or noble acts as a kind of assessor to the local Chinese mandarins, who are still regarded as the representatives of a foreign power. The Tibetan name for Li-chiang is Sa-T'am,[317] by which it is also known to the Mo-so.

[Sidenote: LI-SO]

The Mo-so under their different appellations (including Lashi or Nashi[318] and Djiung) are still a very numerous though not a homogeneous race, and perhaps deserve a more careful study than they have hitherto received. I am strongly inclined to think that it is this race which constitutes the predominant element in the population of Muli-land or Huang Lama. We have seen above[319] that the people of that region call themselves Njong, and I conjecture that this is simply a thinly-disguised form of Djiung. The nasal prefix is a quite frequent linguistic peculiarity in Chinese Tibet, and occurs in many Tibetan words. The ordinary word _dro_, "to walk," for instance, is almost invariably pronounced _ndro_.[320] It may be allowed, however, that the people of Muli have identified themselves more closely than their brethren of Yunnan with the predominant Tibetan race, and have come more directly under Tibetan influences in respect of language and religion. For the people of Muli-land are, as we have seen, Buddhists of the Tibetan type, whereas with the Mo-so of Yunnan Lamaism is only a veneer that covers an even more uncouth system of witchcraft and sorcery, founded on the pre-Buddhistic Bon-pa.[321]

The Li-so,[322] judging from their language only, would appear to be rather closer to the Burmese than to the Mo-so. In the Yung-ning district, however, Li-so and Mo-so live together on amicable terms, and both express contempt or hatred for the Lolos. The Li-so are quite as widely scattered as the Mo-so, and may be found, apparently, in the Shan States and the Kachin highlands as well as in Yunnan and Ssuch'uan. They appear to be very closely related to the La'hu of the British Shan States, and they evidently regard themselves as racially distinct from the Shans, for they refuse to ally themselves in marriage with that people.[323] The Li-so language was examined by Prince Henry of Orleans, who found it like that of the La'hu or Muh-sö and that of the Lolos. He records a tradition among the Li-so that they came originally from Nanking, on the lower Yangtse, "which accorded with a similar tradition among the Lolos."[324] The Li-so of Yung-ning, when questioned by me, gave themselves the name of Lu-su.

[Sidenote: THEORY OF INDIAN ORIGIN]

A very interesting contribution has quite recently been made to the literature that bears on the ethnology of China's Far West by the researches of Mr T. W. Kingsmill.[325] That scholar presents a formidable array of evidence from Greek as well as from Chinese sources to prove that the _Sinae_ of the fourth century of our era were the inhabitants of _India extra Gangem_, namely the west side of what we call the Indo-Chinese peninsula, including Burma; that their capital, Thinae, was on the banks of the Irrawaddy, between Bhamo and Mandalay;[326] that they, "if not identical with the widely-extended people of the Shans," had at least a close ethnological connection with that race; that they and kindred races sprang from the great Maurya family of north-western India; and that to them is due the wide prevalence of Indian political influence and Indian art in the greater part of south-western China as well as throughout the Indo-Chinese peninsula and neighbouring islands. According to this view, which certainly receives some support from history, tradition, philology, and much miscellaneous evidence, the Man-tzŭ were originally of Mauryan stock, but allied themselves with the Böd tribes or Tibetans, with whom their migrations had brought them into close contact.[327]

An apparent difficulty in tracing Shans and Man-tzŭ and other tribes to a common origin in north-eastern India consists in the generally recognised affinity between the Chinese and T'ai peoples.[328] According to the commonly-accepted view the Shans sprang from somewhere in north-western China and were gradually pushed southwards as the Chinese race extended itself. De Lacouperie considered that the cradle of the Shan race was "in the Kiulung mountains, north of Ssuch'uan and south of Shensi, in China proper."[329] Mr Kingsmill's theory would perhaps gain more ready acceptance if we premised that the so-called Indian people from whom he supposes the Man-tzŭ and others to have sprung were themselves not of Indian origin but had entered India at some remote period--probably before either Aryans or Dravidians set foot in the peninsula--either as peaceful immigrants or as an invading host, from the countries that lay to the north-east.[330]

[Sidenote: VESÂLI AND THE MAURYANS]

Our knowledge of the early history of the Maurya family is unfortunately exceedingly scanty, and it is impossible to trace it to its pre-Indian home. To confine ourselves to their Indian history, the Mauryans seem to have sprung from the Licchavis, the strongest members of the powerful Vaggian confederation that dwelt near the Lower Ganges, north-east of the kingdom of Magadha. Just before the time of the Buddha--about the seventh and sixth centuries B.C.--a fierce contest for the mastery of northern India was waged between the kingdoms of Magadha and Kosala. This contest, as Dr Rhys Davids points out, "was decided in the time of the Buddha's boyhood by the final victory of Magadha."[331] About 320 B.C. the Mauryan dynasty under Chandragupta (Sandrakottos) overthrew that of Dhana Nanda and seated itself on the throne of Magadha, which, under a strong ruler, became more powerful than ever. From this time onwards till the extinction of Chandragupta's dynasty about 190 B.C., the Licchavi or Mauryan element was the main source of the strength of Magadha, which became the supreme power in the Indian peninsula. The royal adventurer Chandragupta Maurya was the contemporary and rival of Seleukos Nikator. Chandragupta handed on to his son Bindusāra and his grandson Asoka (the famous Buddhist "Emperor of India") the crown of one of the most powerful monarchies the world had known.[332] The capital of the Licchavis (as distinct from the Magadhans) was the city of Vesâli, which was probably situated about 25 miles north of the Ganges, north-east of Pataliputra (the modern Patna), which was the Magadhan capital. Buddhist records give us some remarkable particulars about Vesâli. "A triple wall encompassed the city, each wall a league distant from the next, and there were three gates with watch-towers. In that city there were always 7707 kings to govern the kingdom, and a like number of viceroys, generals and treasurers."[333] In another place we are told that these numerous royal persons were "all of them given to argument and disputation."[334] Allowing for Oriental exaggeration, these assertions certainly seem to imply that Vesâli, and the people whose capital it was, occupied a unique position in the political system of India.[335] There cannot have been many cities, even in that paradise of philosophers, which, in pre-Buddhistic days, or indeed at any other period, harboured thousands of disputatious kings. The so-called "kings," however, were probably only the heads of the free families. Vesâli was really the metropolis of a number of federated republics, the influence of which extended far beyond the boundaries of Hindustan.

The Licchavis may well have pushed eastwards into China and Indo-China long before the Mauryans gave India its first imperial dynasty; though if we find traces of their influence in western China it seems not improbable that this is due to the fact that after their migration to India they succeeded in maintaining a friendly intercourse with their Eastern kinsfolk. In either case, the Licchavis (and through them the kingdom of Magadha[336]) must have possessed, in the days of the struggle against Kosala, an enormous advantage over their rivals in being able to draw an inexhaustible supply of strength from Indo-Chinese countries to which the Kosalans had no access.

I have no space here to discuss the various arguments that Mr Kingsmill adduces to prove that the Man-tzŭ, Lolos and allied tribes, and perhaps the Shans, are the descendants of the Mauryas--some being more or less mixed with the Böd and Kiang[337] elements of Tibet. Suffice it to say that he traces the Maury an element in tribal names and place-names, in decorative and architectural art,[338] in Chinese records and tribal traditions, and by an analysis of the phonetic history of certain Chinese characters.

[Sidenote: NAN-CHAO]

The fatal weakness of the Indo-Chinese tribes appears at all times to have been their lack of cohesive power. At one time it must have seemed as though their empire would rival that of their Indian kinsmen--if kinsmen they were--in Magadha, and for centuries it might have seemed a doubtful question whether they or the Chinese were to be the masters of the vast country we now know as China. The great kingdom of Tien--which included the greater portions of Burma and Yunnan--was for centuries a formidable obstacle in the way of Chinese expansion towards the south, and it is only within comparatively recent years that Chinese suzerainty has been accepted by western and southern Ssuch'uan. The Shan kings of Tien or Nan-Chao sometimes arrogated to themselves the title of Huang-ti (Emperor), and frequently invaded Chinese territory. In 859 of our era one of these emperors besieged Ch'eng-tu, the capital of Ssuch'uan, and left "eighty per cent. of the inhabitants of certain towns with artificial noses and ears made of wood."[339] To this day there are thousands of square miles of nominally Chinese territory in which Chinese law is unknown, and with the administration of which no Chinese official dares to meddle. Had these various tribes--many of whom have the right, according to Mr Kingsmill's theory, to claim a common Mauryan ancestry--produced a few great rulers endowed with a genius for organisation, their history might have been at least as splendid as that of the Manchus and the so-called Mongolians, both of which peoples have given emperors to China. It is indeed possible that a dynasty of Mauryan blood did actually succeed for a few brilliant years in seating itself on the Chinese throne, though the evidence to this effect is far from conclusive. Mr Kingsmill, not content with identifying the Sinae with a people belonging to the same race as the Mauryans, has also found reason for identifying the Seres with the Man-tzŭ: that is to say, with a race descended from Mauryans and Tibetans. He conjectures that the people of the State of Ts'in (Ch'in) were connected with "the Mans in the south" rather than with "the Chinese in the north." He points out that we can trace the word Ts'in[340] and its homologues to an ancient pronunciation _Ser_,[341] and that when Virgil and other Roman writers mentioned the _Seres_ they were making use of a name which had become famous through the brilliant achievements of the Ts'in or Ser, who through the genius or good luck of Ts'in Shih Huang-ti had established a short-lived supremacy over the other peoples of the Chinese empire.[342] That ruler reigned from about 221 to 209 B.C., and therefore was almost a contemporary of the great emperor Asoka, who died only about eleven years before Ts'in Shih Huang-ti began to reign.[343] The famous episode of "the burning of the books" is said to have taken place about the year 213 B.C. It would be curious if it could be proved that, during the same century in which the great Mauryan emperor of Magadha was trying to inaugurate a new epoch of religion and peace by spreading the doctrines of the Buddha throughout southern Asia, another Mauryan ruler was sitting on the throne of China and inaugurating what he believed to be a new era of progress in north-eastern Asia by the destruction of the sacred books of China.

[Sidenote: HUNG WU'S EMPIRE]

The Min-chia, whose characteristic features seem to dissociate them from the Mo-so in spite of their proximity of habitation, are probably connected more or less closely with the Shans. M. Vial refers to them in a passage which I translate as follows. "In 1394, Hung Wu, emperor of the Ming dynasty, caused a map of the empire to be prepared in which the Yangtse is made to form the southern limit of China. In 1400, Chien Wên or Hui Ti, who was Hung Wu's successor,[344] was dispossessed by one of his uncles and withdrew to Yunnan, where he lay hidden for thirty years. A great number of Chinese followed him and established themselves there. They now form the basis of this Chinese population that we call _pên-ti-jên_ or Min-chia. They allied themselves to women of the indigenous race.[345] All these _pên-ti-jên_ say that they came from a place called Kao Shih Ch'iao of the province of Nanking."[346] The war between Hui Ti and his rebellious but too successful relative, the Prince of Yen, is a matter of history; and it is also stated in the Chinese chronicles that when the emperor was overtaken by hopeless defeat he escaped to Yunnan in the garb of a Buddhist monk. No doubt a number of faithful followers accompanied him into exile, but I am not aware of the evidence upon which M. Vial relies for his statement that they are the ancestors of the Min-chia. The Min-chia type is quite un-Chinese in appearance. That most members of the tribe speak Chinese is no strong argument in favour of their Chinese descent. It is a well-known fact that it was the deliberate policy of the Chinese emperors--especially in the early years of the present dynasty--to compel the conquered people of Yunnan to learn the language of northern China; and this policy was so wonderfully successful that at the present day nearly every one in Yunnan speaks a dialect which is easily intelligible to any one who has learned Pekingese. "The natives of Yunnan" as Baber said "were forced to learn the language of the north on pain of death." That a strain of pure Chinese blood must have mingled with that of the numerous races occupying Yunnan goes without saying; the mere presence of large Chinese armies on Yunnanese soil at times when campaigns lasted for a decade or more must of itself have tended to rub off the sharp edges of racial distinctions; but the special characteristics of the Min-chia are too well marked to justify the hasty adoption of the theory that they are the descendants of Chinese refugees from Nanking.

[Sidenote: TRADITIONAL EASTERN ORIGIN]

The number of different tribes who declare that they came originally from Nanking or elsewhere in the east is surprisingly large. I have already[347] referred to a tradition among the Chung-chia-tzŭ that they came from Kiangsi. The Miao-tzŭ of Kuei-chou apparently believe that they came "from the east." Prince Henry of Orleans records that the La'hu and Lolos both declared to him that they "came from Nanking ages ago,"[348] and mentions a similar tradition among the Li-so. That the Chung-chia-tzŭ, Miao-tzŭ, Lolo, La'hu, Li-so and Min-chia should have all come from the neighbourhood of Nanking seems scarcely credible, and the tradition with regard to most of them, if not all, may be dismissed as a fiction. But indeed I am aware of no theory about the Min-chia, or about Lolos, Mo-so, Li-so, Shans and the rest, that settles all difficulties and fits in with all the facts; and if one is tempted to put faith in any of the numerous hypotheses that have been advanced, it is only because a half-truth is not always "the worst of lies," and a permanent suspension of judgment is a source of discomfort to the mind that shuns the cheerless refuge of agnosticism.

FOOTNOTES:

[284] _China: Her History, Diplomacy and Commerce_, p. 9.

[285] 西番 and 土番.

[286] See Rockhill's _Life of the Buddha_, pp. 215-216, and T. W. Kingsmill's article in the _Journal_ of the Royal Asiatic Society (China Branch), vol. xxxvii. pp. 26-27.

[287] བོད་ The last letter of the Tibetan word is not pronounced, but it modifies the phonetic value of the vowel sound. As regards the Chinese character 番 of which the phonetic value in modern Chinese is generally _fan_, we find several cases in which the sound is still _bo_ or _po_. Mr Kingsmill mentions 鄱 _p'o_ (as in the characters used for the P'o Yang Lake). The characters 嶓, 皤 and 播 are similar instances.

[288] སྟོད་ as opposed to སྨད་ (_smad_, pron. _ma_), meaning _lower_, _inferior_.

[289] Often transliterated _Mantse_, and spelt by Marco Polo _Manji_.

[290] Vol. ii. p. 617 (Legge's ed.).

[291] As in _Shu Ching_, vol. ii. p. 345.

[292] 荒服.

[293] _Shu Ching_ vol. i. p. 147.

[294] _Ibid._ vol. i. pp. 42, 44.

[295] Mencius, p. 255 (Legge, 2nd. ed.).

[296] Mencius, pp. 253-254 (Legge).

[297] _Lun Yü_, pp. 295-296 (Legge, 2nd ed.).

[298] _China: Her History, Diplomacy and Commerce_, p. 310.

[299] See Note 40 (p. 437).

[300] _Comptes Rendus_, Société de Géographie, 1898. No. 8, p. 349. But see M. Paul Vial (_Les Lolos_: Shanghai, 1898). If M. Vial's theory of the origin of the word Lolo is correct, it was originally by no means a disrespectful term. He considers that it is a Chinese reduplication of a form of the word _No_ or _Na_, which was the special name of one of the patrician tribes of the Lolos. He admits, however, that the term is now regarded as impolite. He says that the Lolos have now no common name for the whole race, but simply employ the various tribal names as occasion requires. The Chinese characters for Lolo (generally 玀玀) are merely phonetic. The constant use of the "dog" radical in the Chinese characters employed to represent the names of barbarous tribes is an instructive indication of the contemptuous Chinese attitude towards such people. In the word _Man_ the radical is an insect or reptile.

[301] M. Bonin regards them all as of Tibetan origin; but as they separated from the main branch, he says, before the adoption of Buddhism they have preserved on Chinese soil their primitive fetish-worship. "I consider them in consequence," he concludes, "as the _avant-garde_ of the Tibetans."

[302] _Les Lolos_, p. 4. See also the _Gazetteer of Upper Burma_, pt. i. vol. i. p. 615, where it is stated that the Man-tzŭ "have undoubtedly been distinct from the Lolo for centuries, but the balance of opinion seems to connect them with that tribe."

[303] See the _Gazetteer of Upper Burma_, pt. i. vol. i. pp. 272 _seq._ "The relationship of the T'ai to the Chinese races seems unmistakable.... The research, which has not been long begun, points distinctly to the fact that the Chinese and the T'ai belong to a family of which the Chinese are the most prominent representatives."

[304] 重家子, or 重甲子.

[305] Yule's _Marco Polo_ (Cordier's edition), vol. ii. pp. 122-123. Cordier has, however, another explanation.

[306] Introduction to Colquhoun's _Amongst the Shans_, liv.

[307] See _Gazetteer of Upper Burma_, pt. i. vol. i. p. 597.

[308] _Op. cit._ p. 35.

[309] See p. 223.

[310] See _Gazetteer of Upper Burma_, pt. i. vol. i. pp. 597-601. There are numerous settlements of the Miao-tzŭ in the British Shan States, and the _Gazetteer_ says: "It may be hoped that more will come, for they are a most attractive race."

[311] See pp. 239 and 245-246.

[312] The Chinese characters are 摩□. It is tempting, but rash, to connect the word with _Mu-hsö_, which means "a hunter" in the Shan, Wa, Palaung, Rumai and Riang languages.

[313] The Tibetans also call them Jang or Aj'angs (འཇངས་). Surely there is some justification for tracing a connection between this word, as spelt in Tibetan, with the name of the tribe _A-ch'angs_ mentioned in the _Gazetteer of Upper Burma_, pt. i. vol. i. pp. 618-619. But see Sir George Scott's _Burma_, pp. 94-95.

[314] See p. 270.

[315] _Géographie de l'empire de Chine_, by Richards (Shanghai: 1905).

[316] See above, pp. 228-229.

[317] In Tibetan _Sa_ is "earth" or "land," and _t'am_ is "seal" (_sigillum_) or "offering." Possibly the Tibetan is in this case the transliteration of a Mo-so word.

[318] We have seen on pages 249-250 that the plain west of that of Li-chiang is called Lashi-Pa, or Plain of the Mo-so, and that a village therein bears the same name. M. Paul Vial mentions what he calls a _Lolo_ tribe named Ashi, apparently dwelling in the south-east of Yunnan (_Les Lolos_, p. 25). Now only a few miles west of Lashi-Pa, on the road from Li-chiang to Chung-tien, there is a village called Ashi, which gives its name to a ferry on the Yangtse river. It is possible that the sound in both cases was once either Lashi or Nashi, for, when we find from experience that the _L_ and _N_ are interchangeable, it may well be that in some districts inhabited by Mo-so the initial has been dropped altogether. I do not know the derivation of the word Lashio, the British settlement near the Salwen valley, in the North Shan States. There is also a district called Lashi, in British territory, north-east of Myitkyina, the people of which appear to be a connecting link between the Kachins and the Burmese. (See Sir George Scott's _Burma_, p. 70.)

[319] See above, p. 222.

[320] As in the common expression, _ka-li ka-li ndro a_, "walk slowly" or "there's no hurry."

[321] For some account of the Bon religion see Rockhill's _Life of the Buddha_, pp. 205 _seq._, and Sarat Chandra Das's _Journey to Lhasa_.

[322] 力□.

[323] Mr G. C. B. Stirling, quoted in _Gazetteer of Upper Burma_, pt. i. vol. i. p. 588.

[324] _Gazetteer of Upper Burma_, pt. i. vol. i. p. 616.

[325] _The Mantse and the Golden Chersonese_, and _Ancient Tibet and its Frontagers_, by T. W. Kingsmill, in vols. xxxv. and xxxvii. of the _Journal_ of the Royal Asiatic Society (China Branch).

[326] The name still survives in the province of Theinni and in the classical name Tien (滇) for the Chinese province of Yunnan. The connection between Tien and Theinni was pointed out by Terrien de Lacouperie in his introduction to Colquhoun's _Amongst the Shans_, p. xlviii.

[327] The fable is that a Mauryan woman was married to a Tibetan dog and that their progeny were the Man-tzŭ.

[328] See above, p. 275 (footnote 2).

[329] Introduction to Colquhoun's _Amongst the Shans_.

[330] See Note 41 (p. 438).

[331] _Buddhist India_, p. 260.

[332] See Note 42 (p. 439).

[333] Introduction to _Jātaka_, No. 149. (Cowell's ed., vol. i. p. 316.)

[334] _Ibid._, No. 301 (vol. iii. p. 1).

[335] See Note 43 (p. 439).

[336] "The struggle between Kosala and Magadha for the paramount power in all India was, in fact, probably decided when the powerful confederation of the Licchavis became arrayed on the side of Magadha." (Rhys Davids' _Buddhist India_, p. 25.)

[337] For the Kiang element, see Kingsmill, _Journal_ of the Royal Asiatic Society (China Branch), vol. xxxvii. 29 and 34 _seq._ The Kiang appear to have been a branch of the Yüeh-ti or Lunar Race, to which reference is made on p. 49.

[338] It is to the "Mauryan" Man-tzŭ that Mr Kingsmill ascribes the excavation of the caves of Ssuch'uan (see pp. 46 _seq._). He says that they were evidently the work of a people who had made considerable progress in the arts, and that the art in its predominant features approaches more nearly to ancient Indian types than to Chinese (_Journal_ of the Royal Asiatic Society, China Branch, vol. xxxv. p. 93). As I have already stated, there is not much evidence of a strong artistic instinct in the decoration of the caves. I agree with Mr Kingsmill, nevertheless, in ascribing the art, such as it is, to Indian influences.

[339] _Gazetteer of Upper Burma_, pt. i. vol. i. p. 267.

[340] 秦, pronounced Ch'in in modern Pekingese.

[341] In this connection Mr Kingsmill explains that the character _hsiang_ (象), which means "elephant," was also originally pronounced Ser. I have already mentioned a mountain-pass called the Ta Hsiang Ling which is supposed to be named after either P'u Hsien's elephant or Chu-ko Liang. (See p. 117 and Note 14.) To the south of that pass there is another named the Hsiao Hsiang Ling, or Small Elephant Pass, which must be crossed on the way to the Chien-ch'ang valley. Mr Kingsmill would perhaps translate the names of these passes as the Great and Small Passes of the Ts'in or Ser; in which case we may regard Ts'in Shih Huang-ti as being a third claimant to the honour of giving a name to this pass.

[342] See _Journal_ of the Royal Asiatic Society (China Branch), vol. xxxvii. pp. 22-23.

[343] See Note 44 (p. 440).

[344] Hung Wu was the "reign-title" of the first emperor of the Ming dynasty, who reigned from 1368 to 1398. His successor, whose "reign-title" was Chien Wên, ruled from 1399 to 1402. With regard to the Yangtse being taken as the southern limit of China, this statement can only be accepted with an important modification, for all the southern provinces of China, including Yunnan, were at this time regarded as being within the empire, though the fact that they were chiefly inhabited by non-Chinese tribes made it somewhat anomalous to describe them as forming part of China proper. We have seen that Yunnan was annexed to the empire by Kúblái Khan in the thirteenth century. Towards the close of the following century the Yunnanese princes tried to reassert their independence, and the province was again reduced to complete submission by the generals of the emperor Hung Wu himself, who, in spite of his maps, never for a moment intended to relax the imperial hold on that distant province.

[345] By "indigenous race" M. Vial presumably means Lolos or Mo-so.

[346] That is, Kiang-su, the province in which Shanghai is situated. Nanking was at that time the capital of China.

[347] See above, p. 276.

[348] _Gazetteer of Upper Burma_, pt. i. vol. i. pp. 585-586.

##