The English Historical Review/Volume 20/Number 80/China and the Ancient Cabul Valley

by Edward Harper Parker

China and the Ancient Cabul Valley

THERE is no mention of the considerable state which the the Chinese call Ki-pin until the beginning of the first century before Christ. The Chinese had then for a generation or so already discovered the West: they were not only able to assert themselves up to the same mountain limits that bound the empire now, but they had also some knowledge of Parthia—that is, Persia—of the whole Oxus and Jaxartes region, and of a large state which completely barred their way to the Indus valley and India. This state, in imitation of the native sound, they called Ki-pin or 'Shawl-guests.' In naming new states the Chinese either gave them, and still give them, a well-known Chinese name, often with a differentiating word added—as, for instance, we say New South Wales;—or they imitated as closely as convenient the foreign sound by using one or more Chinese characters, either as pure phonetics or as also descriptive of some striking local peculiarity. Thus Ki-pin was noted for its ki, or 'shawls,' 'rugs,' &c.; and to pin, or 'to guest,' meant 'to come to court with presents or homage.'^[1] The earliest political situation known to the Chinese was this: The Hunnish or Scythian tribes (first called 'Turks' in A.D. 550^[2]) had, in the course of their predatory wars, driven one large tribe of their remoter kinsmen from the Chinese frontiers to the Jaxartes.^[3] These emigrants displaced the various Greek and (it is presumed) other Aryan rulers then in possession of the Oxus valley, and gradually worked their way south; forming at last, between the Jaxartes and the Indus, the vast empire of the Kushans, or Indo-Scythians. The Chinese were too late to witness with their own eyes the remnants of Greek active rule, and considered that most of the dynasts thus displaced belonged to a race called Séh or Săk.^[4] The Indo-Scythians in their further conquests southward from the Jaxartes, but before they themselves approached India, drove the Săk rulers from the Oxus valley further south to Ki-pin, In other words, they took possession of the rich valleys, and drove the displaced Săk rulers to the poorer hill tribes for a living.

The positive statements of the Chinese, made before the beginning of our era, show almost conclusively that Ki-pin must then have included most of the mountainous region embracing Candahar, Ghazni, Cabul, Chitral, Jelalabad, Cashmere, and Baltistan. The argument is as follows: One thousand years ago P'i-shan was annexed by Khoten, and the Ta-ts'ing Yih-t'ung-chï, or 'Official Geography,' of the present Manchu dynasty—a work in 500 books published about 1760—tells us that the site of ancient P'i-shan is 'the modern Bishnam, between Khargalik and Isse-Kul.' Although there may be doubts as to what is really meant by Isse-Kul and Bishnam, there is no doubt whatever about Khargalik, which lies south-east from Yarkand, about two-fifths of the distance that Yarkand lies south-east from Kashgar. Now, 2,000 years ago, we are told that from P'i-shan you go east to Khoten 380 li (120 miles), and from P'i-shan you go north-west to Yarkand exactly the same distance. Hence both ancient and contemporary P'i-shan must be the first considerable town north of Shahidula, in the immediate neighbourhood of Sanju. Ki-pin was 3,790 li west-south-west of P'i-shan on the way to U-yih-shan-li. We know from Mr. Stein's recent observations^[5] that from Rawal Pindi to S'rînagur is 200 miles, from S'rînagur to Leh 260 miles, and from S'rînagur, viâ Skardo, northwards to Misgar 500 miles. Hence from S'rînagur viâ Skardo to Shahidula would be about 700 miles; or, say, 900 to 1,000 miles from Rawal Pindi to Sanju. It is well known that Chinese li are calculated as though shorter in distance but longer in time over mountainous roads, ten li being, in fact, under all varying conditions, the German Stunde. Moreover the ancient Chinese Geography known as the Shwei King, or 'Rivers Classic,' a work of which the original dates from the beginning of our era, and which is preserved in an edition of the fifth century, mentions a stream which rises among the Onion Mountains (Muztagh) and flows south-west through Ki-pin. This must be either the Hydaspes (Vitasta or Jhelum) or Gilgit and Hunza branch of the Indus.

But there are other specific grounds for identifying the chief part at least of Ki-pin with Cashmere. According to the Han Shu, which was published A.D. 90,^[6] the land is flat, the climate mild, the low-lying portions being wet and suitable for rice fields. Gold, silver, and copper are plentiful, and are fashioned into utensils. There are humped cattle, water buffaloes, elephants, large dogs, monkeys, and peacocks. Pearls, coral, amber, agate, and glassware are to be found. The people are ingenious carvers, good cooks, and clever at weaving ki (shawls and rugs of hair). In the bazaars silver and gold coins are used, the obverse showing a riding-horse, and the reverse a man's face. The whole of this account refers to the period when the Săk rulers were still in possession, and before the Kushans, advancing over the Hindu Kush, or Paropamisus, carried their conquests up to the Indus. The same record adds that China had her first relations with Ki-pin between 141 and 87 B.C. 'The king frequently robbed our missions. His son, however, sent envoys to us, but was killed by the Chinese envoy sent by us in return after the king in question had unsuccessfully tried to murder the said envoy.' The third king quarrelled with a second Chinese envoy and massacred him and his suite. The reigning Chinese emperor (B.C. 48–32) felt he had not the means adequately to punish this distant offence; so he took no severer steps than to dismiss the Ki-pin envoy from China, and to deliver him safe at the Hien-tu Passes;^[7] political relations were then broken off. During the next reign (B.C. 32–6) it was proposed to renew intercourse with Ki-pin; but, on the ground that the Hien-tu range presented an insuperable physical barrier to the Ki-pin men, who were in consequence quite unable either to injure or to assist China by force of arms, it was resolved to confine future intercourse to admitting Ki-pin trading caravans. By way of further illustrating the difficulties of mountain travel, it is explained that, leaving P'i-shan to your north, you pass through four or five states independent of China; after 2,000 li (666 miles) you reach the Hien-tu range, crossing the Headache Mountains on the way.^[8] Finally it is added most distinctly that neither the Parthians nor the Kushans nor the Ki-pins are in any way vassals to China.

The next dynastic history, the Hou Han Shu, or Continuation of the Han Shu, of which parts were not published until the fifth century of our era, deals with the period A.D. 1 to 200, but tells us very little of Ki-pin. The five Indo-Scythian kingdoms had now been united into one great empire by one of the five rulers, the prince of Kwei-shwang. This union seems to have taken place before B.C. 100; but although the Chinese of that time knew of the quintuple division, they appear not to have heard of the reunion for at least another century. Moreover, although even then they knew that Kuei-shwang (i.e. Kushan) had become the national designation, they persisted in using the ancient Chinese designation of Yüeh-chï or Ta Yüeh-chï (Great Yüeh-chï). About a hundred years after the union the Kushan ruler invaded Parthia, and took from it Kao-fu.^[9] All south-west of Kao-fu to the sea belonged to India. The people of Kao-fu; says the record, were in most respects like those of India, but very unmanly, though rich and good traders. The conqueror, who is identified by orientalists with Kadphises I, then went on to annex the rest of Ki-pin, and his son (supposed to be Kadaphes) extended his sway over India. The Chinese had now become well acquainted with India, and a few years later (A.D. 65) Buddhism was introduced thence into China by way of the Indo-Scythians and Khoten.^[10] A stray passage in the Wei Lioh, as quoted in the Wei Chï (a production of the fifth century which deals with the years 220–264), states that in A.D. 2 an Indo-Scythian king had busied himself with teaching the tenets to a certain Chinese. The Hou Han Shu adds that Kao-fu, or Cabul, being so unwarlike, used to fall alternately into the hands of Parthia, India, or Ki-pin, whichever had happened to be the strongest; so that, until the Kushans of Bactria and Tokhara conquered considerable parts of all the three states enumerated, there were no other 'powers' at hand capable of annexing Cabul.

For a couple of centuries (200–400) Chinese influence in the Western Regions hardly existed. China had split up into the northern, or Tartar-ruled, and southern, or native-ruled, rival empires. The latter only concerned itself with foreign intercourse by sea. The former only consented after mature deliberation to reopen what were characterised as the dangerous and compromising relations with the west. The period from 400 to 600 was, however, particularly rich in Buddhistic travel and missionary work. Large numbers of Hindoo, Indo-Scythian, Ki-pin, and other mixed western bonzes or peripatetic philosophers wandered over China; and almost equally numerous Chinese religious adventurers prowled about Tartary, 'Turkey' (i.e. the comparatively fertile Hunnish and Scythian strip between Siberia and the Chinese empire, after the name 'Turk' came into use), Turkestan, Samarcand, Bactria, Tokhara, and India. More distinctly separate embassies from the West came to China during this period than had ever come before, or have ever come since; but they were trading caravans or missionary bodies; there was no fighting, and but little of active politics.^[11] In 439 the North China mission returning from India and Ki-pin with their escorts happened to meet near Lob Nor, where they were able to render valuable assistance to the ruler of that region, at the moment threatened by the civilised Scythians ruling at Liang Chou.^[12] The history of the North China Tartar-governed empire known as the Wei Shu, and the Pei Shï, a work of the sixth century, treating of the period from 300 to 580,^[13] tell us that in the year 507 K'a-pi-sha (Kapiça) sent complimentary envoys to China along with those of Kashgar, Samarcand, and other neighbouring states. The same history states that in 451, 453, 502, 508, and 517 Ki-pin sent similar envoys; also that in 511 Gandhâra, S'râvasti, K'a-shï-mih (Cashmere), and Pu-liu-sha (Purusapura, or Peshawur) sent missions. These exact dates prove that, so far as the Chinese historiographers knew or officially admitted, Kapiça and Cashmere, both new names to them, existed as states alongside of Ki-pin; and that Ki-pin, however much reduced by Indo-Scythian encroachments, undoubtedly existed as a separate state after its fall, just as after the collapse of Parthia, or Arsac, as the Chinese call it, a petty remnant of the old Arsac state continued to exist at or near Bokhara.^[14]

The history of this Tartar dynasty (the Toba family) tells us little new of Ki-pin, and merely repeats the old descriptions of the Han dynasty given above. It adds, however—what is important in estimating topographical positions—that Ki-pin was surrounded by mountains, and that it was 300 li (100 miles) north and south by 800 li east and west in extent. In 520 the Chinese pilgrim Sung Yũn, after visiting Udyâna, proceeded to Gandhâra, whose king he found at war with Ki-pin, on the left bank of the Indus, elephants being employed in the war. Sung Yün does not mention Kapiça or Cashmere by those names.^[15] Since Kápishi wine was, according to Yule,^[16] well known as early as B.C. 300, we may conclude that it had always been a petty dependency of Ki-pin or some other great power. As the pilgrim Hüan-tsang gives it a circuit of 4,000 li in A.D. 638, and ascribes to it an hegemony over ten other states, it seems likely that in his time part of the old state of Ki-pin had locally acquired the new generic name of Kapiça, to distinguish it from the Turkish portion of Ki-pin.

The seventh century opens with the reconquest of North and South China as one whole under two purely native dynasties, reigning from 581 to 618 and from 618 onwards. The former of these was anxious to reopen communications with the East Roman empire and with India; but it failed in both cases, and its emissaries, as is explicitly admitted, never penetrated beyond the Oxus region in either direction. The Sui-Shu published in 656, which records the history of this enterprising Sui dynasty, gives in chapter 83 an account of a country, hitherto all but unheard of, 1,700 li (570 miles) south of Tokhara, called Ts'ao, which has been identified with Tsâukûta, i.e. Ghazni or Zabulistan, southwest of Cabulistan.^[17] But this account is a mere copy of what had already been stated of the same country in recapitulation by the Tartar history, which, as we have said, includes an account of Ki-pin, copied from the earlier history of the Han family. This Ts'ao state was 700 li south of Bamyân. It is distinctly identified in 580–606 with the Ki-pin of Han times, but nothing is said of its capital city, except that it was over a mile in extent, and moreover it is plainly stated to be, or to have recently been, under the rule of the Indo-Scythian princes of Samarcand and the Oxus region—that is, of the later Kushans or Ephthalites. In other words, although previously to Sui times the Turks and the Persians between them had annihilated the Indo-Scythian or Ephthalite empire, the Sui history, as is the Chinese wont, continues to repeat partly obsolete stories. The productions of Ts'ao in A.D. 600 are described much as those of Ki-pin were in the earliest histories for A.D. 100, including the elephants, the shawls or rugs, the humped cattle, metals, drugs, &c.; but in such a way as to suggest a quite new and independent nomenclature from fresh observations, and no mere copying. The remarks upon religion prove that either Christianity or the Persian 'heresies' had partly, if not wholly, displaced Buddhism there.

The next dynastic chronicle is that contained in the Old and New T'ang Shu, which were published respectively about 750 and 1060,^[18] and relate the history of the T'ang family, 618–906. This reverts to Ki-pin, and calls its capital by the same name as in the earliest times. At the same time it states that this Ki-pin is identical with the Ts'ao of Sui times. The Headache Mountains, already mentioned, are named as being south-west of Chu-kü-po (Kugiar). Whereas the Sui history says Ts'ao was north of the Onion range (Hindu Kush), the T'ang history says south but, as the T'ang history also says Ts'ao was south of Bamyân, that discrepancy may simply mean that the capital of that part of old Ki-pin known as Ts'ao had, during the time when the comprehensive name Ki-pin was in local abeyance, been established or moved further north—for the Indo-Scythians had themselves, under competitive pressure, shifted their capital further west—only to be retransferred to the old site (presumably S'rînagur) when the Turks took over the Ephthalite empire and began to threaten Persia and Ki-pin. It is stated that the Ki-pin rulers were in the habit of impressing the young men of Zabulisian to repel the incursions of the Turks, and that the population of Zabulistan was made up of three races—Turk, Ki-pin, and Tokhara men. West India is said to be conterminous with Persia and Ki-pin, and Udyâna was 400 li east of Ki-pin, which again was 3,000 li north of S'râvastî. These statements seem contradictory; for, according to Bishop Bigandet,^[19] Udyâna was on the Indus, between Cabul and Cashmere, and other western writers of repute have placed it between the Jhelum and the Indus. It is possible, however, to explain this contradiction too, for the T'ang history still places Ki-pin under the Indo-Scythians, and probably the name of the Ki-pin capital and the fact of vassalage to the Ephthalites were mechanically copied from earlier histories. The Sui history says that this particular Ts'ao once, between 605 and 617, sent tribute; but this is probably a mistake for another country of nearly the same name which actually paid it in 614. The T'ang history says that, though thirty states in that region sent tribute to Sui, Ki-pin was not one of them.

The T'ang history gives an entirely new account of Ki-pin and its productions. Amongst other curiosities were animals of the mongoose type, with a natural antipathy to or capacity for destroying snakes. As these creatures were brought to China in 642, and the Persians sent exactly the same kind of animals to China in 638, it seems plain once more that there was a closer relation with Persia than with India on the part of Western Ki-pin. Moreover the presents of crystal cups, glass beads like dates, gold belts and chains, &c., suggest Western rather than Indian influence. The ancient local kings who had ruled under the Ephthalite domination were evidently still ruling down to 642, when the king reported himself to be the twelfth in succession of his house. In 658 the Western Turks,^[20] whose extensive rule had never done more than rub shoulders and negotiate intermarriages with the Hindu Kush states, were displaced by direct Chinese rule. The king of Ki-pin was made (Chinese) military governor of Siu-sien and the other ten provinces belonging to it. Siu-sien (perhaps S'rînagur) was the ancient capital; but it by no means follows that the king was there in 658. This Chinese arrangement continued for over a century; but from the date of western Turkish collapse the kings or military governors of Ki-pin always seem to have borne Turkish names; in fact, the Turks, having split up, appear to have sent off branch adventurers in all directions. In 739 the teghin (old Turkish for 'grand duke') of Zabulistan, south-west of Ki-pin, was created by the emperor of China king of Ki-pin: this teghin bore exactly the same name as the ruler of Gandhara. The celebrated pilgrim Hüan-tsang, a century before this, tells us that Kapiça had annexed Gandhâra. The king of Kapiça's dominions extended to the Indus, and the king himself saw the pilgrim safely into Tsâukûta, which, it is stated, was then ruled by a prince of Turkish stock. It took Hüan-tsang fifteen days to travel south-east, and then east, from the capital of Bamyân to the first frontier of Kapiça. The son of this king of 739 bore the surname of Fu-lin (meaning, according to my view, 'Frank'), whence I assume that intermarriages with Asiatic subjects of the East Roman empire had taken place; for it was the Turkish practice to use tribal and national designations as personal cognomens (e.g. that of Ouigour). In 745 the son—or possibly the brother—of this last named prince was patented king of Ki-pin and Udyâna.^[21] The T'ang history uses the word Ts'ao-kü-ch'a (Tsâukûta) as the older name of the Chinese military governorship at Hoh-sih-na (Ghazna or Ghazni), created in 658. After 710 this small state became subordinate to Ki-pin. It was also called Holatachï^[22] or Otalachï, and its princes were evidently of Turkish stock.

There is henceforward no further mention in Chinese history of Ki-pin as an actual state. But the T'ang history for the first time gives some account of Ku-shih-mih or K'a-shih-mi-lo (both manifest attempts at Kas'mîra or Cashmere). North of it was Puh-lüh, or Polt (Baltistan): it was encircled by mountains forming a ring of 4,000 li (1,338 miles), and thus was almost impenetrable to hostile attack. Tradition said it was once a lake, the 'Dragon Pool' of antiquity, and had only been populated by mankind when the dragon and the waters had retired. The state consisted of or included five other divisions or sub-states—to wit, in all, Taxila, Sim̃hapura, Uraça, Kas'm̃ir, Parṇotsa, and Râdjapura.^[23] Missions to China were sent in 713 and 720, on the latter occasion with presents of foreign medicines. It first became mixed up with Chinese politics in this way: in 723 the Chinese princess married to the king of Tibet, during the Tibeto-Chinese war, applied for asylum to Tchandrâpida, the king of Cashmere; and he, doubtful of his power to resist the vengeful Tibetan arms, passed on the request to the king of Zabulistan.^[24] In 733 the king's brother and successor Muktâpida was invested with a Chinese title, and announced his intention, with Hindoo co-operation, to assist with supplies any Chinese army, however large, that should come to Baltistan in order to turn the Tibetan flank: he said there were five roads by which the Tibetans might come in separate columns, and added that the guardians of the Mahâpadmanâga (Dragon) Pool would feel honoured by the establishment of a Chinese complimentary shrine there. The Cashmerian envoy is here described as a 'Nestorian bonze,' and in 731 it is said that the king of Central India had also sent a Nestorian bonze to China. Now, although the Chinese often seem to confuse Manicheans, Mazdéans, and Nestorians^[25] with 'heretic' Buddhists, still, in view of the fierce religious contests under the Sassanides of Persia, this singular coincidence must not be overlooked, especially when we remember the Chinese contemporary^[26] mention of Têh-sih (? Tersa, or Christians) in Samarcand, and of Nestorians in China, not to speak of the Tsâukûta religious changes already noticed. A few new things are said of Cashmere. The capital, Póh-lo-wuh-lo Pu-lo (Pravarapura), was on the east bank of the great river Mi-no-sih-to (Mr. Stein's Vitastâ); the people wore woollen frieze; there was more snow than wind; and a sort of burning glass or crystal^[27] was exported. Thus Ki-pin and Cashmere existed separately in the eighth century, the latter under Hindoo, the former under Turkish and Persian influences. The word Kapiça appears in no Chinese history, at any date, under any form, except, as already stated,^[28] in the year 507. In the seventh century several Chinese pilgrims travelled from Balkh and Tokhara to Kap19a. One of them, named Tao-lin, went the reverse way from Cashmere to Udyâna and Kapiça. None of them speak of Ki-pin, which had thus been split up into at least two large states, Cashmere and Kapiça. The name Cashmere had to be officially recognised, because the king was able to interfere in Chinese politics, whereas Kapiça was merely a haunt of peripatetic Chinese bonzes. The Chinese principle has been never to recognise new state names until the de facto ruler sues for fresh recognition.

From 900 to 960 China herself once more broke up, and, apart from minor Tartar and Chinese local rulers, a succession of ephemeral dynasties with a strong Turkish strain in their blood held sway over the imperial or central part of China. In 937 the Magadha-S'râvastî state (Bahar) sent a bonze; and in 940 the state of Kasyamit'o (probably intended for Kasyamilo or Kas'mira) sent another by sea with a Buddha's tooth; both these priests had the well-known Hindoo prefix S'ri, or 'lucky,' attached to their names.^[29] In 960 the highly literary native dynasty of Sung for the third time reunited most of China; but the first emperor was conservative in his political ideas, and would have nothing to do with the foreign regions of the south-west (Yün Nan, Magadha, &c.) In 966, however, Chinese bonzes were sent to collect books in India, and they passed through Pu-lu-sha (Peshawur) and Kia-shih-mi-lo (Cashmere).

For three centuries nothing is heard of Cashmere; certain Hindoo bonzes described and the historians record Udyâna as belonging to North India, and they add that Gandhâra lay twelve days' journey further west, beyond which again lay Nâgarahâra, Lampâ, &c. (in the Cabul region). Fâ-hien (Record of Buddhistic Kingdoms) had many centuries before this travelled westwards from Peshawur to Nâgara (in the Jelalabad region). Hüan-tsang (630–5) had found Lampâ (or Lamghân) under the suzerainty of Kapiça. In 1221–2 Genghis Khan and his generals took Ghazni, Jelalabad, and Beruan (north of Cabul). It is not, however, until the time of Mangu Khan, when the Mongols were in process of assimilating both China and the west, that Cashmere is again specifically mentioned: in 1253 an expedition was sent against K'ê-shih-mi-r and Hên-tu-sz-tan. But a distinguished Cashmerian (K'ia-sya-mi-r) family, of the family name K'a-nai, had already heard of Genghis Khan's exploits (before 1227); two brothers of this family were well received by Ogdai Khan (1229–1246); Gayuk Khan (1246–1251) gave one of them a local charge, and Mangu made him military governor of Cashmere. The other, whose 'name' was Nama (i.e. probably Master of Namah, or 'instructor '), was given authority by Mangu over all the Buddhists and Taoists in the empire, a post subsequently held under Kublai by the Tibetan Pagspa. Meanwhile the native ruler of Cashmere stirred up the populace to murder the military governor, this 'traitorous' friend of Mongol domination; hence the chastisement of 1253, the result of which is not stated. The son of the murdered governor served the Mongols at Peking until his death in 1309. One of the posts he held was that of night watch under the 'minister Puh-lo,' which name some suppose (I think erroneously) to mean Marco Polo.^[30] Marco Polo nevertheless describes Pascia (Peshawur), and Kesimur (Cashmere), seven days to the south-east of it; and the said Cashmerian did really serve under the Saracen Achmat, who is mentioned by Marco Polo. Marco certainly never visited either place, and his description seems to have been taken in part from Chinese tradition or history.

They have passes so strong that they have little dread of an invader. . . . In this country are hermits, who observe great abstinence . . . they have abbeys and monasteries . . . from this place you may go to the sea of India, and if we went further we should enter into that country;

and again—

There are wise . . . men or sorcerers called Tebet and Kesimur, which are the names of two idolatrous nations.

Plano Carpini, nearly a generation before Polo, enumerated 'Casmir' among the Mongol conquests, probably alluding to the attempts of the two Cashmerian brothers to bring that country under Mongol influence. In fact in 1257 the 'sultan' Huli of the K'in-shih-mi tribe submitted to the Chinese general Kwoh K'an,^[31] employed by the Mongols when the young Prince Hulagu, Mangu's brother, was conquering the highlands of Persia. In 1259 Mangu despatched a Chinese envoy named Ch'ang Teh^[32] to his brother Hulagu, and this envoy also mentions K'ih-shih-mi as being north-west of India. He says—

They have kept here all these ages the clothes and bowl of Sakya. The monks have a patriarchal look, like the Chinese pictures of the saint Dharma. They live on vegetable diet and spend their whole time in religious contemplation.

On the ancient Mongol map given to the world by the late Dr. Bretschneider^[33] Cashmere is placed outside the empire, whilst K'o-pu-li (Cabul), K'o-tsi-ning (Ghazni), and Badakshan are included in the middle empire of Jagatai, instead of in Hulagu's empire.

The Ming dynasty of native Chinese, which ejected the Mongols in 1368, had extensive relations with Tamerlane's empire, and also with India by sea. In describing Tamerlane's empire of Samarcand, the Ming historians say it is 'the Ki-pin of Han (B.C. 200–A.D. 200) and the Ts'ao of Sui (A.D. 581–618).' But, as Tamerlane takes his name of 'Timur-lenk,' or 'the Lame,' from his wound at the conquest of Candahar in 1398, and as his son's empire at Herat is plainly described too, it is evident what the Chinese meant. Moreover Timur's grandson Pir Mohammed included Cabul, Ghazni, and Candahar in his provinces, if we are to accept Deguignes's statements, repeated by Bretschneider. K'êh-shih-mi-r and Kêh-shih-mi are mentioned casually amongst the tribute-senders to the Ming emperors, but certainly there is no further Chinese information of a trustworthy nature.

I take Ki-pin to be the Cophes or Cophene of Arrian, Strabo, Pliny, &c. The character Ki is used by the Chinese travelling monks for spelling the words Kanishka and Kanudj, which fact finally decides its potential capacity as the short vague Hindoo kă. As to the second character, it only occurs as a final in the name of one other country; that is the combination Sz-pin, which I identify with Mommsen's Sophene, on the Upper Euphrates—exactly the same syllables in 'power.' So far as I know Ctesiphon^[34] lower down the Euphrates is the only rival to Sophene hitherto proposed; but (apart from the unaccounted for 'Kte'), si-phon and so-phen have manifestly no more than equal primâ facie claims to be sz-pin. Again, there was no other considerable country south of the Hindoo Kush besides Ki-pin known to the Chinese of 2,000 years ago; and it is plain that the situation they assign throughout the course of ages to this land in relation to India, Gandhâra, Udyâna, Balti, &c., corresponds perfectly to the traditional site of the Greek and Roman Cophes and Cophene of Alexander's generals, whose personal observation was after all the main evidence utilised by Arrian, Pliny, &c. If we remember that 'Cabul' is even now pronounced 'Cawbl,' with the accent on Caw and that the Chinese syllables used have (by the living tests indicated above) the value Kaw-bu—with the accent invariably on the first syllable—we need have no hesitation in accepting the Kao-fu^[35] conquered from Ki-pin by the Indo-Scythians as Cabul (the Cabura of Ptolemy) conquered from the Sakya princes of Cophene by Kanishka. As to U-yih-shan-li (still possessing the living 'power' of O-ik-san-li), this must be the Alexandreia of Ptolemy—which Pliny and Stephanus seem to have called Cophen too—at or near Candahar.

E. H. Parker.

1. It is very like the Russian use of the Russian word nyemets ('dumb') to signify the German race, first heard of by them under the name Nyemci in 987. See Asiatic Quarterly Review, April 1904, pp. 353-4.
2. Ante, vol. xi. (1896), 429 et seq.
3. Asiat Quart. Rev. July 1902, p. 131 et seqq.
4. The Chinese commentators of the seventh century of our era consider these Săk of B.C. 100 to be the same as the S'ákya or Shakya princes of Buddhistic India (for instance, Yen Shï-ku, Mayer's Chinese Manual, p. 275). Western writers suggest rather that the Saca princes were of Turanian stock.
5. Journal of the Royal Geogr. Soc. December 1902.
6. Chapters 96a and 96b.
7. Hien-tu, or 'Hanging Passages,' is usually considered to mean, by the usual play of phonetic words, Hindu [Kush].
8. These mountains are mentioned 800 years later. See below, p. 631.
9. It is admitted by nearly all inquirers that Cabul or Cabura Bactriana is meant.
10. See 'Chinese Buddhism,' in the Asiat. Quart. Rev. October 1902.
11. It is interesting to know that Kao-fu now bore a new Indian name, 'Yem-fu Ngai,' i.e. Djambu? Ge.
12. Dr. Legge (Fah Hien's Record of Buddhistic Kingdoms, ch. 4–8, Oxford, 1886) gives us an account of this strongly Buddhistic country forty years before this event. He also tells us how his Chinese pilgrim, leaving Khoten, directed his steps towards Ki-pin viâ Khargalik, Tashkurgan, and Iskardo (?), thence via Udyâna and Su-ho-to (Swât?) to Gandhâra.
13. See chapters 102, 97.
14. In like manner the Kushans were still called by the old name, Yüeh-chï; and the newer Kushans or Ephthalites were also alternatively called Eptal and Yüeh-chĩ until the Turks broke them up.
15. Chavannes's Voyage de Song Yun (Hanoï, 1903). Chavannes's Turcs Occidentaux (Imperial Russian Academy, 1903) contains many critical observations on the comparative Chinese statements about most of the important pilgrims.
16. The Book of Ser Marco Polo, i. 162 (2nd ed. 1875).
17. See Eited's Dict. of Buddhism (Hong Kong, 1870) and Chavannes's Tures Occid.
18. See the former, ch. 198, and the latter, ch. 221 a and b.
19. Légende de Gaudama, 1878.
20. See 'The Western Branch of the Turks,' Asiat Quart. Rev. October 1903.
21. There is an official statement that in 765 the king of Ki-pin was made also king of Baltistan, and that he was in 779 defeated by the Tibetans; but this apparently important addition is a misplaced sentence, which has crept by error into one version only: the king in question (Surendrâditya) was really patented king of Balti in 720, and was defeated by Tibet in 734. He was scarcely likely to have been repatented to Balti in 765 in addition to his supposed title as king of Ki-pin.
22. M. Chavannes identifies this with 'Arokhadj,' or Arachosia, which, if correct, clinches the conclusion that Ts'ao, Tsâukâta, Alexandreia, Ghazni, &c., were parts of Cophene or Ki-pin.
23. Mr. Stein and M. Chavannes have done much to elucidate the ungainly Chinese equivalents, which are all etymologically correct. Mr. Stein says Lake Volur is the old Dragon Pool.
24. I.e. Ghazni, originally called Tsâukûta, later Arokhadj, and finally by an unidentified Chinese name, Sie-yih, or Seyit.
25. 'The Nestorian Tablet,' Dublin Review, Oct. 1902; China and Religion, 1905.
26. 'The Early Christian Road to China,' Asiat. Quart. Rev. October 1903.
27. This is also mentioned as one of the products of India, Borneo, and Cochin China.
28. See above, p. 629.
29. For the period 907–60 we refer to the old Wu-tai Shï (Tsin dynasty), published c. 975, ch. 76, p. 6; and for those 960–1126 and 1127–1279 to the Sung Shï, published c. 1350.
30. 'New Facts about Marco Polo,' Asiat. Quart. Rev. January 1904.
31. Chinese Recorder, 1875; Yüan Shï, ch. 125, 149, and passim (published c. 1380).
32. See 'C'hang Teh's Mission to the West,' Chinese Recorder, 1874; to which may be added the Ming Shï, covering the period 1368–1643, ch. 332 (published 1724).
33. North China Branch, Roy. Asiat. Soc. Journal, vol x. Shanghai, 1876.
34. F. Hirth, China and the Roman Orient.
35. A syllable fu (having the same 'power') is used by the Chinese [ante, p. 629, n. 11) to express the Hindoo word Djambu, not to mention the common word Fu-t'u or Budh.