EMPIRE.] PERSIA 593 lun. and made a drinking-cup of his skull, 1 and the great mass of the vanquished people (the great Yue-chi) left their homes and moved westward, and occupied the land on Lake Issyk-kul, driving before them another nomad race, the Sse. The Sse took the road by Utch and Kash- gar, ultimately reaching and subduing the kingdom of Kipin (the Cabul valley), while their old seats were occu pied by the Great Yue-chi, till they in turn were soon attacked by the U.sun, who lived west of the Hiung-nu, and forced to move farther west (160 or 159). The older Chinese account ignores the residence of the Yue-chi at Lake Issyk-kul, which can at most have lasted only for a few years : the later account goes on to say that, moving westward, they conquered the Ta-hia, i.e., the Bactrians. The language of the older narrative has been held to imply that they went by way of Ferghana and remained there for some time ; but in reality it only says that they retired beyond Ferghana and conquered the Ta-hia, thereupon pitching the royal camp north of the Oxus, and so it ap pears that in 159 they moved straight on Sogdiana, reach ing that land just at the time when internal wars were undermining the might of Eucratides. The conquest, how ever, may have been gradual, since Bactria is still named as independent in 140. "When the Yue-chi were already settled in their new homes the king of China sent a certain Chang-kien to urge them to return and help him to clear the caravan-road by thrusting back the Hiung-nu. He was arrested on his way by the latter, but escaped in 129 to Ferghana, and thence was led to the Yue-chi through the land of the Khang-kiu, on the middle course of the Jaxartes. But the Yue-chi were too happily settled in a rich and peaceful land to listen to his representations, and after a year s residence (128-127) he returned to China, which he reached in 126, after falling again into the hands of the Hiung-nu on the way. From him are derived almost all the accounts of the country and its inhabitants given by the Chinese historians. There were, we arc told, settled and agricultural peoples in Great Wan (Ferghana), Ta-hia (Bactria), and An-si (Parthia). All the races from Ferghana to Parthia had deep -set eyes and strong beard and moustache ; their dialects varied, but as they all understood each other all must have been Iranian in speech. Their manners, too, were much alike ; they paid great respect to women, and the men were very complaisant to their wives. This is almost exactly what Bardesanes says of the position of women in his time among the Kushan in Bactria ; 2 but it was quite otherwise in Parthia, where the Oriental seclusion of women was carried to the extreme (Just., xli. 3, 1, 2). They were all knowing traders, and understood the preparation of silk and lac, but not metallurgy till they were taught that art by Chinese agents and deserters. They then imported the precious metals from China and made gold and silver vessels, but not money, being in this respect behind the Parthians. 3 Great AVan probably corre sponds to the Ovapvoi of Ptolemy (though he misplaces them) and the Varena of the Vcndldad ; it was a separate kingdom, with a popula tion estimated at 300,000 souls in the 1st century B.C., and seventy subject cities. The king, probably a native who had risen on the fall of the Greeks, lived in Kuei-shan (probably Khojend, at the mouth of the Ferghana valley), and could call out an army of 60,000 men, lancers, archers, and mounted bowmen. The land was famous for its wine and for horses of divine race which sweated blood, and for the possession of which China went to war with Great Wan in 104-103, and again in 102-98. Lucerne and grapes were exported to China ; the name of the latter, "po-tao, " is held to be the Greek fiorpvs, which would show that the vine was introduced by the Greeks of Alexandria Eschata. South of the Wei or Oxus lies Ta- hia (probably Zend Dahviju, the land *). Here there was no king, but the several cities were the seats of chiefs, a state of things such as Alexander had found in the country and as reappeared under the Turks in the 7th century A.D. Chang-kien estimated the population at a million ; the} were bad and cowardly soldiers, but excelled in trade, and the chief town, Lan-shi, had rich bazaars of many wares. This town must be one of the commercial cities on the river Bactrus, along which lay the trade-route from India to the north (Pliny, vi. 52), i.e., either Baetra or Encratidia (which, according to Ptolemy, vi. 11, 8 [Codd. B., E., Pal. 1], lay lower down the stream on the left bank). In the latter case Lan- shi may stand for "EXAT^es. North of Ta-hia lay the Great Yue-chi, 1 The Lombards had the same custom, learned, no doubt, in the childhood of the race from their Avarian neighbours. 2 See Langlois, Coll. d. hist, dc VAvmenic, i. 84. 3 Ssematsien, in Ritter, vii. 3, p. 642. 4 Certainly not Dalia?, for they were never in Bactria. and west of the latter was An-si towards the Oxus. This was a 160-44. very great country, whose length might be 1000 li (358 miles), and it had 100 cities great and small. The first caravan from China to An-si passed on its way from the east frontier to the capital (called in the 1st century B.C. Fari-teu, i.e., probably Parthau), a dozen walled cities, which lay almost close together, so dense was then the population of the fertile part of Khorasan. The merchants of An-si visited the neighbouring lands with waggons or with ships for distances of Several thousand li. The coinage was silver, with the image of the king, and was called in and restamped on a new accession. 5 Writing was on skins in horizontal lines. Now, though the money as here described fits Parthia, the mercantile character of the race does not at all correspond to that of the Parthian aristocracy. Both here and in the general description given above, which also contains features not applicable to the Parthians, we see that the Chinese did not distinguish the ruling race from their subjects, and mainly described the latter, who were in point of fact very similar to the people of Bactria and Ferghana. As An-si extends to the Oxus the description is taken from the inhabitants of Margiana, a country which -must have been then subject to Parthia. A later Chinese account, referring to the period 24-220 A.D. , places on the east frontier the city Mo-lu or Little An-si, which is plainly the Mourn of the Vendidad, modern Merv- i-rud, and the Greek Antioch 17 Zvvdpos ; An-si is a corruption^ the last name, just as the Persians call the Syrian Antioch Andiv, and so came to be a name for the Parthian rulers of the city. West of An-si, on the western (Caspian) sea, layTiao-chi (Media), an agri cultural country with a dense population, a dependency of An-si, and in part governed by tributary chiefs. Chang-kien is thinking less of the central parts of Media than of Gilan and Mazandaran, for he speaks of the warm moist climate where rice is produced. And in this quarter there were really various petty states ; not only Atropatene but Dilem had its own king, as appears for the year 65 B.C. from Plutarch, Pomp., 36 (where for Evfj.aiwv read AeXu/icu aw), and the Gelfe and Cadusians doubtless stood under their own mountain chiefs as they had done under the later Achfemenians, and did again under the first Sasanians. It is a proof of the solid power of the empire founded by Mithradates that Parthia was able to assert some kind of supremacy over these hardly accessible districts. North of An-si lay Li-kan (Hyrcania), whose wizards, with those of Tiao-chi, had great reputation. It is clear from this whole account that the centre of the empire was still in the old Parthian lands, and that the lower satrapies were viewed as mere dependencies, " outer lands." In the following century the Chinese obtained knowledge of the west by the caravan-route which passed through Kipin (the Cabul valley) to U-ghe-shan-li (Arachosia) ; and now we find a changed state of affairs ; these two countries are bounded on the west by Tiao-chi, whose powerful king has his capital a hundred days journey from the frontier. An-si is now only mentioned incidentally as reached from Arachosia by going first north and then east, which is correct if we take the name in its original sense of the subjects of Parthia in Margiana and its capital Antioch. But the empire of Parthia, which now had its centre in Media and the western lands, is certainly Tiao-chi, a word that is probably connected with the word for "land" in the official lan guage of the Achremenians, old Persian dahydus. As nomadic peoples Chang-kien names the Great Yue-chi in Sogdiana, the Khang-kiu on the middle course of the Jaxartes, and the Yen-tsai in Chorasmia. The Yue-chi could put from 100,000 to 200,000 bowmen in the field ; later they were reckoned at 100,000 warriors and their families. The royal camp had been north of the Oxus even after the conquest of Bactria, but they finally withdrew entirely to this district. Their capital is called Lan-shi; and the name of Ta-hia disappeared before that of " Land of the Great Yue-chi." At the conquest they had a single king ; afterwards they formed five principalities. The fifth of these corresponds to Cabul, so that the division is younger than the Scythian invasion of Asia after the death of Phraates II. Imme diately north of Ferghana, but separated from the Yue-chi in the south and the Hiung-nu in the east by a series of small kingdoms, were the pasture-grounds of the Khang-kiu on both sides of the Jaxartes ; their force was 80,000 to 90,000 bowmen. North-west of these were the Yen-tsai on the Aral, the northern neighbours of the An -si, and east of Hyrcania, that is, in Chorasmia. If there is no error in the writing of the number they mustered but 10,000 warriors ; then again considerable changes had taken place when the Chinese made war on the Khang-kiu in 44 B.C. The small kingdoms south and east of the latter have disappeared, so that the Khang-kiu border on the Hiung-nu and the great Yue-chi ; but the latter have now moved south, and now, too, the Khang-kiu are the northern neighbours of An-si, and not the Yen-tsai ; the latter are their dependants, and a tribute of mouse-skins is even drawn from the kingdom of Yen beyond the Yen-tsai. Such_a tribute cannot have come from any "place south of the Mukhajar moun- 5 On this point the younger Chinese account falls into a confusion with the coins of the kings of Kipin. XVIII. 75
