^{page 285 note 1} Siang shan or Tsung shan are two names identical in origin and vary only dialectically. The original root was Dar, which probably survives to the present day in Dardistan, etc.

^{page 286 note 1} Djou from root Dyo; Kung=Çur, Gr.

^{page 286 note 2} Erroneously translated by DrLegge, , Chinese Classics, vol. iv., as the Kwan hordes.Google Scholar

^{page 288 note 1} Mencius, book i. 2. xiv., says, “Formerly when King T'ae dwelt in Pin (Ban), the barbarians of the north were continually making encroachments upon it. He therefore left it, went to the foot of Mount K'e, and there took up his residence. He did not take that situation as having selected it. It was a matter of necessity with him.”—Legge's Chinese Classics, vol. ii. p. 50Google Scholar; cf. also pp. 31 and 52–3, and King, Shi (Book of Poetry), iii. 1. iii.Google Scholar; Legge's Classics, vol. iv. p. 437.Google Scholar

^{page 289 note 2} Shi-Ki, chap. 110. Cf. also T'sien Han Shoo, translated by MrWylie, A., Journal of Anthropological Institute, vol. ii. no. iii.Google Scholar

^{page 289 note 3} J, the weakest of semi-vowels, is, in Chinese, constantly interchanged with n. Ng final was anciently, in most cases, pronounced r.

^{page 290 note 1} King, Shi (Book of Poetry), ii. 1. vii. kc.Google ScholarLegge's Chinese Classics, vol. iv. p. 259. Dr. Legge, in a note to the above work, gives the Chinese authorities for the connexion of the Himwans with the tribes subsequently known as the Duh-Kiueh.Google Scholar

^{page 291 note 1} When engaged last year (early part of 1874) in preparing some notes on the probably recent elevation of Central Asia, I had occasion to insert some remarks with regard to the tribes handed down in Chinese history as the (in modern Chinese Yuèti). None of the identifications usually accepted for those tribes seemed to tally with the circumstances of the case, and finding myself without external evidence, I suggested from philological motives alone that the word Vidal would be found an approximation to the actual name of the tribes. I was unaware at the time that M. Vivien de St.-Martin had, in a review of the destruction of the Greco-Bactrian kingdom, from historic grounds identified the so-called Yuèti with the White Huns, the Haiáthalah, the Ephthalitæ of the Greek writers, who, about the year 134 B.C., overran that country, and finally destroyed the Grecian dynasty which from the time of Alexander's death had been paramount. Vidhal rather than Vidal should have been the rendering of the old Chinese name, and as this may be considered as identical with the Greek name for the tribe,—the Ephthalitæ, as well as its Arabic form Haiàthalah, it may be interesting to review the grounds on which this reconstruction of the ancient sounds of the characters was founded. The characters made use of to express the name of the tribes have in their present form no meaning; they were at first only used as phonetic elements. To arrive at their archaic sounds it will be necessary, however, to analyse their separate meanings; and endeavour if possible to find their analogues in some language making use of an alphabet. Such we find in Sanskrit,—a language having many relations with primitive Chinese. Taking then the Chinese characters separately, we find the first Yuè ‘the moon,” in Cantonese, the nearest existing dialect to that spoken at the beginning of Chinese history, Üt or Yüet; the second tî, in Cantonese tai, meaning ‘fundamental,’ ‘radical,’ ‘reaching to the ground.’ Comparing these words with Sanskrit, we find the first represented by a word vidhu, with the same meaning. Bopp gives no derivation, nor does he mention any analogues among Aryan languages. It may therefore be fairly accepted as a non-Aryan term for the moon which found acceptance in Sanskrit. The form tai for the second syllable seems to point to a lost liquid termination, l or r, more probably the former. Tai will thus agree in its original sound with the Sanskrit tal ‘condere.’ This root is one of extensive distribution; in Sanskrit it gives tala ‘solum,’ ‘fundus’; in Latin tellus; in Chinese ti ‘the earth,’ tai ‘fundamental,’ etc. From analogy therefore the combination of the two characters would have been pronounced Vidh-tal, Vidal, or Vidhal; thus corresponding with the actual name of the tribes as handed down from other sounds.—See Trans. North China Branch of the Royal Asiatic Society, new series, no. x. Shanghai, 1876.Google Scholar

^{page 292 note 1} Journal of the Anthropological Institute of Great Britain and Ireland, vol. ii. no. iii.Google Scholar

^{page 292 note 2} Journal of Anthropological Institute, vol. ii. no. iii.Google Scholar

^{page 295 note 1} Da-wan, as will appear below, seems to represent the districts adjoining Kashgar and Yarkand.

^{page 295 note 2} Gang-gu apparently the country about Karakul, possibly the Riangkul Pamîr.

^{page 295 note 3} Ta-hia, Bactria, lately overrun by the Tochâri, and subsequently known as Tochâristan.

^{page 296 note 1} I have been compelled, on a careful consideration of the map of Central Asia, to place most of the localities named in Djang-kien's report considerably eastward of their usually accepted positions. Da-wan has usually been identified with Ferghana, but its position between Gang-gu, Da-hia, the Viddhals, and bounded eastwardly by Khoten and Hanmow, compels me to fix it on the site of the modern Kashgar and Yarkand. I am unable to offer any satisfactory clue to the name.

^{page 296 note 2} Hanmow or Hanmi, a state not marked on the maps. It must have lain N.E. of Khoten.

^{page 296 note 3} Yu-tien, the present Khoten. A. Remusat was the first to point out the Aryan origin of the name as Koustana, ‘the breast of the world;’ more correctly, perhaps, Kulstana, ‘the apex,’ from the root kul, in collis, Greek , Chinese etc., and stana, ‘mamma,’ as Kulagiri from Kul and giri ‘mountain.’

^{page 296 note 4} Ho, the Yellow River.

^{page 296 note 5} Lou-lân, apparently the Charchan of Marco Polo; l representing the d or ch. It does not appear on the maps. Vide Yule's Marco Polo, 1st ed. p. 178. In the later Chinese works it appears as Shen shen. Yuen-chwang (Hiouen Thsang), after leaving Khoten, travelled east 300 li to Pimo, thence 200 li to Nijang, then 400 li of desert to Tonholo; from whence 600 li took him to Chemotona in the kingdom of Tsiemo. Thence 1000 li N.E. lay Nafopo belonging to Lau-lan, from which he went on to Shachow. Fah-hian likewise went from Shachow to Khoten by the same tract. Lau-lan and Shen shen, as well as Polo's Charchan, point to the original name as Dardana.Google Scholar

^{page 296 note 6} Kushi, called also Kiushi by the commentator, apparently stands for Aksu. This is so common in names that identification seems impossible.

^{page 297 note 1} Wû sun, the upper valley of the Naryn and possibly Hi.

^{page 297 note 2} Gang-gu would by analogy represent Kara kul, but this would pre-suppose the presence of Turkish names on the Pamir before the Christian era, of which I have found no other traces. Sarikol ‘the yellow lake,’ if this be the true interpretation of the name, would, however, seem to point to a similar antiquity, especially if, as suggested below, we are to identify the Sakarauli of Strabo with the Sarakoolies of to-day.

^{page 297 note 3} Im-tsai, Samarkand. Lower down the name is given more fully as Im-tsai ar-gan; the first two syllables are probably inverted, so that originally it appeared as Sal-m-ar-gand. The phonetic portion of gan, ‘ashield,’ representing Sanskrit kaṇḍ ‘servare.’

^{page 297 note 4} The Gwai-shui is the Oxus, the syllable corresponding with the Veh-rud of the Parsis or the local Wakh.

^{page 297 note 5} The Chinese commentary adds, that the Viddhals had pink and white complexions and were admirably skilled in the use of the bow. The Greek writers called them the White Huns, according to Procopius, from their light complexions.

^{page 298 note 1} Laou-shang, perhaps Oldar or Ildar.

^{page 298 note 2} Now Sha-chow in Kansuh.

^{page 298 note 3} The Dû-gwai-shui seems to be the Surkh-ab, though that name would more regularly have been represented by the syllables Tû-kwai.

^{page 298 note 4} An-sik or Ngan-sak. Ngan is the representation of Greek hence we may assume Arsak as the original pronunciation. The country is probably so called from its line of kings, the Arsacidæ.

^{page 298 note 5} Tiao-chi, Sarangia. Tiaou is the analogue of Greek chi represents the ngia.

^{page 299 note 1} Yok-shui. The Dead Water (Yok, Greek here spoken of is evidently the Hamun or lake of Seistan. The original Yok-shui of Chinese legend referred apparently to an ancient lake in Eastern Turkistan, of which lake Lob is now the representative.

^{page 299 note 2} Si Wang-mu. See Mayers's Chinese Reader's Manual, s.v.

^{page 299 note 3} Da-hia, the country of the Tochâri. The second syllable, hia ‘summer,’ is the equivalent of Sanskrit gharma, with similar meaning, from root ghar, in Greek Xap.

^{page 299 note 4} Lam-shi-ching, Darapsa of Strabo. The phonetic of Lam points to a root lamb. Darampsa was probably the original name.

^{page 300 note 1} Essay on the Geography of the Valley of the Oxus, prefixed to the new edition of Wood's “Journey to the Source of the Oxus.”