THE long-stretched valley of Wakhan to which I descended on May 19, 1906, across the Bari5ghil saddle, had appeared to me from the first one of the most interesting stages of my journey. It was not merely that I touched here the easternmost marches of true Iran and the headwaters of the River Oxus, which ever since my youth I had longed to follow down to regions of fascinating historical interest : I knew also that I stood here on what from earliest times must have been a main route linking Western Asia, and through it the classical world, with innermost Central Asia and thus the Far East.
Nature itself, as it were, seems to have intended Wakhan to serve as the most direct thoroughfare from the fertile regions of Badakhshan to the line of oases along the southern rim of the Tarim Basin ; for along the whole of the Ab-i-Panja Valley from Ishkashim where the Badakhshan route joins in, right up to Sarhad, a distance of close on a hundred and twenty miles, travel is facilitated by the remarkably easy nature of the ground and the presence of cultivation. Beyond Sarhad, it is true, the Ab-i-Panja is confined to a narrow gorge which provides two trying marches. But further on the road lies open past the Little Pamir to the Wakhjir Pass which for a considerable part of the year gives easy access to Sarikol, the westernmost inhabited valley on the headwaters of the Yarkand River. The importance of Wakhan as a thoroughfare from west to east is still further increased by the fact that the two difficult marches above Sarhad can be avoided by the somewhat longer alternative route which ascends the northern branch of the Ab-i-Panja to the Great Pamir and thence crossing the Little Pamir reaches Sarikol by one or another of the passes, all lower than the Wakhjir.
To the position of Wakhan on the most direct routes linking the Oxus and Tarim is owed the relative abundance of early notices dealing with it which can be gathered from the records of Chinese annalists and travellers. The oldest of these is probably furnished by the Annals of the Former
Han Dynasty. These mention Hsiu-mi - as the first of the five territories ruled by Jabgus
(Slain-ku) which belonged to the great Yüeh-chih nation after its conquest of the regions south of the Oxus. That Hsiu-mi is but an earlier transcription of the old name of Wakhan which appears as Hu-mi gf in the Tang Annals is rendered probable by a notice of the Pei-shill. This, while reproducing the statement about the ancient Yüeh-chih territory, distinctly indicates that it lay to the west of So-ch`ê or Yarkand.' According to the Pei-shill's record which dates from the early part of the 7th century, the territory then bore the name of Ch`ieh pei fhp 1. The name of its capital
is still given as Ho-mo %p , the same as in the Han Annals.
The first actual description of Wakhan is given in the accounts of Sung Yün and his fellow pilgrim Hui-shêng who in A. D. 519 passed down the valley on their way from Sarikol to Udyana.2
' Cf. Chavannes, Pays d'occident, p. 44 note, and Marquart, Érànâahr, p. 225, where the early names of Wakhan have been lucidly discussed. The identification of Hsiü-mi with Hu-mi was first suggested by Cunningham, J.A.S.B., xiv.
p. 433. Cf. also Yule, J.R.A.S., N.S. vi. pp. I I s sqq.
2 Cf. Chavannes, Voyage de Song Yun, p. 23 ; also Marquart, Lrrän§ahr, pp. 223 sq.