Sec. iii] IN THE VALLEYS OF SHUGHNAN 879
position. The use of the designation of the ` Five Shih-nis ', with which we meet also in Wu-k`ung's narrative, is accounted for by the division of the territory into five valleys with distinct chiefships, and accords with a traditional notion to be presently mentioned. The reference to the warlike and marauding propensities of the people is perfectly borne out by the reputation the Shughnis have enjoyed down to very recent times. Their raids are still a subject of lively recollection among the people of Wakhan, and there can be little doubt that the present occupation of Sarikol by a population speaking a language differing but very slightly from Shughni is the result of conquest from the side of Shughnan.4 Both raids and outside settlement must largely be attributed to the very limited extent of arable land in the narrow valleys of Shughnan and the absence of adequate grazing grounds. To this cause is due the migratory tendency and spirit of enterprise that the Shughnis undoubtedly display at the present time. Driven forth by the poverty of their homeland, Shughnis proceed annually in numbers to Farghana for temporary work as farm labourers, while plenty of others seek employment as servants both at Kabul and at large centres like Margilan and Kakand in the north. Coming from Wakhan with its rather ` tame ' submissive population, I was particularly struck by the independent and versatile ways of the average Shughni.
The division of the territory into five separate autonomous chiefships, which the Tang Annals Division
specially mention, is a direct result of its geographical configuration. Instead of being confined into
to a single main valley like that of the Ab-i-Panja in Wakhan, the population of Shughnan lives chiefships. in a series of different tracts, each having a distinct character of its own and separated from the rest by high mountains or the equally effective barriers of difficult river defiles. A look at the map shows clearly enough four such natural divisions : the valleys of Ghund and Shakh-dara, the valley of the Panj or Oxus from below Gharan to the border of Rôshan above Kala-i-Wamar, and Rôshan itself. The last appears always to have been closely linked politically with Shughnan proper, and the language spoken there is a dialect but slightly differing from Shughni. I f we assume that the land on either side of the Oxus was counted as a separate tract, which would be natural enough in view of the difficulties presented by the river crossing, we arrive at the five distinct chiefships. Else possibly Gharan may have been included in the reckoning recorded by the Annals.
From inquiries which subsequently were greatly furthered by Tûran Beg, a very intelligent Traditional
nonagenarian of Shakh-dara and a fountain-head of local information, I ascertained that the reckoning of
use of a closely corresponding traditional designation of Shughnan as Haft sadhd-i-Shughnan was still current. Local opinion was not quite agreed as to the tracts exactly counted among these seven ` Sads ', but generally favoured the inclusion of Darmarak, Kala-i-Bar-Panja, Parsheniw,5 Khàruk,6 Shakh-dara, Ghund, Rôshan. There was, however, consensus as to the fact that until the early part of the last century Ghund, Shakh-dara, and Rôshan were ruled by separate Mirs acknowledging but a nominal subordination to the Mirs of Shughnân, whose residence was at Kala-i-Bar-Panja.7 A somewhat similar state of things is likely to have prevailed in the seventh—eighth centuries A. D., from which the record in the Tang Annals dates.
4 I was unable to trace definite traditional knowledge as to the date of this settlement of Shughnis in Sarikol. It is supposed to have taken place ` a very long time ago '. Yet, according to what I learned in Shakh-dara, relations of kinship are still maintained between certain families living there and others settled in Sarikol.
5 Parshenïw is the chief village of the fertile tract along the right bank of the Oxus below Kharuk.
6 This is the Persian form of the name given to the valley below the confluence of the streams of Ghund and Shàkh
dara. The Shughni pronunciation of the name sounded Kharagh. The officially adopted Russian spelling is Khorok.
The village of Khâruk, unimportant before it became the seat of the Russian military and ` political ' authority, was at the time of my visit said to include 6o homesteads. It also boasted of several shops kept by Shughnis and a couple of Bajauri traders.
7 According to Tûràn Beg, tradition remembers five
Mks of Shughnan ', succeeding to each other in direct descent : Shah Wanji, Shah Amir Beg, K6bâd Khan, Abdur-