Sec. iii] IN THE VALLEYS OF SHUGHNAN 883
dara joins the open valley of Tokuz-bulak coming from the Koi-tezek pass, the Russian cart-road was reached. This brought us, a couple of miles farther on, to the smiling meadow land at Warkhéts in the main Ghund valley, where we halted (elevation about io,000 feet).
On September loth a delightful day's march down the Ghund river allowed me to gain some impressions of the middle portion of the great valley of which I had sighted the head, just a month before, from above the outflow of the Yeshil-kul. It showed very clearly the advantages it must have offered for traffic across the Pamirs towards Badakhshan, whenever conditions prevailing in Shughnan allowed it to be used in safety. That such could not always be depended upon in modern times was indicated by the forts of the Shughnan Mirs that we passed at the hamlets of Sardim and Wang, and farther down below Charsim. Before reaching the last-named village (Fig. 416) among verdant fields and arbours, we crossed a formidable barricade of rock debris extending for about a mile and stretching right across the broad bottom of the valley. It had been thrown down at some period by a landslide from the frowning cliffs to the south. Numerous stone breastworks, both on the eastern and western edges of this barricade, proved the defensive use made of it at different times. Several smooth-faced rock fragments bear shallow sgraffiti in Arabic writing containing pious invocations or formulas, besides the usual rude representations of an open hand and of wild sheep. In the local Ak-sakal's house at Charsim I was interested to note a living hall with the ceiling and skylight arranged in the same ancient style as observed at Miragram in Mastûj and fully described below in the case of a Kala-i-Wamar dwelling.18 From Wer, another pleasant village, some six miles below Charsim, we crossed by a very rickety bridge to the right bank of the river and there made our way to Shitam (about 9,000 feet), a village of sixteen households, situated at the mouth of the very steep valley of the same name. By the glacier pass at the head of this we were to cross the high watershed range towards Rôshan.
Necessary preparations here caused delay in the start of the morrow, and I utilized it to listen to such scanty recollections as could be gathered from the greybeards produced as depositaries of Ghund local tradition. This proved distinctly less enduring than in Shakh-dara, not reaching back beyond the time of Shah Abdurrahim. Apart from this last but one of the Mfrs of Shughnan and his son Yûsûf `Ali, there was vague remembrance only of a Mir Salim and a Shah Kirghiz. That at one time Chinese control had extended to Shughnan was, however, known to all my informants. In support of this was quoted a short Persian inscription, said to be engraved at the village of Deh-baste farther down the valley. According to the verbal rendering, which was all I could obtain, it refers in metrical form to the boundary there fixed between Ghund and Sûchan (near Khàruk) ` by order of the H agim of the Khagâ.n-i-Chin'. All my informants agreed in stating that during the rule of the last Shughnan Mirs and the subsequent Afghan occupation the population of Ghund had greatly diminished owing to severe exactions. Emigration to K6kand, Margilan, &c., became particularly prevalent owing to the Mirs' practice of selling women and children as slaves to increase their revenue. In consequence places like Wang, Wer, and Charsim were practically deserted until better conditions were secured through the Russian occupation. That Ghund was still under-populated was proved by the fact that most of the men I measured at Shitam were labourers who had come from the R8shan side.