now known in these valleys. The impression gained from these scanty remains and from local traditions of occasional finds of bronze arrow-heads and strange fragments of weapons, distinctly suggested an early occupation. The site was invested with additional interest by the full view it commanded of the Nisar-gol plateau opposite, where in 1895 Colonel Kelly's successful action was fought and the relief of Chitral Fort decided.
After we had crossed the gloomy river gorge to the right bank, about one mile above Sanôghar, I was shown, at the eastern end of the barren Maidan of Parwak, a small ruin known as Darbatshâlinogh-r and believed to be of great antiquity. Tradition seemed to connect it with a ' Buzurg ' who was famed for his stinginess. The traceable remains consisted of an oblong of ruined walls of solid though rough construction, measuring about seventeen by twenty-one yards, with the exposed masonry rising eight to ten feet above the ground level. Some portion of the stone material had been utilized for the construction of ' Sangars ', probably when the Nisar-gol position was defended by the Chitralis and their Pathan allies. Outside, the lines of a quadrangular outer enclosure were just traceable in places. The whole, situated in a barren spot far removed from cultivation, suggested the remains of a monastery. But there was no time for closer investigation before the falling dusk obliged me to move on to Mastûj.
My single day's stay at one of the hamlets which make up the present Mastûj capital did not reveal anything of antiquarian interest, unless I may thus allude to that fine old chief, Bahadur Khan, who for his loyalty in 1 895 had secured independence as ruler of the Yarkhûn Valley.' Full of vigour in spite of his eighty years, and full of old-world courtesy in his ways, this ruler of some six thousand households seemed to embody in his person an historical past which is now rapidly fading away. The position occupied by this cluster of hamlets at the point where the Yarkhûn receives from the south its first main tributary, the river of Laspur, and where the routes from Wakhan, Chitral, and Gilgit meet, must always have been of importance for trade as well as for warfare and amply accounts for the simple square-towered castle which serves as the chief's usual residence (Fig. i 3). But at the same time a glance at the bare stony plateau over which these hamlets are scattered, sufficed to show me that there was here neither enough arable land nor adequate irrigation facilities to have supported any large settlement within historical times. This negative observation had its value when I came to trace the position of what was the chief place of the Upper Yarkhûn Valley at an early period.
On my march up the valley on May 13 I was able to examine in some detail the old fort of Brep of which I had heard already at Chitral from Waffadar Khan, Diwan-bégi, as dating back to the ` Kalmak ' or Chinese period. In the midst of the little hamlets collectively known as Brep, which are scattered over the alluvial fan formed by the Chikano-gol, some fourteen miles above Mastuj, there rises a conspicuous mound locally known as Noghôro-dôk, ` the fort mound'. Its artificial origin is clearly indicated by the slopes which, wherever scored by small ravines, show the same mixture of clayey earth and pebbles as found in the walls which crown the top. The summit rises to a height of about thirty-four feet above the ground level on the east, where the slope shows an angle of about 410. On the west, where the ground falls gently away towards the river, the elevation is somewhat greater. The circumvallation on the top forms an irregular oblong of which the east and west faces measure 180 and 183 feet respectively. The two shorter sides face to the north-east and south-west, having a length of about 103 and 133 feet respectively. The walls appear to have been constructed of a base or plinth of large uncut slabs, with masonry of sun-dried bricks superimposed. On the east face the masonry of the wall still showed a clear height of nine feet whereas elsewhere it was badly broken or embedded in débris. Of the base some three feet
' See Desert Cathay, i. p. 46, Fig. r8.