752 EXPLORATION TN (!HTNESJ: TURKESTAN. 



touched my topoonipliical interest>^, I could not resist the temptation 

 of pushing westward, at least as far as the Wakhjir Pass, which leads 

 from the Taohdumhash Pamir to tlie headwaters of the Oxus. Camp- 

 inij- cl()S(^ to the summit of the ^Vakhjir Pass (circ. ir),2<>0 feet), 1 visited 

 on July 2 the head of the Ab-i Panja Valley, near the great glaciers 

 which Lord Curzon first demonstrated to be tlie true source of the 

 river Oxus. It was a strange sensation for me in this desolate moun- 

 tain waste to know that I stood at last at the eastern threshold of that 

 distant region, including Bactria and the upper Oxus Valley, which, as 

 a fi(d(l of ex[)loration, has attracted me ever since I was a boy. It was 

 the thi'cshold oidy 1 had reached, and I knew that this time there was 

 no entrance for me into the forbidden land. Notwithstanding its 

 great elevation the Wakhjir Pass and its approaches, both from the 

 west and east, are comparativel}^ easy. Comparing the topographical 

 features with the itinerary indicated by Hiuen Tsiang, the great Chi- 

 nese pilgi'im, 1 am led to conclude that the route which he followed 

 when traveling, aliout A. D. 649, on his return from India, through 

 the \all(\v of Pa-mi-Io (Pamir) into SariUol, actually traversed this 

 pass. 



As I marched down the gradually widening valley of the Taghdum- 

 bash Pamir toward Tashkui-ghan, the chief place of the Sarikol dis- 

 trict, I fully realized the contrast which its expanses of comparatively 

 rich grazing ofi'er to the I'ocky destitution of the Hunza gorges. 

 Increasing numbers of nomadic herdsmen, both Kirghiz and Wakhi, 

 now frequent the valley, which was an utterly deserted waste, and 

 rarely used, even as a route, while there were Hunza raiding parties 

 ready to swoop down from the mountain fastnesses southward. 



I also felt glad to be once more on the track of Hiuen Tsiang, wdiose 

 footsteps T had traced to so many a sacred Buddhist site of ancient 

 Indiii. The position and remains of Tashkurghan were found to agree 

 most closely with th(^ description which Hiuen Tsiang and the earlier 

 Chinese pilgritn, Sung-yun, give of the capital of the ancient Kie-pan-to. 

 The identification of the latter territory with the modern Sarikol, first 

 suggested ])y Sir Henry Yule, was thus fully established. The ruined 

 town, which extends round the modern Chinese fort of Tashkurghan, 

 and still shows a (juadrangular inclosure of crumbling stone walls, 

 "rests on a great rocky crag, and is backed by the river Sita"" (i, e., 

 the Yarkand River), on the east, exactly as the ])ilgrims describe it. 

 As a striking instance of the tenacity of local tradition, it deserves to 

 })e mentioned that I found the curious legend which Hiuen Tsiang 

 relates of the princess imprisoned in ancient days on a rock fastness 

 still clinging to the identical locality of this valley. 



I believe that Tashkurghan, as an historical site, has claim to even 

 greater anticiuity than that implied l)v the notices of Miucn Tsiang 

 and Sung-yun. Nature itself has plaiidy marked it not only as the 



