________________
OCTOBER, 1917]
THIRD JOURNEY OF EXPLORATION IN CENTRAL ASIA
229
Three rapid marches down the Great Pamir River then carried me to Languar-kisht, where we reached the main Oxus Valley, and the highest of the villages on the Russian side of the river. Here, too, everything was done by the Commandant of the Russian frontier post and the local Wakhi headmen to facilitate my journey. My subsequent journey down the Oxus was attended by an abundant harvest of observations bearing on the historical topography, archaeology, and ethnography of Wakhan, which in early times had formed an important thoroughfare between Bactria, India, and the Central-Asian territories controlled by China. But it would cost too much time and space if I attempted here to give any details. It must suffice to mention that the exact survey of a series of ruined strongholds, some of them very considerable extent, acquainted me with numerous features of distinct archæological interest in their plans, the construction and decoration of their bastioned walls, etc. The natural protection offered by unscaleable rock faces of spurs and ravines was always cleverly utilized in these defences. But some idea of the labour, which even thus their construction must have cost, can be formed from the fact that at one of these strongholds, known as Zemr-i-Atish-parast, the successive lines of walls, with their bastions and turrets solidly built in rough stone or in sun-dried brick, ascend the slopes of a precipitous spur rising to an elevation of close on 1000 feet, and have an extent of more than 3 miles.
It is certain that these hill fastnesses date back to pre-Muhammadan times and to a period when this portion of the Oxus Valley contained a population far denser than at present and enjoying a higher degree of material civilization. Their attribution by the present Wakhi people to the "Siahposh Kafirs" merely gives expression to a vague traditional recollection that they date back to times before the advent of Islam, the Siahposh" of Kafiristan south of the Hindukush never having reached the stage of civilization which these ruins presuppose. Some architectural details seemed to suggest a period roughly corresponding to late Indo-Scythian or early Sassanian domination, during which our scanty records from Chinese sources indicate that Wakhan enjoyed a state of relative affluence and importance.
All along the big valley of Wakhan there opened glorious vistas to the south, where towering above narrow side valleys, and quite near, appeared magnificent ice-clad peaks of the Hindukush main range, looking just as early Chinese pilgrims describe them, like peaks of jade. I realized now what an appropriate invention the "popular etymology" was, which in Muhammadan times has connected the old and much-discussed name of Bolor, vaguely used for the Hindukush region, with the Persian billaur, meaning crystal. The effect was much heightened by the unexpectedly verdant appearance which the cultivated portion of Wakhan still presented at that season, in spite of the elevation from $,000 to over 10,000 feet above sea level, and doubly welcome after the bleak Pamirs. It was pleasant to note abundant evidence of how much the resources of the Wakhis on the Russian side of the valley had increased, both in respect of cattle and sheep and of land brought under cultivation, since annexation under the settlement arrived at by the Anglo-Russian Pamir Boundary Commission had removed all trouble from Kirghiz raids and Afghan exactions.
For all these reasons I felt glad that plentiful antiquarian and anthropometric work kept me busy in Wakhan during the first half of September. To this was added a philological task when, on entering that portion of the valley which adjoins the great northward end of the Oxus and is known as the tract of Ishkashim, I could collect linguistic specimens of the hitherto unrecorded Ishkashmi, one of the so-called Pamir dialects which form an