Sec. i] FROM KASHMIR TO CHILAS
The topographical facts just explained justify the conclusion that the choice of the circuitous route represented by the ` Gilgit Road ' was due mainly, if not solely, to the political and military conditions prevailing at the time when the Sikhs, in 1842, first extended their conquest to Gilgit. The story has been recorded by Mr. Drew with his usual accuracy and need not be repeated here." It makes it quite clear that the Sikhs took this line for their advance because the Dard Raja of Astôr had long since been tributary to them, whereas the tribal communities of Childs and the adjoining valleys of the Indus Kôhistdn maintained a sturdy independence and were foes whom it was wise to leave alone. Of the political conditions in these tracts during earlier times we have no direct historical information. But from the indications I shall have to discuss below as regards Darél, and from the Childs tradition that the whole of Shinkdri or the Dard portion of the Indus Kôhistân was under the rule of one Raja during pre-Muhammadan times,'' it seems safe to conclude that the region comprising Childs and the Kôhistan valleys westwards must have been during certain periods far more accessible to peaceful traffic than it has been in recent times.
If we turn once more to the map for the physical features that determined the lines of such traffic between Kashmir and India in the south and that portion of the Hindukush in the north which the Chinese records know by the name of P`o-lü, we cannot fail to realize an important geographical fact. The direct routes available for this intercourse are divided into two distinct and very unequal groups by the huge glacier-clad mountain mass that rises above the Indus valley between Astôr and Childs and culminates in the grand ice peak of Nangaparbat, 26,62o feet above the sea. To the east of it there is practically only a single line of communication, that which the present ` Gilgit Road ' follows across the Burzil pass (13,65o feet), with a branch route over the slightly lower Kamri pass which rejoins it in the main Astôr valley. The few routes from the Kishangangd to be found farther east all lead into Baltistdn or Skardo, situated much higher up the Indus.
To the west of the impassable barrier of the Nangaparbat massif conditions are quite different. Beginning with the Barai pass, which skirts the westernmost spur of Nangaparbat that carries permanent snow, we find quite a series of passes with practicable routes leading from comparatively well-inhabited parts across to the heads of the several valleys constituting Childs.13 These routes
Later change to ` Gilgit Road '.
Routes divided by Nangaparbat massif.
Easier routes west of Nangaparbat.
11 Cf. Drew, Jummoo and Kashmir, pp. 403, 437 sqq.
12 See Biddulph, Hindoo Koosh, p. 16.
13 Starting from the east we have first the route up the Kg valley and across the Barai pass, with a threefold continuation beyond the north foot of the latter at Surugan. One leads straight down the Bünar valley northward to the Indus, as already mentioned, p. 4 ; the second, followed by myself and found practicable for laden animals, though with difficulty in some places, diverges to the north-west and leads across the Fasat pass into a branch of the Niat valley ; the third leads westwards along easy uplands and gives access transversely to every one of the routes as far as the Babusar pass to be mentioned presently.
Next to the Kél—Barai route we have that which starts from Shardi, the ancient sanctuary of the goddess aradd on the Kishanganga (see my note on Rajat. i. 37 ; vol. II. pp. 279 sqq.) and leads up the considerable stream (known to the Sanskrit tradition of Kashmir as Sarasvati, cf. Rajal. i. 37 ; II. p. 282) that there joins the Kishangangd from the north.
This route crosses the watershed by the Kamakdôri pass (14,120 feet) and thence leads down to the Niat valley
by the Biah Nullah. Its practicability is attested by the official proposal made some time ago to the Kashmir authorities to open it up as a mule road and an alternative line of communication with Childs. Its convenience for this purpose was previously demonstrated in another fashion by marauding inroads of Childsis, which obliged the D6gras, in Maharaja Gulab Singh's time, to erect a fort at Shardi for the protection of the valley. In 1891 this danger had not yet disappeared completely, and I found the fort still garrisoned by a small detachment (cf. R. ~ijal. II. p. 284).
To the north-west of the Kamakdbri route there branch off at least two, if not three, side valleys connecting Shardi by practicable routes (Jalkhadgali, Purbia, Sarai passes) with the open grassy uplands at the head-waters of the Kaghan valley. Notwithstanding the convergence upon Shardi of these several routes, I doubt whether any of them could ever have claimed the importance of the Barai route as a means of direct communication with the main Kashmir valley ; for the configuration of the mountains to the south and the narrowness of the Kishanganga valley both above and below Shardi make this place more difficult of access than Kél from Kashmir proper. It is thus for good reasons