148 DB. E. IT. PASCOE O^ THE EARLY HISTORY [vol. lxXV 



It is perhaps idle to speculate far in this direction, but the 

 following considerations are worthy of some attention. An 

 examination of the map will show an interrupted hydrographic line 

 from Pemakoi to Gilgit, represented at the present day by the 

 Tsangpo (perhaps including the Raga tributary and a small branch 

 •of the Chaktak), the Manasarowar Lakes, the uppermost reaches of 

 the Sutlej, the Indus along its Gartang channel and its tributary 

 the Shyok, as far as Gilgit, and perhaps the Gilgit River beyond. 

 It seems more than likely that this line was once a continuous 

 •drainage-line sloping westwards, and occupying either a geosyncline 

 or a line of easy denudation, such as a belt of softer beds or a fault. 

 The marine and fluviatile Xummulitics of Ladak belong to this 

 line, corresponding in position to the Nummulitics of the hydro- 

 graphic line on the other side of the Himalaya, and indicating that 

 the Tibetan line is as old as the Indian. The Pleistocene of the 

 uppermost reach of the Sutlej also belongs to it. Eventually, there 

 is reason to believe, the Tibetan line was occupied by a river which 

 rose somewhere in or near Pemakoi, and either joined the Oxus, or 

 found its way to the Arabian Sea, independently or by way of the 

 Indobrahm. At the present day the Indus at Bunji, where it 

 .abandons its north-westward course and turns southwards, is 

 3400 feet lower than the Tsangpo-Brahmaputra where it leaves the 

 same geotectonic line. 1 From Mayum, a town above Lake Ukrang 

 •east of the Manasarowar Lakes, to Bunji, the average westward 

 gradient is twice as steep as it is from Mayum eastwards to the 

 point where the present Tsangpo turns southwards. In spite of its 

 immense elevation the Tsangpo is a sluggish and navigable river 

 south of Lhasa, and has cut no deep basin for itself in Tibet. 

 The Indus, on the other hand, has cut its bed to a low level on 

 the Tibetan Plateau, the fall in this part of its course being hardly 

 more than 3 feet in a mile ; across the Himalaya it has a remark- 

 ably equable and comparatively gentle fall.' 2 Without attempting 

 to guess the order in which events occurred, there are reasons for 

 'deducing : — 



(1) That this Tibetan River flowed westwards and north-westwards from 

 Pemakoi to Gilgit. 



(2) That its uppermost waters were captured, possibly in turn by the 

 Irrawadi-Chindwin and the Meghna, but finally by the Indobrahm or Brah- 

 maputra, the capturing river at once beginning to cut back up one of its 

 own small tributaries westwards along the already excavated channel of the 

 •old Tibetan River, capturing this channel piece by piece and its tributaries 

 ■one by one, and completely reversing its drainage. 



(3) That, perhaps before this struggle had proceeded far, the Kali G-andak 

 may have captured the upper part of the Tibetan River, and have thus been 

 •enabled to scoop out the extraordinary depression of the Photu Pass, which is 

 •only 250 feet higher than the Tsangpo plains. 8 



(4) That the Sutlej effected a similar capture, part of the captured river 

 ■still existing as the uppermost reaches of the Sutlej ; and that the dwindling 



1 S. G. Burrard & H. H. Hayden, op. jam cit. p. 171. 

 - Ibid. p. 171. 3 Ibid. pp. 147 & 155. 



