258 The Land of Deep Corrosions 
from which the pressure was applied. It is important to 
note that there stretches eastwards from the Tibetan plateau 
a mountain range which reaches substantially across China, 
forming the watershed between the Yellow River and the 
Yang-tze basins; and in fact all the greatest mountain ranges 
of Asia trend east and west. It is only this small area in 
south-eastern Tibet that has for some reason resisted the 
movement, with the results seen. 
Assuming such an origin for these great parallel ridges 
there are good reasons for believing that the lateral move- 
ment came from the west, not from the east, a subject to 
which we must refer presently. 
The Tsanpo river, flowing eastwards in the deep valley 
between the Himalayan and Trans-Himalayan ranges, 
came up against the most westerly of the mountain barriers 
thus raised, and swung away to the south and west; the 
Yang-tze, rising to the east of the last big ridge, followed 
the trend of the mountains till a region was reached where 
the effect of the squeezing was inappreciable, and, no longer 
fettered by local conditions, turned away to the east under 
the predominant influence of the main uplift. But the 
Irrawaddy, Salween, and Mekong, being caught right in 
the midst of the pinched area maintained their southward 
direction throughout their courses. 
In far western Ssu-chuan the main ridges still run 
approximately north and south, but as the Yang-tze has by 
this time made its great eastern bend, the rivers coming 
down from the north drain into it. Were this not the case, 
however—did the Yang-tze like the Mekong continue south- 
wards—these rivers themselves would, I think, have turned 
eastwards independently under the final influence of the 
main uplift. Far western Ssu-chuan seems to represent 
the unyielding barrier referred to above, against which the 
Tibetan ranges pushed in vain, and as such it has been 
crumpled up in some confusion, two sets of movements 
having been superimposed one upon the other. ‘The effect 
of the lateral movement is still sufficiently pronounced to 
determine the courses of the rivers from north to south, 
but not sufficient to obliterate completely the main axes of 
uplift running east and west. 
This is hypothesis. To substantiate such a theory the 
