212 F. Kingdon Ward—The Land of Deep Corrosions. 
this region, namely, the Yalung, the Dayul River (or Wi-ch‘u), 
and the Ngawchang-hka, while the upper portion of the ’Nmai-hka 
attempts a similar evolution and achieves a double bend, each half of 
which is less than a right angle. By the pinching of the rivers 
between rocks subjected to two sets of earth mevements acting at 
right angles to each other, in some cases actually buckling the strata, 
and to cutting back, owing partly to differences of rainfall on the two 
sides of a mountain range and partly to differences of level and of 
grade in the river-beds, which has gradually brought rivers originally 
belonging to separate hydrographic systems into contact, I imagine 
the whimsical courses of these rivers to have been evolved from simple 
courses. Whether this explanation be right or wrong, there can be 
no doubt that the several rivers which exhibit this phenomenon of 
reversed flow owe it to a common cause, and not to some chance freak 
in each case. 
It may be asked, what became of all the rain-water which must 
have fallen on this country previous to the west to east movement 
which, by cutting across and breaching the long axis of the great 
Asiatic divide, allowed the parallel rivers to drain southwards? ‘The 
answer is, much of the region was then occupied by vast lakes into 
which the water was poured. In the case of Hkamti—Loong, to be. 
referred to presently, at the sources of the Mali-hka (or western 
branch of the Irrawaddy), it is sufficiently obvious that this plain was 
once occupied by a lake, and again much further east we have an 
. ancient lake bottom in the ‘red basin’ of Ssuchuan, now the fertile 
Chengtu plain. But there is proof in the sandstones, slates, and lime- 
stone of the border country itself that most. of it was once under 
water. Much of the Mekong—Yangtze divide, for instance, is capped 
by limestone now raised 18,000 or 20,000 feet above sea-level, and 
the same rock reappears on the ranges to the west. North of Likiang 
we cross plateau country partially occupied by insignificant lakes, 
surrounded by sandstone or limestone ranges from the bases of which 
well up hot springs. But whereas the Hkamti plain is only 1,200 feet 
above sea-level, with parallel ridges of sands and clays enclosing 
organic remains still intact, rising 3,000 or 4,000 feet higher in the 
south, the old sea or lake bottoms east of the Yangtze have been 
pushed up into plateaux 8,000 to 12,000 feet above sea-level, and 
regional metamorphism must have obliterated any organic remains 
which may once have existed. Whereas the isolated lakes of Tibet 
are drying up, those of the Yunnan—Tibet border country have been 
drained. 
As already remarked, this border country has, in my opinion, been 
subjected to two sets of crust movement acting approximately at 
right angles to each other in such a manner that the hydrographic 
system set up during the first phase, at a later period became involved 
with and largely obliterated by that resulting from the second phase ; 
further complications were gradually introduced by the cutting back 
of either the primary or the secondary rivers, but especially by the 
former, according to local circumstances. During the first phase 
the Himalayas and the main backbone of China, separating the 
Yangtze and Yellow River basins, were raised up; during the second, 
a 
ae ae ae a 
