216 F. Kingdon Ward—The Land of Deep Corrosions. 
on the rainy sides of the several divides and the gradual shifting of 
the watersheds eastwards. This is inevitable. Thus the ’Nmai-hka 
will tap the Salween, and the Salween the Mekong, south of latitude 
28° (the southern limit of the arid region on the former river). The 
Mali might tap the ’Nmai—the watershed hangs nght over the latter 
river at present, and the former will then become the main stream of 
the Irrawaddy, the ’Nmai disappearing; but an elevation of the 
Salween—Irrawaddy divide, by further concentrating the monsoon 
wind currents within the Irrawaddy basin, might prevent this by 
levelling up the present unequal distribution between the Mali and 
the ’Nmai; but it would also curtail tapping operations further east, 
south of the arid region, just as they have been stopped in the arid 
region itself, where the rivers grind out their gorges in isolated 
grandeur, indifferent to the freaks of their neighbours because 
independent of water save from the distant glaciers of Tibet. 
It may be that the Mali has already tapped tributaries once flowing 
to the ’Nmai, and that it is gradually stealing all its waters on that 
side; certain it is that no rivers flow to the’ Nmai-hka from the west, 
and whereas we reached the crest of the Mali-’Nmai divide the 
second day after leaving the latter river, an endless series of ridges 
and valleys had to be crossed, occupying nine days, before the Mali 
was reached. 
We come now to a brief description of the rocks in the valleys from 
the Yangtze to the Mali-hka, and on the intervening ranges in about 
latitude 28°. In the Yangtze valley slates and schists were noticed 
above, on the left bank, limestone below, but crossing to the right 
bank and ascending the Mekong—Yangtze divide the order was 
reversed—first schists, then limestone, after which came an outcrop 
of granite ; altitude 9,000-10,000 feet. All these rocks were highly 
tilted, sometimes almost vertical, the schists crumpled; dip varying, 
but approximating to N.E. The summit of the divide is capped by 
limestone and a red grit, probably arkose; the former sometimes 
cropping out in irregular cavernous bosses. Granite and schist also 
crop out in places, dip E.S8.E. at angles varying from 45° to 90°; 
altitude 15,000-17,000 feet. 
Descending to Atuntsi we find chiefly mica-schists, and higher up 
limestones and slates. Atuntsi lies in a depression which is evidently 
a syncline. To the west rises a bulky outlier of the main divide, 
15,600 feet above sea-level, composed chiefly of slates, with curious 
pillars of limestone cropping out on its east and west faces. 
In the Mekong valley itself, purple and green slates are seen in 
the river bed, always standing on edge, and the same in the bed of 
the Salween ; the dip seems to decrease as one ascends. In a gorge 
of the Mekong just south of Atuntsi we pass from north to south 
through (1) vertical slates striking N.N.W., (2) granite, (3) lime- 
stone (?) ina few miles. Just south of the last is an outerop of coral 
limestone on the river bank. In the Mekong valley the schists and 
slates are everywhere almost vertical, the strike varying from N.N.E. 
to N.N.W. Altitude 7,000—9,000 feet. 
At 10,000 feet on the Mekong—Salween divide the schists and slates 
met with were vertical, striking almost due south, giving a succession 
