340 P. King don W ard. 
It seems probable that the absence of Larix from the south 
slopes is due to summer, not to winter conditions, for no deciduous¬ 
leaved trees extend to the limit ol trees on that side, whereas deci¬ 
duous-leaved shrubs do; the limiting trees are in fact pronounced 
xerophytes protected against excessive transpiration. In north-facing 
valleys Larix usually shares with Abies the position of limiting 
species, and I have seen it, stunted and wind-torn, at well over 
15,000 feet. But besides exposure to the prevailing wind, other 
abnormal conditions result from the southern aspect, namely the 
early melting of the snows, partly we may suppose owing to the 
warm south-west winds, and partly to these slopes being in full 
sunshine throughout the day. The low altitude of the sun during 
the early spring makes this last difference more pronounced in the 
case of the higher and steeper ridges. 
The effect of this is twofold. In the first place a warm protective 
covering is rapidly stripped from the mountains, thus exposing the 
vegetation to a dry wind at the very commencement of the 
season, naturally the most momentous period in the life-history of 
the plant, since it finds itself for the first time thrown entirely on 
its own resources ; in the second place, the only water-supply on 
which the vegetation can confidently depend till about the middle 
of June, when the rains break, is removed at the precise moment 
when it is most required (compare PI. 7, Fig. 3, with PI. 8, Fig. 1). - 
In the second week of June we crossed the Mekong-Salween 
watershed at an altitude of 14,000 feet, the pass facing E.N.E. and 
W.S.W. (PI. 8, Fig. 2). Ascending the eastern slope we trudged 
through deep snow for three hours : hut at the summit the snow 
ceased abruptly in a high hank, and the western slope, which was 
extraordinarily precipitous, was clear of snow for nearly a thousand 
feet, though a big drift still lay on the gentler slope below. We 
crossed the same pass again in the second week of November, and 
were probably the last to get over, for the snow lay deep on both 
sides as low as 12,000 feet; at the same time there was no snow on 
the Mekong-Yangtze divide below 15,000 feet. The three passes by 
which we crossed the Mekong-Salween divide, ranging from 13,000 
to 15,000 feet, are quite clear of snow for possibly four months in 
the year, and are passable from June to November; on the other 
hand the Run-tsi-la, a pass from the Mekong to the Yangtze, which 
we crossed in mid-October, is over 17,000 feet and can he crossed 
from June to December, being clear of snow for about three 
months. 
