Limits of Plants in North-West Yunnan. 
34i 
Incidentally two other points call for notice. 
In the first place a snow-clad slope is to a large extent protected 
from atmospheric denudation, and, weathering being in excess of 
transport, it presents both a gentler grade and a more broken 
surface than does the exposed slope; this I think in itself affects 
the composition of the flora, permitting many plants to find shelter 
on the one side which is denied them on the other. 
Secondly, while a high pass near the tree-limit is always bare 
of trees for some hundreds of feet, even on the sheltered side, yet 
scattered forest may extend up the valley sides wherever it can find 
foothold, to a height well above the actual col on the north face, 
though not on the south face, a phenomenon undoubtedly due to the 
fact that a well-marked col is raked by a sharp up-valley wind on both 
sides, as 1 have observed on several occasions. 
In July, seeing that the vegetation on either side of a north 
and south facing pass at an altitude of about 16,000 feet was very 
similar near the summit of the pass, but exhibited distinct formations 
to north and south a few hundred feet below, namely alpine grass¬ 
land in the former case, and scree vegetation in the latter, I took 
the soil-temperatures, an up-valley wind blowing from either 
direction. 
On the north side, thirty yards from a large patch of melting 
snow the soil-temperature was 52-7"?., the air-temperature on the 
ground 47'5° P, and in the wind five feet above the ground 43 , 7°F. 
On the south side the same positions gave precisely the same 
readings, showing the similarity of conditions, as regards tempera¬ 
tures which vitally affect the functioning of a plant’s organs, on 
either slope near the summit of a high pass. (The wind-tem¬ 
peratures here noted may of course be neglected so far as the 
existing vegetation is concerned, since at that altitude no plant 
rises more than a few inches above the ground, largely for that 
very reason however). 
This similarity of vegetation does not apply in the case of a 
ridge exposing north and south slopes, for the summit of the ridge 
marks the division between dense scrub vegetation and bare scree. 
On the south side of such a ridge, on July 19th, at an altitude 
of 15,000 feet, 1 obtained 56-3° and 49-5°F.for soil and air respectively, 
while on the north side I got 59-9 n and 58T°F. within a stone’s 
throw of the former readings. It is evident then that a col exercises 
considerable local influence on the formation. 
Exactly the same principle is illustrated in the case of valleys 
