EAST NEPAL AND THE SIKKIM HIMALAYA MOUNTAINS. 113 



'account for^tlie great injury plants sustain from a thin covering 

 of ice on their foliage, even when tlie temperature is but little 

 below the freezing point. 



The power of terrestrial radiation increases with the elevation, 

 as does solar radiation, but not in an equal proportion. At 

 7000 feet there is little radiation during the rains ; the nights 

 are almost invariably cloudy — 3° to 4° is the mean maximum, 

 but it is not on one night out of six that there is any radiation. 

 From October to December the amount is greater = 10° to 

 12°, and from January till May greater still, reaching 15°. 

 During the winter months the effect of radiation is often felt 

 throughout the clear days, dew forming abundantly at 4000 to 

 8000 feet in the shaded bottoms of narrow valleys, into which 

 the sun does not penetrate till 10 a.m., and from which it dis- 

 appears at 3 P.M. I have seen the thermometer in the reflector 

 fall 12° at 10 A.M. in a shaded valley. This often produces an 

 anomalous effect, causing the temperature in the shade to fall 

 aftei" sunrise ; for the mists which condense in the bottoms of 

 the valleys after midnight disperse after sunrise, but long before 

 tlie sun reaches the valleys, and powerful radiation ensues, lower- 

 ing the surrounding temperature. A fall of 1° to 2° after sun- 

 rise of air in the sliade is hence common in valleys in November 

 and December. The excessive radiation of the winter months 

 often gives rise to a curious phenomenon ; it causes the forma- 

 tion of copious dew on the blanket of the traveller's bed, which 

 radiates to the tent roof, and this inside an open or closed tent. 

 I have experienced this at various elevations, from 6000 to 

 16,000 feet. Whether the mininum temperature was as high 

 as 50° or but little above zero, the effect is the same, except 

 that hoar frost or ice forms in the latter case. Another remark- 

 able effect of nocturnal radiation is the curl of the alpine 

 Rhododendron leaves in November, which is probably due to 

 the freezing and consequent expansion of the water in the upper 

 strata of cells exposed to the sky. The first curl is generally 

 repaired by the ensuing day's sun, but after two or three nights 

 the leaves become permanently curled, and remain so till they 

 fall in the following spring. 



Many alpine plants resist a great degree of radiation ; the 

 Cyananthus, especially, I have observed to be uninjured by a 

 minimum temperature of 31'0° lowered 12'0° by radiation, in 

 the month of September, at 15,500 feet; and yet this is one of 

 the most delicate as well as beautiful of Himalayan blossoms. 

 As a general rule, however, the commencement of the Septem- 

 ber and October radiation is the signal for the extinction of the 

 alpine herbaceous vegetation. 



I have elsewhere said that the nocturnal radiation of the 



