106 CLIMATE AND VEGETATION OF 



of the branchlets and foliage is covered by small scales of 

 a rather bright ferruginous-brown. Leaves ^ to ^ of an inch 

 long, pale green. Corolla -^ of an inch across the lobes, of a 

 purple colour. The whole plant is very odoriferous. It appears 

 indifferent to the changes of climate, remaining buried under 

 many feet of snow for eight months of the year ; whilst at other 

 periods the soil around it is heated to 150°. In the most genial 

 weather snow-storms are of frequent occurrence ; they do not, 

 however, injure its blossoms, which remain open until fertilization 

 has taken place. This species attains, I believe, a loftier ele- 

 vation than any other shrub in the world. Its nearest allies are 

 R. setosum and R. Lapponicum, from which latter it differs in its 

 smaller stature and solitary sessile flowers. 



It is now my purpose to enter into some details respecting the 

 temperature and climate of the three zones of the Sikkim Hima- 

 laya, to which the Rhododendrons are chiefly confined. The 

 information on this head is not so satisfactory as is desirable, 

 owing to the want of a series of observations having been con- 

 tinued throughout the year anywhere, except at Darjiling itself, 

 and there for but a few hours daily through one year. The 

 mean annual temperature of that locality at 7000 feet may be 

 considered as very nearly 53°, probably rather below than above 

 it, this result being deduced from the daily maxima and minima, 

 which, I think, gives too high a result, from local causes which 

 I shall hereafter explain. The monthly means, on the other 

 hand, vary much year by year, and with small differences of 

 position, owing chiefly to the variable amount and unequal dis- 

 tribution of the rain-fall and cloud, and the great power of the 

 sun's rays when unobstructed. The wonderful equality of tem- 

 perature throughout the 24 hours from May till October is only 

 disturbed by the sun's rays, which raise the thermometer 20° in 

 a very few minutes, and sometimes for an equally short period 

 of the whole day. During the height of the rains the mean 

 temperature varies but a few tenths of a degree (June 6r2°, 

 July 61*4°, August 61 '7°), and there is no radiation by night of 

 any consequence. Whilst, therefore, a short sunshine raises the 

 thermometer for a few minutes 10° above the mean of the 

 24 hours, the minimum thermometer never falls more than 3° 

 to 4° below the same mean, whence it follows that the mean of 

 the whole day cannot be indicated by that of the maximum and 

 minimum, as shown by a self-registering thermometer. Again, 

 the maximum of heat occui's during the rains very generally 

 before noon, often before 10 a.m., the forenoon being the least 



