INTRODUCTORY ESSAY. 217 



and commence again in the middle of September. A rather 

 sudden rise of temperature attends the vernal equinox, and 

 the summer is comparatively warm, the maximum sometimes, 

 but rarely, reaching 70°. 



At 13,000 feet the mean temperature probably coincides 

 with that of the freezing-point. At 14-15,000 feet the sum- 

 mer months alone are free from night-frosts, the maximum 

 temperature is only 60° in good shade, and the winter is pro- 

 portionately colder than at 13,000 feet ; thaw commences at 

 the end of April, the night-frosts are slight by the end of that 

 month, and the mean of the day rises to 50°. At 15,500 feet 

 it probably freezes during every night of the year. At 30,000 

 to 31,000 feet there is probably perpetual frost in the shade. 



These numbers however give no indication of the heat to 

 which vegetation is exposed, for, owing to the rarity of the 

 atmosphere and cloudless skies, the sun's rays have intense 

 power, increasing with the elevation, raising the (white glass) 

 thermometer exposed to them sometimes upwards of 100° 

 above the mean temperature of the air. This, combined with 

 the fact of the temperature of the soil being always above that 

 of the air, fully accounts for the sudden impulse given in 

 spring to the vegetation even in the loftiest and coldest re- 

 gions. The heat radiated from the naked rocks has also a very 

 powerful effect, especially on the summer crops. 



Extreme aridity is the characteristic of aU Western Tibet. 

 Rain and snow at moderate elevations are scarcely known, 

 and have no further direct effect on vegetation than is due to 

 the moisture of the soil produced by the melting of glaciers 

 and snow-beds. Dew and hoar-frosts are very rare pheno- 

 mena. The snow-level is nowhere below 18,000 feet; in the 

 mountains north of the upper Indus valley it rises to 30,000. 



Owing mainly to the great drought, the soU is in many 

 places covered with an efflorescence of carbonate and other 

 salts of soda, and salt-lakes are of frequent occurrence. Almost 

 all the large bodies of water indeed are more or less saline, 

 some of them intensely so, especially such as have no outlet, 



3/ 



