SECT, xxvi.] HEAT DECKEASES WITH HEIGHT. 291 



ations in the surface of the land, as forests, sandy deserts, 

 verdant plains, rocks, <fec. ; mountain-chains covered with 

 masses of snow, which diminish the temperature ; the re- 

 verberation of the sun's rays in the valleys, which increases 

 it ; and the interchange of currents, both of air and water, 

 which mitigates the rigour of climates ; the warm currents 

 from the equator softening the severity of the polar frosts, 

 and the cold currents from the poles tempering the intense 

 heat of the equatorial regions. To these may be added 

 cultivation, though its influence extends over but a small 

 portion of the globe, only a fourth part of the land being 

 inhabited. 



Temperature decreases with the height above the level of 

 the sea, as well as with the latitude. The air in the higher 

 regions of the atmosphere is much cooler than that below, 

 because the warm air expands as it rises, by which its 

 capacity for heat is increased, a great proportion becomes 

 latent, and less of it sensible. A portion of air at the 

 surface of the earth whose temperature is 70 of Fahren- 

 heit, if carried to the height of two miles and a half, would 

 expand so much that its temperature would be reduced 50 ; 

 and in the ethereal regions the temperature is 90 below the 

 point of congelation. 



The height at which snow lies perpetually decreases from 

 the equator to the poles, and is higher in summer than in 

 winter ; but it varies from many circumstances. Snow rarely 

 falls when the cold is intense and the atmosphere dry. Ex- 

 tensive forests produce moisture by their evaporation ; and 

 high table-lands, on the contrary, dry and warm the air. In 

 the Cordilleras of the Andes, plains of only twenty-five 

 square leagues raise the temperature as much as 3 or 4 

 above what is found at the same altitude on the rapid de- 

 clivity of a mountain, consequently the line of perpetual 

 enow varies according as one or other of these causes prevails. 

 Aspect in general has also a great influence ; yet, according 

 to M. Jacquemont, the line of perpetual show is much higher 



u 2 



