AT THE earth's SURFACE. 415 



The great expansibility and mobility of an elastic fluid, such 

 as the atmosphere, give rise to a peculiarity with regard to 

 the communication of temperature through it. . Its proper 

 conducting power appeare to be inconsiderable, but, whatever 

 it may be, it is principally by the motion of its parts that heat 

 is diffused through its mass. When the temperature of a 

 lower portion of such a fluid is raised it ascends rapidly ; a 

 fresh portion comes in contact with the surface communicating 

 heat ; and by this successive application and retrocession of the 

 air, and this movement of the heated portion, rather than by 

 the direct communication of caloric from one part to another, 

 the temperature of the whole is raised. It is in this manner 

 that heat is diffused from the earth into the atmosphere. The 

 air incumbent on any part of the surface communicating heat, 

 is, by the elevation of its temperature, rarified, v/hence an as- 

 cending current is formed ; and the common opinion with re- 

 gard to this is, that it carries the excess of heat to the higher 

 regions of the atmosphere, and may allow it to be diffused in- 

 to the interplanetary space. It is on the assumption of this 

 that the heat must be supposed, in Mr Playfair's argument, 

 to be discharged from the earth, so far at least as it is convey- 

 ed from the surface by the surrounding elastic medium. The 

 just view of the operation of the atmosphere in diffusing heat, 

 is however more complicated ; other conditions connected 

 with it are to be taken into account, and lead, I believe, to a 

 very different result. 



The air heated at any part of the earth's surface, may, as it 

 ascends, impart, especially at first, a portion of heat to the air 

 with which it comes in contact. Bui as it rises, becoming sub- 

 ject to less pressure, it expands ; by this its capacity for heat is 

 augmented, and its temperature, therefore, falls proportionally. 

 As it recedes from the surface, with its temperature thus con- 



3 G 2 stantly 



