418 ON THE DIFFUSION OF HEAT 



which this could have been attained, than by an atmosphere 

 constituted as ours is, — expanding as it is heated, and falling in 

 temperature from augmentation of capacity as it expands.. To 

 admit of an unlimited conveyance of heat, a solid mass, how- 

 ever subtle, would have been required, or an atmosphere the 

 capacity of which for caloric should not increase with its rare- 

 faction. 



The passage of heat, is thus from the circumference towards 

 the centre of the globe, instead of the opposite direction, and 

 it is prevented from accumulating at the surface only by being 

 conveyed into the mass of earth. 



These conclusions may be applied to the present argument. 

 If the view I Jiave stated be just, we have demonstration that 

 the atmosphere conveys no heat into unlimited space ; our pla- 

 net, in relation to the discharge of caloi'ic from it, is bounded 

 as it were by a wall of non-conducting matter, and no arrange- 

 ment can be imagined more perfect, by which the heat belong- 

 ing to it could be confined. If the heat from the centre, Mr 

 Playfair observes, on arriving at the circumference, were pre- 

 vented from proceeding farther, and if space or vacuity pre- 

 sented to heat an impenetrable barrier, then an accumulation 

 of it at the surface, and at last a uniform distribution of it 

 through the whole mass, would inevitably be the consequence. 

 It appears that the atmosphere does present such a barrier, and 

 the consequence, therefore, necessarily follows *. 



This 



* If the. sphere from the centre of which heat is supposed to be diffused, be 

 ■ not exposed to an unequal external source of lieat, the heat will be communica- 

 ted equally at its circumference to the incumbent almospliere, and produce as- 

 cent and descent of the air, with little of that calculation of it from one part to 

 another, which is the consequence of inequality of temperature at the surface of 

 the solid. But still the same changes of capacity for heat will accompany tlie 



ascent 



