430 ON THE DIFFUSION OF HEAT 



cause. This assumption, however, is preduded by the proof, 

 that heat is not capable of being carried off by the atmosphere, 

 but can escape only by radiation, which is dependent on an 

 elevated temperature. The conclusions seem, therefore, ne- 

 cessarily to follow, that the original temperature of the earth 

 must have been at least as low as the lowest natural tempera- 

 ture, and of course that the mean temperature has been rai- 

 sed. 



Wliile there is this rise, and this equalization of tempera- 

 ture at the surface, there must be a corresponding rise in the 

 internal mass. It is obvious that no permanent rise can take 

 place at the external layer, without a portion of heat being dif- 

 fused to the internal matter at a lower temperature. It is ob- 

 vious also, that a portion of the heat received necessarily must 

 be conveyed to that matter. And as this diffusion is limited 

 at the centre, the temperature must rise until it become equal 

 or nearly so through the whole, and as high, or nearly as high, 

 as that which the solar rays can excite. 



Lastly, the rise of temperature must observe certain limits ; 

 it must continue until that of the surface is such, that as much 

 caloric is discharged from it by radiation, as is received by the 

 solar rays, and it cannot proceed beyond this. The first con- 

 clusion necessarily follows, when it is proved, that radiation is 

 the only mode by which caloric can escape from this planet. 

 The second conclusion is equally evident. Thus the tenden- 

 cy is to an equal and uniform temperature. How far this may 

 be from that which at present exists, it is difficult to deter- 

 mine. If the increase of heat is accompanied v/ith its more 

 equal diffusion, so as to establish nearly the same temperature 

 over the entire surface, it may not proceed much beyond that 

 which now prevails at the hotter parts of the earth ; for at this, 

 when extended over the whole, the quantity of caloric radiated 



may 



