460 ON THE SOURCES OF HEAT AVHICH INFLUENCE CUMATK. 



mean temperature of the whole earth's surface. These formula determine with 

 moderate exactness the temperature of places under the temperate zone, but 

 Mayer's begins entirely to fail for the higher latitudes, while all of them are 

 uncertain in their application to places within the tropics, and to the whole 

 southern hemisphere; also since the computations of Fourier give to high 

 latitudes a degree of cold far greater than these empirical formulas, I despaired of 

 being able to assign a tolerably true mean temperature to the globe's surface ; but 

 this mean temperature will be higher than at first sight might appear, from the 

 circumstance of a degree of latitude and longitude within the tropics enclosing an 

 area so much larger than the same degrees when nearer the Pole. The value of 

 a degree of latitude is invariably sixty geographical miles; but those of longitude 

 regularly diminish from sixty miles on the equator, to fifty-two miles at latitude 

 30, to thirty miles at latitude 60, or half what they were on the equator, and one 

 mile at 80 latitude. Hence it follows that a large proportion of the superficies of 

 the globe is exposed to a tropical climate ; one half of the earth's surface is con- 

 tained within 30 degrees of north and south latitude ; therefore its mean tem- 

 perature at the level of the sea greatly exceeds that of our climate. Meyer's 

 formulas would give it at 65, Farhenheit, but this is certainly too high. .As 

 already stated, it is in vain in the present extent of our meteorological know- 

 ledge to endeavour to arrive at it correctly ; but assuming 60, Fahrenheit, as 

 the mean — which is as high as our data will warrant — then let us also suppose 

 the earth free from all central heat, and that it had existed a sufficient time for 

 its mass to attain the mean temperature of its surface ; then in penetrating into 

 the earth, under any latitude, and being careful to prevent the heat evolved by 

 the compulsion of the air from affecting the experiments, we should find the 

 temperature slowly to approximate to the mean of the central mass, which in 

 such a case would be the mean temperature of the whole superficies ; thus in high 

 latitudes the heat would increase as we descend, while under the equator a 

 diminution of heat would be expected. 



The latent caloric in the air, being reduced by its increased density, tends also 

 greatly to elevate the temperature as we descend under every parallel, particularly 

 where currents of air are circulating ; but the increment of temperature, as we 

 deepen our mines, is so great that it exhausts both of these recources, and, being 

 further confirmed by experiments conducted with the greatest care, where no 

 currents could vitiate the result, the elevation of temperature must be admitted 

 as altogether independent of solar rays ; and we have only left, to account for it, 

 the existence of central heat coeval with the earth's creation. 



The profound computations of M. Fourier led to this result ; the experimental 

 researches of M. Cordier and M. De La Reene tend towards similar conclusions ; 



