309 



that the earth's radius has not diminished 1 decimetre in 

 consequence of cooling, we are obliged to assume, according 

 to the premises stated, that the mean temperature of our 

 planet cannot have decreased 4 -^ in the same period of time. 



The volume of the earth amounts to 2650 millions of cu- 

 bic miles. A loss of heat sufficient to cool this mass ^ 

 would be equal to the heat given off when the temperature 

 of 6,150,000 cubic miles of water decreases 1 ; hence the 

 loss for one day would be equal to 6*74 cubic miles of heat. 



Fourier has investigated the loss of heat sustained by the 

 earth. Taking the observation that the temperature of the 

 earth increases at the rate of 1 for every 30 metres as the 

 basis of his calculations, this celebrated mathematician finds 

 the heat which the globe loses by conduction through its crust 

 in the space of 100 years to be capable of melting a layer of 

 ice 3 metres in thickness and covering the whole surface of 

 the globe ; this corresponds in one day to 7*7 cubic miles of 

 heat, and in 2500 years to a decrease of 17 centimetres in 

 the length of the radius. 



According to this, the cooling of the globe would be suffi- 

 ciently great to require attention when the earth's velocity of 

 rotation is considered. 



At the same time it is clear that the method employed by 

 Fourier can only bring to our knowledge one part of the heat 

 which is annually lost by the earth ; for simple conduction 

 through terra firma is not the only way by which heat escapes 

 from our globe. 



In the first place, we may make mention of the aqueous 

 deposits of our atmosphere, which, as far as they penetrate 

 our earth, wash away, so to speak, a portion of the heat, and 

 thus accelerate the cooling of the globe. The whole quantity 

 of water which falls from the atmosphere upon the land in 

 one day, however, cannot be assumed to be much more than 

 half a cubic mile in volume, hence the cooling effect produced 

 by this water may be neglected in our calculation. The heat 



