307 



one place or another, and that of the 300 active volcanos 

 some are always in action, it would appear that such a lively 

 reaction of the interior of the earth against the crust is in 

 compatible with the constancy of its volume. 



This apparent discrepancy between Cordier s theory of the 

 connexion between the cooling of the earth and the reaction 

 of the interior on the exterior parts, and Laplace s calcula 

 tion showing the constancy of the length of the day, a calcu 

 lation which is undoubtedly correct, has induced most scien 

 tific men to abandon Cordier s theory, and thus to deprive 

 themselves of any tenable explanation of volcanic activity. 



The continued cooling of the earth cannot be denied, for 

 it takes place according to the laws of nature ; in this respect 

 the earth cannot comport itself differently from any other 

 mass, however small it may be. In spite of the heat which 

 it receives from the sun, the earth will have a tendency to 

 cool so long as the temperature of its interior is higher than 

 the mean temperature of its surface. Between the tropics the 

 mean temperature produced by the sun is about 28, and the 

 sun therefore is as little able to stop the cooling-tendency of 

 the earth as the moderate warmth of the air can prevent the 

 cooling of a red-hot ball suspended in a room. 



Many phenomena, for instance the melting of the glaciers 

 near the bed on which they rest, show the uninterrupted 

 emission of heat from the interior towards the exterior of the 

 earth ; and the question is, Has the earth in 25 centuries 

 actually lost no more heat than that which is requisite to 

 shorten a radius of more than 6 millions of metres only 15 

 centimetres ? 



In answering this question, three points enter into our 

 calculation ; (1) the absolute amount of heat lost by the 

 earth in a certain time, say one day ; (2) the earth s capacity 

 for heat ; and (3) the coefficient of expansion of the mass of 

 the earth. 



As none of these quantities can* be determined by direct 



