ESTIMATES AND CAUSES OF CRUSTAL SHORTENING 43 



Loss of heat by convection is accomplished through transfers 

 of magma and transfers of water. 



If calculations are made of the loss of heat due to vulcanism, 

 upon the basis of the volcanic materials brought to the surface 

 of the earth at the present time, it is highly probable that the 

 conclusion would be reached that this quantity is so small that 

 it might be ignored. But one must remember that present vul- 

 canism is no criterion by which to estimate the transfers of 

 material which have resulted during past great periods of 

 regional vulcanism. The transfers of enormous quantities of 

 igneous material by vulcanism from deep within the earth to its 

 outer shell or to the surface of the earth, described on page 48, 

 strongly suggests that convectional currents have been a more 

 important factor in the process of secular cooling that has been 

 supposed. A large part of the heat which is carried toward the 

 surface of the earth by magmatic convection is transferred only 

 a part of the way by the liquid rock. For, as seen (p. 49), 

 the intrusive rocks probably equal or exceed in quantity the 

 extrusives, and deeper transfers may have occurred of which we 

 have no definite knowledge. From the place where the magma 

 is stayed the heat is brought to the surface in two ways. First, 

 a part is transferred by conduction through the overlying mantle 

 of rock. That such conduction occurs is shown by the fact that 

 the temperature gradients in many districts receiving hotter 

 material are higher than the average. Second, another part of 

 the heat is brought to the surface by convection through under- 

 ground waters. In this case the transfer of heat begun by the 

 magma and by conduction is continued by water. 



This brings us to the second agent by means of which the 

 earth is losing heat through convectional currents. Underground 

 water circulation everywhere pervades the outer zone of fracture 

 at the present time, and doubtless has since a solid crust existed. 

 That heat is brought to the surface of the earth hy water is self- 

 evident in the various districts of geysers and hot springs. 



But in estimating the amount of heat which escapes by convec- 

 tional transfer by underground waters, it is not sufficient to con- 



