40 EVOLUTION OF THE EARTH 



be a granite. The separation has taken place after the in- 

 trusion of the magma, the denser material sinking, the lighter 

 rising. There are indications that the process goes forward 

 on a larger scale also, a scale so large that the dark and heavy 

 base is never seen, erosion of later ages being restricted en- 

 tirely to the granite zone. Such a splitting in composition is 

 indicated in that the earlier intrusions in a period of igneous 

 activity are intermediate or basic and the later products are 

 vast bodies of granite. The greater density in the earth's 

 interior suggests a primal density stratification on even a larger 

 scale, which has been discussed under a previous topic. 



But in the outer shell, 50 to 75 miles thick, the density is 

 far from being uniform. In recent years it has been proved 

 by means of precise geodetic measurements on the local in- 

 tensity of gravity and deflections of the vertical that the crust 

 (beneath the continents is notably less dense than that beneath 

 I the oceans. The most of this difference in density exists prob- 

 *ably within the outer 50 miles. The continents stand high, 

 consequently, for the same reason that an iceberg rises above 

 the surface of the sea : it is the position of equilibrium. At a 

 certain depth the downward pressures given by the thicker 

 continental and the thinner oceanic crust are the same and a 

 condition of hydrostatic equilibrium prevails in the sub-crustal 

 shell. 



This condition of equal pressures at a certain depth is called 

 isostasy. It is not inconsistent with a solid and rigid condi- 

 tion of the earth, but does mean that at a greater depth, 

 apparently from 50 to 300 miles or more, hot but solid rock 

 can slowly yield and flow by recrystallization. The process 

 is physically the same as that by which a glacier flows under 

 the slight stimulus of an almost level surface slope. The condi- 

 tion for such ready recrystallization is found in temperatures 

 which are close to those of fusion. At such temperatures 

 molecules under strain pass readily from the solid to the liquid 



