THE STRENGTH OF THE EARTH'S CRUST 513 



a result of the decrease in density, even if the zone of compensation 

 rested on an unyielding base.^ The logical correctness of this argu- 

 ment is not to be questioned, but rather the degree of its appli- 

 cation. The following arguments suggest that shattering or 

 porosity are, however, very subordinate rather than determining 

 factors in the isostatic problem. 



Such a theory does not account readily for the movements 

 needed to maintain isostasy because of erosion and sedimentation. 

 These surface changes of mass suggest a restoration of mass by 

 lateral undertow. Furthermore, the appeal to nature shows that 

 the rocks, once deep-seated, which have become revealed at the 

 surface by erosion, are almost without pore space. The average 

 porosity according to Fuller is o . 2 per cent, but the mean differ- 

 ences in densities between ocean and continent which must be 

 accounted for under the hypothesis of uniform compensation to 

 a depth of 122 km. amount to about 4 per cent. 



Joints are observed to decrease with depth, becoming tighter 

 and more distantly spaced, and the indications given by the lack of 

 general circulation of ground-water through crystalhne rocks, 

 except within joint spaces near the surface, are that at greater 

 depth the joint spaces are negligible. 



In the great compressive movements the whole thickness of the 

 crust must yield, but even this cannot be conceived as producing 

 porosity by granulation sufficient to notably modify the density. 

 A large part of the deformation in the deeper crust must be by a 

 process of recrystalhzation. Assume, however, that granulation 

 is the dominant process. Observation of granulated rocks shows 

 a reduction in size of the crystals, but these broken fragments fit 

 against each other perfectly and without great internal distortion 

 of crystals. In granulated rocks from the zone of flow there is 

 therefore always some amount of recrystallization, sufficient to 

 eliminate that porosity connected with minute shattering and 

 movement of the broken particles. The explanation appears to be 

 as follows: The minute shattering of the minerals tends to give 

 a high pore space, but with a high pore space the amount of contact 



^ G. T. Becker, "Isostasy and Radioactivity," Science, XLI (1915), 157-60; 

 "On the Earth Considered as a Heat Engine," Proc. Nat. Acad. Sci., I (1915), 8:-86. 



