THE STRENGTH OF THE EARTH'S CRUST 445 



by the more important attractions of superficial masses and hidden 

 by the indeterminate nature of much of the present data. There- 

 fore centrospheric heterogeneity is not a hypothesis which can be 

 used to account for the apparent departures from isostasy. It 

 can be at most only a very secondary factor. 



In the last topic of Section B is discussed the relation of the 

 depth of outstanding masses to the various hypotheses regarding 

 the distribution of compensation. The hypothesis of local com- 

 pensation, as is perceived to a certain extent by those who have 

 used it, is in error in supposing that variations in density correspond 

 to every topographic feature and extend uniformly to the bottom 

 of the zone. But these errors, whether they be small or great, are 

 so spread out in depth and their centers of attraction are conse- 

 quently so far removed from the surface that they have little effect 

 on the geodetic observations. Especially is this true in compari- 

 son with those large and concentrated outstanding masses due to 

 bathoHthic invasion or other causes which are found to exist at 

 moderate depths in the outer crust. The reasons then why the 

 deflection residuals and gravity anomalies appear to show so little 

 relation to local surface relief and larger physiographic provinces 

 are threefold: in part because of a regional compensation; in part 

 because the hypothesis of local compensation as here shown masks 

 the error contained in the assumption of perfect and local isostasy; 

 in part because for many regions the ancient heterogeneities of 

 mass hidden within the crust seem in reality to be greater than the 

 heterogeneities of mass visible at the surface in topographic form 

 and created by present gradational actions. 



The results of this chapter converge with the lines of evidence 

 previously considered and confirm them in showing considerable 

 defects from isostasy for areas which are 100 km. or more in 

 radius. This confirmation is to be expected, since it would be 

 indeed remarkable if a crust, competent to carry such loads as the 

 geologic evidence from erosion and sedimentation shows to be 

 imposed, should give geodetic evidence of fairly local and nearly 

 perfect adjustment between the topographic forms, developed by 

 present external processes, and the variations in density imposed 

 by past internal forces. 



