THE STRENGTH OF THE EARTH'S CRUST 551 



agreement with the measurements of gravity anomahes, so far 

 as the positive or negative sign of the center of mass is concerned, 

 yet are not closely related to the large maxima. They are, in fact, 

 in most cases decidedly eccentric to the anomaly contours. The 

 scarcity of these areas is a result of incompleteness of observations, 

 but their ecccii, ,: position and association with areas of moderate 

 anomaly is not an error due to the reconnaissance nature of the 

 studies. These relations indicate that the neighboring regions 

 giving broad domal areas of anomaly are in such cases not due to the 

 dominating control of centrospheric heterogeneity, for in that case 

 the resultants of the deflection residuals would point over broad 

 areas in the general direction of the epicenter of the mass. The 

 degree of discordance between the centers of dominant anomaly 

 and centers of dominant deflection indicates that fuller observations, 

 would produce agreement by adding to the number of such centers. 

 The present data suggest therefore that the areas of broad excess 

 or defect of mass as shown by the anomaly map are due to aggre- 

 gates more or less composite and shallow, so that each part influ- 

 ences individually to some extent the direction of the deflection 

 residuals about it. Special combinations of masses of shallow depth 

 with other masses below the zone of compensation could, however, 

 also account for the effects. The data of the present map of 

 gravity anomalies are therefore largely indeterminate, but the prob- 

 abilities point toward at least the greater part of the outstanding 

 masses lying well within the zone of compensation. In this con- 

 clusion the data agree with the other lines of evidence. 



RELATION OF DEPTH OF OUTSTANDING MASSES TO HYPOTHESES 

 REGARDING DISTRIBUTION OF COMPENSATION 



The measurements of the deflection residuals are very much 

 more detailed than are those of gravity anomalies. The evidence 

 from them is rather conclusive that, for the regions investigated, 

 the excesses or defects of mass which cause those residuals are 

 situated within the zone of compensation and more especially in its 

 outer half or third. Even if centers of outstanding mass were uni- 

 formly distributed, however, with respect to depth, they would 

 lose influence in proportion to the square of their depth. Smaller 



