300 G. R. PUTNAM CONDITION OF THE EARTH'S CRUST 



limit. By comparing the average anomalies for the outer zone, however, 

 Bowie concludes that local compensation is much nearer the truth than is 

 regional compensation to a distance of 104 miles. 



From general considerations, this conclusion may not be improbable, 

 but it should be noted that definite conclusions on this subject, based on 

 the gravity residuals, are difficult because of the nature of the effects being 

 measured. The pendulum measures the vertical component of the at- 

 tractive and other forces which make up the so-called force of gravity. 

 Hayford assumes the center of gravity of the compensation to be about 

 35 miles below the surface of the ground. The effect of a given compen- 

 sation on gravity at a station will be nearly as great when the compensa- 

 tion is displaced a moderate distance horizontally as when it is directly 

 underneath the station, and the compensation beneath a station will have 

 an appreciable effect for a considerable distance surrounding it. 



In a previous paper 17 I suggested the desirability of testing this matter 

 by comparing the anomalies from pairs of stations in the same general 

 locality, but differing considerably in elevation, as in such a comparison 

 distant effects and various uncertainties will be largely eliminated, and, 

 furthermore, it is in such regions that incomplete compensation might 

 be most likely to occur. 



The following, Table III, gives the results for the pairs of stations for 

 which comparable values are available. 18 



17 Science, December 20, 1912, p. 869. Barrell also developed later the same idea. 

 Journal of Geology, 1914, p. 160. 



1S The differences are taken from the results in Coast and Geodetic Survey Special 

 Publication No. 10, p. 100, and No. 40, p. 91, for Cloudland — Hughes. 



