THE GRAVITY OBSERVATIONS OF 1 895 COMPARED 299 



because of the accordance of gravity anomalies, that so favorable results 

 as to small anomalies were obtained with a method of reduction based on 

 such generalizations as a compensation for the average elevation within 

 100 miles of the station, with both attraction and compensation com- 

 puted as indefinitely extended horizontal plates and without any assump- 

 tion as to the depth, thickness, or density arrangement of the compensa- 

 tion. 14 



comparison op Results for regional or local Compensation 



The evidence from the gravity determinations of the last half century 

 is strong that a general condition of isostatic equilibrium exists in the 

 outer portions of the earth's crust. The question of how close is this ad- 

 justment, of how local is the compensation of the topography, is of inter- 

 est. If the compensation is quite complete for each small topographic 

 feature, so that a single mountain or hill or canyon is exactly compen- 

 sated by a less dense or a more dense material beneath, the outer portions 

 of the earth would in detail be in a condition of nearly perfect equilib- 

 rium and would largely be free from stresses due to the supporting of 

 topographic features. On the other hand, if the compensation is more 

 general, such features of moderate extent would be supported by the 

 partial rigidity of the earth's outer materials. 



On my suggestion, in 1912 Hayford and Bowie made a study of this 

 question by computing the anomalies for 45 stations, using regional in- 

 stead of local compensation to three different zones whose outer radii were 

 respectively about 12, 37, and 104 statute miles, and by comparing the 

 anomalies thus obtained with those for complete local compensation, con- 

 cluded that the evidence indicated that the assumption of complete local 

 compensation is nearer the truth than regional compensation, even within 

 the smallest of these three zones. 15 



In 1917 Bowie applied the same method to 124 stations in the United 

 States, and compared the anomalies for stations grouped according to 

 their location. 10 The most significant of these are two groups in moun- 

 tainous regions, one of 22 stations below the general level, the other of 18 

 stations above the general level. Regional compensation for the two 

 smaller zones — that is, to the limit of 37 miles — gives average results 

 practically identical with local compensation, and it is concluded that 

 there is no evidence favoring regional or local compensation within this 



14 These 1895 anomalies were also the result of a first trial only of the method then 

 used without testing any variations in assumptions. 



15 Coast and Geodetic Survey Special Publication No. 10, p. 98. 

 10 Coast and Geodetic Survey Special Publication No. 40, p. 85. 



