3o6 JOSEPH BARRELL 



Instead of a cylinder suppose the mass which produces the 

 deficiency of gravity to approximate more to the form of a sphere. 

 The results are shown in Table XXII. In columns 2 and 3 the 

 sphere is tangent to the surface, a position diminishing the mass for 

 a given anomaly. In columns 3 and 4 the top of the sphere is 

 7 and 14 km. deep respectively. The low density of column 4 

 shows it to be beyond the limiting conditions. The load, though 

 negative in sign, is seen to be equivalent in order of magnitude to 

 the greater volcanic piles; 30 to 60 miles in diameter, 9,000 to 

 14,000 feet in height for rock of density 2.67. The anomaly pro- 

 duced by the unit mass of 100 feet thickness and density 2 . 67, con- 

 sidered here as 100 feet of polar diameter for a spheroid of the given 

 horizontal dimensions, ranges between the low values of 0.0007 

 and o. 001 1 dyne. 



From a consideration of these two tables it is seen that the 

 large anomalies require either a variation of mass equivalent to 

 as much as 5,000 feet of rock extending over some thousands of 

 square miles or to 10,000 feet of rock, more or less, extending over 

 1,000 square miles, more or less. These tables determine the order 

 of magnitude, but the data are not sufficient to permit a more 

 accurate solution of the problem. 



Thus this detailed examination of the anomalies in the region 

 of Seattle shows that the divisor of 0.0030, as taken by Hayford 

 and Bowie, or 0.0024, as considered here the best for general use, 

 is too high for the more limited areas of high anomaly. The latter 

 may be regarded as made up in part of a regional portion for which 

 the divisor of 0.0024 would be applicable and a local portion for 

 which the divisor is probably not over 0.0015. As a mean value, 

 for the more limited areas of large anomaly the amount due to the 

 unit thickness of 100 feet of rock of density 2.67 should apparently 

 not be taken as over 0.0020 dyne. 



In forming conceptions as to the uncompensated vertical stresses 

 existing widely in the earth's crust it is important to know the 

 maximum range of departures from the mean stress as well as the 

 latter. These can be studied well in Fig. 5.' The mean of four- 

 teen maxima of defect of gravity is —0.033, the mean for eleven 



' Fig. 5, p. 153, Part II; also sec Hayford and Bowie, pp. 107-8. 



