408 H. S. WASHINGTON ISOSTASY AXD ROCK DENSITY 



mate truth. We seem, then, to be justified in assuming that the best 

 normative value for the isopiestic depth is 59 kilometers, and we may 

 accept 60 kilometers, about 37 miles, the value of Bowie, as the nearest 

 approximation to the true isopiestic depth. 



It need not be attempted to explain the difference between the value of 

 60 kilometers, arrived at by the normative method and gravity methods 

 from many stations, and that of 96 kilometers, accepted as the best by 

 Bowie. The probability of such a high value as 96 kilometers being the 

 more correct may, however, be roughly tested. We have noted above that 

 M decreases when A 8 increases, A h remaining constant ; so that M will 

 increase as A 8 decreases. We may then substitute in the hypothetical 

 data just dealt with, and which yielded the value M = 58.99, such values 

 for d at the extreme altitudes which will give the minimum difference 

 between them, the densities being at the same time petrologically possible 

 and the general truth of the law of isostasy being assumed. This has 

 been done, increasing the values for Tibet to (8 = 2.70) and Pamir to 

 (8 = 2.75) and decreasing that for the deep Pacific to (8 = 3.10) . It is 

 thought that no higher or lower values can be reasonably accepted for 

 these respective extreme areas. Calculation from these data (the inter- 

 mediate ones being the same as before) yields the value M = 77.07, which 

 is about half way between Bowie's accepted value and mine. That is to 

 say, we have given great advantage to Bowie's value and we get only half 

 way to the higher figure. One might assume still higher and lower values 

 for the densities of the rocks of the extreme heights and depths, and 

 might finally obtain a value of about 96 kilometers; but it seems to be 

 unnecessary to do this, as it would be of no significance. 



It is, however, significant that the depth of compensation as estimated 

 by Hayford and by Bowie at different dates tends to "decrease succes- 

 sively as the amount of data has increased with the lapse of time, and as 

 successively more powerful analyses of the data have been made." 32 This 

 is shown below in tabular form. 



Depth, 

 Date kilometers Reference 



11K)6 141.5 Hayford, Proceedings of the Washington Acad- 

 emy of Sciences, volume 8, page 31. 



1901) (April) 113.0 Hayford, Figure of the Earth, page 146. 



1909 (December) 122.2 Hayford. Supplementary Investigation, page 54. 



1912 113.7 Hayford and Bowie, Special Publication Num- 

 ber 10. pages 10 and 103. 



1921 J __ I Bowie, Special Publication Number 40, page 5. 



} 60 j 



•"•- 1 am indebted to Dr. Hayford for the suggestion of laying emphasis on this trend of 

 the successive estimates. 



