242 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1916. 



Similarly no adequate computations have been made to determine 

 the maximum stress difference due to the mountains. Darwin com- 

 puted the maximum stress difference produced by two parallel 

 mountain ranges, of density 2.8, rising 13,000 feet above the inter- 

 mediate valley bottom, to be 2.6 tons per square inch. Love, for the 

 same mountain ranges, but with isostatic compensation taken into 

 account, computed the maximum stress difference to be 1.6 tons per 

 square inch. In this case the computation indicates that the isostatic 

 compensation reduced the maximum stress difference to but little 

 more than one-half what it would otherwise be. Here, again, both 

 the computed maximum stress differences have been greatly reduced 

 by substituting hypothetical smoothed-out mountains in the place 

 of the actual, irregular, unsymmetrical mountains. 



To the person who is trying to get a true picture of the present 

 state of stress in the earth, two very import<ant facts are made evi- 

 dent by a comparison of the Love and the Darwin computations. 

 First, the existence of isostatic computation greatly reduces the stress 

 differences which would otherwise be produced by the weight of the 

 continents and mountains. Second, the depth at which the maximum 

 stress difference tends to occur is evidently very much less with iso- 

 static compensation than without it. These two conclusions, based 

 upon the differences between the two computations, are apparently 

 reasonably safe even in spite of the same wild assumptions on which 

 both the computations were based. 



Note that even a little information as to the distribution of densi- 

 ties — a little information about isostatic compensation — profoundly 

 modifies the conclusions as to the state of stress in the earth. It 

 should, therefore, be clear why it was so emphatically stated in an 

 earlier part of this address that information as to the distribution of 

 density in the earth is necessary in order to make safe progress in 

 learning other things about the earth. 



Is the earth competent to withstand without slow yielding the 

 stress differences due to the weight of continents and mountains, 

 the isostatic compensations being considered? From the computa- 

 tions by Darwin and Love, considered in the light of the assumptions 

 made by them to simplify the computations, I estimate that it is 

 probable that the actual mountains and continents with all their 

 irregularities of shape and elevation possibly produce stress differ- 

 ences in some few places as great as 4 tons per square inch, and 

 certainly produce stress differences at many places as great as two- 

 tenths of a ton per square inch. The material would certainly yield 

 slowly under such stress differences especially when they persist 

 continuously over long periods of time and throughout large regions. 

 Four tons per inch is the breaking or rupture load for good granite, 



