300 HAYFORD— THE EARTH FROM [April 24, 



normal. The stresses produced in the body of the earth by these 

 changes of load applied over an area only about 30 miles wide are 

 probably confined almost entirely to the first 100 miles of depth. 

 The magnitude of changes of inclination produced at an observing 

 station on the shore by the changing water load would, therefore, 

 be dependent primarily on the modulus of elasticity of the material 

 below and around the bay to a depth of less than 100 miles. The 

 observations might serve, therefore, to determine a modulus of elas- 

 ticity of the surface portion of the earth rather than of the whole 

 earth. 



Turn now to the third of the physical constants which it was 

 proposed to examine, namely the strength. 



Among the forces which we may consider as furnishing tests of 

 strength are: (i) the forces involved in earthquakes, (2) the weight 

 of continents, and (3) the weight of mountains. 



The forces which produce the more intense earthquakes evi- 

 dently cause stress-difi:erences locally which are beyond the break- 

 ing strength of the material. However from earthquakes we may 

 obtain but little information as to the strength of the earth material 

 because the intensity of the stress-differences cannot be reliably de- 

 termined. We know simply that the intensity exceeds the breaking 

 strength of the material, at the points of rupture. 



It is uncertain how great are the maximum stress-differences 

 produced by the weight of continents. One great difficulty in com- 

 puting these stress-differences arises from the fact that the iso- 

 static compensation of continents, now known to exist, reduces the 

 stress-differences much below what they would otherwise be. Love 

 computed the maximum stress-differences thus reduced as .07 ton 

 per square inch. Darwin computed the greatest stress-difference 

 due to the weight of the continents, without isostatic compensation, 

 as 4 tons per square inch. If each of these computations were based 

 upon assumptions which correspond closely with the facts one should 

 be warranted in drawing the conclusion that the maximum stress-dif- 

 ference caused by the actual continents supported in part by the actual 

 isostastic compensation is between .07 and 4 tons per square inch, and 

 that it is much nearer to the smaller than to the larger value. But a 

 close examination of either of these computations shows that it isbased 



