Crosby.] 84 [October 4, 



in that direction. But the main point is that, with the change in 

 the character of the medium, the waves suffer refraction, being 

 bent toward the seismic vertical. The effect of this refraction is 

 to increase the angles of emergence and the steepness of the wave- 

 paths ; and the latter, extended backward, will consequently meet 

 at some point below the focus. That is, in calculating the depth 

 of earthquake foci, allowance should be made for the fact that the 

 elasticity of the crust increases with depth ; and if this point has 

 not been considered, the depths now generally accepted are prob- 

 ably too great. 1 



It is more important for our present purpose, however, to ob- 

 serve that the number and extent of the fractures produced by 

 an earthquake are inversely proportional to the elasticity of the 

 rocks which it traverses ; and that • just as certainly as the elas- 

 ticity increases downward, the fractures (joints) must increase 

 upward. Therefore, if, as seems necessary, we regard earthquakes 

 as the principal cause of jointing, the conclusion is inevitable that 

 this structure has its best development near the surface and dies 

 out at great depths. But, as we have already seen, shrinkage- 

 joints, whether due to cooling, consolidation or crystallization are 

 probably mainly superficial. And a moment's reflection will show 

 that the fractures resulting directly from the compression and 



1 In the interest of science, I wish to direct attention to an error of a somewhat 

 serious character in the masterly treatise on the Neapolitan earthquake of 1857, by the 

 late Robert Mallet. On page 262, of the second volume, it is stated, in effect (and the 

 accompanying diagram shows clearly the same thing), that earthquake- waves, on 

 passing from the solid, unbroken and elastic underlying formations to the imperfectly 

 consolidated, broken and inelastic overlying formations, i. e., from rocks in which 

 their velocity is greater to those in which it is less, suffer refraction in such a manner 

 that they are bent away from the perpendicular. But every student of physics must 

 see at a glance that the refraction is necessarily in the opposite direction, or toward the 

 normal to the surface of the media; and it it very strange that so patent an error 

 should have crept into the work of this eminent and painstaking investigator. 



The amount and direction of the refraction of the earth-waves, are, of course, impor- 

 tant elements in determining the depths of seismic foci. The wave-paths, on account 

 of being bent upward, must, when followed backward, if no correction is made for 

 refraction, seem to converge at some point below the focus, as already explained, and 

 thus the depth is made too great. That Mallett has not applied, or at least has not 

 properly applied, the correction for refraction is evident from the fact that he con- 

 ceived tha wave-paths to be depressed instead of elevated by refraction, and therefore 

 as appearing to converge above, and not below, the focus. Hence his determinations 

 . of the depths of earthquake-foci are probably in excess of the truth. 



