THE EARTHQUAKE PROBLEM HECK 371 



The consideration of these layers brings us to a very important 

 question — the depth of focus or depth beneath the surface at which 

 tlie eartliquake occurs. This is of importance in the case of distant 

 earthquakes, but it is far more important for those occurring near by. 

 Theoretically there should be important differences in the records of 

 two earthquakes at the epicentral distances of, for example, 200 

 kilometers, but with one having a depth of 20 kilometers and the 

 other 40. 



Methods for determining depth of focus are not yet satisfactory, 

 and in order that such determinations may be undertaken, there must 

 be a large number of precise records from near-by stations and time 

 must be accurately measured to the nearest tenth second. The latter 

 requirement is met at very few stations and the former only in parts 

 of Europe and in Japan, though before long the eastern and extreme 



FiGUBB 4. — Paths of preliminary earthquake waves through the upper layers of the 

 earth's crust according to Jeffreys. Most of the energy passes along layers of dis- 

 continuity and the velocities remain practically constant along such a path 



western portions of the United States should meet this requirement. 

 Sieberg * gives for a European earthquake depths ranging from 138 

 to 40 kilometers, according to the method adopted, with 45 to 60 

 as the most probable. Gutenberg arrives at 60 kilometers as about 

 the average depth for moderately severe earthquakes. 



In the case of the Japanese earthquake of September 1, 1923, 

 there were enough records made by strong motion instruments with 

 three components near by so that a simple method, and one inde- 

 pendent of time, could be adopted. The observations ^ were so near 

 the epicenter that it could be assumed that the angle of emergence 

 of the waves gave directly the direction of the focus on the assump- 

 tion of straight-line transmission. The distance of observation points 

 from epicenter ranged from 70 to 132 kilometers. The depths 

 ranged from 35 to 55 kilometers with 48 as the average. This is 



* Sif'berg, Erdliebenlfiiiule. 



^ Mem. Imp. Marino Obs., Kobe, Japan, vol. 1, no. 4, 1924. 



102992—32 25 



