290 ANNUAL REPORT SMITHSONIAN INSTITUTION, 19 3 9 



surface streetcar, but for greater distances we make up for the time 

 spent in going down to the subway and climbing out at the other end 

 by the higher speed with whicli the subway carries us. At some point 

 not too far distant it would be an even race to go via the surface or via 

 the subway ; beyond that point the subway would be the quicker route. 



Let us suppose that we are equipped with properly designed seis- 

 mographs at each station, from the source outward. The records for 

 the very near stations will show only one arrival — the wave through 

 the surface layer. At a certain minimum distance (which depends on 

 the thickness of the upper stratum and also on the angle with which 

 the ray reaches the surface of discontinuity, the latter depending, 

 in turn, on the ratio of the two velocities F^ and Fj) energy will 

 begin to arrive via the lower path; but it will arrive after the 

 surface-transmitted energy and will appear on the record as a sec- 

 ond impulse. Farther out will be a station at which the energy by 

 the two paths arrives simultaneously. Beyond that point, the first 

 impulse will be that arriving via the lower stratum ; but there will still 

 be a second impulse on the records, due to the energy via the upper 

 stratum. The arrival times of the two classes of impulse may be 

 plotted and will result in defining two intersecting straight lines as 

 shown in the diagram. The slope of these lines gives the velocities 

 Vi and V2. The point of intersection G defines the distance at which 

 the arrivals by the two paths are simultaneous. The two velocities 

 enable us to define the angle with which the ray impinges on the 

 surface of discontinuity. Knowing the distance to the point of inter- 

 section, the two velocities and thence the angle of incidence, we can 

 determine the depth of the upper stratum. 



It is now clear why Abbot's velocities were greater than those of 

 Mallet. Abbot was working at greater distances (8 to 13 miles) and 

 was measuring the velocities, at depth, in a high-speed stratum. 

 Mallet was working at distances of about a mile and was measuring 

 the velocity in the upper, slow-speed stratum. 



Considerable time has been devoted to explaining this seismic 

 method of determining the thickness of a surface layer and the veloc- 

 ities in the upper and lower strata because it is so important and far 

 reaching in its application. The method can be extended to more 

 than two layers — three, four, or even more — yielding the thicknesses 

 of the layers and the velocities in each. It can be used to determine 

 also whether the surfaces of discontinuity are parallel to the surface 

 or tilted and it will give the angles of tilt. It is known as the refrac- 

 tion method and will, henceforward, be referred to by that name. 



In attempting to explore the entire thickness of the crust (about 

 ^ miles) a difficulty presents itself — the explosions we can safely and 

 economically use are inadequate to produce energy sufficient to travel 



