Elastic Waves, with Seismological Applications. 75 



runs up to equality with the incident' wave. In the second 

 set of curves this happens at the angle of total reflexion ; 

 for, not only does the A x -wave vanish, but so also does the 

 A'-wave * — which indeed never attains any great significance 

 at the lower incidences. After the critical angle of incidence 

 is passed, however, the energy of the A'-wave soon reaches a 

 maximum, being then of greater value than that of the B x - 

 wave, and gradually falls away to zero, while the energy of 

 the B x -wave as gradually rises to unity. 



In trying in some way to bring these results into corre- 

 spondence with earthquake phenomena, we notice first of all 

 that, if an earthquake is to be regarded as a progressive wave 

 in an elastic solid, the angles of emergence of the waves will 

 generally be small — that is, the angles of incidence large. 

 Hence we need pay but little attention to the state of affairs 

 at the lower incidences. For higher incidences we see that 

 whether the incident wave is condensational or distortional, 

 the energy is reflected either wholly or almost wholly in the 

 distortional wave form. Suppose for example that a disturb- 

 ance begins at some region below the bottom of the sea, 

 say at the point in the figure ; and let us assume that 

 what starts from C is a simple wave of compression — that 

 is a condensational wave. Then to any point P suitably 

 placed, there will come not only a purely condensational but 

 also a distortional wave produced by reflexion from some 

 part of the surface separating the sea and the land. 



It is easy to see, however, that this transformation of con- 

 densational into distortional straining will accompany all 

 similar cases of reflexion at the boundary of two different 

 media, whether the one medium is water or some other sub- 

 stance — air, say, or mud, or rock. Also we may safely assume 

 that during refraction across a boundary separating two 

 media, both being of the category of elastic solids, an incident 

 condensational wave will give rise to a distortional as well as 

 to a condensational refracted wave. In the light of these 

 results, then, it is little wonder that no definite relation has 

 ever been shown to exist between the manner of motion of a 

 particle and the direction of propagation of an earthquake. 



* This seems to be a result as novel as it is curious from a purely 

 theoretical point of view, although it has no special bearing on earth- 

 quake phenomena. [See below, p. 95.] 



