Elastic Waves, with Sei&inological Applications. 11 



vanity of expecting the study ot' surface motions to throw 

 much light on the question of earthquake origin. 



And now let us pass to the discussion of the refracted 

 water-wave. Here a glance at the two sets of curves shows 

 that the incident distortional wave is, at the higher incidences, 

 much more efficient than the condensational wave in creating 

 a progressive disturbance in the water. The angle of refrac- 

 tion can never exceed 42° ; so that even for very high in- 

 cidences the water wave will travel upwards to the surface 

 tolerably directly. Here I think w r e may have the explana- 

 tion of the curious bumpings which have sometimes been 

 felt at sea. These must not be confounded with the so-called 

 tidal waves so frequently the companions of earthquakes, and 

 due almost without a doubt to large displacements of the ocean 

 bottom. What I refer to here are the jerks or shakings 

 (sometimes accompanied by sounds) discussed by Milne 

 in the opening paragraph of chapter ix. of his book on 

 ' Earthquakes/ Sounds of course will be heard if the 

 periodic time of any of the components in the wave-motion is 

 short enough, and if at the same time the intensity is sufficient 

 to give rise to audible sound waves in the air, either directly, 

 or indirectly through the medium of such a solid as a ship. 

 According to Colladon's experiments at the Lake of Geneva, 

 the speed of sound in water at 8°*1 C. is 1435 metres per 

 second. This gives 14*35 metres (or about 8 fathoms) for 

 the wave-length of a wave whose pitch is 100 vibrations per 

 second. A slower vibration will of course give a longer 

 wave-length ; and a quicker a shorter. But enough has been 

 said to show that in such a wave of condensation we have 

 something quite fitted to affect even a large ship as a whole. 



Now all that has been said regarding the transference of 

 vibrations from rock to water will, in a general way, hold 

 true of their transference from rock to air. For all angles 

 of incidence in the rock, the angle at which the refracted ray 

 passes out into the air is very small. Thus, returning to the 

 equation 



m 2 a m 9 



— cosec / v = —f cosec 2 0, 



P P 



and giving m 1 , the wave-modulus in air, the value 1*41 x 10 6 , 

 and p' the value "0013, we find, with the same values as 

 formerly for the rock constants, 



cosec 2 = -00242 cosec 2 6'. 



Hence if d = 90°, 0' = 2° 50' nearly. 



