EARTHQUAKE MOTION TO GREAT DISTANCES. 173 



or less well defined change in the character as well as the amount of the displacement. 

 Of these the third phase is the most readily and constantly recorded, the second less 

 so, and the first is the phase most frequently absent. 



2. The disturbance of the first and second phases being recorded by heavy 

 pendula, possessing great inertia, with greater constancy and concordance than by 

 light horizontal j>endula specially designed to detect surface tilting, we may conclude 

 that the motion is principally of a to-and-fro nature, and that the records are due 

 to the inertia of the pendula, rather than to a tilting of the surface. This conclusion 

 has been come to by previous writers in the case of particular shocks. 



3. The times of arrival of the first two phases, when plotted, form a curve of 

 increase of apparent velocity with distance, consistent with the hypothesis that they 

 represent the times of arrival of elastic waves propagated through the earth at rates 

 which increase with the depth below the surface. 



4. The increase of rate of propagation with depth appears to be a constant 

 function of the depth, at any rate as far as the greatest depth reached by the waves 

 which emerge at a distance of 90 of arc from the origin. Beyond this depth, which 

 may be put at about 0'45 of the radius, there are some indications of a rapid increase 

 in the rate of propagation. 



5. The time curves drawn through the times of commencement of the first and 

 second phases, if continued to the origin, give initial rates of propagation in 

 tolerably close agreement with the probable initial rates of propagation of conden- 

 sational and distortional waves in continuous rock. 



6. We may consequently accept the conclusions, that the first phase represents 

 the arrival of condensational waves, and the second phase of the distortional waves, 

 each of which have travelled along brachistochronic paths through the earth. 



7. The disturbance of the third phase differs from that of the first, or second, 

 phase in that the light pendula with photographic registration are even more 

 sensitive to it than the heavy pendula whose freedom of movement is trammelled 

 by the friction of their mechanical record. From this we may conclude, that the 

 record is due not to inertia, but to a tilting of the instrument as a whole ; a 

 conclusion which is borne out by the nature of the record in those instruments which 

 trace the displacements on a surface moving with sufficient rapidity to give an open 

 record. This is the phase of the long surface undulations, resembling the swell of 

 the ocean, whose character has been recognised and acknowledged since 1894. 



8. The apparent rate of propagation of the waves of this phase shows no sign of 

 varying with the distance from the origin, but is constant at all distances, or at 

 most subject to a very slight and slow change. From this it may be concluded that 

 they are propagated as surface undulations and that, in their case, the true and 

 apparent velocities are everywhere identical. 



9. The rate of propagation is not, however, constant in the case of all earthquakes, 

 but the waves set up by the greatest earthquakes travel at a higher speed than 



