170 MR. R. D. OLDHAM ON THE PROPAGATION OF 



wave motion is something very different to any which has yet l)een investigated, so 

 far as I know, and requiring some form of wave motion whose rate of propagation 

 decreases with an increase of the modulus of elasticity. 



Even if some such explanation as has just been suggested were possible, it does not 

 appear to be the true one. The value of 3 '3 kiloms. per second given by Professor 

 MILNE for an arc of 110 depends solely on the Argentine earthquake, and, as shown 

 by the times tabulated above, an even higher rate might have been adopted. The 

 values for 60 and 80 also are lower than might have been adopted if the time of com- 

 mencement was referred to, while the rate for 20, though lower than that given by 

 the Turkestan earthquake, may represent closely the average rate for that distance. 



In interpreting the data, however, it is necessary to remember the circumstances in 

 which they were obtained. All the data available as yet are from observatories 

 situated in Europe, and consequently we have not observations of the same 

 earthquake at varying distances from the origin, but observations of different 

 earthquakes whose origins were at various distances from the group of observatories 

 at which they were recorded. 



Now the waves of the third phase, whatever the nature of the molecular movement 

 to which they are due, travel along the surface as distinct undulations with a motion, 

 to use Professor MILNE'S simile,* " not unlike the swell upon an ocean." Such being 

 the case, it is not improbable that their rate of propagation may be dependent on 

 their size, as in the case of sea waves ; and this is the more to be expected if, as seems 

 probable, their propagation is partly gravitational. 



If this be the case, the instruments would only be affected by earthquakes 

 originating at very great distances if they were of very great magnitude and capable 

 of setting up the largest waves, which would not only travel furthest, but at the 

 greatest speed. In the case of earthquakes originating at more moderate distances, a 

 certain proportion would be of lesser magnitude, setting up surface waves of lesser 

 size, and travelling at lesser velocities, by which the average of the apparent rates of 

 propagation would be reduced. Close to the origin, moreover, we come into the 

 region where earthquakes are sensible, and the very low rates of propagation recorded 

 in some cases would further lower the average. 



The interpretation which I put on the records is, therefore, that the third phase 

 corresponds to the arrival of a form of wave motion which is propagated round the 

 surface and not through the interior of the earth ; that the rate of propagation in the 

 case of each individual earthquake is practically constant ; and that the true and 

 apparent velocities of propagation are everywhere the same, but that the rate of 

 propagation varies in the case of different earthquakes, being dependent in some way 

 on the size of the waves set up by it. In the case of the greatest earthquakes, which 

 are recorded at distances of 60 and over, this rate of propagation appears to be 

 practically always about 2 '9 kiloms. per second for the principal and largest waves, 



* 'Seismol. Jour. Japan,' vol. 3, 1894, p. 89. 



