Vol. 6] Beid: The Elastic-Rebound Theory of Earthquakes. 435 



instances, be as great as four miles per second, it is quite clear 

 that only the most accurate time observations could serve to 

 determine the starting point of the rupture ; and that the 

 majority of observations would not be accurate enough to show 

 that disturbances did not start simultaneously throughout the 

 megaseismic district. 



At the time of the rupture the rigidity of the rock would not 

 permit very large movements of the two sides of the fault until 

 the fractured surface had greatly increased in size ; but when 

 the large movements came they would cause the severest part of 

 the shock. The friction at the fault would make these move- 

 ments irregular, so that the vibrations sent out would not be a 

 steady, strong series, but would vary so much in intensity that 

 they would produce the effect of strong shocks separated by 

 weaker intervals. At the time of the California earthquake, the 

 severest part of the disturbance did not come until thirty seconds 

 after the beginning of the fairly strong shocks; and it was felt 

 from thirty to sixty seconds. The time necessary for the sides 

 of the fault to reach their positions of equilibrium under the 

 elastic forces, free of friction, would have been only a little more 

 than two seconds. The duration of the severe shocks at any place 

 was partly due to friction on the fault-surface, partly to the 

 time necessary for the extension of the fracture, and partly to 

 the arrival of shocks from more distant parts of the fault. 



The friction of the two sides of the fault when the dislocation 

 is taking place, and their sudden starting and stopping (the 

 latter due largely to the friction), are the causes of the vibrations 

 which are propagated elastically to a distance ; and they all have 

 their origin in the rupture surface. It has been suggested that 

 the origin of the vibrations may lie in a' volume and not on a 

 surface; and that the sudden folding of the rock or the movement 

 of a block as a whole would cause elastic vibrations to emanate 

 from the whole volume moved. This idea seems erroneous. If 

 the rock were sufficiently plastic to fold very rapidly under the 

 compressive forces it would not be sufficiently elastic to send out 

 vibrations; and if the rock yielded elastically to a suddenly 

 applied force, the vibrations would start from the boimdary 

 where the force must be applied. We shall see that blocks do not 



