128 



SCIENCE 



[N. S. Vol. XXIX. No. 734 



ially different from what it was before the 

 earthquake. In places, and especially 

 near the fault, the general stress is less; 

 in other places it is greater. The region 

 of maximum stress is ordinarily shifted, 

 so that when stresses imposed by external 

 force again overtax the resistance, the new 

 point of yielding is at some distance from 

 the last. 



In view of the complexity of the condi- 

 tions and the intricacy of the interaction 

 among strains, it is not to be supposed that 

 the status at any one epoch will ever be 

 exactly repeated. Nevertheless, its main 

 features may recur, and whenever they 

 do a cycle will have been completed. 

 Such a cycle, however, would be indefi- 

 nitely long, and would be too difficult of 

 discovery to be available for purposes of 

 forecast. 



It is conceivable also that in some limited 

 portions of the general district the local 

 'Conditions may give rise to repetitive col- 

 'lapses somewhat independent of the gen- 

 -eral progress of events. In such case the 

 successive earthquakes would originate in 

 the same place and their systematic char- 

 acter could be recognized through that 

 fact. 



There is a class of natural and artificial 

 rhythms in which energy gradually passes 

 into the potential form as internal stress 

 and strain and is thus stored until a re- 

 sistance of fixed amount is overcome, when 

 a catastrophic discharge of energy takes 

 place. The supply of energy being con- 

 tinuous and uniform, the discharges recur 

 with regular intervals. The frictional 

 machine for generating electric sparks in 

 the laboratory is the type; other examples 

 are the geyser, water gurgling from a 

 bottle, and the alternate adhesion and re- 

 lease of the violin bow in contact with the 

 string. The earthquake is a repetitive 

 catastrophe belonging to the same mechan- 

 ical group, and if its mechanism were as 



simple as that of the electric machine its 

 rhythm would be as perfect. If the 

 stresses of an earthquake district affected 

 only homogeneous rock and were always 

 relieved by slipping on the same fault 

 plane, the cycle of events would be regu- 

 lar; but with complexity of structure and 

 multiplicity of alternative points of col- 

 lapse, all superficial indication of rhythm 

 is lost. If rhythmic order shall ever be 

 found in the apparent confusion, it wiU be 

 through an analysis which takes account 

 of the points of origin of all important 

 shocks. 



Alternation.— 'ThQ principle of alterna- 

 tion in the occurrence of earthquakes has 

 already been touched. When a large 

 amount of stored energy has been dis- 

 charged in the production of a great 

 earthquake and its after-shocks, it would 

 seem theoretically that the next great 

 seismic event in the same seismic district 

 was more likely to occur at some other 

 place, and that successive great events 

 would be distributed with a sort of alter- 

 nation through the district. This hypoth- 

 esis I used twenty-five years ago, in pre- 

 dicting that the next slip on the fault 

 at the base of the Wasatch range, instead 

 of occurring in the locality of the last 

 previous slip, would take place at a differ- 

 ent point ;^ and it has been more recently 

 applied by Omori, Hayes and Lawson in 

 forecasting earthquakes on the western 

 coast of the two Americas. These geog- 

 raphers agTce in regarding the entire 

 coast either as a single seismic district or 

 as a portion of a greater district, in which 

 there is interdependence of parts. Omori 

 pointed out that in the period of six years 

 from 1899 to 1905 there were extensive 

 disturbances in Alaska, Mexico, Central 

 America, Colombia and Ecuador; stated 

 that the gap thus left between Alaska and 

 Mexico had led him to anticipate an early 



'Monograph I., U. S. Geological Survey, p. 362. 



