472 



University of California Publications in Geology [Vol. 12 



southward from a loc*us of zero movement near San Francisco, agrees 

 with and explains many facts. In the absence of the outcrop of any 

 vertical fault, functional at that date, we are forced to assume slipping 

 on a fault of low dip which is not known to emerge at the surface. 

 This conclusion and the application of the principle of elastic rebound, 

 enables us to explain practically all the facts without recourse to the 

 notion of distension, which was first suggested by Hayford and Bald- 

 win, tentatively accepted by myself, and finally adopted by Rothpletz 

 and Wood. The distension, as regards at least the causal flow, becomes 

 unreal, and all the phenomena are accounted for by the operation of 

 a persistent northerly subcrustal flow, generating in the overriding 

 crust a state of strain, which is relieved from time to time by rebound 

 slips on faults, both highly inclined and flat, each slip causing an 

 earthquake. The fundamental weakness in the data upon which the 

 discussion is based is the uncertainty as to the immobility of the 

 geodetic base Mocho-Diablo. If it be assumed, with Hayford and 

 Baldwin, that this base was imaffected by the strain creep, then we 

 are forced to recognize a transverse shift of the region supplementary 

 to tlie slip on the San Andreas fault in 1906. If the Mocho-Diablo 

 line upon which the triangulation was based migrated northerly then 

 the apparent transverse shift becomes unreal, and is easily explained 

 by that migration. But if the Mocho-Diablo base participated in the 

 strain creep theii the quantities for the rate of strain creep and 

 northerly rebound on the west side of the San Andreas fault would 

 in several cases be increased, while the quantities for the southerly 

 rebound on the east side of the fault would in most, if not all cases, 

 be diminished. In view of this uncertainty I have sought throughout 

 the i:)aper merely to develop the general consequences of the validity 

 of the elastic rebound theory, rather than to arrive at precise measures 

 of either strain creep or reboimd. 



One gratifying thing about the hypothesis of persistent northerly 

 strain creep as a local manifestation of the principle of elastic 

 rebound is that it is susceptible of verification or disproof. As the 

 work of the U. S. Coast and Geodetic Survey proceeds we will know 

 with exactness the geographical position of many points in the Coast 

 Ranges of California. In the light of the elastic rebound theory of 

 faulting it is no longer permissible to assume that changes in position, 

 such as were discovered after the earthquake of 1906, are due wholly 

 to sudden shifts which occur at the times of earthquakes. The theory 

 teaches that sudden shifts can occur only after the accumulation of 



