356 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1962 



all along the coast from Puerto Saavedra to tlie south end of Isla 

 Chiloe (figs. 2 and 3), with the observed intensity a maximum at the 

 coast, is quite in accordance with these factors. 



The fault movement began in the north and progressed southward. 

 This tended to focus the energy of the surface waves southward and 

 to decrease the period of waves in that direction. Prof. Hugo Beni- 

 oif/ of the California Institute of Technology, indicated that the 

 fracturing was definitely of a progressive type and that the period of 

 the surface waves, as recorded in Peru, was much greater for the 

 waves arriving by the short arc than for those that went all the way 

 around the world before arriving. 



Recent work by Aki [18], using Rayleigh waves, indicates that the 

 aftershocks of the earthquake took place on dextral strike-slip faults, 

 if the faults are parallel to the coast. Benioff, Press, and Smith [19] 

 independently show from seismological observations that the faulting 

 was progressive and that the fault slipped over an extent of 1,200 km. 



Aki makes the interesting comment that the faulting in aftershocks 

 near the ends of the zone is predominately dextral strike-slip, and 

 those with vertical movement lie in the central part of the disturbed 

 zone. This speaks clearly of continued tearing at the ends and of ver- 

 tical readjustment of a secondary nature ; perhaps, in part, isostatic in 

 the central part, from which the horizontal sheer forces have been 

 temporarily relieved. 



It is probable that the first eartliquake took place on the Arauco 

 fault near the town of Guranilahue (fig. 1) at a depth too great to per- 

 mit surface breakage. This quake caused a redistribution in the 

 forces in the earth, causing another part of the fault, just offshore, to 

 move. It is more than likely, considering the distribution of after- 

 shocks, that the great earthquake was produced by slippage along 

 about 1,200 Irni. of the fault. 



The Reloncavi fault probably moved next, and subsequently a series 

 of others in the central valley and coast ranges slipped, giving rise to 

 the many aftershocks. The Reloncavi fault zone seems to mark the 

 eastern side of the area of earthquake occurrence during this series of 

 seisms. 



Following further activity on the Reloncavi fault, or a branch 

 thereof, a fracture was opened that was sufficiently large to permit 

 lake or ground water to seep to the level of the hot rocks below, and 

 steam, hot gas, and rock dust escaped through the fracture on the side 

 of Volcan Puyehue (fig. 2). 



The volcanoes may well be caused by the divergence between the 

 various fault zones ; any tendency of the seaward blocks to drift north- 

 ward would reduce the pressure on the sides of the block containing 



' Personal communication to Pierre Saint-Amand, dated 19G0. 



