314 AisnsruAL report smithsoman institution, 1950 



Dr. Benioff (1951b) has found, in addition, that the accumulated 

 strain-rebound characteristic of all world earthquakes of magnitudes 

 8.0 to 8.6 that have occurred since 1904 exhibits a saw-tooth shape 

 with very nearly linear segments. The serrations decrease regularly 

 with time in amplitude as well as in period. If a proposed interpre- 

 tation is correct, the curve indicates the following conclusions: (1) 

 World earthquakes in this magnitude range are not independent 

 events. They are related in some form of world-wide stress system. 

 (2) From 1908 to 1950 the total secular strain accumulated at a re- 

 markably constant rate. (3) This strain was released in five active 

 periods of decreasing lengths separated by quiescent intervals of very 

 small or no activity. (4) During the active periods the strain release 

 proceeded at approximately twice the rate of the secular strain 

 accumulation. 



Still another type of approach could start with stresses to be ex- 

 pected from theoretically derived forces — such as those connected 

 with contraction or expansion of the earth, or subcrustal currents — 

 and their effect on geological structures. All ideas dealing with 

 forces and structures are by far too controversial to permit the draw- 

 ing of conclusions concerning expected seismicity, but data obtained 

 from earthquakes and artificial explosions are at present the most 

 reliable basis for hypotheses connected with tectonic and structural 

 problems. The distribution of earthquakes and the other observed 

 phenomena in the Pacific arc structures leave little doubt that there 

 is a great difference between the structure of the Pacific crustal layers 

 and the structure of crustal layers in the continents and in the Atlantic. 



A change in surface structure occurs at some point off all Pacific 

 coasts. In the western and southwestern Pacific the boundary is 

 given by the so-called Marshall line or andesite line, which separates 

 the more andesitic material on the continental side from the less 

 andesitic on the Pacific side. This line is known to run to the east 

 of the Japanese, Marianas, and Philippine Islands and crosses the 

 Caroline Islands leaving Yap and Palau on the continental side. It 

 turns sharply to the east near the northwestern end of New Guinea 

 and later passes between Samoa (which is on the Pacific side) and 

 the Tonga Islands (on the continental side). Near Samoa it turns 

 southward and remains to the east of the Kermadec Islands and of 

 New Zealand. Its location in the eastern Pacific is not known, since 

 no islands can be used there for locating the line, but it appears to 

 follow along the coast of North America at a distance which varies 

 from place to place. The andesite line is the intersection of a deep- 

 going surface of discontinuity with the surface of the earth. The 

 difference in structure on its two sides provides one of the reasons 

 for the accumulation of earthquakes along the line. The fact that in 

 many areas a belt of large negative gravity anomalies parallels the 



