SECT. 1] SUB-OCEANIC STRUCTURAL EXPLORATION BY SEISMIC SURFACE WAVES 125 



of the Pacific and parts of the Atlantic Ocean. They also found mean sediment 

 thicknesses in the Pacific ranging from 0.4 km north of Hawaii to 1.1 km south- 

 west of Hawaii. These values are calculated by subtracting the mean water 

 depth from the liquid layer thickness obtained by comparison of experimental 

 data with theoretical models for different thicknesses of this layer. The values 

 generally agree well with those obtained from refraction shooting. Oliver, 

 Ewing and Press also pointed out that their data suggest somewhat greater 

 depths for the mantle in the western Pacific than for the eastern Pacific. Along 

 certain paths (for example, those from Hawaii to the Philippines, Formosa and 

 Chile) the data, by virtue of slightly lower group velocities for waves of period 

 greater than about 20 sec, suggest structures differing somewhat from that 

 t3rpical of ocean basins. These differences might lie in somewhat lower crustal 

 or mantle velocities, or in greater crustal thickness, or a combination of these. 



Berckhemer (1956) showed the close coincidence between Rayleigh-wave 

 dispersion data for paths across basins of the eastern Atlantic Ocean with dis- 

 persion observed by Oliver, Ewing and Press for basins of the western Atlantic, 

 clearly establishing the essential similarity of crustal structure on both sides 

 of the Mid-Atlantic Ridge. Bath (1959) presented group-velocity dispersion 

 data for Love and Rayleigh waves for a wide variety of azimuths from Uppsala. 

 With one exception, data for the ocean basins, though distinctly oceanic in 

 character, could be obtained only by applying large corrections for continental 

 portions of the paths. However, very favorable paths across the Arctic basin 

 yielded well -determined dispersion curves clearly showing oceanic crustal 

 structure. 



Although surface waves suggest anomalies in such areas as the Easter Island 

 Rise and Polynesia, the contrast between dispersion observed over all oceanic 

 paths with that for continents overshadows the relatively small differences 

 cited above, and is evidence for the relative uniformitv of ocean structure. 



Regardless of the crustal model adopted, all theoretical Rayleigh-mode 

 dispersion curves have a minimum group velocity less than 1 km/sec in the 12- 

 14 sec period range (see Fig. 13). This minimum is equivalent to the one which 

 gives rise to the prominent Airy-phase arrival observed in some explosions in 

 shallow water, suggesting that the corresponding Airy phase might be observed 

 when the source is a natural earthquake. No clear-cut identification of such 

 Airy phases has been made in the case of earthquake sources, however, and 

 recent indications are that the reason for the absence of the waves corresponding 

 to the Airy phase and to the shorter-period branch of the Rayleigh-mode dis- 

 persion curve is the depth of natural earthquakes. 



However, waves corresponding to the short-period inversely dispersed 

 branch of the Rayleigh-wave curve and the minimum group velocity have been 

 observed from artificial sources. Oliver and Ewing (1958) discussed such 

 arrivals at Pasadena and Berkeley, and they associated these with the atomic 

 explosion WIGWAM detonated deep in the Pacific Ocean off lower California. 

 Similar waves recorded at Honolulu were generated by the high-altitude 

 explosions TEAK and ORANGE (see Pomeroy and Oliver, 1960), further 



