120 



OIJVER AND nORMAN 



[chap. 8 



a shear-velocity distribution for tlie continental case (curve A, based on un- 

 ])u])Iislied data of Lehmann) that the region of low shear velocity extends to 

 much shallo^^•er depths beneath the oceans and also that somewhat lower shear 

 velocities prevail down to depths of about 400 km. The sensitivity of this 

 method appears to be such that estimates of shear velocity are accurate to 

 about 0.1 km/sec, providing the assumed density distribution and crustal 

 layering are correct. 



0" 



100 



200 



.3.0 



4.0 



5.0 3.0 



4.0 



5.0 3.0 



4.0 



5.0 



E 



■S.300 



Q 



400 



500 



Shear velocity (km/sec) — > 



Fig. 7. Oceanic and continental shear -velocity distributions for the crust and mantle. 

 Curve A, based on data of Lehmann derived from body-wave studies, gives a com- 

 puted dispersion curve which agrees well with continental crust-mantle Rayleigh- 

 wave dispersion data. Curve B gives a computed group -velocity curve, sho-WTi in 

 Fig. 6, which fits oceanic crust-mantle Rayleigh-wave dispersion data. Curve B was 

 chosen experimentally as the minimum departure from curve A which is consistent 

 with observed oceanic Rayleigh-wave data. Curve C was chosen by Landisman, Sato 

 and Ewing (1958) as the best experimental fit to oceanic Love-wave dispersion data. 

 The oceanic water and mud layers used in computation are not represented in curves 

 B and C. Curves B and C indicate that relatively low shear velocities are found at 

 very shallow depths in the sub-oceanic mantle as compared with the continental case 

 represented by curve A. 



Takeuchi, Press and Kobayashi (1959) performed computations on mantle 

 Rayleigh-wave dispersion based on two models of the mantle using a varia- 

 tional method. This did not differentiate between continental and oceanic 

 mantle but explained the general features of the mantle Rayleigh-wave 

 dispersion curve beyond 100 sec period, including the effect of the world-wide 

 existence of the low-velocity region of the upper mantle. Although the accuracy 

 of those calculations was questioned in a subsequent paper (Press and Takeuchi, 

 1960), the derived structural features still appear to be correct. 



