28 



HISTORY OF THE OCEANS 



Rayleigh waves, are limited by a lack of uniqueness in the presence 

 of a zone of lower velocities. In addition, all the theoretical models 

 calculated to date have used various approximations to the 

 Jeffreys-Bullen density distribution, shown in Fig. 10, which is 

 based on the Jeffreys-Bullen velocity structure. This velocity 

 distribution, not shown above 400 km in Fig. 9, does not include 

 a low-velocity zone, and for this reason has been found to be 

 incorrect by every surface wave study performed to date, and has 

 since been altered on the basis of body wave data from explosions, 

 supplemented by reinterpretation of earthquake arrivals (Jeffreys, 

 1958). 



New data from various sources require a revision in the density 

 distribution used for the upper mantle. This revision will affect 

 the calculated velocities for Love and Rayleigh waves differently, 

 and is expected to remove the discrepancy. Also, new spherical 

 earth calculations show that sphericity is important, and must be 

 allowed for in these studies. 



Finally, a few words are required about the degree to which the 

 various models reproduce the observations of surface wave 

 dispersion. Despite the scatter of Rayleigh wave observations for 

 periods less than 90 sec, it seems certain that new computations 



2000 

 Depth In km 



4000 



Fig. 10. Density in grams per cubic centimeter vs deptli in kilometers for the 

 mantle (after Bullen, from Heiskanen and X'ening Meinesz, 1958). 



