SHAPE AND STRUCTURE OF OCEAN BASINS 25 



concluded that "the low velocity zone in the mantle is a worldwide 

 phenomenon." This equality of G wave velocity led him to suggest 

 "that the composition and distribution of temperature are the 

 same for depths greater than about 50 km under continents and 

 oceans." This conclusion was not based on detailed dispersion 

 calculations and involved debatable assumptions about the relation 

 of G waves to continental Love waves. 



Landisman et al. (1959) presented further cases of the calcu- 

 lations reported by Landisman and Sato (1958). They deduced an 

 oceanic mantle structure by fitting a "typical oceanic Love wave 

 dispersion curve." This was built up from the G wave group 

 velocities determined by Sato (1958) for primarily oceanic world- 

 circling paths, plus the oceanic Love wave data of Oliver et al. 

 (1955), Coulomb (1952), and Wilson (1940) for periods less than 

 60 sec. As a result of the calculations for oceanic areas, it was 

 concluded that (1) "the upper mantle beneath the oceans is 

 different from that under continents" and (2) "under oceans, the 

 region of low shear velocities rises to depths of about 50 km." The 

 oceanic mantle structure resulting from this study is shown in 

 Fig. 9 as Oceans VIII. A fuller account of this work will be pub- 

 lished in the near future. 



Takeuchi et al. (1959) applied variational calculus methods to 

 the problem of dispersion of mantle Rayleigh waves. This com- 

 pared computed phase velocities for the Jeffreys-Bullen and the 

 Gutenberg models of velocity distribution with phase velocity 

 curves obtained by integrating the mantle Rayleigh wave group 

 velocities determined by Ewing and Press (1954a,b). They 

 concluded that the better fit obtained from the Gutenberg model 

 "demonstrates the existence of a low-velocity layer in the upper 

 mantle," and that it must be present under oceans and continents. 

 This work gives no information for comparison of the continental 

 and oceanic mantle, since the original data were specifically 

 limited to a period range in which the ocean-continent contrast 

 produces little effect on Rayleigh wave dispersion. 



Dorman et al. (1960) compared data on Rayleigh wave dispersion 

 over continental paths and oceanic paths with computations on 

 eleven models of velocity distribution. They concluded that "the 



