SECT. 1] GRAVITY AT SEA 163 



of the M-discontinuity at about 230 km, the computed curve is not steep 

 enough to fit the observed curve. This indicates that most, or all, of the 

 anomalous mass is at this depth or shallower. 



Fig. 21b shows an alternative mass distribution where most of the anomaly 

 is caused by a thickening of the topmost layer. This is the 5.0 km/sec layer 

 found in the seismic stations on either flank. A good reversed seismic station 

 near the peak could establish whether this type of distribution is most likely. 



Fig. 21c shows a third alternative mass distribution where most of the 

 anomaly is caused by low density in the lower part of the crust beneath the 

 central part of the ridge. Seismic refraction measurements probably could not 

 delineate this structure as the low-density material could be expected to have 

 lower velocities and would thus refract the seismic waves downward rather 

 than upward. Perhaps earthquake surface-wave studies could detect this type 

 of mass distribution. This type of section shows much greater promise of 

 explaining the magnetic anomalies often associated with the rift valley of the 

 ridge. 



Combinations of the mass distributions shown in Fig. 21 can also be found to 

 explain the gravity anomalies over the ridge. It would be difficult to explain 

 the major part of the anomaly, however, by masses deeper than 25 km. 



C. Islands and Seamounts 



Heiskanen and Vening Meinesz {loc. cit) have summarized their results of 

 isostatic studies for volcanic islands in Table II. These are carried out for 

 densities of the island materials of 2.937 and 3.07, and they give comparisons 

 for a value on the island and for a deep-sea station clear of the island influence. 

 The computations are carried out for various degrees of regionality indicated 

 by radius R. They conclude that the best results (closest to zero anomaly) are 

 obtained for a regionality of 232.4 km and for a density between the two ex- 

 tremes for an Airy isostatic assumption, T = 30. They further conclude, on 

 the basis of estimates of the mechanical properties of the crust and the mantle, 

 that the rigid crust is about 35 km thick. In this concept the upper part of 

 the mantle, although showing no density contrast with the mantle beneath 

 it, has different elastic properties, the upper part resisting viscous flow, while 

 the lower part is capable of viscous flow. 



Woollard (1954) has computed a density distribution to account for the 

 gravity anomaly of the islands of Hawaii and Bermuda. Fig. 22 has been 

 adapted from this study. He finds that a density of 2.30 for the Hawaiian Island 

 mass above the sea floor gives the smoothest Bouguer anomaly and accounts 

 for a regional Bouguer anomaly by a gentle depression of the M-discontinuity. 

 For Bermuda he finds the density 2.7 for the island mass gives the smoothest 

 Bouguer anomaly curve. The study by Harrison and Brisbin (1959) of Sea- 

 mount Jasper (see Fig. 11) indicates a density of 2.3 assuming no underlying 

 density contrasts. 



Worzel and Talwani in a pajDer in preparation have made computations for 



