1 .-)<; 



WORZKI, AND HARRISON 



[chap. 9 



being depressed most between 30 and 35°. From 15°S to about 40°S the geoid 

 is raised slightly above the ellipsoid. 



4. The values in the Arctic from 73 (although difficult to analyze because 

 the topography is so poorly known) indicate that the mean anomaly between 

 75° and 90°N is about 0. 



It is considered suggestive that the latitude belt of the greatest geoidal 

 down-warp in the oceans coincides with the latitude of the highest mountain 

 and, hence, the greatest geoidal up-warp on continents. 



Fig. 14. Five-degree average anomalies for "typical" ocean basins compared with latitude. 

 At the top, the percentage of the Earth's surface which is oceanic for each latitude is 

 plotted. (After Worzel and Talwani, 1959.) 



Koryakin (1958) has made a number of sections across the Atlantic Ocean. 

 These sections are based on some of the published submarine gravity measure- 

 ments and some surface-shi]) measurements reported to have been made by 

 Ciajnanov and Koryakin in 1954-55. They estimate that the latter values have 

 a root-mean-square error of + 16-17 mgal. Unfortunately, only one of the five 

 profiles using these values is illustrated, and that on such a small scale that it 

 would be difficult to add these observations to the world data. 



Fig. 15 shows Koryakin's section from Puerto Rico to the shoreline of Africa, 

 near the Canary Islands. His sections are computed on the following assuni])- 

 tions : (1) a water density of 1.03 ; (2) a sediment density of 2.30 ; (3) a crustal 

 density of 2.8 ; (4) a mantle density of 3.3 ; (5) Helmert's 1909 formula of world 

 gravity ; and (6) the mantle interface in the Nares Basin is 10 km below sea- 

 level. As a section generalized from limited data, it gives a good first-order 

 approximation of the crustal section across the ocean provided that one is 



