

UPWELLING x8 9 



a schematic system for the upwelling there, although his data were too widely scattered to obtain as 

 detailed a horizontal picture as Gunther did. 



Defant started by considering theoretically what events would occur in a long straight canal over 

 which a wind was blowing. If two bodies of water of different density are superimposed in the canal, 

 and if the canal runs north and south in the southern hemisphere and has a wind blowing over it from 



STATIONS 

 SEA MILES 

 OFFSHORE ' 

 0- 



WSIO50 

 I 



100- 



300- 



5O0- 



Fig. 37. Distribution of the anomaly of specific volume. Section off the mouth of Orange river on survey II. The probable 

 direction of water-movement in a vertical plane, deduced from the shape of the isosteres, is also indicated. Positions of 

 stations are shown in Fig. 2. 



south to north, then the upper water layer would be set in motion. Under the provisions of Ekman's 

 theory of wind drift, and when the thickness of the upper water layer is greater than the depth of 

 frictional resistance, the lower water layer should remain at rest, while the boundary between the two 

 would take up a slope across the canal. This slope would rise to the right (east, if one faces in the 

 direction of flow of the surface-current) and thus the underlying heavier water would accumulate on 

 the right-hand side of the current. At the same time a transverse circulation would form in the upper 

 layer, which would depend on the velocity and direction of the wind. The upper layer circulation will 

 form a left-handed screw motion. This would be strongest if the wind blew from the east, and sup- 

 pressed if it blew from the west. 



Applying these deductions to the South-west African coast we can regard the coast as the canal 

 bank on the east. The west bank is missing, but this does not matter since the wind which is being 

 applied (the south-east trade) is of limited lateral extent. Defant thus deduced a circulation pattern 

 as in his fig. 7. The circulation is complicated by the fact that in the sea there are not two separate 

 well-defined density layers but a general increase of density with depth. The trade wind induces a 

 transverse circulation similar to that in the canal, a horizontal axis being present above which the water 



9-2 



WOODS 

 HOLE 



MAS.q' 



