12 DISCOVERY REPORTS 



This current must to a large extent replace the slope current of the homogeneous sea. 

 An attempt to decide whether the actual piling up of the water on the left of the wind is 

 greater than that indicated by the density distribution has been made by Sverdrup 

 (1933), but he was forced by a lack of information to leave the question unsolved. He 

 points, however, to some observations from the Strait of Florida and the Newfound- 

 land region which indicate that the deep current is completely accounted for by the 

 density distribution. 



In the Southern Ocean the density distribution appears even to exaggerate the 

 strength of the deep current, and a close examination of the data illustrated in the 

 vertical sections across the ocean (Plates I-XLIII) shows that the slope of the layers 

 is partly due to the existence of continuous movements of Antarctic water sinking 

 towards the north, and of warm deep water climbing towards the south. The problem 

 has so far not been investigated quantitatively, but there is little doubt that the eastward 

 movement is not as strong as the slope of the surfaces of equal density would at first 

 suggest. 



These considerations, as far as they affect the Antarctic surface water, show that the 

 west wind will cause the surface water to flow east and north, and the total transport 

 within the limit of the wind's frictional influence, probably about 60-100 m., will be 

 more northerly than the surface current. The movement will be modified where the 

 depth of the sea varies, in the neighbourhood of land and submarine ridges. 



The northward movement of Antarctic water appears to be due partly to another 

 cause— a density gradient maintained by differences of climate. Since the cold Antarctic 

 water is heavier than the warmer surface waters found farther north thermodynamical 

 considerations demand that it should sink and be replaced by a southward movement at 

 the surface, and the northward movement of the Antarctic water in both the surface and 

 bottom layers may be largely the result of such density differences. 



The Antarctic bottom water is the heaviest type of water formed in the Antarctic 

 regions. It starts as highly saline water cooled to freezing-point on the wide continental 

 shelf south-west of the Weddell Sea, and sinking down the continental slope it mixes 

 with the warmer water of the deep layer and flows away towards the north and east as a 

 bottom current. The earliest theories of oceanic circulation maintained that it climbed 

 to the surface in the equatorial regions and was returned southwards as a surface current, 

 but it is now known that the circulation is not so simple. 



Salinity differences between regions of great and little evaporation give rise to other 

 heavy and light waters; the prevailing wind systems govern the movements of the 

 various waters to some extent and form cyclonic and anticyclonic systems in which they 

 upwell or sink, and the movements in the surface, deep, and bottom layers are also de- 

 pendent on the disposition of the land masses and submarine ridges. Such factors ex- 

 plain the modern scheme of circulation devised by Merz and Wiist (1922J partly 

 illustrated in Fig. 1 . The presence, for example, of a shallow submarine ridge across 

 the Davis Strait, and between Greenland and the north of Scotland through Iceland 

 and the Faeroe Islands, largely prevents the entry into the Atlantic Ocean of an Arctic 



