DEEP CURRENTS: ATLANTIC OCEAN 91 



deep current by a stratum of slightly lesser salinity. Its properties, and the horizontal 

 distribution of salinity, temperature and oxygen content in the deep water south of 

 Africa (Figs. 19-21), show that it belongs to an eastward movement of North Atlantic 

 deep water. In the light of these new observations it is clear that the Mozambique 

 section constructed by Moller (1929, figs. 4-5), though based on fewer data south of 

 the equator, points to the same conclusions. Moller does not regard the highly saline 

 water at the southern end of the section as essentially Atlantic water, but realizes that it 

 may be derived partly from such an origin. The existence of an eastward current from 

 the Atlantic Ocean was first suggested by Merz and Wiist (1922, p. 23) from the 

 temperature chart for the depth of 1500 fathoms given by Buchan (1895, map 13) in 

 the Challenger reports. Wiist also notes the existence of the current in a preliminary 

 account of the Meteor results (1926, p. 250) ; these results suggest that the salinity of the 

 deep water increases towards the south in the region south of Africa, and Wiist supposes 

 that the Atlantic water flows eastwards chiefly south of 40 S, outside the region 

 influenced by the Agulhas current. Our observations (in section 8, Plate XIV, and Fig. 

 19) show, however, that the greatest salinities are found very near the land and indicate 

 that the Atlantic current flows eastwards close round the Cape of Good Hope. The 

 level of the maximum salinity is below the deepest observations on which Wiist bases 

 his preliminary assumption that the deep layer in the region of the Agulhas current has 

 a low salinity (Meteor Sts. 135-7, Wiist, Bohnecke, and Meyer, 1932, p. 96). 



There are not sufficient data to show how far the Atlantic water penetrates towards 

 the east across the Indian Ocean; but the temperature, salinity and oxygen distribution 

 in the region south-east of the Cape of Good Hope suggest that a large part of the current 

 bends towards the south between 30 and 40 E. The low oxygen content of the water 

 east of 40 E is probably a reliable indication of the presence of deep water from the 

 north Indian current. A preliminary examination of the observations made by the 

 'Discovery II' along the line from Marion Island to Socotra (referred to above) sup- 

 ports this conclusion. Although the deep water south of 20 S belongs partly to an 

 eastward movement from the Atlantic Ocean it is not safe to assume that the southward 

 movement of the north Indian deep current has come to a sudden termination ; the low 

 oxygen content of the upper stratum of the warm deep layer south of 20 S suggests that 

 the Indian water continues towards the south above the Atlantic water, and the other 

 properties of the layer are in close agreement. 



In the neighbourhood of the Antarctic Continent the isotherms and isohalines bend 

 towards the west, and a tongue of warm highly saline water extends along the shores of 

 the Antarctic Continent as far as the Weddell Sea. Brennecke (1921, pp. 117 and 124), 

 who first noted a rise in temperature of the deep water towards the south in the southern 

 part of the Weddell Sea, regarded it as an indication of a westward movement from the 

 Indian Ocean, where the observations of the Valdivia expedition had already shown that 

 the deep water had a relatively high temperature within a short distance of Enderby 

 Land. In Brennecke's time it was too early to trace the origin of the current farther 

 back than this, but our data suggest that the current is composed partly of North 



