THE WATER MASSES 177 



wards to the subtropical convergence, a distance of about 600 miles, this is a relatively shallow layer. 

 Under a well-mixed surface-layer, which is rather poorly saline, there is a pronounced subsurface 

 salinity maximum, and below this again the water becomes less saline. 



It seems fairly certain (Deacon, 1936) that this subsurface salinity maximum represents the core 

 of a southward flow within the subantarctic zone. The properties of this water suggest that it is 

 replenished from two sources, from a subsurface highly saline current within the subtropical zone 

 itself, and (to a lesser extent) from subantarctic water sinking at the subtropical convergence and 

 returning to the south. The vertical T-S curve below the subtropical convergence suggests that the 

 central water there is a mixture between water of higher salinity than the surface-water and the 

 antarctic intermediate water (see Fig. 31), showing that it is probably the subsurface current which is 

 involved in the mixture in this region. The subsurface salinity maximum is not generally distributed 

 over the ocean, however, and no doubt elsewhere it is the surface-water which mixes with the inter- 

 mediate water as occurs in the subtropical region itself. 



Above the salinity maximum where present, and generally over the subtropical zone, the surface 

 water is a fairly homogeneous layer, and its movements, governed principally by the trade winds, 

 follow an anti-cyclonic pattern. From the Brazil current flowing southwards down the western side of 

 the South Atlantic, the water is carried to the east with a decreasing southerly component. It appears 

 to be the balanced effect of this movement with the northward drift of the subantarctic surface-water 

 which is responsible for the maintenance of a sharp subtropical convergence. Deacon has already 

 pointed out that there is not much likelihood of this water sinking at the subtropical convergence, and 

 it is probably mostly carried east in a direction more or less parallel to the convergence. As it reaches 

 the eastern side of the ocean, it turns more to the north, and, with perhaps a contribution from the 

 Agulhas current (Dietrich, 1935 a, b), it returns up the eastern side of the Atlantic as the South-east 

 trade wind drift. Across the Atlantic, and forming a line approximately from Angola to Rio, about 

 the 23 C. surface isotherm, there is a convergence of the surface-water, and to the north of this line 

 there is a warm highly saline layer which is very poor in nutrients and sharply divided from the under- 

 lying water by a strong discontinuity layer. This is the tropical surface-water, which owes its existence 

 to the immense amount of heating and evaporation in this region. From the African side the south 

 equatorial current carries this water across the Atlantic to Brazil, somewhat in the form of a left-hand 

 screw. Within itself the layer is homogeneous and well mixed. 



Temperature-salinity relationships of the South-west African waters 

 The typical T-S curve for the offshore stations of the 'William Scoresby ' is very similar to the general 

 pattern described for the South Atlantic. 



None of the 'William Scoresby' stations was in deep enough waters to encounter the Antarctic 

 bottom water which the ' Meteor ' observations showed to fill the depths of the Cape Basin, but at all 

 stations of sufficient depth off the continental slope the North Atlantic deep water was encountered. 

 At WS 976 only, a maximum salinity of 34-91 % at 2420 m. was recorded. This represented the nearest 

 observation to the core of the upper deep water, but at all other stations salinity values around 

 34-80 % , and still increasing at the lowest depths of observation, indicated that the central core of the 

 upper deep water had not been reached. 



Sometimes between the southward flow of the North Atlantic deep water and the northward flow 

 of antarctic intermediate water, a slight temperature inversion is produced, such as was recorded at 

 station WS 996 where the water at 1470 m. with a temperature of 3-05° C. was 0-04° C. warmer than 

 the water at 11 70 m. Deacon has recognized this layer of minimum temperature, but considers that 



