290 DEACON [CHAP. 12 



transport on the north side of the convergence as well as on the south side, but 

 there is probably a balance between wind transport and density considerations 

 which leads to turbulent southward progress in the sub-Antarctic water. There 

 is some evidence of eddies 40 or 50 miles across. A ship steaming north or south 

 across the convergence often finds a single transition from Antarctic to sub- 

 Antarctic water, but if she is on a slanting course she often finds patches of 

 warm and cold water. Wexler (1959), arguing from two detailed bathythermo- 

 graph sections down to only 200 m, suggests that such cold patches are due to 

 up welling and regards them as evidence of divergence in the boundary region, 

 but further bathythermograph sections discussed by Garner (1958) do not 

 support this view, and simultaneous salinity and deep water observations would 

 no doubt confirm that the observed irregularities are better regarded as minor 

 variations introduced by local meteorological conditions into a picture basically 

 controlled by deep-water movements. In some longitudes the convergence 

 seems generally sharper than others and this seems to occur where a low bottom 

 temperature and a steep slope of the deep isotherms indicate a strong flow of 

 bottom water and deep sharp frontal conditions. The steadiness of the mean 

 position of the convergence seems to be a reliable indication that it is not 

 directly dependent on the surface winds. 



E. The Sub -Antarctic Zone 



North of the Antarctic convergence there is a region of well-mixed water 

 attributable to mixing processes associated with the sinking Antarctic water. 

 The descent of this water can be traced by the weak temperature minimum it 

 enforces below the sub-Antarctic water. 1 In 30°W, the meridian used in Fig. 1, 

 the temperature minimum sinks below 500 m within 100 miles of the con- 

 vergence. At first there is little or no decrease in salinity with depth, but as we 

 go north there is soon evidence of a southward movement of more saline water 

 in the sub-surface layer, and the underlying layer occupied by the descending 

 Antarctic and mixed water is clearly distinguished by its low salinity. This 

 salinity minimum can be traced into the Northern Hemisphere and the north- 

 ward flow that it demonstrates has been called the Antarctic Intermediate 



1 It is here, close to the convergence, that Sir James Clark Ross plotted a circumpolar 

 circle of uniform temperature below which he thought the temperature of the ocean was 

 4°C from top to bottom. Following Dumont Durville, he supposed that sea-water like 

 fresh water had its maximum density at 4°C, and that the ocean depths were filled with 

 such water of maximum density. They believed that the surface water farther south was 

 cooler and lighter than water at 4°C because of excess of outgoing radiation and that 

 farther north it was warmer and lighter because it was warmed by the sun. We know that 

 their observations were faulty because their thermometers, not sufficiently protected 

 against pressure, read 1°C too high for each 1000 m they were lowered, and their ideas 

 were wrong because ocean water does not have a point of maximum density above its 

 freezing point. But the surface temperature of 4°C and a correction of 1°C for each 1000 m 

 produce much the same temperature structure as we find today in the latitudes quoted by 

 Ross, and the agreement suggests that the convergence has not shifted in the past 100 

 years (Deacon, 1954). 



