5 8 DISCOVERY REPORTS 



occurrence of alternating streams of subtropical and sub-Antarctic water. Brennecke 

 (192 1, p. 49) has noted the existence of such streams between the Brazil and Falkland 

 currents, and they are even better known south of the Agulhas current (see pp. 59, 75). 

 The finding from time to time of highly saline patches of water to the south of the 

 convergence in other localities suggests that they may occur less frequently all round the 

 Southern Ocean. 



In contrast with the Antarctic convergence, whose existence is determined primarily 

 by deep water movements, the subtropical convergence appears to be formed solely as 

 a result of the opposition of the subtropical and sub-Antarctic surface currents. Several 

 more intensive series of observations are still needed before an indisputable picture of 

 the water movements near the convergence can be obtained, but there are sufficient to 

 leave little doubt of the correctness of the scheme which has been outlined. 



The position of the convergence shows a general agreement with the water move- 

 ments as shown by the most recent current charts. Where the charts suggest that sub- 

 tropical currents penetrate to high latitudes the convergence lies far south, and where 

 the sub-Antarctic drift extends into low latitudes it lies far north. The convergence is 

 also found to be sharpest where the current charts show the surface currents to be most 

 directly opposed. 



In the western half of the Atlantic Ocean the convergence lies in the boundary region 

 which the Brazil current forms with the Falkland current to the west and the main body 

 of the West Wind Drift to the south. Between the Brazil and Falkland currents the 

 temperature and salinity differences are remarkably large. Klaehn (191 1, p. 650) shows 

 that temperature differences of as much as io° C. in 20-30 miles are not unusual, and 

 the observations made by Brennecke (1921, pp. 48-51) give an example of a temperature 

 and salinity difference of as much as 7 C. and 1-4 °/ 00 within a distance of 8 miles. The 

 exceptional magnitude of these differences is largely due to the fact that the two currents 

 flow for a long distance side by side in opposite directions ; the temperature differences 

 are also increased by the upwelling of colder water on the right flank of the Falkland 

 current, and the salinity differences by this current being diluted with coastal water. 

 Between the Brazil current and the main body of the West Wind Drift to the south the 

 differences are not so large: according to Klaehn (191 1, p. 651) the temperature dif- 

 ferences are rarely as much as 4-6 C. 



With the help of a chart showing the mean annual temperature of the surface 

 water in the region between Cape Horn and the River Plate, based on a large collection 

 of data from the log books of merchant and other vessels, Klaehn has proved that the 

 Brazil current can on an average be traced as far as 49 S. The terminal region is, how- 

 ever, one of sudden temperature and salinity changes; the current direction alters 

 repeatedly as the wind varies, and streams of highly saline water, partly or wholly 

 subtropical, alternate with streams of cold and poorly saline sub-Antarctic water. 



In such a region there is no fixed boundary between the northward and southward 

 currents. The observations in the Southern Ocean as a whole, and the few that are available 

 from this region, suggest, however, that the water north of the 11-5° C. isotherm in 



