12 



TEMPERATURE OF THE SURFACE WATERS OF THE ATLANTIC OCEAN 



thus smaller than in the case of B' at 40° S. Unfortu- 

 nately, the material on current displacements in the 

 Dutch atlases for two-degree fields is not adequate to 

 show in a similar manner the structure of the currents 

 in the middle area of the ocean. Even here a similar 

 change of direction in the vicinity of the boundary can 

 be revealed in some places from the arrows indicating 

 current directions for the two-degree fields in the Dutch 

 publications; for example, the changes occurring in 

 February for the area from 2° E. to 14° W. and 45° S. 



^Figure 12. — Diagram of a conver- 

 gence-divergence line according to 

 Bjerknes. 



to 35° S. (fig. 13) such as those that can be recognized 

 south of Africa in figures 9 through 11. 



In the west, off the Brazilian coast, the situation 

 differs. The Falkland and the Brazil Currents meet, 

 causing the space between the subtropical front and 

 the boundary to narrow to such an extent that 

 they practically coincide. The dynamic influences are, 

 therefore, stronger than the climatic ones. However, 

 because of lack of enough observations for all the months 

 of the year, the presentation of an exact picture of the 

 front is impossible. The portrayal of this area in 

 figure 4 has, then, necessarily been generalized. As 

 examples of the formation of the front in the individual 

 months, the isotherms in figures 14 and 15 are recorded 

 with the boundaries, and the currents with the fronts 

 of December and February. Despite the complicated 

 aspect of these individual cases, the further result at this 

 point is that from the Brazilian coast seaward, the 

 thermal boundary and the dynamic front are at first 

 almost identical. A separation does not take place 

 until they reach 40° W., where the water masses of the 

 Brazil Current, turning off eastward, interrupt the 

 subtropical convergence. The subtropical boundary 

 remains essentially south of these stream lines, and 

 the sub-tropical convergence north of them. In ex- 

 amining figures 14 and 15, it should be noticed that 

 the isotherms were drawn from one-degree fields. The 

 current pattern is thus more generalized than the tem- 

 perature distribution, and this generahzation explains 



the fact that the two diagrams do not coincide perfectly. 

 Finally, in this study of the thermal gradients in the 

 area between the polar front and the subtropical bound- 

 ary, which is chiefly controlled by the west wind drift, 

 more fragments of a further boundary at about 45° S. 

 were found. This boundary, which seems at least in 

 places to separate these zones — perhaps most appro- 

 priately called temperate — into a north and a south 

 zone, is shown in figure 4 as A' and can be recognized 

 in the area of the Falkland Current and south of Africa. 

 In addition to the convergence-divergence line (cf. figs. 

 14 and 15), another cause of its formation can be as- 

 sumed to be a turning of the current from a north-south 

 to an east- west direction. This fragment of a further 

 boundary and front within the west wind drift also 

 becomes visible on the monthly charts of the tempera- 

 ture by means of an increased assembling of the iso- 

 therms and is expressed, as is the subtropical boundary, 

 in Profile Vd, which was traversed by the Meteor, as 

 well as in surface temperatures registered at the time. 

 In figures 7 and 8 at A/ between Stations 133 and 134, 

 it can be seen that the temperature suddenly jumps 

 from 6° to 13.5° at 45° S. lat., and at B' between 



Figure 13. — Surface current in the 

 middle South Atlantic Ocean in 

 February. Designation of the fronts 

 is the same as in Figure 4. 



Stations 134 and 135 from about 14° to more than 21° 

 C. at41°S. lat. 



The seasonal fluctuations of the subtropical boundary 

 are slight, similar to those of the antarctic convergence. 

 They lie within the limits of error of the research 

 methods set up on the basis value of the one-degree 

 field average. Table 4 gives the average monthly 

 position of the subtropical boundary and shows 

 that the average fluctuations are not greater than -f 1 

 degree of latitude. Even the seasonal displacements 

 of the convergence, which can hardly be followed other- 

 wise than according to the two-degree field values of 



Table 5. — Position of the southern subtropical convergence according to current displacements in degrees of latitude 



