ANTARCTIC SURFACE WATER 9 



same region. The difference between the surface and the cold strata appears to be 

 increased in the same way at Sts. 1053 and 1054 (section 4, Plate VI), north of South 

 Georgia, but the smaller temperature differences, 1-4 to 2-2° C, indicate that the sub- 

 Antarctic water is mixed to a greater extent with Antarctic water. Large differences 

 between the two strata have also been noted near the convergence in 23 ° W. The 

 evidence of the southward movement at the surface and the factors which are likely to 

 give rise to it will be discussed further in the following section of the report. 



In the regions where there is a current difference between the two strata the Antarctic 

 layer may not become homogeneous in winter; and the difference found between the 

 two strata south of Australia, for example, may persist throughout the winter. There are 

 not, however, sufficient data to show definitely that it does. In the Falkland Sector, 

 where observations have been made in the coldest month — that of August — the layer 

 has been found to be uniform to within at least a short distance of the convergence. 

 Between the Falkland Islands and South Georgia, where a current difference between 

 the two strata is to be most expected, the temperature difference at St. WS 254 (Station 

 List, 1930), was onlyo-n°C, and recent observations during the same month at St. 1390 

 in 22° 15' W show that there the layer was uniform. The existence of a southward 

 movement which will give rise to a temperature difference between the two strata 

 appears, however, to be more likely in winter than in summer (see p. 37), and the 

 absence of strong indications of it in the Falkland Sector is evidence that the great 

 vertical mixing in winter produces a uniform water column in spite of the current 

 difference. 



The contrast between the surface stratum and the rest of the layer seems to be in- 

 creased in certain other localities by the meeting of different Antarctic currents. At 

 Sts. 1029-30 (section 3, Plate IV), just north of the boundary between the Weddell Sea 

 and Bellingshausen Sea currents in the Falkland Sector, the surface water was 1-35 and 

 1-38° C. warmer than the cold stratum. Such a large difference, found as early as the 

 middle of November, cannot be due solely to the warming of the surface stratum of a 

 homogeneous water mass, but must be caused partly by the sinking of the colder 

 heavier Weddell Sea water below the lighter warmer Bellingshausen Sea water. At 

 Sts. 103 1-3, where the Antarctic layer probably contains only Weddell Sea water, the 

 temperature difference between the two strata was only o-6i to 0-50° C. A large differ- 

 ence between the two strata is to be expected wherever the main drift of Antarctic 

 water towards the north-east is joined by a surface current which flows more directly 

 northwards, as in the region east of the Kerguelen-Gaussberg ridge, north-west of the 

 Balleny Islands, and north-east of the Ross Sea (see pp. 31 e£ seq.). 



The nature of the surface stratum in summer is notably influenced by the presence 

 of melting ice. The changes are illustrated by the observations in section 5 (Plates 

 VII, IX) which, south of 6o° S, were all made in narrow lanes through pack-ice. The 

 water in the cold stratum was found to have almost the same properties as it would have 

 in winter. An approximate estimation based on the observations made by Brennecke 

 (1921, pp. 154 et seq.) indicates that in winter the Antarctic layer is uniform down to a 



