16 DISCOVERY REPORTS 



southerly set near Adelaide Island and Alexander 1st Island, but a northward movement 

 was observed in Matha Bay and Marguerite Bay (Bongrain, 1914, p. 6). The French 

 ship in 1908-10 and the 'William Scoresby' and 'Discovery II' in 1930-1 each found 

 that south of the Palmer Archipelago icebergs were much more numerous than in the 

 neighbourhood of the South Shetland Islands. This concentration may be partly caused 

 by the north-easterly current from the Bellingshausen Sea turning towards the coast 

 and back to the south-west. 



The sea between Adelaide Island and Charcot Island has generally been found 

 closely packed with ice, and this may be an indication that the surface water is piled up 

 towards the coast, as it would be on the left flank of a current towards the south-west. 

 The density distribution in the surface water, illustrated by Rouch (1913, pi. ii), also 

 points to this conclusion. The evidence as a whole suggests that the movement towards 

 the west is generally restored south of the Palmer Archipelago, but indicates that its 

 strength may vary considerably. 



The drift of the ' Belgica ' when beset in the ice showed that although the water 

 movements in the southern part of the Bellingshausen Sea were very variable, the re- 

 sultant current was towards the west. The 'Belgica' drifted from 71 30' S, 85 15' W 

 to 70 50' S, 102 15' W, between February 1898 and March 1899, but except during 

 the last two months, when she drifted rapidly westwards, she moved backwards and 

 forwards over the same ground. The total length of the drift was 1706 miles, but the 

 resultant movement towards the west only 333 miles (Wordie, 1921). The average 

 rate of the drift towards the west was only 0-9 mile a day, but during the last two months 

 4-5 miles a day. The meteorological observations made during the drift showed there 

 were more westerly winds in winter but more easterly winds in summer. 



The downward slope of the isotherms and isohalines towards the south gives a further 

 indication of the existence of a coastal current towards the west. The temperature and 

 salinity distribution in 8o° W (section 18, Plate XLI) suggests that the northern boun- 

 dary of the current lies in 68-69 S. The position is, however, not very certain, because 

 the observations in the westerly current were made in summer, and those in the 

 easterly current in winter. In winter the boundary is probably farther south: the 

 observations made by the ' Belgica ' (Arctowski and Mill, 1908) show signs of a divergence 

 region south of 70 S ; in summer it is probably farther north. 



The observations made in sections 16 and 17 (Plates XXXVI I-XL) show that be- 

 tween 98 and 105 W a movement of the surface water to the east is found as far south 

 as 70 S. A comparison of the sections with section 18 indicates that the easterly current 

 extends farther south to the west of Peter 1st Island than it does to the east. In the 

 neighbourhood of the island itself (68° 49' S, 90 32' W) the surface water has been 

 found to flow northwards. The 'Pourquoi Pas?' found a northward current carrying a 

 large number of icebergs (Bongrain, 1914, p. 6), and a northerly set was also experienced 

 by the 'Odd I' (Antarctic Pilot, 1930, p. 89). 



The evidence of a greater current towards the west in longitudes east of Peter 1st 

 Island and of a northward movement near the island itself suggests that the westerly 



