292 DISCOVERY REPORTS 



type (d) will have drifted farther from its source of origin than type (c), and type (c) 

 than type (b) and so on. Water carrying unmelted pack-ice in the centre of the Weddell 

 Sea and far south in the Bellingshausen Sea is conceived as being the "youngest" type 

 of water. It was met with only to a small extent in the season 193 1-2 in the Bellings- 

 hausen Sea, since the edge of the ice lay to the northward in the Drake Passage in the 

 spring of that year. The chief value of this classification for the purpose of this report 

 is that it provides a useful means of distinguishing between the various regions in 

 which the catches were taken and it will be shown that to a limited extent each type of 

 water is characterized by a slightly different fauna. 



During the season 193 1-2 few observations were taken after the middle of January 

 upon which to base conclusions as to the changes in the disposition of the four types of 

 water as the season advanced. The surface temperatures at Sts. 824 and 825 (— o-i8 

 and 2-13° C.) seem to indicate that a retreat southwards of Weddell Sea water 

 carrying pack-ice took place between the middle of December (St. 766) and the middle 

 of January (Sts. 824 and 825). There were no deep observations at Sts. 824 and 825, 

 however, by which this can be confirmed. During the season 1932-3 a line of stations 

 was taken in early February across the eastern end of the Drake Passage (Sts. 1115- 

 1120), another in late February between the Falklands and South Georgia (Sts. 1121- 

 1 13 1), and another in March from South Georgia across the Weddell Sea current to the 

 pack-ice edge in 69° 22' S, 9° 37*5' E (Sts. 1137-1153). Fig. 7 shows the position of the 

 isotherms, again calculated from the average temperature of a stratum 100 m. deep, at 

 the eastern end of the Drake Passage in early February and in the Weddell Sea in March 

 as found on these lines of stations. The figure shows a clearly perceptible southward 

 movement of the isotherms in the Drake Passage in February as compared with the 

 conditions in this area in November of the same summer (Sts. 1014-1020 ; Fig. 6). In 

 the Weddell Sea in March (Fig. 7) this southward movement of the isotherms is even 

 more pronounced, compared with conditions at the end of November and the beginning 

 of December (Fig. 6). It is doubtless connected with the break up and retreat south- 

 ward of the pack-ice as the season advances. Mackintosh (1934, p. 130, fig. 46) figured 

 the position of the pack-ice in the South Georgia-South Sandwich area in successive 

 months during the season 1 930-1, and illustrated its retreat southwards from a position 

 near South Georgia in October to the middle of the South Sandwich group in February. 

 Mackintosh's figure shows that the movement of the ice from the vicinity of South 

 Georgia to the south-east was less pronounced in the season 1930-1 than in the season 

 1931-2 with which we are concerned in this report. In early December 193 1 the pack-ice 

 was near St. 767 (Fig. i a), which was worked among drift ice and bergs in about the 

 latitude of the northern end of the South Sandwich group. By the end of January it 

 had retreated southwards to the position of St. 823 (Fig. i b) in the latitude of Elephant 

 and Clarence Islands. This was very far south of its position in the previous January, 

 as shown by Mackintosh, when the line of the ice-edge ran approximately from the 

 South Orkneys to the northern end of the South Sandwich group. In the season 1932-3 

 pack-ice was met with at the southern end of the South Sandwich group at the end of 



