THE ANTARCTIC ICE-EDGE 291 



Petrel, an abrupt drop in the sea temperature of about i-i*5° C, and a sudden decrease 

 in the swell indicate that the pack is within a few hours steaming. These phenomena are 

 seldom absent and are usually noticed before ice blink is visible. The blink is sometimes 

 not seen until the ice is very close, but it is often a most valuable indication of the local 

 disposition of the pack. These signs of the approach to pack-ice are described by 

 Lieut. L. C. Hill, R.N.R. (commanding officer of the 'Discovery II', 1936-9) in 

 Supplement No. 7 to the Antarctic Pilot. He mentions also the Antarctic Tern, which 

 was only seen at the ice-edge. 



SEASONAL DISTRIBUTION OF THE ICE-EDGE 



GENERAL FEATURES 



The ice extends farthest north in late winter and spring, its edge lying in much the 

 same position in July, August, September and October. During the summer the ice- 

 edge retreats southward and during the winter it advances north again. The range of 

 this advance and retreat varies considerably in different parts of the Antarctic. It is 

 greatest in the Atlantic sector and probably least in the Bellingshausen Sea, though in 

 the vicinity of Adelie Land also there appears to be a comparatively small range. 



The ice generally lies farthest south about February and March. Parts of the con- 

 tinental coast may be found clear of ice as early as January. Thus the ' Discovery II ' on 

 16 January 1938 approached the coast of Adelie Land without seeing pack. But the ice is 

 probably at its minimum at the end of February and beginning of March, and then is 

 perhaps the best chance of finding the coast clear of ice. Except on the Pacific side and 

 perhaps on the west side of the Weddell Sea, it is probable that nearly all parts of the 

 continental coast or fixed barrier are free of ice from time to time in the late summer. 

 In the Pacific sector between about 75° and 140 ° W the ice seems more stagnant and no 

 ship has penetrated it to the land, although several have attempted to do so. At the end 

 of the winter when the ice lies farthest north, its edge in the Scotia Sea region runs from 

 south-west to north-east and reaches to about 55 S to the north of the South Sandwich 

 Group. This latitude is maintained to about 20 E, but beyond that it falls away more 

 and more to the south. East of no°, off Adelie Land and the Ross Sea and in the 

 Pacific sector, it is possible that the ice-edge is always south of 60 ° S except perhaps to 

 the north-east of the Ross Sea, where it tends to reach farther north. 



Plate LXIX gives an indication of the rate at which the ice-edge retreats during the 

 summer, and it will be seen that in the Atlantic sector there is a rapid break up in 

 December. This will be described in more detail below. Plate LXX gives a tentative 

 representation of the advance of the ice-edge in winter, but it is based on scanty material. 



The latitude of the ice-edge on any given date of course varies a good deal from year 

 to year. The extent of this variation is well shown in the charts except in those months in 

 which observations are very scarce. We are not making here any detailed comparison of 

 the ice conditions in different years. The charts speak for themselves in this respect, but 

 it should be remembered that the "spread" of the plottings in any one month may be 



