348 DISCOVERY REPORTS 



The distance to which the pack may extend beyond the group to the north varies 

 greatly from year to year. In the summer of 1904-5, on December 30, the ' Uruguay' 

 met loose and evidently northwardly dispersing pack in 58'' 40' S, 50° 30' W, while she 

 was still some 240 miles north-west of the islands, and on this somewhat indirect evi- 

 dence Mossman assumes that during the winter of 1904, which was a particularly hard 

 one at the South Orkneys, the northern edge of the pack-ice lay some 200 miles to the 

 north of the group. While there is no direct evidence either for or against this assumption 

 there can be little doubt that the ice-edge in 1904 lay very much farther north than it did 

 the previous winter, when from all accounts there appears to have been open sea within 

 ten miles of Saddle Island throughout the period during which the group was beset. 



In severe winters such as that of 1904, the floes on the whole tend to be firmly packed 

 in the immediate vicinity of the group (see p. 338), but in mild winters such as that of 

 1903, when there is open sea within a few miles of the northern coasts, there is usually 

 much movement among the floes, and pools of open water may appear from time to 

 time in the neighbourhood of the islands for as long as they remain in the grip of the ice. 

 Indeed, during the winter of 1903 the only part of the enveloping ice-sheet which did 

 not break up at one time or another was the fast ice covering the small anchorage at the 

 head of Scotia Bay. Owing to the general northerly movement of the ice the floes tend 

 to become more firmly packed against the southern than against the northern coasts, 

 and in consequence, whether the winter be mild or otherwise, the northern as opposed 

 to the southern floes are always relatively unstable. The difi"erence is most marked, how- 

 ever, in mild winters, owing to the prevalence of north-westerly winds whose resulting 

 swells have little difliculty in penetrating the narrow band of pack by which the northern 

 coasts are beset. J. H. Harvey Pirie in The Voyage of the ' Scotia ' states that during the 

 winter of 1903 in Jessie Bay on the north coast of Laurie Island the "land-floe" never 

 held more than about two miles from the Beach. "As late as July 17th", he remarks, 

 "after a north-west wind and swell which broke up the ice, a south-west wind carried 

 out the pack, and there was nothing but open water in sight ; then the pack-ice would 

 drift back, become frozen together, but sooner or later undergo again the same breaking- 

 up and drifting-out process."^ On the south coast, on the other hand, he states that the 

 greater part of Scotia Bay was immovably packed throughout the winter, the firm ice 

 sometimes holding as far out as Ailsa Craig at the mouth of the bay and round Cape 

 Burn Murdoch. Outside of Ailsa Craig, however, there was almost always a strip of 

 open water, beyond which the steady easterly drift of the bergs and pack-ice could be 

 observed from the anchorage where the ' Scotia' lay. 



In any normal winter pressure ridges may occur in the pack-ice,^ to the south of the 

 group in particular, wherever the progress of the advancing ice is stemmed by outlying 



1 According to Argentine observations Uruguay Cove at the head of Jessie Bay rarely if ever remains 

 firmly frozen for more than a week or two at a time, and then only during exceptional seasons when the ice 

 belt extends for a long way to the north of the group. 



2 See Pirie, J. H. Harvey, 1913, Glaciology of the South Orkneys, Trans. Roy. Soc. Edin., XLix, part iv, 

 p. 861. 



