DESCRIPTION OF THE ISLANDS 349 



rocks such as Ailsa Craig. These rocks tend to become foci of pressure, and against their 

 steep sides the ice piles up in ridges as much as twenty feet in height. 



Before the floes become firmly frozen for the winter there is a period of intensive ice 

 action/ when the still unconsolidated pack, drifting to and fro on the tide or hurled 

 hither and thither by wind and wave, grinds and polishes the rocks and boulders of the 

 shore, as well as any off-lying rocks it may contrive to override in its path (Plate XII, 

 fig. 3). The movement of the ice has been stated to produce occasional scoring or 

 striae in the rocks, resulting from the action of the stones which are sometimes frozen 

 or embedded in the under sides of the floes. With the gradual consolidation of the pack 

 the grinding action of the floes is reduced to a minimum. None at all can take place in 

 the depth of a hard winter, when as a rule the pack lies hard and immovable against the 

 land and is consequently inactive. Yet in certain years, as in the mild winter of 1903, 

 when it never really becomes consolidated and there is at all times considerable move- 

 ment among the floes, there can be little doubt that much ice action must take place all 

 round the coasts throughout the period of their envelopment, even during the coldest 

 months. 



Widespread disruption of the ice, the prelude to its final departure for the summer, is 

 caused by ocean swells, especially by those from the north-west. As we have already 

 seen, however, in mild winters when the pack-ice does not extend very far to the north 

 of the group and there is open water close at hand, circumscribed areas of the enveloping 

 ice-sheet may break up sporadically all round the coasts for as long as they remain 

 beset. Northerly and westerly swells also play a part in the dislocation of the floes, 

 but the effect of swells from the eastward or north-eastward must be negligible, or at 

 any rate very slight, owing to the almost complete absence of winds from that direction. 

 Thus while the break up of the ice to the northward may be ascribed to north-westerly 

 or northerly swells, disruption on the south or leeward side of the group may generally 

 be taken as a sign that the seas are clear of obstruction to the westward. Winds from a 

 southerly quarter tend to drive the disrupting ice northwards away from the group. 

 Throughout the month of December in particular there is a marked tendency to south- 

 westerly winds and these must help largely to carry the already loosened ice northwards 

 to scatter and melt. 



If we disregard such exceptional years as 1908, when no winter pack-ice came at all, 

 and the occasional advent of broad streams of ice in summer, it would appear from the 

 available records that pack as a rule may be expected to arrive and subsequently sur- 

 round the group any time between the beginning of March and the end of May, and 

 commence to break up and disperse any time between the beginning of November and 

 the end of January. In 19 10, however, Scotia Bay did not become blocked until July 25, 

 while the winter of 1904 dragged on until the following summer was well advanced 

 before the ice covering the bay finally broke up and dispersed, the actual date of opening 

 being February 6, 1905. According to Mossman an even later date may be assigned 

 to the opening of Scotia Bay following the winter of 1902; for when the ' Scotia' first 



1 See Pirie, J. H. Harvey, 1913, loc. cit., p. 861. 



9-2 



