154 MARION EXPEDITION TO DAVIS STRAIT AND BAFFIN BAY 



^yinte^, drifts south faster in sprin*; tlian do the bergs, hence the 

 latter arrive as the fiehls and floes are disappearing. 



Upon approaching the Tail of the Grand Bank the bergs either 

 wheel abru})tly to the eastward, to drift northward along the north- 

 ern edge of the Gulf Stream, or they follow around to the westward. 

 If the bergs ground on the southern part of the bank, as many do, 

 they usually disintegrate there, and thus do not menace naviga- 

 tion farther south. ^-^ Some bergs, however, which drift westward 

 around the Tail, continue northward beyond the forty-third parallel, 

 where they are usually caught in the large counterclockwise eddy 

 which characterizes the water mass of this region. ^^ (See fig. 101.) 

 Still other bergs of this group may be carried directly north- 

 westward, where they eventually ground well in on the Grand 

 Bank, even as far as 1^0 miles from the Tail, and survive for several 

 weeks before Anally disintegrating. Occasionally a berg is even 

 carried as far west as longitude 54^ or 55^ W. in this way. Shortness 

 of drift, together with a slow rate of travel, probably reflecting the 

 weakness of the northern current, are the outstanding features of the 

 drifts of early April bergs. 



Bergs drifting southward toward the Tail attain maxinuun abun- 

 dance during May, reaching a normal total of 130 south of Newfound- 

 land. At this, the height of the season, they tend to drift along paths 

 h and i (fig. 97), provided the cold current is normally developed. 

 The prevailing drift is thence southwesterly to a region bounded by 

 the forty-second and forty-third parallels and the fifty-first and fifty- 

 second meridians, in which vicinity most of the bergs of May swing 

 sharply to the eastward, paralleling the northern edge of the Gulf 

 Stream. As a modification of this type, the ice may set more to 

 the southward and even southwest of the Tail (see figs. 105 and 106), 

 but in most cases it does not drift dee])ly into the Gulf Stream before 

 being borne olf eastward, and finally northeast, where it rapidly 

 melts. However, in a year when arctic influences are weak, or the 

 Gulf Stream abnormally strong, bergs do not reach the Tail but are 

 deflected eastward along j^aths d or (\ Figure 97. (See also figs 100, 

 101, 103, and 104.) 



June witnesses a decline from the season's maximum iceberg crop. 

 The normal number is 68 bergs south of Newfoundland. The first 

 Aveek of this month may not show any apparent slacking in the 

 stream of bergs, but by the latter part of June, as a rule, the drift of 

 the ice past the Tail of the Grand Bank noticeably diminishes, due 

 to any one or all of the following causes, the relative importance of 

 each still remaining to be determined : 



{a) The encroachment of warm oceanic water northward in the 

 surface layers toward the Tail of the Grand Bank. 



(5) The westAvard encroachment of the inner side of the Gulf 

 Stream toward the eastern slope of the Grand Bank. 



(c) The withdrawal of the pack ice along the coasts and shelves to 

 the northward allowing the bergs to work inshore and strand. 



''^ During the ice season of 1022 we saw several bergs float again after several days 

 delay in on the southwest slope of the bank, so to resume their journey southward. 

 (Sm'ith, 192.3, p. 58.) 



''* In 1021 (see Smith, 1922. Chart H) the large elliptical track taken by :in iceberg off 

 the western slope of the Grand Bank gave the first inkling that such phenomena prevailed 

 in the current system. But several times since then (see Smith, 1927, p. 8(5, and 1927b, 

 p. 70) the dynamic topographic maps of the sea surface have proved conclusively the 

 existence of such a cyclonic depression. 



