SCIENTIFIC RESULTS 159 



the forty-fourth and forty-fifth parallel^^. The current surveys of 

 1926 and 1927 (see Smith 1927 and 1927b) ])roved such is the case 

 when the cold current is weak or the warm abnormally voluminous. 

 In some years this plienomenon is more pronounced than in others, 

 but whenever Avell developed it definitely blocks the path of the 

 berjjs from reachin<r the Tail of the bank, and deflects them out 

 southward and to the eastward. An excellent illustration of this 

 is to be seen in the charts for 1926. (Smith 1927b, pp. 86 and 90, 

 and fi^i 99, p. 155 of this paper.) For several years past this inshore 

 development of the Gulf Stream on the east side of the bank in June 

 and July. insurin<i: safety to the steamship tracks, has been thought 

 to warrant the discontinuance of the ice patrol for the year. 



(e) Tb.e change in alternation in direction of flow between the 

 warm and cold water masses takes place about 25 miles shoreward 

 from the " cold wall '' or temperature wall. (See fig. 122, p. 204.) 

 It is for this reason that icebergs are seldom sighted offshore of the 

 cold-water area south of Newfoundland and that they often drift 

 along parallel to the Gulf Stream, though lying many miles within 

 the offshore boundary of the cold, mixed waters. 



Icebergs around Newfoundland have been trailed by the ice patrol 

 as they drifted many hundreds of miles in the current ; in fact the 

 compilation of individual berg drifts in the area south of the Tail 

 of the Grand Bank constitute the most positive evidence to refute 

 the old belief that the Labrador current flowed southward along the 

 United States coast. (See figs. 102 and 107.) The first attempts to 

 follow the movements of an iceberg near the Grand Bank were by 

 Johnston (1913. p. 23) wdien the Seneca kept in touch with two bergs 

 for over a period of five weeks, finding that they drifted in a large 

 cyclonic vortex off the east side of the Grand Bank. The next in- 

 stance was in 1915 when a berg was followed for 17 days while it 

 drifted 195 miles in a wide semicircle south of the Tail, first to the 

 westw^ard at the rate of 12 miles per day and then eastward in the 

 Gulf stream at 24 miles per day. In 1921 on the ice patrol we began a 

 policy of tracking the bergs wdienever possible, and the composite map 

 of drifts (fig. 102) so compiled has now become very instructive. Some 

 of these follow: AjH-il 11 to May 12. 1921. a large berg drifted from 

 the northeastern edge of the Grand Bank to a point 90 miles south- 

 southwest of the Tail at an average rate of 15 miles per day; thence 

 it was carried eastward on the Gulf Stream for 10 days at the rate 

 of 28 miles per da3^ March 19 to April 15, 1922, a berg drifted west- 

 ward past the Tail of the Grand Bank, then south and east, a total 

 distance of 315 miles at an average velocity of 12 miles per day. 

 Ma}' 2 to 20 three bergs set southwestward past the Tail of the Grand 

 Bank in a large anticlockwise eddy at the speed of 10 miles per day 

 and a total distance of 175 miles. March 16 to April 11, 1924, a 

 berg on the northeastern part of the Grand Bank drifted south into 

 the Gulf Stream a total distance of 450 miles. The rate of drift was 

 slow to the forty-third parallel, but south of the Grand Bank it 

 increased to 24 miles per day. May 19 to June 30, 1925, a berg Avas 

 follow<'d over a distance of 492 miles as it drifted from the forty- 

 fifth j)arallel southward along the east side of the bank to the Tail 

 and tlience southwesterly until turned easterly by the Gulf Stream. 

 May 2() to June 2, 1928, a berg was traced 265 miles in a semicircular 



