than at the Bonavista triangle. This suggests that the water found 

 in the temperature minima at the South Wolf Island section in sum- 

 mer time were formed with the onset of cold weather in the autumn 

 and the water at the temperature minima found at the Bonavista 

 triangle in summer time were formed later in the winter after con- 

 centration of the salt through freezing and removal of about 75 to 

 100 cm. of ice. The reconcilliation of the two essential parts of such 

 a mechanism (mixing or stirring to a depth of some 75 meters or more 

 and the formation of an ice cover which would prevent wind stirring) 

 is not difficult when it is remembered that the ice covering is fre- 

 quently opened by storms through rafting and by the ice being blown 

 seaward. Almost certainly such a mechanism would produce different 

 minimum temperatures in different years and would probably pro- 

 duce different minimum temperatures in different geographical parts 

 of the area of formation during the same year. Such irregularities 

 are in accord with observations in the Grand Banks region. Assum- 

 ing that the season during which such a mechanism would be operating 

 would be limited to about 5 months of the year, September to April, 

 the explanation leaves much to be desired in accounting for what 

 happens in the other 7 months and would seem to require a seasonal 

 variation in volume transport or that the 'area in which such stirring 

 occurs extend over more than the distance traveled in a single year. 



From table 1 the mean heat transport at the Bonavista triangle just 

 prior to the occupation of the South Wolf Island section was 5.98 and 

 the mean volume transport was 3.96 for the 6 years 1948-53. During 

 the same years at the South Wolf Island section the mean heat trans- 

 port was 12.83 and the mean volume transport 5.13. By difference the 

 mean heat and volume transports of that part of the Labrador Current 

 which branches eastward between the South Wolf Island section and 

 the Bonavista triangle are 6.85 and 1.17 which give a mean tempera- 

 ture of about 5.8°. It is considered that this mean temperature is too 

 high but it emphasizes that it is the warmer offshore part of the current 

 which is diverted from the Grand Banks by such branching. 



Not all of the eastern branch of the Labrador Current as it leaves the 

 Bonavista triangle passes through the valley between the Grand 

 Banks and Flemish Cap, but part of it recurves northeastward. This 

 recurvature is of considerable practical importance in determining the 

 degree to which bergs endanger the steamer tracks passing southeast- 

 ward of Flemish Cap. It has been noted qualitatively that as the ice 

 season advances a larger proportion of bergs fail to move southward 

 between the Grand Banks and Flemish Cap. The area of this recurva- 

 ture, northward of Flemish Cap and the northeastern slope of the 

 Grand Banks is one from which comparitively few data are available. 

 Hence, even though the approximate normals for the Bonavista tri- 

 angle as derived above and the approximate normals for section T are 

 on very shaky foundations it is of interest to consider the difference 



65 



