90 



cates that tliis seasonal maximum occmTed at or prior to the first sur- 

 vey, at least a month prior to the usual occurrence of the maximum 

 volume of flow. Although the curves in figure 24 show a steady 

 decrease in volume at section T and a slight secondary maximum late 

 in June at section U, it will be noted that sections T and U were not 

 occupied during the tliird survey and the shaj^e of the curve for sec- 



MARCH APRIL 



Figure 24.— Variation iu the volume of flow of the Labrador Current in the Grand Banks region 

 during the 1939 season. 



tion W indicates the possibility that all three sections experienced a 

 pronounced secondary maximum during May. Some support for this 

 may be found in the reported location of the bergs. It is character- 

 istic that for a short period early in the season some of the berg-i put 

 in their appearance southeast of the 1,000-fathom curve, fanning out 

 from an axis extending southeasterly from about 46° N., 46° W. 

 This characteristic invasion occurred early in the 1939 season and was 

 repeated at about midseason. Tliis may be a coincidence but it is 

 reasonably possible that a considerable increase in the volimie of flow 

 of the Labrador Current is accompanied by a temporary flooding or 



