85 



dynamic heights at stations 2485 and 2488, and the shallow depth of 

 the 10° isotherm at these stations. Tliis emphasizes one of the most 

 important implications of figure 40, namely that the area is one of 

 great variation with respect to currents ; and at the same time suggests 

 that variations in the size of the major mixed-water pool or parts of 

 this pool wliich may be detached may, acting as virtual obstructions, 

 be partly responsible for the Gulf Stream system dividing into the 

 several branches of the Atlantic Current after passing eastward of 

 longitude 50° W. It would seem that the accumulation of a large 

 quantity of mixed water east of the southern part of the Grand Banks 

 is dependent, not only on an active supply of Labrador Current water, 

 but also on the strength with which the Atlantic Current intrudes 

 westward toward the Banks in about latitude 45° N., and thus restricts 

 the northward flow of the mixed water. Evidently the abnormal 

 growth of tliis pool of mixed water is one of the conditions responsible 

 for the occasionally reported extra-southeasterly drift of bergs. 



The southernmost branch of the Atlantic Current recurves south- 

 westward shortly after passing eastward of the longitude of the Grand 

 Banks. It was hoped that the section running southeasterly from 

 the Grand Banks (stations 2471 to 2484) would cross the major por- 

 tion of the Atlantic Current including this southernmost branch. 

 However, it would seem from the isobaths of figure 40 and also from 

 the depth of the 10° isotherm at stations 2481 and 2486 that this sec- 

 tion did not extend far enough to the south to cross the branch that 

 eventually recurves toward Bermuda. There is, of course, the possi- 

 bility that this branch continued as far as station 2486 before re- 

 xjurving, and that that station was part way across the branch in 

 question. Other evidence regarding this question wiU be considered 

 in connection with the volumes of flow across the various sections of 

 the survey. 



At many of the stations between the Tail of the Grand Banks and 

 Flemish Cap the observations extended to 4,000 meters. Although 

 these stations are so widely spaced that it is not possible to deduce 

 from them detailed patterns of circulation, they nevertheless afl'ord 

 the opportunity of studying the major features. Examination of the 

 anomalies of dynamic height at levels below 2,000 decibars with 

 reference to that surface indicated increasing relative velocities to 

 3,500 decibars and a decrease from 3,500 to 4,000 decibars. Figure 

 41 shows the dynamic topography of the 3,500 decibar surface relative 

 to the 2,000 decibar surface in terms of anomaly of dynamic height, 

 with the contour interval one dynamic centimeter. Similarly con- 

 structed deep-level current charts of the Labrador Sea, published in 

 United States Coast Guard Bulletin No. 26 and based on the 1935 

 post-season cruise, showed the deep water of the Labrador Sea moving 

 southward along the American side and being joined by a contribution 



