45 



closely connected as this it seems improbable that these fluctuations, 

 occurring at all three sections, are unrelated. However, if they are 

 related their common cause would seem to be identified with the 

 Atlantic Current since the effect spreads northward. Possibly the 

 reason is a subsurface eddy in which mixed water moving northeast- 

 ward along the border of the Atlantic Current recurved to augment the 

 true Labrador Current in its southward flow. 



In Bulletin No. 29 of this series the temperature-salinity relationship 

 found during the first current survey of 1939 was discussed. All of the 

 surveys made in the Grand Banks region since and including 1934 have 

 now been studied for T-S correlations. As in the case of the single 

 survey reported on in Bulletin 29, the temperature-salinity curves for 

 all stations fell into one of three groups, namely, Labrador Current 

 water, mixed water, or Atlantic Current water. Occasionally the 

 values for the upper levels at a station would fall in one group and the 

 values for the lower levels would fall in an adjacent group if the station 

 happened to be taken near the boundary between two types of water. 

 A mean T-S curve was drawn for each of the three types of water for 

 each survey. From the cm'ves representing individual surveys mean 

 •curves for the different years and for the 7-year period 1934-40 have 

 been constructed. Figure 29 shows the temperature-salinity corre- 

 lation for the three types of water for the five surveys made during 

 1940 as solid lines. Curves representing the 7-year means are shown 

 as broken lines. If the 7-year means are considered to represent 

 normal conditions it is seen from figure 29 that in 1940 in both the 

 Labrador Current water and the mixed water the temperatures were 

 from 0.2° to 0.7° C. warmer than normal in the upper levels where 

 the salinity was below 34.9%o. In the deeper levels in the Labrador 

 "Current water and mixed water as well as in the border of the Atlantic 

 Current at temperatures below 9° C, the salinities were 0.02%o to 

 O.03%o fresher than normal. An approximate depth scale has been 

 shown on each of the six curves. A comparison of the curves for 1940 

 with those representing the 7-year means indicates that shifts in depth 

 partly compensate for shifts in the T-S curve in parts of the curve. 



To show this more clearly figure 30 is presented. Here the values 

 of ct corresponding to the depths shown on the curves of figure 29 have 

 heen plotted as vertical distribution curves. This figure shows that 

 although the upper levels in the Labrador Current had about normal 

 density the deeper levels as w^ell as the Atlantic Current and mixed 

 water were lighter than usual. The relative positions of the 1940 

 curves for Atlantic Current water in both figures 29 and 30 with 

 respect to the curves for the 7-year mean might be open to the explana- 

 tion that in 1940 the measurements extended farther into the Atlantic 

 'Current than the 7-vear mean and that the Atlantic Current remained 



