From their T-S characteristic relationships the Labrador Current 

 water and the Atlantic Current water found in the Grand Banks region 

 are water masses. Usually also in this region these water masses mix 

 in a sufficiently constant proportion so that the mixed water can be 

 regarded as a virtual water mass. The mean T-S relationships for 

 these three water masses for the 10-year period 1948-57 are shown in figure 

 22 in comparison with the conditions found in 1957. The 10-year means 

 are shown as broken lines and the 1957 conditions are represented by 

 solid lines. 



The typical mixed water found in the Grand Banks region is not always 

 present. In some years the proportion of the parent water masses 

 producing the mixed water is variable and the resulting mixed water 

 follows no definite pattern of characteristic T-S relationship. Also 

 occasionally a group of stations will be found to have a consistent pattern 

 differing from the typical mixed water. In 1957, as mentioned above in 

 the discussion of figures 19 and 20, such a group of stations was found to 

 have a T-S pattern which indicated a mixture of the usual mixed water 

 with Atlantic Current water. These observations have been excluded 

 in computing the average T-S values for the mixed water as shown in 

 figure 22. 



The presence of mixed water as a virtual water mass in the Grand 

 Banks region raises questions as to what happens to this mixed water 

 after it is formed. One possibility was that the mixed water formed 

 in the Grand Banks region might supply the intermediate water of the 

 Labrador Sea. The T-S characteristics of the water were examined at 

 the bulk of the stations taken 1951 through 1956 and located in the 

 northward-moving water north of Flemish Cap. No surveys extended 

 into this area in 1948, 1949, 1950, and 1957, so that the 88 stations 

 examined represent all of the postwar observations in the area in question. 

 The resulting individual station curves showed some grouping in the 

 mixed water but more scatter than in the mixed water of the Grand 

 Banks region. The T-S curves fell principally between the Grand 

 Banks curves for Labrador Current water and mixed water with a few 

 stations on the Atlantic Current side of Grand Banks mixed water. The 

 interpretation is that some of the mixed water formed in the Grand 

 Banks region moves northward of Flemish Cap mixing in varying pro- 

 portions with Labrador Current water in the more northern area and 

 that some sinking and seaward spreading of the mixed water goes on in 

 both areas. 



To determine the spread of individual station curves from the 1957 

 characteristic T-S curves an ellipse was constructed for each level for 

 each water mass using probable differences of individual temperatures 

 and salinities from their average values as the semiaxes of the ellipses. 

 The ellipses for a given level for the three water masses were separate 

 from each other except that at 800 meters the mixed-water ellipse was 

 tangent to that for Atlantic Current water, and at 1,500 meters the 



90 



