48 



It should be remembered that in the foregomg discussion and figures 

 29 and 30 wherever the Atlantic Current is mentioned it is understood 

 to mean the northern or outermost part of the current. In figure 29 

 it will be noted that the T-S curve for this water has a characteristic 

 salinity minimum occurring at a temperature of about 6° C. The 

 source of this fresher water is probably the slope water of the Gulf 

 Stream system off the American coast. Whatever its source, the fact 

 that this salinity minimum is a characteristic of the outer border of the 

 Atlantic Cm-rent in the Grand Banks region suggests its possible use 

 as a boundary criterion there. It has the further advantages that it 

 occurs at depths great enough to be beyond the reach of superficial 

 changes and yet at a temperature safely out of range of the mixed 

 water. 



On the basis of the 1940 surveys the boundary criterion was taken 

 as the condition where a temperature of 6° C. corresponded to a 

 salinity] of 34.95%o. For any particular survey, anomalies from this 

 salinity at the 6°-isotherm were then plotted at the geographical loca- 

 tions of all stations at which the 6°-isotherm was present. From this 

 information the course of the horizontal projection of the line con- 

 necting locations of zero anomaly was interpolated. To determine 

 the range over which a straight line interpolation could be used, the 

 anomalies for all 1940 stations were plotted against depth of the 6°- 

 isotherm. The location of the points on this plot indicated that no 

 considerable error would be introduced by straight line interpolation 

 for anomalies positively greater than about — 0.10%o (which occurred 

 at a depth of about 300 meters) . 



The boundary, thus defined, between the Atlantic Current and the 

 Labrador Current or the mixed water, as the case might be, was then 

 determined for each survey from 1934 to 1940, inclusive. As this 

 boundary is serpentine in its course from such characteristic salients 

 as the intrusion of Atlantic Current water toward the Grand Banks at 

 about latitude 45°, and as these salients vary somewhat from time to 

 tune in latitude as well as longitude, the problem of how best to con- 

 sider fluctuations in the boundary arose. Because fluctuations in a 

 direction normal to the current are of far greater significance in ice 

 patrol work than are the fluctuations in latitude of the characteristic 

 salients, the fluctuations were considered as fluctuations in the area 

 bounded by the current boundary on the southeast, the 45th parallel 

 on the north, the 49th meridian between latitudes 45° and 43° on the 

 west, and a rhumb line extending from 43° N., 49° W., tln-ough 42° 

 N., 47° W., to the current boundary on the southwest. Thus a 

 decrease in the area so bounded corresponds to a general transverse 

 movement of the boundary toward the Grand Banks and an increase ui 

 area corresponds to a retreat of the boundary away from the Grand 

 Banks. 



