water stations fell into two correlation groups instead of being scattered 

 from the Atlantic group to the negative salinity-anomaly limit was an 

 unexpected development because the Labrador Current oft" the Labra- 

 dor coast is characterized by such a scattering which arises from the 

 typical temperature minimum with approximately horizontal iso- 

 therms being intersected by inclined isohalines with the salinity de- 

 creasing shorewards. Hence it was surprising to find that the tem- 

 perature-salinity curves for all the individual stations in the Labrador 

 Current in the Grand Banks region occupied during the first survey 

 followed a definite correlation. The third group was composed of 

 stations taken in the area of mixed water between the Labrador Current 

 and the Atlantic Current. This group of mixed-water stations also 

 followed a definite correlation between temperature and salinity, the 

 correlation curve being located between those for Labrador Current 

 water and Atlantic Current water and closer to the former than the 

 latter. The correlation curves for the Labrador Current water and the 

 mixed water are shown in figure 28. An approximate depth scale in 

 meters has been shown on each of the curves. Southwest and south 

 of the Tail of the Grand Banks the width of the area between the 

 Labrador and Atlantic Currents and occupied by the mixed water was 

 less than the interval between stations. In other words the tempera- 

 ture-salinity characteristics changed abruptly from Labrador Current 

 to Atlantic Current in going from one station to an adjacent station. 

 However, east of the Grand Banks the width of the area of mixed 

 water was considerable, averaging about 80 miles. The shape of the 

 mixed-water area may be interpreted as verifying other indications 

 that the region just south of the Grand Banks is one in which the mix- 

 ing of the Labrador and Atlantic Currents is very active. The fact 

 that the temperature-salinity correlation curve for the mixed water is 

 much nearer that for Labrador Current water than for Atlantic Cur- 

 rent water is probably explained by the direction of motion of the 

 mixed water being counter to that of the Labrador Current water and 

 similar to that of the Atlantic Current resulting in a greater further 

 admixture of the former than of the latter. It is emphasized that the 

 foregoing developed from the study of a single survey and that gen- 

 eralizations cannot be made on a sound basis until a number of such 

 surveys have been investigated. 



One respect in which it is particularly premature to draw general 

 conclusions from this single survey is in regard to the part of the Lab- 

 rador shelf or slope traversed by the Labrador Current found in the 

 Grand Banks region. From the standpoint of berg ice this is a most 

 tempting field for speculation because of the composition of the Lab- 

 rador Current of a frigid portion over the shelf and a relatively warm 

 portion over the slope off the Labrador coast, because of the major 

 mortality suffered by any berg crop en route to the Grand Banks, 



