162 



in June it had shrunk to a breadth of 20 miles and a depth of 160 

 meters (86 fathoms), while in October it was only 12 miles across 

 and 75 meters (41 fathoms) deep. 



The fact that there was less Arctic water around the Grand Bank's 

 slopes in October than in April or June lends support to the theory 

 that the cold northern discharge is subject to seasonal variations of 

 measurable proportions, flooding to a maximum during spring and 

 dwindling in summer. It may cease altogether temporarily in 

 autumn or winter (Treasury Bull. No. 10), leaving behind patches 

 of Arctic water (relics), which would be recognizable by their low 

 temperatures imtil absorbed in the general system of eddylike mix- 

 ings which are constantly in process around the Grand Banks (p. 

 146). In October, when these observations were made, the volume 

 of the Arctic water was near the minimum. The only winter ob- 

 servations in the vicinity of the Grand Bank were made by the ice 

 patrol in February, 1921 and 1922 (Treasury Bulls. No. 9 and No. 

 10), when there proved to be no Arctic water either on the south- 

 west slope or even on the eastern, except north of latitude 43° 15', 

 February, 1922. This might point to a progressive dwindling of the 

 current in late summer and in the fall, which possibly culminates 

 in its complete cessation, but in any event the minimum strength 

 of the discharge from the north occurs in late fall and winter. But 

 even then temporary pulsations probably occur, a belief based 

 upon the observations during the period March- July, 1923, when 

 fluctuations in the strength and volume of the Arctic current around 

 the Tail were frequent and irregular. 



It will be noticed on sketch No. 11 that a patch of isolated icy 

 water was found 90 miles northwest of the Tail on the southwest 

 slope, having the same physical character as the Arctic band at the 

 Tail, and was no doubt a part of the latter, which had been pinched 

 off and moved to the westward along the slope. This corresponds 

 to conditions in the previous May (p. 146), when, during short 

 periods of alternate flooding and ebbing of the northern current, 

 isolated patches of polar water were left along the southwest slope, 

 to be engulfed and finally absorbed by mixing with Atlantic water. 

 In the deeper layers, on the contrary, such icy rehcs tend to work 

 to the westward around the' continental slope, as shown bv the drifts 

 of bergs. (See charts "B" and "C") 



The absence of icy water colder than 0° along the mid section on 

 the southwest slope in October is attributable to the mixing process 

 between cold northern water and warm Atlantic water. The 

 equaUzation of temperatures suggested by the section at this place 

 indicates that immediately prior to the taking of the observations a 

 very active interdigitation took place, leaving a water mass completely 



