152 MARION EXPEDITION TO DAVIS STRAIT AND BAFFIN BAY 



Johns, around Cape Race, and as far west as the Miquelon Islands, 

 has been studied by Dawson (1906, p. 21), by Huntsman (1930, p. 

 11), by Matthews (1914, p. 32). by Smith (1924a, p. 131), and by 

 Iselin (1930, p. 3). The outstanding features as affecting the dis- 

 tribution of arctic bergs are (a) the great variations and occasional 

 reversals of the current and (?>) the western and southern limits of 

 the current as marked by the outline of Green Bank and the conti- 

 nental edge. Icebergs have been sighted at various distances to the 

 southwest of Cape Race, but never, to our knowledge, farther west 

 along the coast than Placentia Bay. (See Huntsman, 1930, p. 4.) 



Hundreds of bergs ground on the northern part of the Grand 

 Bank, where processes of accumulation and disintegration proceed 

 throughout the entii'e season. The coldness of these shelf waters pre- 

 serves the ice longer than if it had continued southward; and so in 

 August bergs may still be on the northern part of the bank long after 

 most of the surrounding localities are clear. As they disintegrate 

 and grow smaller bergs may again be carried of!' the bank to resume 

 their southward journey or occasionally they may be driven farther 

 in the shoal waters of the bank itself. The central regions of the 

 latter, as a rule, are free from bergs. 



Branch e. the richest of all the berg streams south of Newfound- 

 land, follows southward in the deep water along the eastern slope of 

 the Grand Bank, as shown on Figure 97 and also by the actual drifts 

 on Figure 102. The behavior of bergs once started along path c 

 depends primarily on the manner in which the water masses of 

 this northern discharge meet and conflict with the warm easterly- 

 moving oceanic masses. The axis of the icy current tends to hug 

 the eastern slope of the Grand Bank and curve westward around 

 the " Tail '' of tlie latter but the outer edge of this flow is continually 

 sending out temporary offshoots, offshore. The mixing zone of the 

 currents which carry the ice has the form of an undulating front 

 extending from southwest of the Tail of the Grand liank to some- 

 what northeast of Flemish Cap. This demarkation is called the 

 cold wall or temperature wall. Whenever and wherever the bound- 

 ary shows a salient, there similarly a vortex tends to develop (similar 

 in many respects to the "polar front '"' of meteorology), but usually 

 the chief point of discharge is near the Tail of the Grand Bank. When 

 the Arctic current swells, the mixing zone between the two waters often 

 moves farther and farther offshore, and in consecjuence the bergs fan 

 out along the northern edge of the Gulf Stream. During late sum- 

 mer, how^ever. according to our observation, the Labrador current in 

 this region probably dwindles and there are times during the autumn 

 and winter, as proved by the observations of the ice patrol (see Smith 

 1923, p. 85 and also fig. 99) when the icy current is not traceable at 

 all to the southward of the Grand Bank. At this season, largely 

 reflecting these oscillations, few icebergs drift far south or east of 

 Xewfoundland, though an occasional berg may be reported south of 

 Newfoundland in any month of the year. They are at a minimum 

 during November, December, and January, and at a maximum 

 during April, May, and June. 



In March there is a marked tendency for the bergs to take the off- 

 shore path (branch d, fig. 97) just south of Flemish Cap. Such a 

 dispersal is due to several factors: {a) The presence of pack ice 

 farther north along tlie American coast, which prevents the bergs 



