SCIENTIFIC RESULTS 179 



total mass transport of the water, considering the entire brea<lth of 

 the sh)i)e. may remain more or less uniform. The speeding up in 

 one place along the slope, or retardation in another may cause the 

 current, as noted on the dynamic topographical ma[)s, to divide 

 longitudinally into a series of bands or belts. In general the current 

 is sAvift Avhere narrow and sluggish where wide; for example, in 

 July. 11128. as illustrated by the conditions met by the Mar/on expe- 

 dition described on page 149. (See fig. 95.) Off Nain, under the con- 

 ditions then existing, only the ice along the continental edge could 

 have been receiving an api)i'eciable southward set, while practically all 

 of the ice then floating off' Belle Isle was making southward progress. 

 That an iceberg observed drifting southward at a moderate rate may 

 suddenly accelerate to double, or sometimes to treble its former 

 rate, is not evidence that the current throuirhout its entire lenirth 



ICEBERGS Caught in pack ice 



I'njrRB 114. — An ici'berg surrouudt'd by pack ice ou the northern part of the Grand 

 Bank, Feliruary. 1921. The majority of the flr.st bergs appearing .s'outh of New- 

 foundland in early season, are sighted relatively wide off-shore, far to the east- 

 ward. An important factor of this seasonal variation in the course of the bergs 

 is the entangling pack ice which, driving before the westerly winds, exerts a 

 pronounced easterly component to the drift of the bergs. (Official photograph, 

 international ice patrol.) 



has also increased but rather suggests a contraction in breadth along 

 the j)articular pathway that the ice is then following and this is an 

 imj)ortant point as yet little appreciated in oceanography. 



The effect of pack ice on the seasonal distribution of icebergs is 

 of two kinds: {a) When driven before the wind it entangles the 

 bergs and tends to carry them along; or (^) it acts as a fender along 

 the shoreward side of the Labrador current during a critical period, 

 from Xovember to July, tending to prevent the icebergs from 

 drifting toward the coast and grounding as they are borne south- 

 ward. This last assum])tion is corroborated by the fact that it is 

 at the beginning of the season, from F'ebruary through March, 

 when the pack-ice belt is widest that the bergs are distributed 

 farthest to the east of the (IraiKl Bank. 



If tile velocity of the soiitli-flowing current along the Labrador 

 and Xewfouiidland continental edges, as sugge.sted by the Marloih 

 expedition observations, of 12 to 14 miles per day be multiplied by 

 the number of days of the normal iceberg season around the Grand 



