938 BULLETIN OF THE BUREAU OF FISHEEIES 



they show can be accepted only as a rough approximation, not in detail. Some 

 smoothing of the curves has proved necessary in the construction of the chart, also- 



Even with this reservation these contours show that the basin of the gulf 

 (potentially, at least) was then the site of one major cyclonal (i. e., anticlockwise) 

 eddy, with its center taking the form of a troughlike depression extending from the 

 Eastern Channel northward and inward toward the ofBng of the Bay of Fundy. 

 It is interesting that this general eddy seems also to have involved the latter, with 

 the surface water drifting inward along the Nova Scotian side, outward next the 

 New Brunswick shore and past Grand Manan. 



The highest velocities then indicated were a drift northward into the gulf along 

 the western slope of Browns Bank and a counter movement outward along the 

 Georges Bank side of the Eastern Channel. With the correction used here for the 

 difference in depth this indraft works out at about 13.5 centimeters per second, 

 equivalent to 0.27 knot, or about 6J^ miles in 24 hours. The calculated velocity 

 for the outdraft around Georges Bank is lower — 0.22 knot, or 534 miles in 24 hours. 

 These velocities, however, are on the assumptions, first, that the water in the center 

 of the Eastern Channel was stationary and, second, that the difference in depth 

 between the trough of the channel and the crests of its two slopes was correctly 

 allowed for in the calculation (p. 934). 



By contrast, the whole western side of the gulf was "dead," dynamically, as 

 late as the middle of March, in 1920, its upper stratum only tending to drift south- 

 ward (anticlockwise) very slowly, except at the mouth of Massachusetts Bay, where 

 greater velocity in this direction is suggested by contour lines more closely crowded 

 (fig. 188). It is interesting to find that the effect of the discharge from the Kenne- 

 bec and Penobscot was most evident in speeding up the southwesterly surface drift 

 some 40 miles out from the land — not close in to the latter, as the surface chart of 

 density for the same date (fig. 187) would have suggested if taken by itself. 



Lower densities at two of the stations in the basin (20054 and 20052) than in 

 the general vicinity are best interpreted as isolated pools, which, if correct, implies 

 sudsidiary clockwise eddies; so, too, a corresponding high appearing on the east- 

 ern edge of Georges Bank on the dynamic chart (fig. 188). While these seem not 

 to have seriously interrupted the general anticlockwise movement, they are inter- 

 esting illustrations of the persistence of such pools, which have drifted off from the 

 general zone of low density next the coast. 



The comparatively dead state of the water over the whole eastern half of 

 Georges Bank at this season also deserves a word. The chart suggests a slow drift 

 southward and so out of the gulf across the western half of the bank at this time, 

 but the contour of the bottom makes it more likely that the surface water was 

 actually moving eastward around its northern edge, because the underlying strata 

 (which in this case supplied the motive power) are necessarily directed by the sub- 

 marine slope, against which any southward drift must strike. Thus, we may con- 

 clude that the dynamic movement of water around the basin was even more definitely 

 eddyUke and anticlockwise in March than the chart (fig. 188) suggests. 



Lacking March data for the region of Nantucket Shoals, the chart fails to show 

 whether a definite dynamic outflow is to be expected around the latter to the west- 

 ward from the gulf at that season. 



