958 BULLETIN OF THE BUKEAU OF FISHEEIES 



It is not SO easy to reconcile the continued drifts of these Cape EUzabeth bottles 

 toward Nova Scotia and the Bay of Fundy with the dynamic contours, for the latter 

 suggest that any driftage from the northern coast of the gulf that reached the central 

 part of the basin would rather be drawn into the circulation around the heavy center 

 in the Eastern Channel, and so be carried outward around the eastern end of Georges 

 Bank. This, in fact, seems to have been the fate of some of the bottles set out off 

 Cape Ann and of most of those set out off northern Cape Cod in 1923 (fig. 176). 

 It seems reasonable, therefore, to conclude that by the end of July or first of August 

 of most years the zone of demarcation between the eastward drift around the southern 

 side of the northern heavy pool and the counter drift around the northern side of the 

 southern pool is located somewhat farther south -than it was in August, 1914 — not 

 far, in fact, from the line of monthly separation laid down on the chart for that year 

 (fig- 203). 



The distribution of density around the eastern slopes of Georges Bank affords 

 a striking illustration of the necessity for taking account of the difference in depth 

 between pairs of adjacent stations in the dynamic calculations, arbitrary though 

 this correction be (p. 934). Without the inclusion of this factor (p. 934), the dynamic 

 head between the low over the Eastern Channel and the high surface over 

 the neighboring part of Georges Bank would have been only about 1 to 2 dynamic 

 centimeters in July, 1914 (except for one station at the extreme edge of the bank — 

 station 10226 — where an isolated pool of low density was recorded). Inclusion of 

 the difference in depth increases this gradient to about 10 dynamic centimeters, 

 working out at a relative velocity of about 0.5 knot out of the gulf around the eastern 

 end of the bank (except as interrupted by a subsidiary clockwise circulation around 

 the light center, just mentioned), which is probably a closer approximation to the 

 truth. 



The dynamic gradient along the southern edge of Georges Bank for July, 1914 

 (fig. 203) , offers an explanation for the fact that none of the bottles from the hues 

 set out off Cape Ann and off northern Cape Cod, which have gone out of the gulf 

 around the eastern end of Georges Bank, have been reported from west of the longi- 

 tude of Cape Cod, when so many set out to the south of the cape have gone in that 

 direction (p. 881 ; figs. 174 and 176). With the dynamic contours turning southward 

 to sea from the eastern end of the banlt, and with the surface gradient rising from 

 longitude 67° to longitude 68°, the March state (fig. 188) is recalled. 



The reasonable expectation with this dynamic distribution is that driftage leav- 

 ing the gulf by this route would circle offshore somewhere abreast the eastern part 

 of Georges Bank, to be carried toward the northeast, finally, with the so-called 

 "Gulf Stream drift." It is probable, also, that at least three bottles that went to 

 England and to Ireland from the Cape Ann and northern Cape Cod lines of 1923 

 (fig. 176) followed this route. 



The whole area of Georges Bank was comparatively dead water in July, 1914, 

 just as in March; consequently no dominant movement is indicated across it either 

 into or out of the gulf, which is corroborated by the evidence of temperatui-e and of 

 salinity. The bank as a whole is therefore made the center of a clockwise type of 

 dynamic circulation in July, just as the inner part of the gulf is of an anticlockwise 

 type. 



