51 



have been plotted in figure 31 and connected by lines which are solid 

 between observations a year or less apart and broken and shaped to 

 conform to the sea-level curve between observations separated by 

 more than 1 year. In the same figure is shown the fluctuation in the 

 difference in sea level, Bermuda-Charleston. Tide gage records from 

 Bermuda do not exist for a part of the interval and for such months 

 a monthly normal based on the mean of observations made from 

 1833-43 and 1934-36 has been used. Present-day monthly normal 

 barometric pressures have been used to correct the old tide gage read- 

 ings and all recent tide gage readings have been corrected for actual 

 monthly mean barometric pressures. 



Figure 31 is based on a very small number of data. Also it is 

 admittedly a far cry from the Bermuda-Charleston part of the Gulf 

 Stream system to the Cape Farewell section by way of the Arctic 

 basin. In spite of this, the correspondence between the two curves 

 shown in figure 31 is too striking to be dismissed as pure coincidence 

 without further consideration. There are two physical interpreta- 

 tions of figure 31 which are plausible. If an increase in the difference 

 in sea level between Bermuda and Charleston is an indicator of the 

 increase in the volume of flow of the Gulf Stream system, it might 

 be accompanied by an expansion of the system, thus supplying water 

 to the Arctic basin from the northeastern North Atlantic at a faster 

 rate, and in turn speeding up the outflow from the Arctic basin along 

 the east coast of Greenland. The other point of view, that held by 

 Iselin,2 is that an increase in volume of flow of the Gulf Stream is 

 accompanied by a contraction of the system. In the northeastern 

 sector such a contraction would mean a southward retreat of the so- 

 called polar front. As the atmospheric pressure centers shift with 

 the polar front, a southward shift w^ould result in wind directions 

 more favorable to the East Greenland Current. That this does not 

 lead to an unstable situation is considered by Schell ^ to be because 

 such a shift is followed by a lessening of the intensity of the atmos- 

 spheric circulation. 



Possibly both the amount of water contributed to the Arctic Basin 

 by the Gulf Stream system and variations in the driving force of the 

 atmospheric circulation are eft'ective in controlling the flow of the 

 East Greenland Current, even though they seem to have opposite 

 effects ; for it will be noted that, although the year-to-year variations 

 shown by the curves in figure 31 have a positive relationship, each 

 of the curves has a long-range trend and these trends are oppositely 

 directed. 



During the 1940 post-season cruise the section from South Wolf 

 Island, Labrador, to Cape Farewell, Greenland, was occupied, except, 



2 Loc. cit. 



3 Scholl, 1. 1. Polar lee as a factor in seasonal weather. Monthy Weather Review. Supplement No. 39, 

 pp. 27-51. (1940) Washington. 



