PHYSICAL OCEANOGBAPPIY OP THE GULF OP MAINE 
939 
In the offing of Cape Sable the dynamic gradient for March, 1920, calls for a 
weak drift clockwise but spreading far offshore toward Browns Bank before eddying 
northward again toward the gulf. Hence, thecold Nova Scotian water that we 
encountered midway out over the shelf (station 20075, p. 1000) did not then tend to 
round the cape, but to veer offshore, which agrees with the distribution of temper- 
ature and salinity at the time. Dynamic evidence also is strong that whatever water 
was then entering the eastern side of the gulf in the upper stratum was drawn chiefly 
from the region of Browns Bank and from the edge of the continent in the offing of 
Cape Sable — i. e., from the source whence the gulf regularly receives its slope water 
(p. 848). 
The dynamic gradients for March are especially instructive along the continental 
slope abreast of the gulf because of the light they may throw on the problem of the 
so-called “Gulf Stream” along this sector. Fortunately, this is made comparatively 
clear for this region (fig. 188) by the considerable difference in density between the 
outer stations on the two cross profiles of the bank— western and eastern (stations 
20044 and 20069). On the eastern profile the gradient (dipping to a low at the 
outermost station) shows a strong drift to the westward along the edge of the bank, 
its calculated velocity being about 0.6 knot, or 14 miles in 24 hours. While this 
calculation depends on the correct allowance for the difference in depth between 
stations, one of which was much deeper than the other, 92 the direction of this gra- 
dient current is well established. A weak continuation of this westerly drift (indi- 
cated by a low in the dynamic contour) extended along the edge of the bank as 
far as the western profile (run three weeks earlier) ; but here this gave place to a much 
steeper counter gradient to high in the next 10 miles offshore, implying a counter 
drift to the east. 
Unfortunately, the difference in depth between the stations on the edge of the 
bank and outside is again so great on this profile (150 to 200 and 1,000 meters) 
that the arbitrary correction employed to take account of it becomes only a rough 
approximation, though the order of this correction (i. e., whether increasing, decreas- 
ing, or even tending to reverse the gradient calculated for equal depths) is in every 
case clear enough (p. 934). When all reasonable allowance is made for this source of 
error, however, the velocity of the easterly drift may safely be set as at least half a 
knot. Fortunately, calculation of the dynamic head between the two outermost 
stations on these two profiles is not subject to this error, both being deep enough 
(1,000 meters) to reach equal density at the lowest levels. Consequently the general 
contour, as laid down for this region in Figure 188, is established, as is the fact that 
the western profile reached out to water of comparatively high temperature and 
salinity in the upper stratum, while the eastern profile did not, though its outermost 
station was still farther out from the edge of the continent. 
So long as the dynamic gradient continues to be of this sort it is evident that 
the superficial drift of warm water along the continental slope, commonly spoken of 
as the “ inner edge of the Gulf Stream, ” is not only to be described as a typical 
gradient current but is to be expected within 15 to 20 miles of the edge of the bank 
between longitudes 68° and 69°. Farther east, however, the contour lines on the 
chart (fig. 188) show it departing farther and farther from the bank, agreeing in this 
•> Station 20068, 200 meters; station 20069, 1,000 meters. 
