PHYSICAL OCEANOGRAPHY OF THE GULF OF MAINE 
955 
possible errors caused by the considerable period of time over which each survey 
extended. The rapidity with which the density of the upper stratum may be 
increased, if the surface be chilled by vertical circulation of any kind, makes it 
unsafe ever to lay any stress on small regional differences where tidal currents cause 
'as much overturning of the water as they do in parts of the Gulf of Maine. 
The accompanying dynamic chart for the summer of 1914 (fig. 203) shows 
the dynamic tendency toward circulation at the surface of the inner parts of the 
gulf and of the waters off Marthas Vineyard for August and of the Georges Bank- 
Browns Bank region for that July. Unfortunately, these two divisions of the pic- 
ture are not strictly comparable because solar warming had been responsible for 
Fig. 202.— Density at 100 meters, July to August, 1914. Corrected for compression 
some slight decrease in the density of the surface stratum from the one month to 
the next, and for a very considerable decrease close to Cape Sable, where stations 
situated close together but occupied 17 days apart differed by 0.4 in density. Never- 
theless, the general dynamic gradient proved so consistent for the gulf as a whole 
for the two months that it has seemed justifiable to neglect the time interval in draw- 
ing the contour lines; the more so since the heaviest centers for July and August 
proved almost exactly equal in dynamic height. 
If the chart, so combined, be indeed typical of the season (as seems likely from 
general knowledge of the temperature and salinity of the region), two centers of high 
density (indicated as “low” on the dynamic chart) are now to be expected — the one 
