936 
BULLETIN OF THE BUREAU OF FISHERIES 
Once the dynamic gradient is established between any two stations, the corre- 
sponding velocity of the water at the one, relative to the other, is calculable by a 
simple formula, described by Smith (1926, p. 31), who also makes clear the correction 
necessary to learn the true velocity if the profile in question does not cut the current 
at a right angle. An alternative method of calculating the velocity, often employed, 
is described fully by Sandstrom (1919); and the Fish Hawk data for Massachusetts 
Bay (June, 1925) have been treated in this way by R. Parmenter (p. 949; figs. 198 
and 199), as an illustration. 
DYNAMIC CONTOURS AND GRADIENT CURRENTS 
FEBRUARY AND MARCH 
At the end of the winter and during the first days of spring, when the general 
equalization of temperature and of salinity (already discussed) makes the upper 40 
meters extremely uniform, regionally as well as vertically (pp. 522, 703), over the 
whole gulf, the distribution of density at the surface would suggest a very quiescent 
state. Thus, the surface chart for February and March, 1920 (fig. 187), shows a 
maximum regional variation of only about 0.4 units over the whole basin, with the 
central part of the latter virtually uniform (at 25.8 to 25.9) from station to station. 
Only the immediate offing of the Kennebec River was then appreciably less 
dense (about 25) at the surface, the Eastern Channel and the region off its mouth 
slightly more so (about 26 to 26.1); and the whole western and central part of the 
gulf, with the coastal belt along Nova Scotia, was then equally uniform at 40 meters, 
though with slightly higher values (26.3 to 26.5) along the eastern side of the basin 
and through the Eastern Channel. 
It is clear that with the water so nearly homogeneous horizontally there is very 
little dynamic tendency toward any general system of gradient currents in the upper 
stratum of the gulf at that season, except that the freshening of the surface by the 
increasing flow from the Kennebec foreshadows the development of a drift westward 
along the coast — a tendency, however, still confined to so thin a surface stratum 
that it did not yet govern. 
Neither does the state of the water at the surface suggest a general dynamic tend- 
ency at that season toward a drift from the east into the gulf past Cape Sable, or 
vice versa, in the surface stratum, the density of the upper 40 meters being com- 
paratively uniform (in horizontal projection) from the cape out to Browns Bank for 
early March. This corroborates the evidence of salinity and temperature that the 
Nova Scotian current did not flood westward past the cape in the spring of 1920 
until later than sometimes happens (p. 832). However, when the density of the deep 
strata is taken into account it becomes obvious that the hydrostatic forces set in 
operation by the banking up of the heaviest water against the eastern slope of the 
gulf (p. 849, fig. 172) must tend to cause a cyclonal or anticlockwise movement of 
the deeper mid strata, carrying with it, as an overlying blanket, the surface stratum, 
itself so nearly quiescent. 
The dynamic chart for February and March, 1920 (fig. 188), gives an indication 
of the stream lines to be expected at the surface under the conditions of temperature 
and salinity then existing, which may be taken as typical of the first two weeks of 
