934 
BULLETIN OF THE BUKEAU OF FISHERIES 
hence, it must be taken into account in some way in calculating the dynamic slope 
at the surface. 
Jacobsen and Jensen (1926) have very fully discussed this question in their 
dynamic study of the Faroe Channel, finding that in most cases this effect of the 
bottom water may be sufficiently allowed for by arbitrarily applying to the dynamic 
gradient between the two stations in question the product of the difference in 
specific volume between them at the deepest level of the shoaler station multiplied 
by half the difference in depth. If the station where the calculation shows the 
surface as highest also has the largest specific volume at the deepest level of the 
shoaler of the pair, the gradient is to be increased by the amount of this correction- 
decreased if the reverse obtains. If the difference in depth be greater than, say, 150 
meters or so, no arbitrary correction of this sort can be relied upon, consequently 
the dynamic gradient can be stated only within very wide limits. The only cure is 
to establish the stations closer together on future cruises. 
The dynamic-contour chart 90 closely resembles an ordinary weather map in its 
general appearance, and it is as easily interpreted in terms of the resultant circula- 
tion. Dynamically, the water tends to flow down the slopes from the parts of the 
picture where the surface stands high to those where it is low, and at right angles to 
the contour lines. Actually, however, this could happen only at the equator. Every- 
where else the effect of the earth's rotation so deflects this motion that the stream 
lines come nearly to parallel the contour lines, which may then be taken as directly 
representing the current, just as the direction of the wind is roughly parallel to the 
isobars on the weather map. 
In the open ocean, where tidal currents are weak, the contour lines may even 
approximate the tracts of the particles of water if approximately constant accelera- 
tion has been established. This, however, does not apply in a region such as the Gulf 
of Maine, where the tidal currents average much stronger than the dynamic tendencies. 
In this case the latter act only to give to the tidal flow a character more definitely 
rotary than would otherwise be the case, or to strengthen the one tide at the expense 
of the other. Here the dynamic-contour lines show only the general advance which 
the water tends to make good in its tidal oscillations to and fro. 
Because in every case the datum plane for the calculation is necessarily the 
underlying water, not the solid bottom of the sea, the motion indicated by the chart 
is not absolute, but is only relative to that of the deepest stratum of water included 
in the picture. If this be motionless, the calculated drift represents the actual 
motion of the surface (or chosen level) relative to the coast line, but not otherwise. 
In the Northern Hemisphere, where moving bodies are deflected to the right, 
the direction of flow, relative to the plane of reference, 91 is to be identified by the 
rule that the gradient current will constantly have the lightest water (i. e., the 
highest surface) on its right hand, the lowest surface on its left, as it veers cyclon- 
ically around the latter. If the surface drift be faster than the bottom drift, as is 
usually the case, this indicated direction of flow will also be the true direction, 
relative to the bottom; so, too, if bottom and surface drifts be parallel, whichever 
80 Dynamic-contour charts may as easily be constructed for any desired depth below the surface of the sea, as described by 
Smith (1926). 
81 In the Gulf of Maine this is the bottom water between the pairs of adjacent stations. 
