DR. HAROLD JEFFREYS ON TIDAL FRICTION IN SHALLOW SEAS. 
255 
great, but a reasonably accurate estimate of it is difficult to make on account of the- 
form of the shallow portion, which has no narrow place that can be called an entrance. 
It is best to treat the main part of the sea and the bays separately. 
In the south of the sea it is stated that the maximum rate of the water, when clear 
of the passes between the Aleutian Islands, is usually about 2\ knots when the depth 
is less than 100 fathoms. In the region satisfying these conditions the depth is in 
most places about 80 fathoms, so that the current farther north, where the depth is 
often only 20 or 30 fathoms, may exceed this. On the other hand there seems to be 
little semi-diurnal tide in the extreme north. In Norton Sound the tide is diurnal, 
presumably because the waves from the south and from Bering Strait neutralize each 
other. At St. Lawrence Islands, near the entrance to the strait, the tide is only 
about a foot in height, confirming this suggestion. Farther south, however, the 
tidal wave from the Arctic must spread out and become inappreciable. At Pribilof 
Islands, 500 km. from the nearest land and surrounded by water 50 fathoms deep, 
the current reaches 2^ knots. At St. Matthew Island, about midway between these 
and St. Lawrence Islands, in water 30 to 40 fathoms deep, and nearly as far from 
land, the current still reaches 2\ knots. There are no other data given for islands 
far from shore, and it seems that we shall not be overestimating the dissipation if we 
take the maximum current to be 2\ knots all over the shallow region bounded on the 
south by the Fox Islands and extending north till half-way between St. Matthew and 
St. Lawrence Islands. The size of this is 1,000 by 700 km., or 7 x 10 5 sq. km. The 
maximum dissipation is therefore 2'74 x 10 12 x 7 x 10 5 x (2'5) 3 , or 3xl0 19 ergs per 
second, and the mean dissipation l‘2x 10 19 ergs per second. 
In Bristol Bay the average velocity seems to be about 3 knots, though the observa¬ 
tions are few. The corresponding dissipation is about l'5x 10 18 ergs per second. In 
Norton Sound the dissipation is probably small, for it is mostly north of St. Lawrence 
Island, and the tide is diurnal. The Gulf of Anadir probably contributes about as 
much as Bristol Bay; for though its area is twice as great, its more northerly 
situation must reduce the current somewhat. In all, then, the average rate of 
dissipation in Bering Sea is about l‘5xl0 19 ergs per second. This estimate is of 
course subject to considerable error, for it depends wholly on a few observations, 
which may not give quite a fair sample of the whole of the sea. The depths around 
the localities considered are fairly typical of the sea as a whole, so that great error on 
this ground is not to be anticipated ; but errors in observing the velocities may be 
greater, and both kinds of error are magnified in importance by the fact that the 
velocity must be cubed when the dissipation is calculated, so that if the true mean 
velocity were only 2 knots instead of 2\ knots the dissipation would be almost halved. 
It does not appear that the velocity increases much towards the coast; in fact the 
velocities near the Alaskan coast seem to be rather smaller than those near the 
islands. Thus an underestimate on this ground is not probable. 
2 N 
YOL. CCXXI.-A. 
