DR. HAROLD JEFFREYS ON TIDAL FRICTION IN SHALLOW SEAS. 
261 
about 1 '4 x 10 18 ergs per second. The remaining straits of the North-west Passage 
probably do not contribute nearly so much to the dissipation, for the energy of the 
entering wave must be mostly dissipated in the channels already dealt with, and 
partly through this and partly on account of the obstructive effect of the islands 
it is not likely that the straits farther north-west are very important, though this 
cannot be regarded as certain. The dissipation in the whole of the North-west 
Passage is thus about 1‘6 x 10 18 ergs per second on an average. Adding this to the 
amount found for the Bay of Fundy, we have for the whole of North America a 
total of 2 x 10 18 ergs per second on an average. 
Summary. 
The mean rates of dissipation in the lunar semi-diurnal tide found in the foregoing 
investigation are as follows :— 
Ergs per Second. 
European waters. 2'4xl0 18 
Asiatic waters: 
South China Sea. Small. 
Yellow Sea. 1‘lxlO 18 
Sea of Okhotsk.0'4 x 10 18 
Bering Sea. 15'QxlO 18 
Malacca Strait . 1’lxlO 18 
North American Waters: 
Bay of Fundy. 0‘4xl0 18 
North-west Passage. I'6xl0 18 
The total thus accounted for is 2'2 x I0 19 ergs per second. I have shown in a previous 
paper that the dissipation required to account for the secular acceleration of the moon 
{which amounts to 9" per century per century) is about l'4x 10 19 ergs per second, so 
that it seems as if there is more dissipation than is required. If this was so it would 
be necessary to seek for a cause that could produce an appreciable secular retardation 
of the moon, and none such is known. A scrutiny of the results so far obtained is 
therefore desirable, with a view to finding out whether any of them have been over¬ 
estimated. One cause of such an over-estimate is easily seen. The data used for the 
Irish Sea, the English Channel, Malacca Strait and the Bay of Fundy refer definitely 
to spring tides alone when the currents are at a maximum. The height of the tide 
adopted in the calculation for the Yellow Sea was also that of the spring tide. In 
the other cases it is not stated whether the currents have average or spring values, 
but if they were determined at springs the requisite reduction is at once obtained. 
The theoretical ratio of the heights of the lunar and solar tides is 2'3 when inertial 
and frictional effects are neglected. This ratio is probably nearly correct in mid-ocean, 
