than the water in situ there and is therefore denser; it sinks below the surface and continues to flow along the bottom either back to the polar regions or towards the equator.
This main scheme is complicated in various ways: (1) by the rotation of the earth, which continually deflects currents of water or air to the right in the northern or to the left in the southern hemisphere; (2) by the conformation of the land masses (as in the case of the equatorial stream which is banked up in the Gulf of Mexico and flows out through the Straits of Florida); (3) by the varying depth of the ocean, for currents tend to flow more readily through deep than in shallow waters (as in the case of the main Atlantic drift, which flows most strongly through the deep channel between Shetland and the Faroe Is.); and (4) by the driving force of the winds acting on the surface of the sea (thus the drift of water from the equator is not N.E., as one might expect, but from E. to W., because of the impelling force of the N.E. and S.E. trade-winds).
All ocean currents vary from year to year in their strength of flow and the main interest of physical oceanography in recent years has been the tracing-put of these variations and the search for the causes. The variations themselves are detected by the method of seasonally repeated hydrographic soundings. Samples of water are collected periodically from a number of places in a large sea-area (the North or Norwegian seas, or the English Channel, for instance) at the surface, bottom and a number of intermediate levels. At the same time temperature observations are made. Stations which are placed in a straight line across a sea are then connected and “sections” are made. These show the magnitudes of the layers of different salinity and temperature beneath the surface, and when a number of sections are compared the differences from season to season and from year to year can be seen. So far only the North Atlantic has been at all well studied and evidence of seasonal and periodic variations extending over a number of years has been obtained in this area. Water drifting into the North Atlantic from the equatorial stream has a relatively high salinity (from 36‰ to 36.5‰) and a high temperature (from 15 °C. to 20 °C.), and when the distribution of salinity from season to season is studied it is seen that the area of dense water (salinity 36‰) extends farther to the N. in Nov. than in March. A large area of the North Atlantic is thus covered with relatively warm and dense water and this would slowly drift N. until it cooled sufficiently to sink beneath the surface. The prevailing W. and S.W. winds, however, drive it towards the N.E., where it impinges on the shallow seas and shore of northern Europe.
Taking such an easily surveyable area as the North Sea, the quantity of relatively warm and dense Atlantic water entering it from year to year can be estimated by the method of hydrographic sections. It can thus be seen that Atlantic water enters the North Sea round the N. of Shetland and (to a far less extent) through the English Channel. The flow culminates about March in each year, when a considerable part of the North Sea is covered with water of 35‰ salinity, but in Nov. the area so covered is very much less. Therefore the inflow waxes and wanes from season to season throughout the year, but it also varies in the same season in different years. There is no doubt about the latter variation, but with regard to its periodicity—that is, the number of years elapsing between one maximum and the next—much still remains to be done.
Farther to the N. of the British Isles the superficial drift of Atlantic water ceases, the temperature having fallen so much that the inflowing water becomes denser than that in situ, so that it sinks beneath the surface. It still flows on, however, as a deep current and it then becomes a factor of immense importance with regard to the fisheries in the regions into which it penetrates. The sinking-down occurs in the Kattegat when the inflowing Atlantic water enters the Baltic as an undercurrent which is both warmer and denser than that on the surface. The same thing occurs as the Atlantic stream rounds North Cape: there it breaks up into branches which are irregularly distributed and, sooner or later, sink below the surface and flow on as submarine currents. Entering the Barents Sea (that is, the area between the ice and the northern coast of Europe), these currents flow along the bottom. The inflowing Baltic undercurrent carries with it herrings and other fish from the North Sea outside, and the submarine current entering the Barents Sea also carries with it such fish as plaice. It is mainly because these fisheries are seasonal that the periodicity has been noticed, and because of the economic interests involved the study of the seasonal and longer periodicities has become very important.
As to the causes of the changes in the strength of the current from year to year much investigation has still to be made. The connexion that seemed to be first established was between variations in the quantity of water transported from the tropical to the sub-polar Atlantic and variations in the intensity of solar radiation. Helland-Hansen and Nansen traced a periodicity in the flow of Atlantic water along the W. coast of Norway: every ten to twelve years this flow appeared to reach a maximum and a graph of the variations showed a certain resemblance to the well-known graphs showing the numbers of spots on the sun from year to year. Not only so, but a similar variation was traced in the productivity of the great Lofoten (Lofoden) cod-fisheries. It was difficult to be sure as to the variations in the actual number of fish caught, but it was easy to show that there was a real variability in the yield of cod-liver oil (an important product of the fishery). Tracing, then, the quantities of oil given per 1,000 fish from year to year, they seemed to establish a connexion between the variation in “condition” of the fish, the variation in the inflow of Atlantic water, and the variation in the number of sunspots from year to year.
The relation appeared, however, to be far more complicated than was at first supposed. Helland-Hansen and Nansen showed later that it was improbable that variations in the northerly drift of Atlantic water could be traced directly to variations in the quantity of heat received by the sea from solar radiation. Of the total quantity of energy incident on the earth about 40% is reflected back from the earth’s atmosphere. Of the 60% that penetrates only about one-third actually heats up the surface of the land or sea and the rest is absorbed by the atmosphere. The heating of the latter causes great differences of pressure, which in turn set up changes of atmospheric circulation. Now it is probable that the main cause of oceanic circulation is the driving force of the winds upon the superficial layers of water; hence periodic and irregular changes in the direction and velocities of ocean currents are probably due to changes in atmospheric circulation traceable to changes in the quantities of heat absorbed from the sun by the earth’s atmosphere.
Later still Hjort showed that the study of the variability in the productivity of a fishery is always a complex matter—far more so than was formerly supposed. It appeared that the quantity of oil contained in the liver of a cod (per unit of weight) increases with the age of the fish. Detailed study of the cod shoals also showed that their composition was continually changing: in some years the shoal is composed of younger or older fish than the average and with this latter variation there are changes in the quantities of oil yielded per 1,000 fish. The changes in the composition of the shoals, as regards the proportions of the various “year-classes,” are to be correlated with oceanographical changes (see below). It is proper, however, to point out at once how very complicated may be the relationships between oceanographical and strictly biological phenomena, though, of course, the latter are ultimately dependent on the former.
Long-range Periodicities in Oceanographical Changes.—More and more the science seeks to discover periodicities and to correlate these with others. In these attempts new methods are elaborated and in their criticism contributory phenomena are discovered. An interesting example is the discussion, by Otto Pettersson, of the effects of long-range fluctuations in the tide-generating force: this memoir was published about 1914, but has only recently become available to English readers.
The tide-generating force is due to the attraction of the waters of the ocean by sun and moon. There are two gravitational fields which sometimes reinforce and at other times diminish each other and the effect is always a resultant one. There are therefore maxima and minima in the value of the tide-generating force, depending on the relative positions of the sun, earth and moon. The orbits of earth and moon are elliptical, so that the earth is sometimes nearer, sometimes farther away from the sun, and the same is the case with the moon in relation to the earth. The orbital planes of earth and moon are inclined to each other at an angle of 50·8° and at two points only in its orbit can the moon be situated in the plane of the ecliptic: the line joining these two points is called the “line of nodes.” A line joining the moon in perigee and in apogee is called the “line of apsides.” Now such a constellation as the following must sometimes exist: the earth is in perihelion; the line of nodes coincides with the line of apsides and both lie in the line joining earth and sun. The line of nodes rotates in a period of 18·612 years and the line of apsides in a period of 8·84 years. Such a constellation can be shown to occur at intervals of about 1,800 years and about those times the tide-generating force will be at an absolute maximum. Working out the calculations, Pettersson finds that the favourable constellation occurred and will occur in 3500 B.C., 1900 B.C., 250 B.C., 1433 A.D., 3300 A.D., and so on. In addition to these there are subsidiary maxima at intervals of 412, 9 and 84–93 years.
Given, then, that the variations in tide-generating force are big enough, the periods when the maxima occur will be critical with regard to oceanographical and meteorological phenomena. About the time of the maxima there must be a longer tidal range (that is, a greater rise and fall than the average); the difference between neap tides and spring tides will also be increased, and as results of these conditions there must be great tidal floods breaking over low-lying coasts and producing extensive denudation. Pettersson further deduces sharp extremes of climate and great temperature contrasts. Far inland he supposes there will be devastating droughts. An effect of the greater tide-generating force will also be instability of the liquid magmas underlying volcanic areas, leading to violent eruptions and earthquakes. There will be great outbursts of polar ice, but this will melt at higher latitudes than in the periods when the tide-generating force is minimal.
