of ice is placed. On the other end A, a piece of metal ab, is placed in contact with the water, and heated at one end by a lamp at a. A little blue colouring liquid is put into the water near the ice, and a little red colour near the hot metal. A continuous circulation in the direction of the arrows is the result. The end A, of course, represents the and D, the polar, regions. The effects of polar cold and tropi cal heat is thus clearly demonstrated, and had the earth been entirely covered with uniformly-deep water, subject to no other disturbing influences, a regular vertical and meridional circulation of the water would be manifest.
Besides unequal densities, other causes may, and most probably are, operating in the production and maintenance of oceanic circulation. The northern and southern branches of the great equatorial currents, in each of the three great oceans, are directly referable to the action of the winds. In the Pacific and Atlantic the trades generally blow in the same direction, and the great equatorial currents sweep along to the west; but in the Indian Ocean the currents north of the equator are regulated by the monsoons, and vary their flow accordingly. Were the earth uniformly covered with water, the effect of the prevailing winds on the surface of the water would be much more marked than it is ; but the intervention of the land masses breaks the continuity of the ocean, and frequently changes the normal direction of the currents.
The bottom layer of water, of almost glacial coldness, is admitted to be an influx from the poles towards the equator. The generally-received opinion is, that this influx is due to the sinking of the surface water in the polar regions, owing to increased density produced by the extreme cold, and the ex pansion of the surface water in the tropics, owing to the extreme heat ; thus creating a difference in level between the polar and equatorial seas. The equatorial waters being higher
than the polar, they naturally flow towards the poles, but, on arriving in higher latitudes, gradually lose all their heat, and finally sink to the bottom. But the over-flow from the equator must be balanced by an under-flow from the poles, and thus the cold bottom-layer of the ocean is simply polar water in transit towards the equator, where it slowly rises to the surface, and again flows towards the poles. And as the process of heating the surface water in the tropics, and chilling it in the polar seas, is constant, the overflow and underflow are constant also.
The influx of cold water into the Indian and Pacific Oceans must be in the former entirely, and in the latter very nearly so, from the Antarctic basin ; but in the North Atlan tic, at least, the cold bottom-water must be principally derived from the Arctic, while that in the South Atlantic flows in from the Antarctic. As we have already observed, this influx of cold water " creeping " along the bottom of the ocean from the poles to the equator, is due to the unequal densities of polar and equatorial waters ; but Sir Wyville Thomson thinks that "the influx of cold water into the Pacific and Atlantic Oceans from the southwards, is to be referred to the simplest and most obvious of all causes, the excess of evaporation over precipitation in the northern portion of the Land Hemisphere, and the excess of precipitation over evaporation in the middle and southern parts of the Water Hemisphere." But we have no data which warrants the supposition that precipitation exceeds evaporation in the Southern Hemisphere ; and, gene rally speaking, the conditions which regulate precipitation and evaporation would incline us to believe that evaporation ex ceeds precipitation in the Water Hemisphere, and that pre cipitation exceeds evaporation in the Land Hemisphere.