The great oceans however cannot be regarded as simple basins of water such as this. The temperature of the surface water varies only approximately with the latitude. It is generally speaking hottest at the equator and coldest at the poles, but surface currents in the intermediate regions produce many irregularities in the surface temperature. Again, although we have no means of knowing what the temperature of the earth is at 1,000 fathoms below the surface of the ocean, it is very probable that in the great oceans the temperature of the deepest stratum of water is considerably lower than the true earth temperature. This is due to currents of cold water constantly flowing from the poles towards the equator. If these polar currents were at any time to cease, the temperature of the lowest strata of water would rise. Although the polar currents cannot be actually demonstrated nor their exact rapidity be accurately determined, the deduction from the known facts of physical geography that they do actually exist is perfectly sound and beyond dispute. A few considerations will, I think, make this clear. If the ocean were a simple basin somewhat deeper at the equator than at the poles, the cold water at the poles would gradually sink down the slopes of the basin towards the latitude of the equator, and the bottom temperature of the water would be constant all the world over. A few hills here and there would not affect the general statement that for a constant depth the temperature of the lowest stratum of water would be constant. But in some places ridges occur stretching across the oceans from continent to continent, and these ridges shut off the cold water at the bottom of the sea on the polar side from reaching the bottom of the sea on the equator side. If A (fig. 1) represents a ridge stretching from continent to continent across an ocean, and the arrow represents the direction of the current, then the water that flows across the ridge from the polar side to the equator side will be drawn from the layers of water lying above the level of the ridge, and consequently none of the coldest water will ever get across it, and from the level of the ridge to the bottom of the sea on the equatorial side the water will have the same temperature as the water at the level of the ridge on the polar side. It follows from this that in places where there are deep holes in the bed of the ocean surrounded on all sides by considerable elevations, the temperature of the water at the bottom will be the same as the temperature of the water on the summit of the lowest ridges that surround them. [Illustration: FIG. 1.—Diagram illustrating the passage of an ocean current across a barrier (A).]