from the equatorial regions. On the other hand, if the water be too cool for the latitude, then the inference is that it has lost its heat in colder climates, and therefore is found in channels which lead from the polar regions. The arrow-heads point to the direction in which the waters are supposed to flow. Their rate, according to the best information that I have obtained, is, at a mean, only about four knots a day—rather less than more. Accordingly, therefore, as the immense volume of water in the antarctic regions is cooled down, it commences to flow north. As indicated by the arrow-heads, it strikes against Cape Horn; and is divided by the continent, one portion going along the west coast as Humboldt's Current (§ 398); the other, entering the South Atlantic, flows up into the Gulf of Guinea, on the coast of Africa. Now, as the waters of this polar flow approach the torrid zone, they grow warmer and warmer, and finally themselves become tropical in their temperature. They do not then, it may be supposed, stop their flow; on the contrary, they keep moving, for the very cause which brought them from the extra-tropical regions now operates to send them back. This cause is to be found in the difference of the specific gravity at the two places. If, for instance, these waters, when they commence their flow from the hyperborean regions, were at 30°, their specific gravity will correspond to that of sea water at 30°. But when they arrive in the Gulf of Guinea or the Bay of Panama, having risen by the way to 80°, or perhaps 85°, their specific gravity becomes such as is due to sea water of this temperature; and, since fluids differing in specific gravity can no more balance each other on the same level than can unequal weights in the opposite scales of a true balance, this hot water must now return to restore that equilibrium which it has destroyed in the sea by rising from 30° to 80° or 85°. Hence it will be perceived that these masses of water which are marked as cold are not always cold. They gradually pass into warm; for in travelling from the poles to the equator they partake of the temperature of the latitudes through which they flow, and grow warm. Plate IX., therefore, is only introduced to give general ideas; nevertheless, it is very instructive. See how the influx of cold water into the South Atlantic appears to divide the warm water, and squeeze it out at the sides, along the coasts of South Africa and Brazil. So, too, in the North Indian Ocean, the cold water again compelling the warm to escape along the land at the sides, as well as occasionally
