Heat increases its volume and changes the specific gravity of the water, which is dilated or condensed in proportion to the change of temperature. In proportion as it cools, water increases in density, and descends into the depths until it reaches a constant temperature of four degrees twenty-five minutes Cent. below zero, which it preserves in all latitudes at the depth of a thousand yards, according to M. D'Urville. If the water continues to cool, and reaches zero, it becomes lighter than it was at four degrees twenty-five minutes Cent., and ascends in a state of congelation--a process which, by an admirable provision of nature, can only take place at the surface. So long as the temperature is above four degrees twenty-five minutes, water is light, and ascends to the surface, while colder water sinks to the bottom. Below four degrees twenty-five minutes the process is reversed; the first phenomenon is always in force under the Equator, the second near the Poles. The evaporation, which is in continual operation in warm seas, forming vast rain-clouds at the expense of the sea, is compensated by unceasing currents of colder water flowing from the Poles. This evaporation has a direct influence, moreover, on the density of sea water, and is pointed out by Dr. Maury as a remarkable instance of the compensations by which the oceanic waters are governed: "According to Rodgers' observations," he says, "the average specific gravity of sea water on the parallels of thirty-four degrees north and south, at a mean temperature of sixty-four degrees, is just what it ought to be, according to saline and thermal laws; but its specific gravity, when taken from the Equator at a mean temperature of eighty-one degrees, is much greater than, according to the same laws, it ought to be--the observed difference being ·0015, whereas it ought to be ·0025. Let us inquire," he adds, "what makes the equatorial waters so much heavier than they ought to be. "The anomaly occurs in the trade-wind region, and is best developed between the parallel of forty degrees in the North Atlantic and the Equator, where the water grows warmer, but not proportionally lighter. The water sucked up by the trade-winds is fresh water, and the salt it contained, being left behind, is just sufficient to counteract by its weight the effect of thermal dilatation upon the specific gravity of water between the parallels of thirty-four degrees north and south. The thirsting of the trade-winds for vapour is so balanced as to produce perfect compensation, and a more beautiful instance than we have here stumbled upon is not, it appears to me, to be found in the mechanism of the universe."