mixing with either the colder or the warmer parts, and also because water conducts heat very badly.
The phenomena are different in the case of sea-water, and also complicated in other ways. The point of maximum density descends as the weight of the salt-water and its richness in dissolved matter increase. The Swedish chemist and hydrographer, Ekman, after long series of experiments relative to this question, has found that this critical temperature may fall to -4° C. (25° Fahr.) in Atlantic water. The properties of a brackish fluid, such as would be drawn from a fiord, would naturally be intermediate between those of a pure and those of a very salt water. Hence the depths of the ocean can not be at 39° Fahr., as some authors still maintain. A slight excess of salt in solution will weight a stratum of water of mean temperature, whereby a cold zone may be superposed upon another zone which is warmer but more saline. The interior of the ocean, as well as its surface, is plowed by numerous currents, some warm, some cold, which meet, mix, and separate again, so that it is very hard to find out by reasoning what experiment alone can teach. A similar variety is shown in the density of water brought up by soundings. The complication is magnified when we reflect that water is not absolutely incompressible, that each thickness of ten metres exercises a vertical pressure nearly equal to an atmosphere, the action of which added to that of the superior parts weighs upon the inferior liquid, so that at about 4,000 metres the pressure is 400 atmospheres. Water must be extremely dense when it it compressed with so much force, and the influence of salinity and temperature must become very small in these unfathomable abysses. The question of submarine temperatures has given rise to many controversies. Some, with Perron, suppose that the great depths are always cold, like the tops of the highest mountains. On the other extreme, the author of "Epochs of Nature" attributed to the oceanic depths a high temperature on account of their nearness to the central fire. Denis de Montfort and Humboldt are of the opinion that below the superficial parts there prevails a constant temperature, peculiar to each station, and corresponding with the mean annual temperature of the place. This view is correct for regions where the depth is not very great, and in certain bodies of water.
The sea, on account of its great specific heat and its feeble conducting power, plays the part of a moderator of temperature something like that of the fly-wheel of an engine as a moderator of force. In winter it is warmer, in summer it is cooler, than the ambient air, and the difference is emphasized the farther we get away from the shore.
In "The Clouds" of Aristophanes, Strepsiades refuses to pay his creditors who hold that the level of the sea is fixed, believing that, as it receives all the water, it must continue to rise indefinitely. The phenomena of evaporation were not very well understood at that time. Even in the seventeenth century. Father Fournier talked of subterra-
