In this class of researches thermometers à déversement are used, the reservoirs of which overflow as the temperature rises; the mercury remaining in the ball shows the maximum attained. Walferdin's registering thermometer and the geothermometer of Magnus are constructed on this principle. Thermometers à minima, of a different construction, are used to determine the temperature of the ocean-depths, which are generally colder than the water at the surface. The many soundings made by the English scientific expeditions established beyond a doubt the fact that the temperature at the bottom of the sea is often but little above zero. This would be explained by supposing the colder water to be carried to the bottom by its specific gravity, the water warmed and dilated by the sun's heat remaining at the surface. The bed of the ocean at large, where the normal temperature is not affected by warm currents, such as the Gulf Stream, may be said to be covered with water at the freezing-point. The water at the bottom of fresh-water lakes is less cold because the maximum density of fresh water is 4°. It results from this that the portions possessing this temperature are carried to the bottom, while the colder or warmer portions rise to the surface. Thus, that portion of the earth's shell that is covered by water remains at a relatively low temperature, in consequence of the stratification resulting from the varying densities of the liquid, but, if it were possible to carry on in the bed of the sea such investigations as have been made on land, an increase of temperature, such as has been proved to exist in the frozen soil of Siberia, would doubtless be found.
The increase in heat as we descend is generally admitted to average 1° in 30 metres. If this rate is constant it is clear that, at a depth of 2,700 metres, the temperature must equal that of boiling water; and that, at a depth of 50 kilometres, the heat must exceed 1,600°, a point at which iron and the greater part of the rocks would melt. This is the principal ground for the argument of those who maintain that the earth's crust is not more than 40 to 50 kilometres thick—or, relatively to its size, of the thickness of an egg-shell compared with the egg. Certain it is that the increase of heat with the depth, confirmed by so many observers, perforce gives a warrant to the idea of a subterranean fire possessing an inconceivable degree of heat; but the question is, At what depth from the surface does this fire exist?
The thermometric observations thus far made are insufficient to decide this question. Among the mines that have reached a great depth may be mentioned those of Kitzbühel in the Tyrol (900 metres); Kutteuberg in Bohemia (1,200 metres); Mouille-Louge (920 metres); and Speremberg (1,260 metres). Why may not borings be made at the bottom of some of these very deep mines, by means of which the bowels of the earth can be still further penetrated?
It is also desirable that the natural cavities in the earth should be utilized for scientific investigation. The accounts contained in the
