The fact that the temperature of the earth increases with the depth necessarily involves a continual loss of heat from its interior by conduction outward into and through the upper crust, according to a well-known law of equilibrium of temperatures. "Hence, since the upper crust does not become hotter from year to year, there must be a secular loss of heat from the earth."[147] Thus it appears that from the surface of the earth and the ocean, from thermal springs, and from three hundred active volcanoes, the internal heat of the globe is incessantly radiated into space and is practically lost. Now this average loss of heat may be at least approximately measured, and data are thereby furnished for determining the probable _age_ of the earth, or, perhaps more correctly, its phase of life. If a man were to find a hot ball of iron suspended in a room, and if he were carefully to observe the distribution of heat in the ball, he would be able easily to determine whether the ball were becoming hotter or cooler. If he found that the inside were hotter than the outside, he would conclude that the ball was cooling, and had therefore been hotter than when he found it. So far common-sense would be his guide; but with the aid of mathematics, and some knowledge of the physical properties of iron and air, he could go much further, and be able to calculate how hot the ball must have been at any given moment, if it had not been interfered with. Thus he would be able to say, the ball must have been hung up less than, say, five hours ago, for at that time the heat of the metal would have been such that it would have been in a state of fusion, and hence not capable of hanging as a solid mass. Precisely analogous reasoning holds with regard to the earth: it is such a ball; it is hotter inside than outside. The distribution of the heat near its surface is approximately known--1° Fahr. of elevation in temperature for 50 British feet of descent.[148] The properties of the matter of which it is composed are approximately known. The temperature at which granite rocks are fusible has been found to be about 7000° Fahr. This must therefore have been the temperature of the earth in its primitive igneous condition. From these data, Sir William Thomson has, by rigid mathematical calculations, reached the conclusion that the consolidation of the earth's crust commenced 98,000,000 years ago.[149] The rates of increase of temperature _inward_ in a great amount of average rock at various periods after the commencement of cooling, from the primitive heat of 7000° Fahr., are estimated by Sir William Thomson as follows: "At 10,000 y'rs after commencement of cooling should have 2° per ft. At 40,000 " " " " 1° " At 160,000 " " " " 1/2° " At 4,000,000 " " " " 1/10° " At 100,000,000 " " " " 1/50° "