Thus, when a solid melts, or when a liquid is rendered gaseous, a considerable amount of heat is spent in the process, which does not become sensible, that is to say, does not affect the thermometer. Thus, in order to melt a kilogramme of ice, heat is required sufficient to raise a kilogramme of water through 80° C., and yet, when melted, the water is no warmer than the ice. We express this fact by saying that the latent heat of water is 80. Again, if a kilogramme of water at 100° be converted entirely into steam, as much heat is required as would raise the water through 537° C., or 537 kilogrammes of water through one degree; but yet the steam is no hotter than the water, and we express this fact by saying that the latent heat of steam is 537. Now, in both of these instances it is at least extremely probable that a large portion of the heat is spent in doing work against the force of cohesion; and, more especially, when a fluid is converted into a gas, we know that the molecules are in that process separated so far from one another as to lose entirely any trace of mutual force. We may, therefore, conclude that although in most cases the greater portion of the heat applied to a body is spent in increasing its molecular motion, and only a small part in doing work against cohesion, yet when a solid melts, or a liquid vaporizes, a large portion of the heat required is not improbably spent in doing work against molecular forces. But the energy, though spent, is not lost, for when the liquid again freezes, or when the vapour again condenses, this energy is once more transformed into the shape of sensible heat, just as when a stone is dropped from the top of a house, its energy of position is transformed once more into actual energy. 75. A single instance will suffice to give our readers a notion of the strength of molecular forces. If a bar of wrought iron, whose temperature is 10° C. above that of the surrounding medium, be tightly secured at its extremities, it will draw these together with a force of at least one ton for each square inch of section. In some cases where a building has shown signs of bulging outwards, iron bars have been placed across it, and secured while in a heated state to the walls. On cooling, the iron contracted with great force, and the walls were thereby pulled together. 76. We are next brought to consider atomic forces, or those which lead to chemical union, and now let us see how these are influenced by heat. We have seen that heat causes a separation between the molecules of a body, that is to say, it increases the distance between two contiguous molecules, but we must not suppose that, meanwhile, the molecules themselves are left unaltered.