The quantity of the heat excited in this manner is exactly proportional to the force with which the two surfaces are pressed together, and to the rapidity of the friction. When this force was equal to the pressure of ten thousand pounds, and when the rod was turned with such rapidity about its axis that it revolved thirty-two times a minute, the quantity of heat generated by the continual rubbing of the two surfaces together was extraordinarily great. It was equal to the quantity given off by the flame of nine wax-candles of moderate size all burning together. The quantity of heat generated in this manner during a given time is manifestly the same, whether the globular vessel D is filled with water, and the surfaces of the two hemispheres rub on each other in this liquid, or whether there is no water in the vessel, and the apparatus by which the friction is produced is simply surrounded by air. The source of the heat which is generated by this apparatus is inexhaustible. As long as the rod C is turned about its axis, so long will heat be produced by the apparatus, and always to the same amount. If the globe-shaped vessel D is filled with water, this water becomes hotter and hotter, and finally begins to boil. I have myself in this way boiled a considerable quantity of water. If this experiment is performed in winter when the temperature of the air is but little above the freezing-point, and if the vessel D is filled with a mixture of water and pounded ice, the quantity of heat caused in a given time by the rubbing together of the two surfaces can be expressed very exactly by the amount of ice melted by this heat. Since the apparatus affords heat continuously, and always to the same amount, we can melt in this way as much ice as we please. But whence comes this heat? This is the contested point, to determine which was the real aim of the experiment. It is certain that it comes neither from the decomposition of the water nor from the decomposition of the air. Various experiments on this point, which I have described at length in my memoir in the _Philosophical Transactions_, are more than sufficient to establish this fact beyond doubt. Just as little does it come from a change in the capacity for heat brought about by friction in the metal of which the hemispheres are composed. This is shown, first, by the continuance and uniformity of the production of the heat; and, secondly, by an experiment bearing directly on this point, by which I am convinced that not the slightest change had taken place in the capacity of the metal for heat. Just as little does it come from the rods which are attached to the hemispheres, for these rods were always warm, the hemispheres communicating heat to them. Much less could this heat come from the air of the water immediately surrounding the hemispheres, for the apparatus communicated heat to both these fluids without cessation. Whence, then, came this heat? and what is heat actually?