Perchance, too, when the wind only blows softly and exerts no great 1 force, the clouds, wafted against each other, will emit fire strong enough to show a gleam, though not to issue from them. Less force is required for lightning than for the thunderbolt. We found above what a glow the friction of certain woods caused. Now when the air, which is interchangeable with fire, [has been changed in full force into fire 2 and][45] undergoes friction, it is credible and even probable that fire is struck out, but of an evanescent and transitory character, as it arises from no solid material and has no fuel in which it can lodge. It therefore quickly passes; its duration is no longer than its route and course; it has nothing to support it when hurled forth into space. [45] These words seem of more than doubtful genuineness. XXIV But how, you ask me, when you philosophers say that it is the nature 1 of fire to rise, does the bolt seek the earth? Perhaps what you said about fire is not true? It seems to take its course down as well as up. Both my statements, I reply, may be true. Fire naturally does rise and mounts if nothing prevents it, just as water naturally gravitates downwards. But water if affected by a force which drives it uphill is pressed up in the direction from which it was precipitated in rain. In like manner the same force as launched the bolt from the cloud causes it to fall to the ground. Something of the same kind happens to these 2 celestial fires as to trees when bent. The topmost branches if slender may be dragged down so as to touch the ground; but when you let them go, they rebound to their original position. You must not regard the condition which an object involuntarily assumes as characteristic of it. If you allow fire to go where it will, it will return to the sky, the abode of all the lightest bodies. But when there is anything to carry it down and divert it from its natural course, that is not a mark of its disposition but a token of its subjection. XXV You and your friends say, an objector interposes, that clouds emit fire through mutual friction when they are moist, indeed wet. How can such clouds produce fire, which is no more likely to be generated by a cloud than by pure water? XXVI