LIV I will at this point revert to Posidonius’ opinion of the cause of 1 thunder. From the earth and its confines are exhaled certain elements, partly moist, partly dry and smoke-like. The latter element remains in the sky as material for lightning, while the former falls in rain. The dry smoky particles that reach the atmosphere will not allow themselves to be enclosed in clouds, but burst their envelope. Thence comes the report which we name thunder. Besides this, anything in the atmosphere itself that is rarefied is at the same time dried and heated up. This also, if it is enclosed, seeks an exit with equal eagerness, and 2 causes a report as it escapes. On one occasion it makes a complete burst, and the thunder is consequently the more violent; on another it escapes by degrees in small portions. Air of this kind, then, by either bursting or flying through the clouds, produces peals of thunder. The rolling of the air enclosed in a cloud is the most potent cause of setting fire to what is struck. LV Thunder is, in short, simply the report of explosions of dry air, 1 which cannot occur unless there is either friction or a rent in a cloud. Posidonius adds that if the clouds merely collide with each other, the kind of blow needed to produce an explosion is given, but not completely; clouds do not meet through their whole extent, but only part with part. And again, soft substances do not resound unless knocked against hard ones; a wave is not heard unless when it beats on the hard shore. But fire, which is soft, says an opponent, when let into water, also a soft substance, produces sound in being extinguished. Well, suppose it is so, it makes for the opposite view 2 which I urge. For it is not really the fire that makes the sound, but the air escaping through the water that is quenching it. Granted that fire is both produced and extinguished in the cloud, it arises from air and friction. Well then, it is urged, may not some of the shooting stars plunge into a cloud and be extinguished? Even supposing that such a thing can and sometimes does occur, it does not remove the difficulty. It is not the occasional chance cause but the natural normal one that we are in search of. Suppose I admit the truth of your contention that occasionally after thunder fires gleam in the heavens much like shooting and falling stars. Yet this does not prove that 3 the thunder was caused by them; it merely shows that the thunder occurred simultaneously with this other phenomenon. Clidemus asserts that a lightning flash is an empty reflection, and not real fire; for in the same way after nightfall a gleam appears from the motion of oars in water. His illustration is not on all fours with the phenomenon. In the latter case the gleam is seen actually within the water; in the former, in the atmosphere, it bursts and leaps out of its element. LVI