They are generally covered over with a thin crust, which is quite black, and have a very rough surface. Internally, they are greyish and of a granulated texture. By the help of a microscope, you may distinguish roundish grains of a grey colour; others, like rusty iron; some, angular pieces of perfectly metallic iron, which are attracted by the magnet; and the rest is an earthy sort of cement in which the others are embedded. Their chemical composition is very uniform, with the exception of some that consist almost entirely of iron; of which sort is the one represented on the preceding page. Their descent appears to be quite independent of the state of the atmosphere, from whence we may infer that the clouds have nothing at all to do with them, but that they come from much higher regions. When they have been found soon after falling, they have always been extremely hot, and, as I told you of one in India, they have been known to set on fire what they have come in contact with on the surface of the earth. There have been four schemes devised to account for the existence of aerolites. It has been imagined that they were substances which had been cast out by volcanoes to immense heights from distant parts of the world; but this is disproved by the fact that no substance of the same composition as aerolites, has ever been discovered amongst the known products of volcanoes. The celebrated Frenchman, La Place, thought they were substances that had been cast out by volcanoes in the moon, with such violence as to send them within the limits of the earth's attraction. Some have conjectured that they are formed in the air by the consolidation of clouds of gaseous matters exhaled from the earth; but according to Sir H. Davy's view of the nature of flame, the light of meteors must arise from the ignition of _solid_ bodies, so that at least they must become solid while they are in the condition of meteors, and long enough before they approach the earth. I will tell you what I think, according to the present state of our knowledge, the likeliest explanation, though I do not say that the reasons in favour of it are very conclusive. You will remember what I told you respecting the probability of falling stars being fragments of matter revolving in orbits, which the earth at certain times comes near in its annual course round the sun. I suppose shooting stars and aerolites to be the same things, only they are shooting stars while they show in the sky, and aerolites after they have reached the earth. [Illustration] CHAPTER VI. BLOODY RAIN.