For nearly three months the craters, of which there were two, were in activity. The larger one threw up showers of ignited ashes and stones to a height apparently of from two hundred to two hundred and fifty feet; and from the smaller crater steam arose with great violence. Whenever the crater could be approached, it was found incrusted with saline matter: and the walk to the edge of the small crater, on the 6th of January, was through a mass of loose saline matter, principally common salt coloured by muriate of iron, in which the foot sunk to some depth. It was easy, even at a great distance, to distinguish between the steam disengaged by one of the craters, and the earthy matter thrown up by the other. The steam appeared white in the day, and formed perfectly white clouds, which reflected the morning and evening light of the purest tints of red and orange. The earthy matter always appeared as a black smoke, forming dark clouds, and in the night it was highly luminous at the moment of the explosion. He concludes this paper on Volcanoes with some observations on the theory of their phenomena. "It appears," says he, "almost demonstrable, that none of the chemical causes anciently assigned for volcanic fires can be true. Amongst these, the combustion of mineral coal is one of the most current; but it seems wholly inadequate to account for the phenomena. However large the stratum of pit-coal, its combustion under the surface could never produce violent and excessive heat; for the production of carbonic acid gas, when there was no free circulation of air, must tend constantly to impede the process: and it is scarcely possible that carbonaceous matter, if such a cause existed, should not be found in the lava, and be disengaged with the saline or aqueous products from the bocca or craters. There are many instances in England of strata of mineral coal which have been long burning; but the results have been merely baked clay and schists, and it has produced no result similar to lava.