Pulverizing the charcoal, in the first instance, by itself, and adding water, during its mixture, from time to time, a measure proposed in 1808, by M. David, and now generally adopted, will prevent such accidents; for it appears, they have not occurred in France, since the adoption of this plan. Some remarks on spontaneous combustion, may be seen in the article on _artificial volcanoes_. M. Sage, (_Journal de Physique_, vol. lxv, p. 423, or _Nicholson's Journal_, vol. xxiii, p. 277), has written on the spontaneous ignition of charcoal, and adduced some facts on the subject; by which it appears, that M. de Caussigni was the first, who observed, that charcoal was capable of being set on fire, by the pressure of mill stones. Mr. Robin, commissary of the powder mills of Essonne, has given an account, in the _Annales de Chimie_, of the spontaneous inflammation of charcoal, from the black berry bearing alder, that took place the 23d of May, 1801, in the box of the bolter, into which it had been sifted. This charcoal, made two days before, had been ground in the mill, without showing any signs of ignition. The coarse powder, that remained in the bolter, experienced no alteration. The light undulating flame, unextinguishable by water, that appeared on the surface of the sifted charcoal, was of the nature of inflammable gas, which is equally unextinguishable.[17] The moisture of the atmosphere, of which fresh made charcoal is very greedy, appears to have concurred in the development of the inflammable gas, and the combustion of the charcoal. It has been observed, that charcoal powdered and laid in large heaps, heats strongly. Alder charcoal has been seen to take fire in the warehouses, in which it has been stored. About thirty years ago, M. Sage saw the roof of one of the low wings of the mint set on fire by the spontaneous combustion of a large quantity of charcoal, that had been laid in the garrets. Mr. Malet, commissary of gunpowder at Pontailler, near Dijon, has seen charcoal take fire under the pestle. He also found, that when pieces of saltpetre and brimstone were put into the charcoal mortar, the explosion took place between the fifth and sixth strokes of the pestle. The weight of the pestles is eighty pounds each, half of this belonging to the box of rounded bell metal, in which they terminate. The pestles are raised only one foot, and make forty-five strokes in a minute. "In consequence of the precaution now taken," M. Sage observes, "to pound the charcoal, brimstone, and saltpetre separately, no explosions take place; and time is gained in the fabrication, since the paste is made in eight hours, that formerly required four and-twenty.