Gunpowder will burn with great avidity in close vessels, or under an exhausted receiver, and we know that the oxygen is already combined with azote in the nitric acid of the nitrate of potassa, and consequently not in a gaseous but a solid state; yet we also know that a great quantity of caloric and light are emitted during the combustion, and nearly all the products are gaseous. The other anomaly is, that as combustion is produced by pouring nitric acid on spirit of turpentine, the oxygen being already combined with azote, caloric and light are evolved by the mixture of the two fluids, from which it is inferred, that oxygen is capable of giving out caloric and light, not only when liquid, but even after combustion. In the instance of gunpowder, in order to explain the combustion which takes place independently of atmospheric air, or any aeriform supporter, "the caloric and light," in the opinion of Dr. Thomson, (_Chemistry_, i, 128) "must be supposed to be emitted from a solid body during its conversion into gas, which ought to require more caloric and light for its existence in the gaseous state than the solid itself contained."--Mr. Lavoisier (_Elements of Chemistry_, p. 157,) observes, that he and M. De la Place deflagrated a convenient quantity of nitre and charcoal in an ice apparatus, and found that 12 lbs. of ice were melted by the deflagration of one pound of nitre. After giving the proportions of acid and alkali in nitre, and the quantity of oxygen and azote in the acid, he observes, that during the deflagration, 145-1/3 grains of carbon have suffered combustion along wit 3738.34 grains of oxygen; and as 12 lbs. of ice were melted, one pound of oxygen burnt in the same manner would have melted 29.5832 lbs. of ice. To which, if we add the quantity of caloric retained by a pound of oxygen, after combining with carbon to form carbonic acid gas, which was already ascertained to be capable of melting 29.13844 lbs. of ice, we shall have for the total quantity of caloric remaining in a pound of oxygen when combined with nitrous gas in the nitric acid, 58.72164; which is the number of pounds of ice, the caloric remaining in the oxygen in that state is capable of melting. In the state of oxygen gas it contains at least 66.66667. M. Lavoisier infers then, that the oxygen in combining with azote to form nitric acid, only loses 7.94502, and that "this enormous quantity of caloric, retained by oxygen in its combination into nitric acid, explains the cause of the great disengagement of caloric during the deflagration of nitre; or, more strictly speaking, upon all occasions of the decomposition of nitric acid." This view of the subject may enable us to explain the production of caloric, in those cases of combustion which cannot be explained on the ordinary principles; and, with regard to gunpowder, the accension of oil of turpentine by nitric acid, and similar cases, we may conclude, as the