We thus perceive how analysis bears upon the case. We can see by such reasoning on the subject, that, theoretically, there can be but _one set of proportions calculated to produce the best and strongest gunpowder_, and that those proportions must depend upon the established and unerring laws of nature. The proportions, then, for gunpowder, by these considerations, will be those in which the carbon will just consume the oxygen of the nitre, and combine with the sulphur as much as will exactly saturate the potassium. This will be effected by an atom each of nitre and sulphur, and three atoms of carbon; or nitre 75·5, charcoal 18·8, and of sulphur 11·8. In the present improved state of chemical science, when the nature of the bodies comprising gunpowder is so well known, as well as the compounds resulting from their action on each other, the proportions we have named may be taken as the best for practice. The charcoal should, in particular, not be less than the nitre, as the smallest portion less than the whole atom would be the same as to leave out the whole atom, in which case there would be no carbonic oxide formed. If, for example, instead of the proportions of nitre 75·5, charcoal 16·2, sulphur 15, the carbon were 16, then there would be 4·2 of carbon left in the residuum, and no carbonic oxide would be formed, since bodies cannot unite but in definite proportions. From these considerations we can perceive the reason why a small proportion of carbonic oxide is always formed during the decomposition of nitre by charcoal; for it will be evident, that as the nitric acid contains five atoms of oxygen, four of these must combine with two atoms of carbon to form two atoms of carbonic acid, while the _odd atom of oxygen_ is compelled to take another atom to form carbonic oxide. But this is not the case in the combustion of gunpowder, as carbonic acid and nitrogen are the principal gases generated. These proportions differ from any other formula yet prescribed; and, though different in a great degree from the proportions laid down by various writers on the subject, the reasons which are here given, as has been seen, are such as carry with them a conviction of their truth: for there cannot possibly be any benefit arising from a greater quantity of any of these materials than is absolutely necessary to form the composition in question; and if the smallest quantity be above what is requisite to consume the whole, that, however small it may be, is highly detrimental to the effective energy of the mass. What we may here call clean gunpowder, such as may be used with confidence for repeated discharges of fire-arms of any description, is of the greatest importance; therefore, it does not appear to us, that any given proportions are so likely to accomplish that object as those before specified. TABLE OF COMPOSITION OF DIFFERENT GUNPOWDERS.