All nitrates, as well as the different hyperoxymuriates, or chlorates, contain oxygen as an essential ingredient in the acid of their respective salts, which is readily given up to inflammable substances. When nitrates are employed for fire-works, they should be free from moisture, or water of crystallization, unless otherwise required. The presence of water may, in many cases, prove injurious to the composition; and, consequently, the effect in those instances, may be influenced by this circumstance. The composition of nitric acid, and the action of carbon in the decomposition of the nitrates, or salts formed by the union of nitric acid with sundry bases, will claim our attention in the article on gun-powder. With respect to the production of colours, some remarks on this subject may be here added. Speaking of colours, Haüy (_Elementary Treatise of Natural Philosophy_, trans. ii. p. 253.) takes into view their formation according to the Newtonian doctrine; and in a note by the translator, several instances are given of the change of colour by oxidizement and other processes. Iron when exposed to heat in contact with atmospheric air gradually absorbs oxygen, and changes its colour. The colours produced depend entirely on the quantity of oxygen, and on the absorption of some of the rays of light, and the reflection of others. See _Iron_. The tempering of steel instruments depends on this property, and also the blueing of sword blades, and many similar operations. The first impression of fire usually developes a blue colour; a second degree produces a yellow; and, if the oxidizement augments, the iron becomes red. The major part of the metals present similar phenomena. In vegetables, the blue colour is formed by fermentation; and many of these colours are susceptible of passing to red by a greater quantity of oxygen, as they depend on the absorption of oxygen. It is thus that the green fecula of indigo becomes blue; turnsol, red by air and acids; and the protoferrocyanate of iron, blue when exposed to the air. When meat putrefies, the first degree of oxygenation decides the blue colour; the red soon succeeds as the process goes on. It would seem that the maximum of oxidation determines the reflection of rays of every kind, in the same proportions as subsist in solar light, of which we have many instances in combustion. The flame of burning bodies exhibits the same phenomena. It is blue when the combination of oxygen is slow; red when it is stronger, and white when the oxygenation is complete. These facts lead to the conclusion, that the combination of oxygen, and its proportions, give birth in bodies to the property of reflecting corresponding rays of light; but, since the combination of oxygen in different proportions ought to change the thickness and density of the component laminæ, and, consequently, to produce variations in the colours, this doctrine is not easily reconciled with the received theory.