Many other experiments might be mentioned, in which the same effects take place, and from which the same conclusions may be drawn. But with respect to the _effect_, whether it be dull, brilliant, or very brilliant, depends more on the quality of the metal, than perhaps, on its subsequent mixture with the other materials. Crude iron, usually called cast iron, seems to possess this property in an eminent degree; but in the experiment with oxygen gas, steel is always preferable, as the combustion is more rapid, and the effect more striking. The difference, which we will not attempt to explain, may depend on the _state_, as well as the _proportion_ of carbon, which enters into crude iron, as well as steel. In one case, the combustion ensues in contact with nitre, and in atmospheric air; in the other, in contact only with oxygen gas. Be this as it may, this inference is conclusive, that, in all cases of the combustion of iron in fire-works, the metal itself unites with oxygen, and the result of the combustion is an oxide of iron; and with respect to the carbon, in both instances, it is converted alike into carbonic acid. So that whether the iron receives its oxygen from the nitre, or from the air, or from both, is immaterial, as the products are the same. When iron is exposed to the atmosphere, it tarnishes, and is gradually changed into a brown or yellow powder, called rust. This change is owing to its combination with oxygen; and its affinity for oxygen is such, that, when the vapour of water is made to pass through an ignited gun-barrel, it is decomposed, the metal becoming oxidized, and the hydrogen, the other constituent of the water, being liberated in the form of gas.