This seems to indicate that the most essential difference between combustible matters and those which are not so, consists in the latter containing only a few or none of the light, ethereal, and oily matters susceptible of an expansive motion, or, at least, if they contain them, that they are fixed, so that they cannot exercise their volatility whenever the force of the fire is not strong enough to surmount the force of adhesion which retains them united to the fixed parts of matter. It may be said that this induction is confirmed by a number of observations well known to chemists; but what appears to be less so, and which, nevertheless, is a necessary consequence of it, is, that all matter may become volatile when the expansive force of the fire can be rendered superior to the attractive force which holds the parts of matter united; for though to produce a fire sufficiently strong it may require better constructed mirrors than any at present known, yet we are certain that fixity is only a relative quality, and that there is no matter absolutely so, since heat dilates the most fixed bodies. Now is not this dilation the index of a beginning separation, that may be augmented with a degree of heat to fusion, and with a still greater heat to volatilisation? Combustion supposes something more than volatilisation; it is not sufficient that the parts of matter be sufficiently separated to be carried off by those of heat; they must also be of an analogous nature to fire; without that, mercury, being the most fluid next to air, would also be the most combustible, whereas experience demonstrates, that though very volatile it is not combustible. Matter is, in general, composed of four principal substances, called _elements_, that is, earth, water, air, and fire. Those in which earth and water predominate will be fixed, and will only become volatile by the action of heat; and those which contain most air and fire will be the only real combustibles. The great difficulty here is clearly to conceive how air and fire, both so volatile, can fix and become constituent parts of all bodies. Fire, by absorbing air, destroys the spring. Now there are but two methods of destroying a spring, either by compressing it till it breaks, or extending it till it loses its effect. It is plain that fire cannot destroy air by compression, since the least degree of heat rarefies it; on the contrary, by a very strong heat the rarefaction of the air will be so great that it will occupy a space thirteen times more extended than that of its general volume; and by this means the spring becomes weakened, and it is in this state that it can become fixed, and unite with other bodies.