We must not, however, conclude that Nature really performs by the means of water all that we do by fire. The decomposition of every substance is only to be made by division, and the greater this division the more the decomposition will be complete. Fire seems to divide as much as possible those matters which it fuses; nevertheless it may be doubted whether those which water and acids keep in dissolution are not still more divided, and the vapours raised by heat contain matters still further attenuated, in the bowels of the earth, then, by the means of the heat it includes, and the water which insinuates, there is made an infinity of sublimations; distillations, chrystallizations, aggregations, and disjunctions, of every kind. By time all substances may be compounded and decompounded by these means; water may divide and attenuate the parts more than fire when it melts them, and those attenuated parts will join in the same manner as those of fused metal unite by cooling. Crystallization, of which the salts have given us an idea, is never performed but when a substance, being disengaged from every other, is much divided and sustained by a fluid, which having little or no affinity with it, permits it to unite and form by virtue of its force of attraction, masses of a figure nearly similar to its primitive parts. This operation, which supposes all the above circumstances, may be done by the intermediate aid of fire as well as by that of water, and is often accomplished by the concurrence of both, because all this exacts but one division of matter sufficiently great for its primitive parts to be able to form, by uniting figured bodies like themselves. Now fire can bring many substances to this state much better than any other dissolvent, as observation demonstrates to us in asbestos, and other productions of fire, whose figures are regular, and which must be looked upon as true crystallizations. Yet this degree of division, necessary to crystallization, is not the greatest possible, since in this state the small parts of matter are still sufficiently large to constitute a mass, which like other masses, is only obedient to the sole attractive force, and the volumes of which, only touching in points, cannot acquire the resultive force that a much greater division might perform by a more immediate contact, and this is what we see happen in effervescences, where at once, heat and light are produced by the mixture of two cold liquors.