The alchemists thought that the most effectual method of separating a complex substance into more simple substances was to subject it to the action of heat. They were constantly distilling, incinerating, subliming, heating, in order that the spirit, or inner kernel of things, might be obtained. They took for granted that the action of fire was to simplify, and that simplification proceeded whatever might be the nature of the substance which was subjected to this action. Boyle insisted that the effect of heating one substance may be, and often is, essentially different from the effect of heating another substance; and that the behaviour of the same substance when heated, sometimes varies when the conditions are changed. He takes the example of heating sulphur or brimstone: "Exposed to a moderate fire in subliming pots, it rises all into dry, and almost tasteless, flowers; whereas being exposed to a naked fire, it affords store of a saline and fretting liquor." Boyle thought that the action of fire was not necessarily to separate a thing into its principles or elements, but, in most cases, was either to rearrange the parts of the thing, so that new, and it might be, more complex things, were produced, or to form less simple things by the union of the substance with what he called, "the matter of fire." When the product of heating a substance, for example, tin or lead, weighed more than the substance itself, Boyle supposed that the gain in weight was often caused by the "matter of fire" adding itself to the substance which was heated. He commended to the investigation of philosophers this "subtil fluid," which is "able to pierce into the compact and solid bodies of metals, and add something to them that has no despicable weight upon the balance, and is able for a considerable time to continue fixed in the fire." Boyle also drew attention to the possibility of action taking place between a substance which is heated and some other substance, wherewith the original thing may have been mixed. In a word, Boyle showed that the alchemical assumption - fire simplifies - was too simple; and he taught, by precept and example, that the only way of discovering what the action of fire is, on this substance or on that, is to make accurate experiments. "I consider," he says, "that, generally speaking, to render a reason of an effect or phenomenon, is to deduce it from something else in nature more known than itself; and that consequently there may be divers kinds of degrees of explication of the same thing."

Boyle published his experiments and opinions concerning the action of fire on different substances in the seventies of the 17th century; Stahl's books, which laid the foundation of the phlogistic theory, and confirmed the alchemical opinion that the action of fire is essentially a simplifying action, were published about forty years later. But fifty years before Boyle, a French physician, named Jean Rey, had noticed that the calcination of a metal is the production of a more complex, from a less complex substance; and had assigned the increase in weight which accompanies that operation to the attachment of particles of the air to the metal. A few years before the publication of Boyle's work, from which I have quoted, John Mayow, student of Oxford, recounted experiments which led to the conclusion that the air contains two substances, one of which supports combustion and the breathing of animals, while the other extinguishes fire. Mayow called the active component of the atmosphere fiery air; but he was unable to say definitely what becomes of this fiery air when a substance is burnt, although he thought that, in some cases, it probably attaches itself to the burning substances, by which, therefore, it may be said to be fixed. Mayow proved that the air wherein a substance is burnt, or an animal breathes, diminishes in volume during the burning, or the breathing. He tried, without much success, to restore to air that part of it which disappears when combustion, or respiration, proceeds in it.

What happens when a substance is burnt in the air? The alchemists answered this question by asserting that the substance is separated or analysed into things simpler than itself. Boyle said: the process is not necessarily a simplification; it may be, and certainly sometimes is, the formation of something more complicated than the original substance, and when this happens, the process often consists in the fixation of "the matter of fire" by the burning substance. Rey said: calcination, of a metal at anyrate, probably consists in the fixation of particles of air by the substance which is calcined. Mayow answered the question by asserting, on the ground of the results of his experiments, that the substance which is being calcined lays hold of a particular constituent of the air, not the air as a whole.

Now, it is evident that if Mayow's answer was a true description of the process of calcination, or combustion, it should be possible to separate the calcined substance into two different things, one of which would be the thing which was calcined, and the other would be that constituent of the air which had united with the burning, or calcining, substance. It seems clear to us that the one method of proving the accuracy of Mayow's supposition must be, to weigh a definite, combustible, substance - say, a metal; to calcine this in a measured quantity of air; to weigh the product, and to measure the quantity of air which remains; to separate the product of calcination into the original metal, and a kind of air or gas; to prove that the metal thus obtained is the same, and has the same weight, as the metal which was calcined; and to prove that the air or gas obtained from the calcined metal is the same, both in quality and quantity, as the air which disappeared in the process of calcination.