He was a firm believer in the existence of phlogiston; but he seems, at least ultimately, to have adopted the view of Scheele, and many other eminent contemporary chemists--indeed, the view of Cavendish himself--that hydrogen gas is phlogiston in a separate and pure state. Common air he considered as a compound of oxygen and phlogiston. Oxygen, in his opinion, was air quite free from phlogiston, or air in a simple and pure state; while _azotic gas_ (the other constituent of common air) was air saturated with phlogiston. Hence he called oxygen _dephlogisticated_, and azote _phlogisticated air_. The facts that when common air is converted into azotic gas its bulk is diminished about one-fifth part, and that azotic gas is lighter than common air or oxygen gas, though not quite unknown to him, do not seem to have drawn much of his attention. He was not accustomed to use a balance in his experiments, nor to attend much to the alterations which took place in the weight of bodies. Had he done so, most of his theoretical opinions would have fallen to the ground. When a body is allowed to burn in a given quantity of common air, it is known that the quality of the common air is deteriorated; it becomes, in his language, more phlogisticated. This, in his opinion, was owing to an affinity which existed between phlogiston and air. The presence of air is necessary to combustion, in consequence of the affinity which it has for phlogiston. It draws phlogiston out of the burning body, in order to combine with it. When a given bulk of air is saturated with phlogiston, it is converted into azotic gas, or _phlogisticated air_, as he called it; and this air, having no longer any affinity for phlogiston, can no longer attract that principle, and consequently combustion cannot go on in such air. All combustible bodies, in his opinion, contain hydrogen. Of course the metals contain it as a constituent. The calces of metals are those bodies deprived of phlogiston. To prove the truth of this opinion, he showed that when the oxide of iron is heated in hydrogen gas, that gas is absorbed, while the calx is reduced to the metallic state. Finery cinder, which he employed in these experiments, is, in his opinion, iron not quite free from phlogiston. Hence it still retains a quantity of hydrogen. To prove this, he mixed together finery cinder and carbonates of lime, barytes and strontian, and exposed the mixture to a strong heat; and by this process obtained inflammable gas in abundance. In his opinion every inflammable gas contains hydrogen in abundance. Hence this experiment was adduced by him as a demonstration that hydrogen is a constituent of finery cinder.