Berthollet had thrown doubt on the universality of Lavoisier's name "oxygen," the acidifier, but he had not conclusively proved the existence of any acid which did not contain oxygen.
The researches of Davy naturally led him to consider the prevalent views regarding acids, bases and salts.
Muriatic (or as we now call it hydrochloric) acid had long been a stumbling-block to the thorough-going Lavoisierian chemists. Oxygen could not be detected in it, yet it ought to contain oxygen, because oxygen is the acidifier. Of course, if muriatic acid contains oxygen, the salts—muriates—produced by the action of this acid on alkalis and earths must also contain oxygen. Many years before this time the action of muriatic acid on manganese ore had been studied by the Swedish chemist Scheele, who had thus obtained a yellow-coloured gas with a very strong smell. Berthollet had shown that when a solution of this gas in water is exposed to sunlight, oxygen is evolved and muriatic acid is produced. The yellow gas was therefore supposed to be, and was called, "oxidized muriatic acid," and muriatic acid was itself regarded as composed of oxygen and an unknown substance or radicle.
In 1809 Gay-Lussac and Thenard found that one volume of hydrogen united with one volume of the so-called oxidized muriatic acid to form muriatic acid; the presence of hydrogen in this acid was therefore proved.
When Davy began (1810-11) to turn his attention specially to the study of salts, he adopted the generally accepted view that muriatic acid is a compound of oxygen and an unknown radicle, and that by the addition of oxygen to this compound oxidized muriatic acid is produced. But unless Davy could prove the presence of oxygen in muriatic acid he could not long hold the opinion that oxygen was really a constituent of this substance. He tried to obtain direct evidence of the presence of oxygen, but failed. He then set about comparing the action of muriatic acid on metals and metallic oxides with the action of the so-called oxidized muriatic acid on the same substances. He showed that salt-like compounds were produced by the action of oxidized muriatic acid either on metals or on the oxides of these metals, oxygen being evolved in the latter cases; and that the same compounds and water were produced by the action of muriatic acid on the same metallic oxides.
These results were most easily and readily explained by assuming the so-called oxidized muriatic acid to be an elementary substance, and muriatic acid to be a compound of this element with hydrogen. To the new element thus discovered—for he who establishes the elementary nature of a substance may almost be regarded as its discoverer—Davy gave the name of chlorine, suggested by the yellow colour of the gas (from Greek, = yellow). He at once began to study the analogies of chlorine, to find by experiment which elements it resembled, and so to classify it. Many metals, he found, combined readily with chlorine, with evolution of heat and light. It acted, like oxygen, as a supporter of combustion; it was, like oxygen, attracted towards the negative pole of the voltaic battery; its compound with hydrogen was an acid; hence said Davy chlorine, like oxygen, is a supporter of combustion and also an acidifier.
But it was very hard to get chemists to adopt these views. As Bacon says, "If false facts in Nature be once on foot, what through neglect of examination, the countenance of antiquity, and the use made of them in discourse, they are scarce ever retracted."
Chemists had long been accustomed to systems which pretended to explain all chemical facts. The phlogistic theory, which had tyrannized over chemistry, had been succeeded by the Lavoisierian chemistry, which recognized one acidifier, and this also the one supporter of combustion. To ascribe these properties to any element other than oxygen appeared almost profane.
But when Davy spoke of chlorine as an acidifier, he did not use this word in the same sense as that in which it was employed by the upholders of the oxygen theory of acids; he simply meant to express the fact that a compound containing chlorine as one of its constituents, but not containing oxygen, was a true acid. When Gay-Lussac attempted to prove that hydrogen is an alkalizing principle, Davy said, "This is an attempt to introduce into chemistry a doctrine of occult qualities, and to refer to some mysterious and inexplicable energy what must depend upon a peculiar corpuscular arrangement." And with regard to Gay-Lussac's strained use of analogies between hydrogen compounds and alkalis, he says, "The substitution of analogy for fact is the bane of chemical philosophy; the legitimate use of analogy is to connect facts together, and to guide to new experiments."
But Davy's facts were so well established, and his experiments so convincing, that before two or three years had passed, most chemists were persuaded that chlorine was an element—i.e. a substance which had never been decomposed—and that muriatic acid was a compound of this element with hydrogen.