Chemistry and electricity give a ready answer to this latter question. Molecules are composites of still smaller bodies, and to get back to the ultimate particle we must go to atoms. All chemical changes resolve themselves into the breaking up of molecules and rearrangement of their constituent atoms. If the opposite poles of a voltaic battery are inserted in a vessel containing water, molecules of water are broken up, bubbles of gas rise at each pole, and if these are collected, the gas at the positive pole is found to be oxygen, and that at the negative pole hydrogen. Nothing has been added or taken away, for the weight of the two gases evolved exactly equals that of the water which has disappeared. But the molecules of the water have been broken up, and their constituents reappear in totally different forms, for nothing can well be more unlike water than each of the two gases of which it is composed. That it is composed of them can be verified by the reverse experiment of mixing the two gases together in the same proportion of two volumes of hydrogen to one of oxygen as was produced by the decomposition of water, passing an electric spark through the vessel containing the mixture, when with a loud explosion the gases reunite, and water is formed in precisely the same quantity as produced the volumes of gas by its decomposition. Can the ultimate particles of these gases be further subdivided; can they, like those of water, be broken up and reappear in new forms? No; there is no known process by which an atom of oxygen can be made anything but oxygen, or an atom of hydrogen anything but hydrogen.

The only thing which is compound in the composition of oxygen is that its molecules consist of two atoms linked together. This appears from the fact that while the weight of oxygen, and therefore that of its molecules, is sixteen times greater than that of an equal volume of hydrogen, and therefore of hydrogen molecules, it combines with it in the proportion not of sixteen, but of eight to one. If, therefore, the molecule were identical with the atom of oxygen, we must admit that the atom could be halved, which is contrary to its definition as the ultimate indivisible particle of the substance oxygen. But if the oxygen molecule consists of two linked atoms, O—O, and the hydrogen molecule equally of two, H—H, as can be proved by other considerations, everything is explained by assuming that the molecule of water consists of two atoms of hydrogen linked to one of oxygen, or H₂O, and that when this molecule is broken up by electricity, its constituents resolve themselves into atoms, which recombine so as to form twice as many molecules of hydrogen, H—H, as of oxygen, O,—i.e. into two volumes of hydrogen gas to one of oxygen.

Taking the single hydrogen atom as the unit of weight as being the lightest known ponderable body, and calling this weight a microcrith, or standard of the smallest of this order of excessively small weights, this is equivalent to saying that the weight of an oxygen atom is equal to 16 microcriths, and as water is composed of one such atom plus two of hydrogen, the weight of its molecule ought to be 16 + 2 = 18, which is in fact the exact ratio in which the weight of a volume of steam, or water in the form of gas, is heavier than an equal volume of hydrogen.

This key unlocks the whole secret of the chemical changes and combinations by which matter assumes all the various forms known to us in the universe.

Thus oxygen enters into a great variety of combinations forming different substances, but always in the proportion which is either 16, or some multiple of 16, such as 32, 48, 64. That is, either 1, 2, 3, or 4 atoms of oxygen unite with other atoms to form the molecules from which these other substances are made.

One atom of oxygen weighing 16 microcriths combines, as we have seen, with two atoms of hydrogen weighing 2, to form a molecule of water weighing 18 mc. In like manner one atom of oxygen, 16 mc., combines with one of carbon, which weighs 12 mc., to form a molecule of carbonic oxide weighing 28 mc.; and two of oxygen, 32 mc., with one of carbon, 12 mc., to form a molecule of carbonic dioxide weighing 44 mc.

The same applies to all elementary substances. Thus hydrogen, two atoms of which combine with one of oxygen to form water, combines one atom to one with chlorine to form the molecule of hydrochloric acid, which weighs 36·5 mc., being the united weights of one atom of chlorine, 35·5 mc., and one of hydrogen, 1 mc. These, with hundreds of similar instances, are the results not of theories as to molecules and atoms, but of actual facts, ascertained by innumerable experiments made independently by careful observers over long periods of years, many of them dating back to the labours of the alchemists of the middle ages in pursuit of gold. The atomic theory is the child and not the parent of the facts, and is indeed nothing but the summary of the vast variety of experiments which led up to it, as Newton’s law of gravity is of the facts known to us with regard to the attractions and motions of matter in the mass. But as Newton’s law enables us to predict new facts, to calculate eclipses and the return of comets beforehand, and to compile nautical almanacs; so the new chemistry, based on the atomic theory, affords the same conclusive proof of its truth by enabling us in many cases to predict phenomena which are subsequently verified by experiment, and to infer beforehand what combinations are possible, and what will be their nature.

The actual existence, therefore, of molecules and atoms is as well-ascertained a fact, as that of cwts. and lbs., or of planets and stars, of solar systems and nebulæ.