The doctrine of atomic linkage was thus brought into the very unsatisfactory state that two apparently incompatible theories held the field. Each was successful in explaining the behaviour of one class of compounds, but neither could be extended to cover the whole ground. It was evident that the true theory still re mained to be discovered, and that when it was discovered the structural theory and the theory of co-ordination would be seen to be two imperfect aspects of one truth.
The position which chemistry had reached at the beginning of the present century was therefore this. It appeared that the links which hold together the atoms in a molecule can be of three kinds: (I) Polar or ionized links, between the oppositely charged ions of a salt. (2) Non-polar non-ionized links, which can exist be tween atoms of a similar or even identical character, and are pe culiarly common among the compounds of carbon. (3) The co ordinated links of Werner, which are not ionized, are capable of uniting apparently saturated atoms, and occur mainly in inorganic compounds.
There was the further peculiarity that an atom was found to be in general capable of forming the same number of links of the first and of the second kind, and that this number depended on its position in the periodic table. The number of links of the third kind which it could form had no relation to its periodic group, and was usually six, sometimes four, while other numbers were comparatively rare.
Atomic Structure and Valency.—These important, but ob viously not final, conclusions were the result of the chemical in vestigation of the structures of molecules. Their further devel opment, and their combination into one uniform system, were rendered possible only by the increase of our knowledge of the physical structure of the atom (q.v.). Of this it is enough to say here that mainly through the work of J. J. Thomson, Ruther ford, Moseley, and Bohr, it was shown that every atom consists of a minute positively charged nucleus, surrounded by a number of negative electrons which are arranged in groups, revolving at greater or smaller distances from the nucleus. The number of electrons (which is also the number of positive charges on the nucleus) is also the ordinal number (from 1 for hydrogen to 92 for uranium) of the element in the natural series (the periodic classification) which agrees in almost every case with the order of the atomic weights; this is known as the atomic number (q.v.) As it thus appears that every atom contains one electron mon than that of the preceding element, the periodic table acquires new meaning. The regular recurrence of a definite set of proper ties after a definite number of places suggests that the number of electrons so added form a new and complete group, and that for example, lithium (3) and sodium (I I) resemble one another because the eight extra electrons of the latter are grouped to gether, leaving a similar arrangement of the chemically activE electrons in both atoms. It may further be assumed that sincE
certain elements—the inert gases such as helium and argon— have practically no power of chemical combination, the stability of the electronic groups in the atoms of these gases is already so great that it cannot be improved by interaction with other atoms; and that other elements combine to form molecules be. cause by so doing they can rearrange their electrons in a more stable manner. The atomic numbers of the inert gases (helium 2 neon i o, argon 18, krypton 36, xenon 54, radon 86) will thus give the total number of electrons in a series of stable groups, and their differences (2, 8, 8, 18, 18, 32) the sizes of individual groups. We may suppose that other elements tend by combination to acquire electronic groups of these or similar sizes.
The first attempts to work out a theory of valency along these lines were published in 1916 by Kossel in Germany and by G. N. Lewis in America. Kossel dealt with polar or ionized links and Lewis with non-ionized.