for neutral atoms. Thus we may consider that the metallic atoms lose very easily one or more electrons and also that there is a somewhat feeble attachment in their case between the neutral atom and the free electron. Hence metals in the mass are conductors because there are plenty of free electrons present in them. On the other hand, in the case of non-metallic atoms the force required to detach one or more electrons from the atom is much greater and conversely the attachment of free electrons for the neutral atom is larger. Accordingly, in non-metals there are few free electrons and they are therefore non-conductors. Moreover the presence of positive and negative atomic ions causes them to link together into more or less complex molecules and they exhibit polyvalency and act as the grouping elements in molecular complexes. This is a very characteristic quality of the elements, sulphur, silicon and carbon.
Helmholtz long ago laid stress on the fact that certain physical and chemical effects could only be explained by assuming a varying attraction of electricity for matter. The same idea followed out leads to an hypothesis of chemical combination and dissociation of salts in solution. Thus a molecule of sodic chloride is the electrical union of a monovalent sodium ion or sodium atom minus one electron with a chlorine ion which is a chlorine atom plus one electron. It may be asked why in this case does not the extra electron pass over from the chlorine to the sodium ion and leave two neutral atoms. The answer is because the union between the electron and the chloride is probably far more intimate than that between the atomic groups. These latter may revolve round their common center of mass like a double star, but the electron which gives rise to the binding attraction may be more intimately attached to the atomic group into which it has penetrated.
Voltaic Action.
Any theory of electricity must in addition present some adequate account of such fundamental facts as voltaic action and magneto-electric induction. Let us briefly consider the former. Suppose a strip of copper attached to one of zinc and the compound bar immersed in water to which a little hydrochloric acid has been added.
All chemical knowledge seems to point to the necessity and indeed validity of the assumption that the work required to be done to remove an electron from a neutral atom varies with the atom. Conversely the attraction which exists between a free electron and an atom deprived of an electron also varies. Accordingly the attraction between atomic ions, that is, atoms one of which has gained and one of which has lost electrons, is different. Upon this specific attraction of an atomic ion for electrons or their relative desire to form themselves into neutral molecules depends what used to be called chemical affinity. Mr. Ruther-
