B.— CHEMISTRY. 31 



by the other. But the same must hold with covalency also. If the 

 covalent link consisted of a single shared electron, this would not be true. 

 If the atom A could form a covalent link with B merely by sharing one 

 of its own electrons with B, this would use up one of the units of B, since 

 it would increase B's electrons by one ; but it would not affect the com- 

 bining power of A. For example, hydrogen (1) is one electron short of 

 the stable helium number 2 ; carbon (6) is 4 short of the stable neon 

 number 10. If in methane CH 4 each hydrogen atom is attached to the 

 carbon by a single shared electron, then if this electron is derived from 

 the hydrogen it will satisfy the carbon, but will leave the hydrogen still 

 one electron short ; if it is derived from the carbon, it will leave the 

 carbon four electrons short of the stable number. In either case the 

 resulting molecule would be unsaturated, whereas it is in fact saturated. 

 It was to meet this difficulty that Lewis assumed that the covalent link 

 consisted of two shared electrons, one derived from each of the two linked 

 atoms. On this hypothesis the carbon in methane shares one of its four 

 valency electrons with each of the four hydrogen atoms, thus increasing 

 the number of each hydrogen to two, and at the same time each hydrogen 

 shares its own electron with the carbon, thus satisfying the carbon. 



We have therefore got an electronic mechanism which will account 

 for the two recognised forms of valency, the ionised and the non-ionised. 

 If these are really the only two forms of linkage which can exist in a 

 molecule, it must be possible to extend them so as to account for co- 

 ordination. This is in fact surprisingly simple, and the solution was 

 foreshadowed by Lewis in his paper of 1916. It is clear that the link 

 which attaches one of the groups of a co-ordination complex to the central 

 atom is of the non-polar type. It is an essential point in Werner's theory 

 that such links are not ionised ; this is how they are distinguished from 

 the links to atoms in the ' second sphere.' Thus in the compound 

 [Pt(NH 3 ) 4 Cl 2 ]Cl 2 the two chlorine atoms outside the bracket enclosing 

 the co-ordination complex are ionised, while those inside are not. The 

 same conclusion is sujjported by the fact that the arrangement of the 

 groups in the co-ordination complex round the central atom can give 

 rise to optical activity ; for this, as we know from organic chemistry, is 

 only possible with groups which are attached by covalent links, that is, 

 by directed forces. We must therefore look for an explanation of co- 

 ordination in the formation of covalencies, that is, of links formed of 

 pairs of shared electrons. But they must arise in some way different 

 from that which we have hitherto assumed, since their numerical relations 

 are different ; their number is not related to the periodic group of the 

 central atom, and also they can be formed with atoms (such as the nitrogen 

 in ammonia or the oxygen in water) which have already completed a 

 stable number of electrons. Now in the normal covalency formation 

 described above it was assumed that one of the two shared electrons of a 

 link came from each of the two atoms concerned. It is obviously possible 

 that both might be derived from one of them ; and the recognition of 

 this possibility is all that is required to provide an electronic mechanism 

 for co-ordination. By means of this extension of the idea of covalency 

 formation we can explain all the peculiarities of co-ordination compounds, 

 of which, as we have seen, the most important are the power of further 



