July. 22, 1920] 



NATURE 



647 



and we know that it is in many cases not that 

 predicted by the "valency volume" law. The 

 law can be modified, however, so as to apply to 

 the majority of crystals so far analysed. It may 

 be shown that we can assign a definite diameter 

 to the sphere representing- the atom, a diameter 

 characteristic of the element in question. Some 

 atoms appear to occupy a small domain in a 

 crystal structure, others a larger space. By find- 

 ing- the distances between the atomic centres in 

 a number of crystals the diameters represented 

 in Fig. 2 have been calculated. This fig-ure sum- 

 marises an empirical relation, which states that 

 the distance between neighbouring atomic centres 

 in a crystal structure is equal to the sum of two 

 constants, characteristic of the atoms concerned. 

 We can therefore picture the crystal structure as 

 a set of spheres packed tightly together, just as 

 Barlow and Pope did ; but in this case the dimen- 

 sions of the spheres are those in Fig. 2, not those 

 given by the valency volume law. 



arrangements, those of the inert gases, are those 

 in which the outer shell has its full complement 

 of electrons. Such forms are very stable; they 

 are characterised by a weak external field. T!ic 

 chemical properties of the other elements repre- 

 sent their tendency to revert to a more st; ble 

 electron system. 



The crystal of potassium chloride, on this point 

 of view, consists of alternate potassium and 

 chlorine ions. The potassium atom is surrounded 

 by nineteen electrons when electrically neutral. 

 Eighteen of these electrons complete the three 

 electron shells, represented, for instance, by the 

 very stable arrangement of argon. The remain- 

 ing electron has no place in the stable system, 

 and there is therefore a tendency for the atom 

 to part with it and become a positively charged 

 potassium ,ion, the nucleus with nineteen element- 

 ary charges being surrounded by eighteen elec- 

 trons. Chlorine similarly tends to gain an elec- 

 tron. The KCl structure may therefore be re- 





I, 



2- 



\ 





iMillllllllllMilli I 



\ 



\ 





1 1 M 1 1 M 1 1 1 1 1 1 1 1 1 1 1 1 1 M 1 1 1 1 1 



Ba 



S 10 lb 20 



Atomic Numbers of the Elements . 



35 40 



The atoms in a crystal are thus packed together 

 as if they were inelastic spheres in contact. This 

 is merely a way of visualising the structure, and 

 must not be interpreted too literally. A ready 

 explanation of the form of the graph in Fig. 2 is 

 afforded by that conception of atomic structure 

 which Stark, Born, Lande, Lewis, and others 

 have helped to build up, and which has recently 

 been so brilliantly summarised in a series of 

 papers by Langmuir. Many independent lines of 

 investigation have led to the conception of the 

 atom as a positive nucleus surrounded by an elec- 

 tron system, in which the electrons are fixed at, 

 or oscillate about, certain definite positions in the 

 atomic structure. This is a view which forms a 

 contrast to the Bohr atomic model, where the 

 electron orbits enclose the atomic nucleus. In 

 the " fixed electron " atom the electrons are 

 arranged in a series of shells surrounding the 

 nucleus, the numbers which complete the succes- 

 sive shells being 2, 8, 8, 18, 18, and 32. Certain 

 NO. 2647, VOL. 105] 



garded as an assemblage of argon shells, with 

 resultant positive and negative charges, which 

 are held together by their charges, and kept apart 

 by some force of repulsion which we must sup- 

 pose to exist between the outer electron systems. 

 The result is the structure in Fig. i where every 

 ion is surrounded symmetrically by the greatest 

 possible number — six — of ions of the opposite 

 sign. 



In the case of two electro-negative elements 

 which are chemically combined, both atoms have 

 a smaller number of electrons than corresponds 

 to stability of the outer shell. Stability is attained 

 by their holding pairs of electrons in common. 

 In this way Langmuir has succeeded in the most 

 striking manner in explaining the complicated 

 valency of such elements as nitrogen and phos- 

 phorus. 



The structure of calcite (Fig. i) is an example 

 of both types of chemical combination. The cal- 

 cium atom, represented by the large sphere, is 



