DIMENSIONS OF ATOMS AND MOLECULES 51 



F . . . 

 ci . . . 



Br . . . 



I . . . 



rubidium potassium. It has been shown recently that the 

 corresponding caesium compounds are not isomorphous with the 

 above salts (Davey and Wick, Proc. Am. Phys. Soc, December 

 1920), so that they cannot be included in this comparison. 



Similar relationships hold for other series of compounds. 

 The substitution of sulphur for oxygen, for example, increases 

 by a nearly constant amount the inter-atomic distance in a 

 number of compounds, and further, this amount is the same as 

 that in the case of the substitution of fluorine by chlorine, the 

 elements which follow oxygen and sulphur in the periodic table. 

 An attempt to summarise this relationship has been made by 

 the author, which may be exemplified by the graph in Fig. i. 

 The abscissae in the graph are the atomic numbers of the 

 elements. The ordinates represent the " diameters " of the 

 atoms, using this term in a very special sense. The crystal is 

 regarded as composed of a set of spheres (the atoms) packed 

 tightly together, so that the distance between two neighbouring 

 atoms is equal to the sum of their radii. It is found that, by 

 a suitable choice of radii, we can suppose each atom to be of 

 constant size in all the structures of which it forms part. Atoms 

 such as those of the metals potassium, calcium, and so forth 

 have large diameters assigned to them, since they occupy 

 isolated positions in the crystal structure. The electro-negative 

 elements combined together are closely associated, and must 

 be represented by small spheres when building a model of 

 the atomic arrangement. To a very fair approximation, the 

 distance between any two atoms in a crystal structure may be 

 predicted by adding together the radii of the atoms as given 

 by this graph. 



In each period of the atomic series, the diameters in the 

 graph approach a lower limit. These limiting values increase 

 from one period to the next, their values being : 



Second period (ending in neon) . 1-30 x 10 ~' cm. 



Third period (ending in argon) . 2-05 x 10 ~' cm. 



Fourth period (ending in krypton) 2-35 x io~' cm. 



Fifth period (ending in xenon) . 2-70 x 10 -' cm. 



It cannot be assumed, a priori, that the inter-atomic distances 

 in an isolated molecule will be the same as those in a group of 

 atoms which form part of an extended crystal structure, for the 



