172 Prof. W. L. Bragg on the 



of iron pyrites, the iron atoms being at the corners A, C, H, 

 and F. One diagonal of the cube, the diagonal AG in the 

 figure, is an axis of threefold symmetry, and the sulphur 

 atom lies at some position along this axis. Each corner of 

 the cube is a centre of symmetry. If a sulphur atom 

 is centred at the point Si, there will be a corresponding 

 atom at the point S 2 where S 1 Gr = S 2 Gr. A pair of sulphur 

 atoms is thus associated with each corner of the cube not 

 occupied by an iron atom, since one threefold axis passes 

 through every corner. On the conception of the atoms as a 

 set of spheres packed together, it will be seen that there are 

 two possible positions for the sulphur atom. It may lie at 

 the centre of the cube, where it is symmetrically surrounded 

 by four iron atoms at A, C, H, and F. Alternatively, it may 

 move along the diagonal until it is at G, on the other side 

 of the plane HFC, where it will be packed between the 

 three atoms of iron at H, 0, and F and the corresponding- 

 sulphur atom at S 2 . 



The structure of metallic iron has been determined by 

 Hull *. The iron atoms are situated on a cube-centred 

 lattice, the side of the cube having a length of 2'86 A, and 

 the distance between the centres of neighbouring iron atoms 

 is 2*17 A. If this length is taken as a first approximation 

 to the diameter of the sphere representing the iron atom, and 

 spheres of corresponding radius are described with their 

 centres at H, 0, and F, the centre of the sulphur atom is 

 thereby fixed. The atom must touch the three iron atoms and 

 also the other sulphur atom at the point G, as shown in fig. 1 1>. 

 The ratio SiG/AG which determines the position of Si can be 

 calculated to be 0*22/1. This ratio was originally determined 

 by the author as O20/1. A more exact determination by 

 Ewaldf, based on the Laue-photograph of Pyrites, gave the 

 value of the parameter as 0226/1. 



The exact correspondence of the position of the sulphur 

 atom calculated in this way with that found by experiment 

 is not to be expected, the diameters of the spheres repre- 

 senting the atoms cannot be regarded as absolutely fixed. 

 It will be seen, however, that the conception of the atoms as 

 spheres packed together does lead in this case to an approxi- 

 mate value for the parameter not far from the true one. 

 The diameter of the sulphur atom is given by the distance 

 S X S 2 , which is equal to 2*05 A. Each sulphur atom is sur- 

 rounded by three iron atoms and a sulphur atom, each iron 

 atom by six sulphur atoms. 



* Hull, Phys. Rev., 10 Dec. 1917. 

 t Ewald, Phys. Zeit. April 15, 1914. 



