254 



PROF. W. H. BRAGG ON X-RAYS AND CRYSTAL STRUCTURE. 



radiation, and transmitting the remainder. All the reflected portions which unite to 

 travel along BC will be in the same phase if 



n\ = 



-AB = A,D-AB = DN = 2d sin 0, 



where n is any integer. In the figure n = 1. Only when this relation is satisfied is 

 there any sensible reflection. The greater the number of planes concerned in the 

 action the more abruptly does the effect disappear if 6 is slightly varied. 



We speak here of reflecting planes and represent them as surfaces. But the 

 general result is exactly the same if we replace a plane by a thin layer containing 

 scattering particles ; and it still holds exactly if we suppose a continuous succession 

 of thin layers of varying density forming the periodic structure to which we have 

 referred. 



The atoms of a crystal are distributed in an orderly manner. They can be thought 

 of as arranged on series of parallel planes ; and this can be done in many ways. A 

 natural face is always parallel to a series of this kind. 



An atom possesses the power of scattering X-rays to an extent which appears to 

 depend mainly upon its mass. It is not quite clear how this power is distributed 

 within the atom ; one would, expect that both nucleus and electrons share in it, and 

 that it extends more or less over the whole atom. As the diameter of the atom is of 

 the same order of magnitude as the spacings of the crystal planes this last consideration 

 is of importance, and I propose to return to it later. For the present we may observe 

 that if the scattering power were confined to one central point in each atom we 

 should be able to represent the "reflection" by a plane containing atoms as if it 

 occurred in a reflecting surface, as in fig. 2. If, on the other hand, we suppose the 

 scattering power to be distributed through the atom, and if we take a section through 

 the crystal perpendicular to the direction x (see fig. 2). the scattering power of the 



substance in the layer Sx will be a periodic function of x. In either case the formula 

 n\ = 2d sin 6 gives the only values of Q at which reflections can occur. 



Let us take a numerical instance. The atoms of rock salt are now known, as the 

 result of the investigations we are considering, to be arranged as in fig. 3. The 

 plane of the paper is parallel to a cube face. The atoms are represented as circular, 

 and as having a definite boundary ; but that is merely because we must give some 

 form in a drawing. In reality we know very little about such things. 



