APPLICATION OF THE MOLECULAR HYPOTHESIS 95 



between adjacent molecules are exercised equally in all directions, 

 with the result that the molecules are all in haphazard, unordered 

 positions relative to one another and no particular arrangement 

 of particles in space can persist. Substances of this character 

 are therefore called amorphous (Greek, without form) . In crystals, 

 on the other hand, since each substance shows an individual struc- 

 ture, the forces between adjacent particles must be exercised in 

 definite directions. 



By using crystals of different substances as X-ray gratings 

 (see p. 548), W. H. and W. L. Bragg (1914) have been able to 

 show that crystals are composed of particles arranged in rows, the 

 spacing of these rows with respect to one another determining 

 the geometrical form of the crystal. They have also succeeded 

 in proving that the particles, which so arrange 

 themselves in definite patterns in crystals, are 

 not molecules, much less aggregates of molecules, 

 but atoms of the constituent elements of the sub- 

 stance. Thus when common salt (sodium chloride), 

 which crystallizes in cubes, is examined by reflect- 



O -Na, -Cl. 



ing X-rays from an appropriate plane, it is found F IG . 44 

 to consist of alternate rows of sodium atoms and 

 chlorine atoms, the rows being arranged in space in such a 

 way as to build up a cubical structure. The framework of 

 a unit cube of sodium chloride is shown in Fig. 44. The actual 

 length of a side of this cube is approximately only one hundred 

 millionth of an inch! However, by imagining other cubes to 

 be packed all round this, with atoms of sodium and of chlorine 

 placed at their alternate corners, the reader will obtain for him- 

 self an idea of the ultimate structure of a crystal of common salt. 

 Most substances crystallize in less simple geometrical forms and 

 consequently possess a much more complex structure. The frame- 

 work of the carbon atoms in a diamond, for example, is indicated 

 in Fig. 45. 



In a substance in the crystalline state, therefore, definite molec- 



