136 



left-handed. If such crystals as, for instance, those of dextro- and 

 laevogyrate sodiumc-hlorate, are dissolved, giving an optically in- 

 active solution, the supposition must necessarily be made that a 

 rearrangement of the atoms during the process of solution takes 

 place, producing an equal number of both kinds of enantiomorphous 

 molecules, or perhaps a quite different species of them, superposable 

 with their mirror-image. This is intimately connected with the fact 

 that the notion of the fundamental domain is a purely mathematical 

 one, and, therefore, with respect to the endless periodical repetition 

 of equal parts throughout the regular structure, the gathering 

 together of certain atoms into complexes is within wider limits a 

 quite arbitrary, purely mathematical fiction. The notion of "mole- 

 cular complex" is in the crystalline state, therefore, formally without 

 significance ; which, however, does not mean that the connections 

 between the constituting atoms, as involved in the study of the 

 properties of the chemical molecule, should have completely dis- 

 appeared. Only they need not be considered for the mathematical 

 description of the crystalline, periodical arrangement: that is all. 

 The specific character of the crystal-structure lies in the fact 

 that all atoms of the same kind are equivalent in the architecture 

 of the crystal, and that for the mechanical equilibrium finally reached 

 therein, the total action of each atom is as if it were an autonomical 

 individual. Atoms of different kinds can, moreover, always be grouped 

 together so as to form complexes which, similarly and infinitely 

 repeated in an absolutely regular and periodical way according 

 to the special symmetry-properties of the whole structure, will 

 fill up space to produce the remarkable masterpiece of nature, 

 that we call a crystal. 



However, it is exactly this very general character of the theory 

 which makes its application to concrete cases rather difficult. The 

 whole number of symmetrical arrangements thus found amounts 

 to no less than two hundred and thirty, the symmetry of which can 

 be grouped in the same 32 classes as we have previously found to 

 be possible for crystals (Chapter F). 



A considerable number of possible structures belongs therefore 

 to each of these 32 classes; and as for the explanation of physical 

 phenomena the precise arrangement of the constitutive atoms is 

 the point of interest which this general theory leaves totally out of 

 consideration, the chance of its successful application for the purpose 

 of explaining crystallographical and crystallophysical phenomena 



