444 SCIENTIFIC THOUGHT. 



42. An analogy has been pointed out l between the atomic 

 between theory in chemistry, by which Dalton explained the fixed 

 atoraiciaws s i m pl e an( l multiple proportions of the combining weights 



of various substances, and the molecular theory of crystal- 

 line structures, by which the fundamental forms of crystals 

 are defined and the accessory forms derived from them. 

 It has been found that if once a crystal has been defined 

 by a fundamental plane referred to three axes at fixed 

 angles, all other planes or faces can be defined by simple 

 multiples of the numbers which belong to the fundamental 

 plane, and which are called the parameters of the crystal. 

 This fundamental rule or law of crystallisation, termed by 

 Haliy the law of derivation, stands thus in the same rela- 

 tion to the corpuscular theory of the structure of bodies 

 as the law of fixed multiple proportions stands to the" 

 original atomic view of matter, and it is thought that it 

 may in the future lead to important results. 2 



43. Another very remarkable discovery had been made by 

 phism. Mitscherlich in 1 8 23. 3 This is the property which various 



compounds possess of crystallising in the same forms, 

 although they contain different elements such elements 

 being, however, joined together by similar formulae. The 

 elements are, as it were, interchangeable. This phe- 



1 See Ostwald, " Allgemeine 

 Chemie,' vol. i. p. 870. 



2 A question arises in this connec- 

 tion as to the accuracy of the crys- 

 tallographic law of the fixity of the 



a somewhat disturbed manner" (loc. 

 cit., vol. i. p. 890). This I under- 

 stand to mean that, if disturbing 

 circumstances could be removed, 

 the law of the fixity of angles and 



angles. In respect of this Ostwald , the simple multiples of the indices 



says : " On examining the validity I would obtain with the same accur- 



of the fundamental laws of crystallo- acy as do the combining numbers 



graphy, it becomes evident that and their multiples in chemical 



they are only approximate, or per- j combinations. 



haps more correctly, that there exist 3 See supra, chap. ii. p. 191 and 



numerous circumstances which per- note, 



mit them to show themselves only in | 



