199 



Also combinations of poly symmetrical and real polymorphic 

 changes may occur with the same substance 1 ). 



Cases of this are: propylamine-stannichloride 2 ) : (C z H 1 .NH 2 ) 2 SnSl 6 \ 

 diethylamine-platinichloride : [(C 2 H 5 ) 2 NH 2 ] 2 PtCl 6 ; diethylamine-stanni- 

 chloride: \(C 2 H 5 ) 2 NH 2 ] 2 SnCl 6 , which even in its external habit 

 completely simulates a cubic crystal ; tripropylamine-platini- 

 chloride : [(C 3 H 7 ) 3 NH] 2 PtCl 6 ; tetraethyl-ammonium-stannichloride : 

 [N(C 2 H^4\ 2 SnCls, where the lamellar structure of the monoclinic, 

 but pseudocubic crystals is very distinctly recognisable ; tetrapropyl- 

 ammonium-platinichloride: [(C 3 ^ 7 ) 4 A^] 2 P^C/ 6 ; and tetramethyl-ammo- 

 nmm-platimMoride;[N(CH 3 )4\ 2 PtCl 6 , where, however, a rather similar 

 case perhaps occurs to that of the isopropylamine-platinichloride pre- 

 viously mentioned. While in the case of real polymorphism, metastable 

 states may occasionally occur under the influence of retardative 

 circumstances, it must be clear from what has been said about poly- 

 symmetrical changes in general, that of such metastable conditions 

 there can be no question here. 



12. From these and many other researches it has gradually 

 become clear, that even if a crystal be apparently a homogeneous 

 individual, only in rare cases may it be considered as a really homo- 

 geneous thing. According to Mallard's views and those of a number 

 of other investigators, the molecular arrangements which are charac- 

 teristic of crystalline matter, do not necessarily possess the perfect 

 homogeneity involved by Hessel's and Bravais' theories 3 ). Its 

 constituting and identical molecules are, therefore, not always parallel 

 to each other, but they may have different orientations in space, 

 depending on the special symmetry of the crystalline substance. 

 In very numerous cases it is built up from lower symmetrical masses, 

 according to the general laws of twin-formation. 



The fact, that it is just those space-lattices whose dimensions are 

 such as to make them appear to possess an approximate symmetry, 

 which show most conclusively that tendency to aggregate into ap- 

 parently higher symmetrical complexes, whose t winning-elements 

 correspond with the approximate symmetry-elements of these simu- 

 lated higher symmetrical complexes, was certainly first recognised 

 in its general significance by Mallard. But from this to his later 



!) P. Groth, loco cit., p. 7. 



2) A. Ries, Zeits. f. Kryst., 39, 50, 55, 56, 66, 69, 73, (1904); 36, 360, (1902); 

 49, 513, (1911). 



3 ) J. Beckenkamp, Statische und Kinetische Krystalltheorien, I, p. 194, (1913). 



