280 



hypothetical: thus for magnesium-sulphate -\- 1 H 2 0, one of the 

 seven molecules must be considered as "constitution"-water, and 

 the oxygen-atom is thought always to take the place of two 

 coordination-loci, being, in their opinion, a substitute of a dyad 

 character. When the atom is thought to be in the centre of eight 

 coordination-places, distributed in space as the corners of a cube, 

 they demonstrate that of the three possible arrangements for 

 the atom-complex: (SO 4 , H 2 O)", two will have the symmetry Cj 

 and 5, (7 and //: fig. 772), but only one that of the axial group 

 > 3 ; /// in fig j/2. 



In their opinion, to (MgSO if H 2 0)+6 H 2 could, therefore, only 

 be attributed the configuration ///; and in an analogous way 

 they deduce for the CIO' A -, resp. BrO' 3 -ion a configuration quite 

 analogous to that of Werner's triethylenediamine-salts (Z) 3 ), in 



OH 



OH 



HO 



w 



L 



)H 



III. 



which ions the central halogen-atom should then be heptavalent. 



Evidently the success of such applications of Werner's 

 theory will not be very great, as long as no exacter data about 

 the configuration of inorganic salts are at our disposal. 



However, the idea that the optical inactivity of the solution 

 should be caused by a very rapid autoracemisation during the 

 process of solution of the salt, is used by them in a rather 

 suggestive way to explain the dimorphism of NaCl0 3 , discovered 

 by Mallard 1 ). This author observed a second unstable modifica- 

 tion of the cubic salt, isomorphous with the rhombohedral 

 sodium-nitrate. If now NaCl0 3 be supposed to be a racemic 

 compound, its rhombohedral crystals being pseudo-racemic inter- 

 growths of d-, and /-forms of cubic symmetry, the ditrigonal 

 second form of sodium-chlorate found by Mallard, would then 



E. Mallard, Bull, de la Soc. Miner., 7, 352, (1884). 



