238 William E. Ford, 



Such a change in molecular volumes from one end of an isomor- 

 phous series to the other is apparently a common thing. The 

 various series of isomorphous salts that have been most care- 

 fully studied show this, although the fact does not seem to have 

 been previously emphasized. For instance Rettgers^ made 

 studies of the following different series, all of which he proved 

 to be strictly isomorphous. In each case the end members show 

 different molecular volumes. 



K^SO.CMol. vol. = 65.4)— (NHJ^SO.CMol. vol. = 74.5) 

 K2SO,.MgSO,.6H20(Mol. vol. = 197)— 



(NH,),SO,.MgSO,.6HoO(Mol. vol. = 220) 

 (NHj2SO,.MgS04.6H20(Mol. vol. = 220)— 



(NH,)oSO,.FeSO,.6H,0(Mol. vol. = 244) 

 Tl.,SO,.AU(SOj3.24H,.0(Mol. vol. = 463)— 



ICSO,.Al,(SOj3.24H.O(Mol. vol. = 541) 



In considering the crystal structure of the mixed-crystals of 

 such a series we must assume, either that they contain in varying 

 proportion two different molecules which have different molec- 

 ular volumes and therefore different space lattices, or that the 

 molecular volumes of the two molecules must progressively 

 change and in any one specimen be identical. It does not seem 

 possible to determine which of these two suppositions is the 

 correct one but certainly the second seems more reasonable and 

 logical. It can be conceived that the entrance of an isomorphous 

 molecule into the composition of a salt would so affect the whole 

 structure as to force a change in the volume of the unit cell of 

 the crystal net-work. And this change would naturally increase 

 with the rise in the percentage of the second molecule. This 

 change in the spacing of the molecular layers in the crystal does 

 not necessarily involve any change in their angular relations. 



Another question arises in this connection. If we assume this 

 change in the molecular volumes with change in composition, is 

 it necessarily uniform from one end of the series to the other? 

 In other words, does the presence of one per cent of an isomor- 

 phous molecule have the same effect proportionately upon the 

 crystal structure as 25 per cent would have? The study of the 

 molecular volumes of the dolomite analyses might suggest that 



Zs. Phys. Chem., 3, 497, 1889; 6, 235, i{ 



