DOUBLE SALTS 155 



remained for Retgers x to make it absolutely definite, and to 

 show how it could be applied to argue backwards as to the 

 kind and degree of relationship between different elements. 

 He asserted that two simple salts which form a series of mixed 

 crystals cannot form a double salt, and vice versa, that mixture 

 and combination are mutually exclusive, and that this is 

 true without reservation, without any exceptions whatever. 

 Potassium and rubidium and ammonium form isomorphous 

 mixtures, but do not form double salts ; potassium and sodium 

 are not isomorphous, but they do form double salts, such as 

 for instance Seignette salt, the double potassium-sodium tartrate, 

 or Scacchi's salt, the double sodium-ammonium racemate, which 

 is a double salt in a double way (Table II.). Sulphates and 

 selenates form isomorphous mixtures, but not double salts, 

 whilst NaN0 3 combines with Na 2 S0 4 to a double salt. Hence, 

 on the basis of the absolute validity of this relationship, the 

 fact that we know of a number of double salts between 

 potassium and silver, but of none between sodium and silver, 

 leads to the inference of a greater relationship between silver 

 and sodium than between silver and potassium, and points to 

 the probability of isomorphism between silver and sodium, 

 which is proved by the existence of a series of mixed nitrates 

 (and chlorates) of these two metals. 



2. Moreover it follows, as a direct consequence of the 

 recognition of chemical difference as the essential factor in 

 the formation of double salts, that the degree of this difference 

 ought to be of influence on the relative stability of the com- 

 binations produced. That this is actually the case is shown by 

 the numerous instances in which it has been found that the most 

 stable of all the alkali double salts are those of caesium, itself the 

 most electro-positive of all the metals ; and in entire agreement 

 with such incidental and unconnected observations is the result 

 of an investigation on the double chlorides of ferric iron and 

 the alkali metals, which well shows the connection between the 

 stability of the double salt and the gradation in the electro- 

 positive character of the alkali metals. 2 Between o° and 6o° 

 NaCl could not be made to combine with FeCl 3 ; KC1 yielded 

 one double salt, viz. FeCl 3 . 2KCI . H 2 0, whilst CsCl gave two 



1 Retgers, " Uber chemische Verbindungen isomorpher Korper," Zs. physik- 

 Chem. 15, 1894, p. 529. 



2 Hinrichsen and Sachsel, loc. cit. 



