156 SCIENCE PROGRESS 



such salts, viz. FeCl 3 . 2CsCl . H 2 and FeCl3.3CsCl.H2O. It 

 should be noted, however, that no numerical relations emerge 

 from this research, though it allows us to trace the effect on 

 the power of double-salt formation of the substitution of allied 

 elements for one another. 



In a later and very extensive investigation of the series of 

 double salts formed between sodium sulphate and the class of 

 sulphates called "vitriols," the realm of the quantitative is being 

 approached. 1 The principle of this comparative study recalls 

 that followed by Mr. Tutton in his extensive researches on the 

 influence on crystalline form produced by the substitution in 

 a series of salts of one isomorphous element for another. 

 The special feature of this investigation of the class of double 

 sulphates, Na 2 S0 4 . M"S0 4 . 4.H 2 (Table IV.), lies in the fact 

 that the determination of the transition point and the transition 

 interval for each double salt made it possible to examine under 

 strictly comparable conditions what is the influence of the 

 properties of each special vitriol on the properties of the double 

 salt into the composition of which it enters. Thus it could be 

 shown that whenever the vitriol crystallises with seven molecules 

 of water, as happens in the majority of cases (Mg, Zn, Fe", Co, 

 Ni), the double salt crystallises with four molecules; whilst 

 copper and cadmium, whose stable hydrated sulphates contain 

 five and one-half molecules of water respectively, give double 

 salts containing only two molecules ; and finally manganese, 

 whose heptahydrate even at ordinary temperatures passes into 

 a monohydrate, gives a tetrahydrated double salt, which 

 however passes into the dihydrate below 40 , that is, at a 

 temperature considerably lower than that of the dehydration 

 of the other tetrahydrated double salts of the series. Con- 

 nections have also been traced between the solubility of the 

 vitriol and the extension of the transition interval, as well as 

 between the electrical potential of the metal in the vitriol and 

 the transition point of the double salt ; but without the repro- 

 duction of a large number of numerical data these connections 

 cannot be followed out satisfactorily. The results, moreover, 

 are not very striking ; and the chief reason for having said even 

 thus much about this research is to show the method followed 

 in a chemical study of double-salt formation. 



3. The third and last of my arbitrary subdivisions of the 



1 Koppel, loc. cit. 



