74 E. DIVEES AND T. HAGA. 



Ba3(NS207)o + H20, shows the normal sodium salt behaving with a 

 salt of barium somewhat as it does with zinc, manganese, copper, and 

 other salts, that is, as dipotassiam salt and potassium hydroxide 

 (V- 47). 



Fremy, by treating the Kg salt with barium chloride, got a 

 preci|)itate, gelp.tinous at first, afterwards crystalhne, the composition 

 of which, as given by him, may be expressed (1) pretty closely by 

 (HO)3]]:!,Ks(NS20,)„ or (2) l)y BaKNS.O; as an approximation suf- 

 iiciently near to Fremy 's analytical results, considering the lialjility to 

 impurity of his preparations, and his imperfect methods of analysis : — 



Calculated. Found. 



(1) (2) (Fremy) 



Barium 39-29 37-43 39-90 



l^tassium 11-19 10-65 11-55 



Sulphur lG-07 17-49 15-80 



He must have kept the potassium salt in excess when precipit- 

 ating. He found that baryta-water gave the same salt. Although 

 much reliance cannot, we admit, be placed on Fremy's analytical 

 results as closely accurate, we are not inclined to adopt the simpler 

 formula rather than the other of the two we have calculated ; fjr oiu- 

 own results, here and in the case of strontium and lead salts, where 

 we have been sure of the comparative purity of our preparations and 

 accuracy of our methods of analysis, have taught us that compound 

 oximidosulphonates are generally of complex composition. Claus, 

 after experimenting for himself, pronounced Fremy's results to be 

 worthless. He found that after ten minutes or so, the barium pre- 

 cipitates always began to decompose and to contain sulphate. We 

 have not met with this difficulty, the barium salts (except the tAvo- 

 thirds normal salt, as already noticed) proving to be stable, and free 

 from sulphate when analysed. As the barium salts when acidified are 



