J. G. Dinwiddie — Hydrofluoric and Fluosilicic Acids. . 425 



ing potassium fluosilicate. In e the precipitate weighed four 

 or five per cent too much. It was decomposed by ammonia 

 and the solution was filtered to get rid of asbestos and then 

 acidified with hydrochloric acid and evaporated. The residue 

 was heated with hydrochloric acid to drive off all silica and 

 fluorine and to convert the entire residue to potassium chloride. 

 The residue of potassium chloride so obtained was found to 

 weigh several per cent more than if the potassium had been 

 present in the precipitate only as potassium fluosilicate. 



The precipitate of potassium fluosilicate formed in the pres- 

 ence of free hydrofluoric acid contains an excess of potassium 

 salt as well as free hydrofluoric acid, and it is evident that the 

 fluosilicic acid in a mixture of the two acids cannot be esti- 

 mated by determining the potassium in the precipitate or in 

 the filtrate. 



Since the fluosilicic acid carries down with it free hydro- 

 fluoric acid and fluorides when it is precipitated in the presence 

 of these, it was determined, if possible, to precipitate the hydro- 

 fluoric acid as some insoluble fluoride, and, having thus removed 

 it from the solution, to determine the fluosilicic acid remaining. 



Accordingly, to a mixture of fluosilicic and hydrofluoric 

 acids was added a neutral solution of calcium chloride sufficient 

 to precipitate as calcium fluoride all of the fluorine present as 

 hydrofluoric acid. Next was added potassium chloride suffici- 

 ent to form with the fluosilicic acid, potassium fluosilicate, and 

 to the mixture was added alcohol to make the final solution 

 consist of 50 per cent alcohol. The precipitate obtained under 

 these conditions was found very difficult to filter, so the mix- 

 ture was made up to a definite volume and after allowing to 

 settle the clear supernatant liquid was decanted through a dry 

 filter. Aliquot parts of the total volume were then titrated 

 with standard alkali. 



If all of the hydrofluoric acid were actually precipitated as 

 calcium fluoride and all of the fluosilicic acid as potassium 

 fluosilicate, then for each molecule of HF originally present 

 there should be in the filtrate one molecule of hydrochloric 

 acid, and for each molecule of fluosilicic acid two molecules of 

 HC1, according to the reactions, 



2HF + CaCl 2 > CaF 2 + 2HC1 



H a SiF 8 + 2KC1 > K 2 SiF 6 + 2HC1. 



Upon titration of this filtrate, therefore, the alkali required 

 should be equal to that which would be required in water solu- 

 tion for the hydrofluoric acid originally present, plus one-third 

 of that for the fluosilicic acid originally present. Thus the 

 difference between the original water titration and the titration 



