Mr. G. J. Knox's Researches on Fluorine. 193 



the action was allowed to continue for the space of two hours ; 

 at the end of which time the Htmus was found to be reddened, 

 and the gold not acted upon, but a large quantity of sub- 

 fluoride of iron formed. 



In the next experiment I made use of a piece of char- 

 coal, from which the iron had been removed by boiling it in 

 nitric acid ; in this experiment there was no subfluoride ot 

 iron formed, but the vessel was found to contain fluosilicic 

 acid gas. 



In a third experiment a piece of charcoal was employed, 

 which had been previously freed from all metallic impurities 

 and from silica, by being first boiled in pure nitric acid, and 

 afterwards in hydrofluoric acid. Employing this purified 

 charcoal as the positive electrode, I obtained no immediate 

 action upon the litmus paper; but after the action had con- 

 tinued for two hours, it was found to be completely bleached, 

 while the gold had undergone no sensible action. That the 

 bleaching was not due to the action of the vapour of hydro- 

 fluoric acid was ascertained, by leaving litmus paper for se- 

 veral hours in the neck of a platinum retort, from which hy- 

 drofluoric acid was distillinfj. 



The battery was now kept in action for fifteen hours, at 

 the end of which time the vessel being examined, the litmus 

 had disappeared, and the gold-leaf showed signs of having 

 been strongly acted upon, having assumed a dark brownish 

 colour, and having gathered itself into little balls, as if it 

 had undergone the action of heat. The platinum wire was 

 acted upon in those parts where it was in contact with the 

 charcoal, but nowhere else. 



When the platinum wire forming the positive electrode 

 passed through the stopper to the bottom of the vessel, the 

 hydrogen, in place of rising through the perforation in the 

 stopper, as in the former instance, rose now into the receiver, 

 where, upon applying a light, it exploded, showing that it 

 does not enter into combination with fluorine without the aid 

 of heat. The presence of the vapour of hydrofluoric acid in 

 the vessel prevented me from determining by other experi- 

 ments how far fluorine was a supporter of combustion. 



To determine the colour of the gas, a stopper of fluor- 

 spar similar to the former was made to fit one of the trans- 

 parent fluorspar receivers formerly described. The gas 

 evolved in the receiver appeared colourless. 



As the action of the gas upon glass could not be deter- 

 mined, owing to the presence of the vapour of hydrofluoric 

 acid, I fused in a bent tube of German glass (such as is used 

 in organic analysis) fluoride of lead. The wire holding the 



Phil. Ma^. S. 3. Vol. 16. No. 102. March 1840. O 



