FLUORINE. 267 



free I did not fail to try that reaction, and we shall see later that it 

 succeeded perfectly. 



After these first experiments upon silicon fluoride, I took up the 

 investigation of the compounds of fluorine with phosphorus. 



Thorpe discovered the compound PF5, the pentafluoride of phospho- 

 rus. I prepared the trifluoride PF3, and I gave all my attention to the 

 reactions which might lead to its decomposition. I made the experi- 

 Dient of which Humphry Davy had dreamed, of burning phosphorus 

 trifluoride in oxygen, and I found that there was no formation of phos- 

 phoric acid with liberation of fluorine, as the English scholar had 

 expected, but that the trifluoride and the oxygen united to form a new 

 gas — phosphorus oxyfluoride. 



Here is a new example of the ease with which fluorine yields prod- 

 ucts of addition. 



I next tried, but without avail, the action of the induction spark 

 upon phosphorus trifluoride. The pentafluoride of phosphorus discov- 

 ered by Thorpe has, however, been decomposed by very strong sparks 

 into the trifluoride and fluorine. * 



This experiment was made in a glass tube over mercury. You will 

 see that at once fluoride of mercury and fluoride of silicon were formed. 

 There was no hope, under these conditions, of preserving the fluorine, 

 even when it was diluted by an excess of pentafluoride. I then thought 

 of another reaction. 



We have known, since the researches of Fremy, that the fluoride of 

 platinum, produced during the electrolysis of alkaline fluorides, decom- 

 poses at a high temperature. Having found that the fluorides of phos- 

 phorus are easily absorbed by hot platinum sponge, with the final pro- 

 duction of platinum phosphide, I thought that this method of preparing 

 fluoride of platinum might lead to the isolation of fluorine. Heating 

 gently at first, the absorption of phosphorus fluoride should give a 

 mixture of phosphorus and platinum fluoride; and, the qiuantity of 

 the latter being sufficient, a subsequent increase of temperature should 

 disengage fluorine. These experiments and others analogous to them 

 were tried under conditions most favorable to success. They yielded 

 interesting results, which were not, however, sufficiently sharp to settle 

 the question of the isolation of fluorine. 



While still pursuing the above-mentioned studies, I prepared the 

 trifluoride of arsenic, which had already been obtained by Dumas in 

 great purity. I determined its physical constants, together with some 

 new properties, and investigated with great care the action of the elec- 

 tric current upon it. 



The fluoride of arsenic, liquid at ordinary temperatures (a binary 

 compound, formed by a solid, arsenic, with a gas, fluorine) seemed to be 

 admirably suited to electrolytic experiments, 



I was obliged at four different times to interrupt these researches 

 upon arsenious fluoride, the manipulation of which is more dangerous 



