240 ANNUAL OF SCIENTIFIC DISCOVERY. 



when they are anhydrous they give neutral fluorides and pure hydrofluoric 

 acid ; in many experiments they may replace hydrofluoric acid. I have 

 employed the salt of potassa for the production of a new organic com- 

 pound, which is somewhat interesting ; I speak of the hydrofluoric ether 

 from common alcohol. I prepare this ether by submitting to distillation, 

 in a platium apparatus, a mixture of sulphovinate and hydrofluate of flu- 

 oride of potassium. I thus obtained gaseous hydrofluoric ether, which, in 

 its general properties, resembles the corresponding compound of wood- 

 spirit, discovered, as is well known, by MM. Dumas and Peligot. 



The second class is composed of neutral and hydrated fluorides : these 

 bodies are characterized by the ease with which they are decomposed into 

 oxides and hydrofluoric acid, when we endeavor to remove the water which 

 enters into their composition ; they behave exactly like real hydrofluates. 

 Thus, crystallized fluoride of silver, which belongs to the class of hydrated 

 fluorides, disengages hydrofluoric acid, and produces oxide of silver when 

 dried even in vacuo ; when hydrated fluoride of silver is heated, it disen- 

 gages hydrofluoric acid and oxygen, and leaves a residue of very pure 

 silver; in this case, then, it acts like a hydrofluate of oxide of silver. 

 Fluoride of mercury, which is likewise hydrated, is decomposed by heat 

 in the same manner as the preceding salt, disengaging hydrofluoric acid, 

 mercury and oxygen. 



The third class comprehends the anhydrous fluorides. These salts are 

 undecomposable by heat, and may be, according to the nature of the metal 

 which they contain, decomposed by oxygen, hydrogen, chlorine, sulphuret 

 of carbon, and steam. 



I confess that I attach great importance to this division of the fluorides 

 into three classes ; it is because they were not aware of it, that the obser- 

 vers who have preceded me have often committed serious mistakes in the 

 study of the fluorides. Thus M. Louyet thought that he could isolate 

 fluorine by decomposing fluoride of mercury by chlorine with heat ; as 

 fluoride of mercury belongs to the second class, and is hydrated, it behaves 

 in all its reactions like a hydrofluate. The gas of M. Louyet was conse- 

 quently a simple mixture of oxygen and hydrofluoric acid. 



After having studied and classified the principal fluorides, my attention 

 naturally turned to those which, from their nature, might lead to the prep- 

 aration of fluorine. 



I first carefully studied the fluorides formed by the difficult oxidizable 

 metals, hoping that by the action of heat, or some other agent, they might 

 disengage fluorine. My researches in this direction have not given any 

 satisfactory result. 



In fact, I found, to my great surprise, that hydrofluoric acid would not 

 combine either with the oxides of gold or of platinum. 



Fluoride of silver, when hydrated, behaves like a hydrofluate, and, with 

 heat, only disengages oxygen and hydrofluoric acid ; when anhydrous, it 

 is undecomposable. 



