CHEMICAL SCIENCE. 239 



The first part of my memoir relates to the preparation of pure anhydrous 

 hydrofluoric acid. I prepare this acid by a- new method, by submitting 

 to distillation, in a platinum still, hydrofluate of fluoride of potassium. 



Anhydrous hydrofluoric acid, when thus obtained, is gaseous at the 

 ordinary temperature, but may be condensed by a mixture of ice and salt. 

 It has then the appearance of a very fluid liquid, volatilizing when 

 removed from the refrigerating mixture, acting very powerfully on water, 

 diffusing in the air white fumes, whose intensity may be compared to 

 those of fluoride of boron. Contrary to M. Louyet's assertion, anhydrous 

 hydrofluoric acid attacks glass rapidly. 



I have likewise obtained anhydrous hydrofluoric acid by decomposing 

 in a platinum tube, by means of dry hydrogen, fluoride of lead, which I 

 had placed in a charcoal boat, so as to avoid the action on the platinum of 

 the reduced lead. 



To avoid all the errors made by my predecessors in the study of impure 

 fluorides, and in endeavors to isolate fluorine, I have always used, in my 

 investigations, an acid obtained from an absolutely pure, crystallized 

 hydrofluate of fluoride of potassium. I have thus obtained sometimes 

 new fluorides, sometimes fluorides whose characters had been given by 

 Berzelius. 



Thus my memoir contains a complete study of the fluorides of zinc, 

 iron, and lead, which I obtained in a crystallized state. I have produced 

 protofluoride of tin in very clear and voluminous prisms. I have likewise 

 obtained bifluoride of mercury in well- denned crystals. 



Fluoride of silver, which was considered uncrj-stallizable, may, on the 

 contrary, be deposited from a concentrated solution in crystals, whose form 

 presents the greatest regularity. 



I shall now give some of the consequences resulting from this general 

 study of the fluorides. 



All the fluorides which I have analyzed have been obtained directly by 

 uniting the pure acid with the anhydrous or hydrated metallic oxides. 



Hydrofluoric acid does not react on all the oxides which are attacked by 

 hydrochloric acid. Thus, I found it impossible to combine hydrofluoric 

 acid with auric acid and peroxide of platinum ; finding that in this case 

 hydrofluoric acid behaved like an oxyacid, I endeavored to ascertain 

 whether hydrofluoric acid, which had been long called fluoric acid, did not 

 in reality contain oxygen. These experiments, which presented almost 

 insurmountable difficulties, are described in my memoir. I shall here 

 only mention that they confirmed the constitution of hydrofluoric acid 

 admitted by all chemists, and that they give, in my opinion, the character 

 of a rigorous demonstration to it, which has hitherto been wanting. The 

 result of my researches is, that the fluorides should be divided into three 

 classes, and each of these classes forms a collection of important general 

 properties. 



The first class comprehends the acid fluorides or hydrofluates of fluo- 

 rides; these compounds are very easily formed, heat decomposes them, and 



