FLUORINE. 269 



I should add that in order to lessen the vapor tension of the hydro- 

 fluoric acid the liquid was cooled by means of methyl chloride, which 

 by rapid evaporation produces a cold of — 50° C. 



It was necessary to modify the apparatus, and especially the closing 

 of the U tube. Stoppers of fluorspar smoothly ground did not give 

 me good results. The gum lac or gutta-percha which surrounded them 

 was rapidly attacked by the gas produced at the positive pole. It was 

 necessary, therefore, to resort to a closure by means of platinum screws, 

 and after much groping the experiment was thus arranged. 



The platinum U tube was closed by screw stoppers. Each stopper 

 was formed by a cylinder of fluorspar, carefully inserted in a hollow 

 cylinder of platinum, whose outer surface carried the screw thread. 

 Each stopper of fluorspar was penetrated by a square rod of plati- 

 num. The lower ends of these rods, wliich served as electrodes, dipped 

 into the liquid. Finally, two branches of platinum, soldered to the 

 two limbs of the U tube below the stoppers, but above the level of the 

 liquid, allowed the gases generated by the action of the current to 

 escape. 



In order to obtain pure anhydrous hydrofluoric acid, one begins by 

 preparing the fluohydrate of potassium fluoride, taking all the precau- 

 tions indicated by Fremy. Having obtained this salt in a state of 

 purity, it is dried over the water bath at a temperature of 100° C; 

 and afterwards the capsule containing it is placed in vacuo in presence 

 of sulphuric acid and of caustic potash fused in a silver crucible. The 

 acid and potash are replaced every morning during fifteen days, and 

 the vacuum in the bell jar is always maintained to a pressure of about 

 1 centimetre of mercury. 



During this desiccation it is necessary to pulverize the salt from time 

 to time in an iron mortar, in order to expose fresh surfaces. When the 

 fluohydrate no longer contains water, it falls into fine powder, and can 

 then be used for the preparation of hydrofluoric acid. It is to be noted 

 that well-made fluohydrate of potassium fluoride is much less deli- 

 quescent than the normal fluoride. 



When the fluohydrate is thoroughly dry, it is quickly transferred to 

 a platinum alembic, which has been dried at a red heat a little while 

 before. It is heated gently for an hour or an hour and a half, in order 

 that decomposition may begin slowly; and the ^first portions of the 

 hydrofluoric acid formed, which may contain traces of water remaining 

 in the salt, are rejected. The platinum receiver is then attached to 

 the retort, which is heated more strongly, but still so as to effect the 

 decomiiosition of the fluohydrate somewhat slowly. The receiver is 

 surrounded by a mixture of ice and salt; and from this point all the 

 hydrofluoric acid is condensed as a clear liquid, boiling at 19.5° C, very 

 hygroscopic, and, as we know, fuming abundantly in presence of the 

 moisture of the air. 



During this operation the platinum U tube, dried with the utmost 



