180 HALL AND SMITH— COLUMBIUM. [May . 9 . 



until the mass fumed strongly and maintaining the temperature 

 until the excess of sulphuric acid had been almost completely driven 

 out. Several hours were required for this. It is necessary in order 

 to get rid of the hydrofluoric acid. The residual mass was boiled 

 with water to extract the bases which dissolved as sulphates. The 

 insoluble hydroxides were thoroughly washed and dissolved in 

 hydrofluoric acid. The first crop of crystals, obtained by evapora- 

 tion with potassium hydroxide, was removed and the mother-liquor 

 then evaporated to dryness with sufficient potassium hydroxide to 

 change all of the metallic acids into double fluorides. A portion 

 of these crystals (first crop) was dissolved in water and the tan- 

 talum removed by adding dilute potassium hydroxide to the solu- 

 tion, which, after the formation of a permanent precipitate, was 

 boiled for some time. The precipitate consisted mainly of potas- 

 sium tantalum oxyfluoride. It was filtered out and the filtrate 

 evaporated to dryness. The residue was baked for some time at 

 2oo°. By this procedure some hydrofluoric acid was expelled and, 

 on taking up the residue with water and boiling, more potassium 

 tantalum oxyfluoride separated. By repetition of this process all 

 of the tantalum was removed from the solution. The only test relied 

 upon for the detection of tantalum was the solution of this pre- 

 cipitate in a drop of hydrofluoric acid and evaporation to crystalli- 

 zation. If needles separated their solubility in water was used to 

 ascertain whether they were potassium tantalum fluoride or potas- 

 sium columbium oxyfluoride. It is true that this test consumes 

 considerable time, yet it is the only satisfactory means of deter- 

 mining with which of the metals the chemist is dealing. The for- 

 mation of a precipitate by protracted boiling of a dilute solution of 

 potassium tantalum fluoride is not conclusive, for Krliss and Nilson 

 (Ber. 1 88 1, 1676) have shown that potassium columbium oxy- 

 fluoride deposits under like conditions a small amount of a salt 

 containing less fluorine. Further, the double fluoride must be 

 recrystallized several times, so that it will be sufficiently free from 

 acid that tantalum, if it is present in small amounts, may be pre- 

 cipitated by boiling. 



Having freed the double fluoride from tantalum it was dissolved 

 in water and hydrogen sulphide conducted through its solution. 

 A slight precipitate of platinum sulphide was obtained. The fil- 

 trate from it was evaporated to dryness and the residue baked. 



