THE 



AMERICAN JOURNAL OF SCIENCE. 



[THIRD SERIES.] 



Akt. XLIII. — The Quantitative determination of Niobium; by 

 T. B. Osborne. 



Niobium occurs in nature almost always in connection with 

 tantalum and titanium, in the presence of which it has been 

 heretofore impossible to obtain more than an approximate de- 

 termination on account of the close analogy of the behavior of 

 these three elements, when occurring together, toward reagents. 



Marignac proposed in 1866 (Arch, de Sc, xxv, p. 17) the 

 process now generally employed for the determination of nio- 

 bium, but as he says in describing his method, only approxi- 

 mate results can be obtained by its use. His method depends 

 on the difference in solubility of the potassium fluorine salts of 

 tantalum, niobium and titanium. Tantalum forms a salt of the 

 composition TaF 5 2KF dissolving at ordinary temperatures in 

 150-200 parts of water acidulated with hydrofluoric acid and 

 crystallizing in fine needles. Niobium on the other hand 

 forms a salt of the composition NbOF 3 2KFH 2 0, dissolving in 

 twelve parts of water and crystallizing in scales isomorphous 

 with the corresponding titanium salt; The potassium titanium 

 fluoride dissolves in ninety-six parts of water. 



Marignac adds bifluoride of potassium to the solution of the 

 fluorides of tantalum, niobium and titanium, and concentrates 

 till scales of TiF 5 2KF or NbOF 3 2KF appear. The TaF 6 2KF 

 is filtered and washed with the aid of a pump until the wash- 

 ings no longer give any orange red precipitate with solution 

 of galls after standing two hours. On account of the solubilitv 

 of the TaF 6 2KF the mother liquor containing the NbOF 3 2KF 

 must always be a saturated solution of TaF 6 2KF and likewise 

 a loss must occur on washing, so that even if the process of 



• Am. Joue. Sci.— Thied Seeies, Vol. XXX, No. 179.— Nov., 1885. 

 21 



