T. B. Osborne — Quantitative determination of Niobium. 333 



The foregoing series of determinations show that the reduc- 

 tion of the niobium is quite constant for a given strength of 

 acid depending not on its amount but on its strength. The 

 determinations were made by dissolving NbOF 3 2KF in water 

 and adding the indicated amount of acid. The first eight deter- 

 minations, where 50 c. c. of dilute acid were used, give an 

 average of 34*78. With 40 c. c. the same amount of water wa3 

 used to dissolve the fluoride, the strength was consequently 

 less and the average lower, being 34*20. With 30 c. c. 33*56 

 was found. When concentrated acid was used the amount of 

 water required to dissolve the fluorides was small in compari- 

 son with the amount of the acid in solution, so that the amount 

 of water used exerted but little influence on the reduction, 

 but the increased strength of acid raised the average to 37*80. 

 The results obtained when concentrated acid was used to dis- 

 solve the fluoride are higher still, the average being 38*9£. 



Thus we have 



For 30 c. c. HC1 sp. gr. VI 33-56 per cent Nb 2 6 



40 c. c. " " 34-20 " " 



50 e. c. " " 34-78 " 



Concentrated HC1 37-80 " 



Fluorides dissolved in cone. HC1 38 - 94 " " 



It appears therefore that the amount of reduction increases 

 with the concentration of the acid. The formula for an oxide 

 corresponding to the reduction in the last case would be Nb 8 18 . 

 Marignac sought to obtain a formula for this oxide in the same 

 way and obtained Nb 3 5 . This is very nearly equal to ]STb 8 13 

 since, for eight atoms of niobium there would be thirteen and 

 a third atoms of oxygen according to his formula. It is proba- 

 ble that neither of these oxides represents the oxide formed, 

 but that a partial reduction has taken place as will appear from 

 the results obtained from niobic acid in the following manner. 

 Ten grams of JSl"bOF 3 2KF and five grams of TaF 6 2KF were 

 accurately weighed and heated with sulphuric acid in a plati- 

 num dish till all the hydrofluoric acid was removed. The acid 

 solution was poured into water and precipitated with ammonia, 

 washed by decantation and then thoroughly dried at 100° C. 

 Weighed portions were then ignited and the total amount of 

 mixed oxides of tantalum and niobium determined. The per- 

 centage of niobic oxide in these mixed oxides was calculated 

 from the relative amount of the fluorides of each metal taken. 

 In this way it was found that the mixture of tantalic and niobic 

 agicl precipitated and dried at 100° contained 47*36 per cent of 

 Kb 2 5 . Weighed portions were then dissolved in as small an 

 amount of hydrofluoric acid as possible and 50 c. c. of concen- 

 trated acid added and the reduction carried on at a temperature 



