180 THE TANTALUM LAMP, 



fact, a conductor even wlien cold. This observation induced me to 

 try Avhether vanadic acid could not be electrol^ytically decomposed. 

 In this I succeeded, but the melting point of the vanadium obtained 

 proved too low for the purpose in view. Since the metals niobium 

 and tantalum are members of the vanadium group, niobium having 

 an atomic weight double that of vanadium, while the atomic weight 

 of tantalum is double that of niobium, it was thought that one or both 

 of these metals might prove to have the desired qualities. On ex- 

 perimenting with niobium on the lines adopted for vanadium it ap- 

 peared that this metal has a considerablv higher melting point than 

 (hat of vanadium, l)iit not, however, sufficiently high. Moreover, 

 some of the niobium filaments which I made had a very strong ten- 

 dency to break up when heated by the electric current. 



Tantalum was tried next. I reduced potassium tantalo-fltioride 

 in the manner prescribed l)v Berzelius and Rose and found that the 

 finely divided tantalum so produced became fairly coherent on roll- 

 ing, so that by this treatment metallic strips of it could be made. It 

 was also attempted to work tantalum oxide into the shape of a fila- 

 ment by mixing it with j^araffin and to reduce it directly into the 

 form of a metallic thread. In these experiments there was observed 

 for the first time a minute globule of molten tantalum, and this 

 globule was of sufficient toughness to permit hammering and draw- 

 ing into wire. Following out this observation, tantalum powder was 

 melted in a vacuum, and then it was found that the highly heated 

 metal ])arte(l Avith the gases it contained. In this manner I pro- 

 duced my first filaments of pure metallic tantalum, which were, 

 however, very small. When these had been used in lamps with 

 promise of good results, an attempt was made to devise a definite 

 process of purification. The potassium tantalo-fluoride was reduced 

 to metallic powder; this powder contains a small proportion of oxide 

 and of hydrogen which is absoi'bed during the reduction. When the 

 powder was melted in a vacuum the oxide and absorbed gas disap- 

 peared and a reguline metal remained; on carefully remelting this it 

 became so pure that no apprecial^le impurities could be detected in it. 



The chemical properties of this pure tantalum are very remark- 

 able, and some of them are of such a nature as to lead me to suppose 

 that nol)ody other than myself has ever had metallic tantalum in 

 his hands. AYhen cold, the material strongly resists chemical re- 

 agents; it is not attacked by boiling hydrochloric acid, aqua regia, 

 nitric acid, or sulphuric acid, and it is also indifferent to alkaline 

 solutions; it is attacked solely by hydrofluoric acid. Following the 

 behavior of steel, when heated in the air it assumes a yellow tint at 

 about 400° C, and the tint changes to dark blue when the tantalum 

 is exposed for some time to 500° C, or for a shorter time to 000° C 

 Thin wires of the substance burn with Ioav intensity and without 



