AND A NEW NIOBIUM CHLORIDE. 189 



traces of oxygen, the very hygroscopic pentachloride was 

 prepared in a narrow tube, and this was transferred bodily 

 into the wider tube in which the preliminary heating with 

 hydrogen takes place. 



The metal thus obtained was analyzed by placing a 

 weighed quantity in a porcelain boat and heating it in a 

 glass tube in a current of air. The metal takes fire and 

 burns brightly to the pentoxide, a small trace of oxychloride 

 being usually deposited, existing as impurity in the metal. 

 In one instance, when the metal was allowed to oxidize 

 very slowly, the formation of an indigo-blue-coloured lower 

 oxide was first observed ; this afterwards passed into the 

 yellow pentoxide on further heating. 



The following analyses were made with metal obtained 

 in two distinct operations : — 



gram. 



I. Weight of metallic niobium taken o"3963 



„ niobium pentoxide obtained ... 0-5430 



This corresponds to 03810 of metal; or the substance 

 contains 96*14 per cent, of niobium. 



gram. 



II. Weight of metallic niobium taken 0-3900 



„ niobium pentoxide obtained ... 0-5403 



This corresponds to 0*3790 of metal ; or the substance 

 contains 97*18 per cent, of niobium. 



That the substance thus prepared is not, as has been 

 supposed, a hydride, is shown by the fact that the second 

 of the above specimens on analysis was found to contain 

 only 0*27 per cent, of hydrogen. 



Metallic niobium possesses a steel-grey colour and bright 

 metallic lustre. It is remarkable that it is not attacked by 

 hydrochloric or nitric acids, either dilute or concentrated, 

 and aqua regia acts only very slightly upon it. Strong sul- 



