1824.] M. Arfwedson on Uranium. 255 



colour of the liquid changes to a yellow. It must now be decom- 

 posed by means of carbonate of ammonia added in excess, which 

 will take up the oxide of uranium mixed with oxides of cobalt 

 and zinc ; but leaves a great quantity of oxide of iron undis- 

 solved. Should the solution even contain a portion of earth, 

 which did not happen in my experiments, almost the whole of it 

 would be separated mixed with the oxide of iron. The filtered 

 solution is afterwards made to boil, and the boiling is continued 

 as long as the carbonate of ammonia is disengaged. A portion 

 of the oxide of cobalt remains in the solution, which acquires a 

 faint reddish colour ; but another portion of it is precipitated 

 along with the oxide of uranium ; which contains likewise the 

 zinc. The precipitate is collected on the filter, washed, and 

 dried. It is then to be heated to redness ; by which it loses its 

 yellow colour, and becomes dark-green. In this state it must 

 be macerated for some time in dilute muriatic acid, which dis- 

 solves the oxides of cobalt and zinc, together with a small por- 

 tion of peroxide of uranium. This portion was probably united 

 as an acid with the two bases. Pure protoxide of uranium 

 remains undissolved. If the muriatic solution be precipitated 

 with caustic ammonia in excess, we obtain oxide of uranium 

 combined with oxides of cobalt and zinc. From 38-i- parts of 

 pechblende treated in this way, I obtained about 25 parts of 

 protoxide of uranium. This amounts nearly to 65 per cent, 

 which is 15 parts less than the quantity stated by Klaproth. 



Metallic Uranium and Protoxide of Uranium. 



The experiments hitherto made to obtain uranium in the 

 metallic form have been all conducted in charcoal crucibles with 

 or without additions. It is consequently probable, even if the 

 oxide of uranium operated upon had been quite pure, that the 

 metal obtained contained charcoal, or some other substance 

 derived from the fluxes employed in the reduction. In which 

 case the properties of the product might differ materially from 

 those of the pure metal. Fortunately chemists have found out 

 a method of avoiding these inconveniences in the reduction of 

 metals ; for it is now known that several metallic oxides may be 

 deprived of their oxygen by means of hydrogen gas. I deter- 

 mined, therefore, to try whether oxide of uranium could not be 

 brought to the metallic state by this method. If I succeeded I 

 obtained naturally the proportion of oxygen in the oxide deter- 

 mined with the requisite exactness. 



The apparatus employed in this process was a piece of a com- 

 mon barometer tube, blown into a small globe about the middle 

 part. This tube was in the first place weighed, and then a 

 portion of finely pulverized protoxide of uranium which had been 

 heated to redness was introduced into the globular part of the 

 tube. Before determining the weight of this powder, the glass 



