URANIUM. 11 



from this number : If protoxide of uranium be a 

 compound of 1 atom oxygen and 1 atom ura- 

 nium, we have 



3-f)907o7 : 100 : : 1 : 27'095 = atomic weight of uranium. 



This is a little higher than 26 the result 

 which I obtained from the analysis of the triple 

 salt described above. If 26 be the true atomic 

 weight of uranium, as I think it is, then 100 

 parts of this metal, when converted into prot- 

 oxide, should combine with 3'846l parts of oxy- 

 gen, which is only a little higher than Arfwed- 

 son's last result. Indeed, I conceive it to be 

 almost impossible to come to the exact truth by 

 Arfwedson's methods. I have repeated the whole 

 of his experiments with great care ; but my re- 

 sults differ fully as much from each other as the 

 difference between 3*73 and 3'84 : and some of 

 my results gave the quantity of oxygen as high 

 as 3-84. 



4. M. Arfwedson made a set of experiments Arfwed. 

 likewise, to determine the quantity of oxygen in S 

 peroxide of uranium, which I shall state before 

 giving an account of those which I myself made 

 with the same object in view. 



(1.) He formed uraniate of lead by mixing 

 together solutions of permuriate of uranium and 

 muriate of lead, and mixing the liquid with 

 caustic ammonia. A yellow precipitate fell, con- 

 sisting of the two oxides chemically combined. 



