CONTRIBUTIONS FROM THE CHEMICAL LABORATORY 

 OF HARVARD COLLEGE. 



A REVISION OF THE ATOMIC WEIGHT OF IRON. 



SECOND PAPER. — THE ANALYSIS OF FERROUS BROMIDE. 



By Gregory Paul Baxter. 



Presented by T. W. Richards, October 14, 1903. Received October 3, 1903. 



Four years ago a determination of the atomic weight of iron, made 

 in this laboratory by reduction of the oxide in hydrogen, led to the value 

 55.883 (O = 16.000).* Since the atomic weight in use at that time, 

 56.02, differs so materially from the above value, it seemed advisable to 

 redetermine the constant in question by a radically different method. 



Bromides may be analyzed with the greatest ease and exactness, pro- 

 vided they can be obtained in a state of purity. Sirce ferric bromide is 

 far too unstable at high temperatures to permit thorough drying of the 

 salt, ferrous bromide was chosen as the substance for analysis. This 

 salt rapidly oxidizes in the presence of moist air ; but the ease with 

 which ferric bromide is decomposed by heat into ferrous bromide and 

 bromine is an advantage in the preparation of ferrous bromide, for it 

 proved possible completely to free the latter salt from ferric impurity by 

 subjecting it to a high temperature in an atmosphere of dry hydro- 

 broraic acid gas. 



Ferrous bromide was prepared by heating metallic iron in a current 

 of dry nitrogen and hydrobromic acid gases, at a temperature sufficiently 

 high to sublime the resulting salt. When the salt was removed from 

 the sublimation tube, it necessarily came in contact with moist air, and 

 thus become covered with a thin coating of ferric salt. It was then sub- 

 jected to prolonged heating in a current of dry nitrogen and hydrobromic 

 acid gases. In order to determine whether the ferric salt was thus 

 decomposed, samples of the product were dissolved in a freshly boiled, 

 acid solution of amnionic sulphocyanate. If the hydrobromic acid and 



* Richards and Baxter, These Proceedings, 35, 253. 



