RICHARDS AND CUSHMAN. — ATOMIC WEIGHT OP NICKEL. 99 



While it is obvious that certainty is to be obtained only by securing 

 unanimous results from a number of compounds and methods, it is 

 equally obvious that a single method, well worked out, is far better than 

 twenty incomplete ones ; hence the present work was confined for the 

 present to a single compound. 



In our choice of material, we were guided by experience already 

 gained in this Laboratory. The advantage of the bromides as typical 

 compounds in atomic weight determinations has been discussed by one of 

 us in preceding publications, and requires no further mention. The 

 rather meagre current descriptions of the properties of anhydrous nickel 

 bromide did not seem encouraging, but the progress of the research 

 showed that it is possible to dry this salt, weigh it, and dissolve it in 

 water in a perfectly normal condition. Since this was the case, nickelous 

 bromide was naturally chosen as our starting point. 



The Preparation and Properties of Nickelous Bromide. 



Finely divided nickel, when heated to a red heat in a stream of dry 

 bromine vapor, readily forms the bromide, which sublimes at bright 

 redness. The color of the sublimate varies from a pale straw-yellow to 

 a dark bronze-brown, according to the state of aggregation. At a red 

 heat in the presence of traces of air or moisture, the salt loses traces of 

 bromine with the formation of bright green nickelous oxide, unless much 

 hydrobromic acid is present. We have never obtained evidence that an 

 oxybromide is formed under these conditions. 



The sublimed bromide is almost insoluble in cold water, but solution 

 soon becomes apparent to the eye in water at 50^. In water at 90° the 

 salt dissolves less slowly, but a gram still requires an hour or two for its 

 complete solution. When originally free from oxide, the sublimed bro- 

 mide dissolves in water even at the boiling point to form a solution of 

 perfect clearness. According to Berthemot,* a solution of nickelous 

 bromide left for some time in contact with air deposits some flakes of the 

 oxide. We have never met with the slightest evidence of the truth of 

 this statement in the case of a dilute solution. If, however, the nickelous 

 bromide contained only a slight admixture of the oxide, this oxide might 

 escape observation until it had settled out upon the bottom of the vessel. 

 Dilute nitric acid does not materially hasten the solution of the bromide. 

 The sublimed crystalline salt is hygroscopic in character, although not 

 nearly to the extent which we had been led to expect. From several 



* Anal. Chim. Pliys., [2.], XLIV. 389. 



