RICHARDS AND CUSHMAN. — ATOMIC WEIGHT OF NICKEL. 101 



to conduct the sublimation in porcelain. In order to collect the product 

 a small porcelain tube, used as tlie receiver, was fitted, telescope fashion, 

 inside of the larger tube containing the substance to be vaporized. In 

 this way the pure crystals could be removed without contamination. 

 Since nickelous bromide is decomposed either by oxygen or by water at 

 a red heat, unless a very large excess of hydrobromic acid gas is present, 

 both oxygen and water must be excluded with scrupulous care during its 

 final sublimation. At first we were not able to accomplish this complete 

 exclusion, so that most of the bromide used in the preliminary series of 

 analyses contained traces of green crystalline oxide, which were carefully 

 collected and weighed. In making the final calculation, the weight of 

 this oxide was subtracted from the total, in order to obtain the weight 

 of the bromide. Subsequently, with improved apparatus, the bromide 



O 



-> c-^PBLI 



Fig. 1. Section of Apparatus for Sublimation. 



A = glass tube for admitting bromine vapor. = outer porcelain tube. I = inner 

 porcelain tube. T = glass outlet tube. F = perforated Fletcher furnace. 

 K = boat containing nickel. S = Sublimed bromide. 



was obtained in a state of satisfactory purity ; it gave a perfectly clear 

 solution in water, leaving nothing to be desired. With the improved 

 arrangement the oxide itself was quickly converted at a dull red heat 

 into the bromide, by means of a stream of hydrobromic acid gas. Hence 

 it was possible to purify nickelous bivMuide contaminated with oxide by 

 simply treating it for a few minutes in this fashion ; in practice, the 

 scheme worked very well. 



Chemical literature contains no satisfactory determination of the spe- 

 cific gravity of the anhydrous salt ; hence, in order to reduce our weigh- 

 ings to the vacuum standard, the following determination was made. 

 3.3196 grams of pure dry nickelous bromide displaced upon one occasion 

 0.6162 gram and upon another 0.6156 gram of rectified toluol at 28°. 

 Since the specific gravity of the toluol at this temperature was 0.860, 



