108 



PROCEEDINGS OP THE AMERICAN ACADEMY. 



milligram. The balance has already been described in detail;* the 

 weights were of course compared and standardized with great care, and 

 were used for no other work during the progress of tiiis. 



Having been weighed with accuracy, the nickelous bromide was dis- 

 solved in pure waim water in a flask, and i'rom this was transferred to 

 the large beaker flask in which the precipitation was to take place. The 

 platinum boat in which the salt had been treated remained invariable in 

 weight, showing that it had not been attacked by hydrobromic acid at a 

 high temperature. 



<^ B 



Fig. 3. Apparatus for igniting Nickelous Bromide in any desired 



Mixture of Gases. 



The use of rubber was confinerl to tlie first part of tbis train, wliere it could do 

 no barm (A B C D E F and A M N P). 



As has been said already, the salt used in the preliminary series was 

 contaminated with a small amount of nickelous oxide, which was filtered 

 off and weighed. The amount of this impurity is given simply to show 

 that the slight irregularity of the results was not dependent upon the 

 adulteration ; the weights of nickelous bromide given are those left after 

 the subtraction of the weight of the oxide. All the bromine contained 

 in the solution was precipitated in these seven analyses by means of an 



* These Proceedings, XXVI. 242. 



