442 PROCEEDINGS OP THE AMERICAN ACADEMY. 



The sodic carbonate was maintained in slight excess, and this excess 

 finally neutralized by some of the pure hydrobromic acid described 

 above. The salt was recrystallized twice with thorough draining. Very 

 pure material may be made in this way, if the ammonia, bromine, and 

 sodic carbonate are all pure. 



The hydrobromic acid used for the fifth and last sample of sodic 

 bromide was made by bubbling electrolytic hydrogen through bromine, 

 and passing the mixture through a hot hard glass tube. The bromine 

 used was in the first place purified by distillation from a bromide, and 

 a further attempt was made to purify it by passing the vapor when 

 mixed with hydrogen before heating over a counter current of a solution 

 of a pure bromide (made from a portion of the acid itself) flowing over 

 beads in tall glass columns. All the apparatus was constructed of glass, 

 without rubber connections, and the current was run slowly ; but it is 

 doubtful if great gain was effected by contact with the solution. The 

 reason for the slow purification of a vapor by this means is undoubtedly 

 the fact that the reaction can take place only on the limited surface 

 between the gaseous and liquid phase. 5 



A large quantity of acid thus prepared was neutralized by a new 

 specimen of sodic bicarbonate even purer than before. The bicarbonate 

 was made by precipitating a solution of pure often recrystallized sodic 

 carbonate in platinum through the long continued action of carbon 

 dioxide. The easiest method consists in placing the open dish under a 

 bell-jar and keeping the latter full of the gas at a pressure slightly 

 greater than that of the atmosphere. The carbon dioxide was gener- 

 ated by heating pure dry sodic bicarbonate in a large bomb-shaped 

 vessel, and washed once with a dilute solution of the same salt. The 

 sodic bromide thus prepared served for the last two final determinations 

 given below, after having been recrystallized three times. 



In order to remove all doubt as to the purity of salt, a sample prepared 

 in this way was analyzed by the precipitation of argentic bromide with 

 all the precautions of an atomic weight determination, and the result 

 was compared with a similar analysis of a sample made with hydro- 

 bromic acid known to be pure. The latter was kindly given by 

 Professor Baxter and Mr. H. L. Frevert, having been made by them 



5 As a matter of fact, one sample of acid prepared in this way was found by 

 Professor Baxter, who kindly interested himself in the matter, to contain appreci- 

 able traces of chlorine. Evidence is given later showing that the particular 

 sample V of sodic bromide prepared from acid thus made was free from this 

 impurity, however. 



