30 PROCEEDINGS OF THE AMERICAN ACADEMY 



and fourth columns : the arguments indicating the absence of this 

 insidious impurity were discussed at length in the first part of the 

 paper.* 



The agreement between the individual results is as close as could 

 reasonably be expected, when one considers the small amounts of 

 material vised in some cases. It may be concluded, then, that a 

 hundred parts of silver correspond to about 137.747 parts of anhy- 

 drous baric bromide, no matter what may be the method used for 

 its preparation. If the salt contains an impurity, it is strangely 

 constant in amount. 



The Atomic Weight of Barium. 



From the results which have just been given, the atomic weight 

 of barium is very readily computed. In the following table are given 

 the values corresponding to the three standards at present in use. 



From the Ratio of Silver to Baric Bromide. 

 If Silver = 107.93, and Bromine = 79.955, Barium = 137.426 

 If Silver = 107. 66, and Bromine = 79.755, Barium = 137.083 

 If Silver = 107. 12, (Oxygen = 15.88,) Barium = 136.396 



Greatest variations from the mean, •< 



' I _ 0.040. 



From the Ratio of Argentic Bromide to Baric Bromide. 



If Argentic Bromide = 187.885, Barium = 137.431 



If Argentic Bromide = 187.415, Barium = 137. 089 



If Argentic Bromide = 186.476, Barium = 136.401 



Greatest variation from the mean, ±0.054. 



It is not very difficult to explain the reason for the difference 

 between this new value, 137.43, and the old one, 137.10. The in- 

 complete knowledge regarding the end point of chlorine reaction in 

 1858 is probably responsible for a part of the difference, and a 

 portion more may possibly be explained by the impurities which 

 were assumed to be inessential. But it has already been said that 

 a discussion of the results of thirty-five years ago can be of little 

 value. The only true solution of the question is the experimental 

 one. In the near future, I hope to continue the investigation which 

 is herewith commenced, as well as to begin a similar research upon 

 strontium and calcium. 



Cambridge, December 24, 1892. 



* This paper, page 12. 



