24 PROCEEDINGS OF THE AMERICAN ACADEMY 



Analyses 6, 7, 9, and 12 the cloudiness of the solution was so 

 slight as to be inessential, while in Analyses 2, 5, 8, 10, 11, 14, 

 18, and 19 the neutral solution of the baric bromide was abso- 

 lutely clear. 



The method of weighing was precisely similar to that employed 

 with anhydrous cupric sulphate.* Afterwards the salt was dissolved 

 in the purest boiled water, and the traces of baric hydrate and car- 

 bonate present were determined in the manner which has been 

 already described, f The appropriate correction having been ap- 

 plied to the last weight of the baric bromide, the solution was 

 diluted and transferred to a glass-stoppered Erlenmeyer flask. To 

 this was added about the corresponding amount of accurately weighed 

 silver, which had been dissolved in the purest nitric acid with all 

 possible precautions. I The argentic nitrate solution had been 

 freed from the lower oxides of nitrogen by long heating at 100° 

 in an inclined flask, and both solutions were quite cold at the 

 moment of precipitation. Daylight was carefully excluded during 

 this operation, as well as during the subsequent ones. After violent 

 shaking, the precipitate was allowed to settle, and the excess of 

 bromine or silver was determined by titration. It is well known 

 that even here there is a slight difference between the end points, 

 although the possible error is very much less than with the chloride. 

 In the table below the mean between the two extremes is recorded; 

 and in general the observations and method of treatment corre- 

 sponded essentially with those since published for the late J. L. 

 Hoskyns Abrahall in the account of his interesting work on the 

 atomic weight of boron, § to which the reader is referred. Tlie end 

 point was always determined by titrating backward and forward 

 until there could be no doubt of its accuracy, a cubic centimetre of 

 each of the solutions emploj^ed corresponding to one milligramme 

 of silver. In the final experiments the solutions were weighed as 

 well as measured. For these experiments, a dark room was built, 

 and provided with an arrangement, essentially similar to that de- 

 scribed by Stas, || for condensing a beam of yellow light upon the 

 surface of the liquid in the flask, leaving the precipitate in 

 darkness. 



In some cases the baric bromide was poured into the argentic 



— — ■ — — — ■ . . — .— < 



* These Proceedings, XXVI. 243, 252. t This paper, page 12. 



X These Proceedings, XXV. 198. 



§ Edited by T. Evvan and P. J. Hartog, J. Chem. Soc. Proc, 1892, p. 663. 

 II Aronstein's translation, page 45. 



