384 PROCEEDINGS OF THE AMERICAN ACADEMY. 



perature below the softeniug poiut of soft glass. At the moment 

 when it was desired to close the bottle, it was only necessary to ele- 

 vate the desiccator tube from the horizontal to the vertical position, 

 and the hot stopper fell automatically into the equally hot bottle. 

 The desiccator tube was now closed above, and allowed to cool at least 

 four hours in the balance room. It is needless to say that before 

 taking the final weighing of the bottle its stopper was loosened. 



Having thus obtained as nearly as possible the true weight of the 

 typical salt of strontium, the remainder of the analysis was conducted 

 in a manner essentially similar to that adopted in the case of baric 

 bromide.* Since it is unnecessary to describe again most of the pre- 

 cautions, nothing will be noted below excepting those particulars in 

 which the details of the work differed from those already given. 

 Two analyses, which were vitiated by knowu errors, are omitted from 

 the tables. 



The Ratio of Silver to Strontic Bromide. 



First Series. — In this series a slight excess of silver was taken, dis- 

 solved, and diluted with at least a hundred times its weight of water, 

 and added to the strontic bromide in a glass-stoppered flask. After 

 the usual long continued shaking, the precipitate was collected upon a 

 Gooch crucible, and the excess of silver in the evaporated filtrate and 

 first five or six wash waters was determined after Volhard's method. f 

 Upon subtracting this small excess of silver from the total, the 

 amount corresponding to the strontic bromide remains. This method 

 is not a very satisfactory one, the final result being probably too low, 

 because of loss of a portion of the slight excess of silver. 



Second Series. — Here the end point of the reaction was determined 

 by titration after the method of Abrahall,t very weak solutions of 

 silver and hydrobromic acid being used to titrate backwards and for- 

 wards. The mean readinji was taken in each case, and the method of 

 procedure resembled exactly the work with barium. These results 

 are much more trustworthy than the last. In several cases the sample 

 of strontic bromide was first analyzed by this method, and subsequently 

 an excess of silver nitrate was added and the preceding method was 

 applied. 



Third Series. — For this series a new method was devised. Accord- 



* These Proceedings, XXVIII. 24. J These Proceedings, XXVIII. 24. 



t These Proceedings, XXIX. 66. 



