OP ARTS AND SCIENCES. 31 



precipitate, we may refer to Vol. XI I. page 124, of these Proceedings. 

 Tlie results were as follows : — 



For the second set of experiments bromide of silver precipitated, 

 washed, and dried as described (loc. cit.), was melted in a platinum cru- 

 cible, and then reduced by a voltaic battery under dilute sul{)huric 

 acid. This process was devised and perfected by Mr. L. P. Kiunicutt, 

 Assistant in this Laboratory, who had the great kindness to conduct 

 the reduction in the followinsc determinations. 



These results show conclusively the great accuracy of Mr. Kirmi- 

 cutt's process, which he will describe in detail in another place. The 

 mean of the two sets of results gives for the per cent of silver 57.445, 

 the theoretical per cent (Ag = 108 and Br = 80) being 57.446. If 

 we throw out No. 2 of second series, which is obviously less trust- 

 worthy than the other two determinations of the same series, on 

 account of the comparatively small amount of material used, the mean 

 of the remaining five determinations corresponds absolutely to theory, 

 and the total result, therefore, leaves no doubt whatever as to the 

 absolute purity of the materials employed. 



Bromide of cadmium was now prepared by dissolving pure carbon- 

 ate of cadmium in pure hydrobromic acid, and subliming the product 

 previously dried at 200° in a current of pure and perfectly dry carbonic 

 dioxide gas. The carbonic dioxide was prepared by mixing bicarbon- 

 ate of soda and sulphuric acid with a large volume of water in a strong 

 generator, and drawing off the gas under pressure through appropri- 

 ate washers and driers; and the apparatus used for sublimin"- the 

 bromide of cadmium was similar to that described in these Proceed- 



