50 SIXTH REPORT — 183C. 



when present in a M'atcr, together with chlorine, is staled in my 

 work on the Atomic Theory *. 



It is nothing more than an application of the method sug- 

 gested by M. Gay-Liissac for calculating the proportions of soda 

 and potass, to the case of bromine and clilorine, and labours in 

 common with it under the objection, that the inference is de- 

 duced, not from a single experimeiit, but from a comparison 

 of at least two ; and that a very trifling inaccuracy in either, 

 being multiplied in the calculation founded on them, vitiates 

 the whole result. 



It VI ould be well, therefore, if a direct method of determining 

 the same could be hit upon ; and for this reason I set down one 

 suggested by Lowig, which has already found a place in Professor 

 Johnston's Report on Chemistry, published in the first volume of 

 our Reports. 



The dried mixture of chloride and bromide is to be heated in 

 a stream of chlorine, so long as any bromine appears to be dis- 

 engaged. The chlorine and bromine which pass over are re- 

 ceived into a solution of caustic potass, by which chloride of 

 potassium and chlorate of potass, together with bromate of potass, 

 are produced. 



Having neutralized the potass with nitric acid, nitrate of silver 

 is added to precipitate the chlorine and the bromic acid. 



The precipitate, after being washed, is introduced moist into 

 a bottle, and barytic water added. A soluble bromate of barytes 

 is thus formed, whilst the chloride remains untouched. The 

 solution being p(mred off, the excess of barytes is separated by 

 carbonic acid, and the bromate of barytes is thus left in a state 

 of purity. 



Dr. Osannt has lately suggested another mode of separating 

 these two principles. 



It depends on the greater volatility of chlorine than bromine, 

 and on the cii'cumstance, that chloride of silver becomes of a vio- 

 let colour after exposure to light, whilst bromide of silver is 

 rendered greyish black. 



He therefore expels the chlorine and bromine by means of 

 sulphuric acid, slowly distils over the two, and makes them pass 

 into a solution of nitrate of silver. The precipitate is from time 

 to time tested by exposure to light, and when found to assume 

 the appearance belonging to bromide of silver, that which comes 

 over is set apart, and reckoned as such. 



In order to obviate the objection, arising from the circum- 

 stance, that there is an intermediate period when the chlorine 



* Introduction to the Atomic Theory, p. 89. The same method was followed 

 by Dr. Ure in his analysis of the Ashby water; Phil. Transactions, 1834. 

 f PoggcnAorff's Annalcn, 1831. 



