Analysis of Sea Water. 17 



should^still remain in it. The ether was directly introduced into 

 a glass bottle, containing a solution of caustic potash fully suffi- 

 cient to discolor the ether, when after evaporation and ignition it 

 was dissolved in water, and carefully neutralized with hydro- 

 chloric acid. The concentrated solution was mixed with a few 

 drops of an ammoniacal solution of chloride of silver prepared 

 thus : one part of a concentrated solution of chloride of silver in 

 ammonia, mixed with one part of ammonia and one part of wa- 

 ter. A few drops of this mixture produced no turbidness in a 

 solution of chloride of sodium, but indicated a very minute quan- 

 tity of bromine. When no further turbidness was produced by 

 an additional drop of this ammoniacal solution of the chloride of 

 silver, the fluid under treatment, which was kept in an open ves- 

 sel, was heated in a sand-bath until the ammonia was almost 

 evaporated. A few drops of the test were again added, until it 

 no longer produced turbidness, when the glass vessel was again 

 placed in a sand-bath, until the fluid, after having been heated, 

 gave no further indication of bromine ; it was then tested again 

 with chlorine. When the proportion of the chlorides to the bro- 

 mides is not too large, scarcely a faint yellowness will be produ- 

 ced; if, however, it is, the bromine must again be separated by 

 chlorine and ether, and the before-mentioned process repeated, 

 when the last traces of bromine will be separated as bromide of 

 silver, which is to be treated like the iodide of silver before it 

 is weighed. In this manner I have been able to detect the 

 smallest proportion of an iodide and bromide when accompanied 

 by a great quantity of chlorides, and have also been enabled to 

 separate them and to ascertain their respective quantities. Should 

 the quantity of iodine be much larger than that of bromine, it 

 would be requisite to evaporate a little of the ammonia ; and al- 

 though the addition of the ammoniacal solution of chloride of 

 silver, employed as a test for iodine, no longer produces turbid- 

 ness, it is still necessary to add another drop of the precipitating 

 fluid, in order to ensure the separation of every trace of iodine. 

 This is the more important, as the iodide of silver is not entirely 

 insoluble in ammonia ; and although the quantity dissolved might 

 be exceedingly minute, still this repetition is necessary in an ac- 

 curate analysis. The same precaution must be observed in the 

 separation of bromine, as bromide of silver is to some extent 

 soluble in ammonia, for it is obvious, that by the addition of the 



Vol. XXXVIII, No. 1.— Oct.-Dec. 1839. 3 



