Analysis of Sea Water. 15 



when the before-mentioned test readily indicated a slight trace 

 of iodine. 



With respect to the quantity of iodine in sea-water, it is evi- 

 dently very minute, 174 pounds troy not containing one grain. 

 This is remarkable when we consider the comparatively large 

 quantity of iodine and bromine present in sea plants and animals, 

 hence we must conclude that these principles are concentrated 

 by vital action. 



Bromine, when present in fluids, is easily detected by chlorine, 

 which produces a yellow color. If present in very minute quan- 

 tity the fluid must first be concentrated. But when iodine is 

 present we cannot apply this test, as bromides and iodides are 

 both decomposed by it ; and we cannot separate them, even by 

 means of ether, as iodine is soluble in that menstruum, and also 

 possesses greater coloring properties than bromine. From these 

 causes this test is useless when iodine is present, and is only cer- 

 tain when we are previously assured of the absence of that sub- 

 stance. 



The following process for the separation of iodine, chlorine, 

 and bromine in fluids containing these substances in very small 

 quantities has given me satisfactory results, as I had anticipated 

 by previous experiment. The fluid while boiling was mixed 

 with a sufiicient proportion of caustic potash ; my object in this 

 was to decompose the earthy salts, and at the same time prevent 

 the iodine and bromine from being dissipated by heat. The fil- 

 tered fluid was then evaporated to dryness and ignited, and the 

 resulting mass, after having been dissolved, concentrated, and 

 neutraUzed with hydrochloric acid, was carefully mixed, drop 

 by drop, with an ammoniacal solution of chloride of silver pre- 

 pared by mixing one part of a saturated solution of recently pre- 

 cipitated chloride of silver in ammonia with one of liquid am- 

 monia (sp. grav. 0-935) and two parts of water. If to a concen- 

 trated solution of chloride of sodium containing one thirtieth 

 part of a bromide, we add a few drops of this ammoniacal solu- 

 tion of chloride of silver, the solution will remain clear ; but if 

 the most minute particle of an iodide be present, it will be ren- 

 dered turbid. 



To the fluid under examination I added gradually, drop by 

 drop, the solution of ammoniacal chloride of silver, leaving time 

 between each successive addition for the precipitate of iodide of 



