of Mineral Waters. 2!i5 



fully boils. The gas evolved is to be collected in the usual 

 way, in a graduated jar over quicksilver, and submitted to the 

 -following examination :— 



6. Throw up a small quantity of solution of potassa, which, 

 if carbonic acid be present, will absorb it, and the quantity will 

 be shown by the diminution of bulk. 



7. Introduce the remaining air, or a portion of it, into a small 

 bent tube, containing a bit of phosphorus ; heat it so as to 

 kindle the phosphorus, and note the diminution of bulk when 

 cold. It is proportional to the oxygen present, and, if equal to 

 one-fifth of the whole bulk, the gas may be regarded as at- 

 mospheric air *. 



8. If sulphuretted hydrogen be present it may be separated 

 by strong alcoholic solution of iodine, which rapidly absorbs it, 

 and scarcely takes up more than its own volume of carbonic 

 acid gas. Chlorine, added to a mixture of sulphuretted hydrogen 

 and carbonic acid, will also produce the absorption of the former 

 if a little water be present ; but it cannot be conveniently used 

 over mercury. 



9. During the ebullition it not unfrequently happens that a 

 precipitation ensues, indicating that the substances thrown 

 down were dissolved by carbonic acid; and in that case they 

 should be separated upon a filter A, after which the remaining 

 water may be evaporated to dryness in a glazed porcelain basin; 

 the dry residue transferred to a silver capsule, and perfectly 

 desiccated at a temperature not exceeding 500°. B. 



The precipitate A may consist of carbonate of lime, of car- 

 bonate of magnesia, or of oxide of iron ; or it may be a mixture 

 of the three ; dissolve it in dilute muriatic acid, and add oxalic 

 acid which throws down oxalate of lime ; separate this by fil- 

 tration, and saturate the filtrated portion with carbonate of 

 ammonia, which precipitates the peroxide of iron, and having 

 removed this, evaporate the residuary mixture, and expose the 



• In separating oxygen a solution of nitric oxide in protosulphate of iron 

 may sometimes conveniently be employed, but itdoes not give so accurate 

 a r«ftult as the action of phosphonw. 



