apkil — sept. 1858.] Brine-Springs of Cheshire. 113 



sodium, calcium, magnesium, chlorine and sulphuric acid. Of the 

 bodies occurring in less quantity, a minute trace of potassium was 

 found, a small quantity of iron with a little alumina and consider- 

 able traces of manganese, not, however, in proportions capable of 

 determination. Bromine was ascertained to exist in rather large 

 quantity, the unconcentrated brines becoming distinctly yellow 

 upon the passage of a few bubbles of chlorine ; this element was, 

 therefore, determined by the method recommended by Fehling,^ 

 which depends upon the fact, that if a solution of chloride of 

 sodium containing a small amount of bromide be insufficiently 

 precipitated by nitrate of silver, the precipitate nevertheless con- 

 tains all the bromine present ; the bromine in the mixed precipi- 

 tate of chloride and bromide of silver is then determined by the 

 passage of chlorine over the fused mass in the usual way. The 

 results yielded by this process, which were carefully and frequent- 

 ly repeated upon very different quantities of brine, agreed among 

 themselves in the closest manner. The existence of iodine was 

 proved in all the waters, but the quantity present was minute in 

 the extreme. In one case also an indication of fluorine was ob- 

 tained by the apparently etching action exerted upon a watch- 

 glass in the application of the test now usually employed : these 

 markings, however, on the parts uncovered by wax, although quite 

 obvious when the newly-cleaned glass was breathed upon, were 

 perfectly imperceptible when the same watch-glass was taken up 

 a day or two later and re-wiped for the purpose of fresh examina- 

 tion.! Since that experiment, M. Nickles's observations"! on the 

 fallacies incidental to this method of testing have appeared ; and 

 this, doubtless was a case such as he describes, in which the vapour 

 of any acid, or even of water, may exert such an action (not che- 

 mical, I presume, but physical), and fix upon glass any design at 

 first traced upon the layer of wax. The search for the remaining 

 substances above mentioned was unsuccessful, and they were there- 

 fore presumed to be absent. 



* Fresenius, Quant. Analysis, p. 345. 2d Eng. edit. 



t I accounted for this at first by imagining that a thin film of wax might be 

 impressed by the needle upon the glass, which was not immediately removable 

 by the first cleansing. 



% Comptes Rendus, Mareh 30, 1857, p, 679. 



Vol. xx. o. s. Vol. iy. n. s. 



