253 An Attempt to analyse Cast Iron, 



Vauquelin's proposal for precipitating the oxide of man- 

 ganese with hvdro-siilphuret of alkaH may not at all suc- 

 ceed, and so much the less, as Berlhollet states that the 

 oxide of manganese with the sulphuretted hydrogen pro- 

 duces soluble combinations. The same is the case with 

 alkaline sulphurets. What first is added, precipitates no- 

 thing; afterwards, with greater quantities of alkaline sul- 

 phates, there will precipitate a mixture of sulphur, oxide of 

 manganese, and magnesia. I had long before experienced 

 that sal-amn)()niac when boiled with magnesia is decom- 

 posed to a triple salt; theretore, in the hope of turning to 

 account this disposition of the maenesia for separating 

 it from the oxide of manganese, I boiled a mixture of 

 magnesia aid oxide of manganese (precipitated together by 

 caustic potash and edulcorated) with sal-ammoniac in 

 greater quantity than would have been required for forming 

 a triple salt with the whole mass. When on a continued 

 digestion no more ammonia was disengaged, the mLxlur© 

 was filtered, and precipitated with caustic potash. The 

 precipitate contained so much oxide of manganese that it 

 blackened in drying. I therefore varied the experiment, 

 so that a mixed solution of muriate of manganese, muriate 

 of magnesia and sal-ammoniac were treated with caustic 

 ammonia in great excess, and placed under a bell with 

 oxygen gas ; the mixture was taken out after 72 hours. 

 It had deposited a great deal of black oxide of manganese. 

 After filtration it was precipitated with caustic pUash ; the 

 precipitate was found black after drying, and had therefore 

 contained oxide of manganese. Thereafter I endeavoured 

 to separate them, partly with neutral tartrat of ammonia, 

 and partly with super-tartrat of ammonia; but neither of 

 them precipitated thereby. Alallate of ammonia did not 

 precipitate the magnesia, but yielded a small precipitate 

 with manganese sidts. It did not however precipitate fully, 

 and therefore could not be employed. A solution of borax 

 does not pr^cii)itaie magnesia, but it precipitates solutions 

 pf manganese. If we mix a solution of magnesia and 

 manganese wiih borax, there will be no precipitation ; and 

 should we have precipitated the manganese by itself with 

 borax, and then add a solution of sulphat of magnesia, 

 the precipitate redissolves. In short, ihe only method of 

 tolerable satisfaction that I have found for separating them 

 was this: — after having sufficiently burnt the mixture of" 

 magnesia and manganese, for the purpose of expelling the 

 carbonic acid out of the magnesia, they were weighed, dis- 

 solved in diluted sulphuric acid, which left a great portion 



of 



