MANGANESE. 529 



the greatest difficulty. Hydrogen and charcoal, at a red heat, 

 reduce the superior oxides of this metal to the state of pro- 

 toxide, without eliminating the pure metal at that temperature; 

 but at a white heat, charcoal deprives this metal of its whole 

 oxygen. The following process is recommended by M. John for 

 the reduction of manganese : it illustrates the chief points to be 

 attended to in the reduction of the less tractable metals. Instead 

 of a native oxide, an artificial oxide of manganese, obtained by 

 calcining the carbonate in a well closed vessel, is operated 

 upon. This oxide, which is preferred from being in a high 

 state of division, is mixed with oil and ignited in a covered 

 crucible, so as to convert the oil into charcoal. After several 

 repetitions of this treatment, the carbonaceous mass is reduced 

 to powder, and made into a firm paste by kneading it with a 

 little oil. Finally this paste is introduced into a crucible lined 

 with charcoal (creuset brasque), the unoccupied portion of which 

 is filled up with charcoal powder. The crucible is first heated 

 merely to redness for half an hour, to dry the mass and decom- 

 pose the oil, after which its cover is carefully luted down, and 

 it is exposed for an hour and a half to the most violent heat of 

 a wind-furnace, that the crucible itself can support without 

 undergoing fusion. The metal is obtained in the form of a 

 semiglobular mass or button in the lower part of the crucible, 

 but not quite pure, as it contains traces of carbon and silico.i 

 derived from the ashes of the charcoal. By igniting the metal 

 a second time in a charcoal crucible, with a portion of borax, 

 John obtained it more fusible and brilliant, and so free from 

 charcoal, that it left no black powder when dissolved in an 

 acid. 



Manganese is a greyish white metal having the appearance of 

 hard cast iron. Its density according to John is 8.013; while 

 M. Berlhier finds it to be 7-05, and Bergman made it 6.850. 

 From its close resemblance to iron, manganese may be expected 

 to be susceptible of magnetism, but its magnetic powers are 

 doubtful. Peclet has endeavoured to show that manganese can 

 assume and preserve magnetic polarity from the temperature 

 4 up to 70? but that it loses it again at higher temperatures. 

 The small difference between the atomic weights of iron, man- 

 ganese, cobalt and nickel, which are respectively 3.3!). 2, .J-15.I), 

 369 and 369.9,, is remarkable, attended as it is by a great ana- 

 logy between these metals in many other respects. 



M M 2 



