of the Oxides of Manganese. 101 



The quantity of water was determined as usual by means of 

 the chloride of calcium, and amounted to 1*12 per cent. * 



On exposing 23*746 grains of this oxide to a white heat, 

 the loss proved to be 3*064 grains or 12*90 per cent. Sub- 

 tracting 1*12 for water, there remain 11*78 as the loss of 

 oxygen. 



Accordingly, 100 parts of the pyrolusite were resolved into 



Red oxide 84*055 



Oxygen 11*78 



Water 1*12 



Baryta 0*532 



Silica 0*513 



100*000 

 Now, omitting the water, baryta, and silica as accidental im- 

 purities, the remaining 97*835 parts lose 11*78 parts, or 12*04 

 per cent of oxygen in being converted into the red oxide. On 

 the supposition that pyrolusite is composed of one equivalent 

 of manganese and two equivalents of oxygen, it should lose in 

 passing into the state of red oxide exactly 12*122 per cent of 

 oxygen, a quantity which corresponds closely with the result 

 of analysis. It is therefore an anhydrous peroxide of man- 

 ganese. 



I have analysed another columnar variety of pyrolusite, 

 which has a density of 4*819, and of which the individuals ra- 

 diate from a common centre. I brought it with me from Ger- 

 many, and believed it to be from Ihlefeld, as the ticket indi- 

 cated ; but Mr. Haidinger, after carefully inspecting several 

 large cabinets in Germany, has been unable to discover any 

 similar specimen which is known to have been found in that 

 place. Its locality therefore is doubtful. 



This variety is less pure than the foregoing. Before being 

 washed, it yields chlorine on the addition of sulphuric acid ; 

 and after the muriates have been removed by distilled water, 

 the neutral solution in muriatic acid gives traces of lime with 

 oxalate of potash. It contains silica and baryta nearly in the 

 same proportion as the first variety. 



The following is the result of my analysis : 



Red oxide 85*617 



Oxygen 1 1*599 



Water 1*566 



Silica 0*553 



Baryta 0*665 



Lime a trace. 



100*000 



Subtracting 



