268 M. Arfwedson on [April, 



products formed, when protoxide of mangauese is heated to 

 redness in the open fire, and when nitrate of manganese is 

 exposed to a red heat, constitute one and the same oxide ; and 

 that the different colours (for the former is brown and the latter 

 black) are owing to the different states of mechanical division in 

 which the two bodies are. But I have always obtained these 

 two products, very different from each other in appearance, even 

 when both were reduced to powders of the same fineness. The 

 former always gives with sulphuric acid a weak amethyst-red 

 solution ; while the latter forms with the same acid a solution 

 having a deep grass-green colour. These different properties 

 appeared to require a more accurate investigation of the chemi- 

 cal constitution of these two bodies. With this view 1 have, at 

 the desire of Prof. Berzelius, made a set of experiments, of which 

 I shall now give a short account. 



(1.) A neutral solution of pure muriate of manganese was pre- 

 cipitated by bicarbonate of potash. The carbonate of manga- 

 nese obtained was snow-white. After being washed with boiled 

 and cold water, and dried in the vacuum of an air-pump with 

 sulphuric acid, it remained without any change in its colour. 



The carbonate of manganese thus prepared was put into a 

 spherical cavity blown in the middle of a barometer tube, 

 through which a current of hydrogen gas was made to pass. As 

 soon as the gas had driven out all the atmospherical air, the 

 manganese was heated by applying a spirit-lamp to the glass 

 ball containing it ; and the heat was continued as long as any 

 water and carbonic acid gas continued to be evolved. The 

 powder by this process acquired a fine pistachio-green colour, 

 which continued till the apparatus had cooled, and the atmo- 

 spherical air was admitted to it, when it became greyish-green. 

 A little portion of this powder treated with muriatic acid dis- 

 solved without the least effervescence, showing that the carbonic 

 acid had been completely expelled. 



A gramme of protoxide prepared in this way was heated in a 

 glass capsule over a spirit-lamp. The matter immediately took 

 fire, and burnt to a dark-brown oxide, which weighed 1*0735 

 gramme. Thus the increase of weight on 100 parts of protoxide 

 was 7*35 ; and as the protoxide is considered as containing 

 2T9 per cent, of oxygen, the brown oxide must be a com- 

 pound of 



Manganese 72-784 100-00 



Oxygen 27-216 37-39 



100-000 137-39 



, The 1-0735 gramme of brown oxide was boiled with nitric 

 acid to dryness, and heated to redness. The brown colour of 

 the oxide gradually changed to black, with a copious evolution 

 of nitrous gas, and the weight of the powder became 1-195 



