956 An Allempt to analyse Cast Iron. 



washing and burning weighed 0-0 125, and was silica, 

 which as near as ran be corresponds to the silica left un- 

 dissolved in the before described experiment on a larger 

 scale. The filtered solution was mixed with nitric acid, 

 and boiled to the full oxidation of the iron ; after which it 

 was saturated as perfectly as possible with carbonate of 

 potash, and then prtcipitated with benzoate of potash : the 

 precipitate was separated on the filter, and the filtered so- 

 lution, mixed with the inspissated water of edulcoration, 

 was precipitated with caustic potash. The deposit soori 

 blackened. It weighed after washing and burning 335 

 of a grainme. Dissolved in muriatic acid, and precipitated 

 with prussiate of potash, there subsided prussiate of man- 

 ganese, which was collected on the filter. From the per- 

 colated fluid, caustic potash precipitated a small quantity 

 of magnesia, which when burnt weighed O'Ol of a gramme. 



In the 6- 925 grammes of oxide remaining after the 

 oxidation of five grammes of cast iron (No. 3.) there was 

 therefore 0"325 of oxide of manganese, 0"01 of a gramme of 

 magnesia, and (according to the before described larger 

 scale of assay, from which I could be tolerably sure of its 

 charge of silex) 0-0325 of silica. As to the quantity of 

 oxygen in both these earths, it could very well be totally 

 excluded, since the whole quantity of both of them toge- 

 ther does not fully amount to I per cent. But I know 

 from experiments of my own (though not yet in a state for 

 publication) that the silica contains more oxygen than the 

 red oxide of iron; and that the magnesia contains less. 

 In the calculation I will let the one make up for the other, 

 since the error will be but an insignificant fraction. 



The case of the oxide of manganese is not so indifferent. 

 Fourcroy states that 100 parts of manganese take up 68 

 parts of oxygen for forming 168 parts of black oxide of 

 manganese. I do not know whether he in this place 

 means the crystallized black oxide produced by nature, 

 and which on burning gives oxygen gas ; but which also I 

 do not know that any chemist hitherto has produced arti- 

 ficially without the aid of electricity. Bergman estimates 

 the quantity of oxvgen in the brown oxide at 26, and the 

 jnetal at 74 per cent. I reduced a small portion of car- 

 bonate of manganese free from iron, and obtained a regulus 

 which after a careful cleansing of its surface weighed 

 0*5075 of a gramme. It was not in the least attracted by the 

 magnet. Dissolved in pure nitric acid in a weighed glass 

 phial, it yielded, afier the evaporation of the salt and its 

 calcination in the jJiial,0'7225 of a gramme of black oxide of 



manganese. 



