An Attempt to analyse Cast Iron. 255 



or, what amounts to the same, that the oxide of manganese 

 and the sihca had not contained oxygen in such a propor- 

 tion to the oxide of iron as to have much influenced the 

 resuh, then 10-837 grammes oxide correspond to 7'342 

 grammes of metalHc mass ; after which there remain 0'357 of 

 a gramme of loss, which in this experiment can be nothing 

 else than carbon, and being divided among 7"9 grammes of 

 cast iron, gives 45 per cent, of carbon. This must evi- 

 dently be too much, when we see that the oxide obtained 

 could not possibly be oxidated to maximum in the whole 

 of its mass. 



3. Since the result of the preceding experiment was at 

 all events very indecisive, I repeated it with five grammes of 

 coarsely bruised cast iron, which was dissolved completely 

 in so much nitric acid, that the solution, when viewed 

 against the flame of a candle, was found clear at bottom, 

 without any iron being left undissolved. Evaporated iti 

 the same vessel to dryness, and burnt, there were left fi-QSS 

 grammes of oxide of iron. But before I undertake to de- 

 termine the quantity of carbon iu this experiment, I must 

 first mention an attempt made to discover more exactly the 

 quantity of manganese. 



4. Five grammes of the same cast iron were dissolved in 

 an apparatus for disengaging gas, with the assistance of 

 very gentle heat. The gas was received over rain-water, 

 and amounted to 65 Swedish decimal cubic inches ; or, 

 minus 1 cubic inch as much as from 4 grammes of mal- 

 leable iron dissolved in the same apparatus, and conse- 

 quently diluted with an equal volume of atmospheric air 

 from the vessel. The gis was burnt in oxygen gas, and 

 yielded r03 gramme of carbonate of lime, the carbonic 

 acid being detained by lime-water. But iu this experiment 

 the whole apparatus was become coated with a fetid oil. 

 which also collected on the water in the apparatus in the 

 form of iridescent pellicles, and which admitted be- 

 ing washed off with alcohol, after the solution was 

 poured out from the phial in which the solution was per- 

 formed. We thus find it impossible that the whole quan- 

 tity of carbon in the iron could be contained in the gas; 

 besides that, on the combustion of the gas, the last portion* 

 were certainly prevented from complete deflagration by the 

 copiously formed carbonic acid, which also the altered co- 

 lour of the flame appeared to indicate. 'Ihe 103 of car- 

 bonate of lime obtained, answers precisply to 2| per cent, of 

 carbon. The muriatic solution was filtcn;d, and left on 

 the filter a gray, bulky, pulverulent mass, which after 



washing 



