554 An Attempt to analyse Cast Iron. 



the linen. Five grammes of this powder I mixed with 60 

 grammes of pure and recently crystallized nitre, which held 

 no mechanical intermixture of combustible substances. 

 The mixture was put in a retort of glass luted to a tubu- 

 lated receiver, from whence the disengaged gas was con^ 

 ducted through lime-water in several bottles. The retort 

 was heated in a sand cupel. A little before the commence- 

 ment of the ignition, gas was disengaged, which made the 

 lime-water turbid ; an effect which lasted as long as any 

 disenoagement of gas took place, and which continued 

 slowly during the ienition for four hours ; after which it 

 ceased, and the apparatus was lei't to cool. Carbonate of 

 lime had deposited only in the first and second bottle. The 

 lime-water, which in both was still caustic, was decanted, 

 and the lime was repeatedly washed with water. It was 

 then separated from the bottles with muriatic acid, from 

 which it was precipitated with carbonate of ammonia. The 

 mass which remained in the retort was alkaline, and con- 

 tained at the bottom a black, somewhat brownish oxide of 

 iron. It was dissolved in water freed from oxide of iron 

 by filtration, and precipitated with a much diluted solution 

 of muriate of lime. The precipitated muriate of lime 

 mixed with that obtained from the lime-water, weighed 

 after washing and drying r6 gramme, which (when 100 

 parts carbonate of lime contain 43'6 of carbonic acid) cor- 

 respond to0697 parts of carbonic acid. These (since car- 

 bonic acid contains 28*46 per cent, of carbon) correspond 

 to 0*]9S3 of carbon; .vhich divided among five grammes 

 of cast iron, produces G'OSgG of carbon per gramme, or 3*y6 

 per cent. The dark oxide of iron obtained was difficultly 

 soluble in muriatic acid, and gave no signs of the disen- 

 gagement of hydrogen gas. 



2. 7*9 grammes of the same cast iron were dissolved in 

 nitric acid in a weighed glass phial ; the solution was of a 

 thick consistency, black-brown and opake. It was evapo- 

 rated to dryness, and burnt in the phial. The cast iron had 

 gained 2"65 grammes in weight, and left 10'35 grannnes of 

 imnure oxide of iron. On taking out the oxide, there was 

 yet found a rcgulus piece of iron which weighed 7573, 

 and after the solution in nitric acid, and after the ignition 

 of thcnitrate of iron, weighed r057 gramme; so that the 

 total oxide of iron obtained was 10'837 grannnes, of which 

 all could not be considered red oxide, since the stratum 

 nearest the rcgulus must have been black oxide, and the 

 exterior crust of the regulus niust also have contained some. 

 If we now assume that all were oxide of iron in this assay, 



or. 



