CHEMICAL SCIENCE. 221 



in the same manner, and the comparatively very small proportion of carbon- 

 aceous residue given by it also examined. And these were compared with 

 the residue also obtained from cast-iron. If the acid be strong, and heat be 

 used, and the voltaic arrangement be not used, the results are very different. 

 Gaseous nitrogen, in very minute quantity, is given off along with the hy- 

 drogen, some muriate of ammonia is formed in the solution, and but little 

 nitrogen left in the residue. 



Effecting the combustion of each of these residues by aid of the soda lime 

 process, in the usual manner, the following results were obtained: 1. The 

 residue of the malleable iron contained no nitrogen whatever. 2. That from 

 the cast-iron always showed the presence of nitrogen, but in very minute 

 and invariable quantities; an average of results would seem to confirm the 

 analysis of Marchand. 3. In the steel residue there was invariably detected 

 a considerable quantity of nitrogen. 



The analysis of this carbonaceous residue gave C= 0'G3, X= P 2-1; im- 

 purities, 0'13 == 100. The direct analysis of this sample of steel, using the 

 soda lime process, gave, in 100 parts of steel, C== O'GS, X== OTJ; that is, 

 that any 100 parts of steel contained about one-fifth per cent, of nitrogen 

 associated with about three times its weight of carbon. The proportion of 

 nitrogen in the residue was greater than in the steel itself a result proved 

 afterwards to be due to the absorption of nitrogen or formation of ammonia 

 in the act of drying the residue. The direct analysis of the sample of iron 

 gave no nitrogen whatever that of the cast-iron only a trace. This steel 

 contains, therefore, about one-fifth per cent, of nitrogen, and by other trials 

 good steel always gives about this quantity, but inferior steel much less. It 

 is obvious that the residue is an azotized carbon, out of which fact arises 

 some important considerations. But confining these, for the present, to the 

 cases of the malleable iron and the steel, it would appear that the difference 

 in chemical composition between these two is not less remarkable than the 

 difference between their respective physical properties ; but in what precise 

 manner these two elements produce these differences, or the form in which 

 they exist together in the steel, we can only as yet conjecture theoretical!}'. 



When malleable iron is placed in a crucible along with some of this azo- 

 tized carbon, the intractability as to fusion of the metal is soon overcome, 

 and it melts at a white heat, and cast-steel is the product. 



"\Vhen there is prepared a mixed precipitate of oxide of iron and of oxide 

 of manganese, and this is reduced to the metallic state by passing over it, at 

 a high temperature, hydrogen gas that is, in the usual manner, made into 

 spongy iron, plus manganese; this, placed in a covered crucible, readily 

 melts, and gives an exceedingly hard alloy, but one that does not, possess all 

 the properties of real steel. 



But when with spongy iron itself there is mixed some ferrocyanide of man- 

 ganese, and then it is exposed to a full heat in a covered crucible, then the 

 button that is produced has all the properties of steel. 



The same follows on substituting ferrocyanide of iron, or any of the achy- 

 drous alkaline ferrocyanides. 



The addition in the hot and closed crucible of bitartrate of potash to the 

 pure iron does not give steel; but the addition of the double salt' of tartrate 

 of potash and ammonia (which is a kind of improvised mode of making 

 cyanogen), gives a button of steel. 



Another series of instructive facts, in proof and in illustration of these 



19* 



