ON THfi OXIDES OF IRON. Qy^ 



addition of weight from the coinbinaton of oxigen, than 

 iron ill a purer state. Now as Bucholz did not analyse his 

 iron separately, it is difficult to decide on his account, fii 

 the analysis there is an element very difficult to be deter- 

 mined m the present state of our knowledge, this is, the 

 tjiinntity of oxidule remaining. Now as, from these consi- 

 derations, we should regard the highest result as nearest ap- 

 proaching the truth, this leads to the adoption of 45. 



The proportion of oxigen in the black oxide appears to Black oxid«. 

 vary between 30 and 31*8 of oxigen to 100 of iron. It is 

 30 in reducing the black oxides of the valley of Aost: it is 

 40 in reducing the black oxide of the isle of Elba with 

 hidrogen gas. On dissolving the black oxide in nitric 

 acid, the oxidule increases 10 per cent in passing to the 

 state of red oxide. From this experiment it is inferred, 

 that the black oxide contains 30 of oxigen to 100 of iron, 

 supposing the red oxide to contain but 42 of oxigen. If, 

 however, this oxide be estimated at 45, agreeably to our ex- 

 periments, the black oxide will contain 31*8 to 100 of iron. 

 Lastlvj it amounted to 31*38 in the experiment where La- 

 voisier oxided iron by exposing it red-hot to the vapour of 

 water in a copper tube ; and to 30*38 according to Bucholz. 



There appears some difficulty in choosing between the re- Reasons for 

 sultsof 30 and 31*8 of oxigen to 100 of iron in the oxi- [^1'"!"//^^^^^^^^^ 

 dules, yet we think the higher ought to be taken, because, tion, 

 1st, the proportion derived from the oxidulated ores of the 

 valley of Aost is liable to various uncertainties, respecting 

 the identity of the mineral fused and the mineral analysed : 

 adiy, carbon always combines with the regulus, and it is 

 difficult to determine its quantity precisely : 3dly, it seems 

 more proper to ealculate the difference of proportion between 

 the black oxide and the red, by taking the latter at 45 than 

 at 42*. 4thly, of all the experiments, in which the black 

 oxide has been obtained, that where the vai)our of water 

 was employed is the most direct : 5tbly, considering the im- 

 purity of iron, the highest term of oxidation is always to be 

 preferred, when its state is well determined. 



From all these considerations we shall presume, that the or 76 iron to 

 black oxide is composed of 31*8 oxigen to 100 of iron; *'* **'^'Cen. 



Vol. XXVni.— April, ISli. T whence 



