134 M% Berthier on Forge Scaler. [Aug-. 



blue or black colour; secondly, a thick layer of iron, of an 

 uniform colour, inclining to olive ; thirdly, a layer with shades 

 of black and ohve, which soon passes to the pure, and slightly 

 metallic black, of the scales. I examined the olive-coloured 

 part, with the idea that it might probably contain a mixture of 

 metallic iron and protoxide ; but 1 found it composed wholly of 

 reduced iron of the utmost purity, and there is every reason to 

 think that it is even perfectly free from carbon. When treated 

 with muriatic or sulphuric acid, it dissolves without leaving 

 any residuum, and hydrogen gas is disengaged to the last. - 

 The last portions dissolved have the same aspect as the whole 

 mass. When fused in a black lead crucible, either alone or 

 with the addition of an earthy glass, instead of losing weight, 

 as would happen if it contained protoxide, it increases from 

 0*01 to 0*02. The portion with shades of black and ohve be- 

 haves like a mixture of metallic iron and for2;e-scales ; in the 

 moist way, red oxide is always found in it. This fact proves 

 that metallic iron exerts no action en the oxide of the scales, 

 and consequently that it is impossible to obtain the protoxide 

 by heating any oxide with iron. The bluish coat of the buttons 

 seemed to me to be steely iron, or to have passed to the state 

 of steel, by the absorption of a certain quantity of carbon ; but 

 I have not positively ascertained the fact. 



The cementation of the peroxide of iron presents as inter- 

 esting, and more varied results, as the cementation of the forge- 

 scales. If the mass be not very large, as long as any red oxide 

 remains in the centre, no metallic iron is produced at the 

 surface, but only black oxide. When the heat has been kept up 

 a sufficient time, we find in the centre only magnetic oxide, 

 and we may observe towards the surface, as in the cementation 

 of the forge-scales, the bluish steel layer, the layer of olive- 

 coloured iron, and the layer shaded with olive and black. The 

 magnetic oxide in the centre is variable in its composition ; in 

 one experiment I found in it 0*48 of peroxide, and 0*52 of prot- 

 oxide, and in another 0*60 of peroxide and 0*40 of protoxide. 

 Since the native magnetic oxide contains 0*69 of peroxide and 

 0*31 of protoxide, it is obvious that the oxide in question must 

 be a mixture, in variable proportions, of the magnetic oxide of 

 the forge-scales, and native magnetic oxide. 



It appears, from what we have seen above, that the peroxide 

 of iron is changed by cementation, first, into an oxide similar 

 to the native magnetic oxide, and that as soon as this change 

 has been effected, its reduction begins from the surface to the 

 centre, the process going on in such a manner that, in propor 

 tion as metallic iron is produced at the surface, the deutoxide 

 of the forge-scales is formed in the interior of the mass, to its 

 centre; but these proportions diminish from the surface to this 

 point. Lastly, when the cementation is very far advanced, the 



