1825.] M. Berthier on Forge Scales, ISl 



tain period of their formation : nevertheless, although they 

 become soft by an intense heat, we are unable to fuse them, 

 completely. It is probable that their fusion is effected by the 

 local heat which is developed at the moment when the incan- 

 descent iron combines with the oxygen of the air, and which 

 must necessarily be very intense, but as it is quickly dissipated, 

 the matter soon becomes solid, and assumes a crystalline struc- 

 ture, if not cooled too suddenly. A similar phenomenon is seen 

 in the combustion produced by striking fire with the flint and 

 steel, in cupellation, and in several other instances. 



The forge-scales are very magnetic. When reduced to grains 

 of the size of a pin's head, they adhere to a magnet like me- 

 tallic iron. Their specific gravity is 3*5 ; but, as they always 

 contain some blebby cavities, this must be too low. Their 

 powder is of a dull greyish black colour. 



This oxide has hitherto been supposed to be the same as the 

 native magnetic oxide, and that obtained by passing aqueous 

 vapour over iron heated to redness. Having had occasion for 

 some very pure oxide of iron, for some experiments on the sili- 

 cates of that metal, I used the forge-scales, supposing them to 

 be so ; but I soon perceived that they do not contain so much 

 oxygen as the magnetic oxide, which at present is considered 

 as a deutoxide. For instance, when, for the purpose of pre- 

 paring a proto-silicate of iron, I employed calculated propor- 

 tions of forge-scales and iron filings, a certain portion of the 

 metal always remained unoxidated ; and when I reduced forge- 

 scales by cementation in a black lead crucible, I constantly 

 obtained a heavier button than the pure native oxide, similarly 

 treated, would have given. I was therefore induced to inquire 

 into the true composition of these scales ; and the results of my 

 experiments demonstrate that they are a new oxide, which, 

 from the quantity of its oxygen, ranks between the protoxide 

 and the native magnetic oxide. 



This oxide does not produce any peculiar salt; it is decom- 

 posed by the acids into protoxide and peroxide, exactly like 

 the deutoxide, and this property affords a very simple method 

 of analyzing it, and is the one vi'hich I adopted. I treated the 

 forge-scales with pure muriatic acid, in which they dissolve 

 very readily, even in the cold, if the acid be concentrated, and 

 the temj^erature of the liquid becomes considerably raised. I 

 diluted the solution with water, and threw down the peroxide* 

 by gradually pouring in carbonate of ammonia 'till the liquid 

 was discoloured. This process is not attended with any diffi- 

 fculty ;-4t gave me froin 0*34 to 0-36 of peroxide, according to 

 the scales I employed, and which I obtained from different 

 fotges, both from the tilt-hammer and the flatting-mill. The 

 purest gave the largest quantity of peroxide. When I only 

 obtained -34, I always observed, at the moment of solution^ a 



k2 



