g24 OXIDATIONS OF IRON. 



Magnetic from oxide succeed the green, when the 100 grains had absorbed 



oxlgen. 265 ° f from 30 t0 36 of oxi S en - 



This experiment, which I repeated several times, always 



afforded me the same results, except that sometimes I found 



Sometimes a a few hundredth parts of red or of green oxide. It is 



little red oY obviouwly impossible however, from various circumstances, 

 green oxide . ' ' 



■with it. that eve/y particle of the iron should be equally exposed to 



the action of the air and the caloric. 

 Re<Taru1 not On carrying the calcination so fair, that 100 grains had 



•284 eiC W1 taken ap 38 of oxigen, the precipitates were entirely red, 



without exhibiting any trace of magnetic oxide. From 



Oxigenation ra- this term to that of 45 or 50 the oxiffenalion was very slow, 



ther slow, ° J ' 



and would have been impracticable without an increase of 



but carried as heat: by raising the fire however, and renewing the air by 



faras -36 nearly. * , , ,, T • i *u • *■ / 



means of a pair of bellows, 1 carried the oxigenation as far 



as 56. Thrs operation is very tedious and tiresome; but an 

 apparatus to save the trouble of renewing the air might ea- 

 sily be contrived, if the excess of oxigen were of any advan- 

 tage in physic or the arts. 

 All the oxides I treated this oxide at 56 with acids, and afterward p re- 

 after the mag- dpltated- 41 by the alkalis and alkaline earths : I likewise 

 netic alike. l J ' 



added to its solutions prussiates, gallates, and phosphates ; 



to see if I could discover any properties distinguishing it 



from the oxides that had preceded it. My trials however 



were in vain, as I might have expected, since those at 38, 



40, 45, and 48 had afforded me no characters to distinguish 



them from one another. 



Redness and The only difference I observed among these oxides was, 



magnetism not ,, . , . , . , , , ,, . 



owing to pro- tnat the red colour became more decided, and the magnetism 



portion of weaker, in proportion as the oxidation advanced: but 

 these properties depend more on the difference of the com- 

 pactness or density of the oxides, than on their proportion 

 of oxigen. * 

 Different salts This conformity of the properties of oxides, among which 

 owing to dif- the difference of the proportion of oxigen extends as far us 20, 

 ferenceof the or even 40 hundredths, as I shall show, proves the error of 

 requiring the formation of a different salt as characteristic 

 of each degree of oxigenation. Jn fact too extensive an 

 influence on the oxides of iron, and I believe on all the me. 



* See the Note subjoined at the end of this Memoir. 



tallic 



