1314.] Qh the Cause of Chemical Proportions. 357 



care with wlikh I made these experiments, and the agreement 

 between them, induce me to believe that they approach as near 

 accuracy as a direct experiment can do. Other chemists have since 

 published analyses of this oxide. Gay-Lussac, for example, found 

 that 100 of iron combined with 42*35 of oxygen ; but he seems to 

 have paid no attention to the carbon present, nor to my experiments, 

 though they had been published in Paris several months before his. 

 (Ann. de Chim. Nov. ISll, p. 163.) The confidence to which 

 every thing is entitled that bears the name of this celebrated che- 

 mist, may perhaps throw some doubts respecting the accuracy of 

 my analysis, Howe%'er, till some other chemist has verified our 

 results, I hope I may be excused, in consequence of the great car© 

 which 1 bestowed in order to obtain exact results, if I consider my 

 ■own numbers as approaching most nearly to the truth. 



By other experiments I have shown that the oxygen in the black 

 oxide is to that in the peroxide ?s 2 to 8. Hence it follows that the 

 two oxides of iron are Fe + 2 O, Fe + 3 O. The volume of iroa 

 then weighs 693"(i4. 



M. Gay-Lussac has discovered that the black oxide, formed when 

 iron is exposed at a high temperature to the vapour oi' water, con- 

 tains more oxygen than the common black oxide of iron. Accord- 

 ing to the numbers given by Gay-Lussac, iron in that oxide is 

 combined with 1^- as much oxygen as in the common black oxide. 

 He considers it as a particular oxide capable of becoming the basis 

 of salts. Yet as its sulphate is decomposed by alcohol in such a 

 manner that the persulphate of iron is dissolved and the green sul- 

 phate left untouched, as the succinates and benzoates precipitate 

 from it persuccitiate and perbenzoate of iron, and as the caustic 

 alkalies precipitate from it persubsulphates of iron, before precipi- 

 tating the green sulphate, it is clear that this substance cannot be 

 considered as a particular oxide, but as a combination of the black 

 and peroxides of iron. M. Proust has long since proved that 

 common prussiate of potash is a triple prussiate of potash and black 

 oxide of iron. When we precipitate metallic solutions by this 

 prussiate, the potash is exchanged for the metallic oxide, which con- 

 stitutes witii the black oxide of iron a triple prussiate. When the 

 potash is exchanged for peroxide of iron we have as usual a triple 

 prussiate containing the two oxides of iron for its bases : and it i,s 

 clear thiH; in such a case the peroxide of iron exists in the salt just 

 in the same state that any other metallic oxide would. Hence it 

 follows that triple salts containing these two oxides as bases really 

 exist. 



Gay-Lu3sac pretends that the Swedish minerals attracted by the 

 magnet contains his new oxide. I have examined several of these 

 tninenils witiiout ever finding any proofs of liis assertion. I have 

 pulverized these minerals, and have extracted the part attracted by 

 the magnet under water, in order to separate it from the matrix and 

 the red oxide of iron meclianically mixed. When 1 treated what 

 thii magruit had attracted with less diluted muriatic acid tlian was 



