1821.] the Composition of Prussiates. 443 



corrosive sublimate. These proportions being neither as 1 

 to 2 J nor as 2 to 3, I thought it requisite to examine whether the 

 caustic potash did not in some way decompose cyanogen. I, 

 therefore, analyzed prussian blue by digesting it for a long time 

 with bicarbonate of potash. The decomposition was complete^ 

 the bicarbonate left 31 parts of oxide of iron. The yellowish 

 solution was supersaturated with nitric acid, evaporated to dry- 

 ness, and the dried mass was exposed to a red heat. 



This mass being washed with water left 23 parts of oxide of 

 iron undissolved. All these numbers agree together, and the 

 two portions of oxide of iron obtained in each experiment are in 

 the proportion of three atoms of oxide of iron and four atoms of 

 the same oxide; that is to say, = 2935-29 : 3913*72. The 

 analysis then proves decidedly, that in prussian blue prepared in 

 this manner, the oxide of iron contains twice as much oxygen as 

 the protoxide, and consequently its composition is proportional 

 to that of other cyanurets or ferruginous hydrocyanates ; this is 

 also demonstrated by the substances from which prussian blue is 

 precipitated, remaining neutral. But what is then the blue mass 

 which is produced by the oxidation of the white hydrocyanate 

 of iron ? This compound cannot be neutral, for it does not meet with 

 any acid to saturate it, nor can it contain oxide of iron mechani- 

 cally mixed ; first, because I have often seen it totally dissolve 

 in water ; and secondly, because the hydrate of oxide of iron 

 must alter the fine blue colour to green, as happens if the solu- 

 tion in which it is formed contains an excess of the ferruginous 

 salt. 



The experiments which I have made to analyze prussian blue 

 by means of combustion have not given me as decisive results 

 as I expected. A quantity of prussian blue dried at a high 

 temperature gave 58 per cent, of oxide of iron by combustion. 

 Another portion of the same prussian blue, weighed an instant 

 after the former, was burned with oxide of copper, taking the 

 same precautions as already mentioned. The water, carbonic 

 acid, and azotic gases, obtained, indicated 45*6 parts of carbon, 

 and 53*1 of azote = 98*7 parts of cyanogen. The prussian blue 

 employed contained such a portion of metallic iron as would 

 combine almost precisely with this quantity of cyanogen to 

 become cyanuret ; and by adding together the weight of the 

 iron, cyanogen, and of the water obtained, the weight amounted 

 almost precisely to that of the prussian blue employed, without 

 leaving any thing for the weight of the oxygen of the oxide of 

 iron, which ought to exceed the hydrogen of the hydrocyanic 

 acid, if the compound had contained excess of base. 



1 repeated this experiment with prussian blue precipitated 

 from a neutral solution of muriate of deutoxide of iron, and dried 

 in vacuo, without heat. It left after combustion 54*66 per cent, 

 of red oxide of iron ; when burned with oxide of copper it 



