166 Dr. Williamson on the 



iron and reproduce the red cyanide. It might indeed be re- 

 plied to this that the red cyanide is reduced by the protoxide 

 in proportion as it is set free : — 



KO y3K } + 3Fe ° = 2 ( Cf y 2K ) + Fe 3 4 . 

 I described above an experiment of Liebig, showing that by 

 the action of protoxide of iron the red prussiate is reduced to 

 the yellow, but I find that the reverse process takes place with 

 equal facility, that is, the peroxide of iron decomposes the 

 yellow cyanide, forming the red one. A small quantity of 

 perchloride of iron was added to a boiling solution of ferro- 

 cyanide of potassium, and the mixture heated for some minutes. 

 The liquid filtered off clear gave a deep blue precipitate with 

 protosulphate of iron : — 



3(Cfy 2K)\ _ f2Cfy 3K1 KQ 

 Fe 2 3 J-iCfy2Fei + 3KU ' 



I have now endeavoured, as far as possible, to state the two 

 ways of considering these compounds, and will leave the de- 

 cision to more competent judges. 



I next endeavoured to apply the method which had proved 

 so serviceable in obtaining Turnbull's blue free from potas- 

 sium, viz. digesting the precipitate with an excess of the iron 

 salt, to the preparation of prussian blue of equal purity. A 

 weak solution of yellow prussiate was poured into a great 

 excess of perchloride of iron, and the mixture allowed to stand 

 for some hours exposed to a gentle heat.. This precipitate, 

 after being completely washed in the usual manner, gave, on 

 analysis, to 0*707 peroxide of iron, - 057 sulphate of potash, 

 which numbers are in the proportion of 27 iron to 1 potas- 

 sium. If this were to be considered as a peculiar compound, 

 it must be, — 



3(3Cfy4Fe) + 3Cfy/|? e 



All my endeavours to obtain a precipitate free from potas- 

 sium were unavailing, as long as this element was present at the 

 formation; I had therefore at last no alternative left but that 

 of precipitating by pure hydroferrocyanic acid. This acid was 

 prepared by mixing a solution of yellow prussiate, from which, 

 by boiling, all atmospheric air had been expelled, with about 

 an equal volume of muriatic acid which had been similarly 

 freed from air, and precipitating by aether. After it had been 

 filtered off and washed with aether, the acid only needed to 

 be dissolved in absolute alcohol and again precipitated by 

 aether to be obtained perfectly pure. It was then dried, dis- 

 solved in water, and precipitated with an excess of perchloride 

 of iron. The precipitate thus formed may be distinguished 



