Blue Compounds of Cyanogen and Iron. 163 



of washing with distilled water in a vessel from which air 

 was excluded. The precipitation had been performed in 

 a tall glass cylinder, which was completely filled by the li- 

 quids, and could be perfectly closed by a glass plate ad- 

 justed to its ground edge. In this vessel the precipitate was 

 washed by subsidence and decantation, during the first period 

 of which the water ran off colourless, or only slightly coloured 

 by the remaining quantity of red prussiate. As soon, how- 

 ever, as the greater part of the soluble salts had been washed 

 away, the liquid no longer deposited all the blue compound 

 suspended in it, but even after standing several hours ran off 

 with a blue colour. In the supposition that I had here a so- 

 lution of prussian blue, I threw it upon a filter and found 

 that it ran through unchanged. Unwilling however, upon 

 the simple testimony of this fact, to conclude that it was a 

 solution, I transferred it to a filtering apparatus, so contrived 

 that by means of a hydrostatic pressure, which could be va- 

 ried at pleasure, it was driven through a sixfold layer of fil- 

 tering paper supported by a linen cloth. This experiment 

 proved that the blue compound was indeed not in solution, 

 but merely in a state of suspension, so fine as to pass through 

 the pores of a simple filter, for it was completely separated 

 by the denser mass of the compressed folds of paper, and a 

 colourless liquid passed through. The washing was continued 

 until the water filtered through in this manner left no visible 

 residue on evaporation on polished platinum. Subsequent 

 experiments showed that this operation may, without affect- 

 ing the result, be performed more quickly and easily by the 

 use of hot water, as in this case the precipitate assumes a 

 denser form and deposits more easily from the liquid. 



I. 1*706 grm. of this substance gave 0*828 peroxide of 

 iron and 0*189 sulphate of potash, corresponding to 33*6 per 

 cent, of iron and 4*96 per cent, of potassium. 



II. 1*870 grm. gave 0*909 peroxide of iron and 0*221 sul- 

 phate of potash, corresponding to 33*7 per cent, of iron and 

 5*3 per cent, potassium. 



The average of these numbers, namely 33*6 iron and 5*1 

 potassium, are in the proportion of 9*3 equivalents iron to 1 

 potassium. It will be shown later how these numbers are to 

 be considered ; we see in the meanwhile that they indicate a 

 different composition from that of ferridcyanide of iron and 

 potassium. 



I next endeavoured to precipitate in such a manner as . if 

 possible to replace all the potassium of Gmelin's salt by iron, 

 which was done by forming the precipitate in an excess of 

 protochloride of iron, and digesting it for several hours with 

 that salt. The substance thus prepared was washed out in 



