108 Dr. Thomson on [August, 



accurate. The average quantity of sulphate of potash obtained 

 from 100 grains of the triple prussiate, Avas 80-5 grains, indicat- 

 ing 43*9 grains of potash ; but this sulphate was contaminated 

 with peroxide of iron, which accounts for its too great weight. 

 I am disposed, therefore, to rely on the analysis by means of 

 nitric acid, and to conclude that 100 grains of crystallized triple 

 prussiate of potash contain 41-64 grains of potash. 



(3.) Iron. — It is exceedingly difficult to collect the whole of 

 the. iron which this salt contains ; because a portion of it is 

 volatilized when heat is applied to a mixture of triple prussiate 

 and nitric or sulphuric acid. Even the heat evolved by the bare 

 contact of nitric acid is sufficient to volatilize a sensible quantity 

 of iron ; for when I poured nitric acid on the salt from a phial, 

 the whole acid in the phial acquired a blue colour, and deposited 

 a sensible quantity of prussian blue. I made ten successive 

 experiments with considerable care ; but the first three results 

 did not correspond well with each other. The remaining seven 

 were conducted in retorts, the beaks of which were plunged 

 into receivers, or dipped into the mercurial trough, while the 

 gas evolved was collected in glass jars. These precautions 

 enabled me to determine the portion of iron which escaped 

 during the process. The smallest quantity of peroxide of iron 

 which 1 obtained in this way from 100 grains of the salt was 

 I9'2 ; and the greatest quantity was 22-5 grains. The mean of 

 seven experiments gives '21*33 grains. 



After the evidence brought forward by Mr. Porrett, I think 

 that no reasonable doubt can be entertained that the iron in this 

 salt is in the metallic state. If we take the mean of the preced- 

 ing experiments as the true quantity of oxide of iron, the metallic 

 iron in 100 grains of triple prussiate of potash will be 14-22 

 grains. But in the present case I am rather disposed to take 

 the maximum quantity as the nearest approximation to the 

 truth ; because the collection of the whole iron is attended with 

 peculiar difficulty, and because the experiment which gave only 

 19 - 2 grains was a solitary one, all the others giving 21, 22, or 

 22*5. Now 22'5 grains of peroxide of iron is equivalent to 15 

 grains of metallic iron, which, therefore, I consider as the quan- 

 tity of iron contained in 100 grains of the triple prussiate. 



(4.) Remaining Constituents of the Acid. — To determine the 

 other constituents of the complicated acid of this salt, I had 

 recourse to a method originally suggested by Gay-Lussac, and 

 which has been since employed by Vauquelin, Berard, Prout, 

 and, perhaps, other chemists. It seems to answer very well for 

 the analysis of bodies composed of azote, united to carbon and 

 hydrogen. The method is to mix a determinate quantity of the 

 substance to be analysed with fresh calcined peroxide of copper, 

 to put the mixture into a glass tube, and expose it to a neat 

 gradually raised to redness. To the glass tube containing the 

 mixture, another tube is luted filled with dry muriate of lime. 



