1321.] the Composition of Prussiafes, 221 



opinions,"* to which it appears to me he should have added, 

 " with the exception of those cases in which the discordance is 

 the result of an ill-made, incorrect experiment,'^ whenever the 

 science derives a truly negative advantage. 



Mr. Porrett also soon endeavoured to prove Thomson's analy- 

 sis inaccurate. He, on this occasion, analyzed the salt of 

 potash by a solution of tartaric acid in alcohol, and determined 

 the quantity of potash by the supertartrate of potash produced. 

 The following is the result of his analysis : 



Ferruginous hydrocyanic acid 50*93 



Potash 35-48 



Water 13 



He had found that the ferruginous prussic acid might be insu*» 

 lated by this operation, and that it gave cubic crystals by spon- 

 taneous evaporation. In an analysis of this salt by oxide of 

 copper, he obtained four volumes of carbonic acid gas for one of 

 azote ; whence he concluded that the acid is composed of one 

 atom of azote, four atoms of carbon, one atom of hydrogen, and 

 one atom of iron. He attributed the deficiency of carbonic acid 

 in Thomson's results to his having employed too small a propor*- . 

 tion of oxide of copper. 



Lastly, Mr. Porrett published a further memoir on this subject, 

 in which he again corrected his former results as follows : 



Ferruginous hydrocyanic acid ij * V2-r0 r ^^*^^ 



Potash 41-68 ' " 



Water 13-00 : 



Repeating the experiment of burning the salt with oxide of 

 copper, he constantly obtained four volumes of carbonic acid gas 

 for one of azote ; whence he concluded that the acid is com- 

 posed of four atoms of carbon, one atom of azote, one atom of 

 hydrogen, and half an atom of iron ; but this half atom of iron 

 not being consistent with the atomic views, he conceived his 

 experiments to be sufficiently exact and positive, to decide, con- 

 trary to conclusions derived from less complicated and easier 

 experiments made directly on iron and its oxides, that the weight 

 of the atom of iron is not half what it has heretofore been admit- 

 ted to be, or one-fourth of that laid down in my tables ; whence 

 he concludes that the protoxide of this metal is composed of two 

 atoms of base and one of oxygen, and the deutoxide of four of 

 base and three of oxygen. 



M. Vauquelin also made many researches on the prussiates; 

 and his treatise on the subject is full of interesting facts ; but he 

 did not employ himself in determining the proportions of its 

 elements, which is the principal object of the present memoir. 

 He found that prussian blue, contrary to what Gay-Lussac had 



^ ♦ Annals of Philosophy, vol. xii. p, 112» ^ .' 



