1821.] of the Atoms of Copper, Zinc, Iron, Manganese, S^c. 249 



number made choice of by Prout. Berzelius's number is cer- 

 tainly too small. 



It deserves to be mentioned that the whole of the iron cannot 

 be precipitated from a solution of protosulphate of iron by means 

 of phosphate of soda, in what proportion soever the saline solu- 

 tions are mixed. The supernatant liquid contains both oxide of 

 iron and phosphoric acid ; for it strikes a blue with prussiate of 

 potash, and is precipitated white by nitrate of lead. Indeed I 

 know no neutral salt capable of throwing down the whole of the 

 iron from a solution of this salt. 



It is not very easy to procure protosulphate of iron perfectly 

 free from peroxide of iron. The best way seems to be to dissolve 

 the crystals in hot water, to filter the solution, and impregnate it 

 with sulphuretted hydrogen gas. The crystals formed in such a 

 liquid are pure protosulphate of iron. The mark by which the 

 purity of this salt, or at least its freedom from peroxide of iron, 

 may be recognized is, that prussiate of potash precipitates it 

 white instead of blue. 



IV. Manganese, 



The only person who has endeavoured to determine the 

 weight of an atom of manganese experimentally before me is 

 Berzehus ; and his determination (when reduced) differs but 

 little from mine ; for 3'5o785, though a complicated number, 

 exceeds 3'5 only by a very small fraction. I shall describe the 

 experiment which I consider as demonstrating that the true 

 weight of an atom of protoxide of manganese is 4*5. Sulphate 

 of manganese is easily obtained by dissolving carbonate of man- 

 ganese in sulphuric acid, and evaj3orating the solution sufficiently. 

 It forms beautiful rhomboidaJ flesh-coloured prisms which are not 

 sensibly altered by exposure to the air. After a few preliminary 

 trials on this salt, I concluded that it is a compound of 



1 atom sulphuric acid = 5*0 



1 atom protoxide of manganese. .. = 4*5 

 5 atoms water = 5-625 



15-125 



To verify this conclusion, I dissolved 15-125 grs. of the dry 

 crystals in water; 13*25 grs. of dry chloride of barium w^ere 

 dissolved in another portion of water ; and the two liquids were 

 mixed together. After the precipitate of sulphate of barytes 

 had subsided, the supernatant liquid w^as examined by means of 

 muriate of barytes and sulphate of soda. It was not affected by 

 either of these reagents. It follows from this, that 15-125 grs. 

 of sulphate of manganese contain exactly 5 grs. of sulphuric 

 acid. The remaining 10*125 grs. consist of water and protoxide 

 of manganese. Now Dr. John obtained from 15-125 grs. of 

 sulphate of manganese 5*345 grs. of water. I myself, from the 

 same quantity of salt, extracted 5*48 grs. of water. John's 



