12 
Proceedings of the Royal Society of Edinburgh. [Sess. 
III. — On the Reduction of Ferric Iron (1) by Sulphurous Acid 
and (2) by Zinc Dust. By Alexander Charles Cumming 
and E. W. Hamilton Smith. 
(MS. received November 9, 1911. Read December 4, 1911.) 
I. Conditions for Reduction by Sulphurous Acid and Sulphites. 
So many papers have appeared on this subject that some apology seems 
desirable before making an additional contribution. The amount of 
published work on reduction with sulphurous acid is in itself an indication 
that many workers have found difficulties. It has been shown that the 
reduction does not take place in presence of large excess of hydrochloric or 
sulphuric acid, but the reduction will still occur while the reaction of the 
solution is strongly acid. On the other hand, Hillebrand (“Analysis of 
Silicate and Carbonate Rocks,” U.S. Bulletin , 442, p. 113) states if the 
solution after addition of sulphite is red in colour, it is too alkaline and 
acid must be added. 
Some time ago we noticed that sulphur dioxide, passed into a boiling 
ferric alum solution which contained also a small quantity of free acid, did 
not reduce the iron even after several hours’ treatment. Sulphur dioxide 
passed into a hot solution caused rapid and complete reduction, if ammonia 
had previously been added until slight precipitation occurred. While this 
was perfectly satisfactory from an analytical point of view, the reason was 
by no means obvious. A pure ferric sulphate solution is acid in reaction on 
account of hydrolysis, and remains acid after addition of the small amount 
of ammonia necessary to cause slight precipitation of ferric hydroxide. 
Furthermore, the precipitate was at once redissolved by the sulphurous 
acid, and the solution after the reduction contained free sulphuric acid, 
formed in the reaction, in excess of that formed by hydrolysis. Some 
further investigation as to the relation between degree of acidity and 
reduction therefore appeared desirable. 
Solutions used . — The following experiments were performed with the 
same solutions throughout. The permanganate was decinormal, and an 
approximately decinormal ferric solution was prepared from ferric ammon- 
ium alum. Neither of the solutions altered in concentration during the 
course of the work. A few preliminary experiments had shown the 
ammonium sulphate had no influence on the reaction. Two sulphurous 
