Determination of Pyrite and Marcasite. 383 



cipitation of iron even when 5 g. sulphuric acid instead of 4 

 g.* per liter is contained in the standard solution. If correc- 

 tion is made for the precipitated iron, no change in concentra- 

 tion has taken place, i. e., there is no loss of water. It should 

 be stated that in an actual oxidation experiment no iron is 

 precipitated, since the ferric iron is pretty rapidly reduced to 

 the ferrous state. The following results prove these state- 

 ments : 



C, is the original concentration of the iron solution in grams 

 of KMn0 4 solution per 100 cc of the standard. 



C 2 is the concentration after the blank test, to determine 

 which 100 cc solution is then reduced with S0 2 and titrated. 



I II ill 



C 1 37-06 37'06 36-54 



p j 37-09 37-08 36'52 



2 | 37*03 37-04 36*44 



(c) Reduction of the ferric solution during the blank test. — 

 Stokes states that no reduction occurred during a blank deter- 

 mination. We found a small reduction due perhaps to some 

 organic matter associated with the ammonium salt. It 

 amounted to only 0*1 g. of the very dilute standard KMn0 4 

 solution per 100 cc . In actual oxidations it is probable that 

 the error would be less on account of the presence of a 

 competing reducing agent. 



(d) JZrrors in the determination of the iron. — These are very 

 slight as previous results would indicate. In standardizing 

 the ferric solution, 100 cc are measured in a calibrated flask at 

 25°. The solution is washed into a 500 cc flask, reduced in a 

 stream of washed sulphur dioxide, the excess of which is 

 boiled out in a stream of carbon dioxide in which the solution 

 is finally cooled. The titration is then made with a weight 

 burette, using a permanganate solution containing 1*5 g. of 

 the salt per liter. The following determinations*!- of the con- 

 centration of four different solutions show the accuracy of the 

 method : 



I II III IV 



36-60 



36-51 



36*50 



36-69 



36 62 



36-60 



36-51 



36-69 



36-63 



36-59 



36-46 







36-58 



3646 





Sulphur dioxide seems to carry with it mechanically some 

 non-volatile reducing agent. This occasions a relatively large 

 blank, but a very constant one. Blank tests were made as fol- 



* The quantity prescribed by Stokes. 



f The unit in these determinations is the gram of the above standard per- 

 manganate solution. 



