248 THE ROYAL SOCIETY OF CANADA 



In experiments 9, 10, 13 and 14, 3 grams of cupferron made up 

 to a 6% solution were used for each precipitation. In the case of 11 

 and 12 this amount was increased to 5 grams, but apparently without 

 any eff'^ct upon the amount of iron remaining in solution. 



The above results show that the low values for the weight of the 

 iron precipitate are chiefly due to the iron remaining in the mother 

 liquor, increased still further by the solvent effect of the wash water. 

 The weight of iron remaining in the solution varies but little from the 

 mean value of the several determinations, so that a legitimate cor- 

 rection might be applied to the weight of the main precipitate in the 

 case of an experiment carried out under known conditions. With 

 regard to the amount of iron compound dissolved by the acid wash 

 water, this is surprisingly constant, for a^given concentration of acid, 

 and not too large, so that a correction could surely be applied here. 

 For the three estimations in which the concentration of acid was 5 cc. 

 cone, hydrochloric acid to 100 cc. of water, the greatest difference in 

 the weight of iron oxide dissolved was only 0*0002 gr. so that the cor- 

 rection is here pretty closely defined. 



It was thought worth while to try the effect of replacing the 20 

 cc. of hydrochloric acid in the main solution by an equal volume of 

 sulphuric acid. This was done in the case of the three estimations which 

 follow. The iron in the wash water and mother liquor was determined 

 as before. The acid concentration of the wash water corresponded to 

 5 cc. of cone hydrochloric acid to 100 cc. of water. The volume of this 

 solution used varied as set forth below. The final washing in each 

 case was with 25 cc. of water containing ammonium hydroxide. 



