10 ART. 0. G. N. POTDAR : ON THE PARTITTON 



As to the numerical value of the partition coefficient the 

 general mean of all the results obtained was 289. This is 

 perhaps too low, because even a small amount of the lighter 

 phase contaminating the heavier tends to decrease the observed 

 partition coefficient, and as we have seen this source of error is 

 particularly difficult to eliminate. If we omit the three lowest 

 values (Nos. 3, 8, and 9) the average is raised to 309, which 

 may not be very far from the truth. We may therefore affirm 

 that at about 540° silver is about three hundred times more 

 soluble in zinc than in lead. 



The principal feature of Parkes' process of desilverisation 

 is the partition of silver between solid zinc and liquid lead. In 

 the first place argentiferous lead is melted and heated to a little 

 above the melting point of zinc when a certain amount of the 

 latter metal is added and the mixture stirred. The melt is then 

 allowed to cool slowly. Scum consisting principally of zinc and 

 containing silver (together with gold, copper, etc.) rises to the 

 surface and is continually skimmed off. The temperature is 

 allowed to fall nearly to the melthig point of lead. Usually a 

 second and sometimes a third or farther addition of zinc is 

 made, the temperature being raised each time. In this way the 

 desilverisation may be made practically complete. 



The amount of zinc remaining in the desilvered lead depends 

 upon the temperature at which the process is brought to an 

 end ; because the solubility of solid zinc varies considerably with 

 the temperature. The following diagram will serve to make the 

 matter clear. It is purposely not drawn to the scale, in order 

 to bring out the essential features more clearly. Ä and jP are 

 respectively the melting point of lead and zinc. B and JE are 

 eutectic points, while C represents the temperature and the com- 



