12 ART. 'J. — Ct. N. POTDAß : ON THE PARTITION 



paratus described above, I found the solubility of solid zinc in 

 melted lead to be 0.55^. Assuming the curve BC to be a 

 straight line the composition of the point B is estimated at 

 about 0.2 9ö of zinc, the temperature being but very slightly 

 below 325°, the melting point of lead. 



Hence if the last scumming in Parkes' process takes place 

 between the temperatures 325° and 360° as it appears to be the 

 case in practice, the amount of zinc remaining in the desilvered 

 lead will lie between 0.2 to 0.6^0. That this amount is so 

 small is imjwrtant for the economical success of the process. 

 The quantités of zinc added in excess of this are available for 

 the extraction of silver. 



The recent exhaustive study of zinc-silver alloy by Petrenko'^ 

 in Tammann's laboratory has shown the existence of a series of 

 compounds and solid solutions. But in the temperature range 

 which comes into consideration in Parkes' process none of the 

 compounds has a realm of independent existence, so that we 

 shall have a, series of surfaces of solubility of zinc-silver solid 

 solutions in lead, provided no ternary solid phase is formed. 

 Hence it is clear that the j^artition coefficient must vary with 

 the concentration, and the variation itself must show sudden 

 changes at certain points. Yet for very dilute solid solutions 

 the partition coefficient must be practically constant ; for the 

 law of dilute solutions also applies to solid solutions. 



I have tried to test this conclusion experimentally, but Avitli- 

 out success. The temperature chosen was 358°, the boiling point 

 of mercury. The three metals were melted together in a bath 

 of sulphur vapour, then transferred into one of mercury vapour 



1) Zeitä. aiiurg. Chem., 48, (1906), Ö47. 



