904 



the intensity of colour of the very diluted salt solution is inconsider- 

 able compared to that of the benzopurpurin, this error could be 

 neglected. 



The solutions of ferrous sulphate had to be prepared anew for 

 each determination, because after some time, in consequence of oxi- 

 dation, precipitation of ferric basic sulphate look place. As this was 

 not entirely to be prevented the figures for this salt are given with 

 some reserve, of the other salts standard solutions were made con- 

 taining 7,, millimol. per cent of which respectively 5, JO, 15, 20 

 and 40 cc. were used. For the rest we worked as is described 

 under B p. 898 with the exception that a temperature of 70° was 

 chosen. 



First we see that zinc and also cadmium in some degree join 

 these metals. Remarkable is the rapidly ascending course of the 

 curve for zinc, a thing we had already found with zinc chloride in 

 the concentrated solution of benzopurpurin. As yet we cannot decide 

 whether this is based on accidental deviations or whether the higher 

 temperature is the cause of it. As regards the metals of the iron 

 group itself, we see that the precipitating faculty increases.according 

 to the atom number of the metals, except in the case of manganese; 

 but the differences between manganese and iron are very small. 

 We investigated also the chlorides of manganese, cobalt and nickel 

 with which the same succession was found : Mn <^ Co <[ Ni with 

 almost the same figures. 



This result therefore agrees with what we found about the succes- 

 sion zinc, cadmium, mercury, viz. that the precipitating faculty is 

 not connected directly with the atomic weight of the elements, but 

 with a peculiar chemical property, for instance the electro affinity ; 



fa- 



~^i> 



io 



to 





^'<p 



"■^^ ^ salt i^n, -^rz :r>-tJ'-C 



