905 



but not exclnsivel}' with tliat either, because otherwise the action of 

 zinc and magnesium could not be well understood (for the rest see 

 fig. 2). 



E. In the course of oiii- preliminary experiments we had observed 

 that lithiumsulphate exercised ft. smaller action than sodium sulphate; 

 we have repeated these experiments in a quantitative way on the 

 chlorides and supplemented them with NH.Cl and KCl. Except those 

 for lithium the figures are so near to each other that the deviations 

 fall within the range of the experimental error. 



Nevertheless we suppose the succession Li <^ NH^ <^ Na <[ K to 

 be correct, in accordance with the increasing electro affinity (table 

 V and fig. 3). 



Si> 



i'& 



'Jb 



:2 ^ 



S./^ 



T^ó. 



y<p 



salt 



'^Tt*^ 



Also with the alcaline metals a limit was soon reached, which is 

 situated at db 33 7o. therefore considerably lower than that of the 

 alcaline earths (BaCl^ := 60 7.) and at about the same height as that 

 of the other metals of the 2"<^ group, with the exception of mercury, 

 manganese and iron, which are situated much lower. 



With the fixation of the acid benzopurpurin we have kept our 

 attention fixed on the metal. Because we were forced to examine 

 some chlorides in order to obtain a survey of the alcaline eaiths, 

 we were able to compare the action of a few chlorides with that 

 of the sulphates. 



Though the graphs indicate a shifting of the action towards lower 

 concentrations, we did not in any case meet with essential diffe- 

 rences, which was to be expected. 



