( 041 ) 



li, is rcuuirkjihlc llial (lie hxw of diliilioii provi's lo appl.y here 

 lip to fairly higli coiiccntralioiis. Il would llicrcfbre be inleresting 

 to carry on the series towards higher concenli-alions to see where 

 the deviations begin to appear. 



That the freezing point method is infei-ior (o the vapour tension 

 method and the boiling point method with a view to accni-acy 

 folIo\vs also from the faet, that no constant values for K can be 

 calculated from the observations of Loomis, Jones and Getman, as 

 api)ears from the folloAving table. 



T A B L E IX. 



KNO3. 



From the determinations of the freezing point. 



The fact, however, that Biltz-) obtained concordant results for 

 solutions of caesium nitrate by means of the freezing point method 

 justities the hope, that Avhen the experiments are made very carefully, 

 also by this method the law of dilution will prove to hold tor 

 KNOj-sohitions. 



I have agreed with Dr. Biltz that he will examine the behaviour 

 of chlorates, perchlorates and })ermanganates Avilh regard to the law 

 of dilution and I shall investigate the nitrates. 



The above salts manifest little tendency for complex formation and 

 are therefore the most suitable material for the above mentioned 

 purpose. 



Febr. 1904, Ainsti'nlam. 



Chf'inhiJ Lahorafory of the University. 



1) Phys. Rev. 3, 279 (1890). 



2) loc. cit. 



