792 



sideratioii of the curve for KNO, seems to indicate al first only a 

 single large maximmn at about 0.00030 ii. KOH, but when we fix 

 our attention on the course of the lines joining the limiting concen- 

 trations, we find that there are really five apices, namely at 

 0.00010 n. HNO„ at 000010 n. KOH, at 0.00030 n. KOH, al 

 0.00050 n. KOH and at 0,00065 n. KOH. They are arranged very 

 symmetrically on either side of a middle axis. 



I have found nothing in colloidal chemical literature permitting 

 of a direct comparison with our case. Perhaps the reason is that 

 the behaviour of colloids towards salt solutions at various H-ion 

 concentrations has generally not been examined systematically, or 

 where this was done, the quantities of acid and base added were 

 much larger. 



Nevertheless the occurrence of these summits is not in confiict 

 with present views on the behavioui- of amphoteric colloids. The 

 proteins for instance might be able to combine with one or with 

 more molecules of acid or base. lu this way compounds might be 

 formed which could behave very differently towards salts. No 

 agreement has been reached about the conditions determining the 

 existence of these various protein-acid and protein-base compounds. 

 T. B. Robertson '), however, considers that he has proved that their 

 stability is exclusively determined by the H-ion concentration of the 

 solution. If this view is correct, then the presence of the summits 

 in our curves might perhaps be susceptible of explanation. 



The concentrations, in which the salts acted, were small. The 

 highest amount was found for Ca(NO,),. i.e. 0.40 normal. Very small 

 values were found for phosphate. For binary electrolytes the con- 

 centrations were 0.10 — 0.20 normal. The greatest contrast was formed 

 by plurivalent cations and plurivalent anions (calcium salt-phosphate). 



We can also deduce something from figures 1, 3, and 4 about 

 the different behaviour of univalent and plurivalent ions. The curve 

 for potassium sulphate (fig. 1) has a vertical direction between 

 0.00015 n. H,SO, and 0.00100 n. KOH. Here we have a bivalent 

 anion together with a univalent cation ; the curves for KCl (fig. 4) 

 and KNO, (fig. 3) lun obliquely downwards to the left. Here there 

 is a univalent cation with a univalent anion. Experiments which 

 were made with Ca(NO,),, showed clearly that the curve which 

 might be plotted from it, would be still much more oblique, but 

 in the same direction as that for KCl and KNO,. In this case a 

 bivalent cation accompanies a univalent anion. If the direction is 



1) T. B. Robertson, Die Physikalische Chemie der Proteïne. Dresden 1912. 



