1904.] Theory of Amphoteric Electrolytes. 279 



accounted pure in the ordinary analytical sense no less than twenty- 

 four recrystallisations were required to reduce the conductivity to a 

 constant value unaffected by further recrystallisation. The solvent 

 water employed during the last crystallisations and in the determina- 

 tion of the conductivity was purified in the manner described by 

 Walker and Cormack.* The constant molecular conductivity of 

 asparagine found with this water at v = 16 was 0'096. This experi- 

 mental number is still somewhat too great owing to the impossibility 

 of removing all conducting impurity from the water, the error due to- 

 this cause being probably of the dimensions of one-tenth of the total 

 conductivity.! In view of the experimental error, then, the agreement 

 between the values calculated and found is as close as could be 

 expected. 



The results in the case of cacodylic acid have been published by 

 Mr. Johnston elsewhere, J and are also in satisfactory accordance with 

 the theory, which has thus been tested by conductivity and hydrolytic 

 experiments in a range of k a from 5 x 10~ 9 to 1-6 x 10~ 5 , and of k b /K 

 from 32 to 1100. 



An interesting theoretical point arises in connection with cacodylic 

 acid, the formula of which is generally written (CH 3 ) 2 AsO.OH. This 

 formula does not come under the general formula H.X.OH, which is 

 applicable to the amino-acids, inasmuch as the only replaceable 

 hydrogen atom it contains is that of the hydroxyl group. The distinc- 

 tion between these two classes of amphoteric electrolytes has already 

 been pointed out by Ostwald. From the standpoint of the theory just 

 put forward, however, it is unnecessary to consider it, for as long 

 as the unionised substance is not subject to molecular association, 

 the theory is applicable without modification to either class. The 

 freezing-point determinations of Zawidzki show that this condition i& 

 fulfilled for cacodylic acid. 



With regard to the freezing-point depression and correlated 

 phenomena exhibited by non-associating amphoteric electrolytes, the 

 following points may be noted. Since the production of HX + and 

 XOH- from H.X.OH in the first class, or of Y + and YQ- from 

 Y.OH in the second class, involves no change in the total number 

 of molecules, this type of dissociation is not connected with any 

 alteration in the freezing point of the solution. The production of 

 H + and OH~ on the other hand increases the total number of 

 molecules, and in consequence the freezing-point depression. We may 

 say briefly, then, that the acidic or basic ionisation of an amphoteric 

 electrolyte increases the freezing-point depression or any derived 



* ' Journ. Chem. Soc.,' vol. 77, p. 8, 1900. 



f Compare Walker and Cormack, loc. cit., p. 18. 



J ' Berichte Deut. chem. Ges.,' vol. 37, p. 3625, 1904. 



' Zeitschrift fur Elektrochemie,' vol. 6, p. 36, 1899. 



