278 Prof. J. Walker. [Oct. 19, 



If, in the table on p. 274 we interchange k a and fa, the values of uv 

 remain unaltered. We are now, therefore, in a position to consider 

 the whole series of amphoteric electrolytes with a constant product 

 k a fa, beginning with a practically simple acid (i.e., with a substance 

 whose fa is of small dimensions compared with K) and ending with a 

 practically simple base (i.e., with a substance whose k a is small com- 

 pared with K). As k a diminishes and fa increases, the total ionisation 

 falls off rapidly at first, thereafter more slowly until through a com- 

 paratively large range it is practically constant at the minimum value 

 which is actually reached when k a = fa. At this point we are dealing 

 with an absolutely neutral substance. As k a still further diminishes 

 .and fa correspondingly increases, the ionisation begins to increase 

 slowly at first, and the substances considered become more and more 

 T)asic in character. Finally the ionisation increases rapidly, and we deal 

 At last with what is practically a simple base. 



In applying the preceding results it is essential to bear in mind 

 that they have been obtained on the assumption that K may be 

 neglected in comparison with k a u in dealing with an amphoteric acid, 

 and with fau in dealing with an amphoteric base. Should, therefore, 

 the preponderant constant or the total concentration be very small, 

 due allowance must be made for this neglect. 



Since the publication of my previous paper, Mr. John Johnston, 

 B.Sc., has made in my laboratory some experiments on the conductivity 

 of asparagine and of cacodylic acid, two well-marked amphoteric 

 electrolytes of character somewhat different from the amino-benzoic 

 -acids already investigated. 



Asparagine is of special interest, inasmuch as both its acidic and 

 basic constants have been determined by Winkelblech from hydrolysis 

 experiments on the two types of salts to which it gives rise. The 

 values found by him were k a = 5'2 x 10~ 9 and fa/K 150. Calcu- 

 lating from these values and adopting /^H 4 ", XOH~ = 355 and 

 /* W HX + , XOH~ = 66, in accordance with the method previously 

 described, we obtain as the molecular conductivity at v = 16 the 

 number 0'087. The value obtained directly by Walden* was 0'541 

 Winkelblech f also determined the molecular conductivity, and found 

 numbers some 40 per cent, smaller than those of Walden, although he 

 still considers them too great. The reason of the discrepancy is the very 

 small value of the conductivity of asparagine, any conducting impurity 

 either in the substance itself or in the solvent water having a very 

 great relative effect. At my request Mr. Johnston carefully purified 

 a specimen of asparagine by repeated recrystallisation from water 

 until the conductivity showed no further diminution on repetition of 

 the process. He ascertained that with a specimen which might be 



* ' Zeitschrift fur physikal. Ohera.,' vol. 8, p. 483, 1891. 

 t Loc. cit., vol. 36, p. 563, 1901. 



