Theory of Electrolytic Dissociation. 221 



in taste and more toxic towards plants than they ought to be 

 according to their content of hydrogen ions, as determined 

 by the methods in vogue*. 



That the hypothesis of Arrhenius cannot be harmonized 

 with the facts established in the domain of thermal chemistry 

 has repeatedly been emphasized by various writers, and has 

 been well summed up by Reychler in his treatise on physico- 

 chemical theories. The conclusion reached by Reychler is 

 that " above everything else we notice that the hypothesis of 

 free ions is in opposition to thermochemical observations." 



It is a well-known fact that the dissociation theory cannot 

 be harmonized with the law of mass action, and that this 

 constitutes a strong argument against the hypothesis. For 

 solutions of weak organic acids in water, to be sure, Ostwald 

 succeeded in computing a series of constants by means of the 

 formula expressing his well-known dilution law, growing out 

 of an attempt to apply the law of mass action to electrolytic 

 dissociation. But an examination of the constants obtained 

 for these weak acids reveals the fact that the agreement is 

 after all only rough ; furthermore, the attempt to apply the 

 dilution formula to electrolytes par excellence has utterly 

 failed, and yet the advocates of the dissociation theory con- 

 tinue to calculate the so-called solubility product, which 

 frequently involves the application of the law of mass action 

 to strong electrolytes. Attempts made by Lincoln to apply 

 the dilution formula to electrolytes in non-aqueous solutions 

 of about the same order of conductivity as that of some of 

 the solutions of weak organic acids in water, have also proved 

 futile. How can the formula possibly hold in cases where 

 the molecular conductivity remains practically constant, or 

 decreases as the dilution increases ? Rudolphi and also van't 

 Hoff have sought to amend the formula by changing the 

 exponents so that constants would result from the observed 

 data in special cases. While thus better " constants " were 

 obtained in these instances, the altered formulae proved 

 entirely unsatisfactory in other cases ; and then, too, it must 

 be added that the alterations made the formulae simply 

 empirical in character, without any theoretical basis. 



The difficulties met with in applying the Nernst formula 

 Eor calculating the electromotive forces of galvanic cells have 

 been discussed in special papers to which the reader is re- 

 ferredf. It may be remarked here, however, that in the work 

 on electromotive forces especially, many of the solutions 



* Compare Jour. phys. ('hem. iv. p. 33 (1900) ; iv. p. 207 \ iv. p. 53£; 

 iv. p. 553. 



t Journ. pkys. Chem. iii. p. .379 (1899) ; also iv. p. 709 (1900). 



