160 Prof. J. Traube: 
molecules, though we might infer from the electric conduc- 
tivity that the dissolved body was strongly dissociated. Again, 
the concordance between conductivity and affinity (? Ajinr- 
tiitskonstant) of acids and bases is often missing, and on the 
other hand it is known that the bases follow one another in 
the same series of affinity as in water in other solvents, 
where an electrolytic dissociation cannot occur *. 
These and other reasons show that the doubts against the 
fundamental idea of Arrhenius are justified ; notwithstanding, 
the success of that hypothesis is so enormous that we may 
ask : How can it happen that a theory based upon false 
assumptions has had such an influence upon the development 
of physics and chemistry ? 
I think the answer is not difficult. 
The reason for the considerable progress of the theory by 
Arrhenius in comparison with the former theories, especially 
that of Clausius, is to be found in the circumstance that 
Arrhenius has proved the existence of two sorts of particles 
in the solution of an electrolyte which have different values 
in reference to the most different properties, and above all 
that he has calculated the relative quantity of both particles. 
The calculation of the coefficient of dissociation is the undis- 
putable merit of the Swedish author, but the interpretation of 
that coefficient we should certainly reject. 
By comparing Arrhenius’s theory with the former theory 
of Clausius, we find the following difference in the respective 
assumptions :— 
Both Clausius and Arrhenius assume dissociations in a 
solution of electrolytes, but according to Clausius every 
dissociation of a molecule is accompanied by an association 
of another even in very diluted solutions, while Arrhenius 
assumes nothing but dissociations in diluted solutions. The 
idea of Clausius was in agreement with all the views of the 
chemists, not so the idea of Arrhenius, but the last one seemed 
to explain in a very simple manner the osmotic and electro- 
lytic behaviour of the solutions. 
Let us try if we cannot obtain the advantages of Arrhenius’s 
theory by returning to and improving Clausius’s assumptions. 
While at first, after the advance of the osmotic theory, 
the effect of the solvent often was quite neglected, there is 
to-day nobody who does not admit the great influence of the 
solvent on the properties of solutions. That influence is 
quite specific for every solvent. The different dissociating 
power of the solvents, the different quantity of heat produced 
or disappearing especially at the solution of a salt which may 
* Goldschmidt & Salcher, Zeztschr. Phys. Chem. xxix. p. 89 (1899). 
