364 KEPOET— 1886. 



statement deduced from tliis hypothesis is worth quoting (see also Wiedemann, 

 'Elek.', vol. ii. p.924).] 



(20) Every hotly which acts chemically by double decomposition on an electro- 



lyte is itself an electrolyte, as ivell as the products of the decomposition. 

 Therefore tuater, alcohol, aldehydes, Sfc. ^-c, are electrolytes, and 

 therefore conductors. 



[The a priori manner in which this statement is made is striking ; so is the 

 note immediately following.] 



' On this subject people have disputed for some time. They often attribute 

 the feeble conduction observed in these bodies to traces of saline impurity.' ' 



§ 5. Eelation between conductivity and chemical power of acids 



and bases. 



(21) The molecular conductivity of the active part of an acid {in dilute 



solution) is constant, and independent of the nature of the acid. 



For if the chemical formula of an acid is IIR, its molecular conductivity, accord- 

 ing to Kohlrausch, is h + r. Now the molecidar conductivity of dilute salts with 

 the same metal, MR, MR', &c., is the same ; hence r = r'- &c. And therefore it 

 follows that for acids also the molecular conductivity of HR, HR', &c., is the 

 same. A corollary is — 



(22) The more a dilute add solution conducts, the greater is its active part. 



Similar propositions may be stated for bases. 



[The author then goes on to consider the idea of coefficients of activity more 

 particularly in the light of the Williamson-Clausius view, and he gives reasons for 

 supposing molecular conductivity and coefficient of activity to be closely related, 

 and indeed proportional, to one another. He then says, let thon be defined to he 

 equal. A list of molecular conductivities for acids, bases, and salts from the 

 results of Kohlrausch is supplied, and it is pointed out that the best conducting 

 acids are the strongest. Appeal is made to the results of Berthelot, Thomson, and 

 Ostwald in support of this.] 



'Thus we believe we have proved that for acids and bases galvanic activity is 

 accompanied by chemical activity,' and probably non-electrolytes have no direct 

 •chemical action. Cf. Gore on the inactivity of anhydrous HCl. 



§ 6. Double decomposition. 



[This section is the key to the whole, and I had best translate it in full.] 

 Suppose two electrolytes, AB and CD, are dissolved in an inactive solvent. 

 Circular currents pass in the solution [i.e., Williamson-Clausius atomic interchanges 

 occur], whereby the bodies A D and B C are formed. If the coefficients of 

 activity of all four bodies A B, C D, A D, and B C were equal, they would all be 

 present in the same quantity, viz., half an equivalent of each, provided the original 

 substances were in equivalent proportions. And in any case an equilibrium would 

 establish itself, the rate of formation of each substance being equal to its rate of 

 destruction. Now the coefficient of activity of a body indicates that fraction of the 

 body which at any one time participates in the circular currents \i.e., is ready for 

 interchange]. If then this coefficient for A B is a, and if m equivalents of AB 

 exist in solution, am expresses the number of equivalents of A B which at any 

 moment take part in the circular currents. 



One could figure to oneself the process of double decomposition in the following 

 manner. The ions of active molecules rotate round one another. Then, consider- 

 ing the molecules A B, the ion A moves with a certain velocity in the neighbour- 

 hood of B until it comes near another ion, B', after which it follows B'. The 

 molecule A B exists until this happens. 



' See § 9 of letter on p. 387. 



