ON ELECTEOLTSIS IN ITS PHYSICAL AND CHEMICAL BEARINGS. 375 



In a word, we believe experiment has verified the following- rule deduced from 

 the present theory : — 



(40) If an acid HR acts on a salt MR', and if a and ^ are the activity- 



coefficients of the salts MR' and MR, the coefficients of relative affinity 

 between the acids HR and HR' vary almost proportionally to the 

 square root of the quotient aj^. 



In § ] 3 of the first part we remarked that the molecular conductivities of diflerent 

 salts approach each other, in such a way that they would seem for extreme attenu- 

 ation to advance towards a common limit. It follows that the quotient a/j3 

 approaches more and more to unity, and that the different values of a//3 become equal. 



This is why the relative affinity Q beco7nes more and more constant as dilution 

 increases— a, relation Ostwald thinks he has already perceived with dilutions less 

 extreme. 



The activity-coefBcients of acids and bases also approach equality as dilution 

 increases. 



(41) Consequently the coefficients of relative affinity of acids _ approach more 



and tnore to unity as dilution increases. 



M. Ostwald has also on this subject made some measurements which confirm 

 the justness of the above proposition. His researches on the magnitude of the 

 coefficient of relative affinity of HjSO^ indicate that at extreme dilutions this 

 coefficient is * almost equal to that of HNO3, at least it rises to '9 ' (that of HNO^ 

 bein^ 1), while at moderate dilutions it is '5 or less. 



To the invariability of Q for acids and bases M. Ostwald adds a fact of great 

 importance. The molecular conductivities of feeble acids and bases increase rapidly 

 with dilution, i.e., the value of d' in weak solution is much greater than that of 

 S in strong ; so Q increases also, if we compare either of them with a strong acid 

 or base, which is nearly unaffected by addition of water, for — 



Q = A /^ for small dilution, 



Q'=V 



aS' 



for large dilution. 



Hence — 



(42) The relative affinities of feeble acids and bases (compared with a strong 



acid or a strong base) increase considerably as dilution goes on increasing. 



This is the reason for not attributing to these ' coefficients of relative affinity,' 

 or ' avidities' as M. Thomsen calls them, a too great importance. By believing 

 them constant one may seem to incline strongly towards regarding them as funda- 

 mental niunbers. Besides, the determinations of them by different workers do Jiot 

 agree well with each other. 



(43) As the activity coefficients of different acids change a little unequally 



with temperature, the relative affinity of two acids ought also to vary 

 with temperature. 



Thus the activity-coefficient of H.,SO^ increases more slowly than that of 

 monobasic acids (HCl or HNO3), whence it evidently residts that its affinity, relative 

 to either of these acids, diminishes as the temperature rises. We can give only a 

 rough calculation. The temperature-coefficient of conductivity for HCl or HNOj 

 exceeds by 4 per cent, that of H2SO4 at 20° G. ; so Q, which is proportional to square 

 root of conductivity, will vary by about 2 per cent, per degree at this temperature. 



The temperature-coefficient for H^SO^ is still less at higher temperatures, so its 

 relative affinity ought to diminish faster at these temperatures. Experience agrees 

 with this as well as one can expect. 



M. Ostwald has found the following figures for the relative affinity of HCl to 



Between 0° and 40° Q varies about -16 per cent, per degree, while rough calculation 

 indicated '2 per cent. ; at higher temperatures the variation is greater, as expected. 



