DILUTK SOLI TIo.NS AT THE FKKKXING POINT. 323 



temperature coefficients of the conductivity of electrolytes vary greatly, as either 

 the concentration or the temperature is changed. The numbers which follow show 

 tli is clearly for the case of l>ariuni chloride. The measurements were made by 

 determining the conductivity of the solutions between and 18 at intervals of 

 about 2 ; the figures were carefully plotted on a diagram, and the value of the 

 t;uigent to the curve estimated at several points. The following temperature 

 coefficients were thus obtained : 



Solution I. Concentration = 0'0004 gramme-equivalent per thousand grammes of 



solution. 



Temperature 

 Temperature. Resistance. coefficient per cent. 



1 766-3 3-11 



5 680-0 2-89 



9 608-5 270 



13 549-0 2-51 



17 498-5 2-39 



Solution II. Concentration = 0'000018 gramme-equivalent per thousand grammes 



of solution. 



Temperature 

 Temperature. Resistance. coefficient per cent. 



l c 12418 3-29 



5 10927 3-02 



10 9433 2-81 



15 8272 2-51 



20 7327 2-37 



Thus the temperature coefficient increases as the dilution gets greater, and also as 

 the temperature falls. 



This variation is also shown in some numbers given by KOHLRAUSCH* for the 

 conductivity of potassium chloride solutions of different concentrations. For a 

 normal solution, the conductivity at is 0'06541, and at 18 is 0'09822, the ratio 

 of these numbers being 1 '50. For a solution whose concentration is -j^ normal, the 

 corresponding figures are 0'000776 and 0-001225, giving a ratio of 1'58. 



Thus, in each case, the difference in conductivity between and 18 is greater for 

 the more dilute solution. Now the conductivity of a solution depends on two 

 factors : (1) the amount of ionization, (2) the velocities with which the ions move 

 while active. These velocities are increased by heating the solution, owing to a 

 decrease in what we may term the ionic viscosity, and at extreme dilution, when, 

 in such salts as potassium or barium chloride at any rate, the ionization is practically 



* ' Annalcn der Physik und Chemie,' N.F. vol. 64, p. 441, 1898. 



2 T 2 



