1904] On the Mechanism of Electrolytic Conduction. 281 



increasing 5-fold between and 18 and 80-fold between and 

 100 C. ; the form of the conductivity-temperature curves again 

 appears to be determined mainly by molecular changes rendered 

 .apparent in the rapid changes of viscosity which accompany changes 

 of temperature. Moderately dilute solutions give curves that are 

 inflected between and 100 ; a regular curve can be drawn connect- 

 ing the temperature of inflection with the percentage of sodium 

 hydroxide in the solution : this temperature reaches a minimum, at 

 48 J C., in the case of a normal (4 per cent.), solution, but rises to 

 100 C. when the concentration is raised to 30 per cent. 



(2) The inflected conductivity-temperature curves can be represented 

 by the simple cubic formula 



where T is the temperature of inflection. As this formula is unsatis- 

 factory when applied to curves that are not inflected between and 

 100 C., a formula was devised which was based upon the three 

 factors of concentration, ionisation and ionic mobility, on which 

 the variations of specific conductivity depend. The influence of 

 temperature on the density, and therefore on the concentration, has 

 been experimentally determined and can be directly allowed for ; in 

 order to express the influence of temperature on the other two factors, 

 it has been assumed that the variation of ionic mobility with tempera- 

 ture may be expressed by a formula similar to that of Slotte for the 

 variation of fluidity 



and that the decay of ionisation with rising temperature may be 

 expressed by a formula similar to that of Abegg and Seitz for the 

 decrease in the dielectric constant 



D/D O = -. 



The formula = (1 + bt) n e~ at , which has been used to express 



*o po 



the influence of temperature on conductivity, has the advantages that 

 it is applicable to conductivity-temperature curves of all kinds, that a 

 definite physical meaning can be given to each of the constants, and 

 that it gives expression, not only to the inflection now under considera- 

 tion, but also to the maximum conductivity and the second inflection 

 in the general conductivity-temperature curve.* What is, perhaps, of 

 ven greater importance, is the fact that it has been found possible, 

 with the help of this formula, to make some approximation to resolving 

 the two opposing influences which determine the form of this curve. 



(3) The density tables for caustic soda, in general use until about 

 10 years ago, were based on the early measurements of Dalton and of 



* Loc. cit., p. 52. 



