26 Mr W. N. Shaw, On the relation between [Nov. 25, 



concentration but depending on the solvent and possibly also on 

 temperature. 



The physical interpretation of this equation would be that 

 the effect of temperature upon the conductivity of the solution 

 is of a two-fold nature, (1) the conductivity is indirectly increased 

 by the increasing fluidity of the liquid and (2) it is diminished 

 by some alteration of the properties of the solvent which does 

 not affect the fluidity. In electrolytes at ordinary temperatures 

 the first effect is predominant, but on very great rise of tempera- 

 ture (without secondary alterations of condition) the second effect 

 might become very great compared with the first. Thus if the 

 solution (with the salt) were volatilized, the conductivity of the 

 gaseous mixture might be only a small fraction of the conductivity 

 of the solution though the viscosity might have become much 

 less. Integrating equation (1) with regard to temperature (as- 

 suming <x to be constant) we get 



/o % 



or ' h t =j«f t e-°<t-^ (2). 



Jo 



k 

 In equation (2) ~ expresses the relation between conductivity 



J° 



and fluidity at a standard temperature*. If this relation be a 



complicated function of the concentration, as it appears to be, 

 there is no reason to infer a general simplicity of relation between 

 k and / from the fact of their having temperature coefficients 

 which are connected by a simple relation. If we were possessed 

 of experimental data that would enable us to refer the properties 

 of one class of electrolytes to concentration and temperature as 

 variables, in a manner somewhat similar to that in which the 

 properties of gases are referred to pressure and volume, further 

 insight into the nature of the relation might be obtained. 



The evident relation between viscosity and resistance has not 

 yet been satisfactorily accounted for. The hypothesis that the 

 motion of the ions, if these be atoms, is opposed by frictional 

 resistance which can be measured as viscosity for ordinary motion 

 of the liquid seems to be regarded as dubious, although Kohl- 

 rausch"f* has shewn " that the supposition of mechanical and electro- 



* Since this paper was read I have seen a paper of Arrhenius (Zeitschr. fur 

 Phys. Ghem. Band. iv. Heft. 1, July 1889), in which an equation practically identical 

 with (2) is deduced directly from the theory of dissociation. On that hypothesis a 

 would be the fractional temperature coefficient of dissociation, and being negative 

 would imply, as Arrhenius points out, a temperature of maximum conductivity 

 beyond which the temperature coefficient of conductivity would be negative. 



t B. A. Report, 1886, p. 343; Wied. Ann. vol. vi., p. 207, 1879. 



