1889.] Viscosity and Conductivity of Electrolytes. 23 



liarities which arrest attention. As an example of very normal 

 type, I have in fig. 1 reproduced the curve (k) of conductivity 

 and concentration in gramme equivalents per litre of NH 4 C1 from 

 Wiedemann's Electricitdt, vol. I. p. 610. And to compare with it 

 I have plotted a curve, from observations (f) of Grotian*, of the 

 fluidity of solutions of NH 4 C1 for the same range of concentration. 

 There is no striking resemblance but rather the reverse. In fact, 

 the curves for conductivity and concentration are generally roughly 

 parabolic in shape with a vertex of maximum conductivity, the 

 equations of a series of the curves are given by Wiedemann"]*, and 

 the curves for fluidity are according to ReyherJ and Arrhenius§ 

 very exactly represented by the equation y = A x , where y is the 

 viscosity relative to water, x the number of gramme equivalents 

 per litre, and A a constant not differing much from unity. Such 

 curves are of course quite different from the conductivity curves, 

 and yet, when tables of conductivities and fluidities of different 

 solutions are compared, the connexion is very striking; and the 

 obvious suggestion is that both fluidity and conductivity are very 

 complex phenomena, but that for a given solution they are both 

 dependent ultimately in some way upon temperature and con- 

 centration, and are so related that, other things being unaltered, 

 if the fluidity be increased the conductivity is consequently im- 

 proved. It therefore seems more hopeful to compare, as Grotian 

 has done, fractional rates of variation of the two quantities 

 rather than the quantities themselves. Grotian has chosen the 

 rate of variation with temperature, and in parallel columns on pp. 

 949 — 952 of vol. II. of Wiedemann's Electricitdt are values of 10 </> 



and 10V where 6 = ( -f- ) — and K = (- Tr ) j~ , for different con- 



W 22 /, 8 \dtJnh 



centrations of solutions of a large number of salts. The suffixes 

 indicate temperatures, / is fluidity relative to pure water at 10° C, 

 k conductivity referred to mercury, and t temperature. A further 

 advantage of this method of comparison is that the units in which 

 the conductivity and fluidity are measured do not affect the 

 result. 



In order to get a better general view of the comparison of 

 the two magnitudes <f> and k, I have examined the curves plotted 

 from all the results recorded in Wiedemann, and the general 

 parallelism of corresponding curves is very striking. The set 

 of lines representing either one of the properties for the solu- 

 tions of the different substances traverse the paper at a very 

 great variety of inclination to the axes ; some of them are nearly 



* Pogg. Ann., vol. clx., p. 259, 1877. 

 t Electricitdt, vol. i., p. 600. 



X Zeitschrift fur Phys. Chem., vol. n., p. 744, 1888. 

 lb., vol. i., p. 285, 1887. 



