﻿On the Viscosity of Liquids. 459 



These values practically cover the range for which coin- 



T 

 cklent values exist for the two liquids; in fig. 1 rTr is plotted 



J-o 



against T, and a straight line is drawn so as to lie closest to 



the points* 



This method of plotting is similar to that employed by 

 Ramsay and Young in connecting together the vapour- 

 pressures of different liquids. 



On a former occasion (Phil. Mag. June 1907, p. 724) 

 I showed that in so far as that curve may be taken as 

 perfectly straight, it must he possible to represent the 

 vapour-pressure of each of the substances concerned by the 

 formula g 



f(p) =A + 7 j T ; 



where /Qt?) stands for a function of p which must be the same 

 for both liquids, and A and B are constants characteristic of 

 each liquid. 



Applying the same reasoning to the present problem, we 

 see that in so far as the curve such as fig. 1 may be taken 

 as perfectly straight it must be possible to represent the 

 viscosities (//,) or the mobilities (M = 1//a) by the formula 



F(M) =/<» = A + |. 



A formula of this type fits even such an extreme case as 

 that of the viscosity of pitch as some, at present unpublished, 

 results obtained by Mr. G. B. Jeffery show. 



The case in which mercury and water are the two liquids 

 compared is specially interesting, because water is a strongly 

 associated liquid while mercury is considered to be non- 

 associated. In the corresponding vapour-pressure problem 

 it is usually concluded that a deviation from straightness 

 indicates that one of the vapours " undergoes dissociation 

 on heating " or " that there is molecular association in the 

 liquid state"" *. It would seem that the degree of curvature 

 cannot be taken as a criterion of the degree of association in 

 the analogous viscosity problem. 



Similar curves have been obtained for bromine and water, 

 oil of turpentine and water, ethyl ether and benzole, ethyl 

 acetate and benzole, in each case with similar results. 



wessure 



2. Mobility and Vapour- p, 



The discovery of the above similarity between a problem 



in connexion with vapour-pressures and one in connexion 



with viscosities (or mobilities) suggested that other relations 



* Young, ' Stoichiometry," p, 145. 



