﻿642 Prof. J. A. Pollock on a 



vapour-pressure for a plane water surface at the same 

 temperature, and n is the ratio of the latent heat of vapori- 

 zation of water to that of the adsorbed fluid surrounding 

 the core of the ion. m is the mass of the adsorbed fluid, or 

 in the present application, the mobility of the ion reduced to 

 constant air density. The suggestion was also made that a 

 clue to the condition of the adsorbed fluid might be obtained 

 in the value of n found necessary in any instance for the 

 reduction of adsorption observations. 



In the case of the large ions, unity, to an accuracy perhaps 

 greater than one in a hundred, is the value of n which gives 

 the best fit to a line of the points representing the mobilities, 

 determined at different temperatures, when plotted against 

 vapour-pressures. In other instances mentioned in the paper 

 n is also unity. No heat change due to a variation of sur- 

 face energy is involved in the value of n, so in these cases 

 where n = l, as the heat per unit mass necessary to annul a 

 temperature change due to the mere alteration of state is the 

 same as that required to keep the temperature constant when 

 water evaporates, I think we may definitely conclude that 

 the molecules in the contained or adsorbed fluid are in the 

 same condition of aggregation as those of water. 



Now with regard to the intermediate ion, its mobility also 

 depends on the hygrometric condition of the air ; this ion, 

 then, as well as the large one is composed, partly at least, of 

 water molecules. But the mobilities are as sixty to one, 

 so, if no other cause of difference exists, the moisture 

 forming the ions must be in very different states in the two 

 instances. 



If such is the case, the observations of the mobilities of the 

 intermediate ion should only fall into line in relation to 

 the vapour-pressure if reduced according to the formula 

 (PilP2)m= (Pi/P2) 1/W ? W ^h some value of n greater than unity. 

 Unfortunately the results are not accordant enough to enable 

 the value of n to be determined in this way with any 

 accuracy, and as {~P 1 /P 2 ) 1/n , within the limits of the observa- 

 tions, is so nearly equal to unity for even small integer 

 values of n, all that can be done is to compare plots of the 

 measures in the extreme cases, firstly when n is taken as 

 unity, and secondly when it is put equal to some large 

 number. In the first instance, when n = 1, by the preceding 

 formula the mobility will be constant if p 1 / / P l = ^» 2 /P 2 . Ac- 

 cordingly the mobilities are to be plotted against the relative 

 humidities, as in fig. 3, and if the points fall into line it is 

 to be taken as evidence that the adsorbed fluid is in the 

 liquid condition. On the other hand, when n equals some 



