242 PROFESSOR JOHN GIBSON 



For concentrated aqueous solutions of good electrolytes the relationship between 

 A H and r expressed in these units is, over wide ranges of concentration, accurately- 

 expressed by the equation 



\ M = a + bF (1) 



where a and b are constants for each electrolyte. 



To avoid confusion, the units adopted by Kohlrausoh will be referred to as volume 

 units and the units here proposed as mass units. 



If the data for concentrated solutions of good electrolytes given by Kohlrausch and 

 Holborn be translated from volume units into mass units, and A M be plotted against T, 

 straight lines are obtained in almost every case. Fig. 1 shows the graphs obtained in 

 this way for HN0 3 , H 2 S0 4 , KOH, NaCl, and for comparison the corresponding graphs 

 in volume units. The graphs obtained by using volume units are in thin lines and 



those obtained by using mass units in thick lines. It is important to notice that y = — 



and Aji = As, so that the adoption of mass units instead of volume units does not affect 

 the numerical statement of the relationships which have been established for dilute 

 solutions : for when 8=1, as is practically the case in solutions more dilute than 0"1 

 normal, y coincides with n and A M with A. In such dilute solutions, whether we refer 

 to unit volume of the solution or to unit volume of the solvent, or to unit mass of the 

 solution or to unit mass of the solvent, the numerical expression of the experimental 

 values is practically the same. 



But in concentrated solutions the difference between the volume of a given solution and 

 the sum of the volumes of the solvent and the solute taken separately often represents 

 a very high internal pressure, and moreover this internal pressure varies greatly from 

 one electrolyte to the other, so that, by taking equivalent quantities in equal volumes, 

 wc by no means establish comparable conditions. Even in an ideal case of a solution 

 of a binary electrolyte without any such internal pressure, the concentration must be 

 of the order of ^V normal or less, if the condition of the solute in the solution is 

 to be at all comparable with that of a gas at ordinary pressure. There appears 

 therefore no logical reason in favour of volume units as against mass units, and, as 

 stated above, the expression for the relationship between the concentration and 

 conductivity becomes at once more simple and more useful when the mass units are 

 adopted instead of volume units. 



In order to test the applicability of equation (l) as closely as possible, the data 

 given by Kohlrausch and Holborn for concentrated solutions of good electrolytes, 

 in so far as they are sufficient for the purpose, were translated from volume units 

 into mass units, and from the new data thus obtained the constants a and b of 

 equation (1) were calculated by the method of least squares. The results of 

 these calculations are given in Table A, along with the data from which they are 

 derived. 



