1 2 THE TRUE VALUE OF a OF VAN DER WAALS' 



vaporization by means of a modified equation of Crompton and of 

 Dieterici. Namely, L ==[F c /fV c —6 c )]ET LnfdjDj. This gives the 

 value of b c and by substitution in van der Waals' equation at 

 '/'. , a may be found. 



The sixth method is from the surface tension, which is the direct 

 measure of the cohesion. The relation between cohesion and surface 

 tension is given by the formulae of Young and Eötvös. 



The seventh method is to compute a from the pressure of a 

 gas or liquid when it is heated while the volume is kept constant. 

 It is obvious that this rise of pressure will be less the greater the 

 internal pressure or cohesion. By holding the volume constant the 

 distance between the molecules does not change and so there is 

 no change of the cohesion with temperature, provided, of course, 

 that the mass factor of the cohesion does not change with the 

 temperature. So we may find a from the expression (TdVjdT) L . 

 when the volume is the critical volume. 



None of these methods involves any assumption as to the other 

 constant b and they are in every way preferable, it appears to 

 me , to the method of van Laar who has used solely a computation 

 made from van der Waals' equation. The latter is first solved on 

 the basis that a and b are both constants and then a correc- 

 tion, of a hypothetical nature made to it. I have never felt any 

 reliance on this method and the results obtained by it are not to 

 be compared in reliability with the values found from these direct 

 measurements in the way I have indicated. 



We may preface our work with the statement that a for any 

 substance is constant and does not vary with temperature, except 

 as it is affected by association of the molecules or their dissociation. 

 There is no variation if the molecule remains intact. This opinion, 

 which has been almost consistently upheld by van der Waals, is 

 shown to be correct by the Eötvös surface tension law. a is the 

 mass factor of the cohesion ; the factors it contains are shown in 

 this paper; and there is no more reason for assuming that it varies 

 with the temperature than for assuming that gravitational mass 

 varies with the temperature. Van der Waals has shown the im- 

 possibility of a varying with the temperature in the manner 

 supposed by Clausius; and this work, showing that a depends 

 only on the number of molecules, their gravitational mass and 

 number of valences, clearly supports his conclusion. On the other 

 hand b clearly varies with the volume, and probably also, tho 

 to a less extent, with the temperature. 



