( 94 ) 



of' energy 



I'' 



the quantity q^^ = 0. Bnt llien, this is a, very special case, which is 

 probably only reached by approximation in reality. In any case y 

 cannot be = even then, and even though the "internal" varia- 

 tion of energy = at the absolute zero-point, it is not at the ordinary 

 temperatures (because of the term yBT). 



And where for water, acetic acid etc (also in the va|»our) real 

 association is assumed, it is no more than consistent in my opinion 

 to assume this "real" association in (til cases by analogy. 



Whether we consider the matter from a kinetic or from a thermo- 

 dynamical point of view, we always come to the same results, in 

 my opinion. If at a certain moment we could tix the state in the 

 whirl of the molecular movements — we should always see a certain 

 jiumber of groups, where two molecules are in each other's imme- 

 diate neighbourhood (and stay there for some time, however short 

 it be) ; where three, four, or more molecules happen to be together, 

 etc., etc.^j. In the same way the real association is thought also 

 thermodynamically. The priucii)le of the "moltile e(|uilibrium" involves 

 that a certain number of the formed double molecules break up 

 again into simple molecules in a certain time etc. And the known 

 thermodynamic prmciples are applied to the "state of equilibrium" 

 which has set in in this way. 



So association ; but besides variation of volume caused by the 

 association. For again : without assigning some value to Nj, we do 

 not arri\ e at the solid state. The theory developed in V and Yl has 

 proved this convincingly in my opinion. 



And now it is, indeed remarkable, that in his theory of quasi- 

 association van der Waals tloes assume contraction in the \alue of 

 a — which is supposed constant in our theory (see above) — but 

 no change in the value of h .'^). 



No doubt VAN DER Waals will have had a good reason for this 

 contraction in the value of a, — the matter, however, has not 



1) In connection with this we may refer e.g. to the theory of "Schwarmbiidung" 



of V. SCHMOLUCHOWSKI. 



2) See these Proc. June 1910, p. 119-121 (with regard to // p. 121); also 

 Nov. 1910, p. 494. With reference to the value of />, van der Waals owns that 

 a6 will not be = 0, and even malies the supposition that A/; will probably be 

 nearly always positive. 'But then what about the melting-point lines running to 



the left?). Notwithstanding this he assumes provisionally /^j = 0. 



