( 704 ) 
the theorem of heat is not fulfilled yet, for then we get in the 
neighbourhood of 7’=— 0: 
No 
; 0(2),—1 : 0(y,.— 
So ( (Ms =) for 7’—O0, and a fortiori 7 (1 1) become 
’ 07 5 oT 5 
zero, and the theorem of heat further imposes the condition, that 
the expression (4) becomes zero, or that: 
— = Er MR log 2 + SvH + Rp logy (A) 
(ty 
SoH SMR boy 4 SRvibog ys ee 
u. 
And now it has become clear why we could say above that our 
result is independent of the special form of the equation of state. 
For even if we should be willing to accept that the deviations 
which the real substances show from the simple suppositions made 
here, have always exactly that value that prevents the difference of 
entropy (2) from becoming infinite, this difference will never become 
zero as the theorem of heat requires, unless a relation like (5) is 
fulfilled. 
According to this expression we could calculate the difference of 
the constants of the entropy of the reacting substances before and 
after the reaction from quantities which are perfectly determined by 
the equation of state. If we, however, consider in what way these 
arbitrary constants of entropy have been introduced into equation 
(3), this result becomes altogether absurd. 
5. It seems to us that this has sufficiently refuted the “theorem 
of heat” for so far as we have to understand by it the opinion that 
0A 0 
ta and & must be equal in the absolute zero. Nor can the 
conclusion drawn from it, equation (1), be maintained. Nernst himself 
would probably have seen this, if instead of starting from the dif- 
dink N ; 
ferential equation for ae? which van ’r Horr has given, he had 
started from the integral relation for Ink, already drawn up by 
Gisss in his well-known paper. For then he would have known 
what the physical meaning is of the constant of integration occurring 
in his equation (1), viz. that this contains besides the specific heats 
the constants of entropy of the reacting substances, and so that 
exactly the same essential objection exists to his equation (1), as we 
have just advanced against equation (5). But then he would also 
have seen that if he wanted to calculate the chemical equilibria, he 
should not put the question: How do the substances behave at the 
