( 744 ) 
And for the entropy 4. 
fy ip m Pace 3 oe : 
yan, + ike “+ 5 log (2a@m,) + 5 loy (227@m,) 
+ n, log Aan, + n, log tan, 
—n,, logn,, —7,, logn,, — n,, logn,, — n,, logn,, — No, log n,, 
— Ny, — Nyy — Nyy — No, log 2 — n,, log 2 
= (n,, = Ni, LE Nis ate Nos = No) log Li 
+ n,, log w,, + n,, log w‚, + n,, log w‚s- 
12. If we now eompare equation (6) with Gipps’s thermodynamically 
derived equation, we find the above supposition (p. 707) quite 
confirmed. For apart from the variability of the specific heats which we 
have excluded by our suppositions about the constancy of z and the 
character of the bond between the atoms, the two equations agree 
entirely, only in the thermodynamic equation the indefinite quantity 
YvH occurs, here Er logw. So the ‘theorem of heat” reduces to 
this thesis, which is self-evident on the suppositions of BoLTzMann, 
that the “chemical volume” w of a definite chemical substance is a 
quantity characteristic of this substance, with which it appears in 
each of the reactions in which it takes part. So it is also these 
volumes whose size can be determined from the equation of equili- 
brinm. And it is clear from what has been said above (p, 706) that 
from reactions for which the number of molecules does not change, only 
relative values of these volumes can be found for different substances. 
If the number of molecules does change, absolute values can also be 
ealeulated for them, as indeed, has already been done by BorTzMann *). 
If we “finally ask whether Nernst’s numerical calculations have 
supplied us with new data for the knowledge of the chemical volumes, 
we dare not answer in the affirmative. First of all it has, namely, 
already appeared from calculations by Dr. Scurrrer, that the arbi- 
trariness which remains in the choice of the specific heats is such 
that even for the most closely examined reactions we can vary the 
value ef rH within a very wide margin without coming into 
conflict with the experiments. Therefore it may justly be doubted if 
not other values than those given by Nernst, would agree with the 
experimental data. Moreover, however, Nernst’s values contain not 
only the rH, but also the Yrc,. So a calculation of the volumes 
under consideration, which wants to give more than the order of 
magnitude is possible only when one takes the difference in specific 
1) Gastheorie II § 66. 
