iat 
virtue of the above Cia > Cuas. We will take 15 for this difference, 
which will probably be high enough. The curve which corresponds 
to this has for equation 
a 
7 
In order to reach this purpose as well as possible, I shall draw 
3 
5 
. Sparta pl calle yy 
logi > log T + ¢. 
the curve through the two most favourable observations (2"¢ and 
3'¢ of table 9). We then tind « — 6390, c= + 34.415. From this 
we find for the point of transition loy p — + 0.344 (see fig. 5). This 
is therefore the lowest value that can be reached, all allowances being 
made. From this calculation and fig. 5 we see clearly the slight 
influence of the term with the specific heats. 
Even with the most favourable assumption of errors of observa- 
tion and of the data that are little or not known, it appears impos- 
sible to explain JoHNson’s experimeuts by the above assumption. I 
have ascertained by determination of curves of heating and cooling 
that between 180° and the temperatures of the experiments of eva- 
poration there is no second transition point to be found. 
13. If JoHNson’s experiments are correct, the difference in ther- 
modynamic potential of the solid NH,Cl in humid and dry state is 
still to be explained. There is no explanation to be found in the 
phenomena of allotropy discussed in this paper. 
Before the cited investigation by Smirn it was pretty generally 
accepted that ammonium chloride in gas form practically splits up 
entirely into NH, and HCl. If this were so, the partial tension of 
the unsplit part could only be very slight. If during the drying the 
homogeneous dissociation did not take place, we should have to 
find a vapour tension which is very small. Yet Baker has carried 
out vapour density determinations with non-dissociated vapour at 
atmospheric pressure ! 
Now it has appeared in Smirn’s investigation that the dissociation 
is far from perfect; the partial tension has therefore a pretty great 
value. From the calculations of § 12 it appears that the value of 
the vapour tension of p-NH,Cl on the assumption that the homo- 
geneous dissociation does not take place, is indicated by: 
logp = — 
and the pressure of «-NH,Cl, the transition heat mentioned in § 10 
being taken into account, by: 
4835 
— + 10,281. 
Tah 
logp = — 
