1499 
tions of the ammoniumchloride by a series of calorimetric experi- 
ments, in which Mr. G. Meyer assisted me; I gladly express my 
indebtedness to him here. 
A number of glass tubes was filled with ammoniumehloride and 
fused to; the quantity of glass and ammoniumehloride was deter- 
mined by weighing. The tubes were successively heated to 156° 
(boiling point of bromo benzene), 177° (ortho-dichloro benzene), 194° 
(dimethyl aniline), and 212° (nitrobenzene), and quickly conveyed to 
the calorimeter. The two first-mentioned temperatures lie below, the 
two others above the transition point. When we conveyed it to tlie 
calorimeter the glass broke sufficiently through the sudden cooling 
to bring about a very rapid dissolving of the ammoniumchloride. 
Hence in these experiments the heat of solution was always included 
in the measurement, and in order to avoid corrections for the heat 
of dilution we have always chosen the quantity of water in the 
calorimeter so that the obtained solution contained one gramme 
moleeule of NH,Cl to 320 gramme molecules of water. In consequence 
of the rapid solution the temperature had generally established itself 
already after 1 or 2 minutes, and the time, over which corrections 
have to be applied for radiation, was, therefore, only small, which 
enhanees the accuracy of the determinations. The specific heat of 
the glass was determined by separate experiments by heating up to 
156°, resp. 212°, and on the assumption that the true specific heat 
changes linearly with the temperature between 156° and 212°, the 
mean specific heat between room temperature and a temperature 
lying between the two said temperatures could then be found by 
linear interpolation. In every experiment with the tubes of ammo- 
niumehloride the heat which the glass alone yielded, was therefore 
known; if we subtract this heat from the found amount, we keep, 
in the experiments of 156° and 177°, the heat which «-NH,CI yields 
when it is cooled down from the said temperatures to that which 
the calorimeter possesses after the experiment, and then dissolved 
isothermieally. From these values the mean specifie heat of ammo- 
niumebloride between 156° and 177° can, therefore, be calculated, 
and by extrapolation to the transition point (184°.5) it is also to be 
found what heat @-NH,Cl would yield when it was cooled down 
from the transition point. 
The experiments in which heating above the transition point took 
place, yield, after correction for the heat of the glass, the heat which 
B-ammonium chloride gives off when it is cooled down to the transition 
point, transforms isothermically to the «-form, and this then falls 
to the calorimeter temperature and goes into solution isothermically. 
