( 917 ) 
the Cailletet tube as above. If we wish to diminish the quantity of 
carbonic acid, we leave A, after it has been carefully opened, open 
till the vessel D has filled with carbonic acid under the desired 
pressure. We then close the two cocks, exhaust D, and deliver the 
carbonic acid above B again into the Cailletet tube. 
The further method of observation does not call for a detailed 
discussion. We shall only remark that the temperature was obtained 
by an ordinary waterbath with toluol regulator, and the pressure 
measured with a spring manometer of the model constructed by 
SCHAFFER & BUDENBERG for accurate measurements for the Amsterdam 
Laboratory. As we intend shortly to publish a separate communication 
on the experiences with these instruments, it may suffice here to state 
that they are excellently adapted to reach the required accuracy of 
no more than 0.1 per cent, provided they are used in connection 
with a pressure-balance which enables us to test them from time to 
time with very little trouble. 
As a first mixture we investigated a mixture of the concentration 
xv = 0.042. It appeared already at the outset that the system urethane- 
carbonic acid fulfils the condition that at equal temperature the three- 
phase-pressure lies below the pressure of saturation of carbonic acid. 
For a temperature of 26°.1 we find a three-phase-pressure of 65.15 Kg. 
per cM*. = 62.95 atms., while Krrsom gives 64.4 atms. as pressure 
of saturation for 25°.55. At 32°.05 we found 74.35 Ke. per cm? 
= 71.85 atms. Kersom gives 72.93 for 30°.98 (critical point). The 
purity of the prepared mixture. appeared from the constancy of the 
three-phase-pressure which from the beginning to the end of the 
existence of the three phases only increased about 0.25 Ke. per cm’. 
Also for the other mixtures to be discussed presently this condition 
was fulfilled. Perhaps even a better warrant of the absence of 
impurities (air) was furnished by the fact that the different values 
for the three-phase-pressure obtained for different mixtures at the 
same temperature, harmonized perfectly. Thus we found: 
Kg. per cM.’ 
at 31°.9 for mixture I 74.10 mixture III 74.20 IV 74.20°) 
34 .9 HI 79.15 VI 79.10 
35 .9 Fer e0:S0 [IT 80.85 VI 80.85 
ar el I+ 82.85 III 82.80 
') Cf. the table containing all the three-phase-pressures noted down by us. They 
always give the mean value, i. e. the value for which none of the three phases 
has almost disappeared. In the text we have taken those values for a comparison 
which could be reduced to the same temperature with the slightest inter-, 
or extrapolation, in which 0.15 increase per 0°.1 is assumed. But also for nearly 
