180 Proceedings of the Royal Society of Edinburgh. [Sess. 
is, therefore, no longer in a region of varying pressures ( Proc ., p. 436). 
The bath liquids, the mode of operation, and the corrections for submerged 
depth and capillary ascension are the same as before. Most of the 
advantages of this method of determining vapour pressures have already 
been enumerated (l.c. and Proc., 434). The chief of these is that the vapour 
pressures of solids and of substances which interact with mercury may 
be determined, since any suitable material may be used as confining 
fluid. The apparatus has the advantage over the static isoteniscope 
{Proc., 523) that here the confining fluid may have a considerable 
vapour pressure of its own, and the substance may be soluble in the 
confining fluid. Even interaction of the substance with the confining 
fluid does not necessarily interfere with the success of the measure- 
ments. 
In the determinations given below, the bath, stirrer, gauge, and platinum 
resistance thermometer used with the static isoteniscope were employed. 
The temperatures, on the thermodynamic scale (S.B.P. assumed 445°), are 
accurate to ±0 01° below 120° (benzene), and to ±0’1° above 120° (NH 4 C1). 
Vapour Pressures of Benzene . — In previous determinations, which are 
numerous, guarantees, either of the purity of the benzene or of the 
exactness of the experimental method, or both, are lacking. Sulphuric 
acid was used as confining fluid, and the values found were : — 
t. 
p. 
t. 
p. 
t. 
p. 
65 ° 
463 
85 ° 
879 
105 ° 
1542 
70 
551 
90 
1018 
110 
1751 
75 
650 
95 
1180 
115 
1983 
80 
757-5 
100 
1348 
120 
2240 
Vapour Pressures of Ammonium Chloride . — In the previous determina- 
tions of Horstmann the temperatures are uncertain ±5°. In those of 
Ramsay and Young great experimental difficulties were encountered, and 
considerable divergences exist amongst the data, so that one may doubt 
.whether the errors of ±1° at the lower temperatures and ±2° at the 
higher were not considerably exceeded. F. M. G. Johnson’s measure- 
ments afford no assurance whatever of accuracy. At 335°, an error 
of ±1°. corresponds to an error of ±15 mm. in the pressure. The 
existing values appear, therefore, to stand in need of confirmation or 
revision. 
In our measurements, the mixture of potassium and sodium nitrates 
