74 SIR WILLIAM RAMSAY 



at the bottom into a jet impinging on the thermometer just 

 above the wool. We found that this worked very well, and that 

 in the case of water, benzene, etc., the liquid on the cotton- wool 

 could be frozen, so that both boiling points and volatilising points 

 could be determined with the same apparatus. [Phil. Trans. 

 175, 37 (1884).] 



In the case of water we used an apparatus with two vertical 

 tubes provided with thermometers, etc., so that we could have 

 ice on one thermometer and supercooled water on the other, 

 the pressure being necessarily the same in both cases. We were 

 thus enabled to verify Prof. James Thomson's theory of the 

 vapour pressures of solid and liquid in the case of water, ben- 

 zene, acetic acid and camphor. But to make the proof complete, 

 it was advisable to ascertain definitely whether the statical 

 (barometer tube) and dynamical (boiling or volatilising point) 

 methods gave the same results. Accordingly determinations of 

 vapour pressure were made with a barometer tube of a form 

 specially adapted for the complete removal of air either dissolved 

 in the liquid or adhering to the walls of the barometer tube. 

 [Phil Trans. 175, 461 (1884), also Phil. Mag. 23, 61 (1887).] 



The barometer tube was heated by the vapour of a pure liquid 

 boiling under known reduced pressures. From these pressures 

 the temperatures were ascertained from tables of vapour pressure 

 previously compiled. 



In order to have at our disposal a wide range of temperature, 

 we required to know the vapour pressures of a series of stable 

 liquids, easily obtainable in a pure state. The substances we 

 adopted were carbon bisulphide, ethyl alcohol, chlorobenzene, 

 bromobenzene, aniline, methyl salicylate, bromonaphthalene 

 and mercury. [Trans. Chem. Soc. 47, 640 (1885).] Regnault 

 had already determined the vapour pressures of carbon 

 bisulphide, ethyl alcohol and mercury, and we accepted his 

 values, though it was found subsequently that considerable 

 corrections were required in the case of mercury. We had to 



