188 Messrs. J. P. Kuerten and W. G. Robson on 



mercury was introduced, and the tube and bulb were then 

 filled with liquid. All air was then expelled by continued 

 boiling under reduced pressure. The liquid was boiled away 

 from the narrow compression-tube until the required amount 

 was left, when a sudden tilting of the tube caused the mercury 

 to fill the vapour-space. The second bulb prevents the liquid, 

 while boiling is going on, from getting into the tubes leading 

 to the air-pump. 



The bulbs were then cut off and the tube introduced into 

 the compression apparatus. In the case of water and ether it 

 was necessary to introduce some water, boil out the air, and 

 then as quickly as possible fill the tube with ether and 

 proceed in the manner described. In the time occupied by 

 the introduction of the ether, there is some chance of a little 

 air dissolving in the water. Precaution was therefore taken 

 after filling with ether to boil away a knyer of some milli- 

 metres of the watery liquid. A stirrer, consisting of a piece 

 of iron wire provided with enamel knobs, was used inside the 

 compression-tube. We experienced a good deal of trouble on 

 account of the action of the water on the glass tube. This 

 action began at a temperature of about 160° C., and at 200° C. 

 gradually made the tube opaque. The action on the enamel 

 knobs was even stronger ; for our later experiments on ether 

 and water we therefore enclosed the iron wire in a little glass 

 tube. Even this precaution did not remove the difficulty, for 

 when the experiments were nearly finished, the water had 

 eaten a hole into this little tube. By prolonged heating of the 

 mixture at 235° C , under a pressure of 100 atmospheres, the 

 watery layer diminished rapidly, and finally disappeared while 

 a thick white coating was formed on the glass. Curiously 

 enough, the vapour-pressures at lower temperatures were not 

 much affected by this action on the glass, except of course 

 after the lower layer had entirely disappeared. Notwith- 

 standing these difficulties, the pressures which we found from 

 independent observations with different fillings, before and 

 after the application of the high temperature, agree very well, 

 and may therefore be considered quite trustworthy. Some 

 results obtained near the critical point, which we have diffi- 

 culty in reconciling with the theory, have been probably 

 occasioned by an impurity produced by the action of the 

 water on the glass These irregularities consist in the ap- 

 pearance of the critical phenomena, not at the real critical 

 point for the upper liquid and the vapour (201° C), but also 

 at temperatures down to 198° C. This may be explained by 

 the presence of a third substance in small quantity. We 

 never noticed this irregularity with pure substances ; but 



