90 Scientific Proceedings, Royal Dublin Society. 
Andrews himself, although he observed a considerable rise of 
pressure during the condensation of carbon dioxide, attributed this 
rise to the residual air which he was unable to remove from the gas, 
and he estimated the amount of air present at =4, to 7955 of the 
total volume. Andrews clearly stated his belief that if the carbon 
dioxide had been quite pure, the pressure would have remained | 
constant during condensation ; and he discussed the effect which 
would be produced by a small amount of permanent gas. Amagat 
(Mémoires sur l’élasticité et la dilatabilité des fluides, Ann. Chim. 
Phys., 1893), in describing the PV — P diagram, constructed from 
his observations with carbon dioxide, speaks of that part of the 
isothermals which corresponds to liquefaction as straight; and 
Kuenen (Phil. Mag., 1897, vol. xliv., 179), who observed a rise of 
pressure of 0°44 per cent. during the complete process of liquefac- 
tion of carbon dioxide at 14°-95, and of only 0°21 per cent. at 25°-85, 
attributed the slight deviation of his values from those of Amagat 
to the presence of a very small amount of residual air. 
Again, in the last of a series of papers by Ramsay and myself 
on the thermal properties of liquids (Phil. Trans., 1892, vol. 1834, 
p- 107), water is described as differing'from the other liquids (ethyl, 
ether and methyl, ethyl and propyl alcohols), inasmuch as its vapour- 
pressure was found to depend to some extent on the amount of 
substance present (in a given volume).’ 
From this it is clear that there are other experimenters who 
consider that for normal substances the evidence is in favour of the 
formula p = c at constant temperature. 
In addition to ether, the three alcohols, and water, I have— 
either alone, or with the assistance of Miss EH. C. Fortey, or Mr. 
G. L. Thomas—determined the vapour-pressure of twenty-seven 
carefully purified liquids through a wide range of temperature. 
From the ordinary boiling-point to the critical point, the vapour- 
pressures were determined in an apparatus similar in principle to 
that employed by Andrews, but differing from it in detail. The 
liquid in the experimental tube was heated by the vapour of a pure 
liquid boiling under known (usually reduced) pressure, and the 
vapour-pressure was read on one or other of two air-manometers, 
which were kept at a nearly constant temperature by a current of 
1 The rise of pressure during the condensation of steam was attributed to the 
hygroscopic nature of glass. 
