THE PROPERTIES OF WATER AND OF STEAM. 
113 
III. The Densities of the Saturated and Unsaturated Vapour of Water at Different 
Temperatures and Pressures. 
The method pursued was (1) to find the weight of the small quantity of water in 
the tube by ascertaining the values of the products of pressure and volume at large 
volumes and at high temperatures. If these products, for slight decrease of volume, 
and at any one temperature, were sensibly constant, and if the products at different 
temperatures bore to one another the same ratio as that between the absolute 
temperatures, it was assumed that the vapour-density of the steam was normal; if 
hydrogen be taken as unity, the density is nine. The weight could then be calculated 
by the equation given in the Appendix, p. 122. (2) The volume of the gas was then 
diminished by regular decrements at temperatures rising by 10° at a time, and the 
pressure corresponding to each volume was read. From these data the volume of 
1 grm. in cubic centimetres could be calculated. When the volume was sufficiently 
diminished condensation ensued. 
Two distinct sets of experiments were made : one with a large tube at low 
temperatures, and one with a strong pressure tube at high temperatures. The actual 
experimental data for high temperatures (120° to 270°) are given on pp. 124-127 
of the Appendix ; those at low temperatures on pp. 128-130. Reasons are there 
adduced to show that when the pressure approaches the vapour-pressure condensation 
on the glass occurs. This causes the pressure to be lower than it should be for a 
given volume, and renders the results very difficult to interpret. Such adhesion 
appears to be slight at large volumes where the vapour is unsaturated; but, as it 
occurs gradually, the effect is that on decreasing volume some water is left below the 
mercury, and as there is no clue to the amount thus trapped the weight is diminished 
by an unknown amount. But it is reckoned as undiminished, and hence in calcu¬ 
lating the volume of 1 grm. from the actual measurements the volume ascribed is 
too small, and too small by an unknown quantity. At the same time the pressure is 
lowered by the adhesion to the glass, and so the curve is distorted as it approaches 
the pressure of the saturated vapour. 
Under the circumstances it appeared best to follow the experimental results as 
closely as possible. However interesting from a theoretical point of view the 
absolute expansion of water-gas may be, in practice it is always in contact with a 
surface ; and although it is probable that the material, as well as the extent of the 
surface, would influence the amount of adhesion, yet an indication of the behaviour of 
steam in contact with glass cannot fail to be of use in considering the practical case 
of steam in contact with iron. Moreover, the error is to a great extent eliminated by 
making use of the volumes of 1 grm. of saturated steam calculated from Regnault’s 
heats of vaporization. 
MDCCCXCII.—A, 
