IN THE PRESENCE OF DUST-FREE AIR AND OTHER GASES. 
279 
allowed to rise quite to the top of A. There is very little risk of contamination by 
air at this stage, because the piston is slowly floating upwards, and the water Ailing 
the narrow space round the piston has therefore a comparatively great downward 
velocity. Diffusion of the air upwards through this water is thus prevented. 
Then T^ is closed, and some of the gas to be investigated drawn into B by lowering 
Ph and opening Tg. It Is then driven into A by opening and raising To being 
closed. This pumping process is repeated till the pressure in A is rather in excess of 
what is required. 
The gas which remains in B is driven out through Tg, which is then closed, and T, 
opened for a moment, so that the narrow-bore tube and stop-cock are filled with 
water. It is necessary, of course, for this purpose that Rj should be raised high 
enough to overcome the pressure in A; it is, in fact, now kept permanently fixed at 
such a level throughout all subsequent operations. 
The excess of water always remaining above the piston in A at this stage is now 
allowed to escape by applying sufficient pressure below to drive the piston up a little. 
The pressure is applied by pumping air into C, by means of the mercury compressing 
pump shown at the extreme left of the diagram. The plug G is then pushed up and 
the piston thus allowed to fly back into the constriction. 
To prevent contamination of the gas, the glass taps T\ T^ Tg are lubricated with 
water only. The only one which requires to be used after the above operations are 
completed is Tj. Since this is filled with water under considerable pressure, there is 
no danger of air gaining access through it. 
Method of producing expansion of any desired amount. 
With this apparatus, direct volume measurements were not made, but the relative 
volume change was deduced from measurements of the initial j^ressure, and the 
pressure exerted by the saturated gas at the same temperature when occupying the 
increased volume. The final volume Uo, being that of A when the jDiston is at the 
bottom, is always the same ; and the corresponding pressure p^ at the temperature 
of the room shows only comparatively small changes resulting from variations 
temperature, and from the solution of the gas by the small quantity of water which 
is run through the apparatus. 
This final pressure p^ was measured in the following way. A mark was made in 
the wall of A approximately on a level with the top of the j)iston, when this was 
in its lowest position. 
A telescope was then fixed in a clamp about one metre oft’, and its height adjusted 
till the mark appeared to coincide with the top of the piston. By means of the 
compressing pump air was driven into C till the pressiu'e was sufficient to drive up 
the piston a little. Then S 3 being left open and S^ closed, the tap Sg, regulating the 
flow of mercury in the pump, was closed and the reservoir R 3 lowered and fixed in 
