GASES BY THE ACTION OF RONTGEN RAYS AND OTHER AGENTS. 
407 
P 2 , it will be noticed, is the pressure, not at the nioinent when the expansion is com¬ 
pleted, but after the temperature has risen to its original value. 
As the initial pressure. Pi, in these experiments is always approximately equal to 
the atmospheric pressure, it is sufficient for many purposes to take Pi — P 2 , or p, as a 
measure of the expansion without further reduction. 
To make, for example, v<ilv\ = 1'25, Pi •— P 2 must be equal to 15 centims. of mercury 
if the barometric pressure = 760 millims. and the temperature = 15° C. ; while, as 
long as the atmospheric pressure lies between 740 and 780 millims., and the tempera¬ 
ture lies between 10° and 25° C., Pi — P 2 for the same expansion will always lie 
between 14‘4 and 15'4 centims. 
The gas with which the apparatus is to be charged is introduced through the stop¬ 
cock, T 3 , a side tube on the cloud chamber A being connected to a water air-pump, so 
that a stream of the gas at low pressure may pass through the aj)paratus. (At this 
stage a sufficiently low pressure must, of course, be maintained below the plunger, P, 
to prevent it rising out of the mercury in B.) The side tube is afterwards sealed off, 
and when sufficient gas has been generated to bring the pressure nearly up to that ol 
the atmosphere, T 3 is closed. 
The stopcocks, T 2 , T 3 , were lubricated witli water only and protected by mercury 
cups. The mercury in B, as well as that covering the indiaru])ljer stopper over which 
A is slipped, is prevented from coming in contact with the gas in the apparatus by a 
layer of distilled water. 
In most of the experiments in which large expansions were required tlie expansion 
apparatus had the form described above. Many of the experiments with air, however, 
were performed with an -expansion apparatus resembling that described in the ‘ Camb. 
Phil. Soc. Proc.’ {loo. cit.). In it the plunger works in water instead of mercury, and 
is made to fit the outer tube like a piston, instead of working on an internal guide 
tube. The only advantage of the form with mercury is the absence of any risk of 
contamination of the gas in the apparatus by air which, when the plunger works in 
water, may gain entrance by solution and diftusion through the latter. The mercury 
apparatus is also suited for experiments with other liquids than water; such experi¬ 
ments, have, however, not yet been made. 
Both these forms of apparatus give results almost identical with those obtained by 
means of the apparatus used in the earlier experiments (‘Phil. Trans.,’ loc. cit.). No 
considerable error appears to be produced by the yielding of the indiarubber when 
struck by the plunger, or by the momentum acquired by the air in the narrower part 
of the tube. 
It will be convenient to mention here the methods used in preparing the gases 
required for the experiments. Oxygen was prepared by heating potassium per¬ 
manganate. Hydrogen was obtained from palladium, which had previously been 
charged with the gas, obtained from the purest zinc and dilute sulphuric acid, the gas 
being passed through potassium permanganate solution before reaching the palladium. 
