216 
Proceedings of the Royal Society of Edinburgh. [Sess. 
It was therefore necessary that we should have some form of apparatus 
that could be easily, accurately, and quickly manipulated, and that would 
give definite degrees of expansion to any desired amount. Fig. 1 shows the 
general arrangement of the different parts of the apparatus used, though not 
in their relative positions. T is the test-flask in which the cloudy con- 
densation takes place, and which is illuminated by the incandescent gaslight 
L in the dark lantern D, the light being concentrated in T by means of a 
globe of water C. Three tubes pass through the stopper S of the test- 
flask T. One of these tubes connects the flask with the pump P, and is 
provided with a stopcock J. The pump has a capacity of 152 c.c. The 
middle tube connects with what we will call the expander E, while the 
third tube, which is provided with a stopcock K, goes to the source from 
which we wish to draw, by means of the pump P, the air to be tested in T. 
F is a cotton-wool filter which is introduced when required ; N is a lens 
for observing the condensation in T ; M is a box, blackened inside, sur- 
rounding the test-flask to cut off all light that might interfere with the 
observation of what is taking place in the flask. 
The expander, which is the new part of the apparatus, is shown in 
horizontal section in fig. 2. It is practically a pump working in water. 
The barrel B of the pump is fitted with a piston H of two cupped 
leathers, and is enclosed in the small tank G filled with water, the top end 
of the pump opening into the tank. By this arrangement no air can leak 
past the piston and interfere with the tests. In order to get definite 
