of Demonstrating the fonisation of Arr. 685. 
communication with the vacuum chamber is the same as in 
the smaller apparatus described in previous papers, but of 
brass instead of glass ; its mode of working will be under- 
stood from the figure. An indiarubber stopper held tightly 
by the pressure of the atmosphere against the end of-the 
tube leading to the vacuum-chamber V can be suddenly 
pulled away by a spring released by the trigger arrangement 
shown. In this way the sudden motion of the piston, and 
the consequent expansion of the air in the cloud chamber, 
are effected. 
The final volume of the air after expansion is always 
the same, for the piston is then held against the indiarubber 
at the bottom, owing to the low pressure below. To vary the 
expansion the initial volume has to be varied. The air before 
expansion is always at very nearly atmospheric pressure— 
really at a pressure less than that of the atmosphere by the 
pressure required to balance the weight of the partially im- 
mersed piston, the air below the piston being always before 
expansion at atmospheric pressure. The amount of any 
expansion is determined by adjusting the pressure after open- 
ing communication with the gauge M, before allowing the 
piston to rise after the previous expansion, a sufficiently long 
interval being allowed to elapse for the temperature to return 
to that of the surroundings; the constancy of the pressure 
serves as a test of this condition being fulfilled. The pressure 
thus determined will be the same as the pressure after the 
next expansion has taken place and the temperature has 
again become steady. The ratio of the air-pressure before 
expansion (7. ¢., of the whole pressure in the cloud chamber less 
the saturation pressure of the aqueous vapour) to that after 
expansion will then be equal to v,/v,, the ratio of the final to 
the initial volume. To adjust the final pressure to any desired 
value, air can be admitted from the atmosphere through the 
cotton-wool filter W, or removed by opening communication 
with the water-pump through the tube P. 
The drops resulting from expansion are illuminated by a 
narrow beam of light converging to a focus at the centre of 
the cloud chamber ; the source was in most cases an are or 
lime-light ; the effects were, however, quite easily observed 
with the light from an ordinary luminous flame. It was 
found convenient to coat the outside of the glass with black 
enamel over half the circumference, leaving, however, a 
vertical strip about one cm. wide in the middle for the light 
to enter; the glass immediately opposite this slit was also 
blackened over a width of a few cms. Any drops produced 
were then well seen on looking towards the centre of 
