ELECTROMAGNETIC UNIT OE ELECTRICITY TO THE ELECTROSTATIC UNIT. 601 
of introducing the water a tube with a funnel top provided with a tap. The funnel 
could always be kept partially full of water by means of the tap, so that no air was 
introduced. The cylinders were slightly tilted so that the exit tube was at the 
highest place. As the last few grammes were poured in the air was gradually swept 
along by the advancing water and driven up the exit tube. The small bubble which 
sometimes remained under the exit tube was easily removed by agitating it by a fine 
wire introduced through the exit tube. 
The amount of water remaining in the tap and the amount which rose up the exit 
tube were ascertained afterwards by detaching the tap and tube, and finding the 
weight of water required to fill them to the same extent as when the cylinders were 
filled, 
While the water was still about 3 cm. from the glass the air was exhausted from 
the cylinders by a water pump; the pressure of the air was reduced to about 50 mm. 
of mercury. This had the effect of removing a good many bubbles, and we may hope 
that the experiment was free from error in this respect. The wax employed for 
fixing the top on to the cylinders stood this difference of pressure perfectly. A 
drying tube was placed between the cylinders and the tube leading to the pump in 
order to catch any v/ater which might be carried off in the pumping. The increase in 
weight of the drying tube was found to be not more than ’02 or '03 gramme, and this 
has been allowed for. The weighings have also been corrected to a vacuum. 
The annexed sketch shows a section of the cylinders by a plane through their axis, 
and through the tap and exit tube. 
A is tbe flat metal plate. 
B the glass plate. 
MDCCCXC.-A. 
4 H 
0 the funnel and tap. 
D the exit tube. 
