produced by the Explosion of Hydrogen and Oxygen. 473 
the wire was determined by an observation of the electro- 
meter in the way fully described below. Finally the 
pinchcock E was opened and the pressure read on the 
manometer. From these observations the fall of pressure 
caused by the explosion is easily deduced, when the ratio of 
the volume, V, of the apparatus of fig. 2, in which the 
explosion takes place, to the volume, v, of the apparatus in 
fig. 1 between the stopcocks C, D, and E is known. This 
ratio was determined as 1°4 in the apparatus used. 
It was found in the preliminary experiments that the effect 
of the explosions was to break down the insulations, owing, 
no doubt, to water-vapour being thrown on to the insulators. 
It was to obviate this that the bottles K, K’ were designed, 
the phosphorus pentoxide in which served the double purpose 
of protecting the insulators and of absorbing the water- 
vapour. P, P! were further protected by small circular 
ebonite shields placed on the wire, as shown in the figure, 
close to, but not touching, the glass tubes. These shields, 
in addition to intercepting the water-vapour, screened the 
insulators from fragments of phosphorus pentoxide, which 
were in some cases blown into the cylinder by the explosion. 
The preliminary experiments had led us to believe that 
the electricity produced was a small charge, which could be 
best measured by the deflexions of an electrometer connected 
with a condenser of small capacity ; but with the improved 
apparatus it was found that a condenser of the order of a 
microfarad was required. The following was the method 
adopted for measuring the quantity of electricity generated. 
Fig. 3. 
EARTH 
ee 
EARTH 
The quadrants in metallic communication with the wire W 
were connected with one terminal of a condenser C, the other 
terminal of which was connected with the slider of a high- 
resistance potentiometer, as shown in fig. 3. The electrodes 
