electric spark] have disappeared, as you see: their place is vacant, and
fresh gas has gone in. Water has been formed from them; and if we repeat
our operation [repeating the last experiment], I shall have another
vacancy, as you will see by the water rising. I always have an empty
vessel after the explosion, because the vapour or gas into which that
water has been resolved by the battery explodes under the influence of the
spark, and changes into water; and by-and-by you will see in this upper
vessel some drops of water trickling down the sides and collecting at the
bottom.

We are here dealing with water entirely, without reference to the
atmosphere. The water of the candle had the atmosphere helping to produce
it; but in this way it can be produced independently of the air. Water,
therefore, ought to contain that other substance which the candle takes
from the air, and which, combining with the hydrogen, produces water.

Just now you saw that one end of this battery took hold of the copper,
extracting it from the vessel which contained the blue solution. It was
effected by this wire; and surely we may say, if the battery has such
power with a metallic solution which we made and unmade, may we not find
that it is possible to split asunder the component parts of the water, and
put them into this place and that place? Suppose I take the poles--the
metallic ends of this battery--and see what will happen with the water in
this apparatus (fig. 20), where we have separated the two ends far apart.

[Illustration: Fig. 20.]

I place one here (at A), and the other there (at B), and I have little
shelves with holes which I can put upon each pole, and so arrange them
that whatever escapes from the two ends of the battery will appear as