If we had two large water tanks, one of which could be emptied only by allowing the bottom to fall completely out, and the other by means of a narrow pipe, it is easy to see which would be the more useful to us as a source of water supply. If both tanks were filled, then from the first we could get only a sudden uncontrollable rush of water, but from the other we could get a steady stream extending over a long period, and easily controlled. The Leyden jar stores electricity, but in yielding up its store it acts like the first tank, giving a sudden discharge in the form of a bright spark. We cannot control the discharge, and therefore we cannot make it do useful work for us. For practical purposes we require a storing arrangement that will act like the second tank, giving us a steady current of electricity for a long period, and this we have in the accumulator or storage cell. A current of electricity has the power of decomposing certain liquids. If we pass a current through water, the water is split up into its two constituent gases, hydrogen and oxygen, and this may be shown by the apparatus seen in Fig. 12. It consists of a glass vessel with two strips of platinum to which the current is led. The vessel contains water to which has been added a little sulphuric acid to increase its conducting power, and over the strips are inverted two test-tubes filled with the acidulated water. The platinum strips, which are called _electrodes_, are connected to a battery of Daniell cells. When the current passes, the water is decomposed, and oxygen collects at the electrode connected to the positive terminal of the battery, and hydrogen at the other electrode. The two gases rise up into the test-tubes and displace the water in them, and the whole process is called the electrolysis of water. If now we disconnect the battery and join the two electrodes by a wire, we find that a current flows from the apparatus as from a voltaic cell, but in the opposite direction from the original battery current. [Illustration: FIG. 12.--Diagram showing Electrolysis of Water.] It will be remembered that one of the troubles with a simple voltaic cell was polarization, caused by the accumulation of hydrogen; and that this weakened the current by setting up an opposing electro-motive force tending to produce another current in the opposite direction. In the present case a similar opposing or back electro-motive force is produced, and as soon as the battery current is stopped and the electrodes are connected, we get a current in the reverse direction, and this current continues to flow until the two gases have recombined, and the electrodes have regained their original condition. Consequently we can see that in order to electrolyze water, our battery must have an electro-motive force greater than that set up in opposition to it, and at least two Daniell cells are required.