A much more sensitive arrangement can be made by taking a small box of wood through the bottom of which pass two nickel wires which are parallel to one another, but not in contact. In this box is placed a small quantity of very finely powdered metallic nickel or nickel filings, and if the quantity of these filings is properly adjusted, it is possible to make an arrangement which possesses the property that there is no conductivity between the two nickel wires under ordinary circumstances, but that they become conductively connected to one another the moment an oscillatory electric spark is made in the neighbourhood. We shall speak of this contrivance as an _electric wave indicator_, and we shall employ it in subsequent experiments to enable us to detect the presence of an electric wave. We must then return for a moment to the consideration of the production of electrical oscillations in circuits of various kinds. I trust it has been made plain to you that if two metallic surfaces, separated by a non-conductor such as air or glass, are acted upon by an electromotive force, the non-conductor becomes electrically strained. Another way of stating this is to say that a positive charge of electricity exists on one metal surface, and a negative charge on the other. The only objection which can be raised to expressing the facts in this manner is that it fastens attention rather upon the conductors than upon the insulator, which is the real storehouse of the energy. If these two metal surfaces are then connected together by a conductor of low resistance, the charges disappear by a series of oscillations, and the result is an electric current in the conducting circuit connecting the plates, which rushes backwards and forwards in the circuit, but gradually diminishes in strength until it completely dies away. You may picture to yourselves the electrical effect as analogous to the following experiment with two air-vessels: Supposing we have two strong steel bottles, into one of which we compress a quantity of air, and in the other we make a vacuum by pumping out nearly all the air. These vessels would correspond with two conductors, one charged with positive electricity and the other with negative. Imagine these vessels connected by a wide pipe in which is placed a tap or valve, which can be opened suddenly so as to permit the air to rush over from the full vessel to the empty one. If this is done, it is a matter of experience that the equality of pressure between the two vessels is not at once established, but in virtue of the inertia quality of the air, it only takes place after a series of oscillations of air in the pipe. In rushing over from the full vessel to the empty one the air, so to speak, overshoots the mark, and the state of the vessels as regards air-pressure is exactly interchanged. The air then rushes back again, and it is only after a series of to-and-fro movements of the air in the