I have here a wooden frame across which are strained some wires about a quarter of an inch apart (see Fig. 79). If we hold this frame or grid in front of the radiator so that the direction of these wires is at right angles to the direction of the radiator rods which carry the balls, we find that the grid is quite transparent to the electric radiation, but if we turn the grid round so that the wires of the grid are parallel to the radiator rods, we find at once that the grid becomes perfectly opaque. The same experiment can be prettily shown by means of a paper of pins. Here are some large carpet pins arranged in rows in paper, and if I hold this paper of pins in between the radiator and receiver with the pins parallel to the radiator, it is perfectly opaque to the electric ray, but if I turn it so that the pins are at right angles, it is quite transparent. The same experiment succeeds with a paper of ordinary pins, but not so well with a paper of midget pins. [Illustration: FIG. 79.] The explanation of this action of a grid is as follows: You have already seen that an alternating current in one electric circuit can produce another alternating current in a secondary circuit placed parallel with the first. It is not difficult to show, either experimentally or from theory, that when the primary current is an electrical oscillation—that is, a very rapid alternating current—the current in the secondary circuit is also an electrical oscillation of the same frequency or rapidity, but that the currents in the two circuits, primary and secondary, are always moving in opposite directions at the same moment. Accordingly, if we hold a grid in front of the radiator, the wires of the grid have what are called _induced oscillations_ set up in them, and these induced oscillations themselves create electric radiation. Accordingly, it is clear that if a grid of this kind is held near to a radiator with the wires of the grid parallel to the radiator rods, we have two sets of radiations produced which, at any point on the side of the grid furthest from the radiator rods, must neutralize one another, and therefore destroy each other’s effect. Hence it is possible to cause the electric radiations proceeding from two electric circuits parallel with each other to destroy one another at a distant point; and we may, therefore, make use of the same arguments as in the case of a similar experiment with light to prove that this electric radiation must be a wave-motion.