_Secondly_, a supposition of this kind has no permanent value; it is rather a feeler, by which we trace out our way in the darkness, than any actual vision itself. In default of an actual realization of what the electrical relations are we can treat them by means of a supposition. But we must be ready at any moment to give up the supposition if it does not harmonize with the facts. And in the first case does the idea of a real image hold good about the simplest possible actions? If we push our fist towards a glass the image is that of a fist moving in the opposite direction. Now, suppose a pressure exerted on a wall, as, for instance, a hard stone hitting it. The wall undergoes a displacement, but not as a whole—only that part of it where the stone hits. And this displacement is followed by the image displacement, for the wall in the part where it has been hit and pressed back moves forward, and by its reaction throws the stone off. Every case of action and reaction is a case of a motion and its image motion. If a bullet strikes the wall and goes with such velocity that it lodges in it, then the motion of the ball and the image motion of the wall destroy one another, and the result is a shattering of the wall in the path of the bullet. Now in the case of a simple displacement of this kind there is a rule by which we can form the image displacement. Take a point on the wall, and about this point as a centre turn the displacement half way round, so that it does not come to be itself again, but is opposite to itself. By this turning, the displacement becomes the image of itself; a movement into the wall becomes a movement out from the wall; and these follow one another if the wall is not injured. It should be noticed that the displacement is moved round this point, using a direction which is _not_ in the displacement itself. The displacement goes straight into the wall. The turning motion, which we suppose, needs another direction than this. Now suppose, instead of a simple displacement like this, we take a displacement involving two directions, as in the case of a wave disturbance—it will be found that the conditions are just the same. If a wave movement falls on a medium which it does not destroy or move as a whole, the displacement calls up its image displacement. And the image displacement can be found, as before, by twisting the displacement round so as to become opposite to itself—by twisting it half-way round. But in this case, too, a direction must be used which is not used in the displacement itself. [Illustration: Diagram II.] Let us look at the wave disturbance more closely. The horizontal central line in Diagram II. will represent the positions which a number of particles occupy when at rest. That is, let us suppose there to be a number of particles lying in a series forming this line.