Returning to the voltaic cell, we may imagine that the atoms of the zinc which are immersed in the acid are trying to turn themselves into ions, so that they can travel through the solution. In order to do this each atom parts with two electrons, and these electrons try to attach themselves to the next atom. This atom however already has two electrons, and so in order to accept the newcomers it must pass on its own two. In this way electrons are passed on from atom to atom of the zinc, then along the connecting wire, and so to the copper plate. The atoms of zinc which have lost their electrons thus become ions, with power of movement. They leave the zinc plate immediately, and so the plate wastes away or dissolves. So we get a constant stream of electrons travelling along the wire connecting the two plates, and this constitutes an electric current. The electron theory gives us also a clear conception of magnetism. An electric current flowing along a wire produces magnetic effects; that is, it sets up a field of magnetic force. Such a current is a stream of electrons, and therefore we conclude that a magnetic field is produced by electrons in motion. This being so, we are led to suppose that there must be a stream of electrons in a steel magnet, and this stream must be constant because the magnetic field of such a magnet is permanent. The electron stream in a permanent magnet however is not quite the same as the electron stream in a wire conveying a current. We have stated that the electrons constituting an atom move in definite orbits, so that we may picture them travelling round the core of the atom as the planets travel round the Sun. This movement is continuous in every atom of every substance. Apparently we have here the necessary conditions for the production of a magnetic field, that is, a constant stream of electrons; but one important thing is still lacking. In an unmagnetized piece of steel the atoms are not arranged symmetrically, so that the orbits of their electrons lie some in one plane and some in another. Consequently, although the electron stream of each atom undoubtedly produces an infinitesimally small magnetic field, no magnetic effect that we can detect is produced, because the different streams are not working in unison and adding together their forces. In fact they are upsetting and neutralizing each other’s efforts. By stroking the piece of steel with a magnet, or by surrounding it by a coil of wire conveying a current, the atoms are turned so that their electron orbits all lie in the same plane. The electron streams now all work in unison, their magnetic effects are added together, and we get a strong magnetic field as the result of their combined efforts. Any piece of steel or iron may be regarded as a potential magnet, requiring only a rearrangement of its atoms in order to become an active magnet. In Chapter VI. it was stated that other substances besides iron and