[Illustration: FIG. 16.—Vacuum tube with cathode rays and a shadow-producing cross. P and N, conducting wires for the electric current; _a_, cathode; _b_, anode and shadow-producer; _c_, _d_, the shadow.] [Illustration: FIG. 17.—Vacuum tube, where a bundle of cathode rays are deviated by electric forces. _A_, anode; _K_, cathode.] It has previously been said that air is an insulator for electricity, a statement which is, in general, true; however, as has also been said, electric sparks and arcs can pass through air. Moreover, it has been discovered that exhausted air is a very good conductor, so that a strong current can pass between two metal electrodes in a glass tube where the air is exhausted, if the electrodes are connected to an outer conductor by metal wires fused into the glass. In these vacuum tubes there are produced remarkable light effects, at first inexplicable. When the air is very much exhausted, to a hundred thousandth of the atmospheric pressure or less, strong electric forces (large difference of potential between the electrodes) are needed to produce an electric discharge. Such a discharge assumes an entirely new character; in the interior of the glass tube there is hardly any light to be seen, but the glass wall opposite the negative electrode (the cathode) glows with a greenish tint (fluorescence). If a small metal plate is put in the tube between the cathode and the glass wall, a shadow is cast on the wall, just as if light were produced by rays, emitted from the cathode at right angles to its surface (cf. Fig. 16). The English physicist, Crookes, was one of the first to study these cathode rays. He assumed that they are not ether waves like the light rays, but that they consist of particles which are hurled from the cathode with great velocity in straight lines; they light the wall by their collisions with it. There was soon no doubt as to the correctness of Crookes’ theory. The cathode rays are evidently particles of negative electricity, which by repulsions are driven from the cathode (the negative electrode). A metal plate bombarded by the rays becomes charged negatively. Let us suppose that we have a small bundle of cathode rays, obtained by passing the rays from the cathode _K_ (cf. Fig. 17) through two narrow openings _S₁_ and _S_. It can then be shown that the bundle of rays is deviated not only by electric forces, but also by magnetic action from a magnet which is held near the glass. In the figure there is shown a deviation of the kind mentioned, caused by making the plates at _B_ and _C_ respectively positive and negative; since _B_ attracts the negative particles and _C_ repels them, the light spot produced by the bundle of rays is moved from _M_ to _M_₁. The magnetic deviation is in agreement with Ørsted’s rules for the reciprocal actions between currents and magnets, if we consider the bundle of rays produced by moving electric particles as an electric current.