On Rays of Electric Force1

Heinrich Hertz

Immediately after I had succeeded in proving that the action of an electric oscillation propagates in space in the form of a wave, I attempted to strengthen this action and make it perceptible at greater distances by placing the excited conductor in the focal line of a large parabolic concave mirror. These experiments yielded no result, and it became clear to me that their failure was due to the unsuitable relation between the length of the wave used, which was 4–5 m, and the dimensions of the mirror that I employed.

Recently I noticed that the experiments I had described earlier could easily be carried out with oscillations occurring approximately 10 times faster, i.e., with waves 10 times shorter than those used before. I therefore again turned to the use of a concave mirror and achieved considerably greater success than I had expected. I was able to obtain distinct rays of electric force and, with their aid, to perform all the elementary experiments that are carried out with rays of light and heat. These experiments are described below.

Apparatus

The method of obtaining short waves was quite similar to the method by which long waves were excited. The primary conductor used in the experiments was arranged as follows: imagine (see Figs. 1 and 2 and their explanation at the end of the article) a cylindrical copper body 3 cm in diameter and 26 cm long. In the middle it is cut and provided with a spark gap, the poles of which are formed by two spherical surfaces with radii—

som in 2 cm. The length of the conductor was approximately equal to half the wavelength corresponding to the oscillation arising in the straight conductor. From this alone it was already possible to draw an approximate conclusion about the magnitude of the period of the oscillations.

It is necessary that the poles of the spark gap be frequently polished and protected during the experiment from secondary discharges, in which oscillations cannot arise. The appearance and sound of the sparks readily make it possible to judge whether the spark gap is in a satisfactory condition. The discharge was brought to both halves of the conductor by means of two wires covered with gutta-percha; these wires were soldered on both sides near the spark gap1. As the inductor, instead of the large Ruhmkorff apparatus, I used a small Keiser and Schmidt apparatus, which made it possible to obtain between the points a spark 4.5 cm long. The apparatus was supplied by three accumulators, and between the balls of the primary conductor sparks 1–2 cm long could be obtained. During the experiments the length of the spark gap was 3 mm.

For detecting the electric force in space, use was made of small sparks produced by it in a secondary conductor. Sometimes, as the secondary conductor, a circular conductor was used, having its own frequency of oscillation approximately equal to the frequency of the primary conductor. The radius of the circle was 7.5 cm; the circle was made of copper wire 1 mm thick. One end of the wire ended in a brass ball several millimeters in diameter; the other end was sharpened and could be set at a very small distance from the brass ball by means of a micrometer screw insulated from the wire. Of course, here only sparks a few hundredths of a millimeter long were obtained; with some practice it is possible to estimate the intensity of the process not so much by the length of the sparks as by their brightness.

The circular conductor is unsuitable for placing in the focal line of the mirror. Therefore, the greater part of the work was carried out with another secondary conductor arranged as follows: two straight wires, each 50 cm long and 5 mm in diameter, were set in one straight line, their ends facing one another being at a distance of 5 cm. From these ends ran two wires (15 cm long, 1 mm in diameter), parallel to one another and perpendicular to the first wires; they were connected to the spark gap, arranged in the same way as in the case of the circular conductor.

It would have been simpler to place the spark gap directly between the straight wires, but in that case it would have been impossible to observe it in the focal line of the mirror, without—

curve covering the aperture of the mirror with its body. For this reason the construction described above was chosen.

Ray exciter

If the primary conductor is placed in a large free space, then, with the aid of a circular conductor, one can observe near it all those phenomena which were observed for slower oscillations and were described earlier1. The greatest distance at which sparks were still observed in the secondary conductor was 1.5 m, and, when the primary spark gap was in very good condition, even 2 m. It was possible to intensify the effect in one direction by placing, on the opposite side of the primary radiator at a suitable distance, a flat conducting wall parallel to the radiator. Namely, if the distance was very small, or exceeded 30 cm by only a little, then the wall had only a harmful influence: it produced a considerable intensification at a distance of 8–15 cm, a less considerable intensification at a distance of 45 cm, and had no influence at still greater distances. We have already explained this phenomenon earlier and were able to conclude that the primary oscillations correspond in air to a wave whose half-length is equal to 30 cm. One could expect a considerable intensification of the effect by replacing the flat wall with a concave mirror having the form of a parabolic cylinder, the focal line of which would coincide with the axis of the primary conductor. If the concave mirror is to concentrate the rays properly, it is expedient to make its focal distance as small as possible. If, however, the action of the direct wave is not to be destroyed by the reflected one, then the focal distance must not be much less than a quarter of the wavelength. Therefore I chose a focal distance equal to 12½ cm and made a concave mirror, taking a zinc sheet 2 m long, 2 m wide, and ½ mm thick; this sheet was fastened to a wooden frame of the proper form. Thus the height of the mirror was 2 m, the width of its aperture 1.2 m, and its depth 0.7 m. The primary conductor was placed in the middle of the focal line. The wires supplying the discharge were passed through the wall of the mirror. The inductor and the elements were thus located behind the mirror and had no harmful influence. If the oscillations around the mirror are now investigated with the aid of a secondary conductor, then behind the mirror and at its sides absolutely no actions are observed; in the direction of the optical axis of the mirror, however, sparks can be noticed at distances of the order of 5–6 m. Sparks can be observed also at greater distances, up to 9–10 m, if on the path of the wave issuing from the mirror a flat conducting wall is set perpendicular to the direction of its propagation. In this case the waves reflected by the wall at certain points intensify the arriving waves. At other points the two waves

weaken one another. With the aid of a rectilinear conductor one can observe, in front of the wall, distinct maxima and minima, while with the aid of a circular conductor one observes interference phenomena characteristic of standing waves, which were described earlier. I succeeded in detecting four nodal points, situated at the wall itself and at distances of 33, 65, and 98 cm from it. Thus, with a considerable degree of approximation, one may regard half the wavelength as equal to 33 cm, and the period of oscillation as \(1.1 \cdot 10^{-9}\) sec. (the author takes as the period a time equal to half the period adopted by us now. H. M.), if the velocity of propagation is assumed equal to the velocity of light. In the wires a half-wavelength of 29 cm was obtained. Thus, even for these short waves, the velocity of their propagation in wires proves to be somewhat less than in air; but the ratio of the two velocities is very close to the theoretical value, equal to unity, and differs from it less than in our experiments with long waves. This remarkable phenomenon calls for further study1.

Since the phenomena are observed only near the optical axis of the mirror, we may say that an electric ray emerges from the mirror.

Next, I made a second concave mirror, quite similar to the first, and placed in it a rectilinear secondary conductor in such a way that both wires, 50 cm long, coincided with the focal line, while both wires leading to the spark gap passed out by the shortest path through the wall of the mirror, from which they were insulated. Thus the spark gap was located exactly behind the mirror, and the observer could adjust it and examine it without distorting the propagation of the waves. I assumed that, if the device captured the ray, I would succeed in tracing it at still greater distances; and I became convinced that I was not mistaken. In the space at my disposal I was able to observe sparks from one end to the other. The greatest distance at which I traced the ray (for this it was necessary to open a door) was 16 m.

According to the results of the experiments on reflection described below, one may be certain that, in open space, sparks can be obtained at distances up to 20 m. But for further experiments such great distances are not required, and it is practically convenient if the secondary sparks are not too weak. Therefore, for most experiments the most suitable distance is 6–10 m

Now we shall turn to the simplest phenomena that can easily be obtained with the aid of the ray. In all cases where no special reservation is made, the focal lines of both mirrors should be regarded as being arranged vertically.

Rectilinear Propagation

If, on the straight line joining the mirrors, one places, perpendicular to the direction of the ray, a screen made of zinc sheet 2 m high and 1 m wide, then the secondary sparks disappear completely. An equally complete shadow is cast by a screen made of tinfoil or gold leaf. If the assistant intersects the ray, the secondary spark gap darkens as soon as the assistant enters the space of the ray, and lights up again as soon as he leaves it. Insulators do not stop the ray; it penetrates through a wooden wall or a wooden door, so that one may observe, not without surprise, the occurrence of sparks inside a closed room.

If two conducting screens 2 m high and 1 m wide are set up symmetrically to the right and left of the ray (perpendicular to its propagation), they have no influence on the secondary sparks, provided the width of the slit formed by them is not less than the aperture of the mirrors, i.e. 1.2 m. If the slit is made narrower, the sparks weaken and go out when the width of the slit becomes less than 0.5 m. If the width of the slit is made equal to 1.2 m, but the slit is placed to the side of the straight line joining the mirrors, the sparks go out. If the optical axis of the emitting mirror is turned to the right or left from its initial position by about 10°, the secondary sparks weaken; at a rotation of approximately 15° they go out. If the ray has sharp geometrical boundaries, while their shadow does not, then phenomena corresponding to diffraction should be observed. However, I did not succeed in observing maxima and minima at the edge of the shadow1.

Polarization

From the very method of obtaining the ray one can conclude with complete certainty that the ray is formed by transverse oscillations and is linearly polarized in the optical sense. But we can also confirm this by experiments. If the receiving mirror is rotated about the ray until its focal line, and with it the secondary conductor, are no longer horizontal, then it may be noticed that the secondary sparks weaken more and more, and, when the focal lines of both mirrors are crossed, disappear completely, even if the mirrors are placed very close to one another.

Both mirrors play the role of polarizer and analyzer of the polarization apparatus.

I made an octagonal frame 2 m high and 2 m wide and stretched copper wires 1 mm thick on it; all the wires were parallel to one another and were spaced every 3 cm. If the focal lines of both mirrors are set parallel and the grating is placed between them perpendicular to the ray, so that the direction of the wires is perpendicular to the direction of the focal lines, then the presence of the grating has no effect on the secondary sparks. But if the grating is set so that its wires are parallel to the focal lines, then it completely arrests the ray. Thus, with respect to the transmitted energy, the grating behaves like a tourmaline plate acting on a rectilinearly polarized optical ray.

If the focal line of the receiving mirror is set horizontally, then, as was stated, no secondary sparks arise. When a grating whose wires are vertical or horizontal is introduced, the sparks are likewise absent. But if the grating is set so that its wires make an angle of 45° with the horizontal (either of these two angles is possible), then secondary sparks appear. Evidently, the grating resolves the incident oscillation into two components and transmits only that component which is perpendicular to the direction of its wires. This component makes an angle of 45° with the focal line of the second mirror and, being resolved once more by it, acts upon the secondary conductor. This phenomenon is entirely analogous to the appearance of illumination in the dark field of two crossed nicols when a properly oriented tourmaline plate is placed between them.

With regard to polarization, one further observation should be made: by means of the instruments at our disposal at present, it is possible to investigate only the electric force. There is no doubt that its oscillations (when the primary conductor is vertical) take place in the vertical plane passing through the ray, and are absent in the horizontal one. According to the data obtained in the study of slowly varying currents, there can be no doubt that the electric oscillations are accompanied by oscillations of magnetic force, arranged in the horizontal plane passing through the ray and absent in the vertical plane. Thus the polarization of the ray consists not in the fact that the oscillations occur only in the vertical plane, but rather in the fact that electric oscillations arise in the vertical plane and magnetic oscillations in the horizontal one. Therefore the question of in which of the planes the oscillations of our ray occur cannot be decided without specifying which oscillations are meant: electric or magnetic. This also explains the fruitlessness of the old optical discussions, as was first clearly pointed out by Koláček².

Reflection

We have already demonstrated the reflection of waves from conducting surfaces by observing the interference of the incident and reflected waves; moreover, reflection was used in the construction of our concave mirror.

Now we can separate the two systems of waves from each other. First I placed both concave mirrors in a large room side by side in such a way that their openings faced in one and the same direction, and their axes intersected at a point about 3 m from the mirrors. In this arrangement the spark gap in the receiving mirror remained dark. Next, I set up a flat vertical wall of zinc sheet, 2 m high and 2 m wide, at the point of intersection of the axes of the mirrors, with the wall standing perpendicular to the bisector of the angle formed by the axes. In this case intense sparking is observed in the receiving mirror, caused by the ray reflected from the wall. The sparks disappear when the wall is turned about a vertical axis by approximately 15° to one side or the other from its initial position; this proves that the reflection is regular, not diffuse. If the wall is moved away from the mirrors while keeping the point of intersection of their axes on the surface of the wall, the sparks slowly weaken. I was able to detect sparks when the wall was moved 10 m away from the mirrors, so that the waves traveled a path of 20 m. This method can be successfully used for comparing the speed of propagation in air with other (slower) speeds of propagation, for example in a cable.

To obtain reflection of rays at an angle of incidence considerably different from zero, I directed the ray parallel to the wall of a room in which there was a double-leaf door. In the adjoining room, into which this door led, I placed the receiving mirror in such a way that its optical axis passed through the middle of the door and intersected the direction of the primary ray at an angle of 90°. If a vertical flat conducting wall is then placed at the point of intersection, forming angles of 45° both with the ray and with the axis of the receiving mirror, then sparks arise in the secondary conductor, not disappearing even when the door is closed. If the reflecting wall is turned by approximately 10° from the correct position, the sparks go out. Consequently, the reflection is regular, the angles of incidence and reflection being equal to one another. To prove that the path of propagation of the disturbance from the source to the plane mirror and from there to the secondary conductor coincides with that described, it is enough to place a shielding screen at various points of this path. In that case the secondary sparks always disappear; but when the screen is placed arbitrarily in other parts of the room, it has no effect. With the aid of a circular secondary conductor it is possible to determine the position of the wave plane in the ray: it turns out to be perpendicular

the ray both before reflection and after it; thus upon reflection the wave plane turns through \(90^\circ\).

Until now the focal lines of the mirrors had been vertical, and, consequently, the plane of oscillation (of the electric force) was perpendicular to the plane of incidence. In order to obtain reflection in which the oscillations lie in the plane of incidence, I arranged the focal lines of both mirrors horizontally. In this case the same phenomena were observed as before, and no difference in the intensity of the reflected ray in the two cases could be detected. If, however, the focal line of one of the mirrors is vertical and that of the other horizontal, then no secondary sparks are observed. Thus, on reflection, no change occurs in the inclination of the plane of oscillation relative to the plane of incidence, at least for the two positions of it considered above. In the general case this assertion may prove to be incorrect. In particular, it remains unclear whether the ray preserves rectilinear polarization after reflection. The interference arising in front of the mirror, where the two systems of waves intersect, and giving characteristic effects in circular conductors, may perhaps make it possible to solve the questions, well known to opticians, concerning changes of phase and amplitude upon reflection.

Let us mention one further experiment on reflection from electrically anisotropic surfaces. Both mirrors were set up as in the first of the experiments with reflection; but in front of them, as the reflecting wall, there was placed the above-mentioned grating of parallel copper wires. It turned out that the secondary sparks are extinguished if the wires intersect the direction of oscillation at right angles; but if the wires are parallel to the direction of oscillation, then the sparks ignite. Thus the analogy between our grating and a tourmaline plate is preserved only for the transmitted part of the ray. The non-transmitted part is absorbed by a tourmaline plate, but is reflected by our grating1. If, in the last experiment, the focal lines of both mirrors are crossed, then upon reflection from an isotropic wall no sparks arise in the secondary conductor. But I have convinced myself that sparks can be obtained upon reflection from an anisotropic wire grating if the latter is set so that its wires form an angle of \(45^\circ\) with both focal lines. The explanation of this experiment is evident from the preceding arguments.

Refraction

To investigate the question of the refraction of a ray in its passage from air into another insulating medium, I made a large—

a prism from the so-called hard resin (an asphalt-like mass, Hartpech). The cross-section of the prism was an equilateral triangle, the side length of which was \(1.2\) m, and the refracting angle was close to \(30^\circ\). The height of the whole prism, whose refracting edge was vertical, was \(1.5\) m. Since the prism weighed about \(1.2\) tons, and therefore was difficult to move, I made it out of three parts placed one upon another, each \(0.5\) m high. The mass was poured into wooden boxes; since they did not interfere with the experiments, the mass was not removed from them. The prism was set up on a stand of such dimensions that the middle of the refracting edge was at the same height as the primary and secondary spark gaps. After I had satisfied myself that refraction existed, and had approximately estimated its magnitude, I carried out the following experiment. The emitting mirror was placed at a distance of \(2.6\) m from the prism, opposite one of its refracting surfaces, in such a way that the mean line of the beam coincided as far as possible with the center of gravity of the prism, and the refracting surface made an angle of \(65^\circ\) with the beam (reckoned from the side opposite the refracting edge). At the refracting edge of the prism and at the opposite surface two conducting screens were installed, eliminating the possibility of propagation of the beam in any direction other than through the prism. On the side of the emergent beam, a circle of radius \(2.5\) m was marked out on the floor, the center of which coincided with the center of gravity of the prism. Along this circle the receiving mirror was moved, its aperture always being directed toward the center of the circle. If the mirror is set on the continuation of the direction of the incident beam, no sparks are obtained in it; consequently, in this direction the prism gives a shadow. But sparks arise when the mirror is moved toward the base of the prism, and the beginning of their appearance is observed when the mirror is displaced from the above-mentioned position by \(11^\circ\). Subsequently the intensity of the sparks increases up to an angle of rotation equal to \(22^\circ\), and then again begins to decrease. The last, barely perceptible sparks are observed at an angle of rotation of approximately \(34^\circ\). If the mirror is set in the direction corresponding to the greatest action, and moved away from the prism along the radius of the circle, the sparks can be traced over a distance of \(5\) or \(6\) m. An assistant standing in front of the prism or behind it causes the sparks to die out; this proves that the beam reaches the secondary conductor through the prism, and not by some other path.

In a subsequent experiment the arrangement of the prism was kept unchanged, but the focal lines of both mirrors were made horizontal. No changes were observed in this case. To a refracting angle of \(30^\circ\) and a deviation of \(22^\circ\), approximately corresponding to the minimum deviation, there corresponds a refractive index equal to \(1.69\). The optical refractive index for resinous bodies lies between \(1.5\) and \(1.6\). The inaccuracy of our measurements and the insufficient purity of the substance used do not

allow one to ascribe any substantial significance to this discrepancy1.

We have called the phenomenon we have investigated rays of electric force. Perhaps they could have been called light rays of very great wavelength. In any case it seems very probable to me that the experiments described prove the identity of light, heat rays, and electrodynamic wave motion. I think that it is now possible boldly to make use of all the advantages which the admission of this identity gives both for optics and for the theory of electricity.

Explanations of the figures.

To facilitate the repetition and extension of these experiments, I give in Figs. 1 and 2 schematic representations of the apparatus I used, without claiming any durability for them, but evaluating them only from the point of view of the convenience of reproducing the experiments described above. In Fig. 1 the exciting mirror is shown in section and in plan. It is clear from the figure that the base of the mirror consists of two horizontal frames of parabolic form, \(a\), and four vertical uprights, \(b\), which are fastened to the frames with screws and support them. The reflecting sheet is clamped between the frames and the uprights and is attached to them with a large number of screws. The uprights project beyond the sheet both above and below, which facilitates carrying the mirror.

Fig. 1.

In Fig. 2a the arrangement of the primary conductor is shown on an enlarged scale. Both metallic parts enter, with friction, into two sleeves made of thick paper, surrounded by rubber tape. In turn, these sleeves are fastened to a small board by means of four sealing-wax supports, and the small board is attached to a rail connected with the main frames of the mirror (Fig. 1). The lead-in wires

water, enclosed in gutta-percha, enter two holes drilled in the balls of the primary conductor. This device ensures the necessary mobility of the separate parts of the apparatus and can be dismantled and reassembled within a few minutes, which is necessary in view of the need for frequent polishing of the balls of the spark gap. Where the supply wires pass through the mirror, a bluish glow arises during the discharge. To eliminate its effect on the spark gap (under the influence of the glow, the occurrence of oscillations is made more difficult), a screen \(S\), made of a smooth wooden plate, is used.

Fig. 2.

Finally, Fig. 2b shows the secondary spark gap. Both parts of the secondary conductor are fastened to a strip by means of sealing-wax spacers and rubber tape. From the inner ends of both parts run lead-out wires, enclosed in glass tubes, passing through the mirror and approaching one another. At the end of the upper wire there is a small brass ball. To the end of the lower wire is soldered a piece of watch spring, bearing the second pole—a copper point. The point is deliberately made of a softer metal than the ball. Without this precaution it is easily pressed into the ball, and the small sparks arising in this depression elude observation. From the drawing it is clear how the point is moved by means of a screw pressing on the spring, but insulated from it by a glass plate. The peculiar bend of the spring is made in order to achieve very small displacements of the point, which could not be obtained by using only a screw.

Undoubtedly, the apparatus described can be considerably modified, and these modifications will not impair the results of the experiment. On friendly advice I tried to replace the spark gap in

in a secondary conductor by means of a frog’s leg sensitive to current; however, it turned out that this method, so sensitive in other experiments, is not suitable in the present case1.

Literature

1. H. Hertz, Wied. Ann., 34, 155, 1888; 34, 551, 1888; 34, 610, 1888.
2. F. Koláček, Wied. Ann., 34, 676, 1888.
3. König, Wied. Ann., 37, 651, 1889.
4. Lodge and Howard, Phil. Mag., 27, 48, 1889.
5. R. R. Ritter, Wied. Ann., 40, 53, 1890.
6. Dragoumis, Nature, 39, 548, 1890.
7. Boltzmann, Wied. Ann., 40, 399, 1890.
8. Klemenčič, Wied. Ann., 42, 416, 1891.
9. H. Rubens and R. Ritter, Wied. Ann., 40, 55, 1890.

1. At the present time it has proved possible, by various methods, to carry out an objective observation of the phenomena described. Ritter[^5] succeeded in using a frog’s leg. Dragoumis[^6] used a Geissler tube. Boltzmann[^7] proposed a very convenient method in which a victorelectroscope with a leaflet is used. Klemenčič[^8] used a thermoelement. A very clear and elegant method was proposed by Rubens and Ritter,[^9] who made use of a bolometer for demonstrating the experiments and for a series of further, very fruitful investigations. ↩↩↩↩↩↩↩↩