On Short and Very Short Electric Waves*

A. Scheibe, Berlin

Since short and very short electric waves entered radio engineering—which, for many reasons, was impelled to extend the scale of the waves used toward shorter waves—they have attracted great attention because of their technical and scientific interest. Technology acquired the possibility of increasing the range of radio transmitting stations by using a directed beam of rays with a relatively small transmitter power. From the scientific point of view, the possibility emerged of investigating the state of the upper layers of the atmosphere (the ionosphere).

The specialized literature devoted to short and very short waves has become very extensive. Kohl¹, in his review of methods for obtaining very short waves, already in 1930 cited 130 original papers; Försterling and Lassen², in 1931, referred to 56 original papers concerning questions of the propagation of short waves. In the journal Hochfrequenztechnik und Elektroakustik in 1933 alone, no fewer than 37 original papers and abstracts devoted to short waves were published. With such productivity of research, even a specialist working in this field can only with difficulty survey the available material; for specialists working in other fields, such a review is entirely inaccessible.

In the present review, the most important information on the generation, propagation, and use of short and very short waves is set forth in general outline and, as far as possible, briefly. This review is intended not for specialists, but for physicists interested in these problems.

For greater clarity it is necessary to introduce a division of the whole range of wavelengths into several separate parts.

1. Long waves—from 2,000 m and above.
2. Medium waves—approximately from 100 to 2,000 m.
3. Short waves—from 10 to 100 m.
4. Very short waves—from several centimeters to 10 m.

* Report at the colloquium of the German State Physico-Technical Institute, published in the journal Physikalische Zeitschrift 35, 206, 1934; translated by N. N. Malov.

The value 2,000 m has been chosen somewhat arbitrarily; it separates the region from 100 to 2,000 m, in which the waves of broadcasting and government stations lie, from the region of waves used in commercial long-distance communication. The electric waves of groups 1 and 2 propagate chiefly along the earth’s surface (Zenneck waves, ground waves). However, already here, in the short-wave part of group 2, phenomena of fading of reception are observed, testifying to the influence of the upper layers of the atmosphere on wave propagation. The waves of region 3 propagate according to laws determined exclusively by the electrical state of the upper layers of the atmosphere. In radio communication they are used for communication over great distances. Waves shorter than 10 m, belonging to group 4, propagate according to optical laws. Therefore their propagation does not depend on the state of the upper layers of the atmosphere. It should be pointed out, however, that the boundaries 100 and 10 m are somewhat arbitrary.

In the specialized literature the waves of group 4 are called ultrashort, decimeter waves, and microwaves (as the wave is shortened).

It should be recalled that the original experiments of Hertz and Marconi were carried out with short (decimeter and meter) damped waves; subsequently the development of radio engineering proceeded in the direction of long waves; the machine transmitters of powerful stations operated on wavelengths of 18 km and more.

The widespread use of tube transmitters led to an extremely dense filling of the wave range being used, so that by the time of the broad development of radio broadcasting for radio amateurs only wavelengths shorter than 100 m remained free. The undoubted merit of radio amateurs is that they again drew the attention of science and engineering to the problems of practical use of waves shorter than 100 m, which served as an impetus to the development of short-wave technology. It should be noted, however, that although in wireless communication short waves can in many cases replace long waves, a complete cessation of the operation of long-wave stations is impossible, since complete reliability of continuous communication on short waves has not yet been ensured.

Generation of Short and Ultrashort Waves

1. General remarks. The excitation of short and ultrashort electric waves is carried out mainly with the aid of a tube generator with feedback; the basic circuit diagram of its connection is shown in Fig. 1.

The oscillatory circuit of the anode circuit is here coupled inductively to the grid circuit. At wavelengths longer than 100 m there are no difficulties in obtaining, by means of this circuit and a single circuit, sufficient oscillation energy over a very wide range of wavelengths, so that, for example, for obtaining waves

with wavelengths from 100 to 300,000 m one can use only two circuits, differing only slightly from one another in their dimensions. In the wavelength interval from 100 to 30,000 m one can also use the circuit shown in Fig. 3.

In the region of waves shorter than 100 m, a tube generator changes its energy sharply when the wavelength is changed; here it is no longer possible to obtain equal energies over a wide variation of the wavelength range, so that, for example, even for the simplest measurements in the region from 10 to 100 m one has to use at least four circuits of different dimensions.

Besides the difficulties connected with obtaining sufficient oscillation power when changing the wavelength, there are also difficulties caused by the inability of the tube generator to radiate very short waves. This is easily clarified by briefly considering the dependence of the wavelength and the coupling coefficient in a tube generator.

Fig. 1. Transformer coupling.

a) The wavelength \(\lambda\) of a tube generator is, to a close approximation, determined from the constants of the oscillatory circuit by the following relation:

\[ \lambda = k\sqrt{LC}, \]

where \(L\) and \(C\) represent the inductance and capacitance of the circuit. Thus, theoretically, a percentage decrease of \(\lambda\) must correspond to a decrease of

\[ \frac{\Delta L}{2} \quad \text{or} \quad \frac{\Delta C}{2}. \]

In practice, however, as \(L\) or \(C\) is decreased, the influence of the inductance of the wires connecting the circuit with the electrodes of the tube, and also of the capacitance between the anode, the grid, and the cathode of the tube and of the unavoidable parasitic capacitances of the circuit, begins to make itself felt ever more sharply. The latter may be considered as being connected in parallel with the capacitance of the oscillatory circuit (the capacitances shown by the dotted lines in Fig. 1).

b) To excite undamped oscillations in a tube generator it is necessary to observe a definite ratio of the alternating voltage \(U_g\), induced in the grid circuit, to the anode alternating voltage \(U_a\). This ratio is determined by the permeability of the tube \(D\), the resistance of the tube \(R_i\), and the resistance of the anode circuit

\[ R_a = \frac{L}{CR}, \]

and is equal to

\[ D\left(1+\frac{R_i}{R_a}\right). \]

This quantity (called the feedback factor \(K\)) at maximum oscillation power reaches the value \(2D\). Since \(D\)

if in ordinary generator tubes it is close to 0.1, then, in order to obtain oscillations under the most favorable conditions, the alternating grid voltage should be about one fifth of the alternating anode voltage. With a decrease in wavelength, owing to the increasing conductivity of the grid–cathode capacitance connected in parallel with the grid coil, and also because of the decrease in \(R_a\), it is very difficult to obtain this optimal value of the feedback coefficient.

c) In the cases indicated above it was assumed that the transit time of the electron from one electrode of the tube to the other is negligibly small in comparison with the duration of the oscillations of the alternating current. However, for waves shorter than \(1\ \text{m}\), this is no longer the case. The difficulties in obtaining oscillations by means of a tube are thereby increased, since the electron current and the controlling grid voltage no longer coincide in phase. Holmann \(^{3}\) attempted to derive the corresponding relations by means of so-called ultradynamic oscillation characteristics, which would make it possible to connect oscillations obtained by feedback with Barkhausen–Kurz oscillations.

Thus all attempts at the maximum shortening of the wavelength were reduced to finding ways of maximally decreasing parasitic inductances and capacitances and of creating conditions for attaining the optimal value of the feedback coefficient. A whole series of different—one might say individual—circuits was developed which, however, proved applicable only in a very narrow range of waves. In addition, it was necessary to introduce corresponding changes into the design of tubes intended for the generation of short waves.

Short-wave tubes \(^{4}\), such as, for example, the RS-229\(_g\) tube of the “Telefunken” company (Fig. 2), are distinguished by the fact that the anode and grid leads are sealed into the glass at a considerable distance from each other and from the cathode leads, whereby the parasitic capacitances between the electrodes are considerably reduced in comparison with the capacitances arising when the grid and cathode are brought out in the usual manner to the tube base. In addition, the grid lead is designed in such a way that it can safely withstand considerable grid currents, reaching values of \(20\text{—}30\ a\).

Figure 2a shows a Russian short-wave tube of type G-120 with a power of \(5\ \text{kW}\). Its grid lead is located on the side of the glass bulb behind the plates of the oscillatory-circuit capacitor; the cathode is brought out at the upper part of the tube, while the anode, which is a massive copper cylinder, is brought out of the glass bulb and placed in a tank serving to cool the anode with running water.

2. Some feedback circuits. Already in the region of medium wavelengths it proves expedient to replace the circuit of inductive or transformer feedback shown in Fig. 1 by another circuit, in which the grid coil is ...

part of the coil of the oscillatory circuit, feedback is produced by inductive voltage division (Fig. 3).

At short wavelengths the inductive voltage division may be replaced by capacitive division (Fig. 4), which has the advantage that the adjustment of feedback becomes more convenient, and the series connection of the capacitances contributes to a further reduction of the wavelength. To eliminate the settling of charge on the grid, it is connected to the cathode by a high leakage resistance.

Fig. 2. Short-wave lamp RS 229 g of the Telefunken firm.

Fig. 2a. Short-wave lamp G-120 of the Svetlana factory.

Fig. 3. Circuit with inductive voltage division.

Fig. 4. Circuit with capacitive voltage division.

At wavelengths shorter than 10 m, the circuit shown in Fig. 5 works very well, in which feedback between the anode and grid circuits is effected through the internal capacitance of the tubes. The wavelength is varied by lengthening and

shortening the rectangular circuits of the anode and grid, made of tubes that slide into one another, as in an ordinary trombone. With a length \(AB = 22.5\) cm and \(CD = 18.4\) cm, with a small amplifier tube, a wavelength of about 5 m is obtained.

A modification of this circuit, in which a Lecher system of two parallel wires is used as the oscillatory circuit connected between the anode and the grid, makes it possible to obtain waves somewhat shorter than 1 m.

In the circuits considered, a very unpleasant feature is the formation of oscillations in the wires connecting the anode and cathode with the power source. Considerable effort has to be made to reduce the absorption of high-frequency energy in the supply wires to a minimum by inserting suitable chokes in them. The shorter the wavelength, the less benefit these chokes provide.

Fig. 5. Short-wave circuit with internal feedback.

Fig. 6. Push-pull circuit.

A considerable reduction in the absorption of high-frequency energy in the supply wires is achieved by using two tubes connected in a push-pull circuit. Both tubes (Fig. 6) oscillate in opposite phases, owing to which, with a symmetrical design of the generator—for example, when Lecher systems are included in the anode and grid circuits—the supply wires can be connected to easily found nodal points of the high-frequency alternating voltage. In this case the chokes in the supply wires are retained. With the aid of such a circuit it is also possible to obtain waves shorter than 1 m. At a wavelength of 1.3 m, using an REN-904 tube, a high-frequency power of 1.5 W was obtained5.

According to Marconi, using a push-pull circuit it is possible to obtain waves down to 26 cm. The calorimetrically measured power of a similar decimeter generator reaches 3.5 W.

A three-meter Esau generator6 is built on an analogous principle; its circuit is shown in Fig. 7. The oscillatory circuit consists of series-connected elements: \(L_1, L_2, C_2, L_3, C_1\). The anode-grid capacitance of the second tube of the push-pull circuit is here replaced by the capacitor \(C_2\). The feedback coefficient is set ...

by moving the tap of the anode voltage \(b\) along the wire rectangle. The switching-in of the galvanically coupled antenna is likewise carried out by means of a sliding contact, set in the most favorable position. The inclusion of chokes, especially in the filament circuit, is necessary. The high-frequency oscillation power when using a Telefunken lamp of type RS-207* reaches 700 W. With Russian lamps of type G-54 from the Svetlana factory, when using this circuit, it is possible to obtain up to 400 W of oscillatory power at a wavelength of 3.5 m. The G-120 lamp, connected according to the Esau circuit, makes it possible to obtain considerable oscillatory power at a wavelength of 2.7 m.

3. Barkhausen–Kurz generators. As was indicated above, by means of feedback circuits it is possible to obtain waves shorter than 1 m. Thus we enter the region of waves where the period of oscillation of the generator becomes comparable with the time of motion of electrons between the electrodes of the lamp. Under these conditions

Fig. 7. Short-wave Esau circuit.

Fig. 8. Barkhausen–Kurz circuit for very short waves.

a transition to the excitation of oscillations by the Barkhausen–Kurz method is possible. The circuit for this connection is shown in Fig. 8. The most remarkable feature of this circuit is the absence of an external oscillatory circuit. Oscillations are most easily excited in a lamp possessing cylindrical symmetry and having a positively charged grid and a negatively charged anode. Measurements show that the wavelengths are determined by the radii of the electrodes and by the applied voltages. With an unchanged anode voltage, the following relation is obtained between the grid voltage and the wavelength:

\[ \lambda^2 E_g = \mathrm{const}. \]

The occurrence of oscillations can be explained by the oscillatory motion of electrons near the grid under the action of the accelerating

* This lamp, designed for operation at short waves, has a nominal power of 1.5 kW.

of the grid field and the retarding field of the anode. The period of the resulting electrical oscillations must coincide (under this interpretation) with the duration of one oscillation of the electron.

In reality the phenomenon is considerably more complicated than is represented by the above simplest hypothesis, which undoubtedly contains a grain of truth, since the time of motion of the electrons between the electrodes is commensurable with the period of the oscillations.

Without dwelling on a detailed consideration of this question, we shall point out some phenomena that gave impetus to the appearance of a large number of theoretical and experimental works devoted to Barkhausen–Kurz oscillations.

The author of the present report found that, in addition to the ordinary waves, called “long waves,” “short waves” are also excited simultaneously, their length being approximately half that of the former. Of great importance is the nonharmonic relation of the short waves to the long ones. Potapenko discovered the existence of still shorter—“dwarf waves,” whose length could reach only 3.5 cm. These phenomena cannot be explained by a simple oscillatory motion of the electrons relative to the grid. Gill and Morrell, as well as other authors, found that, in addition to Barkhausen–Kurz waves, waves also arise that do not depend on the radius of the electrodes and on the magnitude of the applied voltages, but are determined by the size of an external coupled oscillatory circuit, for example, a Lecher system.

Fig. 9. Kohl’s tube for very short waves. On the left is the schematic diagram; on the right is a view of the tube.

A tremendous number of attempts were made to explain these phenomena. Let us point out that, in Kohl’s opinion, even for Barkhausen–Kurz waves in pure form there exists a tuned oscillatory circuit which, of course, should be sought inside the tube itself. This idea led to the construction of a special tube for 14–30-cm waves with an oscillatory circuit placed inside the tube. Fig. 9 shows the arrangement of such a circuit, consisting of a wire loop and a small capacitor situated between the spiral anode and the grid.

Of great interest is the attempt of Meller^7 and his pupils to consider the processes that cause partial absorption of the energy of electrons and the appearance of “damped” electrons. A detailed consideration of this attempt, as well as of a number of others,

goes beyond the scope of the present survey. We refer those interested to the original works.

The energy of the oscillations obtained in the lamp by the Barkhausen–Kurz method is negligible; usually it amounts only to fractions of a watt. Therefore attempts have repeatedly been made to increase the energy of the oscillations. This is achieved, for example, by using

Fig. 10. Parallel connection of two Barkhausen—Kurz generators. On the left—the voltage distribution; on the right—the connection diagram.

a special baseless lamp8. In this lamp, between the grid and the cathode there is a Lecher system tuned to the frequency of the electron oscillations, so that the leads from the supply source of the grid and anode may be connected to the nodal points of the oscillatory voltage. With great success the Lecher system is also used for supplying power to the filament of the lamp. The author of the present survey9 proposed a parallel connection of several lamps, in which the connections between the electrodes are made in the form of Lecher systems. In Fig. 10b a circuit for two lamps is shown. In Fig. 10a the voltage distribution along the Lecher system is depicted, to the nodal points of which \((q\) and \(p)\) the supply wires are connected. This method has a very favorable effect on the energy of the oscillations, which can be brought to values considerably greater than twice the energy of the oscillations obtained with a single lamp. Another method of increasing the energy of the oscillations, proposed by Marconi10, consists in combining five decimeter-wave generators by means of reflectors into one com-

Fig. 11. Short-wave receiver.

…parabolic generator. Of course, the individual generators must be synchronized to a common wavelength.

1. Receivers according to Barkhausen–Kurz[^11]. The reception of very short waves is practically facilitated by the fact that the generator is modulated by low-frequency oscillations. Reception can be carried out with the aid of a receiver whose circuit is analogous to the Barkhausen–Kurz generator. A considerable improvement in reception is achieved by Holmann by means of an audion called “braking” (Bremsaudion) (Fig. 11), whose principle of operation is analogous to that of an ordinary audion. The sensitivity of reception is determined by the magnitudes of the anode and grid voltages, regulated in the appropriate manner.

Propagation of waves

1. Short waves: 10–100 m. Numerous observations of remarkable phenomena arising in the propagation of short waves—such as, for example, the achievement of extremely large ranges, the formation of a dead zone near the transmitter, the dependence of the range on the time of day, fading and multiple reception of one and the same signal, etc.—have led to the idea that the propagation of waves is determined by an ionized layer surrounding the earth. The existence of such a layer at great heights had been assumed by Kennelly and Heaviside long ago on the basis of observations of the propagation of long waves; to explain the phenomena observed in short-wave communication, the assumption of the existence of this layer is absolutely necessary. The experimental material of radio engineering, and meteorological and astronomical data on the heights of auroras, on the height at which meteors flare up, on the phenomenon of twilight, zodiacal light, the distribution of gases in the atmosphere, etc., make it possible to draw certain conclusions about the structure of the upper layers of the atmosphere which coincide with the ideas obtained in attempts to interpret the processes of propagation of short waves. Försterling and Lassen made a detailed calculation of the path of rays of short electric waves through the ionosphere and showed that abnormal phenomena—such as, for example, increased range, repeated bending of the signal around the earth, double refraction, etc.—can be explained in this way.

The ionization of the upper layers of the atmosphere may be caused by corpuscular radiation, ultraviolet light, or cosmic rays. To explain the phenomena of wave propagation in the most general terms, it is sufficient to ascribe the ionization to the sun’s ultraviolet radiation[^12]. However, an explanation of the extremely rapidly changing phenomena of reception fading, which presuppose rapid changes in the ionization of the active layer, cannot be attributed only to rapid fluctuations in the intensity of ultraviolet radiation; attempts have been made to seek its cause in changing corpuscular…

...polar radiation. Observations of short waves, including observations during the solar eclipse of 1933,^13 which were to decide the question of ultraviolet or corpuscular radiation, yielded material speaking in favor of corpuscular radiation.

As regards the height of the ionized layers of the atmosphere—the ionosphere, or the Kennelly–Heaviside layer (KHL)—on the basis of numerous measurements of height^14 by means of electric waves, one must admit the existence of two separate layers, concerning whose properties the following notions are at present available.

Fig. 12. Changes in electron density in the upper ionized layer as a function of height and time of day.

In the figure: vertical axis—“Height (km)”; horizontal axis—“Electron density \((N_e)\)”; curve labels—“10 h,” “6 h,” “3 h,” “1 h”; annotation—“After sunset”; annotation—“Day—stationary state.”

a) The lower KHL consists of an atmosphere of ionized nitrogen and oxygen; it is situated at an altitude of 100–150 km and has a maximum electron concentration reaching \(10^5\) electrons/cm\(^3\) at an altitude of 120 km. Electric waves whose length is less than 18 m pass through this layer even when incident upon it obliquely. In the opinion of Försterling and Lassen, the reflection of longer waves from this layer is manifested chiefly in the magnitude of the attenuation of long waves.

b) The upper KHL consists of an ionized hydrogen atmosphere; it is situated at an altitude of 200–800 km and has a maximum electron concentration (about \(1.3 \cdot 10^6\) electrons/cm\(^3\)) at an altitude of about 400 km. This layer, which waves shorter than 10 m are still capable of penetrating at oblique incidence, determines the course of short-wave rays; we shall consider this influence below, basing ourselves on the ideas of Försterling and Lassen.

Calculation gives the distribution of electron density in the upper KHL as a function of the height of the layer and the time of day, similar to the distribution shown in Fig. 12. This dependence of the electron concentration on the time of day determines the curvature of the ray in the upper KHL; it shows that the radius of action of a ray of unchanged wavelength varies during the day. In addition, the radius of action depends on the depth to which the ray penetrates into the KHL. This means that the radius of action depends on the angle at which the ray is radiated by the antenna. This dependence, for a definite wavelength and a definite concentration of the KHL, i.e., for a definite time of day, is shown in Fig. 13. The horizontal ray propagates farthest. The closer the emitted ray is to the vertical, the closer the ray reflected to the earth approaches the generator (for example, the point of incidence of ray 6). Finally, with still...

With more vertical radiation (ray 9 and the following ones), the rays penetrate through the layer. The distance between the generator and the point of incidence on the earth, б, determines the radius of the dead zone around the transmitter,

Fig. 13. Path of rays as a function of the angle of radiation.

within which reception is completely absent, if there is no horizontal radiation. Thus the radius of the dead zone depends on the wavelength and the time of day. The dependence on wavelength and on the angle of radiation is shown in Fig. 14, which indicates, for example, that with a 20-meter wave the dead zone during the day has a radius of about 1,500 km (angle of radiation about 30°); with wavelengths of 30 m and more the dead zone is absent altogether.

Fig. 14. Dependence of the radius of the dead zone on the angle of radiation and wavelength (in daytime).

Thus, in order to ensure uninterrupted communication, the wavelength used must change with the time of day, and the difficulties

in the choice of wavelength will be all the greater since, at large distances, on the intermediate sections we have entirely different times of day. In practice it was found that during the day waves from 14 to 18 m are suitable, at twilight—from 18 to 25 m, and at night from 25 m and longer.

The maximum theoretically possible range of action is, for a 20-meter wave, 5,000 km. This result seems to contradict practical data, which indicate the possibility of covering considerably greater distances. Försterling and Lassen overcame this contradiction by allowing for the existence of different attenuation coefficients on different sections of the ray path, and the possibility of multiple reflections of the ray between the SKh and the earth’s surface. With multiple reflection, shown in Fig. 15, it can be shown that at the most distant points of reception field values arise that exceed by approximately 10,000 times the field value that would be obtained with simple single reflection of the ray from the SKh.

Fig. 15. Path of the ray under multiple reflection.

Fig. 16. Splitting of the ray under multiple reflection.

Observations during communication over long distances, in particular between Geltow (near Nauen) and Buenos Aires¹⁵, revealed the existence of multiple reflections. In work on this line, splittings of single signals into 5 and even 10 separate signals, rapidly following one another, were observed. This phenomenon may be explained by the fact that, under multiple reflection, owing to different angles of radiation of the antenna, different paths of rays are obtained, such as those shown in Fig. 16. The difference in the path length of the rays corresponds to the difference in time at the reception of the signal.

Observation of a double reverse signal, described in detail by Kveck and Mögel[^16], judging by the magnitude of the time difference, can likewise be explained by multiple reflections between the SKh and the earth’s surface. In Fig. 17 an oscillogram of this phenomenon, recorded at Geltow, is shown. In addition to signal \(a\), which arrived from Buenos Aires by the shortest path, a signal arriving in the opposite direction \((a_1)\) was recorded, followed after \(0.138\) sec by a second signal \((a_2)\). In the opinion of Försterling and Lassen, this signal cannot be explained by a single prolonged stay of the ray in the SKh, but must be attributed to at least a 10-fold reflection from the SKh and the earth’s surface, with the signal having once more gone around the terrestrial globe.

Both examples cited show how the assumption of the existence of an ionized layer influencing wave propagation can explain the enigmatic phenomena of propagation. In view of the extraordinary difficulty of the problem (it is enough to recall that the upper layers of the atmosphere are accessible only to indirect observations), it is quite understandable that so far it has not been possible to give an exhaustive explanation of all the phenomena.

Fig. 17. Double indirect signal.

Thus, for example, in addition to the above-described indirect signal, there are also observed signals which, apparently, have undergone some reflection in the dead zone and were received as an echo of the transmitted telegram; there has also been observed a “world echo” with a time difference of up to 30 sec. Experiments have been undertaken with the aim of explaining these phenomena; however, it will probably be necessary to accumulate a large amount of experimental material before these complex questions can be resolved. For obtaining large ranges of action in communication on short waves, the SKh plays an enormous role, but sudden rapid changes in the state of this layer cause extremely unpleasant sharp fluctuations of the field strength at the receiving point; the cause of these changes in the state of the SKh should be sought in corpuscular radiation and in disturbances of the earth’s magnetic field. The field strength at the receiving point is determined by the intensities possessed by the signals arriving at the receiver by different paths; not all rays are radiated by the antenna at the same angles, and, moreover, the length of the ray path depends on the frequency.

If along the path of the ray the height of the SKh or its absorbing capacity changes, then under favorable conditions all rays may arrive at the receiving point with the same weakening of intensity,

which will cause a general decrease in the strength of reception; in this case one speaks of a general fading of the transmission. To this general fading one may oppose various measures for combating it in the receiver itself. But the state of the SKh may also change in such a way that only individual frequencies—for example, the carrier frequency or the side frequencies—experience the influence of these changes. In this case the strength of reception changes only for these frequencies. This phenomenon is called selective fading. Its cause may be, for example, the influence of the earth’s field on the rotation of the plane of polarization of the wave under consideration. Another fading phenomenon—local fading (Nahschwund)—is caused by the interference of the ray arriving from the Kennelly–Heaviside layer and the ground ray arriving directly from the transmitter; this phenomenon may be observed in long-distance communication. The partial weakening of the ray arriving from the Kennelly–Heaviside layer is also called local fading.

Besides combating fading in the receiver itself, considerable assistance is provided by installing special transmitting and receiving antennas that make it possible, at short distances, to obtain reception free from fading; in this case waves shorter than 10 m prove to be very favorable. It must be borne in mind, however, that this advantage is purchased at the cost of a reduction in the range of action, which in many cases is undesirable.

2. Very short electric waves. Waves shorter than 10 m penetrate through the SKh even when falling obliquely. Thus no ray reflected toward the earth is obtained here. The direct ground ray is also unsuitable for communication, since these short waves are absorbed even by insignificant obstacles, such as trees, houses, etc., as a result of which, even at high powers, the intensity of reception falls below the limit of receiver sensitivity already at a distance of 1–2 km. Experiments show that, for example, a three-meter transmitter installed at the foot of a hill 69 m high was audible only 8 m below the summit (on the opposite side of the hill)\(^{17}\).

Thus very short waves are suitable for communication only in those cases where there are no obstacles of any kind between the transmitter and the receiver and where a straight line, not distorted by the curvature of the earth’s surface, can be drawn between them\(^{18}\). Electric rays propagate rectilinearly, like light rays. But, unlike the latter, electric rays up to 10 cm in length are capable of penetrating through fog, making it possible to carry out communication under weather conditions. Thus the transparency of a medium for very short waves is not connected with its optical transparency.

Owing to the smallness of their wavelength, very short waves can be concentrated more easily into directed beams than short waves. For this purpose reflecting antennas may be used, but for the shortest waves it is preferable to employ parabolic mirrors. As a result of the insignificant scattering and the possibility of installing analogous concentrating devices

... at the receiver; even with insignificant powers, a fairly large range of action can be obtained.

Investigations of propagation, carried out with a transmitter assembled according to Esau’s circuit and having a wavelength of 3.2 m[^19], installed on a tower situated at a height of 1140 m, showed that the practical range of action proved to be 110 km, which corresponded to the theoretical value of the radius of visibility. Beyond this lay a diffraction zone 6–15 km wide. It is noteworthy that changing the transmitter energy in the ratio 80:1 had no substantial effect on the strength of reception, which is entirely in accord with the laws of optics.

Marconi[^20], working with a transmitter of 10 W power, giving a wavelength from 57 to 26 cm, installed on the seashore on a mountain 750 m high, established communication with a steamer 97 km away, which corresponded to optical visibility. Beyond this lay a diffraction zone extending up to 130 km. In this zone reception was accompanied by strong fading phenomena.

We see that in both cases a diffraction zone is observed, in which fading phenomena exist. Thus the absence of fading is observed only in the zone of optical visibility. The occurrence of the diffraction zone may be explained by the bending of rays by layers of evaporation located above the earth’s surface. Changes in the state of these layers apparently cause fading phenomena.

Thus the range of action of very short waves, within which reliable reception is ensured, is determined by the radius of optical visibility. We shall illustrate the possibility of using these waves with several examples.

For military purposes, communication between small military units on very short waves is extremely suitable, all the more so since they provide reliable reception only over short distances, and the possibility of concentrating them helps preserve the secrecy of transmission.

For navigation purposes, both in water and in air communication, a number of methods have been proposed for using very short waves[^21]. The ability of the rays to penetrate through fog makes them very applicable for guiding ships when entering a harbor by means of radio signals, and also for indicating the direction of landing for airplanes. As an example, let us consider the radio beacon developed by the Lorenz firm[^22] (Fig. 18). A dipole antenna, oscillating at half the wavelength, is continuously excited by a generator. In the plane of the radiating dipole are placed two reflecting dipoles (reflectors), which, in a definite rhythm, are closed and opened in such a way that while one is closed the other is open. The closed reflector deflects the beam of the main dipole toward the open reflector. If the switching occurs so that for the left antenna the rhythm — · — · ... is observed, then the right antenna is closed in the same rhythm at...

the interruption time of the left one, so that for it the rhythm obtained is . — . — . If the airplane approaches from the left of the beacon, it hears the signal . — . —, while if it approaches from the right, the signals have the rhythm — . . — .; finally, if it flies exactly perpendicular to the plane in which the antennas are located and moves straight toward the middle antenna, then both signals combine into a continuous signal ————. In this way the pilot determines the correct direction for landing. With a 7-meter transmitter of 2 watts power the range reaches 4–5 km, and the aperture angle of the beam determining the correct direction is 2–3°. In addition to indicating the direction of landing, it proves possible, by using an appropriate distribution of the field of the emitted waves in combination with a field-strength meter, to indicate to the airplane the landing point, which ensures the correct angle of descent of the airplane²³.

Fig. 18. Radio beacon for determining the direction of an airplane’s landing.

This example is sufficient to prove that the use of very short waves in radio communication is not only possible, but also necessary.

Finally, within the scope of the present survey one can only point to the applicability of short waves for the investigation of the molecular structure of matter (dispersion, absorption), and also for medical purposes (selective heating of solutions and tissues)²⁴.

Use of the Thermal Effect of Short and Very Short Waves*

Short waves can be used in therapy by heating the organism in the field of the capacitor of a short-wave circuit. This heating, in comparison with ordinary diathermic heating (at \(\lambda = 300\text{–}600\ \mathrm{m}\)), presents known advantages: first, it is possible to heat the object without direct contact with the plates of the capacitor, since the capacitive resistance of the air gap at such high frequencies is very

* This paragraph was written by the translator.

little; thanks to this, the heating of open wounds, abscesses, etc., becomes possible, which is inaccessible to ordinary diathermy. Secondly, the distribution of heat in the tissues of the organism differs sharply from the distribution of heat at diathermic frequencies. Indeed, regarding tissue as a semiconductor with dielectric constant $\varepsilon$ and electrical conductivity $\sigma$, and assuming that the tissue is taken in the form of a plane-parallel layer of area $S$ and thickness $d$, situated in a homogeneous field, it is easy to show that the amount of heat released in it will be proportional to the quantity

\[ q=\frac{1}{\sigma}\cdot \frac{1}{1+\left(\frac{\varepsilon}{2\lambda}\right)^2}; \tag{1} \]

for long waves the term in parentheses is small, so that heat generation is determined practically only by $\sigma$; for short waves, however, the influence of $\varepsilon$ begins to be felt; for an unchanged value of $\varepsilon$, the maximum heat generation will occur in that tissue whose electrical conductivity satisfies the relation:

\[ \sigma=\frac{\varepsilon}{2\lambda}. \]

Fig. 19. Heating of electrolytes in the electric field of short waves.

Experimental verification of equation (1), carried out by measuring the heating of aqueous solutions with a gradually changing concentration of the dissolved substance (NaCl, KCl, etc.), gives quite satisfactory results;^24,25 this is evident from Fig. 19, where the rise in temperature is plotted along the ordinate axis, and concentration along the abscissa axis; the dashed curves represent theoretically calculated values of heating, while the solid curves represent experimental data for two different wavelengths. The displacement of the experimental curve relative to the theoretical one is explained by the fact that, in constructing the theoretical curve, the values of $\varepsilon$ and $\sigma$ were taken from tables and were not determined experimentally.

Proceeding from data on $\varepsilon$ and $\sigma$ for various component parts of the organism, obtained at lower frequencies (measurements at very short waves have not yet been made), it can be shown that the maximum heating of a number of tissues and organs lies precisely in the region of very short waves.^26 Calculation of heating^27 by formula (1) leads to the values of heat generation shown in Fig. 20. From this figure two principal facts are clearly revealed: a change in the distribution of heat generation and a decrease in its absolute magnitude as the wavelength is shortened. Of course,

such a calculation is approximate, since it completely fails to take into account the possibility of the existence of a dispersion region of variation of the dielectric constant of substances possessing dipole molecules, whereas there are a number of grounds for expecting the possibility of dispersion precisely in this wavelength interval[^26].

Fig. 20. Heat release in the constituent parts of an organism at different wavelengths.

The study of the heating of dielectrics in the electric field of very short electric waves makes it possible to test Debye’s theory and thus to broaden our knowledge of the structure of matter[^29].

As regards the biological action of short and very short waves and their therapeutic use (the latter field requires powers of the order of tens, even hundreds, of watts; thus waves shorter than 3 m have not yet found application in therapy), this subject lies outside the scope of the present survey; those interested in this question are referred to the original works[^30], [^31], [^32], [^33].

Literature

1. K. Kohl, Erg. d. exakt. Naturw. 9, 275, 1930.
2. K. Försterling und H. Lassen, Z. techn. Phys. 12, 453, 502, 1931; see also H. F. Techn u. El. Ak. 42, 158, 1933.
3. H. E. Hollmann, Sitz-Ber. d. Preuss. Akad. d. Wiss., Phys.-Math. Kl. 6, 293, 1933. Ref. in H. F. Techn. u. El. Ak. 42, 32, 1933.
4. On lamps for \(\lambda < 1\) m, used in feedback circuits, see B. J. Thompson and G. M. Rose, J. Proc. Inst. Rad. Eng. 21, 1707, 1933; Zus. Ber. über Röhren s. W. E. Rühle, Telef.-Ztg. 13, Nr. 61, S. 5, 1932.
5. A. Esau und W. Köhler, H. F. Techn. u. El. Ak. 41, 153, 1933.
6. H. Wechslung, H. F. Techn. u. El. Ak. 31, 176, 1928.
7. H. G. Möller, E. N. T. 7, 293, 1930; E. W. Helmholtz, E. N. T. 10, 181, 1933.
8. A. Schreibe, H. F. Techn. u. El. Ak. 27, 1, 1926.
9. RS 296; Rühle a. a. O.
10. G. Marconi, Alta Frequenza, Märzheft 1933.
11. H. E. Hollmann, H. F. Techn. u. El. Ak. 42, 89 u. 185, 1933; see also Thompson u. Rose, a. a. O.
12. On the influence of sunspots see H. Mögel, Telef. Z. 14, Nr. 65, S. 27, 1933.
13. John T. Henderson, H. F. Techn. u. El. Ak. 42, 79, 1933.
14. On the measurement of heights see V. G. Golbai and J. Zenneck, H. F. Techn. u. El. Ak. 40, 77, 1932, ebenda 41, 77, 1933.
15. For a survey see O. Böhm, Telef. Z. 10, Nr. 53, S. 9, 1929.
16. E. Quäck u. A. Mögel, E. N. T. 6, 45, 1929.
17. K. Stoye, H. F. Techn. u. El. Ak. 35, 235, 1930.
18. F. Schröter, E. N. T. 7, 1, 1930. — W. Hahnemann, ebenda S. 18. — F. Schröter, E. N. T. 8, 431, 1931.
19. F. Gerth u. W. Scheppmann, H. F. Techn. u. El. Ak. 33, 23, 1929.
20. A. a. O.
21. W. Hahnemann, a. a. O.
22. F. Kramar, E. N. T. 9, 469, 1932.
23. E. Kramar, E. N. T. 10, 451, 1933.
24. N. N. Malov, H. F. Techn. u. El. Ak. 42, 190, 1933.
25. J. Pätzold, Z. Hochfrequenz 36, 83, 1930.
26. J. Pätzold, Z. techn. Phys. 13, 212, 1932.
27. N. Malov, Physik. Z. 34, 883, 1933.
28. P. Debye, Polar Molecules.
29. H. Bleck, Physik. Z. 34, 721, 1933.
30. E. Schliephacke, Kurzwellentherapie, Berlin 1932.
31. E. Raab, Die Kurzwellen in der Medizin, Berlin 1933.
32. Pflomm-Arch. klin. Chirurgie, 166, 251, 1931.
33. N. Malov, Kurortology and Physiotherapy, 12, No. 1–3–5, 1934.