On the Application of Visible and Invisible Rays in Communication Technology and Traffic Safety*

Gerhard Tresch

General Principles

1. The waves employed and their excitation.
2. Receiving devices.
3. Propagation of waves and the influence of atmospheric conditions upon it.

Special Applications

1. Optical telephony.
   Reception of modulated radiation.
   Details of operation.
2. Detection of thermal radiation.
3. Other applications.

Bibliography

Introduction

The possibility of transmitting messages by means of light rays was used in a primitive form even in antiquity. Over time, and especially in the nineteenth century, thanks to advances in the manufacture of lenses and mirrors, this method of communication was improved to such an extent that it became possible, over distances within the line of sight, to carry out uninterrupted telegraphic communication with sufficient speed. In this case reception was performed exclusively by direct observation with the eyes. New paths were taken only when it became possible to modulate the radiation of light sources, for example that of an arc lamp; at the same time there also appeared the possibility of te—

* Phys. ZS 1931. Translated by E. M. Brumberg.

phone communication. For such reception it is necessary to have a device that makes it possible to transform modulated radiation into the form of audible transmission. By that time the technique of electrical transmission, especially wireless transmission, had achieved such great success that the problem of carrying out communication without wires seemed thereby to have been solved. The further development of optical means of transmission receded into the background, in view of the fact that optical telephony has a limited range of action. Only recently, when small distances (several kilometers) began to be covered wirelessly, using for this purpose small, light devices with a very limited radius of action (ultrashort waves), intended for special purposes, has greater attention again begun to be paid to transmission by means of light. In the times of the primitive forms of optical transmission, visible light alone was used for communication. Now, however, when the eyes are no longer used for reception, it has become possible also to use the ultraviolet and infrared parts of the spectrum; moreover, invisible light offers certain advantages.

Devices for detecting infrared rays may find application in another field as well. Thus, for example, they may be used to detect bodies that, owing to a difference in temperature, emit a somewhat different region of infrared rays in comparison with surrounding objects (steamship funnels). Such devices may be employed as an auxiliary means for ensuring the safety of communications.

GENERAL PRINCIPLES

1. Applicable wavelengths and their excitation

The shortest-wave rays used up to the present for purposes of optical communication are the ultraviolet rays employed by Zickler, with a wavelength of about \(0.35\,\mu\) (voltaic arc). The upper limit of the wavelengths

suitable for communication purposes lies at about \(2\,\mu\). Longer waves, obtainable only under complicated experimental conditions (Rubens’ method, mercury lamp), cannot be used also because, along with the difficulty of exciting them, for this region of the spectrum there are as yet no sufficiently inertia-free receivers. For work in the applicable spectral region from \(0.22\) to \(2\,\mu\), the following radiation sources are used: for the ultraviolet part, a mercury arc lamp and a carbon arc; for the visible and infrared parts, a carbon arc, an incandescent bulb, and gas-discharge lamps. Individual regions of the spectrum, needed for special purposes, are isolated with the aid of appropriate light filters[^12]. Thus, by using a mercury lamp and a filter made from a solution of nickel oxide[^26], one can obtain exclusively ultraviolet rays, and with a carbon arc and a Schott glass filter RG7, exclusively infrared rays.

For purposes of the safety of communications, a wavelength of \(2\,\mu\) is not a limiting one, since a heat detector need not necessarily employ an inertia-free instrument. The longest waves that can be detected with the aid of heat detectors are emitted by bodies at low temperature, for which the maximum of the emitted energy lies at about \(10\,\mu\).

2. Receiving Devices

For detecting the radiation used for communication and heat signaling, use is made of the photoelectric effect (photocells and selenium cells) and of the thermal action of radiation (bolometers and thermopiles). Cells used for communication must, along with sufficient sensitivity, possess as little inertia as possible. For this reason, integral receivers (bolometers and thermoelements), which possess extremely great inertia, are completely unsuitable for telegraphy. Of the existing photocells—on whose construction and details we shall not dwell here, since there is extensive specialized literature on this question (see Gudden[^13])—

one should choose the appropriate type depending on the wavelength used. The unequal sensitivity of the individual types of photocells to different parts of the spectra is determined chiefly by the material of the cathode. Thus, for example, the resonance wavelength for potassium lies approximately at \(0.44\,\mu\), and for cesium at about \(0.52\,\mu\). An increase in the sensitivity of the apparatus toward the red end of the spectrum is achieved by using sodium cells filled with gas. The greatest wavelength to which such cells are still sensitive corresponds to approximately \(0.8\,\mu\), while the wavelength for which photocells still possess practically sufficient sensitivity corresponds to approximately \(0.6\,\mu\).

A further increase in sensitivity toward the long-wave end of the visible spectrum was obtained thanks to the discovery of the property of the crystalline modification of selenium to change its electrical resistance under the action of light. In the first experiments in the field of phototelephony (Simon, Ruhmer), selenium photocells were used exclusively as receivers. The properties of various kinds of selenium cells are described in detail in the works of Ruhmer \(^{45,47}\) (Fig. 38) and Bernard \(^{3}\).*

Since, in recent times, much attention has been devoted to the purposes of communication specifically by means of ultrared rays, new elements of the selenium type have arisen, thanks to the systematic work of Kutz \(^{16}\), Koblenz \(^{7}\), Majorana and Tedesco \(^{29}\), Michelson \(^{32}\), Schreter \(^{53}\), and Zewig \(^{58}\), which have their maximum sensitivity in the infrared region. In this connection it turned out that the most suitable elements for this purpose are those made of crystalline materials, such as, for example, molybdenum sulfide, selenium-tellurium, and metal-sulfur compounds.

A new type of photocell was indicated by Schottky \(^{51,52}\) and Lange \(^{22,23}\) (see also Graffunder \(^{14}\)). The element consists of a copper plate coated with a layer of cuprous oxide \(\mathrm{Cu}_2\mathrm{O}\), which is not—

* On the mechanism of action of the selenium cell on the basis of new views, see the work of Reichenstein \(^{37}\).

opposite electrode (cuprox). If light falls through this electrode onto the copper oxide layer, photoelectrons are liberated, as a result of which a photocurrent appears.

The circuit of Lange’s photoelement is shown in Fig. 1. This element, unlike others, can also operate without an additional voltage. Its maximum sensitivity lies at approximately \(0.8\,\mu\), and the sensitivity limit at \(6.6\,\mu\). Fig. 2 presents the spectral sensitivity of various photoelements. For good transmission of speech in telephony it is necessary that the changes in the receiving element follow, with sufficient

Fig. 1. Circuit of Lange’s photoelement \(^{22}\)

Fig. 2. Spectral sensitivity of various types of photoelements:

— potassium element according to Cykler \(^{72}\),
— selenium element according to Streeter \(^{58}\),
— cuprous-oxide element according to Lange \(^{22}\),
— thallium element Osram T 52 \(\}\) according to Zworykin \(^{58}\),
— thallium element “Cema” \(\}\) according to Zworykin \(^{58}\).

speed the oscillations of light. This property is called the inertia of the photoelement and is determined by the time, in seconds, required, counting from the moment of illumination, for the ...

so that the current in the photocell circuit reaches its greatest magnitude. In Fig. 3 the increase of current is shown for various types of cells under sudden illumination. In photocells the current reaches its maximum value already after \(10^{-8}\) sec/cm (Schreter and Lyubchinskii \(^{54}\)). Elements of the selenium and thallium types possess the greatest inertia, which increases together with the thickness of the layer.

Fig. 3. Current curve under sudden illumination, according to Zworykin \(^{58}\).

Fig. 4. Dependence of the sensitivity of various cells on frequency:

— photocell \(E = 80\) V;
— — photocell \(E = 90\) V;
— · — selenium cell \(E = 100\) V.

Practically sufficient clarity of telephone transmission was achieved even with the aid of a selenium cell having an inertia of \(10^{-1}\) sec. A consequence of inertia under variable illumination is a decrease in the amplitude of the photocurrent with increasing frequency of the light oscillation; this phenomenon is manifested to the greatest degree in cells possessing considerable inertia. In Fig. 4 this dependence on frequency is shown for individual types of cells. A further condition for achieving correctly sounding telephone transmission is proportionality between illumination and photocurrent.

Fig. 5. Relation between photocurrent and illumination, according to Zworykin \(^{58}\).

This proportionality is strictly fulfilled for photoelements and cuprous elements in their working range, whereas in the thallium element there is a slight decrease in slope at high illumination intensities, and in selenium elements, on the contrary, there is a strong decrease in slope under weak illumination (Fig. 5).

The maximum limit in the practical application of the above-mentioned photoelements is \(2 \mu\). Beyond this limit one must resort to integral receivers (bolometers, thermopiles). On bolometers and thermopiles, see the books and articles by Lecomte \(^{24}\), Rawlings and Taylor, Schaefer and Matossi \(^{49}\), Fege \(^{65}\), and Westphal \(^{66}\).

Since in heat detectors radiation up to \(10 \mu\) is detected, only bolometers and thermopiles can be used here. Their inertia in the present case is of no significance, since here the reception of unmodulated radiation is involved.

3. Propagation of waves and the influence of atmospheric conditions upon it

The theoretically possible maximum radius of action of optical communication, including ultraviolet and infrared rays, is determined, in the absence of absorption in the intervening medium and without taking into account the transmitter power and receiver sensitivity, exclusively by the rectilinear propagation of light. The range depends only on the elevation of both stations and can be calculated from the following expression:

\[ r=\sqrt{2\rho\,(h_1+h_2)}=3.55\sqrt{h_1+h_2}\ \text{km}, \]

where \(\rho\) (the radius of the Earth) \(=6.4\cdot10^6\ \text{m}\), and \(h_1\) and \(h_2\) must be expressed in meters. Thus, for example, with one station located at NN and the other at a height of \(1000\ \text{m}\), this range would be approximately \(110\ \text{km}\). Within this purely geometrical limit, the range is determined, without taking account of absorption by the medium, on the basis of the minimum required for the operation of the receiving element

illumination. According to Zickler1, for the potassium photocell used by him in combination with a biconvex lens of diameter 11 cm, this quantity has the following value:

1. Normal, good audibility, minimum illumination

\[ E g_{(\min)} = 4 \cdot 10^{-3}\ \mathrm{Lx}. \]

1. Weak, still intelligible audibility, minimum illumination

\[ E g_{(\min)} = 4 \cdot 10^{-4}\ \mathrm{Lx}. \]

1. Limit of audibility, no longer intelligible, minimum illumination

\[ E g_{(\min)} = 2 \cdot 10^{-5}\ \mathrm{Lx}. \]

Taking as a basis the illumination threshold for good, normal audibility, we obtain for various arc lamps the ranges indicated in Table 1.

TABLE 1

Dependence of range on transmitter power

(searchlight with Herda-Beck carbons)

Current strength	Mirror diameter in cm	Luminous intensity in HK	Range in km
15	25	\(6 \cdot 10^6\)	38
30	35	\(15 \cdot 10^6\)	61
60	60	\(75 \cdot 10^6\)	137
225	110	\(560 \cdot 10^6\)	374
300	200	\(2000 \cdot 10^6\)	707

If we now take into account the absorption of the intermediate medium, then between the possible range \(r\) and the range in the presence of absorption (the ideal range of action \(ri\)) there exists the following relation, used in calculating a searchlight:

\[ \frac{r}{\sqrt{\left(1-\frac{a}{100}\right)^2}}, \]

where \(a\) is the absorption in percent per 1 km, and \(r\) is the range in km. For the minimum illumination of the receiver, taking into account the range in the presence of absorption, the following relation is obtained:

\[ E g_{(\min)} = \frac{\left(1-\frac{a}{100}\right)^2 I g}{10^6 r^2}, \]

where \(I g\) denotes the intensity of the transmitter in Hefner candles. Calculated from this relation, as a function of absorption, are the ranges for various energies

of the transmitter, taking into account the minimum illumination necessary for good audibility, \(E_g{}_{(\min)} = 4 \cdot 10^{-3}\ \mathrm{Lx}\), are presented according to Zickler in Fig. 6 (\(I'\) and \(I''\) indicate the operating range with poor audibility or at the threshold of audibility).

Fig. 6. Dependence of operating range on absorption at various intensities of transmission according to Zickler
(numerical data, see table).

No.	Carbons Ø mm	Carbons Ø mm	Arc \(I_g A\)	Arc \(E_g V\)	Mirror Ø cm	Light intensity \(I_g H<\)	\(L_{\infty}=\dfrac{I_g}{S}\) at \(F=1170\), \(\cos \varphi=0.48\)
Experimental lamp with Herza–Beck carbons	Experimental lamp with Herza–Beck carbons	Experimental lamp with Herza–Beck carbons	Experimental lamp with Herza–Beck carbons	Experimental lamp with Herza–Beck carbons	Experimental lamp with Herza–Beck carbons	Experimental lamp with Herza–Beck carbons	Experimental lamp with Herza–Beck carbons
I	6	6	30	40	20	\(1 \times 10^6\)	6 A
Searchlight with Herza–Beck carbons	Searchlight with Herza–Beck carbons	Searchlight with Herza–Beck carbons	Searchlight with Herza–Beck carbons	Searchlight with Herza–Beck carbons	Searchlight with Herza–Beck carbons	Searchlight with Herza–Beck carbons	Searchlight with Herza–Beck carbons
II	3	3	15	45	25	\(6 \times 10^6\)	3 A
III	6	6	30	55	35	\(15 \times 10^6\)	6 »
IV	9	8	60	60	60	\(75 \times 10^6\)	12 »
V	16	14	225	90	110	\(560 \times 10^6\)	45 »
VI	185	16	300	100	200	\(1000 \times 10^6\)	60 »

If, at the receiver, the lens of 11 cm diameter, for which all the above calculations have been made, is replaced by a larger one, or by mirrors of larger diameter, then the operating range increases in proportion to the surface area of this lens or mirror.

The magnitude of the loss of radiation energy when passing through an intermediate medium depends on atmospheric conditions. This magnitude is greatest during rain and fog. But even with good optical visibility, losses already appear that depend on wavelength. In practice, however, for infrared and visible waves these losses are appreciable only at very great ranges, whereas in ultraviolet light the losses are noticeable even at short distances. According to the measurements of Dawson, Granath, and Gelbert8, for various ultraviolet waves the following distance \(d\), in km, is required for the light intensity to be reduced to \(1/100\) of its original value:

\(\lambda\) in \(\mu\)	0.28	0.25	0.22	0.205
\(d\) in km	22	5	0.51	0.20

Above \(0.3\,\mu\), losses under good visibility are also negligible for ultraviolet light. During rain and fog, however, even at short ranges, for all waves up to and including infrared rays, losses are of decisive importance, so that their influence had to be investigated experimentally. The principal influence on losses during rain and fog is exerted by the absorbing and scattering properties of the particles. Which influence is decisive in each individual case depends on the size of the particles and their number per unit volume, i.e. on the type of precipitation. For the ratio of the magnitudes of various types of precipitation, see Table 2. Absorption does not depend on the size of the particles and their number, and is determined exclusively by the water content per unit volume.

The effect of scattering* varies, depending on the ratio between the particle diameter and the wavelength. If the size of the particles is of the same order as the wavelength, then Lord Rayleigh’s law applies, according to which the intensity

* See also the theoretical works of Mie33, Debye9, Stratton63.

scattered light increases inversely proportional to the fourth power of the wavelength. In addition, the amount of scattering increases further with an increase in the number of particles per unit volume. Particles of the indicated order of

TABLE 2

Ratio of quantities for various kinds of precipitation

Kind of precipitation	Precipitation per hour	Diameter of drops in mm	Amount, mg of water per cm³ of air
Light fog	minimum	0.01	6.0
Fog	0.05	0.10	55.5
Drizzling rain	0.25	0.20	92.6
Light rain	1.00	0.45	138.9
Moderate rain	4.00	1.00	277.8
Heavy rain	15.00	1.50	833.3
Very heavy	40.00	2.10	1851.9
Downpour	100.00	3.0 to 5.0	5401.4

magnitude occur in the thinnest fog. In this case, by employing long waves, especially infrared light, one can obtain a considerable increase in transmission range. For large particles, such as occur in dense fog and light rain, Lord Rayleigh’s law is no longer valid. The losses even begin to increase* with increasing wavelength until the wavelength, in order of magnitude, again becomes comparable with the size of the particle and Lord Rayleigh’s law comes into force again, which can be achieved with infrared rays. If the particles are too large in comparison with the wavelength (for example, in heavy rain), then the laws of geometrical optics for reflection and refraction in water drops, which do not depend on wavelength, are decisive for the scattering effects. To what extent absorption and scattering act jointly, since they depend on different quantities, is rather difficult to state in a general form. For individual cases an answer can be obtained by experimental investigations.

Granath and Gelbert^15 investigated, during a heavy

* According to the measurements of Rudolph^41, Gelbf and Schering^13.

of fog for waves from 0.4 to 3 μ, the dependence of the intensity of the transmitted radiation on the wavelength.

The thickness of the layer at which the intensity decreased to \(1/100\) of the initial value was as follows for different wavelengths:

\[ \frac{0.4 \quad 0.5 \quad 0.6 \quad 1.0 \quad 2.0 \quad 3.0\ \mu} {710 \quad 843 \quad 910 \quad 970 \quad 980 \quad 980\ m}. \]

Fig. 7. Transmittance of a cloud of small particles according to Anderson¹.

It follows from this that, for the given concentration and size of the fog particles, the strong superiority of infrared rays expected according to Rayleigh’s law, due to the effect of absorption, will be compensated to a considerable degree. Since systematic experiments with ordinary fog are very difficult because of continuously changing conditions, Anderson¹ carried out various laboratory tests with artificial fog of different concentration and particle size. He investigated fog consisting of small and large particles for various regions of the spectrum selected by means of a filter; as the basis of the curves (Figs. 7 and 8) he used the percentage transmittance of dark-red light

(0.7 to 12 μ) for various concentrations, which was plotted on the abscissa axis, while the ordinate axis showed the transmittance in percent for other colors.* The curves show, first of all, that for all concentrations, for both types of fog, the transmittance of infrared rays is considerably greater than for visible light. Further, the transmittance from infrared to visible light in both cases increases with increasing concentration, since absorption increases with increasing concentration more slowly than the effect of scattering. The transmittance for infrared rays is greater in a fog of small particles (Fig. 8) than in a fog of large particles (Fig. 7), since in the latter case absorption is stronger.

Fig. 8. Transmittance of a cloud of large particles according to Anderson¹.

Whereas in Granath and Gelbert only a slight superiority is seen in the range of action of infrared rays in comparison with visible rays during

* With this method of plotting the graph, the curve for dark-red color is a straight line.

fog, this difference is much greater in the artificial fog studied by Anderson, and especially large at higher concentration. For example, in a fog consisting of large particles in a high concentration, with a transmittance for saturated-red light of 5%, the transmittance of visible light is only 2%, whereas for infrared (1.05 to 2.7 μ) the transmittance is 40%.

The slight difference in transmittance found by Granat and Gelbert occurred in artificial fogs to a much greater degree even at the weakest concentration; it follows from this that the fogs investigated here were much denser than, generally speaking, real ones.

On the basis of the conclusions set forth above—that with infrared rays under unfavorable atmospheric conditions it will nevertheless be possible to obtain the greatest ranges of action—attention has recently again turned to the development of optical telephony using infrared rays.

SPECIAL APPLICATIONS

1. Optical Telephony

Optical transmission is preferable in cases where communication is to take place only between two stations. To achieve the greatest range of action, an intense transmitter is necessary, the radiation of which could also—so as to achieve greater intensity—be made directional. By means of very strong directionality one can ensure that, even when visible radiation is used, the possibility of eavesdropping by a third station will be excluded.

The first optical telegraph apparatus suitable for practice,* in which directionality of action was achieved first by a system of lenses and then by mirrors, was proposed by Mangin[^31] in 1870; at that time he succeeded in attaining a range of action of 25 km. Figure 9 shows the apparatus constructed by him.

* For earlier work see Hennig[^19].

With the aid of these apparatuses, the so-called optical telegraphs, only a comparatively slow rate of telegraphy can be achieved, since reception is carried out by the eye. To increase the speed of telegraphy an acoustic receiving device is necessary. For this purpose the emission of the rays must be modulated, which is most simply done by means of a disk with an aperture rotating in front of the light source. Then the signals, as in optical telegraphy in general, will be given by interrupting the stream of light.

Fig. 9. Optical telegraph according to Manzhen31.

It is impossible to use, instead of this, the direct switching on and off of the light source, since, owing to its inertia, the signals will be indistinct. In other respects the same methods of modulation of telegraphic transmission are applied as will be indicated below for telephony. The first device for modulating a light source in the rhythm of voice vibrations belongs to Bell4 and is mentioned here for completeness. The light of an arc lamp is made to fall on a mirror membrane, which reflects it in the direction of the receiver. One speaks into the rear side of the membrane and, owing to its deformations, the divergence of the reflected pencil of rays changes in the rhythm of speech. Similar “electromechanical” types of modulation were also applied recently by Majorana30 and by Roll and Matz40. A further device suitable for practical application is the speaking arc lamp proposed by Simon59, 60, the construction of which is shown in Fig. 10.

A constant current of the battery feeding the arc is, by means of a transformer, superposed with current oscillations in the microphone circuit. In this way, in the arc flame, oscillations in volume arise in rhythm with the frequency of speech, and the flux of the emitted light proves to be modulated by the speech frequency. In the listening receiver, it causes a change in the resistance of the selenium photocell, which makes it possible to receive the transmitted speech by telephone.

Ruhmer^42,43 also describes a modulating device for a gas flame. He supplies voltage oscillations

Fig. 10. Simon’s optical telephone^59. \(F\)—arc, \(M\)—microphone, \(S_1, S_2\)—transformer, \(P\)—parabolic mirror, \(Z\)—selenium element, \(T\)—telephone.

to the flame from the microphone circuit through a transformer, one pole, connected to a platinum plate, being introduced directly into the flame, while the other is connected to the base of the Bunsen burner. However, this method did not receive practical application because of the insignificant intensity of the gas flame and the small fluctuations in intensity. Likewise, the known modulation of a flame (according to König) by speaking into a membrane that closes a gas-filled volume was not applied in practice. For transmission by light telephony and for obtaining good performance, the following conditions must be fulfilled.

1. The greatest intensity of the rays, in order to achieve the maximum range.

1. The greatest possible fluctuations of intensity, in order to achieve the best sound strength.

2. As far as possible, good proportionality between the fluctuations of the current in the lamp and the fluctuations of light caused by them.

3. Possible independence from frequency, in order to obtain a correctly sounding transmission.

To satisfy the first requirement, experience from the field of searchlight technology may be used. When a mirror is used, the energy density at the receiver can be increased by \(10^4\) times in comparison with a directed transmitter. If the receiver is also provided with a mirror, an amplification of \(10^8\) times can be obtained. For a given intensity of the lamp’s radiation, the range is essentially determined by the size and optical qualities of the searchlight employed. Ruhmer\(^{46}\) used mirrors 60 cm in diameter for the transmitter, and Simon\(^{60}\)—90 cm. If optically high-grade mirrors are available,* then one can make do with more modest dimensions, approximately 25 or 30 cm. For telephony with infrared or ultraviolet rays, mirrors with a silvered surface are necessary. In some modulating devices it is more advantageous to direct the energy of the transmitter by means of a system of lenses than by means of mirrors. Of all sources of radiation, in terms of intensity, the carbon arc is the most convenient. When Hertz-Beck carbons are used, with the same direct-current strength, an intensity of radiation 43–75% higher is obtained than when pure carbons are used. And, as indicated in Fig. 11, the maximum intensity at 75% lies at \(0.48\,\mu\).

Fig. 11. Spectral distribution of the luminosity density of a Hertz-Beck arc lamp relative to a normal arc lamp according to Däkler.

* For more detail on mirrors, see Sonnefeld\(^{62}\).

In telephone transmission, chiefly at wavelengths of 0.48 μ, an arc with these carbons gives the greatest effect in comparison with pure carbons, if the receiver consists of a photocell whose maximum sensitivity lies at the same wavelength (0.48 μ). Fig. 12 shows the spectral distribution of the sensitivity of two potassium photocells of the firm Pressler in Leipzig. In addition, the above-mentioned carbons have a relatively small luminous surface, owing to which the light energy can be rationally utilized. If, however, it is desired to use red and infrared rays as the principal waves for telephony, then Hertz-Beck carbons offer fewer advantages in comparison with pure carbons.

Fig. 12. Spectral distribution of the sensitivity of a potassium photocell of Pressler according to Zickler: I — vacuum element, II — gas-filled element.

Fig. 13. Circuit of a speaking arc according to Duddell[^10].

Fig. 14. Transmitter of optical telephony with an arc lamp according to Zickler1.

Fluctuations in the intensity of the light, which must be as large as possible in order to achieve good transmission, depend on various

factors. The decisive role here is played primarily by the amplitude of the alternating current superposed on the direct current of the arc. In Simon’s device shown in Fig. 10, the alternating current also passes through the battery and the rheostat for the arc, as a result of which its amplitude is weakened. Duddell^10 therefore modified this connection as shown in Fig. 13. In this arrangement the alternating current, for which the choke blocks the path through the battery and the resistance, passes through the condenser directly to the arc.

In modern transmitters used in optical telephony, for example those of Thirring^64 and Zickler^72, the alternating current is supplied to the arc through an amplifying device, as indicated in Fig. 14.

The original assumption made by Brown^5 and Simon^59, according to which the oscillations are reduced to oscillations of the Joule heat present in the arc \((i^2 w)\), entailed the necessity of increasing the direct current of the arc in order to increase the oscillations of intensity. Since the oscillations of light, according to this assumption, are proportional to the oscillation of heat \(dQ = 2iwdi\), Simon found an improvement in sound strength when the direct current was increased, whereas Ruhmer^46 did not confirm this observation. Systematic investigations of the dependence of variations in the light both of the crater and of the arc itself on the strength of the direct current feeding the arc for various frequencies have recently been carried out again by Jaumann^21 (see Table 3).

TABLE 3

Dependence of light oscillations on current strength

1. Crater of a Görz-Beck carbon

Current strength at a voltage of 30 V	Light oscillations at 170 Hz	Light oscillations at 1593 Hz	Light oscillations at 9190 Hz
40 A	0.0786	0.0696	0.0318
30 A	0.0693	0.0504	0.0242
20 A	0.0665	0.0385	0.0216

2. Arcs with Hertz–Becke Coals

Current strength at a voltage of 30 V	Changes in light at	Changes in light at	Changes in light at	Changes in light at
	191	1593	5630	11 270
40 A	0.0370	0.0279	0.0161	0.0115
30 “	0.0451	0.0219	0.0153	0.0111
20 “	0.0496	0.0246	0.0144	0.0103

From this table the results obtained by Jaumann are evident. When the direct current in the arc is increased from 30 A by one percent, the changes in luminous intensity take the following values (taking that for 30 A as 1).

Hertz–Becke coals

Crater

Period	169.8	1593	9190
Change	0.210%	0.915%	0.635%

Arc

Period	191	1593	5630	11 270
Change	0.42%	0.22%	0.065%	0.160%

For the general range of frequencies required for the transmission of speech, the magnitude of the light oscillations at all the frequencies investigated increases less than proportionally to the strength of the current, so that the achieved increase in the light oscillations by strengthening the direct current does not appear economically advantageous.

For the correct transmission of sound it is necessary that there be proportionality between the light oscillations and the alternating current superposed on the arc. The extent to which this is observed was systematically investigated by Jaumann. In the frequency range from 100 to 50,000 hertz he was unable to establish any systematic deviations from proportionality. Linearity was, however, maintained with deviations of up to 10%.

The next requirement for correct sound transmission is the greatest possible independence of the light oscillations from frequency. This independence was also investigated by Jaumann in the interval from 100 to 50,000 hertz, whereby he ис-

both the crater and the various regions of the arc for pure carbons and Herck-Beck carbon.

The greatest fluctuations of light for both sorts of carbons are observed in the crater. In the upper parts of the arc almost no fluctuations in luminous intensity are observed, despite the fact that they are the places of greatest intensity of the arc. However, with Herck-Beck carbons the magnitude of the fluctuation of luminous intensity reaches, in the part of the arc adjoining the crater, half the magnitude of the fluctuations of luminous intensity in the crater itself, whereas in the arc from pure carbons this magnitude is considerably smaller. The results are given in Table 4. In the second column is indicated the amplitude of the fluctuations of luminous intensity at an effective strength of the superposed alternating current of 1 A.

The magnitude of the constant light is indicated in each case, since the luminous intensity, despite the constancy of the magnitude of the strength and voltage of the direct current, probably changes as a consequence of the inhomogeneity of the carbons. This change of the constant light at the same time causes changes in the light fluctuations; however, reduction of the corresponding values to some single value of the constant light was not possible, because the regularity of these changes was not precisely established.

A better idea of the dependence on frequency can be obtained by introducing “relative fluctuations of luminous intensity”:

\[ K = \frac{ \dfrac{\text{amplitude of fluctuations of luminous intensity}} {\text{constant light}} }{ \dfrac{\text{amplitude of alternating current}} {\text{direct current}} }, \]

i.e., the dependence between the relative fluctuations of light and the relative fluctuations of current. The influence of the above-mentioned random fluctuations changed in proportion to the constant light, as happens approximately in the arc. Figure 15 presents the dependence of these “relative fluctuations of light” on frequency, and it is seen that for the arc itself this dependence has a smoother course than for the crater. A clear idea is given by the logarithmic repre-

...vibration (Fig. 16). Beginning with 1000 periods, \(k\) decreases approximately in proportion to the square root of the frequency; between 0 and 1000 this decrease proceeds still more slowly.

This comparatively slight dependence on frequency makes it possible to use the arc for the transmission of speech.

Fig. 15. Dependence on frequency of the relative oscillations of the light of the arc according to Jaumann \(^{21}\).

The unsuitability of high frequencies in transmission by means of the light arc can be mitigated by choosing the dimensions of the capacitor when connected according to Duddell’s circuit (Fig. 13), since, when this capacitance is reduced, the amplitudes of the low frequencies may be weakened. In any case, the above-mentioned dependence on frequency facilitates the transmission of speech and improves it in comparison with Taring’s incandescent-lamp transmitter \(^{64}\), in which, even with the thinnest metallic wires, the amplitude of the light variations falls with increasing frequency together with \(1/f\).

In his experiments Jaumann used as a receiv...

TABLE 4

Dependence of light oscillations on the frequency of the crater and arc of carbon arc lamps according to Jaumann

Frequency	Pure carbons — crater — constant light	Pure carbons — crater — light oscillations per 1 A	Pure carbons — arc — constant light	Pure carbons — arc — light oscillations per 1 A	Beck carbons — crater — constant light	Beck carbons — crater — light oscillations per 1 A	Beck carbons — arc — constant light	Beck carbons — arc — light oscillations per 1 A
135	0,910	0,0823	—	—	—	—	—	—
142	—	—	—	—	—	—	0,699	0,0399
191	0,943	0,0860	0,111	0,01625	2,16	0,0779	—	—
254	—	—	0,109	0,01481	2,09	0,0757	0,764	0,0456
270	0,997	0,0889	—	—	—	—	—	—
381	0,944	0,0821	0,114	0,01344	2,15	0,0749	0,682	0,0386
563	0,936	0,0804	0,117	0,01260	2,04	0,0693	0,628	0,0377
796	0,950	0,0705	0,117	0,01192	2,11	0,0641	0,775	0,0413
1127	0,940	0,0688	0,136	0,01178	2,15	0,0676	0,752	0,03805
1593	0,990	0,0545	0,131	0,00873	2,20	0,0583	0,731	0,03535
2252	1,290	0,0461	0,144	0,00789	2,−0	0,0501	0,771	0,0307
3185	1,110	0,0354	0,147	0,00754	2,18	0,0429	0,830	0,0274
4510	0,943	0,0301	0,141	0,00727	2,35	0,0396	0,780	0,01977
6500	0,889	0,0246	0,133	0,00525	2,46	0,0342	0,780	0,01508
9190	0,970	0,0218	0,144	0,00543	2,51	0,0288	0,771	0,01218
11 270	1,120	0,0245	0,144	0,00545	2,00	0,0219	0,730	0,0103
15 830	—	0,01805	0,147	0,00512	1,95	0,0164	0,916	0.0127
16 790	—	0,01725	—	—	2,05	0,0152	0,715	0.00715
22 520	—	0,01670	0,128	0,00493	—	—	—	—
29 100	—	—	0,143	0,00297	2,31	0,0201	0,650	0.0101
7130	0,967	0,0222			—	0,0	—	—
9190	1,110	0.0224			1,86	0.0178	0,620	0,00720
11 270	1,178	0,0221			2,02	0,0200	—	—
15 830	1,150	0,0195			2,25	0,0219	0,770	0,00786
22 520	1,075	0,0163			1,89	0,0145	0,805	0,00696
29 100	1,117	0,01505			1,94	0,0120	0,850	0,00673
35 600	1,158	0,0126			1,93	0,01058	0,752	0,00518
50 400	0,917	0,00931			1,98	0,01080	0,916	0,00540
					—	0,00815	—	0,00385

…of a receiver with a potassium photocell with maximum sensitivity in blue light. Experimental investigations

Fig. 16. Dependence of the relative oscillations of light on frequency (logarithmic scale), according to Jaumann.

in other spectral regions are not yet available. However, on the basis of Jaumann’s theoretical considerations one may conclude that the principal results are also valid for other regions of the spectrum, in particular for the infrared.

Along with the ordinary arc, burning at atmospheric pressure, one may also use an arc in a rarefied gas. Thus, Majorana²⁶ used a quartz mercury lamp as the transmitter.

Fig. 17. Arrangement for modulation of a mercury lamp according to Majorana²⁶.

The modulation circuit shown in Fig. 17 is assembled according to the principle of the Duddell circuit, used for an arc lamp. By means of single-stage amplification of the oscillations of the microphone current, the arc can be modula-

modulated so that its current will fluctuate between \(1/6\) and \(1/5\). This degree of modulation is the limiting one for obtaining good sound and for achieving stable operating conditions.

Of decisive importance is the gas pressure in the discharge tube; at excessively low pressure, according to Majorana, the arc is difficult to modulate. The mercury arc lamp, in comparison with the open arc lamp, has the advantage of greater constancy and greater ease of handling. It should especially be preferred in the case when one has to work with ultraviolet rays, if there is no need to attain great operating distances.

Fig. 18. Glow-discharge lamp according to Praign \(^{35}\).

Likewise, regardless of the range of wavelengths, a glow-discharge lamp is indeed suitable for use as a telephone transmitter. In this case it is possible, by suitable filling with gas, to obtain for transmission certain spectral lines, namely those to which the photoelement employed is especially sensitive. Thus, for example, Schröter \(^{53}\) used a helium lamp, whose resonance line lies in the infrared region near \(1.08\,\mu\), where the maximum sensitivity of the photoelement used by him as a receiver is approximately located. The practical use of the glow-discharge lamp for communication purposes became possible thanks to the development of lamps with high surface brightness, the so-called “light syringes.”

For comparison it should be said that a carbon incandescent lamp of \(3.5\ \mathrm{W}\) at HK has a light density equal to

71 N/cm², whereas the positive crater of a voltaic arc (with pure carbons) gives 15,500 N/cm².

Schroeter used the connection shown in Fig. 19 for modulating a helium lamp. For all the types of modulation so far enumerated it is characteristic that the intensity of the radiation source oscillates in the rhythm of speech. Therefore, as a light transmitter only such sources are suitable whose intensity of oscillation under modulation by speech currents has extremely little inertia. In ordinary incandescent lamps this requirement would be fulfilled only for very low-frequency current oscillations, approximately up to 50 hertz, so that they are not suitable for transmitting speech.

Fig. 19. Modulation of a helium lamp according to Schroeter ⁵³.

Modulation of the luminous flux can be carried out in such a way that the light source is supplied with a current of constant strength, and the change in intensity occurs only along the path of the beam. For this purpose it seems possible to use electromagnetic rotation of the plane of polarization (the Faraday effect) and the Kerr effect. In practice, up to now only the Kerr effect has been used, which makes it possible to control changes in intensity without expending additional power. Fig. 20 shows such a modulation device according to Schroeter, already known from telephotography and image transmission. During speech

Fig. 20. Modulation by means of a Kerr condenser according to Schroeter ⁵³.

in the microphone, voltage oscillations arise at the terminals of the Kerr condenser, which cause an inertia-free change in the electric double refraction in the dielectric liquid (in most cases nitrobenzene). In order to operate in the rectilinear part of the characteristic of the Kerr condenser, so that small voltage oscillations produce large changes in double refraction, it is necessary to apply an additional voltage of the order of 100 V. The inertia of this device is considerably less than that of all those described above.

The disadvantage of this device is its small light output, since in an ordinary Kerr condenser, in order to attain large fields, the distance between the electrodes must be small (several tenths of a millimeter). To eliminate this drawback, a condenser with several electrodes^20 was developed, as shown in Fig. 21.

Fig. 21. High-power Kerr condenser according to Ilberg^20.

Another type of modulation, which, unlike the Kerr cell, permits full utilization of the luminous intensity of the lamp, was proposed by the Zeiss firm in Jena^69.

From the light source a (Fig. 22), by means of the lens b, a beam of light is directed at the angle of total internal reflection onto the surface \(C_1\) of the glass body \(C\). The reflected light falls on the lens d and emerges as a parallel beam. If a second glass hemisphere is pressed against the surface \(C_1\), then the incident beam of light passes through the surface of contact without deviation, and through

...no more light passes through the lens \(d\). Between these two limiting cases, as experiment shows, it is possible continuously to vary the intensity of the light emerging from \(d\), by moving the glass body \(i\) from one extreme position to the other, i.e., from complete contact with the body \(C\) to the position in which total internal reflection occurs.

If, however, the glass body \(i\) is fastened to the diaphragm of a telephone and the magnetic windings \(h_2\) and \(h_1\), together with the microphone and the battery, are connected in a circuit in the indicated manner, then when one speaks into the microphone a modulation of the light emerging from \(a\) is obtained.

Fig. 22. Diagram of an optical telephone according to Zeiss[^69] for reflected light.

Fig. 23. Path of rays in the Zeiss arrangement for transmitted light.

The distance between \(i\) and \(c\) is chosen so that, at the maximum amplitude of the diaphragm, the body \(i\) is in complete contact with the surface \(C\), so that no light at all passes through \(d\).

Instead of the light reflected at the surface of the glass body, it is also possible to modulate transmitted light (Fig. 23). In this case the maximum intensity is obtained with complete contact of the two glass bodies, and complete extinction at such a distance that, together with \(C_1\), total internal reflection begins.

In contrast to modulation by means of a Kerr condenser, which operates without inertia up to approximately \(10^8\) hertz, the method described, in which certain (though small) masses must be moved, possesses great inertia. At speech frequencies, however, this drawback is still of little consequence. The choice of the type of transmitter and modulation in each individual case depends on the special requirements of the problem posed.

Reception of modulated radiation. At present, photoelectric cells are used as receivers of modulated radiation, and of various types, depending on the wavelength range employed. Integral receivers, such as bolometers and thermocouples, are unsuitable for receiving modulated radiation because of their inertia.

Fig. 24. Receiving device for ultraviolet telegraphy according to Zickler \(^{70}\).

Before the use of photoelectric cells for the purposes of optical telegraphy and telephony, Zickler \(^{70}\) indicated an original method of receiving secret telegraphy by ultraviolet rays; using a searchlight as the transmitter, he achieved ranges of up to \(1300\) m. This device did not acquire practical significance (Fig. 24), and it is mentioned here only for completeness. Zickler’s method is based on the fact that a spark that has not yet occurred can be induced by illuminating the spark gap \(e_1 e_2\) with ultraviolet rays.

In the rhythm of the ultraviolet telegraph signals falling on the spark gap, a spark jumps across, which can be heard with the aid of a coherer. To increase the sensitivity, the spark gap was enclosed in a closed vessel with reduced pressure (200 mm Hg). Ultraviolet light was admitted onto the elec-

through a quartz lens \(l\), through a quartz window \(p\). Such a device thus represents a primitive form of a photocell. For testing, a glass plate was placed in the path of the ultraviolet light.

A significant advance in the development of receivers for telephony became possible after the invention of amplifying cathode tubes. Before this, selenium cells alone were used. The receiving device with a selenium cell used by Simon is shown in Fig. 10. Various investigators, for their experiments, for the most part made them themselves by laboratory methods, hoping to obtain greater sensitivity and lower inertia. The quality of the cell was always determined by its efficiency and range. The greatest sensitivity of a selenium cell was achieved by Ruhmer^47; using large mirrors, he achieved transmission over a distance of up to 15 km.

When radiotelegraphy amplifiers developed into apparatus suitable for practical operation, they were also applied in phototelephony in conjunction with selenium cells and photocells.

When selenium cells are used without amplifiers, the oscillations in the circuit of the cell must be made so large as to exceed the minimum values still sufficient to actuate the telephone. By using even a single amplifier, one can get by with considerably smaller oscillations in the circuit of the cell and thereby substantially reduce the load on the cell. This gives an advantage, since with a decrease in load the noises produced by the cell decrease—noises probably caused by a constant change in resistance as a result of internal molecular rearrangement. In order to obtain distinct audibility, this internal noise must be as small as possible in comparison with the noise obtained in the telephone by means of the modulated oscillations. The latter condition will be fulfilled all the more, the weaker the intensity of the diffuse light falling on the cell and the more considerable

the amplitude of the light oscillations of the transmitter. In view of this Tiring^64 used elements placed by him in the focus of the receiving objective, with a very small receiving surface (about 1 mm). Such an element, having a resistance in the absence of illumination of about one megohm, withstands a load of only about twenty volts. Therefore Rumer, in order to obtain stronger oscillations and, consequently, a greater load, had to use elements with a larger surface. For better utilization of the incoming energy the elements were set in the mirrors extrafocally. Tiring used a four-tube amplifier. In this case the lower limit of the current

Fig. 25. Receiver of optical telephony with a photoelement according to Zickler^72.

caused by the oscillations of light must not lie too low, since otherwise audibility may be reduced owing to interfering noises of the amplifier. This occurs, according to Tiring, when the resistance of the cell exceeds approximately ten megohms.

Photoelements in conjunction with a single-tube amplifier were first used by Majorana^26 in his ultraviolet telephony. Like Tiring, he used a collecting lens in the receiver, increasing the diameter and focal length as the receiver was moved away from the transmitter.

Fig. 25 shows the receiving apparatus according to Zickler^72, with a Pressler potassium photoelement, for which the distribution of spectral sensitivity has already been given in Fig. 12. For better utilization of the photosensitive surface, photoelement 2 is set extrafocally with respect to lens L. The element is protected from extraneous light by a metal case having a tube

ГЕРГАРД ГРЕСКИ

Table

Survey of the installations used up to the present time

No.	Author	Light source	Wave region	Isolation of wave region by means of	Method of modulation	Microphone amplifier
1	Bell (4) 1880	Arc lamp	Long-wave visible light	Photoelement	(?)	—
2	Zickler (70, 71) 1898	Arc lamp 60 A	Ultraviolet $\lambda < 0.35\ \mu$	Photoelement	(?)	—
3	Simon (59, 60) 1901	Arc lamp 3 A	Long-wave visible light	Photoelement	Speaking arc	—
4	Ruhmer (46) 1902	Arc lamp	Short-wave visible light	Photoelement	Speaking arc (Duddell)	—
5	Ruhmer (47) 1904	”	Long-wave visible light	Ph.-el. (max up to $0.438\ \mu$)	”	—
6	Thirring (64) 1920	”	Long-wave visible light	Ph.-el. (max up to $0.438\ \mu$)	Speaking arc	With amplifier
7	Thirring (64) 1920	Glow-discharge lamp with a discharge of large intensity, 5 W	Long-wave visible light	Ph.-el. (max up to $0.438\ \mu$)	As with the speaking arc	”
8	Majorana (26) 1928	Mercury arc lamp	Ultraviolet up to $0.365\ \mu$	Nickel-oxide filter	Speaking arc (Duddell)	1 lamp
9	Zickler (72) 1928	Arc lamp at various current strengths	Short-wave visible light up to $0.48\ \mu$	Ph.-el. (max up to $0.438\ \mu$)	”	3 lamps
10	Majorana (27) 1920	Glow-discharge lamp, 500 W	Infrared ($\lambda \sim 1\ \mu$)	Manganese-oxide filter	Reflection from a moving mirror	—
11	Rolla and Mazza (40) 1930	Arc lamp 75 W	Infrared (from $0.8$ to $0.4\ \mu$)	Ph.-el. and filter	Telegraph speaking arc	With amplifier
12	Proter (53) 1930	Glow-discharge lamp 100 ”	Infrared	Filter	Kerr condenser	”
13	” ”	Glow-discharge lamp filled with helium	” (up to $0.8\ \mu$)	”	As with the speaking arc	”
14	Proter (56) 1930	Arc lamp 600 W	Infrared ($\lambda > 0.745\ \mu$)	Infrared filter Wratten No. 87	(?)	—

Table 5

…of installations and transmission ranges in optical telephony

Transmitting optics	Receiving optics	Receiving photoelectric cell	Amplification	Range of operation in km
Mirror or lens	Mirror	Selenium photoelectric cell	—	0.25
Metal mirror 80 cm Ø, 20 cm focal length	Quartz lens 4 cm Ø, 15 cm focal length	(?)	—	1.3
Mirror 90 cm Ø, 32 cm focal length	Lens 30 cm Ø	Selenium photoelectric cell	—	2.5
Mirror 35 cm Ø	Mirror 50 cm Ø, 7 cm focal length	Sensitized, cylindrical form. Selenium photoelectric cell, 23 mm long, 18 mm Ø	—	7.0
Mirror 60 cm Ø	Mirror 90 cm Ø	Cylindrical photoelectric cell	—	15.0
Mirror 35 cm Ø	Lens 22 cm Ø, 30 cm focal length	Selenium photoelectric cell, surface 1 mm²	4 lamps	9.0
Mirror 35 cm Ø	”	”	4 lamps	Several km
Quartz lens 50 cm Ø, 50 cm focal length	Quartz lens 50 cm Ø, 50 cm focal length	Photoelectric cell	1 lamp	16 km
Mirror of various diameters	Lens 11 cm Ø	Potassium photoelectric cell	3-fold amplification, Loewe	Calculated, see Table 1 and Fig. 6
Lens 30 cm Ø	Lens 30 cm Ø	Thallium photoelectric cell	Amplifier	10.2
Lens 30 cm Ø, lens or mirror	Lens	”	Amplifier	18; fog (medium) 15 km; fog (medium) 10 km
Lens	Mirror	Selenium-tellurium or thallium photoelectric cell	Amplifier	Not specified
”	”	”	Amplifier	Not specified
”	”	Selenium-tellurium photoelectric cell	4 lamps	28.0

in the form of a tube. The front wall of this tube is a mirror inclined at \(45^\circ\) to the optical axis and having in the center an aperture about \(1\) mm in diameter.

Observing through the upper aperture, the lens is set so that the resulting image of the transmitter disappears exactly in the aperture in the mirror \(S\), and, owing to this, the light falls on the photoelement.

Fig. 26. Optical-telephony receiver with a photoelement according to Zinckler. Electrical part.

The circuit for connecting the electrical part of the receiver is shown in Fig. 26. The amplifier used is a three-stage Loewe amplifier, with which the best possible amplification of transmission relative to the noises produced by the photoelement is achieved.

In Fig. 27 is shown a receiving device for infrared radiation, used recently by Schreter\(^{53}\), with the application of a selenium-tellurium photoelement and a concave mirror.

The amplifier leads must in all cases be correctly connected to the photo-cell used (cf. also Schreter and Ilberg\(^{55}\)).

Range. In practice, when selecting apparatus, the transmission range attained with their aid cannot always serve as a measure of suitability. The necessary expenses must also be taken into account. Thus, in many cases, and especially in the case of—

Fig. 27. Receiver for ultraviolet telephony according to Schreter\(^{53}\).

more suitable in portable apparatus, a transmitter with an arc lamp is preferable, both with respect to the energy consumed and in view of the applicability in this case of large concave mirrors.

Table 5 gives a survey of the results known in the literature and obtained up to now, and also gives the expenditures and indicates the wavelengths used. It should be noted, however, that on the basis of the data presented no comparisons can be made with respect to the range obtained, since the atmospheric conditions, which are sometimes of decisive importance, are not known in all cases.

2. Receivers of Thermal Radiation

Whereas, in the optical method of transmission, the choice of the transmitting unit has a decisive influence on the range, in receiving thermal rays the range is increased exclusively by increasing the sensitivity of the receiving device.

Since here the question is solely one of receiving invisible temperature radiation, the issue can only be the sensitivity of the receiver in the infrared region. Especially important is sensitivity to the longer thermal waves, since the radiators generally have a relatively low temperature. The inertia of the integral receivers used—bolometers and thermoelements—does not play an essential role, since in the present case only the detection of unmodulated radiation is required.

The corresponding installation is required, on the one hand, to detect the smallest possible amounts of incident energy and, on the other hand, to permit accurate determination of the direction to the radiator. For this purpose, in addition to high sensitivity of the cell, good optics are also required. In the present case the use of lenses is excluded, since the only suitable lenses here, large lenses made of rock salt and fluorite, are too expensive. It is therefore necessary to resort to mirrors with silvered surfaces; the latter is

necessary for the best transfer of energy in the given wavelength region.

The energy concentrated on the cell increases in accordance with the increase in the surface of the mirror, as a result of which mirrors with as large an aperture as possible should be used. The mirrors have to be given a parabolic surface, since spherical mirrors of high luminosity do not give a good image.

In general, for equal diameters, a mirror with a greater focal distance is superior in optical qualities to a mirror with a short focus, which is especially clearly manifested for mirrors of very large diameter. With equal optical qualities for mirrors of different sizes, the sensitivity increases in proportion to the surface of the mirror. Table 6 gives a comparison of parabolic mirrors with a silvered surface, having different diameters and focal distances, according to data obtained by the author.

TABLE 6

Mirror No.	Diameter	Focal distance	Ratio of amplifications (calculated according to surface ratio)	Ratio of amplifications (measured)
1	250	140	1	1
2	350	150	1.95	2.47
3	600	250	5.74	4.52
4	600	480	5.74	8.20
5	1180	480	22.20	11.0

The choice of receiver is determined by the surface sensitivity of the cell and by the size of the collecting surface. The relative surface sensitivity of two bolometers or thermocouples is found from the deflections obtained on the galvanometer when equal surfaces of both cells are illuminated by rays of identical intensity. Such a comparison shows that a bolometer with a smaller collecting surface has, in general, greater surface sensitivity. Thus a receiving device equipped with a bolometer with a smaller collecting surface-

ness, is the most sensitive, provided that the image obtained of the emitter does not exceed the collecting surface of the bolometer.

In addition, this achieves great sharpness of pointing. In cases where considerable sharpness is not required, in order to make it easier to locate the beam it is advisable to use bolometers with a larger receiving surface.

For measurements requiring great sensitivity and accuracy, direct-current galvanometers should be used. The most suitable is the Zeyss loop galvanometer, which has sufficient sensitivity and, owing to its low internal resistance, works well with these photocells. An essential advantage of it is good damping, which makes it possible to work with it even during rapid movements of the mirror. In some cases it is preferable to point at the emitter, using a telephone as the indicator. When working with a bolometer, in this case one should use exclusively an alternating-current bridge; moreover, it is advisable to include an amplifier in the corresponding branch before the telephone. When using a thermoelement, attempts were made, in order to make auditory operation possible, to introduce a mechanical interrupter.

3. Other applications

A further application, when suitable sources of infrared rays are found, thermal seekers may find in navigation on water and in the air. Thus, for example, one can equip the entrance to a harbor, or else flight routes, with similar infrared beacons and carry out the detection of these beacons in weather with poor visibility, especially in fog, by means of thermal seekers. Short-wave infrared rays are especially suitable for this purpose, since in this case a photoelement can also be used as the receiver. It is advisable, for recognizing the beacons, to modulate their radiation with an audio frequency.

The first experiment in applying this idea for purposes of aerial

was made in navigation by K. Müller[^31][^38], who improved apparatus for automatic direction finding. The purpose of such an instrument is the continuous monitoring of the course when moving in fog along an air line illuminated by signals (infrared beacons), and the instantaneous indication of deviations from the course. For this purpose a rotating concave mirror is used, with a thallofide photoelement placed at its focus, sensitive to infrared rays, which during flight, as it were, probes the terrain on both sides of the airplane within a certain angle. The cell is supplied with alternating current in such a way that during the positive half-period the mirror scans along the terrain situated on the right side of the aircraft, and during the negative half-period—on the left. The current from the cell, after suitable amplification, is fed to a direct-current instrument. The instrument will not indicate deviations if the areas on both sides of the aircraft have the same brightness. But if in one of them there is an additional source of radiation, for example an infrared beacon, then the instrument will show a deviation in a definite direction, which can be used to determine the direction of flight. With the help of a second system, having oscillation in a perpendicular plane, one can, along with orientation in the direction of flight, also achieve orientation in altitude.

In all the applications of heat rays set forth so far, for the purposes of servicing communications, the matter concerned the search for an invisible radiator and the observation of changes in its position. It is also possible, under appropriate experimental conditions, by means of infrared rays to prove the presence of non-radiating bodies. For this purpose the given body is intersected by a beam of rays directed from the transmitter to the receiver, and one observes the change in the current obtained from the receiving cell. The indicated changes can be used to actuate a relay, an alarm signal, and the like.

In a similar manner it is also possible to make use of this for guarding the entrance to a harbor in darkness and in fog, or for signaling for the purpose of protecting premises.

Conclusion

The development of light telephony in recent times, owing to the introduction of cathode lamps and the improvement of receiving cells, has led to apparatuses practically capable, in some respects, of competing with electrical short-wave installations. By utilizing the infrared spectrum for the transmission of information, the aim once pursued was to achieve completely invisible transmission and to attain range through fog. The advantages that will be found in the use of longer waves than those employed up to now cannot, for the time being, be made use of for telephony, since as yet there are no non-inertial receiving cells suitable for this region of waves. The question also still comes up against the impossibility of rationally using cold temperature radiators as transmitters, to which the described methods of modulation cannot be rationally applied.

Literature

1. S. H. Anderson, The penetration of light through fog. The Aviation 38, 930—936, 1930.

2. O. v. Auwers u. Kerschbaum, Über Sperrschichtphotozellen, Ann. d. Phys. 7, 129/175, 1930.

3. G. G. Barnard, The selenium cell, its properties and application, London 1930.

4. Bell, Proc. Am. Assoc. Adv. Scient. S., 115—136, 1880.

5. F. Braun, Notiz über Thermophonie, Ann. d. Phys, 65, 358—360, 1898.

6. T. W. Case, Phys. Rev., 15, 289—292, 1920.

7. W. W. Coblentz, Wien-Harms, Handb. Exp. Phys, 9, i. T. S. 411.

8. L. H. Dawson, L. P. Granath a. E. O. Hulburt, The attenuation of ultraviolet light by lower atmosphere, Phys. Rev., 34, 136—139, 1929.

9. P. Debye, Der Lichtdruck auf Kugeln von beliebigem Material Ann. d. Phys., 30, 57—136, 1909.

10. W. Dudell, The Electrician, 46, 269—273, 310—313, 1900.

11. A. Esau and W. Hahnemann, Report on experiments with

electric waves of about 3 meters: their propagation and use, Proc. I, R. E. 18, 471—489, 1930.

1. M. Fritz-Schmidt, G. Gehlhoff u. Thomas, Die physikalischen Eigenschaften der Gläser in Abhängigkeit von der Zusammensetzung IV. Die Durchlässigkeit der Gläser im ultraviolettensichtbaren und ultraroten Gebiet, Z. tech. Phys., 11, 289—326, 1930.

2. G. Gehlhoff u. H. Schering, Über ein neues Photometer sehr hoher Empfindlichkeit und einige Anwendungen, Z. tech. Phys., 1, 247, 1920.

3. W. Graffunder, Über den lichtelektrischen Effekt bei Trockengleichrichtern, Physik Z., 31, 375. 1930.

4. L. P. Granath u. E. O. Hulburt, The absorption of light by fog, Phys. Rev., 34, 140, 1929.

5. G. Gresky, Verkehrssicherung und Nachrichtenübermittlung mit Ultrarotstrahlen, Radio-Helios 7, 69, 1930.

6. G. Gresky, Wärmestrahlungsmessungen mit verschiedenen integrierenden Empfängern mit und ohne Hohlspiegeln.

7. A. Gudden, Lichtelektrische Erscheinungen, Berlin 1928.

8. R. Hennig, Die Entwicklung der Telegraphie und Telephonie, Wissen und Können, Leipzig 1908.

9. W. Ilberg, Die Verwendung trägheitloser Lichtrelais beim Fernsehen, Fernsehen, 1, 97, 1930.

10. J. Jaumann, Die Frequenzabhängigkeit der Lichtschwingungen des tönenden Lichtbogens, Physik Z., 59, 386, 1930.

11. B. Lange, Über eine neue Art von Photozellen, Physik. Z. 31, 139, 1930.

12. B. Lange, Über eine neue Art von Photozellen (2 Mitteilung), Physik. Z. 31, 964, 1930.

13. F. Lecomte, Le spectre infrarouge, Paris 1928.

14. Q. Majorana, Telefonia ottica mediante radiazioni ultraviolette ad ultrarosse, Linc. Rend., 9, 924, 1929.

15. Q. Majorana, Attenuazione dessimetrica di energia raggiante in un mezzo assorbente. Atti di Torino, 64, 359, 1928/29.

16. Q. Majorana, Telefonia ottica mediante radiazione visibili ad invisibili, Rend. di Bologna (NS), 31, 118, 1926/27.

17. Q. Majorana, Su di un nuovo metodo di telefonia ottica con luce ordinarion i con luce ultraviolette, Linc. Rend, 5, 726, 1927.

18. Q. Majorana e Todesco, Preparauine delle cellule fotoelettriche al tallio, Linc. Rend., 8, 9, 1928.

19. Q. Majorana, Dalla radiotelefonia alla telefonia ottice invisibile Cim (NS), 6, 9, 1929.

20. A. Mangin, Mémoire sur le système de télégraphie optique de la défense der Paris, Etude de divers dispositifs optiques destinés à projeter la lumière électrique sur les objets éloignés, Ostwalds Klassiker, 219, übersetzt von A. Sonnefeld.

21. Michelsen, Infrarotempfindliche Zellen, Z. techn. Phys., 11, 511, 1932.

1. G. Mie, Beiträge zur Optik trüber Medien, speziell kolloidaler Metallösungen, Ann. d. Phys., 25, 377, 1908.

2. C. Müller, Vergl. Tätigkeitsber. d. Phys. Techn. Reichsanst. 1929, Z. Einstr. K., 50, 356, 1930.

3. M. Pirani, Fortschritte und Entwicklungsmöglichkeiten auf dem Gebiete der Leuchtröhren, E. T. Z., 51, 889, 1930.

4. F. I. G. Rawlius and A. M. Taylor, Infrared analysis of molecular structure, Cambridge 1929.

5. Reichinstein, Der galvanische Polarisationseffekt der festen lichtelektrischen Selenzellen, Naturw., 18, 685, 1930.

6. Chr. Ries, Das Selen, München 1918.

7. Rothaas, Ultrarote Strahlen, V. D. I. Z., 73, 1859, 1929.

8. L. Rolla and L. Mazza, Sistemi di telegrafia e di telefonia per mezzo di fasci di radiazione infrarosse, Linc. Rend., 11, 19, 1930.

9. A. Rudolph, Über die Durchlässigkeit von Nebel für Lichtstrahlen von verschiedener Wellenlänge, Physik. Z., 5, 36, 1904.

10. E. Ruhmer, Neue Sende- und Empfangsanordnung für drahtlose Telephonie, Physik Z., 2, 339, 1901.

11. E. Ruhmer, Über das sprechende Licht, Physik Z., 2, 325, 1901.

12. E. Ruhmer, Der sprechende elektrische Flammenbogen und seine Verwendung zur drahtlosen Telephonie, E. T. Z., 22, 196, 1901.

13. E. Ruhmer, Über die Empfindlichkeit und Trägheit von Selenzellen, Physik, Z., 3, 468, 1902.

14. E. Ruhmer, Neuere Versuche mit Lichttelephonie, E. T. Z., 23; 859, 1902.

15. E. Ruhmer, Das Selen und seine Bedeutung für die Elektrotechnik mit besonderer Berücksichtigung der drahtlosen Telephonie, Berlin 1902 and E. T. Z., 25, 1021, 1904.

16. I. Runge and Sewig, Über den inneren Photoeffekt in kristallinen Halbleitern, Physik Z., 62, 726, 1930.

17. Cl. Schäfer and F. Matosse, Das ultrarote Spektrum, Berlin 1930.

18. G. Schmidt, Elektrische Sicherheitsvorrichtungen gegen Einbruch und Feuer, Elektrot. Anz., 47, 816, 1930.

19. W. Schottky, Über den Entstehungsort der Photoelektronen in Kupfer-Kupferoxydul Photozellen, Z. tech. Phys., 11, 358, 1930.

20. W. Schottky, Über den Entstehungsort der Photoelektronen in Kupfer-Kupferoxydul Photozellen, Physik Z., 31, 913. 1930.

21. R. Schröter, Hertzsche und infrarote Strahlen als Nachrichtenmittel, E. N. T., 7, 1, 1930.

22. F. Schröter and G. Lubszynsky, Untersuchung zur Frage der Trägheit gasgefüllter Photozellen, Physik. Z., 31, 897, 1930.

23. F. Schröter and W. Ilberg, Vergleich von Selen und Photozellen, Physik Z, 30, 801, 1930.

24. F. Schröter and F. Michellsen, Infrared sensitive cells; from: Photo-electric cells and their applications, London 1930.

1. R. Seeliger, Die Physik der Gasentladungen, Leipzig, 1928.

2. R. Sewig, Die lichtelektrischen Eigenschaften der Thalliumzellen, Z. tech. Physik, 11, 269, 1930.

3. H. Th. Simon, Über den sprechenden Flammenbogen und seine Verwendung zu einer Telephonie ohne Draht, Physik Z., 2, 253, 1901.

4. H. Th. Simon and M. Reich, Flammen und Flammentelephonie, Physik Z., 3, 278, 1901/02.

5. H. Th. Simon, Mörende Flammen und Flammentelephonie, E. T. Z., 22, 510, 1901.

6. A. Sonnefeld, Die Hohlspiegel, Berlin, 1925/26.

7. U. A. Stratton, The effect of rain and fog on propagation of very short radio waves, Proc. I. R. E. 18, 1064, 1930.

8. Thirring, Neue Apparate für Lichttelephonie, Physik Z., 21, 67, 1920.

9. W. Voege, Neue Apparate zur Strahlungsmessung, Physik Z., 21, 288, 1920.

10. W. Westphal, Wärmestrahlung, Handb. d. Astrophysik., 3, Berlin, 1930.

11. E. Wiegand, Widerstandsvergleichung mit Wechselstrom unter Verwendung von Verstärkerröhren, Ann. d. Phys., 4, 781, 1930.

12. W. Wien, Wärmestrahlung, Wien-Harms, Handb. d. Exp. Phys., Leipzig, 1929.

13. C. Zeiss, Improvements in Apparatus Effecting Fluctuations of the Intensity of a Pencil of Light Rays, Engl Patentanm., 317, 318, v. 15, V. 30.

14. K. Zickler, Lichtelektrische Telegraphie, E. T. Z., 19, 474 and 487, 1898.

15. K. Zickler, Weitere Versuche über die lichtelektrische Telegraphie, E. T. Z., 19, 826, 1898.

16. K. Zickler, Das Problem der Phototelephonie, E. u. M., 46, 769 and 793, 1923.

1. 72. ↩↩