MARINE UNDERWATER SIGNALING AND SOUND TRANSMISSION UNDERWATER1

K. V. Drysdale.

Almost all branches of hydromechanics have important technical applications, and only acoustics, until recent years, remained a purely academic discipline to which only a few scientists devoted serious attention. Bells, gongs, whistles, sirens, and musical instruments have been used since ancient times both for entertainment and for signaling, but their design and improvement proceeded almost always empirically and with little participation by physicists.

The war brought sharp changes in this field, as in many others. Acoustics is now not only an important branch of communications engineering, but is beginning to be applied even in engineering practice, providing a new means of transmitting force. This may be judged from the work of M. Konstantinesco, who applies the acoustic method for drilling rock and for riveting, and shows how it can be used to set motors and other mechanisms in motion. Few fields of science are now opening so broad a field for new inventions as acoustics.

Acoustic signaling is especially important for navigation, since sound is the only form of energy that can be transmitted through water without significant absorption. The relatively high electrical conductivity of water makes it almost opaque to light and to electric waves.

The present article sets forth the principles of acoustic signaling underwater, and also briefly touches on certain other problems important for navigation, such as, for example, determining the position of sound sources and determining depths.

As is well known, sound consists of oscillatory disturbances in material media—gases, liquids, or solid bodies—and is a process of purely mechanical nature. A sounding

the body, vibrating, produces rapidly successive compressions and rarefactions of the surrounding medium, which propagate in all directions with a certain velocity. It is important to note that, in the propagation of sound in liquids and gases, the motion of the particles is always longitudinal with respect to the direction of the wave, in consequence of which sound does not exhibit the phenomena of polarization characteristic of light.

Fundamental Properties of Sound

Sounds are divided into musical tones and noises. Musical tones are fully characterized by the number of vibrations of the fundamental tone, the amplitude of vibration, and the form of the curve; these three factors determine the pitch, intensity, and timbre of the sound.

Noises differ from musical tones by the absence of a regular character of vibrations and consist of a series of vibrations of various pitch and intensity. From the point of view of acoustic transmission and reception, speech consists to a considerable extent of noises, since the vibrations characterizing speech are in many cases irregular in character, for example, all consonants. Sounds produced by motors, ships, etc., also have the character of noises. The detection and recognition of such sounds presents considerable difficulties, since almost all transmitting and receiving apparatus have a more or less “selective” character, i.e., they respond more strongly to certain definite frequencies and are relatively insensitive to others. Everyone knows that the telephone and the gramophone reproduce some sounds more loudly than others; in acoustic signaling, as in radiotelegraphic transmission, it is considerably more advantageous to use “tuned” systems, but in that case the amplification of certain frequencies is always obtained at the expense of the weakening of all the others.

Velocity of Propagation

Exact knowledge of the velocity of propagation of sound is very important for acoustic signaling, especially in questions of determining position and distances. The velocity of sound in different media is very different, since it depends on the density and elasticity of the medium and, in addition, on its composition, pressure, and temperature. Here we shall dwell mainly on the velocities of sound in air and in sea water, although in calculating transmitting and receiving apparatus it is also necessary to know the acoustic properties of other substances.

For air at a temperature of \(t^\circ C\), the velocity of sound is equal to

\[ v = 331.3 + 0.552\,t \ \frac{\mathrm{m}}{\mathrm{sec}}. \]

For sea water, according to the latest measurements of A. Wood,

\[ v = 1449.6 + 4.21\, t + 0.037\, t^2 \ \frac{m}{sec}, \]

with a salt content of 35 parts per thousand; the velocity increases by \(1.13 \frac{m}{sec}\) when the salinity is increased by 1 per mille.

Thus for the speed of sound at \(20^\circ\) in air we have \(342.5 \frac{m}{sec}\), and for sea water of normal composition—\(1520 \frac{m}{sec}\), that is, the speed in sea water is approximately 4.5 times greater than the speed in air.

For the wavelength at a frequency of 500 periods we obtain in air \(\lambda = 68.6\) cm, and in sea water: \(\lambda = 303.8\) cm.

As we shall see below, the transmission of long waves in water presents considerable difficulties.

PROPAGATION OF SOUND THROUGH VARIOUS MEDIA.

As Newton first showed, the speed of sound in a given medium can be calculated if its bulk elasticity \(k\) and density \(\rho\) are known. It is easy to derive that the speed of propagation of sound will be

\[ v = \sqrt{\frac{k}{\rho}}. \]

But if one takes into account that, during rapid vibrations of sound frequency, the local heatings arising upon compression and the coolings upon expansion do not have time to dissipate, then in the formula written it is necessary to introduce the adiabatic elasticity instead of the elasticity \(k\) at constant temperature, i.e. the isothermal one. For gases at normal atmospheric pressure the isothermal elasticity is equal to the pressure, i.e. about \(10^6\) dynes per \(cm^3\); the adiabatic elasticity is 1.41 times greater. The density of air is \(0.00129\) g \(cm^3\). Thus the speed of sound in air will be:

\[ v = \sqrt{\frac{1.41 \cdot 10^6}{0.00129}} = 33003 \ \frac{cm}{sec}, \]

which agrees closely with the value found experimentally.

Theory makes it possible to calculate the amount of sound energy carried by a sinusoidal wave, and also to clarify the phenomena occurring when sound passes from one medium into another, which is of primary importance in the question of underwater signaling. It can be shown that between the pressure \(P\), caused by the vibrations (the variable excess over the mean pressure), and the velocity \(v\) of the moving particle at a given point of the medium there exists the relation \(P = Rv\),

where \(R=\sqrt{k\rho}\). This relation is similar to Ohm’s law in electricity, and the quantity \(R\), after Brillouin1, is called the “acoustic resistance” of the medium. The energy \(w\) that has passed through a unit area of the wave front will be:

\[ \frac{1}{2} P_{\max} V_{\max}, \text{ i.e. } \frac{1}{2R} P_{\max}^{2} \text{ or } \frac{R}{2} V_{\max}^{2}. \]

For a sinusoidal wave with frequency \(n\) periods per second, i.e. with angular frequency \(\omega=2\pi n\), we have \(V_{\max}=\omega a\), where \(a\) is the amplitude of displacement; thus we obtain \(w=\frac{\omega}{2}aP_{\max}=\frac{R}{2}\omega^{2}a^{2}\) ergs through \(\text{cm}^{2}\) per second.

For ordinary seawater, for which \(k=2.2\cdot 10^{10}\) dyn/cm\(^2\) and \(\rho=1.028\), \(R=\sqrt{k\rho}=24\cdot 10^{4}\), so that for a frequency of 500 and with an amplitude of \(0.1\) mm, the power flowing through \(1\ \text{cm}^{2}\) will be 7 watts.

When sound passes from one medium into another, it can be shown that, for different acoustic resistances of the two media, a certain reflection will always occur at the intermediate surface. If by \(r\) we denote the ratio of the acoustic resistance of the second medium to the resistance of the first, then we obtain the following expressions for the pressures and particle velocities at the boundary of the first and second media:

\[ P_{2}=\frac{2r}{r+1}P_{1};\quad P'_{1}=\frac{r-1}{r+1}P_{1};\quad V_{2}=\frac{2}{r+1}V_{1};\quad V'_{1}=\frac{r-1}{r+1}V_{1}, \]

where \(P_{1}\) is the pressure and \(V_{1}\) the particle velocity in the original wave;

\[ \begin{array}{cccccc} \text{“} & P_{2} & \text{”} & \text{”} & V_{2} & \text{”} \quad \text{in the transmitted wave} \\ \text{“} & P'_{1} & \text{”} & \text{”} & V'_{1} & \text{”} \quad \text{in the reflected wave.} \end{array} \]

If the second medium has a high resistance in comparison with the first, i.e. \(r\) is very large, then

\[ P_{2}=2P_{1};\quad P'_{1}=P_{1};\quad V_{2}=0;\quad V'_{1}=V_{1}, \]

i.e. the pressure at the boundary surface is equal to twice the pressure of the original wave, while the velocity, equal to \((V_{1}-V'_{1})\), is equal to 0, since the motions in the incident and reflected waves are equal and occur in opposite directions. Thus, the wave is wholly reflected back into the first medium; no passage of sound into the second medium occurs.

On the contrary, if the second medium has a very small acoustic resistance in comparison with the first, i.e. \(r\) is very small, then \(P_2=0\); \(P_1'=-P_1\), \(V_2=2V_1\), and \(V_1'=V_1\). In this case the total pressure at the boundary \((P_1+P_1')=0\), and the velocity \((V_1-V_1')=2V_1\), but here too there is again no transmission of sound into the second medium, since \(P_2=0\); the wave is wholly reflected, with the direction of the velocity of the particles reversed. In the first case the boundary surface is analogous to the “fixed” end of a vibrating string or rod, or to a closed pipe; in the second, to a “free” end or an open pipe.

If the two media have equal acoustic resistances, i.e.

\[ r=1,\quad \text{then: } P_2=P_1;\; P_1'=0,\; V_2=V_1;\; V_1'=0; \]

the wave passes into the second medium without reflection. This case is ideal both for a receiver and for a transmitter of sound.

If the acoustic resistances of two media are not equal, then it is easy to show that the ratio of the energy in the transmitted wave to the initial energy, which may be called the coefficient of sound penetration \((\eta)\), will be:

\[ \eta=\frac{4r}{(r+1)^2}; \]

for the reflected wave the ratio of energies will be:

\[ \left(\frac{r-1}{r+1}\right)^2. \]

We have seen that for water \(R_1=14\cdot 10^4\), and for air \(R_2=40\) (see table on p. 211), so that:

\[ r=\frac{R_2}{R_1}=2.86\cdot 10^{-4} \]

and

\[ \eta=\frac{4r}{(r+1)^2}=0.0011, \]

i.e. a little more than \(0.1\%\) of the energy penetrates from water into air. This immediately indicates the difficulty in underwater listening, since a sound arriving through the water, before reaching the eardrum, must pass from water into air.

For steel \(R \approx 395\cdot 10^4\), so that in passing from water into steel \(r\) is approximately equal to 28, and the coefficient of penetration will be about \(13\%\), whereas from steel into air it will be only \(0.004\%\). Thus, for sound coming from water through the wall of a ship into ...

internal air, the transmission coefficient will be 13% of 0.004%, i.e. 0.00052%; here it is assumed that the ship’s walls are not sufficiently thin to serve as a membrane, which, of course, gives greater sound transmission than with thick walls. Thus, in some cases the loss of energy is very great, and this has led to shipboard listening devices having to be attached either directly to the ship’s hull or in special tanks with water, fastened to the hull from the inside, as will be described below.

The following table of the acoustic properties of various media is taken from Brillouin’s work.

Medium	Bulk elasticity \(k\) per mm\(^3\)	Density in C.G.S. units	Velocity \(v=\sqrt{\dfrac{k}{\rho}}\) m/sec	Acoustic resistance \(R=\sqrt{k\rho}\), C.G.S. units
Steel	\(2.10^4\)	7.8	5100	\(395.10^4\)
Cast iron	\(0.95.10^4\)	7.0	3680	\(258.10^4\)
Brass	\(0.65.10^4\)	8.4	2780	\(234.10^4\)
Bronze	\(0.32.10^4\)	8.8	1910	\(168.10^4\)
Lead	\(0.06.10^4\)	11.4	725	\(82.5.10^4\)
Wood — teak	\(0.16.10^4\)	0.86	4300	\(37.10^4\)
Wood — pine	\(0.09.10^4\)	0.45	4470	\(20.10^4\)
Wood — beech	\(0.06.10^4\)	0.80	2740	\(22.10^4\)
Water	\(2.10^2\)	1.0	1410	\(14.10^4\)
Rubber	\(<1\) (depends on the grade)	1 (approx.)	\(<100\)	\(<1.10^4\)
Air	\(1.40.10^{-2}\)	0.0013	328	\(0.004.10^4\)

It should be noted that the magnitude of the acoustic resistance for pine or beech is close to its magnitude for water and, consequently, sound passes from water into wood and back without strong reflection.

Receivers Perceiving Pressure and Displacement

From the theory of the propagation of oscillations set forth above, it is evident that sound can be detected either by the changes in pressure that it produces, or by the displacements that it produces; in exactly the same way, a source of electricity can be detected either by the electric voltage or by the current (displacement) that it produces. Acoustic receivers, therefore, may be divided into receivers detecting pressure, analogous to electrical voltmeters, and receivers detecting displacement, corresponding to ammeters.

This subdivision cannot be considered strictly scientific, since a receiver cannot be excited by pressure alone or by displacement alone. We have seen that the energy passing per second through a unit area of the wave front is equal to \(\frac{1}{8}P_{\max}V_{\max}\), whence it is clear that, for the perception of energy, it is necessary that the receiver perceive both the pressure and the velocity of displacement. An ideal pressure receiver would in reality be a reflector with a fixed wall, while an ideal displacement receiver would be a reflector with a free wall; and we have seen that in both of these cases no energy is transmitted, but all of it is reflected back.

However, the distinction between pressure and displacement receivers may be considered useful, just as the distinction between voltmeters and ammeters is useful. A voltmeter is chiefly an instrument for measuring electric voltage, although it requires an insignificant current. An ammeter is chiefly an instrument for measuring current, although it requires a small voltage drop. Similarly, a pressure receiver has a relatively rigid membrane, which undergoes only small displacements under the action of oscillations; conversely, a displacement receiver is provided with a very yielding membrane. The distinction between the two kinds of receivers plays a large role in the construction of directional receivers, in view of the fact that pressure oscillations in a homogeneous medium are identical in all directions, whereas displacements occur along the line of wave propagation. In this case a pressure receiver, when rotated in different directions, gives no difference, whereas a displacement receiver will show a maximum when it is directed straight toward the source of sound.

It is obvious that the best results with respect to the sensitivity of a receiver are obtained under the condition of complete absorption of all the energy incident upon it, and this can occur only if the given variable pressure on the membrane produces the same displacement as takes place in the external medium; that is, the energy passes wholly into the receiver without being reflected.

Questions concerning the construction of underwater receivers and hydrophones will be examined further on, but here it will be appropriate to dwell on the basic ideas underlying their design. The simplest of such receivers, analogous to a simple ear trumpet, is a long metal tube with a membrane at the lower end; it is often called a Broca tube. If such a tube is immersed in water, then sound from the water is transmitted through the membrane into the air inside the tube, and the observer hears it at the free end of the tube. The receiver absorbs rather much energy, but is not very sensitive, since it gives no possibility of amplifying the sound, and is not especially convenient, since listening through long bent tubes is difficult and possibly

most, by several meters under water. Therefore, in all modern hydrophones the sound is first converted into the form of electrical oscillations by means of microphones or magnetophones, from which the current is led to an ordinary telephone receiver placed in a convenient listening position.

In view of the great sensitivity, microphones are usually employed. Corresponding to pressure and displacement receivers there are two types of microphones. The first of these, the so-called microphone with a solid backing (solid back type), is one in which a certain number of carbon grains are enclosed between a metal or carbon plate forming the membrane (or attached to the membrane) and a massive, immobile piece of carbon lying behind it. When the membrane is subjected to pressure, it compresses the carbon grains and increases their conductivity, as a result of which the battery gives a larger current through the microphone and the telephone; the sound is reproduced in this case thanks to changes in pressure. In a microphone of the “button” type (button type1) the carbon grains, on the contrary, are enclosed in a light metal box or capsule, covered by a small membrane, and this entire system is fastened to a large membrane; the oscillations of this membrane set the whole capsule in motion and carry with it the carbon grains, which are thereby subjected to pressure owing to the inertial forces that arise. In this case changes in the resistance of the microphone are caused by the motions or displacements of the membrane.

The usual type of simple hydrophone is shown schematically in Fig. 1; its photograph is given in Fig. 18. It consists of a heavy case in the form of a round hollow disk; the internal cavity is closed by a metal membrane, at the center of which a small “button” microphone is fastened. This hydrophone is very sensitive, but has no directional action.

Fig. 1.

Directional reception and transmission.

The problem of directional transmission and reception in acoustic signaling is a very important one. In acoustic transmission, as in radiotelephony and telegraphy, a great difficulty lies in the fact that sound waves, like electromagnetic waves, tend to propagate more or less uniformly in all directions, as a result of which their intensity rapidly decreases, according to the law of inverse proportionality to the square of the distance; moreover, always

possibly interfering action of various stations is possible, and the secrecy of transmission is not ensured. In reception, even a sensitive apparatus that detects the presence of sound at great distances is of relatively little value if it gives no indication of the direction or position of the source of the sound. This shows that the question of directional transmission and reception is of just as much importance as the question of designing powerful transmitters and sensitive receivers. It is therefore clear that directional transmission and reception are the subject of exceptional attention.

Fig. 2.

Binaural, or two-ear, method of directional listening.—Our ears constitute an extremely perfect system of directional reception. When we hear a sudden noise, we instinctively turn toward the direction of the sound, and if we were blind we would nevertheless be able, with a considerable degree of accuracy, to determine the direction from the sound. This ability is due to the fact that our two ears are situated on opposite sides of the head at a distance of about 15 cm from one another, and the sound reaches one ear sooner than the other, provided only that the sound source is not in the plane perpendicular to the line connecting both ears, i.e., in front, behind, or above. Our ears are extremely sensitive to an insignificant difference in time, and since the interval

time depends on direction, becoming the longer the more the source lies to the side, then we are able to determine the direction surprisingly accurately—provided, of course, that both ears are equally sensitive. This property of the ears underlies the so-called binaural method of determining direction, developed both for aerial and for underwater reception. For example, if two funnels (Fig. 2) are fastened to a horizontal bar and a rubber tube of definite length is led from each of them to the ear, then we can detect and determine the direction of an airplane from the noise of its motor with considerable accuracy, since the funnels amplify the sound, and the sensitivity of determining direction can be increased by increasing the distance between the funnels. When a sound is heard and the observer turns the bar carrying the funnels in the direction indicated by the arrow, it then seems to him that the source of the sound passes from one ear toward the other and, finally, becomes established behind the head. This position is called the position of binaural equilibrium, and when such equilibrium has been attained, the bar will stand at a right angle to the direction of the sound source.

The same principle, of course, can also be applied to underwater listening by means of two receivers; but in this case, taking into account that the speed of sound in water is four and a half times greater than in air, the distance between the receivers must be increased in the same ratio in order to obtain the same time difference and, consequently, the same discrimination of directions as with the aid of two ears. These considerations compel the use of a rather large distance between the two receivers, and rotating them underwater becomes difficult; to overcome this difficulty in determining direction, the so-called binaural compensator is used.

Fig. 3. Binaural method with a rectifying compensator.

Returning to our pair of funnels in Fig. 2, let us suppose that the sound source is to the right of the median plane, and that the tubes from the funnels are made not of equal, but of different length, calculated so that the added length of the tube from the right funnel is equal to the excess of the distance from the sound sources to the left funnel. It is clear that in this case the delay of the sound in reaching the more distant left funnel is balanced by the additional delay obtained...

thanks to increasing the length of the tube between the right funnel and the ear, and, despite the fact that the source of the sound lies to the side of the median plane, binaural equilibrium will occur. Thus, the direction of the sound source can also be determined with the two receivers in a fixed position by introducing an additional device that makes it possible to vary the length of the auditory tube. Such a device is called a binaural compensator; its simplest form is shown in Fig. 3. Two equal tubes from the receivers are brought to the two ends of a long straight tube \(AB\), consisting of three parts, of which the middle one can slide farther into one or the other of the outer ones. The middle part of the tube

Fig. 4a. American compensator. General view.

is interrupted at the center and, through branching rubber tubes of equal length, is connected with the right and left ear. When the central part of the tube is in the middle position, the paths of the sound to both ears are equal, and binaural equilibrium is obtained only if the sound source lies in the median plane; if the source lies to the right of the median plane so that the sound reaches the right funnel earlier, then, by moving the central part of the tube to the left, we increase the path of the sound from the right funnel and decrease the path from the left, and can restore equilibrium. If the apparatus is furnished with a specially calculated scale, then from the reading of the divisions of the scale one can immediately obtain the direction of the sound source; it is found from the relation \(2d = b . \sin \theta\), or \(\sin \theta = \frac{2d}{b}\), where \(b\) is the distance

between the horns, \(d\) is the displacement of the central tube from the middle position, and \(\theta\) is the angle between the direction toward the sound source and the median plane.

For the greatest convenience in determining direction by this method, the American firm Automatic Telephone Company has constructed a circular compensator. Fig. 4a shows the general view of this compensator; Fig. 4b explains its construction. In a fixed circular plate two concentric circular grooves are cut, which, when closed from above by a second plate, are transformed into two circular passages. The upper plate can rotate relative to the fixed lower one; it is provided with two shutters, which divide the circular grooves into two non-communicating parts, while at the same time cross-passages are arranged in the shutter, establishing communication between the inner and outer parts of the groove both on the right and on the left side. Sound from the two horns, entering the two ends of the outer groove, passes through them to the shutter, then passes through the cross-passages into the inner groove and proceeds to its ends, to which are attached tubes leading to the ears. It is evident that, when the upper plate is rotated, the difference in the sound paths changes by an amount four times greater than the distance through which the shutter has been moved; the position of the pointer on the scale directly indicates the direction toward the sound source.

Fig. 4b. Principle of construction of the American compensator.

The binaural method is of very great importance, and below we shall see its numerous applications, which is why we have dwelt in such detail on the description of the method. However, there exist other methods of directional reception, which must also be mentioned.

Method of sum and difference.—In the case of electrical receivers, the binaural method may be replaced by the so-called method of sum and difference. Suppose that our two horns on the bar are replaced by two ordinary, completely identical microphone receivers (Fig. 5), \(M_1\) and \(M_2\), and that these receivers are connected with two telephone transformers \(T_1\) and \(T_2\), whose secondary windings may be connected in series. It is evident that if

the sound source lies in the median plane, that is, the sound reaches both receivers simultaneously, then they will be excited equally and will produce equal electromotive forces in the secondary windings of the transformers. If the secondary windings are connected so that the electromotive forces act in the same direction, a loud sound will be obtained in the telephone; but if they are switched so that the electromotive forces are oppositely directed, silence will be obtained in the telephone. If such a position of the receivers is found, and then the source is moved to one side or the other of the median plane, the sound will reach the receivers at different times and complete absence of sound in the telephone will not be obtained; the sound from the source will be the louder, the farther it lies from the median plane. By turning the bar with the receivers until the sound disappears completely or, at least, until the sound is at a minimum, one can find the direction of the sound source in the same way as in the two-ear method; in the case when the electromotive forces have the same direction and the sound lies in the median plane, we obtain a maximum of sound, and when the bar is turned—an attenuation of it.

Fig. 5.

This method of sum and difference has an advantage over the two-ear method, since the differing sensitivity of the observer’s ears can very considerably affect the ability to determine direction. Therefore many observers prefer the second method. But, on the other hand, this method again requires introducing rotation of the receiver system or introducing a compensator between the two receiver systems, which introduces an undesirable complication. However, in some cases the method of sum and difference, as applied to electrical receivers, is of enormous importance, since it makes it possible to eliminate the difficulty which must be overcome before receivers can be used for two-ear determination of direction. It is clear that, in order to obtain absence of sound in the telephone when connected for the difference of the electromotive forces, the necessary condition is the complete identity of the acti-

…operation of both receivers, but this is very difficult to achieve using ordinary microphones, since their diaphragms differ greatly from one another. Indeed, if two such receivers are placed side by side and the same sound is received by them, it very often turns out that the receivers give only a small difference when connected to the sum and difference of the electromotive forces; it is clear that such receivers are unsuitable for the two-ear method, which requires complete identity of the receivers. By replacing the ordinary carbon or metallic diaphragms with rubber ones, it is possible to achieve much greater uniformity; and, in order to select suitable paired receivers for the two-ear method or for the sum-and-difference method, they must first be tested in the laboratory, for which the sum-and-difference method is very suitable.

Directional receivers.—It has already been pointed out that the change of pressure in an acoustic beam has no directional character, but displacements of the medium occur in the direction of propagation, which makes it possible, by means of a displacement receiver, to determine the direction of the sound. This principle has in fact almost never been applied to directional reception, but Smith (B. S. Smith) developed a displacement receiver consisting of a small hollow sphere containing a magnetophone inside and having neutral buoyancy in water. Such a sphere oscillates as though it were part of the water and, consequently, exerts the greatest influence on the magnetophone when its axis is situated in the direction of propagation, and exerts no influence at all when its axis is perpendicular to that direction.

Fig. 6.

The most commonly used form of directional receiver, shown in Figs. 6 and 7, belongs to the type of balanced hydrophones. Such a hydrophone is similar in construction to non-directional hydrophones (Fig. 1), except that instead of a massive hollow metal box it has only a massive brass ring with a diaphragm in the middle, in the center of which there is a cavity enclosing a “carbon” microphone. It is clear that if such a hydrophone is placed so that

Fig. 7.

so that its plane lies in the direction of propagation of the sound, then both sides of the membrane will be subjected to equal and simultaneous pressures, and as a result the membrane will remain motionless; the observer will not hear the sound. If, however, the hydrophone is turned with its plane toward the source, then the rear side will be screened by the body of the hydrophone, the sound will reach it later and with less force than the front side, so that as a result a definite pressure on the membrane will be obtained. Thus, when such a hydrophone is rotated about an axis, the sound will have a minimum when the plane of the hydrophone is directed at the source, and will reach a maximum when turned through a right angle; the strength of the sound at various angles of rotation is shown in the polar diagram of Fig. 7. A similar effect is given by the directional hydrophone of Morris—Sykes (Morris—Sykes), shown in Fig. 8; it has on two sides two identical membranes, connected at the center by a small rod on which the microphone is fastened. When the hydrophone is set edgewise to the sound, both membranes receive identical pressures tending to move them in opposite directions, as a result of which the rod with the microphone remains at rest and the observer hears nothing.

Fig. 8.

The described types of directional hydrophones operate quite satisfactorily, since they give a very sharp minimum of sound, but they do not fully meet the requirement of directionality, because it is evident that the minimum of sound will be obtained equally no matter which edge we turn the hydrophone toward the source of the sound; thus the sound source may be assumed to be in either of two diametrically opposite directions. Because of this property such apparatuses are called bilateral hydrophones (duolateral); but it turns out that this difficulty can be avoided and a bilateral hydrophone converted into a unilateral hydrophone (unilateral), by placing

Fig. 9.

in several inches on one side of the apparatus a “false wall,” shown in Fig. 9. This wall may be made of several layers of wood or metal, or may have a cavity filled with shot, and is intended to shield the rear wall of the hydrophone from sound. Such a hydrophone gives a maximum of sound when the unprotected side is turned toward the source, and a minimum of sound when it is turned away from the source; the strength of the sound received in different directions is represented on the polar diagram, which clearly shows that in determining the direction there can be no error of 180°, for which reason this apparatus is called a one-sided hydrophone. Such a hydrophone, however, does not give as accurate a determination of direction as a two-sided hydrophone, which has a very sharp minimum, and therefore it is most advantageous to connect the two hydrophones as a pair, placing them at right angles to one another on the same vertical rod. When the two-sided hydrophone gives a minimum of sound, the one-sided one gives a maximum, and we can determine the direction of the source unambiguously and with extraordinary accuracy.

Besides the described methods of directional reception, there are others as well; we shall mention the method of Mazon and Pierce, based on the principle of acoustic accumulation discovered by A. W. Porter; in this method a large flat surface is used for reception; further on the Walser apparatus will be described, in which sound is collected into a focus by means of a lenticular device.

As for directional transmitters, one must first of all recall the general principle applicable to all kinds of radiation, namely, that transmission and reception are mutually reciprocal processes; a good receiver is also a good transmitter; a directional receiver can be converted into a directional transmitter with the same distribution of radiation intensity in different directions. For example, if instead of listening by means of two horns connected by equal tubes with the ears, we bring these tubes to a powerful sound source, so that the sound emerges from the two horns exactly as it was formerly perceived by them, then an observer located in the median plane will hear a loud sound; as he moves away from the median plane to one side or the other, the sound will seem weaker to him. In the same way, by setting the membrane of a one-sided hydrophone into vibration, we shall obtain radiation of sound chiefly in one direction. Generalizing this principle, we may say that it is possible to send a beam of sound in a definite desired direction.

DESIGN OF UNDERWATER TRANSMITTERS AND RECEIVERS

We shall now turn to the description of the instruments used in practice for underwater signaling, and shall consider separately: a) transmitters, b) receivers, and c) instruments for directional operation.

Underwater Transmitters or Sources of Sound.

The simplest form of underwater transmitter is the underwater bell, which has long been used as an auxiliary means in navigation. It was originally designed by Edmunds (H. Edmunds) in 1878, but in practice it was seriously applied to navigation only in 1898, when A. J. Munday and Professor Elisha Gray formed the Gray Telephone Company (1899) and began to use a bell, struck under water, and a telephone receiver for underwater signaling. After Gray’s death in 1901, the work was continued by Munday, who formed the Submarine Signal Co. After numerous trials, the type of bell shown in Fig. 10 was developed. It consists of a bronze bell weighing 100 kg and having in water a pitch of 1215 vibrations; the bell is struck by a hammer actuated by compressed air. For supplying compressed air and removing the spent air a double hose is used; the blows of the hammer are regulated by means of a special valve consisting of a small diaphragm that directs a powerful stream of air into the striking mechanism. This type of bell is usually used on floating beacons, in which case it is simply lowered overboard to a depth of 5–6 meters. On beacons where electric current can be used, a similar bell is employed, actuated by an electric mechanism; it is suspended on a tripod support about 7 m high and about 6 m wide, standing directly on the bottom in a convenient place at a distance of about a kilometer or more from the beacon. In this case the hammer is set in motion by means of a circular iron armature attracted by six electromagnets mounted on a common core; the poles are covered with copper caps to avoid delay under the action of residual magnetism. A four-core cable is led to the bell; through two wires a current of \(3 \tfrac{1}{2}\) amperes is supplied to set the apparatus in motion, and through the other two is conducted the current from a telephone apparatus placed inside the mechanism, which makes it possible to check the proper operation of the apparatus by listening to its sound by telephone. The first such electric

Fig. 10.

A bell was installed near Egg Rock, close to Boston Harbor in the United States, and at the present time a large number of both pneumatic and electric bells are in operation along the coasts of Great Britain and America.

Electromagnetic Transmitters

In view of their simplicity of servicing and control, electromagnetic transmitters are at present the most widespread; they are manufactured in large sizes and make it possible to transmit hundreds of watts of acoustic energy. They may be divided into two classes: a) undamped transmitters and b) intermittent or impact transmitters, corresponding respectively to undamped and spark transmitters in wireless telegraphy.

a) Undamped electromagnetic transmitters.—In all such transmitters an alternating current is used, having a frequency corresponding to the natural period of the oscillatory system, and this current is used either to excite an electromagnet with a laminated core, which acts upon a membrane, or else is passed through a coil attached to the membrane and situated in a strong constant magnetic field, as a result of which alternating forces are developed that set the membrane into vibration. These two types may be called transmitters with a “moving iron core” and transmitters with a “moving coil,” by analogy with the corresponding types of electromagnetic measuring instruments.

Transmitters with a moving core have been developed chiefly in Germany; Fig. 11 shows one of the most widespread types, constructed by the firm Signalgesellschaft in Kiel. The membrane \(D\) is provided at its center with a thickening, to which is attached a casting connected with a laminated iron core \(C\) in the shape of the letter \(E\), situated almost in contact with a second laminated core \(C'\) of the same shape lying above it.

Fig. 11.

The exciting coil is placed on the inner branch of the core, as in ordinary transformers, and produces a strong attraction of the two halves of the core at each passage of the current in one direction or the other. Thus, the frequency of oscillations of the mechanical force is twice the frequency of the current. The upper half of the core is not fixed immovably; it is connected with the lower part by means of four vertical steel tubes \(T\) with steel rods inside, the length of these tubes and rods being chosen so that their natural period of longitudinal oscillations is equal to the natural period of the membrane. The membrane is bolted to a conical shell, in which there are openings for bringing the cable inside. A transmitter of this type weighs about 5 kg and has a membrane about 45 cm in diameter; it radiates acoustic power of from 300 to 400 watts, with a mechanical efficiency on the order of 50%.

A serious objection to transmitters with a moving core is their inherently low electrical efficiency, due to their large self-induction, which entails the necessity of a large wattless exciting current. The matter can be improved by using a large capacitor in parallel or in series with the exciting coil, but this is not a fully satisfactory solution.

In view of this, the second type of transmitter, with a moving coil, is often preferred, especially in America; the principle of its construction is clear from Fig. 12. A coil of wire, through which an alternating current flows, is attached directly to the membrane; it moves because it is situated in a strong constant field formed in the circular gap of a “pot-shaped” electromagnet excited by direct current. This type has a relatively small self-induction and therefore a high efficiency; its construction presents certain mechanical difficulties, since the coil of wire is not a solid body and therefore is capable of causing considerable damping and losses.

Fig. 12.

Exciting coil
Moving coil

This difficulty was very elegantly overcome by Fessenden in America, and his transmitter is probably the most economical and powerful of all electromagnetic transmitters. The principle of construction of this transmitter is exactly the same, but instead of attaching the coil directly to the membrane so that it moves together ...

with it, Fessenden uses a stationary coil, which induces currents in a solid copper cylinder, as in the secondary winding of a transformer, and this copper cylinder is already attached to the membrane. Fig. 13 shows a section of Fessenden’s transmitter, in which the direct-current electromagnet is made with two poles; it encloses the copper cylinder attached to the membrane. Alternating current flows through a stationary coil wound on the inner iron core, the coil being wound in its two halves in opposite directions, corresponding to the different poles of the magnet; this coil induces powerful currents in the copper cylinder, and these currents, cutting the strong field of the magnet, produce electrodynamic forces directed along the axis of the cylinder and having the same frequency as the alternating current that supplies it. Such a device is mechanically very rigid; at resonance with the natural frequency of the membrane, which is usually made 1050 ∿, a large power and efficiency are obtained. Transmitters of this type give an acoustic radiation of 500 watts and more and can transmit signals under water over distances up to 500 km. Morse characters can be transmitted by any of these transmitters by means of a suitable key.

Fig. 13.

b) Intermittent or impact transmitters.—We have already mentioned underwater bells, which represent the prototype of the intermittent underwater transmitter and which can be actuated electrically. An even simpler type of impact transmitter is an impact sound source with a membrane, designed by B. Smith; it has the advantage over a bell that the impact mechanism lies entirely inside the apparatus, and no part of it operates in the water. Fig. 14 gives a section of such an instrument; it is furnished with an ordinary steel membrane with a heavy mass at the center, which is struck by a hammer. This hammer is drawn in when direct current passes through the exciting coil, overcoming the force of a spiral spring, and when the current is suddenly interrupted the spring pushes the hammer, which gives the membrane a single sharp blow and then rebounds, leaving it free to vibrate. In this way a series of very powerful sound oscillations is obtained, although it lasts only a very short time owing to the great damping of the membrane in water.

A similar powerful impact transmitter can be put into operation pneumatically by compressed air, with a frequency of blows of about

hundreds per second; this type of transmitter can be used for signaling by Morse code, employing a special pneumatic key.

For the purposes of acoustic depth sounding, the author specially designed a simple transmitter which communicates to the water a series of separate impulses without producing rhythmic oscillations. Here the elastic membrane is completely eliminated and replaced by a square laminated plate, which is attracted to a magnet with a laminated core in the form of the letter \(E\), fed by direct current passing through a coil placed on the middle branch of the yoke. The attractive force is \(\frac{B^2}{8\pi}\) dynes per square centimeter, where \(B\) is the magnetic induction in gausses; thus, if \(B = 15000\), the force obtained is about 14 kg per square centimeter, which, for a pole area of 140 sq. cm, gives a force of about 2 tons. In order to transmit this force to the water, the pole tips and the plate are provided with grooves, and strips of rubber are placed in them, compressed when attracted by the magnet. When the electromagnet is connected to a network with a voltage of 100 volts, the current rises comparatively slowly owing to the large self-induction, and the plate is slowly attracted; but when the current is suddenly interrupted, the reaction of the rubber gaskets pushes the plate with an initial force of about 2 tons and communicates to the water a single sharp blow, similar to an explosion. The operation of this transmitter will be explained in connection with questions of acoustic depth sounding.

Labels in Fig. 14: Membrane; Nozzle; Hammer; Solenoid; Air gap.

Fig. 14.

Underwater Sirens.

In England and Germany a whole series of various kinds of underwater sirens of considerable power has been proposed; in all of them there is a plate or disk, through the openings of which water is driven, causing the disk to rotate. By giving the openings definite dimensions, one can, of course, always make the water rotate the disk, but this method is inconvenient from the standpoint of signaling, since the resulting gradual increase in speed produces a change in the pitch of the sound.

Because of this, the disk is usually rotated independently, at a constant speed, by means of an electric motor, while signaling is carried out by turning on and off the jet of water forced under high pressure. However, sirens have not come into wide use, since electromagnetic transmitters are much more convenient; therefore we shall not dwell on sirens in greater detail. There are many other forms of acoustic transmitters, but those described above are used for acoustic signaling and impact transmission almost everywhere. For the purposes of sonic reconnaissance, small explosions are often used.

RECEIVERS OR HYDROPHONES.

C-tube.

The first and simplest underwater acoustic receiver should be considered the already mentioned Brock tube with a membrane stretched across its lower opening. The Americans improved this device by replacing the membrane with a thick-walled rubber balloon or a flat cushion and named it, in honor of Dr. Coolidge—its inventor—the C-tube; it is shown in Fig. 15. This tube is very sensitive, but the amount of sound energy imparted to the air inside the balloon, owing to the principle set forth above of the transition of sound from one medium into another, proves to be extremely small, which forces the observer to listen through as short a tube as possible and makes this method not particularly convenient.

Fig. 15.

Metal tube
Rubber balloon

The Americans were the first to appreciate the sensitivity and convenience of using microphones; they place microphones in hollow rubber balloons and use a combination of three such balloons, suspended on a tripod frame, for directional reception. Figure 16 shows a double C-tube for binaural reception. As has already been explained, listening with two ears from two different receivers makes it possible to find the direction to the source of sound, but it is obvious that in this case an error of 180° is always possible, as in a two-sided hydrophone, since sources located symmetrically with respect to the line connecting the two

hydrophone, will give the same difference in the paths of the sound wave. By using three receivers placed at the corners of an equilateral triangle and listening to them in pairs in turn, the direction can be determined with complete accuracy. The microphones used for these purposes must first be matched by the sum-and-difference method.

Fig. 16.

Magnetophones.

Magnetophones are less sensitive than microphones, but for underwater reception they have the advantage that they are free from the caprices of carbon microphones, and it is much easier to match them in pairs for the two-ear method. Since their low sensitivity can easily be compensated by means of modern amplifiers with cathode tubes, which cannot be used with microphones because of the noise caused by the carbon grains, an apparatus equipped with magnetophones can be made just as sensitive. The Fessenden transmitter can be used as a powerful magnetophone receiver by exciting its electromagnet and listening to the coil through which, during transmission, an alternating current is passed; it is also usually used as a receiver in signaling, which

Fig. 17.

convenient also in the respect that the membranes of all transmitters of the same type are tuned to one and the same tone. However, this sharp tuning makes such an apparatus unsuitable for receiving two signals.

One of the most sensitive magnetophone devices for reception on ships is Smith’s magnetophone (Fig. 17). It consists of a massive lead casing (4), fastened to the side of the vessel, into which is inserted a thick rubber membrane (2) in contact with the water. Immediately behind this membrane there is mounted an ordinary Braun telephone (3)¹), so that sound, having passed through the rubber membrane into the air inside the apparatus, sets the membrane and the telephone tongue connected with it into vibration and produces induced currents in the coils. This type of receiver, in combination with a three-tube amplifier and a high-impedance telephone, gives excellent reproduction of ordinary sounds; if four such receivers are fastened in pairs, fore and aft, to the hull of the vessel, then the shielding action of the hull makes it possible to determine the direction of the source from the relative intensity of the sound in the 4 receivers—for convenience the apparatus is provided with a 4-pole switch between the receivers and the amplifier. The ship’s noises are considerably weakened by fastening the lead casing of the receiver to plates with rubber bushings; the great inertia of the lead ring (4) prevents the perception of noises originating from the ship’s hull.

Fig. 18.

The described forms of underwater receivers are the most commonly used in practice.

¹) In this telephone the current passing through the coils of a polarized magnet attracts an elastic iron tongue, which is fastened to a light aluminum membrane set into vibration.

Ed.

Construction of Hydrophones

Fig. 18 shows the simplest form of a non-directional hydrophone, the diagram of which was given in Fig. 1; in it a massive hollow bronze body is fitted on one side with a membrane, at the center of which is mounted a microphone with a “massive support” (Fig. 19 represents a double-sided hydrophone with two membranes, shown schematically in Fig. 8).

In order to be able to listen while the ship is under way and to reduce as much as possible the noise of the ship and the movement of the water, hydrophones with a rubber bladder, or other more complex types, are enclosed inside fish-shaped bodies and towed in the water at some distance astern; for binaural listening a combination of several such bodies is used. At present, however, efforts are being made to place the receiving apparatus on board, introducing good acoustic insulation from the hull.

Fig. 19.

In Germany a method of reception by means of reservoirs of water fixed inside the vessel is widely used; it was first introduced by the firm Submarine Signal Co. Fig. 20 shows the arrangement of two such reservoirs with hydrophones inside. With this system, large losses of sound due to reflection at the boundary between water and air are avoided, as was indicated above.

An extremely interesting and sensitive form of directional receiving apparatus was constructed by Lieutenant of the French Navy Walser; in it the sound is collected at a focus, as in a camera obscura, and the direction is determined from the position of this focus. For this purpose a plate (“patch”) is attached to the ship’s hull, consisting of a steel shield with spherical curvature, about 1.1 m in diameter; part of it is visible in Fig. 21. This shield has a bol—

Reservoir
Microphone

Fig. 20.

a large number of openings \(B\), which are covered by thin steel membranes \(C\). All these membranes, arranged on a spherical surface, collect the sound and direct it to the focus, lying at a distance of 1.5 m. For detecting the sound there is a funnel \(D\), with an auditory tube attached to it, rotating on a lever about a vertical axis, so that it can be moved in order to seek the focus and determine the direction of the source of the sound. Usually 2 such apparatuses are installed on both sides of the vessel in its forward part, and the observer is placed between them, inserting the auditory tubes from

Fig. 21.

two funnels in the ears; in this way he can track the position of the source from both one side and the other and determine the direction on the scale.

INSTRUMENTS FOR DETERMINING DIRECTION.

Sound ranging.

One of the most important acoustic tasks during the war was sound reconnaissance for determining the location of guns and underwater explosions. There are two methods for determining location, which may be characterized by the names: the “several-stations” method and the “radio-acoustic” method; of these, only the first was used in the war (despite its lesser convenience), since it does not require the joint action of a transmitter and a receiver.

Sound ranging by the several-stations method.—The several-stations method is based on the fact that sound waves propagate spherically from the source of sound, as from a center. If 3 or more receivers are placed on a circumference described from the source of sound as from a center, then the sound reaches all the receivers simultaneously; thus, if the signals at all receivers are received at one and the same moment, it follows that the source lies at the center of the circle passing through the receivers. But when the source lies in some other position, the signals will be received at different moments and, by measuring the difference in reception times at the different stations, the position of the source can be established by calculation or graphically.

A simple scheme in Fig. 22 explains this method. Let the points \(A, B, C, D\) denote 4 receivers lying at precisely determined points, and let \(P\) be the sound source whose position is to be determined. Draw a circumference with center at \(P\) through the receiver \(A\); it is evident that, when the sound has reached \(A\), it must still travel the distance \(bB\) in order to reach receiver \(B\), and the distances \(cC\) and \(dD\) to the receivers \(C\) and \(D\). Thus, the sound will arrive at \(B, C\), and \(D\) after intervals of time

Fig. 22.

\[ t_1=\frac{bB}{v},\quad t_2=\frac{cC}{v}\quad \text{and}\quad t_3=\frac{dD}{v}. \]

later than at \(A\). Having measured the intervals of time \(t_1\), \(t_2\), and \(t_3\), and multiplying them by the speed of sound, we obtain the perpendicular distances of the receivers \(B\), \(C\), and \(D\) from the circle having its center at \(P\) and passing through the receiver \(A\) nearest to it. If circles are described about \(B\), \(C\), and \(D\) with radii representing, on some scale, the distances \(bB\), \(cC\), and \(dD\), and a circle is then drawn tangent to the three described, its center will determine the position of the sound source \(P\). To determine the intervals of time during the war, Einthoven’s multi-string galvanometer was used almost everywhere; four of its strings are connected with four microphones or hydrophones, and the fifth—with an electric clock or a tuning-fork interrupter, in order to give an exact time scale. The image of the strings is focused on a strip of bromide paper, which is drawn by means of a motor first through the galvanometer camera and then through the developing and fixing bath, and comes out of the apparatus ready for washing and drying; the necessary readings can be made immediately after the appearance of the strip. To facilitate readings of the time intervals, in front of the light source a wheel with thick and thin spokes is made to rotate, by means of a “phonic motor,” synchronous with a tuning fork, so that across the paper there is obtained a series of lines at intervals of tenths and hundredths of a second.

Fig. 23 represents a reproduction of the record of a sound-ranging station of this kind, on which reception from 4 receivers was registered; in Fig. 24 is shown the Einthoven camera for recording. The receivers used in this case were simple microphones, mounted on membranes screwed to watertight reservoirs, which were suspended on tripods standing on the sea bottom at precisely determined points; the microphones were connected by cables with the observation station.

In view of the importance of sound ranging as a means for determining the position of a ship during fog, much effort was made

for the further improvement of the method and for the elimination of the photographic apparatus. A great improvement in this direction was made by A. B. Wood and J. M. Ford at the Admiralty experimental station. They proposed an instrument—which may be called a “phonic chronometer”—showing intervals of time directly on a scale with an accuracy of up to one thousandth of a second. The principle of the construction of this instrument is very simple and is clear from Fig. 25. On a vertical axis there rotates, at the constant speed of 150 revolutions per minute, a special motor fed by current from a tuning-fork interrupter giving 25 interruptions per second; at the top of its axis is placed a round vessel containing mercury, which is intended to maintain constancy of the speed of rotation; the vessel is provided with a rim which, together with it, rotates about the axis. Around this disk there is placed

Fig. 24.

a known number of recording mechanisms (in Fig. 25 there are three of them); each of them consists of a light aluminum wheel mounted on a vertical axis, the rim of which is at a distance of hundredths of a millimeter from the rotating disk. The axis of this little wheel can be drawn inward by means of an electromagnet, and then the aluminum wheel comes into contact with the rotating disk and begins to rotate at a speed of 10 revolutions per second; each of the electromagnets is provided with two identical windings in opposite directions; when current is passed through both windings, just as when there is no current, the electromagnet does not act, but when current is passed through one of the windings it draws the axis of the small wheel inward. To the axes of the aluminum wheels there is attached a light aluminum pointer, which rotates over a scale having one hundred divisions; thus, each division corresponds to one thousandth

fraction of a second; two other indicators, connected with the axis by a gear transmission, indicate longer intervals of time up to 10 seconds.

For use in sound ranging, the electromagnet windings of the chronometer are connected with hydrophones according to a special circuit. In this case the hydrophone consists of a diaphragm with a point contact, which opens under the action of sound and remains open until it is closed again by a special device. Four hydrophones may be used; then the circuit is arranged so that the first hydrophone immediately breaks the circuits of one half of the windings of all the electromagnets and starts all the chronometers simultaneously, and when the sound reaches the following hydrophones, it breaks the second winding of the corresponding electromagnet, after which the wheel of the chronometer moves away from the disk and, under the action of the brake, is stopped at once. Thus, three dials will show us the intervals of time between the moment of arrival of the sound at the first hydrophone and at each of the three others. One thousandth of a second corresponds in sea water to a distance of about 1.5 m; from the quantities found it is easy to construct a diagram according to Fig. 21 and determine the position of the source of sound.

Fig. 25.

For quickly determining position, the author proposed an instrument which was called a sound-ranging locator. A similar device was proposed by Dadourian in America1.

A ship wishing to determine its position during fog calls by radio the nearest sound-ranging station, which offers it the opportunity to explode an underwater charge. From a record by means of an Aitken camera or from the readings of a phonic chronometer, the position of the ship is determined on the chart and communicated to it by radio.

Sound ranging by the radio-acoustic method.—This method was proposed by Prof. Joly and will probably be of enormous importance in navigation; up to now this method has been little developed, since

...it was of little importance in military conditions. In the experiment of Colladon and Sturm in 1826, the speed of sound in water was determined by means of a simultaneous blow on an underwater bell and a flash of powder in the air. Knowing the distance between the source of the sound and the observer and the interval of time between the flash and the sound of the bell, it is easy to calculate the speed of sound in water, assuming that light propagates instantaneously. Conversely, if the interval of time and the speed of sound are known, one can calculate the distance, as is done,

Fig. 26.

for example, in determining the distance of lightning, by measuring the interval of time between the flash of lightning and the clap of thunder. The advantage of the underwater method consists, first, in the small absorption of sound in water and, second, in the absence of water currents similar to wind, which could noticeably affect the speed of sound.

Unfortunately, a flash of light cannot be used during fog, but instead radio-telegraphic signals can be employed, which are little affected by fog and which propagate with the same speed as light. At a lighthouse or at another specified point

is produced simultaneously: an underwater explosion and a short radio signal, and a ship equipped with a receiving radio station and a directional hydrophone can determine the distance from the beacon by measuring the time interval between the radio signal and the sound of the explosion. Since the speed of sound in seawater is approximately 1.5 km per second, the distance can be determined to an accuracy of up to \(1/3\) km by means of a simple stopwatch, and the direction of the station can be found either with a directional hydrophone or with a radio direction finder, without establishing communication between the ship and the station. If the beacon or beacon vessel from time to time sends radio signals and simultaneously strikes an underwater bell, then all ships in the vicinity can, at each such signal, determine their position by using a directional hydrophone; and if the signals of two stations are heard simultaneously, then this becomes possible even without directional hydrophones. The successes achieved in determining direction by means of radio goniometry have made the use of sound ranging in navigation less important, but it must be said that determining direction by radio is sometimes inapplicable, for example at sunrise and sunset; moreover, on steel vessels radio goniometry gives considerable errors in determining direction because of the strong influence of the ship’s mass on the front of the electromagnetic wave. Since the use of hydrophones on ships is coming into ever wider use for the purposes of underwater signaling and so forth, the possibility of applying the radio-acoustic method for determining position is a fact of great importance.

Cable Pilot.

This device, strictly speaking, has no relation to acoustics, but it should be mentioned in a few words as an auxiliary means for guiding ships in foggy weather in harbors and channels. In this case the ship must proceed along a perfectly definite path only a few meters wide, and sound-ranging methods are obviously unsuitable. For the construction of a cable pilot, a cable is laid along the seabed in the desired direction, through which a current of sound frequency, 500 cycles, is passed; the ship is equipped with a “search” coil connected to an amplifier and a telephone; if the ship goes sufficiently close to the cable, then the magnetic field of the cable induces in the coil an alternating electromotive force and produces sounds in the telephone. By arranging two inclined coils on both sides of the hull of an iron or steel ship, it was found that the sound is louder when the telephone is connected to the coil nearer the cable, so that the ship can keep a course along the cable and maintain itself at a perfectly definite distance on one side of it. Ships proceeding toward one another in this way do not ris-

will collide. This idea was first proposed by Stevenson in Edinburgh in 1893 and was put forward again during the war by Captain Mason. The pilot-cable method is now in use both in England and in America. From Southampton harbor to Spithead and farther out to sea the Admiralty has laid a pilot cable 27 nautical miles long.

Acoustic Determination of Depth

Another purely acoustic device that affords considerable assistance in navigation is the sonic method of determining depths by means of reflection (echo) from the bottom. If a ship produces an explosion near the surface, the sound propagates downward and, having been reflected from the bottom, can be heard as an echo; each second of the interval between the explosion and the echo corresponds to a depth equal to half the speed of sound, that is, about 750 meters. Work in this direction was carried out by Marti in France, Behm in Germany, and officers of the American fleet, and they obtained very accurate results. The apparatus used consists of a high-speed chronograph, which marks the interval of time between the explosion of a detonator or other small charge beneath the ship and the reception of the echo reflected from the bottom. Behm’s method was developed by the Anschütz firm, which manufactures apparatus under the name “Echo-lot,” giving excellent results and allowing depth to be determined with an accuracy of up to 0.3 m while the ship is at full speed and in a seaway. The sound source is the explosion of a detonator, produced pneumatically and ignited by pressing a key when it has moved several feet away from the hull. On the other side of the hull, protected by it from the direct action of the explosion, a receiver is installed; it consists of an ordinary carbon microphone, closing the circuit of the time-marking device, shown schematically in Fig. 26. It is designed to register small intervals of time with an accuracy of up to thousandths of a second and consists of a disk rotating on an axis, which is brought by a key into the zero position and held there, against the force of a spring, by attraction by an electromagnet. At the moment of the explosion the circuit of the electromagnet is broken, and the disk receives an impulse from the spring and begins to rotate uniformly; under the action of the sound received by the microphone, a brake is pressed against the disk and stops it. The angle of rotation of the disk is thus proportional to the interval of time between pressing the firing key and the return of the echo; a small mirror on the disk axis directs a beam of light onto a transparent scale, divided directly into depths, and allows the required depth to be read off at once. The entire apparatus can operate from several dry cells. Experience has shown that a rock with a surface area of 2 m² already gives

sufficiently strong echo, and the apparatus can detect it. Bem began his investigations in 1912, and the Ansco company improved his apparatus until 1920, when he left his work at the firm.

Detection by Echo of Ships and Obstacles.

By means of sound ranging, pilot cables, and depth measurement by sound, navigation during fog can be made far safer and more correct; nevertheless there remains a great danger of collision on the open sea between vessels, and the possibility of striking rocks, icebergs, and shipwreck debris. As for ships, there is the possibility of reducing the danger by signaling with sirens; but owing to unexpected reflections and refractions of sound in fog there is always a source of errors and danger. Underwater signaling eliminates these difficulties almost entirely, and as hydrophone devices are introduced the risk of collisions between ships in motion becomes ever smaller. With a good receiving hydrophone having directional reception, an ordinary steamship can easily be detected, and its direction determined from the noise of its engines, at distances of up to several kilometers. But in the case of rocks, icebergs, and wreckage, which emit no sounds, the only method of detection during fog is the use of echo. Unfortunately, the echo from a small rock or ship at a respectable distance is too weak and can easily be confused with the echo from the bottom. However, Fessenden, by means of his powerful electromagnetic transmitter, succeeded as early as 1916 in obtaining echoes from distant objects; directional transmitters and receivers make it possible to amplify the echo, reduce interfering sounds, and determine the direction and approximate distance of the object sought. The method of detecting dangerous obstacles by means of echo was proposed as early as 1912, soon after the sinking of the Titanic, by L. Richardson, and it may be hoped that this method will finally remove the last of the serious dangers in seafaring.

Acoustic Transmission of Power.

Before concluding our survey, it is still necessary to point out the astonishing achievements of M. Constantinesco, who succeeded in demonstrating the possibility of using acoustic energy for a number of mechanical tasks. For the purposes of underwater signaling, quantities of acoustic energy are employed which, though greater than what has hitherto been understood as sound energy, still rarely exceed several hundred watts. Constantinesco boldly foresaw the possibility of transmitting large quantities of energy in water by means—

by means of variable pressures having an acoustic frequency. For many years it has been customary to illustrate the phenomena of alternating electric currents by means of hydraulic analogies, and the author of the present article has written a book in which such analogies are used to construct a complete theory of the subject; however, until Constantinesco expressed his idea about the possibility of transmitting alternating pressures in water for practical purposes, this idea occurred to no one. Now, as soon as this idea appeared, it became clear that the entire theory of the processes taking place was already available by analogy with the theory of electricity. In an astonishingly short time Constantinesco invented generators, motors, and transformers designed for a large amount of power transmitted through hydraulic pipes in the form of acoustic waves with a frequency of about 50 cycles. The generator is a simple valveless high-pressure pump; the motor is constructed in an analogous way. By using three pistons with cranks set 120° apart from one another, it is possible to obtain a three-phase acoustic generator and use it to rotate a motor. The first commercial use of Constantinesco’s apparatus was for drilling rock and riveting, for which this method is especially suitable, since here it is possible to use a simple piston in a cylindrical cage without any valves, while the transmission of power takes place by means of a specially shaped flexible hydraulic hose, analogous to an electric cable. It would be no exaggeration to say that Constantinesco’s ideas open up an entirely new field for technology, and that their development may be of very great importance.

ADDENDUM ¹).

ULTRA-SONIC SIGNALING.

In recent years, the technology of underwater signaling has been enriched by new apparatuses working on a completely different principle than before. We have in mind the underwater quartz emitters proposed by the French physicist P. Langevin and developed by him jointly with the engineers Florisson, Marti, and Tuli ²).

The quartz emitter is based on the piezo-electric properties of this substance. If a thin plate is cut from a quartz crystal (Fig. 27) in such a way that its plane is parallel to the optical axis \(Z\) and normal to the binary or electrical axis \(X\) (there are three such axes in the crystal), and if metal—

¹) Compiled by the translator S. N. Rzhevkin.

²) La Nature, No. 2572, 21 Jul. 1923, and No. 2667, 16 Mai, 1925.

metallic plates, subject it to compression or stretching in the direction of the \(Y\) axis, then an electric charge will appear on the plates,

\[ Q = K \frac{a}{l} P \; CGS \; E, \]

where
\(a\) is the length of the plate parallel to the optical axis,
\(l\) is the thickness of the plate,
\(p\) is the acting force in kg,
\(k\) is the piezo-electric constant of quartz \(= 0.0677\).

Under tension and compression the signs of the charges are opposite; likewise under compression and tension in the direction of the \(X\) axis.

Conversely, when the plates are electrified, a change in the dimensions of the crystal is obtained (compression or stretching, depending on the sign of electrification). The change in dimensions occurs both along the \(X\) axis and along the \(Y\) axis. In absolute magnitude the changes in the dimensions of the crystal are very small, but the forces developed are very considerable.

Fig. 27.

If the plates of the lamina are charged with alternating current, the crystal begins to perform elastic vibrations; the waves propagate along the \(X\) axis and along the \(Y\) axis. Under suitable conditions the vibrations of the crystal are readily transmitted to the surrounding medium; in water, moreover, the radiation from quartz is much more intense than in air, owing to the great closeness of the elastic properties and to the resulting small reflection at the boundary of the two media.

The greatest intensity of elastic vibrations in quartz can be achieved in the case when one half-wave of the elastic vibration is accommodated along the crystal (or, in general, an integral number of half-waves, as in an open organ pipe or a string), and we have resonance of the exciting vibrations with the natural period of the crystal along the \(X\) or \(Y\) axes. Calculation shows that this can occur only at very high frequencies, of the order of hundreds of thousands per second. If the crystal is clamped between thick metal plates, for example steel ones, then a longer standing wave is established in such a system. In practice it has proved convenient to use frequencies from 40,000 to 100,000 oscillations per second.

These frequencies are certainly above the limit of sensitivity of the ear, and therefore they may be called ultrasonic frequencies. The detection of such oscillations is possible by converting them into electrical form and subsequently detecting (rectifying) them, much as is done in radiotelegraphy.

The use of high, ultrasonic frequencies is, it is true, associated with more considerable absorption in water, but on the other hand it makes it possible to use the advantage of directional transmission and concentration of the sound beam, thanks to which it is possible to attain a very great range. Low-frequency oscillations spread from the source in all directions in the form of a spherical wave, whereas high-frequency oscillations, having a wavelength smaller than the diameter of the membrane, are radiated in the form of a plane wave, as a beam in the direction of the normal to the radiator, with very little divergence.

Fig. 28.

Langevin’s powerful underwater radiator consists of many quartz plates arranged in the form of a mosaic and clamped between thick steel disks (Fig. 28). Experiment has shown that such a radiator in a steel mounting gives off 625 times more energy than a single quartz without a mounting, and makes it possible to attain radiated powers of several hundred watts.

The method of ultrasonic signaling is applicable for communication of ships with one another and with the shores, for underwater telephony (by means of modulation of the high frequency by the speech current), and for other purposes.

The most important application of the quartz radiator has been found in determining sea depths by the echo method. If \(t\) denotes the interval of time (in seconds) between the sending of a short sound signal and the arrival of the echo reflected from the bottom, then the depth of the sea is equal to

\[ h=\frac{t \times 1480}{2}\ \text{meters}. \]

The apparatus for determining depths consists of an ordinary spark radio transmitter, in parallel with the capacitance of whose oscillatory circuit a quartz radiator is connected. When the key is pressed, the transmit-

[[unclear: beginning of word]] the emitter sends into the water a series of high-frequency sound waves, following one another at an audible frequency.

The receiver is the same quartz emitter: under the action of sound, charges arise on its plates, and it excites the transmitter’s electrical circuit tuned to resonance; the oscillations are amplified and recorded automatically on a moving tape.

Fig. 29.

By recording on the tape the instant of the signal and the instant of the echo, it is possible to determine immediately and accurately the depths of the sea beneath the ship, as well as the distance to underwater rocks and other obstacles. Existing apparatus makes it possible to determine depths from 5 m to 400 m. Greater depths, down to 4000 m, require more powerful emitters, types of which are being developed. Fig. 29 gives an example of a continuous recording of depths on a tape as the ship moves. Such a recording, obviously, gives a complete picture of the profile of the sea bottom and makes it possible to carry out this work at the speed of the vessel.

1. H. Dadourian. Phys. Rev. August, 1919. ↩↩↩↩