REPRINT OF SECTION 
 
 I? J WIKKLI-SS TI-;i K< I; I'!! . \ MI 
 
 I 2 ^s I WIRELESS TELEPHONY 
 
 I 2 
 I 6 
 
 5 
 
 FROM 
 
 RACTICAL PHYSICS 
 
 BY 
 
 ROBERT ANDREWS MILLIK 
 
 AND 
 
 HENRY GORDON GALE 
 
 IN COLLABORATION WITH 
 
 WILLARD R. PYLE 
 
 GINN A>i -1'A.NY 
 
 BOSTON NEW YORK CH 
 AT1.AJ4TA DALLAS ! KANCBCO 
 
 COPV HI
 
 This book is DUE on the last date stamped below 
 
 TK 
 5742 
 
 Southern Branch 
 of the 
 
 University of California 
 
 Los Angeles
 
 
 M 
 
 
 .RADI ATK > X S 
 ELECTRICAL RADIATIONS 
 
 421 
 
 485. Proof that the discharge of a Ley den jar is oscillatory. 
 We found in 408, p. 346, that the sound waves sent out 
 Wy a sounding tuning fork will set into vibration an adjacent 
 'ork, provided the latter has the same natural period as the 
 ler. Following is the complete electrical analogy of this 
 riment. 
 
 Let the inner aiid outer coats of a Leyden jar A (see Fig. 455) be 
 [ connected by u loop of wire cdef, the sliding crosspiece de being arranged 
 
 so that the length of the loop may be altered at will. Also let a strip 
 
 of tin foil be -brought over the edge of this jar from the inner coat to 
 [within about 1 millimeter 
 I of the outer coat at C. Let 
 I the two coats of an exactly 
 
 .similar jar II be connected 
 jwith the knobs// and n' by 
 
 a second similar wire loop 
 
 of fixed length. Let the 
 
 two jars be placed side by 
 ! .side with their loops par- 
 'allel, and let the jar B be 
 ; successively charged and discharged by connecting its coats with a 
 I static machine or an induction coil. At each discharge of jar H through 
 
 the knobs // and //' a spark will appear in the other jar at C, provided 
 
 the crosspiece de is so placed that the areas of the two loops are equal. 
 
 When de is slid along so as to make one loop considerably larger or 
 
 smaller than the other, the spark at C will disappear. 
 
 The experiment therefore demonstrates that two electrical 
 circuits, like two tuning forks, can be tuned so as to respond to 
 each other sympathetically, and that just as the tuning forks 
 will cease to respond as soon as the period of one is slightly 
 altered, so this cl<'<-tri<' rcxona !><< disappears when the exact 
 symmetry of the two circuits is destroyed. Since, obviously, 
 this phenomenon of resonance can occur only between systems 
 which have natural periods of vibration, the experiment proves 
 that the discharge of a Leyden jar is a vibratorv, that is, an 
 
 51652 
 
 . 455. Sympathetic electrical vibrations
 
 422 
 
 oscillatory, phenomenon. As a matter of fact, when such a 
 spark is viewed in a rapidly revolving mirror, it is actually found 
 to consist of from ten to thirty flashes following each other at 
 equal intervals. Fig. 456 is a photograph of such a spark. 
 
 In spite of these oscillations the whole discharge may be 
 made to take place in the incredibly short time of t 00 * 
 of a second. This fact, coupled 
 with the extreme brightness of 
 the spark, has made possible the 
 surprising results of so-called 
 
 instantaneous electric-spark pho- 
 
 7 rr,, , . .. FIG. 456. Oscillations of the 
 
 tography. The plate opposite electric spark 
 
 page 425 shows the passage of 
 
 a bullet through a soap bubble. The film was rotated continu- 
 ously instead of intermittently, as in ordinary moving-picture 
 photography. The illuminating flashes, 5000 per second, were 
 so nearly instantaneous that the outlines are not blurred. 
 
 486. Electric waves. The experiment of 485 demonstrates 
 not only that the discharge of a Leyden jar is oscillatory but 
 also that these electrical oscillations set up in the surrounding 
 medium disturbances, or waves of some sort, which travel to a 
 neighboring circuit and act upon it precisely as the air waves 
 acted on the second tuning fork in the sound experiment. 
 Whether these are waves in the air, like sound waves, or dis- 
 turbances in the ether, like light waves, can be determined by 
 measuring their velocity of propagation. The first determina- 
 tion of this velocity was made by Heinrich Hertz (see oppo- 
 site p. 102) in 1888. He found it to be precisely the same as 
 that of light, that is, 300,000 kilometers per second. This 
 result shows, therefore, that electrical oscillations set up waves in 
 the ether. These waves are now known as Hertzian waves. 
 
 The length of the waves emitted by the oscillatory spark 
 
 of instantaneous photography is evidently very great, namely, 
 
 . about Vo^ooV.ooV = 30 meters, since the velocity of light is
 
 ELECTRICAL RADIATIONS 423 
 
 300,000,000 meters per second, and since there are 10,000,000 
 oscillations per second ; for we have seen in 382, p. 323, 
 that wave length is equal to velocity divided by the number 
 of oscillations per second. By diminishing the size of the jar 
 and the length of the circuit the length of the waves may be 
 greatly reduced. By causing the electrical discharges to take 
 place between two balls only a fraction of a millimeter in 
 diameter, instead of between the coats of a condenser, elec- 
 trical waves have been obtained as short as ,3 centimeter, 
 only ten times as long as the longest measured heat waves. 
 
 487. Detection of electric waves. In the experiment of 485 
 we detected the presence of the electric waves by means of a 
 small spark gap C in a circuit almost identical with that in 
 which the oscillations were set up. The visible spark may be 
 employed for the detection of waves many feet away from 
 the source, but for detecting the feeble waves which come in 
 from a source hundreds or thousands of miles away we must 
 depend upon sounds produced in an extremely sensitive tele- 
 phone receiver, as explained in the next section. 
 
 488. Wireless telegraphy. Commercial wireless telegraphy 
 was realized in 1896 by Marconi (see opposite p. 316), eight 
 years after the discovery of Hertzian waves. The essential 
 elements of a tuned wave-train, or " spark," system of wireless 
 telegraphy are as follows: 
 
 The key K at the transmitting station (Fig. 457, (1) ) is depressed 
 to allow a current from the alternator .4 to pass through the primary 
 coil P of a transformer 7\, the frequency of the alternations in practice 
 being usually about 500 cycles per second. The high-voltage current 
 induced in the secondary 5 charges the condenser C l until its potential 
 rises high enough to cause a spark discharge to take place across the 
 gap s. This discharge of C l is oscillatory ( 485), and the oscillations 
 thus produced in the condenser circuit containing C,, 3, and L l may, in 
 a low-power short-wave transmitting set, have a frequency as high as 
 1,000,000 per second. An oscillation frequency much lower than this 
 is generally used and is subject to the control of the operator through
 
 424 
 
 INVISIBLE RADIATIONS 
 
 the sliding contact c, precisely as in the case illustrated in Fig. 455. 
 The oscillations in the condenser circuit induce oscillations in the aerial- 
 wire system, which is tuned to resonance with it through the sliding 
 contact c'. 
 
 U) 
 
 Aerial 
 Wires 
 
 Transmitting Station 
 
 a 
 
 jrth 
 
 FIG. 457. Transmitting and receiving stations for wireless telegraphy 
 
 As long as the key 7v is kept closed (assuming a 500-cycle alternator 
 to be used), 1000 sparks per second occur at s, and therefore a regular 
 series of 1000 wave trains (Fig. 458) pass off from the aerial every 
 second and move away with the velocity of light. If the oscillations 
 which produce a wave train have a frequency 
 of, say, 500,000 per second, each wave in the 
 
 , 300,000,000 
 wave train has a length or 
 
 Direction of 
 Propagation 
 
 500,000 ' 
 
 600 meters ; and if these wave trains are 
 produced at the rate of 1000 per second, 
 they follow each other at regular distances 
 of 300,000 meters, that is, nearly 200 miles. 
 The waves sent out by the aerial system 
 of the transmitting station induce like os- 
 cillations in the distant aerial system of the 
 receiving station (Fig. 457, (2) ), which is 
 
 tuned to resonance with it. In case the receiving aerial must be tuned 
 to respond to very long waves, the switch O is closed to cut out the 
 condenser C 2 , and the inductance, or loading coil, B l is used ; whereas, 
 to tune to very short waves, the switch O is opened and the variable 
 
 FIG. 458. One wave train 
 from oscillatory discharge
 
 One of the most notable developments of the war was the directing of a squadron 
 of airplanes in intricate maneuvers by wireless telephone either from the ground 
 or by the commander in the leading plane. The upper panel shows the pilot and 
 the observer conversing with special apparatus designed to eliminate plane noises, 
 ' and the lower panel shows President Wilson talking by wireless to airplanes
 
 
 
 "\ 
 
 
 
 CINEMATOGRAPH FILM OF A Bui 
 
 iD THROUGH A SOAP BUBBLE 
 
 The flight of the missile may be followed easily. It will be seen that the bubble 
 
 breaks, not when the bullet enters, but when it emerges. (From " Moving Pictures,' 
 
 by F. A. Talbot. Courtesy of J. B. Lippincott Company)
 
 ELECTRICAL RADIATIONS 425 
 
 condenser C 2 is brought into use, the loading coil not being utilized. 1 
 The oscillations in the aerial circuit of the receiving station induce 
 exactly similar ones in the detector circuit, which is tuned to resonance 
 with the receiving aerial by means of L 2 , E v and C 3 . The so-called 
 detector of these oscillations may be simply a crystal of galena D in 
 series with the telephone receivers R. This crystal, like the tungar 
 rectifier of 374, has the property of transmitting a current in one 
 direction only. 2 Were it not for this property the telephone could not 
 be used as a detector, because its diaphragm cannot vibrate with a fre- 
 quency of the order of a million; and even if it could, it would produce 
 sound waves far above the limit of 
 hearing. Because of this rectifying 
 property of the crystal the receiver 
 diaphragm is drawn in only once 
 while the oscillations produced by a 
 given wave train last, this effect being 
 due to the rectified pulsating current 
 which passes in one direction through 
 the receivers and then ceases until the 
 oscillations due to the next spark ar- FlG 459 United Stateg navy 
 rive. Since 1000 of the intermittent standard radio receivers 
 
 wave trains strike upon the aerial each 
 
 second, the operator at the receiving station hears a continuous musical 
 note of this pitch as long as the key K is depressed. The working of the key, 
 however, as in ordinary telegraphy, breaks the regular series of wave 
 trains into groups of wave trains, so that the short and long notes heard in 
 the receivers (Fig. 459) correspond to the dots and dashes of telegraphy. 
 The receiving circuit, when tuned as shown in Fig. 457, (2), is highly 
 selective ; that is, it will not pick up waves of other periods. The loading 
 coils B l and B 2 , as well as the two variable condensers C 2 and C 8 , are 
 usually omitted from small amateur receiving sets; but when this is 
 done, the receiving set is less selective and less sensitive. The resist- 
 ance of the receivers is so high, usually from 1000 to 4000 ohms, that 
 
 1 In the diagram an arrow drawn diagonally across a condenser indicates 
 that, for the sake of tuning, the condenser is made adjustable. Similarly, an 
 arrow across two circuits coupled inductively, like the primary and secondary 
 of the "oscillation transformer" T n , indicates that the amount of interaction 
 of the two circuits can be varied, as, for example, by sliding one coil a longer 
 or shorter distance inside the other. 
 
 2 Crystal detectors have been largely superseded by the "audion " for both 
 wireless telegraphy and wireless telephony.
 
 426 INVISIBLE RADIATIONS 
 
 they do not interfere with the oscillations of the condenser system 
 across which they are placed. The receiving station shown in Fig. 457, (), 
 may also be used for receiving wireless-telephone messages. The simplified 
 circuit of an audion receiving station is shown opposite page 441. 
 
 Although the spark, or wave-train, system of wireless teleg- 
 raphy is still widely used, the " continuous wave " system is 
 rapidly displacing it. Just as sound waves differing slightly 
 in frequency combine to produce the phenomenon of beats 
 ( 396), so electrical oscillations differing in frequency give, 
 when combined, a " beat effect." For instance, if electrical 
 oscillations of, say, 30,000 per second and 31,000 per second 
 combine, beats will occur at the rate of 1000 per second, 
 which is a frequency within the limit of hearing. The elec- 
 trical oscillations mentioned above have a frequency beyond 
 the limit of hearing and hence are said to have^ radio fre- 
 quency ; but the beats being within the range of hearing 
 have an audio frequency. Now let us assume that there is 
 at the transmitting station an alternating-current generator 
 which throws into the aerial powerful undamped oscillations 
 of 30,000 per second; and suppose further that at the receiv- 
 ing station there is an oscillation generator which maintains 
 relatively weak oscillations of 31,000 per second in the local 
 receiving aerial. These weak oscillations produced in the 
 receiving aerial by the local generator make no sound in the 
 receiver, being above the limit of hearing ; but whenever, and 
 as long as, the operator at the transmitting station depresses 
 his key, waves come in at the rate of 30,000 per second, 
 strike against the receiving aerial and develop therein weak 
 oscillations which combine with those already present to make 
 1000 beats per second. These beat effects are rectified by a 
 crystal or by a vacuum tube and passed through the receiver. 
 The listener, therefore, hears long and short musical sounds 
 just as he does when receiving by the spark system. The 
 beat method of receiving is called the heterodyne system.
 
 ELECTKICAL KADIATIONS 427 
 
 489. Modulated continuous waves.* The vibrations consti- 
 tuting articulate speech are exceedingly complex, as may be 
 seen from an inspection of the full-page halftone opposite 
 page 346. Because of this complexity it is impossible to trans- 
 mit speech by means of discontinuous waves (Fig. 460) such 
 as are employed in the system of spark telegraphy described 
 in the preceding section. The parts of the voice lost because 
 
 Direction of propagation 
 
 -4 4 -4 -4 4 4 4 -4- 
 
 FIG. 460. A series of wave trains 
 
 of the gaps between the wave trains would render the language 
 unintelligible. Theoretically the voice could be transmitted 
 by continuous electromagnetic waves having the frequencies 
 of voice vibrations, but such a method is entirely impracti- 
 cable on account of the enormous length of aerial needed 
 to produce such long waves and the tremendous amount of 
 power which would be required. Therefore, the only satis- 
 factory method thus far developed is to transmit speech 
 
 FIG. 461. Continuous, or carrier, waves of radio frequency 
 
 on continuous, or " carrier," waves (Fig. 461) having a 
 frequency (radio frequency) above the limit of hearing. 
 
 At the sending station the continuous waves (Fig. 461) 
 are " modulated " by the voice at the transmitter ; that is, 
 the sound waves of the voice act upon the apparatus in such 
 a way as to alter the otherwise uniform amplitude of the series 
 of continuous waves (Fig. 462). These "modulated" con- 
 tinuous waves on reaching the aerial of the receiving station 
 produce corresponding oscillatory currents in the wires of the 
 
 * The pupil should master 374, 375, 376, 485, 486, 487, and 488 before 
 
 reading the six sections following.
 
 428 
 
 INVISIBLE RADIATIONS 
 
 aerial. By means of a crystal or a vacuum tube, the oscilla- 
 tory currents are rectified into a series of unidirectional elec- 
 trical currents, or pulses, somewhat after the manner indicated 
 
 FIG. 462. Modulated radio-frequency waves 
 
 in Fig. 463. These variable pulses of radio frequency, on 
 reaching the telephone receivers of the listener, produce dia- 
 phragm vibrations of low frequencies (audio frequencies), which 
 
 FIG. 463. Rectified oscillations 
 
 rarely go outside the limits of 100 and 3000 vibrations per 
 second. They are represented by the irregular line in Fig. 464. 
 The vibrations of the diaphragms of the receivers, therefore, 
 
 FIG. 464. Audio-frequency variations 
 
 correspond to the vibrations of the voice of the speaker at 
 the distant transmitting station. 
 
 490. Method of producing continuous waves. One of the 
 most important of the different means of producing high- 
 power continuous waves is by use of the Alexanderson high- 
 frequency alternator (see on opposite page). This is an 
 alternating-current dynamo made in various powers up to 
 200 kilowatts (= 268 horse power), the rotor in some of the 
 machines having the very high speed of 20,000 revolutions 
 per minute. For transoceanic telegraphy these machines cause 
 currents of from 600 to 1200 amperes to oscillate in the sending 
 aerial. This powerful sustained oscillation of electrons in an 
 aerial produces continuous electromagnetic waves (Fig. 461).
 
 ij H i
 
 ELECTRICAL KADIATIONS 429 
 
 491. The vacuum tube. There are several devices by which 
 the voice waves may modulate, or vary the amplitude of, the 
 carrier waves, the most important being the highly exhausted 
 " vacuum tube " (see Fig. 465, the halftone opposite p. 441, 
 and the drawing and legend opposite p. 33). 
 
 In attempting to reach an understanding of an " audioii " 
 amplifier or other form of vacuum tube, it is well to remember 
 
 -Filament (in center) 
 
 Grid (surrounding filament) 
 
 te (surrounding grid 
 front half cut away) 
 
 Terminals 
 
 FIG. 465. A popular form of vacuum tube used in radio receiving 
 
 that a current of electricity is a stream of negative elec- 
 trons which, when passing through a vacuum, move with 
 enormous velocity (thousands of miles per second ( 498)), 
 but when passing along a wire (ordinary conduction) move 
 quite slowly (a few centimeters per second). Now we found 
 in studying the tungar rectifier ( 374) that these negative 
 electrons escape freely from an incandescent filament under 
 certain conditions. When the battery B (Fig. 466) has its 
 -I- terminal connected to the plate P of the vacuum tube and
 
 430 
 
 INVISIBLE RADIATIONS 
 
 FIG. 466. A two-electrode 
 vacuum valve 
 
 its terminal to the filament F, no current can flow across 
 the vacuum so long as the filament is cold. When, however, the 
 filament is maintained at incandescence by a battery A, the 
 negative electrons escape from it and are drawn in. a steady 
 stream across the vacuum by the 
 attraction of the + plate P. This 
 flow of electrons from filament to 
 plate constitutes what is considered 
 by convention to be a current of 
 electricity flowing the opposite way, 
 namely, from plate to filament. We 
 now see how battery A, by keeping 
 the filament in a state of incandes- 
 cence, merely establishes and main- 
 tains one of the conditions under which battery B may discharge 
 a steady current through the vacuum. No electronic flow 
 from the cold plate to the filament is ever possible, because 
 cold bodies do not, except in rare instances (see pp. 441 ff.) 
 eject electrons from themselves. The vacuum tube can 
 therefore be utilized as a 
 vacuum valve, or rectifier, for 
 evidently, if a source of alter- 
 nating current be substituted 
 for the direct current source 
 (battery 5), the vacuum valve 
 would transmit current in one 
 direction only, half of each 
 cycle being held in check. 
 
 If a screen of fine wire G, 
 known as a " grid," be introduced between the filament and 
 the plate of Fig. 466 (see Fig. 467) and the grid be main- 
 tained at a sufficiently high potential by a battery C, the 
 - electrons are repelled back into the incandescent filament 
 and cannot escape from it, and thus the electronic flow is 
 
 FIG. 467. A three-electrode vacuum 
 tube
 
 ELECTRICAL RADIATIONS 
 
 431 
 
 Condenser 
 
 The lamp does not burn 
 
 completely checked; that is, no current flows across the vacuum. 
 If now the potential of the grid be varied, say, from zero 
 to the amount required to stop the electronic flow, the current 
 from battery B through the vacuum is thereby varied from 
 the possible maximum in Fig. 466 to zero. Variation of the 
 grid potential, therefore, affords us a means of controlling 
 and of varying the flow 
 of current through a 
 vacuum tube. Indeed, 
 it is found that sliylit D. C. generator^ 
 changes in the grid volt- 
 age produce surpris- 
 ingly great changes in 
 the current through the 
 tube ; that is, the tube 
 is an amplifier. 
 
 492. Transfer of energy 
 through a condenser. In 
 Fig.468,(l),theE.M.F. 
 of the direct-current dy- 
 namo causes a rush of . 
 
 , . , The lamp burns 
 
 electrons out of one side 
 
 of the condenser while 
 electrons to an equal 
 extent rush into the other side. The sides of the condenser 
 are thus charged + and and they remain so as long as the 
 dynamo runs. It is evident that under these conditions there 
 is no flow of current and that consequently the lamp does 
 not burn. If, however, an alternating-current dynamo is used 
 (Fig. 468, (2)), the alternating E.M.F. causes an alternat- 
 ing rush of electrons which charges the condenser first one 
 way and then the opposite way. It is clear, then, that with 
 an alternating-current dynamo, lamp, and condenser thus 
 arranged we may have an alternating current through the 
 
 A. C. generator 
 
 Condenser 
 
 FIG. 468. Energy transferred through a 
 condenser
 
 48-2 
 
 INVISIBLE RADIATIONS 
 
 lamp which will cause it to light up. Condensers of variable 
 capacity are widely used in the circuits of wireless apparatus 
 as aids in tuning, and they permit passage of electrical energy 
 in the manner explained above. 
 
 493. The receiving station. Fig. 469 represents a "regenera- 
 tive" receiving circuit capable of receiving long or short 
 waves. When the modulated waves (Fig. 462) reach the 
 tuned aeriaj of the receiving station, they develop therein 
 feeble electrical oscillations which induce oscillations in L a 
 of the tuned grid circuit. This varies the potential of the 
 
 /Aerial 
 
 Fi<;. 469. A regenerative receiving circuit 
 
 grid 6? 2 , thus causing corresponding changes in the strength 
 of the electronic current flowing from the incandescent fila- 
 ment F 2 to the plate P 2 and thence back through the plate 
 coil PC. The plate circuit is so tuned with respect to the 
 grid circuit that these current variations in the plate coil 
 react inductively on the coil L 2 connected with the grid cir- 
 cuit to strengthen the original grid-circuit current. This 
 intensifies the variations in potential at the grid, which in 
 turn intensifies the variations in strength of the electronic 
 current from filament to plate, and this still further intensi- 
 fies the variations in potential at the grid, and so on, up to
 
 KLKCTUICAL RADIATIONS 
 
 483 
 
 the limit of the electron supply in the tube. This is thr 
 Armstrong r<<<i<>)icr<ilin' )>rhi<-ipl^ by which very feeble oscilla- 
 tions produced by the incoming waves may be amplified and 
 then, used to intensify the original oscillations. The energy 
 for regeneration comes from the battery B . When the tube is 
 in use the grid tends to accumulate a negative charge which, 
 as we have seen ( 491), would tend to block completely 
 the action of the tube. Therefore, a high-resistance grid leak 
 r is shunted around the condenser (' to permit the return 
 
 Variometer 
 
 FIG. 470. A two-variometer tuned-plate-circuit for receiving short waves 
 
 of such a detrimental accumulation of electrons to the fila- 
 ment F^ by way of r and L Z . The telephone receivers used 
 in wireless work contain thousands of turns of very fine wire 
 wound upon iron and because of the consequent "choke- 
 coil " effect, or impedance, of these coils for high-frequency 
 changes in current strength, the racfoo-frequency variations 
 (Fig. 463) of the plate current pass largely by way of the 
 variable condenser (7 8 , while the slower awdfo-frequency varia- 
 tions (Fig. 464) of the plate current pass readily through 
 the receivers to actuate the diaphragm. 
 
 Fig. 470 shows a two- variometer circuit for the reception 
 of short waves. A variometer is a variable inductance used
 
 434 INVISIBLE RADIATIONS 
 
 for tuning and it consists of two coils in series, one of which 
 revolves within the other. If current is passed through the 
 variometer when the inner coil is turned so that its magnetic 
 field combines with that of the other coil to make the greatest 
 resultant magnetic field, the inductance of the variometer is 
 found to have its greatest value and the adjustment is then 
 for the longer waves, or slower oscillations. If the inner coil 
 is now turned through 180, the resultant magnetic field is 
 at minimum strength ; and, because of the small inductance, 
 the variometer is adjusted to the shorter waves. Intermediate 
 positions of the inner coil are used for wave lengths lying 
 between- these limits. Complete tuning is accomplished by 
 use of the two variometers, the two variable condensers and 
 the sliding contact on the aerial coil. 
 
 494. The transmitting station. The vacuum tube may be 
 used not only as a rectifier, a detector, a modulator, and an 
 amplifier, but under certain conditions as a generator of oscil- 
 lations varying over an extremely wide range of frequency 
 from less than 1 oscillation per second to 300,000,000 or 
 more per second. Nearly all present-day " broadcasting " is 
 done by use of vacuum-tube generators. For high-power 
 long-distance transmission banks of vacuum-tube amplifiers 
 may be used to throw into an aerial an aggregate power of 
 many hundreds of kilowatts. Indeed, at the present time 
 rapid progress is being made in the experimental construc- 
 tion of power tubes each one of which is capable of giving an 
 amazing output. The life of a vacuum tube is generally from 
 1000 to 5000 hours, whereas a high-frequency alternator, such 
 as the Alexanderson, will last for many years. 
 
 It is entirely beyond the scope of this book to explain the 
 actual details of a wireless-telephone transmitting station. 
 However, the method used at present in high-power long- 
 distance transmission is indicated in Fig. 471 and may be 
 outlined as follows: Air vibrations produced by the voice
 
 ELE< TKK.'AL RAD1ATK >.\ S 
 
 485 
 
 make variations in the current of the primary circuit of the 
 telephone transmitter ( 376). This induces corresponding 
 E. M. F.'s in the secondary circuit, which impresses audio- 
 frequency variations of potential upon the grid of a vacuum- 
 tube modulator. The resulting changes of audio frequency 
 in the current of the plate circuit of the modulator corre- 
 spondingly affect the output of the high-frequency oscil- 
 lation generator. This modulated radio-frequency output is 
 
 Three-electrode vacuum-tube 
 high-frequency oscillation generator 
 
 \Aerial 
 
 (^illation 
 Transformer 
 
 Telephone 
 Transmit te 
 
 Vacuum-tube modulator 
 FII;. 471. High-power long-distance wireless-telephone transmitting station 
 
 amplified by a bank of three-electrode power tubes and is 
 then delivered to the aerial through an oscillation trans 
 former. In broadcasting stations (see opposite p. 429) a 
 weaker and somewhat simpler arrangement of tubes is used. 
 
 XOTK. The following reference books will prove helpful to teachers and 
 to those pupils who desire a more complete understanding of "wireless " : 
 (1) BUCIIER, Practical Wireless Telegraphy, Wireless Press, 326 Broad- 
 way, New York City ; (2) GOLDSMITH, Radio Telephony, Wireless Press. 
 326 Broadway, New York City; (3) H.USMANN and others, Radio Pimm- 
 Receiving, Van Nostrand Co., 8 Warren St., New York City; (4) Mn:i - 
 CKOKT, Principles of Radio Communication, John Wiley and Sons, 432 
 Fourth Ave., New York City ; (5) SCOTT-TAGT.AKT, Thermionic Tubes in 
 Radio Telegraphy and Telephony, Wireless Press. 326 Broadway, New 
 York City ; (6) Elementary Principles of Radio Telegraphy and Teleph- 
 ony (Radio Communication Pamphlet 1), 79 pages, illustrated, 10 cents, 
 Superintendent of Documents, Government Printing Office, Washington, 
 D.C., 1922.
 
 436 
 
 INVlSUiLK RADIATIONS 
 
 Although transoceanic telephonic communication has been su< 
 cessf'ully and repeatedly accomplished (see opposite p. 441). 
 no regular service for such communication has yet be 
 established. 
 
 495. The electromagnetic theory of light. The stud 
 electromagnetic radiations, like those discussed in the pre- 
 ceding paragraphs, has sho\vn not only that they have the! 
 speed of light but that they are reflected, refracted, and! 
 polarized, in fact, that they possess all the properties of light 
 waves, the only apparent difference being in their greater 
 wave length. Hence modern physics /<'</></* li'jlit <ix <tn i-/i',-f,-<>- 
 magnetic phenomenon; that is, light waves are thought to be; 
 generated by the oscillations of the electrically charged parts' 
 of the atoms. It was as long ago as 1864 that Clerk-Maxwell, i 
 (see opposite p. 102), of Cambridge, England, one of the 
 world's most brilliant physicists and mathematicians, showed : 
 that it ought to be possible to create ether waves by means 
 of electrical disturbances. But the experimental confirmation 
 of his theorv did not come until the time of Ilert/.'s experi- 
 ments (1SSS). Maxwell and Hertz together, therefore, share 
 the honor of establishing the modern electromagnetic theory 
 of light. 
 
 ('ATHOI)K AM) IvONTGKN It AYS 
 
 496. The electric spark in partial vacua. Let n and // ( I i- . I TIM ! 
 
 be the terminals of an induction coil or static machine : <- andjfj electrode-, j 
 sealed into a glass tube lid or 80 
 centimeters Ion-: //. a rubber 
 till"- leading to an air ]>uni]> by 
 which the tube may be ex- 
 hausted. Let the roil be started 
 ln-t'ore the exhaustion is begun. 
 A spark will pass between a and 
 //, since nh is a very much shorter 
 
 Fn.. 47^. Discharge in partial vacua 
 
 path than rf. Then let the tube be rapidly exhausted. When the pres- 
 sure has been reduced to a few centimeters of mercury, the discharge 
 
 51652