UC-NRLF E. E N T DAY Sag |trtm*r0 WIRELESS TELEGRAPHY AND WIRELESS TELEPHONY WIEELESS TELEGEAPHY AND WIEELESS TELEPHONY AN ELEMENTARY TREATISE BY A. E. KENNELLY, A.M., Sc.D. Professor of Electrical Engineering In Harvard University WITH EIGHTY-FOUR ILLUSTRATIONS NEW YORK MOFFAT, YARD AND COMPANY COPYRIGHT, 1906, 1909, BY MOFFAT, YARD & COMPANY Published, November, tgob Reprinted with additions, February, 1909 Reprinted, May, 1909, February, 1910 PREFACE TO SECOND EDITION. Since the first edition of this book went to press, much development has taken place in wireless telegraphy. The range distance to which messages have been carried has in- creased. The number of ship and shore sta- tions has approximately doubled. The num- ber of wireless messages transmitted has greatly increased. A very notable service re- cently rendered by wireless telegraphy has drawn public attention forcibly to its value as a means of protecting lives at sea. Early on the morning of January 23rd, 1909, the east bound White Star Liner " Republic" was in- jured by collision with the west bound Italian Liner " Florida," about fifteen miles south of Nantucket light ship, in a dense fog. A hur- ried wireless general call for assistance brought several vessels to the rescue and, in particular, the White Star Liner " Baltic," that happened to be sixty-four nautical miles west of the scene of accident. So dense was the fog that the 271151 PREFACE Baltic steamed for twelve hours over a zigzag course of some two hundred nautical miles be- fore reaching the helplessly injured and drift- ing vessel. Even then the search would prob- ably have been futile, if wireless messages be- tween the ships and the shore station at Sia- conset had not assisted the meeting. About sixteen hundred persons were ultimately trans- ferred to the Baltic from the Republic, which shortly afterwards foundered in deep water. No loss of life occurred except in the actual collision. In spite, however, of the above achievements of wireless telegraphy during the past two years, the development of wireless telephony has been still more pronounced. During that time, the range distance of wireless telephony has remarkably increased; so that, although far below that of wireless telegraphy of the present time, it is comparable to that of wire- less telegraphy ten years ago. It has, there- fore, seemed desirable to add to this book sev- eral chapters on wireless telephony, while bringing the subject of wireless telegraphy up to date. Wireless telephony differs from wireless telegraphy in details rather than in fundamen- PREFACE tals ; but as an achievement of the human race, the transmission of the voice to great distances on the ripples of electromagnetic waves is, in one sense, a far greater extension into free space of the range of individual personality than any form of wireless telegraphy thus far attempted. Although it has been the endeavor to pre- sent to the reader the fundamental and es- sential principles both of wireless telegraphy and wireless telephony, rather than any de- scription of particular systems or inventions; yet in treating of wireless telephony, much has been taken directly from the publications of Prof. R. A. Fessenden, who has done so much to extend the knowledge and practice of wire- less telephony in America. The author also desires to express his in- debtedness to the writings of Messrs. B. S. Cohen, G. M. Shepherd, and Ernst Ruhmer. Cambridge, Mass., February, 1909. CONTENTS CHAPTER PAGE I INTRODUCTORY i II WAVES AND WAVE-MOTION 3 III MAGNETISM AND ELECTRICITY .... 14 IV ELECTROMAGNETIC WAVES GUIDED OVER CONDUCTORS 29 V ELECTROMAGNETIC WAVES GUIDED BY SINGLE WIRES 36 VI RADIATED ELECTROMAGNETIC WAVES GUIDED BY THE GROUND 40 VII UNGUIDED OR SPHERICALLY RADIATED ELECTROMAGNETIC WAVES ... . 64 VIII PLANE ELECTROMAGNETIC WAVES . . 75 IX THE SIMPLE ANTENNA OR VERTICAL ROD OSCILLATOR 97 X ELECTROMAGNETIC WAVE-DETECTORS OR WIRELESS TELEGRAPH RECEIVERS . . 124 XI WIRELESS TELEGRAPH WORKING ... 149 XII TUNED OR SELECTIVE SIGNALING ... 173 XIII MEASUREMENTS OF ELECTROMAGNETIC WAVES 191 XIV INDUSTRIAL WIRELESS TELEGRAPHY . . 198 XV CONSIDERATIONS PRELIMINARY TO WIRE- LESS TELEPHONY 209 XVI THE PRINCIPLES OF WIRE TELEPHONY . 225 XVII PRINCIPLES OF WIRELESS TELEPHONY . 239 vii CHAPTER I INTRODUCTORY ELECTRICITY is not in its infancy, populai impression notwithstanding. Electric applica- tions, such as the telephone, the wire telegraph, the electric lamp and the electric motor, are very familiar in modern life and have been for a num- ber of years. Electricity has reached adoles- cence in these directions. But wireless teleg- raphy, the most recent of electric applications, is, perhaps, in its infancy. It is only about ten years old. There is something fascinating about the way in which electricity works. It is as swift as it is stealthy. Electric impulses move over wires at enormous speeds and yet the action is invisible and inaudible, appealing to no sense directly. A telegraph wire runs overhead say from New York to Buffalo and the New York sending operator closes the circuit at his key. Almost instantly, and certainly within about one-tenth of a second, the little lever of the receiving in- strument at Buffalo responds. The electric im- 2 WIRELESS TELEGRAPHY pulse reaches its destination perhaps as rapidly as we can turn our thoughts from one end of the wire to the other. We cannot see the electric impulse rush along the wire, we cannot hear it travel. We can only picture the transfer in our imagination. We can see the wire and we know it travels along that. If we cut the wire anywhere, the electric current is stopped. When we turn to wireless telegraphy, how- ever, we are deprived even of the consolation of a guiding wire to aid the imagination. The phenomenon of the wire telegraph is a mystery, but a familiar one to which the wire is a clue. The new phenomenon of the wireless telegraph is a yet more elusive mystery with no clue, at first sight. Nevertheless, we shall see in the sequel that there is not so much difference between the two cases as might at first be sup- posed. The relations between wireless teleg- raphy and wire telegraphy resemble the rela- tions between sound distributed in the open air, and sound channeled, or confined, within a speak- ing tube. CHAPTER II WAVES AND WAVE-MOTION SINCE wireless telegraphy is carried on by means of electro-magnetic waves, it is desirable to ex- amine into the nature of those types of waves with which we are more familiar, before taking up the consideration of the less familiar electro- magnetic waves. Free Ocean Waves A wave is a progressive disturbance, or a dis- turbance which moves along through some kind of medium. The most familar type of wave is the disturbance in the level of water by some displacing force, such as wind, or the splash of a falling body. Even in calm weather, we usually find a wave-motion, commonly called swell, on the surface of the ocean. We then find displacements of the ocean level, alternately up and down, or high and low, moving along the surface, These waves have a certain direction, say from west to east, in the horizontal plane or plane of ocean level. They also have a cer- 3 4 WIRELESS TELEGRAPHY tain speed in this direction. On the deep ocean this speed may be, say, 12 meters per second (39.4 feet per second, or 26.8 miles per hour). Of course this does not mean that in calm weather a cork, or lifeboat, floating on the sea, is moved along by the swell at this high speed. The cork, or the boat, bobs up and down on the swell, with only a very small movement in the swell's direc- tion. But it means that a torpedo boat would have to steam in this case at a speed of 26.8 miles per hour from west to east in order to keep abreast of any one particular roller in the swell. It is to be observed that the waves advance through and over the water, without carrying the water bodily along to any marked extent. Another noteworthy point in the waves of ocean swell is that there is commonly a fairly definite and uniform wave-length. This wave- length is measured in the direction of wave- motion, and may be taken as the distance from crest to crest of successive waves. This wave- length in open ocean swell is often about 100 meters (328 feet or 109.3 yards). A tightly stretched string 328 feet long, held east and west, would in such a case just span the de- pression between successive roller crests. The waves of the ocean manifestly contain energy. That is to say, work had to be done WAVES AND WAVE-MOTION 5 originally by the winds to produce them, and the waves are capable of doing a good deal of work before being brought to rest. Attempts have been made, in fact, to obtain work from waves at suitable points on the sea-coast by means of engines operated by the rise and fall of the waves. Sound- Waves in Air A type of wave of which we are constantly receiving impressions through our ears, but which is more difficult to analyze than the ocean wave, is the sound-wave in the atmosphere. Waves of sound are invisible and hence the difficulty we experience in becoming familiar with their forms, speed and other properties. We learn that sound in air is a disturbance in its density and pressure which moves through the air at a definite speed. If we fire a pistol in the air, the explosion in the barrel displaces the air, or compresses it, in the immediate neigh- borhood of the discharge. The zone of com- pression moves off into surrounding air with substantially the same speed in all directions, if the air is calm, and is followed immediately by a zone of expanded air; just as a hollow or de- pression follows a hump or elevation in an ocean wave. 6 WIRELESS TELEGRAPHY Fig. i gives a diagrammatic view of a single sound-wave of compression and dilation, shortly after the wave has moved off from the explosion FIG. i. Diagram Indicating a Single Sound Wave Shortly after Leaving Its Source. or origin of the disturbance, o. The wave is hemispherical in form, and the diameter of the hemisphere is a a, at the instant considered. The external portion of the hemispherical wave shell contains compressed air, indicated by the concentric semi-circles. The internal dotted portion, immediately following, contains ex- panded air. After a brief interval of time the wave will have expanded to the contour indi- cated in Fig. 2, where the diameter of the hemis- FIG. 2. Diagram Indicating a Single Sound Wave Which Has Separated Itself from Its Origin by Five Wave-Lengths. pherical shell is b o b. The length of the wave is the distance b c, measured radially from the front to the back of the wave, or from the out- WAVES AND WAVE-MOTION 7 side to the inside of the hemispherical shell. If this wave-length be, say 100 meters (328 feet or 109.3 yards), then the distance o c, by which the wave has already removed itself from the origin, is 500 meters (1640 feet or 546.7 yards). Another brief interval of time would bring about the condition indicated in Fig. 3, where the wave cd FIG. 3. Diagram Indicating a Single Sound- Wave Which Has Separated Itself from Its Origin by Eleven Wave-Lengths. has expanded hemispherically to the diameter c o c. The length of the wave, c d remains as it was in the earlier stages, assumed as 100 meters (328 feet or 109.3 yards); but the distance o d which the wave has now placed between itself and the origin is uoo meters (3609 feet or 1203 yards). Observers in balloons floating in the air at points such as e or / on the wave front, would hear the sound of the explosion simul- taneously with observers on the ground at points c c. 8 WIRELESS TELEGRAPHY Speed oj Sound-Waves The speed with which the sound-waves moves radially outward in all directions is approxi- mately 333 meters per second (1090 feet per second, or 746 miles per hour), depending slightly upon the temperature and humidity of the air. We Cannot see the hemispherical shell of disturbed air expanding; but we can picture the process to the mind's eye. In an hour, the expansion would carry the radius of the hemis- phere to a distance of 1200 kilometers (746 miles). But the density of the atmosphere would be infinitesimally small at an elevation amounting to this distance, and sound being a disturbance of air cannot travel where the air ceases to exist. Consequently, the hemispheri- cal form of the wave must disappear at great dis- tances, for lack of air above the origin Dilution of Intensity in Sound- Waves The energy residing in the sound-wave would be the same in the successive states of Figs, i, 2, and 3, if there were no expenditure of energy in friction during the motion. That is to say, we may suppose that a certain part of the energy in the explosion at o was stowed away in the wave of compressed and dilated air. But the space WAVES AND WAVE-MOTION 9 occupied by the wave in the stage of Fig. 2 with radius o a, 150 meters (164 yards), would be 13.6 millions of cubic meters; in the stage of Fig. 2, 382 millions of cubic meters, and in the stage of Fig. 3, 1660 millions of cubic meters (corres- ponding to 17.8, 500, and 2160 millions of cubic yards respectively). It is evident, therefore, that the energy in the wave is constantly spread out into more space, or diluted, as the wave expands; so that the energy in a given volume, such as i cubic meter, is constantly diminishing. This is another way of saying that the intensity of the wave diminishes as time goes on, and the radius of the wave increases. The loudness of the explosion as noted by an observer at a in Fig. i would be considerably greater than that noted by an observer at b in Fig. 2; or again, than that noted by an observer at c in Fig. 3. If we were to place a sufficiently sensitive re- cording barometer anywhere in the neighbor- hood of the explosion, and carefully observe the barometer record as the noise of the explosion occurred, we should expect to find that the barom- eter would record, just before the explosion, a horizontal straight line a b, Fig. 4, corresponding to the reading of the barometer at that time, say 760 millimeters (29.92 inches) of mercury. When the sound of the explosion arrived, the io WIRELESS TELEGRAPHY barometer would rise very slightly to b c, then fall to e, and then return to the normal straight line / g. The elevation from b to c would mark the degree of compression in the first half of the wave, and the depression from d to e would similarly mark the following dilatation. The crest height k c, or k t e, would measure the amplitude of the disturbance or amplitude of the wave. If the barometer were located close to FIG. 4. Ideal Barometric Record at the Time of Pas- sage of the Noise of an Explosion. the origin of the explosion, as at a in Fig. i, the amplitude of the pressure disturbance record, and the amplitude of the recorded sound, would be relatively large. If, on the other hand, the barometer were placed further from the origin, as at c in Fig. 3, the amplitude both of the recorded disturbance and of the sound-wave at the barometer would be smaller. Ordinary sound-waves possess so little energy, or have so small an amplitude, that recording barometers show no trace of them. Expressing the same thing in another way, the impression- WAVES AND WAVE-MOTION n producing mechanism of the ear is far more sensitive to the disturbances of pressure in sound- waves than the ordinary barometer. Very powerful explosions are capable of pro- ducing sound-waves of sufficient intensity to be observed at great distances. The great ex- plosion of the volcano Krakatoa, near the Sunda Straits, in the year 1883, is stated to have been heard at distances greater than 3200 kilometers (2000 miles). The outgoing wave affected baro- meters all over the world, and left traces on recording barometers. This wave is stated to have traveled at a speed of 1130 kilometers per hour (700 miles per hour), to have swelled at the antipodes to Krakatoa, in 18 hours, and to have spread out again over the globe. It was not finally lost sight of until it had passed around the globe several times. When the first out- going wave passed over Singapore, a port dis- tant about 830 kilometers (516 miles) from the origin of the disturbance, the gas-holder of the town is stated to have leaped into the air several feet up and down. Sound- Wave Trains If instead of producing a solitary explosion at the origin, and a corresponding solitary wave moving off radially therefrom, we produce a 12 WIRELESS TELEGRAPHY rapidly and rhythmically repeated disturbance; as in blowing air through an organ-pipe, or forcing air through a syren-wheel, a succession of waves is produced, and the sensory effect produced on the ear is that of a tone or musi- FIG. 5. Diagram Representing the Succession of Sound- Waves Emitted from the Origin O When a Simple Musical Note Is There Produced. cal note. A succession of outgoing waves is diagrammatically represented in Fig. 5. The shaded areas there correspond to zones of com- pressed air, and the intermediate unshaded areas to zones of dilated air. Eight complete waves are indicated. If the note sounded be the deepest E of the double-bass viol, making 40 complete vibrations a second, then the length WAVES AND WAVE-MOTION 13 of each wave will be the fortieth part of 333 meters, the distance which sound travels in a second, because 40 complete waves will occupy the space covered by advancing sound in one second. Each wave will, therefore, be 8.33 meters (27.34 feet or 9.11 yards) in length, and the length o d of eight wave-lengths will have been covered in one-fifth of a second from the commencement of the sound. The curve at D o or o D' gives the trace-record that we should expect a recording barometer would give at d after all of the eight waves passed by, on the supposition that this pure, musical note was sustained uniformly for eight complete cycles of the disturbance. If, however, the string of the double-bass, instead of being excited by the bow, were plucked by the finger in such a manner that the vibrations died away, then the record of the supposed sensitive barom- eter at d might indicate the curve d, o of de- caying amplitudes. CHAPTER III MAGNETISM AND ELECTRICITY HAVING paved the way for a consideration of electromagnetic waves by a few outlines of sound waves in air, we may now fitly turn attention to magnetism and electricity. Wind and Its Energy Everyone is familiar with the fact that wind is an active or disturbed state of the atmosphere, a movement of the air. We ordinarily understand wind to be a uniform movement of the air in any one given direction, and we ordinarily under- stand by eddies or gusts, twisting or vortical movements of the air, but, in general, wind may include both linear movement and vortical move- ment, since one cannot occur in the atmosphere without involving the other. The material for the creation of a wind is always present, for this material is the air itself. We only need to ener- gise the air in a particular way, to make it move forward. Energy must be expended in pro- ducing a wind ? and energy resides in the wind. 14 MAGNETISM AND ELECTRICITY 15 If we employ a hand-fan to produce a local breeze, we must expend muscular energy, or do work on the fan, to force the air into motion, and the air once set in motion contains energy or can do work by moving, for example, light obstacles in its path. Consequently, we may say that wind is air, plus energy given to it in a particular way. Air is a material fluid. It forms an ocean on the surface of this earth, and we live at or near the bottom of this air-ocean. Air gravitates, or pulls upon the mass of the earth. Each indi- vidual atom of air gravitates, and the sum total of all the individual pulls exerted on the earth amounts to a pressure of about i kilogramme per square centimeter of surface (14.25 Ibs. per sq. inch). The Invisible Ether It is generally believed that all space, includ- ing the interior of solid bodies, is permeated by an immaterial fluid called the universal ether. The ether is just as invisible as air. Whether it consist of matter or not, it is immaterial in the sense that it apparently does not gravitate. It does not directly appeal to any sense, but it is much easier to assume its presence everywhere than to deny its existence. If we take a vacuum- 1 6 WIRELESS TELEGRAPHY tube, i.e., a sealed glass tube from the interior of which the air has been almost entirely re- moved, it can be shown experimentally that sound cannot move across the interior of the tube, but light passes across it, and so do radiant heat and gravitational force. We can- not believe that these activities are transmitted through absolutely empty space. Something must transmit them, for they are transmitted at definite speeds. This something is named the ether. Beyond its powers of transmitting energy, hardly anything is yet known about the ether. Its structure, and the manner in which it permeates space, are still unsolved riddles. Nature oj Electricity and Magnetism As soon as the ether is postulated to be a uni- versal fluid or medium in which all matter swims, so to speak, many things may be accounted for which otherwise we could not even attempt to explain. Electricity and magnetism, for ex- ample, may be accounted for in a general way. Just as wind is, we have seen, a particular ener- gized condition of the circumambient air, so both electricity and magnetism are particular energized conditions of the universal ether, which underlies the air and everything else. It is not BO easy, however, to define the nature and rela- MAGNETISM AND ELECTRICITY 17 tions of these particular energized conditions. We cannot at present say, for example, that electricity is the same kind of motion of the ether that wind is of the air. If we do not yet apprehend the nature of the ether itself, how shall the task be undertaken of denning its ener- gized conditions? The energized conditions might be statical, and involve no motion of ether, like the energy of a stationary coiled-up spring; or they might be dynamical, and involve modes of motion of the ether. In any event, it seems clear from the known laws of electromagnetism that there is a definite mutual relation between the two energized conditions of ether, electricity and magnetism, such that as soon as either is defined the other also is immediately determined. In mathematical language, one is the "curl" of the other. If, for instance, electricity should be a definite kind of tension or static stress long- wise, then magnetism would be a definite kind of twist or crosswise static stress, and reciprocally. Or, if electricity should be a simple, straight- forward motion, or streaming, of the ether, then magnetism would be eddy motion or vortical rotational motion, or spin, of the ether, and re- ciprocally. It is surprising how much is known concerning the laws of action and behavior of electricity and magnetism, considering the little 1 8 WIRELESS TELEGRAPHY that is known of their absolute fundamental nature. We can control electricity and magne- tism remarkably well, considering that we do so from beyond a hitherto impenetrable veil that does not admit of perceiving the things directed. It is evident that whatever may be the precise nature of electricity and magnetism, the widely admitted postulate of the universal ether requires that the material for either or both is omnipresent. Just as the material for wind is always present in the circumambient air, and all we need is the application of energy to the air in a particular way in order to produce a wind, so the material for electricity or magnetism is universally present, and all we need is the application of energy to the ether in particular ways. Consequently, electricity and magnetism may be regarded as the ether, plus particular forms of energy. Magnetic Flux and Its Properties If we consider an ordinary, permanent horse- shoe magnet, such as is indicated in Fig. 6, we find that all around it, and particularly between its poles N and s, there is a certain invisible ac- tivity which possesses both direction and inten- sity. In the illustration, the direction is roughly indicated by the broken lines with arrow-heads, and the intensity by the relative crowding or con- MAGNETISM AND ELECTRICITY densation of these lines. The direction of this magnetic activity in the air between the poles is seen to be from the north pole N to the south pole S. This is strictly speaking a pure con- vention. It might have been originally agreed to draw all the arrows in the opposite direction. All that is certain is that there is a definite polarity about the sys- tem, and that the actions pertaining to the north pole are distinctly in- verse to the actions per- taining to the south pole, magneticians all agree- \ ""- '" / ing upon the direction """*. .*.*'* shown. The north pole . - , i . i . r ,, FIG. 6. Diagram of Mag- lS the pole which, if the netic Flux in the Space magnet were freely sus- Between the Poles of a Permanent Magnet, pended, would seek for, or point approximately toward, the north geo- graphic pole of the earth, or the earth's magnetic pole near the Greenland end of the earth's axis. That is, the N end is the north-seeking end. As roughly indicated in Fig. 6, the magnetic activity, or magnetic flux as it is called, is densest or most intense, in the air between the opposing 20 WIRELESS TELEGRAPHY pole-tips, or where the air-space separating the poles is shortest. As we leave this region, the magnetic flux becomes thinner, or weaker. A peculiarity about this flux is that it always re- turns back upon itself in closed loops or chains. In other words, magnetic flux is always con- tinuous and re-entrant. At first sight it appears to be discontinuous, because it seems to com- mence at one pole and end at the other. But it can be shown experimentally that the flux con- tinues through the interior substance of the steel magnet, and each loop, such as N A S, completes a circuit B C D E F within the substance of the magnet. Provisional Hypothesis as to Nature of Magnetic Flux Although the real nature of this magnetic flux is not yet known, yet it may help us to follow the actions of electromagnetic waves later on, if we assume, for the purposes of description, that magnetic flux is a streaming motion of the ether. On this assumption, a permanent magnet is a force pump which draws the ether in at the south pole, through the substance of the steel in the interior, and forces it out at the north pole. With no friction, this streaming would not neces- sarily absorb energy, and we know that perma- MAGNETISM AND ELECTRICITY 21 nent magnets may be designed to retain their magnetism without sensible diminution for an indefinitely long time. Energy oj Magnetic Flux Although magnetic flux does not need energy to be expended in order to keep it going, yet energy has to be expended to create it. That is, magnetic flux contains energy, or has energy always associated with it. As long as the mag- neitc flux persists, the energy resides quiescent with it. When the flux disappears, its energy disappears also. Consequently work must be done to create magnetic flux, and magnetic flux is able to do work or give up its energy when it disappears. Between the opposed pole-tips N S, we may consider, on the above hypothesis, that the stream of ether is densest, and receding from this region the stream gets weaker. The streaming is steady both at any particular point for all con- sidered time, and for all points at any one time. The magnet ether-pump is steady. The pump- ing action is due to activities in the molecules of the steel. Each molecule of iron is supposed to be a little individual ether pump, by virtue of /hternal activities as yet only dimly guessed at. When the horseshoe is magnetized, all of the mo- 22 WIRELESS TELEGRAPHY lecules are caused to align themselves in parallel directions, or to face the same way, whereby they all pump the ether in the same general di- rection. Within the iron molecules, the pump- ing activities are believed to be electric; but into these we need not enter. The point here to be observed is that in the air-space outside of the magnet, the steady magnetic flux produces no electric action. In this air-space we find mag- netism but not electricity. If, however, we move the magnet, and with the magnet the system of magnetic flux pertaining thereto, there will be electric action produced where the magnetic flux lines are carried through space. If, for example, the magnet be lifted bodily toward the observer without twisting, feeble electric forces will be brought into play in directions lying within the plane of the horseshoe. In the region between the poles these electric forces will be directed, during the motion, in the direction from A to D. The intensity of these electric forces will be pro- portional to the speed with which the magnetic flux moves sidewise. If the magnetic flux moves longwise, or parallel to itself, there is no electric action set up, but if the magnetic flux moves sidewise, or crabwise, there is electric action set up. It is on this action that all dyna- mos depend; namely, upon relative sideways MAGNETISM AND ELECTRICITY 23 motion between magnetic flux and an electric conductor to pick up and utilize the induced electric force. Steady and stationary magnetic flux is thus unaccompanied by electric action, but unsteady, varying, or sidewise-moving mag- netic flux sets up electric action. Electric Flux and Its Properties Turning now to electricity, Fig. 7 represents a vertical metallic rod, and terminal balls, sup- . --. Si .->*.. ^ * 4 FIG. 7. Diagram of Electric Flux in the Space Between an Electrified Rod and Disk. ported by an insulating holder not shown, in air above the center O, of a horizontal, insulated metallic disk COD. This insulated pair of conductors may be electrically charged either by a frictional electric machine, an influence ma- chine, a spark coil, or a voltaic battery. That 24 WIRELESS TELEGRAPHY is, the charge may be communicated from any suitable electric source. The charge will be re- tained, because the rod is insulated from the disk, and if the insulation could be made perfect, the charge would be retained indefinitely. Let us suppose that the rod is positively charged and the disk negatively. In the air-space between the electrified rod and disk there is an invisible influence which possesses both direction and intensity. It is called electric flux. This electric flux is dia- grammatically represented in Fig. 7 by the little arrows which proceed, by convention, from the positive charge to the negative charge. The arrows are drawn on little lines of points, instead of little broken lines as in Fig. 6, in order to dis- tinguish them from lines of magnetic flux. Be- tween A and O, where the air-space is shortest, the electric flux is most densely crowded, or its intensity is greatest. As the separating air- space increases, the flux density weakens. Energy o] Electric Flux Energy always resides in the electric flux, so that each and every part of the region permeated by the electric flux represented in the illustration contains energy. The energy is not uniformly distributed. It is greater per unit volume at a MAGNETISM AND ELECTRICITY 25 point like F than at a point like G. It is stowed away in proportion to the square of the electric flux density, so that in two regions one of which has double the flux density of the other, there will be four times more energy per cubic centi- meter, or cubic inch of space, in the former than in the latter. As long at the electric flux per- sists, this energy resides therein or accompanies it, and when the flux disappears the energy has disappeared. This energy is communicated to the ether in the insulating air between rod and disk at the time of their charge. Provisional Hypothesis as to the Nature of Electric Flux In conformity with the provisional hypothesis already adopted for magnetic flux, stationary electric flux may be assumed to be an elastic twist or stress in the ether; so that the whole system of ether tends to revolve clockwise about the rod A B as axis, when looking down on the disk from above. The screw or twist will have maximum intensity along the central line O A, and is resisted by the elastic rigidity of the ether. The elastic energy of the twist is the total energy of the electric flux as summed up throughout the entire electric field, or permeated insulating 26 WIRELESS TELEGRAPHY region. The amount of electric energy that air can be made to hold without breaking electrically, or disrupting into a spark discharge, depends upon the atmospheric pressure and upon the shape of the opposed electrified surfaces. At ordinary atmospheric pressures, and parallel opposed surfaces, the most favorable form, the energy that air can hold is limited to about 480 ergs per c. c., or i foot-pound per cubic foot; i.e., the work done by lifting one pound one foot high, to the cubic foot of air space under powerful electric stress. Electric flux at rest differs from magnetic flux at rest in the fact that the former is discontinu- ous while the latter is continuous. The magnetic flux, as we have seen, always forms closed loops or chains in space. Steady electric flux, on the contrary, always starts from a positive charge and ends on a negative charge. In the case of opposed conductors, the charges always reside on their surfaces, and thus the electric flux al- ways starts on the surface of the positive con- ductor and ends on the surface of the negative conductor. Tensions in Electric and Magnetic Fluxes The electric flux, like magnetic flux, always possesses the property of pulling along its own MAGNETISM AND ELECTRICITY 27 direction at the same time that it pushes side- ways. The curved arrow lines of Fig. 7 merely indicate the direction and the relative crowding of the electric flux from point to point of this particular electrified system; but if we suppose that each of these lines is a little elastic thread, exerting a certain mechanical tension, and if we also suppose that each such elastic thread exerts a repulsion sideways against its neighbors, or tries to secure all the elbow-room it can, we get an idea of the static forces which reside in such a stationary electric flux. Thus, the line H G K, in addition to its own tension, pushes sideways against the adjacent lines h g k and L M D. The resultant effect is a tension, or attractive pull, between the rod and the disk, or the familiar attractive force between oppositely electrified bodies. As long as the electric flux remains steady and stationary, no magnetism, or magnetic force is produced. There will be a tendency to move any electrified object, such as a pith-ball, along the electric flux in Fig. 7, but there will be no tendency to affect the direction of a magnetic compass-needle. If, however, the electrified system be moved bodily sidewise, without losing the charge, feeble magnetic forces will be de- veloped in directions at right angles both to the 28 WIRELESS TELEGRAPHY moving electric flux and to the direction of motion. Just as sidewise-moving magnetic flux gener- ates electric flux, so sidewise-moving electric flux generates magnetic flux. CHAPTER IV ELECTROMAGNETIC WAVES GUIDED OVER CONDUCTORS Automatic Movement 0} Electric Flux over Conducting Surfaces IN order to bring electric flux into movement, it is not necessary to move a charged system of conductors, the flux will set itself in motion if an opportunity is offered to let it slide upon a pair of conductors. Electric Current Over a Pair o] Wires If we bring a long pair of parallel insulated metallic wires M N, P Q, Fig. 8, simultaneously into contact, one with the rod and the other with the disk, as indicated in the figure, the electric flux immediately takes advantage of the exten- sion of the system thus offered and glides away, guided by the wires. It may be considered that the sidewise repulsion of the flux tends to make it spread its boundaries in this way, whenever possible. The electric charge moves out along the wires hand in hand with the electric flux, the positive charge spreading along the upper wire 29 30 WIRELESS TELEGRAPHY M N, in Fig. 8, and the negative charge along the lower wire P Q. The electric flux runs along with these charges, always bridging over between the positive and negative charges. The phenom- enon of the movement of two parallel moving ?;;;* 4 i i i i * * 4 at I ii i 4 * i 4 FIG. 8. Electric Flux Wave Moving Over Pair of Parallel Insulated Wires. charges with the moving electric flux between them and linking them, constitutes an electric current, or electric discharge. The effect of bringing the two parallel wires into contact with the charged electric system of Fig. 7, is, therefore, to let the charge escape over the wires, and to set up thereby an electric cur- rent over the wires. The current rush takes place in the form of a wave. Electric flux and its associated energy move off the disk into the insulating air-space between the wires. Creation of Magnetic Flux by Moving Electric Flux We have already seen that sideways-moving electric flux generates magnetic force and flux. ELECTROMAGNETIC WAVES 31 As soon, therefore, as the electric flux begins to move, half of the electric flux energy disappears and is replaced in the form of magnetic energy. Instead of having stationary electric flux in the confined insulated system of Fig. 7, we have moving electric flux, and magnetic flux associated therewith, or advancing with it. Consequently, although we can have either stationary electric flux alone, or stationary magnetic flux alone, we cannot preserve them independently when they move freely in an insulator. Any wave of elec- tric disturbance is an electromagnetic wave, be- cause in it electric and magnetic fluxes are tied up together. Electromagnetic Wave Guided by a Pair o) Parallel Wires The distributions of the electric and magnetic fluxes in the advancing wave of Fig. 7 are illus- trated in Fig. 9, where M P are the sections of the two parallel wires in a plane at right angles to their length. The electric flux-paths are indi- cated, as in previous instances, by lines of points with arrow heads; while the magnetic flux-paths are indicated by broken lines with arrow heads. The wave is supposed to be receding from the observer, and the upper wire M is carrying the positive charge, while P, the lower wire, carries WIRELESS TELEGRAPHY the negative charge. The electric flux, there- fore, emerges from the surface of the wire M and terminates upon the surface of the wire P. If the metal of which the wires are composed be supposed to conduct perfectly, the electric flux will skim over the sur- faces of the wires and not penetrate into their mass. The more im- perfect the conductivity of the wires, the more deeply the moving elec- tric flux will penetrate into them. The magnetic flux at netic Fluxes in Electro- the center O of the loop bv Two Parallel Wires, has the direction D O C, perpendicular to the loop. At all other points the magnetic flux-paths are circular in this plane of cross-section, or cylindrical with regard to a length of the wires. Both the electric and the magnetic flux-paths are systems of circles, and it is to be noticed that at every point they intersect each other perpendicularly. That is, any one circle crosses all the intersecting circles at right angles. It is also to be observed that where the electric FIG. 9. Electric and Mag magnetic Wave Guided T\v ' ave Receding by W Observer. from ELECTROMAGNETIC WAVES 33 flux runs densest, so does the magnetic flux. The densest electric and magnetic fluxes are found close to either wire. Both the fluxes get weaker as we recede from the wires. In fact the inten- sity of the electric flux in any single pure electro- magnetic wave is always and everywhere numeri- cally equal to the intensity of the magnetic flux at the same point and time, when each is meas- ured in its appropriate units. At an indefinitely great distance from the loop of active wires the density of the fluxes is nil. Speed 0} Electromagnetic Waves The speed of sound waves in air we have seen to be in the neighborhood of 333 meters per second, (1090 feet per second or 746 miles per hour). But the speed of a free electric wave in air is enormously greater, being approximately 300,000 kilometers per second (186,400 miles per second), or 7^ times around the world in a second. This is also the speed at which light travels in air. That is to say, no difference has yet been determined between the speed of electro- magnetic waves in air and the speed of light. Energy is carried in the advancing electro- magnetic wave indicated in Figs. 8 and 9. The energy is the energy residing in all the electric flux that moves on, plus the equal amount of 34 WIRELESS TELEGRAPHY energy in all the associated and interlinked magnetic flux. This energy is carried away from the original stock of electric energy in the air- space of the electrified system in Figs. 6 and 7. The energy was originally bound up in the sta- tionary electric flux. The advancing electric and magnetic fluxes in the wave robbed the charged system of flux and of energy and trans- ported that energy whithersoever they went. Summing up the conditions which we have noted in the guided electromagnetic wave of Figs. 8 and 9 we may state them as follows: An electric or magnetic disturbance associated with a pair of insulated aerial wires propagates itself along the wires at the speed of light. The wave consists of electric and magnetic fluxes, which are always perpendicular to each other and to the direction in which the wave is moving. If the two wires are parallel, the fluxes are dis- tributed cylindrically; i.e., circularly in any plane perpendicular to the wires. The energy in each flux is the same, and the intensity of the two fluxes is the same at every point. The energy per unit volume varies as the square of the in- tensity of the moving fluxes. The advancing wave conveys this energy with it. On the sur- faces of the wires are opposite electric charges, moving with the flux, and supporting the ends of ELECTROMAGNETIC WAVES 35 the electric flux. The entire series of associated phenomena is an electric current. Guided electromagnetic waves properly belong to the domain of ordinary wire telegraphy, or to the transmission of electric power by wires. As such, they lie outside of the province of this en- quiry. It may suffice to observe that the steady electric current found in any electric circuit oper- ated by a dynamo, or a voltaic battery, is merely the sum of what is usually a large number of superimposed electromagnetic waves of the type above considered. These waves are kept stream- ing out of the dynamo, and are also reflected back from the distant end, or other parts, of the circuit; so that after a brief interval of time we have a complex aggregate of waves present. We may now proceed to the study of semi-guided electromagnetic waves, or waves in which the electric flux is held at one end only on an insu- lated artificial conductor, or is guided by but a single wire. CHAPTER V. ELECTROMAGNETIC WAVES GUIDED BY SINGLE WIRES Electromagnetic Wave Guided by a Single Wire and the Ground IF we place the disk C O D of Fig. 7 upon the level surface of the ground, taking pains to secure good electrical conductivity in the adja- cent soil, and charge the vertical rod A B, while supporting the same insulated above the center O, there will be but little change effected in the charge or distribution of the electric flux by rea- son of the grounding of the disk. From an electrical point of view, we shall merely have in- definitely extended the area of the conducting disk. If we now bring a single very long insu- lated horizontal wire M N, like an ordinary tele- graph wire into contact with the charged rod, as indicated in Fig. 10, the electric charge and elec- tric flux will immediately rush out at light-speed over this wire in an electromagnetic wave. The electric flux will be guided by the wire M N on its 36 ELECTROMAGNETIC WAVES 37 positive ends; but the negative ends will be un- constrained, or left loose to themselves. In this sense, the wave is only singly guided. If the surface of the earth G G be assumed to conduct perfectly, the electric flux will skim over this sur- face, and a negative charge will also distribute . -r\\i i * j 4 t i i 1 1 1 1 I f*s*S>$& iiiiiM*** 1 * FIG. 10. Single- Wire Guided Electromagnetic Wave. itself over the same, advancing with the electric flux. The electric flux under these circumstances will spread out over the surface of the ground G G in such a distribution as would be effected if the ground were removed and in its place a second wire were run parallel to M N and as far beneath the level surface G F G as the first wire M N is above it. The distribution is indicated in Fig. n, where M is the section of the wire in a plane at right angles to its length, and G G is the conducting surface of the ground. The wave in this case is supposed to be advancing towards the observer. The lines of points show the paths of electric flux issuing from the positive charge moving along the wire M. They terminate at a 38 WIRELESS TELEGRAPHY negative charge distributed over G G, and mov- ing over the same with like rapidity. N is the position of the " image" wire, which, in the absence of the ground G G, would be able, as in Fig. 9, to produce the same distribution of fluxes above the level G O G, as is developed with the FIG. ii. Electric and Magnetic Fluxes in Electro- magnetic Wave Guided by a Single Wire Over a Conducting Ground Surface. Wave Advancing Toward Observer. conducting ground. If the ground conducted perfectly, the electric flux would not penetrate below the surface; but would slide frictionless over the same. In practice, the conductivity of the ground is never perfect and the fluxes pene- trate beneath the surface to a greater or less ex- tent, with a corresponding expenditure of energy in the soil. Nevertheless, the conditions are ELECTROMAGNETIC WAVES 39 ordinarily regarded as a slight deviation from the condition of perfect earth conduction as indicated in the Figure. The magnetic flux is established in cylindrical distribution, as shown, by the motion of the 'electric flux at the light-speed. The two fluxes have equal densities and energies at any given point, and between them they transport to a dis- tance, along the wire, the energy originally bound up in the stationary electric flux of the charged system of Fig. 7. The process thus initiated pertains to single- wire telegraphy, the usual type of wire telegraphy. The currents employed in telegraphy consist of such electromagnetic waves, either singly, or in superposition by confluence. CHAPTER VI RADIATED ELECTROMAGNETIC WAVES GUIDED B\ THE GROUND Radiation of Electromagnetic Waves by an Electric Disturbance or Explosion THE electromagnetic waves considered in the last two chapters were set in motion by bringing a wire, or a pair of wires, into electric connection with the charged electric system, and allowing the electric flux to overflow from that system on to and along the wires. But electromagnetic waves may also be set in motion by sudden dis- turbances of an electric charge. In such cases the emitted waves are likely to be radiated in all directions in a manner resembling the expansion of a sound wave in air as outlined in Chapter II. Let us suppose the rod and disk system to be charged, as indicated in Fig. 7, after the disk has been placed horizontally upon the surface of good conducting ground. Instead of allowing a wire to approach the rod and discharge it, let the system be discharged by a spark between A and O (Fig. 7). Let us assume that the discharge is 40 RADIATED ELECTROMAGNETIC WAVES 41 completed by a single spark of extremely short duration; so that the entire system of electric flux collapses precipitately. The flux near the axis A O is the first to disappear into the spark, then the longer and outlying flux. Last of all, the longest and furthest reaching flux issuing from B, will run down the rod and vanish at the gap A O. If the discharge be delayed, or the charge allowed to dribble slowly across the gap A O, as, for example, by the action of rough and oxydised opposing surfaces, the collapse of the entire umbrella-shaped electric flux system of Fig. 7 may take place without any appreciable electro- magnetic disturbance. The energy stored away in the flux will be expended in heating the path of discharge, and, when the process is complete, the flux having disappeared, the discharge stops, and there is no aftermath. If, however, the collapse of the umbrella flux system in Fig. 7 is sudden, the rapid descent of the flux down the rod and into the gap A O sets up magnetic forces and magnetic flux. We have already seen that when electric flux moves, it establishes magnetic flux in a direction across itself, and also across the direction of motion; also that energy is imparted to the magnetic flux at the expense of the electric flux energy. If we 42 WIRELESS TELEGRAPHY consider, for example, the particular flux path L M D in Fig. 7, it is evident that during the brief interval of collapse it tends to run at light- speed into the successive positions H G K, h g k, and so on to A O, when it disappears into the spark discharge. But the movement of this flux element will set up magnetic flux directed in planes parallel to the disk, and pointing towards the observer. On the other side of the rod A B, the similar inrush of electric flux will beget a magnetic flux in horizontal planes, i.e., planes parallel to the disk, but directed away from the observer. Considering all the actions that occur simultaneously, it will be evident that a concen- tric ring system of magnetic flux will be set up, as in Fig. 12, around the rod A B as axis, each ring being in a horizontal plane. Part of the energy of the original electric flux of Fig. 7 is delivered to this ring distribution of magnetic flux; so that when the discharge is sudden, a lesser total energy tumbles into the spark than when the discharge is slow. Fig. 12 indicates, in plan view, the ring dis- tribution of magnetic flux accompanying the collapse of the umbrella electric distribution of Fig. 7, on the passage of a spark at O A. The eye of the observer is supposed to be situated immediately over the disk C D and rod B. The RADIATED ELECTROMAGNETIC WAVES 43 magnetic flux streams are all directed clockwise, and they lie in various horizontal planes. According to our provisional theory, we found in Chapter III, at page 25, that the original charge of the rod-and-disk system of Fig. 7 gave FIG. 12. Plan View Diagram of Magnetic Flux Distri- bution Accompanying Collapse of Electric Flux Around Charged Rod. a right-handed or clockwise screw-twist to the surrounding ether, as viewed by an observer looking down on the disk. The ether may be supposed to have taken a right-handed, or clock- wise, elastic set or strain, under the stress of the electric flux, which stress is supposed to be re- 44 WIRELESS TELEGRAPHY sisted by the elastic rigidity of the ether. The stress is a maximum at the axis O A. If the ether gives way, by the disruption of the aii particles at this point, the rigidity fails to oppose the stress, and the ether flows clockwise bodily in the magnetic stream lines of Fig. 12. It comes to the same thing, therefore, whether we view, in imagination, the collapsing electric flux during the sudden discharge of the system, and watch the magnetic flux rings spring into exist- ence with the downward electric motion, or whether we view in imagination the ether give way under the screw twist of the original charge, and watch the flow of the ether in obedience to that twist when the spark occurs at the axis. Electromagnetic Wave Generated by Sudden Collapse of Electric Flux Distribution The ring magnetic flux, as in Fig. 12, accom- panies the collapsing electric flux down the rod; it also sets up, by reaction, an external wave of upward and outwardly rising magnetic ring flux in the counter-clockwise direction. This rising magnetic flux sets up in its turn electric flux across itself. The direction of this rising shell of electric flux is indicated in Fig. 13 at A C and A D. It is directed from the rod to the disk, as in the original charge distribution of Fig. 7, RADIATED ELECTROMAGNETIC WAVES 45 This hemispherical shell of downward electric flux, and ring magnetic flux, expands radially outwards in all directions at the light-speed. The collapsing ring magnetic flux of Fig. 12, when it reaches the disk, is reflected back and up FIG. 13. Electric Flux Induced by the Ring Distribu- tion of Magnetic Flux in Fig. 12. the rod, still clockwise in direction, but moving upwards immediately behind the shell C B D. It sets up an electric flux in the directions indi- cated at A, Fig. 13. The two concentric hemi- spheres of electric and magnetic fluxes detach themselves from the rod in the manner dia- grammatically indicated in Fig. 14. In the external hemispherical shell w P w, the electric flux is downwards and the magnetic flux lies in counter-clockwise rings centered on the polar axis B P. In the internal hemispherical shell x Q x, the directions are reversed, the electric flux being upwards, and the magnetic clockwise. In a very brief interval of time after the dis- charge of the system by the passage of the spark, we have complete disappearance of electro- magnetic charges, fluxes and energy in the rod- 46 WIRELESS TELEGRAPHY jnd-disk system, while a hemispherical electro- magnetic wave moves off radially with the light- speed, the radius of the hemisphere being theoret- ically 300,000 kilometers (186,400 miles) after one second of time. The thickness of the *//./ !V ? .*" * V, * '<*\ % * I* '. '.' FIG. 14. Vertical Cross-Section and Plan of Single Expanding Electromagnetic Wave. double-layered hemispherical shell remains con- stant, but since the energy in the shell also re- mains constant, in the absence of absorption, the density of the fluxes and their energy per unit of volume rapidly diminish. In other words the energy per cubic meter of space in the wave rapidly diminishes. RADIATED ELECTROMAGNETIC WAVES 47 The feet of the electric flux lines skim over the surface of the ground, assumed to be perfectly conducting. At the external edge w w w w, a negative ring charge runs out radially over the ground surface at the light- speed. At x x, a similar positive ring charge runs out radially at the same speed. This succession of running electric charges, linked together by loops of elec- tric flux, constitutes a single cycle 0} alternating current flowing along the ground. From its external aspect, the expanding hemi- spherical electromagnetic wave has electric flux distributed along meridians of longitude, sym- metrically disposed with respect to the polar axis B Q P. The magnetic flux is distributed in circles of latitude, the smallest circles being near the pole, and the greatest near the equator or ground surface. Resemblances Between Solitary Explosion Waves of Sound and Electromagnetism The solitary hemispherical electromagnetic wave of Fig. 14 bears some resemblance to the solitary hemispherical sound wave in air of Fig. 3. Each consists of a double layer, the disturbance in the external layer being of the opposite sign to that in the internal layer. On the other hand, there are notable differences. There is an 48 WIRELESS TELEGRAPHY enormous difference in speed (nearly a million to one). The electric wave has a polar node at P Q and the sound wave has none. The electric wave is propagated in the ether, the sound wave in a gaseous substance. If the earth's surface is not perfectly conduct- ing, and in practice it is far from being perfectly conducting, the electric flux will penetrate to some extent into the soil, carrying also magnetic flux with it. The fluxes which sink in this way expend their energy in warming the soil very slightly and the hemispherical wave is thereby drained of a part of its energy, or is subjected to frictional losses in running along the ground. The energy per cubic meter of wave shell will thus diminish more rapidly than would be ac- counted for by the mere increase in bulk of the expanding wave shell. Electromagnetic Wave-trains The discharge of a rod-and-disk electrified system does not ordinarily give rise to but a single electromagnetic wave, such as is depicted in Fig. 14. On the contrary, the discharge gen- erally gives rise to a series, or train, of successive waves of diminishing amplitude, each feebler than its predecessor. The rate of diminution depends upon the amount of heat expended in RADIATED ELECTROMAGNETIC WAVES 49 the discharge spark, and to a lesser extent upon other structural details, but the amplitude of oscillation usually falls to one half, or loses fifty per cent., in about two complete swings, or after two successive waves have been thrown off. If the spark remained uniform, the amplitude would in such a case fall to one half again in two more swings, or to one quarter of the original amplitude in a total of four swings, or to one eighth in six swings and so on. Consequently, the amplitude of the successive waves emitted by a simple rod oscillator soon dwindles into in- significance. Analysis of Oscillatory Current on Rod and the Generation of Waves It may be of interest to consider in some detail the process of emitting a train of hemispherical electromagnetic waves from a rod-and-disk sys- tem laid on a perfectly conducting ground and suddenly set into electric oscillation by a spark discharge. Such a system may be briefly de- scribed as a simple vertical oscillator. Reference is made to Fig. 15, where the vertical rod is repre- sented in nine successive stages of the process of manufacturing and shipping half of an electro- magnetic wave, and two such diagrams would be 4 I i I 4 4 4 4 ft t ft K -I f f t I I t I f t IT 4? If 41 I i i i I i i i I I I I f I I t 4 4 i i i i i -H4 f f t I f I I bo !l O 2 o 8 J 3 002 1.1 -ff*- 1 H 41 If KH'^'d i e *-S I- B* I IS1 i 5- 50 RADIATED ELECTROMAGNETIC WAVES 51 required to illustrate the delivery of one com- plete magnetic wave into free space. Instead of commencing with the insulated rod just prior to the spark discharge, as in Fig. 7, it is more con- venient to commence with the condition indi- cated at A, Fig. 15, where the electric flux has just completed its movement at light-speed up to the top of the rod, or has climbed up to the sum- mit of the conductor. The flux arrows touching the rod are pointing inwards, indicating a nega- tive charge on the surface of the rod. Ascending electric flux is marked by solid arrows and de- scending flux by broken or dotted arrows. Accompanying the upward movement of the converging flux which has culminated in the con- dition at A, there will be, as previously stated, an associated magnetic flux. The direction of this magnetic flux is indicated by the device of a small circle and a small upright bar, the former on the left-hand of the rod, and the latter on the right hand. The circle may be looked upon as the feathers, or heel, and the small bar as the point, or barb, of an arrow in a horizontal plane bent into a semicircle about the rod on the side remote from the observer. The device is illustrated in detail in Fig. i5a, where P Q is a vertical section, and p q a plan view, of a converging plane of electric flux terminating on the rod E F at its WIRELESS TELEGRAPHY FIG. i5a. Sectional Ele- vation and Plan of *a Plane Electromagnetic center and sliding up- wards from E to F over the rod's surface, like a ring over a peg. When, as shown in Fig. 1 5a, the small circle c is on the left hand, and the bar b on the right hand, the ring magnetic flux c f d b' is disposed clockwise about the rod as v i ewe d from above. Counter clock-wise movement of Directions of Fluxes, magnetic flux calls for Key Diagram to Fig. 15. . . , the circle on the right hand and the bar on the left. Rules for Memorizing the Directions oj Motion, and oj Electric Flux and oj Magnetic Flux in Any Single Free Electromagnetic Wave. There is a simple law connecting the directions of electric and magnetic fluxes in any simple plane electromagnetic wave. It may be ex- pressed mnemonically in either of the following ways: (i) Draw an arrow in the direction in which the electric flux points. Let the head of the arrow be supposed to be the head of a man who RADIATED ELECTROMAGNETIC WAVES 53 runs in the direction in which the flux is running. Then the man's side-extended right hand will point in the direction of the magnetic flux. This rule is illustrated mnemonically in Fig. 16. The FIG. 16. Memory Picture for Recalling Directions of Electric and Magnetic Fluxes in Waves. point of the Greek warrior's sword is supposed to be magnetized and supporting iron nails. (2) When electric flux, converging upon a con- ducting rod or cylinder, as in Fig. 15 a, points its arrows inwards like the V's in the face of a clock (See Fig. 17), then if the flux is moving like the light that makes them visible from the clock towards the observer, the magnetic flux M will point circularly in the direction of the Motion of 54 WIRELESS TELEGRAPHY the clock hands. Reversing either the electric arrows E, E, E, or their movement towards the observer, reverses the direction of the magnetic flux; but reversing both, leaves the magnetic flux M pointing clockwise. Release of Electromagnetic Wave from Simple Rod Oscillator Applying either of the rules to the upwardly moving and inwardly pointing horizontal electric flux on the rod at A, Fig. 15, it will be evident that the magnetic flux is directed clockwise, as viewed from above. The moment the elec- tric flux reaches the top of the rod it is reflected Re- FIG. 17. Mnemonic Clock Diagram of Relative from the free end. flection of electric flux Directions of Electric Flux E and Magnetic Flux M in an Electro- from a free end always the direction magnetic Wave Coming Towards the Observer requires of the magnetic flux to be reversed, leaving the electric flux arrows unchanged in direction, but moving backwards, or retreating. This relation is a consequence of either of the above rules. Referring to Fig. 15 a, the flux which has reached RADIATED ELECTROMAGNETIC WAVES 5^ the top of the rod at light- speed must keep on moving, and since it can go no further up- wards, it commences to descend at light-speed. In Fig. 15, descending electric flux is shown by dotted line arrows and ascending flux by heavy line arrows. The reversal of the direction of motion of magnetic flux around the rod, in changing from going up to coming down, delivers a blow, by inertia, to the surrounding external ether. In other words, the jerk required to reverse the magnetic flux around the rod between stages A and E of Fig. 15, sets up a counter- jerk or kick in the surrounding ether. The kick sets up two free waves traveling in opposite directions. The electric fluxes in these two waves are mutually opposite ; but the magnetic fluxes conspire clock- wise. These relations are indicated by the ex- ternal pairs of arrows at A, which start into existence at the instant of reflection of the central wave from the top of the rod. At the instant of time represented at B, the second stage, there has been a movement of the flux gliding over the rod, and also a movement in each of the two external free waves. Taking these in order, the gliding flux has commenced to move down at the top, or to double back upon itself, the three lower layers still climbing, but the 56 WIRELESS TELEGRAPHY leading layer descending. The electric flux ar- rows point inwards at all parts of the rod, but the magnetic fluxes are opposed at the top, or tend to neutralize there. In the external waves, the ascending one has advanced one stage, and its wave-front is at a. The descending one has reached the conducting disk, or ground, at the base, and has been reflected from this surface. Reflection at a normal conducting surface en- tails reversal of the electric flux arrows but no reversal of the magnetic flux arrows. Conse- quently the lowest layer of the free external de- scending wave at A has turned its arrows from outwards to inwards, and the ring is moving bodily upwards. At the next stage, indicated at C, the leading half of the wave on the rod has doubled back on the following half, the electric flux all pointing inwards, and the magnetic fluxes completely neutralizing. There is no resultant magnetic flux at this instant around the rod. In the free external waves, the ascending front has reached b. Half of the external wave which commenced moving downwards at A, has doubled back upon itself and is ascending. Continuing this process to E, we find that the ekctric flux slipping on the rod is now all moving downwards and the magnetic flux is all counter RADIATED ELECTROMAGNETIC WAVES 57 clockwise. This means in ordinary language that the electric current in the rod is at this in- stant a maximum in the downward direction. The electric flux arrows all point inwards, so that there is a negative electric charge all over the rod. In the external ether, the direction of motion is altogether upward, with the electric flux inward and the magnetic flux clockwise. The front of the emitted wave has now reached d. A new spark will now cross the air-gap at the base of the rod, not shown in the Figure, and the electric flux will pass over the conducting spark column to the ground at the base. It is reflected back from there with reversal of electric flux ar- rows, but persistence of magnetic flux; so that there is no kick or disturbance generated this time in external space. At F, the head of the wave conducted to the ground over the rod has turned around and is ascending. The external free wave has reached e and is clear of the ground. At G, there is complete neutralization of elec- tric fluxes, there being no resultant charge on the rod at this instant. But the current wave, as gauged by the conspiring magnetic flux, is at its maximum, and directed upwards. In the last stage, at I, the flux gliding along the rod has reached its full development in the 58 WIRELESS TELEGRAPH^ upward direction, and is about to be reflected back from the free end at the top. This will in- volve a reversal of magnetic flux and a new shock to the surrounding ether, but in the opposite di- rection to the shock delivered at A. A new pair of oppositely moving external waves is thereby created, as indicated at I. After eight more stages have been passed, the external wave will have completely deployed, and the length of this emitted wave will be just four times the length of the rod. The half wave con- tained between h and i, at I, is twice the length of the rod. Reviewing the various stages, it will be evident that the electric flux reaches its maximum result- ant value near the top of the rod; while the magnetic fluxes, on the contrary, are always in opposition at reflections from the top and reach their maximum near the bottom. This is an- other way of saying that the electric charge, and electric 'voltage or potential, develop maximum amplitude in oscillation at the top of the rod, and the electric current at the base. The diagram of Fig. 1 5 must not be interpreted too literally. The actual flux distributions are somewhat more complex, and the radius of the emitted wave at I, say, is not exactly equal in all directions to the height h I. The emitted wave RADIATED ELECTROMAGNETIC WAVES 59 becomes sensibly hemispherical, however, after the radius has acquired the length of one-half wave. It is sufficient to observe that there are two sparks for each complete electromagnetic oscillation, and one complete oscillation of elec- tric pressure and current on the rod is accom- panied by the emission of one complete free FK>. 1 8. Train of Seven Hemispherical Electromag- netic Waves of Decaying Amplitude Liberated by a Rod Oscillator at Center. hemispherical wave into space. The energy contained in the fluxes of the wave are drawn from the energy of the fluxes oscillating up and down the rod, which are thereby constantly being weakened, and reduced in density. A diagrammatic vertical cross- section of seven complete hemispherical waves is seen in Fig. 18, as emitted from a simple rod and grounded disk oscillator o at the center. The first wave has attained the radius o w, about 28 rod-lengths from the center; while the seventh wave has just 60 WIRELESS TELEGRAPHY been released. The first wave was the most powerful and is represented in the heaviest lines. Each successive wave is weaker and weaker. It should be remembered, however, that the heavi- ness of the lines in the illustration only relates to the strength of each wave at the moment of its release, or at the moment when it passes a given point; for as each wave expands, its energy per cubic meter or cubic foot rapidly diminishes, be- cause the volume occupied by the wave rapidly increases. Consequently, the energy per cubic meter in the first and strongest wave, by the time it has reached the position w x w, may be much less than the energy per cubic meter in the shell of the last and feeblest, but most condensed, wave at the center, O. At any point, R, the advance of the wave is in the radial direction O R, at the speed of light. The directions of electric and magnetic fluxes in this train of waves is indicated by the devices already used in Figs. 15 and Deviation oj Electromagnetic Waves from the Hemispherical Form Owing to the Curva- ture oj the Earth According to the theory above outlined, the hemispherical waves of Fig. 18 would travel over a perfectly flat conducting ground surface at RADIATED ELECTROMAGNETIC WAVES 6l light- speed and the polar radius o x would be 300,000 kilometers (186,400 miles) long after one second. In practice, when such waves are thrown out by a rod oscillator we have to deal with a moderately conducting spheroidal world surface. This changes the shape of the waves and makes them less geometrically simple. The waves will conform to the curvature of the earth and sea, the successive rings of positive and negative charge running out in all directions over the earth's surface at light-speed, and the feet of the electric flux shells gliding along with them. The waves continue to weaken in intensity per unit of volume as they run, both on account of expanding volume, and owing to sinking into the imperfectly conducting earth surface at their feet, i.e., by frictional dissipation of energy in the ground. Condensation of Fluxes, and of Their Energy, Towards the Equatorial Zone Although the direction of movement of these waves is always radially outwards, or perpendicu- lar to the wave front, yet the density of the elec- tric flux in each wave is not uniform all over the surface of the hemisphere. It is greatest near to the ground, and least near to the pole. This is roughly indicated in Fig. 14. Some of the up- 62 WIRELESS TELEGRAPHY ward electric flux in the inner hemispherical shell turns back to the earth a short distance above the surface. The higher we rise in the shell, the less flux we find in it, and when we reach the polar axis Q P all the flux has ceased. The same condition necessarily attaches to the associated magnetic flux in the wave. The result is that the energy contained in unit volume of either or of both fluxes near the ground is greater than it would be according to simple uniform distribu- tion, by about 60 per cent. In other words, the flux densities and energy in any hemispherical electromagnetic wave are greatest at the ground or equator and dwindle towards the pole. Relations Between Wave-length, Frequency, and Periodic Time From the relation that the length of the waves emitted by a simple vertical rod oscillator is four times the length of the rod, we can readily find the duration of each wave as it passes any point on the ground or in the air above. Let us sup- pose that the oscillator is kept supplied with electric energy so as to keep sending out waves for just one second of time, and that the length of the rod is, say, 25 meters (27.34 yards). Then the length of each wave would be 100 meters (109.36 yards) measured along any radius. But RADIATED ELECTROMAGNETIC WAVES 63 in one second of time, the radius of the outermost wave would have reached to 300,000 kilometers (186,400 miles) and this distance would cover 3,000,000 such wave-lengths. It is clear then that the oscillator must have emitted three mil- lions of waves in the one second of time con- sidered, and also that the time occupied by the wave to pass a given point would be of a second. This is stated in the customary phraseology by saying that the frequency of oscil- lation is 3,000,000 cycles per second and that the periodic time of such waves is &o'oo t innr second. In a similar manner, if we know any one of the three quantities wave-length, frequency, or periodic-time of an electromagnetic wave in air, we can instantly assign the other two, because the velocity of propagation is the light- speed in air of 300,000 kilometers (186,400 miles) per second. It is believed that there is hardly any difference between the speed of light in air and in free ether space devoid of air. CHAPTER VII UNGUIDED, OR SPHERICALLY RADIATED ELECTRO- MAGNETIC WAVES Generation of Spherical Electromagnetic Waves by the Discharge oj a Double-rod Oscillator IF we take a pair of conducting rods A B, C D, and suitably support them insulated in line with each other as indicated in Fig. 19, then on charg- ing the system electrically, with the rod A B, say, positive, the electric flux lines will permeate all the surrounding air in the distribution roughly depicted. So long as the insulation is maintained there will be no magnetic flux. If, however, we raise the electrification to a point at which the air breaks down between the two opposed extremi- ties B C, the electric flux system collapses and runs in towards the spark. At the same time magnetic flux is generated in rings around the axis A D by the inrushing electric flux. The sudden generation of magnetic flux gives a shock to the surrounding ether which sends off a spherical electromagnetic wave into surrounding space at light-speed. 64 UNGUIDED ELECTROMAGNETIC WAVES 65 The contours of five successive spherical waves is given diagrammatically in Fig. 20, with refer- ence to the pair of discharging rods at o, the cen- ter of disturbance. If the two electrified rods ' * > *&" ^ 4* * > ~ + ... *' . * / . * rf r ^ v N * * J ' - : " N - ^ \ I / ' i }fA*\\ **'*' * ; * * i.'mfc\* * * ; * * FIG 19. Electric Flux Between Oppositely Charged Conducting Rods. have no source of energy to maintain their oscil- lations except the original charge, the successive outgoing waves carry that energy away, while some of the remainder is dissipated in heat in the spark and also in the surfaces of the rods, so that the oscillations rapidly die away. It is not im- possible, however, to supply electric energy to 66 WIRELESS TELEGRAPHY the rods as fast at it is radiated externally and dissipated locally, so as to maintain the oscilla- tions indefinitely, although it is very difficult to do this experimentally. In such a case, the wave FIG. 20. Diagram of Section through the Polar Axis of a Train of Five Spherical Electromagnetic Waves Emitted by a Double-Rod Oscillator at the Center. train would go on extending and expanding in- definitely at light-speed. If the rods could some- how be placed jn free space, remote from the earth and all conductors, the spherical waves would keep on moving radially outwards. In practice, on this earth, the waves must almost immediately strike the surface of the ground on one side at least, and be reflected there, not to UNGUIDED ELECTROMAGNETIC WAVES 67 speak of the influence of neighboring walls, trees, etc., so that the pure spherical form cannot be maintained. Figure 20 shows that the polar axis P P is in the line of the rods at the center o. On this axis the electromagnetic fluxes and energies disappear. At and near the equator Q Q, the fluxes are densest and their energies are a maximum, for any given radial distance from the center. The length of each wave is four times the length of either rod, as in the hemispherical waves considered in the last chapter; or it is twice the length of the double- rod oscillator A D of Fig. 19. Spherical Electromagnetic Waves Identical with Long-wave Polarized Light Physicists are now agreed that such an oscilla- tor as above described, if kept supplied with energy for radiating electromagnetic waves, would emit light. That is to say, such electro- magnetic waves constitute light; although not ordinary light such as is recognized by the eye. The main difference between such waves and ordinary light lies in the wave-length. The human eye is able to recognize as light electro- magnetic waves whose length lies between the 68 WIFELESS TELEGRAPHY limits of 0.4 micron* (-gu^nr inch), in violet light, and 0.8 micron (-^.TOT inch) in red light. Electromagnetic waves which are either shorter or longer than this are not directly visible; al- though they may still be objectively regarded as light. Modern Electric Theory That All Matter is Ultimately Electricity Ordinary matter, such as a piece of match- wood, is believed to be made up of ultimate par- ticles called molecules, too small to be seen by the microscope. Molecules are chemical combina- tions or chemical groups of atoms. Atoms are the supposed ultimate particles of elementary substances, or the smallest pieces of such elemen- tary substances which can exist separately as such. Atoms in their turn are now supposed to be each constructed of much more minute or ultra-ultra-microscopic electrical charges, called electrons, there being a definite number and or- ganization of electrons to each atom of an ele- mentary chemical substance. Consequently, all the matter in the universe is ultimately con- structed, according to this theory, of definitely * The micron is the term used in microscopy for the one-millionth part of one meter from the Greek mikros, small. It is usually designated by the Greek letter UNGUIDED ELECTROMAGNETIC WAVES 69 organized electric charges, or of electricity. An atom of hydrogen, for example, is supposed to comprise about 800 electrons in some definite organized orbital or planetary movements, an atom of oxygen about 10,000 electrons in a dif- ferent grouping of orbits ; and so on, for other elementary substances. When atoms are heated, as for example, the hydrogen and carbon atoms of wood, by setting fire to a match, the electric charges or electrons within the atoms are regarded as being forced into rapid and violent oscillation, whereby elec- tromagnetic waves are radiated off. Since these atomic oscillators are of ultra-microscopic di- mensions, so too are the lengths of some of their electromagnetic waves. Those waves whose lengths lie between 0.4 and 0.8 micron are per- ceived by our eyes as light. A pair of little rod oscillators, as in Fig. 19, each about 0.2 micron l n g (12 0*0 60 mc h) excited into sustained radia- tion, would give off waves of red light, the longest waves by which the retina of the eye is affected. Virtual Ultra-microscopic Oscillators in Heated Matter and Their Emitted Waves Looked at in another way, the shortest electro- magnetic waves that have yet been produced bv 70 WIRELESS TELEGRAPHY the discharge of electric rods or spheres are a few centimeters, or inches, in length. In order to make visible light in the same manner, we should have to use ultra-microscopic particles as discharging bodies. On the other hand, the waves employed in wireless telegraphy usually vary between 100 meters (109.4 yards) and 10,000 meters (6.21 miles) in length. The latter would include in one wave length 25,000,000,000 waves of violet light, the shortest detected by the human eye. The velocities of all electromagnetic waves be- ing apparently the same, whatever .their length, their frequencies differ in a similar range. The frequency of a lo-kilometer (6.21 miles) wave would be 30,000 cycles per second, or each oscil- lation would occupy 3 ^ second in execution. But the frequency of violet light is 760 millions of millions per second. Each color of the spec- trum has its own frequency and corresponding wave-length. There is one other difference between visible light and spherical electromagnetic waves pro- duced by electric discharge between conductors as in Fig. 19. This is in regard to the directions of the poles of the waves. In Fig. 20, the polar axis of the waves always lies in the line of the rods, no matter how far the waves may extend into UNGUIDED ELECTROMAGNETIC WAVES 71 space. There is no energy emitted along this axis. Let us suppose that the rods are ultra- microscopic, so that they are enabled to emit waves short enough to affect the eye, and that their energy of oscillation is somehow sustained. Then the point o in Fig. 20 would be a luminous point, shining with one particular color of the spectrum. The brightness of the point would, however, be greatest in the equatorial plane Q Q, and it would dwindle to zero as we moved the eye to the polar axis. An eye at Q would see the shining point; but an eye at P would see nothing. This is contrary to experience with glowing ma- terial points. A lighted match or glowing point, sends out rays in all directions. The discrepancy is accounted for by the fact that the polar axis of the electric disturbance in a luminous point, supposed to be due to oscillat- ing electrons, is constantly shifting its direction in space. One wave may have its polar axis vertical, but after a few more have passed by, there will be a wave with its axis horizontal, and later again the axis will be vertical; so that in a^ single second of time including millions of mil-! lions of waves, the atomic electric oscillators will have turned into all directions and will have made many gyrations. Consequently the eye will have received in that time many waves in 72 WIRELESS TELEGRAPHY their equatorial zone and also many in their polar zone, so that the average effect will be the same in every direction. We need only suppose the rod oscillator of Figs. 10 and 20 rotated about the spark center in all directions at great speed during the emission of a long train of waves, to see, in imagination, the effect that would be pro- duced upon the eye of an observer at any distant point. Electromagnetic waves or light waves in which the polar axis remains fixed in space, as in Fig. 20, are called plane- polarized waves. The plane of polarization is the equatorial plane Q O Q ? parallel to which all the magnetic flux-paths are disposedc Ordinary, or non-polarized, light may be artificially plane-polarized by optical methods. We may say then that ordinary visible light consists of electromagnetic waves of sustained amplitude t.e. 9 not merely a few decaying oscil- lations within a certain sharply limited range of small wave-lengths or high frequencies, and with the polar axis in all directions in rapid succes- sion. Ordinary daylight contains almost all the wave-lengths within the visible range, showing that vast numbers of atomic oscillators of differ- ent " lengths " are simultaneously operating in the glowing solar surface and are mingling or superposing their electromagnetic waves. These TJNGUIDED ELECTROMAGNETIC WAVES 73 waves reach us in about 500 seconds after they leave the atomic electric oscillators in the solar atmosphere. Solar Wireless Telegraph Waves, in Broad Sense, Necessary to Lije oj Human Beings In a certain sense, therefore, every shining star in the heavens is constantly sending out spherical electromagnetic waves within the range of visual perception, besides probably many longer waves, outside of that range. In this particular sense we are constantly receiving wireless telegraph waves from every visible orb, and the message received is not news but light. Moreover, since all animal energy is derived from plants, and all plants build up their substance from the energy contained in the sunlight they receive, it follows that all our muscular energy is derived indirectly from wireless telegraph waves received from the sun. Union of Optics with Electromagnetics All of the phenomena of light, reflection, re- fraction, polarization, interference, etc., which have been within the special study of Optics for many decades, have in recent years been imitated, on a relatively large scale, by electro- magnetic waves set up by the discharge of elec- 74 WIRELESS TELEGRAPHY trifled conductors. In fact, a few of the proper- ties of optical waves which are difficult to detect, by reason of the excessively short optical wave- length, are more easily studied and revealed in electromagnetic waves in the electrical labora- tory. Classification 0} Types oj Electromagnetic Waves Summing up the conclusions reached in the last few chapters, we may say that discharges between two rods or conductors set up spherical waves. Discharges between a conductor and a plane conducting surface, such as the ground approximates, set up hemispherical waves. Waves guided between a pair of parallel wires; and between an aerial wire and the ground are cylindrical waves, moving end- wise. The waves employed in ordinary wireless telegraphy are initially hemispherical waves conforming to, or guided by, the spheroidal earth. CHAPTER VIII PLANE ELECTROMAGNETIC WAVES Hemispherical Waves oj Large Radius Are Virtually Plane at Any One Point ANY small section or piece cut from the front of a hemispherical wave is practically flat, or plane, when the wave is remote from its origin, just as the earth's spherical surface is practically flat, or plane, at any one point on the ocean, be- cause its radius is relatively so large. Conse- quently, any hemispherical wave advancing over the surface of the earth or sea may be regarded as plane locally. It comes along like an invisible upright wall. A section of a single such wave is shown in Fig. 21, taken along the line of march V V. G G represents the surface of the ground. The elec- tric flux rises perpendicularly at P P, or very nearly so. If the earth conducted perfectly, the electric flux would rise strictly vertical from it. Imperfect conductivity causes a wave to lean over, or bend forward slightly, as it moves, so that a perpendicular to the wave front would no 75 76 WIRELESS TELEGRAPHY longer lie parallel to the ground but would point into it. For practical purposes, however, we may take the electric flux as perpendicular. In the front half of the wave, we have taken the flux P P as pointing upwards, corresponding to a moving positive charge on the ground be- neath ; while in the rear half these conditions are reversed. This relation of directions depends upon the direction of the fluxes in the oscillator at the moment that this particular wave was born. The amplitude of the current waves on the ground are indicated by the curved line f p o n r. The line of zero current is the line z f o r z. The direction of the magnetic fluxes is also indicated, by circles where the flux is directed away from the observer and by short horizontal bars where towards the observer. The wave front has reached F F, while the rear of the wave is at R R. The wave-length is, therefore, the horizontal distance F R. At the central vertical plane O O, midway between the positive and negative developments of the wave, the fluxes are zero and their energies are con- sequently zero. The fluxes are densest at the central planes P P and N N, and their energies in a given volume are also a maximum at these planes. PLANE ELECTROMAGNETIC WAVES 77 In practice, there will usually be a train of successive waves moving over the ground in place of the solitary wave of Fig. 21. A wave FIG. 21. Section of a Single Electromagnetic Wave Along Line of Advance and Near to Surface of the Ground. train in wireless telegraphy does not usually con- tain many waves, and their amplitude succes- sively diminishes; so that the final waves in the train are extremely feeble. Analysis oj a Single Wireless Telegraph Wave At and Near the Earth A section of the wave in the plane of P P, Fig. 2i, is given diagrammatically in Fig. 22. It ap- pears as a number of parallel equidistant, ver- 78 WIRELESS TELEGRAPHY tically rising, electric flux lines, crossed at right angles by a number of parallel equidistant hori- zontal magnetic flux lines. This means that both the electric and the magnetic fluxes have uniform intensity in this plane. The charge moving upon r f- *f *f- 4- f- *f 4. 4. 4. 4. 4. 4. 4, 4. 4. 4. 4. 4. 4 V 4. 4. 4. 4 V 4. 4.4.4.4.4.4.4.4.4. 4-4-4-4- -L FIG. 22. Diagrammatic Section of Plane Vertical Electromagnetic Wave Parallel to Wave-Front and Advancing Towards Observer, with Electric Flux Rising Vertically from Positive Charge on Ground and Magnetic Flux Horizontal. the surface of the ground below the wave is posi- tive. Since the ground is not a perfect conduc- tor, the fluxes penetrate into it to some extent. This causes a certain amount of energy to be expended in the penetrated layer of soil as heat, derived from eddy currents, of parasitic electric currents, in the soil. The energy expended at PLANE ELECTROMAGNETIC WAVES 79 the foot of the wave has to be paid for from the stock of energy residing in the wave as a whole; so that energy is fed downwards as the wave runs along, causing a weakening of the moving fluxes, in addition to the weakening caused by the simple hemispherical expansion of the wave. Electric and Magnetic Forces Embodied in the Wave and Moving Therewith If we could compel the wave to stand still for a moment, instead of running by the observer at light-speed, we should expect to find that a positively electrified pith ball would be urged upwards by the upwardly pointing electric flux of the wave as depicted in Fig. 22; while a deli- cately poised magnetic compass needle would tend to align itself along the lines M m in the wave front. In any ordinary wave, however, these electric and magnetic forces would be of very feeble magnitude. The fact that they are able to produce recognizable effects as they pass by is due to their enormous speed, the speed of light in the medium. Transparency of Electric Non-conducting Obsta- cles to Long-wave Light When a plane electromagnetic wave, many meters or yards long, running along the surface 80 WIRELEvSS TELEGRAPHY of the ground, strikes a brick wall, or a wooden- frame house devoid of metal, it passes through these obstacles with very little disturbance. This means that if our eyes were capable of responding to these waves, so that they produced the sensation of some type of color, such non-conducting obsta- cles would be transparent to that color of light, and we could look through a brick house or a wooden house without difficulty, when objects were illuminated by these waves, or in this imaginary type of color. If, however, the ad- vancing waves strike an electrically conducting obstacle, such, for example, as a simple vertical metallic rod, indicated in Fig. 23, the obstacle will either absorb or reflect the waves and will cast a shadow beyond it, the shadow being, of course, invisible to us; since the waves are invis- ible; but the shadow can be determined and mapped out by suitable electric apparatus; or by what may be called an artificial eye. Shadow Cast by a Vertical Electric Conductor in the Path of an Electromagnetic Beam The electric flux only is indicated in Fig. 23 advancing from left to right, over the ground G G. At fl, it is about to strike the vertical metallic rod A B, connected with the ground. Such a vertical might be a leaden water-pipe, or PLANE ELECTROMAGNETIC WAVES 81 a copper wire. At b the wave has passed the vertical and a gap has been thereby torn in the wave. The lower edge of the wave at the rent is bent backwards and the subsequent direction of movement of this edge, being always perpendicu- lar to the local surface, is downwards as well as 1 t I 1 f 1 1 1 J 1 1 4 * 1 i * * * I 1 * ' * 1 r 4, 4 4 l | 4 1 1 i i i 1 4 1 i! I I If 4 1 | 4 4 1 t 4 i i i t t t i 1 t 4 t I ill i i i FIG. 23. Diagram Indicating the Electromagnetic Shadow Cast by a Vertical Conductor in the Path of an Advancing Plane Wave. to the right. In the successive positions c d e /, the wave is spreading down from above, and at m the rent in the wave net has been repaired, or the shadow behind A B has been rilled up, the energy of the flux put into the patch being drawn from the remainder of the wave. It is to be understood, however, that the diagram of Fig. 23 makes no pretensions to geometrical accuracy, and the exact contour of the shadow thrown by a conducting obstacle is not yet determinable with precision. 82 WIRELESS TELEGRAPHY After the wave has struck the rod A B, a dis- turbance is reflected back from the rod into the region A B n G, at the same time that the shadow * > rr * * * * t * * lil t f FIG. 24. Diagrammatic Sections in Elevation and Plan of Electromagnetic Wave Striking a Vertical Con- ductor while Advancing Towards Observer. is cast beyond. This reflected disturbance is, however, omitted from the illustration. Fig. 24 presents a sketch, both in elevation and in plan, of the actions occurring when the wave strikes the vertical conductor. The wave is sup- posed to be advancing towards the observer. It will be seen that the electric flux, which is every- PLANE ELECTROMAGNETIC WAVES 83 where distributed as in Fig. 22 (reversed), before it strikes the vertical A B, is drawn in on each side to the rod and converges on the same, con- tinuing to run down the rod for a little while after the wave has passed. The magnetic flux is shown in the plan at the base of the illustration. Before the wave reaches the rod, as at m m, the magnetic flux lies in a horizontal straight line, parallel to the wave front. As soon as the wave strikes the rod, the magnetic flux bends around it clockwise, and also descends the rod at light- speed. S S is the shadow cast by the rod B, or the space denuded of magnetic flux for a certain distance behind B. Looking at the action from another standpoint, we may, in the light of our provisional electro- magnetic theory, consider that the electric flux advancing over the ground brings a local right- handed torsional stress upon the ether, which, by electric rigidity, resists the stress and limits the flow of ether to that small amount found in the wave front as indicated by the horizontal magnetic flux lines M M, Fig. 24. At soon as the electric flux strikes the conducting rod A B, the elastic rigidity of the ether is lost, owing to the action of the conductor and the electrons residing in it. The ether at the rod gives way before the stress, and flows bodily around the rod 84 WIRELESS TELEGRAPHY in dense magnetic flux streams. On such a hypothesis, a conductor behaves like a gap in the ether, and the advancing electromagnetic wave pours electric and magnetic fluxes spirally or vertically down into the gap as it goes by. Comparisons Between Reflection of Short and Long Waves of Light Whatever hypothesis we adopt to assist the mind's eye in depicting the process, we must expect to find the action similar to that which occurs when half-micron electromagnetic waves, i.e., visible light, strike an opaque obstacle. There is a reflected wave train thrown back by the obstacle. There is also a shadow cast be- hind it, and there is energy absorbed into the substance of the obstacle. The width of the shadow cast by a parallel beam of light is appar- ently no wider than the obstacle; whereas in Fig. 24, the shadow cast is indicated as being many times the width of the vertical rod. But it has to be remembered that if the optical shadow of a rod were one quarter of a wave- length wider than the rod, the difference would be only about a sixth of a micron ( Tg ^ o0ft ) and quite insignificant; whereas if the rod A B had the same height as the simple rod oscillator which originally emitted the wave, a shadow hav- PLANE ELECTROMAGNETIC WAVES 85 ing a breadth of a quarter wave-length would be as broad as the height A B, or the distance N N in Fig. 24, and would be equal to the height A B, a very appreciable distance. Gashes Torn in Electromagnetic Waves by Verti- cal Conductors on the Earth It is evident that upright metallic rods, such as lightning-conductors, tear rents in any passing electromagnetic wave running along the ground. On the other hand, a conductor parallel to the ground, such as a trolley-wire, or an overhead telegraph wire, does not sensibly affect a passing electromagnetic wave. Looked at in another way, a vertical rod is cut by the rushing hori- zontal magnetic flux at light-speed, and acts like a single-wire dynamo, moving through a very feeble magnetic field at the speed of light. Again, a vertical rod picks up a certain difference of electric potential between the electric flux at its top and at its base. In either of these ways, the rod becomes the seat of an electric impulse or electromotive-] or ce during the brief interval in which the wave is passing by it. But if we place the rod horizontal, instead of vertical, the electric flux in the wave will cut the rod perpendicularly and the magnetic flux, in cutting, only acts upon the thickness of the rod; so that the electromo- 86 WIRELESS TELEGRAPHY tive force set up therein by the passing wave will be insignificantly small, and will be directed transversely or across the diameter of the hori- zontal rod. Accordingly, when a single electromagnetic wave hits a vertical rod, a rent is torn in the wave, and the breadth of the rent, although not yet accurately known, may, perhaps, be a quarter of a wave-length. The energy which resided in the wave within the region torn out, is available for setting up electric currents in the rod, after allowing for what is lost by reflection and second- ary radiation. It may be readily imagined that when an electromagnetic wave strikes a steel bridge, or a steel sky-scraper office-building, it casts a long shadow, and a relatively large quantity of energy is torn out of the wave. Trees also, and shrubs too, in lesser degree, have been found to be feebly conducting, and it is believed that they absorb energy from waves passing them. This fact taken in connection with the imperfect conduc- tivity of dry soil, in comparison with sea water, accounts for the considerably greater distance at which electromagnetic waves can be transmitted and detected over the ocean than over land. The signaling distance range over the sea is, roughly, double the signaling distance range across country. PLANE ELECTROMAGNETIC WAVES 8; Elementary Analysis of Electric Oscillations Set Up in a Vertical Conductor by the Passage of Waves It is important to notice the principal events that occur in the neighborhood of the vertical rod after it has been struck by the onrushing electric wave. Fig. 25 indicates diagrammati- cally nine successive stages in half a complete cycle of these events. The line of crossed ar- rows immediately under the letters ABC . . . . I, represents the directions of electric and magnetic flux in the advancing wave over the rod. Thus at A, the conditions are those indicated in Figs. 23 and 24; namely, the electric flux is pointing downwards and the magnetic flux pointing to the right, as viewed by an ob- server who sees the wave advancing towards him. At E and F these fluxes are in the act of reversing through zero, corresponding to a plane such as O O in Fig. 21. At I, the fluxes have completely reversed. Underneath each diagram of a rod section in Fig. 25, there appears a plan view showing the direction of magnetic flux in the wave just before striking the rod, and also of magnetic flux en- circling the rod. Thus, at A, the magnetic flux in the air is at full development towards the right 88 WIRELESS TELEGRAPHY hand of the observer, while around the rod it is clockwise. At C, the clockwise magnetic flux encircling the rod has reached full development, or the electric current over it is a maximum. Between E and F the magnetic flux in the air reverses or passes through zero. At G, the magnetic flux encircling the rod passes through zero. At I, the magnetic flux in the air- wave has developed completely in its reverse, or left- handed direction. Examining the rod at A, it will be seen that the electric flux of the passing wave has con- verged upon it, as already seen in Figs. 23 and 24. This flux immediately starts to run down the rod to ground, as indicated by the long dotted arrow. The instant it begins to run, the electric flux reverses direction, or assumes the outward direction shown at B,ihe magnetic flux remaining clockwise, as viewed from above. As soon as the flux reaches the conducting ground at the base of the rod, it is reflected thence upward, with a new reversal of electric, and maintenance of magnetic, flux direction. At E, the stream of flux on the rod is about to reach the top. At the top, the flux reverses magnetically, or is reflected downward, with persistence of inward electric flux. At G, the magnetic flux is half clockwise and half counter-clockwise, representing zero of 8 9 QO WIRELESS TELEGRAPHY current, but maximum electric potential. At 1, the flux is in full descent again, with counter- clockwise magnetic field. Resonance in Electric Conductors Struck by Wave-trains We have purposely chosen the length of the rod as one quarter of the length of the plane wave advancing through the air. This brings about such an adjustment of the motion of flux over the rod as enables the next succeeding wave to add to, or increase, the movement. If we ex- tended the diagram of Fig. 25 through eight more such phases we should return to the original con- dition at A, when the flux in the next wave would not only repeat the cycle, but would also increase the amplitude. If the rod conducted perfectly, and also the ground at its base, each wave as it arrived through the air would add to the fluxes running up and down the rod, on the familiar principle of the child's swing, whose oscillations may be increased by timing the pushes to the natural period of oscillation. In this case, how- ever, the swing of the rod is adjusted by its length, so as to be in rhythm to the train of arriving waves. Such a condition of coincidence between the times of arrival of the successive wave- crests, PLANE ELECTROMAGNETIC WAVES 91 and the natural time of electric oscillation of the rod, is called electric resonance. If we could obtain a very long train of uniform advancing waves and adjust the length of the vertical rod into resonance therewith, retaining perfect conduction, the fluxes running up and down would increase indefinitely, were it not for secondary radiation. That is, the rod, excited in this way by arriving waves, would become an oscillator in its turn, and discharge the energy it received in a new series of radiating waves, as in Fig. 15. In practice, however, the waves re- ceived through the air have such feeble amplitude, they decay so soon, the number in a train is so small, and the conductivity of the rod and ground base is so far from being perfect, that even when the rod length is adjusted into resonance, the currents developed over the rod, as in Fig. 25, are comparatively feeble. The secondary radiation is, therefore, insignificant. If the length of the rod is in the opposite con- dition to that required for resonance, the fluxes generated thereon by the first wave will be op- posed, instead of aided, by the fluxes generated in the second, and so on. Consequently, there will be comparatively feeble currents set up on the rod. If, however, the length of the rod is adjusted for resonance, there will be a building 92 WIRELESS TELEGRAPHY up of electric current on the rod, unless the arriv- ing wave train is too short, or unless the electric obstruction and want of conductivity in rod and ground suppress the development. In order to adjust the rod to the resonant con- dition, it is not always necessary to alter its height. The virtual length can be altered by the insertion of a suitable form of conductor or wire at the base, between rod and ground, in a manner to be described later. In such a manner the time of oscillation of a rod can be altered without chang- ing the actual height. Resume o] Conditions Attending the Impact of Waves Against Vertical Conductors Summing up the above results, we find that a vertical conductor connected with good conduct- ing ground, and set up anywhere in the path of a train of electromagnetic waves, will have alternat- ing, or to-and-fro electric currents set up on it, the energy contained in these currents being the energy in the up-and-down moving fluxes, which energy is drawn from, or scooped out of, the arriving electromagnetic waves as they pass by. These alternating currents in the rod are capable of being built up, or successively increased in strength, by the impulses of the successive waves, if there be resonance, i.e., if the natural time of PLANE ELECTROMAGNETIC WAVES 93 oscillation of the rod be the same as the periodic time of the arriving waves. For a simple vertical rod, devoid of inserted apparatus, this will be when its' height is one quarter of the wave-length, and therefore equal to the height of the simple vertical rod oscillator which is capable of originat- ing such waves. In other words, if the arriving waves have been produced by a distant simple rod oscillator, resonance will require the heights of the oscillator and of the receiver verticals to be equal. Resonance would be capable, theo- retically, of setting up an indefinitely great ampli- tude in a perfectly conducting receiver rod, set in perfectly conducting ground, with a constantly maintained succession of waves at the oscillator, were it not for secondary radiation of waves from the receiver. In practice, however, the alternat- ing currents set up at the receiver may be ma- terially increased by bringing the receiving rod into resonance, but the development is arrested by imperfect conduction at the receiver, as well as by discontinuity of the oscillations, or small trains of waves at the oscillator. Moreover, the insertion of a receiving instrument into the path of the vertical receiver rod also causes energy to be absorbed, and interferes with the production of resonant increase of oscillations. 94 WIRELESS TELEGRAPHY Energy oj Electric Oscillations, or Oscillating Currents, Set Up in a Vertical Receiver The energy which is available for producing such electric currents by any wave at the receiver depends upon the energy in the entire hemispheri- cal wave at that moment. It will evidently be but a very small fraction of the total energy of the wave, when the receiver is far from the oscillator, since the area of the wave which can come into contact with the receiving rod, or into its region of influence, is so small. If we suppose that the receiver has a height of a quarter wave-length, for resonance, and that the effective breadth of area from which energy is drawn, as in Fig. 23, is also a quarter wave-length, then the fractional part of the wave's energy available for producing electric current at the receiver, is the square of the height of the receiver rod, divided by the superficial area of the hemisphere occupied by the entire wave at the instant it strikes the rod. For example, if we suppose that a certain wave in a series emitted by an oscillator contains at the moment of shipment an amount of energy equal to i kilogramme-meter (7.24 foot-pounds, or the work done in lifting one pound through a vertical height of 7.24 feet), then a quarter- wave vertical receiving rod at a distance of 30 kilometers PLANE ELECTROMAGNETIC WAVES 95 (18.6 miles), with a height of say 31.6 meters (103.6 feet) might perhaps absorb energy from the wave as it passed, over a height of 31.6 meters and a breadth of 31.6 meters, or an area of wave surface amounting to 1,000 square meters (10,760 square feet). But, neglecting the curva- ture of the earth, the area of a hemisphere 30 kilometers (18.6 miles) in radius would be 5,655 millions of square meters (60,800 millions of square feet); so that the electromagnetic energy capable of being drawn on to the rod would be CTTiTTO* a kilogramme- meter, or 1 7 ergs. The received energy should be about 60 per cent, greater than this, because of the greater density of flux and energy in the equa- torial zone of the transmitted wave, i.e., near the earth's surface. On the other hand, how- ever, a distinct reduction would have to be made for the energy wasted in transmission along the surface of the soil, by reason of the earth's im- perfect conductivity, or for other vertical con- ductors, such as metallic structures, or trees, intervening between oscillator and receiver. Our knowledge is still very imperfect as to the effective surface area drawn upon by a vertical rod, and also as to the amount of energy drained from the feet of an advancing hemispherical wave by reason of the earth's imperfect conductivity. 96 WIRELESS TELEGRAPHY This loss is known to be greater after dry weather than after rain. It is, however, clear that accord- ing to the working theory outlined above, the energy capable of being received by any such ver- vertical rod increases as the square of its height, assuming that the resonant condition is main- tained, and also inversely as the square of the distance between the sending and receiving sta- tions. The total energy available for producing alternating electric currents at the receiver will be the sum of the successive fractional amounts drawn from each single wave in turn, assuming that the successive effects can be prevented from canceling or annulling each other, by the adjust- ment of the receiver to the resonant condition. CHAPTER IX THE SIMPLE ANTENNA OR VERTICAL ROD OSCILLATOR The Antenna and Transmitting Apparatus WE have already arrived at the conclusion from preceding chapters that wireless telegraphy ordinarily employs hemispherical electromagnetic waves emitted from a vertical rod oscillator or antenna, in short trains of from two to thirty waves of successively diminishing amplitude. On or near the ground, or sea level, and at a great distance from the transmitting station, these waves are for all practical purposes plane waves, advancing over the conducting ground, or sea, with the speed of light, in a direction radial to the sending station, after allowing for the curvature of the earth. We now proceed to consider the essential elements of the antenna, and of the apparatus employed, at the trans- mitting station. The simplest type of vertical antenna or rod oscillator is represented in Fig. 26. It consists essentially of a single vertical metallic wire A B, 97 WIRELESS TELEGRAPHY suspended from an insulator I, which is sup- ported from a wooden mast structure indicated in dotted lines. This vertical wire, air- wire, aerial, or antenna, is insulated from the ground FlG. 26. Essential Elements of a Simple Vertical Os- cillator, or Antenna, for Emitting Hemispherical Electromagnetic Waves. by the air-gap G, so long as it is electrically in- active. The lower terminal of the air-gap com- municates with a metallic plate P sunk in moist earth, or below low-tide level, if on the seashore. Sometimes bare wires w w are laid out radially from the ground wire in various directions, at or THE SIMPLE ANTENNA 99 near the surface of the ground, so as to improve the local conductivity of the soil, and help to form a good electric mirror at the ground surface s s s, from which the waves may be reflected back and up the antenna, not only at the conductor, but in its vicinity, as indicated in Fig. 15. The length of the wave emitted by a simple vertical wire antenna as shown in Fig. 26 is be- lieved to be very closely four times the height of the antenna A B G S. Thus, if the antenna had a height of 30 meters (32.8 yards), above perfect ground, the length of the waves sent out would be 120 meters (131.2 yards). The number of such waves which would cover the distance travelled by light in one second would be AHJfcfJJJUUL= 2,500,000; so that there would be two and a half millions of such waves occupying one second, if the oscillator could be kept at work for that time. This means that the frequency of the waves would be 2,500,000 cycles per second, or the time occupied by any one complete wave to pass a given point would be 3,5 o o;o o o tn second. If we call the one millionth part of a second, one microsecond for convenience of description, then one complete wave would pass off in -$=% micro- second. Since each wave contains both a posi- tive and a negative impulse, either impulse would pass by in \ of a microsecond. 1C*, WIRELESS TELEGRAPHY The Large Activity of an Antenna Owing to this extremely short period of oscil- lation, antennas are remarkable for their activity or power. The amount of energy which can be stowed away in a simple vertical antenna as electric- flux energy in the surrounding ether, by charging it to a suitably high voltage, is compara- tively small, being usually not more than 20 gramme-meters (0.14 foot-pound), or the work done in lifting 20 grammes to a height of one meter. When this energy is released, by the discharge of the antenna across the spark-gap G, Fig. 26, part of this energy is expended in the heat of the spark and in heating the surface of the conducting antenna. The remainder is available for radiation as a series of electro- magnetic waves. Perhaps not more than 3 gramme-meters (0.022 foot-pound) of energy will be shipped off in any single wave. Nevertheless, this energy is shot off by this particular antenna in -^5- part of a microsecond, and the average rate of power radiation during this brief interval will thus be 7,500,000 gram-meters per second, or 7,500 kilogramme-meters per second, or about 100 horse-power. With the aid of auxiliary apparatus, an antenna may be capable of radiating electromagnetic wave THE SIMPLE NTENA' J JjJ rot energy at the rate of hundreds - of but only for a few microseconds at a time, so that its average power in one second, or in one minute, during its operation, may be only a small fraction of a horse-power. An antenna of the simple type shown in Fig. 26, looks like a very simple and innocent machine; but, when thrown into electric vibrations, it may throw out as much power as it takes to operate a high- speed electric locomotive; only it does not keep the power up. The case is somewhat similar to that of a revolver, which is being fired, say, three times per second. At each explosion the power of the machine is relatively very great; but between shots the power falls to nil; so that the average for one second, the power of the machine, or its mean rate of throwing energy off, is comparatively low. In practice, single-wire antennas are seldom used, and multiple- wire antennas are customary. The purpose of employing a plurality of conduc- tors is two-fold. In the first place, the larger surface of the antenna gives more electric flux in the air when charged, and this increases the stock of energy held by the antenna prior to re- lease and radiation. In the second place, the larger surface permits of a more free active radia- tion or discharge of electromagnetic waves into WIRELESS TELEGRAPHY ir, Independently of the amount of energy to be released. Cylindrical Antennas Fig. 27 represents the cylindrical type of verti- cal antenna, one having four parallel vertical t wires and the other sev- enteen. Any convenient number may be used. The wires are usually soldered to two or more metallic hoops H H, which not only strength- en the structure mechan- ically, but also keep the oscillations symmetrical electrically. The diam- eter of these hoops may range from 30 centime- ters to several meters (i foot to several yards). If the component vertical wires are not further than, say, half a meter apart (19.7 inches) these bird-cage cylinders are almost equivalent, electrically, to complete sheet- cylinders of metal. The bird-cage cylinder of multiple parallel wires is of course far superior FIG. 27. Types of Cylin- drical Frame Vertical Oscillators. THE SIMPLE ANTENNA 103 mechanically to a complete sheet-cylinder or large pipe, both in cost, lightness and freedom from wind-pressures. These cylindrical metallic frames may be supported by suitable insulators from a mast-arm at I, their lower extremities G leading to spark gaps. In some instances a cylindrical antenna is formed of a rigid vertical steel tube, bolted to- gether in sections and supported on insulators at the base. The tube or cylinder is prevented from falling by guy- ropes running in various directions, and in which insulators called strain- insulators, because they are subjected to tension, are inserted at some suitable point or points. Harp, Fan and Inverted Cone Antennas Other forms of antenna in use are outlined in Figs. 28 and 29. The former indicates the harp type. This is conveniently supported between two wooden masts, as indicated in dotted lines, but a single mast may serve if the wires are sus- pended from a horizontal arm. The five wires shown are connected at the top and at the bot- tom by horizontal wires. The entire conducting ' frame is supported by insulators at 1 1 1 1. The harp is connected to the spark-gap by the wire G, The fan-shaped antenna of Fig. 29 is some- times used on board ship. The stout wire I I is 104 WIRELESS TELEGRAPHY strung between the masts B A and D C, being supported by end-insulators I I. The descend- ib FIG. 28. FIG. 29. FIGS. 28 and 29. Types of Harp-Shaped and Fan- Shaped Antennas. ing wires are each connected to the top-wire I I above and to the central point below; whence a wire G runs to the spark- gap and ground, or on a steamer to the metallic frame of the hull. At some stations a ser- ies of fans are connected together into an inverted cone, as seen in Fig. 30. FIG. 30. Typeof Inverted Here four masts support Cone Antenna. . a metallic rectangle, through insulators not shown in the diagram. Metallic wires drop from the rectangle at inter- THE SIMPLE ANTENNA 105 vals to the central point O, whence a wire runs across a spark-gap to ground. Whatever the form of the antenna, cylindrical, harp-shaped, fan-shaped or conical, the object sought, already mentioned, is to increase the electric flux, and electric energy associated there- with, in the charge of the antenna, and also to facilitate the emission of the waves into space at the recoils from the upper end or ends of the antenna. Electric Oscillations on Antennas Skin Deep The thickness of the individual wires forming the antenna is of secondary importance It is the surface of the wires which is of principal con- sideration. The high-frequency electric cur- rents, or oscillations, running up and down the antenna, are not able to penetrate below a cer- tain skin depth into the conductor, say i mm. The higher the frequency, the less the penetra- tion, and the thinner the effective conducting skin. The wires are usually of copper, and about 4 mm (J inch) in diameter. Other things being equal, the higher the antenna, of whatever form, the more electric flux, charge, and energy it will hold; so that the power it can release is greater. At the same time the length of the wave tends to be greater. 106 WIRELESS TELEGRAPHY Sources o) Energy jor Feeding to an Antenna The source of electric energy for charging the antenna is generally an induction coil, or spark coil, excited either by a dynamo, or by a voltaic battery. If a voltaic battery is used, it is com- monly a secondary ^ or storage battery ? charged by, and receiving energy from, a dynamo. Con- sequently, while it might be possible to use any electric source of energy, such for example as a frictional machine; yet, in practice, the energy is furnished by a dynamo driven by water-power, steam-power, or gas. An ideal form of dynamo exciter would be an alternating- cur rent dynamo which generated to-and-fro electric currents, or currents of successively reversing directions, with a frequency precisely that required for setting up resonance in the antenna. If such a very high- frequency dynamo could be constructed conveni- ently, it would be capable of keeping the antenna in full oscillation indefinitely. That is, if the radiating power of the antenna were say 300 kilo- watts or 400 horse-power, it would be possible to connect a dynamo of at least 300 kilowatts capa- city (400 H P) to the antenna, and keep it con- stantly in action at that rate. Such a dynamo would have, however, to generate alternating currents with a frequency either of millions, or, THE SIMPLE ANTENNA 107 at least, many thousands of cycles per second; whereas the dynamos used in electric lighting and power transmission ordinarily only generate alternating currents with a frequency of sixty FIG. 31. Induction Coil for Generating a High Voltage. (60) cycles per second. This frequency is at least hundreds of times too low for direct exci- tation. Under present conditions it is customary to charge the antenna by an induction coil of some kindo When the energy is supplied by a storage battery, an induction coil is used resembling that shown in Fig. 31. This apparatus, which is essentially a powerful spark-coil, has a central 108 WIRELESS TELEGRAPHY core ot iron, in the form of a bundle of iron wires. There are two coils, or windings, of insulated wire placed on the iron core. These two wind- ings are carefully insulated from the core, and from each other. One is the primary winding, consisting of comparatively few turns of coarse cotton-covered copper wire. The other is the secondary winding of very many turns of fine silk- covered wire. The primary wires are led out at p p and the ends of the long fine secondary winding are connected to the discharge knobs s s. When a strong current is flowing steadily through the primary winding, supplied by an external storage battery, there will be no electric impulse, or electromotive force, in the secondary. There will, however, be a powerful stationary magnetic flux distribution surrounding the pri- mary current, and linked with the secondary coil. If the primary current be now suddenly inter- rupted, the magnetic flux linked with the coils will collapse and disappear. In so doing, how- ever, its movement generates a brief but very powerful electric impulse in the secondary wind- ing, constituting a powerful electromotive force, or a high voltage, i.e., a voltage capable of jumping across a considerable distance of air- space. Other things being equal, the length of the air-gap across which a spark will jump is an THE SIMPLE ANTENNA 109 indication of the magnitude of the electromotive force or voltage producing the spark. Similarity of Process of Transferring Energy in Induction Coil to Wireless Transmission The conditions which accompany the trans- mission of electric power from the primary to the secondary winding, a distance of a few milli- meters or centimeters (a few tenths of an inch up to an inch or two), resemble those which ac- company the transference of electric energy from the sending to the receiving antenna. Whereas, however, in the latter case, the distance between the primary and secondary wires is relatively very great, and the energy is transferred from one place to the other stowed away in a wave or series of waves; in the former case of the induc- tion coil, the wave has no room to develop a separate existence, but the electromagnetic fluxes are linked with both circuits throughout the process. For the same reason, the efficiency of the transmission is enormously greater in the in- duction coil than in the wireless case. Nearly all of the electric energy leaving the primary winding is absorbed by the secondary winding. On the contrary, nearly all of the electric energy leaving the primary antenna goes off into space, or else is ultimately absorbed in the ground, and 110 WIRELESS TELEGRAPHY hardly any is absorbed by the secondary antenna. In the rough calculation given on page 95, Chapter VIII, it appeared, for example, that only i part in 5,655,000 of the energy liberated by the oscil- lator, or sending antenna, was picked up by the receiving antenna, under the conditions there considered. Elements o] Sending Apparatus for Producing Electromagnetic Waves The elements of the connections at a wireless- telegraph sending station are illustrated in Fig. 32. A A is the antenna, or the wire connecting therewith. C is the induction coil. The pri- mary circuit is marked in full lines and the second- ary in broken lines. The primary circuit com- prises the primary winding of the coil C, the voltaic battery B, a hand key K, and an electro- magnetic vibrator or interrupter V. The vibra- tor may be a separate piece of apparatus included in the primary circuit ; or it may form part of the induction-coil mechanism as shown. It is essen- tially a vibrating circuit-maker-and-breaker like the vibrator of the ordinary electric bell. Its purpose is to interrupt the primary circuit auto- matically and rhythmically, as long as the key K is depressed. The vibrator V may give inter- ruptions at the rate of say 200 cycles per second. THE SIMPLE ANTENNA ill It dlso gives a musical note or tone in the sur- rounding air, corresponding to its frequency of vibration. At each interruption of the primary circuit at the vibrator V, there is a sudden elec- tric impulse generated in the secondary circuit, ! ! _ f J j ^ !A < o B )' . takj i j FIG. 32. Elements of Connections at Sending Station. and this travels up the antenna at light- speed. If the spark gap g did not break down, there would be a reflection of the impulse from the top of the antenna, accompanied by an electro- magnetic impulse or radiation into space; but there would be no succession of waves and no 112 WIRELESS TELEGRAPHY proper development of electromagnetic wave emission. If, however, the impulse on its return from the top of the antenna is able to break down the air-gap g in a spark discharge, the electric oscillation continues, and will go on in a succession of sparks, each feebler than its predecessor and each accompanying a half-wave of radiated en- ergy thrown off into space. After a certain number Fio. 33. Diagram of Electric Impulses Delivered to Antenna by Induction Coil. of sparks have passed, depending upon the length of the gap and other conditions, the impulse re- maining is no longer able to follow up by another spark and the train of dwindling oscillations ceases. Fig. 33 shows diagrammatically the succession of electric impulses generated in the secondary coil under the action of the vibrator V, or at the rate which we have assumed to be 200 per second. The electromotive force rises to E with a sudden jump at each interruption of the vibrator, and then changes more slowly to a THE SIMPLE ANTENNA 113 smaller magnitude in the opposite direction. The sudden kicks a E are those which excite os- cillations. Ten of these kicks are indicated in one-twentieth of a second. Each impulse occurs in one two-hundredth of a second, which is F of a second or 5,000 microseconds. Discontinuity oj Electromagnetic Flashes from a Coil-Fed Antenna Fig. 34 indicates the surges or oscillations set up in the antenna at every kick or impulse a E FIG. 34. Diagram of Oscillations Set Up in the Antenna at Each Electric Impulse in Secondary Coil. in Fig. 33, as set up by the vibrator V of Fig. 32. The illustrations show six complete waves A, C, E, G, I, K or twelve successively reversed im- pulses of steadily diminishing amplitude. The l\4 WIRELESS TELEGRAPHY last impulse L is supposed to be too feeble to create another spark at the gap g of Fig. 32, so that the series of oscillations comes to an end at L. Each complete or double oscillation is represented as occupying one microsecond, cor- responding to an emitted wave-length of 300 meters (328 yards) or a height of simple rod oscillator equal to 75 meters (82 yards). It is evident that the whole series of six complete oscillations only lasts for 6 microseconds, and since the kicks, or stimuli, from the induction coil, only occur by assumption at intervals of 5,000 microseconds, there is evidently a long interval of darkness and inactivity between the little flashing intervals in which the antenna is giving out waves of invisible light, or long- wave polarized light. The period of darkness is in this instance 832 times as long as the period of light. If the key K in Fig. 32 were held steadily down, and the vibrator V were thus kept at work, a dis- tant eye assumed capable of seeing this long- wave light would see the antenna shine out, in a certain unknown color, for 6 microseconds in every 5,000, like a flashing lighthouse which sent a beam over the sea for 6 seconds every eighty- three minutes. Although the numerical values here assumed may vary in practice through a considerable range, yet they are fairly represen- THE SIMPLE ANTENNA 115 talive, and a wireless telegraph sending antenna in full activity is many times more intermittent than the longest period flashing light house in the world. The observer with the hypothetical eye capable of perceiving the long electromagnetic waves of wireless telegraphy would always see the flashes in the direction of, or "on the true bearing," as a sailor would say, of the sending antenna, but the flash would appear tangential to the surface of the ocean, or in the true level horizontal plane, as distinguished from the actual visible horizon which is depressed somewhat below; so that the observer would expect to see the luminous image of the antenna thrown up, as though by mirage, to the level of his eye or to the horizontal plane at his level. Long after the antenna ceased to be visible by ordinary short-wave light, which moves in radial straight lines, he would expect to see the flash of the antenna by the bending of the long waves around the conducting curved surface of the sea. Nature and Use of Auxiliary Condenser at Sending Antenna An important piece of apparatus auxiliary to the antenna when set in oscillation is a Ii6 WIRELESS TELEGRAPHY ''condenser." It consists of an expanded pair of opposed conducting surfaces, such as tin foil, sepa- rated by relatively thin sheets or intervals of insu- lating material, such as glass, mica, oiled paper, oil or compressed air. A simple condenser may be formed of a sheet of window glass, coated on each side with tin foil, except near the edges. The thinner the glass and the larger its surface area, the greater the electric charge it will hold, or the more electric flux and electric flux energy it will stow away in the glass, for a given magni- tude of charging voltage, or in technical lan- guage, the greater becomes its capacity. Another well-known form of condenser is a glass bottle, coated on the inside, as well as on the outside, with tin foil. It is the glass walls of this bottle or Leyden jar, which receive the electric flux and flux energy when the jar is charged. The greater the surface of the jar and the thinner its wall, the greater will be its capacity. Looking at an antenna as a condenser, or Leyden jar, the surface area of the conductor or conductors composing it may be considerable; but the slab of insulating air between the antenna and the ground is on the average many meters thick. Consequently, the capacity of an antenna is relatively small. An antenna 50 meters (54.7 yards) high, even if made up of numerous wires, THE SIMPLE ANTENNA may have no more capacity than a single Leyden jar of ordinary size. A diagrammatic view of the electric flux stowed away in the dielectric of a charged condenser is shown in Fig. 35, where the upper conducting plate is represented as being charged positively. The flux density increas- es with the thinness of the insulating slab and also with the charging voltage. A limit to the thinness of the insulator is set, however, by the electric strength of the material, which ruptures, or breaks down in spark discharge, if a certain electric intensity is exceeded. Air at an ordinary pressu/e and temperature has a strength (between parallel planes) of about 4 kilovolts (4,000 volts) per millimeter (101,600 volts per inch), glass 8 kilo- volts per millimeter, mica 25 kilovolts per milli- meter, and so on for other substances. The electric strengths are affected by the purity of the material, its temperature and other condi- tions. FIG. 35. Diagrammatic Section of a Charged Condenser Formed by Two Parallel Plates, Showing the Distribu- tion of Electric Flux in the Insulator Between Them. WIRELESS TELEGRAPHY Adjustment of Auxiliary Condenser Circuit to Consonance with Antenna A condenser is often connected in parallel with the antenna at the sending station in the manner CO FIG. 36. Condenser Connected in Parallel with Antenna to Re-enforce Oscillations. indicated in Fig. 36. By this means the capac- ity of the insulated system may be much increased so that it will receive a much greater electric charge from the induction coil, with correspond- ingly increased electric flux and electric flux THE SIMPLE ANTENNA 1 19 energy. When the system of Fig. 36 is dis- charged at the spark-gap, the energy released in radiation may be considerably increased, owing to the presence of the condenser and its electric flux contents. On the other hand, however, the auxiliary circuit containing the condenser should be adjusted to the length of the antenna, in such a manner that the two shall oscillate together, or in synchronism. In other words, the time of oscillation of the condenser circuit A C B G should be adjusted to the time of oscillation of the antenna; otherwise, the oscillations of the two will mutually interfere, and cancel each other; so that if the condenser circuit is not tuned into synchronism with the antenna, the radiation into space may be weakened, instead of being enhanced, by the presence of the con- denser, in spite of the greater stock of energy available. The tuning of the condenser circuit may be accomplished not only by altering the size of the condenser, but also by altering the length and disposition of the wire connecting the condenser to the induction coil. If this wire be arranged in a coil A or B, Fig. 36, of several turns, the effect of the turns is to increase the virtual length of the wire in rapid proportion, because the magnetic flux generated around any one turn 120 WIRELESS TELEGRAPHY links also, more or less, with the other turns. By suitably adjusting the number of turns of wire in the condenser circuit, the free period of oscillation discharge of the condenser can be ad- justed into synchronism with that of the antenna; so that the latter can thereby be thrown into re- enforced oscillation and wave emission. Loaded Antennas It is also possible to alter the virtual length of the antenna, by connecting a coil of a few turns of wire in its circuit as indicated in Fig. 37. Such an antenna is called a loaded antenna to distin- guish it from the simple or unloaded antenna of Fig. 26. An antenna loaded by a simple coil as in Fig. 37 always behaves as though its length were increased. That is its wave-length is in- creased, or its frequency of oscillation is reduced. Whereas, therefore, a simple antenna, say 25 meters (27.3 yards) in height, would throw off waves 100 meters (109.4 yards) long, or with a frequency of 3,000,000 cycles per second, or with a period of -J of a microsecond; the same antenna loaded with a coil might readily increase its wave- length to a kilometer or more (1094 yards) with a frequency of 300,000 cycles per second. At the same time, however, the radiating power of THE SIMPLE ANTENNA 121 the antenna, or the energy it can throw off in a single wave is likely to be greatly reduced by loading. It is thus possible to adjust the antenna and the condenser into synchronism by altering the CJ p IG> 37. Coil Inserted in Series with Antenna to Increase Its Virtual Length. number of turns of wire in circuit with either. It is also possible to lengthen the wave emitted by an antenna, within certain limits, by loading it with an appropriate coil, releasing the energy in a longer train of feebler waves instead of a very short train of stronger waves. 122 WIRELESS TELEGRAPHY The length of the wave commonly employed in ordinary wireless telegraphy varies from say 100 meters to 10 kilometers (109 yards to 6.2 miles) corresponding to frequencies between 3,000,000 and 30,000. A common wave-length would be, say, 300 meters (328 yards) with a fre- quency of i,ooOyOoo cycles per second. It may be observed that the presence of the induction coil I in Figs. 32, 36 or 37 does not appreciably affect the virtual length of the antenna, because it is in parallel to the spark-gap G and not in series therewith, like the coil L of Fig. 37. If the induction coil I were throvn in series, thereby adding its virtual length to the antenna, the frequency would be insignificantly low. On account of the very large number of turns in the secondary coil I, or, as it is technically expressed, on account of its large s el] -induction, the oscillations set up on the antenna pass across the gap G and are unable to find their way through the wire of the coil. It is also possible to insert a condenser into the .path of the vertical antenna with the tendency of increasing the frequency of oscillation, or of diminishing the virtual height of the antenna. The actions of a condenser and a coil are in this respect opposite to each other* THE SIMPLE ANTENNA 123 Danger o] Secondary Internal Reflections Oc- curring In a Loaded Antenna, When, however, any sudden obstacle or appa- ratus, such as a ceil, a condenser, a resistance, or a discontinuity of any kind is inserted in the path of an antenna, there is a tendency to set up reflec- tions of the oscillations at the discontinuity. These reflections may break up the rhythm and diminish the amplitude of oscillation. Conse- quently, care has to be taken so to introduce dis- continuities into an antenna as to minimize the detrimental effect of. internal partial reflection; or the benefit gained by the insertion of the dis- continuity, as in adjusting the frequency to resonance, may be more than offset by the shattering of main oscillations into minor and discordant ripple trains. CHAPTER X ELECTROMAGNETIC WAVE-DETECTORS, OR WIRE- LESS TELEGRAPH RECEIVERS Voltage Detectors and Current Detectors SINCE the human eye is incapable of respond- ing to the long-wave flashes, or electromagnetic waves, given off by a wireless telegraph antenna, an artificially constructed eye has to be used in order to detect and respond to them. The plan followed is to place an antenna at any suitable place in the path of the waves, so that oscillating electric impulses may be set up in this antenna, and then to permit these impulses to act upon some electromagnetic apparatus connected either directly in the path of the antenna, or indirectly, by the aid of a little induction coil. The receiv- ing instrument must therefore be affected either by the oscillatory voltage, or by the oscillatory current in the antenna. A voltage detector may be theoretically any apparatus which responds to electric potential difference; such as an electroscope or a pair of 124 ELECTROMAGNETIC WAVE-DETECTORS 125 diverging gold-leaves. In practice, however, it consists of a little instrument called a coherer. A current detector may be of any of the various types which are used to indicate the presence of high-frequency alternating currents. There are a number of different receivers of which we need only consider the prominent types. In practice, there are three well-known types: namely, the thermal, the electrolytic and the electromagnetic. Coherers Coherers are illustrated typically in Figs. 38, 39 and 40. In Fig. 38, we have a sealed glass tube T T about 4 cms. (i| inches) long and of 2j mm. (-fa inch) bore. Near the middle of the tube are two metallic plugs, P P, often made of silver. These are connected to the external wires WW by sealed-in platinum connections. The plugs P P do not touch, but are separated by a small gap about J mm. (-^-th inch) wide. The tube may be partially exhausted of air; but the gap between the plugs P P contains fine metallic powder, or metallic dust, which lies loosely in the little crevasse. The loose metallic particles bridging across between the plugs P P have the property of offer- ing an obstruction, or very high resistance, to the flow of current from a single voltaic cell. In 126 WIRELESS TELEGRAPHY other words, the gap is almost an insulator to this feeble voltaic electromotive force. If, how- ever, a higher electromotive force be applied across the gap for even a very minute interval of time, its effect is to break down the insulator and FIG, 38. FIG. 41. FIGS. 38, 40 and 41. Types of Coherers. to allow the voltaic cell to send a continuous current. The sudden higher electric impulse changes the resistance offered by the bridge of metallic dust from a very high to a relatively low value. The exact nature of the action which takes place when the electric impulse operates, and ELECTROMAGNETIC WAVE-DETECTORS 127 when the resistance of the gap breaks down, is hard to determine with certainty. It has been much discussed and unanimity has not yet been reached upon the matter. It suffices for present purposes, however, to say that the extra voltage brought to bear across the gap of metallic par- ticles causes them to weld together, or to cohere electrically, thus converting a very bad joint in the local circuit of the voltaic cell into a fairly good one. Connections Between Coherer and Antenna The simplest method of connecting the coherer of Fig. 38 with the receiving antenna is indicated in Fig. 39. A B C S G is the antenna path to ground. It is cut at C, and the coherer inserted by means of the wires W W, Fig. 38. A local circuit EMC, indicated in broken lines, con- nects a suitable low voltaic electromotive force, such as a single voltaic cell, to the coherer ter- minals through an electromagnetic receiver, rep- resented as an ordinary wire-telegraph sounder. Prior to the arrival of electromagnetic waves, the gap of filings in the coherer interposes a high resistance in the local circuit, as well as in the antenna path. Consequently, no appreciable current flows through the sounder M, the arma- ture lever of which remains released against its 128 WIRELESS TELEGRAPHY upper stop under the action of a spiral spring. As soon as an electromagnetic wave, or wave- train, of suitable intensity passes the antenna, an oscillating electromotive force will be set up along the antenna and across the coherer gap. The FIG. 39. Essential Elements of Coherer- Connections When Receiving Signals. coherer will instantly break down in insulation, and will cause the metallic filings to cohere. The oscillations will not discharge through tne local circuit owing to the self-induction or choking effect of the coil on the magnet M. The voltaic cell E will now be able to send a current through ELECTROMAGNETIC WAVE-DETECTORS 129 the local circuit EMC and excite the electro- magnet M of the sounder, the armature lever of which will descend with a click, thus giving evi- dence of the arrival of the wave. Mechanical Decoherence The current in the local circuit would continue indefinitely after the bridging of the coherer gap by the first wave, if means were not provided for decohering, or restoring the coherer to its original insulating state. This may be done by giving a tap or light mechanical agitation to the coherer tube, thus shaking up the filings in the gap and breaking up the recently welded bridge between the plugs P P, Fig. 38. In practice, the arma- ture lever of the sounder M, may be arranged to deliver a light tap to the coherer tube at the same moment that it produces its click. This tap restores the original condition of the coherer, interrupts the local circuit and cuts off the excita- tion from the sounder magnet M, which promptly releases its armature lever under the influence of the spiral spring; so that the apparatus is again ready to respond to the next electromagnetic wave. The connections of the local circuit are usually somewhat more complex than Fig. 39 "shows; but the principle remains essentially the same. 130 WIRELESS TELEGRAPHY The sounder M, for instance, is not directly actuated by the local circuit of the coherer, be- cause it needs a relatively strong current, which is unsuitable. A delicate electromagnet called a relay (see Figs. 60 and 61) is, however, placed in the circuit at the point M, and the armature lever of the relay closes another local circuit through a more powerful voltaic battery and the sounder. A feeble current through the local circuit of the coherer is thus enabled to send a suitably strong current through the sounder. An auxiliary electromagnet, actuated also by the relay, is often applied to the sole duty of tapping the coherer, or causing it to decohere, after the relay and sounder have responded. It is evident that when the apparatus is in working order, signals consisting of short and long groups of electromagnetic waves will be able to spell out corresponding short and long opera- tions of the electromagnet M, or dots and dashes of the Morse alphabet. Fig. 40 represents a modified form of coherer in which there are two gaps g g. In each gap there is a little globule of mercury. The end plugs C C may be of carbon, and the central plug a little cylinder of iron. This form of coherer has the advantage that it is self-decohering, or auto-decohering. That is to say it needs no ELECTROMAGNETIC WAVE-DETECTORS 131 blow or agitation to restore the status quo after the passage of a wave. It normally possesses a high resistance. The electric oscillation or surge in the antenna breaks down this resistance momentarily and permits a current to flow through a local voltaic circuit. Immediately af- ter the passage of the wave the high internal resistance is restored. The reason for this re- markable action is concealed in the general obscurity of the whole subject of coherence, but is perhaps connected with the liquid state of the substance in the gaps. Another form of coherer is indicated in Fig. 41. It consists of a small insulating vessel or reservoir V, containing mercury. The mercury is brought into very light contact with the thin edge of a metallic disk P, kept rotating by clockwork. There is a very thin film of insulating oil on the surface of the mercury and the effect of the film is to insulate, or electrically separate, the metallic disk from the mercury. The thin film, may, however, be broken down, or electrically dis- rupted, by a relatively feeble voltage in excess of that used in a local voltaic circuit. The wires W W connect the device with the local circuit and the antenna, as in Fig. 39. On the arrival of a signal, theT)il film is broken and contact es- tablished between the disk and mercury; but 132 WIRELESS TELEGRAPHY the revolution of the disk almost instantly re- stores the oil film and decoheres the device. Hot-wire Receivers Coherers depend, as we have seen, upon the electromotive force or voltage of the surge set up in the antenna to force a discharge across the gap of imperfectly contacting matter, in order to give passage to a local vol- taic current. Among re- ceivers which depend, however, upon the mag- nitude of oscillating cur- rent set up through them when they are inserted in the path of the receiv- ing antenna, without interrupting the same, we have the hot-wire receiver. One form of this device is presented to view in Fig. 42. A pair of parallel brass strips AB, CD, are fastened near to each other and side by side, by an insulat- ing block F of hard rubber. Leading-in wires W W are soldered to these strips above, at A and C. Between the lower adjacent corners is sol- dered a little piece of silver wire e, bent into the form of a sharp V. This silver wire may be about 3 millimeters (0.12 inch) long and about FIG. 42. Hot-Wire Receiver, ELECTROMAGNETIC WAVE-DETECTORS 133 0.076 millimeter (0.003 inch) in diameter. A cross section of this silver wire is indicated in Fig. 43, at A B C. At or near the center q is a thin filamentary wire of platinum, like the wick inside a paraffin candle. The diameter of the FIG. 43. Cross -Section of , . . , . , Composite Wire. platinum wick is about i. 5 microns (0.0015 milli- meter or 0.000,06 inch), or about one-fortieth of the diameter of a thin hu- man hair. A platinum wire so fine is only ob- tained by thickly coating an ordinary size of plat- inum wire with silver, FlG 44 ._view of Loop of and then drawing down Sensitive Fine Wire tin- der Microscope, the thick composite wire through successively diminishing dies. As the silver wire gets thinner and longer, so also does the internally held wick or filament of platinum. After the little V loop of silver candle-wire has been soldered to the brass plates at B and D, Fig. 42, the device is carefully lowered into a bath of nitric acid, in such a manner that the point of the V loop is submerged in the acid, which immediately attacks and dissolves the sil- 134 WIRELESS TELEGRAPHY ver chemically, leaving the platinum wick un- injured. The process is aided by a feeble electric current from a local voltaic cell, is watched under the microscope, and is arrested at the proper stage. The appearance in the microscope of the V loop after the silver has been dissolved off the tip is shown in Fig. 44, where A B and C D are the 76-micron or 0.076 millimeter silver wires, and e f g the 1.5 micron platinum filament, hanging in a short loop. The device is then ready for use and is conveniently protected from injury by placing it in a short glass bottle or test-tube. The connection of the little hot-wire device with the receiving antenna is illustrated in its simplest elements at Fig. 45. A B is the antenna, connected to ground through the hot-wire at H. A local voltaic circuit, in broken lines, connects a feeble electromotive force, such as a single voltaic cell, through the telephone receiver T and the hot-wire H. Prior to the advent of electro- magnetic waves, a steady current flows through the local voltaic circuit, producing no sound in the receiver T. This current serves to warm the fine platinum wire, the electric resistance of which is appreciable, but constant at any con- stant temperature. As the temperature of the platinum is increased, however, the resistance increases. ELECTROMAGNETIC WAVE-DETECTORS 135 If now an electromagnetic wave or wave-train strikes the antenna B A, an oscillating current will pass through the fine platinum filament H, and will heat the same appreciably, being super- posed upon the steady current from the voltaic FIG. 45. Connection of Hot- Wire Receiver with Receiving Antenna. cell E. The antenna is prevented from dis- charging to ground through the telephone T; or by the path AB TE SG, owing to the self- induction, or choking effect, of the telephone. Practically all the discharge goes through H. The momentary increase in the heat and tern- 136 WIRELESS TELEGRAPHY perature of the filament H causes its resistance to be momentarily raised, and this reacts upon the local voltaic current, diminishing the same momentarily. The telephone T responds audi- bly to the sudden alteration of current, which lasts as long as the waves or groups of waves are passing, and ceases the moment the waves cease to arrive. The sensitiveness of the device is due to the small dimensions of the fine filament. The oscillating electric currents received through the filament from the antenna are very feeble when the antenna is far from the sending station; but the cross- section of the filament being only about 2 square microns (-g-grj- of one millionth of a square inch), even a very feeble electric current will be condensed to a relatively appreciable cur- rent density at the filament, thus giving rise to appreciable heating in a mass of metal only about 2,000 cubic microns in volume (^-sV^-th of one millionth of a cubic inch). Electrolytic Receivers The device, represented in outline by Fig. 46, consists of a small vessel C containing a suitable solution, such as dilute nitric acid. A candle wire w, of the kind above described in connection with Figs. 43 and 44; i.e., a silver wire of about 76 microns in diameter (0.003 inch) with a ELECTROMAGNETIC WAVE-DETECTORS 137 platinum wick about 1.5 microns in diameter (0.000,06 inch), is immersed to a suitable depth, perhaps a quarter millimeter (o.oi inch) in the solution. The acid dissolves off the silver, so that the filament of plat- inum is immersed in the solution, offering thereto an immersed surface area of about 1,200 square microns (2,660,060 th of a square inch). The wire w is fastened to the lower end of a brass screw hav- ing a milled head ss, the screw passing through a brass support P P. The depth of immersion of the fine platinum filament can be adjusted by turn- ing the milled head in one or the other direction. A solution of an acid or alkali traversed by an electric current is called an electrolyte, and the current is carried only by atoms, or groups of atoms, which are separated out from the solu- tion. In other words electric conduction through an electrolyte is accompanied by chemical de- composition of the electrolyte. The connection of the electrolytic receiver with the receiving antenna is essentially the same as FIG. 46. Electrolytic Receiver. 138 WIRELESS TELEGRAPHY that of the hot-wire receiver represented in Fig. 45. That is, a local voltaic current is provided containing a small electromotive force and a telephone receiver. Prior to the advent of the electromagnetic waves, the cell E, Fig. 45, sends a feeble steady current through the telephone and the electrolytic receiver. This current causes minute bubbles of gas to be liberated from the fine immersed platinum filament, but the telephone T gives no sound. As soon as an electromagnetic wave strikes the antenna, the oscillating current set up passes through the electrolytic receiver, and heats the same in the minute constricted mass of liquid immediately surrounding the fine platinum filament. The effect of the heat so liberated is two-fold. In the first place it momentarily raises the temperature of the pellicle of solution immediately surround- ing the filament, thereby reducing the electric resistance of the device; for electrolytes, unlike metals, improve in conductivity when heated. In the second place the gas is liberated more freely from the surface of the fine filament as the tem- perature is increased. In technical language, the momentary warming effect of the oscillating current causes the filament to be partly depolar- ized. Owing to both of these actions, the appar- ent resistance offered to the local current from ELECTROMAGNETIC WAVE-DETECTORS 139 the voltaic cell is temporarily diminished and the sudden increase of current produces a sound in the telephone. Immediately after the passage of the wave, the heat is dissipated by conduction into the solution, and the original resistance and counter electromotive jorce of polarization in the device are restored. The sensitiveness of this device is attributable to the constriction of the conducting path from antenna to ground into a minute volume of liquid, having an 'extremely small cross-section. In this tiny volume there exists a very appreciable electric resistance and also an appreciable elec- trolytic back voltage, like that of an opposing voltaic cell. The thermal effects of even a very feeble oscillating current from the antenna are here condensed into so small a volume that the temperature of that small volume can be appre- ciably raised. The rise of temperature has a powerful effect both on the localized resistance of the constricted liquid path and on the back voltage of the virtual opposing voltaic cell con- tained in the device. Another form of the electrolytic receiver is represented in Fig. 47. A vessel, V V V, such as a small glass tumbler, is filled with an elec- trolyte such as dilute nitric acid, or dilute caustic soda, to the level L L. Two metallic surfaces 140 WIRELESS TELEGRAPHY or electrodes dip into the electrolyte. One of these, indicated at E, may have any convenient size and shape, connecting with a leading-in wire W 2 ; or this wire may itself dip into the solution and form the electrode E. The other FIG. 47. Simple Form of Electrolytic Receiver. electrode like w, in Fig. 46, has extremely small surface area. A very simple way of preparing such a small area of electrode is illustrated in greater detail at the lower part of the Figure. A glass tube, a b c of any convenient dimensions, say 7.5 cms. (3 inches) long, 3 mm. and i mm. (about J and -fa inch) in external and internal ELECTROMAGNETIC WAVE-DETECTORS 141 diameters respectively, is slipped over a short length of copper wire. This copper wire is welded at one end in the flame of a Bunsen burner to a few centimeters (one inch, say) of fine platinum wire having a diameter of about 0.05 mm. (0.002 inch). In the illustration, the weld or junction between the copper and plati- num wires is shown at J. The tube is then heated to the softening point over the fine plati- num wire, and the softened walls are squeezed tightly over the platinum wire with a small pair of tongs, so as to seal in the wire hermetically for a short distance, say i cm. (f inch). The tube after cooling is now scratched with a file across the seal and is broken sharply across with the fingers. The break if properly made will leave the ruptured platinum wire very slightly project- ing beyond the glass. A few strokes, with a fine file, will file the platinum wire flush with the glass, thus presenting as at d, Fig. 47, an exposed disk of platinum of the diameter of the fine wire at the end of the seal b c, and with a surface area of about 2,000 square microns or T,7rdr,innr square inch. The depth of immersion of the glass tube in the electrolyte is of no consequence so long as the end of the fine platinum wire is fairly covered by the solution. Neither does the distance between 142 WIRELESS TELEGRAPHY the two electrodes in the vessel V V, make any appreciable difference. In other words, the seat of the actions occurring in the apparatus is in the constricted liquid path immediately covering and including the minute exposed area. The resistance of the device is practically all located within a millimeter (^ inch) of the exposure, and the voltaic counter electromotive force of polar- ization is located at the exposure; so that all the rest of the electrolyte merely provides an out- ward escape for the electric current that has passed through the intensely constricted region of influence over the minute area. This form of electrolytic receiver is probably the simplest to construct of all the receivers used in wireless telegraphy. Like most of the other devices it is patented. It is not capable of ad- justment in surface exposure area, like the appa- ratus of Fig. 46, and if the minute exposure gets dirty or clogged, it has to be thrown away and a new one substituted. It is, however, capable of great sensitiveness and the materials for its con- struction are very easily obtained. Electromagnetic Receivers There are various types of electromagnetic receivers, but that illustrated in Fig. 48 is gener- ally admitted to be the most sensitive. It con- ELECTROMAGNETIC WAVE-DETECTORS 143 sists of a flexible band b b b b of iron wires pass- ing over the grooved pulleys L L, which are steadily driven by clockwork. The band of iron wires moves through a glass tube 1 1, on which is placed a winding of insulated wire with external connections W W, leading to the antenna on one side and to ground on the other. This winding, T FIG. 48. Magnetic Receiver. inserted in the antenna path, forms the primary winding of a little induction coil, in which the moving band of iron wires is the core. The secondary winding S, placed over the middle of the glass tube is connected to a telephone re- ceiver, T. The band of iron wires in passing through the tube makes its procession in front of the two fixed permanent horseshoe magnets M M. These 144 WIRELESS TELEGRAPHY are so arranged with regard to strength and direction of polarity, that the band of iron emerges from the tube with its internal magnet, ism reversed in direction from that with which it enters. Iron has a curious magnetic property when its magnetism is cyclically reversed. If the mag- netism is established along a band of iron wires in one direction, then when the process of de- magnetization and reversal is started, the change of magnetic flux in the iron takes place very slowly at first, until a certain stage of magnetic instability is reached, and then the magnetic flux reverses with great swiftness. The action may be compared to that of a ball moving alternately from side to side on the deck of a rolling ship at sea. If the deck is flat and plane, the ball will swing regularly from side to side. If, however, the deck be somewhat bowed, rising in the middle like a turtle-back, the ball will be slow to return on each roll until it gets to the top of the turtle- back and then it will run down with great speed. The function of the two magnets M M in the magnetic receiver is to bring the iron wire core of the induction coil into the unstable magnetic condition during its passage within the primary winding connected in the antenna path. Under these conditions, if any electric oscillations come ELECTROMAGNETIC WAVE-DETECTORS 145 through the primary coil from the antenna, they will be able to shake out the magnetic flux in the enclosed band and precipitate its reversal. The rotating mechanism brings the magnetic flux to the edge of the precipice, as it were, and the feeble electric currents are able to push it over. The sudden change of magnetic flux inside the secondary coil s, sets up an electric impulse that will produce an audible sound in the telephone T. The above form of magnetic receiver is thus essentially an induction coil with the antenna path passing through the primary winding and the delicate receiving telephone in the secondary. The induction is increased in sensitiveness by the aid of the constantly renewed magnetic instabil- ity in the iron core, under the action of the per- manent magnets. Comparison 0} Receivers Comparing the behavior of the various types of receiver, it is to be noted that the coherer is the only one which permits of a permanent record being obtained. The coherer, as outlined in Fig. 39, operates an electromagnetic receiver of the Morse type. Such a receiver is able to record the message in dots and dashes inked upon the surface of a long strip of paper, coiled on a roller in the apparatus and moved by clockwork, A 146 WIRELESS TELEGRAPHY particular form of Morse inkwriter is seen in Fig. 49. It will be observed that the armature consists of a split soft iron tube fastened to a rocking lever in such a manner as to be attracted downward when the black coated electromagnet FIG. 49. Morse Inkwriter. is excited. The lever throws up a disk against a moving band of paper not shown. The disk is kept rotating by clockwork and dips into an inkwell. On the other hand, however, the speed at which signals can be received and recorded by means of the coherer is distinctly lower than that obtainable with non - recording receivers. A speed of 15 to 20 words a minute is considered good with a recording receiver. With some non- recording receivers this speed may be doubled. ELECTROMAGNETIC WAVE-DETECTORS 147 In regard to sensitiveness, the coherer has hitherto proved much inferior to the others. The most sensitive is the electrolytic receiver and next to that the magnetic. Both of these use the telephone as the receiving instrument. Telephone Receivers A convenient form of telephone receiver, illus- trated by Fig. 50, is such as telephone operators employ. A leather-covered steel band L L goes over the head and supports the receiver R R close to one ear. The band is fastened to the receiver by the thumb-screw s. The covered wires w w serve to connect the receiver with the antenna system. Fig. 51 shows the parts of the receiver disassem- bled. B B is a hard rub- be r box with a screw cover C. Inside the box are three pairs of half- ring steel permanent magnets, NS, NS. In the center, a pair of soft iron pole pieces are sup- ported, receiving their polarity from the magnets NS and wound with many turns of fine silk- covered copper wire connecting with the leading wires ww. D is the ferro-type disk of steel FIG. 50. Head Tele- phone. 148 WIRELESS TELEGRAPHY which is clamped around its edge between the box and the cover, so as to be held over but not quite touching the poles at the center. The sensitiveness of the electrolytic and mag- netic receivers is at least partly attributable to FIG. 51. Head Telephone Disassembled. the great sensitiveness of the telephone which they employ as their intermediary with the human brain. The telephone, as is well known, is extraordinarily sensitive in detecting feeble electric currents undergoing rapid variations CHAPTER XI WIRELESS TELEGRAPH WORKING Alternate Use o] an Antenna jor Sending and Receiving IN order to carry on simple wireless telegraphy between a single pair of stations, remote from all other wireless telegraphists, it is evidently neces- sary to have an antenna at each station. The dimensions required for the antennas will depend upon the distance between them. For sending messages from one room to another in the same building, the antennas may be a few centimeters or inches long. For sending messages between adjacent buildings, or buildings separated only by a few kilometers or miles, the antennas need only be a few meters or yards high. For dis- tances of hundreds of kilometers or miles, large and tall antennas are at present necessary. The object of the sending station in long-distance wireless telegraphy is to throw out as long a train of powerful waves as possible, while that of the receiving station is to employ as sensitive a re- ceiver as possible. 149 150 WIRELESS TELEGRAPHS One and the same antenna is used at a station for sending and receiving alternately. The con- nection is changed from the sending to the receiv- ing apparatus by a switch as indicated in Fig. 52. tife FIG. 52. Diagram of Switch Connections from Sending to Receiving. The switch S, has a metallic blade or lever-arm which is pivoted at c and may be turned into con- tact with the point d for sending, or with the point r for receiving. A is the antenna, or the wire leading thereto, and G the ground-connec- tion. In the position shown, the switch is turned to the sending position and the antenna is con- nected to the spark-gap k. This is excited by the induction-coil I and vibrator V, when the WIRELESS TELEGRAPH WORKING- 151 key K in the primary circuit is depressed. There is also an auxiliary circuit X, consisting in this case of a pair of condensers with a coil between them, to increase the stock of radiation-energy prior to discharge. The receiving apparatus is indicated as consisting of a coherer H, working into a relay R, (see Fig. 61) through the local circuit H R b. The relay in its turn operates the sounder M through a second local circuit, including the voltaic cell v. It is necessary to make sure that the delicate receiving apparatus is completely disconnected from the antenna while the latter is being used for sending. Some- times an automatic switch is used which will not permit the induction coil to be excited unless the receiver is cut off. In some installations the coherer and its immediate connections are shut up in a metal box to keep accidental waves from impinging upon the coherer. The sending key used in wireless telegraphy differs only in mechanical details from the ordi- nary Morse key of wire telegraphy. An ordinary Morse key is illustrated in Fig. 53, and a form of wireless telegraph key in Fig. 54. The wire- less key has to send a much stronger current to the induction coil than that which an ordinary wire - telegraph key controls, so its contact is larger and stouter. Moreover, there is a pos- 152 WIRELESS TELEGRAPHY sibility of the operator receiving a severe electric shock from the induction coil; so the insulating handle is made more massive. There is apt to be some sparking at the key contact on breaking FIG. 53. Ordinary Wire- FIG. 54. Form of Wireless Telegraph Morse Key. Telegraph Morse Key. circuit, and certain forms of key are designed to overcome this, in some cases by breaking the contact under oil, and in others by breaking the contact between the poles of an electromagnet. Morse Alphabets or Codes In sending signals, the contacts of the key are made by the operator in conformity with the Morse code. There are unfortunately two tele- graph codes the American Morse code, or that in almost universal use on the North American continent, and the international Morse code, or continental Morse code, in use in practically all other parts of the world. The American Morse code was introduced in the early days of the Morse system in the United States. In Europe, WIRELESS TELEGRAPH WORKING 153 about the same time, a number of different al- phabets or codes sprang into existence, in differ- ent places. The frequent transition and inter- communication of messages among European countries was soon hampered by differences in Morse alphabet, so that, by international con- vention, the present Continental code was ar- rived at and adopted. Unfortunately, the date of this convention was prior to the introduction of Atlantic cables and fast ocean steamers so that America was not a party to the conference. The two codes are presented in Fig. 55, so far as concerns the use of the English language. It will be seen that the signals for a, b, d, e, g, h, i, k, m, n, s, t, u, v, w, and 4 are common to both codes; while the signals for c, f, j, 1, o ? p, q, r, x, y, z, i, 2, 3, 5, 6, 7, 8, 9, o, . , ?, and ! are different. In both codes the dot is the standard element of length. A dash has the length of three dots, and the space separating dots or dashes in a letter are of dot length; except in the American letters c, o, and z which are called spaced letters, and in which there is an extra space of two dots' length. The American 1 is a six-dot length dash, and the Americai) zero is a nine-dot length dash. The space separating adjacent letters is three dots long and the space separating words, six dots long. i54 WIRELESS TELEGRAPHY The American code is shorter on the average, in the signaling of ordinary English, by about 5 per cent.; that is to say 95 dot elements of American code will be equivalent in the forma- tion of letters to 100 elements of international code, so that the American code is swifter by about this amount. On the other hand, the spaced American letters are a source of possible errors, if the signaling is not in the best condi- tion, and some operators who are practically acquainted with both codes maintain that, owing to the care needed in sending spaced letters, there is no sensible difference in swiftness be- tween the codes. It is certain that a person con- versant solely with one of these codes is quite unable to read messages sent in the other. Listen- ing to the other code in such a case is like listening to foreign unknown speech. It is greatly to be regretted that this confusion of telegraphic lan- guage exists, and wireless telegraphy has tended to make the confusion more evident. Most of the ships carrying wireless telegraph outfits talk international Morse. Only those like the Fall River Liners, on the American coast, talk Amer- ican. Some ships can talk in either code. Each dot contact of the sending key must be accompanied by at least one discharge of the induction coil and, therefore, by at least one WIRELESS TELEGRAPH WORKING 155 Cmttuintil American. fettiVunhl JU**. A - w B - ... X C ... y D . - z E - i F .-. 2 G 3 H - 4 6" -- J 6 . K 7 L .. - 8 M ; - 9 O P J . R i_. *. sn ..._- 6 - t T - ; U -- - I V "" & FIG. 55. Continental and American Morse Telegraph Codes. 156 WIRELESS TELEGRAPHY spark, or train of oscillating spark discharges at the spark-gap &, Fig. 52. The dashes, on the other hand, need to set up a series of discharges. A dot may thus be accompanied by a single inter- ruption at the vibrator and a dash by three or more such interruptions. Alternations 0} Sending and Receiving. During the time that the message is being sent, the operator is unable in the ordinary wireless system to receive a message, or to know whether the receiving operator has been able to take the signals. In ordinary wire telegraphy, as prac- ticed in the United States, the receiving operator can "break" the line circuit, and thus notify the sender that the message is not being taken. But the receiver of a wireless telegraph message cannot ordinarily stop the sender, and the sender goes on until either the message has been con- cluded, or until he deems it prudent to turn his switch and let the receiver send back encourage- ment to proceed. For the same reasons, it is not uncommon for the sender to repeat a message, as soon as he has finished it, in order that the distant receiving operator may check what he has written against the repetition, so as to avoid mistakes. Attempts have already been made to send and receive messages simultaneously by the WIRELESS TELEGRAPH WORKING 157 same antenna, but such operations can only be regarded at present as in a subsidiary stage. As soon as the difficulties in the path of ordinary, simple, simplex to-and-fro wireless telegraph signaling have been overcome, it will be time to attend to the development of more intricate methods. The Insulation of Antennas The insulator or insulators which support the sending antenna have to be maintained in good order, because they must withstand high voltage without appreciably leaking or sparking. Where- ever the antenna wire or wires come in contact with any substance, an insulator has to be used, especially where there is a chance of moisture from rain, dew or fog. The antenna may itself be an insulated or rubber-covered wire. This is, however, of but little service except in preventing atmospheric discharges from driving wind and snow. On the other hand, the insulation of an antenna used solely for receiving messages need not be so carefully maintained. In order to send messages, powerful voltages must be used and insulated; but in order merely to receive mes- sages, or to listen to what is going on in the neighboring ether so far as concerns long electro- magnetic waves, only feeble voltages are pro- 158 WIRELESS TELEGRAPHY duced ; so that while insulation is proper, it need not be safeguarded elaborately. A bare copper wire fastened around a branch oi a tree and touching the boughs or trunk at several places on the way down, may often enable messages to be received. Such a wire could hardly be used for sending messages to a distance. Heights of Antennas Antennas are carried to various heights, as already mentioned. Since the trouble and ex- pense of construction increase rapidly above a heigh;, of 30 meters (32.8 yards), very high masts are only installed for very long distance trans- mission. The greatest height to which they have been thus far carried is 128 meters (420 feet), in the form of a steel chimney. On board ship the mast height usually limits the elevation of the antenna to about 30 meters (98.4 feet). A widely extending antenna is not so useful for receiving as for sending, except when the waves are much longer than four times the mast height ; because although side expansion no doubt affords an increase of catchment area for passing waves, there is little doubt that the area from which energy is absorbed by a receiving antenna is fairly wide (see page 84), even when the antenna consists of but a single vertical wire. On the WIRELESS TELEGRAPH WORKING l>9 other hand, increasing the height of a receiving antenna increases the energy that can be scooped out of a passing wave-train approximately as the square of the elevation, provided that resonance is maintained; so that increase of height always aids long-distance reception. The ordinary height of a shore antenna mast is about 45 meters or 150 feet. If a mast is ob- served on the seashore with a little wooden house at the base, it is highly probable that it is con- structed for wireless telegraph purposes. If a group of several wires can be seen to festoon from the masthead to the house, the existence, either in the present or past, of a wireless tele- graph station may safely be assumed. Although the usual method of installing an antenna is to have the same insulated at the top and well grounded at the base, yet other methods are possible and are occasionally employed. For example, the antenna may consist of an arch or vertical loop, either one or both ends of which may be grounded. Again, the ground connec- tion of an antenna may be dispensed with en- tirely, and an insulated metallic plate used in its place, the plate being preferably supported parallel to the ground and at a height of about 2 meters (6.5 feet) above it. Such a plate is equivalent to an air condenser inserted in the 160 WIRELESS TELEGRAPHY antenna path near the ground connection. Or, the ground may be dispensed with, and a hori- zontal wire may be run at a short distance above the ground. A number of such variations of installation have been suggested or used at differ- ent times; but while they introduce differences of action in detail, they usually conform to the same broad, fundamental principles of action as the ordinary lightning-rod type of vertical wire antenna with grounded base. Power Required for a Wireless Telegraph Sending Station In order to receive wireless signals, no power has to be expended beyond the almost infinitesi- mal amount supplied by voltaic cells in the local circuit or circuits of the receiver. But in order to transmit signals to a distance, very appreciable power must be supplied. For distances under 100 kilometers (60 miles), the power supplied does not usually exceed 3 kilowatts (4 horse- power), and is sometimes considerably less. As the distance increases, the power absorbed is likewise increased. A few stations use 50 kilo- watts (67 horse-power) or more. As we have already seen, the antenna radiates the power by jerks, or with long intermissions; so that al- though in sending with the key depressed, the WIRELESS TELEGRAPH WORKING 161 power supplied to the primary winding of the induction coil may be; say, 2 kilowatts (2 horse- power), the antenna may be varying in power between 200 kilowatts and o. In the neighborhood of a city, or of an electric lighting system, the power required may be taken from electric-light mains. The same facility may be afforded on vessels electrically lighted. Otherwise, a small gas, oil or steam- engine has to be installed, in order to generate the power required. Contrast Between Power Required jor Wire and Wireless Telegraphy The contrast between wire and wireless teleg- raphy is nowhere more marked than in respect to power. In wire telegraphy, over a distance of, say, 200 kilometers (120 miles), the power re- quired for the transmission of messages is per- haps 8 watts at the generator (5.9 foot-pounds per second, or 0.0107 horse-power). In wireless telegraphy, working over the same distance, 8 kilowatts (10.8 horse-power) may be expended, or a thousand times more. Even then, the wire- less system is only rendered practicable by the enormously greater sensitiveness of the receiving apparatus it employs. In the one case, the electromagnetic waves are guided in a single 1 62 WIRELESS TELEGRAPHY beam to their destination by a wire, and the only serious loss is by leakage over the insulators along the line, or by imperfect conduction and internal warming of the line wire. In the other case, the power is scattered in every direction and only an extremely small fraction can be picked up and utilized at the distant receiving station. Never- theless, the fact that a wire can be dispensed with is so wonderful, that we may be glad to obtain wireless telegraph communication at any rea- sonable cost, and not be over anxious to cavil at the waste of energy. Spreading 0} Waves jrom the Sending Station We have already seen that electromagnetic waves employed in wireless telegraphy start off as hemispherical waves, expanding at light- speed in all directions. The hemispherical formation fairly commences about a half-wave length from the sending antenna. Inside that radius the wave is much more complex in form. Theoreti- cally, the expansion goes on indefinitely, allowing for the curvature of the earth. Practically, very little has yet been determined about the matter, except that signals from a sending antenna have been detected in balloons at moderate elevations above the earth not far from the antenna, and also by observers on the earth at a distance not WIRELESS TELEGRAPH WORKING 163 exceeding 5,000 kilometers (3,000 miles) in any direction. No receptions of messages have been yet reported over a greater distance than this. FIG. 56. The Boston Hemisphere of the Globe in Stereographic Projection; or the Hemisphere with Boston as Pole. Fig. 56 represents the spreading of the waves from the sending station according to the existing theory. The sending station is sup- posed to be located at Boston, Mass., which is, 164 WIRELESS TELEGRAPHY therefore, placed at the pole of the earth, for the purpose of this discussion. Each of the dotted circles i, 2, 3, . . . 9 represents an over- sea distance of ten degrees, or 600 nautical miles (i,m kilometers or 690 statute miles). In pass- ing from one circle to the next, the waves will take -2To~th of a second, or 3,700 microseconds. The complete journey from Boston at the pole, to the equator on circle 9, or one-quarter round the world, would occupy -^j-th second. It will be observed that the outgoing waves strike New- foundland, Bermuda and the Great Lakes, about the same instant, -^-yc"^ f a second after being emitted. Cuba and the Greenland coast would be struck almost simultaneously in -g-f-g-ths sec- ond. When the waves reach England, in about S-fg-ths second, they will also be striking Alaska, the Pacific coast, Norway, France, Spain, Mo- rocco and the Brazilian shore. They will also nearly have reached the North Pole. After j-fg-ths second, Japan, Thibet, Persia, Egypt, the Gold Coast, Argentina and Peru will be reached almost simultaneously. The corresponding de- velopment of the waves, in section through the center of the globe and the sending station, is presented in Fig. 57. Boston would occupy the position P, and a point on the ocean south of Perth, Australia, would occupy the antipodes P'. WIRELESS TELEGRAPH WORKING 165 The expanding waves would occupy the succes- sive shells in, 222, 333, and 444, the last being attained in ^th of a second. The equatorial circle Q 222 Q corresponds to the circle 9 in FIG. 57. Hypothetical Expansion of Wireless Tele- graph Waves Over the Globe. Figure 56. If will be observed that after the wave passes the equatorial circle Q Q, it narrows its circle, and when it reaches the antipodes P', it has gathered all its feet to this point. This should mean that the weakening of the waves, which occurs by expansion at the outset, dimin- ishes slightly on the surface of the globe after 1 66 WIRELESS TELEGRAPHY passing the equator, and the signals at the an- tipodes P' should be relatively stronger than at other points in that region. All this is, however, as yet entirely inferential, because the waves have not yet been detected beyond the first posi- tion 1,1, about one-eighth of the distance around the globe, or half way to the equator. Moreover, it is uncertain as to what happens to the expanding waves in the upper regions of the earth's atmosphere, where no balloon can ascend to take observations. Air is an excellent insulator at ordinary pressures; but when air is exhausted from a vacuum-tube, the dregs of air remaining within can be made to conduct electric discharges better than sea- water. The pressure of the air is a maximum at the earth's surface, and dwindles indefinitely as the earth's surface is departed from. At an elevation of from 20 to 100 kilometers above the earth (12 to 60 miles), depending upon local conditions, there must be strata of rarified air having as low a density as that produced in such vacuum-tubes. It is un- certain whether very feeble electric forces can evoke conduction in such rarified air. That is to say, it is quite possible that conduction may occur in laboratory vacuum-tubes after the rari- fied air has been modified, or ionised as it is called by the action of a relative powerful elec- WIRELESS TELEGRAPH WORKING 167 trie flux, and that in the absence of such degrees of electric flux-density, the rarified air would fail to conduct. If there is a shell of rarified air above the earth at a height of, say, 70 kilometers (roughly 40 miles) which suddenly conducts like sea-water, then the electric flux would cease to expand into the space beyond, but would skim along the interior wall of this conducting shell. The waves would then be confined to expansion sideways between the earth's conducting surface below and the internal wall of the concentric globe of rarified air above. This would tend to reduce the attenuation or weakening of the waves markedly; because beyond the radius of 70 kilo- meters, the expansion would continue in but two dimensions longways and sideways instead of three dimensions including height. On the other hand, however, there might be layers of rarified air which conducted even under very feeble electric flux, and yet the conduction might be gradual instead of sudden. If the conduction increased gradually to a maximum through many kil )meters of ascent, there would be loss of energy by reason of imperfect conduction and eddy- currents in the transitional layers, so that the benefit due to ultimate confinement of the waves within the rarified shells, might be more than lost by waste of energy in this partial conduction. 1 68 WIRELESS TELEGRAPHY The whole subject of wave contour at great dis- tances must remain in abeyance until sufficient measurements have been made of the wave- strengths at different distances to enable the con- tours to be inferred. Very little information is yet obtainable as to relative wave-strengths at great distances, partly owing to the difficulty of measuring extremely feeble wave-intensities and partly owing to atmospheric variations, to be referred to later. At short distances, i.e., 100 kilometers (about 60 miles) or less, the few re- sults obtained appear to indicate that the energy received diminishes roughly as the inverse square of the distance, or in conformity with simple hemispherical expansion. Experimental Apparatus for a Range 0} a Few Meters For simple experimental and demonstrative purposes, apparatus is readily obtainable that will produce recognizable signals at very short range and with an insignificant expenditure of power. The apparatus of Fig. 58 consists of a small induction coil with vibrator, to which the primary current from an external voltaic battery is admitted by the Morse key K. The terminals T t are for connection to a voltaic battery. The secondary winding of the induction-coil charges WIRELESS TELEGRAPH WORKING 169 the insulated double-rod system R R. By this means short waves are thrown off which strike FIG. 58. Simple Form of Wireless Sending Apparatus for Transmitting to a Distance of a Few Metres. FIG. 59. Receiver for Experimental Wireless Telegraph Set of Few Metres Range. the receiver in the vicinity. A form of receiving apparatus capable of being used with such a set is indicated in Fig. 59. Here the horizontal 170 WIRELESS TELEGRAPHY glass-tube coherer C is connected to the binding- posts 3, 3 on the wooden base. Short projecting wires may be clamped in these to help seize the passing waves. A voltaic cell is also connected through binding posts 2, 2 with the coherer, through the relay R. This relay is actuated when the coherer responds, and closes a local circuit through the vibrating electric bell B and FIG, 60. Simple Neutral Relay. another voltaic cell connected to binding-posts i, i. The ringing of the bell not only gives the signal, but also agitates the coherer and restores its normal insulation, after the passage of the wave. The relay R is represented in greater detail in Fig. 60. A pair of electromagnetic coils M are wound with silk-covered fine copper wire con- nected at the ends to the main terminals T T, The armature of the relay is free to move about a WIRELESS TELEGRAPH WORKING 171 horizontal axis through a small play, set by the two uppermost opposing screws. A spiral spring adjusted in tension by the set-screw S, draws the armature away from the electromagnetic poles, when these are unexcited by electric current in the coils. The armature lever then rests against an insulating stop. On being attracted by the magnet poles, the lever _ FIG. 6 1, Polarised Relay, strikes the contact point C, thus completing a local circuit through the terminals L L, and wires to the same underneath the base. In long-distance wireless telegraphy with the coherer, a more delicate form of relay is gener- ally employed, such as that shown in Fig. 61. In this form the delicate lever plays, under glass cover, about a vertical axis and the adjustment is provided by the screw on the side. The simple apparatus of Figs. 58 and 59 for sending messages across a hall, or from one building to another near by ? is of some practical importance, because a wireless telegraph station equipped to receive messages over a distance of hundreds of kilometers or miles may not disdain 172 WIRELESS TELEGRAPHY to employ an even still simpler apparatus of the same character, for testing purposes. An ordi- nary vibrating electric bell, such as that shown at B in Fig. 59, when excited by a voltaic cell, rapidly makes and breaks its circuit at the con- tact point of the vibrator. Each such interrup- tion is usually accompanied by a little spark at the vibrator contact, and a feeble electric wave, or very brief wave-train is thrown off from the circuit. Such an apparatus may be installed, say, on the wall of the wireless telegraph station, and excited by pressing an ordinary push button at the receiving operator's desk. The operator wishing to ascertain whether his receiving appa- ratus is in order, may do so in the absence of any incoming signals, by pressing the bell-button. The feeble electromagnetic waves thrown off from the bell wires are thus enabled to attach themselves to the antenna wire or wires, and so to produce a feeble signal that the operator can recognize. CHAPTER XII i TUNED OR SELECTIVE SIGNALING SYSTEMS The Problem 0} Selective Signaling IN the last chapter we considered wireless tele- graph working between two stations to the exclu- sion of all others within the working range. But the earth's atmosphere is no longer an Eden with but a single pair of occupants. In most parts of the civilized world, the actual problem is how to communicate with the station that is wanted, and yet to keep out of communication with dis- interested stations. Nature of Interference When a ship carrying an ordinary untuned coherer receiving set occupies a position at which wave signals are passing from only one sending station, the coherer is able to transmit those signals correctly to the Morse sounder, or ink- writer, in its local circuit. The same may be true if there are two sending stations working simultaneously, one of which produces much 174 WIRELESS TELEGRAPHY more powerful wave signals, at the ship's posi- tion, than the other. The adjustment of the coherer may be such that the signals which are feebler, either owing to greater distance, or weaker transmitting apparatus, are unable to interfere, and only the stronger signals are re- corded on board the ship. If, however, there are two or more stations in the neighborhood sending signals simultaneously, and the inter- secting waves from these stations are about equally strong, the ship's coherer tends to re- spond to all of the signals, or to give an unintel- ligible mixed record. It is true that if the waves from the different competing sending stations have different lengths, that length of wave which most nearly conforms to the quadruple of the equivalent ship's antenna height will preponder- ate in strength. Nevertheless, in an untuned system, the other waves are likely to interfere. This is partly because, as we have already seen, a simple sending antenna, not tuned in connec- tion with an auxiliary discharging condenser, tends to emit very short wave-trains, that are virtually but solitary waves with tails to them; and partly because a simple coherer, at the base of a simple receiving antenna, does not admit of much resonant building up of voltage, even with long wave-trains. TUNED OR SELECTIVE SIGNALING 175 Auditory Selection This interference and jumbling together of signals from different sending stations in the neighborhood was soon found to constitute a menace to wireless signalling on any extensive scale, especially with the coherer type of receiv- ing instrument. With other types of receiver which operate through a telephone, less trouble from interference is liable to be felt. This is for the reason that in a telephone receiver the signals usually have a buzzing sound, or possess a defi- nite semi-musical tone. The pitch of the tone corresponds to the frequency of the induction coil vibrator at the sending station; or, as it is termed, the group- frequency; i.e., the number of impulses per second delivered to the sending in- duction coil when the sender's key is held down (see Fig. 33), or to the number of groups of waves emitted per second. As a general rule, different sending stations do not employ just the same group frequency, or pitch of vibrator, so that the characteristic buzz or tone of the signals heard in the receiving telephone is different for differ- ent stations. When, therefore, a number of sta- tions are sending messages simultaneously in the neighborhood, an untuned receiving telephone set will render them all audible at once. If they 176 WIRELESS TELEGRAPHY have all exactly the same tone, it would be im- possible to make anything of the jumble of signals; but if, as usually happens, the tones are appreciably different, the conditions resemble the jumble of sounds maintained in a reception room, when many individuals are speaking close by at once. It is often possible, with a little effort, to focus attention on one particular succession of tones and mentally to read the signals they con- tain, to the exclusion of all the others. Need for Resonant Selection There is, however, a limit to the possibilities of deciphering one tone of telephonic signals from among a crowd. If a powerful antenna near a civilized seashore is connected untuned to a sensitive telephone employing receiver, a regular babel of signals is frequently to be heard. Some of these are from shore stations nearby, others from distant shore stations, and yet others from ships at sea. As we look upon the surface of a large lake, or of the sea, we usually discern waves or ripple trains, which are crossing, superposing, or intersecting in endless variation. A calm, unruffled water surface is the exception. The same conditions now apply to the atmospheric ether in civilized districts, so far as concerns long TUNED OR SELECTIVE SIGNALING 177 electromagnetic waves. The atmospheric ocean is rarely quiescent. Adjustment to Resonance In order to bring about sharply selective signaling, it is necessary that the receiving an- tenna and apparatus connected therewith should be tuned to one definite wave-length, so as to respond to waves of that length exclusively, and also that the sending station desiring to commu- nicate solely with that receiver should be tuned to emit as long wave-trains as possible, possess- ing this particular wave-length. Theoretically, it should be possible to adjust an antenna into resonance to a given wave-length within any de- sired degree of precision, so that if the waves received were one per cent, too short, or too long, they would fail to actuate the receiver. There is, however, a limit to precision in practical tuning for various reasons. About five per cent, above or below is usually regarded as satisfactory. That is, a receiving antenna and apparatus can be arranged to respond to a given wave-length of arriving signals, and not ordinarily to respond to waves five per cent, shorter or longer. This' means that the receiver would not ordinarily recognize, or report in the telephone or recording iy8 WIRELESS TELEGRAPHY apparatus, any passing waves outside of these limits. There are various ways of connecting, sending and receiving antenna circuits in order to tune them. One such way is indicated in Fig. 62 for the sending apparatus, and in Fig. 63 for the B FIG. 62. Particular Set of Tuned Sending Connections. receiving apparatus. By means of a suitable switch, or group of switches, the change may be made from one set to the other, that is from send- ing to receiving, with one and the same antenna. Referring to Fig. 62, A B is the antenna or wire leading thereto, C an adjustable coil for virtually TUNED OR SELECTIVE SIGNALING 179 altering the equivalent height and wave-length of the antenna. A high-frequency induction coil of relatively few turns without any iron core, is indicated at L, the secondary winding L 3 being connected to the antenna, and the primary L x to the secondary terminals of the spark coil S, FIG. 63. Particular Set of Tuned Receiving Connections. through adjustable condensers c c*. The key K is in the primary circuit of the spark coil S. The circuit s, c, L x , c f is adjusted to oscillate electrically at the required frequency and the antenna path B, A, C, L a is also adjusted to 180 WIRELESS TELEGRAPHY oscillate at the same frequency. It generally results that there are at least two different fre- quencies, and not merely one frequency of oscil- lation set up in such a system; but one of the frequencies is taken as the effective or working frequency, and the others are regarded as in- effective or merely parasitical. Turning now to the receiving connections of Fig. 63, A B is the antenna wire and C an adjust- able coil as before. L is a high-frequency induc- tion coil. In the secondary circuit of this coil is an adjustable small coil 1, an adjustable con- denser c, and the receiver r, connected with the local telephone, or recording instrument. The antenna path is adjusted to oscillate at the re- quired frequency, and the secondary circuit L 2 , 1, c, r is also adjusted to oscillate at this frequency Under these circumstances the whole receiving system is tuned to resonance with the single re- quired frequency, or wave-length. Simultaneous Sending and Receiving with Aid of Differential Resonance When tuning is carefully and effectively car- ried out, it may enable remarkable results to be accomplished. For instance, it has been found possible to receive messages over an antenna at, say, the foremast of a steamer, and at the same TUNED OR SELECTIVE SIGNALING 181 time to send messages in another wave-length, from an antenna at the mainmast, to a different and perhaps very distant station. It is evident that this result would not ordinarily be possible with simple untuned apparatus at both antennas, nor would it be possible with tuned apparatus, if the frequency and wave-length of both the sending mast and the receiving mast were the same. Increase oj Transmission Range by Means of Resonance The advantages of tuning are found not only in the elimination of interference from extraneous signaling stations, but also in increase of sensibil- ity and effective signaling range. The tuning of the sending station connections permits of in- creasing the length of the train of waves follow- ing each discharge of the spark coil. The sym- pathetic tuning of the receiving station connec- tions permits of building up a resonant current strength in the receiver, due to the successive additions of impulses from the successive waves in the train. By this means, signals which would be too faint to be recognizable if they de- pended upon the impulse of a single wave, be- come recognizable by the cumulative impulses of a number of successive waves. There is good 1 82 WIRELESS TELEGRAPHY reason, therefore, for expecting that when a suit- able high-frequency dynamo is developed for the continuous maintenance of power in the sending circuit and antenna, a further great increase in effective range of transmission will be rendered possible. Reduction of Atmospheric Disturbance by Means oj Resonance Another advantage derived from tuning is in the direction of minimizing the influence of atmospheric discharges. An antenna is a sort of lightning rod, and the taller and more exten- sive it is, the better atmospheric discharger it tends to become. The atmosphere contains electrically charged particles or free electric charges. Their presence may be accounted for in several ways that need not here be discussed. Consequently, an antenna is apt to receive a perpetual stream of little electric discharges from the layers of air near its top, to the ground at its base. This action is quite distinct from the much more powerful discharges which may occur over the antenna in the presence of a thunder- storm in the vicinity. During such a thunder- storm it is usually necessary to stop signaling and to keep the antenna grounded. The little atmospheric discharges become objectionably TUNED OR SELECTIVE SIGNALING 183 noticeable in the receiver, and sometimes give rise to false signals. These continuous atmos- pheric disturbances are stated to be more notice- able and troublesome in the tropical than in the temperate zones, but they vary in intensity from day to day and hour to hour. On some occa- sions they are almost entirely absent, and on other occasions they are markedly prevalent. Distant thunderstorms and atmospheric dis- charges likewise produce noises in the receiving telephone, interfering more or less with received signals. Tuning of the antenna connections is capable of reducing atmospheric disturbances, although perhaps not of eliminating them en- tirely. Multiple Wireless Telegraphy by Means of Resonance The results of tuning have even been carried further. It has been found possible to send two messages simultaneously over one and the same sending antenna, by connecting the antenna to two different coils, and auxiliary oscillating send- ing circuits, or to two different sections of one and the same coil. Again, it has been found possible to receive two messages simultaneously over the same receiving antenna by a correspond- ing connection to two oscillatory receiving cir- 184 WIRELESS TELEGRAPHY cults. This means that an antenna system may be arranged to emit two frequencies, or wave- lengths, simultaneously and independently. In the same way, one antenna may be arranged to resonate to each of two different frequencies or wave-lengths. By connecting a plurality of such oscillating circuits with an antenna, it is theoreti- cally possible either to send or to receive an in- definite number of messages simultaneously, each in a definite appropriate wave-length. Up to the present time, however, but little use has been made of this possibility. A multiple system of wireless telegraphy is manifestly more compli- cated and difficult to maintain than a single sys- tem. Limitations of Commumcability Through Resonance Along with the advantages which pertain to a tuned or selective signaling system, there is one evident disadvantage, namely, loss of communi cability. It is all very well for a ship which is tuned to receive waves say 300 meters (328 yards) long, or at a frequency of one million cycles per second, to be able to carry on communication with another ship, or a shore station, that uses the same wave-length ; but a third station having a different tuning and producing waves, say 400 TUNED OR SELECTIVE SIGNALING 185 meters long, may desire to communicate with the ship, and not be able to do so, either from not knowing the particular wave length to which the ship responds, or from not being able to alter his tuning thereto. It is partly for this reason that the ordinary wireless equipment on board ocean- going ves^ls is not sharply tuned. It is desirable that they should be able to speak to all comers within normal short range. German Practice Under the auspices of the German government a dozen stations along the German coast are all tuned to emit waves of 365 meters (398.5 yards) in length, corresponding approximately to a fre- quency of 820,000 cycles per second. These stations have a signaling range of about 200 kilometers (125 miles) between each other, or about 120 kilometers (75 miles) from any one shore station to ships in its neighborhood. The ships are also tuned to this wave-length. Under these conditions the shore stations take prece- dence. When a shore station calls, ships within range are instructed not to speak unless called. Ships are also instructed not to call a shore sta- tion needlessly, or beyond the normal 120 kilo- meter (75 miles) range. By the observance of such regulations mutual advantage is subserved, i86 WIRELESS TELEGRAPHY as well as the keeping of ethereal peace. A wire- less-telegraph etiquette is thus gradually becom- ing evolved. Difficulties in the Way of Using Reflectors or Lenses }or Wireless Light It would clearly be of great advantage if, in- stead of radiating waves from a sending station to all the thirty-two points of the compass at once, there were some convenient means of channeling the waves into a beam, capable of being sent into any desired direction. This would not only save wasted power, but would also save needlessly stirring up the ether into noisy signals in outlying regions where other parties are trying to talk. At first sight it would seem that because we can readily accomplish this result with short-length luminous waves, as in the search-light beam for instance, we ought like- wise to be able to accomplish it with the long waves of wireless telegraphy. The trouble is, however, that optics leads to the general law that both reflectors and lenses must be large with respect to the wave-length of the waves they bend into parallelism. Thus, with half-micron waves of light, a tiny mirror, no larger in diameter than one millimeter (1-2 5th inch), would cover 2,000 wave-lengths. . If the mirror had a diameter of TUNED OR SELECTIVE SIGNALING 187 less than half a micron (-51$-$-$ inch) or less than a wave-length, it would fail to serve properly as a reflector. In the same way, when we deal with, say, 4oo-meter (437 yards) waves of invisible polarized light, it would take a metallic surface of more than 400 meters (437 yards) square to make a serviceable reflector, and such sizes are pro- hibitive. A sending station situated in front of a very steep overshadowing cliff, of fairly con- ducting surface, might have its shore side shel- tered and its sea-going waves strengthened by the reflecting surface of the cliff; but such locali- ties are not always forthcoming; nor perhaps is the surface soil sufficiently conducting to provide in most cases a satisfactory reflector. It is well known that a ship getting behind such a cliff may be sheltered on that side from arriving signals, or in other words that such cliffs may throw long electromagnetic shadows beyond them. On the other hand, a vessel lying in a harbor enclosed by hills which do not rise abruptly, but slope to the harbor, will often receive wireless messages from a direction over the hills, the waves in such cases running down over the slope. Although no marked degree of success has hitherto attended the erection of reflecting ver- ticals, or mirror surfaces, behind sending an- tennas, yet care has to be taken that a receiving 1 88 WIRELESS TELEGRAPH* antenna is suspended reasonably clear of high conductors capable of casting a shadow, Thus a receiving antenna wire suspended behind a steamer's funnel of steel, or immediately behind a steel mast, would be likely to have its signals much weakened, if not entirely absorbed. For this reason an antenna is usually suspended from a rope between two masts, in such a manner as to hang free from both. Some experimental progress toward the space direction of emitted waves has been recently an- nounced by the addition of a horizontal an- tenna on the top of, and proceeding from, a vertical antenna. This construction is equiva- lent to taking say an 8o-meter (87.5 yards) an- tenna, and instead of setting it entirely erect, placing 30 meters (32.8 yards) vertical and then carrying the remaining 50 meters (54.7 yards) out horizontally in a straight line overhead. In such a case the radiation is partly polarized in the horizontal plane, and partly in a vertical plane, the resultant being a sort of combination, or elliptical polarisation. Such waves are subject to different degrees of attenuation in different direc- tions. It is found that the radiated waves are strongest in the direction opposite to that of the horizontal offset; so that this offset in the an- tenna should be made away from the direction TUNED OR SELECTIVE SIGNALING 189 in which it is desired to send the waves. In the direction of the offset, the waves are weak and they are weakest of all in a direction nearly at right angles to the offset. In time, it is to be hoped and expected that all large vessels will carry wireless telegraph appara- tus, and that all lighthouse stations along the ocean shores will be equipped with invisible as well as visible beams. The invisible beams would be perceptible by electromagnetic appara- tus far out at sea, and also in foggy or hazy weather. This system of coast protection for arriving vessels is already commencing. Desirability of Means for Determining Wave Directions at Sea When a wirelessly equipped steamer ap- proaches the coast of Europe or of America, she is apt to be apprised of the proximity of land in thick weather by the reception of signals from coast wireless stations. It is not easy to deter- mine, however, the direction whence the signals come, or the bearing of the station. It would be advantageous to be able to orient the received signals, or to determine their direction of advance. By suspending two or more antennas far apart, near the bow and stern of the vessel respectively, and bringing them into communication through IQO WIRELESS TELEGRAPHY suitably tuned apparatus, it might be possible by swinging the ship to find the approximate bearing of the shore station. When the antennas were in line with the station, the signals would reach a maximum (or minimum), and when the line joining them was at right angles to the bearing of the station, the opposite condition should occur. Such a procedure is, however, objection- able, since it requires the ship to be stopped for observations, or at least turned erratically off her course. It seems likely that the first method to be found available will be for each shore station to erect a system of radiating wires, and to deter- mine the direction of the ship by tests conducted upon alternate wires or pairs of these wires. The shore might then inform the vessel of her bearing, which would be the next best to the ship's deter- mining the shore station bearing herself. CHAPTER Xin MEASUREMENTS OF ELECTROMAGNETIC WAVES Importance of Determining Wave-lengths WE have already seen in Chapter VI that the length of the waves thrown off by a simple verti- cal rod oscillator, discharging into a perfectly conducting level ground,% is four times the height of the rod; so that in the case of a simple un- loaded antenna of the type represented in Fig. 26, we can form a close estimate of the wave- length emitted. But antennas are usually loaded by coils, condensers, or other apparatus, so that it becomes impossible to estimate their emitted wave-length with any degree of reliability,: More- over, the wave-length, which is not of great im- portance in simple untuned signaling, becomes of great practical importance in selective signal- ing by the aid of resonance. If we can readily measure the length of the waves thrown off by an antenna, we can proceed in a rational and straightforward way to tune the system, and also to produce any required wave-length within reach. 191 IQ2 WIRELESS TELEGRAPHY Method jor Determining Wave-lengths. The wave-length of an oscillating system is measured by bringing a portable, adjustable oscillation system into electromagnetic communi cation with it, and adjusting the portable system until resonance is observed therein. At reso- nance, the wave-length of the tested system and of the portable system is the same. The wave- length of the portable system is determined, or computed, from its adjustments, and the wave- length of the tested system therefore becomes known. The simplest kind of an adjustable oscillating system consists of a circuit containing a coil of wire and a condenser. Either the coil may be adjustable in its length and virtual number of turns, or the condenser may be adjustable in the extent of its effective surface, or both may be adjustable independently. The electromagnetic length of a coil is conveniently measured and stated in centimeters or meters. The capacity of a condenser is also conveniently stated in the same units of length. When the oscillating cir- cuit is adjusted to resonance, the length of its wave is the circumferential length of that circle whose radius is the geometric mean (square root of product) of the coil meters and condenser MEASUREMENTS OF WAVES 193 meters. Thus, if the circuit was in resonance when the coil was adjusted to 1,000 meters, and the condenser to 40 meters, the geometric mean of these two lengths would be 200 meters and a circle with this radius would have a circumference of 1,257 meters (1,375 yards) which is the wave- length of the resonant circuit. Means 0} Determining Resonant Condition in Portable Circuit In order to recognize when the portable oscil- lation circuit has been adjusted to resonance, a galvanometer, ammeter, or current-indicator suit- able for high-frequency to-and-fro currents, or oscillations is placed in the circuit. Fig. 64 shows one form of such an ammeter. It consists FlG . 6 4 . Hot-Wire Am- essentially of a short meter or Current Meas- length of fine platinum wire, whose ends are secured to fixed supports. The center of the suspended wire is fastened to a thread that runs over a delicately supported pulley, carrying a pointer or index. If a rapidly oscillating electric current passes over the fine platinum wire, the wire is thereby heated and 194 WIRELESS TELEGRAPHY elongates. The elongation is detected by the sagging of the thread fastened to the center, and the degree of elongation, duly magnified, is caused to indicate the effective strength of the current. Care has to be taken that the current passing is not strong enough to overheat or melt the platinum wire. If such an instrument is inserted in a resonant oscillating circuit, the current indicated will become a maximum when resonance is produced. Thermo- galvanometer A much more sensitive galvanometer or cur- rent indicator depending also upon the heat pro- duced by current in a fine wire, is seen in Fig. 65. A loop of fine platinum wire is supported close beneath, but not touch- ing, the soldered junction of two wires of different metals, forming a circuit FIG. 65 Delicate Ther- which is delicately sus- mo-Galvanometer. . . pended in a permanent magnetic field. When current passes through the fine platinum wire, some of the heat produced is communicated to the soldered junction. This raises the temperature of the junction and pro- MEASUREMENTS OF WAVES 195 duces a steady thermo-electric force or voltage, as long as the elevation of temperature persists. The voltage sets up an electric current in the local circuit of the suspended loop, and the loop twists or tends to deflect sideways about the sus- pension, in a manner and to a degree which is greatly magnified by an attached mirror and a beam of light reflected therefrom. By means of this thermo- galvanometer very feeble oscillating currents may be measured. Some of the best measurements yet published concerning the strength of signals received at different distances from the sending station have been obtained by its use. Disk Galvanometer. Another form of high-frequency galvanometer is indicated in Fig. 66. It consists of a coil of insulated wire having comparatively few turns, and supported at the center of the instrument, connected in the circuit under test. Inside this fixed coil is delicately suspended a little silver disk s, to which is attached a small mirror m, the two being carried on a fine quartz fiber about 30 cms. (12 inches) long. The suspension is illus- trated separately in the figure. When a high- frequency current alternates in the fixed coil, a feeble alternating current of the same frequency .96 WIRELESS TELEGRAPHY is set up in the silver disk, and the electromagnetic attraction between the current in the coil and the current in the disk causes the disk to twist, or FIG. 66. Oscillating-Current Galvanometer. deflect about the axis of suspension. The deflec- tion is magnified and rendered measurable by a beam of light reflected from the mirror m. The portable oscillation circuit is brought into electromagnetic communication with the tested circuit, either by actual contact at two close points, or by means of a little induction coil of very few turns. MEASUREMENTS OF WAVES 197 Wave-measuring Helix Another type of portable resonant circuit is made in the form of a helix, or close spiral, of fine wire wound upon an insulating rod or cylinder. Such a circuit has a greatly reduced or minified wave-length. That is, a wave which would occupy a length of say 50 meters (54.7 yards) of straight wire, would perhaps occupy only 15 centimeters (6 inches) of this wire wound in a curl or helix. When the antenna to be tested is in excitation, one end of this helix is presented near to but not touching the antenna, and the length of helix which will set up resonance is ascertained by running a grounded conductor along it. Resonance is in this case detected by the formation of a glow or brush discharge, either from the end of the helix itself, or within a small vacuum- tube placed adjacent thereto. A scale marked along the rod then enables the observer to read off the wave-length directly. CHAPTER XIV INDUSTRIAL WIRELESS TELEGRAPHY The Ocean, the Kingdom of Wireless Telegraphy WIRELESS telegraphy has already come into widely extended use over the ocean. It has not come as yet into extended use over land. The reasons for this are evident : Wire telegraphy has already held undisputed sway overland. Wherever there has been developed urgent demand for the telegraph, the wire has been run to meet it. But a moving ship cannot keep up wire communica- tion with the land, except in the rare instances where a ship is employed in laying a submarine cable. Consequently, wireless telegraphy has absolutely undisputed sway over the surface of the ocean in reaching ships at a distance, or in reaching ships near by, when visual signals can- not be read, as at night, or in fog. Already, wireless telegraphy has done splendid work in maintaining communication with ships at sea. In a certain sense wireless telegraphy has re- moved the sea, because the sense of isolation in a vessel out of sight of land is almost entirely lost 198 INDUSTRIAL WIRELESS TELEGRAPHY 199 when messages are received on board through the ethereal medium of the air. In a psychological sense, distance has been destroyed. Moreover, since the sea is the conducting medium or broad conductor for guiding wireless telegrams, we may say that the sea has ceased to divide countries and now connects them. Prospects 0} Submarine Cable Telegraphy It has been much debated whether wireless telegraphy would render submarine telegraph cables useless and cable property valueless. Up to the present time it has not done so, and there is no immediate prospect of its doing so. Tak- ing, for example, the islands of Great Britain and Ireland, these are electrically connected with North America by thirteen submarine cables, and with South America by three more. Al- together there are about sixty wires connecting Great Britain with other countries. These cables are pouring messages into the islands at an average rate each of, say, fifteen words per minute, day and night continuously. It would be an enormous undertaking to replace these cables by wireless telegraphy, without any refer- ence to all the other parts of the world served by submarine cables. Up to this date, wireless telegraphy has probably aided submarine cables 200 WIRELESS TELEGRAPHY by bringing telegraph messages to and from ships at sea, for transmission by cable, more than it has injured cable telegraphy by sending mes- sages over the sea that would otherwise have gone underneath. On the other hand, however, it must be remembered that wireless telegraphy is still very young, and that it has already made far more progress in its brief lifetime than did wire telegraphy in the corresponding period of its life, sixty to seventy years ago. If wireless telegraphy continues to advance in the future at the rate it has maintained in the past, it may be that at some distant future time submarine cables will cease to be laid, and their work surrendered to their wireless rival. There has been already at least one case where wireless telegraphy has supplanted a submarine cable, and that is between the U. S. army stations of Fort St. Michaels and Safety Harbor, Cape Nome, across Norton Sound on the coast of Alaska, a distance of about 177 kilometers (no miles). The cable having been repeatedly broken by ice on this ice-bound coast, the tele- graph service has been carried on for about two years continuously by wireless telegraphy. There are also several short sea distances be- tween islands, or between islands and mainland, which have recently been covered by wireless INDUSTRIAL WIRELESS TELEGRAPHY 201 telegraph equipments instead of by submarine cable. One of these is between Port Blair on the Andaman Islands and the mainland, Burma, a distance of 491.1 kilometers (305.2 miles) under the auspices of the Indian government tele- graph department, the traffic averaging about ninety messages per month each way. On the other hand, a number of important long sub- marine cables have been laid since the introduc- tion of wireless telegraphy. Peculiar Difficulties Incident to Wireless Telegraphy. One of the difficulties that long-distance wire- less telegraphy has had to deal with, which will probably always have to be expected, is disturb- ance from thunderstorms in the vicinity of a sta- tion. Such visitations are not rare during sum- mer seasons, and usually call for a temporary suspension of traffic. A curious lesser difficulty that long-distance wireless telegraphy has to meet is the effect of sunlight upon the atmosphere. Messages can be sent and signals received much further by night than by day. The effect of sunlight on the at- mosphere is apparently to make the air foggy for the long waves of invisible wireless light. The nature of the action has not yet been clearly 202 WIRELESS TELEGRAPHY demonstrated; but it is supposed that the en- ergy of sunlight, or short-wave light, in im- pinging upon the ocean of air, either disrupts many air-atoms and ionizes them, or else in- jects streams of ionized matter from the sun, thereby leaving floating electric charges in the air. The passage of electric waves through ionized air causes work to be done in displac- ing the electrons, or electric charges, and such energy is absorbed from the stock of energy in the waves. The waves therefore become enfeebled, or absorbed, in a manner suggest- ing the action of fog upon ordinary light. The degree of enfeeblement during daylight hours is not uniform, and varies from day to day in a most fluctuating and apparently er- ratic manner. This means that in order to carry on wireless telegraph service during the worst atmospheric daytime conditions, there must be a considerable reserve of power over and above that necessary during the night time under the best conditions. It has been found that the atmospheric ab- sorption of electromagnetic wave energy oc- curs more generally and more powerfully in the tropical zone than in the north temperate zone, presumably on account of the greater INDUSTRIAL WIRELESS TELEGRAPHY 203 relative intensity of solar radiation upon the atmosphere in the tropics. It is also reported that the absorption depends in a marked de- gree upon the length of the electromagnetic waves and falls off very rapidly for lengths of wave exceeding 3 kilometers (1.86 miles) ; so that for wave-lengths of 3.75 kilometers (2.33 miles) and upwards, corresponding to frequencies of 80,000 cycles per second and under, the atmospheric absorption is compara- tively small. There is still uncertainty as to the nature of the atmospheric conditions which produce absorption; but the great and sudden changes in the strength of transatlantic sig- nals, which reveal themselves in a few min- utes of time, suggest the presence of invisi- ble masses of ionized air, cloudlike in form, which may hover in the body of the upper atmosphere, springing into existence under the influence of sunlight and disappearing when the ionizing influence of sunlight is re- moved. It is also thought that there may be a close connection between the degree of at- mospheric absorption and the amount of mag- netic variation of the compass needle, judg- ing from a comparison of records for the daily variation of transatlantic electromag- 204 WIRELESS TELEGRAPHY netic wave mean absorption, and for the reg- ularly tabulated daily variation of the mag- netic needle. The subject of atmospheric absorption of passing electromagnetic waves * is a promis- ing field for future investigation bearing upon meteorology and terrestrial magnetics. Transoceanic Wireless Telegraphy Regular transatlantic wireless telegraph transmission has been introduced between Clifden on the coast of Galway, Ireland, and Glace Bay on the shore of Cape Breton Is- land, Nova Scotia, an oversea distance of 1930 nautical miles (3570 kilometers or 2220 statute miles). Messages exchanged between these terminal antennas are transmitted by ordinary wire telegraphy to other places on each side of the Atlantic ocean. It is stated that about 5000 words a day are regularly transmitted across the ocean in this way. Extent of Use of Wireless Telegraphy on Vessels A large number of naval vessels of different governments are now equipped with wireless * See an important paper on " Wireless Telephony " by R. A. Fessenden, Proceedings of the American In- stitute of Electrical Engineers, July, 1908. INDUSTRIAL WIRELESS TELEGRAPHY 205 telegraph apparatus. Wireless telegraphy played a conspicuous part in the naval maneu- vers of the Russo-Japanese war. By its means a blockade was sustained of Port Arthur for many months by a Japanese fleet at a safe anchorage a considerable distance away. Wireless telegraphy equipment has been placed on board steamers of the following lines crossing the Atlantic ocean: The Anchor Line Co. The Cunard Steamship Co. The Norddeutscher Lloyd Co. The American Line Co. The Allan Line Co. The Atlantic Transport Co. The Compagnie Transatlantique. The Red Star Line. The White Star Line. The Hamburg- American Line. The Belgium S. S. Line. The Scandinavian American Line. The Navigazione Generale Italiana. The Austro- American Line. About 116 vessels of this transatlantic fleet are so equipped. There are two powerful wireless transmitting stations on the shores of the North Atlantic Ocean, one at Cape 206 WIRELESS TELEGRAPHY Cod, Mass., and the other at Poldhu in Corn- wall. It is becoming customary for a vessel with a long-distance equipment to maintain communication across the Atlantic with one of these stations until she establishes communi- cation with the other; so that at no time is she outside the range of communication. A wirelessly equipped vessel leaving New York is in communication with the station at Sea Gate, N. Y. This is carried until it is exchanged for communication with Babylon, L. I. This is carried until Sagaponack, L. I. Next Nantucket, Mass., is taken, then Sable Island, and finally Cape Race. Communica- tion for the exchange of messages is thus maintained for about seventy hours after leav- ing port, each of these stations being in per- manent wire communication with the rest of the world. It is stated that in the year prior to Jan. 31, 1906, these American stations sent and re- ceived, with ships, altogether 15,000 messages comprising over 200,000 words. Number of Land Wireless Stations According to a report of the U. S. Navy Department there are this year (1908) about 468 land wirdess stations in different parts of INDUSTRIAL WIRELESS TELEGRAPHY 207 the world, either erected or projected, namely : Belgium 2 Argentina . . . . 5 Hong-Kong .. I Denmark .. . . . n Brazil .11 China o Germany .71 Hawaii 7 France T8 Chili . 7 Japan Great Britain. . 46 Colombia T Dutch E. Ind.. 5 Holland 6 Costa Rica. . . . 2 Asiatic Russia. i Spain TO Mexico 6 Egypt . /i Portugal T Panama . 2 r^-^ Morocco c Gibraltar I United States. III Mozambique .. 2 Italy 22 Trinidad . . . . . I Tripoli T Malta I Porto Rico. . . . 2 Canaries I Montenegro . . I San Domingo. . 2 Ecuador 2 Norway q Tongking . . . . . 2 Formosa i Austria-Hung . 7 Uruguay . 2 T Finland 2 Zanzibar 2 Korea Switzerland .. . I Australia . 5 Nicaragua 2 Roumania 6 Cuba 10 Peru c; Russia 15 Tobago . T Philippines . . . 4 Sweden 4" li India . . 3, Qi M -p-l ^ Turkey 5 Burma . T According to the same navy department list, there are at this date some 340 mercantile vessels equipped for wireless telegraphy, in- cluding the Atlantic liners already referred to, and carrying flags of the following countries: Belgium 10 Germany 38 France 8 Great Britain 86 Holland 10 Italy 10 United States 141 Canada 27 Japan 10 The total number of recorded ship and shore stations is thus about 800, exclusive of many warships of various nations, 208 WIRELESS TELEGRAPHY Each station possesses a definite " code- name " or " call-letter," which is usually a group of two, or even three, letters, such as BA (Babylon, L.I). Stations are called by their code names and they sign their messages with them. This is even more necessary in wireless than in wire telegraphy; because the distant station may be far beyond visible range, and be otherwise unidentified. It is customary for a steamer to pick up communication with another steamer at a dis- tance of say 150 kilometers (79.8 nautical miles, or 93 statute miles) and to carry on communication until the distance between the ships is, say, 250 kilometers (133 nautical miles, or 155 statute miles). Occasionally, however, messages are exchanged between a ship and shore at much greater distances than these. CHAPTER XV CONSIDERATIONS PRELIMINARY TO WIRELESS TELEPHONY WIRELESS telephony depends upon the same principles as wireless telegraphy; but differs therefrom in details connected with the nature and requirements of the electric telephone transmitter and receiver. It is necessary, therefore, to have a sufficiently clear under- standing of the nature and mode of operation of electric telephony with wires, in order to follow understandingly the modified nature and mode of operation of electric telephony without wires. The word telephony is derived from two Greek words signifying the far-off transmis- sion of sound. Nature of Sound All material substances are capable of being compressed and dilated; i.e., of being altered in density, by the application of suitable forces 209 210 WIRELESS TELEGRAPHY to them; although the facility for being thus compressed and dilated varies enormously in different kinds of matter. For instance, gases, such as the air, readily admit of being com- pressed or dilated, as in the manipulation of a concertina; while, on the other hand, many liquids, such as water, require so much more force to compress or dilate them that, until the year 1762, it was supposed that water was incompressible. When a body is subjected to a rapid rhyth- mic variation in its density, it is said to vibrate, as when the metal in a bell is forced into rapid small alternations of compression and dilation by a hammer blow; or when the framework of a building is set into slight vi- bration by the passing by of a rapidly moving railway train. Such rapid vibrations, commu- nicating themselves to the ear, usually by way of the air and the external ear orifice, excite in our consciousness the sensation of sound. From this standpoint, sound is a par- ticular mode of sensation excited by vibratory disturbances of density in neighboring ma- terial substances. But it is also customary to call a vibratory disturbance sound, when this disturbance is capable of exciting the sensation of sound. Consequently, sound may mean ELEMENTS OF WIRELESS TELEPHONY 211 either the particular mode of physiological sensation received through the ear, or the phys- ical disturbances in a medium, say air, such as might give rise to the sensation. Difference Between Musical Sound and Noise When the vibratory disturbance, or sound, in a material medium is non-rhythmic and ir- regular in repetition, the sensation produced is called non-musical sound, or noise, as when coal is emptied from a cart into a cellar. The impacts of the many falling lumps of coal with the cellar floor, or with each other, set the coal and the surrounding air into powerful in- coordinate vibration of a jangling character. When, however, the vibratory disturbance is rhythmically repeated, the sound sensation produced is more or less musical. If a gong is struck, its metallic mass is thrown into rapid vibrations which may be readily felt by the hand, and which are rhythmic in charac- ter. These rapid and rhythmic vibrations are communicated to the air surrounding the gong, and produce alternate compressions and dilatations in the air. Such local disturbances in the density of the air do not remain fixed in the space where they are produced; but move off. in all directions at a definite speed. 212 WIRELESS TELEGRAPHY When these moving air pulses impinge upon the eardrum of a listener, they cause the eardrum to vibrate, and communicate to the listener the sensation of a musical sound. The Nature of Plane Waves of Sound in a Speaking Tube A few examples of the spherical expansion of sound-waves have been considered in Chapter II. We may here examine the par- ticular case of sound transmission within and along a straight pipe, such as a speaking-tube. In Fig. 67, A A', B B', represents a short length of a speaking-tube containing air, through which a simple single musical note is being transmitted acoustically. The actual process is quite invisible; but we may repre- sent the density and pressure of the air in dif- ferent parts of the tube by the relative prox- imity of transverse lines. Where the density and pressure are above normal, as at C and C', the lines are heavy and closely crowded. Where, on the contrary, the density and pres- sure of the air are below the normal, the lines are dotted and are also separated out. At the particular instant selected, as represented in the figure, the air at the points C C' is in the state of maximum compression; while at ELEMENTS OF WIRELESS TELEPHONY 213 D D' the air is in the state of maximum dila- tation. Between NV and N 2 the air is mo- mentarily compressed; while between N 2 and N 3 it is dilated or rarified, to a correspond- ing degree; but with ordinary intensities of sound, the differences of pressure between the compressed and dilated portions of the air are remarkably small. It is difficult to present clearly to the eye the variations of density and crowding in the air by means of the crowding together of transverse lines as in the upper part of Fig. 67. A much more convenient diagram for this purpose is given in the lower part of the figure; where the straight line OO' repre- sents normal or undisturbed air-density and pressure ; while deviations above this line stand for compressions, and deviations below the line for dilatations or rarefactions. Thus, the wavy line d % c n 2 d' n 3 c' indicates the condition of pressure and density of all the air in the length of pipe A B, at the instant selected. The sound waves Nj. C N 2 D N 3 C do not stand still but move along the tube, from the speaker to the listener, at a steady speed, which in free air is about 335 meters per second (noo feet per second); but which in a nar- 214 WIRELESS TELEGRAPHY row tube may be somewhat less, owing to friction with the walls, say 315 meters per second (1032 feet per second). The air in the tube does not move bodily. When a speaker blows through a speaking tube, the air moves bodily along it, and may actuate a whistle when escaping through the distant end. But when he speaks into the tube, the FIG. 67. Diagram Indicating the Comparative^Densities of the Air in a Speaking Tube at a Particular In- stant When Transmitting a Single Pure Musical Tone. sound of his voice may be carried through the air without any bodily motion of that air. Each air particle joins in the vibratory mo- tion, and oscillates slightly to and fro in the direction of the tube's length, about its mean position of equilibrium. In other words, when sound is transmitted, it is the dis- turbance in density which moves bodily along in waves and not the parts of the substance in ELEMENTS OF WIRELESS TELEPHONY 215 which the disturbance exists. The individual parts merely vibrate, being alternately closer .together, and further apart, than in the quies- cent state. Intensity or Loudness of Musical Tones Musical tones, when steadily maintained one at a time, differ from each other in intensity, and in pitch. The intensity or loudness of the sound sensation produced by a simple musical tone depends upon the amplitude of the vibration producing that sensation ; that is, upon the maximum excursion of the air par- ticles in their vibration from their mean po- sition of equilibrium. In Fig. 67 the ampli- tude would be measured by the distance o c. The loudness of a sound sensation increases with the amplitude of vibration; although not in simple proportion. The sound of a steamer's whistle is often piercingly loud to a listener standing on the deck immediately in front of it; but becomes fainter with distance. This means that the amplitude of vibration of the particles of air is relatively great near the whistle; but becomes smaller as the dis- tance from the whistle increases, and as the area of the sound-wave surface expands. The human ear is so sensitive to some sounds, 216 WIRELESS TELEGRAPHY that, according to accepted measurements, the sound of a whistle has been detected in air when the amplitude of disturbance at the listener's ear can only have been about i mil- limicron (i m/x or ^_ inch.) Pitch of Single Musical Tones The pitch of a single musical tone depends only on the number of complete vibrations, or cycles, of disturbance per second. A note of low pitch, like that of a deep bass voice, pos- sesses relatively few vibrations per second; while a note of a high pitch, like that of a soprano voice in its upper register, possesses relatively many vibrations per second. The number of complete to-and-fro vibra- tions, or cycles, of vibration per second, exe- cuted in a single musical tone is called its fre- quency, as in the case of electric vibrations referred to on page 63. A note of high pitch is, therefore, a high-frequency note, and a note of low pitch a low-frequency note. The hu- man ear is able to hear sounds whose frequen- cies lie between about 16 cycles per second in the bass and about 16,000 cycles per second in the high treble; or over a range of some ten octaves, the limits of pitch audibility vary- ing to some extent with different individuals. ELEMENTS OF WIRELESS TELEPHONY 217 The usual pianoforte keyboard is from A t of 27 cycles per second, in the bass, to c 5 of 4100 cycles per second, in the treble, or about 7^ octaves. The ordinary range of pitch in the singing voice is somewhat less than 2 oc- taves, a man's baritone compass being com- monly from A of 108 to f of 316 cycles per second, and a woman's soprano compass from c' of 256 to a" of 854 cycles per second. The fundamental tone of a man's speaking voice is usually in the neighborhood of 150 cycles per second with a wave length in air of about 2 meters (6.56 ft.), and that of a woman's voice near 300 cycles per second with a wave length of about i meter (3.28 ft.). Purity of Musical Tone Contrary to what might be supposed at first thought, a pure musical tone in the sense of a single simple musical tone, cannot be produced by the human voice, is very diffi- cult to produce artificially, and w r hen produced, is not particularly pleasing to the ear. A musical tone produced by a trained voice is found to be not a single simple musical tone of the desired pitch ; but a harmonious associa- tion of feebler higher pitch tones with the tone of desired pitch. The wave form of a sim- 218 WIRELESS TELEGRAPHY pie musical tone is indicated in the line d c d' c' of Fig. 67. A close approximation to such a tone may be produced by mounting a tuning fork on a suitably shaped hollow cham- ber, or resonator. A flute may also be made to produce a fair approximation to a single FIG. 68. Composition of a Simple Musical Tone with an Overtone of Eight Times Its Frequency and One-Fifth of Its Amplitude Into a Resulting Com- posite Musical Tone. musical tone. Ordinarily, what we describe as a single musical tone, is the association of that tone with a number of fainter tones of higher pitch; so that the wave shape is ren- dered complex. To take a simple example, ELEMENTS OF WIRELESS TELEPHONY 219 Fig. 68 shows at o A B c D the wave-form of a certain pure musical note, say middle c' of the piano, with an amplitude Q B. Above this appears the wave-form of another pure musical note a b c d of eight times the fre- quency, and corresponding therefore to the triple octave, or c"" above the treble clef, with an amplitude q b, one-fifth of Q B. If both these pure musical notes are sounded to- gether, the resulting wave-form is shown at A' B 7 C' D'. A trained ear listening to the composite note might detect both the funda- mental tone of O A B C D and the fainter overtone or harmonic o a b c d, which would blend together harmoniously. Even an un- trained ear might detect that the quality of the composite tone A' B' C' D' was different from that of the simple tone A B C D. The wave shape of a composite musical tone may be altered in three ways-: (1) By changing the number of associated overtones. (2) By changing the relative amplitudes of associated overtones. (3) By changing the relative positions, or " phases," of the overtones. The quality of the tone as appreciated by the ear will be affected by changes (i) and 220 WIRELESS TELEGRAPHY (2), but not by (3). In regard to change (3), it may be observed that in Fig. 68, the overtone has the negative, or downward, amplitude a at the moment when the funda- mental tone has the positive, or upward, am- plitude A; so that the composite tone wave is diminished in amplitude at A' and C'; but increased at B' and D/ If, however, the ripple train abed were advanced though half its wave-length, or changed in phase by half a wave, with respect to the fundamental wave O A B C D, the composite tone would have the same quality to the ear, but its wave form would have increased amplitude at A' and C', with diminished amplitude at B' and D'. It usually happens that a source of musical tones, such as a horn, trumpet or harp, pro- duces, along with each fundamental tone, an association of a number of fainter overtones, whose frequencies are usually all simple mul- tiples of the fundamental frequency. The number and relative prominence of these overtones give the distinguishing quality of the note produced by each instrument. Thus, a flute sounding middle c', produces relatively few and feeble overtones. The fundamental tone is heard almost pure. On the other ELEMENTS OF WIRELESS TELEPHONY 221 hand, the same note sounded on a violin would be accompanied by a large number of overtones, of successive frequencies 2, 3, 4, etc., times that of the fundamental note. As a general rule, the higher the frequency of an overtone, the smaller its amplitude; so that beyond a certain frequency the overtones tend to become inappreciable. Occasionally, how- ever, particular overtones, such as the 7th or Qth-frequency overtones, may be more prominent than their neighbors. The shape and physical conditions of the violin sound- ing-board tend to accentuate some overtones more than others. The reason, therefore, that the note, say " middle c'," sounds quite differently when sung by a voice, piano, or violin, lies mainly in the differences of associa- tions of overtones, and in the corresponding wave shapes of the composite tones. Moreover, the same sustained musical note sung by a trained singer, and by an untrained singer, may be very different, in spite of the fact that each may be producing essentially the same fundamental tone. In the untrained voice, there is likely to be a wavering, or .un- steadiness, of pitch, or of amplitude, or of both, due to imperfect muscular control. There may also be an unmusical roughness, 222 WIRELESS TELEGRAPHY or noise, included in the tone, owing to the imperfect interaction of the vocal chords in the larynx. There is also likely to be an un- pleasing association of overtones, both in re- gard to their relative amplitudes, and to their number; while the particular association of overtones may be varying, or wavering, from moment to moment in an unpleasant manner. In the note of the trained singer, we are likely to find, on the contrary, a sustained steadiness, either in uniformity, or in graded change, of loudness, an absence of roughness, or extra- neous unmusical sound (noise), and also a pleasing association of evertones, brought about by the habitual formation of the vocal cavities so as to resonate with, and reenforce, harmonious components. Similarly, the dif- ference in the quality of the same notes pro- duced by a player successively, and with the same skill, on different violins or pianos, de- pends mainly upon their respective sounding- boards, and the resonating influence of these on the overtones. Some particular blendings of overtones in regard to number, or relative amplitudes, are more pleasing to the ear than others. The skill of the instrument maker is shown in the resonating qualities he is able ELEMENTS OF WIRELESS TELEPHONY 223 to bestow upon the instrument when it leaves his hands. Tones in the Speaking Voice We have already seen that the wave-forms of the sounds in the singing voice are com- plex in character, owing to the large number of different single tones that are ordinarily contained therein; but the sounds of speech are still more complex. In the sounds of speech we find vowel-sounds and consonant- sounds, as well as inflections and cadences of tone. The vowel-sounds are of a quasi-musi- cal character, and the musical quality of a speaker's voice depends in large measure upon them. The inflections and cadences of speech are mainly variations in the fundamental tones of the vowel-sounds. The consonant-sounds are of different kinds, such as labial, dental, guttural sounds; but are mainly quick, sudden and explosive. The more prolonged vowel- sounds connect and are terminated by the more sudden consonant-sounds. The defi- niteness and intelligibility of speech resides principally in the consonant-sounds. Speech, deprived of its consonants, becomes a mere droning, or caricature of song. 224 WIRELESS TELEGRAPHY Some of the consonant-sounds are feebler, or have smaller amplitude, than vowel-sounds. This is particularly the case with sibilants, such as s, z, ss t etc. One of the most difficult words for a phonograph, gramophone, or tele- phone to repeat is "specie" Owing to the large range of frequency in consonant-sounds, and their frequent lack of amplitude, it is more difficult to reproduce articulate speech than vowel-sounds, or music. It may be possible for a phonograph, or tele- phone, to reproduce recognizable musical tones, when the reproduction of recognizable speech is impossible. CHAPTER XVI THE PRINCIPLES OF WIRE TELEPHONY IN the ordinary process of electric teleph- ony by means of wires, the speaker talks in front of a " transmitter " such as that shown at T in Fig. 69. The es- sential elements forming this transmitter are indicated in Fig. 70. M is a hard rubber mouth- piece, usually a FIG. 69. Ordinary Desk Set of a y Telephone Receiver and Trans- perforated grid mitter as Used in Wire Teleph- at the base, to ony * prevent a pencil, knife or other pointed in- strument from being pushed in, to the detriment of the delicately adjusted parts 225 226 WIRELESS TELEGRAPHY beyond. An aluminum circular diaphram D is supported around its edge, and held in a soft rubber groove or gasket. At the center C, the diaphram is hollowed out to form a circular chamber. In this chamber are placed two carbon disks F and R, separated by granules of hard carbon. The front disk F is carried by the diaphram D. The rear disk R is fastened rigidly to a pin at the center of the solid metal back B. A thin mica disk A is clamped between the diaphram and the rear disk, so as to close the chamber flexibly and _ ^^ maintain a mois- 1 iPi ibtate* ture-tight seal. The front and rear disks are connected b y wires to the termi- nals of the trans- mitter. When the speak- er's voice is directed towards the trans- FIG. 70.- Sectional View of mitter > the sound Principal Parts of a Tele- waves in the air en- phone Transmitter. ter the funnel or mouthpiece M, and impinge upon the aluminum diaphram D, which is set into vibration corres- ponding to the vocal vibration. The diaphram ELEMENTS OF WIRELESS TELEPHONY 227 flexes and buckles to and fro very rapidly, as indicated diagrammatically by the dotted white lines in the figure. Since the solid metal back plate B is practically rigid, the rear carbon disk R stands fixed, and resists the vibratory force. Consequently, the particles of hard carbon C lying between the vibrating front carbon disk F and the stationary rear carbon disk R, are subjected to alternating compression and relaxation of pressure. These vibratory changes in pressure accompany the vibrations of the diaphram D, which as we have seen, follow the air vibrations of the waves of sound arriving from the speaker's lips. The pow- dered carbon F has the peculiar and valuable property that, when lying loose and uncom- pressed, it offers considerable resistance or ob- struction to the passage of an electric current; whereas, when compressed and compacted, this obstruction is in considerable part re- moved. Consequently, a voltaic battery, con- nected to the transmitter, is able to send more current through the powdered carbon F each time that the diaphram D is moved inwards to compress the carbon, but is compelled to send less current each time that the diaphram D is moved outwards to release pressure on the carbon. Each vibratory motion of the 228 WIRELESS TELEGRAPHY disk thus produces a corresponding vibratory impulse of electric current in the wires carrying the current from the battery to the transmitter. It is as though the vibrating disk governed a little throttle-valve, by which electricity was alternately admitted to and cut off from the circuit comprising the battery, the wires, the transmitter, and any other in- struments included therewith. It is easily understood that the successive vibratory movements of the diaphram are im- mediately followed by similar successive elec- tric current impulses along the wires con- nected to the transmitter, owing to the cor- respondingly varying electric resistance of the carbon particles F. The electric current impulses move along the conducting wires as invisible electromagnetic waves at very great speed. If the diaphram D behaved perfectly, it would faithfully repeat in its vibratory movements each and all of the vibratory movements of the impinging air particles. In other words, a perfectly acting diaphram D would follow all of the faintest ripples on the back of the most complex wave-forms per- taining to the incident vocal sounds. As- suming such perfect behavior on the part of the diaphram, the wave-forms of the vibra- PRINCIPLES OF WIRE TELEPHONY 229 tory pressure communicated to the powdered carbon c would be the exact counterparts of the wave-forms of the vocal sounds uttered by the speaker. The effect of the corre- sponding changes in electric resistance in the carbon would be to produce electric currents whose wave-forms would all correspond with those of the vocal sound-waves. As a mat- ter of fact, however, the diaphram D is never perfect in its behavior. It tends to develop favorite vibrations of its own, con- sidered as a flat bell or gong, and it distorts more or less in its actual vibrations, the wave- forms of the air vibrations impinging on its surface. Nevertheless, the vibrations of the diaphram D follow those of the incident sound- waves sufficiently nearly for practical tele- phonic purposes, and the electric current waves, which closely correspond to the dia- phram's vibrations, represent the vocal sound- waves fairly well. By means of suitable delicate electromag- netic mechanism, the electric current waves in a wire telephone circuit can be made to photograph themselves, if care is taken to make them relatively powerful. With this object in view the circuit must be compara- tively short : that is, it must not include many 230 WIRELESS TELEGRAPHY miles of wire, the instruments must be ad- justed as delicately as possible, and the speaker must place his lips close to the mouthpiece of the transmitter and speak in a full clear tone. Many persons fail to make themselves clearly heard in ordinary wire telephonic con- versation, because they talk into the circum- ambient air, instead of talking into the trans- mitter. A low tone of voice, with the lips nearly touching and fully opposite to the transmitter mouthpiece, is likely to be more effective in making the distant listener under- stand, especially in long-distance telephony, than loud shouting with the face directed away from, or to one side of, the transmitter. Fig. 71 presents three " oscillograms " or photographs of the electric current wave- forms in the articulation of the three syllables cur, pea, and tea* Except for the vibratory imperfections of the transmitter diaphram, above referred to, these electric wave pictures may be regarded as portraits of the sound- waves in the voice of the speaker that uttered those syllables. Beginning at A on the top line, the interval A B represents a small frac- * From a paper on " Telephonic Transmission Measurements" by B. S. Cohen and G. M. Shepherd, Proceedings of the Institute of Electrical Engineers, London, May, 1907. PRINCIPLES OF WIRE TELEPHONY 231 tion of one second of time, during which the speaker uttered the syllable cur. First comes the c consonant or sound of k, as a train of about twenty small high-frequency waves of very complex form. Then there is a brief pause, during which the muscular adjustments appear to be made for the following vowel- sound ur, and finally we have about eight p as in pea P FIG. 71. Photographs of Electric Current Waves in the Transmission of Three Particular Vowel-Sounds. complete fundamental waves of the vowel- sound, judging by the recurring sharp peaks below the line, with numerous associated overtones that distort the fundamental wave almost beyond recognition. We can imagine that if the outline A B were accurately cut into the surface of a wax phonograph cylinder, the passage of the reproducing stylus over the 232 WIRELESS TELEGRAPHY indented surface might cause the instrument to repeat this syllable cur. Similarly with the syllable pea, as recorded along the line c D. First comes the con- sonant sound, then a brief pause, and then about eight waves of the fundamental vowel sound with a clearly visible prominent over- tone ripple of perhaps four times the funda- mental frequency. Again, at E F, in the syl- lable tea, there is first the brief explosive con- sonant, then a pause containing apparently a feeble high tone, or a group of high-frequency tones, and finally the vowel-sound which somewhat resembles the vowel-sound at C D. Changes of Wave-form in Telephonic Trans- missions Over Long Wires. When electromagnetic waves are delivered to a pair of conducting wires, in ordinary wire telephony, by the action of the speaker's voice on his transmitter, two changes occur in these waves as they are carried over the wires to the listener at the receiving end: namely (i) a diminution in the amplitude, or strength, of the waves, and (2) a different diminution in waves of different frequency. The first change is a mere weakening, like that of sounds heard at great distances in air. It is PRINCIPLES OF WIRE TELEPHONY 233 called attenuation. The second change means that the different frequency components in composite sounds are attenuated differently, so that the shape of the current waves arriving at the receiving end of the line is different from that of the outgoing waves at the sending end. In general, tones suffer more attenua- tion the higher their frequency. That is, the a FIG. 72. Oscillograms of Singing Voice at Sending and Receiving Ends of a Moderately Long Tele- phone Line. 4 FIG. 73. Oscillograms of Singing Voice at Sending and Receiving Ends of a Considerable Length of Telephone Line. fundamental tones are not weakened so much as the over-tones. The result is that the char- acter of the transmitted sound is altered dur- ing the electric part of the transmission. The relative influences of attenuation and distortion in wire telephony are fairly well presented in Figs. 72 and 73, .which are taken 234 WIRELESS TELEGRAPHY from the same paper as the last illustration. In Fig. 72, A B is the oscillogram of the elec- tric current waves at the sending end of a telephone line, produced by a fairly high note sung into the transmitter by a girl's voice. The corresponding oscillogram a b beneath, shows the electric current waves at the re- ceiving end of the line. The line was not of great length from a telephonic standpoint. In the oscillogram A B, there are 19 funda- mental waves, judging by the lower peaks. These correspond to the frequency of the note sung. There is also a prominent ripple of three times the fundamental frequency, and there are, besides, other overtones discernible of yet higher frequency. In the oscillogram a b from the receiving end, there are the same number of waves, but the amplitude is re- duced. That is, there is evidence of consider- able attenuation. Moreover, there is evidence of some distortion, because the outlines of the received waves are not merely smaller than at A B, but they are also smoother and rounder, indicating that the ripples have been attenuated more than the fundamental. This is more clearly shown in Fig. 73, where C D and c d are the oscillograms at the sending and receiving ends of a fairly long telephone PRINCIPLES OF WIRE TELEPHONY 235 circuit when the syllable oo was sung into the transmitter. Here 14 waves of fundamental frequency may be detected. The waves re- ceived at c d are not merely attenuated. If only attenuated, they would retain the exact shape of the waves of the sending end, on a smaller scale of amplitude. They have also been distorted. The sharp overtones and peaks in c D are absent in c d. The received waves have more of the fundamental and less of the overtones in their composition. They approach more nearly to the type of simple fundamental wave appearing in Fig. 67. To a listener on such a telephone circuit, the voice of the speaker might be intelligible; but would probably sound quite differently. It would be altered in character and would prob- ably sound " drummy." This is a well known condition pertaining to wire-telephony over circuits that are electrically very long and distorting. On arriving at the receiving end of the tele- phone circuit, the invisible waves of electric current are enabled to reproduce corre- sponding sound-waves by passing through the coils of fine insulated copper wire in a tele- phone receiver. One form of receiver suit- able for wearing on the head, has already been 236 WIRELESS TELEGRAPHY described in connection with Figs. 50 and 51. A particular form of hand receiver is seen partly disassembled in Fig. 74. The hard rubber shell SS' receives the connecting wires at the narrow end, and clamps the thin ferro- type disk 'or diaphram D between its broad end S', and the hard rubber screw cover R. Inside the shell h held the magnetic system, FIG. 74. Internal Parts of a Tele- phone Receiver. consisting of a pair of hard steel permanently magnetized bars a b, c d connected at a c by an iron yoke-piece and terminating at b d in a circular bridge-piece g g of non-magnetic metal. On the poles b d are mounted soft iron strips forming the cores of two small electro- magnet coils, which are wound with many turns of fine silk-covered copper wire. When PRINCIPLES OF WIRE TELEPHONY 237 assembled, the diaphram D is clamped close to, but out of contact with, the soft iron pole- pieces of the electromagnets. These are kept magnetized under the influence of the permanent bar magnets a b, c d; so that the diaphram D is steadily attracted or pulled magnetically towards the soft iron poles when no current passes through the instrument. If now a current passes through the electromag- net coils in one direction, the magnetic pull of the permanent magnet is strengthened, thus tending to bend down or buckle the diaphram D, near its center. If, however, a current passes through the coils in the opposite direc- tion, the magnetic pull of the permanent mag- net is weakened, and the elasticity of the dia- phram D tends to flatten the diaphram or diminish its bending down at the center. Each wave, or superposed ripple, of electric current sets up a corresponding up and down movement of the center of the diaphram, the edge of which is held fixed between the ring on S' and an opposing ring on the cover R. Rapidly succeeding current waves thus throw the diaphram into vibrations, which, if the system were perfect, would be identical with those of the transmitter diaphram at the sending end of the circuit, The electromag- 238 WIRELESS TELEGRAPHY netic vibration of the receiver diaphram D sets in vibration the air over the diaphram, and when the cover R is pressed against the listener's ear, the sound-waves are led through the air directly from the diaphram D to the eardrum. Unless the electric current waves are much stronger than are ordinarily em- ployed in telephony, the amplitude of vibra- tion of the diaphram D is so small as to be imperceptible except with the aid of very deli- cate instruments. It is generally less than i micron ( ^~ inch). Nevertheless, within this small range, the vibratory movement of the receiver diaphram corresponds to that of the transmitter diaphram under the influence of the speaker's voice, after allowance has been made for the electrical and mechanical imperfections of the system. CHAPTER XVII PRINCIPLES OF WIRELESS TELEPHONY IN order that long-distance wireless teleph- ony may be possible, by means of electro- magnetic waves conducted over the earth's surface, it is necessary that an antenna at the sending station should radiate waves that are definitely related to the sound-waves emitted from the speaker's lips, and that an antenna at the receiving station should pick up these waves and utilize them in such a manner as to reproduce these sound-waves. An ideally simple arrangement would be that the transmitter, actuated by the speaker's voice, should generate alternating electric im- pulses supplied directly to an antenna, that the antenna should radiate the energy of these impulses in electromagnetic waves, the wave- forms of which would be identical with those of the actuating vocal sound-waves, and that the receiving mast-wire should be connected to ground through a receiving telephone, and operate the same by the electric current im- 239 240 WIRELESS TELEGRAPHY pulses produced by the passage of the waves. Such an arrangement is, however, impracti- cable, because the power of the human voice is insufficient to- generate electromagnetic waves capable of producing audible sounds at any considerable distance. Moreover, the fre- quencies that are serviceable in transmitting speech are relatively low, not necessarily ex- ceeding 2000 cycles per second, and seldom exceeding 5000 cycles per second. An an- tenna does not radiate electromagnetic waves to any considerable extent until the frequency is raised to at least tens of thousands of cycles per second. Such frequencies extend beyond the limits of audibility. The general plan that is adopted is to supply electric power to the sending antenna under such conditions as will permit of sustained radiation of electromagnetic waves. This power supply is modified in some manner by a transmitter, under the action of the speaker's voice. The electromagnetic waves radiate out, carrying with them the vocally imposed modifications, and the distant re- ceiving antenna, in the path of these waves, is able to make their modifications audible as articulate sounds in the connected receiving telephone. PRINCIPLES OF WIRELESS TELEPHONY 241 Methods of Maintaining Continuous Radiation It has already been pointed out in Chapter IX that when a sending antenna is supplied with power from an induction coil, operated through a vibrator, the radiation of electro- magnetic waves from the antenna is likely to be markedly discontinuous. For instance, if the vibrator delivers 200 electric impulses per second to the antenna, the latter may radiate a brief train of waves at each 2OOth part of a second, with relatively long intervening gaps of quiescence. It is manifest that such a type of radiation is ill adapted for wireless te- lephony, because during the utterance of any one syllable at the transmitter, there will be only a few groups of waves, emitted from the antenna in jets, with relatively long inter- vening pauses. In order to transmit articu- late speech, the radiation from the antenna must be continuously sustained; or, if discon- tinuous, the discontinuities must be relatively brief. Two methods have recently been developed for continuously sustaining the radiation from a sending antenna. The first method employs the electric arc. The second method employs a high-frequency, alternating-current, dynamo machine. 242 WIRELESS TELEGRAPHY The Singing-Arc Method of Setting Up Sustained Oscillation. One variety of the arc-lamp method is rep- resented in its simplest elements by Fig. 75. A dynamo D supplies direct current to an arc lamp A, through suitably adjusted electric resistances R R', and " choking coils " C C'. The main function of the resistances is to steady and control the current supplied to the lamp; while the choking coils prevent rapid cur- rent oscillations from traversing the branch C R D R' C' of the system. Connected in parallel with the arc lamp is a branch cir- R D A --; R' C 1 FIG. 75 Arrangement for Maintaining Continuous Os- cillation of an Antenna With the Aid of a Voltaic Arc. PRINCIPLES OF WIRELESS TELEPHONY 243 cuit c P c', containing condensers c c' and a coil P, which also forms the primary winding of an induction coil, having its secondary S in the sending antenna. The condenser capacity and self-induction of the branch c P c' are such as to favor the production of suitable high-frequency oscillations. The mast wire M is also tuned to the same frequency, with the aid of the adjustable coil L, as described on page 1 20, in connection with Fig. 37. The arc lamp A serves to excite the oscillating-current branch c P c' into sustained oscillatory action, in a manner about to be described, and these oscillations are imparted to the synchronously tuned mast wire M, through the induction coil; so that the mast wire is kept in con- tinuous electric oscillation, and therefore, in steadily sustained radiation of electromagnetic waves. The energy carried off by these waves is supplied by the dynamo D and by its prime mover, say a gas-engine or steam- engine. The action of the arc lamp by which it ex- cites oscillatory currents in the electrically tuned branches c PC'- M L S G, is somewhat complex in detail; but, in outline, is simple enough. The solid cylinders, which support the arc between their tips, offer comparatively 244 WIRELESS TELEGRAPHY little resistance to the passage of electric cur- rent; but the vividly incandescent column of metallic vapor, constituting the arc, offers a considerable resistance, which depends in mag- nitude upon the strength of the current in the arc. If the current is feeble, the arc is a thin band of incandescent vapor, and offers a relatively high resistance. If the current through the arc is increased, the arc itself swells and broadens, while, at the same time, its resistance is lowered. In other words, a stout arc, carrying a strong current, conducts electrically better than a thin arc, carrying a weak current. If the arc lamp A is started with the oscilla- tory current branch c P of removed or inter- rupted, a steady current will flow from the dynamo D through the arc lamp and the coils R C, R' C'. The arc will burn with a fairly steady flame, as in an ordinary street arc lamp. There will be no tendency to set up high-frequency alternating currents in the sys- tem. When, however, the oscillatory branch c P c' is applied to the arc, the condensers in this branch take a sudden charge, or brief cur- rent impulse, which is deflected from the arc, because the choking coils C C' do not permit a sudden change of current to occur in the PRINCIPLES OF WIRELESS TELEPHONY 245 supply circuit. The sudden diminution of current in the arc instantly causes the arc to shrink and rise in resistance; thus tending to throttle the current in the supply circuit D R C A C' R'. But the choking coils resist this sudden throttling of the current, and, -in their endeavor to keep the current steady, they force electricity into the condensers, where it can go for the moment, after the conductive path through the arc is obstructed. The con- densers thus become overcharged. Their electric elasticity speedily arrests the action, and forces electricity back through the arc, since the choking coils C C' resist all sudden changes. The current now builds up in the arc, and, as it does so, the arc column swells and conducts better. The arc resistance being thus reduced, the condensers over-discharge, being aided in this by the electromagnetic inertia of the induction coil primary winding P ; while, at the same time, a strong oscillatory impulse is delivered from ground G to the mast wire M. The brief over-discharge of the condensers soon terminates, the current in the arc falls to the normal steady value, its resistance rises, and current is thereby de- flected again into the condensers so as to charge them, at the same time inducing a 246 WIRELESS TELEGRAPHY reverse impulse in the mast wire in synchro- nism with its natural period of swing. In this manner a substantially steady flow of current is delivered by the dynamo D, but alternately in throbs or impulses to the arc A and to the oscillatory branch c P c', the frequency of these impulses being determined by the natural period of the latter branch. The amplitude of the current oscillations that can be imparted in this way to the oscil- latory current branch, and to the mast wire, depends, other things being equal, upon the change in resistance of the arc A with change of current. If the arc changes greatly in resistance for a given change in current, the action above described will be powerful ; while, if the arc changes but little in resistance, the action will manifestly be but weak. What is needed, therefore, is an arc that is very sensi- tive in its resistance to changes of current. This sensitiveness is found to depend partly on the condition of the solid cylinders sup- porting the arc between their tips, and partly on the condition of the gas in which the arc is formed. PRINCIPLES OF WIRELESS TELEPHONY 247 Means Resorted to for Increasing the Sensi- tiveness of the Singing Arc It has been found that the sensitiveness of the arc can be increased by substituting for the upper carbon rod a water-cooled metallic cylinder, and also by substituting for atmos- pheric air some other gaseous medium in which the arc is allowed to burn. Both hydrogen and illuminating gas have been em- ployed. Moreover, it has been found advan- tageous to employ a plurality of sensitive arc lamps instead of a single arc lamp, in order to augment the action. In some cases, these arcs are all connected in series, while in others, they have been all connected in parallel. The Singing Arc The sensitive type of voltaic arc flame above described is called the " singing arc," by reason of a curious and interesting inverse property which it possesses. We have seen that, when properly adjusted, the arc auto- matically charges and discharges the associ- ated oscillatory branch c P c', by altering its volume and conducting power, in accordance with rapid variations of current strength. Such variations of breadth and volume in the 248 WIRELESS TELEGRAPHY arc flame set up corresponding vibrations in the surrounding air; so that the arc is able to emit sounds. In the case of high-frequency alternations, suitable for keeping an antenna in electric oscillation, the sound would prob- ably be inaudible, being above the limits of audibility; but if the frequency is sufficiently low, the arc can be made to give a fairly loud tone. In fact, if a properly adjusted arc lamp is supplied with a direct current, which has passed through a suitably designed carbon telephone transmitter, and musical sounds im- pinge upon the transmitter diaphram, the arc will be able to reproduce them, even though the arc may be at a great distance from the transmitter. In such a case, the transmitter produces rapid variations in the current sup- plying the arc, in conformity with the incident sound-waves. These current variations pro- duce corresponding fluctuations in the volume of incandescent vapor in the arc, which there- by exerts, in its turn, corresponding fluctua- tions in the pressure upon the surrounding air, and so produces sounds in the same. An arc lamp can thus be made to sing and reproduce music. An arc so adjusted is called a sing- ing arc. It is even possible to recognize vo cal sounds reproduced by the arc, but the ar- PRINCIPLES OF WIRELESS TELEPHONY 249 ticulation is seldom clean In the case -of adjusting an arc to reproduce sounds, the property of resistance variation in the arc vapor accompanying the sound is not brought into service; whereas in the application of the singing arc to exciting an antenna into oscil- lation, this property is of the first importance. Exciting Sustained Oscillation in an Antenna by Means of a High-frequency Alternator The second method, mentioned above, for continuously sustaining the radiation from a sending antenna, employs a specially con- structed high-frequency alternating-current dynamo, or alternator. The alternators which are used in America, for electric lighting and power transmission ordinarily generate either 60 cycles per second, or 25 cycles per second, the former frequency being suitable for street arc-lighting, and the latter for power trans- mission. The natural frequency of an un- loaded antenna, 50 meters (164 feet) in height, is in the neighborhood of 1,500,000 cycles per second ; so that this would be the proper frequency that an alternator should generate in order to be in simple synchronism, or in resonance, with the antenna. By load- ing the antenna, however, as described in 250 WIRELESS TELEGRAPHY Chapter IX, it is possible to reduce the natural frequency of the antenna, or the frequency of its free oscillation and radiation. By this means, it has recently been found possible to bring the natural frequency down to a limit which specially constructed high-frequency alternators can attain. FIG. 76. High-Frequency Turbo-Alternator. Fig. 76 illustrates a high-frequency turbo- alternator set. On the right hand side at T is a de Laval steam turbine which, with the aid of step-up gear, drives the shaft of the alternator A at a speed of about 16,000 revo- lutions per minute. The revolving element, or rotor, of the alternator, comprises a pair of PRINCIPLES OF WIRELESS TELEPHONY 251 steel disks, in the periphery of which 300 radial groves are cut. One of these disks is shown in Fig. 77. Between the grooves of two such revolving disks is mounted a thin stationary armature frame with 600 radial slots, and a coil in each pair of adjacent slots, the coils being then connected in series. Each revolution of the disks, with their 300 polar teeth or projec- tions, produces 300 cycles of alternating electromotive force in the armature; so that Fl 77- One of ! the Two Revolving Field Poles of at the Speed of l6,OOO the High-Frequency Al- r. p. m. there will be ternaton generated a frequency of 4,800,000 cycles per minute or 80,000 cycles per second. If now the armature is connected between the sending antenna and ground, and the antenna is tuned to this frequency, the alternator will be producing electric impulses in step with the natural electric oscillations of the antenna, and the system will be brought into full swing. The power developed elec- trically at the alternator terminals will be all radiated out from jthe antenna in electromag- 252 WIRELESS TELEGRAPHY netic waves, after deducting the heat losses which occur by the up and down movement of the electric current in the antenna. It has been found that the radiated power from a large and powerful sending antenna, when excited to resonance in the above manner, represents a load on the high-frequency alternator such as would be produced by a simple non-induc- tive resistance of 8 or 10 ohms. That is, the antenna, when in full oscillation, behaves as though it were grounded at the top of the mast through such a resistance. Receiving Circuit Connections When, by either of the above methods, or by some other arrangement, a sending antenna is excited into steady radiative action, a cor- responding steady emission of electromagnetic waves takes place at this antenna. Any re- ceiving mast within effective range will then be able to pick up a steady electric disturb- ance in the antenna, caused by the continued passage of these waves as they run by. The disturbance will take the form of an alternat- ing electromotive force, as described in Chap- ter VIII, and the frequency of the alternation will be identical with that of the sending an- tenna. The electrical effect of this alternat- PRINCIPLES OF WIRELESS TELEPHONY 253 ing disturbance will be a maximum when the receiving antenna is tuned into resonance with that frequency, by suitably adjusting its load of inductance, or capacity, or both. It may be possible to use any type of sensitive and rapidly acting wave detector in the receiving antenna circuit connected with a receiving telephone. An electrolytic receiver, such as that described in connection with Fig. 46, is found to answer the purpose satisfactorily, and the receiving connections may be such as are indicated in Fig. 63. Under these condi- tions, a high-frequency received current will pass through the receiver r, and a current of the same frequency will also be set up in the coils of the receiving telephone connected to the receiver. The frequency of even a heavily loaded antenna is, howevef, far above the highest frequency that the ear can detect; so that nothing is heard in the receiving tele- phone, although a very considerable high-fre- quency alternating current may be maintained flowing up and down the receiving antenna through the receiver r. In order to break this silence, it is necessary to modify the oscilla- tions of the sending antenna in accordance with the vocal sound-waves of the speaker, and to cause these modifications in the emitted 254 WIRELESS TELEGRAPHY waves to manifest themselves in the receiving telephone. Fig. 78 indicates diagrammati- cally a method of accomplishing this result. The rapid oscillations O A B C D E F repre- sent either the high-frequency alternating cur- rents supplied steadily to the sending antenna A B D E F FIG. 78. Diagram Illustrating the Production of Audi- ble-Sounds by the Modification in Amplitude of Ultra-Audible Frequency Currents. when no telephonic transmission occurs, or the high-frequency waves which are steadily being radiated, under that condition, from the send- ing antenna. If now the carbon transmitter at the sending station can be made to alter the amplitude of these outgoing waves, in accordance with the diagram O'A'B'C'D'- PRINCIPLES OF WIRELESS TELEPHONY 255 E'F', which performs periodic variations of thirty times lower frequency, then the tele- phone connected to the receiving antenna, as indicated in Fig. 63, may be regarded as giving vibrations of its diaphram corresponding to the variable amplitude high-frequency waves A'B'C'D'E'F'. This high-frequency may be be- yond the limits of audibility ; but the amplitude, wavering at the lower frequency, may pro- duce an audible effect corresponding to the simple musical tone wave a b c d e f. Conditions Sufficient for the Transmission and Reproduction of Speech In order, therefore, that articulate speech may be transmitted from the sending to the receiving antenna, and rendered capable of recognition by a listener at the latter, it is sufficient that an alternating current of ultra- audible frequency be steadily produced in the receiving antenna, and its apparatus, when the speaker is silent, and that when the speaker talks into the transmitter, the latter shall con- trol the amplitude of the high-frequency waves, so that the complex wave-forms of the vocal tones may be developed in the shapes of the waves of amplitude. Under these condi- tions, if the degree of amplitude affected is 256 WIRELESS TELEGRAPHY sufficient, and the distance between the send- ing and receiving stations is not too great, the receiving telephone may reproduce the vocal tones of the ^speaker with sufficient power to make speech recognizable. In the case of Fig. 78, only a pure musical note could be expected to become audible, but if the transmitter had produced any association of tones within, its compass, however com- plex, the same association might be expected to be bound up in the rapidly varying amplitudes of the successive outgoing waves and might be expected to be re- produced by the receiving tele- phone. One method of controlling the amplitude of the high-frequency out-going waves, in accordance with lower frequency vocal sounds, employs a carbon trans- mitter in the main sending an- tenna path as indicated in Fig. FIG. 79. Trans- 79- In this case, the antenna A'"" iTn^a M havin S been adjusted into Branch. resonance with the high-fre- quency alternator A, behaves substantially like a non-inductive resistance to ground, that is, it M PRINCIPLES OF WIRELESS TELEPHONY 257 virtually closes the circuit of the alternator upon the radiation resistance of the antenna plus local connections. The carbon transmitter T, in the main circuit, alters the resistance of this circuit in conformity with the vocal sound-waves im- pinging upon its diaphram. The amplitudes of the high-frequency alternating current, and of the electromagnetic waves emitted from the antenna, are thus caused to fol- low the wave forms of the speaker's vocal tones. It is not, however, necessary that the transmitter should be inserted in the main circuit. The trans- mitter may be placed in a circuit which is inductively connected with the main circuit, through the medium of an induction-coil such as is shown at I in Fig. 80. A condenser has also been used in the local transmitter circuit instead of the voltaic battery. In such cases, the variation of FIG. 80. Trans- resistance in the transmitter cir- mitter in In- . . . ,. ductively cuit when the transmitter dia- Connected phram is disturbed by sound- Circmt - waves is reflected inductively into the main an- tenna path, and serves to control the ampli- 258 WIRELESS TELEGRAPHY tudes of the emitted radiant waves. Another plan has been to place the high-frequency al- ternator in the primary circuit of an induction coil, the secondary circuit of which is connected with the antenna in the manner indicated in Fig. 36, care being taken to bring both the local oscillation branch A c B, and the antenna, into syntony with the high-frequency alternator, and with each other. This causes a steady stream of radiated waves of the same fre- quency to be thrown out from the sending mast and the amplitudes of these radiant waves is modified in conformity with the vocal tones of the speaker, by means of a trans- mitter, connected either in the antenna branch, or in some branch inductively associated with it. The distance to which wireless telephony can be practically carried depends upon the amount of electric power that can be con- trolled in amplitude of current by the trans- mitter at the sending antenna and upon the limiting minimum of electric power which can be picked up from the radiated waves by the distant receiving antenna and utilized to oper- ate the listener's telephone. The amount of power which the ordinary carbon transmitter controls in wire telephony may be less than PRINCIPLES OF WIRELESS TELEPHONY 259 one watt (equivalent to the power expended in lifting about J4 lb. I foot high per second) ; whereas, in wireless telephony, much larger powers^must be developed and controlled by the transmitter, on account of the scattering of this power in all directions from the sending antenna. The power that the transmitter is required to handle in long-distance wireless telephony may be hundreds or even thou- sands of watts. The ordinary carbon trans- mitter of wire telephony could not carry such an amount of power without becoming overheated and deranged. A special form of carbon transmitter designed to carry and con- trol alternating currents up to 15 amperes is shown in Fig. 81. The mouthpiece M leads the incident sound-waves to a metallic dia- phram which carries a short metallic rod fas- tened at its center. The rod passes through a hole in a metallic terminal plate and terminates in a platinum-indium spade. Vibrations of the diaphram cause the spade to vibrate in a chamber packed with carbon particles, and having for walls two metallic terminal plates, which are separated by the white insulating ring. The metallic terminal plates are con- nected to the terminals T T and the spade to the terminal t. Water, admitted by the open- 260 WIRELESS TELEGRAPHY ings W W, is circulated through the two terminal plates to keep them from being over- FIG. 81. A Form of Water-Cooled Carbon Transmit- ter Employed in Wireless Telephony. heated by the powerful alternating currents passing through the carbon. Freedom of Wireless Telephony from Distortion It seems evident, both from the principles and the practice of wireless telephony, that although the radiated impulses which carry the vocal tones are subject to marked atten- PRINCIPLES OF WIRELESS TELEPHONY 261 nation, being weakened, not only by the absorp- tion of the waves into the surface of the im- perfectly conducting land and sea, but also by the expansion of the waves into ever-in- creasing areas; yet they are not subject to the distortion which accompanies their transmis- sion in wire telephony, particularly when the wires are placed close together in an un- derground cable. In other words, the sound- waves of wireless telephony get fainter as the range of transmission is increased up to the limits at present existing; but all of the tones transmitted become fainter in the same pro- portion; so that there is no indistinctness pro- duced by the alteration of tone quality. Range of Wireless Telephone Transmission The greatest distance reported at present for the transmission of recognizable wireless te- lephony in America is from Brant Rock, Mass., to Washington, D. C, a distance of 657 kilo- meters (408 miles). In Europe, the greatest reported distance has been from Monte Mario at Rome, Italy, and a vessel off the coast of Sicily near Trapani, an over-sea distance of over 500 kilometers (300 miles). Wireless telegraphy is still young, but wireless telephony 262 WIRELESS TELEGRAPHY is younger still; so that the limits of range to which the human voice can be carried, on electromagnetic waves, are by no means yet set. It would seem that the limits lie with the amount of current and power which can be handled by the transmitter, assuming that no further improvements are made in direct- ing the outgoing beam of electromagnetic ra- diation, in the antennas, or in the delicacy of the receiving instruments. In one sense, the extension of the present range from a few hundred kilometers to the antipodes, or half way around the world (20,000 kilo- meters, or 12,000 miles), would be less won- derful than the already accomplished feat of reproducing recognizable speech at the range now attained; because the extension of the range of speech to the antipodes is a matter of degree; whereas the achievement of wire- less telephony to a range of even 100 kilo- meters (60 miles), is a wonderful acquisition in kind. It is only reasonable to expect, however, that the range of possible wireless telephony will be less than, and gradually increase to- wards, the range of possible wireless telegra- phy, because, in wireless telegraphy, the prob- lem of communication is limited to producing PRINCIPLES OF WIRELESS TELEPHONY 263 any recognizable type of signal that can be repeated in successive periods of dots and dashes, whereas, in wireless telephony, the problem of communication involves the more complex condition of reproducing, at the re- ceiving antenna, waves that have been succes- sively modified in a long succession, substan- tially in accordance with the sound-waves of a speaker's voice, either at the sending antenna, or at a station connected electrically with the sending antenna. Selectivity of Wireless Telephony Just as it is possible to select at a wireless receiving telegraph station one particular series of waves emitted from a particular send- ing station, to the exclusion of other sending stations, by some method of tuning; so it is possible to select at a wireless receiving tele- phone station, one particular series of waves emitted from a particular sending station, to the exclusion of other sending stations. For instance, if A, B, C and D, are four wireless" telephone stations, so located as- all to lie within each other's range of influence, and if A desires to speak with B exclusively ; while C desires to speak with D exclusively, it will suffice, for A and B to communicate in the fre- 264 WIRELESS TELEGRAPHY quency of say 80,000 cycles per second (wave- length 3.75 kilometers), and for C and D to communicate in the frequency of say 90,000 cycles per second (wave-length 3 1-3 kilo- meters). That is, not only the generating source (arc lamps or alternator) at A would be tuned to 80,000 ^; but the sending an- tenna system of A, and the receiving antenna system of B, would also require to be tuned to this frequency. When suitably tuned in this manner, waves of frequency 90,000 "\ would fail to be detected by B's receiver, and, therefore, all variations in the amplitudes of such waves, capable of reproducing speech, would be cut off from B's telephone. The telephone at B would only hear the speaker A, to which it was adjusted in syntony. With sharply-tuned antenna systems, it would be possible for a number of such telephonic con- versations to be carried on selectively, each employing a powerful series of independently acting electromagnetic waves. Simultaneous Speaking and Listening With the arrangements above described, it would be necessary to employ a switch to change the antenna connections of a wireless telephone station from sending to receiving, PRINCIPLES OF WIRELESS TELEPHONY 265 i.e., from speaking to listening, alternately. With ordinary wire telephony, such a switch was required in a very early stage of the art; but no switch is at present needed for this purpose, the transmitter and the telephone be- FlG. 82. Connections for Simultaneous Speaking and Listening. Duplex Telephony. ing always in the circuit simultaneously dur- ing conversation; so that it is possible both to speak and to listen at the same time. An arrangement of connections has been devised for effecting the same result for wireless te- lephony, and is indicated in Fig. 82. The an- tenna A, with its tuning coil 1, is permanently 266 WIRELESS TELEGRAPHY connected to ground G, through the secondary windings of four induction coils, i, 2, 3, 4 and an artificial antenna L c R, consisting of a suit- ably adjusted combination of inductance, ca- pacity and resistance. The high-frequency alternator H is connected to the transmitter T, through the primary windings of the four induction coils. Under these conditions, if no sounds are delivered to the transmitter, a steady high-frequency alternating current is sent through the four induction coils to both antennas. The real and artificial mast wires are thus both thrown into full electric oscil- lation. If the artificial antenna is properly adjusted so as to balance the real antenna, the four induction coils will mutually neu- tralize each other's influences, and no current will flow through the dotted system 5, c, 6, c', which connects the points e, f, and which con- nects with the receiving telephone t. Again, if the speaker talks into the transmitter T, the amplitudes of the high-frequency alter- nating currents will be varied in accordance with his vocal tones, but the power will be equally divided between the real antenna A and the artificial antenna L C R. The power in the real antenna will be expended in radi- ated electromagnetic waves, after deducting in- PRINCIPLES OF WIRELESS TELEPHONY 267 cidental losses in heating the mast wires. The power in the artificial antenna will be ex- pended in heating the resistance R, which rep- resents a radiation resistance, after deducting incidental losses in heating L and C. The real antenna may attain a height of 100 meters (328 feet), or more and may cover a con- siderable area of ground surface. The arti- ficial antenna is a small affair that may be put inside a cupboard of one cubic meter space (26.4 cubic feet). If electromagnetic waves are received at the real antenna A, of the same frequency as that to which it is tuned, they will develop an al- ternating current of that frequency passing to ground from the antenna through the points e f, and the dotted receiving system between them. Some of the current will pass also through the artificial antenna L C R; but will do no harm except in weakening the effect on the receiver. The divided primary receiving system 50, 6c', is called an interference pre- venter. All of these four elements are inde- pendently adjustable. The secondary receiv- ing system connects the two secondary coils through the liquid barretter or wave detector B, which is also connected to the voltaic bat- tery v through the receiving telephone t. By 268 WIRELESS TELEGRAPHY suitably differentiating the two oscillating- current receiving branches, it is possible to tune the secondary system to respond loudly to the selected frequency, and to the practical exclusion of all others. The result is that the receiving telephone t is prevented from re- ceiving any part of the locally generated high- frequency currents passing through T, owing to the differential balance between the four coils 1234 and between the two antennas A and L C R. It will be silent to those cur- rents, whether the transmitter T is spoken to or not; but the receiving telephone t is able to receive the incoming electromagnetic wave disturbances reaching the real antenna, be- cause these disturbances are not destroyed by the differential balance. In this manner, both speaking and listening may continue simul- taneously, as in ordinary wire telephony, al- though there is some weakening of the re- ceived currents, and also half the power avail- able for sending out waves is absorbed locally as heat in the artificial antenna. That is, more power must be used with the arrange- ment of Fig. 82, for the same limiting range of recognizable telephonic communication, than with alternate speaking and listening. The system of connections indicated in Fig. PRINCIPLES OF WIRELESS TELEPHONY 269 82 is likewise available for duplex wireless telegraphy; that is, for the simultaneous send- ing and receiving of messages at one and the same antenna. It has been found that the ar- tificial antenna, once adjusted to balance the real antenna, requires less change from. day to day than does the " artificial line " em- ployed in duplexing an ordinary wire tele- FIG. 83. Adjustable Condenser and Induction Coil Forming Elements of Interference Preventer. graph line. This is .apparently due to the fact that changes of weather have more influence in changing the electric conditions of a line hundreds of kilometers in length, than in changing the electric conditions of an an- tenna. An adjustable air-condenser and an adjusta- ble induction-coil for use in an interference preventer is shown in Fig. 83. 270 WIRELESS TELEGRAPHY Relaying Telephonic Currents to and From an Antenna In telephoning wirelessly from a ship to a ship, or between a wireless shore station and a ship, the persons conversing together are FIG. 84. Automatic Telephone Relay. close to their respective antennas; but when one of the persons is on shore at some place telephonically connected with, but remote from, the wireless telephone station on the sea coast, and wishes to speak to a person on a ship within range, it is necessary for his con- PRINCIPLES OF WIRELESS TELEPHONY 271 versation either to be repeated by the operator at the coast station acting as intermediary; or to be repeated automatically to and fro by relays at the coast station. Fig. 84 shows a relay designed for this duty. It consists essentially of a telephone receiver in which a little vibratory tongue is substituted for the usual vibratory diaphram. The tongue dips into a trough containing carbon particles with water-cooled walls, arranged substantially as in the transmitter of Fig. 81. The apparatus is in fact a telephone receiver directly operat- ing a carbon transmitter. The receiver is con- nected at the coast station to the incoming telephone line wire circuit, and transmits di- rectly to the antenna. Another relay receives from the antenna, and transmits back to the wire telephone circuit. INDEX OF SUBJECTS PAGE A Activity of antenna.... 100 Air, Sound- Waves in. . 5 Alphabets, Morse. ...".. 153 Alternating-current ac- companying radiated wave 47 Alternating-current dy- namo 105 Alternator, High - fre- quency 249 Ammeter * 193 Amplitude of audible sound 216 Amplitude of receiver diaphragm vibration.. 238 Amplitude of sound- waves 10 Antenna 97 Antenna, Activity of... 100 Antenna, Artificial in duplex telephony or telegraphy 269 Antenna, Essential ele- ments of 98 Antenna, loaded, Inter- nal reflections in 123 Antenna, Radiation re- sistance of 252 Antennas, Cylindrical... 102 Antennas, Fan 103 Antennas, Harp 103 Antennas, Height of... 158 Antennas, Insulation of 157 Antennas, Inverted-cone 103 PAGE Antennas, Loaded 120 Antennas, Multiple-wire 101 Antennas, Single-wire.. 101 Arc, The singing 247 Artificial antenna in du- plex telephony or te- legraphy 269 Atmospheric absorption and magnetic varia- tion of compass, Pos- sible connection be- tween 203 Attenuation of tele- phonic currents 233 Audible sound, Ampli- tude of 216 Auditory selection. ..... 175 Auxiliary condenser. ... 115 B Branch circuit, Oscilla- tory 242 Capacity of condenser. . 116 Circuit, Local, of re- ceiver 127 Circuit, Portable. ... . . . 193 Changes of wave form in telephonic trans- missions over long wires 232 Classification of elec- tromagnetic waves , , , 74 274 INDEX PAGE Coherers 125 Coil, Self-induction of. . 122 Comparison of receivers 145 Compass of tones 220 Condensation of fluxes toward equatorial zone 61 Condenser, Auxiliary.. . 115 Condenser, Nature of . . 116 Conditions sufficient for the wireless transmis- sion and reproduction of speech , . . . 255 Connections for duplex telephony 265 Consonance, Electric... 118 Continuous radiation, Methods of maintain- ing 241 Counter - electromotive - force of polarization. 139 Current-detectors 124 Cylindrical antennas 102 D Decohering, Mechanical 129 Depolarizing of elec- trodes 138 Detectors, Current 124 Detectors, Electrolytic.. 125 Detectors, Electromag- netic 125 Detectors, Thermal 125 Detectors, Voltage 124 Deviation of waves from hemispherical form 60 Diameter of telephone relay 270 Difference between mu- sical sound and noise. 211 Diminution of intensity in sound-waves 8 PAGE Directing wireless tele- graph waves 186 Discontinuity of coil- fed antenna oscilla- tions 113 Disk galvanometer 195 Distortion of telephonic wave currents over wires 233 Duplex wireless tele- graph, Connections for 265 Dynamo, Alternating- current 105 Earth's Surfaces, Im- perfect conduction of 48 Eddy currents in sur- face of earth or sea. . 78 Electric and magnetic fluxes, Tensions in. .. 26 Electric and magnetic forces in moving waves 79 Electric consonance.... 118 Electric conductorsj Resonance in 90 Electric Field 25 Electric Flux and its properties 23 Electric Flux, Energy of 24 Electric Flux, move- ment over conductors 29 Electric Flux, Proper- ties of 23 Electric Flux, Provi- sional hypothesis con- cerning nature of. ... 25 Electric oscillations, Skin depth of 105 Electric Resonance. .... 91 INDEX 275 PAGE Electric strength of in- sulators 117 Electricity and magne- tism, Nature of 16 Electric theory of mat- ter 68 Electrodes 140 Electrolyte, Definition of 127 Electrolytic detectors. . . 125 Electromagnetic detec- tors 125 Electromagnetics, and optics 73 Electromagnetic waves and polarized light. .. 67 Electromagnetic waves Classification of 74 Electromagnetic wave- detectors 124 Electromagnetic - wave, Energy of 33 Electromagnetic waves, Hemispherical 59 Electromagnetic waves, Measurement of 191 Electromagnetic wave, Nature of 31 Electromagnetic waves, Plane 75 Electromagnetic wave, Radiation of 40 Electromagnetic waves, Speed of 33 Electromagnetic waves, Spherical 64 Electromagnetic wave- trains Electromotive force 85 Energy in waves 4 Energy of Electric Flux 24 Energy of Electromag- netic waves 33 Energy of Magnetic Flux 21 PAGE Energy of received elec- tric oscillations 94 Energy of Wind 14 Equatorial zone, Con- densation of fluxes toward 61 Ether, Assumption of The 15 Fan antennas., 103 Free Ocean Waves 3 Freedom of wireless te- lephony from distor- tion 260 Frequency, Group 175 Frequency, Relations to wave length and peri- odic time 62 Galvanometer, H i g h - frequency 195 Gashes in waves, torn by vertical conduc- tors 85 German practice 185 Group frequency 175 Guided electromagnetic waves 35 H Harmonics 219 Harp antennas 103 Heights of antennas... 158 Helix, wave-measuring. 197 Hemispherical electro- magnetic waves 59 High-frequency alterna- tor 249 High-frequency galva- nometer 195 Hot-wire receivers 132 INDEX PAGE Image, Electric 38 Impact of waves against vertical conductors ... 92 Imperfect Conduction of earth 48 Induction coil for high voltage 107 Industrial wireless te- legraphy 198 Insulation of antennas. 157 Insulators. Electric strength of 117 Intensity of musical tones 215 Intensity in sound- waves, Diminution of 8 Internal reflections in loaded antenna 123 Interference, Preventer 267, 269 Inverted-cone antennas. 103 Ionizing of rarefied air. 166 K Krakatoa explosion. .... 1 1 Light, Plane-polarized.. 72 Loaded antennas 120 Local circuit of receiver 127 Loudness oJf musical tones 215 M Magnetic and electric fluxes, Tensions in. . . 26 Magnetic and electric forces in moving waves.., 79 PAGE Magnetic Flux and its Properties 18 Magnetic Flux Created by Moving Electric Flux 30 Magnetic Flux, Energy of .. 21 Magnetism and Elec- tricity, Nature of 16 Matter, Electric theory of 68 Measurements of elec- tromagnetic waves... 191 Mechanical decohering. 129 Methods' of maintaining continuous radiation. 241 Micron, a unit of length 68 Microsecond, Unit of time 99 Mnemonic rules for di- rections of fluxes in waves 52 Morse alphabets 153 Morse inkwriter 146 Multiple-wire antennas. 101 Multiple wireless teleg- raphy 183 Musical sound and noise, Difference be- tween 211 Musical tones, Intensity or loudness of 215 Musical tones, Pitch of 216 Musical tones, Purity of 217 N Nature of sound 209 Number of land wire- less stations in 1908. . 206 Non-conductors, Trans- parency of , , . . , 79 INDEX 277 PAGE o Ocean steamers, Wire- less telegraph equip- ment on 205 Ocean Waves, Free.... 73 Optics and electromag- netics 93 Oscillating-current gal- vanometer 196 Oscillating currents in receiving verticals 94 Oscillator, Simple ver- tical 49 Oscillatory branch cir- cuit 242 Oscillograms of wire telephone currents. . . . 230 Overtones 219 Overtones, Phase of.... 219 P Periodic time, relations to wave-length and frequency . 62 Phase of overtones 219 Pitch of single musical tones 216 Plane electromagnetic waves 75 Plane-polarized light... 72 Plane waves of sound in a speaking tube... 212 Polarized light and elec- tromagnetic waves... 67 Portable circuit 193 Possible connection be- tween atmospheric ab- sorption and magnetic variation of compass. 203 Power required for wireless telegraph sending 160 Preventing interference ..................267,269 Purity of musical tones 217 PAGE R Radiation of electro- magnetic waves 49 Radiation resistance of antenna 252 Range of recognisable wireless telephonic transmission .... .258, 261 Received electric oscil- lations, Energy of.... 94 Receivers, Comparison of ... 145 Receivers, Wireless tel- egraph 124 Relations between wave- length, frequency, and periodic time 62 Relay for use in te- lephony 270 Relay, Electromagnetic 130, 170 171 Resonance in electric conductors . . . . 90 Resonant selection 176 Selective signalling 173 Selectivity of wireless telephony 263 Self-induction of coil.. . 122 Sending apparatus, ele- ments of no Sending keys 151 Shadow cast by electric conductors 80 Short-range apparatus.. 168 Signalling range on sea and land 86 Simple vertical oscilla- tor 49 Simultaneous sending and receiving 180 Simultaneous speaking and listening.. , . . 264 INDEX PAGE Singing arc, The 247 Singing arc, Method of sustaining oscillation. 242 Single-wire antennas. .. 101 Skin depth of electric oscillations 105 Sound, Nature of 209 Sound-wave Trains u Sound-waves, A m p 1 i - tude of 10 Sound-waves, Diminu- tion of intensity in.... 8 Sound-waves in air 5 Sound-waves, Speed of. 8 Speaking and listening, Simultaneous 264 Speaking tubes, Plane waves of sound in.. . . 212 Speaking voice, Tones in 223 Speed of electro-mag- netic waves 33 Speed of sound-waves. . 8 Spherical electromag- netic wave 64 Spreading of wireless- telegraph waves 162 Sunlight, Effect on wireless telegraphy of 201 Telephone Receiver, Principles of 236 Telephone receivers 147 Telephone transmitter, Principles of 225 Tensions in electric and magnetic fluxes 26 Thermal detectors 125 Thermo-electric force. . 195 Thermo-galvanometer . 194 Tones in the speaking voice .... 4 223 PAGE Total number of re- corded ship and shore stations 207 Transmitter employed in wireless telephony. 260 Trans-oceanic wireless telegraphy 204 Transparency of electric non-conductors 79 Tuned signalling 173 U Unguided electromag- netic waves 64 Vertical conductors, Im- pact of waves against 92 Vertical oscillator, Es- sential elements of. .. 98 Vertical oscillator, Sim- ple 49 Voltage 108 Voltage-detectors 124 W Wave-detectors, Elec- tromagnetic 124 Wave form in telephon- ic transmissions over long wires 232 Wave-length, Relations to frequency and peri- odic time 62 Wave-lengths, Determi- nation of 191 Wave-lengths, in ocean waves 4 Wave-measuring helix.. 197 Wave trains, Electro- magnetic 48 Wind energy 14 INDEX 279 PAGE Wireless stations in 1908, Number of land 206 Wireless-telegraph in- terference 173 Wireless-telegraph re- ceivers 124 Wireless- telegraphy equipment on ocean steamers 205 Wireless-telegraphy, In- dustrial 198 Wireless-telegraphy on vessels 202 Wireless-telegraph sta- tions, Number of 204 Wireless- telegraph waves, Spreading of. . 162 PAGE Wireless telephone cir- cuit connections 256 Wireless telephonic range 258, 261 Wireless telephony, Freedom from distor- tion 260 Wireless telephony, Se- lectivity of 263 Wireless telephony, transmitter employed in 260 Wireless transmission and reproduction of speech, Conditions sufficient for 255 UNIVERSITY OF CALIFORNIA LIBRARY BERKELEY Return to desk from which borrowed. This book is DUE on the last date stamped below. ENGINEERING LIBRARY APR 1950 D 21-100m-9,'48(B399sl6)476 24190 271151 UNIVERSITY OF CALIFORNIA lylBRARY